{"kind":"task","effective_mode":"full","benchmark":{"kind":"benchmark","effective_mode":"full","slug":"longbench-v2","formal_name":"LongBench v2","introduction":"LongBench v2 evaluates deep understanding and reasoning over long contexts through multiple-choice questions. Its official description lists 503 questions spanning tasks such as single-document and multi-document QA and code-repository understanding.","introduction_ja":"","introduction_en":"","category":"Category not supplied","task_count":null,"acquisition_status":"Acquisition status not supplied","official_url":"https://huggingface.co/datasets/zai-org/LongBench-v2","indexing_mode":"noindex","profile":{"resources":[],"task_format":"","scoring":"","metric":"","size":"","answer_access":"","license":"","citation":"","maintainer":"","released":"","why_hard":"","related":[]}},"task_id":"b94047eb-347c-5057-9dcd-006584ca3629","task_key":"train--66f00f1a821e116aacb252cf","task_revision_id":"3","upstream_id":"66f00f1a821e116aacb252cf","short_description":"Considering both ‘Social Identity and Intergroup Conflict and Conflict…","config":"","split":"train","body":"{\"choice_A\":\"Both texts agree that group narratives are essential to forming social identity and conflict, with ‘Social Identity and Intergroup Conflict and Conflict Reduction’ emphasizing the integration of power structures into conflict resolution strategies, while ‘Social Identity: International Perspectives’ highlights the need for addressing historical trauma through narrative transformation as a key part of mediation.\",\"choice_B\":\"Both texts agree that group narratives shape social identity and conflict, but ‘Social Identity and Intergroup Conflict and Conflict Reduction’ highlights the importance of integrating power dynamics into conflict resolution strategies, while ‘Social Identity: International Perspectives’ focuses on the need for addressing collective trauma through storytelling as part of mediation.\",\"choice_C\":\"Both texts acknowledge that group narratives shape intergroup conflict, with ‘Social Identity and Intergroup Conflict and Conflict Reduction’ focusing on the impact of power dynamics on narrative formation, while ‘Social Identity: International Perspectives’ stresses the importance of reshaping narratives through trauma-informed mediation approaches.\",\"choice_D\":\"Both texts recognize that group narratives play a central role in escalating conflicts, with ‘Social Identity and Intergroup Conflict and Conflict Reduction’ arguing for the integration of socio-political power relations in conflict resolution, while ‘Social Identity: International Perspectives’ focuses on using collective trauma healing to transform these narratives through mediation.\",\"context\":\"List o\\nf\\n Tables and Figures \\nList o\\nf\\n Contributors \\nPre\\nf\\nace \\nContents \\nPart I \\nRepresentations of Self and Group \\nRegarding the Relationship Between Social Identity and \\nPersonal Identity \\nJean-Claude Deschamps and Thierry Devos \\nVII \\nIX \\nXVll \\n2 \\nSocial Representations in Personal Identity \\n1 3  \\nWillem Doise \\n3 \\nThe Personal-Social Interplay: Social-Cognitive Prospects on \\nIdentity and Self-Others Comparison \\n24 \\nCarmencita Serino \\n4 \\nSelf \\n-We-Others Schemata and Social Identifications \\n44 \\nMaria Jarymowicz \\nPart II Establishing Group Identity \\n53 \\n5 \\nA Developmental View of the Search for Group Identity \\n53 \\nStephen Worchel \\n6 \\nThe Role of Prototypicality in Group Influence and Cohesion: \\nContextual Variation in the Graded Structure of Social \\nCategories \\n75 \\nPenelope Oakes, S. Alexander Haslam and John C. Turner \\n7 \\nGroup Beliefs as an Expression of Social Identity \\nDaniel Bar-Tal \\n8 \\nSocial Identity and National Identity \\nJose Miguel Salazar \\nPart III Identity and GrouplIntergroup Process \\n9 \\nSocial Identity and Intragroup Differentiation as Subjective \\n93 \\n1 14 \\n124 \\nSocial Control \\n124 \\nJose M. Marques, Dario Paez and Dominic Abrams \\n\\n\\nVI \\nSocial identity: international perspectives \\n10 \\nCategorization and Social Influence \\n142 \\nJuan Antonio Perez and Gabriel Mugny \\n11 \\nSocial Identity and Interpersonal Relationships \\n154 \\nMiguel Moya \\n12 \\nSocial Identity and Aspects of Social Creativity: Shifting to \\nNew Dimensions of Intergroup Comparison \\n166 \\nSteve Hinkle, Laurie A. Taylor, Lee Fox-Cardamone and \\nPamela G. Ely \\n13 \\nMere Categorization and Competition: A Closer Look at \\nSocial Identity Theory and the Discontinuity Effect \\n180 \\nStephen M. Drigotas, Chester Insko and John Schopler \\nPart IV \\nThe Impact of Culture on Identity and Categorization \\n199 \\n14 \\nDiscrimination and Beliefs on Discrimination in Individualists \\nand Collectivists \\n199 \\nJ. Francisco Morales, Mercedes Lopez-Saez and Laura Vega \\n15 \\nConstructing Social Identity: The Role of Status, Collective \\nValues, Collective Self-Esteem, Perception and Social \\nBehaviour \\n211 \\nDario Paez, Cristina Martinez-Taboada, Juan Jose Arrospide, \\nPatricia Insua and Sabino Ayestaran \\nRe\\nf\\nerences \\nIndex \\n230 \\n255 \\n\\n\\nList of Tables and Figures \\nTables \\n4.1 The social self as a structural basis of different forms of \\nsocial identifications \\n4S \\nS.1 Perceived differences in ingroup and outgroup similarity \\n70 \\nS.2 Minority influences in groups over time \\n72 \\n1l.l Meta-contrast ratios in self-perception and the four stimuli \\nused by Moya and Peplau \\n162 \\n13.1 Max. reI. as a function of matrix type (PDR versus PDC), \\ntrials and groups versus individuals, experiment 2 \\n190 \\n14.1 Summary of some differences between individualists and \\ncollectivists \\n202 \\n14.2 Relation between identification with the goup and intergroup \\ndifferentiation as a function of individualism/collectivism and \\nautonomous/relational orientation \\n203 \\nIS.1 Perceived ingroup and outgroup heterogeneity \\n216 \\nIS.2 Perception of ingroup and outgroup by salience versus \\nnon-salience of social comparison \\n217 \\nIS.3 Perceived social desirability of attributes for engineers and \\npsychologists \\n218 \\n15.4 Attributes for engineers and psychologists \\n220 \\nIS.5 Factorial analysis of the 'most important social identity \\nscale' (Garza & Herringer, 1986) \\n224 \\nIS.6 Differences between engineers and psychologists as a \\nfunction of their degree of collectivism \\n227 \\nFigures \\n1.1 Personal and social dimensions in psychosocial identity \\n11 \\n3.1 Mean scores (x 100) of each content category elicited in self/ \\ningroup versus self/outgroup similarity judgements ( N  = 121) \\n39 \\n4.1 Selfˮwe schemata distinctness and the degree of individuals' \\ninvolvement with outgroups and dissimilar others: the \\ntheoretical model \\nSO \\n4.2 Recall of children's problems \\nSI \\n4.3 Motivation to solve children's problems \\n51 \\n4.4 Readiness to host different nationalities \\nSI \\n\\n\\nVlll \\nSocial identity: international perspectives \\n4.5 Tolerance of German minority activity \\n5 1  \\n4.6 Positivity of evaluation of Slovak and Irish paintings \\n52 \\n4.7 Positivity of evaluation of Japanese painting \\n52 \\n5.1 Stages of group development in existing groups \\n58 \\n5.2 Nature of desired interaction with outgroup as a function of \\ngroup life \\n68 \\n6. 1 A categorial hierarchy of vegetation \\n78 \\n6.2 A categorial hierarchy of Europeans \\n79 \\n9. 1 Manipulation of group uniformity and group members' \\nstatus: distribution of ingroup and outgroup members' \\nresponses \\n1 34 \\n9.2 Evaluations of attractive and unattractive ingroup and \\noutgroup members as a function of group uniformity \\n1 35 \\n9.3 Example of ranking of characters A-F used to build modal \\nand deviate members \\n1 36 \\n9.4 Effects of norm awareness on judgements of modal and \\ndeviate ingroup and outgroup members \\n1 37 \\n9.5 Effects of ingroup and outgroup accountability on \\njudgements of modal and deviate ingroup and outgroup \\nmembers \\n1 38 \\n9.6 A two-step process of inter- and intragroup judgement \\n140 \\n1 1 . 1  Interpersonal and intergroup demands according to \\nStephenson (1984) \\n1 58 \\n1 1 .2 Example of a hypothetical identity taken from Deaux's \\nformulation and Moya's data \\n161 \\n12. 1 \\nIngroup favouritism on different comparison dimensions as a \\nfunction of outgroup status \\n1 73 \\n12.2 Ingroup favouritism on different comparison dimensions as a \\nfunction of outgroup status \\n1 75 \\n1 3. 1  A Prisoner's Dilemma Game matrix (PDG) \\n1 88 \\n1 3.2 A Prisoner's Dilemma max. reI. matrix (PDR) and a \\nPrisoner's Dilemma confounded matrix (PDC) \\n1 89 \\n1 3.3 A Prisoner's Dilemma Game alternative matrix (PDG-Alt.) \\n195 \\n\\n\\nList of Contributors \\nDominic Abrams is Professor of Social Psychology and Director of the \\nCentre of the Study of Group Processes at the University of Kent, UK. His \\nmain research interests are intergroup conflict and differentiation, social \\nidentity processes and social self-regulation. He is co-author and co-editor \\n(together with M.A. Hogg) of four books: Social Identifications (London: \\nRoutledge, 1988); Social Identity Theory (Hemel Hempstead: Harvester \\nWheatsheaf, 1990); Group Motivation (Hemel Hempstead: Harvester \\nWheatsheaf, 1993); and Social Identity and Social Cognition (Oxford: \\nBlackwell, forthcoming, 1998). \\nJuan Jose Arrospide is currently Assistant Professor at the Department of \\nSocial Psychology, School of Psychology, Basque Country University, \\nSpain. He has done research on the topic of social identity, group dynamics \\nand social representions. He is co-author (together with S. Ayestaran and \\nC. Martinez-Taboada) of a chapter on 'New Techniques for the Study of \\nGroups' ('Nuevas Tecnicas en el estudio de los grupos') included in \\nMethods and Techniques o\\nf\\n Research in Social Psychology ( Mhodos y \\nTecnicas de Investigacion en Psicologia Social, Madrid: Eudema, 1994). \\nSabino Ayestaran is currently Professor of Social Psychology at the \\nDepartment of Social Psychology, School of Psychology, Basque Country \\nUniversity, Spain. His research interests focus on social identity and group \\ndynamics, with special attention to intra- and intergroup conflict. He is the \\neditor of a volume on The Group as a Social Construction (El Grupo como \\nConstruccion Social, Barcelona: Plural, 1996). \\nDaniel Bar-Tal is Professor of Psychology at the School of Education, Tel­\\nAviv University, Israel. His research interests focus on political and social \\npsychology, studying societal beliefs regarding conflict, delegitimization, \\nsecurity, patriotism and siege mentality. He is the author of Group Belie\\nf\\ns\\n \\n(New York: Springer Verlag, 1990) and Societal Belie\\nf\\ns\\n o\\nf\\n Ethos (London: \\nSage, in press) and co-editor of Social Psychology o\\nf\\n Knowledge (with W. \\nStroebe, A.W. Kruglanski and M. Hewstone; New York: Springer Verlag, \\n1988), Stereotyping and Pre\\njudice (with C. Grauman, A.W. Kruglanski and \\nW. Stroebe; New York: Springer Verlag, 1989), Patriotism in the Lives o\\nf\\n \\nIndividuals and Nations (with E. Staub; Chicago: Nelson-Hall, 1997) and \\nConcerned with Security (with D. Jacobson and A. Klieman; Greenwich, \\nCT: JAI Press, in press). \\n\\n\\nx \\nSocial identity: international pers\\npectives \\nJean-Claude Deschamps, born in France, is currently Professor of Social \\nPsychology at the School of Social Sciences, Lausanne University, \\nSwitzerland. He has studied social identity and intergroup relations, the \\nprocesses of causal attribution and social representations. He is co-author \\n(with W. Doise) of a chapter on 'Crossed Categories Membership and \\nIntergroup Relations', included in H. Tajfel (ed.), Differen\\nSocial Groups (Cambridge: Cambridge University Press, 1978) and (with A. \\nClemence) of a volume on Everyday Explanation (L'ex\\nplication quotidienne, \\nFriburg: Delval, 1987). \\nThierry Devos is Assistant Professor at the School of Social Sciences, \\nLausanne University, Switzerland. He has studied social identity, causal \\nattributions and social representations. He is co-author (with J.-c. \\nDeschamps and L. Comby) of 'Assymmetries in Judgment of Ingroup \\nand Outgroup Variability', European Review o\\nf\\n Social Psychology (1996). \\nWillem Doise, born in Belgium, is Professor of Social Psychology at the \\nSchool of Psychology, University of Geneva, Switzerland. His research has \\nfocused on the link between mind and society, specifically social represen­\\ntations and the role of socio-cognitive conflict in mental development. He \\nhas published several books, among them Groups and Individuals \\n(Cambridge: Cambridge University Press, 1978) and Levels o\\nf\\n Explanation \\nin Social Psychology (Cambridge: Cambridge University Press, 1986). \\nStephen M. Drigotas is currently Assistant Professor of Psychology at the \\nSouthern Methodist University, Dallas, Texas, USA. His recent publications \\ninclude: 'On the Peculiarities of Loyalty' (together with G.A. Whitney and \\nC.E. Rusbult), Personality and Social Psychology Bulletin (1995), 'The Role \\nof Identifiability in the Reduction of the Interindividual-Intergroup Dis­\\ncontinuity Effect' (together with J. Schopler et al.), Journal o\\nf\\n Personality \\nand Social Psychology (1995), and 'Willingness to Sacrifice in Close Rela­\\ntionships' (together with P.A.M. Van Lange et al.), Journal o\\nf\\n Personality \\nand Social Psychology (in press). \\nPamela G. Ely is Assistant Professor and Chair of Psychology at Saint \\nAndrews College in Laurinburg, North Carolina, USA. She also directs the \\nHuman Services Program of the Saint Andrews at Sandhills program. Her \\nresearch focuses on the relationship between individualism/collectivism and \\ngroup and intergroup processes, including group-oriented instructional \\nmethods. Her most recent publication is 'Recognizing Group Diversity: \\nIndividualist-Collectivist and Autonomous-Relational Social Orientations \\nand Their Implications for Intergroup Processes' (with R.J. Brown et al.), \\nBritish Journal o\\nf\\n Social Psychology (1992). \\nLee Fox-Cardamone is Assistant Professor of Psychology at Kent State \\nUniversity - Stark Campus in Canton, Ohio, USA. Her research interests \\ninclude political activism, intergroup processes and group decision-making. \\nHer most recent publications are 'Grassroots Political Actions as an \\n\\n\\nList o\\nf\\n contributors \\nXl \\nIntergroup Phenomenon' (with S. Hinkel et al.), Journal o\\nf\\n Social Issues \\n(1996), and 'Recognizing Group Diversity: Individualist-Collectivist and \\nAutonomous-Relational Social Orientations and Their Implications for \\nIntergroup Processes' (with R.J. Brown et al.), British Journal o\\nf\\n Social \\nPsychology (1992). \\nS. Alexander Haslam is a Senior Lecturer in Social Psychology at the \\nAustralian National University. His main research interests are in the social \\npsychology of stereotyping and social judgement with particular emphasis on \\nthe role which group memberships play in those processes. This work \\naddresses issues of \\nsocial categorization, social identity and prejudice, dealing \\nspecifically with topics of stereotype consensus, perceived group homo­\\ngeneity, illusory correlation, prejudice and extremism. His recent publica­\\ntions include Stereotyping and Social Reality (with P.J. Oakes and J.C. \\nTurner; Oxford: Blackwell, 1994), The Social Psychology o\\nf\\n Stereotyping and \\nGroup Lif\\ne\\n (with R. Spears, P.J. Oakes and N. Ellemers; Oxford: Blackwell, \\n1997), The Message o\\nf\\n Social Psychology (with C. McGarty; Oxford: \\nBlackwell, 1997), and Doing Psychology (with C. McGarty; London: Sage, in \\npress). \\nSteve Hinkle is an Associate Professor of Psychology at Miami University \\nin Oxford, Ohio, USA. His main research interests are in motivational \\naspects of social identity theory, the relationship between individualism/ \\ncollectivism and aspects of group and intergroup processes, and the psy­\\nchology of political action. Recent publications include 'Grassroots \\nPolitical Action as an Intergroup Phenomenon' (with L. Fox-Cardamone \\net al.), Journal o\\nf\\n Social Issues (1996) and 'Recognizing Group Diversity: \\nIndividualist-Collectivist and Autonomous-Relational Social Orientations \\nand Their Implications for Intergroup Processes' (with R.J. Brown et al.), \\nBritish Journal o\\nf Social Psychology (1992). \\nChester Insko is Professor at the Department of Psychology, University of \\nNorth Carolina, USA. Research interests are intergroup relations, inter­\\npersonal attraction and attitude change. Recent publications include \\n'Memory for and Experience of Differential Competitive Behavior of \\nIndividuals and Groups' (with M.J. Pemberton and J. Schopler), Journal \\no\\nf\\n Personality and Social Psychology (1996), and 'Differential Distrust of \\nGroups and Individuals' (with J. Schopler) and 'Entitativity, Loss of Per­\\nsonal Control, and Evolution' (with J. Schopler and C. Sedikides), both \\nincluded in C. Sedikides et al. (eds), Intergroup Cognition and Intergroup \\nBehavior (Hillsdale, NJ: Erlbaum, in press). \\nPatricia Insua is Assistant Professor of Language and Thinking in the \\nDepartment of Basic Psychology, School of Psychology, University of the \\nBasque Country. She is co-author, with D. Paez and J. Marques, of a chapter \\non 'Social Cognition' in the book edited by J.F. Morales et al. entitled Social \\nPsychology (Psicologia Social) (Madrid: McGraw Hill, 1994). Her main \\n\\n\\nxu \\nSocial identity: international perspectives \\nresearch interests in social psychology are social representations, attitudes, \\nAIDS and drug-addiction processes, and they also embrace general \\npsychological processes of language, cognition and emotional problems. \\nMaria Jarymowicz has been Professor of Psychology at the School of \\nPsychology of the University of Warsaw and Head of the Laboratory \\n'Studies on Personal and Social Identity' since 198 1 .  Her main research \\ninterests focus on ego and self in social perception, self-we-others schemata \\ndistinctiveness and personal/social identity differentation - in relation to \\negocentric/non-egocentric perception, the individualism-collectivism rela­\\ntionship and the affective basis of social stereotypes. Her most representative \\npublications (both as editor) are To Know Self \\n- to Understand Others (Delft: \\nEburon, 1993) and Beyond Egocentrism in Self- and Social Perception (Poza \\nEgocentrycznQ PerspektywQ Widzenia Siebie i Swiata, Warsaw: Wydaw­\\nnictwo Instytutu Psychologii PAN, 1994). \\nMercedes Lopez-Saez is Assistant Professor of Social Psychology at the \\nDepartment of Social Psychology of the Spanish Open University at \\nMadrid (UNED). Her main research interests include gender stereotypes, \\nsocial influence and persuasion. Recent publications include 'Cultural and \\nIndividual Processes Involved in Gender Stereotyping: An Empirical \\nApproximation to Career Election' ('Procesos Culturales e Individuales \\nImplicados en la Estereotipia de Genero: Una Aproximacion Empirica a la \\nEleccion de Carrera'), Revista de Psicologia Social (1994), 'Gender Stereo­\\ntypes in the Spanish Population: Looking Toward the Future' (with 1.F. \\nMorales), in L. Amancio and C. Nogueira (eds), Gender, Management and \\nScience (Braga: Institute of Education and Psychology, 1995), and The \\nElection o\\nf a Typically Feminine or Masculine Career f\\nrom a Psychosocial \\nPerspective (La Eleccion de una Carrera Tipicamente Femenina 0 Masculina \\nDesde la Perspectiva Psicosocial), Madrid: Ministry of Science and \\nEducation, 1995). \\nJose M. Marques is Professor of Social Psychology at the University of \\nPorto, Portugal. Research interests centre mainly on the field of intergroup \\nrelations, specifically the role of intergroup processes in generating group \\ncohesiveness and enhancement of social identification. Recent publications \\ninclude 'The Black Sheep Effect: Social Categorization, Rejection of \\nIngroup Deviates, and Perception of Ingroup Variability' (with D. Paez), \\nEuropean Review o\\nf\\n Social Psychology (1994), 'The Representation of \\nEmotions in Groups: The Relative Impact of Social Norms, Positive­\\nNegative Asymmetry and Familiarity on the Perception of Emotions' (with \\nD. Paez and A.I. Vergara), European Journal o\\nf\\n Social Psychology (1996), \\nand 'Collective Memory Processes Associated with Traumatic Experiences: \\nSocial Sharing, Emotional Climate and the Transgenerational Transmission \\nof Information in the Case of the Portuguese Colonial War' (with D. Paez \\nand A.F. Serra), in 1. Pennebaker et al. (eds), Social Memory and Collective \\nTrauma (Hillsdale, Nl: Erlbaum, 1997). \\n\\n\\nList o\\nf\\n contributors \\nX \\n111 \\nCristina Martinez-Taboada is Assistant Professor at the Department of \\nSocial Psychology, School of Psychology, Basque Country University, \\nSpain. He has done research on the topic of social identity, group dynamics \\nand social representions. He is co-author (with S. Ayestarin and J.J. \\nArrospide) of a chapter on 'New Techniques for the Study of Groups' \\n('Nuevas Tecnicas en el Estudio de los Grupos') included in Methods and \\nTechniques o\\nf\\n Research in Social Psychology ( Mhodos y Tecnicas de \\nInvestigaci6n en Psicologia Social, Madrid: Eudema, 1994). \\nJ. Francisco Morales is Professor of Social Psychology at the Department \\nof Social Psychology, of the Spanish Open University at Madrid (UNED). \\nHe has done research on leadership, social identity and its relation to \\nindividualism-collectivism. Recent publications include 'Three Portraits of \\nSocial Psychology in Europe', Proceedings o\\nf\\n the Tenth Meeting o\\nf\\n Cheiron \\n( 1992), 'Leadership in Two Types of Health-Care Organizations' (with F. \\nMolero), in J.M. Peiro et al. (eds), Work and Organizational Psychology: \\nEuropean Contributions o\\nf\\n the Nineties (London: Taylor & Francis, 1995), \\nTreatise o\\nf\\n Social Psychology I: Basic Processes ( Tratado de Psicologia \\nSocial I: Procesos Basicos (with M. Moya); Madrid: Sintesis, 1997) and \\nSocial Psychology ( Psicologia Social) (with M. Moya et al.); Madrid: \\nMcGraw-Hill, 1 994). \\nMiguel Moya is Professor of Social Psychology at the School of Psychology \\nof the University of Granada, Spain. His main research interests are \\ngender, interpersonal and intergroup relations, prejudice and stigma. \\nRecent publications include Treatise o\\nf\\n Social Psychology I: Basic Processes \\n( Tratado de Psicologia Social I: Procesos Basicos), (with J.F. Morales; \\nMadrid: Sintesis, 1 997), Psychology and Gender (Psicologia y Genero) (with \\nM.e. Gomez; Granada: Universidad de Granada, 1997), and Social \\nPsychology (Psicologia Social), (with J.F. Morales et al.; Madrid: \\nMcGraw-Hill, 1994). \\nGabriel Mugny is Professor of Social Psychology at the Faculty of Psy­\\nchology and Educational Sciences, University of Geneva, Switzerland. His \\nmain research interests are social influence, intergroup relations, social \\nidentity, cognitive development and social representations. Recent \\npublications include Psychologie Sociale et Developpement Cognitif (with \\nW. Doise; Paris: Armand Colin, 1997), La Psychologie Sociale, Vol. i: \\nRelations Humaines, Groupes et In\\nfluence Sociale (with D. Oberle and J.L. \\nBeauvois; Grenoble: Presses Universitaires de Grenoble, 1995), The Social \\nPsychology o\\nf Minority Influence (with J.A. Perez; Cambridge: Cambridge \\nUniversity Press, 1991) and In\\nfluences Sociales (with J.A. Perez et al.; \\nNeuchatel: Delachaux et Niestle, 1993). \\nPenelope Oakes born in England, is currently Senior Lecturer in Social \\nPsychology at the Australian National University in Canberra, Australia. \\nShe has published widely on stereotyping and group processes, including \\n\\n\\nXlV \\nSocial identity: international pers\\npectives \\nStereotyping and Social Reality (with A. Haslam and J.C Turner; Oxford: \\nBlackwell, 1 994). She is also co-editor of The Social Psychology o\\nf\\n Stereo­\\ntyping and Group Lif\\ne (with R. Spears et al.; Oxford: Blackwell, 1997). \\nDario Paez, born in Chile, is currently Professor of Social Psychology at \\nthe Department of Social Psychology, School of Psychology, Basque \\nCountry University, Spain. He has done research mainly on social identity, \\nemotions and health, collective memory and social representations. He has \\npublished 'Inhibition, Alexythimia and Health' (with N. Basabe et al.) in J. \\nPennebaker (ed.), Emotion, Disclosure and Health (Washington: APA, \\n1995), and Collective Memory o\\nf\\n Political Events (with J. Pennebaker and \\nB. Rime; Mahwah: Erlbaum, 1997). \\nJuan Antonio Perez is Professor of Social Psychology at the University of \\nValencia, Spain. His main current research interests are in minority \\ninfluence and racism against gypsies. Recent publications include Influences \\nSociales (with G. Mugny et al.; Neuchatel: De1achaux et Niestle, 1993) and \\n'The Conflict Elaboration Theory of Social Influence', (with G. Mugny), in \\nE. Witte and J. Davis (eds), Understanding Group Behavior, Vo!' 2 \\n(Hillsdale, NJ: Erlbaum, 1990). \\nJose Miguel Salazar is Professor of Psychology at the Central University of \\nVenezuela, where he pioneered social psychology studies in the late 1950s. \\nCurrently he is Editor of the Revista Interamericana de Psicologial \\nInteramerican Journal o\\nf\\n Psychology. His main interests centre on issues of \\nnationalism, national identity and supra-nationalism, and he has done \\nresearch on the relationships of nationalistic attitudes to behaviour, Latin \\nAmerican identity and value changes, and, more recently, the construction \\nof national and quasi-national identities. He has published Psychological \\nFoundations o\\nf\\n Nationalism (Bases Psicol6gicas del Nacionalismo, Mexico: \\nTrillas, 1987) and numerous articles and chapters on the subject. \\nJohn Schopler is Professor of Social Psychology, Department of \\nPsychology, University of North Carolina, Chapel Hill, USA. His main \\nresearch interests are in intergroup relations, social power, interdependence \\ntheory and conflict resolution. His recent publications are 'The \\nDiscontinuity Effect in Interpersonal and Intergroup Relations: Generality \\nand Mediation' (with CA. Insko), European Review o\\nf\\n Social Psychology \\n(1992), 'The Role of Communication in Interindividual-Intergroup \\nDiscontinuity' (with CA. Insko et al.), Journal o\\nf\\n Con\\nflict Resolution \\n(1993), 'Individual-Group Discontinuity: Further Evidence for Mediation \\nby Fear and Greed' (with CA. Insko et al.), Personality and Social \\nPsychology Bulletin (1992), 'The Survival of a Cooperative Tradition in the \\nIntergroup Discontinuity Context' (with CA. Insko et al.), Motivation and \\nEmotion ( 1994), 'The Role of Identifiability in the Reduction of \\nInterindividual-Intergroup Discontinuity' (with CA. Insko et al.), Journal \\no\\nf\\n European Social Psychology (1995), 'Memory for and Experience of \\nDifferential Competitive Behavior of Individuals and Groups' (with M.J. \\n\\n\\nList o\\nf\\n contributors \\nxv \\nPemberton and CA. Insko), Journal o\\nf\\n Personality and Social Psychology \\n(1996), and Intergroup Cognition and Intergroup Behavior (with C \\nSedikides and CA. Insko; Hillsdale, NJ: Erlbaum, in press). \\nCarmencita Serino is Lecturer in Social Psychology at the Department of \\nPsychology of the University of Bari, Italy. Her main research topics are \\nsocial categorization, intergroup relations, similarity judgement, self-others \\ncomparison and social representations. Her recent publications include \\nAsymmetry Phenomena in the Interpersonal Comparison (with L. Arcuri; \\nNaples: Liguori, 1992), 'Information Processing and Self-Others Compar­\\nison: Social Cognitive Extensions of a Non-Metrical Approach to Simi­\\nlarity', in M.C Hurtig et al. (eds), Studies on the Self and Social Cognition \\n(New York: World Scientific Publishing, 1993), 'Mental Illness in Psycho­\\nsocial Research: Processes of Explanations and Implicit Personality \\nTheories' ('La Malattia Mentale Nella Ricerca Psico-Sociale: Processi di \\nSpiegazione e Teorie Implicite della Personalita'), in G. Bellelli (ed.), The \\nOther Illness: How Society Thinks about Mental Illness (L'Altra Malattia: \\nComme la Societa Pensa la Malattia Mentale, Naples: Liguori, 1994), \\nSingular Worlds: Socio-Cognitive Perspectives on Identity ( Mondi al \\nSingolare: Prospective Socio-Cognitive Sull'identita) (with J.C Deschamps; \\nNaples: Liguori, 1996), and 'Similarities and Differences in the Social \\nRepresentations of the Handicapped' (,Somiglianza e Differenza nella \\nRappresentazione Sociale dell'Handicap', in G. Petrillo (ed.), Social Psy­\\nchology o\\nf\\n Health ( Psicologia Soziale della Salute, Naples: Liguori, 1997). \\nLaurie A. Taylor is the Manager of Community Evaluation Processes at the \\nUnited Way of the Greater Dayton Area in Dayton, Ohio, USA. Her \\nresearch interests include group decision-making, intergroup perceptions \\nand identity formation, conflict management and applied outcome \\nevaluation. She has most recently published 'Recognizing Group Diversity: \\nIndividualist-Collectivist and Autonomous-Relational Social Orientations \\nand Their Implications for Intergroup Processes' (with R.J. Brown et al.), \\nBritish Journal o\\nf\\n Social Psychology (1992). \\nJohn C. Turner, born in England, is Professor of Psychology and Head of \\nDivision, Division of Psychology, Australian National University. His \\nresearch interests are in social identity, intergroup relations, group \\nprocesses and social cognition, particularly from the perspective of self­\\ncategorization theory. He collaborated with Henri Tajfel at Bristol in the \\n1970s in the development of social identity theory and is the originator of \\nself-categorization theory. Recent publications include Stereotyping and \\nSocial Reality (with P.J. Oakes and S.A. Haslam; Oxford: Blackwell, 1994), \\n'Social Identity, Interdependence and the Social Group: A Reply to Rabbie \\net aI.' (with R.Y. Bourhis), in W.P. Robinson (ed.), Social Groups and \\nIdentity (Oxford: Butterworth Heinemann, 1996), and 'The Socially Struc­\\ntured Mind' (with P.J. Oakes), in CA. McGarty and S.A. Haslam (eds), \\nThe Message o\\nf\\n Social Psychology (Oxford: Blackwell, 1997). \\n\\n\\nXVI \\nSocial identity: international pers\\npectives \\nLaura Vega has done graduate studies in Social Psychology at the Depart­\\nment of Social Psychology of the Spanish Open University at Madrid \\n(UNED). She is currently working for her doctoral dissertation and doing \\nresearch on the relations between individualism and identity differentiation. \\nShe is co-author of a paper on 'Individualism, Collectivism and Social \\nIdentity' (,Individualismo, Colectivismo e Identidad Socia!'), included in a \\nmonographic volume on 'Social Identity' published by the Revista de \\nPsicologia Social ( 1992). \\nStephen Worchel is Dean of the College of Arts and Sciences and Professor \\nof Psychology at the University of Southern Maine. His research areas \\ninclude intergroup relations, group dynamics, conflict resolution and cross­\\ncultural psychology. Recent publications related to social identity include \\nthe following: 'The View from Below: Intergroup Relations from the \\nPerspective of the Disadvantaged Group' (with H. Rothgerber), Journal o\\nf\\n \\nPersonality and Social Psychology ( 1997); 'A Dynamic Approach to Group \\nDynamics', in P. Nye and A. Brower (eds), What's Social about Social \\nCognition? (Newbury Park: Sage, 1996); 'Changing the Stereotype of \\nStereotypes' (with H. Rothgerber), in R. Spears et al. (eds), The Social \\nPsychology o\\nf\\n Stereotyping and Group Lif\\ne (Oxford: Blackwell, 1997); and \\nCon\\nflict between People and Groups (with J. Simpson; Chicago: Nelson­\\nHall, 1993). \\n\\n\\nPreface \\nThis volume deals with the issue of identity, the concern over which is far \\nfrom being something new. In fact, it can be traced to classical Antiquity. \\nIdentity is present in the famous 'Know Thyself' precept, inscribed on the \\nfrontispiece of the Delphi Temple. It is also the target of many of the \\nfamous Socratic questions and the stimulus for Plato's well-known reflec­\\ntions. It can be considered, without exaggeration, an issue which has \\noccupied humankind since the beginning of civilization. \\nIt also focuses on categorization or, more explicitly, on the analysis of \\nhow we perceive other people (stereotypes) and how we decide to behave \\ntowards them (prejudice, discrimination), and, in general, on the under­\\nstanding of our readiness to come to grips with the social environment. All \\nthese problems have a long tradition, but they cannot be taken as old­\\nfashioned or as belonging to the past, since they are pervasive in our \\ncurrent, so-called 'civilized' societies. \\nAdmittedly, it would be illusory to try to find in the Socratic-Platonic \\nanalysis of the Delphic formula the point of departure of a current \\npsychosocial analysis of the personal and social self. Only in the last 20 \\nyears or so have social psychologists provided such an analysis, framed in \\nthe recent, but already well-known theories of social identity, and social \\nand self-categorization. It is the purpose of this volume to offer an updated \\nand balanced view of these fairly new developments, in the expectation that \\nit will foster a better understanding of some of the more urgent and central \\nproblems of our time. \\nThe latest elaborations of the theories and the most recent research \\ninspired by them are especially attended to. A detailed summary of them \\nwould exceed the length usually allowed to any preface. However, we can, \\nat least, highlight three of their main contributions, closely linked to three \\nsalient features of the theories mentioned. \\nIn this respect, it should be stressed that one of the most important assets \\nof the social identity and self-categorization theories is the building of an \\noriginal theoretical corpus which has provided social psychology with new \\ntheoretical insights. \\nUndoubtedly, before Tajfel's formulation of social identity theory, there \\nwere several psychosocial approaches to the study of identity. However, \\nperhaps due to their low theoretical coherence, they did not succeed in \\nmaking social identity a crucial concept of the discipline of social psy­\\nchology. The same can be said of self-categorization theory. \\n\\n\\nXVlll \\nSocial identity: international perspectives \\nThis theoretical corpus was not without its appropriate empirical content, \\nsince the theoretical concepts and propositions were elaborated through an \\nintense dialogue with certain psychosocial phenomena demanding explana­\\ntion, such as intergroup discrimination, the effect of stability and legitimacy \\nof status differences, the processes of social comparison and of group \\ncompetition, among others. In providing such an explanation, both theories \\nbecame more and more influential and began to play a leading role within \\nsocial psychology. \\nGoing still a step further, they widened their scope to accommodate \\nother psychosocial phenomena. Let's take the example of the group con­\\ncept. Tajfel himself had suggested the need for its reformulation. However, \\nthis was only done by Turner several years later. The covariation hypoth­\\nesis, formulated by Deschamps, and linking intra-\\nand intergroup \\ndifferentiation, is just another example of the success of the theories in \\ntheir efforts to cover other psychosocial phenomena. Besides the group \\nconcept and the covariation hypothesis, the following processes, among \\nothers, are worth mentioning: influence, group polarization, collective \\nbehaviour, the impact of minorities, prejudice, stereotypes and leadership. \\nThis speaks in favour of the fruitfulness of both theories. \\nThese three features correspond approximately to the stages of develop­\\nment of the theories, but, at the same time, all three have been, and still \\nare, open and subject to changes and modifications. The fact that there is a \\nmutual influence between them cannot be forgotten either. Thus, the \\nintensive study of intergroup discrimination has helped to reformulate the \\nearly concepts, throwing light on the functions of comparison dimensions. \\nBy the same token, the extension of the theory to the group concept has \\nallowed Turner to replace the early interperson-intergroup continuum by a \\nnew, more accurate interperson-group one. \\nIn recent years, the sustained vitality of the theories has allowed them to \\ngain new empirical content with the corresponding increase in the amount \\nof explanations of phenomena of interest for social psychoiogy. As a result, \\nthey have become both more complex internally and more sophisticated in \\ntheir theoretical formulation. The contributors to this volume, among them \\nsome of the original formulators of both theories, have tried to give them \\neven more impulse. They raise new issues, examine critically some crucial \\npoints, and explore new areas of research. Their work will serve to expand \\nthe theories and, we hope, to persuade others to join the effort. \\nThe overall aim of this volume is, then, to demonstrate the breadth of \\nscope of issues related to social identity. Indeed, one of the most unique \\nfeatures of the area is that it has found a warm welcome in a variety of \\ncamps. As such, we begin the volume with chapters addressing identity and \\nthe self. We move from here to contributions that deal with identity in the \\ngroup and intergroup context. And the final contributions venture into the \\nrealm of cultural influences on social identity. \\nBut the breadth of social identity is not only demonstrated by its con­\\ntribution across the social spectrum. It is also a theory that has captured \\n\\n\\nPre\\nf\\nace \\nxix \\nthe imagination of investigators in diverse lands with different cultural \\nperspectives. In an effort to represent this geographical scope, the chapters \\nin this volume represent a mosaic from five continents (Asia, Europe, \\nNorth America, South America, Australia) and seven countries (Israel, \\nAustralia, Spain, Great Britain, the United States, Poland, Venezuela, \\nSwitzerland). Although the specific focus of each chapter represents a \\ncombination of the investigator's interest and his or her cultural founda­\\ntion, the social identity framework provides a common language that \\neclipses geographical, political and cultural boundaries. \\nWe'd like to extend our appreciation to our colleagues who have com­\\nmented on specific chapters. Special thanks is given to Dawna Coutant, \\nJohn Turner and William Webb. Jennifer Ford, Kim Cozzi and Jon Iuzzini \\nprovided invaluable editorial assistance in preparing the manuscript for \\npublication. And above all, we extend our appreciation to Ziyad Marar \\nand Lucy Robinson, who were brave enough to accept such an ambitious \\nproject and whose expertise and gentle guidance moved it from the idea \\nstage to book form. \\n\\n\\n\\n\\nPART I \\nREPRESENTATIONS OF \\nSELF AND GROUP \\n1 \\nRegarding the Relationship Between \\nSocial Identity and Personal Identity \\nlean-Claude Deschamps and Thierry Devos \\nThe problems to be discussed in this chapter can be considered the central \\nissue regarding both identity and a general perspective of social psychology; \\nthe theme of this chapter describes the relationship between two concepts \\ncalled social identity and personal identity. \\nFew concepts are as polysemic as identity. The idea of identity appears in \\nthe first texts of Antiquity (the famous 'Know Thyself' of Delphi) and is \\nused in both everyday language and scholarly discourses; it is also the \\nconcern of many scientific disciplines. There are many synonyms or near \\nsynonyms for identity (one can mention at random: oneself, t, the self, we, \\nself-perception, self-image, self-representation or self-awareness, the ego, \\nand so on). The self can refer to a familiar subject, but also to a social \\nagent's action. It can also refer to active processes and mechanisms that rule \\nbehaviour. Therefore, our remarks are limited to specific aspects of identity. \\nSocial identity and personal identity \\nIdentity is a central concept in social psychology, probably because it is one \\nof the main concerns of that field. As Codol wrote (1979, p. 424), that \\nmajor concern - which has been the subject of very old debates regarding \\nphilosophy and religion as well as ideology and the humanities - simply \\nrefers to the conflict of affirmation and individual necessity versus affir­\\nmation and collective necessity, the search for personal identity versus the \\nsearch for collective identity, what constitutes individual difference versus \\nwhat constitutes similarity to others (same things constitute both), social \\n\\n\\n2 \\nSocial identity: international perspectives \\nvisibility versus conformity, in short the conflict of the individual versus the \\ngroup. And that concern, which can take many forms, refers to the \\nrelationship between the concept of the individual and the collective, often \\nviewed as conflicting. This relationship is central in social psychology. In \\nother words, it is an opposition between the diverse and the homogeneous, \\nor, as Robert Pages notes, it is the recurrent theme of the same and the \\nother. \\nStudies about self-concept and identity also deal with the opposition \\nbetween the individual and the social; and that opposition is codified \\naccording to the distinction made between personal identity and social \\nidentity. This distinction was already made - although not explicitly - at \\nthe beginning of modern psychological thinking. At the end of the last \\ncentury, William James ( 1890) wrote about the distinction between the I \\nand the me and came up with the idea of duality in self-representation. \\nNowadays, this distinction continues to have a great impact on studies on \\nidentity. Mead ( 1934) goes deeper into the question with his idea that the \\nself consists simultaneously of a sociological component (the me) - which \\nwould only be an internalization of the social function - and a more \\npersonal component (the I); the developments of that definition of the self \\nhave gradually led to the distinction between social identity and personal \\nidentity. Of course, everyone agrees that identity is both personal - in the \\nsense that it is 'situated within' a person - and social, inasmuch as its \\nprocesses of individual formation are social, but the individual is indeed \\ncharacterized by two poles - one is psychological and the other sociological \\n- that have no connection to each other. At first, the individual/society \\ndichotomy seemed outmoded, particularly with Mead's idea of conversation \\nbetween the I and the me, according to which the individual is a continuous \\ncreation of society, and society an unremitting creation of individuals. But \\nwe will see that the two poles, the individual and the social, tend to be \\nconsidered opposite. \\nLet us first make clear the concepts of social identity and personal \\nidentity. These concepts are based on the idea that every individual is \\ncharacterized by social features which show his or her membership of a \\ngroup or a category, on the one hand, and by personal features or indi­\\nvidual characteristics which are more specific, more idiosyncratic, on the \\nother. \\nThe former features define the social identity of a person. Belonging to a \\ngroup or to a given social category is the most important. Social identity is \\ncodified as the part of the self which refers to cognitions ensuing from \\nsocial ecological positions (Sarbin & Allen, 1968). Those who have similar \\npositions and common backgrounds have similar social identities, therefore \\nsocial identity does indeed refer to the similarity pole. However, the feeling \\nof belonging to a group and the phenomena of identification are only \\npossible in connection with groups or categories one does not belong to. \\nThus social identity refers to the fact that the individual perceives him- or \\nherself as similar to others of the same background (the we), but social \\n\\n\\nSocial identity and personal identity \\n3 \\nidentity also refers to a difference, to a specificity of that we in connection \\nwith members of other groups or categories (the them). We have then a \\ndouble motion which combines ingroup similarities and intergroup or \\ncategorial differentiation. The stronger the identification with a group, \\nthe more significant the differentiation of that group from other groups \\nwill be. \\nMore personal features or specific character attributes of each individual \\nrefer to personal identity and the idea that each individual is a unique \\ncombination of features which make him different from others, and which \\nexplains his uniqueness and the fact that he is specific. Personal identity is \\nnot well defined. However, it indicates how an individual is aware of his \\ndifference with respect to others. That feeling can only be experienced in \\nrelation to others and personal identity refers to the fact that the individual \\nperceives himself as identical to himself; in other words he is the same in \\ntime and in space, but that is also what specifies him and marks him out \\nfrom others. Personal identity is what makes you similar to yourself and \\ndifferent from others. \\nOf course, one could debate this postulate regarding self-permanence as a \\nrelatively stable and durable concept held by everyone (see, for instance, \\nGoffman [1956] with his descriptions of the individual who plays different \\nroles for different audiences and Gergen [1965, 1982] with his idea of \\nfluidity of the self). It is therefore necessary to speak of identities in the \\nplural, considering that every social agent - whether individual or collective \\n- can actualize, mobilize or produce identities according to the context. \\nThis also leads us to the question of the historical universality of this feeling \\nof identity. But that does not change fundamentally the problem that \\nconcerns us. The fact that the feeling of identity is socially built into every \\nindividual and modulated in every social position and that it is an \\nideological construction does not mean that it is not real and efficient; it is a \\nguide for individuals' actions and it is essential for the functioning of our \\nsociety. \\nNow, we must see how the articulation, the organization, the synthesis of \\npersonal and social aspects of identity are conceptualized. Social identity \\nrefers to a feeling of similarity to (some) others; personal identity refers to a \\nfeeling of difference in relation to the same others. This distinction between \\npersonal identity and social identity is indeed only an example of duality \\nbetween the individual and the collective or difference and similarity. On \\nthe basis of generally accepted definitions of what is conventionally called \\nsocial identity and personal identity, we have to deal with the conflict \\nbetween the individual and the social. This lack of differentiation on a \\ncertain level - in other words social identity - and this differentiation -\\npersonal identity - are considered most of the time as two opposite poles \\nbetween which human behaviour oscillates. This combination of similarities \\nand differences is problematic. It is hard to conceive how one can feel very \\nsimilar and very different at the same time; it seems as though one has to \\nchoose to be one or the other. \\n\\n\\n4 \\nSocial identity: international perspectives \\nSocial identity theory \\nThe studies carried out by the 'Bristol School' emphasize the link between \\nsocial identity and personal identity. In order to understand how the con­\\ncept of identity is theorized and used in that perspective, first of all a certain \\nnumber of developments regarding the process of categorization must be \\naddressed. \\nCategorization and social categorization \\nCategorization refers to psychological processes which tend to organize the \\nenvironment into categories or groups of persons, objects, events (or groups \\nof some of their characteristics) according to their similarities, their equi­\\nvalences concerning their actions, their intentions or behaviour (Tajfel, \\n1972a, p. 272). Categorization divides the environment into groups of sub­\\njects which are or seem to be similar according to certain criteria; these \\nsubjects are or seem to be different from other subjects according to the same \\ncriteria. Categorization plays a specific part in structuring the environment \\nby systematizing (dividing and organizing) and thus by simplifying it. \\nOne of the major effects of categorization and the simplification it implies \\nis that it emphasizes the differences between categories (contrast or cogni­\\ntive differentiation effect) and the similarities within the same category \\n(assimilation or cognitive stereotype effect). Each element of the category is \\ncharacterized by the features shared by all the subjects of that category. \\nThus, a category is a group of elements which have in common one or \\nseveral features. Cognitive apprehension is a simpler apprehension of \\nreality. The emphasis on similarities and differences is one of the aspects of \\nthat simplification. Dealing with a psychological process, there is no reason \\nat first sight to think that this process does not also play a role in the \\nperception of the others, who are characterized by the fact that they belong \\nto social categories. The fact that the elements referred to are persons and \\nthat we deal with social categories does not matter; the process does not \\nchange in that perspective. Actually, a social category is principally a \\ncognitive category, where individuals who are supposed to share one or \\nseveral features are grouped together. Stereotypes are - on a social level -\\nthe expression of the attribution of features shared by different members of \\na group without taking into account the interindividual differences. The \\nclassification used as a basis for the stereotypes minimizes the differences in \\nfeatures among the members of a social group. As Doise noted in 1976, \\ncategorization is then a psychological process that simplifies the perception \\nof physical and social worlds and the way the individual organizes the \\nsubjective perception of his or her environment. \\nBut apart from the homology between the perception of physical environ­\\nment and social environment, social categorization - unlike physical stimuli \\ncategorization - implies that subjects are themselves inside a system of \\ncategories. They are subjects and objects of categorization at the same time. \\n\\n\\nSocial identity and personal identity \\n5 \\nSocial categorization minimizes the in-category differences and exaggerates \\nthe differences between categories. In addition, these differences are used for \\nevaluation. When people hold the representation of a dichotomous universe \\nwhere you can belong only to one category, they end up having a dis­\\ncriminating attitude towards the members of other categories. \\nSocial categorization and identity \\nHow can discrimination between groups be explained? Unlike in Sherif's \\ntheory of the objective conflict of interest between groups (see, for instance, \\n1966), competition is not a sufficient explanation for discrimination between \\ngroups. For Tajfel (l 972a) the idea of belonging to two different groups \\nleads to discrimination in favour of the group one belongs to. But then, one \\nmust refer to concepts other than competition to explain the phenomenon \\nof discrimination. This is when the concept of identity is introduced. It is \\nimportant to recall the main points of social identity theory (see in \\nparticular Tajfel & Turner, 1979). \\nAccording to Tajfel ( l972a), social identity is conceptualized as being \\nconnected to the individual's knowledge of belonging to a certain social \\ngroup and to the emotional and evaluative signification that results from \\nthis group membership. Thus, it is through their belonging to different \\ngroups that individuals acquire a social identity defining their specific posi­\\ntions in society. But belonging to a given group contributes to the devel­\\nopment of a positive social identity only if the characteristics of that group \\ncan be compared favourably to other groups. In comparing the groups, \\nindividuals tend to establish a difference which is in favour of their own \\ngroup. \\nFrom 1 974 on, Tajfel adds an a priori distinction between two opposite \\npoles in social behaviour. At one extreme (which can probably not be found \\nin such a 'pure' form in 'real life'), we find interactions between two or more \\nindividuals. These individuals are totally determined by their interpersonal \\nrelations and by their individual characteristics and not affected at all by the \\ndifferent groups or social categories they belong to. At the other extreme, \\nthere are interactions between at least two individuals (or groups of \\nindividuals) which are totally determined by their respective membership of \\ndifferent groups or social categories and not affected at all by inter­\\nindividual relationships among the concerned persons. Here again, there are \\nfew chances of encountering such 'pure' forms of this extreme in 'real' social \\nsituations (Tajfel & Turner, 1979). \\nThe two poles refer to interpersonal attitudes on the one hand, and to \\nintergroup behaviour on the other. As Turner (198Ja) noted, personal \\nidentity corresponds to interpersonal behaviour - which means differ­\\nentiation between the self and the others; and social identity corresponds to \\nintergroup behaviour - which means differentiation between groups or \\nbetween 'we' and 'them'. \\n\\n\\n6 \\nSocial identity: international perspectives \\nWhen identification with a group is emphasized, one changes from the \\ninterpersonal pole to the intergroup pole. The introduction of this distinc­\\ntion between interpersonal and intergroup phenomena is important for \\nthings are not going to be as simple as in Tajfel's early theory, according to \\nwhich this distinction motivates a need for a positive difference in favour of \\none's own group compared to other groups, the creation of that difference \\nand the emphasis on it. What really matters it the need for self-esteem, the \\n'desire' for a positive self-evaluation, the need for individual positivity. The \\nbasic postulate is that 'individuals tend to maintain or increase their self­\\nesteem; they try to reach a positive self-concept' (Tajfel & Turner, 1979, \\np. 34). Thus, the individual tries to preserve or reach a positive self-image, \\nand if she cannot evaluate herself according to the interpersonal behaviour \\npole - which lies in evaluating oneself directly by comparing oneself to the \\nothers - she can satisfy her desire for positive self-evaluation through social \\ncompetition between groups (which must be considered as a groups com­\\nparison) with this tendency to introduce a positive difference in favour of \\none's own group in relation with other groups. \\nIn fact, social identity theory is founded on two bases: one is cognitive \\nand the other is motivational. As mentioned before, it is cognitive insofar as \\nthe categorization process leads the subjects to overestimate intergroup \\ndifferences and to underestimate ingroup differences. It is motivational \\ninsofar as what motivates discrimination is the need for self-esteem - or \\nself-respect, according to Emler and Hopkins (1990). \\nTurner (1975a, 1975b), who belongs to that school of thought, concludes \\nthat subjects act in terms of differentiation between groups only when this \\ndifferentiation is the only means for them to achieve a positive self­\\nevaluation, a positive identity. If they can act in terms of a differentiation \\nbetween self and others, they do not resort to the intergroup 'bias'. This only \\noccurs when identification with the group provides for a positive identity. In \\nother words, categorization on its own would not be enough to observe a \\ndifferentiation between groups; the tendency to establish a positive differ­\\nentiation between self and others underlies the social categorization process. \\nTo put it differently, if the individual accedes to a positive identity through \\nhis identification with a group, he establishes differentiation between groups \\nbut no longer tends to establish differentiation with the other members of his \\ngroup. If the individual is able to differentiate himself from the others \\ndirectly - acceding in this way to a positive self-evaluation - he no longer \\nestablishes differentiations with other groups which he knows. \\nThe intention here is not to develop the limits of these studies or to \\nquestion aspects of social identity theory (see Deschamps, 1984, 1991, for \\nmore details). However, it must be underlined that on the basis of this \\nasserted dichotomy between interindividual and intergroup behaviour, one \\ncan say that the stronger social identity is, the less important is personal \\nidentity, and the more prominent a personal identity is, the less the \\nindividual needs a social identity, since social identity and personal identity \\nsatisfy the same need for a positive self-image. According to the social \\n\\n\\nSocial identity and personal identity \\n7 \\nidentity model, we can predict that when the salience of belonging to a \\ngroup increases, every subject identifies him- or herself more and more with \\nthe in-group and that will reduce differentiation between the self and the \\ningroup while exacerbating differentiation between groups. In other words, \\nwhen identification with a group increases, one changes from the inter­\\npersonal pole to the intergroup pole, and vice versa. This dichotomy solves \\nthe problem of differences and similarities since only one pole is possible, \\nswitching from one to the other. This is indeed what Turner refers to when \\nhe introduces the concept of 'depersonalization'. For him, the factors which \\nintensify the salience of the ingroup/outgroup categorization tend to \\nincrease identification (similarity, equivalence, interchangeability) between \\nthe self and the other members of the group (and differentiations with \\nmembers of another group). Consequently, these factors depersonalize the \\nself according to stereotypes determining the group. Depersonalization \\nrefers to the 'auto-stereotypy' process through which individuals come to \\nconsider themselves above all as interchangeable models of a category \\nrather than as unique individuals who are distinct from the others (Turner, \\n1987, p. 50). As we can see, both analyses - the one based on difference and \\nthe one based on similarity - depend on each other negatively. The dicho­\\ntomy postulated between interindividual and intergroup behaviour, which \\nactually refers to the duality of self-awareness and the opposition between \\npersonal identity and social identity solves - apparently - the problem of \\nsimilarities and differences while involving two mutually exclusive poles. \\nSelf-de\\nfinition and its diff\\nerent levels \\nTurner ( 1987) tries to explain the opposition between the psychological \\naspect of identity - which refers to an individual's unity with its constel­\\nlation of specific features - and the sociological aspect of identity - which \\nrefers to the idea of group membership and similarities - by distinguishing \\nthree levels in self-definition with his self-categorization theory: \\n(a) a supra-order level where the self is defined as a human being: it refers \\nto a human identity based on comparisons between the species (simi­\\nlarity with the human race, differences with other forms of life); \\n(b) an intermediate level of self-definition where the self is defined as a \\nmember of a group (ingroup similarities and intergroup differences): it \\nrefers to a social identity based on intergroup comparisons (in-species); \\n(c) a subordinate level of self-definition where the self is defined as a \\nunique being (differentiation between the self and the others inside a \\ngroup); it refers to a personal identity based on interpersonal (in-group) \\ncompansons. \\nHowever, when individuals situate themselves at a level of categorization, \\nthey obviously have to occlude the two other levels. As a matter of fact, \\nthere is a fundamental antagonism between the salience of one self-categorization \\nlevel and the other levels: the salience of one of the categorization levels engenders \\n\\n\\n8 \\nSocial identity: international perspectives \\nin-class similarity and interclass differentiation which reduces or inhibits the \\nperception of in-class differences and interclass similarities on which inferior and \\nsuperior levels of categorization are based. (Turner, 1987, p. 49) \\nTurner adds that if, for simplification purposes, the supra-orderly categor­\\nization level is ignored, the (social) intermediate level and the (personal) \\nsubordinate level are linked by a reverse relation. Obviously, the intro­\\nduction of these levels does not solve the opposition between the collective \\nand the individual problem. Their relations remain negatively dependent \\n(self-definition in terms of categorial membership - both on the supra­\\norderly and the intermediate levels - versus individual specificity). \\nIn fact, as a whole these studies give the impression that it is difficult to \\nget rid of this opposition between the individual and the collective, and that \\nwe are trapped with a dichotomous analysis according to which similarity \\nand difference, social identity and personal identity, are two poles which \\ndepend on each other negatively. Having questioned the link between \\ningroup resemblance and intergroup divergence, we shall now present \\nanother series of studies in order to go beyond this opposition between \\nindividual and collective forms. \\nThe categorization model, like social identity theory or self-categorization \\ntheory, states that the same factors increase differences between groups and \\nsimilarity or homogeneity inside the groups. However, this especially \\napplies to outgroups and the way they are perceived and treated. The \\nindividual's own group would be perceived in a much more heterogeneous \\nway. In fact, a certain number of studies carried out over the last 1 5  years \\nillustrate the fact that the individual believes that his or her own group is \\nmore differentiated than other groups. This means that the group one \\nbelongs to can be perceived as heterogeneous by its members. This tendency \\nto consider one's own group as relatively less homogeneous than a group \\none does not belong to (or the fact that a group tends to be considered more \\nheterogeneous by its own members and not so by individuals who are not \\npart of it) has been called the 'outgroup homogeneity effect'. This phenom­\\nenon has been explained in many ways (some explanations are based on \\nmotivation, in terms of beliefs, and refer to the self as a special category, \\nrecalling the way information about ingroup and outgroup is processed and \\ncodified in the memory; see Ostrom & Sedikides, 1992). But this is not \\nessential here. What is important is that the concept according to which \\ndifferentiation between groups necessarily means convergence within groups \\nhas been questioned and is no longer the rule. \\nCovariation between interindividual and intergroup differentiation \\nhypothesis \\nA number of studies carried out in the 1970s already stated that emphasis \\non similarities inside a group does not necessarily mean emphasis on \\n\\n\\nSocial identity and personal identity \\n9 \\ndifferences between groups (for instance, Deschamps, 1972- 1973; \\nDeschamps & Doise, 1978). These observations, which were astonishing \\nat a time when categorization effects were taken for granted, indicated \\n(Deschamps, 1979) that the simple representation of belonging to two \\ndifferent groups could cause discrimination between groups in favour of \\none's own group but it would also engender differentiation between the self \\nand the rest of one's group. When the representation of a dichotomized \\nworld is not prominent or relevant in a situation, there is no discrimination \\nbetween groups - which is obvious. Moreover, in certain conditions, the \\ndifferentiation between self and others could be less important than that \\nbetween the self and the rest of the membership group when belonging to a \\ngroup and dichotomy in groups are significant criteria in defining a \\nsituation. \\nThis has led us to a new perspective on similarities and differences: the \\nintergroup and in-group differences covariation. Instead of considering \\ninterindividual and intergroup differences as two extremes of a continuum, \\nand that they are in this way mutually (at least partly) exclusive, we assume \\nthat in certain conditions the stronger the identification with the group, the \\nmore important is interindividual differentiation within the group. Codol \\n(for instance, 1975) underlines this issue with what he calls the phenomenon \\nof 'superior conformity of the self': the more an individual conforms to the \\nstandards of a group (and the more she identifies herself with that group), \\nthe more she will tend to consider herself different from the other members \\nof the group, believing that she corresponds to the standards better than the \\nothers. \\nHere is the central concept of the interindividual and intergroup differ­\\nentiation model, in a few words. A general process of cognitive centrism \\nappears when individuals are induced with the representation of a dichoto­\\nmized world, divided into two mutually exclusive categories. According to \\nthis representation, both ingroup favouritism or intergroup differentiation \\n(which can be called sociocentrism) and autofavouritism or differentiation \\nbetween self and others (which can be called egocentrism) would increase \\nwhen categorization is emphasized. \\nThe aim of a certain number of our studies was to test this postulate \\n(see Deschamps, 1982a, 1983- 1984, 1984; Deschamps & Lorenzi-Cioldi, \\n198 1 ;  Deschamps & Volpato, 1984). The experiments carried out in these \\nstudies are similar to the ones used in Tajfel's minimal group paradigm and \\nrefer to the changes introduced into that paradigm by Turner in 1975. An \\noutline of the surveys will suffice here (Deschamps, 1984). One of them was \\ncarried out in a girls' college. Using a sample of 88 girls, aged between 16 \\nand 20, half of the subjects were divided into two groups (categorization \\ncondition), apparently according to their preference for the works of two \\nmodern painters (in reality, it was an arbitrary way of splitting them). The \\nother half was not divided into two groups and did not belong to any \\ngroup in terms of their artistic preferences (non-categorization condition). \\nIn accordance with our hypothesis, it was predicted that differentiation \\n\\n\\n10 \\nSocial identity: international pers\\npectives \\nbetween self and others as well as intergroup differentiation would be more \\nsignificant in the categorization condition than in the non-categorization \\ncondition. \\nConcerning differentiation between groups, the results confirmed our \\nexpectations: discrimination in favour of the ingroup in the categorization \\ncondition and of course no discrimination in the non-categorization condi­\\ntion. Concerning differentiation between self and others, in the categoriza­\\ntion condition, the differentiation is much more significant than in the \\nnon-categorization condition. In other words, when individuals are simply \\ninduced to the representation of an environment including their group and \\nanother one (which leads of course to differentiation between groups), it \\nincreases significantly autofavouritism 'bias' or differentiation between self \\nand others, compared to a non-categorization condition (without categor­\\nization induction). Emphasizing intergroup differentiation in an experiment \\n(in this example, by inducing or not the subjects with the representation of a \\ndichotomized world) is sufficient to increase the differentiation subjects set \\nbetween themselves and others. \\nBut more observations must be addressed. The contrary is also true. If \\ndifferences between self and others within a group are emphasized in an \\nexperiment, differentiation between groups increases simultaneously. \\nAnother survey was carried out in a high school with 1 12 boys and girls \\naged between 14 and 1 5  (see Deschamps & Volpato, 1984). As in the \\ncategorization condition of the experiment mentioned above, the subjects \\nwere placed into two groups, apparently according to their preference for \\nthe works of a contemporary music composer. Half of the subjects, the ones \\nin the individualization condition, expected that each individual would be \\ngiven exactly the same remuneration, according to the minimal group \\nparadigm. The other half, the ones in the fusion condition, expected that all \\nmembers of a group would receive the same remuneration, corresponding to \\nan average of points which all the subjects would give to the different \\nmembers of that group. Our expectations proved to be right. In the fusion \\ncondition, differentiation between self and others and discrimination in \\nfavour of one's own group are less than in the individualization condition. \\nWhen differentiation between self and others is emphasized - by focusing \\nnot on the groups as a whole, but on the fact that individuals are char­\\nacterized by their group memberships - differentiation between groups \\nincreases in the same way. \\nIn addition, the covariation model seems to apply especially to indi­\\nviduals belonging to privileged groups. Not only do differentiation between \\ngroups and ingroup homogeneity not necessarily go together, but the \\nsimultaneous variation of differentiation inside a group and between groups \\nalso depends on the relative status of the groups in question (Deschamps, \\n1 982b; see also Lorenzi-Cioldi, 1988). These are the conclusions of other \\nstudies. For instance, it has been shown (Deschamps, 1977) that when boys \\nhad to evaluate themselves as well as other boys and girls for different \\nassignments, they discriminated between boys and girls and between \\n\\n\\nSocial identity and personal identity \\nSI: x--x \\nPI: x---------------x \\nSI: x---------------x \\nPI: x---------------x \\n\\\" \\n\\\" \\n\\\" \\n\\\" \\n\\\"\\nSI: x-------x / \\n/ \\n/ \\n/ \\n/ \\nSocial identity \\nPI: x-------x \\naxis (SI) \\n-----Compromise -----\\nSI: x --x \\nPI: x --x \\n/ \\n,, W \\n. i:- / \\n,,<9\\n‚ \\n(similarity) \\nn\\no / \\n,,ƒ\\n-? \\n\\\"\\\"\\ni-\\n{!>\\n/ \\n\\\"\\n'.90,,.\\n \\n,(\\n..,0/ \\n'- -? \\n'- °\\n0 ' \\n/ \\n,, œ  \\n/ \\n,,'.9-\\nKey \\nSI: x ---------------x \\nPI: x--x \\nx--x \\nrestricted range of identity \\n(personal and/or social) \\nx -------x \\nmedium range \\nx---------------x \\nexpanded range of identity \\n(personal and/or social) \\nFigure 1 . 1  Personal and social dimensions in psychosocial identity \\n1 1  \\nthemselves and the other boys. Girls would discriminate between boys and \\ngirls, but unlike the boys they would not tend to discriminate between other \\nmembers of their own category. \\nIn conclusion, the results of these studies briefly mentioned underline the \\nrelevance of the simultaneous variation of differentiation between groups \\nand between self and others. However, it does not mean that under certain \\ncircumstances there is an opposition between personal identity and social \\nidentity. Obviously, relationships between the individual and the collective \\ncan be examined according to situations, cultures and societies. Never­\\ntheless, one must at least consider the possibility of simultaneity between \\nsimilarity and difference. Therefore, the postulated axis defining the inter­\\nindividual relations and behaviour pole versus the intergroup relations and \\nbehaviour pole should be reconsidered. In that prospect, similarity and \\ndifference, social identity and personal identity must no longer be con­\\nsidered as two poles of the same continuum which are negatively dependent. \\nThese different elements must rather be considered as two disconnected \\ndimensions that can be 'orthogonalized' in some way (see Figure 1 . 1). \\nConsequently, models based on Tajfel's work and the covariation \\nhypothesis can be explained at the same time. \\n\\n\\n12 \\nSocial identity: international pers\\npectives \\nNote \\nThis chapter was translated from the French by Mehr-Afarin Khosrowdad. It takes up and \\ndevelops arguments which have been put forth in some of the previous texts and quotes several \\npassages from an article published in 1991 by the first author in Cahiers Internationaux de \\nPsychologie Sociale called 'Identites, Appartenances Sociales et DiITerenciations Individuelles'. \\n\\n\\n2 \\nSocial Representations in Personal \\nIdentity \\nWillem Doise \\nIntroduction: Self as social representation? \\nSocial representations can be defined as organizing principles of symbolic \\nrelationships. They permit the positioning of the individual in relation with \\nsignificant social objects. The self certainly is such an object embedded in a \\nnetwork of relationships. Given its central position in communicative \\nsystems, personal identity is not just an idiosyncratic entity, it is one of the \\nmost important organizing principles of symbolic relationships. \\nIn order to validate the basic assumption of personal identity as a social \\nrepresentation, existence of common knowledge about personal identity \\nshould be evidenced together with systematic positioning of individuals in \\nthe frame of this common knowledge and the anchoring of such positioning \\nin other socio-psychological meaning systems. Indeed, important compo­\\nnents of all studies on social representations are the definition of a common \\nfield, the systematizing of individual positioning in that field, and the \\nanchoring of these positionings in related fields of representations (Doise, \\nClemence & Lorenzi-Cioldi, 1993). \\nSeveral investigators have already examined aspects of personal identity \\nin the frame of social representation theory (see, for instance, Durand \\nDelvigne, 1992; Palmonari, Carugati, Ricci-Bitti & Sarchielli, 1979; \\nRavaud, Beaufils & Paicheler, 1986; Tap, 1985; Zavalloni & Louis­\\nGuerin, 1984; Zeegers, 1988). Nevertheless, personal identity is more often \\nstudied in other theoretical frameworks and this may explain why books \\ncontaining prototypical social representation studies do not mention \\ninvestigations of personal identity (see, for instance, Abric, 1994; Doise & \\nPalmonari, 1986; Duveen & Lloyd, 1990; Jodelet, 1989). A remarkable and \\npioneering exception is the study by Codol (1 984a; see Farr & Moscovici, \\n1984). In this study Codol analysed the system of relationships connecting \\nrepresentations of self, others, group and task. \\nAnother possible reason for the paucity of research on personal identity as \\nsocial representation is the limited definition of the concept of personal \\nidentity. Generally, identity is conceived of as a very peculiar set of opinions, \\njudgements, evaluations, attitudes, manifested by a person towards him- or \\n\\n\\n14 \\nSocial identity: international perspectives \\nherself. Although research from different theoretical traditions bears on the \\ngeneral processes governing identity development and construction (Damon \\n& Hart, 1982; Erikson, 1977; L'Ecuyer, 198 1 ; Marcia, 1966) and on \\nsituational regularities enhancing saliency of identity aspects (Codol, 1990; \\nKuhn & McPartland, 1954; McGuire & McGuire, 1988), the study of \\nindividual identity as socially organized still remains marginal. \\nNevertheless, identity, as other social representations, can be studied as a \\ncognitive organization, oriented by a meta-system of social regulations. In \\norder to illustrate this thesis I shall successively show (a) that common \\nknowledge on identity exists, (b) that organizing principles rule individual \\npositioning in relation with this common knowledge frame, and (c) that \\nspecific social insertions and experiences modulate such positioning. In \\ndeveloping my position, I draw on selected evidence from studies on social \\nrepresentation theory and a reinterpretation of traditional research on social \\nidentity. My main objective is to convince the reader that studying identity \\nas a social representation is feasible. \\nCommon knowledge in personal identity \\nPersonal identity is often considered as unique and singular. On the other \\nhand, shared group membership involves common characteristics. Hence, \\nindividuals would be considered similar to the extent that they belong to the \\nsame or similar categories. In this sense group membership would result in \\nhomogeneity and individual differences would correspond to personal char­\\nacteristics which are not shared by other group members. Such considera­\\ntions fit with Turner's conception of a functional antagonism between \\nthe salience of different levels of self-categorization, the group level and the \\nindividual level: 'The salience of one level produces the intra-class simi­\\nlarities and inter-class differences which reduce or inhibit the perception of \\nthe intra-class differences and inter-class similarities upon which lower and \\nhigher levels respectively are based' (Turner, 1987, p. 49). \\nRecent research, however, suggests that the problem of similarity and \\ndifference is not so simple: commonness could also be studied at the indi­\\nvidual level and specificity at the group level. Let us consider some of the \\nresearch evidence. \\nRavaud et al. (1986) investigated school classes attended by physically \\nhandicapped pupils as well as classes of \\nnon-handicapped students. Students \\nin both types of classes were asked to describe on the same scales members \\nof their class (themselves included), as well as handicapped pupils in general \\nand non-handicapped pupils in general. Factorial correspondence analysis \\nyielded very contrasting descriptions of the general group of handicapped \\nand non-handicapped students. However, self-descriptions of handicapped \\nand non-handicapped pupils were very similar. In addition, when the \\ndescriptions targeted specific handicapped or non-handicapped individuals, \\nthere was a high degree of similarity. In terms of social representations, \\n\\n\\nSocial representations in personal identity \\n1 5  \\ndifferences are found at the category level, but they do not manifest \\nthemselves when individuals are asked to describe individual members of \\nthe categories. Commenting on these results, the authors contest the gener­\\nality and importance of the so-called 'deductive effect' of social categ­\\norization. If such an effect were strong, distinctive traits of categories as \\nsuch should be attributed to single members of each category. \\nThis is not the case: \\nIt is a rather striking observation that stereotypes are not very important in \\nstructuring representations of individuals considered to be bearers of the stereo­\\ntypes. The abstract individual described by the valid pupils is totally opposed to \\nthe descriptions of the concrete individuals and the image that valid pupils \\nconstruct of real handicapped ones and that appears in their answers seems not to \\nbe mediated through stereotypical representations. This is a strong challenge for \\nassumptions that consider stereotypes to be a necessary tool, a measure applied to \\nall members of a stereotyped group. (Ravaud et aI., 1986, p. 1 80) \\nHowever, it should be noted that the same research reports results showing \\nthat real discrimination can coexist with a blurring of stereotypes. When the \\nstudents were asked to choose others for a party at their home or as work \\npartners, non-handicapped and handicapped pupils clearly prefer non­\\nhandicapped companions. Such discrimination does not reveal itself in the \\nconcrete images pupils give of themselves and of each other. \\nIn the same line of thought, research carried out on adolescent identity by \\nPalmonari and colleagues is relevant (Palmonari et aI., 1979; Palmonari, \\nPombeni & Kirchler, 1989). Their results show striking similarities between \\ndescriptions of self given by adolescents belonging to different formal and \\ninformal groups. These descriptions reflect their openness to the social \\nworld, their commitment, disengagement and dissatisfaction. \\nSimilar results were obtained by Doise and Lorenzi-Cioldi ( 1 99 1 )  study­\\ning Swiss and second-generation immigrant pupils. These adolescents gave \\nhighly contrasting descriptions of Swiss in general and foreigners in general. \\nHowever, when describing themselves and their friends, Swiss and immi­\\ngrant pupils gave quite similar 'concrete' descriptions. Moreover, variations \\nbetween subjects of the same category in self-descriptions are not more \\nimportant than variations between them in the description of group \\nstereotypes. To state it differently, self-images are as stereotypical as group \\nimages. \\nThe results of other investigations (Nakbi, 1990; Nakbi & Arnal­\\nDuchemin, 1987) support similar conclusions. These studies gathered not \\nonly self-images and other-images of male and female students, but also \\nideal images and self-images attributed to others. Multidimensional analyses \\nof these data revealed a very similar structure of these representations for \\nmales and females. \\nThis research demonstrates a high degree of similarity between self­\\ndefinitions of members of different groups. General societal norms affect \\nthese definitions much more than do specific group memberships. This is \\n\\n\\n16 \\nSocial identity: international perspectives \\nalso a conclusion that can be drawn from research by Cowan and Hoffman \\n(1986) and by Inoff, Halverson and Pizzigati (1983). These investigators \\nexamined the self-descriptions of children, finding striking similarities in the \\nself-descriptions of boys and girls, despite the fact that conventional gender \\nstereotypes are used in the description of others. It was assumed that these \\nchildren did not apply gender stereotypes to themselves because they tend to \\nascribe favourable traits to themselves, independently of the gender conno­\\ntation these traits may convey. A similar motivation underlies the self­\\ndescriptions of Genevan pupils: Swiss and immigrant students assigned \\npositive traits to themselves that are part of the stereotypes of Swiss \\nnationals and foreigners. This tendency also corresponds to the inclination \\nof individuals to ascribe to themselves numerous and contrasted charac­\\nteristics (Sande, 1990). \\nThe content of self-representations varies from culture to culture and \\nepoch to epoch. Commonality in self-definitions in the Netherlands was \\nstudied by Zeegers (1 988), who analysed the contents of 'lonely heart \\nadvertisements' in Dutch newspapers from 1945 to 1986. Three styles of \\nself-presentation appeared successively in these personal ads. During a first \\nphase starting in 1945, a majority of the ads contained traits such as serious \\n(degelijk), tidy (net), educated (beschaa\\nf\\nd), that is, transsituational evalu­\\native characteristics. During a second phase, from 1965 on, personality \\ntraits adapted to more specific situations appeared, carrying information \\nabout lifestyle considered important for developing commitment to a \\nrelationship that should above all result in friendship. After 1975, the \\nadvertiser's description of his or her identity became still more specific, \\nincluding more personal traits and declarations of interest in particular \\nactivities. \\nIt would be remiss not to mention that there are individual differences in \\nthe presentation of self-images of members in a group. Experimental results \\nalso show that it is easy to entice individuals to differentiate themselves \\nfrom other individuals within as well as between groups using characteristics \\ndeemed to underscore one's originality. However, such differentiations often \\ninvolve an accentuation of conformity to prevailing norms, the so-called \\n'superior conformity of the self' (Codol, 1975). \\nNormative regulations strongly intervene in self-presentations. More \\ngenerally, linguistic systems offer definite dimensions for describing the self. \\nThe English language contains numerous terms for individuals to describe \\ntheir personality. Allport and Odbert (1936) identified more than 4,000 such \\nterms. Various techniques have been used to systematize this set of traits. \\nBy asking individuals to use samples of these adjectives to describe them­\\nselves or other persons, or to indicate similarities between the traits, \\ninvestigators consistently demonstrated the existence of a common structure \\n(Goldberg, 1 990; Peabody & Goldberg, 1989). This structure shall be \\ndescribed in the next section, but it is important to note here that according \\nto various authors this structure corresponds to a common-sense psychology \\nembedded in daily language, much more than to the results of strict \\n\\n\\nSocial representations in personal identity \\n17 \\nscientific observation (Leyens, 1983; Semin, 1987). The important point in \\nthis discussion is that individuals use common knowledge systems to \\ndescribe themselves. \\nIdentity as individual positioning \\nThe images that individuals construct of themselves, although highly similar, \\nnonetheless vary from one person to another. In many studies five dimen­\\nsions are considered necessary and sufficient to describe these interindividual \\nvariations. According to Goldberg (1990): 'These \\\"Big-Five\\\" factors have \\ntraditionally been numbered and labeled as follows: (I) Surgency (or \\nExtraversion), (II) Agreeableness, (III) Conscientiousness (or Dependabil­\\nity), (IV) Emotional Stability (vs. Neuroticism), and (V) Culture. Alterna­\\ntively, Factor V has been interpreted as Intellect' (p. 1217). \\nIn fact these organizing principles of between-individual variation deal \\nwith the way people behave towards each other. A person is involved daily \\nin multiple social relationships and the way s/he participates in these \\ninteractions is being constantly evaluated. If common language, well before \\npsychological expert language, contains an abundant range of terms to \\nqualify individuals, this variety primarily serves the purpose of differentiat­\\ning between individuals according to their behaviour in all kinds of social \\ninteractions. \\nIt is impossible to estimate the number of people who have been asked, at \\nleast once, to describe themselves on a personality questionnaire. If these \\nindividuals were able to comply with such a request it is because they \\ndisposed of common references, of comparison frames allowing them to \\ncompare their own way of behaving with the way prototypical persons \\nbehave. Apparently, a stable organization of the description of interactional \\nsystems can be traced back in ancient historical documents. \\nAdamopoulos (1982) examined the descriptions of social interactions \\nbetween figures in epic texts of various historical epochs (Homer's Iliad; an \\nOld English poem of a fight with a dragon: Beowulf, and a story of a young \\nrecruit's adventures during the American Civil War: The Red Badge o\\nf\\n \\nCourage). In each text, dyadic interactions were retrieved and the type of \\ninteraction between each dyad was classified. Finally, a matrix was con­\\nstructed for each text linking types of interactions and pairs of persons. For \\neach matrix a principal component factor analysis was performed. Three \\ncommon factors were obtained for the interpersonal structures in the three \\ndocuments: affiliation, power (status) and formality (versus intimacy), \\nalthough the latter factor appears less clearly in the most ancient text. \\nAdamopoulos and Bontempo (1 986) repeated the procedure for The \\nOdyssey and The Song o\\nf\\n Roland, and were able to retrieve the same three \\nfactors in both documents. \\nThese results can be compared to those obtained by Wish (1976; Wish, \\nDeutsch & Kaplan, 1976), who analysed similarity judgements between 25 \\n\\n\\n1 8  \\nSocial identity: international perspectives \\nsocial relationships (for instance: between husband and wife, between prison \\nguard and inmate, between two enemies, between two political opponents). \\nThe main aim of the study was to identify the dimensions underlying the \\nconceptions of these different social relationships. \\nFour dimensions described the judgements of prototypical relation­\\nships. Dimension I contrasted competitive versus cooperative relationships: \\ndivorced couple, personal enemies, political opponents versus good friends, \\nhusband-wife and team mates. Dimension 2 involved unequal versus equal \\nrelationships: guard-inmate, interviewer-interviewee, teacher-pupil versus \\nteam mates, political opponents, good friends, personal enemies and cousins. \\nDimension 3 concerned socio-emotional-oriented versus task-completion­\\noriented relationships: parent-child, brother-sister, husband-wife versus \\ninterviewer-job applicant, seller-buyer and political opponents. Dimension \\n4 included intense versus superficial relationships: parent-child, husband­\\nwife, good friends, psychotherapist-patient versus chance encounters, \\ninterviewer-interviewee and employee-employer. Individual differences in \\nthe use of these common dimensions were also investigated. \\nThese representations of social relationships offer important coordinates \\nfor individuals who have to position themselves in self-descriptions. In this \\nrespect McAdams (1985) proposes a theory of self as composed of internal \\npersonified and idealized images. This Imago theory postulates that identity \\n'is a story complete with setting, scenes, characters, plot, and recurrent \\nthemes. Identity formation is the process of constructing a self-defining life \\nstory. The main characters in the story are imagoes' (McAdams, 1985, \\np. 127). References are classical authors such as Erikson (1977), James \\n(1 892), Jung (1943), Klein (1948), but also Markus's (1984) ideas on \\n'possible selves'. The symbolic relationships between these prototypical \\nimages remind us of Adamopoulos's analyses: \\nThe two major thematic lines around which the content of identity can be \\norganized are agency (power/mastery/separation) and communion (intimacy, \\nsurrender, union; Bakan, 1966). Initial investigations of life stories of college \\nstudents and men and women at midlife suggest that the power motive (Winter, \\n1973) and the intimacy motive (McAdams, 1980) are two independent personality \\ndispositions - both assessed via the Thematic Apperception Test or TAT - that \\nserve as significant predictors of salience in life stories of thematic lines of agency \\nand communion, respectively. (McAdams, 1985, p. 128) \\nFurthermore, McAdams presented a taxonomy of 12 major deities of the \\nGreek pantheon according to their high or low degree of 'agency' and \\n'communion', a taxonomy that he also applied to the analysis of identities \\nof 50 respondents on the basis of personality tests and of a life story \\ninterview. \\nIndividuals define themselves in relation to each other, referring to rep­\\nresentations of their participation in social interaction settings. The research \\ntradition in which I locate my own work emphasizes the importance of \\ngroup and category membership in self-definitions. However, we have \\nalready seen that individuals often do not assign for themselves the same \\n\\n\\nSocial representations in personal identity \\n19 \\ncharacteristics they ascribe to their groups. Notwithstanding such findings, \\nit remains true that differences exist between individuals and that some \\npeople more readily attribute typical group characteristics to themselves. \\nThis is especially true for gender characteristics. Each time individuals are \\ninvited to express their degree of agreement with the relevance of masculine \\nand feminine traits to their self-descriptions, factors opposing both kind \\nof traits can be found in the data (for research with French-language \\nsubjects, see Lorenzi-Cioldi, 1994; Tap, 1985). This effect certainly cannot \\nbe explained by mere chance that this organizing male-f\\nemale principle \\nactualizes the opposition between 'agency' and 'communion' described by \\nMcAdams. In his taxonomy of the pantheon, deities with high 'agency' and \\nlow 'communion' were masculine (Zeus, Hermes and Ares) whereas deities \\nwith opposed characteristics were feminine (Demeter, Hera and Aphrodite). \\nThe organizing principles of identity regulate symbolic relationships \\nbetween social agents and, therefore, correspond to the definition of social \\nrepresentations (Doise, 1986). But the fact that positioning in relation with \\neach other actualizes dimensions of 'agency' and 'communion' does not \\npreclude the importance of other dimensions. \\nVerkuyten (1990) has analysed the identity components of a nationwide \\nsample of Dutch and Turkish adolescents living in the Netherlands (N = \\n2,710 Dutch and 237 Turkish). Some significant differences were found in \\nself-descriptions between the Dutch and Turkish youngsters, but these \\ndifferences explained no more than 1 % of the overall variance. More \\nimportantly, in order to analyse relationship between self-esteem and five \\ncomponents of the self-concept (ethnic identity, academic ability, sports, \\npopularity, body image), a standardized multiple regression analysis with \\nglobal self-esteem as dependent variable was used for respondents of Dutch \\nand Turkish origin separately. The results were highly similar. For the \\nTurkish and the Dutch adolescents, body image had the greatest impact on \\nself-esteem and ethnic identity, while academic ability and sports had the \\nlowest. The similarity of results for both samples suggests that general \\nnorms are very important in shaping self-esteem, and that general societal \\nvalues transcend, in this case, cultural barriers, probably as a consequence \\nof the homogenizing effect of common schooling. \\nMakris-Botsaris and Robinson (1991) adapted for Greek youngsters \\nscales elaborated by Harter (1985) for measuring different components of \\nself-evaluation and perception in the USA. They compared factorial \\nstructures obtained in Greece and in the USA with youth aged II to 14. In \\nboth countries they found a highly similar five-factor structure of scholastic \\ncompetence, social acceptance, athletic competence, physical appearance \\nand behavioural conduct. Correlations between subscales corresponding to \\nthese factors and global self-worth were comparable in Greece and the \\nUSA, the correlation between physical appearance and self-worth always \\nbeing the highest. \\nCommon knowledge reflects itself in the way individuals define themselves, \\nothers, and their relationships. Application to oneself of this knowledge is \\n\\n\\n20 \\nSocial identity: international perspectives \\nhighly normative and results in significant interindividual resemblances. \\nNevertheless, between-individual variations are equally important and they \\nare organized along dimensions which are also consensually defined and \\nwhich refer to characteristics of the multiple relationships in which indi­\\nviduals take part. \\nAnchored identities \\nThe anchoring of social representations has mainly been studied in three \\ndifferent ways. First, the relationships with general beliefs and values (such \\nas belief in a just world or egalitarianism) have been investigated. Such \\nvalues and beliefs are considered general to the extent that they supposedly \\norganize symbolic relationships in various domains. Social representations \\nhave also been studied as anchored in the views individuals develop in \\nstructuring their social environment, as for instance the representations they \\nhold concerning relationships between social groups and categories, such as \\ngender. Finally, a third way of studying anchoring investigates how group \\nmemberships or social positions held by individuals influence their represen­\\ntations, the general hypothesis being that shared social insertions lead to \\nspecific interactions and experiences that modulate social representations. \\nThe first way of studying anchoring mainly uses correlational methods in \\norder to determine if links exist between adherence to given beliefs or \\nopinions. Such a method was used in the studies reported at the end of the \\nprevious section showing that important aspects of youth's identity, namely \\nself-esteem, were linked to perceptions of their physical appearance and less \\nto evaluation of their popularity, academic or sportive achievement. \\nAnchoring in the views on the social environment is directly studied in \\nresearch on psychological androgyny linking personal sexual identity to the \\nactualization of a gender scheme (see, for instance, Bem, 198 1 ,  1985). \\nIndeed, what else is the gender scheme if not a 'naive' theory on the \\nrespective positions of males and females in society? \\nMore generally, representations held by individuals on the nature of \\nsocial groups intervene in the structuring of personal identity. Lorenzi­\\nCioldi (1988) studied this intervention experimentally. He distinguished two \\nconceptions of the social group: the collection group and the aggregate \\ngroup. Members of dominant groups in our society tend more to adhere to \\na conception of their group as a collection of individuals, each having their \\nspecificity. These individuals define themselves through their personal \\ncharacteristics, apparently not linking themselves to their group. However, \\nmembers of more dominated groups tend to consider themselves as less \\ndifferentiated from each other, as more homogeneous, and they attribute \\nmore directly to themselves characteristics of the group as a whole. \\nLorenzi-Cioldi (1988, 1994) illustrated the complexity of the links between \\ncollective and aggregate groups, gender groups, within- and between-group \\ndifferentiation on common and specific dimensions. His data supported the \\n\\n\\nSocial representations in personal identity \\n21 \\ngeneral hypothesis of a homology between the representations of groups \\nwhich are apparently not linked. Groups of men, of original and singular \\npersonalities, of competitors, on the one hand, and groups of women, \\nindividuals with similar personalities and cooperators, on the other, are \\nrespectively associated with respect to identity dynamics of their members. \\nSocial representations of groups, membership groups as well as other \\ngroups with which one interacts, intervene in actualizing personal identity. \\nAs social interactions often involve groups occupying asymmetrical posi­\\ntions in a representational field, one is easily led to the conclusion that \\nidentities of interacting individuals differ. This was not the starting point of \\nmy theorizing on personal identity as a social representation. Initially, I \\ninsisted on the similarity between the identities of members of different \\ngroups, such as autochthonous and immigrants. I will return to this point \\nafter discussing the relevance of the third way of studying anchoring in \\nresearch on identity as a social representation. \\nThis approach to studying anchoring is still more directly concerned with \\nthe problem of differences between groups. Results presented by Durand \\nDelvigne (1 992) can be used to illustrate this form of anchoring. She \\nanalysed self-descriptions of men and women, of different ages and social \\nconditions, and found important similarities for different groups of \\nrespondents in factorial structures of answers to a French version of the \\nsexual role inventory. Notwithstanding this similarity, a striking difference \\nalso characterized the structures of men and women, the opposition between \\nthe feminine and masculine poles being much stronger in the factorial \\nstructures of responses obtained from women than in those of men. Durand \\nDelvigne summarized these results in the following way: \\nThis result is to be considered in relation with a hypothesis prevailing in relevant \\nliterature and stating that women more than men are referred to their social \\nidentity. Social models would be available for defining them, whereas men would \\nbe more easily individuated. For women, gender category would be a necessary \\nmarker of identity (Hurtig and Pichevin, 1986; Lorenzi-Cioldi, 1988). (Durand \\nDelvigne, 1992, p. 1 82) \\nHence, differences exist in the organizing principles of male and female \\nidentities and they result in different self-descriptions. This is particularly \\ntrue when comparing the self-descriptions of males and females of high and \\nlow social status. Masculinity scores are high for individuals of higher social \\nstatus, independent of their gender, and femininity scores are especially low \\nfor men with a higher status. Androgyny scores are high for all groups of \\nhigher status independent of gender. Considering these results together with \\nmany others that are accumulating in the area of gender studies, one cannot \\nbut subscribe to Durand Delvigne's thesis: \\nGender, as identity, is not inherent to a sex group, but refers to social status. \\nOtherwise said, gender orientation of women is related to the opportunities they \\n\\n\\n22 \\nSocial identity: international perspectives \\ndo have or do not have to overcome the social definition of themselves. In this \\ncase, and related to identity, gender is not a dynamic set of attributes produced by \\nsexual category membership. It is a social marker linked with the asymmetric \\nposition of dominant and dominated individuals. (Durand Delvigne, 1 992, p. 223) \\nGender opposition does not represent sex opposition. It is a social con­\\nstruction of hierarchical relationships linked with sex group reports and \\nmodulating personal identity through social differentiation processes. \\nNevertheless, the existence of similarities of personal identities across \\ngroup boundaries is real. This apparent similarity may correspond to \\nwithin-group differences along lines that were not analysed in the studies \\nreported previously. Another explanation for the repeated findings of \\nbetween-group similarities in identity research is suggested by some findings \\nof Lorenzi-Cioldi (1988) and Amancio (1989), who report that identity \\nprocesses are differently actualized as a function of intergroup encounter \\nsituations and of available comparison dimensions. In line with such \\nfindings, one can expect that a request to describe oneself on scales without \\nexplicit gender connotation or without any intergroup comparison will \\ninduce descriptions that privilege the individual as abstracted from his or \\nher group memberships. Otherwise stated, it is possible that questionnaire \\ninstructions convey social representations of the 'collection' type. Depriving \\nthe individual of his or her representations of membership in 'aggregate' \\ngroups focuses his or her attention on representations of self as an indi­\\nvidual in search of originality and distinction. But such a search itself is \\nhighly normative and results in rather similar identity descriptions. \\nThe predominance of such normative self-definitions should not prevent \\nus from studying less dominant ones. Reicher and Emler (1986) investigated \\nsuch identities in their work on the origin of adolescent delinquency. \\nAccording to them, behaviours such as vandalism, theft or physical \\naggressions are more often admitted by early to mid-teens than by any other \\nage-group. There is evidence that such self-report scores are related to the \\nactual level of involvement in delinquent activities (Singh, 1979). More \\nimportantly, such self-reports are analysed by Reicher and Emler (1986) as \\npart of a reputation-building strategy. At this age life chances are con­\\nsidered to be settled by one's educational success or failure, and a young \\nperson's identity is characterized by its relation to his/her school career. \\nNegative attitudes towards school authority are not infrequent and are \\ngeneralized to other forms of institutionalized authorities, such as police and \\nlegal authorities. This means that important aspects of the social contract, \\nnamely that authorities will protect individuals from anarchy and redress \\nviolations of rights, are no longer considered valid: \\nThus delinquency can be seen in part as a strategy for coping with the lack of \\ninstitutional protection. In the first place, delinquency acts as redress. It resolves \\ngrievances through attacks upon enemies, upon the police force and upon school \\nand public property. Second, it offers its own kind of protection through the kind \\nof reputation it creates. (Reicher & Emler, 1986, pp. 30-31) \\n\\n\\nSocial representations in personal identity \\n23 \\nA self-presentation and social reputation is actively promoted in order to \\nsignify to others that one is the kind of person who stands up for one's \\nrights. Another objective of the social construction of such a representation \\nis to obtain support from others sustaining the same oppositional identity \\n(see also Emler & Reicher, 1 995). \\nConclusion: A definition of personal identity \\nPersonal identity can be considered to be a social representation, an organ­\\nizing principle of individual positioning in a field of symbolic relationships \\nbetween individuals and groups. This positioning is related to pre-existing \\ninsertions in a network of social relationships. \\nNot all social relationships in which we take part are equally present in \\nour minds. Depending on situations, various relationships become salient \\nand give rise to different identity dynamics. Many self-descriptions only \\nactualize general societal norms inducing individuals to define themselves in \\nrelation to each other on consensual dimensions, such as physical appear­\\nance for youth. Other situations actualize social representations more \\ndirectly anchored in group and category memberships. Memberships inter­\\nvene more intensively in conditions that are now well studied (Fiske & \\nNeuberg, 1 990; McGuire & McGuire, 1988; Pavelchak, 1 989). Studying \\npersonal identity as a social representation should contribute to a better \\nunderstanding of identity processes. \\n\\n\\n3 \\nThe Personal-Social Interplay: \\nSocial-Cognitive Prospects on Identity \\nand Self-Others Comparison \\nCarmencita Serino \\nIntroduction: 'Personal' and 'social' in social identity theory \\nIn his analysis of social categorization, Tajfel describes intra-categorial \\nassimilation and inter-categorial differentiation as two directly related and \\nsimultaneously occurring processes. The minimal group paradigm (for \\nexample, Billig & Tajfel, 1 973; Tajfel, Flament, Billig & Bundy, 1 971) is \\nintended to show that these effects usually occur even when group member­\\nship is an extremely abstract and irrelevant experience for the subjects. In \\nmost of these experiments, the assimilation/differentiation processes manifest \\nthemselves by means of some form of favouritism: the mere categorization \\ninto different groups would elicit subjects' 'behaviours' (that is, giving \\nmoney, rewards, and so on) which are clearly intended to favour their own \\ngroup and to enhance the distance between ingroup and outgroup. \\nIndeed, Tajfel aimed at providing a purely cognitive basis for intergroup \\ndifferentiation (Hewstone, Hantzi & Johnston, 1 991). On the other hand, he \\nalso pointed out the existence of a collective level of reality which cannot be \\nreduced to the sum of merely individual phenomena: under certain circum­\\nstances, individuals may act essentially as members of a particular group, \\neven independently of their strictly individual tendencies or characteristics. \\nSocial identity theory (SIT) was later developed by Tajfel and Turner \\nalong these lines (for example, Tajfel & Turner, 1 979; Turner, 1 978). \\nAccording to Tajfel ( l978b), 'at least in our kinds of societies, an individual \\nstrives to achieve a satisfactory concept or image of himself'. This \\nindividual will show, then, a tendency either to confirm or to change his/her \\ncondition of group member, depending on whether a particular group can \\ncontribute (or not) to the positive aspects of his/her identity. In other words, \\nwhen his/her need for self-positive evaluation is not complied with, an \\nindividual will tend to leave the group (unless some objective conditions or \\nparticular values and norms interfere with this tendency). \\nWe may notice that, despite the original orientation towards a purely \\ncognitive approach, motivational factors are mostly put forward by the SI \\ntheorists. A true 'axiom related to the individuals' need for positive self-\\n\\n\\nThe personal-social interplay \\n25 \\nevaluation' (Deschamps, Clemence & Roux, 1983) might in fact underlie \\ntheir assumptions. \\nIndeed, the idea according to which a fundamental need for self-positivity \\nis at the basis of social identity and group membership could be easily \\nverified in short-lived groups (such as the 'minimal' ones), wherein joining a \\ngroup is a rather extrinsic and fortuitous (if not totally exploitable) experi­\\nence. By contrast, when real social groups are involved, individual and \\nsocial identities appear to be much more intertwined. Indeed, belonging to \\nparticular social groups is in general a deep-rooted aspect of persons' \\nidentity, a real 'way of being', which also affects the ways in which all social \\nknowledges are filtered and reconstructed. ' As we will argue below, \\nbelonging to some particular social groups may even be the necessary \\ncondition allowing subjects to develop a sharper and better defined self­\\nimage, to perceive themselves as being quite specific and highly distinctive \\nindividuals. \\nIn this vein, Tajfel and Turner's approach to the personal-social inter­\\nplay is worth further analysis. Their idea of 'continuum' fails adequately to \\nrepresent the dynamic connection of individual and collective processes. The \\npersonal-social continuum seems to imply an essential, unsolved dichotomy \\nbetween individual and collective processes, personal and social identity, \\nself\\n/others and intergroup comparisons. \\nAlternative views on the personal-social continuum \\nThe close interdependence of intra-psychic and social psychological pro­\\ncesses in identity formation is recognized by Zavalloni (Zavalloni, 1973, \\n1983; Zavalloni & Louis-Guerin, 1984). Her research suggests: (a) that even \\nnegative stereotypes about the ingroup are possible: these do not weaken, \\nhowever, the individual feeling of being a group member; (b) that, on the \\nother hand, the strong identification with a group does not exclude the self/ \\nothers differentiation. The intra-categorial differences can be as marked as \\nthe inter-categorial ones. Individuals can make a fine-grained distinction \\nbetween the characteristics of their own group and the degree to which these \\ncharacteristics refer to themselves (Serino, 1988a; Zavalloni, 1 983). Thus, \\nthe conditions in which group membership is made salient do not lead, \\nnecessarily, to an intra-categorial assimilation process. \\nOn the whole, the idea of a necessary and positive relationship between \\nintra-categorial assimilation and inter-categorial differentiation was not \\nalways completely confirmed (cf. Eiser, 1983; Tajfel & Wilkes, 1963). \\nBy contrast, some experiments (Deschamps, 1982a) suggest a different \\nhypothesis. In these studies, the greater the social identification, the greater \\nis also the self/others differentiation occurring inside the membership group. \\nOn the other hand, when the self/others differentiation is emphasized, the \\nintergroup differentiation is also found to become stronger. In other words, \\nintragroup and intergroup differentiations do not appear as being \\n\\n\\n26 \\nSocial identity: international pers\\npectives \\nnecessarily located at the opposite poles of a continuum. Sometimes, they \\nappear as being consistent, simultaneous judgements which can 'vary \\ntogether'. This is usually referred to as the 'covariation hypothesis'. \\nThe 'covariation' model was developed in the frame of a greater attention \\nto comparison processes occurring in 'natural' sets, where the different \\ngroups are often characterized by a different social status. Natural groups \\nand their functioning already represented a matter of central concern in \\nTajfel's perspective (see for instance, Tajfel, 1 982b). Yet, in his research, \\nprominent attention was given to the minimal experimental conditions of \\nsocial categorization. \\nIt was mostly Doise's model of the 'categorial differentiation' which led \\nto a systematic employment of natural groups (for example, males/females; \\nhigh school students/apprentices) in the experiments (Doise, 1 976). This \\nyielded a special focus on the asymmetrical functioning of categorial differ­\\nentiation, due to the place taken by different groups in our society. In this \\nvein, Deschamps ( l982b) focused on the relationships between 'dominant' \\nand 'dominated' social groups. According to him, subjects belonging either \\nto a 'dominant' social group or to a 'dominated' one would show important \\ndifferences in perceiving intergroup similarities and in 'favouring' behav­\\niours, as well as in the ways of representing themselves. Social identity \\nwould vary as a function of the 'symbolic and material capital' available to \\ndifferent subjects. In particular, the members of a 'dominant' social group \\nwould show a greater tendency to perceive themselves as highly distinctive \\nindividuals endowed with unique idiosyncratic traits. The members of a \\n'dominated' group, instead, would usually show a tendency to perceive \\nthemselves as somewhat undifferentiated components of a 'collective object' \\n(Deschamps, 1 980; Lorenzi-Cioldi, 1 988). \\nIn a study already mentioned above, Zavalloni suggested that there may \\nbe deeply different identities, depending on the social status of the mem­\\nbership group. In her view, a meaningful example is offered by the asym­\\nmetrical development of men's and women's identities. The two genders are \\nseen as the 'most basic identity groups, not only because they were the first \\nto be learned . . . but also because they have traditionally englobed all \\nothers' (Zavalloni, 1 983, p. 2 1 5), thus becoming the basis of a social \\nhierarchy codified in the culture. The dynamic interplay of men's and \\nwomen's identities implies a power relationship, analysed with reference to \\nits most subtle and elusive aspects. Language itself, for instance, reveals that \\nonly 'male categories' are usually associated with the positive cultural \\nconnotations of the concept of humanity (Zavalloni, 1 983, p. 217). \\nThere is an essential convergence in these different studies. Although \\nadopting different methods and starting points, they emphasize the close \\ninterdependence between comparison processes, self-representations and \\nindividuals' concrete positions in a hierarchical social context. They suggest \\nthat the different levels of self/others comparison (namely intraindividual, \\ninterindividual, intergroup levels) are closely articulated with each other \\n(Doise, 1 988). \\n\\n\\nThe personal-social interplay \\n27 \\nIn this view, Deschamps's studies on 'covariation' and those on the \\ncomparison between 'dominant' and 'dominated' social groups should be \\nconsidered as a whole. The covariation hypothesis suggests that, instead of \\nbeing opposed along a continuum, the personal 'pole' and the social one \\ncan be seen as 'similar' (Doise, 1 988, p. 1 05). Yet, in our opinion, this \\n'similarity' is not to be understood merely as a form of isomorphism, that is, \\nas a 'structural homology' between interpersonal and intergroup com­\\nparisons. A deeper and more concrete connection between different levels \\nand aspects of self/others comparison can be underlined. This rests on the \\nfact that social conditions may affect not only intergroup comparisons but \\nalso the forms of individuals' membership, as well as the structure and \\ncontents of personal identity. The links between interpersonal and inter­\\ngroup comparisons can be further analysed along these lines. \\nWe will briefly consider now two main questions singled out in this \\nsection. These refer, respectively, (a) to the relationships between similarity \\nand difference and (b) to the interplay of cognitive and social processes in \\nsocial comparison. \\nSimilarity relationships and self/others comparison \\nSome of the studies mentioned above suggest that there may be several \\nways of connecting similarity and difference with each other in social \\ncomparison. This point can be further highlighted by referring to Codol's \\nexperiments on the 'superior conformity of self' behaviour (Codol, 1975). \\nThis is also described as the primus inter pares (PIP) eff\\nect: the subjects of \\nCodol's experiments, set in a highly normative situation, tended to present \\nthemselves as conforming, better than the others, to the shard norms. \\nActually, they based their own self-distinctiveness on the claim for a 'simi­\\nlarity', deemed to be a value in this case. Considering these effects, we may \\nnotice that an entangled co-occurrence of similarity and difference appears \\nat the intra-categorial level. This being similar and different at one and the \\nsame time (Codol, 1 982) sheds light on the dynamic interplay (if not, on the \\nvery 'genetic' relationship: see, for instance, Luria, 1 976) between personal \\nand social identity. It is only by starting from some adhesion to a group, to \\nits norms and values (that is, by starting from a certain conformity) that it \\nbecomes possible to develop the very idea of one's own uniqueness, and to \\nset out an image of oneself as a highly distinctive individual. \\nFurther insights into the complex structure of the self/others comparison \\ncan be derived from a particular theoretical approach: that singled out by \\nTversky (1977) in his analysis of similarity. Tversky goes beyond the \\ntraditional 'metrical' approach, wherein similarity and difference are \\nusually described just as proximity/distance between points along a linear \\ndimension. He opposes this view with a 'non-metrical' approach: in this \\ncase, similarity is understood as a global judgement (that is, a J\\neature­\\nmatching process), taking into account both common and distinctive \\n\\n\\n28 \\nSocial identity: international perspectives \\nfeatures of the objects to be compared. This author demonstrated that \\nsimilarity judgements usually take a specific direction. In general, one of \\nthe two objects is in fact 'prominent' (that is, more 'salient' or 'proto­\\ntypical' than the other). It then becomes the reference point (the 'model') \\nwhich the other term (that is, the 'variant') is usually referred and \\ncompared to. By systematical variations of the comparison reference point, \\nTversky showed that perceived similarity is higher in this standard form of \\ncomparison than in the inverse condition (that is, when the model is judged \\nin reference to the variant). \\nFollowing these studies on the 'asymmetry phenomena', several authors \\nattempted to verify whether (and in what conditions) the self is a major \\nreference point in interpersonal comparison (Holyoak & Gordon, 1 983; \\nSrull & Gaelik, 1 983). Codol (1984b, 1 987)2 suggested that interpersonal \\ncomparison too is characterized by a cognitive asymmetry, depending on \\nthe different 'salience' of the terms (self, other) under consideration. Codol's \\nfindings show that, in general, perceived similarity is higher when the others \\nare compared to the self than in the inverse condition. In practice, it is not \\nthe same to state: 'I am similar to the others' as to state: 'the others are \\nsimilar to me'. A person would tend to claim others' similarity ('the others \\nare similar to me more than 1 am similar to the others') as well as hislher \\nown difference ('I am different from the others more than the others are \\ndifferent from me'). \\nStarting from these results, several meanings of similarity and difference \\ncan be pointed out (Serino, 1 988b): similarity may mean 'conformity', but \\nalso 'representativeness'; difference may mean 'specificity' but it may refer \\nto 'deviance' as well. These different meanings are the result of a social \\nconstruction and can be 'recoded' depending on the comparison direction \\n(Serino, 1 989). \\nA 'non-metrical approach' to similarity may yield important conse­\\nquences for the analysis of social categorization. One of these consequences \\ncould be the greater attention to the 'productive' aspect of categorization. In \\nfact, a feature-matching process involves a number of questions concerning \\nthe ways in which similarity judgement is performed whenever a few simple \\n(and generally a priori provided) dimensions are not available as yardsticks \\nfor comparison. The multidimensional character of comparison is then \\nemphasized: in this light, similarity and difference may even appear as \\ncompatible and co-occurring judgements, rather than inversely related and \\nopposed along a continuum. \\nA 'metrical approach' to similarity? \\nThe way in which concepts are operationalized may affect the way in which \\nparticular questions are focused on. Thus, when considering social identity \\ntheory, we may observe that (a) by focusing, mainly, on 'favouritism', it \\nappears as shifting to a motivational analysis of group membership, and \\n\\n\\nThe personal-social interplay \\n29 \\nthat (b) by considering personal and social identity as inversely related \\nlevels, the true meaning and the subjective relevance of an identity group for \\nthe individuals risk being somewhat neglected. \\nStarting from Tversky's analysis, we might even observe that Tajfel's \\napproach to similarity seems to lie on some 'metrical' assumptions. Right \\nfrom his earliest studies in this field (Tajfel, 1 959; Tajfel & Wilkes, 1 963), \\nsome similarity/difference judgements were inf\\nerred, being based on the \\nperception of some 'proximity' or 'distance' between points along dimen­\\nsions (such as weight or length) already provided by the experimenter. The \\nauthors were quite aware that this procedure needed to be further \\ncontrolled. Wilkes and Tajfel (1966) could in fact demonstrate that there \\nexists a connection between the 'distance of an actor from the series of \\nstimuli and its categorization as the \\\"same\\\" or \\\"different from\\\" the series' \\n(Tajfel, 1 98 1 ,  p. 92). Yet, these experiments cannot explain how similarity \\njudgements are performed whenever comparison is not based on a few \\nsimple units or dimensions which are already available and/or directly \\nprovided to perceivers. \\nSome general issues of research on social categorization might be traced \\nback to this 'metrical' view on similarity. For instance: \\n(a) special attention is paid to the 'deductive' face of categorization, that \\nis, to its consequences: categories are understood as a condition, at the \\norigin of the categorizing activity, more than as a product of this \\nactivity; \\n(b) judgement is split up into as many similarity relationships as there are \\ncomparison dimensions; \\n(c) similarity and difference appear as being two opposite and mutually \\nexcluding forms of relationship between the terms to be compared. \\nThe same characteristics are also shared by social identity theory, as IS \\nhighlighted by the following SIT assumptions: \\nAs a consequence of the outstanding categorization (whether 'inter­\\npersonal' or 'intergroup'), subjects will act either as mere individuals or \\nas group members. \\n2 \\nA more direct or explicit involvement of the self in the comparison \\nwould change the whole situation: the previous categorization would be \\ncancelled and a different one would be introduced. Thus, the existing \\nopposition between two groups is replaced by a more general (oneself\\n/ \\nanother) opposition, wherein membership of either category ends up \\nbecoming somehow irrelevant. \\n3 \\nWhen an intergroup categorization is activated, the typical accentuation \\nof intra-categorial (that is, self/others) similarity will be observed. Con­\\nversely, an accentuation of oneself/another difference is likely to occur \\nonly when intergroup categorization is not salient. \\n\\n\\n30 \\nSocial identity: international perspectives \\nAs a result, two main issues arise in the frame of social identity theory. \\nFirst of all the close interdependence between personal and social aspects of \\nidentity is often overlooked. Second, self\\n/ingroup and self\\n/outgroup \\nconditions are not sufficiently differentiated from each other. These are \\njust described as two, more or less similar, examples of 'interpersonal' \\ncomparison. \\nTo summarize, the continuum model, based on a necessary co-occurrence \\nof intra-categorial assimilation and inter-categorial dif\\nf\\nerentiation, is not the \\nonly way of describing the relationships between personal and social \\nidentity. In our view, even the analysis of 'covariation' could be further \\ndeveloped. Not only can intra- and intergroup comparisons appear to be \\n'similar' rather than 'opposite' poles (as in Deschamps's experiments show­\\ning that intra- and intergroup similarities can 'go together'). As a matter of \\nfact, a special form of 'covariation' also occurs when the self takes a \\nprominent (prototypical) role, thus appearing, simultaneously, as being \\nboth similar to and dif\\nf\\nerent f\\nrom the other ingroup members. Here, too, \\nsimilarity and difference are linked to each other: yet, in this case, they 'co­\\nvary' in a quite unexpected way. \\nThe relationships between similarity and difference are not totally \\nexplained by an 'all or none' approach (Deschamps & Serino, 1 996), and \\nthis may yield some consequences for identity research. \\nSelf-categorization theory \\nMore recent developments of Turner's analysis were aimed at dealing with \\nsome of the problems pointed out above. Although an essential continuity is \\naffirmed between social identity theory and Turner's self-categorization \\ntheory, nonetheless some important novelties are also introduced. In par­\\nticular: (a) the attention given to motivational factors is reduced; (b) the \\n'productive' face of categorization is emphasized; (c) similarity and \\ndifference are more closely connected with each other. \\nOf course, an exhaustive description of this theory exceeds the scope of \\nour chapter. Thus, we will just illustrate, briefly, the three points stated \\nabove. \\nFocus on cognitive processes \\nTurner integrates the prototypicality theory with the analysis of social \\ncategorization. \\nThe prototype concept was singled out by Rosch (1973, 1 975, 1978) in the \\nframe of her analysis of categories as 'fuzzy' sets. According to Rosch, \\nclass-inclusion can hardly be conceived as a rigid all-or-none process. In \\nmost cases, in fact, classification appears as a probabilistic process. More or \\nless typical examples are gathered together in the same category; therefore, \\n\\n\\nThe personal-social interplay \\n3 1  \\ncategories take a hierarchical structure, being organized around a central \\n(prototypical) element. Categories are organized, in turn, into taxonomies \\n(that is, hierarchical categorial systems), with different levels of abstraction \\nand inclusiveness (Rosch, 1 978). \\nTurner draws on these ideas when developing his self-categorization \\ntheory. According to him, the self is (also) a cognitive structure, and self­\\ncategorizations can be described as 'a part of a hierarchical system of \\nclassification' (Turner, 1 987, pp. 44-45). In this view, 'all social com­\\nparisons with others depend upon the categorization of the others as part of \\na self-category at some level of abstraction' (Turner, 1 987, p. 48). \\nConforming to Rosch, the self-categorization process gives way to a \\nnumber of superordinate and subordinate self-categories, with increasing \\nlevels of abstraction and inclusiveness. Turner explains this point by singling \\nout three different levels, describing 'one's human, social and personal \\nidentity respectively'. A very abstract and inclusive level of self­\\ncategorization is the one opposing 'human beings' to other forms of life \\n(inter-species comparison). At an intermediate level (intergroup compar­\\nison), the self-concept describes the individual as a member of particular \\nsocial groups. The most concrete and least inclusive level of categorization \\nwould be based on interpersonal comparisons, wherein a person is \\ndistinguished from the others by means of hislher particular personality, of \\nhislher unique and strictly individual characteristics. \\nThe productive f\\nace o\\nf\\n categorization \\nThe 'productive' character of social categorization is put forward by \\nfocusing on those processes directed to select, among different levels of \\nabstraction/inclusiveness, the more pertinent ones in the given comparison \\nconditions. For this reason, the question of salience is particularly \\nemphasized (Oakes, 1 987; Oakes & Turner, 1 986). According to Turner and \\nhis co-workers, the salience of a given categorization can hardly be \\ndescribed as the effect of an automatic perceptual prominence of certain \\nstimuli. Instead, it depends on the interaction between accessibility and fit \\n(see Bruner, 1 957). Accessibility deals with the degree of readiness with \\nwhich certain pieces of information are activated, while fit 'ties \\ncategorization firmly to reality' (Oakes, Turner & Haslam, 1991) in that \\nit refers to an adequate and veridical representation of social situation. In \\nthis view, the salience of a given categorization is highly dependent on \\ncontext, as well as on perceivers' purposes and goals. \\nUnlike Rosch, Turner suggests that the basic categorization level (that is, \\nthe level at which the best 'compromise' between information generality and \\nconcreteness is achieved) is a 'highly contextual variable rather than fixed' \\n(Turner, 1987). \\nCantor and Mischel (1979) already emphasized that the relation­\\nships between different levels of abstraction is flexible rather than rigid in \\n\\n\\n32 \\nSocial identity: international perspectives \\ncategorial systems concerning social reality. These authors too applied \\nprototypicality theory to the analysis of persons' perception, focusing on \\ncontext and perceivers' purposes as crucial aspects of classification. They \\npointed out a number of analogies, but also some important differences in \\nthe ways categorial judgements are performed in the social versus non-social \\ndomain (see also Holyoak & Gordon, 1 984; Wyer & Podeschi, 1 978). For \\ninstance, the different levels of abstraction seem to be 'interchangeable' \\n(Cantor & Mischel, 1 979) in the social domain. Moreover, class-inclusion is \\nmostly based on a bipolar opposition, that is, on the incompatibility \\nbetween different categories, rather than on merely descriptive typologies \\n(Cantor & Mischel, 1979, p. 33). These differences between social and non­\\nsocial categorial systems emphasize the evaluative character of classification \\nin the social domain. Elsewhere (Serino, 1990) we addressed this question \\nby referring to Billig's (1985) analysis of 'particularization'. In a well-known \\nexample, Billig describes the process by which one general category (for \\nexample, Christianity) is split up into two distinct sub-categories (black and \\nwhite Christians). Notice that the new distinction does not describe only the \\nshift from a wide, superordinate category to less abstract and inclusive ones. \\nThe difference between black and white christians, in fact, may also entail a \\nbasic opposition between a lay and primitive religiousness and a deep and \\nmature one. From this viewpoint, particularization yields an evaluative \\ndiscrimination allowing the new ingroup to take its place at the top of an \\nextremely general hierarchy (that opposing 'true' religiousness to super­\\nstition). The reference to a value system entails both particularization and \\ngeneralization, as two interrelated aspects of categorizing. The reference to \\na value system also explains why, although the level of abstraction is not \\nimmutable, the 'inclusiveness of the category remains stable' (Cantor & \\nMischel, 1 979, p. 1 6). In fact, even the introduction of a very abstract level \\nof classification can be aimed at enhancing interpersonal or intergroup \\ndifferentiation! \\nOur discussion converges on Turner's position. According to him, \\ncategorization and particularization are two interdependent aspects of the \\nsame process (Turner, 1 988). Turner's description of 'normative fit' as an \\nimportant component of salience also emphasizes, as we do, the social \\nmeaning of categorization. He argues, for instance, that categorizing some­\\none as 'Catholic' is not only the result of a comparison with other categories \\n(cf. Oakes et aI., 1 991); it also deals with behavioural and normative \\ncontents of that person's identity. \\nIndeed, social categories are highly meaningful and personally relevant \\nfor individuals. This might account for 'people's tendencies to perceive \\nthemselves and others in terms of psychologically opposing forces' (Cantor \\n& Mischel, 1 979) which seem to affect the judgements of prototypicality in \\nthe social domain. In this light, proto \\ntypicality itself might carry a par­\\nticular meaning: when social categories are involved, most 'prototypical' \\nfeatures would be in fact those endowed with a strong (positive or negative) \\nvalue. \\n\\n\\nThe personal-social inter\\nplay \\n33 \\nSimilarity and dif\\nf\\nerence as interdependent judgements \\nIn self-categorization theory, similarity and difference are linked to each \\nother in a more complex fashion. Difference always implies a 'higher level \\nidentity', so that no comparative relation in terms of difference could be \\nassessed between two stimuli if they were not suitable to be included in \\nthe 'same category' at a higher level of abstraction (Turner, 1 987). Accord­\\ning to Turner, categorization follows the meta-contrast principle. This \\nentails an inverse relationship between intra- and inter-categorial com­\\nparisons.3 Therefore, it is consistent with the idea that the perception of \\nintra-categorial similarities and inter-categorial differences is at the basis \\nof categories formation and functioning. In this view, the salience of a given \\ncategory and even the prototypicality of its prominent examples can be \\nunderstood as a function of the \\nmeta-contrast perceived between intergroup and intragroup differences. As the \\ndifferences perceived between individuals belonging to different (implicit) groups \\nincreases, compared to the differences between individuals belonging to the same \\n(implicit) groups, then the more salient will the social categorization become . . .  \\ni.e. members will tend to be perceived less as individually different persons and \\nmore as unitary social groups. (Oakes et aI., 1991, pp. 126-127) \\nTurner (1987, p. 49) claims there is a 'functional antagonism between the \\nsalience of one level of self-categorization and other levels', assuming that \\n'there tends to be an inverse relationship between the salience of the \\npersonal and of the social levels'. In this sense, Turner clearly draws on \\nTajfel's idea of 'continuum'. \\nPrototypes: similar and different \\nThe meta-contrast ratio also provides a simple operational description of \\nthe prototype concept. As Turner suggests, a given stimulus will tend to be \\nperceived as more exemplary or representative of the category as a whole \\nthe higher the meta-contrast ratio: the higher the mean perceived difference \\nbetween the target stimulus and the out-group members over the mean \\nperceived difference between the stimulus and other ingroup members, the \\nmore prototypical the target stimulus. \\nYet, by referring to the meta-contrast, only the comparative aspect (the \\n'horizontal' dimension) of prototypicality is emphasized. Being aware of \\nthis, Turner himself recently focused on some aspects of the prototype \\nwhich cannot be described in this way. \\nIn our view, the prototype concept also contains the idea that \\nsimilarity and difference 'go together'. A prototype in fact differs from \\nthe other items because it better represents an abstract image or an ideal \\nof the common category; an ideal which allows that category to be \\n\\n\\n34 \\nSocial identity: international pers\\npectives \\ndefined and distinguished in turn from the other categories. Similarity \\nand difference, conformity and uniqueness, are tightly intertwined in the \\nprototype concept. \\nThis close connection of seemingly opposite judgements can be observed \\nalso when referring to the self-concept. On the whole, Codol's studies \\nmentioned above seem to suggest that the self may act as a sort of proto­\\ntype (Hardoin & Codol, 1 984) in interpersonal perception. By varying the \\norder of two different description tasks (self-description, description of \\nanother stimulus-person), Codol (1987) also shows that common and dis­\\ntinctive traits carry a different weight across different conditions. Subjects \\nstarting with the self-description task used the same self-descriptive traits for \\ndescribing the others. By contrast, when the other-description came first, \\nsubjects tended to use quite different traits in the subsequent description of \\nthemselves. (Different results were obtained when the other person was the \\nsubject's 'best friend'.) According to Codol, these results would suggest that, \\nin general, 'the others' are assumed to belong to the same category as the \\nself, whilst the reverse is not true. \\nMore generally, there would be a dynamic interplay between similarity \\nand classification. As Tversky (1977) argued, similarity can determine the \\nway in which objects are classified, but it also depends on the adopted \\nclassification. \\nWhile developing his analysis of salience, Tversky suggests that each \\npertinent feature is weighted according to both its 'intensity' and its 'diag­\\nnostic value' (this refers to the 'kind of classification' introduced by that \\nfeature). Of course, these two different components are assumed to be both \\npresent and tightly intertwined in any matching process. Nonetheless, we \\nmay easily imagine some conditions where either intensity or diagnosticity \\nis, in turn, particularly emphasized. \\nIn some cases, when two objects are to be compared, they are already \\nassumed to belong to the same category. Comparison is just aimed at \\nsingling out the reasons for an overall accepted similarity. When a common \\ncategory is involved (even implicitly), comparison might give rise, mainly, \\nto the (differential) intensity of the shared (categorial) features. \\nIn other conditions, similarity is the final result of a more inductive \\nprocess. Here, judges are searching for an appropriate category including \\nboth objects (thus making the comparison possible): greater attention would \\nbe paid, then, to the simple presence/absence of several (either common or \\ndistinctive) features, whose 'diagnostic value' is foregrounded. \\nWhen considering in this light Turner's analysis of the different self­\\ncategorization levels, it seems that similarities and differences are judged \\nmainly by referring to the diagnostic value of particular traits. Thus, at the \\ninterpersonal level, individuals are deemed to make comparisons only on \\nthe basis of their unique, exclusive personality traits. These are obviously \\nopposed to other traits (which are not recognized as pertaining to them­\\nselves) in a kind of judgement giving rise, mainly, to the simple presence/ \\nabsence of the critical features. \\n\\n\\nThe personal-social interplay \\n35 \\nHowever, we are dealing here with an abstract idea of the individual. \\nIndeed, when self and others are perceived in terms of 'psychological \\nopposing forces', little attention will be paid to common characteristics. \\nNonetheless, interpersonal comparisons are often performed in a frame of \\nalready accepted common categories, relevant to each member and to his/ \\nher sense of personal identity. In these conditions, greater attention would \\nbe paid to some characteristics which are present, to a different degree, in \\nthe different (more or less representative) ingroup members. \\nIn this way, we can also point out an important difference between self\\n/ \\ningroup and self/outgroup comparisons: these are conditions where a \\ndifferent focus on either intensity or diagnostic value of the relevant features \\nmight be observed. \\nResearch on intergroup relations already demonstrated that the ingroup \\nmembers are perceived as varying along specific dimensions more than the \\nmembers of an outgroup are (Park & Rothbart, 1 982); actually, it seems \\nthat 'ingroupers' are compared in terms of different intensity of the same \\nfeatures, while 'outgroupers' are not. \\nAt the intragroup level, the self/others comparison might be based on the \\ndifferent intensity of common characteristics, more than on the different \\nsalience of features unique to each individual. Again, some aspects of the \\npersonal-social interplay seem to escape an explanation in terms of 'con­\\ntinuum', and we can imagine an interpersonal level of self-categorization \\nwhere some shared social characteristics do not stop being salient and \\nimportant. \\nIn most cases, a special form of particularization is likely to occur at the \\nintragroup level: it would be based on dimensional discrimination about \\nsome crucial categorial characteristics which are shared, to a different \\nextent, by oneself and by the other ingroup members. This process seems to \\nhave been described, for instance, in a study by Montei! ( 1992): this shows a \\nparticular form of 'individualization', based on the individual appropriation \\nof the positive characteristics of the group. \\nOn the other hand, a generalization process is required when self and \\nother are (perceived as being) quite different. 'Similarity' can be assessed, in \\nthis case, only by shifting to a very inclusive category (see also Turner, \\n1 987). At this level, people may share, at least, the fact of being 'indi­\\nviduals', endowed with their own personality. Thus, personal traits, which \\nare usually considered as being the most genuine and concrete identity \\nissues, would be typically elicited in some very general and abstract inter­\\npersonal conditions. \\nStarting from the prototype concept, personal identity and intragroup \\nrelations appear as an amazing, intertwined connection of similarity and \\ndifference, unicity and conformity, assimilation and distinctiveness (Codol, \\n1 982). It would be hard to reduce this complex interplay to the frame of a \\npersonal-social bipolar opposition. In this light, what about the 'functional \\nantagonism' defended by Turner? Is it possible to reconsider the same idea \\nin a diff\\nerent perspective? \\n\\n\\n36 \\nSocial identity: international perspectives \\n'Functional antagonism': different levels of categorization or different \\nmodes of processing information? \\nAccording to Turner, different contents become salient at different (intra­\\ngroup, intergroup) levels of categorization. His idea of the functional \\nantagonism between personal and social levels of self-categorization also \\nentails the opposition of personal versus social features. This dichotomy can \\nhardly be overcome just be postulating, as Turner does, a 'midpoint' of the \\npersonal-social continuum 'where the self-perception is likely to be located \\nmuch of the time' (Turner, 1 987). \\nWe would refer to a different form of functional antagonism, where \\ndifferent modes of processing information are applied to a substantially \\nunitary social reality. Cognitive research on person perception has pointed \\nout a fundamental opposition between categorial and personalized \\ninformation, top-down and bottom-up ('stimulus-related') processes (for \\nexample, Asch & Zuckier, 1 984; Fiske & Pavelchak, 1 986; Pratto & Bargh, \\n1991). \\nBrewer (1988), for instance, proposed a 'dual model', wherein which route \\nis chosen depends, mainly, on the perceiver's purposes and goals. Similarly to \\nTurner, Brewer suggests that 'the level of abstraction at which new \\ninformation is initially encoded is expected to be different for category-based \\nand person-based processing'. It is worth emphasizing, however, that exactly \\nthe same information 'can be processed in a top-down (de-personalized) \\nmanner (resulting in category-based cognitions), or in a bottom-up (per­\\nsonalized) fashion, that results in person-based representations'. Thus, \\n'personalization refers to a different kind of cognitive structure in which the \\nindividual becomes the basis for organization of all relevant information'. \\nWhen the personalized view is activated, perceivers are dealing with \\nsalient information which can hardly be reduced to strictly individual \\nproperties. Personalization does not lead, necessarily, to focusing on \\nparticular contents or on isolated individuals. Self-involvement may also be \\nlinked to the increasing awareness of a number of complex interactions. \\nIn other words, personalization could give way to an extremely sensitive \\nappraisal of all pertinent variables and of their concomitant variations. This \\nwas suggested by a recent study (Serino & Mazzara, 1 995), bearing on \\npersons' perception in particular social contexts. Subjects (high school \\nstudents) were presented with a slide showing either a class-room (C) or a \\nfast-food restaurant (F) and an ambiguous figure entering the situation. \\nThis figure was described either as 'a person' (N) or as a 'handicapped \\nperson' (H). Among other tasks, subjects had to describe themselves and the \\nstimulus-person by means of a list of 1 5  bipolar dimensions. The order of \\nthese descriptions was systematically varied: half the subjects started by the \\nself-description, later describing the other person; the other half described \\nthe stimulus-person first, while the self-description was performed later. On \\nthe whole, three independent variables were then introduced: stimulus­\\nperson (NIH); situation (C/F) and re\\nf\\nerence point (either 'oneself' or 'the \\n\\n\\nThe personal-social interplay \\n37 \\nother person', assuming that the term described first also serves as a 'model' \\nin the subsequent description). \\nAccording to our results, the description of the other person was mostly \\nbased on categorial information (that is, whether that person was presented \\nas being 'normal' or 'handicapped': F(1 5, 120) = 4.87; P = .001). The self­\\ndescription was performed in a quite different way. Different effects of the \\nthree independent variables are observed in reference to each particular self­\\ndescriptive trait, so that, for instance, some aspects of the self-description \\nseem to be affected by the situation, other aspects depend mainly on the \\nre\\nf\\nerence point, while, in other cases, the interaction of all three independent \\nvariables yields significant effects.4 \\nIn short, the self-description seemed to be more mutable, supple and \\n'personalized'. The special nature of this 'personalization' is worth empha­\\nsizing, however. It does not lie only on the reference to personal traits \\nindependent of the context. Personalization might also spring from an \\nextremely fine-grained analysis of all variables involved in a given social \\ncontext. \\nAsymmetries and prototypes in intergroup relations \\nIn the previous paragraphs we challenged the idea that personal identity \\nrefers, necessarily, to an isolated individual and to hislher unique person­\\nality traits. However, the close interdependence of personal and social \\nidentity can also be met from a different perspective. \\nAs we have already suggested, being a member of particular groups may \\ndeeply affect the individuals' experience, their ideas about social reality, \\nothers and themselves. In an intergroup context, usually characterized by \\nseveral status differences, even self-concept, self\\n/others comparison and \\npersonal identity could be defined in different ways, depending on indi­\\nviduals' and groups' position in the social hierarchy. Thus, a number of new \\nquestions arise: What structure does the self\\n/other comparison take when \\nthe 'others' are ingroup or outgroup members, respectively? Do the nature \\nand the social status of these different 'others' affect comparison modes and \\ndirection? What are prototypes in an intergroup frame? \\nBy referring to the asymmetry phenomena described above, we may even \\nwonder whether there is any link or overlapping between the objective, \\nsocial asymmetry of intergroup relations and the cognitive asymmetry of \\nperceived similarity. By linking the cognitive asymmetry of similarity \\njudgement to the structure of concrete intergroup relations, the close inter­\\ndependence between mental and social processes could be better highlighted \\n(see also Boltanski, 1982). \\nIn a study by Hurtig, Pichevin and Piolat (1 992) this connection is \\nhighlighted. These authors focused on sex and age as factors affecting \\nasymmetry phenomena. Their study centred on similarity judgements \\nbetween pictures presenting two people of different sex and age. Indeed \\n\\n\\n38 \\nSocial identity: international perspectives \\nthese two variables were considered as defining different social categories, \\nhierarchically related to each other (younger/older; male/female). Conform­\\ning to the hypothesis, a strong asymmetry effect was found, due to the sex \\nof the ref\\nerence point person. When the male photo was the reference point, \\nthe similarity perceived between the two stimulus-persons was higher. The \\nauthors argued that the male figure showed a great 'salience' (or 'proto­\\ntypicality' degree) than the female figure. Moreover, when same-sex pairs \\nwere compared, similarity was perceived as being lower in male pairs than \\nin female pairs: that is, the more salient (prototypical) stimulus also \\nappeared as fitting a highly articulated and best differentiated social \\ncategory. \\nIn the same vein, I focused on the self/other asymmetry occurring when \\ndifferent (advantaged/disadvantaged) social groups are involved. After a \\npilot-study, the powerful asymmetry between Northern and Southern \\nItalians was taken as an example of relationships between 'dominant' and \\n'dominated' social groups. The Northerners were considered as the upper, \\ndominant group. \\nIn the first of these studies (Serino, 1 988b), 358 subjects belonging to the \\ntwo groups (Northern and Southern Italians) had to perform a comparison \\nbetween themselves and the Northerners or between themselves and the \\nSoutherners, later providing, by free explanation, the reasons for their \\nprevious judgement. The comparison reference point (oneself or others) was \\nsystematically varied, as in other experiments on the asymmetry effect.5 A \\n2 x 2 x 2 design was set up, taking as independent (between subjects) \\nvariables the kind o\\nf\\n judgement (in terms of similarity or difference), its \\ncontent (comparison with the ingroup or with the outgroup) and the \\ncomparison re\\nf\\nerence point (oneself or the others, respectively). \\nAccording to our results: \\nA certain asymmetry (that is, an independent effect of the reference \\npoint variable) was observed. \\n2 \\nThis effect was found mainly in comparisons involving the group of \\nhigher status (that is, the Northerners). \\n3 \\nSelf\\n/ingroup and self\\n/outgroup comparisons tended to take an opposite \\n'direction'. In the self\\n/ingroup comparison, the reference point was the \\ngroup rather than the self (for example, 'I am similar to the Northerners \\nmore than the Northerners are similar to me'). An opposite (although \\nnon-significant) tendency appeared in comparisons with the outgroupers. \\nOn the whole, the different contents elicited in self\\n/ingroup and self\\n/ \\noutgroup comparisons show that, at the intragroup level, subjects are com­\\npared not only to other individuals, but also to a general norm. Similarity \\nwith the 'ingroupers' is based on a large amount of references to 'member­\\nship'. Figure 3.1 suggests that group-belongingness is dramatically empha­\\nsized in the self\\n/ingroup condition. Thus, on the one hand, we can stress the \\ndifference between self\\n/ingroup and self/outgroup conditions; on the other \\nhand we can observe that it is not a simple comparison between individuals \\n\\n\\nThe personal-social interplay \\n1 20 \\n1 00 \\n• Self/ingroup \\nD Self/outgroup \\n80 \\n60 \\n40 \\n20 \\n0 \\nTR \\nCU \\nHIS \\nBHV \\nSEC \\nMBSH D/ND \\nOTH \\nKey: \\nTR = personality traits; CU = culture and habits; HIS = history; \\nBHV = behaviours; SEC \\n= social and economic conditions; \\nNC \\nMBSH = group membership; D/ND = difference and/or non-difference; \\nOTH = other categories; NC = non-classified \\nFigure 3.1 \\nMean scores ( x  100) o\\nf\\n each content category elicited in \\nself/ingroup versus seljloutgroup similarity judgements (N = 121) \\n39 \\n(just focusing on strictly idiosyncratic personal traits) that we observe at the \\nintra-categorial level. \\nThe above-mentioned study did not provide decisive results. It suggested, \\nhowever, that subjects belonging to the dominant group (that is, Northern­\\ners) were more sensitive to the structure of the comparison: perceived \\nsimilarity and difference varied as a function of the reference point (either \\nself or group) even at the intragroup level. By contrast, Southern subjects \\nappeared mostly concerned with the comparison content: in this case, sig­\\nnificant effects were due, mainly, to the 'type of comparison' variable (that \\nis, whether the Northerners or the Southerners were involved). \\nOn the other hand, the free answers pointed out two different ways of \\nreferring to a common category. The Northerners employed 'we' mostly in \\nrelation to their own group (we, the Northern Italians). By contrast, the \\nSoutherners referred to a more inclusive category, describing the set formed \\nby both groups (we all, Italians from the North and the South). Thus, it \\nseems that the 'dominant' subjects focused mainly on their own category, \\nwhile the 'dominated' subjects centred, mostly, on the relationshi\\nps between \\nthe two categories (Northerners/Southerners). In a sense, our Northern \\nsubjects (that is, the 'dominant' ones) would given an example of \\ncomparison \\nconditions wherein similarity and difference are activated simultaneously at \\nthe intra-categorial level. Conversely our Southern ('dominated') subjects \\nwould set out a kind of comparison where similarity and difference are \\n\\n\\n40 \\nSocial identity: international perspectives \\nactivated simultaneously at the inter-categorial level. A strong tendency to \\novercome the Northerners-Southerners opposition goes hand in hand with \\nthe clear awareness of the difference between the two categories (a difference \\nrecoded as unjustified discrimination, and then rejected). \\nSome of the above results were worth further investigation. To this end \\nnew research on the same topic (Serino, 1 996) was carried out. A group of \\nIII subjects, equally divided into Northerners and Southerners,6 had to \\nassess the perceived similarity between themselves and the Northerners or \\nbetween themselves and the Southerners. The 2 x 2 x 2 design considered \\nas (between subjects) independent variables: \\n(a) the subjects' origin (either from the North or from the South); \\n(b) the comparison content (comparison with the Northerners or with the \\nSoutherners); \\n(c) the comparison re\\nf\\nerence point (either 'oneself' or 'the others'). \\nLet us examine some results: \\nThe comparison content yields significant effects: greater similarity is \\nassessed in comparisons with the Northerners than with the Southerners \\n[F(l ,  101) = 7.76; p = .01]. This happens for all subjects, irrespective of \\ntheir origin. All subjects underline that the Northerners (that is, the \\ngroup of higher status) are 'more similar' to themselves. \\n2 \\nOn the whole, asymmetry phenomena are more marked in comparisons \\nwith the Northerners (tS4 = 2.80; P \\n= .05) than with the Southerners \\n(tSI = - l .65; n.s. ). This tendency may suggest that a group of high status \\nalso entail a more prominent, highly differentiated cognitive category. \\n3 \\nAgain, the subjects from the North and those from the South seem to \\nconstruct their self\\n/others comparisons in different ways. It is mainly the \\nsubjects belonging to the group of high status (Northerners) who grant a \\nprominent role to the self, which is taken as the comparison reference \\npoint. By contrast, the Southern subjects show a greater tendency to \\njudge their own similarity in reference with 'the others'. \\n4 The self\\n/ingroup comparisons appear to be more 'asymmetrical', which \\nmight suggest that ingroup refers, in any case, to a more differentiated \\nand complex category than that describing the outgroup. This is true for \\nall subjects, but, once again, the comparison direction is not the same \\nfor Northern and Southern respondents. For the Northerners, the self/ \\ningroup similarity is higher when it is judged in reference to themselves \\n(t24 = 3.182; P = .004). They appear, then, as particularly centred on \\ntheir own 'self' inside the membership group. By contrast, the Southern \\nsubjects tend to focus on their own similarity, by taking the \\n'Southerners' (rather than 'themselves') as the comparison reference \\npoint ( t26 = -3.38; P = .01). \\nOn the whole, the different direction of these comparisons may suggest \\nthat different meanings of similarity are entailed by Northern and Southern \\nsubjects. In fact, the former seem to emphasize their own representativeness, \\n\\n\\nThe personal-social inter\\nplay \\n41 \\nwhile the latter appear as mostly concerned with their own conf\\normity to an \\nexternal model. Cognitive asymmetry, which is more marked in those \\ncomparisons involving the group of higher status (that is, the Northerners), \\nmay shed light on the close connection between mental and social aspects of \\ncompanson. \\nConclusion \\nIn this chapter, we have focused on personal and social identity by referring \\nto the directional character of similarity (Tversky, 1 977), to the role of \\nprototypes in person perception (Cantor & Mischel, 1 979; Rosch, 1978), \\nand to the different ways of processing social information (Brewer, 1 988). \\nWe argued that the close relationship between individual and collective \\nprocesses, personal and social dimensions, is not always suitably represented \\nby the continuum model described in social identity theory. A dichotomic \\nview of intra- and inter-categorial relations, of personal and social identity, \\nstill appear as the basis of Tajfel and Turner's research. More recent \\ndevelopments of the same perspective (such as Turner's self-categorization \\ntheory) integrate the theory under several important aspects. Yet that \\n'dualistic conceptualization of macro/micro \\nsocial relations' (Condor, 1990) \\nwhich is typically expressed by the notion of 'continuum' does not appear to \\nbe completely overcome. \\nWhen considering self/others comparisons performed in a concrete social \\ncontext, it seems that (a) even personal identity is socially defined, \\ndepending on individuals' and groups' place in a given social hierarchy, and \\nthat (b) even categorial information can be processed in a 'personalized' \\nmanner. From this viewpoint, personal and social processes, individual and \\ngroup identities, show their close connections. \\nIn short, the personal-social interplay is worth deeper analysis. Con­\\ncerning the self/others comparison, several strategies can be observed, where \\nsimilarity and difference are linked to each other in unexpected ways which \\nare beyond the classical connection of intra-categorial assimilation and \\ninter-categorial differentiation. We explained this point by referring to the \\nprototype concept. Prototypes are - at one and the same time - highly \\n'conforming' (to the category standards) and clearly 'distinctive' (in refer­\\nence to more peripheral categorial elements). A special interdependence of \\nsimilarity and difference is then required. We have suggested that, in the \\nsocial comparison, a similar close intricacy of similarity and difference often \\noccurs at the intragroup level. Here, personal identity could be based not \\nonly on one's unique personality traits, but also on the different intensity of \\ncommon characteristics, expressed to a different extent by oneself and by \\nthe other ingroup members. Thus, in some cases, different levels of self­\\ncategorization could be characterized by different modes of judging (for \\nexample, intensity versus diagnostic value), more than by the emphasis on \\ndifferent contents. \\n\\n\\n42 \\nSocial identity: international perspectives \\nThe prototype concept implies a probabilistic view on class-inclusion and \\na non-metrical approach to similarity. We suggested that this non-metrical \\napproach may yield some consequences for the analysis of social \\ncomparison, and we reconsidered some aspects of Tajfel's theory in this \\nlight. \\nIn the same vein, we referred to some studies on the asymmetry \\nphenomena in similarity judgement, showing that these effects may also \\noccur in the social domain. We must admit that it is not yet so clear what \\nasymmetry really means when comparison involves such meaningful social \\n'objects' as the self or an identity group. Nonetheless, these phenomena \\noften reveal that some sophisticated strategies are activated in interpersonal \\nperception and that respondents are sensitive to the context and to the \\n(social, cultural) consequences of their judgement. From this viewpoint, \\nasymmetry may even serve as an indirect measure of a certain stereotype, \\nwhich operates independently of respondents' intention (Greenwald & \\nBanaji, 1995; Serino, 1 996). \\nThe typical procedure of asymmetry research (based on systematic \\nvariations of the comparison reference point) would point out the different \\nmeanings of similarity and difference, which are often at the core of the \\ninteractions and 'negotiations' between different social groups. To illustrate \\nthis point, we referred to a few studies where asymmetry effects have been \\nobserved in similarity judgements performed by subjects belonging to \\ngroups of different status. Social status seems to affect the meaning of \\ncomparison, the ways in which self\\n/others similarity is assessed and the role \\nof the self. \\nFrom this point of view, asymmetry research may contribute to developing \\na typical social-cognitive level of analysis, wherein individual, interpersonal \\nand intergroup processes are closely interrelated. Social categorization and \\nsimilarity judgements are typical 'social-cognitive' processes: cognitive \\nresearch can provide fruitful suggestions in order to understand their \\nfunctioning. Yet the close articulation between mental and social processes \\nstill ought to be further highlighted. \\nNotes \\nI .  In this light, we may consider the extremely rich literature on social representations (for a \\nreview, see Jode1et, 1989). Intergroup differences in the ways of representing some particular \\naspects of social lif\\ne (such as mental illness) were highlighted in a series of studies carried out \\nsimultaneously in Italy and Spain (see Ayestanin, 1985; 8ellelli, 1987, 1994). \\n2. The substantial continuity of Codol's research (from the discovery of the PIP effect to the \\nanalysis of its asymmetry phenomena in interpersonal comparison) is discussed in Serino \\n(1992). \\n3. Hogg and McGarty (1990) suggested that the meta-contrast ratio is more appropriate in \\nsimple unidimensional contexts than in more complex and cross-cutting social frames. \\n4. For this reason, the self-description would hardly be explained by means of just one \\nvariable. Thus, we observe a slightly significant interaction of the three independent variables \\n(F(15, 125) \\n= 1 .69; p = .06) when considering all traits together. However, a number of \\n\\n\\nThe personal-social interplay \\n43 \\nsignificant effects are observed when considering each particular bipolar dimension. Thus, for \\ninstance, some self-descriptive traits (for example, calm/agitated) vary according to the situation \\n(F( \\nI ,  141) = 16.46; p = .0 I); the re\\nf\\nerence point affects the description of oneself as being serious \\nrather than f\\nrivolous (F(I ,  141) = 6.01; p < .01); the judgement of oneself as being gloomy or \\ncheer\\nf\\nul depends on the interaction of all three independent variables (F(I, 141) = 10.47; P = \\n.0 \\nI); and so forth. \\n5. For instance, some subjects had to assess 'their own similarity to the Northerners \\n(Southerners), rather than the 'Northerners' (Southerners') similarity to themselves'. According \\nto the experimental conditions, all kinds of comparisons were perf\\normed, by different subjects. \\n6. The data in the Northern region were collected by N. Cavazza, University of Bologna. \\n\\n\\n4 \\nSelf \\n-We-Others Schemata and Social \\nIdentifications \\nMaria Jarymowicz \\nSearching for the social self: the self and prosocial attachment \\nIn order to examine the social self from a social psychological perspective, \\nwe must refer to traditions in other areas of psychology (Jarymowicz & \\nGreenwald, 1 993), and distinguish between the ego and the self (Allport, \\n1 961; James, 1 890). In social psychology, this distinction is revealed in the \\nview of the self as a sub\\nject (or self as a knower) and the self as an object \\no\\nf\\n knowledge (Epstein, 1980; Greenwald & Pratkanis, 1 984). In other \\nwords, we can describe the self as an active agent or as an object to be \\nunderstood. \\nDifferent conceptualizations of the social self define how the individual \\nis connected to other people in the social environment. There are two \\nparticularly important approaches that deserve attention. On the one hand, \\nthe social self can be described by the individual's social identifications: \\ndifferent forms of syntony with other people and states of prosocial ego \\ninvolvement. These states are accompanied by feelings of attachment, \\nsympathy, affiliation, devotion, respect for others' expectations, common \\ngoals, understanding and acceptance. On the other hand, the social self \\ncan be associated with people's ability to perceive and recognize \\nthemselves, in other words, the ability to create cognitive representations \\nof one's person and self schemata. The content of the self-description \\ncombined with the representation of other(s) constitutes the social self and \\nwe schemata. \\nThe relationship between the ego and the self, or between the ego-identity \\nand self-identity (Erikson, 1956), has not been carefully examined in \\npsychology. The aim of this chapter is to elaborate on the relation between \\nthe self and the other(s) schemata, on the one side, and the states of social \\nidentification, on the other. We are particularly interested in the relationship \\nbetween the distinctness of self-we-other(s) schemata and attitudes towards \\nothers, ingroup versus outgroup members. \\nIn order to understand the process of social identifications, we can begin \\nby questioning the generality of the popular thesis that states that the self­\\nothers similarity leads to prosocial ego involvement and that the perceived \\n\\n\\nSel\\nf\\n-we-others schemata \\n45 \\nTable 4. 1 \\nThe social self as a structural basis o\\nf\\n dif\\nf\\nerent f\\norms o\\nf\\n social \\nidentifications \\nT\\nhe self structure: \\ninternal basis of social identifications \\n'T\\nhe experiential social self' \\naffect and drives mechanisms: emotional \\nempathy, security, affiliation, and other \\nhedonistic motives \\n2 \\n'T\\nhe puhlic self' \\ndependence on group's reinforcements, \\nthe we-as-a-group concept (e.g. 'We - a \\nfamily') \\n3 \\n'T\\nhe collective self' \\ncognitive representations of real social \\ndivisions and memberships, the we-as-a­\\ncategory concept ('We - workers, women, \\netc.) \\n4 \\n'T\\nhe conceptual social sel\\nf' \\ncognitive categorization based on abstract \\ncriteria, the attributive-we concept ('We -\\nhonest, tolerant, etc.) \\n5 \\n'T\\nhe autonomous social self' \\ndifferentiation of two self subsystems: 'me \\nas an individual' and 'me as a member of \\nsocial groups and categories' \\nEgo-involvement: \\nstates of social identifications \\nsyntonic feelings, seeking for contact, \\naffiliation, sensual comfort \\nconformity: respect for particular ingroup \\nnorms and goals \\nfeeling of ingroup (categorial) homogeneity, \\ninternalization of social norms, ingroup \\nfavouritism \\nf\\neeling of we-ness across (and independent \\nfrom) social groups and categories \\nrecognition of diff\\nerent and common values/ \\ngoals, enjoying communality with \\nautonomous others, groups and categories \\nsimilarity to others is the main determinant of attachment to them. This \\nthesis has support in theoretical and empirical work (Byrne, 1 969; \\nReykowski, 1979). However, relationships between self-other(s) schemata \\nand ego involvement are far more complex than this work suggests (Brewer, \\n199 1 ;  Jarymowicz, 1987, 1 994c; Snyder & Fromkin, 1980). In order to \\nunderstand this relationship, it is necessary to analyse different forms of \\nprosocial ego involvement, ranging from the simple drives to affiliate or \\nconform (when prosocial behaviour leads to affect comfort) to exocentric \\nego involvement (when prosocial behaviour is focused on other's states and \\nneeds, regardless of one's own needs: Jarymowicz, I 992a; Karylowski, \\n1982). \\nSocial identifications, expressed by syntony with others and prosocial ego \\ninvolvement, are sometimes considered to be defining characteristics of the \\nsocial self. This view, however, creates more problems than benefits because \\nthere are many different forms of social identifications based on different \\nsocial structures (Jarymowicz, 1 992b, 1994a; see Table 4. 1 ), \\nOn the primary level ( 1 )  drives and affective mechanisms lead to non­\\nspecific affiliation and seeking syntony with others. The social self cannot be \\nexplicitly articulated by the individual, it is more like 'I have experienced it' \\nor 'I f\\neel it'. Identifications are automatic and strong, their objects are \\n\\n\\n46 \\nSocial identity: international pers\\npectives \\npersons who can gratify primary needs. The feelings last as long as the \\nrelationship with particular others endures. \\nThanks to a number of social reinforcers that can be offered by groups, \\nthe process of socialization determines the respect and conformity with \\ngroup expectations (level 2). The social self may be a family, partnerships or \\nclassmates: we as a group is f\\normed (Jarymowicz, 1993a, 1 994b). Identi­\\nfications refer to the concrete group, and terminate when the person no \\nlonger belongs to the group. \\nThe ability to focus on broader social categories gradually develops, and \\nthe social self moves beyond face-to-face contacts (level 3): we as a category \\nis created (Jarymowicz, 1 994b) based on intra-categorial differentiations. It \\ngives rise to syntonical identifications with category members and results in \\ngroup favouritism. Referents of these identifications are stable only so long \\nas the cognitive map of the world reflects social reality. \\nThe human mind can create different types of categorizations as well \\n(Brown & Turner, 1 981), for example those based on abstract criteria. \\nTherefore, the attributive we appears ('We, optimists, art lovers'). This 'we' \\nis formed independently of, and sometimes cross-cuts, social groups and \\ncategories (level 4). The identifications may refer to other people, regardless \\nof their social membership. They are modified as new, cognitive, abstract \\ncategories of understanding the world (for example, acquiring the concept \\nof 'loyalty' enables one to discern between loyal and disloyal people). \\nSocial categorization and social comparison processes are not the only \\nways the social self is identified. Individuals may define themselves beyond \\ncomparisons with others or membership to particular categories. Thanks to \\nintra \\npersonal differentiation and formation of personal evaluative \\nstandards, one's self can be recognized as someone realizing the pursuit \\nof certain superior values. Thus social identifications do not refer to a \\ncategory of people but to common goals (level 5). When values are per­\\nceived as common, identifications are likely to last a long time. When goals \\nare common, identifications last as long as they are relevant. \\nThe distinction, we suggest, reflects the sequence of developmental stages \\nin which people can use new cognitive categories and evaluative standards \\n(Golab & Reykowski, 1 985; Hoffman, 1 989; Kohlberg, 1976; Piaget, 1 975; \\nReykowski, 1 989; Ziller, 1964). The stages also determine new areas and \\nobjects of people's identifications (Tap, 1 979): from social groups or \\ncategories of own membership (levels 1 -3), to categories going beyond \\nsocial divisions (level 4) or categorization as such (level 5). \\nHow can we define the relationship between the social self, social identity \\nand social identifications? In our approach the social self is considered not \\nas a subjective state of identification with other people, but as a subsystem \\nof self-knowledge. Prototypical elements of the social self constitute indi­\\nviduals' social identity (and, consequently, determine its content). On the \\nother hand, social syntony and prosocial ego involvement are considered as \\nstates of social identifications, states of ego which depend on self-we-others \\nschemata and thus on social identity. \\n\\n\\nSelf\\n-we-others schemata \\n47 \\nSelf-we-others schemata distinctness as a basis of identity \\nSelf-specific traits as personal identity attributes \\nOur definition of identity is based on two groups of premises. The first \\nreflects the traditional way of identity conceptualization: identity is con­\\nsidered as a result of recognition of the self (including language-mediated \\neffects of 'talks with the self'; see Mead, 1934). Identity components are \\nonly these elements of the self which are explicitly indicated by a subject \\nanswering the question 'Who am I?' \\nThe other group of identity-definitional criteria refer to empirical data. \\nAccording to the well-known studies by Markus (Markus, 1977; Markus \\n& Smith, 1981), traits described by participants as (a) very important \\nand (b) highly characteristic of the self are processed differently (for \\nexample, they are processed quicker). Markus labelled them schematic \\ntraits. \\n2 \\nOur own studies on social comparison processes (Jarymowicz, 1991a) \\nshowed that traits of the self perceived as most clearly distinguishing the \\nself from the others were processed differently (for example, egotism was \\nfound only for these traits of the self which were the most highly \\ndistinct). Our definition of identity refers to self-descriptive traits, traits \\nwhich are not only characteristic of the self (and can be also charac­\\nteristic of the other), but also specific - distinguishing the self from the \\nother. Such traits of the self, which are characteristic and distinguishing \\nat the same time, we have labelled identity traits. \\nIn our definition (Jarymowicz, 1991a, p. 693), personal identity is a \\nsubsystem of self-knowledge, which is constituted by the traits perceived as \\nthe most characteristic of and the most specific (distinct) to the self. \\nSelf\\n-we-others schemata distinctness as a basis o\\nf\\n personal versus \\nsocial identity \\nAccording to Byrne's studies (Byrne, 1969), interpersonal similarity leads to \\nattraction. On the other hand, studies on identity have shown that, in some \\nconditions, self/we-others similarity may be aversive and distinctness \\nattractive (Brewer, 1991; Codol, 1979; Jarymowicz, 1982, 1987; Snyder & \\nFromkin, 1980). Seeking for explanations of the incoherent data concerning \\nself/we-others similarity/distinctness and attractiveness, we found that the \\nperceived dissimilarity was not a good measure of self/we distinctness from \\nthe other. People's estimation of the degree of similarity or dissimilarity is \\ndetermined not only by the ability to differentiate between own person or \\nown group and other people, but also by motivation to perceive self or we \\nas similar or not similar to others (Suls & Miller, 1977). In order to create a \\nbetter measure, we focused on the real cognitive distinctness of self/we­\\nothers schemata, instead of the perceived similarity (depending often on \\nsubjects' wishful thinking). \\n\\n\\n48 \\nSocial identity: international pers\\npectives \\nThe new operationalization was based on three procedural requirements. \\nIndividuals' attention on the self, the we or the other has to be focused \\nwithout inducing social comparison processes (which are likely to \\nactivate social norms and evaluation). \\n2 As was shown by Holyoak and Gordon (1983), the self is the usual \\nreference point in social perception, and usually has stronger impact on \\nperception of the other than does other on the self (see Hardoin & \\nCodol, 1 984) - to focus attention on the self, the we and the others, a \\nreverse order has to be applied. \\n3 \\nTo focus the subject's attention on descriptive and not evaluative \\ndimensions of a particular object, nouns instead of adjectives have to be \\nused (as the latter may help induce affective or evaluative associations). \\nThe so-called Questionnaire o\\nf\\n Social Perception consists of a list of 70 \\nnouns, enumerating dimensions such as elegance, intelligence, tolerance (all \\npresented in a positive manner). It contains three parts, presented to \\nparticipants subsequently. The first part of the Questionnaire concerns the \\ncategory of others. Participants read the list of dimensions and mark those \\nwhich 'are on their mind when they think about other people'. Afterwards, \\nout of the already indicated nouns, they mark l O  'which are considered the \\nmost frequent'. The second part examines the 'we' category. This category \\nis first activated by an additional operation: participants are requested 'Try \\nto think about those people who are on your mind when you use the word \\nWE' and describe the chosen 'we' in an informal way. Then they read the \\nsame list of 70 nouns, and choose 10 dimensions 'most frequently used \\nwhen thinking about people considered We'. The third part examines the \\nself category: subjects choose 10 dimensions 'most frequently considered \\nwhen thinking about oneself'. \\nThe three tasks refer to schematic dimensions (labelled analogous to \\nMarkus, 1977) of three objects. The distinctness of schemata is defined, \\nhowever, by prototype, specific, distinctive dimensions, that is, only those \\ndimensions which subjects ascribed to one and only one object. Measure­\\nments built upon such material are likely to reflect the distinctness of self­\\nothers, we-others and self-we schemata. \\nAssuming that social identity may be derived from intergroup (or rather \\nwe-others) differentiation, in our approach the content of social identity is \\ndetermined by distinctive dimensions of the we related to the others, that is, \\nthose dimensions which are marked by a subject as referring to the we and \\nnot marked as referring to others in general. \\nFurther on, assuming that personal identity may be derived from intra­\\ngroup (self-other members) differentiation (or distinguishing the self from \\nsimilar ones; see Festinger, 1954), the content of personal identity is deter­\\nmined by distinctive dimensions of the self related to the we, that is, those \\ndimensions which refer to the self and not to the we). \\nWe assume that the level of the distinctness of the self (from the we) and \\nthe we (from the others) is related to the stage of the identities' development \\n\\n\\nSelf\\n-we-others schemata \\n49 \\nand their status, that is, their role in the regulation of individuals' func­\\ntioning: the higher the level of the distinctness, the greater the impact on the \\nindividual's functioning. \\nFinally we posit (and present relevant data; larymowicz, 1 991b, 1993b): \\n(a) social identity is likely to prevail in persons with a high level of we­\\nother schemata distinctness and a low level of self-we schemata \\ndistinctness; \\n(b) personal identity is likely to prevail in persons with a high level of self­\\nwe schemata distinctness and a low level of we-others schemata \\ndistinctness. \\nBy 'prevail' we mean both the higher level of development of a particular \\nscheme and its relatively greater impact on information processing, evalu­\\nation processes or behaviour. It is particularly clear when the social self and \\npersonal self do not overlap. This prediction was empirically verified and is \\npresented elsewhere (Jarymowicz, 1993b, I 994a). \\nSelf-we schemata distinctness as a basis of social identifications \\nThe distinctness o\\nf\\n the self as a precondition o\\nf\\n understanding and \\nacceptance o\\nf\\n outgroups and dissimilar others: a theoretical model \\nA number of theoretical arguments and empirical results contribute to the \\nwell-known thesis: similarity of an other to the self and the membership of \\nthe same social category facilitate identifications, and result in ingroup \\nfavouritism. In contrast, we focus on identifications with dissimilar others \\nwho do not belong to the same social group or category and study empirical \\nfindings concerning the relationships between self \\n-we-others schemata \\ndistinctness and intergroup attitudes. \\nThe following predictions are proposed (larymowicz, I 992a, I 994c); \\nThe strong syntonic identification with one's own group leads to ingroup \\nfavouritism and, in certain conditions, to outgroup discrimination. \\n2 \\nThe lower the level of self-we schemata distinctness, the stronger the \\nidentification with one's own group and ingroup favouritism, and \\nconsequently the lower the ability to decentrate and to recognize the \\noutgroup's perspective, and the lower the possibility of accepting others' \\ngoals and values. \\n3 \\nThe higher the level of self-we schemata distinctness, the weaker the \\ninfluence of a reference group on individuals' functioning (including the \\ntendency to ingroup favouritism and outgroup discrimination). \\nWe propose a theoretical model of the relationship between self-we \\nschemata distinctness and possible identifications with dissimilar others. \\nTwo groups of premises are considered. \\n\\n\\n50 \\nSocial identity: international pers\\npectives \\nDegree of \\nothers-oriented \\ninvolvement \\nHigh \\nLow \\nLevels of self-we \\n'--_----_e_----+-\\ndistinctness \\nLow \\nMedium \\nHigh \\n0>----<0\\n \\nAbility to recognize others' (outgroup members') states \\n• \\n• \\nMotivation for engagement \\nFigure 4.1 \\nSelf\\n-we schemata distinctness and the degree o\\nf\\n individuals' \\ninvolvement with outgroups and dissimilar others: a theoretical model \\n(a) instrumental (cognitive) premises - an ability to decentration and to \\nrecognize others' perspectives; \\n(b) motivational premises - a readiness to focus on others' perspectives, to \\nthink what is beneficial to them, and to be involved. \\nThe model is presented in Figure 4. 1 .  \\nAccording to the model: \\nThe low level of self-we schemata distinctness relates to lack of ability \\nto distinguish perspectives of one's own group from those of other \\ngroups, and to recognize others' states, needs, goals - which is a \\nnecessary precondition of others-oriented involvement. \\n2 The high level of self-we schemata distinctness relates: \\n(a) to an ability to distinguish perspectives of one's own group from \\nthose of other groups, and to recognize others' states, needs, goals \\n- which is a necessary but not sufficient precondition of others­\\noriented involvement; and \\n(b) to different levels of pre \\nsocial motivation (which depends on factors \\nother than self-we distinctness) - thus, degree of others-oriented \\ninvolvement varies in different situations. \\nThis model has been confronted with empirical data. Below some results \\nare presented. \\n\\n\\nSelf\\n-we-others schemata \\n6 .5 \\n(5.7) \\n-0 \\n,g 5.0 /;:21 \\n-\\neo \\no \\nu \\nl \\nQJ \\nQJ \\nl \\n.0 \\n(/) 3.5 \\nE E \\n::> \\nQJ \\n(3 1) \\nz :o  2.0 \\n0 \\nCi \\nL \\nLow \\nMedium \\nHigh \\nSelf-we distinctness \\nFigure 4.2 Recall o\\nf\\n children's \\nproblems (Szuster-Zbro\\njewicz, \\n1 990) \\n80 \\n(78 1) \\n-\\n(/) 75 \\u0002 5J \\no \\n(/) \\nQJ \\nQJ \\nQJ .<: 70 \\nm -o  \\nOl eo  \\nQJ \\nQJ \\no \\nl \\n65 (68 9) \\nL \\nLow \\nMedium \\nHigh \\nSelf-we distinctness \\nFigure 4.4 \\nReadiness to host \\ndiff\\nerent nationalities ( Cylwik, \\n1 992) \\nThe model and empirical data \\n35 \\n-\\nc \\no 0 \\nill\\n \\n'';::;\\n \\nQJ \\neo \\nl .Œ \\nOl +-'  \\nQJ \\n0 \\no E \\n25 \\n(24 3) \\n': \\\"\\u0004 \\nL \\n(5 m \\nLow \\nMedium \\nHigh \\nSelf-we distinctness \\nFigure 4.3 \\nMotivation to \\nsolve children's problems \\n(Szuster-Zbro\\njewicz, 1990) \\n90 \\nQJ \\nU \\nC \\no U 70 \\nc O  \\n6 V 50 \\nE $ \\n<i   \\n30 \\n(81 5) \\nL\\n(35 7) \\nLow \\nMedium \\n(37 2) \\nHigh \\nSelf-we distinctness \\nFigure 4.5 \\nTolerance o\\nf\\n \\nGerman minority activity \\n(Pawluczuk, 1992) \\n5 1  \\nParticipants of experiments were focused on others - presented either as \\nanonymous 'others' or as 'they' (people of different social or national \\nmembership). In all groups self \\n-we-others schemata distinctness was \\nmeasured, as well as one of the following manifestations of others-oriented \\ninvolvement: \\n(a) recall of children's problems, mentioned prior to the experiment \\n(Figure 4.2); \\n(b) motivation to solve children's problems (Figure 4.3); \\n(c) readiness to host at home students of different nationalities (Figure \\n4.4); \\n(d) tolerance of the activity of the German minority in Poland (to preserve \\ntheir national tradition) (Figure 4.5); \\n(e) positivity of evaluation of a painting, the artist being presented as a \\nforeigner (Slovak, Irish or Japanese) (Figures 4.6-4.7). \\nAs can be seen, presented data are consistent with the model. \\n\\n\\n52 \\nSocial identity: international perspectives \\nHigh 31 \\n(30.7) \\nHigh 36 \\n(35.1) \\n-\\n29 \\n-\\n33 \\no ˆ \\no ˆ \\nQ) S \\nQ) 'S; \\n‡ :-e 27 \\n(28 2) \\n‡ :-e 30 \\n0l\\nCf)\\n \\n0l Cf)  \\nQ) 0 \\nQ) 0 \\n0\\n0.\\n 25 \\no \\n0. 27 \\n(25 8) \\n(28 3) \\nLow \\n--( \\nLow \\n--((27 3) \\nLow \\nMedium \\nHigh \\nLow \\nMedium \\nHigh \\nSelf-we distinctness \\nSelf-we distinctness \\nFigure 4.6 \\nPositivity o\\nf\\n \\nevaluation o\\nf\\n Slovak and Irish \\npaintings (Sobczyk, 1992) \\nFigure 4.7 \\nPositivity o\\nf\\n \\nevaluation of Japanese painting \\n(Sokolowska, 1992) \\nThe follow-up study established that participants with a low level of self­\\nwe schemata distinctness were least likely to favour outgroups (similar \\nresults were obtained by Jarymowicz, 1994c; Kobuszewska, 1 989; \\nSzuster-Zbrojewicz, 1988). The only inconsistency found was the motiva­\\ntional pattern presented in Figure 4.3 (Szuster-Zbrojewicz, 1 990). When \\nsubjects had to declare more than just good will, persons with a low level \\nof the schemata distinctness were least likely to favour strangers. \\n2 \\nThe results of participants with a high level of the schemata distinctness \\nare less homogeneous: in some studies a rectilinear rather than a \\ncurvilinear relationship is found. \\nConclusion \\nThere are a number of conclusions that can be drawn from this discussion. \\nLet us outline them briefly. \\nEgocentrism or ingroup orientation IS related to universal human \\nproperties. \\n2 \\nThe differentiation of cognitive schemata is a necessary precondition to \\ngo beyond own egocentrism: to recognize different social perspectives \\nand understand others. \\n3 \\nSocial self and social identity (if related to social groups or categories -\\nsee Table 4. 1) induce syntony and ingroup involvement. \\n4 The necessary precondition of coexistence with outgroup members is an \\nability to distinguish the self from the we schemata: a formation of \\nsocial and personal identity is necessary to enable the individual to \\nswitch focus and shift attention from one perspective to another. \\nIn summary, then, and coming back to the opening sentences of the \\nchapter, we conclude by stating that we have tried to reconcile different \\ntraditions in the study of self in psychology. We think that from now on \\nthey should not be considered as separate. \\n\\n\\nPART II \\nESTABLISHING GROUP \\nIDENTITY \\n5 \\nA Developmental View of the Search for \\nGroup Identity \\nStephen Worchel \\nI spent several years of my life on the East Coast of the United States. \\nComing from Texas, which had only two types of weather, hot or rainy, I \\nwas awed by the changing seasons on the East coast. Each year, fall \\nfollowed summer and signalled the coming of winter. I could consult my \\ncalendar to find the exact time of the metamorphosis, but nature was a \\nbetter, but less predictable, source of information. Some years the change of \\nseason would be abrupt, easily noticeable and sudden when a winter storm \\nwould arrive and take control of my world for months. Other years, the \\nchanges of season were gradual, silent, almost secret; a few leaves might \\ndrop from a tree to whisper the advent of fall, but the full message took \\nweeks. This four-act play would be performed every year. \\nThe drama of changing seasons has a message that carries far beyond the \\nwoodlands. The Navaho Indians believe that human behaviour is a mirror \\nthat reflects the forces of nature. If we wish to understand the behaviour of \\nhumans, we should look first to understand the natural environment. \\nIndeed, there may be many messages about group and intergroup behaviour \\nin the winds of seasonal change. \\nI begin with this retreat into nature because it so sharply contrasts with \\nthe approach that characterizes much of the study of group and intergroup \\nbehaviour. Typically, social psychology searches for e\\nf\\nf\\nects and the causes \\nof these effects. Our methodological microscopes focus on a single behav­\\niour, at a single point of time. The context in which that behaviour occurs \\n(the person's previous actions, expectations about the present situation, or \\nanticipations of the future) is seen as a pesky problem and given the \\nignominious title of 'error variance'. We speak of effects such as conformity, \\n\\n\\n54 \\nSocial identity: international pers\\npectives \\ngroup polarization, social loafing, cooperation, social categorization and \\nintergroup competition as if they were stones in a behavioural river that \\nexist in isolation. \\nI want to offer a somewhat different perspective in this chapter. To begin, \\nI want to question the concept of e\\nf\\nf\\nect. Our 'effects' are not effects in the \\nsense that we characterize them. Rather they are processes that can only be \\nunderstood in the context of the preceding events and behaviours, the \\npresent interpretations of the setting, and the individual's anticipation of \\nfuture activities. These processes are intertwined with the flow of behaviour \\nover time and can best be understood and predicted in this context. They \\nare the stream itself, not concrete forms that divert, interrupt and stand \\nalone in the behavioural stream. This is not a new view. Lewin found that \\nthe effectiveness of leadership styles could be described only when one style \\nwas examined in light of the leadership style that had preceded it (Lewin, \\nLippitt & White, 1 939). Schachter (1951) found that reactions to a deviate \\nnot only changed over time, they were influenced by changes in the deviate's \\nbehaviour. If the deviate accepted the group position, he was quickly \\nembraced by other group members. Continued disagreement with the group \\nresulted in a cessation of communication directed towards him and \\nrejection. Sherif and his colleagues (Sherif, Harvey, White, Hood & Sherif, \\n1 961) found that hostility between groups could be reduced only after a \\nseries of cooperative encounters. And McGrath and his colleagues (Kelly, \\n1 988; McGrath, 1 988) have lamented the neglect of time in the study of \\nsocial psychology. \\nI want to build on the foundation laid by these early studies and present \\ngroup and intergroup behaviour as a procession of events, each related to \\nthe other, and each taking on meaning from the context in which it occurs. \\nGroups, I believe, develop through rather predictable stages (or seasons). \\nThe process of group development is not one with a distinct beginning, and \\noften there is no clear end. In fact, many groups 'live on' long after all the \\noriginal members have departed. Rather, the development is a cycle that \\nrepeats itself. Each stage has a predictable impact on the behaviour of \\ngroup members and on the relationship between groups. Actions that are \\naccepted or encouraged at one point in the process will be censored and \\nrejected at other points. Understanding and predicting group and inter­\\ngroup behaviour requires that we understand the developmental process. \\nIndependence versus group membership: a life-long dilemma \\nBefore jumping directly into issues of group and intergroup behaviour, let \\nme be true to my psychology heritage and focus on the individual for a \\nmoment. Humanistic psychology argues that nearly from the moment of \\nbirth we embark on a life-long journey to 'find' our identity (Maslow, 1 954; \\nRogers, 1 95 1). The journey is a strenuous one, but at the heart of it is \\ndetermining our relationship with the groups to which we belong. Defining \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n55 \\nthis relationship throws us into two raging and competing cross-currents. \\nOn one hand is our desire to be accepted and valued by others. Groups \\noffer us support, safety, protection and a reason for existence. While groups \\nare a safe harbour in many ways, they also exact their price. They demand \\nobedience and conformity, they require us to place group needs above \\npersonal needs, they coerce us to give up some of our personal identity. \\nIt is in this latter demand where the counter-currents meet with such fury, \\nbecause paralleling the desire for group membership is the desire to be \\nrecognized as an independent, unique individual. In order to achieve this \\ngoal, we must differentiate ourselves from the groups to which we belong. \\nWe must resist their seductive lure to abandon our independent self and \\naccept the group identity. Our lives are characterized by the constant \\nstruggle of deciding how much of our personal identity we will sacrifice for \\nthe substitute group identity. \\nTajfel and his colleagues (Tajfel, I 972b, 1 978d; Tajfel & Turner, 1986) \\nanticipated this discussion in their development of social identity theory. \\nThey implied that we actually hold two identities (more correctly, one \\nidentity with two opposite poles), a personal identity that includes our \\npersonal characteristics and a social identity derived from our membership \\nin groups. Turner (1982) argued that the identity (personal or social) which \\nwe stress at any moment is a function of the context in which a particular \\ninteraction takes place. Intimate interactions, for example, enhance the \\nsalience of our personal identity, while interacting with people from other \\nidentifiable groups may make our social identity salient (Doise, Deschamps \\n& Meyer, 1978). For example, being the only woman in a group of men \\nmakes salient the social identity, especially of the minority person, and \\ninfluences the nature of the interaction that will take place. Social identity \\ntheory further suggests that we wish to maintain a positive image of \\nourselves, and in order to enhance our social identity we will advantage our \\ningroup in relation to outgroups. \\nSocial identity theory offers unique insight into a host of social and \\nindividual behaviours. I would like to push the social identity approach a \\nstep further and suggest that the desire to emphasize one identity or the \\nother is influenced by more than the social context in which people find \\nthemselves. I want to argue that the individual's own context (history, \\ninterpretation of the present situation, and anticipation of the future events) \\nplays at least as important a role as does the social situations in determining \\nthe salience of personal/social identity. Further, an individual's life can be \\ncharacterized as alternating between placing an emphasis on group identity \\nand emphasizing personal differentiation. I accept the social identity posi­\\ntion that the social situation influences which identity is salient. However, I \\nsuggest that it is the individual's own context that motivates him or her to \\nseek social situations that emphasize one identity or the other. For example, \\ndescriptions of close relationships suggest that individuals often alternate \\nbetween periods of close intimacy with a partner followed by efforts to \\n'touch base' with larger groups of people (Argyle, 1 992; Oldenquist, 1 992; \\n\\n\\n56 \\nSocial identity: international perspectives \\nReis, Nezlek & Wheeler, 1 980). Partners in intimate relationships are often \\nthreatened when they perceive the other 'pulling away' and becoming \\ninterested in the activities of social groups. However, the natural tide of \\nrelationships is a cycle of emphasizing the intimate relationship and \\nemphasizing the membership in other groups. A threat to one aspect of the \\nidentity can lead a person to focus on and attempt to repair that identity. \\nIndividuals who have suffered threats to their social being, such as being \\nfired from a job or going through a divorce, seek social support networks to \\nreaffirm their social identity (Rook, 1 987). On the other hand, when the \\nthreat is to the personal identity, such as the case of experiencing an \\nembarrassing event, the desire to be alone and repair the personal self is \\ndominant (Morris et aI., 1976). Alternating between one identity and the \\nother occurs not only in brief episodes during life. It also characterizes the \\nlarger pattern of human existence; greatest emphasis on establishing \\npersonal identity occurs during early childhood and middle age, while social \\nidentity themes are often most evident in adolescence and in the elderly. \\nFinally, none of our behaviour escapes the embrace of culture, so it is not \\nsurprising that something so central as identity is a product of culture. \\nNumerous investigators (for example, Hofstede, 1 980; Triandis, 1 988) have \\nobserved that cultures place varying emphasis on personal and social \\nidentities. In collective cultures such as those found in Asia, Greece and the \\nMiddle East, the emphasis is on the group and the individual's social \\nidentity is stressed. On the other hand, more individualistic cultures such as \\nthe United States and some countries of Western Europe force the \\nindividual into centre-stage, and personal identity plays a critical role. It is \\nno accident that humanistic psychology with its emphasis on differentiating \\nthe self from the social context is a Western phenomenon. \\nThe development of groups \\nThis has been a rather long preamble to a discussion of groups and \\nintergroup behaviour. However, I wanted to make the point that even at the \\nlevel of personal and interpersonal behaviour there is an ongoing process of \\nidentification that involves emphasizing themes of inclusion and differentia­\\ntion at different points of time. The theme that is emphasized is a function \\nof dynamics occurring within the individual and between individuals, as \\nwell as events that occur within the person's physical and social environ­\\nment. My aim is now to demonstrate that similar processes characterize \\ngroups and the relationship between groups. My interest in group develop­\\nment was initially sparked because so much of the research and theory in \\ngroup and intergroup behaviour was not dynamic in approach, focusing \\nonly on specific behaviours at a single period, and ignoring the history of \\nthe members and the groups. I was concerned that most research on group \\ndynamics began with the group already established, either in the sense of a \\nlaboratory group of strangers who arrived for an experiment or in an \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n57 \\nestablished group in a natural setting. This approach overlooks the process \\nof group formation, how individuals decide to join groups, and how these \\nvariables affect group behaviour. These are important issues that may have \\ndramatic impact on all group and intergroup behaviours. Unfortunately, \\nthese issues are beyond the scope of this chapter. However, although I will \\nnot address them, I want to raise them for future consideration. \\nMy second concern with traditional group research is that group devel­\\nopment and change is rarely addressed. Specifically, after reviewing the \\nliterature I was left with many questions about the dynamics of groups: for \\nexample, how do groups develop and change over time, and how does this \\ndevelopment affect intermember and intergroup behaviour? In order to \\nexamine these questions, I turned to empirical accounts in social science \\nliterature on group dynamics and popular archival accounts of groups. My \\nsearch of the archival literature included descriptions of small groups \\n(families, athletic teams, splinter church groups, gangs), large groups \\n(business organizations, political parties, nations) and social movements \\n(women's rights movements, the civil rights movement, ethnic identity \\nmovements). I included a wide variety of groups because I wanted to \\nidentify both similarities and differences in the process involved in these \\ngroups. I was also guided by a personal interest in the processes involved in \\ngroups gaining independence from larger groups. This interest moved me to \\nconsider the area of group formation, as well as development, and carried \\nme into the fascinating literature on revolution, protests and terrorism. I \\nhave discussed group formation and revolution in earlier papers (Worchel, \\nCoutant-Sassic & Grossman, 1992), and space does not allow me to deal \\nwith this topic here. For the present, then, I will focus more directly on \\ngroup behaviour from the time a group is initially formed. \\nActually, there has been a long history of efforts to examine group \\ndevelopment. The t-group movement beginning in the early 1 950s offered a \\nmarvellous opportunity to observe groups over time and chart their devel­\\nopment and change. Based on data from these and similar groups, investi­\\ngators suggested that groups do move through rather predictable stages. \\nFor example, Tuckman (1 965; Tuckman & Jensen, 1977) identified the \\nstages of forming, storming, norming, performing and adjourning. While \\nthis earlier work focused on the group, Moreland and Levine (1 982, 1988) \\nexamined the stages that individual members go through as they enter and \\nmove through the group. These studies of group and member development \\nare rich and interesting, but it is difficult to determine where they have \\nimpacted group research. Rather than representing a coordinated effort, the \\nstudies and observations have been undertaken in isolation from each other \\nso that it is not possible to identify a unified theory of group development. \\nIn addition, these fragmented observations of groups do not readily lend \\nthemselves to developing specific hypotheses. This is not necessarily an \\nindictment of the approach, because much of the work was designed to \\ndescribe group development, not generate research, but it does help explain \\nwhy this approach has not been incorporated into mainstream research \\n\\n\\n58 \\nSocial identity: international pers\\npectives \\nGroup identity \\nGroup productivity \\nIndividuation \\nNote: During the decay period numerous members may leave the original group to join \\nother groups or form new groups. While the old group re-enters the group identity \\nstage, new groups formed by the spinning off of defectors will enter into their own \\nstage of group identity. \\nFigure 5. 1 \\nStages o\\nf\\n group development in existing groups \\nefforts. Finally, many of these observations dealt with rather specific groups \\n(t-groups, therapy groups), and the applications were rather group-specific. \\nMost of the groups involved had specific beginning and end points, and \\nthey were formed for specific purposes. Therefore, the work rarely dealt \\nwith issues of group formation or examined cycles in development. Even the \\nimportant work of Moreland and Levine follows the individual member \\nthrough one group; there is no discussion of how the experience of \\nmembership in one group impacts the member's choice of additional groups \\nor behaviour in these additional groups. In this sense, most earlier models \\nare linear. As we will see, my observations suggested that group develop­\\nment is cyclical and repetitive. This cyclical nature of groups has important \\nimplications for explaining intragroup behaviour and the continuing rela­\\ntions between groups. \\nMy work on group development used this earlier work as a starting point, \\nand it was designed to deal with the limitations I perceived in these earlier \\napproaches. As a beginning, I was able to identify four stages or periods \\nthat occur in intact groups (assuming group formation as the beginning \\npoint). The stages represented themes that dominated group activity during \\nspecific time periods. The stages are reciprocal in that they affect the \\nbehaviour of group members and the relation between groups, and they are, \\nin turn, affected by member behaviour and intergroup relations. Beginning \\nwith the formation of the groups, the stages are: identification, group pro­\\nductivity, individuation and decay (see Figure 5. 1). \\nIdentification \\nOnce a group is formed, its initial focus is on establishing its identity. \\nEfforts are directed at defining the boundaries of the group, its 'place' in \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n59 \\nrelation to other groups (and history), and the relationship between mem­\\nbers. This focus has profound effects on member behaviour and intergroup \\nrelations. The group establishes a central dogma or theme for the group. \\nAttitudes on these issues are often extreme, clear and uncompromising. \\nMinority positions are not tolerated and minorities are quickly excluded \\nfrom the group. There are strong pressures on members to conform. Groups \\noften adopt a dress code or uniform, a special language and other symbols \\nthat identify the group and mark people as being members. In many cases, \\nefforts are made to 'reach back' in history to establish the legitimacy of the \\ngroup, recognize patrons and previous founders, and locate the marks that \\nwill distinguish the group and identify its 'rightful' members. For example, \\nthe 1980s were a time of rebirth of Maori people in New Zealand. The \\nMaori became especially interested in their rightful place in New Zealand \\nhistory, their language, their customs and dress. There was a renewed push \\nfor recognition of the Maori within New Zealand society, a return of lands \\nto the tribes, and greater powers for self-governance. \\nLeadership during identification is often centralized. Members may be \\nasked to demonstrate their loyalty to the group by making sacrifices or \\nundergoing difficult initiation rites. The boundaries of the group are closed \\nto new members. Recall that the initial stages of the doomsday cult studied \\nby Festinger, Riecken and Schachter ( 1956) was characterized by \\nwithdrawing from public focus and closing the group to new members. In \\nmany cases, groups in a period of identification will physically withdraw, \\ndemanding that members break contact with other groups to which they \\nhave belonged. \\nRelations with outgroups become especially strained at this point. Group \\nboundaries are defined by the relation of one group to other groups. As a \\nresult the group either seeks to isolate itself from other groups or it invites \\nconflict and competition with outgroups. Either of these actions defines \\nboundaries and has a bonding effect within a group. Outgroups are \\npresented as dangerous, aggressive and sinister, thereby forcing members \\ntogether to deal with this outside menace. Similarities within the group \\nmembership and differences between groups are emphasized. In some cases, \\ndemands are made on outgroups which cannot be met, precisely to 'prove' \\nthat the outgroups are evil and unfair. For example, in many cases where \\nethnic groups attempt to 're-establish' themselves and incite member \\nawareness, demands are made for the return of land or other resources that \\nwere taken from the group at an earlier period of history. Even when it is \\ndeemed that the resources were illegally seized, the wholesale return of these \\nproperties would create havoc in the larger society, so that the wholesale \\nreturn is either impossible or will involve a slow, complicated process. The \\nreluctance to return this property or the slow process is then used to 'prove' \\nthe lack of good will in the outgroup. \\nFrom the group's perspective, the aim of the identification stage is to \\nfocus group members on their social identity, as opposed to their personal \\nidentity. Interactions within the group and between ingroup and outgroup \\n\\n\\n60 \\nSocial identity: international perspectives \\nmembers emphasize social identity. At the same time that the group seeks to \\nstrip away the individuating characteristics of its members, it attempts to \\ndifferentiate its own characteristics as a group, separate and different from \\nother groups. \\nGroup productivity \\nWhile efforts during the identification stage increase the cohesion within the \\ngroup, they often interfere with group productivity. Therefore, as identity \\nbecomes established, group members begin to examine the goals of the \\ngroup, especially ones involving group productivity. Although the orien­\\ntation is still on the ingroup, the topic of concern is productivity rather than \\nidentification. This focus creates numerous changes in the group and inter­\\ngroup behaviour. Overall, the emotional climate cools from the passionate \\nstate that characterized the identification process. Members now take a \\nmore analytical approach. During identification, group members were most \\nconcerned with the similarities of the group members. As the concern turns \\nto productivity, members begin to examine the differences between them­\\nselves, but only those differences that may affect the group's task-related \\nactivities. The search is for the specific skills that members may possess that \\nwill help the group towards its productivity goals. Although still empha­\\nsizing social identity, the door is now opened for members to focus on their \\npersonal identity as it relates to group productivity. The boundaries of the \\ngroup become more permeable and new members may be recruited to \\nperform specific tasks. In some cases, the group may search for highly \\nskilled members to fulfil high esteemed roles, but in many cases the new \\nmembers will have a lower status and be expected to perform less attractive \\njobs, until they have 'proven themselves' worthy of group membership. This \\nis often the fate of new immigrants to a country. \\nThe group becomes less antagonistic towards the outgroups, but its \\nrelationship with them remains cautious. In some cases, comparison with \\noutgroups is sought in order to determine how 'productive' the ingroup is. \\nLeadership becomes more task-oriented and the emotional pitch of the \\ngroup is reduced. Minorities can influence the group on task-related issues, \\nbut these minorities will still meet quick rejection and censure if they \\nthreaten group identity. Group failures are attributed to external sources, \\nwhile group success is attributed to internal group factors. \\nIndividuation \\nDuring the productivity stage, members began to differentiate between \\nthemselves based on their task skills. The self-focus expands and accelerates \\nduring individuation. Members begin to assess their contributions to and \\nrewards from group membership. They demand equity based on their \\ncontributions to the group rather than equality based on simple group \\nmembership. Subgroups appear during this stage, based first on skills and \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n61 \\nroles developed in the drive for productivity, but later expanding to include \\nsubgroups based on attitudes, interests and skills that may be unrelated to \\ngroup activity. Individuals become more focused on personal needs and \\ncharacteristics. Interactions within the group become more selective, but \\nmore intimate. Minority voice is not only tolerated, but encouraged. Social \\nloafing (Latane, Williams & Harkins, 1979), in the sense of working less \\nhard for group goals than personal goals, becomes common. Leadership \\nmay be questioned and failures blamed on poor leadership. \\nRelations with the outgroup take on a very different flavour compared to \\nthe earlier two stages. Individuals begin a process of social comparison with \\noutgroup members, as well as with ingroup members. They seek infor­\\nmation about out \\ngroups to use as standards for comparing their own group, \\nand their treatment within the group. There is a tendency to overestimate \\nthe quality of life in the outgroup and the success of the outgroup. But this \\nview is not necessarily used as a threat to the ingroup identity. Rather, it is \\npresented as evidence that the ingroup is not meeting individual needs, and \\nthat changes are needed within the group. Rules, roles and norms of the \\ningroup are questioned, and there is a demand for greater personal freedom. \\nIndividual members begin negotiating with the ingroup to improve their \\npersonal standing. \\nDecay \\nFrom a group identity standpoint, the shift in focus to personal identity \\namong group members poses a threat to group identity. There may be \\nefforts to suppress this move towards individuality, but the group has \\nbecome fragmented and the threat of rejection is less feared by members. \\nThe group becomes less important in the individual's self-concept. Indi­\\nvidual members actively explore defecting to other groups, or subgroups \\nseek to break away from the group and form their new group. Marginal \\nmembers who are not central to the group's identity are the first to leave. \\nLater in this stage, more central members who have 'saleable' skills leave \\nthe group. These losses are viewed with alarm by the remaining members, \\nwho use these losses to demand change within the group. Although the call \\nfor change may result from a recognizable problem, members are often \\nunclear about the specific change they desire. The result is that initial \\nattempts to respond to these demands are often unorganized and ineffective. \\nThe easiest change is to replace the leader(s). There may be rapid succession \\nof leaders as much is demanded from them, but members are unwilling to \\ninvest much power in leadership positions. Bitterness is expressed as \\nmembers feel that they were lured into the group on the basis of false \\npromises and expectations. Anticonformity as a way of demonstrating \\npersonal independence and displeasure will be common. \\nOutgroups perceive the vulnerability of the group and efforts may be \\nlaunched to encourage defection. Alarm over these incursions may be \\nexpressed, but the group is unable to mount a concerted defence. On a broad \\n\\n\\n62 \\nSocial identity: international pers\\npectives \\nscale, the decay of one group often creates conflict and hostility between \\noutgroups as each attempts to profit from the condition. \\nThe cycle \\nThe stage of decay has a dramatic impact on the group. Many members \\nleave the group and a negative emotional state predominates the group. But \\nas the most disaffected members leave, the remaining members begin \\nturning their attention back to the group. In the chaos, a strong leader often \\nemerges, offering the remaining members the hope of re-establishing their \\ngroup. The defections are seen as a group cleansing. Anger and hostility is \\nexpressed towards the defectors, and the group takes some solace in stories \\nof disasters that befell these defectors. Attention is turned to rebuilding and \\nredefining the group. The phoenix rises from the ashes. The group begins \\nthe identification stage once again. Often there is a surge of energy that \\ncomes from rebirth. The cleansing process may reach new heights as the \\ngroup looks internally to define traitors or marginal members who have \\nremained. These people are expelled, often with public ceremony. The \\nfundamentalism that characterizes identification process grips the group. \\nThe reborn group begins differentiating itself from other outgroups, \\nespecially those that have defectors or profited form the demise of the \\ngroup. Members' social identity is emphasized over personal identity. The \\ncycle begins anew. \\nThe issue of inevitability \\nThis brief summary of the process of development raises as many questions \\nas it answers. One of the first questions is whether this progression of stages \\nis inevitable and must proceed uninterrupted in this order. The answer to \\nthis question is 'no'. Although this cycle may be the 'natural' flow of events, \\nit is very clear that the process can be interrupted or short-circuited by \\nnumerous intended and unintended events. \\nAt any point during the process, events may occur that force the group to \\nfocus on specific issues and drive it to a particular stage. The most common \\nsituation is a return to group identification concerns. A clear or perceived \\nthreat from an outgroup, regardless of when it occurs in the development \\nprocess, often makes group identity issues salient. An outgroup may make a \\nbold effort to capture resources (land, material, status) held by a group. For \\nexample, it is not uncommon for employees of a company to rally to the aid \\nof that company in the face of a rival's bid to buy that company. This can \\noccur even when the employees are expressing dissatisfaction with their \\norganization. \\nA wily leader, feeling that his or her power is slipping or that the group is \\ndisintegrating, may seek out or manufacture a conflict with an outgroup. \\nThe result of this conflict is that the group will be thrown back into a stage \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n63 \\nof identification. For example, President George Bush was accused of \\naccelerating the confrontation between the United States and Iraq in an \\neffort to consolidate his power base and tum Americans' attention away \\nfrom a deteriorating economy and onto the issue of national identity and \\npride. \\nA sudden windfall or a dramatic increase in resources may push groups \\ninto a stage of individuation. Such a windfall requires that the group grapple \\nwith the problem of how to divide these resources among the members. \\nIndividual members may then begin to make their claim to these resources \\nbased on their skills, contributions to the group, personal needs or other \\npersonal characteristics. It is interesting to notice that the windfall, in this \\ncase, may hasten the decay of the group, in its present form. \\nFinally, a scandal that arises from within the group may lead to disillu­\\nsionment and decay. Scapegoating and dissatisfaction with leadership can \\nresult in members feeling that the group cannot meet their needs, and result \\nin their defection to other groups. \\nWhile the progression through stages in order is not inevitable, I would \\nlike to offer two observations based on the reviews of group behaviour. \\nFirst, it will take some unusual or unexpected event to derail the orderly \\nprogression. Second, when a group is pushed either back to an earlier stage \\nor forward into a new stage, further group development will progress from \\nthat new stage. For example, if a group undergoing individuation is forced \\nsuddenly into the identification stage, the next step of that group will be into \\ngroup productivity. The hands of the group clock move from the stage at \\nhand, they don't 'catch up' or revert back to assume an earlier cycle of \\ndevelopment. \\nTransition and the nature of development stages \\nA second major issue concerns the nature of the various stages and what \\nforces the group from one stage into the next. These questions captured my \\ninterest as I read hundreds of accounts of groups. Was one stage necessarily \\nlonger than another? How did a group 'decide' that one stage was com­\\npleted and it was time to move to the next stage? Were there clear \\ndistinctions between stages? \\nI continually attempted to fit the literature into a concrete structure, but \\nthe fit was always an uncomfortable one. Like the variety found in the \\ntransitions between seasons, the changes in group stages followed a variety \\nof patterns. In some cases, the transition was abrupt and clear, while in \\nother cases it was more gradual, with the group dealing with two issues (for \\nexample, identification and productivity) simultaneously. Slowly it became \\nclear that the stages were not necessarily distinct, bounded entities, but \\nrather they were periods in which certain issues predominated the group's \\nconcern. Although central issues could be defined with considerable inter­\\nrater reliability, the fact that one issue was dominant did not mean that the \\n\\n\\n64 \\nSocial identity: international perspectives \\ngroup avoided issues characteristic of other stages. Again, going back to the \\nanalogy of seasons, winter in most parts of the world is characterized by a \\npreponderance of cold days. But even in the harshest winters, sunny warm \\ndays may appear for brief periods of time. Such was the nature of the group \\nstages. The group discussion during a period would be ruled by concerns \\nwith a specific issue, but other concerns would also be addressed. However, \\nintroduction of these 'other' issues usually generated briefer discussions that \\noften involved comparatively few members. Thus, a particular stage could \\nbe identified by the topic of discussion and behaviour that dominated the \\ngroup and by the length o\\nf\\n time s\\npent on discussion and the number o\\nf\\n group \\nmembers involved with that issue. \\nSimilarly with the transition from one stage to another. It was often a \\nmatter of degree, but the change could be empirically identified by charting \\nthe topics of group discussion. Such, too, was the case with identifying the \\n'trigger' for the transitions. There seemed to be no single event that moved \\nthe group from one stage to another. In some cases, a group member or \\nmembers would declare that it was time the group moved on to another \\nissue. This occurred both in small groups and in larger groups including \\nsocial movements. What was interesting was that this 'trigger' propelled \\nthe group into another stage only when the group was ready to move. \\nThere were numerous examples in all sizes of groups of when a member, \\noften the leader, requested that the group move on, only to be ignored by \\nthe group, which clung steadfastly to its present issue. In some groups, the \\ntrigger was a deadline or an imposed demand from the outside. Often \\nmembers would indicate that a certain issue, generally productivity, must \\nbe addressed in order to meet a deadline; this moved the group out of \\nidentification concerns. Boredom was a frequent trigger. Several members \\nmight complain that the group was going nowhere or that it had already \\naddressed certain issues and that they were ready for something new. \\nAnother sign of malaise and desire to move on was a lack of participation \\nby members. \\nFailure and success were some of the most interesting events in the groups I \\nexamined. Failure, regardless of the stage, tended to fixate members in that \\nstage for a lengthy period of time. Like the classic Zeigarnik effect (Zeigarnik, \\n1927), groups stayed with or quickly returned to issues that they failed to \\nresolve. However, if these new efforts were not successful, groups tended to \\nmove into decay. \\nSuccess was often as difficult to cope with as failure. Success spelled the \\nend of a particular stage and moved the group, sometimes prematurely, into \\nthe next stage. For example, when workers at a computer company in \\nTexas began organizing to protest decisions by management, they expected \\nto meet stiff opposition. But when management quickly recognized and \\nencouraged the worker organization, the group moved from concerns with \\nidentification to expressing their demands and seeking concessions (produc­\\ntivity). In this case, the identification phase of development was cut short \\nand the workers failed to establish a clear organization. The result was that \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n65 \\nthe union was disorganized and it soon disbanded, leaving many important \\nworker/management issues unaddressed. Success and the lack of outgroup \\nconflict resulted in the failure of the group to develop. \\nFrom a research standpoint, these observations have serious implications. \\nDetermining the precise stage of development must be based on careful \\nobservations of group discussion and behaviour; in this sense, the deter­\\nminations must be post hoc. Groups will diff\\ner in the amount of time spent \\nin each stage of development. Therefore, direct comparisons across groups \\nusing only time of existence as the independent variable will be difficult and \\nmust represent approximations. It is tempting to try to force a more con­\\ncrete calendar onto the group development process. However, this would \\nnot adequately represent the process and such efforts would not help in \\nunderstanding group development. While the fuzziness of the process is a \\nlimitation from an empirical sense, it is not fatal to research efforts, as we \\nshall see. Predictions about the types of behaviour that should dominate a \\ngroup or time period can be based on the model and these predictions can \\nbe tested. Analysis of group discussions and behaviours can identify stages \\nand allow empirical tests of specific hypotheses. And the model can be used \\nto explicate a wide range of intragroup and intergroup behaviours that \\ncannot be easily explained by existing, more static approaches to group \\nbehaviour. \\nGroup identity: some empirical evidence \\nEnough of this talk of 'fuzzy stages' and observation! Let's examine some \\nempirical data. But before undertaking this adventure, one point should be \\nmade. The theory of social identity and its close cousin, self-categorization \\ntheory (Turner, 1 987), deal with the individual. They argue that individuals \\nare driven to develop an identity, and, as a result of this desire, their behav­\\niour influences group and intergroup actions. Individuals are the driving \\nforces and groups are their vehicles. There is a great deal of support for this \\nposition, and I have no desire to attack it. However, my observations of \\ngroups suggest that the theories may be incomplete as they stand. For lurking \\nin the background of the research on group development is the suggestion \\nthat groups, like individuals, strive to develop their identity, separate from \\nthat of other groups and apart from the unique identities of the individual \\nmembers. \\nIndeed, I would suggest that groups are engaged in a struggle with other \\ngroups and with individual members to establish and maintain a group \\nidentity. The group identity often outlives the members. It is displayed in \\nsymbols such as flags, anthems, laws, group structure and costumes that are \\nenduring and not subject to the whims of individual members in search of \\ntheir own identities. Groups strive for their independence from other \\ngroups, and they struggle with group members to keep the identity of the \\ngroup equal to, if not more important than, individual identity. The striking \\n\\n\\n66 \\nSocial identity: international perspectives \\nmotto of the Three Musketeers drives home this point: 'All for one, one for \\nall [the group].' \\nIndeed, the model of group development may be viewed as a model of \\ngroup identity. The stages represent the group's struggle to establish its \\ncharacter, which is threatened by demands from individual members and \\nfrom outside groups. In fact, it is this conflict that ultimately gives the \\ndevelopmental process its energy, its dynamic. Where social identity theory \\nwould suggest that groups change and evolve to meet the demands and \\ndesires of their members, the group developmental approach argues that \\nthere is a more reciprocal relationship between the group and its member; \\nthe needs of individuals affect the development of the group, but the group, \\ntoo, is often the puppeteer, guiding the actions of the members to meet the \\nneeds of the group. \\nThis approach argues that the behaviour of individual members is \\naffected by the stage in which the group happens to be. Most pointedly, \\nwhen developmental patterns or external events challenge the group to \\nestablish a clear identity, individual members will respond to support this \\ngoal. Specifically, this position predicts that in the early stages of group \\ndevelopment, when the focus is on group identity, group members will \\ndirect both their intragroup and intergroup behaviours towards group \\nidentity. These forces stand apart from the need of the individuals to \\ndevelop their personal or social identities. In essence, when the group is bent \\non crafting its identity, individual members are forced to focus on the social \\n(group) side of their identities. \\nDuring the early stages of group development, behaviours within the \\ngroup and between groups should be directed towards securing the group's \\nidentity. On the intergroup side, there should be a rise in intergroup \\ncompetition and conflict designed to create clear group boundaries. Within \\nthe group, individuals will be 'encouraged' to sacrifice their independence \\nfor the sake of creating a unified, clearly distinguishable group identity. \\nSuch actions as high conformity, swift rejection of deviants and perceptions \\nof ingroup will foster this goal. Once the identity of the group is achieved, \\nthe group environment can become one that supports the goals of individual \\nmembers to establish their own unique identities (social and/or personal). As \\na result, latter periods of group development should be characterized by \\nfewer demands for personal sacrifice. Three studies, focusing on very \\ndifferent aspects of group behaviour, can be offered as evidence of this \\nposition. \\nPreferred intergroup interaction \\nRelations between the ingroup and outgroups are important and different at \\neach stage of development, but these relations are most vital during the \\nstage of identification. Nearly every aspect of group identity is defined in \\nterms of the outgroups. As individuals, we may have difficulty defining our \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n67 \\nplace in the universe, but at least we have our skin that defines our physical \\nbeing. Groups, however, have no predetermined boundaries. Hence, not \\nonly are they faced with establishing their physical boundaries, but they \\nmust, in a sense, create their own skin. Indeed, this is often the first step in \\nestablishing the identity of the group. \\nEstablishing group boundaries not only involves determining who is in \\nthe group, but it also requires describing who is not a group member. One \\nconvenient means for making a clear designation of group boundaries is \\nthrough intergroup conflict and competition. Such conflict plays multiple \\nroles in the identification process: it forces individuals to declare their \\nmembership in one group or another; it marks the boundaries of the groups; \\nit creates a common bond and purpose between members of the ingroup; \\nand it forces the group to clarify what it stands for and what it stands \\nagainst. Therefore, it can be predicted that group members will help their \\ngroups establish boundaries by seeking conflict and competition with out­\\ngroups during the early identification stage of group development. \\nOn the other hand, the model of development predicts that intergroup \\nconflict will be increasingly distasteful as the group continues its develop­\\nment. During the group productivity phase, such conflict will distract \\nmembers from task-oriented activities, although it may still serve the \\npurpose of enhancing cohesion within the group. During individuation, \\nmembers are concerned with making social comparisons with members of \\ntheir own group and with members of the outgroup. In attempting to \\nestablish their worth to their ingroup and negotiating rewards, they must \\noften demonstrate their potential value to other groups. Therefore, they \\nneed information from and acceptance by outgroups. Intergroup conflict \\nwould be disruptive to this purpose. And during decay, members are \\nactively seeking alternative groups. Conflict between their present group \\nand the alternative groups would clearly work against their individual goals. \\nIn an effort to examine these predictions from the model, groups of five \\nstudents assembled to work together for three one-hour sessions. They were \\ninformed that other groups would also be working in similar settings during \\nthis period. The groups were allowed to choose from a variety of tasks on \\nwhich to work and they could spend as long as they desired on each task. \\nTheir behaviour was video-taped during all the sessions and they responded \\nto periodic questionnaires throughout. \\nAn initial analysis of the sessions indicated that the groups were quite \\nsimilar in their pattern of development. During the early periods, time was \\nspent getting acquainted, exploring similarities in backgrounds and \\ninterests, establishing norms and roles, discussing the uniqueness of their \\nown groups, and agreeing to best the actions of other groups. As expected, \\nthe next order of business was on tasks: members examined and discussed \\nthe options, chose tasks and began talking about their past experience with \\nsimilar tasks, their contributions to the group work, and their expectations \\nfor the next task. In several of the groups, members jokingly threatening to \\n'send' productive members to the outgroups, and also in the form of jest, \\n\\n\\n68 \\nOnly \\ncooperation \\nt \\nI \\nOnly \\nSocial identity: international perspectives \\ncompetition '--\\n_\\n.L..-\\n_\\n--'--\\n_\\n-'-\\n_\\n---'--\\n_\\n----'-\\n_\\n----\\\"-\\n_\\n---'\\n \\n_\\n_\\n '--\\n_\\n-'---\\n----l\\n \\nBeginning \\nof group \\nEnding \\nof group \\nā - - - - - - - - - - - - - - - - - - - - - - - - - - -\\nTime of group existence \\nFigure 5.2 \\nNature o\\nf\\n desired interaction with outgroup as a f\\nunction o\\nf\\n \\ngroup lif\\ne ( Worchel, Coutant-Sassic & Wong, 1993, p. 84) \\nsome members threatened to defect to the outgroups if their present group \\ndid not treat them well. During the last session, many of the groups engaged \\nin a review of the group efforts, members talked about what they were \\ngoing to do after the experiment, and closure was the topic of the period. \\nMost interesting from the perspective of intergroup relations were \\nresponses on the written questionnaires. Several of these questions queried \\nsubjects about their feelings about their own group and desired relations \\nwith the other groups. As can be seen in Figure 5.2, subjects expressed a \\ndesire to engage in intergroup competition during the early periods of the \\ngroup. As the group aged, their desires changed markedly, and by the time \\nof group closing the desire was for cooperation with outgroups. Feelings \\nabout internal conflict showed the opposite pattern: extreme discomfort \\nduring the early stages and greater comfort by the final period (Figure 5.2). \\nThis is the pattern that would be predicted by the model on the assump­\\ntion that intergroup competition helps differentiate the ingroup from the \\noutgroup, and such differentiation is most critical during early stages \\ndevoted to group formation. The topics of group discussion also fit the \\npredicted pattern, focusing first on group identification and productivity \\nand later on more individual concerns. It should be pointed out that the \\ncyclical pattern of the stages was not observed, but this may have been due \\nto the fact that these groups had a defined end point. \\nThese results are not only interesting from a theoretical standpoint, but \\nthey may also carry a message for applied group dynamics. If groups \\nactively seek out competition and conflict with outgroups at certain periods, \\nit follows that efforts to reduce intergroup hostility during these periods will \\nnot be well received. Reducing intergroup hostility may only be possible \\nafter groups have been able to establish their identity, and do not f\\neel that \\ninteraction with the outgroup will threaten that identity. Sherif et al. (1961) \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n69 \\nfound that intergroup hostility was reduced only after several events \\ninvolving intergroup cooperation towards a superordinate goal. One expla­\\nnation for these results is that several cooperative encounters are necessary \\nbefore hostility is reduced. Another explanation argues that the timing of \\nthe cooperative endeavours is the critical factor in ameliorating this \\nhostility. Sherif's groups may have been most accepting of those final \\ncooperative situations which occurred towards the end of the session. \\nIngroup and outgroup perceptions \\nA number of studies have found that groups see greater homogeneity in the \\ncharacteristics of the outgroup than within their own group (Jones, Wood & \\nQuattrone, 198 1 ;  Linville & Jones, 1980; Park & Rothbart, 1 982). Explana­\\ntions for the effect include the fact that people have more and varied \\ncontact with ingroup members than with outgroup members. Therefore, \\nthey can readily see the variability of behaviour of ingroup members, but do \\nnot experience this variability in the limited contact with outgroup mem­\\nbers. Further, the contact with ingroup members generally occurs in a \\nvariety of situations, while contact with the outgroup is often in a single \\ntype of situation (Quattrone, 1 986). \\nThis interesting effect poses potential problems for the present model of \\ngroup development. The model suggests that during periods of identification \\ningroup members are most interested in finding commonalities between \\ntheir own group so as to create a group identity. However, later in the \\ndevelopmental cycle, especially during individuation and decay, the model \\nsuggests that members will be searching for differences between ingroup \\nmembers, thereby enhancing their own individuality and giving them reason \\nto search for other groups. The basic prediction, then, is that perceptions of \\nhomogeneity within groups will be dependent on the period of group \\ndevelopment in which these perceptions are measured. Perceptions of \\ningroup homogeneity should be greatest during the identification stage and \\nleast intense during individuation and decay stages. \\nIn order to examine this prediction, groups of eight subjects worked on a \\nseries of tasks (Worchel, Coutant-Sassic & Grossman, 1992). They were \\nthen divided into two groups of four, and the two groups competed on \\nother tasks. Subjects were asked to rate ingroup and outgroup members on \\na number of dimensions at different times during the competition (immedi­\\nately at the start of competition, after one competitive encounter and after \\ntwo competitive events). A final condition involved assigning subjects to \\ntwo groups on arrival, never allowing them to work as a single group, and \\nthen measuring their perceptions after one competitive event. Based on the \\nmodel of group development, it was predicted that members would perceive \\ningroups as more homogeneous than outgroups during the early phases of \\ncompetition. This was the situation where groups should be most concerned \\n\\n\\n70 \\nSocial identity: international perspectives \\nTable 5. 1 \\nPerceived dif\\nf\\nerences in ingroup and outgroup similarity \\nTraits \\nCondition \\nIntelligence\\\" \\nFriendliness \\nAppearance \\nCombined-questioned \\nimmediately \\n3.78b,c \\n3.84 \\n1 .24 \\nCombined-questioned \\nafter one task \\n2.97 \\n4.03 \\n0.97 \\nCombined-questioned \\nafter two tasks \\n-3. 1 1  \\n-1 .66 \\n-0.67 \\nNever combined \\n-2.69 \\n-2.07 \\n-2.00 \\nTalkativeness \\n2.47 \\n1 .33 \\n-0.1 2  \\n-1 .25 \\n\\\" Subjects were asked: How similar were the members of your group (outgroup) on the \\nfollowing dimensions: intelligence, friendliness, personal appearance and talkativeness. I = \\nvery dissimilar, 10 = very similar. \\nb Scores were determined by ingroup-outgroup. Positive score indicates ingroup more similar; \\nnegative score indicates outgroup perceived as being more similar. \\nc N = 6 groups in each condition. \\nSource: Worchel, Coutant-Sassic & Grossman, 1992, p. 198. \\nwith identity, especially if the group had just been formed from a larger \\ngroup, Later in the competition, there would be pressures to perceive the \\ndifferences within the group as the members focused on task and member \\nskills and contributions. \\nAs can be seen in Table 5 , 1 ,  the results supported these predictions and \\nthey illustrate the need to consider group development when examining \\ngroup perceptions, Ingroups were perceived as more homogeneous than \\noutgroups during the early stages of competition, but the ingroups were \\nseen as more heterogeneous than outgroups during the latter stages of \\ncompetition or when the groups had existed as separate units from the \\nbeginning of the experiment. It is also interesting that groups perceived \\ngreater absolute differences between the ingroup and outgroup during the \\nearly stages of competition than during later stages. \\nResponses to minorities \\nThere is voluminous literature demonstrating that minorities can exercise \\nconsiderable influence over group decisions (Moscovici, 1 976; Moscovici & \\nMugny, 1 983; Moscovici & Nemeth, 1974; Nemeth & Staw, 1 989), The \\nfocus of much of the research is on the most effective tactic for minorities to \\nfollow (Maass & Clark, 1 984) and on explaining the cognitive impact of the \\nminority (Nemeth, 1 986). Surprisingly, the research has not examined many \\ngroup factors in demonstrating the impact of the minority, In fact, minority \\ninfluence work has been conducted outside of group settings altogether. The \\ntypical experiment examines how subjects respond to learning of a statis­\\ntically rare or unexpected position, While the minority influence paradigm \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n7 1  \\nhad generated exciting data, it is important to bring the research into group \\nsettings. Indeed, minorities are defined by their relation to groups. \\nThe present model of group development suggests that the minority plays \\nan important role in the group process, but that the impact of the minority \\nis influenced by the stage of group development. On one hand, the voice of \\ndissent represents a clear threat to group cohesion and identity. Minorities \\ndifferentiate members within the group, and may present an image of a \\nvulnerable or disorganized unit. On the other hand, minorities can offer \\nvaluable information and new perspectives to the group. Their unmolested \\nexistence signals that dissent, disagreement and dissatisfaction with the \\ngroup are acceptable. Given these functions, it can be predicted that \\nminorities will be more influential and accepted in the group during the \\nstages of productivity, individuation and decay. On the other hand, \\nminorities will be quickly rejected during the identification stage when the \\ngroup is concerned with developing cohesion and uniformity. Furthermore, \\nthe influence of the minority will be dependent on the group's previous \\nexperience with minorities. If the experience has been that minorities are not \\ndisruptive and help move the group towards its goals, minorities will be \\nmore readily accepted than when the group's previous experience with \\nminorities has been negative. \\nMy students and I have completed several studies examining minority \\ninfluence. In one (y.I \\norchel, Grossman & Coutant-Sassic, 1993) we found \\nthat minorities were most influential in groups if they had previously \\ndemonstrated that their positions were correct and the issue of influence \\ndealt with productivity, not group identity. Minorities who were perceived \\nas a threat to group identity were excluded from the group; indeed, they \\nwere viewed as forming another outgroup. It was clear in this study that \\nminorities could exercise greater influence on some issues than others and \\nthat the group focus or activity determined which issues were subject to \\ninfluence. Previous work by Crano (1989) also found that minorities were \\nmore influential on some issues (objective) than others (subjective). This \\nresearch extends these findings by suggesting that when examined in a group \\ncontext the group will determine the issues on which minority opinion will \\nbe accepted. \\nIn a second study (Coutant, Worchel & Grossman, in preparation), \\ngroups of five subjects met together weekly over a period of six weeks. In \\neach session, they were given tasks on which to work. At different points \\nduring the life of the group, various manipulations were executed; the main \\nindependent variable was the time at which the manipulation took place \\n(during the first session, during middle sessions or during the last session). \\nIn the case of minority influence, the groups were given problems to discuss \\nand required to arrive at a group decision about the solution. A confederate \\nin the group played the role of the minority, and after ascertaining the \\nmodal opinion he proposed an extremely different position. The major \\ndependent measure was the influence of the minority on the group decision \\nat different periods of the group's life. \\n\\n\\n72 \\nSocial identity: international perspectives \\nTable 5.2 \\nMinority in\\nfluences in groups \\nover time \\nTime \\nEarly \\nMiddle \\nLate \\n4.24 \\n4.67 \\n4.85 \\nNote: Responses represent the group decision \\non how to deal with a case at hand. Minority \\ntook a position of lO on a scale of 1 - 1 0, \\ntherefore higher numbers indicate closer \\nagreement with the minority position. \\nAs can be seen from Table 5.2, the minority had little influence in the \\ninitial meeting of the group; in fact, the group moved away from his stated \\nposition. The minority was most influential when the influence attempt \\ncame during the closing group session. Experience with these ongoing \\ngroups indicated that the initial sessions were devoted to issues of group \\nidentity, while in the closing session individual members focused more on \\npersonal issues and the ending of the group. Hence, these results are in line \\nwith predictions based on the model of development: minorities will have \\nlittle influence during identification stages and greatest influence during \\nperiods of individuation and decay. In addition to supporting the specific \\npredictions taken from the model, this study and one finding similar results \\n(Coutant-Sassic, 1 991) illustrate the need to consider the factor of group \\ndevelopment when discussing a group 'effect' such as minority influence. \\nWe cannot speak of minority influence as a general effect; rather the influ­\\nence is closely tied to the issue and to the development of the group. In fact, \\nit might be interesting to speculate that the process of influence differs under \\nthe different conditions. The present view of minority influence (Nemeth, \\n1986) is that it involves a single process: creating divergent thinking in the \\nmajority. I would suggest that during the stage of identification when \\nthe minority is viewed as a threat to group identity convergent thinking \\non the majority position results from the presence of a minority point of \\nview. \\nThe task ahead \\nIt is time to take stock of where we now stand. In a general sense, the \\npresent model can be viewed as a model of group social identity. Groups, \\nlike individuals, must establish their own identity. This identity has two \\ndimensions: the identity of the group (similar to the personal identity of \\nthe individual in social identity theory) and the identity of the group as it \\nrelates to other groups in its universe (similar to the social identity of \\nindividuals). The search for identity influences both ingroup and intergroup \\n\\n\\nA developmental view o\\nf\\n the search f\\nor identity \\n73 \\nbehaviour as one identity or the other predominates. By way of summary, I \\nhave attempted to establish a number of points concerning group and \\nintergroup behaviour. First, individuals are engaged in a constant struggle \\nof balancing their desire to belong to groups and the concern of establishing \\ntheir personal independence and uniqueness. Events or situations that \\nthreaten their social identity will lead individuals to emphasize their rela­\\ntionship with their groups. Likewise, threats to their personal identity will \\nmotivate individuals to seek more intimate interactions, making their \\npersonal characteristics more salient. This struggle is felt at the group level \\nby members alternating between being active participants in group activities \\nand withdrawing themselves, physically and/or psychologically, from the \\ngroup. \\nSecond, groups are dynamic units that develop through a predictable \\nseries of stages. The development process is cyclical in that it can repeat \\nitself many times during the life of the group. The repetition occurs as the \\ngroup loses old members and gains new ones. The amount of time the \\ngroup spends in any single stage is dependent on numerous intragroup and \\nintergroup factors. The transition between stages can be abrupt and clear or \\nslow and gradual. Numerous events can trigger the movement from one \\nstage to another, including success, failure, leader behaviour, time pressures \\nand deadlines, and intergroup events. \\nThird, stages can be identified by the issue that dominates group behav­\\niour and discussion and involves the most group members. Stages not only \\ninfluence the focus of group activity, they also impact the interpersonal \\nbehaviours that will characterize the group. For example, minority influence \\nwill be most evident during individuation and decay stages, while con­\\nformity will be most common during the identification stage. Relating \\ndevelopmental stages to individual and interpersonal behaviour offers fertile \\ngrounds for developing specific predictions. \\nFourth, the development of the group is keenly affected by intergroup \\nrelations, and these intergroup relations are also influenced by the group's \\ndevelopment. For example, the group will be most concerned with diff\\ner­\\nentiating itself from outgroups during the identification period. This \\ndifferentiation process will be seen in efforts by the group to instigate \\nconflict and competition with the outgroup. On the other hand, perceived \\nthreats to the group's identity from the outgroup may throw the group into \\nthe identification stage, regardless of when these threats occur. \\nFinally, the developmental approach to group behaviour offers insights \\nthat may be used in intervention strategies. Those strategies that are \\ncongruent with the group's stage of development will be most effective. For \\nexample, it was suggested that prejudice and discrimination could be most \\neffectively reduced by interventions during the individuation or decay stages \\nof development. \\nThese are just a few of the advantages that result from expanding the \\nscope of social identity theory. The theory has a solid base, and it is time to \\nexplore pushing its boundaries. \\n\\n\\n74 \\nSocial identity: international perspectives \\nNote \\nWork on this chapter was supported in part by a research grant from the Texas Coordinating \\nBoard of Higher Education (T \\nARP). The chapter was completed while I was a Visiting \\nLecturer at the Department of Psychology, University of Waikato, Hamilton, New Zealand. I \\nwish to express my appreciation for the support I received from my colleagues at the University \\nof Waikato. I also want to acknowledge the help of several students who have worked with me \\non developing the ideas presented in this chapter: Dawna Coutant, William Webb, Frankie \\nWong, Michele Grossman, Judy Ouellete, Greg Pool and Leah Worchel. \\n\\n\\n6 \\nThe Role of Prototypicality in Group \\nInfluence and Cohesion: Contextual \\nVariation in the Graded Structure of \\nSocial Categories \\nPenelope Oakes, S. A lexander Haslam and John C. Turner \\nThe concept of prototypicality \\nThe concept of proto \\ntypicality was one product of the 'Roschian revolution' \\n(Neisser, 1 987, p. vii) of the 1 970s, which introduced some key new ideas \\nto the analysis of category structure and representation. As is now well \\nknown, Rosch challenged the philosophically based, classical view of \\ncategory representation with her ideas about, and evidence of, extensive \\nvariation in category structure. She and her colleagues (for example, Rosch, \\n1978; Rosch, Mervis, Gray, Johnson & Boyes-Braem, 1 976) found that \\ncomparisons between categories reveal that they vary in their relative \\ninclusiveness, for example 'dalmation', 'dog', 'animal', 'living thing', are \\ncategories of increasing inclusiveness. Those at a higher level are called \\nsuperordinate, compared to lower level categories known as subordinate. \\nAround the middle of the hierarchy of inclusiveness is the basic level (for \\nexample, 'dog', 'chair', 'tree') at which, it is argued, most perceptual activity \\ntakes place (Mervis & Rosch, 1 98 1 ;  see also Lassaline, Wisniewski & \\nMedin, 1 992). More importantly for our present purposes, Rosch also \\nfound that within categories, members vary in their typicality, for example \\nrobins are seen (by American subjects) as more typical of the category 'bird' \\nthan are ostriches, but both robins and ostriches are recognized as sharing \\nmembership in the one 'bird' category. In other words, it appears that \\ncategories have an internally graded structure, rather than members possess­\\ning an even level of shared defining attributes, as in the classical view. It was \\nthis finding that led to the idea of a category prototype (a best example of \\nthe category), and the argument that category membership requires a \\ncertain level of similarity to the prototype. Rather than being tightly \\nbounded, categories seem to take the form of 'fuzzy sets', much looser \\ngroupings which depend upon what Rosch called 'family resemblance', that \\nis, members do not all share a given set of defining features, but are related \\nthrough their similarity to a prototype. \\n\\n\\n76 \\nSocial identity: international perspectives \\nIn her most widely cited paper Rosch (1978) tried to counter some \\npossible misunderstandings of her ideas. In particular, she emphasized that, \\nalthough her work indicated that prototypes served important functions in \\ncategorization, the concept did not 'constitute any particular model of \\nprocesses, representations, or learning' (p. 40). Indeed, she commented: \\nTo speak of a prototype at all is simply a convenient grammatical fiction; what is \\nreally referred to are judgements of degree of prototypicality \\n. . . .  \\nto speak of a \\nsingle entity that is the prototype is either a gross misunderstanding of the \\nempirical data or a covert theory of mental representation \\n. . . .  \\nIn short, proto­\\ntypes only constrain but do not specify representation and process models. \\n(pp. 40-4l )  \\nPrototypicality in self-categorization theory \\nSelf-categorization theory (Turner, 1985; Turner, Hogg, Oakes, Reicher & \\nWetherell, 1987) is often introduced as a theory of processes underlying \\npsychological group formation, and indeed its original mission was to \\nexplain the psychological basis of the group. In effect, however, it attempts \\nmore than this, and one of its most important contributions has been the \\ndevelopment of a new perspective on the functioning of categorization \\nprocesses in social perception and interaction (cf. Brewer, 1988; Fiske & \\nNeuberg, 1990). We shall not attempt to provide a comprehensive overview \\nof the theory here, but it is worth setting down some of the main points \\nconcerning categorization in order to place our discussion of prototypicality \\nin its proper context (for more detailed discussion of categorization, see \\nOakes, 1996). \\nIn discussing the role of categorization in social perception, the theory \\napplies and develops Rosch's ideas about variation in both relative \\ninclusiveness across categories and relative prototypicality within categories. \\nThe theory begins with the assumption that self-conception reflects self­\\ncategorization, the cognitive grouping of the self as identical to some class of \\nstimuli in contrast to some other class of stimuli. Following Rosch (1978), \\nself-categories are assumed to exist at different levels of abstraction related \\nby class inclusion. That is, a given self-category (for example, 'scientist') is \\nseen as more abstract than another (for example, 'biologist') to the extent \\nthat it can contain the other, but the other cannot contain it: all biologists are \\nscientists, but not all scientists are biologists. Going beyond the simple \\ndistinction between personal identity (self as a unique individual) and social \\nidentity (self as a member of a social group) presented in earlier work (for \\nexample, see Turner & Giles, 1981), self-categories both more and less \\nabstract than personal and social identity are envisaged, but for purposes of \\ntheoretical exposition three levels of abstraction of self-categories are \\ndistinguished: the interpersonal (subordinate level of abstraction, personal \\nidentity, self as an individual person), intergroup (intermediate level of \\n\\n\\nThe role o\\nf\\n pro \\nto \\ntypicality \\n77 \\nabstraction, social identity, self as a group member) and interspecies (super­\\nordinate level of abstraction, self as a human being). These are defined not by \\nspecific attributes but by the level at which people are being compared and \\ncategorized. For instance, 'altruism' could function as a cue to individual \\nidentity, to a particular social category, or to being human, depending on the \\ncontext. \\nIn terms of the processes underlying group formation, the central \\nexplanatory idea in the theory is that group behaviour is made possible by \\nand reflects self-concept functioning at the level of ingroup-outgroup (rather \\nthan interpersonal or interspecies) categorizations. Developing TajfeJ's \\n(1978d) idea of an interpersonal-intergroup continuum, self-categorization \\ntheory contrasts the social and personal (ingroup-outgroup and inter­\\npersonal) levels of self-categorization, and hypothesizes that shifts between \\nthese levels are a varying outcome of a continual conflict between personal \\nand social identity as they vary in relative salience. Shifts towards social \\nidentity produce depersonalization of self-perception and behaviour, that is, \\nself-stereotyping, perception of increased identity between the self and \\ningroup members and increased difference from outgroup members (on \\nrelevant dimensions). The theory proposes that it is this process of \\ndepersonalization that makes group behaviour possible and produces its \\nemergent, irreducible properties. \\nIn retaining (from social identity theory) this emphasis on the two levels \\nof personal and social identity, the theory appears to be suggesting that \\nthere are, in Rosch's terms, two 'basic' levels in the cognitive representation \\nof the self and others, the social and the personal. In truth, however, this \\ndistinction is maintained only for purposes of theoretical clarity in \\nexplaining the hypothesized basis of psychological group membership. One \\nof the essential points of the self-categorization analysis is that the appro­\\npriate level o\\nf\\n categorization varies with the context, and that to speak of any \\nlevel as more 'basic' than another is misconceived. \\nTo develop this point further, the theory portrays categorization as a \\ndynamic, context-dependent process, determined (in part) by comparative \\nrelations within a given context (for example, see Haslam, Oakes, Turner & \\nMcGarty, 1 996). This approach is formalized in the principle of meta­\\ncontrast, which is so called because it involves a contrast between contrasts, a \\njudgement of difference between differences. The meta-contrast principle \\npredicts that a given set of items is more likely to be categorized as a single \\nentity to the degree that dif\\nf\\nerences within that set o\\nf\\n items are less than \\ndif\\nf\\nerences between that set and others within the comparative context (cf. \\nCampbell, 1 958; Rosch, 1978; Taj fel \\n, 1969a; Tversky, 1 977). This principle \\nencapsulates but subtly transforms the classic idea that categories form on \\nthe basis of intraclass similarities and interclass differences. It goes further \\nthan earlier work in making clear that such similarities and differences are \\nnot independent and additive, but aspects of the same meta-contrast (cf. \\nMedin, 1 989; see below). The meta-contrast principle is explicit that \\ncategorization is relative to the frame of reference and hence to the contrasts \\n\\n\\n78 \\nSocial identity: international pers\\npectives \\nFigure 6.1 \\nA categoria! hierarchy o\\nf\\n vegetation \\navailable within the salient stimulus field. It depends not just on similarities \\nbetween stimuli (as is so often assumed), but on relative similarities, more \\nsimilarity between certain stimuli than between others. Meta-contrast \\ncontextualizes categorization by tying it to an on-the-spot judgement of these \\nrelative similarities and differences. \\nFor example, consider the system of natural categories in Figure 6. 1 .  \\nMeta-contrast predicts that we would categorize and perceive a given piece \\nof vegetation as, say, a 'tree' (rather than as 'vegetation') to the extent that, \\nin the current comparative context, the differences between trees (oaks, \\nbirches, and so on) are perceived to be less than the differences between \\n'trees' and 'shrubs' - the distinction between trees and shrubs is more \\nmarked, and more relevant, than are the features that trees and shrubs \\nshare as 'vegetation'. Alternatively, the tree might be categorized and per­\\nceived simply as 'vegetation' to the extent that differences between types of \\nvegetation (trees, shrubs, and so on) are seen to be less than the differences \\nbetween vegetation and, say, animals. The salient categorization will be \\n'oak' when perceived differences between individual oak trees are less than \\nthe differences between oaks and birches (or some other comparison \\nspecies). \\nThe meta-contrast principle applies similarly in person perception. \\nConsider the hierarchy in Figure 6.2. Here, we might categorize an \\nindividual as 'Spanish' to the extent that, in the current comparative \\ncontext, perceived differences between individual Spaniards (Carmen, \\nIsabella, and so on) are less than those between Spaniards and Germans. \\nAlternatively, the salient category might be 'European' in a context \\nwhere perceived differences between various European groups (such as \\nSpaniards and Germans) are less than those between Europeans and non­\\nEuropeans. \\nThese ideas can be expressed in terms of a meta-contrast ratio, that is, the \\naverage perceived intercategory difference divided by the average perceived \\nintra-category difference. For example, this might be: \\n\\n\\nThe role o\\nf\\n pro \\nto \\ntypicality \\nFigure 6.2 A categorial hierarchy o\\nf\\n Europeans \\naverage perceived difference between Spaniards and Germans \\naverage perceived difference between individual Spaniards \\n79 \\nThe higher this ratio, the more likely it is that Spaniards will be perceived in \\nterms of their shared national identity. Turner and Oakes (1989, p. 242) \\nprovide specific examples of how meta-contrast can be calculated, as do \\nHaslam and Turner (1 992) and some of the papers exploring prototypicality \\nin more detail that we shall discuss below (e.g. McGarty, Turner, Hogg, \\nDavid & Wetherell, 1 992). \\nNote, however, that meta-contrast provides only a partial account of \\ncategorization. It describes the comparative relations between stimuli which \\nlead them to be represented by a category, but it is also important to take \\ninto account the social meaning of differences between people in terms of \\nthe normative and behavioural content of their actions, and the relative \\naccessibility of particular categorizations (Oakes, 1987; Oakes, Turner & \\nHaslam, 1 991). In general terms, the theory explains the salience of any \\ngiven category as a function of an interaction between its relative accessi­\\nbility (the 'readiness' of a perceiver to use a particular categorization) and \\nthe 'fit' between the category and reality. Meta-contrast is the theory's \\nprinciple of 'comparative fit' (the match between category and the com­\\nparative properties of stimuli), but 'normative fit' (the match between \\ncategory and the content properties of stimuli) is also always inseparably at \\nwork (Oakes et aI., 1991). \\nThis emphasis on categorization as highly variable and context-dependent \\nproduces a concomitant emphasis on the context-dependence of perceived \\nsimilarity and difference, the major outcome of categorization. People who \\nare categorized and perceived as different in one context (for example, \\n'biologists' and 'physicists' within a science faculty) can be recategorized \\nand perceived as similar in another context (for example, as 'scientists' \\nrather than 'social scientists' within a university) without any actual change \\n\\n\\n80 \\nSocial identity: international perspectives \\nin their own positions (Gaertner, Mann, Murrell & Dovidio, 1989). This is \\nthe essence of categorization: it is a cognitive grouping process that trans­\\nfonns differences into similarities and vice versa. We need some psycho­\\nlogically neutral tenn such as perhaps 'distances' to indicate precognized \\nstimulus relations: there are distances between people, but are they \\nsimilarities or differences? Are physicists and biologists similar or different? \\nArising from the comparisons specified in the meta-contrast principle, \\ncategorization subjectively transfonns 'distances' into similarities and differ­\\nences, and from perceived similarities and differences flow, amongst other \\nthings, perceptions of attraction and dislike, agreement and disagreement, \\ncooperation and conflict. In sum, categorization provides the fundamental \\nbasis of our social orientation towards others. Within the science faculty, \\nphysicists might reject and deride biologists, claiming they aren't 'real \\nscientists', but in comparison with social scientists the two groups may \\npresent as inseparable allies. \\nAs changes in the comparative context produce changes in the ingroup­\\noutgroup relationships, they also affect intra-category structure, and this is \\nwhere the concept of prototypicality becomes important. The meta-contrast \\nprinciple can also be used to define the relative prototypicality of members \\nwithin a group (see Turner & Oakes, 1989, pp. 259-265). In general tenns, \\nthe more a group member differs from outgroup members and the less he or \\nshe differs from other ingroup members (that is, the more this person \\nexemplifies what in \\ngroup members share and what they do not share with \\nthe outgroup), the more that individual will be perceived as prototypical of \\nthe group. So, for example, in a communist party, members wishing to be \\nperceived as prototypical communists must differ in politics from members \\nof more conservative, capitalist parties, but they must not be so ultra-left \\nthat they begin to differ significantly from other communists - prototypi­\\ncality depends upon both inter- and intragroup comparisons. One important \\npoint is that, because relative prototypicality depends on (amongst other \\nthings) intergroup comparisons, it will vary along with variation in the \\nintergroup context in which judgements are made. For example, the proto­\\ntypical communist in a comparative context including fascists will be \\ndif\\nf\\nerent from the prototypical communist as compared with liberal \\ndemocrats, or Trotskyists. Self-categorization theory thus agrees with Rosch \\n(1978) that fixed prototypes are 'fictions', and emphasizes context­\\ndependent judgements o\\nf\\n proto \\ntypicality rather than fixed prototypical \\nimages which represent groups as constants across changing contexts (cf. \\nBrewer, 1988; Brewer, Dull & Lui, 198 1 ;  see below). \\nThere are further aspects of the analysis of categorization presented within \\nthe theory (for example, the issue of the relationship between the different \\nlevels of abstraction, the dependence of comparison and categorization upon \\nidentity at a higher level of abstraction), but these issues are not central to \\nour current discussion and are covered elsewhere (for example, Oakes, \\nHaslam & Turner, 1994; Turner & Oakes, 1989). The major point to \\nemphasize here is the context-dependence and fundamentally comparative \\n\\n\\nThe role o\\nf\\n prototypicality \\n81 \\nnature of \\nthe categorization process, and the consequent variability of \\nboth \\nthe level and the content (including the prototypical position) of social \\ncategories. \\nPrototypicality and group polarization \\nGroup polarization refers to the tendency for group interaction (for \\nexample, discussion or some related activity) to move the average of group \\nmembers' responses closer to the extreme of an already favoured position. \\nFor example, after discussing feminism a group of moderately pro-feminist \\nindividuals would tend to become, on average, more pro-f\\neminist than they \\nhad been initially. There is a large and complex literature related to group \\npolarization (Turner, 1 991), and our aim here is not to review that \\nliterature, nor to enter into any debate about the relative merits of different \\nexplanations of the phenomenon, but simply to show how the concept of \\nprototypicality as conceptualized in self-categorization theory has enabled \\nresolution of the apparent contradiction between polarization and other \\nsocial influence phenomena (for more detail see Turner, 1991 ; Turner & \\nOakes, 1989; Wetherell, 1987). \\nGroup polarization has been seen as inconsistent with a central tenet of \\nthe traditional informational influence model of conformity derived from \\nthe findings of Allport, Sherif and Asch - the idea that social norms form \\nthrough a process of interpersonal averaging or convergence. Norms are \\nthought of as reflecting what the group shares, and this has been concep­\\ntualized in terms of a simple conglomeration of what each separate indi­\\nvidual brings to the group, specifically the average of individual members' \\nseparate views. Given this view of norms, specialized theories have had to \\nbe developed to explain polarization as something more than or different \\nfrom conformity to group norms (Burnstein, 1982; Sanders & Baron, 1 977; \\nWetherell, 1 987), because in polarization group members do not converge \\non the average of their individual positions. Something occurring within the \\ngroup moves them towards a position more extreme than the average. \\nIt has long been evident that polarization could be explained as conformity \\nif it could be shown how ingroup norms could sometimes be more extreme \\nthan the mean (for example, Singleton, 1979). To do this, one needs to \\nrethink the interpersonal averaging approach to the development of group \\nnorms, and there are additional reasons for doing this. The interpersonal \\naveraging approach is individualistic (Turner & Oakes, 1986), implying that \\nsocial influence is simply a 'change in individuals induced by individuals' \\n(Kiesler & Kiesler, 1969, p. 26). In contrast, self-categorization theory argues \\nthat group behaviour (including social influence) reflects and is made \\npossible by a change in the level of abstraction of self-categorization, a \\nqualitative shift in both self-conception and the nature of relations with \\nothers. In this sense, group interaction does not involve individuals whose \\nidiosyncratic opinions can be averaged, but group members whose behaviour \\n\\n\\n82 \\nSocial identity: international pers\\npectives \\nwill be guided by the distinctive, emergent, irreducible properties of their \\ngroup. The group norm embodies those group properties, and reflects both \\nwhat group members have in common (on dimensions relevant to the group) \\nand the way in which the group can be distinguished from relevant \\noutgroups. Group norms (and other distinctive group properties such as \\nstereotypes) reflect the comparative relations within which the group defines \\nitself as well as, and as much as, they reflect attributes of group members. \\nMore specifically, the position perceived as normative for a group in a \\ngiven context will be the position that is most prototypical in that context, \\nas defined by the principle of meta-contrast (see above). It will be the \\nposition which is, on average, most diff\\nerent from that of relevant outgroup \\nmembers and least diff\\nerent from that of other ingroup members (obtained \\nby dividing the individual's average difference from outgroup members by \\nhis or her average difference from ingroup members). Of course, the posi­\\ntion least different from that of other ingroup members is the mean, but the \\nmean may not be the position most different from the outgroup. Thus, \\nunder certain intergroup conditions (see below) the prototypical position \\nwill be skewed away from the mean. \\nThe extent to which the mean does not represent the position most \\ndifferent from the outgroup, and hence the degree of discrepancy between \\nthe most prototypical position and the mean, depends on the comparative \\ncontext within which the ingroup defines itself. In general, the ingroup \\nprototype will tend to coincide with the mean where the mean is at the \\nmidpoint of the comparative context, that is, the ingroup is occupying the \\n'middle ground' on some issue relative to relevant outgroups. When ingroup \\nresponses occupy a more extreme position in the comparative context (as a \\nfunction of either ingroup responses shifting towards an extreme, or a \\nchange in intergroup comparison making moderate responses appear more \\nextreme), it becomes more likely that the most prototypical position will \\ntend to be more extreme than the mean. For example, insofar as a group of \\nmoderately left-wing students is seen to occupy the middle ground in \\nstudent politics within their institution (that is, there are other salient groups \\nto both the left and the right of the moderates), the average view in the \\ngroup will be more prototypical. However, if either individual opinions in \\nthe group shift to the left, or the comparative context comes to be domi­\\nnated by, say, fascists and other extreme conservatives, a more extremely \\nleft-wing position within the moderate group will become prototypical. The \\naverage position would no longer adequately represent the difference of the \\nmoderates from the right-wingers. Examples specifying this process of \\ncalculating the prototypical position (and its variation from the mean \\nposition) through the principle of meta-contrast can be found in McGarty et \\nal. (1992), and a number of other sources (for example, Turner et aI., 1987). \\nSeveral studies provide evidence consistent with the self-categorization \\nanalysis of group polarization (for example, Abrams & Hogg, 1990; Abrams, \\nWetherell, Cochrane, Hogg & Turner, 1 990; Hogg, Turner & Davidson, \\n1 990; Mackie & Cooper, 1 984; McGarty et aI., 1992; Spears, Lea & Lee, \\n\\n\\nThe role o\\nf\\n pro \\nto \\ntypicality \\n83 \\n1990; Turner, Wetherell & Hogg, 1989; van Knippenberg & Wilke, 1 988). \\nFor example, McGarty et al. (1 992) present correlational evidence (from a \\ncomputer simulation and two semi-naturalistic group discussion studies) that \\ningroup prototypes do become more polarized as the group becomes more \\nextreme in the social context. They also show that the magnitude of group \\npolarization is a function of the degree to which groups actually shift \\ntowards their prototypes together with the degree to which those prototypes \\nare polarized. \\nOne novel prediction from the self-categorization analysis is that the \\ndirection of polarization can be reversed by changing the comparative \\ncontext, for example moderate left-wingers should move to the left in a \\ncontext dominated by the right wing, but to the right when comparing \\nthemselves with ultra-leftists. Consistent with this prediction, Hogg et al. \\n(1 990) report findings from a decision-making study in which subjects \\nperceived the ingroup norm as riskier than the pretest mean when con­\\nfronted by a more cautious outgroup, but as more cautious than the pretest \\nmean when confronted by a riskier outgroup. \\nIn summary, self-categorization theory explains group polarization as \\nconformity to an extremitized group norm, and explains the extremitization \\nof the group norm through an account of the way in which comparative \\ncontext influences norm formation. The concept of proto \\ntypicality has been \\nimportant here in specifying the position that will be most normative for a \\ngiven group in a given comparative context: the position with the highest \\nmeta-contrast (maximal similarity to ingroup positions and maximal \\ndifference from outgroup positions) is the most prototypical of the group \\nand hence the most consensual and normative. \\nPrototypicality and social attraction \\nPrior to the 'cognitive redefinition' (Turner, 1982) of the social group \\nrepresented by social identity theory and self-categorization theory, the \\nsense of group belongingness and attraction within groups was understood \\nto rest simply on inter\\npersonal attraction between individual group mem­\\nbers. This was the essence of the traditional concept of group cohesiveness \\n(Lott & Lott, 1 965), the affective and motivational 'glue' that held groups \\ntogether. However, in the same way that the interpersonal averaging \\nconception of group norms has been reconsidered in self-categorization \\ntheory, the interpersonal attraction account of group cohesiveness has also \\nbeen challenged (Hogg, 1 987, 1 992, 1 993). It has long been evident that \\ninterpersonal attraction is neither necessary nor sufficient for group \\nbehaviour (for example, Hogg & Turner, 1985a; Turner, Hogg, Oakes & \\nSmith, 1 984; Turner, Sachdev & Hogg, 1 983), but that, at the same time, \\nthere are strong bonds of attraction within groups (for example, Sherif, \\n1967), including 'minimal' groups (for example, Brewer & Silver, 1 978; \\nBrown & Turner, 1979; Doise, Csepeli, Dann, Gouge, Larsen & Ostell, \\n\\n\\n84 \\nSocial identity: international perspectives \\n1972). Self-categorization theory has introduced the idea of depersonalized \\nsocial attraction to distinguish between interpersonal liking (which seems to \\nbe unrelated to group processes) and the intragroup liking that accompanies \\nsalient group membership and group interaction. \\nHogg (1992) summarizes the self-categorization analysis of intragroup \\nattraction thus: \\nthe core idea [is] that self-categorization depersonalizes the basis of interindividual \\nattraction, such that ingroup individuals (and self) are liked in proportion to their \\nperceived group prototypicality. (p. 125) \\nAgain, we have an emphasis on the qualitative distinction between group \\nand individual behaviour. Group cohesiveness and intragroup attraction are \\nbased on group members' liking of each other as group members, not as \\nunique individuals, and it is prototypicality that defines the extent to which \\nindividuals are or are not group members. Recall that prototypicality \\ndepends upon both similarity to other ingroup members and difference from \\nthe outgroup. The closer any individual group member comes to this ideal, \\nmaximally normative ingroup position, the more he or she will be liked by \\nother ingroup members, and it is this qualitatively distinctive depersonalized \\nsocial attraction that holds groups together, rather than the individual \\npersonal attraction discussed in earlier models (Hogg, 1 992, 1 993). \\nHogg and his colleagues have tested a number of predictions from this \\nself-categorization model of depersonalized attraction (for example, Hogg, \\nCooper-Shaw & Holzworth, in press; Hogg & Hardie, 1991 ,  1992). Hogg et \\nal. (in press, Study I )  tested the basic hypothesis that, where group mem­\\nberships rather than interpersonal relationships are salient, liking among \\ngroup members is depersonalized in terms of the group prototype. They \\nstudied six small, mixed-sex interactive work groups within a large tertiary \\neducational institution, who were asked to report patterns of intragroup \\nliking under conditions that either heightened or reduced the salience of \\ntheir work group. They also rated the cohesiveness of the group, described \\nthe group prototype, ranked the prototypicality of fellow members, rated \\ntheir own prototypicality, and rated the subjective clarity of the group \\nprototype. As predicted, it was found that group liking and interpersonal \\nliking were independent, and that group liking was positively associated \\nwith the proto \\ntypicality of both the self and others, perceptions of strong \\ngroup cohesiveness and a clear group prototype. Interpersonai liking, on the \\nother hand, was either unrelated or negatively related to these variables, but \\nit was associated with perceptions of interpersonal similarity. Overall, then, \\nthe study produced clear evidence of depersonalized, category-based intra­\\ngroup attraction as distinct from interpersonal liking based on interpersonal \\nsimilarity. A second study involving 1 5  mixed-sex groups of between 6 and \\n10 introductory psychology students produced similar findings. \\nHogg and Hardie (1991) report a field study of depersonalized attraction. \\nMembers of an Australian football team completed a questionnaire at a \\npractice session which assessed subjective perceptions of prototypical \\n\\n\\nThe role o\\nf\\n proto \\ntypicality \\n85 \\nfeatures of the team, and a rating of self-prototypicality. Team members \\nthen ranked each other in terms of prototypicality, social attraction (elicited \\nunder conditions of high team membership salience) and personal attraction \\n(elicited where interpersonal friendships were emphasized). \\nTeam members' responses indicated significant agreement about the team's \\nprototypical f\\neatures. The most important were a sense of camaraderie \\n(,mateship', and unqualified mutual support), task-oriented mutual encour­\\nagement ('revving-up') and participation in good-natured 'horsing-around' \\nand socializing. As predicted, in the team-salient condition there was a more \\nsignificant relationship between subjects' liking for others (social attraction) \\nand perceived prototypicality than between interpersonal liking and \\nperceived prototypicality, and this effect was most marked for members \\nwho identified most strongly with the team and saw themselves as highly \\nprototypical. Social popularity (that is, being consensually liked as a team \\nmember) was also strongly related to consensual prototypicality, but the \\nrelationship between personal popularity (that is, being consensually liked as \\na personal friend) and prototypicality was weaker. \\nIn summary, Hogg and his colleagues present a strong case for distinction \\nbetween social attraction and personal attraction, with the former uniquely \\nrelated to group processes and involving the perception of self and others in \\nterms of group prototypicality rather than individual characteristics. The \\nmechanisms underlying the relationship between perceived prototypicality \\nand attraction are being investigated further (see Hogg, 1992, 1993): for \\nexample, does mutual self-categorization produce attraction simply through \\nthe accentuation of prototypical similarity, or is group prototypicality in \\nsome way inherently attractive, and might this be because intergroup pro­\\ncesses work to produce ethnocentric prototypes which portray the ingroup as \\nmore attractive than comparison groups (for example, see Dunning, Perie & \\nStory, 1 99 1 ,  for evidence of the self-serving nature of prototypes)? Clearly, \\nthe precise nature of depersonalized attraction is not yet fully understood, but \\nthe fact that there is a distinction between group-prototypical and inter­\\npersonal liking has been established, and further emphasizes the irreducible \\nnature of group processes. \\nThe contextual variability of prototypicality \\nAs we noted above, Rosch (1 978) emphasized that her analysis of the graded \\nnature of within-category structure focused on judgements o\\nf\\n prototypicality \\nrather than any notion of fixed prototypes, and she insisted that prototypes \\nwere not to be taken as a model of category representation or processing. \\nSimilarly, self-categorization theory emphasizes the context-dependence of \\nprototypicality, seeing it as a variable outcome of the categorization process. \\nIndeed, contextual variability is built in to the self-categorization definition of \\nprototypicality, in that it depends upon both within-category similarity and \\nbetween-category difference. Thus, at the very least, prototypicality for a \\n\\n\\n86 \\nSocial identity: international perspectives \\ngiven category will vary as the nature of relevant comparison groups varies, \\nand as the wider context in which those groups are perceived varies. For \\nexample, we have already cited evidence from Hogg et al. (1990) that a group \\ncan perceive itself as prototypically risky or prototypically cautious \\ndepending upon whether it compares itself with a more cautious or more \\nrisky outgroup respectively. \\nMore recently, two experiments conducted by Haslam, McGarty, Oakes, \\nTurner and Onorato (1995) directly tested self-categorization theory's \\nassertion that the relative proto \\ntypicality of extreme outgroup members \\nwould increase to the extent that an intergroup definition of the judgemental \\ncontext was made salient. This prediction follows from the principle of \\nmeta-contrast discussed above: in intragroup contexts individuals embody­\\ning the mean group position should be more representative of the group as \\na whole (since they maximize intragroup similarity), but in intergroup \\ncontexts polarized members of both ingroup and outgroup should be more \\nrepresentative (as they also maximize intergroup difference). \\nIn Experiment I all subjects watched the same video-taped message from \\na woman who discussed the causes of road accident-induced brain damage \\nand the means by which it could be reduced. They were then asked to make \\njudgements about the degree to which she was representative of the group \\nto which she was said to belong. In one of eight independent conditions she \\nwas described as either an extreme or a moderate member of either an \\ningroup (a group that wanted to improve road safety) or an outgroup (a \\ngroup that wanted to ban the sale and consumption of alcohol in order to, \\namongst other things, reduce the road toll). The salience of subjects' own \\nsocial identity was also manipulated so that half the subjects were simply \\ngiven information about the target's position (that is, low salience) but half \\nalso had to state whether or not they themselves were in favour of banning \\nalcohol/improving road safety (high salience). \\nAs anticipated, the perceived proto \\ntypicality of the target varied with her \\nalleged position within her own group (moderate or extreme) and with the \\nextent to which an intergroup relationship between subjects and target was \\nsalient. More specifically, when the target was described as an outgroup \\nmember she was seen as most representative of her group in the condition \\nwhere subjects' own group membership was salient and she was described as \\nan extremist. In other words, in the context of a salient intergroup division, \\nan extremist was more representative of the outgroup than a moderate. \\nA similar pattern also emerged from a second study in which all subjects \\nwatched a video of a discussion in which a group of four people debated \\nissues about crime and punishment and came to a pro-authority conclusion. \\nFor example, the group disagreed with the statement 'criminals should be \\nhelped rather than punished' and agreed with the statement 'the best way to \\nsolve the drug problem is for the present laws to be made much stricter'. In \\none (low salience) condition subjects then simply made judgements about \\nhow representative of the group as a whole were (a) the group member who \\nwas most pro-authority and (b) the group member who was least pro-\\n\\n\\nThe role o\\nf\\n pro \\nto \\ntypicality \\n87 \\nauthority. In two further conditions the intergroup nature of the judge­\\nmental context was made more salient. In the medium salience condition \\nsubjects were asked to stage their own (generally anti-authority) views \\nbefore they saw the video. In the high salience condition they were given a \\nlecture about the psychological basis and negative social consequences of \\nauthoritarianism before stating their own views, and the views of all sub­\\njects were then collated on a blackboard. As predicted, to the extent that the \\nexperimental manipulations served to make intergroup differences between \\nthe subjects and the stimulus group salient, the relative prototypicality of \\nthe extremely pro-authority target increased so that this individual was seen \\nas more representative of the group as a whole than the individual who was \\nonly moderately pro-authority. In higher salience conditions subjects were \\nalso less likely to agree with the extremist and more likely to characterize \\nthe group as a whole in negative terms. \\nTaken together, these studies confirm the point central to self­\\ncategorization theory that the degree to which a person is perceived to be \\nrepresentative of a group is not simply a function of properties of that \\nperson considered in isolation. Instead, the very same individual will be \\nperceived as more or less prototypical of a social category depending on the \\nsocial context within which he or she is defined. More specifically, both \\nstudies show that extreme members of outgroups will be seen as more \\nrepresentative of the groups to which they belong to the extent that \\ndifferences between ingroup and outgroup are made salient and perceivers \\ntherefore conceptualize the judgemental context in intergroup rather than \\ninterpersonal terms. \\nTo what extent does other social psychological work on prototypicality \\ntake into account this context-dependence, and the idea of prototypicality as \\na variable outcome of categorization? In general, prototypes have been \\ndiscussed as the basis for category representation, with a concomitant \\nemphasis on identifying the prototype of a given category and the effects of \\nsimilarity to this prototype on information processing, rather than issues of \\nprototype variability. For example, Brewer and her colleagues were \\namongst the first to apply Rosch's ideas in social psychology in a series of \\nstudies examining prototypes of elderly people (Brewer et a!., 198 1 ;  Brewer \\n& Lui, 1984). This work offers the following view of prototypes: \\nCategories are represented cognitively by prototypes - actual or imaginary \\ninstances of the category that contain attributes most representative of items \\ninside the category and least representative of items outside the category. Once a \\nprototype of a category has been formed, membership in the category is assessed \\nin terms of 'prototypicality' or perceived similarity to the prototypic instance. \\n(Brewer et aI., 198 1 ,  p. 656) \\nThe assumption is that category representation is in terms of prototypes \\n(cf. Rosch, 1978), and that, once formed, prototypes are the fundamental \\nbasis of categorization decisions, that is, it is similarity to the represented \\nprototype that determines whether or not an item is perceived as a member \\nof a category. In later work, Brewer (1 988) has developed this analysis of \\n\\n\\n88 \\nSocial identity: international pers\\npectives \\nprototypes further to argue that categories are represented in terms of \\npictoliteral prototypes, picture-like images of the ideal category member. \\nThus, the idea of contextual variation in prototypicality within categories is \\nnot raised in Brewer's work, and the suggestion seems to be that categories \\nare represented by fixed visual images. \\nLord and his colleagues discuss prototypes in their leadership categoriza­\\ntion theory (for example, Fraser & Lord, 1988; Lord, Foti & De Vader, \\n1 984; cf. Hogg, 1996). They describe leadership prototypes as cognitive \\nstructures (for example, Fraser & Lord, 1988, p. 292), and argue that \\nperception of an individual as a leader depends upon the extent to which he \\nor she is similar to a leadership prototype. Lord et al. (1984) discuss the idea \\nof basic-level leadership prototypes, and suggest that these could be used as \\n'a standard for appropriate behaviour' (p. 359). Clearly, then, the emphasis \\nhere is on the identification of relatively context-independent prototypical \\nstandards against which behaviour is evaluated. Similarly, in their treatment \\nof leadership Rush and Russell (1 988) refer to prototypes as 'easily accessed \\nschemata which serve as abstract representations of the most representative \\n(stereo \\ntypic) member of a particular category' (p. 89). Again, the emphasis \\nis on prototypes as stable, abstract representations of categories rather than \\nvariable outcomes of the categorization process. \\nIn contrast, cognitive researchers have identified a number of problems \\nwith the view that categories are organized around fixed prototypes. \\nBarsalou (1987, 1989) has reported extensive evidence of the contextual \\nvariability of prototypicality. For example, Barsalou, Sewell and Ballato \\n(1986) found a low level of agreement between subjects' typicality judge­\\nments given twice in the same context but at two different times separated \\nby a few weeks. Barsalou and Sewell (1 984) asked American undergraduate \\nsubjects to judge typicality from both the American and the Chinese point \\nof view. They generated graded structures that were, on average, uncor­\\nrelated across categories. For example, from the American perspective the \\nmost typical birds were robin and eagle, but from the Chinese perspective \\nthey were swan and peacock. Barsalou and Sewell also correlated the \\naverage graded structure provided by Emory University undergraduates \\nwith that provided by Emory University faculty for the same categories and \\nfound a low overall correlation (r = .2). However, when asked to provide \\ntypicality judgements from faculty's point of view, the undergraduates \\nproduced a perfect replication of the faculty's own responses, and graduate \\nstudents were able to reproduce exactly both undergraduate and faculty \\ntypicality judgements. Barsalou (1 987) concludes from this work: \\nThe graded structure within categories does not remain stable. Instead a category's \\ngraded structure can shift substantially with changes in context. This suggests that \\ngraded structures do not reflect invariant properties of categories but instead are \\nhighly dependent on constraints inherent in specific situations. (p. 107) \\nSimilarly, Medin (1989) identifies the treatment of categories as context­\\nindependent as a major problem for theories which assume prototype-based \\n\\n\\nThe role o\\nf\\n pro \\nto \\ntyp icality \\n89 \\nrepresentation. He cites the work of Roth and Shoben (1 983) which again \\nshows the contextual variability of judgements of prototypicality. For \\nexample, Roth and Shoben found that tea is judged to be more typical of \\nthe category 'beverage' than milk in the context of secretaries taking a \\nbreak, but the reverse ordering obtained when the context was truck drivers \\ntaking a break. \\nAn intriguing series of \\nexperiments by Medin, Wattenmaker and Hampson \\n(1987) demonstrates how difficult it is for subjects to work with \\ncategorizations based simply on relative similarity to a prototype, or family \\nresemblance. For the first four experiments, stimuli were constructed to form \\ntwo categories within which items were linked by resemblance to a prototype. \\nThis was done in terms of the degree to which individual items had properties \\nwhich matched or mismatched the identified prototype. For example, the \\ncategories might be two groups of cartoon-like animals, and the varying \\nproperties might be the shape of the head, length of the tail, body markings \\nand number of legs (see Medin et a!., 1987, for details). Overall, these \\nexperiments provided virtually no support for the idea of family resemblance­\\nbased categorization. Asked to sort the full set of stimuli into two equal-sized \\ngroups, most subjects simply sorted them on the basis of one feature and \\nignored the identified prototypes completely. \\nIn light of these negative findings, Medin et a!. re-evaluated their assump­\\ntion that the simple matching and mismatching of isolated properties was \\nsufficient to capture what goes on when people make categorization deci­\\nsions. In the remaining three experiments reported in the paper, they gave \\nsubjects information about interproperty relationships which could serve to \\nexplain why integration across separate components of the stimuli was \\nappropriate. For example, in Experiment 7 cartoon-like drawings were again \\nused, with variation in body size, foot type (paws or webbed feet), body \\ncovering (feathers or hairs), ear size and mouth type (beak or mouth). Again, \\ntwo prototypes were identified, associated with a pattern of correlated \\nattributes which suggested that one category might be related to birds and \\nflying. Subjects, given the task of sorting 1 0  figures into two equal-sized \\ngroups, were told that half the animals were flyers and half were non-flyers. \\nThis instruction was designed to give the subjects a basis on which to relate \\nthe separate attributes to each other. It was found that although the majority \\nof the sortings were still unidimensional, a significant minority (9 out of 24 \\nsubjects) now did sort by family resemblance, that is, into categories struc­\\ntured around the identified prototypes. Further, subjects' sortings were \\nsupported by explanations which drew on the information about flying, even \\nwhen they did not follow the predicted family resemblance pattern. Medin et \\na!. attribute their increased success in these latter experiments to the fact that \\nthe further information provided (for example, about flying) both made \\ninterproperty linkages salient and allowed subjects to associate individual \\nitems through their relationship to a higher order theme. \\nThese findings are consistent with Medin's cogent argument (for example, \\nMedin, 1989; Medin & Wattenmaker, 1987; Murphy & Medin, 1 985) that \\n\\n\\n90 \\nSocial identity: international pers\\npectives \\ncategorization cannot be based on judgements of similarity through match­\\ning of isolated attributes. Indeed, he asserts that similarity 'is more like a \\ndependent variable than an independent variable' (Medin, 1989, p. 1474) in \\nthe categorization process (cf. McGarty & Turner, 1992). He has suggested \\nthat perceivers have wide-ranging 'background theories' about their world \\nand how it works, and that these theories play a crucial role in 'category \\ncohesiveness', that is, in defining the way in which categories 'hang together' \\nas coherent wholes, and therefore in determining why we have the \\ncategories we do rather than others. For example, Murphy and Medin \\n( 1985) point out that on the basis of simple attribute matching we could \\nidentify several similarities between a plum and a lawnmower (both cannot \\nhear, both have a smell, both can be dropped . . .  \\n) which might suggest they \\nshould share a category membership, but this is not consistent with our \\ngeneral theories about the world, so we understand that the 'similarities' \\nidentified are not sufficient to determine categorization. On the other hand, \\nwe can construct meaningful categories which bring together items with \\nvery little apparent similarity. Barsalou (1983) provides the example of a \\ncategory comprising children, money, photo albums and pets, which \\nbecomes sensible when one considers it in the context of a knowledge base \\nwhich includes the fact that the category represents 'things to take out of \\none's house in a fire' (see also Rips, 1989, for further discussion of the idea \\nthat similarity is neither necessary nor sufficient to determine category \\nmembership). \\nThese arguments about the insufficiency of similarity have been used to \\nchallenge the classical view of categorization, but Medin (1989) points out \\nthat the more recent prototype theories (and exemplar-based theories) are in \\nfact just as dependent upon similarity as a determinant of categorization. It \\nis similarity to the prototype (or a specific exemplar) that is assumed to \\ndetermine categorization (for example, Brewer et aI., 198 1 ; Lord et aI., \\n1984). Medin argues for knowledge-based rather than similarity-based \\ncategorization, although he sees a role for similarity in contributing to what \\nhe calls 'psychological essentialism', which functions to constrain the \\n'deeper substance' (Medin, 1989, p. 1479) of the knowledge-rich theories \\nwhich drive the categorization process (see Medin, 1989; Medin & Ortony, \\n1989, for further discussion of psychological essentialism). \\nIn summary, cognitive psychologists have argued that theories which \\nassume that categories are represented by fixed prototypes, and that \\ncategorization is a matter of attribute matching to a prototype, cannot deal \\nwith evidence of contextual variability in category content and structure. \\nThey also fall foul of evidence that attribute matching (simple similarity \\njudgement) is insufficient to account for category cohesiveness and cat­\\negorization decisions. Medin (1989) comments: \\nPrototype theories . . . fail to reflect the context sensitivity that is evident in \\nhuman categorization. Rather than getting at the character of human conceptual \\nrepresentation prototypes appear to be more of a caricature of it. (p. 1472) \\n\\n\\nThe role o\\nf\\n pro \\nto \\ntyp icality \\n91 \\nIn this context, it is important to emphasize that self-categorization \\ntheory is not a 'prototype theory' in the sense referred to in Medin's work. \\nAs noted above, the theory sees proto \\ntypicality as a wholly context­\\ndependent outcome of the categorization process, not the basis for category \\nrepresentation, and agrees with Medin that this categorization is a product \\nof perceivers' background knowledge in interaction with current stimuli (see \\nOakes et aI., 1994; Turner, Oakes, Haslam & McGarty, 1994). The simi­\\nlarity and difference judgements involved in the meta-contrast principle are \\nmade in the context of background theories about the ways in which \\ncategory members tend, in general, to be similar and different - this is \\nembodied in the 'normative fit' principle discussed briefly above (Oakes et \\naI., 1991). One of the major contributions of the theory has been its analysis \\nof perceived similarity and difference as outcomes (rather than simple \\ndeterminants) of categorization, and we see perceived proto \\ntypicality as one \\naspect of the way in which categorization contextualizes simple attribute \\njudgements, gives them meaning in relation to other stimuli being judged \\n(Turner & Oakes, 1989; Turner et aI., 1995). \\nConclusion \\nThis chapter has considered the concept of prototypicality from the per­\\nspective of self-categorization theory. We have reviewed ideas and evidence \\nfrom two main areas of self-categorization research in which prototypicality \\nhas played an important role. In polarization research, it has been shown \\nthat polarization can be explained as conformity to the prototypical position \\nwithin the group, a position extremitized through the influence of com­\\nparative context on ingroup definition. Research on intragroup attraction \\nhas established a distinction between social and personal attraction, the \\nformer reflecting attraction within groups based on depersonalization and \\nprototypicality. In both these research areas, prototypicality is seen as a \\ncontext-dependent outcome of the categorization process, as determined by \\nboth comparative (meta-contrast) and normative (fit with the content of \\nbackground theories) factors. Because the principle of meta-contrast allows \\naccurate specification of the most prototypical attributes within a given \\ncomparative context, precise predictions about conformity and attraction \\nwithin groups have been made possible. \\nRecognition of the internally graded, variable structure of categories was \\none of the revolutionary contributions of Rosch's work. However, many \\nsubsequent developments of that work have transformed the dynamic idea \\nof proto \\ntypicality judgements into a concept of idealized 'caricatures' \\nfunctioning as relatively fixed representations of categories, and determining \\nthe categorization process through judgements of item-prototype similarity. \\nSelf-categorization theory maintains a more Roschian perspective in this \\nregard, with its emphasis on prototypicality rather than prototypes, and its \\nrejection of the view of categorization as a matter of \\njudging similarity to a \\n\\n\\n92 \\nSocial identity: international perspectives \\nprototype. In predicting categorization through the interaction of com­\\nparative and normative factors, the theory shares with the work of Medin \\nand others an emphasis on the intellectual (Neisser, 1 987) basis of \\ncategorization. \\nWhilst the theory's conceptualization of prototypicality has proved very \\nfruitful in the research areas reviewed here, it is nonetheless true that more \\nwork remains to be done in order to demonstrate the inherently variable \\nnature of category representation (cf. Barsalou, 1 987, 1 989). This represents \\none significant direction in which our current work is advancing (Haslam, \\nTurner, Oakes, McGarty & Hayes, 1 992; Oakes et aI., 1994; Oakes, Haslam \\n& Reynolds, in press). \\n\\n\\n7 \\nGroup Beliefs as an Expression of Social \\nIdentity \\nDaniel Bar-Tal \\nSocial identity \\nOne of the seminal theories of social psychology which contributes to the \\nunderstanding of intragroup and intergroup processes is social identity \\ntheory (Tajfel, 1978d, 1981 ,  1982b). The theory posits that individuals not \\nonly adopt a personal identity as unique persons, but also form a social \\nidentity which reflects their membership in various groups to which they \\nbelong. Specifically, social identity is defined as 'that part of an individual's \\nself-concept which derives from his knowledge of his membership of a social \\ngroup (or groups) together with the value and emotional significance \\nattached to that membership' (Tajfel, 1 978b, p. 63). Social identities, in the \\nform of such social categories as nationality, religion, gender, profession, \\nethnicity or political orientation, are internalized and constitute a poten­\\ntially important part of the individual's self-concept. They provide meaning­\\nful and significant self-references through which individuals perceive \\nthemselves and the world around. \\nOf special importance is the assumption stating that individuals have a \\nneed to achieve positive social identity. Specifically, it is assumed that indi­\\nviduals strive for positive self-image and that in the course of identifying with \\na group this need translates into a tendency to view one's own group \\nfavourably. This view is achieved not necessarily through positive \\nperception, but rather through comparison on appropriate dimensions. \\nThe comparison allows one to perceive the ingroup as 'better' than relevant \\noutgroups. \\nThe process of social identification is further elaborated in self­\\ncategorization theory (Turner, 1 987), an extension of social identity \\ntheory. The theory suggests that individuals categorize themselves as \\nmembers of social categories, and then define, describe and evaluate \\nthemselves in terms of these categories. In the process of self-categorization, \\na set of cognitive representations of self is formed. Part of the cognitive \\nrepresentations reflect the perceived membership of individuals in groups. \\nSince individuals belong to various groups, they possess multiple rep­\\nresentations of self. The categorization of the self results in a perceptual \\n\\n\\n94 \\nSocial identity: international perspectives \\naccentuation of intra-category similarities and inter-category differences on \\ndimensions believed to be associated with the categorization. \\nTheories of social identity and social categorization focus on the defini­\\ntion of 'we-ness' by group members in terms of 'We belong to a group'. \\nGroup members in this process perceive themselves as group members, \\nidentify themselves as such, and differentiate between their own group and \\nother groups. This primary process of \\ncognitive-emotional nature shapes the \\nsocial identity of individuals. Nevertheless, on the basis of social identity \\nand social categorization theories, a question should be asked: What are the \\nother cognitive representations, in addition to social categorization of the \\nself, which underlie a person's social identity? The answer to this question is \\nprovided in an analysis of group beliefs (Bar-Tal, 1 990). The purpose of the \\npresent chapter is to present group beliefs as a framework for analysing a \\ngroup member's social identity. Specifically, first the chapter will delineate \\nthe conception of group beliefs; second, it will outline their characteristics; \\nthird, it will describe group beliefs' formation; and, finally, implications will \\nbe presented. \\nThe conception of group beliefs \\n'Group beliefs are defined as convictions that group members (a) are aware \\nthat they share, and (b) consider as defining their \\\"groupness'\\\" (Bar-Tal, \\n1 990, p. 36). The first part of the definition suggests that two beliefs have to \\nbe shared by group members. One belief pertains to any content that is the \\nsubject of the group belief and the other pertains to specific knowledge \\nsaying that the former is shared by group members. The second part of the \\ndefinition suggests that group beliefs are those beliefs that group members \\nperceive as defining the essence of their group. It is assumed that they \\nprovide the cognitive basis that group members view as uniting them as one \\nentity. In fact, group beliefs serve as a foundation for group formation and, \\nlater, as a bond for group existence. 'We are exploited', 'Communism is the \\nbest system for human beings', 'Jesus is God's son', 'Iraq is our enemy', are \\nexamples of possible group beliefs. \\nIn this framework, the belief 'We are a group', expressing social identity \\nof group members, is called the f\\nundamental group belief However, in \\naddition to the fundamental group beliefs, group members share group \\nbeliefs of various contents pertaining to a variety of other subjects. These \\nbeliefs are part of cognitive representations which together with self-social \\ncategorization (that is, the fundamental group belief) define social identity \\nof group members and thus underlie their we-ness and uniqueness. \\nIndividuals regard group beliefs as characterizing them as group members \\nand as defining the boundary of the group. On their basis individuals may \\ncategorize themselves and may be categorized by others as group members. \\nAcceptance of group beliefs is one of the important indicators of group \\nmembership. Sharing group beliefs implies, most frequently, that a person is \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n95 \\na group member. Moreover, being a group member signals to people that a \\nperson shares group beliefs. A person joining a group must, at least exter­\\nnally, express acceptance of group beliefs. Thus, on the one hand, group \\nbeliefs unify group members and, on the other hand, differentiate them \\nfrom outgroups. The unification is reflected in the recognition of similarity \\n(Bar-Tal, 1 993). Group members are aware that they share the same unique \\nbeliefs that define their 'groupness' and thus become aware of the similarity. \\nGroup members who may differ with regard to many characteristics are \\naware that they share the same unique beliefs that define their 'groupness' \\nand thus become aware of the similarity. \\nAlso, since in many cases groups try to be differentiated, group beliefs \\noften provide a criterion for differentiation. Group beliefs draw the line \\nbetween the in \\ngroup and outgroups. As Sherif (1951) pointed out: 'From \\nthe point view of inter-group relations, the most important consequence of \\ngroup structuring is the delineation of in-group from out-groups. The devel­\\nopment of in-group and \\\"we-experience\\\" is accompanied by the demarca­\\ntion and setting of boundaries from out-groups' (p. 395). \\nAn experiment by Allen and Wilder (1975) demonstrated that group \\nformation (including the formation of a group belief) is a more potent \\nfactor in discrimination against outgroup than the awareness that the \\noutgroup members have some similar beliefs. Also, belief similarity about \\nthe basis for group classification (that is, formation of a group belief) had a \\nsignificant effect on discriminative behaviour. Special ingroup favouritism \\nwas displayed when the in \\ngroup had similar beliefs. \\nContents of group beliefs \\nGroup beliefs, as implied by the definition, are held by group members, who \\nshare these beliefs and consider them as defining their group. Thus, group \\nbeliefs should be considered as a category of beliefs and should be \\ncharacterized by the same features as other beliefs. First of all, the contents \\nof group beliefs will be discussed. \\nBeliefs in the present conceptual framework, viewed as propositions to \\nwhich individuals attribute at least some truth, constitute units of knowl­\\nedge. This is a wide-scope definition which considers any ideas, thoughts or \\nopinions about any topic as beliefs. They can thus cover various contents \\nwhich can be classified to such categories as 'values', 'ideologies', 'inten­\\ntions', 'goals', and so on. Group beliefs may also refer to different topics or \\nideas that a human mind can think. They may concern such subjects as \\nreligious doctrines, political ideologies, philanthropic goals, human rights, \\nidentity characteristics, prescriptions of behaviour (that is, norms), and \\nothers. As examples, four categories of group beliefs will be presented -\\nthose which refer to group norms, values, goals and ideology. \\n\\n\\n96 \\nSocial identity: international perspectives \\nGroup norms \\nGroup norms, defined as 'an idea in the minds of the members of a group, \\nan idea that can be put in the form of a statement specifying what the \\nmembers or other men should do, ought to do, are expected to do, under \\ngiven circumstances' (Homans, 1950, p. 123), regulate group members' \\nbehaviour and provide criteria for judging it. Norms, as defined, do not \\nnecessarily have to be group beliefs. Norms may regulate group members' \\nbehaviour without being considered as characteristic of the specific group. \\nIn such groups, group beliefs may pertain to contents other than norms. In \\nother groups, at least part of the norms may define their characteristics and \\nbe group beliefs. Nevertheless, when norms function as group beliefs, group \\nmembers believe that their patterns of behaviour are unique to them and \\ncharacterize their membership. \\nIn an extensive description of the Amish society, \\nI Hostetler (1 968) points \\nout several norms that can be considered as group beliefs. That is, Amish \\npeople believe that particular prescriptions of behaviours define their \\n'groupness' and contribute to their social identity. For example, Amish \\nbeliefs prescribe the necessity of separation from the world. Therefore, an \\nAmish man does not marry a non-Amish woman, does not do business with \\nan outsider, and even does not enter into intimate relationships with a \\nperson outside the Amish community. In addition to norms of separation, \\nthe Amish group also perpetuates norms of non-resistance. Group members \\nare forbidden to take part in violence or war, and they withdraw from \\nconflicts or resistance. They refuse to swear any oath, to bear arms, or to \\nhold public offices. \\nIn another example, Ardener (1983) described the group beliefs of \\nCanadian Doukhobors who emigrated from Russia at the beginning of the \\ntwentieth century. They believe that the Perfect Life is attainable on earth \\nthrough the rejection of materialism. They have not recognized human \\nauthority structure and therefore are opposed to state legislation or oath \\ntaking. Their uniqueness has been reflected in deprivation, suffering and \\npoverty, which became, according to Ardener, part of their self-definition, \\ncorresponding to their identification with Christ himself. In this vein, to \\nform their own identity, Doukhobors resorted to nudity as a social artefact \\nfor identifying their group and to the use of arson against institutions and \\nthen against their own property. Ardener suggested that both practices \\nhelped to maintain the boundary of the group and to enhance solidarity in \\nthe face of outgroups. \\nGroup values \\nAn analysis similar to the one presented about norms applies to group \\nvalues. Groups may hold values, but they do not necessarily have to \\nfunction as group beliefs. A value is defined as 'an enduring belief that a \\nspecific mode of conduct or end-state of existence' (Rokeach, 1 973, p. 5). \\nThe former is called an instrumental value, while the latter is called a \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n97 \\ntenninal value. Values guide the selection of the means and ends of specific \\nactions, and serve as criteria by which objects, actions or events are \\nevaluated. Examples of values are freedom, truth, individualism or equality. \\nThey reflect the ideals to which individuals and groups aspire. \\nIt is possible to characterize groups by the values that their group \\nmembers share (Kluckhohn, 1 95 1 ;  Parsons, 1 968). Sociologists have focused \\non shared values, referring to them as value systems in a given society. \\nNevertheless, a given shared value is not necessarily held by every person in \\nthe group. When a sufficient number of group members subscribe to the \\nvalue, it may not only characterize the group, but may also become an \\nimportant detenninant of the group members' behaviour. \\nIn some groups the values, like nonns, may be fonnally fonnulated (that \\nis, in writing) and maintained as group beliefs, while in other groups they \\nare never fonnally defined, but may be considered as group beliefs and \\ncarried latently through the social processes of socialization and influence. \\nHowever, in both types of groups, when values are shared by group mem­\\nbers and viewed by them as characterizing the group, they are group beliefs. \\nIn this case, the value or values define the uniqueness of the group and \\ndifferentiate it from outgroups. \\nTo specify values that characterize a nation is a complex task. Never­\\ntheless, social scientists commonly have suggested, on the basis of either \\ncollected data or their own observations, that self-sufficient individualism \\nis one of the salient characteristics of American society (for example, \\nDiamond, 1976; Gillin, 1 955; Williams, 1970). This value emphasizes the \\nresponsibility and the right of each individual to pursue personal goals with \\nas much freedom from restraints as possible. Americans are socialized to \\ninternalize this value and to believe that it characterizes their society. \\nMany groups present values as part of their credo. In fact, many of the \\ngroup nonns and goals functioning as group beliefs are based on values. \\nThis stress on values can be observed in religious groups, as well as in many \\npolitical groups. An example of a political group that was founded on the \\nbasis of a value is Amnesty International. In its statute, the value of human \\nrights underlies more specific goals. The beginning reads: \\nI .  Considering that every person has the right freely to hold and to express his \\nor her convictions and the obligation to extend a like freedom to others, the \\nobjects of Amnesty International shall be to secure throughout the world the \\nobservance of the provisions of the Universal Declaration of Human Rights, \\nby: \\na) irrespective of political considerations working towards the release of and \\nproviding assistance to persons who in violation of the aforesaid provisions \\nare imprisoned, detained or otherwise physically restricted by reason of their \\npolitical, religious or other conscientiously held beliefs or by reason of their \\nethnic origin, sex, color, or language, provided that they had not used or \\nadvocated violence (hereinafter referred to as 'Prisoners of Conscience'); \\nb) opposing by all appropriate means the detention of any Prisoners of \\nConscience or any political prisoners without trial within a reasonable time or \\n\\n\\n98 \\nSocial identity: international perspectives \\nany trial procedures relating to such prisoners that do not conform to \\ninternationally recognized norms; \\nc) Opposing by all appropriate means the imposition and infliction of death \\npenalties and torture or other cruel, inhuman or degrading treatment or \\npunishment of prisoners or other detained or restricted persons whether or \\nnot they have used or advocated violence. (Amnesty International pamphlet, \\nLondon, 1978) \\nAlthough the Ku Klux Klan is vastly different from Amnesty International, \\nits group beliefs are also based on values. The principal beliefs of the Ku \\nKlux Klan are presented in the following way: \\nWe invite all men who can qualify to become citizens of the Invisible Empire, to \\napproach the portal of our beneficent domain, join us in our noble work of \\nextending its boundaries, and in disseminating the gospel of Klankraft, thereby \\nencouraging, conserving, protecting and making vital the fraternal relationships in \\nthe practice of an honorable clannishness; to share with us the sacred duty of \\nprotecting womanhood; to maintain forever the God-given supremacy of the \\nWhite Race; to commemorate the hold and chivalric achievement of our fathers, \\nto safeguard the sacred rights, privileges and institutions of our civil government; \\nto bless mankind and to keep eternally ablaze the sacred fire of a fervent devotion \\nto a pure Americanism. (Cited in Vander Zanden, 1960, p. 459) \\nThus, the values of womanhood, white supremacy, mankind and American­\\nism are the values that the group intends to protect. \\nGroup goals \\nGroup goals are defined as beliefs of valued or desired future specific states \\nfor the group (Cartwright & Zander, 1968; Etzioni, 1 975). On the basis of \\nthis definition, we can assume that with few exceptions almost all groups \\nhave goals. In most of these cases, goals may serve as group beliefs in \\naddition to other beliefs. Individuals form groups to achieve certain goals. \\nSubsequently, goals are often considered as a raison d'etre for group \\nformation, frequently keep group members together, provide a basis for \\nsolidarity, and give direction for activity. These functions are particularly \\nexplicit in certain voluntary groups that are formed to advance specific \\ngoals (Sills, 1958). Thus it is not surprising that group goals frequently serve \\nas basic group beliefs. They increase the identification of the group \\nmembers with their group and define the boundary for group membership. \\nIn this vein, March and Simon (1958) noted that: \\nThe greater the extent to which goals are perceived as shared among members of \\na group, the stronger the propensity of the individual to identify with the group \\nand vice versa. (p. 66) \\nTwo examples of group goals may provide illustrations of how they \\nfunction as group beliefs. Trans-Species Unlimited (TSU) is a national \\nanimal rights group 'dedicated to the total elimination of animal abuse and \\nexploitation'. This group belief leads TSU to direct actions and legislative \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n99 \\ncampaigns to end such practices as the use of leghold trap and live bird \\nshoots and commercial exploitation of rabbits for meat, fur and research \\nThe Richard III Society was founded in England in 1924 with the \\nfollowing aims: \\n1 .  To promote in every possible way historical research into the life and times of \\nKing Richard III. \\n2. To secure a reassessment of the historical material relating to this period and \\nof the role in English history of this monarch. \\n3. To circulate all relevant historical information to members of \\nthe Society and \\nto educational authorities. (copied form the pamphlet of The Richard III \\nSociety, New Orleans, undated) \\nThe principal goal of this group is to present the truthful history of King \\nRichard III in view of past distortions and biases. \\nGroups differ with regard to the goals that they try to achieve. There are a \\nvariety of goals that can be classified into different categories. One type of \\ncategory has received special attention by political scientists - namely, goals \\nto influence governmental decisions. Groups that have these goals are called \\ninterest groups (Salisbury, 1975), and then the term interest denotes a 'shared \\nattitude toward what is needed or wanted in a given situation, observable as \\ndemands, or claims upon other groups in the society' (Truman, 1 95 1 ,  p. 33). \\nThese shared demands or claims, which reflect goals, serve as a basis for \\ngroup formation and function as group beliefs. The interests can pertain to \\nvarious contents involving either tangible or intangible claims, and concerns \\nof \\neither a particular policy objective or wide societal goals (Salisbury, 1 975). \\nInterest groups can be based on either a few or many group beliefs. In \\nsome groups, the group beliefs refer to one broad topic that is relevant to \\nseveral issues. For example, the Consumer Federation of America reflects a \\nviewpoint that favours the protection of the rights of individual consumers \\nthrough the regulation of corporate activities, and the US Chamber of \\nCommerce reflects a strong belief in the character of the American free \\nenterprise system, opposing excessive government regulation of business. \\nOther groups' interests are formed on the basis of narrow group beliefs \\nreflecting single issues. For example, the National Rifle Association has a \\nstrong belief in the American citizen's unrestricted right to bear arms; the \\nNational Right to Life Committee believes in the sanctity of the foetus. \\nGroup ideology \\nOne of the most discussed concepts for describing a group's uniqueness is \\nideology. Ideology is often referred to as the mental characteristic of \\na group. Indeed, various definitions of ideology refer to its commonality - a \\nstate indicating that ideological beliefs are shared by group members. It \\nreflects their common experience and serves as a basis for group members' \\ncooperation, morale, order and as a rationale for their behaviour. Ideology \\nrefers to an integrated set of beliefs constituting a programme, a theory of \\ncauses and effects, and premises on the nature of humanity and societal \\n\\n\\n100 \\nSocial identity: international perspectives \\norder (Apter, 1964; Lane, 1 962; Shils, 1968). For example, Toch (1965) \\ndefined ideology as: \\na set of related beliefs held by a group of persons. The ideology of a social \\nmovement is a statement of what the members of the movement are trying to \\nachieve together, and what they wish to affirm jointly . . .  \\nthe ideology of a social \\nmovement defines the movement, and contrasts it with other movements and \\ninstitutions. (p. 21) \\nWhen group members hold an ideology, it frequently serves as a group \\nbelief. By definition, it consists of a set of ideas that characterize the way in \\nwhich a group posits, explains and justifies the ends and means of its \\norganized social actions. It usually provides an identity to group members, \\ndefines the group cohesion, and describes its exclusivity; in most cases, an \\nidentification with a specific ideology indicates a membership in a particular \\ngroup (Lane, 1962). \\nThe contents of an ideology can be political, social or even religious. An \\nideology usually refers to images of the desired society and the means and \\nconditions needed to achieve it. For example, the ideology of the Fasci \\nItaliani di Combattimento (the Italian Fascist Party) placed a special \\nemphasis on a nation or a state - the fascist state. The individuals and \\ngroups are secondary to the state, which embodies the culture and spirit of \\nthe people and is considered to be the highest and noblest value. The state \\nprovides an orientation for the individuals - it is only there that they can \\nfulfil their identity. According to the fascist ideology, life is a continuing \\nstruggle for national supremacy. In order to achieve the nation's goals, the \\nstate has to exercise complete authority over all areas of public and private \\nlife, from the education of children to control of the economy. The state has \\nthe right to control and restructure the society. Accordingly, the state \\nrequires total loyalty. There is no place for groups or individuals to object, \\nsince the state is the source of political, economic and moral action \\n(Mussolini, 1 935). These principles served as group beliefs of the Italian \\nFascist Party, and, with its ascendance to power in 1 922, the party tried to \\nturn them into group beliefs for the entire nation. \\nFor seven decades, the principles of communist ideology served as group \\nbeliefs for the members of the Communist Party of the Soviet Union. This \\nideology was the official creed and the most accessible characteristic in \\nevery domain of life in Russia. The basic principles of communism were \\ninculcated in Soviet citizens from the early years of education and main­\\ntained through all the state communication channels. The ideology \\nadvocates a classless social system of equality for all members of society, \\nwhich provides everyone with material and cultural benefits according to \\ntheir growing needs (Triska, 1 962). \\nIn sum, this limited review of the contents did not exhaust all possible \\ncategories of group beliefs. As indicated, group beliefs can have any \\ncontents, and the above categories serve only as relevant examples of the \\nones that are frequently used by social scientists. Group members share \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n101 \\ngroup beliefs that characterize them and differentiate them from other groups. \\nAs illustrations of possible contents, this review was not based on group \\nmembers' reports identifying their group beliefs, though it is possible that \\nthey actually serve as group beliefs. Future research should treat the formal \\nstatements as hypotheses that need to be validated through empirical \\nevidence. \\nCharacteristics of group beliefs \\nGroup beliefs can be characterized by the same features as other beliefs. \\nDegree of confidence and level of centrality are among the most important \\ncharacteristics of beliefs (Bar-Tal, 1990; Bem, 1 970; Kruglanski, 1 989). \\nThese characteristics are of special relevance for group beliefs and therefore \\ndiscussed at some length. \\nCon\\nfidence \\nGroup beliefs are usually held with great confidence because they are con­\\nsidered to be facts and verities. This occurs because group beliefs define the \\nessence of the group and thus the reality of the individuals who view \\nthemselves as group members. In fact, in most cases, individuals voluntarily \\njoin groups, such as religious denominations, political organizations or \\nprofessional associations, on the basis of group beliefs. The act of \\njoining a \\ngroup indicates that group members accept, at least formally, group beliefs \\nand hold them as valid. \\nA reduced confidence in group beliefs may shatter group members' \\nreality, especially in cases when group membership is very important for \\nindividuals or when the change is sudden. The questioning of group beliefs, \\nespecially central ones, may have significant implications for group life. \\nMoving group beliefs from the status of verities to the status of hypotheses \\nby the majority of group members may change the essence of the group or \\nits raison d'etre, or even cause its division or disintegration. It is, therefore, \\nnot surprising that groups often use various mechanisms to maintain high \\nconfidence in group beliefs. \\nIt is recognized, though, that groups differ with regard to their insistence \\non confidence. While orthodox groups insist on high confidence in group \\nbeliefs (Deconchy, 1 984), other groups may be less strict on this matter. The \\nformer groups do not allow any doubts or sanction their expression. The \\nother groups are more tolerant of changes and therefore their group beliefs \\nmay be gradually altered. \\nCentrality \\nThe centrality of group beliefs implies that they are often accessible in \\ngroup members' repertoire and that they are relevant for consideration \\n\\n\\n1 02 \\nSocial identity: international pers\\npectives \\nin making various evaluations, judgements or decisions, including behav­\\niours (Bar-Tal, 1 986). The centrality characteristics of group beliefs has at \\nleast two different aspects. First, the characteristic may refer to the \\nimportance of the set of beliefs for the group members, and, second, it may \\nrefer to the importance of the specific belief within the set of group beliefs. \\nThe former aspect is based on an assumption that some groups are more \\nimportant for their members than others. Individuals are usually members \\nin a number of groups and not all of them are of equal importance to \\nthem. The importance of the group for group members is reflected in the \\ncentrality of group beliefs. That is, when the group is important for group \\nmembers, in most cases the group beliefs are central. For example, it can be \\nassumed that group beliefs are very central in groups such as the Amish \\n(Hostetler, 1 968) and Lubovitcher Hassidim (Levy, 1 975). Group \\nmembership in these groups is the most important characteristic for the \\nmembers and the group beliefs are frequently accessible and often taken \\ninto consideration. In other groups, group beliefs may be unimportant for \\nmost of the group members. \\nThe other aspect of characterizing group beliefs as central refers to the \\nassumption that not all beliefs are of equal centrality for group members. \\nSome group beliefs may be of greater importance than others and thus may \\nbe more central. It means that when a group has a set of elaborated group \\nbeliefs, some of them contribute more to the group characterization and its \\ndefinition than others. The more central group belie\\nf\\ns\\n are considered as \\nprototypic in group characterization and there\\nf\\nore are called basic group \\nbelie\\nf\\ns. Their weight in defining the essence of the group is crucial. Beyond \\nthe simple self-definition 'We are a group', these beliefs provide the group's \\ncredo. For instance, in certain groups, group beliefs that refer to group \\ngoals or identity may be more central than group beliefs that refer to values \\nor group history. In a specific example, it may be assumed that while the \\nMormons have numerous group beliefs, one of the basic ones is the \\nconviction that the Book of Mormon is another Testament of Jesus Christ, \\na religious and secular record of ancient American civilization (O'Dea, \\n1 957). \\nThe centrality characteristic is not a given, but often is maintained by \\nexternal factors. Availability and saliency of group beliefs in the group \\nmembers' environment may influence their centrality. The first element \\nrefers to the frequency with which the content of group beliefs is exposed \\nto group members. The other element expresses the prominence and \\nvividness of the group belief when it is presented to group members. \\nWhen group beliefs are repeatedly presented to group members in a \\nprominent and vivid manner, they are often accessible in group members' \\nrepertoires (Higgins & King, 1981). Groups often keep group beliefs \\naccessible in the minds of the group members by constantly repeating the \\ncontents of group beliefs. Moreover, they even sometimes develop cultural \\nand educational mechanisms to maintain group beliefs as central in group \\nmembers' minds. \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n1 03 \\nFormation of group beliefs \\nFormation o\\nf\\n the f\\nundamental group belief \\nThe crucial phase in the f\\normation of group belie\\nf\\ns\\n is the emergence o\\nf\\n the \\nf\\nundamental group belief indicating group existence (that is, ' We are a \\ngroup'). This phase is preceded by various social perceptual and cognitive \\nprocesses. According to the present view, numerous reasons can cause \\nindividuals to start believing that they constitute a group. Individuals usually \\nhave to f\\normulate at least one belief which provides the antecedent f\\nor the \\nf\\nundamental group belief stating that a group exists (that is, ' We are a \\ngroup'). Such an antecedent belief may later become a group belie\\nf \\nAny belief can serve as an antecedent f\\nor individuals to begin considering \\nthemselves members o\\nf\\n a group. The belief 'We are a group' may be instigated \\nby any of the following beliefs: 'We act interdependently', 'We have the same \\nfate', 'We have the same characteristic', 'We live in the same place', 'We have \\nthe same goal', 'We believe in the same religious doctrine', 'We have the \\nsame ancestors', 'We accept the same ideology', 'We are treated in the same \\nway', and others. The antecedent belie\\nf(s) may be f\\normed on the basis o\\nf\\n \\nactual experience and perception or on the basis o\\nf\\n reliance on inf\\normation \\nwhich comes f\\nrom sources that propagate these belie\\nf\\ns. In the first case, \\nindividuals may actually become aware of the similarity in what may be \\ncalled natural evolvement, whereas in the other case, an epistemic source \\nmay indicate to the individuals that they are similar and point out that \\nsimilarity. \\nFor example, Epstein's (1978) work indicates that ethnic identity is based \\non shared beliefs in common attitudes or property, which differentiate the \\ngroup from the outgroups. The shared beliefs are products of common \\nperceptions of a group's past and interaction between forces operating within \\na group and those coming from outgroups. Tajfel (1981) hypothesized that \\nminority groups tend to emerge as a result of common attitudes or treatment \\nby the outside groups which, on the one hand, facilitates the perception of a \\ncommon fate and, on the other hand, indicates the boundaries operating \\nbetween the ingroup and the outgroups. Several other conceptions stress \\nindividuals' shared needs (for example, Killian, 1 964; Toch, 1 965), such as \\ncommon feelings of frustration, alienation, deprivation, exploitation and \\ninjustice, any of which may serve as a basis for group formation. \\nAn experiment by Zander, Stotland and Wolfe (1960) may be seen as a \\ndemonstration of how the basic group belief is formed. In this experiment, \\nthe investigators created a group composed of female college students. To \\nfoster a group entity, the experimenter provided several antecedent beliefs \\nregarding similarity, proximity, common goals and identity. Specifically, the \\nexperimenter systematically manipulated the perception of 'groupness' by \\nvarying seating arrangements (creating proximity), by pointing to similari­\\nties among group members (creating commonality), and by assigning a goal \\nfor the group (creating competition against another group). Group members \\n\\n\\n1 04 \\nSocial identity: international pers\\npectives \\nwere then asked to choose a name for their group. These antecedent beliefs, \\ntogether with the name of \\nthe group, produced the fundamental group belief \\n'We are a group'. The study also showed that these individuals behaved as \\ngroup members. Thus, the investigators concluded that when a group is \\nformed, large proportions of the self 'become involved in the group and are \\naffected by identification with the group' (p. 475). \\nThe effects of group experiences on the formulation of the fundamental \\ngroup belief are underlaid by the process of perception. It has been sug­\\ngested that principles of perceptual organization contribute greatly to the \\nself-perception of individuals as a group (Asch, 1 952; Campbell, 1958). \\nSpecifically, Campbell (1958) proposed the use of four Gestalt principles \\n(proximity, similarity, common fate and pregnance) to explain why indi­\\nviduals begin to see themselves as one entity. These principles, which lead \\ndiscrete elements to be perceived as parts of a whole, can be viewed as \\ncharacteristics of individuals, and therefore allowing definition of the group \\nand differentiation between the ingroup and the outgroups. Indeed, \\nBrewer's (1979) review of group formation under limited conditions indi­\\ncates that factors such as interdependence, intragroup similarity or shared \\nfate affect the feeling of 'groupness'. Moreover, these factors not only \\ndetermined group formation, but also influenced group members' behaviour \\nin the direction of ingroup bias. \\nA somewhat similar approach to group formation was recently suggested \\nby cognitive social psychologists. They placed their focus on the knowledge \\nthat individuals possess (for example, Pryor & Ostrom, 1 987). Thus, infor­\\nmation about the collective of individuals (that is, their attributes) serves as \\na basis for their self-classification as group members. This information \\ndenotes the similarity of group members and their uniqueness in com­\\nparison to outgroup members. \\nMoreland (1987) extended the perceptual-cognitive perspective to four \\ndifferent bases for self-classification of individuals as group members, and, \\ntherefore, for the formation of the fundamental group belief 'We are a \\ngroup'. One basis derives from the conditions and resources in the environ­\\nment. Factors such as money, time, propinquity or social networks may \\naffect group formation. For example, people who live or work in close \\nproximity may form a group which can be based on common goals or \\ncommon interests. Also, a social network of relationships may facilitate \\ngroup formation, as happens among friends or members of professional \\norganizations. \\nBehavioural basis is evident when people become dependent on one \\nanother for the satisfaction of their needs. Various theoretical approaches, \\nsuch as evolutionary perspective, social exchange theory, social evaluation \\nperspective or psychodynamic theories, suggest that individuals form groups \\nin order to satisfy their various needs. Needs of survival, defence, social \\nsupport, predictability, world understanding, uncertainty reduction, anxiety \\navoidance or personal adjustment are only a few examples of needs which \\ncan be satisfied through group membership. \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n1 05 \\nGroups may also be formed on the basis of affection as reflected in \\nshared feelings. The positive feelings can be expressed in interpersonal \\nattraction of group members, attraction towards group goals or activities, \\nand love and respect of the group's leader. \\nFinally, the cognitive basis indicates that groups are formed when people \\nrealize that they share important personal characteristics such as attitudes, \\nvalues, interest or goals. Summarizing this basis, Moreland (1 987) pointed \\nout that: \\nResearch on personal factors suggests that small groups will form when people (a) \\nhave acknowledged their shared characteristics before, or (b) have done so \\nrecently, or (c) have found it useful to think of themselves in that way. Research \\non situational factors suggests that small groups will form when (d) people are \\nreminded of shared characteristics, or (e) their outcomes seem to depend on those \\ncharacteristics, or (f) the characteristics that they share are unusual in some way. \\n(p. 103) \\nOn the basis of the present analysis, it becomes obvious that social \\ncategorization is the psychological process which underlies group f\\normation in \\nthe minds o\\nf\\n the individuals who are group members. As noted earlier, \\nTajfel's theory of social identity and Turner's theory of self-categorization \\nelaborate on this process (Tajfel, 1978d, 1981 ;  Turner, 1987). \\nIn this vein, several studies demonstrate that individuals form a psycho­\\nlogical group solely on the basis of shared social category, in the process of \\ncomparing themselves to others. In these experiments, performed within the \\nframework of Tajfel's social identity theory and Turner's self-categorization \\ntheory, individuals were categorized as group members and they acted \\naccordingly (for example, Billig & Tajfel, 1 973; Hogg & Turner, 1985b; \\nTajfel, Flament, Billig & Bundy, 197 1 ;  Turner, Sachdev & Hogg, 1983). \\nIn a study by Tajfel et al. (1971), individuals were divided into groups on \\nthe basis of a fairly trivial task (expressing a preference for abstract \\npaintings of Klee and Kandinsky or guessing the number of dots in rapidly \\nprojected clusters). Later, they were put in individual cubicles and asked to \\ndivide points worth money between two persons who were identified as \\neither belonging to their group or not. This minimal group situation was \\nenough to produce ingroup favouritism. Individuals tended to award more \\npoints to members of their group. These results were replicated, even in the \\nexperiment by Billig and Tajfel (1973), when the division into groups was \\ndone on the basis of the toss of a coin. \\nThe important contribution of these studies to the understanding of group \\nformation is the demonstration showing that the perception of the group as \\nan entity does not have to be based on actual experiences, but a mere piece \\nof information may be a sufficient condition for self-labelling as a group. \\nIndividuals may be told that they are given the same label (or category), \\nand this information may cause them to form the belief 'We are a group' \\n(for example, Billig & Tajfel, 1973). \\nIn this respect it was analysed how groups form the fundamental belief \\nwhich underlies the formation of a group. Common experience and \\n\\n\\n106 \\nSocial identity: international pers\\npectives \\nperceived information may influence individuals to form a new shared \\nreality. The appearance of the fundamental belief 'We are a group', which \\nwas the centre of the work of Tajfel and Turner, reflects the emergence of \\nnew social identity and new self-categorization. This is a psychological \\nbeginning of group existence. \\nBut the present conception suggests that social identity is not based solely \\non the fundamental group belief which reflects self categorizations of \\nindividuals as group members. As was presented, additional beliefs called \\ngroup beliefs play a determinative role in the formation and maintenance of \\nsocial identity. These group beliefs establish the raison d'etre for group \\nexistence and provide the rationale for individuals to be group members. \\nFormation o\\nf\\n additional group belie\\nf\\ns\\n \\nOnce the fundamental group belief 'We are a group' is formed, other group \\nbeliefs are usually added, though the contents of some additional group \\nbeliefs may be formed as antecedent beliefs prior to the formation of the \\nfundamental group belief. As already indicated, the contents of group \\nbeliefs can be of various categories and of unlimited scope. Groups differ \\nwith regard to the repertoire of group beliefs that they form. They differ in \\nthe contents and in the quantity of group beliefs that they have. Each group \\nhas its own set of group beliefs. Although groups may have similar group \\nbeliefs, two separate groups must have at least one group belief which \\nmakes them different from each other. With regard to quantity, groups may \\nhave few group beliefs or a very long list of group beliefs. The selection of \\ngroup beliefs depends very much on a group's goal(s), common experiences, \\nideas regarding symbols, emerging norms and values, constructed rational­\\nizations for group formation, perceived similarities, perceived uniting forces, \\nand so on. \\nOne of the few studies of small groups from which the emergence of \\ngroup beliefs can be inferred is the study of norm formation by Sherif \\n(1936). He found that a group of strangers facing an ambiguous stimulus \\ntended to converge their judgements into a uniform norm. That is, common \\nexperience in a group situation served as a basis for establishing a common \\nsocial reality. Sherif concluded from these results that individuals who are in \\ncontact, in a defined place, form common values or norms as a basis for \\ngroup formation. It may be further assumed that some of these norms and \\nvalues may serve as group beliefs, since they may characterize the group \\nand underlie its uniqueness. In this vein are findings obtained by Festinger \\nand Thibaut (1951), who observed that 'belonging to the same group tends \\nto produce changes in opinions and attitudes in the direction of establishing \\nuniformity with the group' (p. 92). \\nAnother line of research by Sherif and his colleagues that contributes to \\nthe understanding of group belief formation was the Robbers' Cave \\nexperiments, which investigated intergroup conflict and cooperation (Sherif, \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n107 \\nHarvey, White, Hood & Sherif, 1 961). The experiments also studied the \\nformation of a group. In one of them, 22 fifth graders were randomly \\ndivided into two groups and brought to separate camps so that group \\nformation would be the consequence of the bonding process within the \\ningroup without the contact of an outgroup. Within seven to eight days the \\nformation of the two groups emerged and stabilized. As a point of interest, \\nit can be noted that the two groups formed various group beliefs which \\ndefined their uniqueness and character. Thus, for example, the first group \\nselected the name 'Tom Hale Rattlers', chose a flat, referred to several \\nplaces as their territory, and developed a norm of 'toughness' (reflected in \\ncursing and avoidance of reporting injuries). The second group called itself \\n'Eagles', selected their own song, and selected places in their territory for a \\ncampfire and swimming hole. \\nAs Sherif et al. (1961) pointed out, these beliefs were produced from \\nscratch. They were formed as a consequence of the interaction processes \\nfound in intragroup relations. They served as unique characteristics of the \\ngroup and defined the substance of the 'groupness'. The particular group \\nname, the flag, the norm or the territory, in addition to the belief 'We are a \\ngroup', defined the essence of the group. For group members, these group \\nbeliefs served as a cognitive basis for their feeling of 'groupness'. The belief \\nthat they are a group, as well as other group beliefs, became part of their \\nreality. It can be said that although the mere division into groups created \\nthe fundamental group belief, the addition of group beliefs not only \\nstrengthened the group identity, but also provided meaning to the we-ness \\nreality. Group beliefs, thus, are an inseparable part of social identity. They \\nprovide the essence to the mere feeling of belonging to a group. \\nIt should be noted though that group beliefs do not have to be formed \\nafter the group was founded. Already it has been indicated that certain \\nbeliefs may function as antecedents for individuals to form a group and \\nlater become group beliefs. Also, individuals formulate beliefs which may \\nserve as a basis for group formation. \\nNumerous political parties, religious groups and volunteer organizations \\nare based on group beliefs which were formulated before the group was \\nformed. In this process, the belie\\nf\\ns\\n are first f\\normulated, and then an attempt \\nis made (sometimes it is even unintentional) to f\\norm a group in which \\nthese belie\\nf\\ns\\n serve as group belie\\nf\\ns. The following examples illustrate this \\nprocess. \\nThe Jehovah's Witness group was founded in 1 872 by Charles Taze \\nRussell in a suburb of Pittsburgh, Pennsylvania. Russell, who was brought \\nup as a Presbyterian, was exposed to ideas of Adventism and Second \\nAdventism until he formulated his own beliefs. These beliefs, which were \\nlater developed and modified by his successor, Joseph Franklin Rutherford, \\noriginally served as a basis for the formulation of a new denomination \\n(Stroup, 1945). \\nCannon (1973) described how the Black Panthers formulated their group \\nbeliefs: \\n\\n\\n1 08 \\nSocial identity: international pers\\npectives \\nOne afternoon in October, 1966, Huey told Bobby Seale, 'We need a program. \\nWe have to have a program for the people. A program that people can under­\\nstand. A program that the people can read and see, and which expresses their \\ndesires and needs at the same time.' That night they sat down in the North \\nOakland Poverty Center, where they were working, and wrote out the l O-point \\nprogram which is still the basis for all Panthers' actions. (p. 339) \\nThe programme concerned freedom for the black community, full \\nemployment, the end of exploitation, decent housing, true education, the \\nexemption of blacks from military service, the end of police brutality, free­\\ndom to black prisoners, establishment of black courts and self-determination \\n(Cannon, 1973). \\nIn the present conception, an important question should be asked: Which \\ncommon belie\\nf\\ns\\n (that is, shared belie\\nf\\ns\\n by group members) become group \\nbelie\\nf\\ns? In principle, many beliefs can become group beliefs in the group \\nmembers' repertoire. In reality, however, relatively few beliefs achieve this \\nstatus. Three conditions are suggested as an explanation as to which shared \\nbeliefs become group beliefs. One condition refers to the functionality of \\ngroup beliefs. Belie\\nf\\ns\\n which are f\\nunctional f\\nor group f\\normation and main­\\ntenance become group belie\\nf\\ns, since it is assumed that the need to belong and \\nthe need to f\\norm social identity are important needs which underlie human \\nbeings' cognitions and behaviours. In this context, of special importance are \\nbeliefs which differentiate between the ingroup and outgroups, since they \\nallow demarcation of the group boundary and provide significant infor­\\nmation about the group. \\nThe differentiating beliefs indicate the similarity of group members which \\noverrides the differences between them. That is, through the psychological \\nprocess of categorization, group beliefs simultaneously make the group \\nmembers uniquely similar, yet distinctively different from other groups. \\nThey provide the individuals with a system of orientation towards their \\nown group and other groups. This system leads to the perception and \\njudgement of individuals of the same category as more similar to one \\nanother than they really are and enhances the perceptual and judgemental \\ncontrast between individuals not belonging to this category (Tajfel, 1 978d, \\n1 981). An experiment by Allen and Wilder (1 979) demonstrated this \\nphenomenon. They created two groups, ostensibly on the basis of artistic \\npreferences (which served as group beliefs), and then had group members \\ncomplete an opinion survey in the manner they thought another member of \\ntheir own (or the other) group would respond to the items. They found that \\nindividuals assumed that another member of their group would express \\nbeliefs more similar to their own (previously assessed) opinions than would \\nan outgroup member. This difference occurred even for belief items \\nirrelevant to art. \\nIn addition, those beliefs which provide information about the group \\nbecome group beliefs. It is no accident that group beliefs consist of contents \\nwhich contain such elements as group history, group goals and common \\ncharacteristics of the group. These contents provide information for group \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n1 09 \\nmembers about themselves and their commonalities. They provide the \\nraison d'etre for group formation and, later, for group maintenance. Also, a \\nfulfilment of additional functions by a belief may especially strengthen its \\nstatus as a group belief. Thus, for example, beliefs that help to achieve \\ngroup goals, raise the self esteem of the group members or strengthen \\nfeelings of security are likely to become group beliefs. \\nAnother condition pertains to the saliencey o\\nf\\n the belie\\nf\\ns\\n which become \\ngroup belie\\nf\\ns. Belie\\nf\\ns' saliency re\\nf\\ners to those belie\\nf\\ns\\n which draw special \\nattention o\\nf\\n group members because o\\nf\\n their prominence and distinctiveness. \\nIn combination with the first condition, these beliefs are efficiently and \\nswiftly absorbed and demarcate the boundary between the ingroup and the \\noutgroups. In some groups the salient group beliefs may pertain to physical \\nappearance, the most perceivable cues. Special clothing, ornamentation or \\neven skin colour may serve as a salient basis for defining one's own group. \\nIn other groups, salient beliefs may refer to individual's needs (Toch, 1 965), \\ngroup goals or common experiences. \\nFinally, in order f\\nor belie\\nf\\ns\\n to become group belie\\nf\\ns, epistemic authorities \\nof group members have to consider them as such, and support them. \\nEpistemic authorities are those knowledge sources who exert determinative \\ninfluence on the formation of knowledge (Bar-Tal, Raviv & Brosch, 1 991; \\nKruglanski, 1989). Group members attribute high confidence to beliefs \\ncoming from epistemic authority, consider these beliefs as truth, and adopt \\nthem as part of their own repertoires. In our case, the concept applies \\nmostly to leaders who perpetuate beliefs to become group beliefs. Political, \\nintellectual, religious, social and cultural leaders determine to a \\nconsiderable extent which beliefs become group beliefs and influence \\ngroup members to accept them as such. They frequently select the group's \\ngoals which may serve as group beliefs, formulate ideologies or religious \\ndoctrines, decide what events in the past should be remembered as group \\nbeliefs for the future, select the symbols for the group, and decide on the \\nattributes that characterize group members (for example, Y. Bar-Tal, \\n1989). \\nIt is usually in the formative phase of group development that group \\nbeliefs are established. This is a period of malleability during which group \\nbeliefs are formulated and reformulated. With time, group beliefs emerge in \\na relatively stable formulation, although always with a possibility of change. \\nIn any event, as long as the group exists, it will always be faced with the \\ntalk of disseminating group beliefs and maintaining them among group \\nmembers. \\nThese processes are necessary, since group beliefs serve as a foundation \\nfor group existence. The contents of group beliefs provide the raison d'etre \\nfor a group's existence and delineate its uniqueness vis-a.-vis other groups. \\nThe fundamental group belief indicates that the group exists. In addition, \\ngroup members may add other group beliefs of a wide scope of contents. \\nGroup members share group beliefs, are aware of this sharing and believe \\nthat group beliefs define their 'groupness', that is, social identity. \\n\\n\\n1 10 \\nSocial identity: international pers\\npectives \\nImplications \\nThe proposed conception implies that group beliefs, as all beliefs, exist in \\nthe individual's mind. It does not suggest a view of group beliefs as a special \\nentity of the 'group mind' category. Beliefs are related to individuals. \\nGroups, organizations, societies or any other collectivity do not hold beliefs \\non the collective level - only individual members in the aggregate groups \\ndo. But the present conception describes a widely recognized phenomenon \\nthat group members share beliefs and these beliefs may be viewed as \\nproviding the basis for group members' social identity and as defining the \\nessence of that group. Moreover, the presented approach suggests that \\ngroup beliefs are more than a mere sum of group members' personal beliefs \\n(see also Bar-Tal, in press). Sharing of a belief by group members and a \\nrecognition that it characterizes them provide group beliefs with \\ndistinguished properties. \\nFirst of all, group beliefs serve as a basis for group existence. They \\nprovide a common basis with which individuals can identify and through \\nwhich they can define their membership in the group. All groups have to \\nhave group beliefs, and a necessary condition for a formation of any group \\nis formulation of group beliefs. This is in order that individuals should feel \\nthat they have something in common which distinguishes them from \\noutgroups. They need at least one group belief - fundamental group belief. \\nTajfel's minimal group paradigm has demonstrated this effect. Using the \\nminimal group paradigm, numerous studies have shown that the mere \\ncategorization of persons as ingroup or outgroup - that is, formation of a \\nfundamental group belief - is sufficient to establish group identity and to \\ninfluence group members' behaviour in the direction of triggering ingroup \\nfavouritism and outgroup bias (for example, Billig & Tajfel, 1973; Tajfel et \\naI., 1971). But in a real life situation group beliefs are needed in order to \\nprovide individuals with reasons for their group membership. Group beliefs \\nsupply the basis for cementing social identity. \\nSecond, group beliefs tend to arouse high confidence in their content. The \\nperception of group members that a group belief is shared and characterizes \\nthem causes them to ascribe it high validity. The recognition that a belief is \\nshared by group members reduces the subjective feelings of uncertainty and \\nincreases the perceived validity of this belief (see Festinger, 1954; Jones & \\nGerard, 1967). Therefore, group beliefs are usually considered by group \\nmembers as facts, truths or verities. Another factor which contributes to this \\nview is the perception that group beliefs define the essence of the group. \\nBeliefs of this type cannot be held with low confidence, since low confidence \\nmay shatter the basic foundation of group existence. Therefore, shared \\nbeliefs which define the essence of the grouping are held with high \\nconfidence. \\nOne consequence of holding group beliefs with high confidence is their \\nrelative freezing in the group members' repertoire. Group beliefs do not \\nchange easily, since freezing implies closure. That is, group members do \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\nI II \\nnot easily entertain alternative hypotheses to group beliefs, but tend to \\ncollect information which validates them. The change of group beliefs is of \\nspecial difficulty when they are central. The centrality of group beliefs \\nindicates their importance for the group's life. Central group beliefs may \\nbecome part of the group's culture and tradition, and then they are espe­\\ncially enduring. Group members pass them from generation to generation \\nand rarely change them. \\nThird, since group beliefs define the essence of the group, group members \\ntend to organize their personal beliefs in accordance with them. Group \\nbeliefs provide the frame of reference for other beliefs. This process takes \\nplace especially in cases when group beliefs are central in group members' \\nrepertoire. Then, group members draw implications from group beliefs and \\nadd new personal beliefs consistent with them. In extreme cases of an \\northodox group, group beliefs serve as anchors around which other beliefs \\nare organized. \\nFourth, group beliefs provide the group members with special powers. \\nThe mere perception that group members share group beliefs indicates \\nstrength. Converse (1964) suggested in his classic analysis of beliefs that the \\nnumber of people associated with a particular belief system is an important \\nfactor in political system: \\nclaims to numbers are of some modest continuing importance in democratic \\nsystems for the legitimacy they confer upon demands; and much more sporadi­\\ncally, claims to numbers become important in nondemocratic systems as threats \\nof potential coercion. (p. 207) \\nIn addition, the power of group beliefs is derived from the perception of \\nunity and commonality that may characterize group members who are \\naware of sharing beliefs. Knowledge that group members share a given \\nbelief expresses a unity of the group and indicates common fate. Group \\nmembers derive strength out of this knowledge. They feel united, belonging \\nand integrated within the group. One consequence of these feelings of \\nstrength is that group beliefs may serve as a basis for demands, desires or \\ngoals of group members. Leaders of the group usually take into con­\\nsideration the group's beliefs when they make decisions that affect the \\ngroup's life. Being aware that group members share a belief, and are \\ninfluenced by the belief in their group behaviour, leaders pay special \\nattention to group beliefs. Group beliefs reflect the direction that group \\nmembers desire to take in their behaviour. Therefore, leaders frequently \\nmake decisions regarding group behaviour that correspond to group beliefs. \\nFifth, group beliefs may determine the attitudes and behaviours of an \\noutgroup towards the group. Group beliefs are important information for \\noutgroups about any given group. They characterize the group and may \\nimply even the behaviour that the group may take. Group beliefs may \\nindicate possible goals, ideology, values, history, norms or characteristics of \\nthe group. This information enables acquaintanceship with the group and in \\nturn influences the type of intergroup relations that may develop. \\n\\n\\n1 12 \\nSocial identity: international pers\\npectives \\nFinally, if one believes that group members act according to their beliefs, \\nthen group beliefs should be considered as an important source of under­\\nstanding group behaviour. As Krech, Crutchfield and Ballachey (1962) \\nsuggested: \\nMan acts upon his ideas. His irrational act no less than his rational acts are guided \\nby what he thinks, what he believes, what he anticipates. However bizarre to \\nthe behavior of men, tribes or nations may appear to an outsider, to the men, to the \\ntribes, to the nations, their behavior makes sense in terms of their own world view. \\n(p. 1 7) \\nGroup beliefs provide the cognitive basis to many group behaviours. They \\nmay serve as reasons, goals, explanations or justifications for group behav­\\niour. In other words, group beliefs may function as guiding forces for a \\ngroup, and, therefore may determine the direction, intensity and persistence \\nof group behaviour. This is one of the most important implications of group \\nbeliefs. \\nConclusion \\nGroup belief conception provides cognitive and social perspectives to the \\nstudy of a group. The existence of a group is a social reality for group \\nmembers. Group beliefs provide the basis which allows group members to \\nview the group as a social reality. The group, then, is a product not only of \\nstructural characteristics, environment, situational conditions, motivational \\ntendencies or social influence, but also of personal cognitive processes. \\nThese processes determine the essence of the group, since group beliefs \\nprovide the contents which serves as a basis for group formation and group \\nmaintenance. \\nThis conception carries the notion of social identity one step beyond the \\nTajfel and Turner theories. It suggests that social identity is not based \\nmerely on the categorization process, but also on beliefs which provide the \\nrationale for the group existence. The fundamental group belief indicating \\n'We are a group' and additional group beliefs of different contents which \\ndefine the nature of the groupness allow individuals to feel, think and act as \\ngroup members. These beliefs underlie the uniqueness of the group and \\ndiff\\nerentiate it from outgroups. \\nIt is important that behavioural scientists study the contents of group \\nbeliefs and the process through which group members acquire them, become \\naware of sharing them, and are affected by them in their behaviour. This \\nline of interest will, on the one hand, add a cognitive aspect to the study of \\na group and, on the other hand, liberate beliefs from their individual closet \\nto extent their social meaning. In this respect, the present approach provides \\nthe bridge between the individual's level of analysis and group level. Group \\nmembers, as individuals, are the ones that acquire beliefs which shape their \\nreality. But when group members become aware of the fact that their beliefs \\nare shared by other group members, these beliefs acquire special quality. \\n\\n\\nGroup belie\\nf\\ns\\n as an expression o\\nf\\n social identity \\n1 1 3 \\nThat is, they may have important cognitive, affective and even behavioural \\nimplications for the group members as individuals and for the group as a \\nwhole. \\nNote \\nI .  Most of the examples provided in the following review are drawn from formal statements \\nprepared by the groups. Although it is possible that they reflect group beliefs, group members \\nwere not examined to indicate their responses. Therefore, the presented 'group beliefs' should \\nbe treated with caution. They only serve as illustrations for presenting the conception. Further \\nresearch among group members may determine the actual status of these beliefs. \\n\\n\\n8 \\nSocial Identity and National Identity \\nJose Miguel Salazar \\nOf all the variants of social identity, national identity is perhaps the one \\nthat has had the most dramatic impact on historical events, with both \\npositive and negative consequences for humankind. \\nThe great romantic-nationalist movements of the nineteenth century, as \\nwell as the movements of national liberation in the first half of the twentieth \\ncentury and the most recent break-up of multinational European states, are \\nclear examples of the past and present role played by this type of identity. \\nIdentity with national or ethnic categories is evidently also a reference point \\nused by Tajfel in the development of his theory, as inferred from the \\ncharacter of some of his early work (Tajfel, I \\n969b, I \\n970a). \\nEven though the eighteenth and nineteenth centuries have been called the \\nera of nationalism (Kohn, 1 962), this designation could also apply to the \\npresent century, as there is ample evidence of creative manifestations and \\nexcesses of the national idea in our time. In the aftermath of the Second \\nWorld War, nationalism and independence were bywords of the movements \\nto reaffirm identities in countries which were colonies in the previous \\ncentury, particularly in Asia and Africa. Yet not only in the so-called third \\nworld is evidence found of the ascent of this type of identity; it is increasingly \\nnoticeable in the awakening of subdued nationalisms in European countries, \\nand in the breaking-up of previously functioning multinational states. Also \\nin the Americas there are clear manifestations of its increasing importance \\nfrom Canada, to Mexico, to Peru; and even in the Pacific area with the \\nawakening of the aboriginal populations of Australia, silent for centuries. \\nWhoever suggests that the era of nationalism is finished and that we are \\nin a new era of universal humanity must be disconnected from his or her \\nsurrounding reality. The allegations of 'proletarian internationalism' which \\nproposes that the proletariat has no country, as well as of 'cosmopolitan­\\nism', which indicates that as a consequence of transnationalization now it is \\nthe bourgeoisie that has no country, are both questionable. On the contrary, \\nthe so-called process of globalization and the development of the 'global \\nvillage', based on technological improvements, constitutes a good example \\nof the dialectics of the historical-cultural process, since it has had the \\ncurious effect of reinforcing ethnicity and national identities. \\nIt has been claimed that nations and nationalism only acquire reality in \\nrelation to the state and that historically these are intimately connected with \\n\\n\\nSocial identity and national identity \\n1 1 5 \\nthe ascent of the bourgeoisie, and should be considered transient phenomena \\nonly referred to as a stage of historical development (that is, Kedourie, 1 960; \\nKohn, 1 944; Seton-Watson, 1965). Yet it is possible to differentiate between \\nnationalism, or more precisely 'modern nationalism', clearly associated with \\nthe state, and 'national sentiment', which is more ancient and of a more \\ngeneral and less political character (Hertz, 1 944). \\nThe emphasis on the nation-state is important from the perspective of \\nelaborating a political history, but in the study of national phenomena from \\na socio-psychological perspective it is impossible to ignore this other facet. \\nHence in this chapter we shall begin by considering the affective back­\\nground of national phenomena. After that, in order to gain a better under­\\nstanding of national identity as such, some developments within the social \\nidentity approach will be examined in order to answer some questions about \\nthe problem in hand. \\nConsiderations about national sentiments \\nThe definitions of what a 'nation' is have been very varied. The term \\nappears in a fifteenth-century Bible translating a Hebrew word that means \\nat the same time 'ascendance' and 'posterity'. It is also used to refer to a \\nnatural quality, a native quality. In the medieval universities a 'nation' \\nreferred to a group of students from a common region. And the idea of \\nbelonging to a group based on descendance, differentiable from others, was \\npresent and important in history long before the appearance of modern \\nnationalism (Smith, 1 99 1 ;  Chap. 2). \\nEven in the case of empires as ancient as Egypt and China, national \\nsentiment was important, at least for members of some strata. Evidence of \\nthis is found in some hieroglyphs about a Theban prince who expelled the \\nhyksos from Egypt around 1 580 Be; they say: 'I will grapple with them, and \\ncut open their belly! I will save Egypt and overthrow the Asiatics' (cited by \\nSmith, 1 991, p. 46). Even in Greece, being an Athenian or a Spartan had \\nemotional and behavioural implications; the feeling of being an alien, \\ndramatized in Euripides's Medea, must surely reflect a reality understood by \\nthe audience of the drama. \\nOn the other hand, the biological basis of national sentiment, and more \\nconcretely of ethnocentrism, has been considered by social scientists \\ninterested in socio-biology. Campbell (1965) in his provocative essay 'Ethno­\\ncentric and Other Altruistic Motives', elucidates the thesis that ethnocentric \\nmotivation and behaviour is directed to the protection and survival of the \\nmembers of the ingroup, leading to sacrifices and even death. This act by \\ndefinition is an altruistic type of behaviour, going against personally selfish \\ninterests. More recently Reynolds, Falger and Vine (1987) have published an \\ninteresting book, The Sociobiology o\\nf\\n Ethnocentrism, that gathers more up­\\nto-date contributions from this same perspective. \\n\\n\\n1 16 \\nSocial identity: international pers\\npectives \\nThere has been great insistence that 'nationalism' is a phenomena of \\nmodernity, and that it is incorrect to consider it something more trans­\\ncendental. This is a way of dodging considerations derived from the \\nbiological condition of humankind and the fragmentary but unquestionable \\nevidence of the antiquity of 'national sentiments'. There is no doubt that \\nfeelings of attachment to the nation-state are a relatively recent phenom­\\nenon since the objects of attachment ('modem' nations and the 'nation­\\nstate') are also recent, but to deny the attachment to what is perceived as \\none's own (be this a city-state, as in Greece, or a tribe in the Amazonian \\nforest) seems exceedingly restrictive. To do this does not take into account \\nsome basic considerations about the human condition. \\nIn speaking of a 'nation', reference to the fact of 'birth' is important. \\n(Although in some languages this relationship has been lost or never existed, \\nit is very clear in Spanish [nacer = to be born; nacion = nation] and in other \\nRomance languages.) There is a determination given by belonging to a \\ngroup in which one was born, an ascribed role that confers an important \\nanchor to the individual. So although the term 'nationalism' in effect \\nappears in common usage around the period of the French Revolution, the \\nterm 'nation' is much older. The concept of nationalism is endowed with \\nclear political references to the nation-state, which is not necessarily the case \\nfor the concept of nation. \\nThere are groups that in the accepted sense constitute nation-states, \\nbut there are other conglomerates, based on the same principle of birth, \\nthat may generate loyalties and attachments as strong as, or sometimes \\neven stronger than, the nation-state. The following three cases may be \\nmentioned: \\n(a) the positive affect towards a regional group within an existing state \\n('regionalism'); \\n(b) the attachment to a grouping that ignores existing geographical borders, \\nas in the case of the Kurds of the Middle East or the Waiyu in the \\nGoajiro Peninsula that Venezuela and Colombia share (,stateless \\nnationalism'); \\n(c) the positive feeling towards supra-national groupings, as in the case of \\nLatin America ('supra-nationalism'). \\nWe consider that there are certain commonalties in the psychological \\ncontents of the sentiments expressed in these cases, as well as in the case of \\nstate nationalism. And it is this broadly defined national referent which is in \\nour mind when we examine the question of 'national sentiment' in this \\nchapter. \\nSo what is the basis of this national sentiment? If we take a close look, we \\nfind that it is based on four elements: (a) territoriality, (b) shared culture, (c) \\nhistorical memory of genealogical communality and (d) the existence of a \\nnation-state. \\n\\n\\nSocial identity and national identity \\n1 17 \\nTerritoriality \\nArdrey (1967) extensively develops the point of the importance of territory \\nin human behaviour, going back to the biological base of animal terri­\\ntoriality and making use of the concept of territorial imprinting. He defines \\nwhat he calls the 'biological nation' as 'a social group containing at least \\ntwo mature males which holds as an exclusive possession a continuous area \\nof space, which isolates itself from others of its kind through outwards \\nantagonism, and which through joint defense of its social territory achieves \\nleadership, cooperation, and capacity for concerted action' (p. 1 91). His \\nargument is centred on an evolutionary position and observations of \\nphylogenetic ancestors, which from his point of view give evidence of the \\nantiquity of the phenomenon 'nation'. Of course, like many other animal \\nanalogies, this may be questioned, especially if it is sustained that language \\nand culture introduce a qualitative break with evolutionary continuity. Yet \\nthe observations continue to have interesting implications. \\nThe excesses of the approaches centred on geographical determinism, in \\nvogue in the nineteenth century, and its later rejection, have made us lose \\nsight of the importance of the geographical environment in relation to \\nbehaviour. Even though the geographical referent is more tangible and \\nobjectifiable than other referents of nation, there are few authors who give it \\nthe relevance it deserves, one exception being Doob (1964), who bases his \\nstudies about patriotism in Tyrol on the attachment to the 'Heimat'. Also \\nthere has been some recent rethinking about the theme amongst environ­\\nmental psychologists (Altman & Low, 1992; Proshansky, Fabian & \\nKaminoff, 1983). \\nThere are multiple cases in history of the fight to hold on to territories by \\npeople who were born in an area, which may not be the best from the point \\nof view of economic rationality; and there are many well-known examples \\nof the feeling of ostracism of migrants or displaced peoples who must live in \\na territory which is not their own. \\nBut while it is very concrete, the territorial referent is more evident in the \\nregionalism variant of national loyalty. Given geographical variability \\nwithin countries, it is logical that the role of territory and all that is con­\\nnected to it should be more marked in more restricted loyalties. \\nAltogether there is no doubt that territoriality plays an important role in \\nthe development of national sentiments, \\nCulture \\nSharing a culture involves collective ways of solving the problems con­\\nfronted in interacting with the surrounding world, developed and trans­\\nmitted from generation to generation. This level of shared experience is also \\nimportant in the development of national sentiments. \\nLanguage is the supreme creation of culture, so much so that for many \\nanthropologists it is considered its defining trait. In any case it permits the \\ntransmission of collective solutions and their preservation. Language is also \\n\\n\\n1 1 8 \\nSocial identity: international pers\\npectives \\na very early acquisition in human ontogenesis, and so it becomes a basic \\npoint or matrix for the development of feelings of attachment. \\nThere are other elements of culture that become obvious referents of \\nattachment: music, different elements of the so-called 'material culture', \\ncustoms, shared values. They may all be very powerful in awakening \\npositive feelings and loyalties towards the national ingroup. \\nEthnicity \\nGiven the conceptual weaknesses and dangerous political use of the term \\n'race', the concept of ethnicity has become a partial substitute. Smith \\n(I \\n984a, 1984b) develops the idea of 'myths of descent' as the basis of ethnic \\nfeeling and ethnic nationalisms. Through these myths, the members of a \\ncommunity perceive they have a common origin, usually connected to a \\nglorious past. An ethnie is structured around these mythological socio­\\ncultural elements, the association with a common land, elements of \\ncommon culture, and something of great relevance: a 'collective proper \\nname' (Smith, 1 991, p. 21). It must be clear that in talking about myths no \\nattempt is made to disprove the importance of the phenomena, since myths \\nare not to be ignored. As Mariategui (1925), a very important Peruvian \\nthinker, said: 'Myths move mankind through history. Without a myth, the \\nexistence of mankind has no historical sense' (p. 22). Hence considerations \\nabout ethnicity cannot just be discarded, and obviously are connected to the \\ndevelopment and maintenance of national sentiments. \\nState \\nIt is indisputable that the existence of a national state stimulates an \\nimportant type of national sentiment through the systematic use of ideology \\nproducing institutions partially in the hands of the state like the school \\nsystem and the mass media. This is so pervasive that for some the only \\nreally important national sentiments are those related to the state. But this \\nis too restrictive, as regional, ethnic and more recently supra-national \\nidentities may have as much reality, from a psychological perspective. \\nThere is an additional problem in restricting the study of national \\nsentiment to what occurs at the level of the nation-state, and it is that the \\nidea that a nation equals a state is not a viable equation in our present \\nworld. It constitutes, at most, an ideal type, but the great majority of states \\nare really multinational. Some contemplate it explicitly, as Canada or \\nSpain, where there is recognition of the existence of diverse nationalities \\nwithin the state. However, in states such as Venezuela the fiction of a \\nculturally homogeneous state is disturbed with news of massacres of \\nindigenous tribes, such as the Yanomami. Some states that are explicitly \\nmultinational apparently function well, although surprises occur when the \\nlatent centripetal forces gain ascendance over the forces of integration, as in \\nthe painful case of Yugoslavia. \\n\\n\\nSocial identity and national identity \\n1 19 \\nIn defining a nation, there is implicit the existence of one or several non­\\nnations and the existence of limits separating them. When a nation is \\ndefined in terms of a state, it is usually clear what the limits are. But the \\nsame doesn't happen when you base your definition on any of the other \\nelements: the territorial, the cultural or the ethnic. Hence it becomes \\nnecessary to look for theorizings capable of suggesting ways of solving this \\nproblem. \\nSocial identity theory frames the problem of 'borders' in socio­\\npsychological terms. Consequently this theoretical approach is essential in \\nany attempt to understand national phenomena. Hence we shall consider \\nvery briefly some elements of this approach which may help us to \\nunderstand better the question in hand. \\nThe social identity approach and its relevance to understanding \\nnational phenomena. \\nThe identity theory approach is relevant to the problem of defining group \\nborders since it places it at the irrefutable level of human subjectivity. This \\ntheoretical approach, which in its beginnings Tajfel sometimes referred to as \\nCIC theory (categorization-identity-comparison) (Tajfel, 1 98 1 )  has \\nstimulated multiple theoretical and empirical developments, examples of \\nwhich can be found in the other chapters of this book. \\nThe three original basic elements, the existence of a category accepted \\ninternally and externally, the identification with it, and the process of social \\ncomparison, are still very fruitful concepts. The first element becomes \\nconcrete in the form of a label; the second element implies the acceptance of \\nthe category for oneself with all the affective, cognitive and behavioural \\nconsequences. Regarding the third element, the process of social com­\\nparison, it is frequently a positive evaluation of the assumed category, but \\nthis is not always the case. Tajfel (1981, Chap. 1 3) himself presents evidence \\nof studies with children in which there is evidence of devaluation of their \\nown ethnic group. Several studies have produced evidence of devaluation of \\nthe ingroup (Salazar, 1 983a). The postulation of the need to maintain a \\npositive group evaluation seems to have been taken from Sumner's (1906) \\nethnocentrism theory, considered by many as constituting a very basic \\nmeans of understanding intergroup relations (Le Vine & Campbell, 1 972). \\nMany developments have occurred on the basis of the original elements \\nof social identity theory, among these accounts regarding the basis of \\ncategorizations (Billig & Tajfel, 1 973; Turner, Hogg, Oakes, Reicher & \\nWetherell, 1 987), meta-contrast ratios (Turner et ai., 1 987), the acquisition \\nof categories, the hierarchy of identities; the process of depersonalization, \\nthe salience of categories, and so on, are very relevant to understanding \\nnational identity. \\nMaking use of some of these developments, we have attempted answers \\nto some questions related to the understanding of national identities. \\n\\n\\n120 \\nSocial identity: international pers\\npectives \\nQuestions about national identity \\n1 .  What is the motivation that underlies the acquisition o\\nf\\n national identities? \\nKelman (1969) considers that it is based on both perceived instrumentality \\nand a need for transcendence; the latter would give a particular force to the \\nsocial identities related to the national. With a bit of poetic licence, it may \\nbe said that the existence of national communities is based on the desire for \\neternity. So the idea of some to conceive of nationalism as a form of religion \\n(Hayes, 1 926) is not far-fetched; nationalism may be conceived as a lay \\nreligion that offers, like the others, the promise of eternity. \\nThe nation as a community links people not only with a present inhabited \\nby other people they do not and could not possible know, but also with a \\npast and a future. This is possible through depersonalization, derived from \\nthe process of self-categorization. In reality that which transcends, that \\nwhich is eternal, is the group, but when this happens this 'eternity' gets \\ntransferred to the depersonalized individuals that are part of it. \\nThis may be seen as an attempt to postulate a basic religious motivation \\nin human beings. But whether religiosity is something intrinsic to human­\\nkind or is a historical phenomenon is not important. What is argued here is \\nthat national identity is based on an aspect of such religiosity, basically the \\nneed to transcend. \\nIn more mundane terms, the 'transcendence' of a nation may be con­\\nceived of as attainable by the power it exerts or has been able to exert \\nwithin the international context. Hence national identities have their more \\npowerful manifestations related to political states, which are by definition \\nconceived of in power terms. \\n2. On what basis do the perceptions o\\nf\\n similarities and dif\\nf\\nerences arise in \\nthe case o\\nf\\n national identities? \\nIt is known that a dialectical relationship exists between categorizations, \\nexpressed in a label and the perception of similarities. The mere existence of \\na label is not enough, for a base must exist to give justification or backing to \\nassuming the category. It is also evident that the dimensions around which \\nthe comparisons take place vary in terms of the identities considered. \\nHumans' capacity for interaction is limited. Dunbar (1992), making a \\nprojection on the basis of the size of grouping in higher primates, suggests \\nthat 1 50 is the limit of a viable human group, something like the natural \\nsize of a human group. Nevertheless, humankind lives in communities much \\nlarger than that, hence necessarily we live in 'imagined communities' \\n(Anderson, 1983). An ethnic community or a nation, is hence one of these \\n'imagined communities', but even these must be construed around certain \\nelements or dimensions. \\nWhen the question of the basis of national sentiments is analysed, four \\nelements come to the fore: territory, culture, ethnicity and state. These \\nelements may be considered as dimensions along which meta-contrast ratios \\nof similitude-difference may be established. Particularly in the cases of \\nculture and ethnicity, and in a lesser degree in the case of territory, it is \\n\\n\\nSocial identity and national identity \\n121 \\npossible to conceive of continuums in which the meta-contrast ratios \\ngradually change. In the case of the state the categories are discrete and \\nvariations in meta-contrast ratios are less applicable. \\n3. Could we conceive o\\nf\\n levels or hierarchies in the case o\\nf\\n nation-related \\nidentities? \\nIn order to encompass the similitudes between state nationalism, region­\\nalism and supra-nationalism, we may speak of 'nation-related identities'. \\nThese identities tend to occur in concentric circles. The idea of concentricity \\nof 'national-related' identities could be considered as a particular case of \\nhierarchies within the intermediate level, taking as the higher one that of the \\nindividual as member of humanity, and the lowest that of the individual \\nself. \\nWe suggest that there exists a common psychological substratum to \\nidentities centred on a region or a specific regional group, and the one \\ncentred on a specific nation-state or the emergent supra-national identities. \\nWe see this as a matryoshka of identities, one inside the other. \\nThe relative importance of these identities is not permanent and situ­\\national factors play an important role in determining their role or effect on \\nbehaviour. \\n4. How does the process o\\nf\\n social comparison f\\nunction in the case o\\nf\\n \\nnational identity? \\nIn the case of national identities the question of social comparisons is \\nparticularly important. Social comparisons will affect both the salience of \\nthe category as well as its evaluation, and influence the development of a \\npositive social identity. The social comparisons that are established involve \\nother nations or ethnic groups. These comparisons may take place along \\ndifferent dimensions: size of the countries, political systems, state of the \\neconomy, 'well-being' or 'happiness' of the inhabitants, and so on. In \\ndifferent cases the activated meta-contrasts will vary. \\nBut in the activation of 'nation-related' concentric identities, a basic \\nelement to be considered is the degree of transcendence that identity may \\nguarantee without losing the essence of the identity. It may be argued that \\nthe more extensive the reference group, the more positively valued will be \\nthe identity, since the possibility of transcendence is greater. But it is also \\ntrue that the smaller the unit, the greater the homogeneity, and hence the \\nmore solid the meta-contrasts. There must exist an equilibrium point \\nbetween both elements, possibly of transcendence and solidity of identity, \\nwhich constitutes an interesting empirical problem. \\nThere is evidence that when Latin Americans and Venezuelans (or \\nPeruvians or Colombians) are evaluated, the social identity is more positive \\nin the first case (Salazar, 1 983b), and it may be that this is due to an \\nunderlying perception of greater possibility of transcendence at that level \\nthan at the more restricted ones. Something similar seems to be occurring in \\nEurope. \\nBut as has been frequently pointed out, positive social identities may be \\nconnected to negative manifestations of aggressive ethnocentrism. The \\n\\n\\n1 22 \\nSocial identity: international pers\\npectives \\ndevelopment of an identity within a wider context redefines the outgroup, \\nand here the European case is illustrative. The strengthening of the sense of \\nbeing European, a psychosocial correlate to the development of the \\nEuropean community is evident (Echebarria, Elejabarrieta, Valencia & \\nVillareal, 1 992; Fells & Niznik, 1 992). This has led to the channelling \\ntowards non-Europeans of feelings of rejection that in previous periods had \\nbeen directed to other groups within Europe. The racist manifestations \\nagainst Turks in Germany, against North Africans in France, are cases in \\npoint. In Spain the development of \\nthe feeling of being a part of Europe has \\nbeen accompanied by a rejection of Dominicans and other 'South \\nAmericans' that in another definition of the situation, that is, in terms of a \\nHispanic community, were considered less objectionable. \\nThe varying processes of social comparison point to the importance of \\nexploring the consequences that may be derived from placing the same \\nindividual in varying referential frames, be these regional, national or supra­\\nnational. \\n5. In what conditions do dif\\nf\\nerent 'nation-related' identities become salient? \\nTo begin with, it must be accepted that nation-related identities become \\nactivated by field factors, situational factors that affect the salience of a \\nparticular category. In the case of the European community, economic and \\npolitical developments activated a supra-national identity that has been \\nlatent, and submerged under a history of conflicts. Obviously identities are \\npolitically manageable. However, for this to occur they must exist in a \\nlatent form. \\nIn defining the ingroup, there exists an implicit definition of outgroups. \\nYet these definitions are not static. In emphasizing one level of the nation­\\nrelated identities, a group that shares the same identity at another con­\\ncentric level could be defined as an external group. An example makes this \\nclear. The identities of a Caraquefio (from Caracas), a Venezuelan and a \\nLatin American are nation-related and subsist in concentric circles. When \\nmy Caraquefio identity comes to the fore, I consider people from other \\nregions of my country as external groups, with possible behavioural conse­\\nquences. This ceases to be so when my Venezuelan identity gets activated, in \\nwhich case a Colombian, or another national group, gets defined as the \\nexternal group. This externality would disappear when Latin American \\nidentity is invoked, since it encompasses both groups at a higher level of \\nabstraction. \\nFor a Latin American living in Europe, the Latin American identity is \\neasily activated; in a Latin American country different from ours the \\nVenezuelan identity becomes frequently activated; in a gathering taking \\nplace in a small provincial town, the regional identity of coming from the \\ncapital city may come to the fore. It is here that the questions of accessi­\\nbility and ad\\njustment come into play. \\nBut there are qualifications that must be made in relation to the thesis of \\nconcentricity. It is erroneous to think that there exists perfect balance or \\nconcordance within the concentric circles. On the contrary, it must be \\n\\n\\nSocial identity and national identity \\n123 \\nrecognized that sometimes these may appear as contradictory and in effect \\nenter into conflict. In this sense the Basque case is particularly significant, \\nwhere for many this identity is independent of the Spanish one, and hence \\ndoes not function in concentric circles. According to Linz (1986), a high \\npercentage of the inhabitants of the Basque provinces do not consider \\nthemselves Spaniards. Nevertheless the idea of concentric circles probably \\nfunctions with regard to the European identity (Torregrosa & Ramirez, \\n199 1). \\nAnother problematic situation concerns the conflict between national \\nidentity and other types of identity, such as those related to partisan politi­\\ncal loyalties. Here these may take precedence and lead to dramatic situ­\\nations as those of a civil war, as has happened so many times in history. \\nNational identity may also come into conflict with human identity, as \\nwould be the case with a soldier who refused to comply with the order to \\nkill someone on the basis of their national label. \\nThis brings to the fore the question of 'globalization'. The interrelation \\nbetween the different countries in the world has always existed; what is new \\nis the development of technologies that bring us close enough to find out \\nwhat is happening in other places in the world with a speed non-existent in \\nprevious times. But while it may be true that this facilitates the imitation \\nand copying of foreign models, usually from more developed countries, it \\nalso has created consciousness of similitudes and differences, and does not \\nnecessarily lead to homogenizations. A more probable outcome, it seems to \\nme, would be the development of wider communities, on the basis of \\ncommon elements, in which the sense of national identity is not eliminated, \\nbut transferred to a higher level. \\nIn my view, homogenization is not a desirable goal. Rather, I would \\nargue that the recognition and acceptance of differences is preferable to the \\nuniversal assimilation of all to a single pattern. \\n\\n\\nPART III \\nIDENTITY AND GROUP/ \\nINTERGROUP PROCESS \\n9 \\nSocial Identity and Intragroup \\nDifferentiation as Subjective Social \\nControl \\nJose M. Marques, Daria Paez and Dominic Abrams \\nIn analysing collective behaviour associated with religious movements III\\n \\nsixteenth-century Europe, the historian 1.1. Delumeau (1993) wrote: \\nThe Protestant uprising naturally provoked an increasing aversion to heresy, \\nwhich was considered as the ultimate evil . . .  \\n: the traitor inside is worse than the \\nenemy outside. He must be made to leave his hide-out, he must be eliminated as a \\npriority and no other punishment is severe enough for him. (pp. 514-51 5) \\nThis statement illustrates the fact that the deviate members of a large \\ncommunity may be more harshly treated than outsiders, because, while both \\ndepart from the community's values, the former threaten these values more \\nseverely than the latter. Deviates' behaviour and beliefs endanger the \\ncollective value standards more to the extent that these deviates cannot \\nsimply be expelled to the outgroup. \\nPunishment of deviates and traitors appears to be an important com­\\nponent of a community's social regulation system. Another historical illus­\\ntration of this phenomenon is witch-hunting in the Middle Ages (for \\nexample, Hamilton & Rauma, 1995; Yamagishi, 1 995). Witch-hunting, as \\nwell as more recent instances of collective persecution - like the Inquisition \\nin the fourteenth- to sixteenth-century Iberian countries, McCarthyism in \\nthe United States in the 1 950s, or political purges in the Stalinist period \\nin Soviet Union, to mention but a few - often find support from the general \\npopulation, or, at least, from those who identify with the value-system that \\nlegitimates the persecution, even when the persecuted do not represent an \\nimmediate or material danger. This support is sometimes tangible but, most \\n\\n\\nSocial identity and intragroup dif\\nf\\nerentiation \\n125 \\noften, emerges in the form of derogatory attitudes and judgements. It likely \\nstands for the individuals' commitment to the community's standards and \\nself-definition in light of these standards (cf. Doise, Chapter 2, this volume). \\nConcomitant with this social self-definition is the effort to ensure and to \\nlegitimize beliefs on the distinctiveness and relative superiority of the \\ncommunity's standards as compared to the outgroup. This process may also \\nbe viewed as a reaction against social change and as an immediate psycho­\\nlogical response against threatening minorities cf. Perez and Mugny, \\nChapter 10, this volume). It may be a major factor in the maintenance of a \\nclear-cut and secure social identity as well as a guarantee of the individual's \\npsychological well-being as a group member. As sociologist Georg Simmel \\n(1955) put it, in groups structured around their members' attitudes. \\nthe opposition of a member to an associate is no purely negative social factor, if \\nonly because such opposition is often the only means for making life with actually \\nunbearable people at least possible. . . . It allows us to prove our strength \\nconsciously and only thus gives vitality and reciprocity to conditions from which, \\nwithout such corrective, we would withdraw at any cost. Opposition achieves this \\naim even where it has no noticeable success, where it does not become manifest \\nbut remains purely covert. Yet while it has hardly any practical effect, it may yet \\nachieve an inner balance . . .  \\n, may exert a quieting influence, produces a feeling \\nof virtual power, and thus save relationships whose continuation often puzzles the \\nobserver. (p. 19) \\nIn the present chapter, we argue that identification with ingroup norms \\nand values may lead individuals to subjectively engage in processes analog­\\nous to social influence upon actual or represented ingroup instances. These \\nsubjective social influence processes may allow individuals to restore their \\nconfidence in the group and, hence, to re-establish a positive social identity. \\nThe social psychological analysis of small groups and large social \\ncategories \\nSocial psychological research has focused on different factors involved in \\nthe explanation of group processes. Traditionally, this research evolved \\nalong two separate lines which, for simplicity, we may designate as the \\n'small-group approach' and the 'social identity framework' (including self­\\ncategorization theory). The former has focused mainly on processes like \\nproductivity, decision-making, role and status differentiation, cohesiveness \\nor socialization (cf. Brown, 1 988; Cartwright & Zander, 1 968; Levine & \\nMoreland, 1994). Most of this work has been conducted in the realm of \\nface-to-face groups. The latter has emphasized the effects of social cat­\\negorization in the psychological dynamics involved with individuals' sense \\nof group membership. This work has often been conducted with large social \\ncategories and their representations (for example, Tajfel, 1978d). Only \\nrecently have these two lines been subject to integrative efforts (Hogg, 1992; \\nHogg & Abrams, 1 988; Marques & Paez, 1994). \\n\\n\\n126 \\nSocial identity: international pers\\npectives \\nGroup affiliation, social influence and intragroup differentiation \\nResearch in small groups was strongly influenced by the work of Festinger \\nand colleagues. Festinger (1950) proposed that affiliation in groups fulfils a \\nsocial reality f\\nunction, that is, uncertainty reduction through selective \\naffiliation with those who share similar beliefs, and a locomotion function, \\nthat is, facilitation of goal achievement through the association with those \\npeople whose goals are similar to the individual's. Indeed, people in small \\ngroups generate consensual opinions (Festinger, Schachter & Back, 1950) \\nand reject deviates who resist group pressures to comply with the group's \\nposition on a relevant matter of opinion (Schachter, 1951), or who inten­\\ntionally fail to contribute to the group's goal achievement (Jones & \\nDeCharms, 1967; cf. Levine, 1980). Most importantly, evidence shows that \\nincreases in external pressures upon the group, in group uniformity and in \\nthe discrepancy between majority and deviate positions increase the \\nmajority's pressures towards uniformity (Cartwright, 1968; Janis, 1968; \\nLevine, 1 980; Moreland & Levine, 1982; Shaw, 1976). Evidence also suggests \\nthat social reality and group locomotion are ensured within the group \\nthrough two forms of social control, traditionally known as inf\\normational, or \\nprivate, and normative, or public, influence (Jones & Gerard, 1967). How­\\never, research in small-group processes has rarely addressed itself - at least \\nconceptually - to the role played by intergroup relations in processes \\noccurring within groups (Hogg, 1992; Hogg & Abrams, 1988; Marques & \\nPaez, 1994). \\nSocial categorization, social identification and intergroup \\ndifferentiation \\nIn tum, research on large-scale categories has emphasized intergroup more \\nthan intragroup processes. This may be due to the influential work of Tajfel \\nand colleagues (for example, Tajfel, 1978d; Tajfel & Turner, 1979; Turner, \\n1975a). This work started from the assumption that social behaviour may \\nbe conceived of as occurring along one continuum, whose poles are inter­\\ngroup and interpersonal behaviour: \\nWhat is meant by 'purely' interpersonal is any social encounter between two or \\nmore people in which all the interaction that takes place is determined by the \\npersonal relationships between the individuals and by their respective individual \\ncharacteristics. The 'intergroup' extreme is that in which all behaviour of two or \\nmore individuals towards each other is determined by their membership in \\ndifferent social groups or categories. (Tajfel, 1978a, p. 41) \\nWhereas interpersonal behaviour involves the perception of features \\nspecific to the individuals involved, intergroup behaviour implies self- and \\nother-stereotyping. In categorizing themselves as members in a social \\ncategory, people focus on those features that they share with all category \\n\\n\\nSocial identity and intragroup dif\\nf\\nerentiation \\n1 27 \\nmembers. Brown and Turner (1981) advanced this idea under the heading \\nof the self-stereotyping hypothesis. \\nIn brief, the social identity framework proposes that group representa­\\ntions correspond to prototype-like representations, because intergroup \\nsimilarities and intragroup differences are irrelevant in intergroup situations. \\nClearly, this assumption restrains a view of intergroup and intragroup \\nprocesses as simultaneously occurring in social life. Although this is not \\nentirely true (Hogg, 1 992; Turner, 1 99 1), it nevertheless remains that the \\nsocial identity framework (including self-categorization theory) has diffi­\\nculties in dealing with the simultaneous emergence of intergroup and \\nintragroup behaviour. \\nGroup prototypes as denotative norms \\nAmong other well-known assumptions, the social identity framework states \\nthat, in social life, people attempt to achieve cognitive clarity by exagger­\\nating intergroup differences and intragroup similarities (Hogg, 1 992; Doise, \\nDeschamps & Meyer, 1 978; Tajfel, 1 969a, 1 978d; Tajfel & Wilkes, 1 963). \\nBasic to this process is comparative fit (Oakes, Haslam & Turner, 1 994, and \\nChapter 6 this volume; Oakes & Turner, 1 990). Comparative fit stands for \\nthe actual differences and similarities between the stimuli present in the \\njudgement setting that allow the establishment of correlations between \\ncontinuous stimulus properties and discrete categories (cf. also Tajfel, \\n1969a). \\nAccording to Turner, Hogg, Oakes, Reicher and Wetherell (1 987; cf. also \\nHogg, 1 992; Hogg & McGarty, 1 990), comparative fit generates the cog­\\nnitive construction of group prototypes. This process has been formalized in \\nterms of a meta-contrast princi\\nple (Turner et al. 1987; cf. also Hogg, 1 992; \\nHogg & Abrams, 1 988; Hogg & McGarty, 1 990). Given a distribution of \\ngroup member positions on a relevant intergroup dimension (for instance, a \\npersonality attribute or an attitude dimension), the meta-contrast principle \\nallows perceivers to ascertain whether intergroup differences are, on the \\naverage, higher than intragroup differences. Formally, a ratio can be com­\\nputed for each group member, so that the group's most prototypical \\nmember is the one who shows the highest ratio and thus occupies that \\nposition which simultaneously maximizes intergroup differences and \\nminimizes intragroup differences. This position corresponds to the group \\nprototype. From this point, in \\ngroup and outgroup members are subjectively \\nassessed in terms of their respective prototypes. As a result intergroup \\ndifferences are emphasized to the detriment of interpersonal differences \\namong the members of the same group and interpersonal similarities among \\ningroup and outgroup. \\nIn addition, according to self-categorization theory, individuals expect \\ngroup members to match their respective groups' prototypes. Once social \\ncategories become salient, perceivers may judge the 'normative fit' (Turner \\n\\n\\n1 28 \\nSocial identity: international pers\\npectives \\net aI., 1987), that is, the correspondence between the activated categories and \\nthe content properties of stimuli, in the situation. Lack of normative fit \\nwould lead perceivers to switch to more adequate categorization criteria \\n(Oakes et aI., 1994), that is, to alternative intergroup dimensions which best \\nallow the generation of clear-cut intergroup boundaries, and more eff\\nectively \\naccount for the judgemental situation. Different criteria of categorization \\nwill be used to the extent that, depending on the stimulus properties of the \\njudgemental setting, they succeed or fail in cognitively maximizing \\nintergroup differences and intragroup similarities (Haslam & Turner, 1992; \\nTurner et aI., 1987; cf. also Bruner, 1957). To give an example, in a political \\nrally, comparative fit could entail a 'liberal versus conservative' categor­\\nization. But if in that rally one noticed that, regardless of their political \\naffiliation, female speakers supported women's rights whereas men addressed \\nthemselves to other issues, one would likely switch from the initial 'liberaV \\nconservative' dimension to an alternative 'female/male' categorization. \\nIn brief, for self-categorization theory, group prototypes fulfil structuring \\nfunctions. The meta-contrast generates cognitive clarity in that it allows \\nindividuals to perceive the social world in terms of meaningful, differ­\\nentiated units. Concomitantly, individuals deal with emerging intragroup \\ndifferences by changing the criteria of categorization to alternative dimen­\\nsions which hinder such differences in order to steadily maximize intergroup \\ndistinctiveness. They are strongly motivated to maximize intergroup distinc­\\ntiveness, and this requires that social situations be cognitively construed so \\nthat intragroup differences are minimized. \\nBut, according to self-categorization theory, group prototypes also fulfil \\nemotional functions (Hogg, 1992). Following the above-mentioned self­\\nstereotyping hypothesis, self-categorization theory states that individuals \\nidentify with groups to the extent that they perceive a match between the \\nself and the ingroup prototype. When self-stereotyping occurs, the ingroup \\nbecomes fully representative of the self, and the self becomes perceptually \\ninterchangeable with other ingroup members (Turner, 1 981). According to \\nHogg (1992; Hogg & Abrams, 1988), a major implication of this fact is that \\nsocial identification must be logically conceived of not as an outcome of \\ninterpersonal attitudes or similarity among group members, but, rather, \\nas the result of a positive orientation towards the group prototype. As he \\nput it, \\nwhen a specific social identity is the salient basis for self-conception, self­\\nperception and conduct become ingroup stereotypical and normative, perceptions \\nof relevant outgroup members become outgroup stereotypical, and intergroup \\nbehaviour acquires - to varying degrees, depending on the history of relations \\nbetween the groups - competitive and discriminatory properties. (Hogg, 1992, pp. \\n90-91) \\nIn short, when the ingroup is cognitively salient, a positive orientation \\ntowards the self becomes intrinsically equivalent to a positive orientation \\ntowards the group. Therefore, we may see social identification as a positive \\norientation towards the ingroup prototype, as it materializes in the \\n\\n\\nSocial identity and intragroup dif\\nf\\nerentiation \\n1 29 \\nstereotyped self (Hogg, 1 992; Hogg & Abrams, 1 988). Furthermore, as \\nindividuals assimilate themselves to an ingroup prototype, their thoughts \\nand behaviour conform to this prototype. The prototype is thus internalized \\nas a norm defining the appropriate features and behaviour standards of the \\nsocial self (Turner, 1 99 1 ). \\nDenotative versus prescriptive group norms in inter- an intragroup \\njudgements \\nThe above idea is important to our present concerns. From it, we may accept \\na definition of 'group norm' as a descriptive principle of intergroup \\ndifferentiation, or cognitive clarity concomitant with social categorization. \\nBut norms may also be viewed as prescriptive standards, which are closer to \\novert or cognitively exerted social influence (Kenny & La Voie, 1 985). In this \\ncase, norms may be viewed as standards on which ingroup members anchor \\ntheir judgements (Forsyth, 1 995) about other ingroup members in order to \\nascertain, not intergroup distinctiveness, but rather the legitimacy of the \\ningroup's superiority (Marques & Paez, 1994). In this vein, we shall speak of \\ndenotative norms to refer to the former, and prescriptive norms to refer to the \\nlatter kind of process. Notice, however, that prescriptive norms should \\nfunction as a prerequisite for intergroup differentiation. Accepting this point, \\nthe major question is through what kinds of mechanisms do individuals step \\nfrom a descriptive norm, which generates intergroup differentiation and \\nidentification with the ingroup, to a prescriptive norm, according to which \\nsuch identification is regulated and reinforced? We now elaborate on this \\nquestion. \\nThe 'black sheep effect' \\nOur main empirical argument for the distinction between denotative and \\nprescriptive group norms comes from evidence on the so-called 'black sheep \\neffect' (Marques, 1990; Marques, yzerbyt & Leyens, 1 988). This evidence \\n(Marques, 1 990; Marques & Paez, 1994, for reviews), shows that people \\ndifferentiate more strongly between likeable and unlikeable ingroup members \\nthan between likeable and unlikeable outgroup members. For instance, \\nMarques et al. (1988) asked their Belgian subjects to judge attractive and \\nunattractive Belgian and North African students on a series of traits. The \\nresults showed that attractive Belgian students were judged more favourably \\nthan attractive North African students, whereas unattractive Belgian \\nstudents were judged more un \\nfavourably than unattractive North African \\nstudents. Similar results were found by Marques and Yzerbyt (1 988). These \\nresults showed that law students evaluated a good performance of another \\nlaw student more favourably than a similar performance by a philosophy \\nstudent, and evaluated a poor performance by a law student more unfavour­\\nably than a similar performance by a philosophy student. Also, Marques, \\n\\n\\n1 30 \\nSocial identity: international perspectives \\nRobalo and Rocha (1992) found that high-school pupils evaluated more \\nextremely, both negatively and positively, likeable and unlikeable pupils of \\ntheir own school than similar pupils of a rival high school. \\nTaken at face-value, these findings might suggest an interpretation based \\non normative fit. That is, when faced with likeable and unlikeable in \\ngroup \\nand outgroup members, subjects would switch from the initial ingroup­\\noutgroup categorization to one which proved more informative given the \\nstimulus-material they were provided with (that is, likeable and unlikeable \\ncharacteristics or good and poor performance). However, other findings \\nraise difficulties with this explanation. For instance, Marques et al. (1988, \\nExperiment 2) showed that subjects over-rated likeable in \\ngroup members \\nand derogated unlikeable ingroup members, relative to their outgroup \\ncounterparts, only on dimensions relevant to the definition of the ingroup as \\na whole. On other equally socially desirable or undesirable dimensions, but \\nwhich did not contribute to the distinctiveness of ingroup and outgroup, \\nsubjects were equally extreme in their evaluations of ingroup and outgroup \\nmembers. In another study, Marques (1990) showed that subjects differ­\\nentiated more strongly between normative and counter-normative ingroup \\nthan outgroup members, but only when they considered the normative \\ndimensions as directly relevant to the ingroup's image. Moreover, contrary \\nto an interpretation based on a switch from the original ingroup-outgroup \\ndimension to another one (distinguishing, for example, between likeable and \\nunlikeable persons), Marques and colleagues found that ingroup bias (that \\nis, positive differentiation of ingroup from outgroup as a whole) generally \\nco-occurs with higher differentiation between likeable and unlikeable \\ningroup rather than outgroup members. \\nIn discussing evidence on the black sheep effect, Marques and Paez \\n(1994) drew from the general assumptions of the social identity framework. \\nHowever, they raised a problem with this framework as regards the \\nexplanation of the entire set of results. Research on intergroup relations has \\nlong shown that ingroup members are expected to be more likeable than \\noutgroup members (for example, Brewer, 1979). Therefore, it is true that an \\nexplanation based on a search for intergroup differentiation and clarity \\nof group boundaries may partly account for more negative judgements of \\nundesirable (that is, outgroup typical) than of desirable (that is, in \\ngroup \\ntypical) ingroup members. A simple reading of the social identity frame­\\nwork would lead one to state that socially undesirable (that is, non­\\nprototypical) ingroup members are disliked because they contribute nega­\\ntively to the clearness of group boundaries. In tum, socially desirable (that \\nis, prototypical) ingroup members would be liked, since they bolster such \\nclearness (Hogg, 1992). However, this reasoning would fall short of \\nexplaining why unlikeable (that is, outgroup typical) outgroup members are \\nnot preferred to likeable (that is, ingroup typical) outgroup members. \\nIndeed, in assuming that the subjects' basic motivation is to enhance inter­\\ngroup differentiation, one should expect them to express more unfavourable \\nfeelings towards undesirable ingroup and desirable outgroup members. \\n\\n\\nSocial identity and intragroup diff\\nerentiation \\n1 3 1  \\nThese feelings would ensue from a negative reaction to those who threaten \\nthe intergroup boundaries. By the same token, subjects would express more \\nfavourable feelings towards desirable in \\ngroup and undesirable outgroup \\nmembers, who, by definition, bolster such boundaries. But, clearly, this was \\nnot the case. To Marques and Paez (1994), the social identity framework \\ncould account for the black sheep effect if it encompassed the idea that \\npeople derogate socially undesirable ingroup members as part of a cognitive­\\nmotivational strategy to purge from the group those ingroup members who \\nnegatively contribute to positive intergroup distinctiveness. In other words, \\nderogation of unattractive, or deviate, ingroup members, and upgrading of \\nattractive, or normative, ingroup members relative to outgroup members \\nsupports, or at least is concomitant with, intergroup differentiation. Based on \\nthis idea, they proposed an explanation based on what we can designate as \\n'subjective group dynamics'. \\nThe notion of 'subjective group dynamics' \\nThe notion of 'subjective group dynamics' was inspired by an articulation \\nbetween self-categorization theory and theory and research on small-group \\nprocesses (Marques & Paez, 1994). In brief, following self-categorization \\ntheory, self-stereotyping generates full interchangeability between ingroup \\nmembers (including the self) as regards ingroup-defining standards. In other \\nwords, self-stereotyping would generate 'promotional interdependence' \\n(Deutsch, 1968; Lott & Lott, 1965), or subjective interdependence between \\nthe self and other ingroup members. As a result, the value assigned to the \\nsocial self concept depends on the value assigned to the salient ingroup \\nmembers in the judgemental context. In this vein, the self should become \\nfully interdependent from other ingroup members. \\nThe above idea states a cognitive parallel between rejection of deviates in \\nsmall groups and negative evaluations of unlikeable ingroup members in \\nsocial categories which people conceptualize in terms of subjective belong­\\ningness (Hogg & Abrams, 1988; Hogg, 1992; Tajfel, 1978d). Moreover, the \\neffects occurring in this latter case should be strengthened relative to those \\nobserved in face-to-face groups. In face-to-face groups, deviates are \\ndowngraded because they endanger the subjective validity of social reality \\nand the likelihood of group goal achievement. In groups characterized by \\nsubjective membership, social reality exclusively amounts to self-definition. \\nThe only relevant social reality is the positive value assigned to the social \\nself. Any potential jeopardy of this special aspect of social reality would, \\nmost likely, be experienced as more threatening than contingencies arising \\nfrom other aspects of social reality. \\nThe subjective full interdependence between ingroup members allied with \\nthe fact that self-definition engulfs the social reality of the cognitive group \\nwould have logical consequences at the judgemental level. Ingroup members \\nwho are perceived in line with expectations supporting the ingroup's relative \\n\\n\\n1 32 \\nSocial identity: international pers\\npectives \\nsuperiority should generate a state of subjective validity (Turner, 1 991) of \\nthe positive social self. This would explain praise of likeable in \\ngroup \\nmembers as occurs in the black sheep effect. But other ingroup members \\nwhose perceived characteristics or behaviour hinders positive intergroup \\ndistinctiveness would spur strong hostility. This would explain derogation of \\ningroup deviates. \\n'Subjective group dynamics' and the cognitive construction of \\ndeviance \\nRecent research on perceptions of group variability indirectly supports the \\nidea that derogation of ingroup deviates emerges as a way to subjectively \\nsecure perceived group cohesiveness and hence to validate the superiority of \\nthe ingroup relative to the outgroup as a whole. First, there is evidence that \\nwhen social identity is threatened or dominated, individuals tend to perceive \\nthe ingroup as more homogeneous than the outgroup. Lorenzi-Cioldi and \\ncolleagues (for example, Lorenzi-Cioldi, 1988, 1993; Lorenzi-Cioldi, Eagly \\n& Stewart, 1995), for instance, have shown that members of dominated \\ngroups tend more to see themselves as similar to an ingroup prototype (they \\nsee their group as an 'aggregate'), and focus more on the abstract features \\nthat characterize the ingroup. In turn, members of dominant groups tend to \\nsee themselves more as individualized instances of the category (they see \\ntheir group as a 'collection') and focus more on interpersonal differences \\nbetween ingroup members (Doise & Lorenzi-Cioldi, 1990). Consistently \\nwith this phenomenon, evidence collected by Simon and colleagues (Simon \\n& Brown, 1987; Simon & Hamilton, 1994) shows that members of minority \\nor dominated groups consider themselves to be more similar to other \\ningroup members and judge the ingroup to be more homogeneous than do \\nmembers of majorities or dominant groups. In the same vein, Doosje, \\nEllemers and Spears (1 996) have found that members of low-status groups \\nperceive the ingroup to be more homogeneous than the outgroup when their \\nsocial identification is strong. In turn, members of high-status groups, or \\nmembers of low-status groups whose identification is low, perceive the \\ningroup to be more heterogeneous. \\nDerogation of ingroup deviates as a function of lack of perceived \\ningroup cohesiveness \\nClearly, the above-mentioned research allows us to parallel findings in \\nsmall-group settings and evidence found using larger social categories. In \\nboth cases, threats to the group generate a striving on the part of group \\nmembers to increase group cohesiveness, probably as a means to reinforce \\nthe positive aspects of their social identity (Marques & Paez, 1994). In small \\ngroups, derogation of group deviates allows undesirable members to be \\n\\n\\nSocial identity and intragroup diff\\nerentiation \\n1 33 \\npurged from the group. One may thus logically reason that a similar process \\nwould occur when the basis for judgements is membership of cognitive \\ncategories. \\nMarques, Brito, Correia and Serodio (1996, Experiment 2) have recently \\nexplored this idea. They divided their subjects in minimal groups according \\nto a bogus test of imagination, and asked them to fill in an attitude ques­\\ntionnaire. The questionnaire was composed of several statements ordered \\nfrom highly socially desirable to highly socially undesirable, and subjects \\nwere asked to choose, among those statements, the one with which they \\nagreed the most. One week later, subjects were informed of their group \\nmembership (imagination type), and examined responses given by five \\ningroup or outgroup members to the attitude questionnaire. Figure 9.1 \\noutlines the information provided to subjects about ingroup or outgroup \\nmembers in the homogeneous and heterogeneous group conditions. As can \\nbe seen, in the homogeneous ingroup condition, all group members but one \\n(the socially undesirable member) gave exactly the same response as the \\nsubject (item 6). When the ingroup was heterogeneous, two ingroup \\nmembers gave the same response as the subject, two other members gave \\nneighbouring responses, and one gave a socially undesirable response (item \\n2). This was reversed in the outgroup condition. \\nThe subjects were then asked to judge one socially desirable member and \\none socially undesirable member (respectively, members X and Y, in Figure \\n9. 1). Marques and colleagues (1996) predicted that subjects would derogate \\nthe unlikeable ingroup member more strongly in the heterogeneous ingroup \\nthan in the homogeneous ingroup condition, because, in the former condi­\\ntion, the socially undesirable ingroup member would be more threatening to \\nthe validation of the group's identity. In turn, no differences would emerge \\nfrom evaluations of outgroup members. The data supported this idea. As \\ncan be seen in Figure 9.2, subjects derogated the undesirable ingroup \\nmember more than the undesirable outgroup member in the heterogeneous \\ncondition, and derogated the former member more strongly in this con­\\ndition than in the homogeneous ingroup condition. \\nThese results indicate that rejection of ingroup deviates may be a direct \\nfunction of lack of ingroup cohesiveness around valued ingroup standards. \\nWhen the ingroup is cohesive - that is, majority members uniformly adopt \\na valued position so that the group's norms are secured - group members \\nseem to feel the presence of deviates as less threatening to the group than \\nwhen the majority is more heterogeneous, and hence the group's normative \\nsystem is more easily impaired by this presence. \\nDerogation of ingroup deviates and upgrading of outgroup deviates \\nas a prescriptive means to legitimize ingroup norms. \\nIn more general theoretical terms, the above-summarized study suggests that \\nintragroup differentiation is concomitant with intergroup differentiation \\n\\n\\n1 34 \\nSocial identity: international pers\\npectives \\nHomogeneous group \\nHeterogeneous group \\nSocial desirability continuum of \\nIngroup \\nOutgroup \\nIngroup \\nOutgroup \\nstatements \\n1 .  'Homosexuals should be exiled to \\nappropriate facilities from which they \\nA \\nshould never be released so as not to \\nendanger normal people' \\n2. 'Homosexuals should be housed in \\ninstitutions and allowed to leave only \\nY \\nA B C  \\nY \\nB \\non certain special occasions, as long as \\nX \\nX \\nthey do not endanger other people' \\n3. 'Homosexuality is an i llness and \\nhomosexuals should be forced to \\nC \\nundergo medical treatment' \\n4. 'Homosexuality is shameful and gays \\nshould at least try to disguise their \\ntendency' \\n5. 'Although they are normal, gays should \\nC \\nbe helped to find a better way' \\n6. 'Gays, as everyone else, have the right \\nX \\nY \\nX \\nY \\nto choose their own sexual life' \\nA B C  \\nB \\n7. 'Homosexuality is adapted to modern \\nsociety and, therefore, gays should be \\nA \\nappointed to decision-making \\npositions' \\nNotes: \\n1 .  Statement '6' = subject's stand on the issue. \\n2. Distance of ingroup deviate member and outgroup deviate member (D) from their \\ngroups' modal positions = 14.001. \\nFigure 9. 1 \\nMani\\npulation o\\nf\\n group unif\\normity and group members' status: \\ndistribution o\\nf\\n ingroup and outgroup members' responses \\nrather than just the outcome of a switch from one categorization level (for \\ninstance, 'type of imagination') to another (for instance, 'tolerance towards \\nsexual minorities'). Obviously, the notion of 'norm' used in this context is \\nakin to that proposed by Turner and colleagues to designate group proto­\\ntypes, since it would correspond to the definition of the acceptable ingroup \\nposition on a social comparison dimension. However, we also propose that \\nsuch norms encompass a prescriptive component: individuals would be \\nnormative not only as regards the 'denotative' prototypical definition of their \\ngroup (a cognitive parallel, for instance, to Jones and Gerard's [1967] notion \\nof 'informational influence'), but also in terms of a 'prescriptive' demand for \\ncompliance with this prototype (a cognitive parallel to Jones and Gerard's \\n[1967] notion of 'normative influence'). \\n\\n\\nFavourability \\n7 \\nSocial identity and intragroup dif\\nf\\nerentiation \\nMember: \\n-- Likeable \\n1 35 \\n6 \\n7 \\n6 \\n5 \\n4 \\n3 \\n2 \\n- - - Unlikeable \\n5 \\n4 \\n3 \\n2 \\nIngroup \\nOutgroup \\nHomogeneous group \\nIngroup \\nOutgroup \\nHeterogeneous group \\nFigure 9.2 \\nEvaluations o\\nf\\n attractive and unattractive ingroup and outgroup \\nmembers as a f\\nunction of group unif\\normity (Marques, Brito, Correia & \\nSerOdio, 1996, Experiment 2) \\nThe legitimizing role of outgroup 'converts' \\nIn another set of studies, the present authors attempted to show that this \\nprocess is indeed the outcome of an interplay between denotative and \\nprescriptive norms, as defined above. Specifically, whereas denotative norms \\nserve intergroup differentiation purposes, prescriptive norms serve ingroup \\nlegitimization purposes. As a result of prescriptive ingroup norm salience, \\nsubjects would not only downgrade ingroup members who fail to attune to \\nthese norms, but also upgrade outgroup members who attune to them. \\nIn one study, Marques, Abrams and Paez ( 1996, Experiment 3) asked \\ntheir subjects to examine a case of murder, purportedly as part of a study \\non jury decision-making. Subjects were informed that two opposed patterns \\nof decision-making existed (patterns X and Y) and that their membership of \\none pattern would be detected on the basis of their criteria in judging the \\ncrime. Subjects were also to rank the six characters involved, from most to \\nleast responsible. One week later, subjects were informed that, according to \\ntheir reasoning about the murder (but not according to their ranking), they \\nbelonged to one pattern. In one condition (norm condition), subjects were \\ntold that a norm existed according to which those who belonged to their \\npattern should rank the characters exactly as they had done, and that the \\nopposed pattern should show the reverse ranking. In another condition (no \\nnorm condition), subjects did not receive this information. In both con­\\nditions, it was made explicit that the ranking did not indicate to which \\npattern individuals belonged. The only valid criterion was the way the \\nrankings were justified by them. Subjects were then presented with the \\nresponses of five ingroup or five outgroup members. In the ingroup con­\\ndition, these responses were constructed such that: (a) each subject's \\n\\n\\n1 36 \\nSocial identity: international perspectives \\nModal member \\nDeviate member \\nRank \\nSubject's \\nIngroup Outgroup \\nIngroup Outgroup \\nA \\nA \\nB \\nD \\nC \\n2 \\nC \\nC \\nD \\nB \\nA \\n3 \\nF \\nF \\nE \\nE \\nF \\n4 \\nE \\nE \\nF \\nF \\nE \\n5 \\nD \\nD \\nC \\nA \\nB \\n6 \\nB \\nB \\nA \\nC \\nD \\nFigure 9.3 Example of ranking o\\nf\\n characters A-F used to build modal and \\ndeviate members \\nresponse fully matched the ingroup norm; (b) four ingroup members \\ndisplayed exactly the same response; and (c) one ingroup member (deviate \\nmember) displayed a response similar but not identical to the outgroup. In \\nthe outgroup condition, the pattern of information was reversed. To clarify, \\nFigure 9.3 exemplifies the way information was constructed to manipulate \\ningroup and outgroup members' modal or deviate status relative to the \\ngroup norm. \\nAmong other results, judgements of ingroup and outgroup as a whole \\nwere not affected by the norm information. In line with the classical \\ningroup bias, subjects evaluated the ingroup more favourably than the \\noutgroup. That is, the categorization according to the X and Y pattern \\nentailed intergroup differentiation. However, evaluations of modal and \\ndeviate group members varied according to whether subjects were or were \\nnot aware of the ranking norm. In the no norm condition, modal and \\ndeviate ingroup members were always judged more favourably than out­\\ngroup members. But, in the norm condition, modal ingroup members and \\ndeviate outgroup members were judged more favourably than deviate \\ningroup members and modal outgroup members. In other words, when no \\nprescriptive norm existed, subjects behaved according to the same \\nintergroup differentiation principle, in judgements of the whole groups, in \\njudgements of modal members and in judgements of deviate members. \\nHowever, when they were provided with information about such norms, \\ntheir judgements diverged in the direction of the legitimization of the \\ningroup prescriptive norm (see Figure 9.4). \\nThese results suggest that denotative and prescriptive norms simulta­\\nneously operate in judgements of groups and their members. They could not \\nbe accounted for simply in terms of the denotative character of the meta­\\ncontrast principle. Indeed, this principle would state that subjects would \\nprefer normative ingroup members and normative outgroup members to \\ncounter-normative ingroup or outgroup members, because the former \\nreinforce intergroup differentiation whereas the latter hinder such differ­\\nentiation. If, however, another quality of norms (that is, prescriptive norms) \\nserved to legitimate the values underlying social identity, then subjects \\n\\n\\nSocial identity and intragroup dif\\nf\\nerentiation \\nFavourability \\n7 \\n6 \\n5 \\n4 \\n3 \\n2 \\nIngroup \\nOutgroup \\nNorm conditions \\nMember: \\n-- Modal \\n- - - Deviate \\nIngroup \\nOutgroup \\nNo norm conditions \\nFigure 9.4 \\nE\\nf\\nf\\nects o\\nf\\n norm awareness on judgements o\\nf\\n modal and deviate \\ningroup and outgroup members \\n1 37 \\nwould prefer outgroup members, who, by subscribing to ingroup norms, \\nwould hinder intergroup differentiation, to outgroup members, who, by \\nsubscribing to outgroup norms would reinforce such differentiation. \\nWhereas denotative norms serve to differentiate between ingroup and out­\\ngroup as a whole, prescriptive norms serve to legitimate such differentiation. \\nIn support of this interpretation, we found that, in the norm condition, \\nidentification with the group was associated with the derogation of the \\ndeviate ingroup member (r = 0.32) and to the upgrading of the deviate \\noutgroup member (r = 0. 1 8; meta-analytical average r regardless of \\ncorrelation valence = 0.26). Instead, in the no norm conditions, identi­\\nfication with the ingroup was associated with a more favourable evaluation \\nof the modal outgroup member (r = 0.68, p < .02). This result suggests that, \\nin a minimal group situation, subjects' judgements follow the meta-contrast \\nprinciple. They judge group members depending on their membership rather \\nthan on their contribution to the legitimization of ingroup norms, and value \\ningroup distinctiveness to the detriment of prescriptive aspects of normative \\ningroup conformity. However, when prescriptive norms enter the picture, \\nsubjects appear to simultaneously value intergroup distinctiveness and \\ngroup members' normativity. \\nIn another study (Marques, Abrams & Paez, 1996, Experiment 2), instead \\nof manipulating subjects' awareness of a prescriptive norm, we manipulated \\nthe normative components of the judgemental setting: whereas some \\nsubjects were informed that their judgements of modal and deviate members \\nwould be scrutinized by typical ingroup members (ingroup accountability \\ncondition), others were informed that these judgements would be scrutinized \\nby typical outgroup members (outgroup accountability condition). We \\nreasoned that, if judgements uniquely depended on the meta-contrast \\nprinciple, outgroup accountability would make intergroup comparison more \\nsalient. As a result, subjects would upgrade ingroup members relative to \\n\\n\\n1 38 \\nFavourability \\n7 \\nSocial identity: international pers\\npectives \\n6 ~ \\n5 \\n4 \\n3 \\n/ \\n/ \\n2 \\nMember: \\n-- Modal \\n- - - Deviate \\nIngroup \\ntarget \\nOutgroup \\ntarget \\nIngroup \\ntarget \\nOutgroup \\ntarget \\nIngroup accountable \\nOutgroup accountable \\nFigure 9.5 \\nE\\nf\\nf\\nects o\\nf\\n in \\ngroup and outgroup accountability on judgements o\\nf\\n \\nmodal and deviate ingroup and outgroup members \\noutgroup members, regardless of their normative status, and would show a \\nstronger ingroup bias in judgements of the entire ingroup and outgroup. \\nHowever, if judgements were guided by a prescriptive principle, ingroup \\naccountability would enhance the prescriptive status of the judgements. As \\na result, subjects would derogate deviate ingroup members and upgrade \\ndeviate outgroup members more in this condition, while simultaneously \\nshowing a stronger ingroup bias in judgements of ingroup and outgroup as \\na whole. The results supported this idea. Subjects upgraded the ingroup as \\nwhole relative to the outgroup more in the ingroup accountability condition \\nthan in the outgroup accountability condition. At the same time, they \\nderogated the deviate ingroup member and upgraded the deviate outgroup \\nmember more than when they were accountable to the outgroup (see Figure \\n9.5). \\nAgain, correlational data showed that, in the ingroup accountability \\ncondition, the less subjects identified with the ingroup, the less they \\nupgraded the deviate outgroup member. In the outgroup accountability \\ncondition, the correlation was also negative but non-significant (r = -0. 1 1 ,  \\nn.s.). Together with the whole set of results, these data indicate that, in a \\nstrongly normative context (ingroup accountability or high normative \\nawareness), subjects appear more motivated to legitimate in \\ngroup norms by \\nupgrading outgroup members who endorse, and by downgrading ingroup \\nmembers who oppose, ingroup norms. \\nA model of intergroup and intragroup differentiation \\nThe results of the studies summarized above suggest that, as predicted by \\nself-categorization theory, and, specifically, the meta-contrast principle (e.g. \\n\\n\\nSocial identity and intragroup dif\\nf\\nerentiation \\n1 39 \\nHogg & McGarty, 1990), social categorization motivates individuals to \\nengage in intergroup differentiation. But these results also suggest that the \\nprescriptive context in which intragroup judgements arise strongly affects \\nthe subjects' responses. Ingroup members were downgraded when the \\ningroup lacked cohesiveness, as well as when prescriptive norms were made \\nsalient or the situation explicitly evoked the anticipated evaluation of rele­\\nvant others (that is, ingroup members). It thus seems reasonable to suppose \\nthat derogation of ingroup members is a part of a more general process of \\n'subjective group dynamics'. \\n'Subjective group dynamics' and 'referent informational influence' \\nThe relationship between the observed patterns of judgement and the \\nconstructs of social identity as a group's social reality as well as of absolute \\ninterdependence among ingroup members seems straightforward. Social \\ncategorization entails a sense of interdependence between the self and the \\nrepresentation of other ingroup members as a means to maintain a positive \\nsocial identity. In turn, this social identity is built around cognitively rep­\\nresented group characteristics and behaviour standards whose maintenance \\nis pivotal to positive intergroup differentiation. As a result, group members \\nshould be motivated to exert subjective control over ingroup members in \\norder to sustain a sense of legitimacy of the ingroup's superior position \\nrelative to other groups. A threat to such legitimacy, and hence to the self, \\nwould arise with any perceived lack of conformity within the ingroup in \\nlight of the relevant ingroup standards. Subjectively control would be \\nexerted especially over those ingroup members who threaten the overall \\npositiveness of the group and its relative superiority over the outgroup. \\nIn brief, we propose that group judgements may, in circumstances more \\ncomplex than strictly minimal intergroup situations, arise from an 'implicit \\ntheory' of group processes in the individual's mind. With Figure 9.6, we \\nhave attempted to sketch the proposed process. \\nThe basic idea underlying the process depicted in the figure is that \\nderogation of some ingroup members occurs when, following social cat­\\negorization individuals know the normatively appropriate behaviour to be \\nadopted by ingroup members in order to protect a positive intergroup \\ndifferentiation (Abrams, 1990). This corresponds to re\\nf\\nerent inf\\normational \\ninfluence (Hogg & Turner, 1987; Turner, 1991; cf. also Hogg & Abrams, \\n1988). Ingroup members who are perceived as not observing these behaviour \\nrequirements will trigger 'subjective group dynamics'. They will enhance \\nsubjective social control in order to secure the positive ingroup differen­\\ntiation. In the absence of perceived norm violation, the activated contrasting \\ncategory labels will suffice to generate intergroup differentiation. This \\nphenomenon can be explained in terms of a 'denotative' meta-contrast, and \\ncorresponds to the intergroup level, which represents the classical process of \\nintergroup differentiation and social discrimination. \\n\\n\\n1 40 \\nSocial identity: international perspectives \\n1 .  Social categorization and intergroup salience \\nSTIMULUS \\nSETIING \\nFit \\n(Comparative) \\n(Normative) \\nsecure social \\nidentity \\n(ingroup bias) \\ninter-category \\nsalience \\nsocial identity as \\n'social reality' and \\nsuperordinate goal \\nnorm \\nviolation \\n? \\nyes \\ndenotative \\ncognitions \\nM ETA-CONTRAST \\nself-stereotyping and \\nperceived \\ninterchangeability \\nbetween self and ingroup \\n2. Intragroup salience and subjective control \\nperceived absolute \\ninterdependence and \\n'common fate' \\nlegitimization of \\npositive social \\nidentity \\nFigure 9.6 \\nA two-step process o\\nf\\n inter- and intragroup judgement \\nBut when an intragroup distinction becomes salient and threatens inter­\\ngroup differentiation, subjects will derogate ingroup members who do not \\ncomply with the ingroup standards. This corresponds to the intragroup \\nlevel. In this vein, the intragroup level would correspond to a subjective \\neffort intended to legitimize the intergroup differentiation, and to warrant \\n\\n\\nSocial identity and intragroup diff\\nerentiation \\n141 \\nthe ingroup's unifonnity with regard to the prescribed standards that sustain \\ningroup's superiority. \\nConclusion \\nWe propose that in most circumstances nonnative intragroup differentiation \\nmay function as a foundation for intergroup differentiation and a positive \\nsocial identity. This process can be viewed as the psychological equivalent \\nof public actions of nonn enforcement, like those we mentioned at the \\noutset of this chapter. In this sense, nonn violation is functional to group \\nlife, to the extent that the social control mechanisms it evokes help to define \\nrelevant criteria of group membership. Because the prescriptive nonn only \\nappears as such once it has been violated (Forsyth, 1995), deviance plays a \\nbasic role in defining the group's objectives and interests by generating \\n(subjective or objective) social control (cf. Hamilton & Rauma, 1995). \\nMoreover, active participation in the punishment of deviance may reinforce \\nthe individual's commitment to the nonn, thus reinforcing social identi­\\nfication. In purely psychological tenns, this process might function more or \\nless as predicted by Bern's (1967) self-perception theory: when they lack a \\ndefinite attitudinal position about an object, individuals may infer their \\nattitude from their behaviours ('if I have punished a deviate person I infer \\nthat I have a clear attitude in favour of the nonn'). Moreover, punishment \\nrituals towards deviates may be viewed as part of an initiation process, to \\nreinforce the positive evaluation of the group (cf. Zimbardo & Leippe, \\n1991 \\n). \\nIt is obvious that the presently proposed process is still in need of \\ntheoretical and empirical scrutiny. But we hope that the ideas outlined in this \\nchapter may serve as a valuable contribution to the study of the processes \\noccurring in small and large groups alike, and a starting point for research­\\ning into the cognitive processes underlying the social construction of nonns \\nand deviance in daily life. \\n\\n\\n1 0  \\nCategorization and Social Influence \\nJuan Antonio Perez and Gabriel Mugny \\nA relevant dimension in studies of social influence is whether or not the \\ntargets of influences see themselves as sharing the same social category as \\nthe influence source. The most commonly advanced hypothesis is that the \\neffects of categorization (whether at the cognitive level of accentuating \\nintragroup similarities and intergroup differences, or at a more motivational \\nlevel of ingroup favouritism and/or outgroup discrimination) should facili­\\ntate the influence of an ingroup source and diminish that of an outgroup \\nsource. However, it appears to be difficult to draw any general conclusion \\ngiven that in practice experimental effects have not all consistently been in a \\nsingle direction, especially if manifest and latent influences are distinguished \\n(Moscovici, 1980). The first objective of this chapter is to review those \\nstudies which have varied the degree of identity between source and target, \\nand the second is to present the principal conceptualizations of the links \\nbetween categorization and influence. \\nThe effects of categorization on influence \\nThere is first of all a series of studies which have manipulated the cat­\\negorization (in \\ngroup or outgroup) of sources having a majority status \\n(Doise, Gachoud & Mugny, 1986; Perez, Mugny & Moscovici, 1986; \\nVolpato, Maass, Mucchi-Faina & Vitti, 1990). Ingroup majorities have \\nmore (direct and immediate) influence than outgroup majorities. In those \\nstudies which have also considered indirect dimensions of influence, this \\ningroup majority advantage disappears in interaction with other factors. \\nStudies in which subjects are simultaneously exposed to minority and \\nmajority positions do not contradict and indeed confirm these findings. Thus, \\nin the study by Perez and Mugny (1987), the in \\ngroup majority achieved \\ngreater direct influence than the outgroup majority, but at the indirect level \\nthis simple effect disappeared in interaction with the categorization of \\nthe minority. In the Clark and Maass (1988b) study, categorization of the \\nmajority was not varied - it was an in \\ngroup - but it could be seen that it had \\nmore influence at the public level than either an ingroup or an outgroup \\nminority. At the private level an interaction could again be observed with the \\ncategorization of the minority as either ingroup or outgroup. \\n\\n\\nCategorization and social influence \\n143 \\nIn studies where the categorization involved is based on arbitrary criteria \\n(such as whether one over- or underestimates the number of pupils attend­\\ning secondary school, whether one is in the same classroom, and so on) the \\nin \\ngroup has more influence at the public level than the outgroup. This \\n'simple categorization' effect can even be observed when the source has a \\nminority status (cf. two studies by Martin, 1988b, 1988c, and one by \\nPapastamou et aI., cited in Mugny & Perez, 1991a, pp. 24-27), as well as \\nwhen the minority or majority character of the source is not made salient \\n(Wilder, 1990, study I). \\nIn two studies on attitudes towards abortion, in which gender was used as \\nthe basis for categorization, it was found that a male minority group had \\nmore influence than a female minority group. In the Maass, Clark and \\nHaberkorn (1982) study they constituted an ingroup source (although when \\nthe variance was decomposed the effect did not reach significance), while in \\nthe Perez and Mugny (l 985b) study they represented an outgroup. This \\ngender effect may be specific to the theme of abortion because in a task \\ninvolving estimation of increases in study grants, Martin (l 988b) found that \\na female minority group produced more public influence among other \\nwomen than did a male minority group, while men were more influenced by \\na male source. \\nIf one turns to consideration of \\nanother variable, the distance between the \\nattitude of the target and that of the source, one finds similarly variable \\nresults. In the Clark and Maass (1988a) studies on homosexual rights, an \\noutgroup (homosexuals) produced more influence than an ingroup source \\n(heterosexuals) among those subjects most favourable to homosexual rights; \\nthe ingroup source, however, produced more influence than the outgroup \\nsource among subjects who had the least favourable attitudes. Mackie, \\nWorth and Asuncion (1 990) observed that the ingroup had a stronger \\ninfluence than the outgroup but only among subjects who were most \\nopposed and when the message contained a 'strong' argument. Using the \\ntheme of xenophobic attitudes, Mugny, Kaiser and Papastamou (1983) \\nfound the opposite interaction: the outgroup (a group of foreigners) had \\nmore influence than the ingroup (a group of fellow compatriots) on the \\nmost xenophobic subjects; an in \\ngroup minority had more influence than an \\noutgroup minority on those subjects whose own position was already closest \\nto the anti-xenophobic message. \\nIn four studies in which perception of the source was 'individualized' or \\n'personalized', ingroup and outgroup achieved the same degree of influence \\n(Doise et aI., 1986; Martin 1988a; Papastamou et aI., cited in Mugny & \\nPerez, 1991a; Wilder, 1990). \\nNegotiating style (cf. Mugny, 1982) tends to interact with the identity of \\nthe source. In one study on 'green' attitudes, Mugny and Papastamou \\n(1982) found that when influence targets are led to see themselves as similar \\nin identity to the source of influence, the latter has more influence when the \\nargument is made in a rigid style than when it is made in a more flexible \\nstyle, when subjects are led to see themselves as dissimilar, it is the source \\n\\n\\n144 \\nSocial identity: international perspectives \\nwith the flexible style that is more influential. In a study on attitudes \\ntowards foreigners, Mugny and Perez (1985) observed the same interaction \\neffect. In the same vein, Mugny, Kaiser, Papastamou and Perez (1984) \\ndiscovered that a source sharing the same nationality arguing in favour of \\nrights for foreigners had more influence if the argument was based on \\nradical socio-political principles, while a source consisting of foreigners had \\nmore influence when the same case was argued from humanitarian \\nprinciples. \\nAnother problem concerns the effects of categorization in different \\ninfluence conditions. In studies in which the school to which the source \\nbelongs is varied (same school as the subjects or a rival school), results when \\nthey have been significant have shown that the in \\ngroup has more impact at \\nthe public level than the outgroup (Mackie et aI., 1990; Martin, 1 988a; \\nWilder, 1990). Martin observed that at the private level the outgroup had as \\nmuch influence as the ingroup, both kinds of source producing more change \\nthan that of a control condition without a persuasive message. \\nIn some studies measures of influence at both a direct and an indirect \\nlevel have been available. Thus, in a paradigm based on musical prefer­\\nences, Aebischer, Hewstone and Henderson (1984) found that an outgroup \\nmajority (from a different and devalued education establishment) had more \\nindirect influence than a valued ingroup majority source, while there was no \\ndifference in direct influence. In a study of environmentalist attitudes (Perez \\n& Mugny, 1985a), it was found that when the persuasive message was seen \\nto have derived from a single individual selected at random and not \\nrepresentative of the group, direct influence was greater than when the \\nsource was seen to represent the opinion of the entire group. However, at an \\nindirect level the effects were reversed: the message representative of the \\ngroup had more impact than the unrepresentative message. In a study \\ninvolving attitudes on abortion and contraception (Perez & M \\nugny, 1 986a), \\nthe connotation of a source was manipulated (the source was characterized \\nin terms of entirely positive attributes or entirely negative attributes), while \\nthe source was also an ingroup minority (same age-group as the subjects: \\nyoung people) or an outgroup minority (different category: adults between \\n30 and 50 years old). The minority outgroup with positive connotations was \\nmore influential at the direct level than the outgroup with negative \\nconnotations, and more influential also than the ingroup with a positive \\nimage. At the level of indirect influence the most significant effect was that \\nof the negative-image in \\ngroup, as compared to the positive-image ingroup. \\nIn two studies Volpato et al. (\\\\ 990) measured the effects of categorization \\non creativity. They led subjects to believe that a committee either in their \\nown town (ingroup) or in another town (outgroup) had produced two \\ncommuniques, one by the majority and the other by a minority, respectively \\nopposing or favouring the establishment of a final exam at the completion \\nof baccalaureat studies. They were exposed only to the argument in favour \\nof this proposal. The results showed that a smaller number of subjects \\nindicated they were in favour of the source's proposition when this was \\n\\n\\nCategorization and social in\\nfluence \\n145 \\npresented as originating from an ingroup minority. The proposItion was \\nmore widely accepted when it was attributed to an ingroup majority or \\nindeed an outgroup minority. Although the direct influence of an ingroup \\nminority was shown to be less than in these other two conditions, in these \\ntwo studies this same condition produced a higher proportion of subjects \\nwho invented new solutions to solve the problem. \\nFinally, on the matter of differentiated influence, Mugny et al. (1 983) \\nreported that under certain conditions an outgroup arguing on humanitarian \\ngrounds had greater delayed influence than the same outgroup using a socio­\\npolitical argument. \\nThese, in summary, are the effects which have been observed of categor­\\nization on social influence. It must be acknowledged that their diversity and \\ncomplexity is such that the most prudent conclusion would be to admit that \\nthere is no systematic effect associated with categorization as such, nor any \\nstraightforward relation between categorization and type of influence. The \\nhypothesis that there is an ingroup bias in influence has therefore to be \\nregarded as inadequate; the variables moderating the effects of categoriza­\\ntion appear to be as numerous as the paradigms. Perhaps this disparity of \\nmanipulations reflects a belief that 'simple categorization' (cf. Tajfel, \\nFlament, Billig & Bundy, 1971) is sufficient to produce intergroup effects, \\nand, by analogy, differences in social influence. This leads us to consider \\nwhat are currently the most elaborate explanations in this field, and to ask \\nto what extent they can account for this diversity. \\nIndependence and heterogeneity of the source \\nWilder invoked independence and heterogeneity of the source (1 977, 1978, \\n1 990; cf. also Harkins & Petty, 198 1 ,  1983, 1987) to account f\\nor conformity \\nin restricted groups, and more recently to explain why one should expect an \\ningroup to have more influence than an outgroup. The idea is that the \\nindividuals who comprise a group have more influence if they are perceived \\nas independent entities than if they are perceived as a collective entity. The \\nreasons advanced for this are, on the one hand, that information provided by \\na group is treated with less care (for example, in a less personalized and less \\nspecific manner), and, on the other, that members perceived as a group are \\nconsidered less credible because they are attributed less independence. \\nRecently, Wilder (1990) has used this argument to predict that an ingroup \\nwill have more influence than an outgroup. His reasoning is that the target \\nstarts from the presupposition of a greater homogeneity among outgroup \\nmembers than among the members of his or her own group. The target \\nwould recognize finer distinctions within the ingroup which would lead to a \\nperception of members of the ingroup as more independent of one another. \\nThe same author argued that his is not a motivational explanation of \\ningroup superiority (it does not appeal to a motive of ingroup favouritism or \\nof discrimination against the outgroup), but rather a cognitive explanation, \\n\\n\\n146 \\nSocial identity: international perspectives \\nbased on perception of the homogeneity or heterogeneity of members of the \\ninfluence source. A series of four experiments illustrates this model (Wilder, \\n1990), and to these may be added two studies by Mackie et al. (1990) which \\nshow that the message attributed to an ingroup is given greater attention \\nthan one attributed to an 'other' group. \\nOne difficulty posed by this analysis is the implication that an ingroup is \\nnothing but an aggregation of separate individuals. Thus the comparison \\nmade is between an aggregate of individuals and an outgroup rather than \\nbetween an ingroup and an outgroup. This does not take into account recent \\ndevelopments in the theory of levels of self-categorization (Turner, Hogg, \\nOakes, Reicher & Wetherell, 1987), according to which three such levels may \\nbe distinguished: a superordinate level at which individuals see themselves as \\nidentical to all other human beings; an intermediate level at which the social \\nidentity of the individual is defined by the assertion of differences between \\ngroups or categories of membership and non-membership; and finally a \\nsubordinate level which defines personal identity in terms of categorization \\nof the self as a unique being, distinct from all others including members of \\none's own group. From this perspective, 'personalized' perception is not \\nsynonymous with ingroup categorization. If this argument is conceded, it \\nwould appear that Wilder's model, rather than explaining the ingroup­\\noutgroup difference, is more suited to accounting for the superior influence \\nof a personalized source (whether belonging to an ingroup or an outgroup) \\nas compared to that of a source perceived in more 'depersonalized' or \\nhomogeneous terms. \\nThe results of a study (Perez, Mugny & Navarro, 1991) justify our stress \\non the importance of these distinctions. In this study subjects were required \\nto react to a message either in personal terms (subordinate level of se1f­\\ncategorization), or in 'social terms', which is to say at the intermediate level, \\nin one of two ways: either to react as a representative of their membership \\ngroup (in this case young people) or as a member of the outgroup (adults). \\nThe results show that direct influence is greater when the message is treated \\nin either personal or ingroup terms. These findings are consistent with the \\npredictions of Wilder's model. However, at the indirect level, while influence \\nremains strong in the personalized condition, it disappears in the ingroup \\ncondition, an effect which confirms the importance of not confounding \\nthe personalized level with the intergroup level. Additionally, indirect influ­\\nence was found in reaction to the message in intergroup terms. This set of \\nresults is interpretable in terms of the notion of ingroup paralysis: under \\nthe conditions provided subjects conform to positions prototypical of the \\ningroup, which has the effect of paralysing examination of other positions \\nfrom a personal point of view or from any other point of view which is not \\nthat of \\nthe ingroup. In this case the ingroup introduces a resistance to change \\nand innovation. In a series of studies Sanchez-Mazas, Perez, Navarro, \\nMugny and Jovanovic (1993) have confirmed that when subjects are led to \\nthe self-perception that in terms of their attitudes they are identical to their \\nown group, this paralyses any change in their attitudes. \\n\\n\\nCategorization and social in\\nfluence \\n147 \\nInfluence according to self-categorization theory \\nSelf-categorization theory treats influence as an intergroup phenomenon \\n(cf. Turner, 198 1  \\na, 1991). The fundamental factors are the level of self­\\ncategorization and the ways in which individuals seem to operate to \\ndetermine the validity of their opinions and behaviour. In the first place \\nindividuals begin by anticipating that their judgements and actions in a \\ngiven task will coincide with those of other members perceived as similar to \\nthe self and who are regarded as a comparison group appropriate to the \\nsituation. Individuals will be confident that their own judgements are valid \\nto the degree that similar others are observed to express the same opinions \\nor behave in the same way in the situation. \\nUncertainty and loss of confidence in their judgements and behaviours \\nincrease when individuals find themselves in disagreement with other group \\nmembers whom they perceive, despite the disagreement, as similar to \\nthemselves, which is to say as an ingroup. Insofar as this disagreement \\ncannot be attributed to differences between the self and the group this \\ntheory predicts a change in judgements or behaviours in the direction of \\nthose of the group or of those which best represent the ingroup consensus; \\nthose which are most prototypical. \\nNo such expectation exists when the others are perceived as different, \\nwhich is to say as an outgroup. Disagreement in this case can be attributed \\nto differences in group membership, according to the principle of categorial \\ndifferentiation (Doise, 1976). Consequently, disagreement with the outgroup \\nproduces little uncertainty, and it is not anticipated that individuals in these \\ncircumstances would change their judgements or their behaviour. \\nIn the application of his theory to minority influence, Turner (1991) \\nclaims in addition that for a minority to exert influence it must present \\nitself, or be perceived, as forming part of the ingroup (and not an out­\\ngroup), which is to say a subgroup that defends a pro-normative position \\nwithin the ingroup. He predicts a reduction in direct influence if the \\ndisagreement the minority introduces in the group is polarized to the point \\nof turning it into an intergroup, and not simply an intragroup, conflict. \\nThis theory tends also to account for differences between direct and \\nindirect influence in terms of the level of self-categorization which obtains at \\nany particular moment. Thus a minority categorized as an outgroup will \\nhave less influence, whether direct or indirect. In order for minorities to \\nproduce direct influence they must be perceived as part of the ingroup. On \\nthe other hand, if a minority produces a conversion effect (indirect influ­\\nence), this will be because the comparison is made within a broader context \\nin which the minority is perceived as an integral part of the ingroup, and \\nbecause at this superordinate level of categorization, its alternative norm is \\ncongruent with the norms and values of the ingroup. \\nThis theory offers one of the most complete explanations for the tendency \\nof an outgroup to have less influence than an ingroup. An initial problem, \\nhowever, is that the theory tends to predict both a direct and an indirect \\n\\n\\n148 \\nSocial identity: international perspectives \\ninfluence of an ingroup majority and does not take account of the latent \\nparalysis effect which appears to be precisely the effect of overt conformity \\n(Sanchez-Mazas et aI., 1993). A further difficulty is that it does not allow \\nfor the fact that an ingroup minority produces less direct influence than an \\noutgroup minority or one which is not explicitly categorized as an ingroup. \\nAdditionally, the explanation it offers for conversion does not seem to be \\nentirely convincing. In particular, in several studies, it has been found that \\nin the same situation the same source which does not produce a direct \\ninfluence does produce indirect influence. Is one supposed to believe that in \\nthis situation the influence source is perceived simultaneously at two levels \\nof categorization and that, for example, when faced with an item reflecting \\ndirect influence subjects see the source as an outgroup, but when they \\nencounter an indirect item they perceive the same source as part of the \\ningroup? There is a similar difficulty in seeing how it is possible that at a \\npublic level the influence source could be categorized as an outgroup but at \\na private level as an ingroup, or how a public context could prime an \\nintergroup context while in a private context priming a superordinate \\nintragroup context. These possibilities certainly cannot be excluded but no \\nresearch has yet supported such dynamics. \\nDissociation theory \\nThe need to relate different levels of influence to one another is taken into \\nconsideration by dissociation theory, which we will now present. Briefly, \\nthis theory makes a general distinction between two levels of influence: first, \\nmanifest influence, which includes public, immediate and direct influence; \\nand, second, latent influence, which covers influence at the private, delayed \\nand indirect levels. It provides a basis for relating processes that account \\nfor manifest influence (and which the subject can control) to the processes \\nunderlying latent influence (which the subject is less able to control \\nconsciously). \\nManif\\nest in\\nfluence and positive social identity \\nManifest influence is a function of normative and informational dependence \\n(Deutsch & Gerard, 1955). The importance of normative dependence relates \\nto the individual's motivation to acquire, to present or to preserve a positive \\nself-image or self-concept. Given that one function of manifest influence is \\nprotection of a positive social identity, one is influenced at this level by \\nsources able to exercise the strongest normative pressures or in situations in \\nwhich modifications in judgements allow one to avoid costs or gain more \\ncredit than if one maintained one's independence. Susceptibility to manifest \\ninfluence also flows from the uncertainties felt by the subject, such that the \\nmore the source is able to engender uncertainty in the subject (by virtue of \\nthe former's credibility, ingroup identity or majority status), the g-:eater the \\nsource's influence. \\n\\n\\nCategorization and social influence \\n149 \\nLatent influence and social constructivism \\nBy latent influence we do not have in mind only private influence but all \\ninfluence reflecting constructivist activity. Thus, for example, it includes the \\ncase in which subjects change on dimensions which can be inferred through \\nthe operation of a given mental activity. This concerns the attention the \\ntarget pays to what the source says (to its message, its judgement, its \\nattitude, and so on), the deployment of an innovative form of thought \\nwhich allows the target to derive or discover new functions and ramifi­\\ncations on the basis of the information provided, to establish new connec­\\ntions between the content given and other contents, as well as the \\nelaboration of a new representation of the question. Some authors have \\ndiscussed these kinds of activity under the heading of validation processes \\n(Butera, Legrenzi, Mugny & Perez, 1 990-1991; Moscovici, 1980), others \\nunder the heading of divergent thinking (Nemeth, 1986). Huguet, Mugny \\nand Perez (1990-I \\n991) also use the concept of decentration. Within the set \\nof paradigms dealing with attitudes and opinions, these activities have been \\ndescribed as inferences about the organizing principle underlying a position \\n(Perez & M \\nugny, 1986b). In all these cases, however, the same basic \\nhypothesis underlies these various designations: mental activity is required if \\na latent mode of influence is to arise out of social interaction, and it will \\noccur only in the context of certain kinds of social relationship and not \\nothers. \\nPostulates o\\nf dissociation theory \\nTo the degree that manifest influence is directly observable by others or is, \\nultimately, something of which subjects themselves are conscious, then \\nsubjects' identity is at stake whenever they are influenced by any source, and \\ndissociation theory recognizes that identification is more probable in the \\nface of an ingroup source. This hypothesis indicates that a message \\nemanating from an ingroup is more likely to be accepted, even if it is not \\nread or processed, than a message originating with an outgroup source. \\nThis more positive attitude towards whatever originates in the ingroup \\ncan, especially when the subject is highly involved or motivated, give rise to \\na higher degree of attention to a message attributed to an ingroup and thus \\na more elaborated processing of the message as compared to one from an \\noutgroup. Faced with an ingroup, because identification is more intense the \\nmessage can then be accepted with a less defensive attitude and the subject \\ncan feel more motivated to pay attention to it. One may therefore predict \\nthat in certain conditions there will be direct but also indirect influence. This \\nwill only happen, however, if the situation does not induce a paralysis and if \\nidentification with the ingroup is positively valued and poses no conflict of \\nidentification. \\nIn order for the ingroup to have indirect influence, one basic condition \\nmust be satisfied: subjects must perceive a divergence between their own \\nposition and that of their group. One risk for an ingroup is that if \\n\\n\\n1 50 \\nSocial identity: international perspectives \\nindividuals perceive the group to be in agreement with their own position, \\nand this accords with an anticipated homogeneity of attitudes, they will \\ndevelop a socio-cognitive paralysis, the consequence of which is to inhibit \\nthe activity necessary for the emergence of \\nindirect influence. The attitudinal \\nproximity which individuals often establish through their manifest \\nconformity to the ingroup can therefore by this same process limit the \\nlatter's impact to a pattern of influences similar to compliance (positive \\nmanifest influence combined with zero latent influence), as has been \\nobserved in a number of studies (cf. Perez et aI., 1991; Sanchez-Mazas et aI., \\n1 993). One way of motivating subjects to process the ingroup message is \\ntherefore to lead them to perceive a normative conflict between their own \\nposition and that of their group, which flows not from a simple self­\\ncategorization but from an intragroup conflict. \\nAnother difficulty for the ingroup arises in the case of identification \\nconflict (cf. Mugny et aI., 1983; Mugny & Perez, 1991b), in which indi­\\nviduals are aware of belonging to a group which has little positive value and \\nmay even be negatively valued, or in which they risk being identified with \\nsuch a group. In these cases, identity with or similarity to this ingroup \\nthreatens the maintenance or construction of a positive social and personal \\nidentity. Would individuals let themselves be publicly and manifestly \\ninfluenced by a socially discredited group posing such a conflict of identi­\\nfication? In fact it is not enough to respond that in all probability they \\nwould not let themselves be influenced; it is also important to recognize \\nwhat would happen at a latent level in such a situation. A basic postulate of \\ndissociation theory is that the more that socio-cognitive activities and \\nattention are occupied in creating a disidentification or social differentiation \\nfrom a source of influence, the less these are concerned with the influence \\nmessage. In the event that individuals come to give any attention to the \\nmessage and to analyse it, this is undertaken on the basis of a motivation to \\nconstruct a differentiation so that a negative processing of the message tends \\nto predominate. \\nLet us imagine a situation in which the target of influence does not wish \\nto be identified with the source of influence and that in one case this \\ninvolves an ingroup source and in another an outgroup source. \\nIn the situation where the influence source is perceived to be an ingroup \\nwith which the influence target does not wish to be seen to be identified (for \\nexample, the case of a discredited minority), the available data suggest that \\nin \\ngroup favouritism is indeed not the dominant strategy. Studies by \\nMarques (1990) on what he has described as the black sheep phenomenon \\nprovide a direct illustration of the way in which a negatively valued ingroup \\ncan be the object of more discrimination than an equally negatively valued \\noutgroup. The reason for this is that the more intense the identification is \\n(for example, by virtue of a strong identity or similarity between individual \\nand group), the more effort and activity will be required, should the \\noccasion arise, to try to disidentify. This is the reason why ingroup influence \\nmay be zero at the latent as well as at the manifest level. \\n\\n\\nCategorization and social influence \\n1 5 1  \\nIt can thus prove to be more positive for social identity to discriminate \\nagainst an ingroup rather than an outgroup, or to favour the outgroup to \\nthe detriment of the ingroup. Indeed, an outgroup source can paradoxically \\nhave a net advantage, particularly at the latent level, over an ingroup \\nsource. Given that the ties of identification are less close with an outgroup, \\nit seems reasonable to predict that disidentification with such a group \\nrequires less activity and effort, because in this case a ready-made differ­\\nentiation is provided; it is founded on differences of group membership. The \\nmental activity generating latent effects will no longer be confounded with \\nthe activity of differentiation, the two being immediately dissociated. \\nA major hypothesis of dissociation theory is that even in situations \\ninvolving intense conflict a source categorized as an outgroup will be more \\nlikely to produce dissociation, to the degree that social differentiation is \\nassured by the existence of a pre-established categorization, or that the \\ndifferences are so great that to shift in the direction of the source's responses \\ncarries no risk of a 'categorial confusion' (Lemaine, 1974). This dynamic \\nis also involved when the space for comparison of subject with source is \\nmultidimensional, such that there is no negative interdependence between \\nsource and target of influence (Perez & Mugny, 1990). But this does not \\nimply that an outgroup source will have direct influence. In reality, at the \\ndirect level it is the categorization processes that operate, as it were quite \\nnaturally. However, to the degree that this social differentiation is dissoci­\\nated from message reception, conversion effects towards outgroup positions \\ncan appear. If for one reason or another the subject does not come to \\ndissociate reception of the message from construction of this differentiation, \\nthe outgroup will no longer have any significant indirect effect (Butera, \\nHuguet, Mugny & Perez, 1994). \\nConversion dynamics and the nature of the task \\nLet us acknowledge that at the psychological level the influence situation \\nproduces a dissociation of the relation with the source from the processing \\nof the message. One question remains then unanswered, namely: what can \\nmotivate the subject, particularly when faced with an outgroup, to process \\nthe content of the message? \\nA first possibility is that epistemic expectations intervene which are \\nspecific to the nature of task to which the outgroup's divergent judgement \\nrelates. For example, in objective, non-ambiguous tasks (of the type used by \\nAsch, 1956) for which only one correct response, whatever its provenance, is \\nassumed to be possible, the influence mechanism of an outgroup generally \\nrests upon the motivation of the subject to re-establish uniformity of \\njudgements. In such a situation and given the impossibility of changing the \\nsource, it is the subject who ends up changing, always at the latent level, in \\norder to reconcile social differentiation and epistemological beliefs con­\\ncerning the necessity of consensus (Butera et aI., 1994). \\n\\n\\n1 52 \\nSocial identity: international perspectives \\nWhen tests of ability are involved (for example, estimating distances) and \\nsubjects presuppose that there do exist objectively correct answers but do \\nnot know what they are and so feel uncertain of their ability, outgroups can \\nachieve an influence by the triangulation effect already described: \\noutgroups' judgements can be perceived as valid by virtue of their socially \\nindependent origins (cf. Goethals, 1972). \\nWhen the tasks involve opinions, differentiation from the outgroup at the \\nlevel of attitudes itself constitutes an epistemological requirement (Perez, \\nMugny, L1avata & Fierres, 1993) which accounts for manifest discrimina­\\ntion. The indirect effects of an outgroup (in particular when this is a \\nminority) arise from the fact that dissociation allows a centration on the \\ncontents and initiates a debate about ideas (Billig, 1985). Examination of a \\ncontrary argument, even if this is in order to resist it, implies in particular \\nthat one increases one's exposure to information from the other. Here then, \\nas in the case of denial (Falomir, Mugny & Perez, 1996), a process of social \\ncryptoamnesia may be engaged through which, paradoxically, ideas may be \\ninternalized whose social origins may be forgotten as a result of dissociation \\n(cf. M \\nugny & Perez, 1991 \\nb). \\nCertain social minorities (for example, black people) may also constitute \\noutgroups protected by the Zeitgeist in such a way that discrimination \\ntowards them is socially censured, at least for a significant proportion of \\nsociety. In this case, a double effect can be produced. At a manifest level \\nthere may be on the one hand, a greater influence (reflecting social \\ndesirability) which does not, however, generalize to the latent level, and, on \\nthe other, an effect close to socio-cognitive paralysis. Paradoxically it is \\nwhen subjects discriminate against minorities that they experience a conflict \\nthe resolution of which can provoke a change in latent attitudes (cf. Perez, \\nMugny, L1avata & Fierres, 1993). The occurrence of racist, xenophobic or \\nsexist acts can therefore be in conflict with norms and values of justice and \\nsocial equality accepted more generally by subjects. The self-reproach \\narising from this discrimination can thus act as a mechanism for change in \\nlatent attitudes. \\nConclusion \\nIt is clear that intergroup dynamics, at least with respect to social influence, \\nare not confined to the habitually recognized effects of social categorization \\n(accentuation of intragroup similarities and intergroup differences). They \\nalso involve strategies for constructing or maintaining a positive social \\nidentity, and can additionally disfavour the in \\ngroup or favour the outgroup, \\nin particular as a function of whether or not the social context threatens the \\ntargets' personal and social identities in their relations with salient entities in \\nthe categorial field. \\nOf the two factors that are fundamental to understanding social influence \\nin an intergroup context, one certainly is the dynamic of identification or \\n\\n\\nCategorization and social in\\nfluence \\n1 53 \\ndifferentiation, but the other is the conflict created by the divergent position \\nof a source of influence. Simply categorizing a source as ingroup to produce \\nidentification is insufficient to guarantee influence, whether at the manifest \\nor the latent level. It is the divergence and conflict which determine matters \\nin one direction or the other and these alone are able to account for latent \\nprocesses of constructivism. \\nIn all cases, divergence creates a conflict by reason of epistemological \\npresuppositions about the task, in particular expectations about consensus \\nor dissensus. Categorization therefore is not alone in playing an active role, \\nprocesses of influence have their own autonomy by virtue of the relational \\nbut also socio-cognitive conflicts which they introduce. From this point of \\nview, dissociation theory seems to provide a tool which allows us, to a \\ngreater degree than models offering a single-process view of intergroup \\neffects, to organize the multiple dynamics which can arise from the inter­\\naction between categorization and social influences. \\nNote \\nThis text was prepared with the financial support of the Swiss Fond National de la Recherche \\nScientifique, the Spanish Comision Interministerial de Ciencia y Tecnologia (CICYT, SEC95-\\n0628) and the French Centre National de la Recherche Scientifique. \\n\\n\\n1 1  \\nSocial Identity and Interpersonal \\nRelationships \\nMiguel Maya \\nAttraction and interpersonal relationships \\nThe emotional-sexual relationships established between men and women \\nresulting from attraction (from here on referred to as 'relationships between \\ncouples') can, at first, be considered as being situated close to the inter­\\npersonal end of the interpersonal-intergroup continuum, already mentioned \\nin previous chapters of this book. This can be deduced both from the \\nclassical research carried out in social psychology and the formulations of \\nsome authors classified as social identity theorists. \\nThe traditional approach towards studying attraction and interpersonal \\nrelationships has always placed special emphasis on considering them from \\nthe perspective of the individual, without bearing in mind the social or \\nintergroup dimension. This is reflected not only in the most commonly used \\nphrase to define this study area - 'interpersonal attraction' - but also in the \\napproaches and theories used to explain these phenomena. \\nThe main theoretical focal points used to explain attraction are the \\ntheories of cognitive consistence, exchange and interdependence (Kelley & \\nThibaut, 1978) and those based on the principles of association and \\nreinforcement (Byrne, 1971). All these, as well as the factors used to \\nmeasure attraction (proximity, similarity, reciprocity, among others), refer \\nfundamentally to peculiar characteristics of individuals involved in the \\ninteraction. \\nSome authors working within the social identity theory framework have \\nalso argued that this particular type of male-female relationships can be \\nplaced close to the interpersonal end of the continuum (Hogg & Abrams, \\n1988; Huici, 1984). There is a tendency to consider male-female relation­\\nships as intergroup when the treatment shown towards people belonging to \\nanother sex group is undifferentiated, which implies minimal or almost non­\\nexistent knowledge of them. In this case it is likely that stereotyped percep­\\ntion of the outgroup will appear, in addition to intergroup conflict and \\ningroup favouritism. On the other hand, interpersonal male-female rela­\\ntionships tend to be characterized by a one-to-one relationship, a married or \\nheterosexual couple. \\n\\n\\nSocial identity and interpersonal relationshi\\nps \\n1 55 \\nRelationships between couples as intergroup relationships \\nThis heading could also be termed: 'Interpersonal relationships as inter­\\ngroup relationships', which would be no mere play on words. Our intention \\nhere is to show that despite the fact that relationships between couples have \\nbeen traditionally referred to as 'interpersonal relationships', and have thus \\nhad to be placed close to the interpersonal end of the continuum, there in \\nfact exist a number of powerful intergroup aspects within them. Indeed, \\nthese have been acknowledged by the above-mentioned authors. \\nFor example, Huici (1984) argues that the existence of power and status \\ndifferences between women and men in society, the consensual agreement \\non differential perceptions of each sex and the virtual universality of the \\ndiscrimination based on sex tend to locate interactions between members of \\ngender categories close to the intergroup end of the continuum. Hogg and \\nAbrams (1988) or Williams and Giles (1978) come to similar conclusions. \\nHere we would like to emphasize that the consideration of these power \\ndifferences between sexes or the existence of outgroup discrimination \\n(features closely related to the intergroup extreme) does not necessarily need \\nto appear associated with anonymous or collective interactions or those \\nwhich occur in the professional or public spheres. An interaction between \\ntwo people, although charged with positive affection, with completely idio­\\nsyncratic characteristics and situated within the sphere of intimacy, may still \\nreveal powerful intergroup features. \\nThe definitions of intergroup behaviour given by both Sherif (1 967) and \\nTajfel (1978a), clearly show that what distinguishes interpersonal from \\ngroup or intergroup behaviour is not the number of people implied. Take, \\nfor example, the hypothetical case of the businessman X who turns down \\nthe woman Y's application to work in his office because he considers that \\nwomen are less intelligent and capable than men. If this case involved only \\na man and a woman, it would appear to be clearly based on an intergroup \\nrelationship, given that the businessman's behaviour (male) with regards to \\nthe candidate (female) was exclusively concerned with her belonging to the \\nabove group and not with her personal qualities. As Brown (1 988) shows, in \\norder to establish at what point of the interpersonal-intergroup continuum \\nis situated an interaction between two people from different social \\ncategories (such as male and female), we need to study in close detail the \\nnature of that very interaction. If the behaviour and attitude of the man \\ntowards the woman (or vice versa) is that stereotypically associated with \\neach gender, we can then begin to talk of intergroup behaviour. If this turns \\nout to be impossible, the interaction's idiosyncratic features will emerge in a \\nmore interpersonal relationship. \\nIn most societies (although our data refer only to the Western world), \\nboyfriends and girlfriends, wives and husbands and fathers and mothers \\nbehave differently, although coinciding to a great extent with the stereo­\\ntypes and roles that society associates with men and women (Peplau, 1983). \\nA good example of this is that commonly heard saying: 'I can't do that, \\n\\n\\n1 56 \\nSocial identity: international perspectives \\nbecause I'm a married man (woman) .' In this way, we can safely say that \\ncouples, even those characterized by high intimacy levels, also reveal \\nimportant intergroup aspects. We will now go on to discuss our approach in \\nmore detail, referring to two areas in which uniformities can be seen in both \\nmale and female behaviour in a relationship between couples. \\n(1) Conception o\\nf\\n a couple. The majority of studies, carried out with middle­\\nclass American college students, reveal great similarities in the expectations \\nand conceptions that boys and girls have of couples and heterosexual \\nrelationships in general, although some differences have been found. While \\nwomen from all cultural and economic backgrounds generally grant a great \\nvalue to affection and companionship in their relationships, only male \\ncollege students - but not those from a working-class background -\\ncoincided with these feelings (Peplau & Gordon, 1985). Likewise, women \\ngrant a great importance to intimate communication in their relationships \\nwith men (Parelman, 1983). In general, men tend to have more traditional \\nbeliefs with regard to their relationships with women, in keeping with male \\nroles and stereotypes (Peplau & Gordon, 1985). This result is as expected \\ngiven that the traditional concept of a couple would appear to benefit men \\nas opposed to women. \\n(2) Pro\\nf\\nessional development and economic res\\nponsibility. In accordance \\nwith traditional gender stereotypes, economic responsibility for both the \\ncouple's and family's welfare falls on the shoulders of the male. Greater \\npriority is therefore given to his professional career. This is clearly reflected, \\nalong with other signs, in the proportion of men and women who work \\noutside the home. \\nKotkin (1 983) found that 70% of married women at university had, in his \\nstudy, sacrificed something (for example, economic support, change of \\naddress, renouncement of own career) to aid their partner's university \\nstudies or professional life, while only 10% of men had made the same \\nsacrifice for their female partners. It will, therefore, come as no surprise that \\nwomen intending to study at university or work hard at a professional career \\nare often less committed to their couple (Rubin, Peplau & Hill, 1981). \\nHowever, aside from the priority given to men's professional aspirations \\nover women's, we are also concerned with the fact that simply getting \\nmarried and having a family can act as an obstacle for women's professional \\nsuccess. For example, when Harrison, Moore and Rucker (1985) studied the \\nbiographies of more than two thousand eminent men and women who had \\nappeared in various editions of Who's Who, they found that it was less likely \\nfor successful women to marry and have children, and more likely that they \\nwould get divorced. In the same way, Etaugh and Riley (1983) have shown \\nthat single women have higher income levels, from both employment and \\neducation, than their married counterparts, while the reverse model is true in \\nthe case of men. \\n\\n\\nSocial identity and inter\\npersonal relationships \\n1 57 \\nAnother research area which would appear to reinforce the notion that \\nwomen retreat from professional life within a couple deals with the image \\nmen have of successful women (in keeping with the traditional gender \\nstereotypes), especially those working in typically male fields. In general, \\nthese studies portray a negative image, meaning that, to male eyes, these \\nwomen would appear to be less attractive as partners (Pfost & Fiore, 1990). \\nThus, it would seem that although couples can be categorized as inter­\\npersonal (as they involve two people, possess strong levels of intimacy and \\naffection and can vary greatly in nature), they also possess strong intergroup \\nelements: they emphasize the similarities of male attitudes and behaviours, \\non the one hand, and the similarities of female attitudes and behaviours, on \\nthe other, and in doing so they follow the traditional gender stereotypes. The \\nfact that stereotyped concepts of gender are used in couples to evaluate and \\nestablish their behaviour (more so than other concepts) is hardly surprising, \\ngiven that a great many of these stereotypes refer to male-female rela­\\ntionships. Thus, it would seem perfectly acceptable for a traditionally \\nminded male to ask his wife to 'take the initiative' with regard to domestic \\nproblems and their children's education, or even in the realm of her work \\noutside the home, but not in her direct dealings with him. \\nRelationships between interpersonal and intergroup aspects \\nThere exist three fundamental ways of connecting the interpersonal and the \\nintergroup: (1) as orthogonally related; (2) as qualitatively different aspects \\nof reality; and (3) as parts of a continuous flow. We shall examine each in \\nturn. \\n1. Orthogonality o\\nf\\n both dimensions \\nStephenson (1984) was one of the first researchers who, from within social \\nidentity theory, established the possibility that the interpersonal and inter­\\ngroup need not be placed at opposite ends of a single dimension but rather \\nact as two independent dimensions. In this way, a relationship could be \\nplaced at a given point in a bi-dimensional space, resulting from the \\nintersection of both dimensions, as shown in Figure 1 1 . 1 .  \\nIt is possible that there exist certain emotional-sexual relationships \\nbetween men and women in which neither interpersonal nor intergroup \\naspects are of great importance. This would be the case in an incipient \\nrelationship in which contact between the two members of the couple is \\nrather low - the fact that both interpersonal and intergroup requirements \\nare low does not imply that they are of equal importance. In this case, it is \\nlikely that the intergroup aspect would predominate over the interpersonal. \\nOther relationships can reflect a high intergroup requirement and a low \\ninterpersonal one: marriages of convenience provide a good example, or \\n\\n\\n1 58 \\nSocial identity: international pers\\npectives \\nINTERGROUP DEMANDS \\nHigh \\n'Compromise' \\n'Conflict' \\nHigh \\nINTERPERSONAL \\nDEMANDS \\n'Collusion' \\n'Apathy' \\nLow \\nINTERGROUP DEMANDS \\nLow \\nINTERPERSONAL \\nDEMANDS \\nFigure 1 1 . 1  Inter\\npersonal and intergroup demands according to Stephenson (1984) \\nthose marriages typically found in the Victorian age, in which intimacy was \\nvirtually non-existent and almost the entire relationship and the behaviour \\nof the couple had been already established by social criteria. Relationships \\nwith a high interpersonal and a low intergroup component would be those \\nin which the couple challenges established social criteria concerning behav­\\niour and attitudes expected from men and women involved in a love affair \\nand create a complete different and unique relationship. Finally, it should \\nalso be possible to find relationships in which both the interpersonal and \\nintergroup aspects are present. A graphic example of this could be the \\nrelationship portrayed in Liliana Cavanni's film The Night Porter. This \\nstory concerns an ex-Nazi official and a former female prisoner who main­\\ntain a passionate relationship that reflects (to a great extent) intergroup \\nrequirements since a great part of their behaviour is stereotypical (domi­\\nnance, aggression, and so on, in the male and submission, passivity, and so \\non, in the female). However, at the same time, interpersonal aspects are \\nequally important: the relationship is deeply idiosyncratic and strange, \\nconsidering they belong to entirely separate groups, their romance thus \\nbeing distinguished from the millions of male-female relationships estab­\\nlished during the Second World War. \\nAn example of the simultaneity of interpersonal and intergroup aspects \\nappears in our study (Moya, 1987) on social episodes. Following Forgas's \\n\\n\\nSocial identity and interpersonal relationshi\\nps \\n1 59 \\napproach (1 979) a number of subjects (60 female and 36 male psychology \\nstudents, 9 male and 13 female student nurses and 23 housewives) were \\nasked to indicate the perceived similarity among a set of situations rep­\\nresenting usual interactions with members of the opposite sex and rate these \\nsame situations on 1 5  bipolar scales. Both the situations and the scales had \\nbeen obtained in a previous session and were all representative of the \\nheterosexual interactions of each group. Examples of these interactions \\ninclude 'talking on the telephone with a boyfriend/girlfriend', 'having sexual \\ncontact', or 'studying with colleagues'. Here, our interest lies in those \\nepisodes in which the member of the opposite sex is a partner. For example, \\nin the case of the housewives, these episodes included: 'speaking with my \\nhusband', 'having a drink with my husband', 'making love with my \\nhusband', and in the case of the student nurses, 'eating with my partner', \\n'having arguments with my partner', 'making love with my partner' and \\n'waking up next to my partner'. \\nThe results show that the subjects perceive these episodes as intergroup \\ninteractions, in the sense that they consider their behaviour here as being \\nvery similar to that of members of the same sex (remember that uniformity \\nis one of the fundamental characteristics of intergroup behaviour). Thus, on \\nthe scale of 0 to 9 points whose extremes were 'I behave differently to how a \\nmember of my sex would' and 'I behave in a similar way . .  \\n.', the scores \\nobtained in those episodes involving a couple were the following: 7.3 in \\nmale psychology students and 7.7 in female; 4. 1 5  in male student nurses and \\n6.3 in female; and 6.92 in housewives. At the same time, however, these \\ninteractions were perceived as being highly interpersonal, since the subjects \\nsaw themselves as expressing great intimacy, emotional involvement and \\ndignity. They also believed that they expressed themselves spontaneously in \\nthe relationship and felt treated as persons. This is confirmed by the results \\nof a multidimensional scale analysis (INDSCAL by Carroll & Chang, 1970) \\nof the ratings of the interaction episodes on the 1 5  bipolar scales. See, for \\nexample, the episodic space obtained by the housewives, where two \\ndimensions proved sufficient to represent the space. \\nThe first dimension was defined by the scales 'passive-active' (-.86), \\n'intimate-non-intimate' (-.79), 'boring-interesting' (-.71), 'intense-super­\\nficial' (-.58) and 'frustrating-rewarding' (-.51). To our way of thinking, this \\ndimension clearly refers to the emotional aspect of the episodes - which \\ncould be termed intimacy - and is related with the person's degree of \\nemotional involvement in the situation. The episodes 'chatting with our \\nchild', 'making love with my husband', 'playing with our child' and \\n'interactions with friends' were situated at one extreme, while 'speaking with \\nthe children's teacher' and 'going to the doctor' were placed at a midway \\npoint on the dimension. Finally, at the other extreme were positioned the \\nshortest interactions with virtual strangers (shopkeeper, baker, gas man). The \\nsecond dimension was defined by aspects of dignity, equality and control \\nperceived by the person in the episode, the scales defining this dimension \\nbeing: 'relaxed-tense' (-.89), 'pleasant-unpleasant' (-.8), 'cooperative-\\n\\n\\n1 60 \\nSocial identity: international perspectives \\ncompetitive' (-.8), 'free-coerced' (-.79), 'I feel treated as a person-I don't \\nfeel treated as a person' (-.77), 'I know how to behave-I don't know how to \\nbehave' (-.75), 'I express myself as I am-I don't express myself as I am' \\n(-.74) and 'reciprocated-not reciprocated' (-.71). At one extreme of this \\ndimension were located the following episodes: 'a man is pursuing her', \\n'quarreling with our child', 'going to the doctor', 'speaking to the children's \\nteacher' and 'speaking to an office worker'. At the midway point more \\nsuperficial episodes were placed ('going to the hairdressers', 'chatting with a \\nneighbour') and at the other extreme were situated 'making love with my \\nhusband' and the relationships with children and neighbours. \\nAs can be seen, there is no interpersonal-intergroup dimension along \\nwhich the interactions of these housewives with men can be placed sub­\\njectively so that interpersonal episodes could be placed at one extreme -\\nrelationship with partner or with child - and the intergroup ones at the \\nother - a man is pursuing her, going to the hairdresser. An interpersonal \\nrelationship (for example, with their child) is perceived in an entirely \\ndifferent way depending on whether the housewife is playing or arguing \\nwith himlher. \\n2. The interpersonal and intergroup as qualitatively diff\\nerent aspects \\no\\nf identity \\nKay Deaux (1992, 1993) has recently established that social and personal \\nidentities, rather than being clearly distinguishable, are in fact deeply inter­\\nrelated. According to Deaux, social identities can be defined as those roles \\nor categories within which a person feels represented. Personal identity \\nwould refer to those features or forms of behaviour which a person feels \\nadequately describe himlherself; these characteristics are often intimately \\nassociated with one or more social identities. In this way, a male X can \\nregard himself as being very much a typical man, similar to the majority of \\nmen in his social surroundings (social identity), and at the same time \\nconsider himself as 'affectionate' (personal identity). This latter character­\\nistic is probably more associated with his belonging to the category of \\nfather, grandfather or spouse than to that of a military officer (imagine that \\nthis male belongs to all these categories or plays all these social roles). In \\nthis way, personal identity is defined, at least in part, by our membership of \\na group, at the same time as these group memberships will, in themselves, \\nbe charged with personal meaning. \\nIt should also be borne in mind that identity can be perceived as a \\nhierarchical structure in which groups of identities are related to categories \\nof characteristics or attributes. For example, the identity of a female Y \\nwould be exemplified as shown in Figure 1 1 .2. The upper part of the figure \\nshows social identities and the lower - as if it were a mirror - the com­\\nponents of a personal identity. \\nOne of Deaux's primary concerns surrounds this hierarchical structure of \\nsocial identity. The higher a determined identity, the more behavioural \\n\\n\\nI \\nWoman I \\n1 Maintains interpersonal 1 \\nrelationships \\n1 \\nProfessional & 1 \\nindependent \\n1 \\nClose 1 \\n1 Superficial 1 \\n1 Professional 1 \\n1 \\nWith a progressive 1 \\nideology \\nI \\n1 Involved in 1 \\nHousewife 1 1 Popular 1 \\n1 University 1 \\n1 \\nBoss 1 \\n1 \\nFeminist 1 1 \\nLesbian 1 \\na couple \\nwith men \\nstudent \\n- Worried about \\n- Devoted to others 1- Attractive 1 \\n1- Ambitious 1 1- Ambitious 1 \\n- With social \\n- Independent \\nher appearance \\n- Kind \\n- Sexy \\n- Emotional \\n- Aware of her \\n- Attractive \\nfeelings \\n1- Kind \\n1 \\n- Committed \\n1- Nice \\n1 \\n- Good-looking \\n1- Helpful to \\nothers \\n1 \\n- Works hard \\n- Competitive \\nworries \\n- With many \\nI \\nI \\n- Ambitious \\ngirlfriends \\nI \\nI \\n1- Works hard \\n1 \\n- Follows her own interests \\n1- Fights against injustice 1 \\n- Active \\n1- Follows her \\nown interests 1 \\n- Sensitive \\n- Practical \\n- Affectionate \\n- Responsible \\nFigure 1 1 .2 \\nExample o\\nf\\n a hypothetical identity taken f\\nrom Deaux's f\\normulation and Maya's data (Maya, 1 989; Maya & Peplau, in \\npreparation) \\nR \\n'\\\" \\nis' \\n--\\n) \\n \\nS. \\nI:l \\n;::\\n \\n:::.\\n \\nS· \\nő \\n~ \\n(1) \\n' \\n;::\\n \\nI:l \\n--\\n€ \\n::r \\nT. \\n;::\\n \\n} \\n.;:;. \\n'\\\" \\n0\\\\ \\n\\n\\n1 62 \\nSocial identity: international perspectives \\nTable 1 1 . 1  Meta-contrast ratios in self-perception \\nand the f\\nour stimuli used by Moya and Pep/au \\nUCLA \\nGranada \\nPerson involved in a couple \\n1 .09 \\n1 .25 \\nUniversity student \\n1 .22 \\n1 . 1 3  \\nProfessional \\n1.03 \\n0.88 \\nCommitted to their family \\n0.95 \\n1.04 \\nrepercussions this will have for an individual. Although it is likely that \\ncertain exceptional identities occupy high positions in almost all individuals \\nin a society (such as being male or female), it is also possible that there exist \\ngreat individual or subcultural differences. Maintaining her position in a \\ncouple may, for a certain woman be thought of as belonging to an entirely \\nsecondary category included in, or subordinate to, her membership of \\nanother group, such as that of a high social status. On the other hand, for \\nanother woman, being part of a couple could be considered as the most \\ngeneral category, into which all other memberships or social roles fall. It is \\nworth remembering that any characteristics associated with one social \\nidentity can be associated with others and that this association between \\nsocial identities and personal characteristics can vary from one individual to \\nthe next or from one social group to another. \\nSome of these points were verified in recent research (Moya & Peplau, in \\npreparation). A total of 204 subjects (95 men and 1 09 women) took part \\nin the first test, all students of an introductory course in psychology at the \\nUniversity of California at Los Angeles (UCLA). In the second test, 45 men \\nand 1 25 women participated, all third-year psychology students at the \\nUniversity of Granada. All subjects were asked to rate themselves on 22 \\nbipolar scales (9 represented instrumental attributes, 9 expressive and 4 \\nneutral). They were then asked to rate four different types of people of the \\nsame sex on the same 22 scales (university student, person involved in a \\nlasting relationship, a committed family man/woman and a professional). \\nWith these scores, 4 meta-contrast ratios were calculated, one for each \\nstimulus, following Turner, Hogg, Oakes, Reicher and Wetherell's proposal \\n(1987). If the meta-contrast ratio is lower than I, the stimulus in question is \\nperceived as being a member of the outgroup. The higher above I, the \\ngreater the stimulus is perceived as a member of the ingroup. The meta­\\ncontrast ratios obtained by the American and Spanish students appear in \\nTable 1 1. 1 .  \\nUCLA students only considered persons committed to their family as \\noutgroup members (RMC lower than 1), while for those from Granada, \\nonly the professional was considered as such. Likewise, both sets of students \\nperceived themselves as being closer to the person involved in a relationship. \\nAn analysis of variance was performed using as the dependent variable \\nthe meta-contrast ratio. The independent variable was subjects' current \\ninvolvement in a relationship and had two values: involvement versus non-\\n\\n\\nSocial identity and interpersonal relationshi\\nps \\n1 63 \\ninvolvement. For UCLA students involved in a relationship, the meta­\\ncontrast ratio with the person committed to the family was .89. For those \\nnot involved, the corresponding figure was 1 .01 F(I ,  1 55) = 7.26 was \\nsignificant at p < .01 .  No similar differences appeared in the Granada \\nsample. \\nThe importance of taking into account the specific contents that \\nindividuals associate with each social identity has also been shown in other \\nresults of this research. According to our hypothesis, those involved in an \\nemotional relationship with a member of the opposite sex rated themselves \\nas higher in the expressive items than those not involved. On the other \\nhand, and contrary to expectations, this was also the case with instrumental \\nitems (although here the differences did not reach statistical significance). \\nThese tendencies were found in both the American and the Spanish sample. \\nIn our judgement, these results reveal the existence of different social \\nidentity configurations in two distinct social mediums (the American and \\nthe Spanish) since being a university student or a committed family member \\nseems to occupy different positions in identity structure, while the relation­\\nship between these identities in each sample also presents certain differences \\n(as in the opposition present in the American sample between those \\ninvolved in a relationship or not). \\n3. The continuous flow o\\nf\\n identities \\nTurner and coIleagues (1987) conceive of the existence of different levels of \\nabstraction in the categorization of '1', separated from each other and \\ndiscretely formed, in such a way that the individual passes from one to \\nanother according to how the situation varies. It is therefore the dynamic \\ncharacter of identity which is emphasized here, an aspect that has been \\npresent in social identity theory but which is rarely reflected in practical \\nresearch carried out under its inspiration (Condor, 1989). Tajfel (1982a) \\nemphasized how social identifications could be subject to temporal fluctu­\\nations: 'the psychological existence of a group for its members consists of a \\ncomplex sequence of appearances and disappearances, of factors emerging \\nand blurring into thin air' (p. 485). Turner considered his model of identity \\nas neither static, fixed, global or corrected, but totaIly the reverse: 'the \\ncategories of \\\"Self\\\" are not absolute, but rather \\\"Self\\\" is dynamic, relational, \\ncomparative, fluid, context-specific and variable' (Turner, 1988, p. 1 14). \\nThus, it is possible that the interactions maintained by members of a \\ncouple change continuously from some category memberships to others (as \\nconceived by self-categorization theory) or fluctuate incessantly along the \\ninterpersonal-intergroup continuum. For example, deciding what to do one \\nSunday afternoon, it might occur to a boy to invite to the cinema the girl he \\nhas just started seeing, to then pay for her ticket and attempt some physical \\nor sensual contact (foIl \\nowing certain stereotypical criteria). He suddenly \\nfinds out that the girl cannot stand going to the cinema and would much \\n\\n\\n1 64 \\nSocial identity: international perspectives \\nrather have an afternoon chat in a quiet cafe, he thus rapidly changes his \\nplans. In this way, a form of behaviour which commenced with certain \\nintergroup components has quickly found itself being affected by the \\nparticular characteristics of the people involved. \\nFinal comments \\nThis chapter's purpose has been two-fold. On the one hand, it has \\nattempted to show that the study of intergroup relations and the processes \\nand structures related to identity can be particularly enhanced if the \\ninterpersonal relationships, in which the individuals are involved, are taken \\ninto account at the same time (for example, relationships between couples). \\nOn the other, it has indicated how the study of interpersonal relationships \\ncan be enhanced if we simultaneously bear in mind identity and intergroup \\naspects. We believe that an articulation of intergroup and interpersonal \\nprocesses could be more useful in the understanding of certain phenomena \\nthan considering these two aspects as antagonistic or contrary. \\nHowever, this articulation between interpersonal and intergroup aspects \\nseems relatively complex. We have made reference to some approximations \\nthat contain interesting suggestions regarding how these two components \\nshould be articulated. Yet, rather than coming up with definitive answers, \\nthese approximations only suggest paths down which one can advance, thus \\ngranting an important role to empirical research. \\nFocusing now on the case of couples, the arguments we have put forward \\nshow that a greater understanding of them could be attained if, in addition \\nto considering the specific dynamics of the relationship, other factors were \\nalso taken into account. These could include the social identity structure of \\nthose involved in a couple, as well as the content of their different social \\nidentities. Nevertheless, the study of these identities demands approxima­\\ntions which consider their relational character (Rosenberg & Gara, 1985). \\nSo, although identities of a general character exist (such as 'being a man' or \\n'being a woman') and guide the behaviour of individuals without involving \\nother people (for example, they serve the individual when deciding how to \\ndress or what hobbies to take up), there are others (such as 'being in a \\ncouple' or 'being a father') which cannot be understood without taking into \\naccount a relationship with other people (for example, his/her partner or \\nthe specific composition of the family as in the mentioned examples). \\nConsider, for example, the different conception of self felt by a 'woman \\nin a relationship' before and after her partner had been paralysed in an \\naccident. \\nIn conclusion, this approach merely connects up with an old tradition of \\nsocial psychology (Morales, 1989) which has included: the battle over \\nstructural and process aspects of identity (for example, Manford Kuhn and \\nRalph Turner versus Blumer and Goffman within symbolic interactionism); \\nthe extraordinarily discerning analysis of social and personal identity carried \\n\\n\\nSocial identity and inter\\npersonal relationships \\n1 65 \\nout by, for example, Goffman (1 968); and establishing the impossibility of \\nunderstanding 'self' or the identity, disregarding the specific social inter­\\nactions maintained by individuals. \\nNote \\nI .  The content of Figure 1 1 .2 has an illustrative purpose. The structure of the identity \\nwithin has been taken from Deaux (1993) and the content of the different identities has been \\nelaborated from the results of research carried out by Moya (1989) and Moya and Peplau (in \\npreparation). \\n\\n\\n1 2  \\nSocial Identity and Aspects of Social \\nCreativity: Shifting to New Dimensions of \\nIntergroup Comparison \\nSteve Hinkle, Laurie A. Taylor, Lee Fox-Cardamone and \\nPamela G. Ely \\nMost accounts of social identity theory (SIT) outline three response patterns \\nwhich groups and group members can make to unfavourable social \\nidentities: individual mobility, social competition and social creativity (see \\nHogg & Abrams, 1988; Tajfel & Turner, 1986). Perhaps the earliest dis­\\ncussion of social creativity can be found in two of Tajfel's 1978 publications \\nconcerning group differentiation (l 978d) and the social psychology of \\nminorities (l978e). In these works, Tajfel discussed low status or minority \\ngroups' response to their unsatisfactory social identities. When social cir­\\ncumstances supported leaving the ingroup, social mobility or moving on to \\nanother, presumably high-status, ingroup was proposed. When mobility was \\nnot sociologically or psychologically likely, Tajfel proposed three \\npossi \\nbili \\nties. \\n(a) to become, through action and reinterpretation of group character­\\nistics, more like the superior group; \\n(b) to reinterpret the existing inferior characteristics of the group, so that \\nthey do not appear as inferior but acquire a positively valued distinc­\\ntiveness from the superior group; \\n(c) to create, through social action and or diffusion of new 'ideologies' new \\ngroup characteristics which have a positively valued distinctiveness \\nfrom the superior group (Tajfel, 1978c, pp. 93-94). \\nLater, this list was elaborated to include a fourth possibility: \\n(d) group members may also seek positive distinctiveness by 'changing the \\nout-group (or selecting the out-group) with which the in-group is \\ncompared - in particular, ceasing or avoiding to use the high-status \\nout-group as a comparative frame of reference' (Tajfel & Turner, 1979, \\np. 43). \\nIn more recent discussions of SIT, the first of these is typically referred to as \\n'social competition', while the latter three are, generally, considered as \\n\\n\\nSocial identity and aspects o\\nf\\n social creativity \\n1 67 \\naspects of 'social creativity' (see Hogg & Abrams, 1988; Tajfel & Turner, \\n1986).' \\nThe purpose of this chapter is to take a detailed look at one of these \\naspects of social creativity: the establishment of new group characteristics or \\ndimensions of intergroup comparison as a means to institute positive \\ndistinctiveness for the ingroup. First, theoretical issues pertinent to this \\naspect of social creativity will be summarized. The chapter will then review \\npast research pertinent to establishing positive ingroup distinctiveness \\nthrough the use of new dimensions for intergroup comparison. New, as yet \\nunpublished, research we have conducted with several other colleagues will \\nthen be presented. The chapter will conclude with some evaluative com­\\nments on this aspect of SIT and SIT research along with discussion of \\napplications and issues for future research. \\nSome theoretical issues \\nLegitimacy and stability \\nAt the core of many SIT predictions are the notions of status legitimacy and \\nstability. The relationships between groups can be perceived as legitimate or \\nillegitimate, stable or instable (Tajfel 1978a). In general, Tajfel saw legiti­\\nmacy and stability as covarying. That is, according to SIT, legitimate and \\nstable intergroup relations tend to co-occur as do illegitimate and instable \\nintergroup relations (Tajfel, 1978a; for contrasting perspective, see Turner & \\nBrown, 1978, who treat legitimacy and stability as independent aspects of \\nthe intergroup situation). SIT goes on to associate social competition with \\nillegitimate/instable intergroup relations. Hence, the social creativity \\nstrategies, including establishing positive ingroup distinctiveness on new \\ndimensions of intergroup comparison, are postulated by SIT as being most \\nrelevant to legitimate and stable intergroup contexts (Tajfel, 1978b). Here, \\nwhere groups with threatened or negative social identities see the status quo \\nas justified and/or not likely to change, direct social competition to alter the \\nstatus quo on prevailing comparison dimensions is not apt to be viewed as \\nlikely to succeed or justifiable. Hence, the creation of new comparison \\ndimensions as a path to positive social identity is a strategy of obvious \\nmotivational significance. \\nGroup status \\nAnother theoretical issue posited as relevant to the use of new intergroup \\ncomparison dimensions is group status. While SIT is not as clear as it could \\nbe, it appears as though Tajfel believed that the use of new comparison \\ndimensions was most likely for low-status groups in legitimate/stable status \\ncircumstances, but most likely for high status groups when they perceive \\ntheir advantaged position as illegitimate and/or instable (Tajfel, 1978c). \\nHowever, not all of the early SIT theorizing is consistent on this. For \\n\\n\\n1 68 \\nSocial identity: international perspectives \\nexample, Caddick (1982) sees illegitimacy and instability as instigating the \\nsearch for new dimensions of distinctiveness for both high- and low-status \\ngroups. It also appears that, irrespective of illegitimacy and instability, early \\nSIT research viewed the use of new comparison dimensions as being more \\ncharacteristic of low- than of high-status groups (Brown & Ross, 1982). \\nSelection o\\nf\\n new comparison dimensions \\nPerhaps one of the most useful and detailed early theoretical discussions in \\nSIT of the use of new dimensions for intergroup comparison is that of van \\nKnippenberg (1978). In particular, he offers some valuable ideas regarding \\nthe selection of new comparison dimensions. Assuming legitimacy and \\nstability, existing status differences will continue and be difficult to deny. As \\na result, van Knippenberg proposed that new comparison dimensions must \\nbe capable of enhancing a group's social identity, but without affecting \\nalready existing status dimensions. That is, in selecting new comparison \\ndimensions to effect a more favourable social identity, the new dimensions \\nwill afford the ingroup positive distinctiveness while being uncorrelated with \\nthe existing status dimension where the ingroup does not compare \\nfavourably. \\nLemaine, Kastersztein and Personnaz's (1978) position is somewhat \\ndifferent from those of Tajfel (1978c) and van Knippenberg (1978). The \\nlatter two authors take no position regarding the relative importance of new \\nversus old dimensions of comparison. In contrast, Lemaine et al. clearly \\nhold that groups will strive to make new bases for intergroup comparisons \\ndominant. For Lemaine et a!., groups will not be content to simply \\nacknowledge their own inferiority on existing status dimensions while \\nclaiming superiority on a new status dimension. Rather, the ingroup will go \\nfurther and strive to make the new basis for comparison of greater \\nimportance and, thus, the dominant basis for status comparisons. In a later \\npaper, van Knippenberg (1984) modifies his earlier stance in a manner \\nconsistent with the views of Lemaine et a!., at least with respect to the \\nclassic conflictual intergroup situations emphasized by SIT. He also pro­\\nposes socially cooperative intergroup relations where groups coexist in a \\nclimate of mutual appreciation. Here, van Knippenberg (1 984) proposes \\nthat groups can and will acknowledge both their own superiority and \\ninferiority relative to other groups on various comparison dimensions and \\nthat all such dimensions can be valued by all groups. He does, however, \\nnote that ingroups and outgroups may weight the relative importance of the \\nsundry dimensions differently. \\nNew comparison dimensions versus other responses to un\\nf\\navourable \\nidentities \\nAnother important theoretical issue concerns the prepotency of turning \\nto new bases for intergroup comparison relative to other responses to \\nunfavourable identities such as mobility, social competition and other social \\n\\n\\nSocial identity and aspects o\\nf\\n social creativity \\n1 69 \\ncreativity strategies, including shifting the locus of comparison to new \\noutgroups and re-evaluating the valence of existing, unfavourable com­\\nparisons. Tajfel (1978c) clearly sees mobility as the strategy of choice in \\ncircumstances where mobility is viewed as possible and desirable (that is, \\nwith a social mobility ideology; Hogg & Abrams, 1988; Tajfel & Turner, \\n1979). However, there is little in his writings to clearly order social com­\\npetition and social creativity into a hierarchy of strategic choice. Certainly, \\nhe says virtually nothing concerning the relative strength of the various \\nforms of social creativity. In contrast, Ball, Giles and Hewstone (1984) \\npropose that mobility attempts will precede social creativity strategies, \\nwhich, in tum, will take place prior to social competition. However, Ball et \\nal. say nothing about the ordering of the various forms of social creativity. \\nIn general, the lack of theory regarding use of the various social creativity \\nstrategies is an unfortunate omission in SIT. \\nProcess issues \\nFinally, what does SIT say relevant to the process of creating new bases for \\npositive intergroup distinctiveness? First, group members must somehow \\ndevelop a shared perspective on the attributes serving as new bases for \\nintergroup comparisons and the ingroup's standings on these new attributes \\nmust, in fact, be positive (Tajfel, 1978c). The next step involves persuading \\nthe outgroup to accept the new bases for intergroup comparison and the \\ningroup's positive standing on these attributes. That is, consensual accept­\\nance by both the in- and outgroup is necessary for the new dimensions of \\nintergroup comparison to fully validate the ingroup's positive social identity \\n(Tajfel, 1978c). Lemaine et al. (1978) add to SIT theorizing about such \\nprocesses. They hold that the initial intragroup aspect of this process will, \\nideally, take place without the knowledge of the outgroup. Such seclusion \\nshields the ingroup seeking new bases for intergroup comparisons from \\ncriticism and helps reduce uncertainty regarding the new comparisons. \\nPast empirical research: a selected review of the literature2 \\nIn spite of the complex SIT theorizing concerning the use of new com­\\nparison dimensions as a response to negative social identities, most of the \\nrelevant empirical literature has yet to address the theoretical postulates \\noutlined above. As we shall see, the dominant characterization of this \\nliterature is simply the demonstration that groups with threatened social \\nidentities will, at times, tum to new bases for making intergroup com­\\nparisons. \\nIn all likelihood the most frequently cited empirical work is that of \\nLemaine and his colleagues (Lemaine et aI., 1978). In one study, children at \\nsummer camps participated in a building competition where some groups \\nwere relatively disadvantaged by being given inadequate materials. While \\n\\n\\n1 70 \\nSocial identity: international perspectives \\nunable to build as large or elaborate huts as other groups, disadvantaged \\ngroups began developing gardens, thereby establishing a new basis for \\nintergroup comparisons more favourable to their ingroups. \\nIn a review, Hinkle and Brown (1990) cited 12 papers, using both \\nlaboratory and field contexts with subject populations from nurses to \\nuniversity lecturers; all demonstrated that the same group when making \\nintergroup comparisons on multiple dimensions will evince in \\ngroup \\nfavouritism on some of the dimensions, outgroup favouritism on others, \\nand no favouritism at all on still other dimensions. The simultaneous \\noccurrence of ingroup and outgroup favouritism on differing dimensions is \\nconsistent with the notion of using new bases for intergroup comparisons to \\nestablish positive social identities when comparisons on other dimensions do \\nnot favour the ingroup. In addition to the studies reviewed by Hinkle and \\nBrown, other similar results are reported in Capozza, Bonaldo and \\nDiMaggio (1982). The diversity of subject populations and paradigms \\nacross these various researches points to the robustness of this form of \\nsocial creativity. \\nPerhaps one of the most interesting early SIT studies bearing on the use \\nof new dimensions for making intergroup comparisons is Brown and Ross \\n(1982). The study's primary emphases are on both social competition and \\nsocial creativity. In particular, Brown and Ross's consideration of creativity \\nemphasizes the re-evalution of the importance of various comparison \\ndimensions rather than the topic of the present chapter, the use of new \\ncomparison dimensions. Nonetheless, while the latter issue is not focal to \\nBrown and Ross, they do present some important relevant data. Before \\ndiscussing their findings, some background regarding the paper's hypotheses \\nand design is necessary. \\nBrown and Ross discuss illegitimacy and instability as inducing a 'threat \\nto identity' or insecure social identity which, in response, produces either \\nsocial competition or social creativity ('the strategies . . .  \\nwere not thought \\nto be mutually exclusive', p. 1 62). Their study's primary manipulation was \\npresented in terms of high, moderate or low threats to identity, though \\ncareful examination of their procedures indicates that the study's primary \\nindependent variable can just as easily be termed 'status illegitimacy'. A \\nsecond important independent variable in this research was a manipulation \\nof ingroup status as being either superior or inferior. One of the \\nexperiment's primary predictions was 'that increases in differentiation along \\n. . .  \\nalternative dimensions would . . .  \\nvary in proportion to the threat to \\nidentity experienced, and that these changes would be most prominent in \\nthe lower status groups' (p. 1 62). \\nWhen ingroup bias on alternative dimensions of intergroup comparison is \\ncompared before and after implementation of the status illegitimacy \\nmanipulation, the data clearly show that inferior status groups in the high \\nand moderate illegitimacy (that is, threat to identity) conditions manifested \\nsubstantially more ingroup favouritism on the alternative comparison \\ndimensions. This shift towards greater ingroup bias on the alternative \\n\\n\\nSocial identity and aspects o\\nf\\n social creativity \\n1 7 1  \\ndimensions did not occur for inferior status groups in the low illegitimacy \\ncondition, nor did it occur for superior status groups in any of the three \\nconditions of status legitimacy.3 \\nThese results are partially supportive of early SIT theorizing in that the \\nuse of new comparison dimensions to attain positive group distinctiveness is \\nmore characteristic of low- rather than high-status groups. On the other \\nhand, Brown and Ross's results contradict Tajfel's (1978c) view that this \\nand other forms of social creativity would be most prominent for low-status \\ngroups when there are legitimate/stable status relations. Similarly, their \\nfindings are also inconsistent with Caddick's (1982) prediction that the \\nvarious social creativity strategies would occur for both high- and low-status \\ngroups when the intergroup status structure was both illegitimate and \\ninstable. It appears from Brown and Ross's data that the most empirically \\ntenable theoretical position for SIT is that social creativity, at least in the \\nform of using new dimensions for making ingroup-favouring intergroup \\ncomparisons, is most likely to be observed in low-status groups under \\nconditions of status illegitimacy and/or instability (though the specific \\neffects of illegitimacy and instability as distinct variables remain to be \\ntested). \\nOther interesting data from Brown and Ross bear on the issue of the \\nvaluing of new dimensions of intergroup comparison. Recall Lemaine et \\nal.'s (1987) and van Knippenberg's (1984) hypotheses that groups shifting to \\nnew dimensions of intergroup comparison to attain positive distinctiveness \\nwould also see these new dimensions as of greater importance than previous \\nbases for intergroup comparison. In general, Brown and Ross found that \\nwhile inferior status groups were, with status illegitimacy, quick to use \\nalternative comparison dimensions to evince ingroup favouritism, there was \\nlittle evidence to support the notions that these new dimensions' importance \\nincreased or that they were viewed as being of greater importance than the \\noriginal intergroup comparison dimensions. This result is consistent with \\nsimilar findings reviewed by Hinkle and Brown (1990). Indeed, the existing \\nevidence seems, if anything, more consistent with the view that low-status \\ngroups are willing to salvage damaged or unfavourable social identities by \\nshifting to or emphasizing new in \\ngroup-favouring dimensions of com­\\nparison even while acknowledging that the latter are of lesser importance. \\nAn example of this might be the losing sports team that emphasizes its own \\nsuperiority in terms of better sportsmanship (see Lalonde, 1992) or more \\nenjoyment of the game, even while willingly owning that it is more \\nimportant to win. \\nCaddick (1982) presents similar findings with respect to illegitimate status \\ndifferences and the occurrence of ingroup favouritism on new dimensions of \\nintergroup comparison. An illegitimate status difference led to increased \\nexpression of ingroup favouritism on several comparison dimensions other \\nthan the original basis for intergroup comparison. Interestingly, and in \\ncontrast to Brown and Ross, ingroup favouritism on the new comparison \\ndimensions occurred for members of both low- and high-status groups. \\n\\n\\n1 72 \\nSocial identity: international perspectives \\nCaddick develops a thought-provoking analysis of this result based on the \\nfact that in his experimental procedure the high-status group is clearly \\nresponsible for the illegitimacy of the initial comparison between the \\ngroups. Caddick suggests that this may have compromised the high-status \\ngroup's social identity through guilt feelings induced by violation of values \\nsuch as fairness or justice. To compensate for their compromised favourable \\nsocial identities, the high-status groups adopted new bases for making \\ningroup-favouring intergroup comparisons. \\nSome new empirical findings \\nOur own investigation of social creativity issues began rather unexpectedly \\nthrough an effort to replicate findings from a study by Condor, Brown and \\nWilliams (1987). One component of Condor et al.'s experiment involved \\ngroups of unacquainted subjects working on a laboratory task to generate \\nideas for recruiting students to their university under an instructional set \\nemphasizing task issues such as efficiency and productivity. Following this \\nexercise, subjects watched a short video of another group engaged in the \\nsame endeavour and then made in- and outgroup ratings on task-oriented \\ndimensions such as the group's competence and energy and socio-emotional \\ndimensions such as the groups' friendliness and warmth. Condor et al. \\nfound much greater ingroup favouritism on the task-oriented dimensions of \\nintergroup comparison than on the socio-emotional dimensions. This was \\nexpected since the task-oriented dimensions are more closely associated \\nwith the experimental instructional set than the socio-emotional dimen­\\nsions. \\nHinkle and Crook (1 987) attempted to replicate Condor et al. as closely \\nas possible and found a complete reversal of their findings. That is, signifi­\\ncantly greater levels of ingroup favouritism occurred on the socioemotional \\ndimensions than on the task-oriented dimensions. Why? There were only \\ntwo important differences between the procedures of Condor et al. and \\nHinkle and Crook. First, the subject populations differed, with the latter \\nstudy using American university students while the former's subjects were \\nstudents from a British university. While it is virtually impossible to rule out \\ndifferences in the populations as an account for the reversal of findings, one \\nof the results from Hinkle and Crook points to the comparability of the two \\nsubjects populations. Their subjects saw the task-oriented dimensions as of \\ngreater importance than the socio-emotional dimensions. Thus, Hinkle and \\nCrook's findings appear to be a social creativity effect in that their subjects \\nare manifesting the most ingroup favouritism on a secondary, less important \\ndimension of intergroup comparison. \\nThere was a second difference in the procedures of Condor et al. and \\nHinkle and Crook. Rather than presenting the outgroup via a video, the \\nlatter simply asked subjects to think of an average group of university \\nstudents working on the same task and rate this 'imagined' outgroup. \\n\\n\\nE \\n(/) \\n'.\\\" \\n'§ \\n0 \\n> \\n† \\na. \\n:::l\\n \\ne \\nCJ) \\n.E: \\nSocial identity and aspects o\\nf\\n social creativity \\n35 \\n30 \\n25 \\n20 \\n1 5  \\n1 0  \\n5 \\n0 \\n-5 \\nBelow \\naverage \\nComparison dimension \\n• Task-oriented \\no Socio-emotional \\nAverage \\nOutgroup status \\nWell above \\naverage \\nFigure 12. 1 \\n/ngroup f\\navouritism on dif\\nf\\nerent comparison dimensions as a \\nf\\nunction of outgroup status (Hinkle, Fox-Cardamone, Taylor & Crook, \\n/988) \\n1 73 \\nAssuming a rather high level of achievement motivation amongst these \\nuniversity student subjects, it may be that their vision of an 'average' group \\nis of one whose performance does indeed threaten that of their own group. \\nThis interpretation corroborates the view that Hinkle and Crook's subjects \\nare engaging in a social creativity process. If the outgroup's task perform­\\nance is imagined to be quite good relative to ingroup performance, subjects' \\nshift to the less important socio-emotional dimensions to manifest ingroup \\nfavouritism may be seen as an effort to offset the unfavourable intergroup \\ncomparison on the task-oriented dimensions. \\nExperimental demonstrations o\\nf\\n social creativity e\\nf\\nf\\nects \\nTo follow up on the interpretation outlined above, Hinkle, Fox-Cardamone, \\nTaylor and Crook (1988) replicated Hinkle and Crook's experiment and \\nextended it by adding two additional experimental conditions. While some \\ngroups of subjects were asked to think of an average group of university \\nstudents as the outgroup stimulus, others were asked to think of below \\naverage or well-above average groups. It was predicted that below average \\noutgroups would not threaten the distinctiveness of the ingroup on the \\nimportant, task-oriented dimensions. Hence, subjects in this condition were \\nexpected to manifest greater ingroup favouritism on the task-oriented \\ndimensions than the socio-emotional dimensions. In contrast, the prediction \\nfor the average and well-above average outgroup conditions was that such \\noutgroups would threaten the ingroup's distinctiveness on the important task­\\noriented dimensions. As a result, social creativity would lead to greater \\ningroup favouritism on the secondary dimensions of intergroup comparison, \\nthat is, the socio-emotional dimensions. As can be seen in Figure 12. 1 ,  this is \\nexactly what was found. \\n\\n\\n1 74 \\nSocial identity: international pers\\npectives \\nInterestingly, Hinkle et al. collected data on the perceived importance of \\nthe task and socio-emotional comparison dimension a\\nf\\nter subjects had made \\ntheir ingroup and outgroup ratings. There was no evidence for any change in \\nthe importance of the dimensions. The task-oriented dimensions were still \\nseen as being significantly more important than the socio-emotional \\ndimensions in all three experimental conditions. Thus, the social creativity \\nseen in the average and well-above average outgroup conditions does not \\nappear to actually compensate for unfavourable comparisons on the task­\\noriented dimensions. In these conditions subjects are not saying, 'Your group \\nis better at the task, but it is more important to get along well and we are \\nbetter at that.' Rather, they appear to be saying, 'While your group is better \\nthan ours at the task, at least we get along better.' The social creativity seen \\nin this study does not appear to compensate for an unfavourable social \\nidentity, but simply to take a step towards salvaging the ingroup's social \\nidentity. As with earlier research reviewed above, this result is inconsistent \\nwith Lemain et al.'s (1978) and van Knippenberg's (1984) views that the shift \\nto new, ingroup-favouring dimensions of intergroup comparison would \\nnecessarily be accompanied by perceiving the new dimensions as of greater \\nvalue or importance than those upon which the ingroup did not compare \\nfavourably. \\nThe Hinkle et al. research can be criticized for the use of an imagined \\nmanipulation of outgroup status. Recently, two of us (Laurie Taylor and \\nSteve Hinkle) conducted an experiment similar to those reported above, but \\nincorporating a more concrete manipulation of outgroup status. Again, \\ngroups worked under a task-emphasis instructional set to generate ideas for \\nthe recruitment of students to their university, and the ingroup and out­\\ngroup were rated on both task-oriented and socio-emotional dimensions of \\nintergroup comparison. However, rather than asking students to imagine \\noutgroups of varying status levels, outgroup status was manipulated by \\nproviding subjects with a list of some of the recruitment ideas generated by \\nthe other group. These lists were prepared in light of pilot test data such \\nthat they contained only the lowest (below average outgroup status) or \\nhighest (above average outgroup status) rated ideas from the pilot study. In \\nthe below average outgroup condition, ingroup favouritism was greater on \\nthe task-oriented dimensions than on the socio-emotional dimensions. In the \\nabove average outgroup condition, a social creativity effect was again \\nobserved: greater ingroup favouritism on the socio-emotional comparison \\ndimensions than on the task-oriented dimensions (see Figure 12.2). \\nConclusions, applications and directions for future theory and \\nresearch \\nGeneral conclusions \\nThe strongest and most consistent finding across numerous studies using \\nvarious populations and paradigms is that social creativity in the form of \\n\\n\\nE \\nIf) \\nok \\n0;::\\n \\n::J \\n0 \\n> \\n† \\nQ. \\n::J \\n2 \\nCJ) \\n.E: \\nSocial identity and aspects o\\nf\\n social creativity \\n30 \\n25 \\n20 \\n1 5  \\n1 0  \\n5 \\n0 \\n_\\n_\\n_\\n_ \\n--j Comparison dimension \\nBelow average \\n• Task-oriented \\nD Socio-emotional \\nAbove average \\nOutgroup status \\nFigure 12.2 Ingroup f\\navouritism on dif\\nf\\nerent comparison dimensions as a \\nf\\nunction of outgroup status (unpublished data f\\nrom Taylor & Hinkle) \\n1 75 \\nresponding to a negative social identity with ingroup favouritism on new \\ncomparison dimensions is a reliable and robust phenomenon. Furthermore, \\nthe experimental investigations of Hinkle et al. and Taylor and Hinkle \\ndemonstrate that low ingroup status causes greater ingroup favouritism on \\nnew comparison dimensions than on the original dimensions upon which \\nthe ingroup does not fare well. \\nStatus, legitimacy and stability \\nPast this broad conclusion, the existing research findings bearing on the \\ntheoretical issues raised at the outset of this chapter are limited. The issue of \\ngroup status in conjunction with its legitimacy and stability has received the \\nmost attention and the early SIT theoretical formulations have received \\nmixed support. The evidence from studies such as Brown and Ross (1982) \\nand Caddick (1982) supports the notion that low-status groups are more apt \\nto manifest group favouritism on alternative dimensions of intergroup \\ncomparison than high-status groups. How legitimacy and stability enter the \\ntheoretical equation has been far less clear. Caddick's results indicate that \\nillegitimacy increases ingroup-favouring comparisons on new dimensions \\nfor both high- and low-status groups. Brown and Ross found this to be the \\ncase for low-status groups only. Clearly, other variables must be moderating \\nthe relationship between illegitimacy, group status and the occurrence of \\ningroup favouritism on new comparison dimensions. Relevant theory and, \\nmore importantly, research are needed. Caddick's speculation concerning \\nthe responsibility of a high-status group for the existence of an illegitimate \\nintergroup status structure appears to be a potential moderator worthy of \\nfurther empirical attention. \\nAt present, there appears to be no support for Tajfel's ( I978c) early \\nproposal that low-status groups would be particularly likely to use new \\n\\n\\n1 76 \\nSocial identity: international perspectives \\ndimensions of intergroup comparison when status relations are legitimate \\nand stable. Perhaps with a legitimate/stable status structure it is very \\nunlikely for low-status group members to perceive any change in the status \\nquo relative to superior groups, including the use of new comparison \\ndimensions. It seems plausible that another social creativity strategy, a shift \\nto different comparison outgroups, may be the more likely response to an \\nunfavourable social identity under a legitimate/stable status structure. A \\nsuperordinate issue to any SIT discussion regarding legitimacy and stability \\nis the need for more thoroughly elaborated theorizing and research on the \\nrelationships between legitimacy and stability. \\nImportance o\\nf\\n comparison dimensions \\nThe other theoretical issue which has received some empirical attention has \\nto do with the importance or value placed on new intergroup comparison \\ndimensions. Recall that both Lemaine et al. (1978) and van Knippenberg \\n(1984) hypothesized that a shift to ingroup favouritism on new intergroup \\ncomparison dimensions would be accompanied by the new bases of com­\\nparison acquiring greater value or importance than the original comparison \\ndimensions. The intuitive strength of this postulate makes it quite difficult to \\nabandon. However, there is substantial evidence to the contrary. Groups \\nappear to frequently evince ingroup favouritism on new, less valued com­\\nparison dimensions while acknowledging that they do not measure up to the \\noutgroup in other, apparently more important, comparisons. This is an \\nintriguing result calling out for more research to further illuminate this \\nfinding and perhaps identify moderator variables specifying when new \\ncomparison dimensions will or will not be seen as being of greater import­\\nance than the initial comparison dimensions. \\nVan Knippenberg's (1984) discussion of socially cooperative intergroup \\nrelations in contrast to SIT's traditionally competitive intergroup relations \\nseems to offer a good deal of theoretical promise with respect to this as it \\nseems plausible that the psychological importance of placing greater value \\non new intergroup comparison dimensions may only be relevant to com­\\npetitively oriented intergroup relations. \\nSimilarly, Hinkle and Brown's (1990) distinction between autonomous \\nand relational processes of social identity construction may also be theor­\\netically relevant (see also Brown, Hinkle, Ely, Fox-Cardamone, Maras & \\nTaylor, 1992; Hinkle, Brown & Ely, 1992). Groups with autonomous \\nidentity construction orientations may establish entirely satisfactory social \\nidentities through seeing themselves favourably on relevant comparison \\ndimensions independent of whether other groups rank below or above them. \\nAs a result, the relative value or importance of various comparison dimen­\\nsions may not be psychologically relevant with the autonomous orientation. \\nAnother potentially useful approach to developing more research and \\ntheory on the importance of different comparison dimensions takes a \\ntemporal perspective. It seems plausible that the immediate response to an \\n\\n\\nSocial identity and aspects o\\nf\\n social creativity \\n1 77 \\nunfavourable intergroup comparison is, to use a term introduced earlier, \\nto salvage a positive social identity. If our research group fares disas­\\ntrously in the latest government league tables of research quantity and \\nquality, the first step is to establish a toehold towards re-establishing a \\nfavourable social identity. One way to do this is by claiming high status \\nfor the ingroup on any dimension, even an inferior one. 'Yes, the quality \\nand quantity of our group's research is not on par with others, but we are \\ndoing some limited, innovative work and our teaching is very good. And, \\nyes, we recognize that research innovation and teaching quality really are \\nnot the sine qua non for research groups.' This salvages the group's esteem \\nin the short term and provides the foundation for subsequent develop­\\nments such as actually improving the group's research productivity or \\nconvincing relevant entities in the social structure that research innovation \\nand teaching quality ought to be the more important bases for intergroup \\ncomparIson. \\nThe processes o\\nf\\n social creativity \\nThere is, to our knowledge, no research on Tajfel's ( l978c) and Lemaine et \\nal.'s (1978) proposals that selection of new comparison dimensions proceeds \\nfirst as an intragroup process and, according to Lemaine et aI., surrep­\\ntitiously and, then, as an intergroup process where the ingroup attempts to \\npersuade the outgroup regarding the importance of the new comparison \\ndimension and the ingroup's superior status. The comments in the previous \\nsection on temporal aspects of social identity processes again seems \\nrelevant. It may be that groups with unfavourable identities must first take \\nsteps to salvage positive distinctiveness, then work within the group to \\ndevelop genuinely favourable intergroup comparisons, and, finally, shift to \\nthe issue of persuading other groups as to the validity and importance of the \\nnew bases of comparison. Research paradigms without a longitudinal \\ncomponent are unlikely to detect or shed light on such processes. Unfor­\\ntunately, longitudinally oriented studies are virtually non-existent in the SIT \\ntradition. \\nHinkle and Brown's (1990) notion of autonomous orientations may also \\nbe pertinent. For autonomous groups, since they are interested in the \\ningroup in comparison to abstract standards rather than in comparison to \\nother groups, the need to persuade external audiences of the validity of new \\nintergroup comparisons is apt to be irrelevant. \\nAn interesting and closely related issue is raised by van Knippenberg \\n(\\\\ 984), who argues that responses to unfavourable social identities are not \\nalways directed towards immediate enhancement of the group's social iden­\\ntity. Specifically, van Knippenberg discusses the role of self-presentational \\nprocesses in the expression of ingroup or outgroup favouritism. Self­\\npresentational issues seem equally relevant to social creativity processes \\nincluding the selection of new bases for making intergroup comparisons. \\nAbrams and Hogg (1988; also Hogg & Abrams, 1990) have outlined similar \\n\\n\\n1 78 \\nSocial identity: international pers\\npectives \\nideas which go well beyond social identity and self-presentation as motiva­\\ntional issues relevant to the expression of ingroup or outgroup favouritism. \\nStrategy choice in response to unf\\navourable social identities \\nWhile there are a few studies on mobility versus either social competition or \\nsocial creativity (or both, for example, Ellemers, Wilke & van Knippenberg, \\n1993; Jackson, 1993; Wright, Taylor & Moghaddam, 1990), there is no \\nresearch and virtually no relevant theory on selection of the various social \\ncreativity strategies. These are important theoretical issues. SIT's inability \\nto predict the specific form of response to an unfavourable social identity is \\na significant omission and diminution of its utility. While there may be good \\nhistorical reasons for why these issues were not well addressed in the early \\nyears of SIT's development, the contemporary status of SIT would be \\nenhanced by progress on this front. \\nPotential applications \\nWe are unaware of any genuinely applied literature on the use of new \\nintergroup comparison dimensions. However, existing studies and theor­\\netical discussions are rife with possibility. Perhaps the richest potential for \\napplication resides in issues of group morale, cohesion and motivation. \\nHow might poorly performing sports teams, business organizations and \\nother similar entities deal with attendance problems of low group morale \\nand motivation? The generation of new, favourable dimensions of inter­\\ngroup comparison offers one strategy. While the existing research indicates \\nthat a losing team's development of the view that they enjoy what they are \\ndoing more than their competitors does not fully psychologically \\ncompensate for its poor won-lost record, which such a team is still apt \\nto see of premier importance, it still may serve as a starting point for \\nrebuilding team spirit and motivation. While the theoretical basis for this \\nhas not yet been tested, the hypothesis that social creativity results in \\nenhanced cohesiveness and, subsequently, better group performance is an \\nintriguing one. Were notions such as this begun to be examined in applied \\ncontexts, other interesting issues would certainly come quickly to the fore. \\nFor example, in all previous research of this area, the creation or ingroup­\\nfavouring use of new comparison dimensions has always arisen spon­\\ntaneously from the studies' inherent intragroup dynamics. In contemplating \\napplications, it is quite likely that the instigation to the use of new \\ndimensions of intergroup comparison would result from the intervention of \\na group's leadership such as managers, administrators, coaches, and so on, \\nor outside sources such as consultants. There is little if any helpful existing \\nresearch or theory. Links to work in organizational psychology and the \\nleadership literature would likely be beneficial in determining whether or \\nnot leadership efforts to stimulate social creativity would be effective or how \\nsuch endeavours could be best implemented. \\n\\n\\nSocial identity and aspects o\\nf\\n social creativity \\n1 79 \\nFinal comments \\nIn conclusion, research on social creatIVIty in the fonn of selecting new \\ncomparison dimensions for making ingroup-favouring evaluations has \\nplayed a small, but important, role in SIT research, a role which appears to \\nbe increasing in its significance and potential applications. The robustness of \\nthis fonn of social creativity is well documented and some important com­\\nplexities have begun to be addressed. The existing empirical findings provide \\nsupport for certain SIT precepts, but not others. Our own view on this is \\nthat in instances where the literature has not supported SIT it has served the \\nvaluable function of precipitating modifications of the theory that have \\nenhanced its power and viability. However, as should be clear fonn earlier \\ndiscussion in this chapter, there are many interesting issues still in need of \\nboth conceptual and empirical attention. SIT can only be enriched as these \\nissues stimulate further work in the study of intergroup processes. \\nNotes \\n1 .  Note that other authors label or group responses to unfavourable social identities \\nsomewhat differently (for example, Taylor & Moghaddam, 1987). In preparing the present \\nchapter, the authors have endeavoured to follow as closely as possible the conceptual \\norganization offered in Tajf\\nel's early statements of SIT. \\n2. Turner and Brown (1978) is a frequently cited paper on legitimacy and stability and their \\nrelationship to social competition and social creativity. However, in the present context, their \\nmeasures are of limited relevance. The creativity indices concern the number of new possible \\ncomparison dimensions generated by the SUbjects. However, subjects did not actually make any \\ningroup or outgroup ratings on these new dimensions. Hence, the study's results do not bear \\ndirectly on the use of new bases of comparison to establish favourable social identities. \\n3. Note that Brown and Ross themselves do not emphasize this pattern in their results. In \\nfact, they do not discuss it at all. However, the described pattern is very clear from examination \\nof the data in their Tables 2 (p. 168) and 5 (p. 1 7 1 ). \\n\\n\\n1 3  \\nMere Categorization and Competition: A \\nCloser Look at Social Identity Theory and \\nthe Discontinuity Effect \\nStephen M. Drigotas, Chester Insko and John Schopfer \\nThe word 'group' can have many meanings. It can be used to represent \\nanything from face-to-face interaction to aggregates of unacquainted \\npeople. Its imprecision is further enhanced by the fact that one may be \\nassigned to many groups (based on sex, race, nationality, family, and so on) \\nwithout any feelings of belonging or particular affinity for any one of them. \\nBecause of this apparent amorphousness, researchers through the years have \\nattempted to delineate when an aggregate of individuals becomes a group. \\nHorwitz and Rabbie (1989) credit Lewin (1948) for early theoretical interest \\nin perceived collectivity as the defining nature of groupness. Lewin main­\\ntained that interdependence of fate is the determinant of who belongs to \\nwhich group, not simply degree of similarity-dissimilarity of features. \\nCampbell (1958) called the shift from perceived aggregate to perceived \\ncollectivity 'entitativity' and suggested various Gestalt mechanisms for the \\nperception of many as one. At this point we feel that it would be premature \\nto settle on what precisely is meant by the term 'group'. \\nThe behaviour of individuals within groups can be as diverse as the \\nconsiderations of what constitutes a group. Historically some social scien­\\ntists have focused on the supposedly animalistic nature of crowd behaviour \\nas providing possible evidence for a group mind unfettered by moral \\nconstraints (cf. Allport, 1924; Durkheim, 1 898; LeBon, 1 896; McDougall, \\n1920). On the other hand, after extensive observation McPhail (1991) argues \\nthat the notion of 'madding' crowd is a 'myth'. Still within contemporary \\nsocial psychology there is evidence that under some circumstances \\nindividuals within groups behave in a manner that is less than praiseworthy. \\nFor example, individuals within groups are more likely socially to loaf \\n(Latane, Williams & Harkins, 1979), give more shocks to a 'learner' \\n(Zimbardo, 1969), and not help in an emergency situation (Latane & \\nDarley, 1970). \\nOur approach to the general problem of group behaviour has been \\nspecifically to compare interindividual behaviour with intergroup behaviour. \\nConsistent with Roger Brown's (1954) suggestion of a discontinuity between \\nindividual and group behaviour, we have adopted the term 'discontinuity' \\n\\n\\nMere categorization and competition \\n181 \\nfor the tendency of intergroup behaviour to be more competitive, or less \\ncooperative, than interindividual behaviour. Within the context of certain \\nmatrix games there is now an abundance of evidence for a descriptively \\nlarge discontinuity effect (for a review see Schopler & Insko, 1992). What is \\nless clear is the reason for the discontinuity effect. \\nI While different theorists \\nhave suggested divergent causes for the competitiveness of groups, the \\npurpose of the present chapter is to compare and contrast two of the major \\napproaches to understanding the phenomenon: social identity theory \\n(Tajfel, 1970b, 1978a, 1978b, 1982c; Tajfel & Turner, 1979, 1986; Turner, \\n198 1  \\nb) and the research tradition which has specifically investigated the \\ndiscontinuity effect (Insko & Schopler, 1987; Insko, Pinkley, Harring, et aI., \\n1987; Insko, Pinkley, Hoyle, et aI., 1987; Insko, Schopler, Hoyle, Dardis & \\nGraetz, 1990; Insko et ai., 1988; Insko et aI., 1992, 1993; McCallum et ai., \\n1985; Schopler & Insko, 1992; Schopler, Insko, Graetz, Drigotas & Smith, \\n199 1 ;  Schopler et aI., 1993). \\nSocial identity theory \\nAn overview \\nTajfel's social identity theory (SIT) is based upon the assumption that we all \\nhave a need for favourable self-identity, and that the status of the groups to \\nwhich we belong helps us achieve such positive self-identity. Moreover, \\nTajfel and his associates claim that such positive evaluations are made on a \\ncomparative basis - we nourish our positive self-identities when our group \\ndoes better than another group (Tajfel, 1978b, 198 1 ,  1982c; Tajfel & Turner, \\n1979, 1986; Turner, 1981b). Thus, intergroup competitiveness is directly \\nlinked to the maintenance of self-identity. The SIT theorists also claim that \\nmere categorization of subjects into 'groups' results in the orientation \\ntowards own-category favouritism. An abundance of research (Billig & \\nTajfel, 1973; Tajfel & Billig, 1974; see Brewer, 1979 and Diehl, 1 990, for \\nreviews) has demonstrated that when subjects are categorized into one of two \\nmutually exclusive groups and respond to a non-correspondent allocation \\ntask, they typically make choices favouring an own-group member over an \\noutgroup member - ostensibly in the pursuit of bolstering self-identity \\nthrough the relative advantage for own group. \\nA discontinuity perspective on social identity theory \\nA serious problem inherent in social identity theory is its inability \\ntheoretically to account for the diff\\nerences between intergroup interaction \\nand interindividual interaction. Research involving the discontinuity effect \\nhas long demonstrated that group interactions can in many circumstances \\nbe more competitive than individual interactions. However, as noted \\nelsewhere (Insko & Schopler, 1 987; Schopler & Insko, 1992), individuals \\nshould also receive positive bolstering of self-esteem from doing better than \\n\\n\\n182 \\nSocial identity: international pers\\npectives \\nanother individual. But if this is true, why from a social identity theory \\nperspective should individuals not be just as competitive as groups? It is true \\nthat social identity theory was developed to address intergroup relations. \\nThe point is, however, that there is nothing in social identity theory itself \\nthat indicates why the processes inherent in two-on-two relations should not \\nalso be present in one-on-one relations. If besting another group should \\nbolster one's self-esteem and one's self-definition, the same should also be \\ntrue of besting another individual. Why then, from a social identity per­\\nspective, are not interindividual relations just as competitive as intergroup \\nrelations? Furthermore, if it is true that individuals receive a self-definitional \\nboost from besting another individual, by implication social identity theory \\ndoes not adequately account for between-group competitiveness. Social \\nidentity theory's self-definitional premise is plausible, but does not by itself \\naccount for between-group competitiveness. The theory is not so much \\nimplausible as incomplete. \\nThe minimal group paradigm \\nInvestigations of social identity theory's prediction that mere categorization \\nis sufficient to evoke relativistic social comparisons have used a minimal \\ngroup paradigm in which subjects are categorized and then asked to allocate \\noutcomes to own- and other-category members. The most common tech­\\nnique for categorization involves dividing subjects into sets, or groups, on the \\nbasis of their fabricated preferences for the paintings of Klee versus those of \\nKandinksy. Subjects are asked to rate pairs of Klee-Kandinsky slides and are \\nthen assigned to different rooms based upon their supposed preferences. \\nOnce in the separate rooms subjects are asked individually to select one \\ncolumn of payoff points from a matrix, and thus assign a payoff amount to \\nan own-group member and an other-group member. While Tajfel and his \\nassociates use a somewhat complicated 'pull' procedure for scoring (see \\nTurner, 1978a, for a complete description of the assessment procedure), only \\nthe four resultant orientations are discussed here. They include (a) maximum \\ningroup profit (MIP), (b) maximum difference in favour of the ingroup \\n(MD), (c) fairness (F) and (d) maximum joint profit (MJP). It is the \\nmaximum difference choice that represents the greatest relative advantage \\nand best exemplifies the need for ingroup favouritism. Their research has \\ngenerally concluded that MD is the most powerful predictor of choice, \\nalthough evidence for the other four choices does exist. In fact, Turner, \\nBrown and Tajfel (1979) summarized their results as indicating that the MD \\nchoice is only non-significantly more influential than MIP and MJP. \\nNevertheless, the SIT theorists hold strong to their contention that between­\\ngroup competitiveness is the result of the need for positive social identity \\nthrough the relative advantage of ingroup performance. \\nAs indicated below, there is a question whether or not mere categoriza­\\ntion is sufficient to produce relativistic social comparison. Beyond that there \\nis a definitional matter as to whether aggregates of individuals should be \\n\\n\\nMere categorization and competition \\n1 83 \\ndefined as groups to the extent that they engage in relativistic social \\ncomparisons. One could interpret Tajfel and his associates as believing, first, \\nthat entitativity is created (on the independent variable side) by categor­\\nization, and, second, that entitativity is indexed (on the dependent variable \\nside) by relativistic social comparisons. While Tajfel and his associates do \\nnot use the term 'entitativity', they do believe that mere categorization does \\ninduce 'genuine awareness of membership in separate and distinct groups' \\n(Tajfel, 1 978a, p. 35). \\nThe alternative allocation task experiments \\nA series of experiments using the Multiple Alternative Matrices (MAMs) \\nwere conducted during the early eighties to assess directly the maximizing \\ndifference orientation exhibited within the minimal group paradigm. \\nBornstein et al. (1 983a) developed the procedure which enables the meas­\\nurement of seven different social orientations. The labels for the orientations, \\ntaken from the analyses of Kelley and Thibaut (1978), are differentiated \\nbased upon whether the orientations favour own group, other group or \\nneither group. The orientations favouring own group include maximizing \\nown gain (max. own), maximizing relative own gain (max. reI. own), and \\nmaximizing joint gain favouring own group (max. joint own). The orien­\\ntations favouring the other group include maximizing other's gain (max. \\nother), maximizing other's relative gain (max. reI. other), and maximizing \\njoint gain favouring other (max. joint other). Finally, the orientation \\nfavouring neither group minimizes the difference between group outcomes \\n(min. diff.). \\nThe obvious advantage of this approach is its delineation of interpersonal \\nmotives. Each orientation represents a mutually exclusive motivation for \\ninteraction. Thus, the overlapping orientations problem inherent in the \\nminimal group paradigm is easily resolved through use of the MAMs and a \\nbetter test of SIT is possible. The max. reI. own choice best represents SIT's \\nmaximum difference (MD), max. own is similar to maximum ingroup profit \\n(MIP), min. diff. is similar to fairness (F), and max. joint best represents \\nmaximum joint profit (MJP). \\nPoint values for own versus other group can be assigned to represent the \\nseven orientations without motivational overlap. In other words, the rep­\\nresentations of the seven orientations are mutually exclusive as to the \\nmotivation for choosing them. Therefore, the specific predictions of SIT \\nconcerning categorization and self-esteem gained through maximizing \\nrelative advantage could be tested in an unconfounded manner through the \\nuse of the Multiple Alternative Matrices. \\nA series of studies using the MAMs was conducted to test the viability of \\nSIT in predicting actual motivational behaviour in the minimal group \\nparadigm. Because the details of these studies have been presented in \\nprevious articles and a chapter (Insko & Schopler, 1 987), only the highlights \\nare presented here. \\n\\n\\n1 84 \\nSocial identity: international pers\\npectives \\nBornstein et al. (1983a) conducted a series of three experiments using the \\nMAMs that represented a substantial challenge to social identity theory. In \\nthe first two studies, college students were given a cover story concerning \\ntheir participation in an experiment as part of a larger investigation of \\nartistic preferences. The rest of the experiment followed the Tajfelian pro­\\ncedure exactly except that subjects made monetary choices for one own­\\ngroup and outgroup member using the multiple alternative matrices. The \\nresults indicated that the subjects were not choosing max. reI. own more \\nthan would be expected by chance (across the seven orientations and 10 \\ntrials), and they actually chose it significantly less than would be expected \\nby chance. Recall that max. reI. own is analogous to MD, the main strategy \\nbelieved to be evoked by mere categorization. Instead, subjects significantly \\ntended to choose both max. joint own and min. diff. (with significantly \\nmore min. diff. for females than males). \\nIn the second experiment subjects were also asked to go back through the \\nmatrices and indicate a second preference on each matrix, and then a third \\npreference. Subjects did not select max. reI. above chance for either the \\nsecond or third preferences. Furthermore, the second experiment also \\nincluded a condition in which subjects responded using the Tajfel matrices \\ninstead of the MAMs. Results using these matrices indicated support for the \\nMD choice, as in previous studies using the less differentiated matrices. \\nThe third experiment was conducted using junior high students (9th and \\n10th grade) from a middle-class US suburb. No cover story was used. \\nResults again indicated no significant deviation from chance for max. reI. \\nown. As in the first two experiments, min. diff. was selected significantly \\nmore than chance. However, unlike in the first two experiments, max. own \\nwas also selected significantly more than chance. \\nGiven the slightly different results for junior high students, Wittenbraker \\n(1983) conducted an experiment using 5th, 9th and 12th grade students \\nfrom the same school system used in the previously mentioned third study. \\nHe used the same mere categorization procedure except that coloured bar \\ngraphs were used to represent outcomes for Klee versus Kandinsky \\nmembers. Once again, results indicated a significant effect for min. diff. \\nThere was also a replication of the max. own effect for 9th grade males. \\nFinally, there was a significant effect for max. joint own for both 5th and \\n12th graders. \\n2 \\nSocial identity theory's response \\nIn response to the Bornstein et al. series of experiments, Turner (1983a, \\n1 983b) raised three objections. The first two were procedural in nature while \\nthe third constituted a major theoretical challenge. The two procedural \\nobjections were that the cover story distorted the results and that in the \\nMAMs format subjects allocated money to just one own group and one \\noutgroup. Two different studies (Bornstein et aI., 1983b) were conducted in \\nresponse to these objections. The first compared the old cover story with a \\n\\n\\nMere categorization and competition \\n1 85 \\nnew cover story and a condition with no cover story at all. The results \\nindicated no effect for the three-level manipulation of cover story. The \\nsecond experiment contrasted a procedure using allocation to one own \\ngroup and one other group to one using allocation to different own group \\nand outgroup members. The results indicated non-significant effects for \\nconstancy of recipient. Furthermore, both studies again found significant \\neffects for both min. diff. and max. joint own and a non-significant effect \\nfor max. reI. \\nThe major theoretical objection to the use of the MAMs concerned what \\nTurner called the maximizing assumption. Turner maintained that the \\nMAMs differentiate between social orientations only to the degree that \\nindividuals wish to maximize a value reflected by a particular orientation. \\nInstead, subjects may make a specific choice not because of its social \\norientation as reflected by the maximizing assumption, but because of a \\ndifferent orientation altogether. For example, a subject may choose max. \\nown not out of a desire to maximize own profits but instead out of a desire \\nto win to a lesser extent than is possible through max. reI. own. Thus, the \\nMAMs would reflect social values only to the degree that subjects adhere to \\nthe maximizing assumption. \\nIf one is able to cast doubt on whether subjects actually adhere to the \\nmaximizing assumption, reinterpretation of previous results using the \\nMAMs allows for possible support for social identity theory. Recall that \\nresearch with adult subjects found an abundance of evidence for max. joint \\nown but no evidence for max. reI. own or max. own. Obviously, max. reI. \\nown and max. own both involve advantaging of own group. But what \\nabout max. joint own? If the choice is interpreted to reflect a concern with \\nboth own-group and outgroup outcomes, the evidence of desire for \\nsuperiority is ambiguous at best. However, Turner argues that subjects may \\nchoose max. joint own not out of concern for maximizing both own and \\nother outcomes but out of a desire to win to a lesser extent than is possible \\nin max. reI. own. Thus, if the maximizing assumption is rejected, both the \\nmax. reI. own and max. joint own choices support the social comparison \\nprediction of social identity theory. \\nInsko, Pinkley, Harring, et al. (1987) conducted a study designed to \\ndetermine whether subjects actually invoked the maximizing assumption \\nwhen choosing either max. own or max. joint own. An identical format was \\nused to test the subjects except that the MAMs were presented one at a \\ntime. If and when a max. own or max. joint own was selected the subject \\nwas given a questionnaire containing an open-ended item and a closed­\\nended item. The open-ended item simply asked subjects to state their reason \\nfor their immediately prior choice. The closed-ended item forced subjects to \\nchoose between the maximizing assumption and the Turner rationale. For \\nmax. own these items were 'to gain the most amount of money possible' \\nand 'to win by less than the biggest margin'. For max. joint own these items \\nwere 'to gain the highest combined amount of money for everyone, with my \\ngroup receiving slightly more', and 'to win by less than the biggest margin'. \\n\\n\\n1 86 \\nSocial identity: international perspectives \\nThe results for both the open-ended and closed-ended items indicated \\noverwhelming support for the maximizing assumption. For example, 1 8  of \\nthe 21 subjects who chose max. joint own had open-ended responses \\nconsistent with the maximizing assumption (p < .0006 by binomial test). \\nThe MAMs versus the Ta\\nj\\nf\\nelian pull procedure \\nIf subjects do accept the maximizing assumption of the MAMs, why do the \\nMAMs yield different results from the pull procedure? Currently, we do not \\nhave a definitive answer to this question. We can, however, point to two \\nobvious differences in the assessment procedures. First, because the array of \\nalternatives on the MAMs are unconfounded, it should be relatively clear to \\nthe subject that he or she is selecting among distinctive choices. \\nSecond, unlike the pull procedure, the MAMs simultaneously present \\nsubjects with the full array of possible preferences to be assessed. Thus a \\nsubject is able to make a choice with complete knowledge of all of the \\npossibilities that are available. We find it plausible that in such a situation \\nthe subject's ability to indicate his or her clear preference will be optimized, \\nthat is, that the choice will be an informed choice. \\nConsider, for example, a subject who selects max. reI. own on the \\nMAMs. It should be apparent to the subject that the choice is clearly \\ndifferent from max. own, max. joint own, max. other, and so on. The \\nsubject should clearly understand that a max. reI. own choice is not a choice \\nto maximize own category outcomes, not a choice to maximize joint \\noutcomes, not a choice to maximize other category outcomes, but rather a \\nchoice to maximize the relative difference between own- and other-category \\noutcomes. If it is advantageous to use an assessment procedure which \\nmaximizes the subject's understanding of what is being chosen, why should \\ninvestigators continue to rely on the pull procedure? \\nSIT summary \\nTaken together, this series of experiments using the MAMs provides a very \\ndissatisfying picture of the contention that mere categorization is enough to \\ninstil relativistic competition through social comparison mechanisms. First, \\nfive different studies using a variety of populations, including adults, found \\nthat mere categorization does not lead to maximizing relative choices when \\ndifferentiated social orientation matrices are used instead of the traditional, \\nmotivationally confounded, Tajfelian matrices. Instead, subjects tend to \\nchoose max. joint own. Furthermore, research has shown that subjects do \\nappear to use the maximizing assumption inherent in the interpretation of \\nthe MAMs. In other words, with closer inspection of social motives, mere \\ncategorization does not produce the relative advantage competition that \\nTajfel believed to characterize real groups. Given this revelation, the \\nquestion then becomes: what exactly is needed beyond mere categorization \\nto produce intergroup competition? \\n\\n\\nMere categorization and competition \\nThe discontinuity effect \\nThe possibility o\\nf\\n schema-related belie\\nf\\ns\\n concerning the \\nappropriateness o\\nf\\n competitive behaviour \\n1 87 \\nAs mentioned earlier, the discontinuity effect refers to the tendency for \\nintergroup interactions to be more competitive than interindividual inter­\\nactions. Within the domain of investigated matrices, the effect appears to be \\nrobust and reliable. Moreover, research concerning the discontinuity effect \\nhas helped identify what is needed beyond mere categorization to produce \\nintergroup competition. We specifically hypothesize that all of the evidence \\ntriangulates on the suggestion that competition flows from circumstances \\nimplying the believed appropriateness of competition. We, furthermore, \\nspeculate that a set of such beliefs in the appropriateness of competition \\nconstitute a schema. By a schema we mean a set of beliefs serving to organize \\nand guide memory for past events and expectations regarding future events. \\nIn postulating such a schema we are clearly going beyond the evidence. More \\ndefinitive evidence for a schema would be provided by data indicating that \\nperceived effects are magnified or altered in memory. We have no such data; \\nrather, we have data which suggest that experienced competitiveness can \\nproduce further competitiveness - even when the further competitiveness \\ninvolves a different opponent. \\nCompetitive experience with one opponent and subsequent \\ncompetitive relations with another opponent \\nInsko, Pinkley, Harring, et ai. (1987) conducted an experiment comparing \\nintergroup interactions versus interindividual interactions in a Prisoner's \\nDilemma Game (PDG; see Figure 13.1) followed by the usual mere cat­\\negorization procedure. Previous research using the PDG had indicated that \\ngroups behaved competitively while individuals behaved cooperatively in \\nthe identical setting (McCallum et aI., 1985). Thus, the behaviour of \\nsubjects who experienced competition (in group-on-group interaction) or \\ndid not experience competition (in one-on-one interaction) on a previous \\ntask could be compared in the mere categorization task. It was hypothesized \\nthat the prior experience would carry over to the categorization task, \\nresulting in increased levels of competitiveness (max. reI. own or max. own). \\nThe results indicated that there was more competition between groups than \\nbetween individuals with the PDG and that this behaviour carried over to \\nthe mere categorization procedure in the form of increased max. own (but \\nnot max. reI.) on the MAMs. \\nReciprocation of competitive responses, or escalation of conflict with the \\nsame opponent, has been documented by others (for example, Kelley & \\nStahelski, 1970). It is important to realize, however, that the Insko, Pinkley, \\nHarring, et ai. result occurred despite the fact that the subjects were led to \\nbelieve that the subject composition of the Klee and Kandinsky categories \\nin the mere categorization situation was not necessarily the same as the \\n\\n\\n188 \\nSocial identity: international pers\\npectives \\nCooperation \\nCompetition \\nCooperation \\nCompetition \\n$0.57 \\n$0.66 \\n$0.39 \\n$0.66 \\nFigure 13.1 \\nA Prisoner's Dilemma Game matrix (PDG) \\nsubject composition of the groups in the PDG situation, and thus that the \\nobtained results were not a simple result of reciprocation. Why, then, did \\ncompetitive behaviour carry over from one situation to the other? One \\npossibility is that there was a perseveration of competitive behaviour. We \\nbelieve, however, that such an explanation is at best an oversimplification. \\nRather we interpret the results as suggesting that subjects shifted their \\nbeliefs regarding what was appropriate or normative. As a result of prior \\ncompetitive group-on-group interaction, subjects believed that when sub­\\nsequently categorized in a different group it was appropriate to behave \\ncompetitively towards another category or group. \\nTo some readers it may appear that in postulating that competitive \\nexperience results in a change in beliefs regarding the appropriateness of \\ncompetition we are going beyond the evidence, and indeed we are. Why do \\nwe need to postulate anything other than simple perseveration of com­\\npetitive behaviour? There are two answers to this question. First, the reader \\nshould note that there is a wealth of published research indicating that \\nbehaviour change can result in attitude and belief change (for a review of \\nthis literature see Chapter 1 1  in Eagly & Chaiken, 1993). In addition the \\nevidence for this linkage is particularly strong in those situations in which \\nthe choice to engage in the behaviour is 'free' or apparently unconstrained, \\nand that appears to be the case in the above-described research. Second, the \\nreaders should note that competitive experience also involves competitive­\\nness in the opponent, and a series of studies by Insko et aI. (1 992) relating \\nto max. reI. indicates that it is an opponent's competitiveness which is \\nparticularly important in producing subsequent competitive behaviour. \\nThe design of the Insko et aI. (1992) studies used a three-choice matrix \\n(labelled a PDR matrix; see Figure 1 3.2) which reflected three different \\nsocial orientations: max. joint, max. own and max. reI. The max. own \\nchoice on this matrix is suggested by Campbell's (1965) realistic conflict \\ntheory, which implies that the dominant social motive in intergroup conflict \\nis maximizing 'real' outcomes. The max. reI. choice obviously reflects the \\n\\n\\nMax. joint \\nMax. own \\nMax. reI. \\nMere categorization and competition \\nMax. joint \\nMax. own \\n$0.99 \\n$1 44 \\n$1 . 1 4  \\nConfounded \\nmax. rei. \\n$1 .44 \\n$1 . 1 4  \\n$0.84 \\n$0.84 \\n$1 44 \\nMax. rei. \\n$0.84 \\n$0.69 \\n$0.99 \\n$1 . 1 4  \\nConfounded \\nmax. reI. \\n$1 44 \\n$0.84 \\n$1 . 1 4  \\n$0.99 \\n$0.99 \\n$0.99 \\n1 89 \\nNote: The third choice in the PDR matrix is max. rei. The third choice in the PDC \\nmatrix is confounded max. reI., which is max. rei. that is confounded with max. own. \\nFigure 13.2 A Prisoner's Dilemma max. reI. matrix (PDR) and a \\nPrisoner's Dilemma con\\nf\\nounded matrix (PDC) \\nTajfelian notion of maximizing relative advantage to bolster social identity. \\nThe results of the first study supported both Campbell and Tajfel's theory \\nconcerning group behaviour: relative to individuals, groups made both \\nmore max. own and max. reI. choices. There was some hint, however, that \\nthe max. reI. behaviour only occurred after the completion of the practice \\ntrials, which were played for points instead of money against the other \\ngroup (or individual). \\nOne possible reason for this initial reluctance to choose max. reI. might \\nrelate to tactical considerations. Some groups may try tactically to 'set up' \\nthe other group by not choosing max. reI. until the later trials. Therefore, in \\nexperiment 2 we eliminated this confound by having subjects play their \\npractice trials against an experimenter, and invoked a procedure whereby \\neach group (or individual) played one trial with each of three other groups \\n(or individuals). This procedure was designed to eliminate possible tactical \\nconsiderations because subjects knew they would only interact once with \\neach other group (or individual). The procedure had the added benefit of \\n\\n\\n190 \\nSocial identity: international pers\\npectives \\nTable 13.1 \\nMax. rei. as a f\\nunction o\\nf\\n matrix type (PDR versus PDC), \\ntrials and groups versus individuals, experiment 2 \\nTrial I \\nTrial 2 \\nTrial 3 \\nGroups \\nIndividuals \\nGroups \\nIndividuals \\nGroups \\nIndividuals \\nPDR \\nPDC \\n0.08 \\n0. 12 \\n0.08 \\n0.63 \\n0.17 \\n1 .38 \\n0.18 \\n0.28 \\n0.50 \\n1 .06 \\n0.00 \\n0.27 \\nNote: Because four groups (or individuals) interacted on any given trial, max. reI. scores have a \\npotential range from 0 to 4. \\neliminating a possible interpretation of the results as being due to recipro­\\ncation of competition, or the escalation of conflict with an opponent. \\nIt is likely that the max. reI. that occurred after the practice trials reflects \\nthe fact that a number of interactions are needed to instil the belief that \\ncompetition is appropriate within the context of the PDR matrix. A further \\ncomplication is that subjects may have been initially unwilling to choose \\nmax. reI. out of a reluctance to reduce own outcomes. The max. own choice \\nproduced higher own outcomes than the max. reI. choice (see the PDR \\nmatrix in Figure 1 3.2). Experiment 2 addressed this problem by using, in \\naddition to the PDR matrix, a new matrix, the PDC matrix. On the PDC \\nmatrix own outcomes are identical (or confounded) for the max. own and \\nmax. reI. choices (see Figure 13.2). Thus, for this matrix the selection of \\nmax. reI. imposed no penalty on own outcomes. \\nThe results, once again, supported both Campbell and Tajfel: groups \\nselected max. own and max. reI. more often than individuals. However, \\nthere were significant max. reI. sequence effects for both the PDR and PDC \\nmatrices (see Table 1 3.1). On the PDC matrix, the tendency for groups to \\nchoose max. reI. more than individuals was non-existent on the first trial, \\nincreased on the second trial, and then levelled off on the third trial. On the \\nPDR matrix, the max. reI. effect was non-existent on the first and second \\ntrials, but was there on the third trial. There were no sequence effects for \\nmax. own on either matrix. There was a cooperation effect inverse of that \\nfor max. reI. for both matrices, but this effect can be regarded as a \\nconsequence of the shifts in max. reI. \\nThe findings for both studies are consistent. Max. reI. selections did not \\ncommence until some experience with the matrices had occurred. We should \\nnote here that, at least in study 2, this could not have been the result of \\ntactical considerations that arise when repeatedly facing another group. \\nThese effects occurred even when subjects knew they would face each other \\ngroup only once. \\nIt is also important to recognize that the increased competitiveness cannot \\nbe the result of simple reciprocation of conflict. Again, subjects knew they \\nwould face each other group only once. Therefore, there was no chance for \\nretaliation against a group that competed. Reciprocation and escalation of \\nconflict seem unlikely explanations of the competition sequence effects. \\n\\n\\nMere categorization and competition \\n191 \\nA further analysis of the tria1-by-trial patterns in the data indicated that \\nthe tendency to chose max. reI. (confounded with max. own) on the PDC \\nmatrix on the second trial occurred most obviously for those groups whose \\nopponents on the first trial had chosen max. own. Having been hit with a \\ncompetitive choice from the first opponent, groups, but not individuals, hit \\ntheir second opponent with a more competitive (but not own outcome \\nreducing) max. reI. choice. On the other hand, the tendency to chose pure \\nmax. reI. on the PDR matrix on the third trial occurred for those groups \\nwho had engaged in mutual max. own responses with their second trial \\nopponents. Having exchanged mutual max. own choices with their second \\nopponent, groups, not individuals, hit their third opponent with pure max. \\nreI. (max. reI. that did reduce own outcomes and was thus unconfounded \\nwith max. own). Such findings are consistent with the general supposition \\nthat prior competitive experience can activate a schema regarding the \\nbelieved appropriateness of a highly competitive, max. reI. response. The \\nevidence suggests that after groups engage in mutual max. own responses \\nthey 'reach for a bigger gun' (pure max, reI.) to use against their next \\nopponent. At that point groups are willing to reduce their own outcomes \\njust to increase the likelihood of doing relatively better than their new \\nopponent. \\nRecall the above-described Insko, Pinkley, Harring, et aI. (1987) experi­\\nment in which participants interacted first in the context of the PDG and \\nsecond in the mere categorization situation. In that experiment there was \\nthe possibility that the max. own choices in the mere categorization \\nsituation were a simple result of response perseveration. Note, however, that \\nin the present experiments (Insko et aI., 1992) the response-perseveration \\npossibility can be ruled out for the simple reason that the max. reI. choice \\ndid not follow prior own max. reI. choice. For the PDC matrix the max. reI. \\nresponse occurred following the max. own choice by the opponent., and for \\nthe PDR matrix the max. reI. response occurred following mutual max. own \\nresponses. The key thus appears to be the experience of competitiveness by \\nthe opponent. \\nThe discontinuity research paradigm \\nStudies investigating the discontinuity effect are typically laboratory experi­\\nments that observe interaction using a Prisoner's Dilemma outcome matrix \\n(or some variation thereof). The research is usually conducted in a \\nlaboratory suite in which same-sex individuals or groups are located in \\ndifferent rooms connected to a central room. After extensive training \\nconcerning the interdependent nature of the PDG, a trial or series of trials is \\nconducted. Subjects are given time to examine a version of the PDG, then \\nhave some time to discuss possible action with the other group (often face to \\nface with representatives but sometimes over a communication system), and \\nthen return to own-group discussion and actual decision-making. In group \\nconditions the decision is typically made by the group as a whole or based \\n\\n\\n1 92 \\nSocial identity: international perspectives \\nupon a majority vote. The decisions are then collected and the subjects paid \\naccordingly. The discontinuity effect has been shown across a large number \\nof studies regardless of: number of trials; size of payoff; gender; intergroup \\ncontact using representatives or all group members; whether the low values \\nin the matrices are positive or negative; the type of matrices used (PDG, \\nPDG-Alt., Mutual Fate Control); and whether there is equal or non-equal \\npower. \\nGroups are not always more competitive than individuals \\nA further issue that merits comment relates to whether groups are 'always' \\nmore competitive than individuals. Clearly in many instances groups are \\ncooperative and individuals are competitive and it would be foolish to \\nmaintain otherwise. Even though the above point has been repeatedly made \\n(for example, Insko et aI., 1993; Schopler & Insko, 1992), it is sometimes \\nasserted that we believe otherwise. For example, Rabbie and Lodewijkx \\n(1994) claim that 'Schopler and Insko (1992) have concluded, based on their \\nresearch with the PDG, that groups are invariably more competitive and \\nless cooperative than individuals' (p. 1 58). Actually, Schopler and Insko \\n(1992) and Insko et al. (1993) argued that Kelley and Thibaut's (1978) index \\nof correspondence provides a basis for predicting when groups will be more \\ncompetitive than individuals. Kelley and Thibaut's index of correspondence, \\nin a situation in which the range of outcomes for the two players are \\nthe same, is the correlation between the outcomes across the cells. With the \\nPDG and with other matrices that we have investigated, this correlation \\nis moderately negative. Clearly, one would not expect discontinuity with \\nmatrices involving totally correspondent outcomes (+ 1 correlation). With \\nsuch matrices both groups and individuals should coordinate their responses \\nand there thus would be no discontinuity effect. Likewise in a zero-sum \\nsituation with totally non-correspondent outcomes (-1 correlation) both \\nindividuals and groups have no alternative to competing, and there would \\nagain be no basis for a discontinuity effect. Given this theoretical argument, \\nthe empirical challenge is to determine whether the circumstances in which \\ngroups are more competitive than individuals are, indeed, situations with \\nsomewhat, but not totally, non-correspondent outcomes. This is an issue \\nthat is currently being investigated. \\nA further consideration that is relevant to the issue as to whether groups \\nare always more competitive than individuals is the fact that our investi­\\ngations of the discontinuity effect have almost always been in situations in \\nwhich individuals and groups communicate with each other prior to making \\nsimultaneous and irrevocable choices. Typically the communication has \\nbeen face-to-face communication, but sometimes the communication has \\nbeen via an intercom. Furthermore, Insko et al. (1993) obtained evidence \\nindicating that in one-trial situations the absence of communication via an \\nintercom produced a larger decrease in cooperation between individuals \\nthan between groups. Further evidence indicated that the presence of \\n\\n\\nMere categorization and competition \\n193 \\ncommunication produced a greater decrease in expected competItIOn \\nbetween individuals than between groups. Such evidence is consistent with \\nthe possibility, as discussed below, that there is an outgroup schema leading \\ngroups not to trust each other and thus for communication between groups \\nto be of relatively low credibility. All of which is to suggest that without \\nsome form of communication the discontinuity effect should be markedly \\nsmaller and may, under some circumstances, be non-existent. Typically, of \\ncourse, relations between individuals and relations between groups do occur \\nin the context of some form of communication. \\nThe theoretical basis f\\nor discontinuity: ingroup (greed) and outgroup \\n(f\\near) schemata \\nAs mentioned earlier, research investigating the discontinuity effect allows \\nfor careful consideration of the roots and nature of competition between \\ngroups. Previously mentioned research has provided evidence for the link \\nbetween the belief in the appropriateness of competition (group schema) \\nand competitive behaviour. Actually there is reason to believe that the \\nschema concerning the believed appropriateness of competition can be \\nsubdivided into two schemata, the ingroup, or greed, schema and the \\noutgroup, or fear, schema. At this point we would like to ruminate more \\nextensively on these two theoretical bases for the discontinuity effect, \\nThe ingroup (greed) schema \\nIf a set of beliefs regarding the appropriate­\\nness of competition exist, it is reasonable to suppose that some of those \\nbeliefs relate to what is appropriate and expected within the ingroup. But \\nwhat might those beliefs be? One possibility relates to the beliefs that are \\npostulated by the altruistic rationalization hypothesis. According to this \\nhypothesis, group members assume that they are able to rationalize their \\ncompetitiveness in intergroup interactions by believing that such behaviour \\nis for the benefit of in \\ngroup members. Thus, such competition is altruistic in \\nregards to other members of the ingroup. The hypothesis bears some \\nresemblance to Snyder's concept of attributional ambiguity (Snyder, Kleck, \\nStrenta & Mentzer, 1979; Snyder & Wicklund, 198 1). Snyder maintains that \\nself-interested behaviour is more likely to occur when there is attributional \\nambiguity as to whether the behaviour is self-interested in nature. Unfor­\\ntunately, there is evidence that when subjects are explicitly given the \\nopportunity to share outcomes, and thus make altruistic rationalizations, \\nthey do not become more competitive. Insko, Pinkley, Hoyle, et al. (1987) \\nconducted two studies employing a condition in which individuals were \\nbound together by outcome interdependence. Within this condition, there \\nwere three simultaneous dyad interactions, with all subjects having a separ­\\nate room, three on each side of the centre hall. However, subjects were told \\nthat, while they would interact alone with an individual on the other side \\nof the hall, their money would be divided equally with the other two people \\non the same side of the hall. According to the altruistic rationalization \\n\\n\\n1 94 \\nSocial identity: international pers\\npectives \\nhypothesis, this outcome interdependence condition should have created \\nmore competition than the individual interaction condition. However, the \\nmeans for the two conditions were virtually identical, thus making the \\naltruistic rationalization hypothesis implausible. \\nA more compelling argument for the creation of a competition-allowing \\ningroup schema revolves around the social support f\\nor shared self-interest \\nthat is available for groups but not for individuals. A belief in the avail­\\nability of within-group support is quite likely based on a wealth of within­\\ngroup experience. We thus postulate that a history of mutual support for \\nshort-term self-interest has produced an ingroup schema that supports \\ngreedy behaviour. Individuals, by definition, are not afforded this support \\nand therefore, apart from fantasised social support, are less likely to develop \\nthis schema. In characterizing social support for shared self-interest as a \\nschema we are explicitly postulating that individuals in groups expect that \\nthey will be supported when they suggest a course of action consistent with \\nthe group's self-interest. Beyond this, however, we also postulate that greed­\\nbased competitiveness is most likely in a situation in which the other group \\nis expected to be cooperative and is therefore vulnerable. (Combativeness \\nthat occurs in a context in which the other group is expected to be com­\\npetitive and is therefore dangerous is more reasonably attributed to fear \\nthan to greed.) Schopler et al.'s (1993) study 2 was designed to demonstrate \\nthe resultant competitive behaviour when there is social support for self­\\ninterest, and an expectation that the other group will be cooperative. \\nConfederates were placed in each group and told to give a single suggestion \\nas to what the group should choose in a standard PDG. The confederates \\nwere told to suggest to the other two subjects either that they should choose \\nX (the cooperative choice) or Z (the competitive choice). The use of \\nconfederates also allowed for manipulation of the feedback concerning what \\nthe other group actually chose. By selecting the confederates as represen­\\ntatives it was possible to give the subjects false feedback that the other \\ngroup was either choosing all X, all Z, or a 50-50 mixture of the two. The \\nfeedback manipulation was conceived as a manipulation of the other \\ngroup's expected behaviour. The prediction was of an interaction between \\nthe feedback and suggestion factors such that only with cooperative \\nfeedback would the cooperative suggestion succeed in reducing intergroup \\ncompetitiveness. Within the context of the repeated cooperative feedback \\nwhere the other group is perceived as vulnerable, the cooperative suggestion \\ntheoretically should have reduced the within-group social support for \\ncombativeness, and thus reduced competitiveness in this condition relative \\nto all the other conditions. The predicted pattern of results occurred. \\nThe outgroup (f\\near) schema While social support may foster greed through \\nthe bolstering of a competition-allowing ingroup schema, belief that the \\nother group will be competitive is also expected to create competition within \\nthe PDG paradigm. Competition as the result of a competitive outgroup \\nschema has been termed the 'fear' hypothesis. \\n\\n\\nMere categorization and competition \\nCooperation \\nWithdrawal \\nCompetition \\nCooperation \\nWithdrawal \\nCompetition \\nFigure 1 3.3 \\nA Prisoner's Dilemma Game alternative matrix (PDG-Alt. ) \\n195 \\nThere is some evidence that individuals do maintain beliefs describing \\nother groups as competitive, aggressive and dishonest in nature. Doise \\n(1969) reported that before any interaction subjects attribute fewer coopera­\\ntive motives to the opposing group than to themselves or their own group \\nmembers. Hoyle, Pinkley and Insko (1989) found that the source of this \\nnegative expectancy was not whether one was alone or in a group, but \\nwhether the opponent was a group or an individual. The expectations for \\nthe outgroup were significantly more competitive and aggressive regardless \\nof whether one was alone or in a group. We do not have definitive evidence \\nthat such expectations, or beliefs, constitute a schema, but we are explicitly \\npostulating that they do. \\nThere is evidence relating the postulated outgroup schema to the discon­\\ntinuity effect itself. This evidence comes from research using a new three­\\nchoice matrix referred to as a \\nDilemma Game (PDG-Alt.) matrix (see \\nFigure 1 3.3). The PDG-Alt. matrix includes a third choice in addition to the \\ncooperative and competitive choices of the PDG matrix. This new choice \\nguarantees intermediate outcomes regardless of the opponent's choice. This \\nnew choice, labelled 'withdrawal', is thus a safe choice that should be \\nrationally preferred when there is fear of the opponent's competitive intent. \\nInsko et al.'s (1990) second study used the PDG-Alt. matrix. The results \\nindicated a discontinuity effect for both competition and withdrawal. Not \\nonly did groups compete and withdraw more than individuals, they com­\\npeted and withdrew to approximately the same extent. Thus, the discon­\\ntinuity effect cannot be explained by greed alone. The negativity of the \\noutgroup schema (that is, the expectation that another group IS more \\ndangerous than another individual) also affects group behaviour. \\n\\n\\n196 \\nSocial identity: international perspectives \\nWhat is not clear, however, is the basis for this negative outgroup schema. \\nOne could suggest a number of hypotheses regarding its development. The \\nfirst could simply be experience. We may develop general negative outgroup \\nschemata based upon negative experiences with specific outgroups. Such \\nexperiences may cloud interactions with any group, just as the competitive \\nPDG experience made competition more likely in the mere categorization \\nparadigm, and just as max. reI. did not emerge until after a few trials using \\nthe PDR and PDC matrices. Consistent with this speculation, Pemberton, \\nInsko and Schopler (1997) obtained evidence indicating that intergroup \\nrelations are recalled as more competitive than interindividual relations, and \\nalso that intergroup relations are experienced as more competitive than \\ninterindividual relations. The experiential data were obtained from 'diaries' \\nthat subjects carried with them for seven days (Reis & Wheeler, 1991). \\nA second hypothesis concerns the role of false consensus. This negative \\nview of outgroups in general may stem from our knowledge of our own \\ncompetitive tendencies within groups. Thus, we may only need to look at \\nourselves and by projection develop a prediction of how the other group \\nmay behave. Further research should expand our understanding of these \\nmatters. \\nRegardless of how the negative outgroup schema develops, its effects on \\nbehaviour are predictable. When the only choice is between cooperation \\nand competition, as in the standard PDG, groups lash out in fear. It may be \\nthat the root of many group behaviours that are intensely competitive in \\nnature are simply the result of fear of the other group and not a reflection of \\ningroup greed. This is certainly apparent when there is the chance to \\nwithdraw from the situation with certainty of outcomes, as in the PDG-Alt. \\nIn such situations groups tend to withdraw as often as they attack. \\nReducing discontinuity through recategorization: preliminary results \\nThere is an 'older' literature and a 'newer' literature on the overcoming of \\ningroup ethnocentricism. The 'older' literature tended to specify remedial \\ninterventions in terms of interactive behaviour between two groups. There \\nwere such global recommendations as joint interactions to attain super­\\nordinate goals (Sherif, Harvey, White, Hood & Sherif, 1961), and there \\nwere the specific behavioural rules contained in Osgood's (1962) Graduated \\nand Reciprocated Initiatives in Tension-reduction (GRIT). The 'newer' \\nliterature focuses on the cognitive mechanisms presumed to be antecedents \\nto intergroup bias, and seems to vary with respect to the directness with \\nwhich the intervention affects relevant cognitions. The most direct methods \\nare those that degrade the categorized representation by such methods as \\nrecategorization (Gaertner, Dovidio, Anastasio, Bachman & Rust, 1993; \\nGaertner, Mann, Murrell & Dovidio, 1989) or criss-crossing category \\nmembership (Brewer, Ho, Lee & Miller, 1987). With the exception of \\nmodelling GRIT (Lindskold, 1983), both traditions tested for a reduction in \\nethnocentrism with measures of individual perceptions of ingroup and \\n\\n\\nMere categorization and competition \\n197 \\noutgroup members. To our knowledge no one has attempted to assess \\nwhether the intervention was successful in increasing such behaviours as \\nintergroup cooperation. We believe the ultimate effectiveness of any tech­\\nnique resides in its ability to create intergroup cooperation between intact \\ngroups. \\nWe have pilot tested some recategorization procedures that, in terms of \\nreduced intergroup conflict, were spectacular failures. The pilot studies \\nbegan with the formation of four three-person groups, who were sub­\\nsequently combined into two rivalrous six-person groups or, in a different \\nversion, two, independent, six-person groups each working cooperatively on \\na task. The recategorization worked well, in the sense of increasing liking \\nfor former outgroup members and creating satisfactory working relations \\nwith them. But, when we regrouped participants into their initial three­\\nperson groups and asked them to interact with previous group members on \\na PDG, their rate of competition was just as large as the rate for groups \\nwhose members were never recategorized. All of which makes us suspect \\nthat if at the end of camp Sherif and his associates (Sherif et aI., 1961) had \\nput the Eagles and Rattlers into a mixed-motive situation, they would have \\nbeen just as competitive as they were in the tournament. Quite possibly \\nSherif and others have underestimated the power of the social setting. \\nConclusions \\nThat individuals gain self-definition and bolster self-esteem by belonging to \\ngroups is certainly a reasonable working hypothesis. Furthermore, we find it \\nplausible that the need to support high self-esteem through the demon­\\nstrated superiority of one's own group may be at least a partial explanation \\nfor between-group competitiveness. This main tenet of social identity theory \\nhas not been questioned. However, the belief that mere categorization into \\ngroups produces such intergroup competitiveness, a second tenet of the \\ntheory, has been questioned. An abundance of research has demonstrated \\nthat when more differentiated social orientations are used, mere \\ncategorization does not produce the relative competitiveness predicted by \\nsocial identity theory. \\nInstead, research detailing the discontinuity effect has suggested that what \\nis need to produce the intergroup competitiveness beyond mere categoriza­\\ntion is a belief that competition is appropriate. There is evidence that \\ncompetitive behaviour increases following competitive experience. This is \\ntrue even when the opponent is a group other than the one with whom the \\nprior competitive experience occurred. Such evidence is consistent with the \\npossibility that beliefs in the appropriateness of competition can produce \\ncompetition between categories. We have speculated that such beliefs flow \\nboth from an ingroup schema relating to expected social support for self­\\ninterested, or greedy, behaviour, and from an outgroup schema relating to \\nthe expected competitiveness, and dangerousness, of an outgroup. \\n\\n\\n198 \\nSocial identity: international pers\\npectives \\nNotes \\nWork on this chapter was supported by a National Institute of Mental Health Grant \\n(5ROI MH53258) to the last two authors. \\nI .  Given the existence of difference in the interaction of two groups and the interaction of \\ntwo individuals, we are inclined to define aggregates of people as groups whenever behaviour \\nbetween these aggregates is different from behaviour between individuals. However, in order to \\navoid circularity it is necessary to use the term 'group' in two different senses. Groups in the \\nsense of aggregates, or sets of individuals, is assumed in the definition of discontinuity. The \\nexistence of discontinuity, however, provides evidence for the existence of groups in the more \\npsychological meaning of the term. \\n2. In seeming conflict with the above results, Platow, McClintock and Liebrand (1990) \\nfound that categorized subjects would select max. reI. on the Bornstein et al. (1983a, 1983b) \\nmatrices. Two differences between this experiment and all, or much, of the above research are \\nthat Platow et al. used points and not money for outcomes, and did not have a cover story. The \\ncombination of points and the lack of a non-game cover story may have led some subjects to \\ninterpret the situation as game-like, and thus to attribute relative value to the points. We \\nbelieve that researchers should be careful not to create a game-like atmosphere and should use \\noutcomes that have value outside of the experimental context. \\n\\n\\nPART IV \\nTHE IMPACT OF \\nCULTURE ON IDENTITY \\nAND CATEGORIZATION \\n1 4  \\nDiscrimination and Beliefs on \\nDiscrimination in Individualists and \\nCollectivists \\nJ. Francisco Morales, Mercedes L6pez-Saez and Laura Vega \\nThe concept of individualism \\nIndividualism refers not to the individual as an empirical subject or \\n'individual specimen of the human species', but rather to the individual as a \\n'moral, independent, autonomous being, . . .  \\nbearer of our supreme values'. \\nIn this second sense, individualism is the prevailing contemporaneous \\nideology of person and society (see Dumont, 1987, pp. 37ff.). \\nAs opposed to individualism, which ascribes to the individual the main \\nvalue in society, for holism the supreme value lies in society as a whole. \\nHistorically, individualism appears in holistic societies as a source of \\nopposition to them and as a sort of supplement. Here, the Indian 'renouncer' \\nis an important landmark. In order to achieve the utmost truth, the \\nrenouncer relinquishes social life and its limitations and devotes himself to \\nhis own destiny. Similarly, in the Hellenistic period, against the self-sufficient \\npolis of \\nPlato and Aristotle, the idea was postulated that in order to become a \\nwise person it was necessary to renounce the world. At that time, non­\\nlearned people were considered as victims of the worldly life. \\nFrom the beginning of the modem era, society is usually conceived as an \\n'association', which suggests that the individuals who belong to it have \\nassociated with each other 'voluntarily', and thus they are prior to the groups \\nand/or the relations they establish among themselves. This conception entails \\n\\n\\n200 \\nSocial identity: international perspectives \\ntwo characteristic features: equality and freedom. There is a belief that one's \\nown selfish interest is the best guarantee of social order and of general \\nsatisfaction for all members of society. \\nNevertheless, quite often, the way many people have used freedom has \\nprevented others from achieving their own. Here is the central 'paradox' of \\nindividualism, according to Camps. Ethics must necessarily be individua­\\nlistic, because it has to preserve the individual. But it must be achieved in \\nsuch a way that it protects everyone at the same time, not some at the \\nexpense of others. When this is not achieved, individualism is equivalent to \\n'anti-ideology', insolidarity, insensitivity to unequalities, lack of interest in \\npublic affairs. And that is what happens to most developed societies, 'whose \\nonly expectation is the perpetuation of group' (see Camps, 1 993, pp. 14f.). \\nCultural aspects of individualism \\nHofstede (1980) reports the results of the HERMES project, which involved \\nthe study of employees from 40 different countries, all of them belonging to \\nthe same multinational business corporation. From the responses given by \\nthis sample of subjects to 14 work goals, this author identified a factor \\nrepresenting a dimension of individualism-collectivism. \\nThe individualism-collectivism factor loaded positively on the following \\ngoals, from highest to lowest loading: (a) enough free time for oneself and \\none's family; (b) freedom to follow one's own way at work; (c) having a job \\nwhich provides a personal sense of performance; and (d) living in a place \\ndesirable to oneself and one's family. The goals with negative loadings, also \\nin a decreasing order, were: (a) opportunities to learn new skills; (b) good \\nphysical working conditions for work; (c) use of skills and abilities at work; \\nand (d) good marginal profits. According to Hofstede, goals with positive \\nloadings have in common that they emphasize personal independence vis-a­\\nvis the organization. On the other hand, goals with negative loadings refer \\nto things the organization can do for the individual. \\nThe factor analysis conducted by Hofstede has been called 'ecological', \\nbecause the loading of each goal is the correlation coefficient between each \\ncountry's factor score and its mean score in that goal. This correlation is \\ncalculated across all 40 countries. From factor scores an Individualism \\nIndex is calculated for each country, ranging from 0 to 100 due to a \\nmathematical transformation. \\nPsychological aspects of individualism \\nHui (1988) postulates that individualism is also an individual tendency or \\npredisposition, since there is, within any culture, subcultural variation \\namong individuals regarding this dimension. This author's work has two \\nmain characteristics. First, the development of a scale to measure the \\n\\n\\nDiscrimination and belie\\nf\\ns\\n on discrimination \\n201 \\nindividualism-collectivism construct, named INDCOL, consisting of eight \\nsubscales. Thus, the construct is considered as multifaceted and is measured \\nin relation to eight specific targets: spouse or partner, relatives, parents, \\nneighbours, close friends, co-workers, acquaintances and siblings. \\nSecond, INDCOL measures beliefs, attitudes, behavioural intentions and \\nbehaviours. As Hui himself notes, this has both theoretical and methodo­\\nlogical consequences. From a methodological standpoint, generality of \\nmeasurement increases, although it is at the expense of a part of the scale's \\ninternal consistency. From a theoretical point of view, individualism­\\ncollectivism is not considered as a unitary disposition, but rather as a \\nsyndrome. In fact, by adding up the scores of the eight subscales, the \\nGeneral Index of Collectivism is obtained. Hui defines this index as the \\ndegree to which the person believes the group, and not the individual, is the \\nbasic unit of survival. \\nBased on INDCOL and other scales developed by himself, Triandis, \\nBontempo, Villareal, Asai and Lucca (1988) produced the Three Factor \\nQuestionnaire. They applied this questionnaire to samples from the United \\nStates and found, as did Hui, that individualism is a multifaceted construct, \\nsince it appears characterized by the following features: (a) priority given to \\none's own goals over those of the ingroup; (b) less attention paid to the \\nviews of the ingroup; (c) self-reliance with competition; (d) distance from \\ningroup; (e) making decisions on one's own before asking others' point of \\nview; (f) less concern for ingroup. Therefore, in the event of a conflict \\nbetween ingroup and individual, individualism will lead the person to do \\nwhat he or she, and not the ingroup, considers appropriate. \\nCorrelates of individualism: self, relationship with in \\ngroup and \\noutgroup, emotional experience \\nBesides the above-mentioned work by Hui (1988) and Triandis et al. (1988), \\nmany others have studied the differences in social behaviour between \\nindividualists and collectivists. Among the best known the following can be \\ncited: Hui and Triandis (1986), Markus and Kitayama (J991a), Triandis \\n(1 993), Triandis, Leung, Villareal and Clack (1985), Triandis et al. (1 986) \\nand Triandis, McCusker and Hui (1990). Table 14.1 summarizes the main \\ndifferences regarding self and relationship with ingroups and outgroups. \\nIn order to better understand the content of this table, two points must be \\nnoted. First, not all ingroups are equally important for collectivists. On the \\ncontrary, there is usually a preference or priority ranking among them. \\nAccording to Triandis (1993), in many societies the order is (a) family, (b) \\nnation, (c) religious group and (d) work group. However, in Japan the \\nreligious group does not exist, and the work group ranks higher. Together \\nwith the priority order of ingroups, we must take into account the existence \\nof vertical collectivism and horizontal collectivism. The former implies \\n\\n\\n202 \\nSocial identity: international pers\\npectives \\nTable 14. 1 \\nSummary of some dif\\nf\\nerences between individualists and \\ncollectivists \\nDiff\\nerence target \\nIndividualists \\nCollectivists \\nContent of self \\nIndividual diff\\nerences \\nSocial categories \\nWay of achieving self-\\n'I can do whatever I want' \\n'I am not a burden to my \\nactualization \\ngroup' \\nBasic unit of survival (belief) \\nIndividual \\nGroup \\nRegulation of behaviour \\nPersonal attitudes and cost-\\nIngroup norms \\nbenefit analysis \\nPersonal goals vs ingroup \\nPersonal > ingroup \\nIngroup > personal \\ngoals \\nDiff\\nerence between ingroup \\nWeak \\nStrong \\nand outgroup \\nIngroup and outgroup \\nOutgroup more \\nIngroup more homogeneous \\nhomogeneity \\nhomogeneous \\nKind of relations \\nHorizontal \\nVertical \\nsubmission to ingroup authority. The latter means a fusion with in \\ngroup \\nidentity. Both kinds of collectivism may coexist, but the prevalence of one \\nor the other depends on the type of society concerned. \\nMarkus and Kitayama (199Ia) designate individualist cultures as 'inde­\\npendent self' cultures, and collectivist ones as 'interdependent self' cultures. \\nAccording to these authors, there are important differences between both \\ntypes of culture as far as emotional experience is concerned, since the \\nfundamental emotions, their sustaining basis and the functions they fulfil \\ndo not coincide. Thus, in 'independent self' cultures, the most important \\nemotions are ego-focused, specifically anger, frustration, pride, guilt and \\nhaughtiness, their basis consists on the subject's internal attributes, and their \\nfunction is the expression of inner feelings. On the other hand, emotions in \\n'interdependent self' cultures are focused on others. Among them shame \\nand sympathy stand out, the basis of these being a special sensitivity to \\nothers, an ability to take the perspective of the other and an active search \\nfor interdependence. Their function is to express the outcome of \\ninterpersonal relationships. \\nIndividualism and social identity: Hinkle and Brown's proposal \\nAs the various chapters in this book have shown, social identity theory \\n(SIT) deals with processes of group comparison and identification, rela­\\ntionships between groups, the formation of psychological group, and the \\nrelative weight of personal and social identity, among many other processes. \\nResults obtained by those who have studied the repercussions of the \\nindividualism-collectivism orientation, summarized in Table 14.2, lead us \\nto expect a convergence between studies on individualism and those carried \\nout within the framework of SIT. \\n\\n\\nDiscrimination and belie\\nf\\ns\\n on discrimination \\nTable 14.2 Relation between identification with the \\ngroup and intergroup dif\\nf\\nerentiation as a f\\nunction of \\nindividualism/collectivism and autonomous/relational \\norientation \\nOrientation \\nOrientation \\nAutonomous \\nRelational \\nIndividualist \\n.03 \\n.33* \\nn = 17 \\nn = 24 \\nCollectivist \\n.47** \\n.79*** \\nn = 25 \\nn = 1 8  \\n* p < . 10; ** p < .05; * * *  \\np < .01 \\n203 \\nHinkle and Brown (1990) were the first to pay attention to this con­\\nvergence. From a theoretical point of view, they state that processes of \\nsocial identity studied by SIT will take place mainly in 'collectivist kind of \\ngroups' (p. 66), since it seems reasonable to expect only collectivist groups \\nor individuals to worry about the outcome of comparisons between ingroup \\nand outgroups. The general underlying idea is that a collectivist orientation \\nseems important for the manifestation of social identity processes (Hinkle & \\nBrown, p. 67). \\nFrom an empirical point of view, Hinkle, Brown and Ely (1992) found \\nthat the relation between identification with the group and intergroup \\ndifferentiation is significantly greater in collectivists than in individualists. \\nThis is a proof, they argue, that only collectivists are really concerned with \\nthe outcome of comparisons between ingroup and outgroups (See Table \\n14.2). \\nIndividualism and identity theory: other perspectives \\nSIT is also a theory of the formation of a psychological group. Since the \\nearlier experiments using the minimal group paradigm, both Tajfel and \\nTurner interpreted the discrimination taking place in that situation as \\nevidence of group formation. Insko and Schopler (1987, p. 2 1 5) note that \\nTajfel considers the effect of mere categorization as a proof of entitativity, \\nmeaning the feeling of shared group membership. As early as 1 98 1 ,  Brown \\nand Turner proposed using an interpersonal-group continuum instead of \\nthe famous interpersonal-intergroup continuum. The reason was, according \\nto these authors, that 'the same processes may tend to distinguish both \\nintragroup and intergroup processes of interpersonal relationships' (Brown \\n& Turner, 198 1 ,  p. 40). In other words, intergroup differentiation implies \\ngroup formation and vice versa. \\nNevertheless, research by Triandis and his co-workers suggests that one \\nof the main differences between individualists and collectivists is the \\n\\n\\n204 \\nSocial identity: international perspectives \\nrelationship they maintain with their own groups. In principle, the idea \\nmust be rejected that individualists find it difficult to join a group. Even \\naccepting the schematic picture of Table 14. 1, the very fact of having an \\ningroup and an outgroup implies group identification, as minimal as this \\nmay be. Even more, it would not be incorrect to predict a greater ease for \\nindividualists to form a group, having in mind that their differentiation vis­\\na-vis outgroup people is weaker. For the same reason, it seems difficult for \\ncollectivists to be able to form a group with people not belonging formerly \\nto any of their few ingroups. \\nAt this point we must mention an interesting result reported by Hinkle, \\nBrown and Ely (1992, p. 107-108). For subjects showing an autonomous \\nas well as individualist orientation, the correlation between identification \\nand differentiation was .03, as can be seen in Table 14.2. However, these \\nsubjects did show positive discrimination in favour of their group. There­\\nfore, we tried to find out: (a) whether individualists show discrimination in \\nthe minimal group paradigm situation; (b) whether there are differences \\nbetween them and collectivists in their showing intergroup discrimination. \\nFirst experiment \\nSubjects were 52 students from adult education courses, aged between 20 \\nand 60, and being of a low socio-economic level. All of them had previously \\ncompleted the Spanish version of Triandis et al.'s (1988) Three Factor \\nQuestionnaire, validated in this country. Two groups of 26 subjects were \\nformed, each consisting of 1 3  subjects above and 1 3  subjects below the \\nmedian of the questionnaire. Thus, both groups were homogeneous in this \\nvariable. One of the groups was put to the standard categorization pro­\\ncedure. This was the experimental or categorized group. The other one, not \\ncategorized, was the control group. Therefore, there were two variables -\\ncategorization and individualism - each having two values. Subjects \\nanswered only once to the six Tajfel matrices and to the 10 Insko matrices. \\nThe MANOV \\nA carried out on Tajfel's strategies showed that the \\nexperimental group (categorized) was more discriminatory than the control \\ngroup (not categorized). Thus the experimental group was significantly \\nhigher in maximum ingroup profit (MIP) over maximum joint profit (M1P): \\n6.17 vs I, F(I, 39) = 1 2.93, p < .01 .  It was also higher in maximum \\n(ingroup) difference (MD): 5.29 vs -.58, F(I, 39) = 1 8.61, P < .01 and in \\nMIP over fairness (F): 5.96 vs -.26, F(I, 39) = 16.96, P < .01 .  On the other \\nhand, the control group was higher in fairness or equity: 7.84 vs 2.38, F(I, \\n39) = 7.78, P < .01 .  \\nThe MANOV \\nA carried out on Insko's strategies yielded similar results. \\nThe experimental group was higher in max. reI. own: .86 vs - 1 .0 \\nI, F(I, 39) \\n= 10.83, P = .06. The control group was higher in max. joint other: .25 vs \\n-.93, F(I, 39) = 9. 1 5, p < .06, and also in min. diff.: 1 .83 vs .61, F(I, 39) \\n-3.76, p < .06. \\n\\n\\nDiscrimination and belie\\nf\\ns\\n on discrimination \\n205 \\nNo effects of individualism were found, nor interaction between cat­\\negorization and individualism. This means that the classic effect of ingroup \\nfavouritism, first found by Tajfel and afterwards by many other researchers, \\nappears again in our study, both for individualists and for collectivists in the \\ntwo types of matrices. \\nIn order to find out whether there were any differences in the kind of \\ndiscrimination showed by individualists and collectivists, within-cell analysis \\nof the preferences in the use of different strategies were carried out, both in \\nexperimental and in control group. No significant differences appeared \\neither. \\nSecond experiment \\nWithin-cell comparisons in the first experiment were based on a small \\nnumber of subjects (13). We carried out a second experiment to increase \\nthat number. We decided not to include the control group, since the effect \\nof categorization seemed beyond doubt, and focused only on the individual­\\nism variable. Ninety-two subjects participated, of whom 47 were indi­\\nvidualists and 45 were collectivists. \\nThe procedure was similar: subjects completed the six Tajfel matrices and \\nthe 10 Insko ones, which provided the first measure of the dependent \\nvariable, named first choice. However, a change was introduced. After the \\nbooklet where they had marked their preferences had been taken away, \\nsubjects were given a second one where they were asked: (a) to complete the \\n16  matrices again, recording their previous choice in each, but explaining \\nwhy they had made it - since there were 1 6  different matrices, subjects had \\nto provide 1 6  reasons; this second measure was called justified choice; (b) to \\nrecord the choice they thought people of their group had made, ingroup \\nchoice; and (c) to record the choice they thought people of the other group \\nhad made, outgroup choice. \\nNext, we will present the results obtained for each of these measures. But \\nbefore we do so, we must stress that the effect of categorization, in the \\nabsence of a control group, is evident for every measure in the prevalence of \\ndiscriminating over non-discriminating strategies. Thus, in the first choice \\nonly the fairness/equity strategy in Tajfel's matrices is higher than discrimi­\\nnating ones: 6.87. But both MIP over MJP (3.89), and MD (3.47), as well as \\nMIP over fairness (3.01) are higher than MJP (.30) and MJP + MIP (.47). \\nSimilar results are found in the other three measures of the dependent \\nvariable. Insko's matrices show a similar pattern, with the exception of, \\nonly, the strategy max. joint own, which is as high as the discriminating \\nstrategies max. own and max. reI. own. \\nResults regarding first choice \\nThe MANOV \\nA carried out on Tajfel's matrices yielded significant values. \\nThe univariate Fs show that individualists are lower than collectivists in \\n\\n\\n206 \\nSocial identity: international pers\\npectives \\nMIP over fairness, 1 .60 vs 4.44, F(I ,  89) = 6.98, p < .01 ,  which is evidence \\nof their lesser degree of discrimination. At the same time, they are higher in \\nfairness, 7.69 vs 6.04, F (1, 89) \\n= 2.00, p < . 1 6. MANOVA was not \\nsignificant for Insko's matrices. \\nHowever, in these latter matrices differences do appear when relative \\npreferences of individualists and collectivists are compared by means of the \\nWilcoxon test. Thus, we find that individualists are different from collec­\\ntivists in that they prefer the strategy of fairness (min. diff.) over max. reI. \\nown and max. joint own; also in their preference for max. joint own over \\nmax. own. Collectivists, on the other hand, choose max. reI. own more \\noften than max. joint own. All this converges in a description of indi­\\nvidualists as more prone than collectivists to fairness and to joint and non­\\ndiscriminating strategies, bearing in mind that this differential tendency is \\nfound within a discrimination framework shared by both kinds of subjects, \\nwhich explains the lack of differences in the MANOV \\nA. \\nResults regarding justified choice \\nThe MANOV \\nA did not yield any significant differences, either in Tajfel's \\nmatrices or in Insko's ones. \\nAs we have mentioned, after subjects had completed the matrices, they \\nwere asked a number of questions. The first one was about the reason why \\nthey had chosen a particular strategy in allocating points between two \\npersons. More specifically, the question was: 'Why did you choose that cell?' \\nEach subject had to record the reason corresponding to each cell. Since \\nthese amounted to 1 6, each subject provided 1 6 reasons to explain his/her \\nchoice-allocating strategies. \\nAfter analysing the reasons given by all subjects, we found they could be \\ngrouped into 10 main categories: \\n1 \\nFairness or equality \\n34% \\n2 \\nEquality to balance \\n4.2% \\n3 \\nEquality when it is the only choice not detrimental to one's \\ngroup \\n0.58% \\n4 \\nMaximum ingroup profit \\n24.8% \\n5 \\nMaximum ingroup difference \\n8.53% \\n6 \\nMaximum ingroup joint profit or maximum joint own \\n1 5.31% \\n7 \\nMaximum joint other \\n3.2% \\n8 \\nMaximum outgroup difference \\n0.77% \\n9 \\nOthers \\n0.84% \\n10 Chance \\n7% \\nSix of these reasons correspond exactly to the strategies of matrices. These \\nare f\\nairness or equality (F) (min. diff.), maximum ingroup pro\\nfit (MIP) (max. \\nown), maximum ingroup dif\\nf\\nerence (MD) (max. reI. own), maximum ingroup \\njoint pro\\nfit (MJP) (max. joint own), maximum joint other (max. joint other) \\nand maximum outgroup diff\\nerence (max. reI. other). \\n\\n\\nDiscrimination and belie\\nf\\ns\\n on discrimination \\n207 \\nThese six strategies tend to be the most often used by subjects. In fact, a \\nhigh percentage of them repeat the reason given for the first cell in the \\nremaining 1 5, that is, does not show 'alternation' in its reasons. In these \\ncases, which we have called 'Use of Single Strategy', or USS, subjects tend \\nto use, almost always, some of these six reasons. \\nOn the other hand, reasons 2, 3 and 9, respectively categorized as \\n'equality to balance', 'equality when it is the only choice not detrimental to \\none's group' and 'others', are subsidiary to the former ones. This means that \\nthey are used only in cases of 'alternation' of strategies, or SA, where \\nsubjects do not want to limit themselves to only one reason and resort to \\nother complementary ones. \\nLet us see an example. Subject 34 uses a perfect alternation of reasons. \\nOn eight occasions he employs reason 4 (MIP) (max. own), which is dis­\\ncriminatory. In the remaining eight, he uses reason 2 (equality to balance), \\nwhich is egalitarian. This subject's words are worth being cited. When \\nhe chooses reason 4, he justifies his choice as follows: 'To give points to the \\npeople having the same code as myself'. When he chooses reason 2, he \\nstates: 'In order not to increase too much the difference in favour of my \\ngroup.' \\nIn the case of this subject, as well as in other similar ones, reason 2 is \\nclearly subsidiary to reason 4. The subject wants to benefit his/her group \\nmore, but does not want to do so in an extreme or radical way. Therefore, \\nthis justice or equality pursued arises from a wish to compensate a previous \\ndiscrimination. \\nReason 3 is even more subsidiary than reason 2. It is usually employed by \\nsubjects with a discriminatory orientation when they meet a Tajfel matrix in \\nwhich they cannot benefit their group more than the opposite group. This \\nis the case of subject 39, who shows a discriminatory pattern of reasons, \\nchoosing 1 5  times reason 4 (MIP) (max. own), and only once reason 3. In \\njustifying his discriminating strategy, this subject states: 'To give more \\npoints to my group.' However, when choosing the equality strategy in \\nTajfe\\\\'s last matrix, he says: 'To give as few points as possible to the other \\ngroup.' Effectively, in this latter matrix it is not possible to discriminate in \\nfavour of one's own group. \\nIt is clear, then, that reason 3 justifies equality in a subsidiary way, as a \\nsort of lesser evil to prevent one's own group from getting less points than \\nthe other. \\nThe last subsidiary reason is 9, which we have called 'others'. Sometimes, \\nsubjects look for justice allocating one group more points in one cell and \\nless in the next one, since they think that a strict equality strategy will be \\ndetrimental to both groups in rendering their whole absolute profit lower. \\nThus, subject 1 8  states: 'I have tried to balance the two groups. Since \\ncompensation is sometimes difficult, it is rather a random criterion.' \\nReason 1 0, 'chance', embodies those reasons not related to the strategy \\nchosen. Five subjects used only this reason and were removed from the \\nanalyses, as was another who chose it most of the times. \\n\\n\\n208 \\nSocial identity: international pers\\npectives \\nNevertheless, the great majority of reasons (82.64%) were distributed \\namong the four main categories, which closely coincided with the main \\nallocating strategies. These four reasons are fairness (34%), maximum \\ningroup profit (24.8%), maximum difference (8.53%) and maximum joint \\nprofit (1 5.3%). \\nAlthough each subject had to justify 1 6 different choices, a thorough \\nanalysis of the pattern of justifications showed that most subjects used only \\none reason for all 1 6  occasions or alternated between two. There were no \\ndifferences between individualists and collectivists in the regularity of use of \\njustification patterns: 74.4% of the former and 73.3% of the latter used one \\nor two reasons to justify their respective choices. \\nWhen only the justifications of these 'regular' subjects are considered, the \\nfour above-mentioned categories explained 96.5% of the reasons used \\n(instead of 82.64% which appeared when all subjects were computed). Dis­\\ncriminatory strategies, on the one hand (maximum ingroup profit and \\nmaximum difference), and non-discriminatory ones, on the other (fairness \\nand maximum joint profit), were added up. This calculation led to the \\ndiscovery that in individualists the use of discriminatory reasons (28.1%) is \\nless than the use of non-discriminatory reasons (69.2%), a difference which \\nis significant. However, in collectivists, though the use of discriminatory \\nreasons is also less (42% and 54%, respectively), the difference is far from \\nbeing significant. \\nIngroup choice \\nBy means of the Wilcoxon test we tried to find out whether there were \\ndifferences between individualists and collectivists in their preferences for \\nany strategy in the various choices. In other words, when individualists \\nproceed from first choice and justified choice to ingroup choice, do they \\nchange their preferences as compared with collectivists? If this is true, the \\nconclusion would be that individualists see ingroup members in a different \\nway from how collectivists see members of their own ingroup. \\nAccepting p < .05 as a criterion of statistical significance, we found that, \\nin effect, there were strong differences between collectivists and individual­\\nists. They certainly coincide in thinking that members of their own group use \\nthe fairness/equity strategy of Tajfel's matrices less often than themselves, \\nwhich means they consider themselves as fairer than their ingroup's mem­\\nbers. But individualists, as opposed to collectivists, think that ingroup mates \\nare higher than themselves in the three Tajfel discriminatory strategies, \\nnamely MIP over MJP, MD, and MIP over fairness. \\nIn Insko's matrices something similar happens. Collectivists consider \\nthemselves as more just than in \\ngroup members. And so do individualists. \\nMoreover, collectivists see themselves as less discriminatory than ingroup \\nmembers in max. reI. own. And so do individualists. But these think they \\nare less discriminatory than ingroup members in max. own, and here they \\nare opposed to collectivists. \\n\\n\\nDiscrimination and belie\\nf\\ns\\n on discrimination \\n209 \\nOutgroup choice \\nAs in the previous choice, the Wilcoxon test was employed to look for \\ndifferences between individualists and collectivists in the use of strategies in \\nthis choice as compared to the two fonner ones. \\nIn Tajfel's matrices, results are exactly the same as in the previous choice. \\nCollectivists consider themselves as more prone to fairness (F) than \\noutgroup members. But collectivists do not think outgroup members to be \\nmore discriminatory than themselves. Individualists, however, do think so \\nin the following strategies: MIP over MJP, MD, and MIP over F. \\nResults are also very similar in Insko's matrices. Both collectivists and \\nindividualists consider themselves as more fair than outgroup members and \\nless discriminatory in max. reI. own. But individualists, as opposed to \\ncollectivists, think they are less discriminatory than outgroup members in \\nmax. own. \\nFirst choice-justified choice and ingroup choice-outgroup choice \\ncomparisons \\nConsidering the six Tajfel strategies and the seven Insko ones, as well as the \\ntwo groups, individualists and collectivists, a total of 26 first choice-justified \\nchoice comparisons and 26 ingroup choice-outgroup choice comparisons \\nwere carried out. None of them yielded significant differences, which, in \\nview of the previous results, suggests that subjects are very homogeneous \\nwhen answering for themselves (whether having to justify their answer or \\nnot) and when they answer in the place of others (be these from in \\ngroup or \\nfrom outgroup). \\nConclusion \\nTwo conclusions can be drawn from the results of our two experiments. \\nFirst, both individualists and collectivists discriminate in favour of their \\nown group in a minimal group situation. As we have seen, there is a slight \\ntendency for individualists to be less discriminatory. However, on the \\nwhole, both kinds of subjects provide evidence of group fonnation, along \\nthe lines suggested by Insko and Schopler (1987). \\nSecond, as far as beliefs on discrimination are concerned, it seems clear \\nthat individualists consider their point allocation to be less discriminatory \\nthan that of collectivists. The reasons they provide to justify their second \\nchoice seem to imply that individualists interpret their allocating behaviour \\nas a search for justice and a wish to share with the other group. On a much \\nlower number of occasions they admit to being motivated by a wish to \\ndominate the opposite group. \\nMoreover, the divergence between individualists and collectivists appears \\nagain in the clear-cut distinction the fonner establish between themselves \\n\\n\\n210 \\nSocial identity: international perspectives \\nand people of ingroup and outgroup. Collectivists tend to consider them­\\nselves as more just in their comparisons to ingroup and outgroup, but this \\nresult can be easily explained as due to social desirability. Individualists also \\nsee themselves as more fair and, at the same time, as less discriminatory. \\nThis explicit and public acceptance of their non-following ingroup norm \\nseems to speak for a claim of independence vis-a.-vis the group, of a certain \\ncritical positioning against it, despite the fact that actual allocating \\nbehaviour is discriminatory and, as such, is a proof that the person feels as \\na member of the group to all purposes. \\nAll the aforementioned leads us to the conclusion already outlined in \\nprevious pages. It is not a matter of individualists experiencing insurmount­\\nable difficulties to form groups or identify themselves with them. Rather, \\nwhat happens is that they do so while trying to preserve their individual \\ninitiative, in this particular case restricted to beliefs. Of course, other inter­\\npretations of our data are possible, this allowing and, at the same time, \\nencouraging future research on a subject which is of interest to understand \\nprocesses mediating behaviour of people in groups and in society as a \\nwhole. \\nNote \\nThis chapter was translated from the Spanish by Elena Gaviria Stewart. \\n\\n\\n1 5  \\nConstructing Social Identity: The Role of \\nStatus, Collective Values, Collective Self­\\nEsteem, Perception and Social Behaviour \\nDaria Paez, Cristina Martinez-Taboada, Juan Jose Arr6spide, \\nPatricia Insua and Sabino Ayestaran \\nSocial identity theory (SIT) and Turner's self-categorization theory (SeT), a \\nmore cognitive version of the former, have stimulated research on group \\nprocesses and are currently some of the most heuristic theories in this \\ndomain. Nevertheless, some psychological aspects of SIT have not received \\nsystematic empirical support. The most important ones are the following: \\n(a) the belief in the relationship between salience of the categorization and \\nintergroup discrimination; (b) the relationship between strong identification \\nand perceptual and behavioural biases which favour the ingroup; (c) the \\nrelationship between discrimination and an increase in self-esteem (Messick \\n& Mackie, 1 989). These results suggest that the relationship put forward by \\nSIT, and also partly by SeT, between salience and favourable behaviour \\ntowards the ingroup, coupled with an increase in self-esteem, is not so \\ngeneral as was once thought. However, we must state that these criticisms \\nare valid only for a more individualistic approach to SIT and SeT. \\nFollowing Turner, Hogg, Oakes, Reicher and Wetherell (1987), psycho­\\nlogical processes of identification, self-esteem, discrimination, and so on, are \\nonly one aspect of intergroup behaviour. Socio-structural position and \\nsocially shared belie\\nf\\ns\\n on the legitimization of power/status differences and \\nindividual/collective strategies to cope with these social differences are two \\nmain explanatory processes of intergroup behaviour. From this perspective \\nit is a mistake to focus only on the intra \\npsychological relationships of \\nidentification, self-esteem and discrimination. \\nThis chapter will analyse certain themes of debate which have arisen with \\nrespect to the concept of social identity: When does it regulate a person's \\nbehaviour and perception? When does it produce a favourable perception of \\nthe in \\ngroup and discriminatory behaviour towards the outgroup? Hinkle, \\nBrown and Ely's (1992) revision of \\nthe literature has shown that there is not \\nalways an association between group identification and ingroup favouritism, \\nand that there is no consistent association between discrimination in favour \\nof one's group and increasing one's self-esteem. Hinkle et al. confirmed that \\n\\n\\n212 \\nSocial identity: international perspectives \\nsubjects who share collectivistic values are those who show more association \\nbetween identification and ingroup favouritism. Crocker and Luhtanen \\n(1990) have found that the variable linked to intergroup discrimination is \\ncollective and not personal self-esteem: it is those subjects with a high \\ncollective self-esteem who show a discriminatory bias in situations which \\nthreaten one's social identity. Abrams (1990, 1 994) has shown the moder­\\nating role of private self-consciousness and strong identification: those \\nsubjects with a strong group identification and high private self­\\nconsciousness are those who show a stronger behaviour regulation on the \\nbasis of one's social identity. We will present Spanish data which confirm \\nand specify the role of these moderating variables in the relationship \\nbetween social identity, perception and social behaviour. Finally low-status \\ngroups' socio-cognitive functioning does not always respond to the logic \\nderived from SIT's statement of the need to defend one's identity: instead of \\nshowing ethnocentrism subjects present centrism or outgroup favouritism. \\nTurner's perspective emphasizes the role of socially shared belie\\nf\\ns\\n and socio­\\nstructural position in order to explain these results. In our case we will focus \\non the role played by individualistic versus collectivistic values and in the \\nrole of collective self-esteem. This is why we will analyse social status and \\nthe predominance of the collectivistic cultural values as factors which will \\nmoderate the regulation of one's behaviour by social identity. \\nWe believe it is specially important to contrast the macro \\nsocial and \\ntranscultural validity of the above-mentioned theories. This is one of the \\nreasons why first of all we will devote special attention to the studies \\nconducted in Spanish and Latin American contexts on the issues of \\ncollective identity, stereotypes and discrimination (Morales & Paez, 1 996). \\nIndividualism-collectivism, status and the importance of personal and \\nsocial identity \\nSIT states that there is a personal-intergroup continuum. On the intergroup \\npole, subjects see themselves as interchangeable members of the category or \\ngroup. On the personal pole, subjects will be defined by their singularity and \\nindividual specificity. It is thought that the coping strategies which the \\nmembers of low- or high-status groups rely on will be located as opposites \\nin this dimension: on one extreme we will find individual mobility, on the \\nother side social competition. Nevertheless, research carried out in Latin \\nAmerica has shown that Chilean workers simultaneously agreed with social \\nconflict strategies and using strategies aimed towards individual improve­\\nment of their siblings' position by means of education (Quevedo, 1 972). \\nThis study did not analyse the correlation between the agreement with both \\nstrategies. Even though this correlation could be negative, and so confirm \\nthe hypothesis which states the opposite nature of both these strategies, the \\nfact that the average agreement with both strategies was high confirms that \\nthey may coexist on a group level. The opposition between the salience of \\n\\n\\nConstructing social identity \\n213 \\nidentification levels which have different inclusiveness (that is, regional and \\nsupra-national) has not been confinned as a general phenomenon. We \\nmust state that some subjects show strong regional identification and a low \\nidentification on a higher level. But other subjects simultaneously show a \\nstrong identification with different levels of inclusiveness (as was the case \\nof a simultaneously strong identification with the Polish and European \\nnational identities, Mlicki & Ellemers, 1996). Finally it is possible that in \\ncollectivistic cultures the difference between social and personal identity is \\nmuch less than in individualistic cultures. SIT posits that an important \\ndimension of identity is the social dimension. Nevertheless, the identity or \\nindividual self-concept may be more or less shaped by social identity. \\nThere is both a trait and situational variability with regard to the \\nintegration of social identity in personal identity or the fonner's relevance. \\nThe importance of social identity and its integration within personal \\nidentity will be mediated by the subjects and the historical context which \\ncontextualizes them. The importance of social identity is stronger in \\ncollectivistic and sociocentric cultures in comparison to individualistic and \\negocentric ones. \\nIndividualism is a set of beliefs, values and cultural practices in which \\npeople manifest an emotional independence from groups and organizations. \\nOn the other hand, collectivism as a cultural syndrome is associated with an \\nemotional dependence towards a group. A collectivistic culture is that in \\nwhich there is a stable relationship between the individual and certain \\ninfonnal groups to which one belongs, a subordination of individual objec­\\ntives to those of the collective, and an adjustment between the person's and \\nthe group's objectives. There is a tendency towards not leaving the group \\neven though the requirements may be high. An individualistic culture is that \\nin which a person has negotiable relationships with various groups. S/he \\nmay abandon these groups when they are very demanding (or potentially \\nmay be so), and the individual's goals are more important than the group's \\ngoals (Morales, Lopez & Vega, 1992; Triandis, 1992, 1994; Triandis, \\nBontempo, Villareal, Asai & Lucca, 1988). Cooperation with the group, \\nsharing resources and spaces with members of certain stable ingroups, col­\\nlective achievements and a strong affective link all characterize collectivistic \\nsubjects. Interpersonal competition, having important private spaces and \\nresources, individual achievements and distancing oneself from the group \\nare all characteristics of an individualistic subject (Hinkle et aI., 1992). \\nFrom the point of view of the importance of social or collective identity \\nfor one's self-image or for the social representation of the individual (see \\nDoise, Chapter 2 in this volume), individualistic cultures stress a concept of \\nthe individual as a separate, autonomous, self-contained and independent \\nbeing. Collectivistic cultures emphasize a representation of the person as an \\nentity centred in his or her relationships with others, connected to, and \\ndependent on, the social context (Markus & Kitayama, 1991 \\nb). \\nThe group's social status also affects the degree to which individuals \\ndefine themselves in tenns of their social or personal identity. High-status \\n\\n\\n214 \\nSocial identity: international perspectives \\ngroups perceive themselves as a set of individuals. Low-status groups define \\nthemselves as an aggregate of subjects who share a series of similar \\ncharacteristics (as Lorenzi-Cioldi et al.'s studies have shown). Members of \\ndominant groups will define their identity mainly on the basis of personal \\nfactors, while members of dominated groups will define their self-concept on \\nthe basis of mainly social factors (Lorenzi-Cioldi & Doise, 1 994). As a \\nsummary we may state that dominated groups and collectivistic cultures will \\nreinforce the relevance of social identity over personal identity. On the other \\nhand, individualistic cultures and dominant groups will pay more attention \\nto, and have a better evaluation of, an individual who possesses resources \\nwhile also stressing the relevance of personal identity over social identity. \\nLow status, salience of the intergroup comparison, outgroup \\nfavouritism and group perception \\nCulture and status have an influence not only on the importance of the \\nsocial categories in the representation of one's self, but also in assessing \\nthe representation of group perception. Collectivistic cultures, which posi­\\ntively value the group, reinforce the perception of intragroup homogeneity, \\nwhile individualistic cultures, which positively value an autonomous subject, \\nreinforce the perception of ingroup variability (Devos, Comby & \\nDeschamps, 1 995). \\nFrom the point of view of intra- and intergroup social perception, we \\nmust remember that there is a tendency towards perceiving the ingroup as \\nbetter in certain attributes. For instance Rodriguez (1992) found that \\nsubjects from Venezuela evaluated themselves better than subjects from \\nColombia in four out of six positive attributes, whilst the Colombians did \\nthe same in five out of six attributes. \\nIn general, subjects differ positively in those dimensions which define the \\ningroup, although they may accept outgroup superiority in another set of \\ndimensions. This social perception which favours the in \\ngroup is modelled \\nby social status: high-status groups tend towards a stronger perceptive \\ndifferentiation in favour of the ingroup than those subjects who hold a \\nlower status, as Mullen, Brown and Smith's (1992) meta-analytic review \\nshowed. \\nIn relation to the variability of stereotypical attributes, the ingroup is seen \\nas more complex, variable and heterogeneous than the outgroup. There is a \\ntendency towards perceiving the outgroup as more homogeneous and the \\ningroup as more heterogeneous. As Mullen and Hu's (1989) meta-analytic \\nreview has confirmed, this effect is further stressed in the case of natural \\ngroups. Status also has an influence in the perception of variability: minority \\nor threatened groups will perceive their group as less variable. This is a way \\nof reinforcing its distinctiveness and cohesion. Results have also shown that \\ngroups perceive themselves as more homogeneous on those attributes or \\ndimensions which define them (Leyens, yzerbyt & Schadron, 1 994). \\n\\n\\nConstructing social identity \\n215 \\n'Cold' socio-cognitive theory (Leyens et ai., 1994) suggests that having \\nhad a direct experience with the group will lead to a more variable, \\nheterogeneous or individualized representation of the ingroup. \\nFrom an SIT perspective, cognitive categorization as a member of a \\ngroup, or the identification with this group, will stress intergroup differences \\nand intragroup similarities. Individualization or aggregation would be a \\nfunction of the defence of one's social identity, on the basis of the existing \\npower and status relationships. Social comparison will influence not only \\nthe perception of differences but also homogeneity. It will not always be the \\noutgroup who will be perceived as less variable, simple or homogeneous. \\nThe ingroup will be perceived as more homogeneous in those attributes \\nwhich define their social identity and in situations in which this identity is \\nunder threat (Leyens et ai., 1994; Marques & Paez, 1 994). \\nIn a study conducted by us we confirmed the predominance of this \\nnormative logic of identity defence, instead of a cognitive logic of experience \\n(see Marques, Paez and Abrams, Chapter 9, this volume). We asked \\npsychology students (N = 109) to evaluate the ingroup's (psychologists) and \\noutgroup's (engineers) variability in stereotypical attributes (Table 15. 1). \\nThese attributes had been generated in previous studies and suggested that \\nbeing friendly and sociable are typical traits awarded to psychologists, whilst \\nbeing competitive, intellectual and having a strong leadership capacity are \\ntypical of engineers. On the basis of these answers we calculated the Pd \\n(perceived difference) and Var (variability). The estimated variabilities are \\nsignificantly different, although not regularly higher in the ingroup. \\nVariability was higher in the outgroup on those attributes stereotypical of \\nthe ingroup and vice versa. In other words, psychologists perceived them­\\nselves as more homogeneously friendly and sociable in comparison to \\nengineers, while this last group was seen as more homogeneously competitive \\nand intellectual, although in this last attribute differences were smaller and \\nother studies have found that it is associated with the psychologists' \\nstereotype. In those variables in which we did not find significant differences, \\nmeans were in the hypothesized direction. Variability was smaller for \\nengineers in leadership and in being normative. \\nThe tendency towards accentuating ingroup favouritism on important or \\nconsensual dimensions seems an increase in competition or salience of the \\ncomparison conditions, as Doise and Sinclair's study (reanalysed by van \\nKnippenberg & Wilke, mentioned in Morales & Huici, 1994, p. 737) has \\nshown. Nevertheless in a study conducted in Brazil by Nascimento-Schulze \\n(1993) on bank clerks, a social comparison situation led to a worse image of \\nthe group than a situation in which there was no social comparison, and the \\nsame happened when the context made group membership more salient. As \\nwe can see, this is contrary to the fact that group salience should activate a \\nprocess of favouritism or defence of the group's image. The author \\ninterprets these results as showing that these groups of bank clerks in social \\ncomparison and salient conditions are aware of their problematic work \\nconditions and of their unstable status, which is reflected in a diminution of \\n\\n\\n216 \\nSocial identity: international pers\\npectives \\nTable 15.1 Perceived ingroup and outgroup heterogeneity \\nVariability means of \\nEngineers \\nPsychologists \\nAttributes \\nMean \\nSD \\nMean \\nSD \\nFriendly \\n0.88 \\n0.56 \\n0.68 \\n0.54 \\nCompetitive \\n0.60 \\n0.61 \\n0.82 \\n0.63 \\nSociable \\n0.8 1  \\n0.72 \\n0.56 \\n0.52 \\nIntellectual \\n0.64 \\n0.56 \\n0.76 \\n0.53 \\n** p < .01 ;  *** p < .001 \\nt (1 1 3) \\n3.86*** \\n-3.17** \\n4.03** \\n-2.72** \\ntheir self-image. This socio-cognitive functioning of groups with unstable \\nand relatively insecure status does not seem to follow an ethnocentric logic. \\nIn a conceptual replication of this research, we have confirmed that in \\nconditions of salience of the comparison subjects increase the superiority of \\nthe outgroup in consensual or stereotypical attributes. In order to manipu­\\nlate the salience of the comparison, we asked subjects enrolled in a group \\ndynamics course (N = 1 1 8) to perform a cognitive creativity task: to list the \\nattributes which characterize a good leadership. Subjects were randomly \\nassigned to a condition of activation of the comparison with the engineer \\noutgroup, or of ingroup comparison. All the subjects had to read a first set \\nof attributes. Depending on the condition, subjects were told that this list \\nhad been written by psychologists (ingroup) or engineers (outgroup). Apart \\nfrom performing the task, subjects also had to estimate ingroup and out­\\ngroup variability on a series of six attributes: leadership, friendly, competi­\\ntive, normative or followers of the norm, sociable and intellectual. On a \\nfive-point scale subjects had to decide, for instance, what percentage of \\nengineers and psychologists had no, some, regular, a high or an extremely \\nhigh intellectual capacity. This cognitive task allowed us to obtain vari­\\nability indices such as the Pd and the Var, whilst also allowing us to infer \\nthe group's mean on that attribute. All the subjects evaluated both psy­\\nchologists and engineers. The F of the between-group interaction with \\nregard to the group that was being evaluated (ingroup or outgroup) and the \\nsalience manipulation was significant (F( l ,  1 1 3) = 4.89; p < .05). Specifically \\npsychologists were attributed a much lower leadership average when the \\nsocial comparison was made salient. This reflects the realistic perception of \\nthe inferior nature, in terms of status, of the psychologists compared to \\nengineers (see Table 1 5.2). \\nIn order to confirm that previous results (psychologists perceived them­\\nselves as more homogeneously friendly and sociable in comparison to \\nengineers; the engineers group was seen as more homogeneously com­\\npetitive in comparison to psychologists) followed a normative or social \\ndesirability logic, we asked a sample of psychology students (N = 9), \\npsychologists and mental health workers (N = 9), and qualified workers, \\nincluding physicians and engineers, and non-qualified workers (N = 1 2) to \\n\\n\\nConstructing social identity \\n217 \\nTable 1 5.2 \\nPerception of ingroup and outgroup by salience versus non­\\nsalience o\\nf\\n social comparison \\nAttributes \\nLeaders \\nSalience of social comparison \\n('between' variable) \\nEngineers \\nPsychologists \\n('within' variable) \\nInferred mean \\nInferred mean \\n3.31 \\n2.71 \\nNon-salience \\n('between' variable) \\nEngineers \\nPsychologists \\n('within' variable) \\nInferred mean \\nInferred mean \\n3.47 \\n3.34 \\n'evaluate to which degree the following characteristics are desirable or \\npositive in each jobs' performance'. As we can see in Table 15.3 both the \\nengineers' higher status and the normative nature for psychologists on those \\nattributes in which there was a smaller variability (friendly, sociable) was \\nconfirmed. Exactly the opposite took place for the attribute 'competitive' \\n(less variability and it was attributed more to engineers). The only exception \\nwas found in the attribute 'intellectual', in which psychologists have a \\nhigher score and that in the first study showed more variability, although \\nthe difference in variability means was the smallest of all. In order to \\nconfirm that the attribution of the traits followed both a professional social \\ndesirability logic as well as a group identification logic, we correlated the \\nsubjects' level of identification with psychologists ('we ask you to indicate \\nyour identification with the professional category termed \\\"psychologists\\\"') \\nwith a composite rating of attributes (friendly, sociable, intellectual) of the \\ngroup. This composite score showed a satisfactory reliability (alpha = .67) \\nand a significant correlation of 0.35; p < .05 with one's identification with \\nthe group of psychologists. This identification was individually associated \\nwith believing that it was desirable that the psychologists should be friendly \\n(r = .36; p < .05), with sociable (r = .27; P < .10) and, although of marginal \\nsignificance, with intellectual (r = . 1 5; P < . 1 0). \\nOn the other hand, identification with psychologists was not associated \\nwith the agreement that psychologists were competitive and normative (r = \\n< .09). However, those features were typical of the engineers. Identification \\nwith psychologists was associated with the leadership attribute (r = .42; P < \\n.05). Bearing in mind one's professional experience, in comparison to \\nstudents, psychologists and mental health experts stated that it was more \\nsocially desirable for a psychologist to have leadership capacity (mean = \\n5.1 1), and both these groups believed in this fact more than did qualified \\nand non-qualified workers (mean = 3.75; F(2, 27) = 4.07; P < .03). Group \\nidentification was linked with a stronger attribution to this group of the \\npositive attributes, especially when these define their professional identity. \\nMoreover, even when the task was a comparison with a group of engineers, \\nthose psychologists and mental health experts more professionally involved \\nattributed positive traits (e.g. leadership) to the ingroup, though it has been \\nshown that these traits tend to be associated with high status groups. In \\n\\n\\n218 \\nSocial identity: international pers\\npectives \\nTable 1 5.3 Perceived social desirability o\\nf\\n attributes f\\nor engineers and \\npsychologists \\nMeans of social desirability for \\nEngineers \\nPsychologists \\nAttributes \\nMean \\nSD \\nMean \\nSD \\nFriendly \\n3.70 \\n1 .61 \\n5.90 \\n1 .26 \\nCompetitive \\n5.70 \\n0.95 \\n4.17 \\n1 .58 \\nSociable \\n4.67 \\n1 .40 \\n5.97 \\n1 .22 \\nIntellectual \\n5.00 \\n1 .19 \\n5.87 \\n0.90 \\nLeadership \\n5.80 \\n1 .3 1  \\n4.30 \\n1 .31 \\nNormative \\n5.50 \\n1 .07 \\n4.70 \\n1 .58 \\nMeans of group status for \\nEngineers \\nPsychologists \\nStatus \\n6.27 \\n0.69 \\n4.00 \\n1 .08 \\nI \\n= not at all; 7 = very much \\n* p < .05; ** p < .01 ;  *** = p < .001 \\nt (29) \\n-7.05*** \\n4.63*** \\n-4.33*** \\n-3.98** \\n3.84** \\n2.68** \\n1 1 .84*** \\nsummary we may state that the lower variability traits associated with \\nthe psychologists are in general those which are seen as socially desirable \\nfor the job of being a psychologist. These traits are also associated with the \\nidentification with this group and involvement in one's professional activity \\n(see Table 1 5.3). \\nWe may conclude that a certain relative ingroup bias is present in social \\nperception and that it follows a normative logic - of defence of a positive \\nsocial identity by means of stressing the positive aspects of the ingroup in \\nthose attributes used to define the group. Moreover, in subordinated groups \\nthis tendency is not shown in social comparison situations. But this does not \\nimply that subordinated groups do not show in general outgroup favour­\\nitism. Bourhis, Gagnon and MOise (1994) have shown how members of \\ndominated groups choose and evaluate more positively dominant outgroup \\nsymbols. In studies conducted in Latin America, different authors have \\nfound that the Venezuelan ingroup was evaluated worse than the European \\n(Italian) and North American outgroups (Montero, 1987, 1990). Even \\nthough the subjects identified themselves with the ingroup (Venezuela), \\nsubjects had a better evaluation of the European groups (England and \\nSpain) and of certain American nations such as Argentina, although they \\nhad a bad image of other American countries such as Colombia. Due to the \\nresults obtained from similar psychosocial, anthropological and historical \\nstudies, which show that in Latin America, Asia and Africa there is a \\ndevaluation of the national ingroup and favouritism towards First World \\noutgroups, the concept of Altercentrism has been coined. Altercentrism is \\nan important theoretical concept as it allows us to pay attention to a basic \\nproblematic found in SIT: how do we explain the socio-cognitive func-\\n\\n\\nConstructing social identity \\n219 \\ntioning of \\nthe dominated groups who do not mobilize themselves, or cannot \\nbe individually included in higher status groups? \\nVarious explanatory processes put forward by SIT suggest how domi­\\nnated groups manage their ambivalent image, showing at the same time \\nfavouritism towards dominant groups. \\nIntragroup social comparison with more disadvantaged subsectors, or, \\nalternatively, an advantaged social comparison with a group of similar and \\npertinent attributes, would be another mechanism which would allow the \\nshowing of favouritism towards the dominant outgroups while at the same \\ntime retaining a partly positive image of the ingroup. Members of sub­\\nordinated ingroups (that is, Venezuelans) accept the superiority of dominant \\noutgroups (that is, North Americans), although at the same time they stress \\ntheir superiority over other groups who hold a similar position to theirs and \\nwhom they undervalue (that is, Colombians) (Montero, 1 987; Salazar, \\n1989). \\nAnother phenomenon that characterizes dominated groups is a type of \\nsymbolic individual mobility - of which Bovarism and the cosmopolitanism \\nof the dominant classes of Latin America is a good example. Bovarism \\n(permanently believing one is living in Paris, London or New York, \\nalthough one is really living in Puerto Hundido) is a form of symbolic and \\nideal identification with the dominant outgroup. Latin Americans, \\nespecially, although not only, those from dominant classes, view themselves \\nas second-class or 'transplanted' Europeans or North Americans. The use of \\nsurnames (for instance, the large amount of people called Johnny, William \\nor Peter Gonzalez), adopting customs and identifying with the values of the \\ndominant cultures constitute a coping strategy of the accepted negative \\nnational identity. Studies carried out in developed countries have shown \\nthat this symbolic individual mobility strategy takes place even though there \\nare no real possibilities of integrating oneself into the dominant group \\n(Wright, Taylor & Moghaddam, 1990, mentioned in Bourhis et a!., 1 994). \\nFinally, low-status or dominated groups use cognitive creativity, redefin­\\ning as important a series of new dimensions in which the ingroup is superior \\n(see Smith & Bond, 1 993). Research conducted by Salazar et a!. on national \\nstereotypes in Latin America has confirmed the existence of a status/ \\ncompetence dimension, a sociability dimension and a third, less important, \\neducation/instruction factor. There is evidence to suggest that the instru­\\nmental or status/power dimensions and the expressiveness or sociability \\nones are transculturally valid dimensions of social perception (Zebrowitz, \\n1990). Various studies have shown that subordinated groups accept the \\nsuperiority of dominant groups on the competence/ability dimensions, or on \\nstatus, power and resources, but they view themselves as superior in the \\nexpressive or sociability dimensions. For instance, nurses with a high status \\nviewed themselves as superior in technical knowledge, while lower status \\nnurses perceived themselves as being more friendly (Smith & Bond, 1993, p. \\n62). Studies conducted in Latin America confirm that the self-stereotypes \\nreaffirm a positive evaluation of sociability attributes: attributes stereo-\\n\\n\\n220 \\nSocial identity: international pers\\npectives \\nTable 1 5.4 Attributes f\\nor engineers and psychologists \\nEngineers \\nPsychologists \\nAttributes \\nInferred mean \\nSD \\nInferred mean \\nSD \\nLeadership \\n3.40 \\n0.83 \\n3.07 \\n0.80 \\n3.13* \\nFriendly \\n2.98 \\n0.66 \\n3.83 \\n0.77 \\n-9.08** \\nCompetitive \\n4. 1 1  \\n0.70 \\n2.78 \\n0.80 \\n14.40** \\nNormative \\n3.56 \\n0.90 \\n2.69 \\n0.84 \\n7.57** \\nSociable \\n3.00 \\n0.66 \\n3.96 \\n0.56 \\n-1 1 .75** \\nIntellectual \\n3.42 \\n0.83 \\n3.27 \\n0.67 \\n6.49** \\n* = p < .01; ** = p < .001 \\ntypically favourable to the Latin American national groups were of the \\nsociable type (that is, happy), while the stereotypical positive attributes \\ntowards European and North American groups were of a more instru­\\nmental nature (that is, practical, workers and so on) (Rodriguez, 1992). \\nOur own research conducted in Spain on psychologists and engineers \\nconfirmed that the perceptive superiority of a lower status ingroup (psycho­\\nlogists) was found on the expressive dimension (they viewed themselves as \\nmore friendly and sociable), whilst they also accepted the superiority of the \\nhigher status outgroup on more instrumental attributes. They perceived the \\nengineers as more competitive, intellectual, and so on (see Table 1 5.4). \\nThe above-mentioned results may be interpreted as the effect of a \\nnormative process, in the sense that subjects try to defend their group by \\nperceiving it as an entity which adjusts itself to the norms and values which \\nrelevantly define its subculture more than the outgroup does (Marques & \\nPaez, 1994; also see Devos et al., 1995). Nevertheless, although low-status \\ngroup members may use symbolic individual mobility, comparison with \\ndisadvantaged groups and cognitive creativeness in order to explain one's \\nambivalent social image, according to the psychological version of SIT, low \\nstatus must have an influence on one's self-esteem. \\nStatus, self-esteem, collectivistic values and social identity \\nStatus is an important variable for SIT due to the fact that it is associated \\nwith the prestige and self-esteem one may infer from one's group \\nmembership. According to SIT, belonging to a social category or group is a \\nsource of self-esteem. Belonging to low-status groups, which possess nega­\\ntive attributes, would motivate a search for individual mobility, cognitive \\ncreativity and social competition. What is implicit in this approach is the \\nidea that subjects from disadvantaged classes and groups should suffer a \\ndeficit in self-esteem. This idea is included within the hypothesis stating that \\na deficit or threat to self-esteem would be an antecedent or cause of the \\ningroup's favourable social perception, of ethnocentric attitudinal prejudice \\nand behavioural discrimination. Personal self-esteem as an evaluative \\n\\n\\nConstructing social identity \\n221 \\ncomponent of personal identity may be differentiated from self-esteem \\nreferring to group belonging or being a part of a certain social category, in \\nother words collective self-esteem. \\nOne answer to the fact that the salience of the categorization does not \\ndirectly lead to differentiation, or that it leads to an increase in self-esteem, \\nis methodological. The relationship between discrimination or intergroup \\ndifferentiation and self-esteem should take place and be proven not on the \\nbasis of personal self-esteem, but on the basis of collective self-esteem, in \\nother words on the basis of the favourable evaluation of the category or \\nsocial group to which one belongs. Nascimento-Schulze's (1993) results with \\nBrazilian workers confirm that there is no relationship between intergroup \\ndiscrimination and an increase in general personal self-esteem. Intergroup \\ndiscrimination produced an increase in the group's evaluation or collective \\nself-esteem. In this sense, the development of collective self-esteem scales is \\na step forward in order to better understand the relationship between social \\nidentity and social behaviour. Crocker and Luhtanen, 1990; Luhtanen & \\nCrocker, 1992) have developed a scale which measures stable individual \\ndifferences in the level of a person's positive social identity (high collective \\nself-esteem) and also a positive personal identity. These authors developed a \\nscale with a private collective self-esteem dimension, a public collective self­\\nesteem dimension, the feeling of being a member of the group, and the \\nimportance of this identity. These scales had structural validity and alpha \\ncoefficients which scored higher than . \\n73 (in fact the public self-esteem scale \\nreached a score of .80). The private collective self-esteem scale and the one \\nreferring to the importance of being a member of the group, which meas­\\nures the most individual aspects of social identity, also correlated signifi­\\ncantly with personal self-esteem scales. The collective self-esteem scores \\nwere positively and coherently associated with the agreement with collective \\nvalues - measured using Hui's individualism-collectivism scale (Luhtanen \\n& Crocker, 1992). \\nOur results with the Spanish translation of this scale also show satis­\\nfactory results in relation to internal consistency. Alpha coefficients were \\nsatisfactory for the private and public collective self-esteem scales, identi­\\nfication with the ingroup and belonging to this same group (Cronbach \\nalphas of .83, .78, .66 and .71, respectively). Moreover, they also showed \\nconvergent validity with Hui's collectivistic values scale (see below). \\nExperimental research found that, using the above-mentioned scale, \\nsubjects high in collective self-esteem, contrary to those with low collective \\nself-esteem, showed more positive biases in relation to their group in \\nsituations in which there was a threat to their group identity. In situations of \\nfailure, subjects with a high collective identity changed their scores in such a \\nway that they mitigated the threat to the group's worth. Personal self-esteem \\ndid not play a regulating role in the answers (Crocker & Luhtanen, 1990). \\nThese results show that we have to differentiate between the need for \\nindividual and collective esteem. Moreover, they also suggest that the SIT \\nstatement referring to the tendency for discriminatory biases which favour \\n\\n\\n222 \\nSocial identity: international pers\\npectives \\nthe ingroup with the aim of obtaining a positive social identity is valid only \\nfor those subjects with a stable tendency towards positively evaluating social \\nidentity or socially sharing a positive definition of collective self. \\nStudies conducted on relative deprivation and political participation \\nreinforce the idea that it is the threat to collective self-esteem, and not the \\nlack of personal self-esteem, which is associated with the regulation of \\nperception and social conduct by social identity. Let us remember that \\ndeprivation or personal dissatisfaction is not associated with socio-political \\nparticipation behaviours, which we may view as examples of social conflict, \\nbut that in fact it is associated with anomie and individual affective \\nalterations. It is sociocentric or fraternal relative deprivation (in relation to \\nthe group) which is linked to political participation (Guimond & Tougas, \\n1 994). On the other hand, research on political participation shows that it is \\nthose subjects who are socially integrated, with a perception of political self­\\nefficiency, but who are also dissidents or critical of the social system which \\nthey view as the cause of the group's disadvantaged position, who mobilize \\nthemselves - and not those subjects who are most isolated, deprived or \\ndisadvantaged (Kinder & Sears, 1985). This is why Crocker, Luhtanen, \\nBlaine and Broadnax (1994) have found that it is those subjects with high \\ncollective (socially shared beliefs or collective affective evaluation with \\nregard to the group and not the person) self-esteem (those who are integrated \\nand efficient) who show more ingroup bias in threatening conditions (in \\nother words, of relative deprivation in which the group does not receive what \\nit expected). As we can see, those subjects who regulate their perception and \\nbehaviour on the basis of their social identity and those who show ingroup \\nfavouritism and outgroup discrimination are the ones who have high col­\\nlective self-esteem in threatening situations. \\nThe differentiation between social and personal identity as well as the \\nrelative independence of public, private and personal collective self-esteem \\nare two mechanisms which may explain why members of dominated and \\nstigmatized classes do not show a deficit in personal self-esteem (Phinney, \\n1 990). \\nAn explanatory process for the lack of deficit in personal and collective \\nself-esteem in a dominated group included in an individualistic culture is the \\nfrequently found phenomenon of the acceptance of the disadvantaged \\nnature of the group, while at the same time rejecting its individual dis­\\nadvantage (Crosby, Cordova & Jaskar, 1994). For instance, Puerto Rican \\nsubjects accepted that Puerto Ricans are dependent, but they themselves are \\nnot like that. Personal self-valuation was mainly positive, while collective \\nself-valuation was mainly negative (Rivera, 1991). As Rivera (1991) stated: \\nThe results of this first study were the pronounced and consistent divergence \\nbetween the collective and personal valuation; the latter was invariably higher than \\nthe collective valuation. These results suggest that participants may have inter­\\nnalized, to some degree, the Puerto Ricans' negative stereotype, depersonalizing \\nthese categories in order to show that the rest of \\nthe Puerto Ricans are that way, but \\nnot themselves, who are attributed positive personal valuations. (pp. 10 1-102) \\n\\n\\nConstructing social identity \\n223 \\nAnother explanatory process for the lack of low personal self-esteem in \\nmembers of dominated groups is that subjects will not link other people's \\nopinions about their group with their own personal self-esteem. Crocker et \\nal. (1994) have found that the relationship between public (how the subject \\nbelieves that other people evaluate hislher group) and private collective self­\\nesteem (how the subject evaluates hislher own group) is medium in \\nindividualistic cultures (Anglo-Americans), positive and strong in collecti­\\nvistic cultures (Asian-Americans) and null in dominated groups (African­\\nAmericans). Let us add that private collective self-esteem is strongly linked \\nin general with personal self-esteem. Nevertheless, in the sample of black \\nsubjects there was no relationship between public and private collective self­\\nesteem. These black subjects strongly differentiate how they believe other \\npeople value their group and how they themselves do so - and probably \\nhow they value themselves as individuals. In this way it is possible to cope \\nwith discrimination and prejudice separating collective negative judgements \\nfrom one's image of oneself. On the other hand, for Asian subjects the \\nvision the other person has is strongly associated with one's own view of the \\ningroup. This suggests that the relationship between personal self-esteem \\nand group self-esteem is more underlined in collectivistic cultures. \\nSelf-consciousness, individualistic-collectivistic cultural values, social \\nidentity and behaviour regulation \\nAs Abrams (1990) has stated, sometimes behaviour is not regulated by \\nsocial identity. Categorization in a group and salience of group membership \\nwill not be sufficient conditions for ingroup discrimination and its function \\nof augmenting self-esteem. Moreover, a person must strongly identify with \\nthe group. Abrams (1990, 1 994) has shown in different studies that group \\nidentification and self-consciousness or focalizing attention on the private \\nand public aspects of one's self converge in order to regulate behaviour. \\nSubjects with strong identification and high private self-consciousness \\nregulated behaviour by means of social identity. \\nThere is correlational evidence which confirms that high identification with \\na group is associated with a good evaluation of this same group. Garza and \\nHerringer (1986) developed an instrument which measured important \\ndimensions in group identity, the evaluation of the group, its emotional \\nvaluation and the stability of group identity. In a replication of this study, we \\napplied the Spanish version of this scale (Paez, Arrospide, Martinez-Taboada \\n& Ayestaran, 1 992). Our sample was composed of 120 subjects, 46% were \\nmen and 54% women with a mean age of 24.7 years. Forty-seven subjects \\nwere students who were starting their participation in a group dynamics \\ncourse, and the rest of the sample were relatives of fifth-year psychology \\nstudents enrolled in the Psychology Faculty of the Basque Country \\nUniversity. Students answered the questionnaires in class while the relatives \\nwere interviewed at home by the students as part of their class curricula. \\n\\n\\n224 \\nSocial identity: international perspectives \\nTable 1 5.5 Factorial analysis o\\nf\\n the 'most important social identity scale' \\n(Garza & Herringer, 1986) \\nFactor \\n2 \\n3 \\nEV Good-bad \\n.88 \\nEV Likes-dislikes \\n.86 \\nEV Positive-negative \\n.87 \\nEM Pleasant-unpleasant \\n.83 \\nEM Comfortable-uncomfortable \\n.80 \\n1M Important-unimportant \\n.67 \\n1M Valid-invalid \\n.80 \\nST \\nPermanent-impermanent \\n.69 \\nST \\nStable-unstable \\n.75 \\nST \\nSecure-insecure \\n.81 \\nST \\nCooperative-competitive \\n.64 \\nST \\nDominant-dominated \\n.81 \\nST \\nWith power-without power \\n.84 \\nEM Amusing-boring \\n.70 \\nST Abstract \\n-concrete \\n.77 \\nST \\nPrivate-public \\n.67 \\nEV = Evaluative dimension; EM = Emotional dimension; 1M = Importance; ST = Stability \\nWe found three clearly differentiated factors in this scale (see Table 1 5.5). \\nThe first factor explaining 47% of the variance was saturated by the four \\ndimensions mentioned by Garza and Herringer (1986): emotion, evaluation, \\nimportance and stability, and also by item number 1 3, which refers to \\ncompetition-cooperation. The second factor was composed of items related \\nto power and dominance, and they explained 9.4% of the variance. The \\nthird factor included those items referring to the abstract versus the concrete \\nand the private versus the public. This final factor explained 8% of the \\nvariance. \\nThese results reaffirm that the evaluation of the group, the emotional \\nresponse to the group, the importance of the group for identity and the \\nstability of the identification with a group are congruently associated. With \\nregard to the relationship between identification and stereotypes, we \\nhave found that in Latin America those subjects who strongly identified with \\ntheir national groups (samples from Venezuela, Chile and Colombia) \\nrejected sloth, while stressing happiness as a description of a Latin American \\nperson when answering closed items. With regard to open answers, these \\nsamples underlined characteristics associated with the Latin American's \\npositive sociability: happy, hospitable. Moreover, they mentioned history \\nand socio-political processes such as oppression and dependency as identity \\nelements. Those subjects who shared a low identification mentioned more \\nfrequently deficitary instrumental attributes, such as sloth and irresponsi­\\nbility, and they also made more mention of underdevelopment, poverty and \\nlanguage as identity factors (Salazar, 1 989). In other words, those people \\nwith a strong identification show perceptive favouritism stressing the \\nstereotype's positive attributes, relativizing the negative ones, while those \\n\\n\\nConstructing social identity \\n225 \\nsubjects who did not identify share the stereotypical negative attributes of \\ntheir social category. \\nAs a summary, we may state that group identification and focalizing \\nattention will interact in order to regulate behaviour. It is easy to believe \\nthat it will be those individuals with a high private self-consciousness and \\ncollective identity who will be more stable in their beliefs and behaviours \\nlinked to obtaining group goals. They will also be more emotionally \\nreactive to stimuli such as group threats. \\nOf course not only situational factors or internal tendencies of group \\nidentity salience and attention directed towards aspects of collective identity \\ninfluence behaviour regulation. Macrosocial phenomena such as collecti­\\nvistic orientation cultures reinforce the importance of collective or social \\nidentity, while individualistic cultures reinforce personal or individual \\nidentity, as we have already seen (Gudykunst, 1988; Smith & Bond, 1 993). \\nCollectivistic cultures emphasize an intensive relationship with ingroup \\nmembers, while in individualistic cultures people behave in a more extensive \\nthan intensive fashion - even with the ingroup members. Consequently, the \\ndifference between behaviour towards the outgroup and behaviour towards \\nthe ingroup is not so important (Triandis, 1994). There is partial evidence \\nwhich confirms that individuals from collectivistic cultures are more sensible \\nto outgroup and ingroup membership in the regulation of their behaviour \\n(Triandis, 1994). People from collectivistic countries show a stronger ten­\\ndency towards cooperation, and people from individualistic-type countries \\ntend more towards competition. Nevertheless, when they must interact with \\noutgroup members, collectivistic persons become more competitive. For \\nexample, Argentinean and Indian children who had to share game rewards \\nwith subjects who did not belong to their group were more competitive than \\nchildren from the United States (who in general are competitive) (Smith & \\nBond, 1 993). \\nIn the following experiment we empirically confirmed the importance of \\ncollectivistic values in behaviour and perception regulation. We asked \\nstudents (N = 48) enrolled in a group dynamics class to perform a cognitive \\ncreativeness task: to list the attributes which characterize good leadership. \\nAll the subjects had to read a first set of attributes supposedly previously \\ngenerated by a group. Depending on the condition, subjects were told that \\nthis list had been generated by psychologists (ingroup) or engineers \\n(outgroup). This first list and the one they created after 10 minutes of \\nindividual reflection and another 10 minutes of group discussion were \\nevaluated as: good-bad, complete-incomplete, unreal-realistic, and \\ntypical-not at all typical. The response scale ranged from I = the first \\nattribute of the pair to 7 = the second attribute of the set. For instance I = \\ngood, 7 = bad. \\nDifferences in scores between their own list and the previous one were \\ntaken as an index of favouritism. The difference was established between the \\nfirst score (the score given to the supposedly other group) and the second \\nscore (given to one's own group). The subtraction was performed in such a \\n\\n\\n226 \\nSocial identity: international perspectives \\nway that a positive difference meant a better evaluation of one's own group. \\nWe subtracted from the scale with highest negative scores (bad \\n= 7; \\nincomplete = 7; not at all typical = 7) the score given to the first set (other \\ngroup) from the score given to the second list (one's own group). For \\nexample, if the first list ( \\nother group) is considered bad = 7 and the second \\nlist (own group) is good = 1 ,  the positive difference of +6 was a sign of this \\nfavouritism. When the scale had a higher score and also a positive \\nevaluation (realistic = 7), the second score was subtracted (own group list) \\nfrom the first list. If the ingroup list was awarded a 7 (very realistic) and the \\nfirst list (other group) was awarded a 1 (not at all realistic), the subtraction \\n(7 - 1) would yield a positive score which would reflect ingroup favouritism. \\nIn other words, a positive score always meant ingroup favouritism. \\nThe scores in Crocker and Luhtanen's collective self-esteem scale, Hui's \\ncollectivistic values and Carver, Antoni and Scheier's ( 1 985) self­\\nconsciousness scale were used as covariables and their influence was \\ncontrolled. The variance and covariance multivariate analysis showed a \\nsignificant effect if the first comparison group was that of engineers or \\npsychologists. On a univariate level the comparison with an ingroup or \\noutgroup had a significant effect for the positive evaluation of the list (F(5, \\n46) = 1 3. 1 ; p < .001) and near significance for the complete list (F(5, 46) = \\n2.74; p < . 10). The other group's list was evaluated as better and more \\ncomplete when it was attributed to psychologists (ingroup), and was \\nevaluated as worse and less complete when it was attributed to the engineers \\n(outgroup). \\nThe private and public self-consciousness variables, as well as the group \\nidentification variables, did not interact with the experimental manipula­\\ntion. Nevertheless the individualistic/collectivistic values did. Subjects with a \\nhigh score in collectivism behaved in a different way depending on whether \\nthe first list was attributed to the outgroup or to the ingroup. The multi­\\nvariate interaction effect was significant (F( l l ,  38) = 2.33; P < .026). The \\nunivariate contrasts indicate that those subjects who share more collec­\\ntivistic values view the first list as more believable (F(1, 48) = 4.88; P < .03), \\nrealistic (F(1, 48) = 6.66; p < .02), complete (F( I ,  48) = 5.75; p < .02), and \\nless typical (F( l ,  48) = 1 1 .04; p < .002) when this list was attributed to \\npsychologists. In other words, when the stimulus was associated with the \\ngeneral ingroup and subjects were collectivistic (with regard to their values) \\nit was evaluated more positively and as more original. Exactly the opposite \\nhappened with subjects low on collectivism; these subjects viewed the list \\nthat was attributed to the outgroup (engineers) as less typical, more \\nbelievable, realistic and complete (see Table 1 5.6). \\nThe lack of self-consciousness and identity effects may be explained in the \\nfirst case because the scale measures the tendency to focus attention on the \\npublic and private aspects of the self - and not necessarily on the collective \\ndimension of the self. In the second case, the scale measures the general \\ntendency to identify with groups in general - and not with the group which \\nwas used in the manipulation (psychologists). Finally the low number of \\n\\n\\nConstructing social identity \\nTable 1 5.6 Dif\\nf\\nerences between engineers and psychologists as a \\nf\\nunction o\\nf\\n their degree of collectivism \\nFirst list \\nEngineers \\nPsychologists \\nDifferences in \\nLow \\nHigh \\nLow \\nHigh \\nevaluation \\ncollectivism \\ncollectivism \\ncollectivism \\ncollectivism \\nBelievable \\n-1 .30 \\n-0. 12 \\n-0. 17 \\n-\\\\ .00 \\nIncomplete \\n\\\\ .50 \\n0.25 \\n-0.80 \\n0.73 \\nRealistic \\n-1 .50 \\n-0.63 \\n-0. 18 \\n-\\\\ .80 \\nNot typical \\n- 1 .30 \\n0.60 \\n0.08 \\n-1.\\\\0 \\n227 \\nsubjects does not allow us to perform a contrast simultaneously choosing \\nsubjects with high group identification and high private self-consciousness. \\nThe above-mentioned research suggests that those subjects who show \\nmore ingroup bias in conditions in which the intergroup comparison is \\nsalient are those who share more the collectivistic values. Moreover, an index \\nof collectivistic values is associated with social identification, collective self­\\nesteem and public and private collective self-consciousness, confirming the \\nrelationship between these processes. \\nApplying the private and public self-consciousness items to the attention \\non social identity's external and internal aspects, we have developed a \\ncollective self-consciousness scale which we validated by comparing it to \\nGarza and Herringer's scale (Paez et aI., 1992). This scale was developed on \\nthe basis of the items extracted from the original private and public self­\\nconsciousness scale - referring to the most important group a person \\nbelongs to. The items and instructions were as follows: Think of the most \\nrelevant group to which you belong (family, friends, and so on). (I) I always \\ntry to analyse myself in relation to my group's objectives; (2) I am interested \\nin how I do things when I do them in my group's name; (3) I reflect a lot \\nupon myself in relation to my group's values and beliefs; (4) In relation to \\nmy group's objectives I am very careful with the image I give to other \\npeople; (5) I am well aware of the image I give in relation to my group; (6) I \\nconstantly think of the reasons for my behaviour in relation to my group's \\nobjectives; (7) I normally think of myself as a representative of my group; \\n(8) I am worried about the image my group gives; (9) When my group gets \\nwhat it wants I am happy; (10) I dream about myself many times in relation \\nto my group; (I I) I do not feel responsible or guilty if a member of my \\nclosest group (family) fails; (12) I do not feel responsible or guilty if a friend \\nfails; (13) When my group does not achieve what it wants I get very \\nannoyed; (14) When a member of my group achieves what s/he wants I feel \\nwell. Questions number 1 1  and 12 are to be inversely scored. The scale was \\npassed with a score of I = totally disagree to 7 = total agreement. In our \\nsample the reliability of our collective self-consciousness scale was .73. This \\ncollective self-consciousness scale had a satisfactory convergent validity - a \\ncorrelation of .39, p < .01 with Hui's individualism-collectivism scale; .22, \\n\\n\\n228 \\nSocial identity: international pers\\npectives \\np < .05 with Garza and Herringer's scale of group identification; and .21 , \\np < .05 with a scale of importance of group membership for one's self (Paez \\net ai., 1992). \\nThese results suggest that the chronic accessibility of the collective \\ndimension of the self, or, in other words, the stable focalization of internal \\nand external attention on the collective or group dimensions of identity, is \\ncoherently associated with a positive evaluation of collectivism, with a \\npositive identification with the groups and with a structure of one's self \\nbased on one's group membership. \\nConclusions \\nThis chapter has examined the conditions and processes by which social \\nidentity regulates social perception and behaviour. We have emphasized the \\nrole of socio-structural variables and of socially shared beliefs in order to \\nexplain this process. This is why we underlined the importance of social \\nstatus and the predominance of collectivistic cultural values as factors which \\nwill moderate the regulation of one's behaviour by social identity. \\nWe stressed the stronger influence which social identity has in collectivistic \\nand sociocentric cultures in comparison to individualistic or egocentric \\ncultures. We also showed the same pattern of results for dominated or low­\\nstatus groups. On the other hand, dominant classes and individualistic \\ncultures reflect more the attention and valuation of a subject who possesses \\nresources, augmenting the salience of personal over social identity. \\nHigh-status groups will stress the personal aspects of identity, the per­\\nception of ingroup individuation and dif\\nf\\nerentiation. Dominated groups will \\nhave a self-concept which will place more emphasis on group membership, \\nstressing the perception of intragroup homogeneity - the same as subjects \\nfrom collectivistic cultures. \\nGroups in general view themselves as superior in comparison to other \\ngroups on those attributes which define them or which are typical of them, \\nand tend to perceive outgroups as less variable or more homogeneous. \\nNevertheless, in threatening situations and in relation to defining or typical \\nattributes of one's collective identity, exactly the opposite takes place: they \\nare perceived as more homogeneous. \\nSocial status was another variable which moderated the relationship \\nbetween social identity and ingroup favouritism. Subjects perceive their \\ningroup as superior to the outgroup on those attributes which define it, \\nalthough they do not necessarily show general favouritism. This favouritism \\nwill be most clearly found among high-status groups. Although in general \\nsubjects represent their ingroup as more heterogeneous than the outgroup, \\nat the same time they state that it is more homogeneous, or less variable, on \\nthose attributes which define it. Low-status or -prestige groups show out­\\ngroup f\\navouritism or altercentrism, while at the same time they view them­\\nselves as superior on specific dimensions, especially sociability and \\n\\n\\nConstructing social identity \\n229 \\nexpressiveness. Our own study on psychologists and engineers in Spain \\nconfirmed that the perceptive superiority o\\nf\\n a lower status ingroup (psy­\\nchologists) was found in an expressive dimension (they viewed themselves as \\nmore friendly and sociable), accepting the fact that the high-status outgroup \\n(engineers) was superior to them in more instrumental attributes. \\nSIT states that group membership is a source of self-esteem, and that \\nbelonging to a disadvantaged group will motivate favourable social com­\\nparison and social competition. In a sense, dominated or low-status groups \\ndo not show a deficit in self-esteem. The differentiation between personal \\nand social identity, as well as the relative independence of private, public \\nand personal collective self-esteem, are two mechanisms which explain why \\nthe members of dominated and stigmatized classes do not show a deficit in \\npersonal self-esteem. \\nAlso, as we have already stated, the relationship between discrimination \\nand the increase in self-esteem is not consistent. The difference between \\nindividual and collective self-esteem partly explains the lack of association \\nbetween identification, discrimination and self-esteem. Studies conducted on \\ncollective self-esteem and on fratemalistic relative deprivation and political \\nparticipation suggest that a threatened high collective self-esteem is one of \\nthe antecedents of a biased social perception and of discriminatory \\nbehaviour. \\nTendencies towards ingroup favouritism are clearer among those indi­\\nviduals who share a strong group identification, high collective self-esteem \\nand collectivistic values. Those individuals who value belongingness to \\ncollectivities also think that their group membership is important and they \\npositively value this membership for their identity. These are the people \\nwho have stronger reactions when their social identity is under threat; they \\nincrease intergroup differentiation, positively compare their ingroup to \\nother groups, and increase the identification and attraction of their group. \\nOur correlational studies confirmed that individuals with a strong group \\nidentification show high private and public group self-consciousness, more \\nstability, a better valuation, a positive emotional reaction, and attach more \\nimportance to belonging to a group for their identity. \\nNote \\nWe wish to thank John C. Turner for his comments on an early version of this chapter. \\n\\n\\nReferences \\nAbrams, D. (1990). How do group members regulate their behaviour? An integration of social \\nidentity and self-awareness theories. In D. Abrams & M.A. Hogg (Eds.), Social identity \\ntheory: Constructive and critical advances. Hemel Hempstead: Harvester Wheatsheaf. \\nAbrams, D. (1994). Social self-regulation. Personality and Social Psychology Bulletin, 20, 473-\\n483. \\nAbrams, D., & Hogg, M.A. (1988). Comments on the motivational status of self-esteem in \\nsocial identity and intergroup discrimination. European Journal o\\nf\\n Social Psychology, IS, \\n3 17-334. \\nAbrams, D., & Hogg, M.A. (1990). Social identification, self-categorization and social \\ninfluence. In W. Stroebe & M. Hewstone (Eds.), European review o\\nf\\n social psychology, Vol. \\n1. Chichester: Wiley. \\nAbrams, D., Wetherell, M.S., Cochrane, S., Hogg, M.A., & Turner, J.e. (1990). Knowing \\nwhat to think by knowing who you are: Self-categorization and the nature of norm \\nformation, conformity and group polarization. British Journal o\\nf\\n Social Psychology, 29, 97-\\n1 19. \\nAbric, J.e. (1994). Pratiques sociales et representations. Paris: Presses Universitaires de France. \\nAdamopoulos, J. (1982). Analysis of interpersonal structures in literary works of three \\nhistorical periods. Journal o\\nf\\n Cross-Cultural Psychology, 13, 1 57-168. \\nAdamopoulos, J., & Bontempo, R.N. (1986). Diachronic universals in interpersonal structures. \\nJournal o\\nf\\n Cross-Cultural Psychology, 1 7, 169-189. \\nAebischer, V., Hewstone, M., & Henderson, M. (1984). Minority influence and musical \\npreference: Innovation by conversion not coercion. European Journal o\\nf\\n Social Psychology, \\n14, 23-33. \\nAllen, V.L., & Wilder, D.A. (1975). Categorization, belief similarity, and intergroup \\ndiscrimination. Journal o\\nf\\n Personality and Social Psychology, 32, 971-977. \\nAllen, V.L., & Wilder, D.A. (1979). Group categorization and attribution of belief similarity. \\nSmall Group Behaviour, 10, 73-80. \\nAllport, F.H. (1924). Social psychology. Boston: Houghton Mifflin. \\nAllport, G.W. (1961). The ego in contemporary psychology. Psychological Review, 50, 471-\\n478. \\nAllport, G.W., & Odbert, H.S. (1936). Trait-names: A psycho-lexical study. Psychological \\nMonographs, 47. \\nAltman, I., & Low, S.M. (Eds.). (1992). Place attachment. New York: Plenum. \\nAmancio, L. (1989). Social differentiation between 'dominant' and 'dominated' groups: Toward \\nan integration of social stereotypes and social identity. European Journal o\\nf\\n Social \\nPsychology, 19, 1-10. \\nAnderson, B. (1983). Imagined communities: Re\\nflections on the origin and spread o\\nf\\n nationalism. \\nLondon: Verso. \\nApter, D.E. (1964). Ideology and discontent. In D.E. Apter (Ed.), Ideology and discontent. \\nLondon Collier Macmillan. \\nArdener, S. (1983). Arson, nudity and bombs among the Canadian Doukhobors: A question of \\nidentity. In G.M. Breakwell (Ed.), Threatened identities. Chichester: Wiley. \\nArdrey, R. (1967). T\\nhe territorial imperative. London: Collins. \\nArgyle, M. (1992). The social psychology o\\nf\\n everyday lif\\ne. London: Routledge. \\nAsch, S.E. (1952). Social psychology. Englewood Cliffs, NJ: Prentice Hall. \\n\\n\\nRe\\nf\\nerences \\n231 \\nAsch, S.E. (1956). Studies on independence and conformity: A minority of one against a \\nunanimous majority. Psychological Monographs, 70, 456. \\nAsch, S.E., & Zuckier, H. (1984). Thinking about persons. Journal o\\nf\\n Personality and Social \\nPsychology, 46, 1230- 1240. \\nAyesteran, S. (Ed.). (1985). Psicologia y enf\\nermedad mental: ldeologia y representacion social de \\nla enf\\nermedad mental. Bilbao: Universidad del Pais Vasco. \\nBakan, D. (1966). The duality o\\nf\\n human existence. Boston: Beacon. \\nBall, P., Giles, H., & Hewstone, M. (1984). Secondary language acquisition: The intergroup \\ntheory with catastrophic dimensions. In H. Tajfel (Ed.), The social dimension: European \\ndevelopments in social psychology, Vol. 2. Cambridge: Cambridge University Press. \\nBarsalou, L.W. (1983). Ad hoc categories. Memory and Cognition, 11, 2 1 1 -227. \\nBarsalou, L. \\nW. (1987). The instability of graded structure: Implications for the nature of \\nconcepts. In U. Neisser (Ed.), Concepts and conceptual development: Ecological and \\nintellectual f\\nactors in categorization. Cambridge: Cambridge University Press. \\nBarsalou, L.W. (1989). Intraconcept similarity and its implications for interconcept similarity. \\nIn S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning. Cambridge: \\nCambridge University Press. \\nBarsalou, L.W., & Sewell, D.R. (1984). Constructing representations o\\nf\\n categories f\\nrom diff\\nerent \\npoints o\\nf\\n view (Emory Cognition Projects Report No. 2). Atlanta: Emory University. \\nBarsalou, L.W., Sewell, D.R., & Ballato, S.M. (1986). Assessing the stability o\\nf\\n category \\nrepresentations with graded structure. Unpublished manuscript. \\nBar-Tal, D. (1986). The Masada Syndrome: A case of central belief. In N. Milgram (Ed.), \\nStress and coping in time o\\nf\\n war. New York: Brunner/Maze!' \\nBar-Tal, D. (1990). Group belie\\nf\\ns: A conce\\nption f\\nor analyzing group structure, processes, and \\nbehavior. New York: Springer Verlag. \\nBar-Tal, D. (1993). American convictions about conflictive USA-USSR relations: A case of \\ngroup beliefs. In S. Worchel & 1. Simpson (Eds.), Conflict between people and peoples. \\nChicago: Nelson-Hall. \\nBar-Tal, D. (in press). Societal belie\\nf\\ns\\n o\\nf\\n ethos: A social psychological analysis o\\nf\\n society. \\nLondon: Sage. \\nBar-Tal, D., Raviv, A., & Brosh, M. (1991). Perception of epistemic authority and attribution \\nfor the choice as a function of knowledge area and age. European Journal o\\nf\\n Social \\nPsychology, 21, 477-492. \\nBar-Tal, Y. (1989). Can leaders change followers' stereotypes? In D. Bar-Tal, C. Grauman, \\nA.W. Kruglanski & W. Stroebe (Eds.), Stereotyping and pre\\njudice: Changing conceptions. \\nNew York: Springer Verlag. \\nBellelli, G. (Ed.). (1 987). La representation sociale de la maladie mentale. Naples: Liguori. \\nBellelli, G. (Ed.). (1994). L'altra malattia: Come la societa pensa la malattia mentale. Naples: \\nLiguori. \\nBern, D.J. (1967) Self perception: An alternative interpretation of cognitive dissonance \\nphenomena. Psychological Review, 74, 1 83-200. \\nBern, D.l. (1970). Belie\\nf\\ns, attitudes, and human a\\nf\\nf\\nairs. Belmont, CA: Brooks/Cole. \\nBern, S.L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological \\nReview, 88, 354-364. \\nBern, S.L. (1985). Androgyny and gender schema theory: A conceptual and empirical \\nintegration. In T.B. Soderberegger (Ed.), Psychology and gender: Nebraska Symposium on \\nMotivation, Vol. 32. Lincoln: University of Nebraska Press. \\nBillig, M. (1985). Prejudice, categorization and particularization: From a perceptual to a \\nrhetorical approach. European Journal o\\nf\\n Social Psychology, 15, 79-103. \\nBillig, M., & Tajfel, H. (1973). Social categorization and similarity in intergroup behaviour. \\nEuropean Journal o\\nf\\n Social Psychology, 3, 27-52. \\nBoltanksi, 1.L. (1982). Les cadres: La f\\normation d'un groupe social. Paris: Minuit. \\nBomstein, G., Crum, L., Wittenbraker, 1., Harring, K., Insko, C.A., & Thibaut, 1. (l983a). On \\nthe measurement of social orientation in the minimal group paradigm. European Journal o\\nf\\n \\nSocial Psychology, 13, 321 -350. \\n\\n\\n232 \\nSocial identity: international perspectives \\nBomstein, G., Crum, L., Wittenbraker, 1., Harring, K., Insko, C.A., & Thibaut, 1. (1983b). \\nReply to Turner's comments. European Journal o\\nf\\n Social Psychology, 13, 369-381 .  \\nBourhis, R.Y., Gagnon, A., & MOise, L.C. (1 994). Discrimination et relation intergroupe. In \\nR.Y. Bourhis & l.-P. Leyens (Eds.), Stereotypes, discrimination et relations intergroupes. \\nBrussels: Mardaga. \\nBrewer, M.B. (1979). Ingroup bias in the minimal intergroup situation: A cognitive­\\nmotivational analysis. Psychological Bulletin, 86, 307-324. \\nBrewer, M.B. (1988). A dual process model of impression formation. In T.K. Srull & R.S. \\nWyer (Eds.), Advances in social cognition, Vol. 1. Hillsdale, Nl: Erlbaum. \\nBrewer, M.B. (1991). The social self: On being the same and different at the same time. \\nPersonality and Social Psychology Bulletin, 1 7, 475-482. \\nBrewer, M.B., Dull, Y., & Lui, L. (1981). Perceptions of the elderly: Stereotypes as prototypes. \\nJournal o\\nf\\n Personality and Social Psychology, 41, 656-670. \\nBrewer, M.B., Ho, H., Lee, 1., & Miller, N. (1987). Social identity and social distance among \\nHong Kong school children. Personality and Social Psychology Bulletin, 13, 1 56-165. \\nBrewer, M.B., & Lui, L. (1984). Categorization of the elderly by the elderly: Effects of \\nperceiver's category membership. Personality and Social Psychology Bulletin, 10, 585-595. \\nBrewer, M.B., & Silver, M. (1978). Ingroup bias as a function of task characteristics. European \\nJournal o\\nf\\n Social Psychology, 8, 393-400. \\nBrown, R. (1954). Mass phenomena. In G. Lindzey (Ed.), Handbook o\\nf\\n social psychology, Vol. \\n2. Cambridge, MA: Addison-Wesley. \\nBrown, R.l. (1988). Group processes: Dynamics within and between groups. Oxford: Blackwell. \\nBrown, R.l., Hinkle, S., Ely, P.G., Fox-Cardamone, L., Maras, P., & Taylor, L.A. (1992). \\nRecognising group diversity: Individualist-collectivist and autonomous-relational social \\norientations and their implications for intergroup processes. British Journal o\\nf\\n Social \\nPsychology, 31, 327-342. \\nBrown, R.1., & Ross, G.F. (1982). The battle for acceptance: An investigation into the \\ndynamics of intergroup behaviour. In H. Tajfel (Ed.), Social identity and intergrou\\np \\nrelations. Cambridge: Cambridge University Press. \\nBrown, R.l., & Turner, 1.C. (1979). The criss-cross categorization effect in intergroup \\ndiscrimination. British Journal o\\nf\\n Social and Clinical Psychology, 18, 371 -383. \\nBrown, R.l., & Turner, 1.C. (1981). Interpersonal and intergroup behaviour. In 1.C. Turner & \\nH. Giles (Eds.), I\\nntergroup behaviour. Oxford: Blackwell. \\nBruner, 1.S. (1957). On perceptual readiness. Psychological Review, 64, 123-151. \\nBurnstein, E. (1982). Persuasion as argument processing. In H. Brandstatter, 1. Davis & G. \\nStocker-Kreichgauer (Eds.), Group decision-making. London: Academic Press. \\nButera, F., Huguet, P., Mugny, G., & Perez, 1.A. (1994). Socio-epistemic conflict and \\nconstructivism. Swiss Journal o\\nf\\n Psychology, 53, 229-239. \\nButera, F., Legrenzi, P., Mugny, G., & Perez, 1.A. (1990-1991). Influence sociale et \\nraisonnement. Bulletin de Psycho \\nlogie, 45, 144-154. \\nByrne, D. (1969). Attitudes and attraction. In L. Berkowitz (Ed.), Advances in ex\\nperimental \\nsocial psychology, Vol. 4. New York: Academic Press. \\nByrne, D. (1971). The attraction problem. New York: Academic Press. \\nCaddick, B. (1982). Perceived illegitimacy and intergroup relations. In H. Tajfel (Ed.), Social \\nidentity and intergroup relations. Cambridge: Cambridge University Press. \\nCampbell, D.T. (1958). Common fate, similarity, and other indices of the status of aggregates \\nof persons as social entities. Behavioral Science, 3, 14-25. \\nCampbell, D.T. (1965). Ethnocentric and other altruistic motives. In D. Levine (Ed.), Nebraska \\nSymposium on Motivation, Vol. 13. Lincoln: University of Nebraska Press. \\nCamps, Y. (1993). Parado\\njas del individualismo. Barcelona: Critica. \\nCannon, T. (1973). Black Panther party program: What we want, and what we believe. In R.R. \\nEvans (Ed.), Social Movements. Chicago: Rand McNally. \\nCantor, N., & Mischel, W. (1979). Prototypes in person perception. In L. Berkowitz (Ed.), \\nAdvances in experimental social psychology, Vol 12. New York: Academic Press. \\nCapozza, D., Bonaldo, E., & Di Maggio, A. (1982). Problems of identity and social conflict: \\n\\n\\nRe\\nf\\nerences \\n233 \\nResearch on ethnic groups in Italy. In H. Tajfel (Ed.), Social identity and intergroup \\nrelations. Cambridge: Cambridge University Press. \\nCarroll, 0.1., & Chang, J.J. (1970). Analysis of individual differences in multidimensional \\nscaling via N-way generalisation of 'Eckart-Young' decomposition. Psychometrika, 35, \\n283-319. \\nCartwright, D. (1968). The nature of group cohesiveness. In D. Cartwright & A. Zander (Eds.), \\nGroup dynamics (3rd ed.). London: Tavistock; New York: Harper & Row. \\nCartwright, D., & Zander, A. (Eds.). (1968). Group dynamics (3rd ed.). London: Tavistock; \\nNew York: Harper & Row. \\nCarver, c., Antoni, M., & Scheier, M. (1985). Self-consciousness and self-assessment. Journal \\no\\nf\\n Personality and Social Psychology, 48, 1 1 7-124. \\nClark, R.D., III, & Maass, A. (1988a). Social categorization in minority influence: The case of \\nhomosexuality. European Journal o\\nf\\n Social Psychology, 18, 347-364. \\nClark, R.D., HI, & Maass, A. (I \\n988b). The role of social categorization and perceived source \\ncredibility in minority influence. European Journal o\\nf\\n Social Psychology, 18, 381 -394. \\nCodol, J.-P. (1975). On the so-called 'superior conformity of the self' behaviour: Twenty \\nexperimental investigations. European Journal o\\nf\\n Social Psychology, 5, 457-501 .  \\nCodol, J.-P. (1979). Semblables et difJerents: Recherches sur la quete de la similitude et de la \\ndifJ\\nerenciation sociale. Thesis, Universite de Provence, Aix-en-Provence. \\nCodol, J.-P. (1982). Diff\\nerentiating and non-differentiating behavior: A cognitive approach to \\nthe sense of identity. In J.-P. Codol & J.-P. Leyens (Eds.), Cognitive analysis o\\nf\\n social \\nbehavior. The Hague: Martinus Nijhoff. \\nCodol, J.-P. (1984a). On the system of representations in an artificial social situation. In R.M. \\nFarr & S. Moscovici (Eds.), Social representations. Cambridge: Cambridge University Press. \\nCodol, J.-P. (1 984b). La perception de la similitude interpersonnelle: Influence de \\nI'appartenance categorielle et du point de ref\\nerence de la comparaison. L'Annee \\nPsychologique, 84, 43-56. \\nCodol, J.-P. (1987). Comparability and incomparability between oneself and others: Means of \\ndifferentiation and comparison ref\\nerence point. CaMers de Psychologie Cognitive, 7, 1- 18. \\nCodol, J.-P. (1990). Studies on self-centered assimilation processes. In J.P. Caverni, J.M. Fabre \\n& M. Gonzalez (Eds.), Cognitive biases. Amsterdam: North Holland. \\nCondor, S. (1989). 'Biting into the future': Social change and the social identity of women. In S. \\nSkevington & D. Baker (Eds.), The social identity o\\nf\\n women. London: Sage. \\nCondor, S. (1 990). Social stereotypes and social identity. In D. Abrams & M.A. Hogg (Eds.), \\nSocial identity theory: Constructive and critical advances. Hemel Hempstead: Harvester \\nWheatsheaf. \\nCondor, S., Brown, R.1., & Williams, J. (1987). Social identification and intergroup behaviour. \\nQuarterly Journal o\\nf\\n Social Science Af\\nf\\nairs, 3, 299-317. \\nConverse, P.E. (1964). The nature of the belief systems in mass pUblics. In D.E. Apter (Ed.), \\nIdeology and discontent. New York: Free Press. \\nCoutant, D., Worchel, S., & Grossman, M. (in preparation). Minority influence over time. \\nCoutant-Sassic, D. (1991). DifJ\\nerential cognitive processing o\\nf\\n a ma\\njority vs. a minority in\\nfluence \\non the same or difJ\\nerent task. Unpublished master's thesis, Texas A&M University. \\nCowan, G., & Hoffman, C.D. (1986). Gender stereotyping in young children: Evidence to \\nsupport a concept-learning approach. Sex Roles, 14, 2 1 1 -224. \\nCrano, W.O. (1989). Judgmental sub\\njectivity, objectivity, and minority influence. Paper \\npresented to the third workshop on minority influence, Perugia, Italy, June. \\nCrocker, J., & Luhtanen, R. (1990). Collective self-esteem and ingroup bias. Journal o\\nf\\n \\nPersonality and Social Psychology, 58, 323-338. \\nCrocker, J., Luhtanen, R., Blaine, B., & Broadnax, S. (1 994). Collective self-esteem and \\npsychological well-being among white, Black and Asian college students. Personality and \\nSocial Psychology Bulletin, 20, 503-51 3. \\nCrosby, F., Cordova, D., & Jaskar, K. (1994). On the failure to see oneself as disadvantaged: \\nCognitive and emotional components. In M.A. Hogg & D. Abrams (Eds.), Motivation: \\nSocial psychological perspectives. Hemel Hempstead: Harvester Wheatsheaf. \\n\\n\\n234 \\nSocial identity: international pers\\npectives \\nCylwik, A. (1992). Przychylnosc wobec innych narodowosci i mniejszosci narodowych a schematy \\nla/My. Unpublished master's thesis, Wydzial Psychologii, Uniwersytet Warzawski, Warsaw. \\nDamon, W., & Hart, D. (1982). The development of self-understanding from infancy through \\nadolescence. Child Development, 53, 841-864. \\nDeaux, K. (1992). Personalizing identity and socializing self. In G. Breakwell (Ed.), Social \\npsychology o\\nf\\n identity and the self\\n-conce\\npt. London: Academic Press. \\nDeaux, K. (1993). Reconstructing social identity. Personality and Social Psychology Bulletin, \\n19, 4-12. \\nDeconchy, J.P. (1984). Rationality and social control in orthodox systems. In H. Tajfel (Ed.), \\nThe social dimension: European developments in social psychology, Vol. 1. Cambridge: \\nCambridge University Press. \\nDelumeau, J.J. (1993). La peur en Occident. Paris: Hachette: Pluriel. \\nDeschamps, J.-C. (1972-1973). Imputation de la responsabilite de I'echec (ou de la reussite) et \\ncategorisation sociale. Bulletin de Psychologie, 26, 794-806. \\nDeschamps, J.-C. (1977). L'attribution et la categorisation sociale. Berne: Peter Lang. \\nDeschamps, J.-C. (1979). Diff\\nerenciation categorielle et differenciation de soi par rapport a \\nautrui. Recherches de Psychologie Sociale, 1, 29-38. \\nDeschamps, J.-c. (1980). L'identite sociale et les rapports de domination. Revue Suisse de \\nSociologie, 1, 1 1 1-122. \\nDeschamps, J.-c. ( l982a). Differenciation entre soi et autrui et entre groupes: Recherches sur la \\n'covariation' entre les differenciations inter-individuelles et inter-groupes. In J.-P. Codol & \\nJ.-P. Leyens (Eds.), Cognitive analysis o\\nf\\n social behavior. The Hague: Martinus Nijhoff. \\nDeschamps, J.-C. (1982b). Social identity and relations of power between groups. In H. Tajfe\\\\ \\n(Ed.), Social identity and intergroup relations. Cambridge: Cambridge University Press. \\nDeschamps, J .-c. (1983-1984). Categorisation sociale et differenciations interindividuelles. \\nBulletin de Psycho \\nlogie, 37, 489-500. \\nDeschamps, J.-c. (1984). Identite socia \\nIe et differenciation categorielle. Cahiers de Psychologie \\nCognitive, 4, 449-474. \\nDeschamps, J.-C. (1991). Identites, appartenances sociales et differenciations individuelles. \\nCahiers Internationaux de Psychologie Sociale, 9-10, 49-6\\\\. \\nDeschamps, J.C., Clemence, A., & Roux, P. (1983). Differ\\nenciations inter-groupes et inter­\\nindividuelles. Paper presented at Congreso della Divisione di Psicologia della SIPS, Padua. \\nDeschamps, J.-C., & Doise, W. (1978). Crossed category memberships in intergroup relations. \\nIn H. Tajfel (Ed.), Diff\\nerentiation between social groups: Studies in the social psychology o\\nf\\n \\nintergroup relations. London: Academic Press. \\nDeschamps, J.-c., & Lorenzi-Cioldi, F. (1981). 'Egocentrisme' et 'sociocentrisme' dans les \\nrelations entre groupes: Recherche experimentale. Revue Suisse de Psychologie Pure et \\nApplique, 40, 108-131. \\nDeschamps, J.-c., & Serino, C. (1996). Mondo al singolare: Pros\\npettive socio-cognitive \\nsull'identita. Naples: Liguori. \\nDeschamps, J.-c., & Volpato, C. (1984). Identita individuale nelle relazioni tra gruppi. Giornale \\nItaliano di Psicologia, 11, 275-301 .  \\nDeutsch, M. (1968). The eff\\nects of cooperation and competition upon group process. I n  D. \\nCartwright & A. Zander (Eds.), Group d\\nynamics (3rd ed.). London: Tavistock; New York: \\nHarper & Row. \\nDeutsch, M., & Gerard, H.B. (1955). A study of normative and informational social influence \\nupon individual judgment. lournal o\\nf\\n Abnormal and Social Psychology, 51, 629-636. \\nDevos, T., Comby, L., & Deschamps, J.-c. (\\\\995). Appartenances sociales et jugement de \\nvariabilite intra-groupe. Cahiers Internationaux de Psychologie Sociale, 27, 9-43. \\nDiamond, M. (1976). The American idea of man: The view from the founding. In T. Kristol & \\nP. Weaver (Eds.), The Americans: 1976, Vol. 2. Lexington, MA: Lexington Books. \\nDiehl, M. (\\\\990). The minimal group paradigm: Theoretical explanations and empirical \\nfindings. In W. Stroebe & M. Hewstone (Eds.), European review o\\nf\\n social \\npsychology, Vol. 1. \\nChichester: Wiley. \\n\\n\\nRe\\nf\\nerences \\n235 \\nDoise, W. (1969). Strategies de \\njeu a I'interieur et entre des groupes de nationalites diff\\nerentes. \\nBulletin du Centre d'Etudes et Recherches Psycho \\ntechniques, 18, 1 3-26. \\nDoise, W. (1976). L'articulation psychosociologique et les relations entre groupes. Brussels: De \\nBoeck. \\nDoise, W. (1986). Les representations sociales: Definition d'un concept. In W. Doise & A. \\nPalmonari (Eds.), L'erude des representations sociales. Paris: Delachaux et NiestIe. \\nDoise, W. (1988). Individual and social identities in intergroup relations. European Journal o\\nf\\n \\nSocial Psychology, /8, 99-1 1 \\\\ .  \\nDoise, W., Clemence, A., & Lorenzi-Cioldi, F .  (1993). The quantitative anal\\nysis o\\nf\\n social \\nrepresentations. Hemel Hempstead: Harvester Wheatsheaf. \\nDoise, W., Csepeli, G., Dann, H.D., Gouge, C., Larsen, K., & Ostell, A. (1972). An experi­\\nmental investigation into the formation of intergroup representations. European Journal o\\nf\\n \\nSocial Psychology, 2, 202-204. \\nDoise, W., Deschamps, J.-c., & Meyer, G. (1 978). The accentuation of intracategory \\nsimilarities. In H. Tajfel (Ed.), Diff\\nerentiation between social groups: Studies in the social \\npsychology o\\nf\\n intergroup relations. London: Academic Press. \\nDoise, W., Gachoud, J.P., & Mugny, G. (1986). Influence directe et indirecte entre groupes \\ndans des choix esthetiques. Cahiers de Psychologie Cognitive, 6, 283-30\\\\ .  \\nDoise, W., & Lorenzi-Cioldi, F .  (1990). Levels of analysis and social identity. In D .  Abrams & \\nM.A. Hogg (Eds.), Social identity theory: Constructive and critical advances. Hemel \\nHempstead: Harvester Wheatsheaf. \\nDoise, W., & Lorenzi-Cioldi, F. (1991). L'identite comme representation sociale. In V. \\nAebischer, J.P. Deconchy & E.P. Lipiansky (Eds.), Ideologie et representations sociales. \\nCousset: Delval. \\nDoise, W., & Palmonari, A. (1986). L'etude des representations sociales. Paris: Delachaux et \\nNiestle. \\nDoob, L.W. (1964). Patriotism and nationalism. New Haven, CT: Yale University Press. \\nDoosje, B., Ellemers, N. & Spears, R. (1996). Perceived intragroup variability as a function of \\ngroup status and identification. Journal o\\nf\\n Experimental Social Psychology, 3/, 410-436. \\nDumont, L. (1987). Ensayos sobre el individualismo: Una pers\\npectiva antropologica sobre la \\nideologia mode \\nrna. Madrid: Alianza. \\nDunbar, R. (1992). Why gossip is good for you. New Scientist, 21 November, 28-3\\\\ .  \\nDunning, D., Perie, M., & Story, A.L. (1991). Self-serving prototypes of social categories. \\nJournal o\\nf\\n Personality and Social Psychology, 6/, 957-968. \\nDurand Delvigne, A. (\\\\ 992). Identite et modides sexues des personnes. These de Doctorat d'Etat \\nde Lettres et Sciences H \\numaines, Paris. \\nDurkheim, E. (1898). Representations individuelles et representations collectives. Revue de \\nMetaphysique, 6, 274-302. \\nDuveen, G., & Lloyd, B. (1990). Social representations and the development o\\nf\\n knowledge. \\nCambridge: Cambridge University Press. \\nEagly, A.H., & Chaiken, S. (1993). The psychology o\\nf\\n attitudes. New York: Harcourt Brace \\nJovanovich. \\nEchebarria, A., Elajabarrieta, F., Valencia, J., & Villareal, M. (1992). Representations sociales \\nde l'Europe et identites sociales. Bulletin de Psychologie, 45, 280-288. \\nEiser, J.R. (1983). Psicologia sociale cognitivista. Bologne: II Mulino. (Original English \\npublication: Maidenhead: McGraw-Hill, 1980.) \\nEllemers, N., Wilke, H., & van Knippenberg, A. (1993). Effects of the legitimacy of low group \\nor individual status on individual and collective status-enhancement strategies. Journal o\\nf\\n \\nPersonality and Social Psychology, 64, 766-778. \\nEmler, N., & Hopkins, N. (1990). Reputation, social identity and the self. In D. Abrams & \\nM.A. Hogg (Eds.), Social identity theory: Constructive and critical advances. Hemel \\nHempstead: Harvester Wheatsheaf. \\nEmler, N., & Reicher, S. (1995). Adolescence and delinquency. Oxford: Blackwell. \\nEpstein, A.L. (1978). Ethos and identity. London: Tavistock. \\nEpstein, S. (\\\\ 980). The self-concept: A review and the proposal of an integrated theory of \\n\\n\\n236 \\nSocial identity: international pers\\npectives \\npersonality. In E. Staub (Ed.), Personality: Basic issues and current research. Englewood \\nCliffs, NJ: Prentice Hall. \\nErikson, E. (1956). The problem of ego identity. Journal o\\nf\\n the American Psychiatric \\nAssociation, 1, 58-121. \\nErikson, E. (1977). Childhood and society. London: Paladin. \\nEtaugh, C, & Riley, S. (1983). Evaluating competence of women and men: Effects of marital \\nand parental status and occupational sex-typing. Sex Roles, 9, 943-952. \\nEtzioni, A. (1975). Complex organizations (rev. ed.). New York: Free Press. \\nFalomir, J.M., Mugny, G., & Perez, I.A. (1996). Le paradoxe du denio In J.-C Deschamps & \\nJ.-L. Beauvois (Eds.), Des attitudes aux attributions. Grenoble: Presses Universitaires de \\nGrenoble. \\nFarr, R.M., & Moscovici, S. (1984). Social representations. Cambridge: Cambridge University \\nPress. \\nFells, J., & Niznik, J. (1992). Conclusion: What is Europe? International Journal o\\nf\\n Sociology, \\n26, 201 -206. \\nFestinger, L. (1950). Informal social communication. Psychological Review, 57, 271-282. \\nFestinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 1 17-140. \\nFestinger, L., Riecken, H.W., & Schachter, S. (1956). W\\nhen prophecy f\\nails. Minneapolis: \\nUniversity of Minnesota Press. \\nFestinger, L., & Thibaut, J. (1951). Interpersonal communication in small groups. Journal o\\nf\\n \\nAbnormal and Social Psychology, 46, 92-99. \\nFestinger, L., Schachter, S., & Back, K. (1950). Social pressures in inf\\normal groups: A stud\\ny o\\nf\\n \\nhuman f\\nactors i\\nn\\n housing. New York: Harper. \\nFiske, S.T., & Neuberg, S.L. (1990). A continuum of impression formation, from category­\\nbased to individuating processes: Influences of information and motivation on attention and \\ninterpretation. In M.P. Zanna (Ed.), Advances in ex\\nperimental social psychology, Vol. 23. \\nNew York: Academic Press. \\nFiske, S.T., & Pavelchak, M.A. (1986). Category-based versus piecemeal-based aff\\nective \\nresponses: Developments in schema-triggered affect. In R.M. Sorrentino & E.T. Higgins \\n(Eds.), Handbook o\\nf\\n motivation and cognition: Foundations o\\nf\\n social behavior, Vol. 1. New \\nYork: Guilford. \\nForgas, J.P. (1979). Social episodes: T\\nhe stud\\ny o\\nf\\n interaction routines. New York: Academic \\nPress. \\nForsyth, D.R. (\\\\995). Our social world. Pacific Grove, CA: Brooks/Cole. \\nFraser, S.L., & Lord, R.G. (l988). Stimulus prototypicality and general leadership impressions: \\nTheir role in leadership and behavioural ratings. Journal o\\nf\\n Psychology, 122, 291-303. \\nGaertner, S.L., Dovidio, J.F., Anastasio, P.A., Bachman, B.A., & Rust, M.C (1993). The \\ncommon ingroup identity model: Recategorization and the reduction of intergroup bias. In \\nW. Stroebe & M. Hewstone (Eds.), European review o\\nf\\n social psychology, Vol. 4. Chichester: \\nWiley. \\nGaertner, S.L., Mann, J., Murrell, A., & Dovidio, J.F. (\\\\989). Reducing intergroup bias: The \\nbenefits of recategorization. Journal o\\nf\\n Personality and Social Psychology, 57, 239-249. \\nGarza, R.T., & Herringer, L.G. (\\\\986). Social identity: A multidimensional approach. Journal \\no\\nf\\n Social Psychology, 127, 299-308. \\nGergen, K.J. (1965). Interaction goals and personalistic feedback as factors affecting the \\npresentation of the self. Journal o\\nf\\n Personality and Social Psychology, 1, 413-424. \\nGergen, KJ. (1982). From self to science: What is there to know? In I.M. Suls (Ed.), \\nPsychological pers\\npectives on the self. Hillsdale, Nl: Erlbaum. \\nGillin, J. (\\\\ 955). National and regional cultural values in the United States. Social Forces, 34, \\n107- 1 1 3. \\nGoethals, G.R. (1972). Consensus and modality in the attribution process: The role of \\nsimilarity and information. Journal o\\nf\\n Personality and Social Psychology, 21, 84-92. \\nGoffman, E. (I 956}. The presentation o\\nf\\n self in everyday lif\\ne. Edinburgh: Social Science \\nResearch Centre - University of Edinburgh. \\nGoffman, E. (\\\\ 968). Stigma. Englewood, Cliffs: Prentice Hall. \\n\\n\\nRe\\nf\\nerences \\n237 \\nGol̗b, A., & Reykowski, 1. (1985). Studia nad rozwojem standardow ewaluatywnych. Wroclaw: \\nOssolineum. \\nGoldberg, L.R. (1990). An alternative 'Description of Personality': The big-five factor \\nstructure. Journal o\\nf\\n Personality and Social Psychology, 59, 1216-1229. \\nGottman, 1.M. (1981). Time series analysis: A comprehensive introduction f\\nor social scientists. \\nCambridge: Cambridge University Press. \\nGreenwald, A.G, & Banaji, M. (1995). Implicit social cognition: Attitudes, self-esteem and \\nstereotypes. Psychological Review, 12, 1-19. \\nGreenwald, A.G., & Pratkanis, A.R. (1984). The self. In R.S. Wyer & T.K. Srull (Eds.), \\nHandbook o\\nf\\n social cognition, Vol. 1. Hillsdale, NJ: Erlbaum. \\nGudkyunst, W. (1988). Culture and intergroup processes. In M. Bond (Ed.), T\\nhe cross-cultural \\nchallenge to social psychology. Newbury Park, CA: Sage. \\nGuimond, S., & Tougas, F. (1994). Sentiment d'injustice et action collective: La privation \\nrelative. In R.Y. Bourhis & J.-P. Leyens (Eds.), Stert!otypes, discrimination et relations \\nintergroupes. Brussels: Mardaga. \\nHamilton, V.L., & Rauma, D. (1995). Social psychology of deviance and law. In K.S. Cook, \\nG.A. Fine & 1.S. House (Eds.), Sociological pers\\npectives on social psychology. Boston: Allyn \\n& Bacon. \\nHardoin, M., & Codol, J.-P. (1984). Description de soi et d'autrui: Influence de l'ordre de \\ndescription sur les categories de reponse utilisees. Cahiers de Psychologie Cognitive, 4, 295-\\n302. \\nHarkins, S., & Petty, R.E. (1981). The multiple source effect in persuasion: The effects of \\ndistraction. Personality and Social Psychology Bulletin, 21, 84-92. \\nHarkins, S., & Petty, R.E. (1983). Social context effects in persuasion: The effects of multiple \\nsource and multiple targets. In P.B. Paulus (Ed.), Basic group process. New York: Springer \\nVerlag. \\nHarkins, S., & Petty, R.E. (1987). Inf\\normational utility and the multiple-source effect. Journal \\no\\nf\\n Personality and Social Psychology, 52, 260-268. \\nHarrison, A.A., Moore, M., & Rucker, M.H. (1985). Further evidence on career and family \\ncompatibility among eminent women and men. Archivio di Psicologia, Neurologia e \\nPsichiatria, 46, 140-1 55. \\nHarter, S. (1985). The self-perception pro\\nfile f\\nor children: Revision o\\nf\\n the Perceived Competence \\nScale f\\nor children. Manual, University of Denver. \\nHaslam, S.A., McGarty, C, Oakes, P.J., Turner, J.C, & Onorato, R. (1995). Contextual shifts \\nin the prototypicality of extreme and moderate outgroup members. European Journal o\\nf\\n \\nSocial Psychology, 25, 509-530. \\nHaslam, S.A., Oakes, P.J., Turner, J.C, & McGarty, C (1996). Social identity, self­\\ncategorization and the perceived homogeneity of in \\ngroups and outgroups: The interaction \\nbetween social motivation and cognition. In R.M. Sorrentino & E.T. Higgins (Eds.), \\nHandbook o\\nf\\n motivation and cognition, Vol. 3. New York: Guilf\\nord. \\nHaslam, S.A., & Turner, 1.C (1992). Context-dependent variation in social stereotyping 2: The \\nrelationship between frame of ref\\nerence, self-categorization and accentuation. European \\nJournal of Social Psychology, 22, 251-277. \\nHaslam, S.A., Turner, J.C, Oakes, PJ., McGarty, C., & Hayes, B.K. (1992). Context­\\ndependent variation in stereotyping I :  The effects of intergroup relations as mediated by \\nsocial change and frame of reference. European Journal of Social Psychology, 22, 3-20. \\nHayes, CH. (1926). Nationalism as a religion. In CH. Hayes, Essays on nationalism. New \\nYork: Macmillan, 1962. \\nHertz, F. (1944). Nationalism in history and politics. London: Routledge & Kegan Paul. \\nHewstone, M., Hantzi, A., & lohnston, L. (1991). Social categorization and person memory: \\nThe pervasiveness of race as an organizing principle. European Journal o\\nf\\n Social Psychology, \\n21, 517-528. \\nHiggins, E.T., & King, G. (1981). Accessibility of social constructs: Information-processing \\nconsequences of individual and contextual variability. In N. Cantor & J.F. Kihlstrom (Eds.), \\nPersonality, cognition, and social interaction. Hillsdale, NJ: Erlbaum. \\n\\n\\n238 \\nSocial identity: international pers\\npectives \\nHinkle, S., & Brown, R.J. (1990). Intergroup comparisons and social identity: Some links and \\nlacunae. In D. Abrams & M.A. Hogg (Eds.), Social identity theory: Constructive and critical \\nadvances. Hemel Hempstead: Harvester Wheatsheaf. \\nHinkle, S., Brown, R.J., & Ely, P.G. (1992). Social identity theory processes: Some limitations \\nand limiting conditions. Revista de Psicologia Social, Monogra\\njico, 99- 1 1 1 .  \\nHinkle, S., & Crook, K.F. (1987). Identification, importance o\\nf\\n comparison dimensions and \\nintergroup diff\\nerentiation. Paper presented at the International Conference on Social Identity, \\nExeter, UK, July. \\nHinkle, S., Fox-Cardamone, D.L., Taylor, L.A., & Crook, K.F. (1988). Identification, outgroup \\nthreat and dimensions o\\nf\\n intergrou\\np comparison. Paper presented at the Invitational \\nConference on Aggression, Conflict and Negotiation, Kill Devil Hills, NC. June. \\nHoffman, M.L. (1989). Empathy and pro \\nsocial activities. In N. Eisenberg, J. Rekowski & E. \\nStaub (Eds.), Social and moral values. Hillsdale, NJ: Erlbaum. \\nHofstede, G. (1980). Culture's consequences: International dif\\nf\\nerences in work-related values. \\nBeverly Hills, CA: Sage. \\nHogg, M.A. (1987). Social identity and group cohesiveness. In l.e. Turner, M.A. Hogg, P.l. \\nOakes, S.D. Reicher & M.S. Wetherell, Rediscovering the social group: A self-categorization \\ntheory. Oxford: Blackwell. \\nHogg, M.A. (1992). The social psychology o\\nf\\n group cohesiveness. Hemel Hempstead: Harvester \\nWheatsheaf. \\nHogg, M.A. (1993). Group cohesiveness: A critical review and some new directions. In W. \\nStroebe & M. Hewstone (Eds.), European review o\\nf\\n social psychology, Vol. 4. Chichester: \\nWiley. \\nHogg, M.A. (1996). Intragroup processes: group structure, and social identity. In P. Robinson \\n(Ed.), Social groups and identity: Developing the legacy o\\nf\\n Henri T\\na\\nj\\nf\\nel. Oxford: Butterworth \\nHeinemann. \\nHogg, M.A., & Abrams, D. (1988). Social identifications: A social psychology o\\nf\\n intergroup \\nrelations. London and New York: Routledge & Kegan Paul. \\nHogg, M.A., & Abrams, D. (1990). Social motivation, self-esteem and social identity. In D. \\nAbrams & M. Hogg (Eds.), Social identity theory: Constructive and critical advances. Hemel \\nHempstead: Harvester Wheatsheaf. \\nHogg, M.A., Cooper-Shaw, L., & Holzworth, D.W. (in press). Studies of group prototypicality \\nand depersonalized attraction in small interactive groups. Personality and Social Psychology \\nBulletin. \\nHogg, M.A., & Hardie, E.A. (1991). Social attraction, personal attraction, and self­\\ncategorization: A field study. Personality and Social Psychology Bulletin, 1 7, 175-180. \\nHogg, M.A., & Hardie, E.A. (1992). Prototypicality, conformity and depersonalized attraction: \\nA self-categorization analysis of group cohesiveness. British Journal o\\nf\\n Social Psychology, \\n31, 45-56. \\nHogg, M.A., & McGarty, C. (1990). Self-categorization and social identity. In D. Abrams & \\nM.A. Hogg (Eds.), Social identity theory: Constructive and critical advances. Hemel \\nHempstead: Harvester Wheatsheaf. \\nHogg, M.A., & Turner, l.e. (l985a). When liking begets solidarity: An experiment on the role \\nof interpersonal attraction in psychological group formation. British Journal o\\nf\\n Social \\nPsychology, 24, 267-281 .  \\nHogg, M.A., & Turner, l.C. (1985b). Interpersonal attraction, social identification, and \\npsychological group formation. European Journal o\\nf\\n Social Psychology, 15, 51-66. \\nHogg, M.A., & Turner, J.e. (1987). Social identity and conformity: A theory of referent \\ninformation influence. In W. Doise & S. Moscovici (Eds.), Current issues in European social \\npsychology, Vol. 2. Cambridge: Cambridge University Press; Paris: Maison des Sciences de \\nI'Homme. \\nHogg, M.A., Turner, J.C., & Davidson, B. (1990). Polarized norms and social frames of \\nreference: A test of the self-categorization theory of group polarization. Basic and Applied \\nSocial Psychology, 11, 77-100. \\n\\n\\nRe\\nf\\nerences \\n239 \\nHolyoak, K.J., & Gordon, P.e. (1983). Social reference points. Journal o\\nf\\n Personality and \\nSocial Psychology, 44, 881-887. \\nHolyoak, K.J., & Gordon, P.e. (1984). Information processing and social cognition. In R.S. \\nWyer & T.K. Srull (Eds.), Handbook o\\nf\\n social cognition, Vol. 1. Hillsdale, NJ: Erlbaum. \\nHomans, G. (1950). The human group. New York: Harcourt Brace. \\nHorwitz, M., & Rabbie, J.M. (1989). Stereotypes of groups, group members. and individuals in \\ncategories: A differential analysis. In D. Bar-Tal, e.F. Grumann, A. Kruglanski & W. \\nStroebe (Eds.), Stereotyping and \\npre\\njudice. New York: Springer Verlag. \\nHostetler, J.A. (1968). Amish society (rev. ed.). Baltimore: Johns Hopkins University Press. \\nHoyle, R.H., Pinkley, R.L., & Insko, e.A. (1989). Perceptions of behavior: Evidence of \\ndiffering expectations for interpersonal and intergroup interactions. Personality and Social \\nPsychology Bulletin, 15, 365-376. \\nHuguet, P., Mugny, G., & Perez, J.A. (1990-1991). Influence sociale et processus de \\ndecentration. Bulletin de Psychologie, 45, 155-163. \\nHui, e.H. (1988). Measurement of individualism-collectivism. Journal o\\nf\\n Research in \\nPersonality, 28, 17-36. \\nHui, e.H., & Triandis, H.e. (1986). Individualism-collectivism: A study of cross-cultural \\nresearchers. Journal o\\nf\\n Cross-Cultural Psychology, 1 7, 225-248. \\nHuici, e. (1 984). The individual and social functions of sex role stereotypes. In H. Tajfel (Ed.), \\nThe social dimension: European developments in social psychology, Vol. 2. Cambridge: \\nCambridge University Press. \\nHuici, C. (1985). Relaciones intergrupales II: EI paradigma del Grupo Minimo. In e. Huici \\n(Ed.), Estructura y procesos de Gru\\npo, Vol. 2. Madrid: UNED. \\nHurtig, M.e., & Pichevin, M.F. (1986). La dif\\nference des sexes. Paris: Tierce. \\nHurtig, M.e., Pichevin, M.F., & Piolat, M. (1992). Sex and age as factors of asymmetry in the \\nperceived similarity between two persons. In L. Arcuri & e. Serino (Eds.), Asymmetry \\nphenomena in interpersonal comparison: Cognitive and social issues. Naples: Liguori. \\nInoff, G., Halverson, e.F., & Pizzigati, K.A.L. (1983). The influence of sex-role standards on \\nchildren's self\\n- and peer-attributions. Sex Roles, 9, 1205-1222. \\nInsko, e.A., Hoyle, R.H., Pinkley, R.L., Hong, G., Slim, R., Dalton, G., Lin, Y., Ruffin, P.P., \\nDardis, G.J., Bernthal, P.R., & Schopler, J. (1988). Individual-group discontinuity: The role \\nof a consensus rule. Journal o\\nf\\n Ex\\nperimental Social Psychology, 24, 505-5\\\\9. \\nInsko, e.A., Pinkley, R.L., Harring, K., Holton, B., Hong, G., Krams, D.S., Hoyle, R.H., & \\nThibaut, J. (1987). Beyond categorization to competition: Expectations of appropriate \\nbehavior. Representative Research in Social Psychology, 1 7, 5-36. \\nInsko, C.A., Pinkley, R.L., Hoyle, R.H., Dalton, G., Hong, G., Slim, R., Landry, P., Holton, \\nB., Ruffin, P.P., & Thibaut, J. (1987). Individual-group discontinuity: The role of \\nintergroup contact. Journal o\\nf\\n Ex\\nperimental Social Psychology, 23, 250-267. \\nInsko, e.A., & Schopler, J. (1987). Categorization, competition, and collectivity. In e. \\nHendrick (Ed.), Review o\\nf\\n personality and \\nsocial psychology: Group processes, Vol. 8. Beverly \\nHills, CA: Sage. \\nInsko, e.A., Schopler, J., Drigotas, S.M., Graetz, K.A., Kennedy, J., Cox, e., & Bomstein, G. \\n(1993). The role of communication in interindividual-intergroup discontinuity. Journal o\\nf\\n \\nConflict Resolution, 37, 108-1 38. \\nInsko, e.A., Schopler, J., Hoyle, R.H., Dardis, G.J., & Graetz, K.A. (1990). Individual-group \\ndiscontinuity as a function of fear and greed. Journal o\\nf\\n Personality and Social Psychology, \\n58, 68-79. \\nInsko, e.A., Schopler, J., Kennedy, J.F., Dahl, K.R., Gratez, K.A., & Drigotas, S.M. (1992). \\nIndividual-group discontinuity from the differing perspectives of Campbell's realistic group \\nconflict theory and Tajfel and Turner's social identity theory. Social Psychological Quarterly, \\n55, 272-291 . \\nJackson, L.A. (1993). Achieving positive social identity. Paper presented at the Invitational \\nConference on Self and Social Identity. Highland Park, FL, June. \\nJames, W. (\\\\890). The princi\\nples o\\nf\\n psychology. New York: Holt. \\nJames, W. (1892). Psychology: A brie\\nf\\ner course. Greenwich, CT: Fawcett. \\n\\n\\n240 \\nSocial identity: international perspectives \\nJanis, I.L. (1968). Group identification under conditions of external danger. In D. Cartwright & \\nA. Zander (Eds.), Group d\\nynamics (3rd ed.). London: Tavistock; New York: Harper & Row. \\nJarymowicz, M. (1982). Perception of self-others dissimilarity and signs of attractiveness or \\naversiveness of diversity from others. Polish Psychological Bulletin, 13, 39-44. \\nJarymowicz, M. (1987). Perceiving one's own individuality: Estimation and attractiveness o\\nf\\n \\none's own distinctiveness f\\nrom others. Warsaw: Wydawnictwa Uniwersytetu Warszawskiego. \\nJarymowicz, M. (l99Ia). The self-distinctive traits as personal identity attributes. CaMers de \\nPsychologie Cognitive, 11, 679-695. \\nJarymowicz, M. (199Ib). La distinctive Soi-Nous-Autre comme base d'identite individuelle et \\nsocia \\nIe. Cahiers Internationaux de Psychologie Sociale, 9-10, 103-1 16. \\nJarymowicz, M. (I \\n992a). Self, We and Other(s): Schemata distinctiveness and altruism. In P. \\nOliner, S. Oliner, L. Baron, D. Krebs & Z. Smoleska (Eds.), Embracing the other: \\nPhilosophical, psychological, and historical perspectives on altruism. New York: University of \\nNew York Press. \\nJarymowicz, M. ( l992b). Introduction to the Social Self \\nSymposium: Searching \\nf\\nor the social self \\nconcept. Paper presented to the EAESP/SESP Joint Meeting, LeuveniLouvain-la-Neuve, \\nJuly. \\nJarymowicz, M. (l993a). Categorial vs cross-categorial social identity and outgroup involvement. \\nPaper presented to the EAESP General Meeting, Lisbon, September. \\nJarymowicz, M. (I 993b). An attempt to operationalize 'social' versus 'personal identity' \\nconstructs. In M. Jarymowicz (Ed.), To know self - to understand others. Delf\\nt: Eburon. \\nJarymowicz, M. (1994a). W stron̘ indywidualnej podmiotowoSci i zblizen z innymi: \\nPodmiotowe podstawy spolecznych identyfikacji. In M. Jarymowicz (Ed.), Poza \\negocentrycznt;l perspektywt;l widzenia siebie i swiata. Warsaw: Wydawnictwo Instytutu \\nPsychologii PAN. \\nJarymowicz, M. (1994b). 0 formach umyslowego ujmowania MY i ich zwilizku ze \\nspostrzeganiem Innych. In M. Jarymowicz (Ed.), Poza egocentrycznt;l perpsktywt;l widzenia \\nsiebie i swiata. Warsaw: Wydawnictwo Instytutu Psychologii PAN. \\nJarymowicz, M. (1994c). Poznawcza indiwiduacja a spoleczne identyfikacje. In M. Jarymowicz \\n(Ed.), Poza egocentrycznt;l perspektywt;l widzenia siebie i swiata. Warsaw: Wydawnictwo \\nInstytutu Psychologii PAN. \\nJarymowicz, M., & Greenwald, A.G. (1993). The anchoring function of the self in social \\njudgments: Introduction. In M.C. Hurtig, M.F. Pichevin & M. Piolat (Eds.), Studies on the \\nself and social cognition. London: World Scientific Publishing. \\nJodelet, D. (Ed.) (1989). Les representations sociales. Paris: Presses Universitaires de France. \\nJones, E.E., & DeCharms, R. (1967). Changes in social perception as a function of the personal \\nrelevance of behavior. Sociometry, 20, 175-185. \\nJones, E.E., & Gerard, H.B. (1967). Foundations o\\nf\\n social psychology. New York: Wiley. \\nJones, E. E., Wood, G., & Quattrone, G. (1981 ). Perceived variability of personal characteristics \\nin in-groups and out-groups: The role of knowledge and evaluation. Personality and Social \\nPsychology Bulletin, 7, 523-528. \\nJung, C.G. (1943). The psychology of the unconscious. In Collected Works. Princeton, NJ: \\nPrinceton University Press. \\nKarylowski, J. (1982). On the two types of altruistic behavior: doing good to feel good versus to \\nmake others feel good. In V. Derlega and J. Grzelak (Eds.), Cooperation and Helping \\nBehavior. New York: Academic Press. \\nKedourie, E. (1960). Nationalism. London: Hutchinson. \\nKelley, H.H., & Stahelski, A. (1970). The social interaction basis of cooperators' and \\ncompetitors' beliefs about others. Journal o\\nf\\n Personality and Social Psychology, 16, 66-9\\\\ .  \\nKelley, H.H., & Thibaut, J.W. (1978). Interpersonal relations: A theory o\\nf\\n interdependence. New \\nYork: Wiley. \\nKelly, J. (1988). Entertainment in individual and group behavior. In J.E. McGrath (Ed.), The \\nsocial psychology o\\nf\\n time. Newbury Park, CA: Sage. \\nKelman, H.C. (1969). Patterns of personal involvement in the national system: A social \\n\\n\\nRe\\nf\\nerences \\n241 \\npsychological analysis of political legitimacy. In 1.N. Rosenau (Ed.), International politics \\nand f\\noreign policy: A reader in research and theory. Glencoe, IL: Free Press. \\nKenny, D.A., & La Voie, L. (1985). Separating individual and group effects. Journal o\\nf\\n \\nPersonality and Social Psychology, 48, 339-348. \\nKiesler, CA., & Kiesler, S.B. (1969). Con\\nf\\normity. Reading, MA: Addison-Wesley. \\nKillian, L.M. (1964). Social movements. In R.E.L. Faris (Ed.), Handbook o\\nf\\n modern sociology. \\nChicago: Rand McNally. \\nKinder, D.R., & Sears, D.O. (1995). Public opinion and political action. In G. Lindzey & E. \\nAronson (Eds.), Handbook o\\nf\\n social psychology. Vol. 2. New York: Random House. \\nKlein, M. (1948). Contributions to psychoanalysis 1921-1945. London: Hogarth. \\nKluckhohn, C (1951). Values and value-orientations in the theory of action: An exploration in \\ndefinition and classification. In T. Parsons & E. Shils (Ed.), Toward a general theory o\\nf\\n \\naction. Cambridge, MA: Harvard University Press. \\nKobuszewska, E. (1989). Mechanizmy altruizmu w warunkach deindywiduacji. Unpublished \\nmaster's thesis, Wydzial Psychologii, Uniwersytet Warszawski, Warsaw. \\nKohlberg, L. (1976). Moral stages and moralization: The cognitive-developmental approach. In \\nT. Lickona (Ed.), Moral development and behavior. New York: Holt, Rinehart & Winston. \\nKohn, H. (1944). The idea of \\nnationalism: A stud\\ny o\\nf\\n its origins and background. New York: \\nMacmillan. \\nKohn, H. (1962). The age o\\nf\\n nationalism: The first era o\\nf\\n global history. New York: Harper. \\nKotkin, M. (1983). Sex roles among married and unmarried couples. Sex Roles, 9, 975-985. \\nKrech, D., Crutchfield, R.S., & Ballachey, E.L. (1962). Individual in society. New York: \\nMcGraw-Hill. \\nKruglanski, A.W. (1989). Lay epistemics and human knowledge: Cognitive and motivational \\nbases. New York: Plenum. \\nKuhn, M.H., & McPartland, T.S. (1954). An empirical investigation of self-attitudes. American \\nSociological Review, 19, 68-76. \\nLalonde, R.N. (1992). The dynamics of group differentiation in the face of defeat. Personality \\nand Social Psychology Bulletin, 18, 336-342. \\nLane, R.E. (1962). Political ideology. New York: Free Press. \\nLassaline, M.E., Wisniewski, E.l., & Medin, D.L. (1992). Basic levels in natural categories: Are \\nall basic levels created equal? In B. Bums (Ed.), Percepts, concepts and categories. New \\nYork: Elsevier Science Publishers. \\nLatani:, 8., & Darley, 1.M. (1970). T\\nhe unres\\nponsive bystander: Why doesn't he hel\\np? New \\nYork: Appleton-Century-Crofts. \\nLatane, B., Williams, K., & Harkins, S. (1979). Many hands make light the work: The causes \\nand consequences of social loafing. Journal o\\nf\\n Personality and Social Psychology, 37, 822-\\n832. \\nLeBon, G. (1 896). The crowd. London: Unwin. \\nL'Ecuyer, R. (1981). The development of the self-concept through the life span. In M.D. \\nLynch, A.A. Norem-Hebeisen & K. Gergen (Eds.), Self concept: Advances in research. \\nCambridge, MA: Ballinger. \\nLemaine, G. (1974). Social differentiation and social originality. European Journal o\\nf\\n Social \\nPsychology, 4, 17-52. \\nLemaine, G., Kastersztein, 1., & Personnaz, B. (1978). Social differentiation. In H. Tajfel (Ed.), \\nDif\\nf\\nerentiation between social groups: Studies in the social psychology o\\nf\\n intergroup relations. \\nLondon: Academic Press. \\nLevine, I.M. (1980). Reactions to opinion deviance in small groups. In P.B. Paulus (Ed.), \\nPsychology o\\nf\\n group in\\nfluence. Hillsdale, Nl: Erlbaum. \\nLevine, 1.M., & Moreland, R.L. (1994). Group socialization: Theory and research. In W. \\nStroebe and M. Hewstone (Eds.), European review o\\nf\\n social psychology, Vol. 5. Chichester: \\nWiley. pp. 305-336. \\nLe Vine, R.A., & Campbell, D.T. (1972). Ethnocentrism: T\\nheories o\\nf\\n con\\nflict, ethnic attitudes \\nand group con\\nflict. New York: Wiley. \\n\\n\\n242 \\nSocial identity: international pers\\npectives \\nLevy, S.B. (1975). Shifting patterns of ethnic identification among the Hassidim. In J.W. \\nBennet (Ed.), The new ethnicity: Pers\\npectives \\nf\\nrom ethnology. St Paul, MN: West Publishing. \\nLewin, K. (1948). Resolving social con\\nflicts. New York: Harper & Row. \\nLewin, K., Lippitt, R., & White, R. (1939). Patterns of aggressive behavior in experimentally \\ncreated social climates. Journal o\\nf\\n Psychology, 10, 271-299. \\nLeyens, J.-P. (1983). Sommes nous tous des psychologues? Brussels: Mardaga. \\nLeyens, J.-P., Yzerbyt, V.Y., & Schadron, G. (1994). Stereotypes and social cognition. London: \\nSage. \\nLindskold, S. (1983). Conflict and conciliation with groups and individuals. In A.P. Hare, V. \\nKent & M. Davies (Eds.), Small groups and social interaction, Vol. 2. New York: Wiley. \\nLinville, P.W., & Jones, E.E. (1980). Polarized appraisals of out-group members. Journal o\\nf\\n \\nPersonality and Social Psychology, 38, 689-703. \\nLinz, J.J. (1986). El con\\nflicto en Euskadi. Madrid: Espasa-Calpe. \\nLord, R.G., Foti, R.J., & De Vader, c.L. (1984). A test of leadership categorization theory: \\nInternal structure, information processing, and leadership perceptions. Organizational \\nBehaviour and Human Per\\nf\\normance, 34, 343-378. \\nLorenzi-Cioldi, F. (1988). Individus dominants et groupes domines: Images masculines et \\njeminines. Grenoble: Presses Universitaires de Grenoble. \\nLorenzo-Cioldi, F. (1993). They all look alike, but so do we . . .  \\nsometimes. Journal o\\nf\\n Social \\nPsychology, 32, 1 1 1-24. \\nLorenzi-Cioldi, F. (1994). Les androgynes. Paris: Presses Universitaires de France. \\nLorenzi-Cioldi, F., & Doise, W. (1990). Levels of analysis and social identity. In D. Abrams \\nand M.A. Hogg (Eds.), Social identity theory, constructive and critical advances. Hemel \\nHempstead: Harvester Wheatsheaf. \\nLorenzi-Cioldi, F., & Doise, W. (1994). Identite personelle et identitee sociale. In R.Y. Bourhis \\n& J.-P. Leyens (Eds.), Stereotypes, discrimination et relations intergroupes. Brussels: \\nMardaga. \\nLorenzi-Cioldi, F., Eagley, A.M., & Stewart, T. (1995). Homogeneity of gender groups in \\nmemory. Unpublished manuscript, University of Geneva. \\nLott, A.J., & Lott, B.E. (1965). Group cohesiveness as interpersonal attraction: A review of \\nrelationships with antecedent and consequent variables. Psychological Bulletin, 64, 259-309. \\nLuthanen, R., & Crocker, J. (1 992). A collective self-esteem scale: Self-evaluation of one's \\nsocial identity. Personality and Social Psychology Bulletin, 3, 302-318. \\nLuria, A.R. (1 976). La storia sociale dei processi cognitivi. Florence: Giunti-Barbera. \\nMaass, A., & Clark, R.D., III. (1984). Hidden impact of minorities: Fifteen years of minority \\ninfluence research. Psychological Bulletin, 95, 428-450. \\nMaass, A., Clark, R.D., III, & Haberkorn, G. (1982). The effects of differential ascribed \\ncategory membership and norms on minority influence. European Journal o\\nf\\n Social \\nPsychology, 12, 89-104. \\nMackie, D.M., & Cooper, J. (1984). Attitude polarization: The effects of group membership. \\nJournal o\\nf\\n Personality and Social Psychology, 46, 575-585. \\nMackie, D.M., Worth, L.T., & Asuncion, A.G. (1990). Processing of persuasive in-group \\nmessages. Journal o\\nf\\n Personality and Social Psychology, 58, 812-822. \\nMakris-Botsaris, E., & Robinson, P.W. (1991). Harter's self-perception profile for children: A \\ncross-cultural validation in Greece. Evaluation and Research in Evaluation, 5, 135-143. \\nMarch, J.G., & Simon, H.A. (1958). Organizations. New York: Wiley. \\nMarcia, J.E. (1966). Development and validation of ego identity status. Journal o\\nf\\n Personality \\nand Social Psychology, 3, 551 -558. \\nMariategui, J.C. (1925). EI hombre y el mito. In J.C. Mariategui, El alma matinal. Lima: \\nAmatuam, 1 950. \\nMarkus, H. (1977). Self-schemas and processing information about the self. Journal o\\nf\\n \\nPersonality and Social Psychology, 35, 63-78. \\nMarkus, H. (1984). Possible selves. Paper presented at the Boston University Symposium for \\nthe Interdisciplinary Study of Personality. \\n\\n\\nRe\\nf\\nerences \\n243 \\nMarkus, H., & Kitayama, S. (l99Ia). Culture and the self: Implications for cognition, emotion \\nand motivation. Psychological Review, 98, 224-253. \\nMarkus, H., & Kitayama. S. (l99Ib). Cultural variations in the self-concept. In J. Strauss & \\nG.R. Goethals (Eds.), The sel\\nf: An interdisc i\\nplinary approach. New York: Springer Verlag. \\nMarkus, H., & Smith, J. (1981). The influence of self-schemata on the perception of others. In \\nN. Cantor & J. Kihlstrom (Eds.), Personality, cognition and social interaction. Hillsdale, NJ: \\nErlbaum. \\nMarques, J.M. (1990). The black sheep effect: Outgroup homogeneity in social comparison \\nsettings. In D. Abrams & M.A. Hogg (Eds.), Social identity theory: Constructive and critical \\nadvances. Hemel Hempstead: Harvester Wheatsheaf. \\nMarques, J.M., Abrams, D., & Paez, D. (1996). Unpublished manuscript. \\nMarques, J.M., Brito, P.N., Correia, M.G., & Serodio, R.G. (1996). Evaluation of attractive \\nand unattractive ingroup and outgroup members as a function of group uniformity. \\nUnpublished manuscript. \\nMarques, J.M., & Paez, D. (1994). The black sheep eff\\nect: Social categorization, rejection of \\ningroup deviates, and perception of group variability. In M. Hewstone & W. Stroebe (Eds.), \\nEuropean review o\\nf\\n social psychology, Vol. 5. Chichester: Wiley. \\nMarques, I.M., Robalo, E.M., & Rocha, S.A. (1992). Ingroup bias and the black sheep effect: \\nAssessing the impact of social identification and perceived variability on group judgments. \\nEuropean Journal o\\nf\\n Social Psychology, 22, 331-352. \\nMarques, J.M., & Yzerbyt, V.Y. (1988). The black sheep effect: Judgmental extremity towards \\ningroup members in inter- and intragroup situations. European Journal o\\nf\\n Social Psychology, \\n18, 287-292. \\nMarques, J.M., Yzerbyt, V.Y., & Leyens, J.-P. (1988). The black sheep effect: Judgmental \\nextremity towards ingroup members as a function of group identification. European Journal \\no\\nf\\n Social Psychology, 18, 1-16. \\nMartin, R. (I 988a). Ingroup and outgroup minorities: Diff\\nerential impact upon public and \\nprivate response. European Journal o\\nf\\n Social Psychology, 18, 39-52. \\nMartin, R. (1988b). Minority influence and social categorization: A replication. European \\nJournal of Social Psychology, 18, 369-373. \\nMartin, R. (1988c). Minority influence and 'trivial' social categorization. European Journal o\\nf\\n \\nSocial Psychology, 18, 465-470. \\nMaslow, A.H. (1954). Motivation and personality. New York: Harper & Row. \\nMcAdams, D.P. (1980). A thematic coding system for the intimacy motive. Journal o\\nf\\n Research \\nin Personality, 14, 412-432. \\nMcAdams, D.P. (1985). The 'Imago': A key narrative component of identity. In P. Shaver \\n(Ed.), Self, situations and social behavior. Beverly Hills, CA: Sage. \\nMcCallum, D.M., Harring, K., Gilmore, R., Drenan, S., Chase, J., Insko, CA., & Thibaut, 1. \\n(1985). Competition between groups and between individuals. Journal o\\nf\\n Ex\\nperimental \\nSocial Psychology, 21, 301-320. \\nMcDougall, W. (1920). The group mind. London: Putnam. \\nMcGarty, C, & Turner, J.C (1992). The effects of categorization on social judgement. British \\nJournal o\\nf\\n Social Psychology, 31, 253-268. \\nMcGarty, C, Turner, J.C., Hogg, M.A., David, 8., & Wetherell, M.S. (1992) Group \\npolarization as conformity to the prototypical group member. British Journal o\\nf\\n Social \\nPsychology, 31, 1-20. \\nMcGrath, J.E. (1988). Time and social psychology. In J.E. McGrath (Ed.), The social \\npsychology o\\nf\\n time. Newbury Park, CA: Sage. \\nMcGuire, W.J., & McGuire, C.V. (1988). Content and process in the experience of the self. In \\nL. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 21. New York: \\nAcademic Press. \\nMcPhail, C. (1991). The myth o\\nf\\n the madding crowd. New York: Aldine de Gruyter. \\nMead, G.H. (1934). Mind, self and society. Chicago: University of Chicago Press. \\nMedin, D.L. (1989). Concepts and conceptual structure. American Psychologist, 44, 1469-1481. \\n\\n\\n244 \\nSocial identity: international pers\\npectives \\nMedin, D.L., & Ortony, A. (1989). Psychological essentialism. In S. Vosniadou & A. Ortony \\n(Eds.), Similarity and analogical reasoning. Cambridge: Cambridge University Press. \\nMedin, D.L., & Wattenmaker, W.D. (1987). Category cohesiveness, theories, and cognitive \\narchaeology. In U. Neisser (Ed.), Concepts and conceptual development: Ecological and \\nintellectual f\\nactors in categorization. Cambridge: Cambridge University Press. \\nMedin, D.L., Wattenmaker, W.D., & Hampson, S.E. (1987). Family resemblance, conceptual \\ncohesiveness and category construction. Cognitive Psychology, 19, 242-279. \\nMervis, CB., & Rosch, E. (1981). Categorization of natural objects. In M.R. Rozenzweig & \\nL.W. Porter (Eds.), Annual review o\\nf\\n psychology, Vol. 32. Palo Alto, CA: Annual Reviews. \\nMessick, D., & Mackie, D. (1989). Intergroup relations. Annual Review o\\nf\\n Psychology, 40, 45-\\n81. \\nMlicki, P.P., & Ellemers, N. (1996). Being different or being better? National stereotypes and \\nidentifications of Polish and Dutch students. European Journal o\\nf\\n Social Psychology, 26, 97-\\n1 14. \\nMonteil, J.M. (1992). Intergroup differentiation and individuation: The effect of social \\ndeprivation. European Bulletin o\\nf\\n Cognitive Psychology, 12, 189-203. \\nMontero, M. (1987). A traves del espejo: Una aproximacion teorica al estudio de la conciencia \\nsocial en America Latina. In M. Montero (Ed.), Psicologia Politica Latinoamerica. Caracas: \\nPanapo. \\nMontero, M. ( 1990). Autoimagen de los venezolanos: Lo positivo y 10 negativo. In H. \\nRiquelme (Ed.), Buscando America Latina. Caracas: Nueva Sociedad. \\nMorales, J \\n.F. (1989). Identidad social y personal. In J. Mayor & J.L. Pinillos (Eds.), Tratado de \\npsicologia general. 7: Creencias, actitudes y valores. Madrid: Alhambra. \\nMorales, J.F., & Huici, C. (1994). Procesos grupales. In J.F. Morales, M. Moya, J.M. \\nFernandez-Dols, C Huici, J. Marques, D. Paez & J.A. Perez (Eds.), Psicologia social. \\nMadrid: McGraw-Hill. \\nMorales, J.F., Lopez, M., & Vega, L. (1992). Individualismo, collectivismo e identidad social. \\nRevista de Psicologia Social, Monogra\\nfico, 49-72. \\nMorales, J.F. & Paez, D. (1996). Estereotipos, discriminacion y relaciones intergrupos en \\nEspana y Latinoamerica. In R.Y. Bourhis & J.-P. Leyens (Eds.), Estereotipos discriminacion \\ny relaciones entre grupos. Madrid: McGraw-Hill. \\nMoreland, R.L. (1987). The formation of small groups. In C. Hendrick (Ed.), Review o\\nf\\n \\npersonality and social psychology: Group processes, Vol. 8. Beverly Hills, CA: Sage. \\nMoreland, R.L., & Levine, J.M. (1982). Socialization in small groups: Temporal changes in \\nindividual-group relations. In L. Berkowitz (Ed.), Advances in experimental social \\npsychology, Vol. 15. New York: Academic Press. \\nMoreland, R.L., & Levine, J.M. (1988). Group dynamics over time: Development and \\nsocialization in small groups. In J.E. McGrath (Ed.), The social psychology o\\nf\\n time. \\nNewbury Park, CA: Sage. \\nMorris, W.N., Worchel, S., Bios, J., Pearson, J., Roundtree, C, Samaha, G., Wachtler, J., & \\nWright, S. (1976). Collective coping with stress: Group reactions to fear, anxiety, and \\nambiguity. Journal o\\nf\\n Personality and Social Psychology, 33, 674-679. \\nMoscovici, S. (1976). Social in\\nfluence and social changes. London: Academic Press. \\nMoscovici, S. (1980). Toward a theory of conversion behavior. In L. Berkowitz (Ed.), Advances \\nin experimental social psychology, Vol. 13. New York: Academic Press. \\nMoscovici, S., & Mugny, G. (1983). Minority influence. In P. Paulus (Ed.), Basic group process. \\nNew York: Springer Verlag. \\nMoscovici, S., & Nemeth, CJ. (1974). Minority influence. In CJ. Nemeth (Ed.), Social \\npsychology: Classic and contemporary integrations. Chicago: Rand McNally. \\nMoya, M.C (1987). Percepcion de es\\npisodios y estereotipia sexual. Tesis Doctoral, Servicio de \\nPublicaciones, Universidad de Granada. \\nMoya, M.C. (1989). Genero, estereotipia y relaciones intergrupales. In P. Ballarin & T. Ortiz \\n(Eds.), La mu\\njer en Andalucia. Granada: Universidad de Granada. \\nMoya, M.C, & Peplau, L.A. (in preparation). 1dentidad de genero y relaciones de pare\\nja. \\nMugny, G. (1982). The power o\\nf\\n minorities. London: Academic Press. \\n\\n\\nRe\\nf\\nerences \\n245 \\nMugny, G., Kaiser, e., & Papastamou, S. (1983). Influence minontaue, identification et \\nrelations entre groupes: Etude experimentale autour d'une votation. Cahiers de Psychologie \\nSociale, 19, 1-30. \\nMugny, G., Kaiser, e., Papastamou, S., & Perez, J.A. (1984). Intergroup relations, \\nidentification and social influence. British Journal o\\nf\\n Social Psychology, 23, 31 7-322. \\nMugny, G., & Papastamou, S. (1982). Minority influence and psycho-social identity. European \\nJournal o\\nf\\n Social Psychology, 12, 379-394. \\nMugny, G., & Perez, J.A. (1985). Influence sociale, conflit et identification: Etude \\nexperimentale autour d'une persuasion 'manquee' lors d'une votation. Cahiers de \\nPsychologie Sociale, 6, 1-13. \\nMugny, G., & Perez, J.A. (199Ia). T\\nhe social psychology o\\nf\\n minority influence. Cambridge: \\nCambridge University Press. \\nMugny, G., & Perez, J.A. (199Ib). EI efecto de criptomnesia social. Anthropos: Revista de \\nDocumentacian Cientifica de la Cultura, 27, 1 10-1 12. \\nMullen, B., Brown, R.J., & Smith, C. (1992). Ingroup bias as a function of salience, relevance \\nand status: An integration. European Journal o\\nf\\n Social Psychology, 22, 103-1 12. \\nMullen, B., & Hu, L. (1989). Perceptions of in \\ngroup and outgroup variabilty: A meta-analytic \\nintegration. Basic and Applied Social Psychology, 10, 233-252. \\nMurphy, G.L., & Medin, D.L. (1985). The role of theories in conceptual coherence. \\nPsychological Review, 92, 289-3 16. \\nMussolini, B. (1935). Fascism: Doctrine and institutions. Rome: Ardita. \\nNakbi, J.L. (1990). Evolution de I'identite psychosociale feminine et masculine de l'enfance a \\nl'age adulte. Cahiers de Psychologie Sociale, 34, 1 -9. \\nNakbi, J.L., & Arnal-Duchemin, M.J. (1987). Dynamique des representations chez les femmes \\net les hommes. Cahiers de Psychologie Sociale, 34, 1-9. \\nNascimento-Schulze, C.M. (1993). Social comparison, group identity and prof\\nessional identity: \\nA study with bank clerks. Revista de Psicologia Social, 8, 69-82. \\nNeisser, U. (Ed.). (1987). Concepts and conceptual development: Ecological and intellectual \\nf\\nactors in categorization. Cambridge: Cambridge University Press. \\nNemeth, C. (1986). Differential contributions of majority and minority influence. Psychological \\nReview, 93, 1-10, 23-32. \\nNemeth, e., & Staw, B.M. (1989). The tradeoffs of social control and innovation in groups and \\norganizations. In L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 23. \\nOrlando, FL: Academic Press. \\nOakes, P.J. (1987). The salience of social categories. In J.e. Turner, M.A. Hogg, P.l. Oakes, \\nS.D. Reicher & M.S. Wetherell, Rediscovering the social group: A self\\n-categorization theory. \\nOxford: Blackwell. \\nOakes, P.J. (1996). The categorization process: Cognition and the group in the social \\npsychology of stereotyping. In P. Robinson (Ed.), Social groups and identity: Developing the \\nlegacy o\\nf\\n Henri T\\na\\nj\\nf\\nel. Oxford: Butterworth Heinemann. \\nOakes, P.J., Haslam, S.A., & Reynolds, K.J. (in press). Social categorization and social \\ncontext: An example from stereotype change. In D. Abrams & M.A. Hogg (Eds.), Social \\nidentity and social cognition. Oxford: Blackwell. \\nOakes, P.J., Haslam, A.S., & Turner, J.e. (1994). Stereotyping and social reality. Oxford: \\nBlackwell. \\nOakes, P.J., & Turner, J.C. (\\\\986). Distinctiveness and the salience of social category \\nmemberships: Is there an automatic perceptual bias towards novelty? European Journal o\\nf\\n \\nSocial Psychology, 16, 325-344. \\nOakes, P.J., & Turner, J.C. (\\\\990). Is limited infonnation processing capacity the cause of \\nsocial stereotyping? In W. Stroebe and M. Hewstone (Eds.), European review o\\nf\\n social \\npsychology, Vol. I. Chichester: Wiley. pp. 1 1 1-136. \\nOakes, P.J., Turner, J.e., & Haslam, S.A. (1991). Perceiving people as group members: The \\nrole of fit in the salience of social categorizations. British Journal o\\nf\\n Social Psychology, 30, \\n125-144. \\nO'Dea, T.F. (1957). The Mormons. Chicago: University of Chicago Press. \\n\\n\\n246 \\nSocial identity: international pers\\npectives \\nOlden \\nquist, A. (1992). Autonomy, social identities, and alienation. In F. Geyer & W.R. Heinz \\n(Eds.), Alienation, society, and the individual. New Brunswick, NJ: Transaction Publishers. \\nOsgood, C.E. (1962). An alternative to war or surrender. Urbana: University of Illinois Press. \\nOstrom, T.M., & Sedikides, C. (1992). Out-group homogeneity effect in natural and minimal \\ngroups. Psychological Bulletin, 112, 536-552. \\nPaez, D., Arrospide, l., Martinez-Taboada, C., & Ayestaran, S. (1992). Identidad social, \\nautoesquematizacion y autoconciencia collectiva: Investigaciones correlacionales en la \\nperspectiva de la teona de la identidad social. Revista de Psicologia Social, Monogrti\\nfico, 3-\\n20. \\nPalmonari, A., Carugati, F., Ricci-Bitti, P., & Sarchielli, G. (1979). Identita imper\\nf\\nette. \\nBologne: II Mulino. \\nPalmonari, A., Pombeni, M.L., & Kirchler, E. (1989). Peergroups and evolution of the self­\\nsystem in adolescence. European Journal o\\nf\\n Psychology o\\nf\\n Education, 4, 3-15. \\nParelman, A. (1983). Emotional intimacy in marriage: A sex-roles perspective. Ann Arbor, MI: \\nUMI Research Press. \\nPark, B., & Rothbart, M. (1982). Perception of out-group homogeneity and levels of social \\ncategorization: Memory for subordinate attributes of in-group and out-group members. \\nJournal o\\nf\\n Personality and Social Psychology, 42, 1051-1068. \\nParsons, T. (1968). The position of identity in the general theory of action. In G. Gordon & \\nKJ. Gergen (Eds.), The self in social interaction. New York: Wiley. \\nPavelchak, M.A. (1989). Piecemeal and category-based evaluation: An idiographic analysis. \\nJournal o\\nf\\n Personality and Social Psychology, 56, 354-363. \\nPawluczuk, B. (1992). Indywiduac\\nja systemu norm a toleranc\\nja wobec obcych. Unpublished \\nresearch report, Wydzial Psychologii, Uniwersytet Warszawski, Warsaw. \\nPeabody, D., & Goldberg, L.R. (1989). Some determinants of factor structures from \\npersonality-descriptors. Journal o\\nf\\n Personality and Social Psychology, 57, 552-567. \\nPemberton, MJ., Insko, C.A., & Schopler, l. (1996). Memory for and experience of differential \\ncompetitive behavior of individuals and groups. Journal o\\nf\\n Personality and Social \\nPsychology, 71, 5, 953-966. \\nPeplau, L.A. (1983). Roles and gender. In H.H. Kelley, E. Berscheid, A. Christensen, l. \\nHarvey, T.L. Huston, G. Levinger, E. McClintock, A. Peplau & D.R. Peterson (Eds.), Close \\nrelationshi\\nps. New York: W.H. Freeman & Co. \\nPeplau, L.A., & Gordon, S.L. (1985). Women and men in love: Gender differences in close \\nheterosexual relationships. In V.E. O'Leary, R.K. Under & B.S. Wallston (Eds.), Women, \\ngender, and social psychology. Hillsdale, Nl: Erlbaum. \\nPerez, l.A., & Mugny, G. (1985a). Categorizacion e influencia minoritaria. Anuario de \\nPsicologia. 32, 100-1 16. \\nPerez, l.A., & Mugny, G. (1985b). Influencia minoritaria sobre las opiniones frente al aborto y \\nlos anticonceptivos. Estudios de Psicologia, 23124, 29-54. \\nPerez, l.A., & Mugny, G. (I 986a). Efectos paradojicos de la categorizaci6n en la influencia \\nminoritaria. Boletin de Psicologia, 12, 65-89. \\nPerez, l.A., & Mugny, G. (1986b). Induction experimentale d'une influence minoritaire \\nindirecte. Cahiers de Psychologie Sociale, 32, 15-24. \\nPerez, l.A., & Mugny, G. (1987). Paradoxical effects of categorization in minority influence: \\nWhen being an out-group is an advantage. European Journal o\\nf\\n Social Psychology, 17, 1 57-\\n169. \\nPerez, l.A., & Mugny, G. (1990). Minority influence, manifest discrimination and latent \\ninfluence. In D. Abrams & M.A. Hogg (Eds.), Social identity theory: Constructive and \\ncritical advances. Heme! Hempstead: Harvester Wheatsheaf. \\nPerez, l.A., Mugny, G., Llavata, E., & Fierres, R. (1993). Paradoxe de la discrimination et \\nconflit culture!: Etudes sur Ie racisme. In l.A. Perez, G. Mugny, et aI., I\\nn\\nfluences sociales: La \\ntheorie de l'elaboration du con\\nflit. Paris: Delachaux et Niestle. \\nPerez, l.A., Mugny, G., & Moscovici, S. (1986). Les eff\\nets paradoxaux du deni dans l'influence \\nsociale. Cahiers de Psychologie Sociale, 32, 1-14. \\nPerez, l.A., Mugny, G., & Navarro, E. (1991). EI efecto de la 'paralisis intragrupal': Niveles de \\n\\n\\nRe\\nf\\nerences \\n247 \\ncategorizaci6n del yo y niveles de influencia social. Anthropos: Revista de Documentacion \\nCientifica de la Cultura, 27, 121-132. \\nPerez, J.A., Mugny, G., Roux, P., & Butera, F. (1991). Influence via la comparaison et \\ninfluence via la validation. In J.L. Beauvois, R.V. Joule & J.M. Monteil (Eds.), Perspectives \\ncognitives et conduites sociales, Vol. 3. Cousset: Delval. \\nPfost, K.S., & Fiore, M. (1990). Pursuit of nontraditional occupations: Fear of success or fear \\nof not being chosen? Sex Roles, 23, 1 5-24. \\nPhinney, J.S. (1990). Ethnic identity in adolescents and adults: Review of research. \\nPsychological Bulletin, 108, 499-514. \\nPiaget, J. (1975). L'equilibration de structures cognitives: Probleme central du developpement. \\nParis: Presses Universitaires de France. \\nPlatow, MJ., McClintock, e.G., & Liebrand, W. (1990). Predicting intergroup fairness and \\ningroup bias in the minimal group paradigm, European Journal o\\nf\\n Social Psychology, 20, \\n221-239. \\nPratto, F., & Bargh, J.A. (1991). Stereotyping based on apparently individuating information: \\nTrait and global components of sex stereotypes under attention overload. Journal o\\nf\\n \\nExperimental Social Psychology, 27, 26-47. \\nProshansky, H.M., Fabian, A.K., & Kaminoff, R. (1983). Place identity: Physical world \\nsocialization of the self. Journal o\\nf\\n Environmental Psychology, 3, 57-83. \\nPryor, J.B., & Ostrom, T.M. (1987). Social cognition theory of group processes. In B. Mullen & \\nG.R. Goethals (Eds.), Theories o\\nf\\n group behavior. New York: Springer Verlag. \\nQuattrone, G.A. (1986). On the perception of a group's variability. In S.G. Worchel & W.G. \\nAustin (Eds.), Psychology o\\nf\\n intergroup relations. Chicago: Nelson-Hall. \\nQuevedo, S. (1972). Revindicaci6n urbana y lucha politica. Revista Latinoamericana de \\nEstudios Urbanos Regionales, 2, 12-43. \\nRabbie, J.M., & Lodewijkx, H.G.M. (1994). Conflict and interaction: An individual-group \\ncontinuum. In B. Markovsky, K. Heimer & J. O'Brien (Eds.), Advances in group processes, \\nVol. 11. Greenwich, CT: JAI Press. \\nRavaud, J.F., Beaufils, B., & Paicheler, H. (1986). Handicap et integration scolaire. Sciences \\nSociales et Sante, 3-4, 167-194. \\nReicher, S., & Emler, N. (1986). The management of delinquent reputations. In H. Beloff (Ed.), \\nGetting into lif\\ne. London: Methuen. \\nReis, H.T., Nezlek, J., & Wheeler, L. (1980). Physical attractiveness in social interaction. \\nJournal o\\nf\\n Personality and Social Psychology, 38, 604-617. \\nReis, H.T., & Wheeler, L. (1991). StUdying social interaction with the Rochester Interaction \\nRecord. In M.P. Zanna (Ed.), Advances in experimental social psychology, Vol. 24. New \\nYork: Academic Press. \\nReykowski, J. (1979). Motywac\\nja, postawy prospoleczne a osobowoSi:. Warsaw: PWN. \\nReykowski, J. (1989). Dimensions of the development in moral values. In N. Eisenberg, J. \\nReykowski & E. Staub (Eds.), Social and moral values. Hillsdale, NJ: Erlbaum. \\nReynolds, V., Falger, V., & Vine, I. (1987). The sociobiology o\\nf\\n ethnocentrism. London: Croom \\nHelm. \\nRips, L.J. (1989). Similarity, typicality, and categorization. In S. Vosniadou & A. Ortony \\n(Eds.), Similarity and analogical reasoning. Cambridge: Cambridge University Press. \\nRivera, A.N. (1991). Psicologia y colonizaci6n. In M. Montero (Ed.), Accion y discurso: \\nProblemas de psicologia politica en America Latina. Caracas: Eduven. \\nRobinson, J.P. (1988). Time diary evidence about the social psychology of \\neveryday life. In J \\n.E. \\nMcGrath (Ed.), The social psychology o\\nf\\n time. Newbury Park, CA: Sage. \\nRodriguez, C.P. (1992). Estereotipos de nacionalidad en estudiantes colombianos y \\nvenezolanaos. Boletin A vepso, 15, 65-74. \\nRogers, C. (1951). Client-centered therapy. Boston: Houghton-Mifflin. \\nRokeach, M. (1973). The nature o\\nf\\n human values. New York: Free Press. \\nRook, D.S. (1987). Social support versus companionship: Effects of life stress, loneliness, and \\nevaluations by others. Journal o\\nf\\n Personality and Social Psychology, 52, 1 1 32-1 147. \\nRosch, E. (1973). On the internal structure of perceptual and semantic categories. In T.E. \\n\\n\\n248 \\nSocial identity: international perspectives \\nMoore (Eds.), Cognitive development and the acquisition o\\nf\\n language. New York: Academic \\nPress. \\nRosch, E. (1975). Cognitive reference points. Cognitive Psychology, 7, 532-547. \\nRosch, E. (1978). Principles of categorization. In E. Rosch & B.B. Lloyd (Eds.), Cognition and \\ncategorization. Hillsdale, NJ: Erlbaum. \\nRosch, E., Mervis, CB., Gray, W.D., Johnson, D.M., & Boyes-Braem, P. (1976). Basic objects \\nin natural categorization. Cognitive Psychology, 15, 346-378. \\nRosenberg, S., & Gara, M. (1985). The multiplicity of personal identity. Review o\\nf\\n Personality \\nand Social Psychology, 6, 87-1 13. \\nRoth, E.M., & Schoben, E.1. (1983). The effect of context on the structure of categories. \\nCognitive Psychology, 15, 346-378. \\nRubin, Z., Peplau, L.A., & Hill, CT. (1981). Loving and leaving: Sex differences in romantic \\nattachments. Sex Roles, 7, 821-835. \\nRush, M.C, & Russell, J.E.A. (1988). Leader prototypes and prototype-contingent consensus \\nin leader behaviour descriptions. Journal o\\nf\\n Experimental Social Psychology, 24, 88-104. \\nSalazar, J.M. (1983a). Bases psicologicas del nacionalismo. Mexico: Trillas. \\nSalazar, J.M. (1983b). On the psychological viability of 'Latin-Americanism'. International \\nSocial Science Journal, 35, 295-308. \\nSalazar, J.M. (1989). Niveles de identificacion y estructura cognoscitiva en relacion con el \\nlatinoamericano. Revista de Psicologia Social, 4, 13-22. \\nSalisbury, R.H. (1975). Interest groups. In F.I. Greenstein & N.W. Polsby (Eds.), Handbook o\\nf\\n \\npolitical science, Vol. 4. Reading, MA: Addison-Wesley. \\nSanchez-Mazas, M., Perez, J.A., Navarro, E., Mugny, G., & Jovanovic, J. (1993). De la \\nparalysie intragroupe au confiit normatif: Etudes dur I'avortement, la contraception et la \\nxenophobie. In J.A. Perez & G. Mugny (Eds.), Influences sociales: La tMorie de l'eboration \\ndu confiit. Paris: Delachaux et Niestle. \\nSande, G.N. (1990). The multifaceted self. In J.M. Olson & M.P. Zanna (Eds.), Self\\n-in\\nf\\nerence \\nprocesses: The Ontario Symposium, Vol. 6. Hillsdale, NJ: Erlbaum. \\nSanders, G.S., & Baron, R.S. (1977). Is social comparison irrelevant for producing choice \\nshifts? Journal o\\nf\\n Experimental Social Psychology, 13, 303-314. \\nSarbin, T.R., & Allen, V.L. (1968). Role theory. In G. Lindzey & E. Aronson (Eds.), Handbook \\no\\nf\\n social psychology, Vol. 1 (2nd ed.). Reading, Mass: Addison-Wesley. \\nSchachter, S. (1951). Deviation, rejection, and communication. Journal o\\nf\\n Abnormal and Social \\nPsychology, 46, 190-207. \\nSchopler, J., & Insko, CA. (1992). The discontinuity effect in interpersonal and intergroup \\nrelations: Generality and mediation. In W. Stroebe & M. Hewstone (eds.), European review \\no\\nf\\n social psychology, Vo!' 1. Chichester: Wiley. \\nSchopler, J., Insko, CA., Graetz, K.A., Drigotas, S.M., & Smith, V.A. (1991). The generality \\nof the individual-group discontinuity eff\\nect: Variations in positivity-negativity of outcomes, \\nplayers' relative power, and magnitude of outcomes. Personality and Social Psychology \\nBulletin, 1 7, 612-624. \\nSchopler, J., Insko, CA., Graetz, K.A., Drigotas, S.M., Smith, VA, & Dahl, K. (1993). \\nIndividual-group discontinuity: Further evidence for mediation by fear and greed. \\nPersonality and Social Psychology Bulletin, 19, 419-43 1. \\nSemin, G.R. (1987). On the relationship between representation of \\ntheories in psychology and \\nordinary language. In W. Doise & S. Moscovici (Eds.), Current issues in European social \\npsychology, Vol. 2. Cambridge: Cambridge University Press. \\nSerino, C (1988a). Biases egocentrici e fenomeni di asimmetria nei confronti Se/Altro. Giornale \\nItaliano di Psicologia, 15, 213-233. \\nSerino, C. (I \\n988b). Strategies et structure de la comparaison sociale: Quelques aspects de \\nl'asymmetrie soi/autrui dans les relations entre groupes. Cahiers de Psychologie Cognitive, 8, \\n627-648. \\nSerino, C (1989). Processi di confronto e argomentazione: In G. Bellelli (Ed.), II metodo del \\ndiscorso. Naples: Liguouri. \\nSerino, C (1990). La comparaison socia \\nIe et Ie structure des systemes categoriels: Quelques \\n\\n\\nRe\\nf\\nerences \\n249 \\nfI!flexions sur la prototypicalite du soi. Cahiers Internationaux de Psychologie Socia \\nIe, J, 73-\\n96. \\nSerino, C (1992). From the asymmetry phenomena to the prototypicality of self: Studies on the \\nstructure of social categories. In L. Arcuri & C Serino (Eds.), Asymmetry phenomena in \\ninterpersonal comparison: Cognitive and social issues. Naples: Liguori. \\nSerino, C (1993). Information processing and self/others comparison: Social-cognitive \\nextensions of a non-metrical approach to similarity. In M.C. Hurtig, Pichevin, M.F., & \\nM. Piolat (Eds.), Studies on the self and social cognition. London: World Scientific \\nPublishing. \\nSerino, C (1996). Similarity: Inter\\npersonal and intergroup comparisons in an asymmetrical \\nf\\nrame. Paper presented at the Small Group Meeting on 'Intergroup Relations: Current Work \\nand Future Perspectives', Catania, Italy, 26-9 September. \\nSerino, C, & Mazzara, B. (1995). Scripts e processi di confronto nella rappresentazione sociale \\ndella persona handicappata: Una ricerca sperimentale. Giornale Italiano di Psicologia, 22, \\n265-292. \\nSeton-Watson, H. (1965). Nationalism. old and new. Sydney: Sydney University Press. \\nShaw, M.E. (1976). Group d\\nynamics: The psychology o\\nf\\n small group behavior. New York: \\nMcGraw-Hill. \\nSherif, M. (1936). The psychology o\\nf\\n social norms. New York: Harper & Brothers. \\nSherif, M. (1951). A preliminary experimental study of inter-group relations. In 1.H. Rohrer & \\nM. Sherif (Eds.), Social psychology at the crossroads. New York: Harper & Brothers. \\nSherif, M. (1966). In common predicament: Social psychology o\\nf\\n intergroup conflict and \\ncooperation. Boston: Houghton Mifflin. \\nSherif, M. (1967). Group conflict and cooperation: T\\nheir social psychology. London: Routledge \\n& Kegan Paul. \\nSherif, M., Harvey, 0.1., White, B.1., Hood, W.R., & Sherif, CW. (1961). I\\nntergroup conflict \\nand cooperation: The Robbers' Cave experiment. Norman, OK: University Book Exchange. \\nShils, E. (1968). Ideology. In D.L. Sills (Ed.), International encylopedia o\\nf\\n the social sciences. \\nVol. 7. New York: Macmillan and Free Press. \\nSills, D.L. (1958). The volunteers. New York: Free Press. \\nSimmel, G. (1955). Conflict: T\\nhe web o\\nf\\n group a\\nffiliations. Chicago: Chicago Free Press. \\nSimon, B. (1994). The perception of ingroup and outgroup homogeneity: Reintroducing the \\nintergroup context. In W. Stroebe and M. Hewstone (Eds.), European review o\\nf\\n social \\npsychology. Vol. 3. Chichester: Wiley. pp. 1-3. \\nSimon, B., & Brown, R. (1987). Perceived intragroup homogeneity in minority-majority \\ncontexts. Journal o\\nf\\n Psychology and Social Psychology, 53, 703-71 1 .  \\nSimon, B., & Hamilton, D.L. (1994). Self stereotyping and social context: The effects of relative \\nin-group size and in-group status. Journal o\\nf\\n Personality and Social Psychology, 66, 699-\\n7 1 1 .  \\nSingh, A .  (1979). Reliability and validity of self-reported delinquency studies. Psychological \\nReports, 44, 987-993. \\nSingleton, R., 1r. (1979). Another look at the conformity explanation of group-induced shift in \\nchoice. Human Relations, 32, 37-56. \\nSmith, A.D. (I 984a). National identity and myths of ethnic descent. Research in Social \\nMovements. Con\\nflict and Change, 7, 95-230. \\nSmith, A.D. (1984b). Ethnic myths and ethnic revivals. European Journal o\\nf\\n Sociology, 25, \\n283-305. \\nSmith, A.D. (199 1). National identity. London: Penguin. \\nSmith, P.B., & Bond, M.H. (1993). Social psychology across cultures. New York: Harvester \\nWheatsheaf. \\nSnyder, CR., & Fromkin, H.L. (1980). Uniqueness: the human pursuit o\\nf\\n d\\niff\\nerence. New York: \\nPlenum Press. \\nSnyder, M.L., Kleck, B.E., Strentna, A., & Mentzer, S.l. (1 979). A comparison of egotism, \\nnegativity, and learned helplessness as explanations for poor performance after unsolvable \\nproblems. Journal o\\nf\\n Personality and Social Psychology, 40, 24-30. \\n\\n\\n250 \\nSocial identity: international pers\\npectives \\nSnyder, M.L., & Wicklund, R. (1981). Attribute ambiguity. In J.H. Harvey, W. Ickes & R.R. \\nKidd (Eds.), New directions in attribution research. New York: Halsted. \\nSobczyk, K. (1992). Schematy Ja/My/I\\nnni a spostrzegany d\\nystans i ustosunkowania wobec \\nnarodowosci blizsze\\nj i dalsze\\nj. Unpublished master's thesis, Wydzial Psychologii, Universytet \\nWarszawski, Warsaw. \\nSokolowska, G. (1992). Tozsamosciowe przes/anki ustosunkowim wobec obcych. Unpublished \\nresearch report, Wydzial Psychologii, Uniwersytet Warszawski, Warsaw. \\nSpears, R., Lea, M., & Lee, S. (1990). De-individuation and group polarization in computer­\\nmediated communication. British Journal o\\nf\\n Social Psychology, 29, 121-134. \\nSrull, T.K., & Gaelik, L. (1983). General principles and individual diff\\nerences in the self as a \\nhabitual ref\\nerence point: An examination of self-other judgments of similarity. Social \\nCognition, 2, 108-121. \\nStephenson, G.M. (1984). Interpersonal and intergroup dimensions of bargaining and \\nnegotiation. In H. Tajfel (Ed.), The social dimension: European developments in social \\npsychology, Vol. 2. Cambridge: Cambridge University Press. \\nStroup, H.H. (1945). T\\nhe Jehovah's Witnesses. New York: Russell & Russell. \\nSuls, J.M., & Miller, R.L. (Eds.). (1977). Social comparison processes: Theoretical and empirical \\npers\\npectives. Washington, DC: Hemisphere. \\nSumner, W.G. (1906). Folkways. New York: Ginn. \\nSzuster-Zbrojewicz, A. (1988). Z badan nad zwiazkami pobodienstwalodn;:nosci Ja-Inni z \\ngotowoSci̕ do egzocentrycznego angaiowania sil;: w sprawy pozaosobiste. In M. Jarymowicz \\n(Ed.), Studia nad spostrzeganiem relac\\nji Ja-I\\nnni: Tozsamosc, ind\\nywiduac\\nja, przynaleznosc. \\nWroclaw: Ossolineum. \\nSzuster-Zbrojewicz, A. (1 990). Formy reprezentacji MY a prze\\njawy pozaosobistego \\nzaangazowania. Unpublished report G-MEN-9/90, Wydzial Psychologii, Uniwersytet \\nWarszawski, Warsaw. \\nTajfel, H. (1959). Quantitative judgment in social perception. British Journal o\\nf\\n Psychology, 50, \\n16-29. \\nTajfel, H. (1969a). Cognitive aspects of prejudice. Journal o\\nf\\n Social Issues, 25, 79-97. \\nTajfel, H. (1969b). The formation of national attitudes: A social psychological perspective. In \\nM. Sherif & c. Sherif (Eds.), Interdisci\\nplinary relationshi\\nps in the social sciences. Chicago: \\nAldine. \\nTajfel, H. (1970a). Aspects of national and ethnic loyalty. Social Science I\\nn\\nf\\normation, 9, 1 19-\\n144. \\nTajfel, H. (1970b). Experiments in intergroup discrimination. Scientific American, 223, 96-102. \\nTajfel, H. (I \\n972a). La categorisation sociale. In S. Moscovici (Ed.), Introduction a la \\npsychologie sociale, Vol. 1 .  Paris: Larousse. \\nTajfel, H. (1972b). Experiments in a vacuum. In J. Israel & H. Tajfel (Eds.), T\\nhe context o\\nf\\n \\nsocial psychology: A critical assessment (European Monographs in Social Psychology No. 2). \\nLondon: Academic Press. \\nTajf\\nel, H. (1974). Social identity and intergroup behavior. Social Science Inf\\normation, 13, 65-\\n93. \\nTajfel, H. (1978a). Interindividual behaviour and intergroup behaviour. In H. Tajfel (Ed.), \\nDiff\\nerentiation between social groups: Studies in the social psychology o\\nf\\n intergroup behaviour. \\nLondon: Academic Press. \\nTajfel, H. (1978b). Social categorization, social identity and social comparison. In H. Tajfel \\n(Ed.), Dif\\nf\\nerentiation between social groups: Studies in the social psychology o\\nf\\n intergroup \\nrelations. London: Academic Press. \\nTajfel, H. (I978c). The achievement of group diff\\nerentiation. In H. Tajfel (Ed.), Diff\\nerentiation \\nbetween social groups: Studies in the social psychology o\\nf\\n intergroup relations. London: \\nAcademic Press. \\nTajfel, H. (Ed.). (1978d). Dif\\nf\\nerentiation between social groups: Studies in the social psychology \\nof intergroup relations. London: Academic Press. \\nTajfel, H. (1978e). The social psychology o\\nf\\n minorities. London: Minority Rights Group. \\n\\n\\nRe\\nf\\nerences \\n251 \\nTajfel, H. (1979). Individuals and groups in social psychology. British Journal o\\nf\\n Social \\nPsychology, 18, 183-190. \\nTajfel, H. (\\\\981). Human groups and social categories: Studies in social psychology. Cambridge: \\nCambridge University Press. \\nTajfel, H. (1982a). Instrumentality, identity, and social comparison. In H. Tajfel (Ed.), Social \\nidentity and intergroup relations. Cambridge: Cambridge University Press. \\nTajfel, H. (Ed.). (\\\\982b). Social identity and intergroup relations. Cambridge: Cambridge \\nUniversity Press. \\nTajfel, H. (1982c). Social psychology of intergroup relations. In M.R. Rosenzweig & L.R. \\nPorter (Eds.), Annual review o\\nf\\n psychology, Vol. 33. Palo Alto, CA: Annual Reviews. \\nTajfel, H., & Billig, M. (\\\\974). Familiarity and categorization in intergroup behavior. Journal \\no\\nf\\n Experimental Social Psychology, 10, 159-170. \\nTajfel, H., Flament, e., Billig, M., & Bundy, R.P. (1971). Social categorization and intergroup \\nbehaviour. European Journal o\\nf\\n Social Psychology, 1, 149-178. \\nTajfel, H., & Turner, J.C. (\\\\979). An integrative theory of intergroup conflict. In W.G. Austin \\n& S. Worchel (Eds.), The social psychology o\\nf\\n intergroup relations. Monterey, CA: Brooks/ \\nCole. \\nTajfel, H., & Turner, J.e. (1986). The social identity theory of intergroup behavior. In S. \\nWorchel & W.G. Austin (Eds.), The psychology o\\nf\\n intergroup relations. Chicago: Nelson­\\nHall. \\nTajfel, H., & Wilkes, A.L. (1963). Classification and quantitative judgement. British Journal o\\nf\\n \\nPsychology, 54, 101-1 14. \\nTap, P. (1979). L'identification: Est-elle une alienation de l'identite? In P. Tap (Ed.), Jdentite \\nindividuelle et personalisation. Toulouse: Privat. \\nTap, P. (1985). Masculin et f\\neminin chez I'en\\nf\\nant. Toulouse: Privat. \\nTaylor, D.M., & Moghaddam, F.M. (1987). T\\nheories o\\nf\\n intergroup relations. New York: \\nPraeger. \\nToch, H. (\\\\965). The social psychology o\\nf\\n social movements. Indianapolis: Bobbs-Merrill. \\nTorregrosa, J.R., & Ramirez, S. (1991). Identidades nacionales y estados plurinacionales: Una \\nperspectiva psicosociol6gica. lnteraccion Social, 1, 123-1 36. \\nTriandis, H.e. (\\\\ \\n988). Collectivism vs. individualism. In G. Verma & e. Bagley (Eds.), Cross­\\ncultural studies of \\npersonality, attitudes, and cognition. London: Macmillan. \\nTriandis, H.e. (1992). Comment on S. Hinkle, R.J. Brown & P. Ely, Social identity processes: \\nsome limitations and limiting conditions. Revista de Psicologia Social, Monogra\\nfico, 1 1 3-\\n123. \\nTriandis, H.e. (1993). Paper for Psychology in the Americas: North-South Exchange, \\nCommunicaci6n Presentada al Congreso Interamericano, Santiago de Chile. \\nTriandis, H.e. (\\\\994). Recherches recentes sur I'individualisme et collectivisme. Cahiers \\n1nternationaux de Psychologie Sociale, 23, 14-27. \\nTriandis, H.C., Bontempo, R., Betancourt, H., Bond, M., Leung, D., Brenes, A., Georgas, G., \\nHui, e.H., Marin, G., Setiadi, B., Sinha, J.B.P., Verma, J., Spangenberg, J., Touzard, H., & \\nMontmollin, G. (1986). The measurement of etic aspects of individualism and collectivism \\nacross cultures. Australian Journal o\\nf\\n Psychology, 38, 257-267. \\nTriandis, H.C., Bontempo, R., Villareal, M.J., Asai, M., & Lucca, N. (\\\\988). Individualism \\nand collectivism: Cross-cultural perspectives on self-ingroup relationships. Journal o\\nf\\n \\nPersonality and Social Psychology, 54, 323-338. \\nTriandis, H.e., Leung, K., Villareal, M.J., & Clack, F.L. (1985). Allocentric versus idiocentric \\ntendencies: Convergent and discriminant validation. Journal o\\nf\\n Research in Personality, 19, \\n395-415. \\nTriandis, H.C., McCusker, e., & Hui, e.H. (1990). Multimethod probes of individualism and \\ncollectivism. Journal o\\nf\\n Personality and Social Psychology, 59, 1006-1020. \\nTriska, J.F. (Ed.). (\\\\962). Soviet communism: Programs and rules. San Francisco: Chandler. \\nTruman, D.B. (\\\\951). The governmental process. New York: Knopf. \\nTuckman, B.W. (1965). Developmental sequence in small groups. Psychological Bulletin, 63, \\n384-399. \\n\\n\\n252 \\nSocial identity: international pers\\npectives \\nTuckman, B.W., & Jensen, M.A.e. (1977). Stages of small group development revisited. Group \\nand Organization Studies, 2, 419-427. \\nTurner, J.e. (l975a). Social comparison and social identity: Some prospects for intergroup \\nbehaviour. European Journal o\\nf\\n Social Psychology, 5, 5-34. \\nTurner, J.e. (1975b). Social categorization and social comparison in intergroup relations. PhD, \\nUniversity of Bristol. \\nTurner, J.C. (1978). Social categorization and social discrimination in the minimal group \\nparadigm. In H. Tajfel (Ed.), Diff\\nerentiation between social groups: Studies in the social \\npsychology o\\nf\\n intergroup relations. London: Academic Press. \\nTurner, J.e. (l98Ia). Towards a cognitive redefinition of the social group. Cahiers de \\nPsychologie Cognitive, 1, 93-1 18. \\nTurner, J.e. (198Ib). The experimental social psychology of intergroup behavior. In J.C. \\nTurner & H. Giles (Eds.), Intergroup behavior. Chicago: University of Chicago Press. \\nTurner, J.C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.), \\nSocial identity and intergroup relations. Cambridge: Cambridge University Press. \\nTurner, J.C. ( l983a). Some comments on the measurement of social orientations in the minimal \\ngroup paradigm. European Journal o\\nf\\n Social Psychology, 13, 351-368. \\nTurner, J.e. (I \\n983b). A second reply to Bornstein, Crum, Wittenbraker, Harring, Insko, and \\nThibaut on the measurement of social orientations. European Journal o\\nf\\n Social Psychology, \\n13, 383-387. \\nTurner, J.e. (1985). Social categorization and the self-concept: A social cognitive theory of \\ngroup behaviour. In E.J. Lawler (Ed.), Advances in group processes, Vol. 2. Greenwich, CT: \\nJAI Press. \\nTurner, J.C. (1987). In J.e. Turner, M.A. Hogg, P.J. Oakes, S.D. Reicher & M.S. Wetherell, \\nRediscovering the social group: A self-categorization theory. Oxford: Blackwell. \\nTurner, J.e. (1988). Comments on Doise's 'Individual and social identities in intergroup \\nrelations'. European Journal o\\nf\\n Social Psychology, 18, 1 1 3-1 16. \\nTurner, J.C. (1991). Social influence. Milton Keynes: Open University Press. \\nTurner, J.e., & Brown, R.I. (1978). Social status, cognitive alternatives and intergroup \\nrelations. In H. Tajfel (Ed.), Diff\\nerentiation between social groups: Studies in the social \\npsychology o\\nf\\n intergroup relations. London: Academic Press. \\nTurner, J.e., Brown, R.J., & Tajfel, H. (1979). Social comparison and group interest in \\nintergroup favoritism. European Journal o\\nf\\n Social Psychology, 9, 187-204. \\nTurner, J.C., & Giles, H. (Eds.). (1981). Intergroup behaviour. Oxford: Blackwell; Chicago: \\nUniversity of Chicago Press. \\nTurner, J.e., Hogg, M.A., Oakes, P.I., Reicher, S.D., & Wetherell, M.S. (1987). Rediscovering \\nthe social group: A self-categorization theory. Oxford: Blackwell. \\nTurner, J.e., Hogg, M.A., Oakes, P.I., & Smith, P.M. (1984). Failure and defeat as \\ndeterminants of group cohesiveness. British Journal o\\nf\\n Social Psychology, 23, 97-1 1 1 .  \\nTurner, J.e., & Oakes, P.J. (1986). The significance of the social identity concept for social \\npsychology with ref\\nerence to individualism, interactionism, and social influence. British \\nJournal o\\nf\\n Social Psychology, 25, 237-252. \\nTurner, J.e., & Oakes, P.J. (1989). Self-categorization theory and social influence. In P.B. \\nPaulus (Ed.), The psychology o\\nf\\n group influence. Hillsdale, NJ: Erlbaum. \\nTurner, J.C., Oakes, P.I., Haslam, S.A., & McGarty, e.A. (1994). Self and collective: \\nCognition and social context. Personality and Social Psychology Bulletin, 20, 454-63. \\nTurner, J.e., Sachdev, I., & Hogg, M.A. (1983). Social categorization, interpersonal attraction \\nand group formation. British Journal o\\nf\\n Social Psychology, 22, 227-239. \\nTurner, J.e., Wetherell, M.S., & Hogg, M.A. (1989). Referent informational influence and \\ngroup polarization. British Journal o\\nf\\n Social Psychology, 28, 135- 147. \\nTversky, A. (1977). Features of similarity. Psychological Review, 84, 327-352. \\nVander Zanden, J.W. (1960). The Klan revival. American Journal o\\nf\\n Sociology, 65, 426-459. \\nvan Knippenberg, A. (1978). Status differences, comparative relevance and intergroup \\ndifferentiation. In H. Tajfel (Ed.), Diff\\nerentiation between social groups: Studies in the social \\npsychology o\\nf\\n intergroup relations. London: Academic Press. \\n\\n\\nRe\\nf\\nerences \\n253 \\nvan Knippenberg, A. (1984). Intergroup diff\\nerences in group perception. In H. Tajfel (Ed.), The \\nsocial dimension: Developments in European social psychology, \\nVol. \\n2. Cambridge: \\nCambridge University Press. \\nvan Knippenberg, A., & Wilke, H. (1988). Social categorization and attitude change. European \\nJournal of Social Psychology, 18, 395-406. \\nVerkuyten, M. (1990). Self-esteem and the evaluation of ethnic identity among Turkish and \\nDutch adolescents in the Netherlands. Journal o\\nf\\n Social Psychology, 130, 285-297. \\nVolpato, C., Maass, A., Mucchi-Faina, A., & Vitti, E. (1990). Minority influence and social \\ncategorization. European Journal o\\nf\\n Social Psychology, 20, 1 19-132. \\nWetherell, M.S. (1987). Social identity and group polarization. In J.c. Turner, M.A. Hogg, P.J. \\nOakes, S.D. Reicher & M.S. Wetherell, Rediscovering the social group: A self\\n-categorization \\ntheory. Oxford: Blackwell. \\nWheeler, L., Reis, H., & Nezlek, J. (1983). Loneliness, social interaction, and sex roles. Journal \\no\\nf\\n Personality and Social Psychology, 45, 943-953. \\nWilder, D.A. (1977). Perception of groups, size of oppositions, and social influence. Journal o\\nf\\n \\nExperimental Social Psychology, 13, 253-268. \\nWilder, D.A. (1978). Homogeneity of jurors: The majority's influence depends upon their \\nperceived independence. Law and Human Behavior, 2, 363-376. \\nWilder, D.A. (1990). Some determinants of the persuasive power of in-groups and out-groups: \\nOrganization of information and attribution of independence. Journal o\\nf\\n Personality and \\nSocial Psychology, 59, 1 202-1213. \\nWilkes, A., & Tajfel, H. (1966). Types de classification et importance du contraste relatif. \\nBulletin du CERP, 15, 77-81 .  \\nWilliams, J.A., & Giles, H .  (1978). The changing status of women in society: An intergroup \\nperspective. In H. Tajfel (Ed.), Diff\\nerentiation between social groups: Studies in the social \\npsychology o\\nf\\n intergroup Behaviour. London: Academic Press. \\nWilliams, R.M. (1970). American society (3rd ed.). New York: Knopf. \\nWinter, D.G. (1973). The power motive. New York: Free Press. \\nWish, M. (1976). Comparisons among multidimensional structures of interpersonal relations. \\nMultivariate Behavioral Research, 11, 297-324. \\nWish, M., Deutsch, M., & Kaplan, S.J. (1976). Perceived dimensions of interpersonal relations. \\nJournal o\\nf\\n Personality and Social Psychology, 33, 409-420. \\nWittenbraker, J. (1983). The etiology o\\nf\\n intergroup discrimination: A developmental approach. \\nUnpublished doctoral dissertation, University of North Carolina, Chapel Hill. \\nWorchel, S., Coutant-Sassic, D., & Grossman, M. (1992). A developmental approach to group \\ndynamics: A model and illustrative research. In S. Worchel, W. Wood & J.A. Simpson \\n(Eds.), Group process and productivity. Newbury Park, CA: Sage. \\nWorchel,. S., Coutant-Sassic, D., & Wong, F. (1993). Toward a more balanced view of conflict: \\nThere is a positive side. In S. Worchel & J.A. Simpson (Eds.), Con\\nflict between people and \\ngroups. Chicago: Nelson-Hall. \\nWorchel, S., Grossman, M., & Coutant-Sassic, D. (1993). Minority influence in the group \\ncontext: How group factors aff\\nect when the minority will be influential. In A. Mucchi Faina \\n& S. Moscovici (Eds.), Minority influence. Chicago: Nelson-Hall. \\nWright, S.c., Taylor, D.M., & Moghaddam, F.M. (1990). Responding to membership in a \\ndisadvantaged group: From acceptance to collective protest. Journal o\\nf\\n Personality and \\nSocial Psychology, 58, 994-1 003. \\nWyer, R.S., & Podeschi, D.M. (1978). The acceptance of generalizations about persons, objects \\nand events. In R. Revlin & R. Mayer (Eds.), Human reasoning. Washington, DC: Winston. \\nYamagishi, T. (1995). Social dilemmas. In K.S. Cook, G.A. Fine & J.S. House (Eds.), \\nSociological perspectives on social psychology. Boston: Allyn & Bacon. \\nZander, A. (1968). Group aspirations. In D. Cartwright & A. Zander (Eds.), Group d\\nynamics \\n(3rd ed.). London: Tavistock; New York: Harper & Row. \\nZander, A., Stotland, E., & Wolfe, D. (1960). Unity of group, identification with group, and \\nself-esteem of members. Journal o\\nf\\n Personality, 28, 463-478. \\n\\n\\n254 \\nSocial identity: international perspectives \\nZavalloni, M. (1973). Social identity: Perspectives and prospects. Social Science I\\nn\\nf\\normation, \\n12, 65-9J. \\nZavalloni, M. (1983). Ego-ecology: The study of the interaction between social and personal \\nidentities. In A. Jacobson-Widdings (Ed.), Identity: Personal and socio-cultural. Stockholm: \\nAlmqvist & Wicksell International. \\nZavalloni, M., & Louis-Guerin, C. (1984). Identite sociale et conscience: Introduction a /'ego­\\necologique. Toulouse: Privat. \\nZebrowitz, L. (1990). Social perception. Buckingham: Open University Press. \\nZeegers, W. (1988). Andere tijden, andere mensen: De sociale re\\npresentatie van identiteit. \\nAmsterdam: Uitgeveri\\nj Bert Bakker. \\nZeigarnik, B. (1927). fIber das Behalten von erledigten und unerledigten Handlungen. \\nPsychologische Forschung, 9, 1-85. \\nZiller, R. (1964). Individuation and socialization: A theory of assimilation in large \\norganizations. Human Relations, 1 7, 341-360. \\nZimbardo, P.G. (1969). The human choice: Individuation, reason, and order versus \\ndeindividuation, impulse, and chaos. In W.J. Arnold & D. Levine (Eds.), Nebraska \\nSymposium on Motivation, Vol. 1 7. Lincoln: University of Nebraska Press. \\nZimbardo, P.G., & Leippe, M.R. (1991). T\\nhe psychology o\\nf\\n attitude change and social influence. \\nNew York: McGraw-HilI. \\n\\n\\nIndex \\nAbrams, D., 82, 124-41, 154, 1 55, 166, 167, \\n169, 177-8, 212, 223 \\nAbric, I.e., 13 \\nAdamopoulos, I., 17 \\nAebischer, V., 144 \\nagency, 18, 19 \\naggregate groups, 20-1 ,  132 \\nAllen, V.L., 2, 95, 108 \\nAllport, F.H., 1 80 \\nAllport, G.W., 16, 44, 81 \\naltercentrism, 218-19, 228 \\nAltman, 1., 1 17 \\naltruistic rationalization hypothesis, 193-4 \\nAmancio, L., 22 \\nAmish society, 96, 102 \\nAmnesty International, 97-8 \\nAnastasio, P.A., 1 96 \\nanchored identities, 20-3 \\nAnderson, B., 1 20 \\nAntoni, M., 226 \\nApter, D.E., 100 \\nArdener, S., 96 \\nArdrey, R., 1 17 \\nArgyle, M., 55 \\nArnal-Duchemin, M.J., 15 \\nArr6spide, 1.1., 21 1-29 \\nAsai, M., 201, 213 \\nAsch, S.E., 36, 81, 104, 151 \\nassimilation effect, 4 \\nAsuncion, A.G., 143 \\nasymmetry phenomena in similarity \\njudgments, 37-41, 42 \\nattraction see interpersonal attraction; social \\nattraction \\nauto-stereotypy, 7 \\nAyestanin, S., 21 1-29 \\nBachman, B.A., 196 \\nBack, K., 126 \\nBakan, D., 18 \\nBall, P., 169 \\nBallachey, E.L., 1 12 \\nBallato, S.M., 88 \\nBanaji, M., 42 \\nBargh, I.A., 36 \\nBaron, R.S., 81 \\nBarsalou, L.W., 88, 90, 92 \\nBar-Tal, D., 93-1 13 \\nBar-Tal, Y., 109 \\nBasques, 1 23 \\nBeaufils, B., 13 \\nbeliefs, group, 93-1 13 \\nBern, D.I., 101, 141 \\nBern, S.L., 20 \\nBillig, M., 24, 32, 105, 1 10, 1 19, 145, 1 52, \\n181 \\nBlack Panthers, 107-8 \\n'black sheep eff\\nect', 1 29-31, 150 \\nBlaine, B., 222 \\nBoltanski, 1.L., 37 \\nBonaldo, E., 170 \\nBond, M.H., 219, 225 \\nBontempo, R.N., 17, 201 , 213 \\nBornstein, G., 183, 184 \\nBourhis, R.Y., 218 \\nBovarism, 219 \\nBoyes-Braem, P., 75 \\nBrewer, M.B., 36, 41, 45, 47, 76, 80, 83, 87-8, \\n90, 104, 130, 181, 1 96 \\nBrito, P.N., 133 \\nBroadnax, S., 222 \\nBrosch, M., 109 \\nBrown, R., 1 32, 1 80 \\nBrown, R.I., 46, 83, 125, 1 27, 1 55, 167, 168, \\n170, 171, 172, 175, 176, 177, 182, 202-3, \\n204, 2 1 1 ,  214 \\nBruner, I.S., 31, 128 \\nBundy, R.P., 24, 105, 145 \\nBurnstein, E., 81 \\nButera, F., 149, 151 \\nByrne, D., 45, 47, 154 \\nCaddick, B., 168, 171-2, 175 \\nCampbell, D.T., 77, 104, 1 15, 1 19, 180, \\n188 \\nCamps, V., 200 \\nCannon, T., 107-8 \\nCantor, N., 31-2, 41 \\nCapozza, D., 170 \\nCarroll, D.I., 159 \\n\\n\\n256 \\nSocial identity: international pers\\npectives \\nCartwright, D., 98, 125, 126 \\nCarugati, F., 13 \\nCarver, C., 226 \\ncategorial diff\\nerentiation see differentiation, \\nintergroup \\ncategorization, 4-5, 32 \\nand intergroup competition, 180-97 \\nrecategorization and the discontinuity \\neff\\nect, 196-7 \\nsee also self-categorization; self­\\ncategorization theory; social \\ncategorization \\nCavanni, L., 158 \\ncentrality of group beliefs, 101-2 \\nChaiken, S., 188 \\nChang, 1.1., 1 59 \\nCIC theory see social identity theory \\nClack, F.L., 201 \\nClark, R.D., 70, 142, 143 \\nClemence, A., 13, 25 \\nCochrane, S., 82 \\nCodol, I.-P., I, 9, 13, 14, 27, 28, 34, 35, 47, \\n48 \\ncognitive centrism, 9 \\ncognitive creativity, 219 \\ncognitive differentiation eff\\nect, 4 \\ncognitive stereotype effect, 4 \\ncollection groups, 20-1 ,  1 32 \\ncollectivism, 201-2, 203, 204, 205-6, 212-14, \\n214, 225, 228 \\nComby, L., 214 \\ncommon knowledge, in personal identity, \\n14-17, 19-20 \\ncommunion, 18, 19 \\ncommunist ideology, \\\\00 \\n'comparative fit' principle, 79, 127, 128 \\ncompetition and conflict \\nintergroup, 5, 66, 67, 68-9, \\\\06-7, 176, \\n180-97 \\ninterpersonal, 181-2, 192-3, 213 \\nconcentricity thesis, 121, 122-3 \\nCondor, S., 41, 163, 172 \\nconfidence in group beliefs, 101, l lO-l l  \\nConsumer Federation of America, 99 \\ncontext, dependence of categorization on, \\n75-92 \\ncontrast or cognitive diff\\nerentiation effect, \\n4 \\nConverse, P.E., I I I  \\nCooper, 1., 82 \\nCooper-Shaw, L., 84 \\ncooperation between groups, 68-9, 106-7, \\n168, 176 \\nCordova, D., 222 \\nCorreia, M.G., 133 \\ncouples, relationships between see \\ninterpersonal relationships \\nCoutant, D., 71 \\nCoutant-Sassic, D., 57, 69, 71, 72 \\ncovariation hypothesis, xviii, 8-l l ,  26, 27, 30 \\nCowan, G., 16 \\nCrano, W.o., 71 \\nCrocker, 1., 212, 221 ,  222, 223, 226 \\nCrook, K.F., 172-3 \\nCrosby, F., 222 \\nCrutchfield, R.S., 1 12 \\nCsepeli, G., 83 \\nculture, 56, 1 1 7-18, 120-1, 199-229 \\nCylwik, A., 51 \\nDamon, W., 14 \\nDann, H.D., 83 \\nDardis, G.I., 181 \\nDarley, I.M., 180 \\nDavid, B., 79 \\nDavidson, B., 82 \\nDeaux, K., 160 \\ndecay of groups, 61-2, 63, 73 \\nDeCharms, R., 126 \\nDeconchy, I.P., 101 \\ndelinquency, adolescent, 22-3 \\nDelumeau, 1.1., 124 \\ndepersonalization, 7, 77, 84, 91, 120 \\nDeschamps, I.-C., xviii, I -l l ,  25, 26, 27, 30, \\n55, 127, 214 \\nDeutsch, M., 17, 131, 148 \\nDe Vader, C.L., 88 \\ndeviates, derogation of, 124-5, 132-4, \\n1 35-41 \\nDevos, T., l - l l ,  214, 220 \\nDiamond, M., 97 \\nDiehl, M., 181 \\ndiff\\nerentiation, 27-30 \\nand groups beliefs, 95, \\\\08 \\nintergroup, 3, 4, 5-7, 8-l l ,  14, 24, 25-6, \\n77, 121-2, 126-41 passim, 199-210, \\n203 \\nintragroup, 14, 25-6, 124-41 \\nand the meta-contrast principle, 77-80 \\nand national identity, 120-1 \\nsee also covariation hypothesis; \\nprototypicality \\nDiMaggio, A., 170 \\ndiscontinuity effect, 180-2, 187-97 \\ndissociation theory, 148-51, 1 52, 153 \\nDoise, W., 4, 9, 1 3-23, 26, 27, 55, 83, \\n127, 132, 142, 143, 147, 195, 213, \\n214, 215 \\ndominant groups, 20, 26, 27, 132, 214, \\n228 \\n\\n\\ndominated groups, 20, 26, 27, 132, 214-20, \\n222, 223, 228 \\nDoob, L.W., 1 17 \\ndoomsday cult, 59 \\nDoosje, B., 132 \\nDoukhobors, 96 \\nDovidio, J.F., 80, 196 \\nDrigotas, S.M., 180-97 \\nDull, V., 80 \\nDumont, L., 199 \\nDunbar, R., 120 \\nDunning, D., 85 \\nDurand Delvigne, A., 1 3, 21-2 \\nDurkheim, E., 180 \\nDuveen, G., 13 \\nEagly, A.H., 132, 188 \\nEchebarria, A., 122 \\nego and the self, 44 \\negocentrism, 9, 52 \\nEiser, J.R., 25 \\nElejabarrieta, F., 122 \\nEllemers, N., 1 32, 178, 213 \\nEly, P.G., 166-79, 203, 204, 21 1 \\nEmler, N., 6, 22-3 \\nemotional experience and individualist! \\ncollectivist culture, 202 \\n'entitativity', 180, 183, 203 \\nEpstein, A.L., 103 \\nEpstein, S., 44 \\nErikson, E., 14, 18, 44 \\nEtaugh, C., 1 56 \\nethnicity, 103, 1 18, 120-1 \\nethnocentrism, 1 1 5, 121 \\nethnocentrism theory, 1 19 \\nEtzioni, A., 98 \\nFabian, A. \\nK., 1 17 \\nfailure of groups, 64, 65 \\nFalger, V., 1 1 5  \\nFalomir, J.M., 152 \\nFarr, R.M., 13 \\nfascist ideology, 100 \\nfear-based competition, 194-6 \\nFells, J., 122 \\nFestinger, L., 59, 106, 1 10, 126 \\nFierres, R., 1 52 \\nFiore, M., 1 57 \\nFiske, S.T., 23, 36, 76 \\nFlament, C., 24, 105, 145 \\nForgas, J.P., 158-9 \\nForsyth, D.R., 129, 141 \\nFoti, R.J., 88 \\nFox-Cardamone, L., 166-79 \\nFraser, S.L., 88 \\nIndex \\nFromkin, H.L., 45, 47 \\nfundamental group belief, 94, 103-6 \\nGachoud, J.P., 142 \\nGaelik, L., 28 \\nGaertner, S.L., 80, 196 \\nGagnon, A., 218 \\nGara, M., 164 \\nGarza, R.T., 223, 224, 227 \\ngender, 19, 20, 21-2, 26, 143 \\nGerard, H.B., 1 10, 126, 148 \\nGergen, K.J., 3 \\nGiles, H., 76, 155, 169 \\nGillin, J., 97 \\nglobalization, 1 14, 123 \\ngoals, group, 98-9 \\nGoethals, G.R., 1 52 \\nGoffman, E., 165 \\nGoldberg, L.R., 16, 17 \\nGordon, P.c., 28, 32, 48 \\nGordon, S.L., 1 56 \\nGotab, A., 46 \\nGouge, c., 83 \\n257 \\nGraduated and Reciprocated Initiatives in \\nTension-reduction (GRIT), 196 \\nGraetz, K.A., 181 \\nGray, W.D., 75 \\ngreed-based competition, 193-4 \\nGreenwald, A.G., 42, 44 \\nGrossman, M., 57, 69, 71 \\ngroup beliefs, 93- 1 1 3  \\ngroup concept, xviii \\ngroup identity, 53-123 \\ndevelopmental view of, 56-73 \\ngroup polarization, 81-3, 91 \\ngroup productivity, 60, 63, 67, 68 \\nGudkyunst, W., 225 \\nGuimond, S., 222 \\nHaberkorn, G., 143 \\nHalverson, C.F., 16 \\nHamilton, D.L., 132 \\nHamilton, V.L., 124, 141 \\nHampson, S.E., 89 \\nHantzi, A., 24 \\nHardie, E.A., 84-5 \\nHardoin, M., 34, 48 \\nHarkins, S., 61, 180 \\nHarring, K., 181, 185, 187, 191 \\nHarrison, A.A., 156 \\nHart, D., 14 \\nHarter, S., 19 \\nHarvey, O.J., 54, 107, 196 \\nHaslam, S.A., 31, 75-92, 127, 128 \\nHayes, C.H., 92, 120 \\n\\n\\n258 \\nSocial identity: international pers\\npectives \\nHenderson, M., 144 \\nHERMES project, 200 \\nHerringer, L.G., 223, 224, 227 \\nHertz, F., 1 1 5  \\nHewstone, M., 24, 144, 169 \\nHiggins, E.T., 102 \\nHill, C.T., 156 \\nHinkle, S., 166-79, 202-3, 204, 21 1-12 \\nHo, H., 196 \\nHolTman, C.D., 16 \\nHolT\\nman, M.L., 46 \\nHofstede, G., 56, 200 \\nHogg, M.A., 76, 79, 82, 83, 84-5, 86, 88, 105, \\n1 19, 125-31 passim, 139, 146, 1 54, 1 55, \\n162, 166, 167, 1 77-8, 2 1 1  \\nHolyoak, K.J., 28, 32, 48 \\nHolzworth, D.W., 84 \\nHomans, G., 96 \\nHood, W.R., 54, 107, 196 \\nHopkins, N., 6 \\nHorwitz, M., 180 \\nHostetler, J.A., 96, 102 \\nHoyle, R.H., 181, 1 93, 1 95 \\nHu, L., 214 \\nHuguet, P., 149, 1 5 1  \\nHui, C.H., 200-1 \\nHuici, C., 154, 155, 215 \\nhuman identity, 31, 77 \\nHurtig, M.C., 21, 37-8 \\nidentification stage of group development, \\n58-60, 62, 68, 73 \\nideology, 99-101 \\nImago theory, 18 \\nINDCOL scale, 201 \\nindependence, 54-6 \\nindividualism, 199-210, 212-14, 214, 225, \\n228 \\nindividualization, 35, 215 \\nindividuation stage of \\ngroup development, \\n60-1, 63, 73 \\ninfluence, 126 \\ninformational, 1 26, 1 34, 1 39-41 \\nof majorities, 142 \\nof minorities, 70-2, 73, 147-8 \\nnormative, 1 26, 1 34 \\nand social categorization, 142-53 \\nInoff, G., 16 \\nInsko, C.A., 180-97, 203, 209 \\nInstIa, P., 2 1 1 -29 \\ninterest groups, 99 \\nintergroup relationships, 66-9, 73 \\ncompetitive, S, 66, 67, 68-9, 106-7, 176, \\n180-97 \\ncooperative, 68-9, 106-7, 168, 176 \\ninterpersonal relationships as, 1 54, \\n1 55-65 \\nuse of new comparison dimensions in, \\n166-79 \\ninterpersonal attraction, 1 54 \\nand group behaviour, 83-4, 85, 91 \\ninterpersonal relationships, S, 6, 55-6 \\ncompetition and conflict in, 181-2, 192-3, \\n213 \\nas intergroup relationships, 154, 1 55-65 \\nJackson, L.A., 178 \\nJames, W., 2, 18, 44 \\nJanis, I.L., 126 \\nJarymowicz, M., 44-52 \\nJaskar, K., 222 \\nJehovah's Witness group, 107 \\nJensen, M.A.C., 57 \\nJodelet, D., I 3  \\nJohnson, D.M., 75 \\nJohnston, L., 24 \\nJones, E.E., 69, 1 10, 126, 134 \\nJovanovic, M., 146 \\nJung, C.G., 18 \\nKaiser, c., 143, 144 \\nKaminolT, R., 1 17 \\nKaplan, S.J., 17 \\nKarytowski, J., 45 \\nKastersztein, J., 168 \\nKedourie, E., 1 15 \\nKelley, H.H., 1 54, 183, 187, 192 \\nKelly, J., 54 \\nKelman, H.C., 120 \\nKenny, D.A., 1 29 \\nKiesler, C.A. and Kiesler, S.B., 81 \\nKillian, L.M., 103 \\nKinder, D.R., 222 \\nKing, G., 102 \\nKirchler, E., 1 5  \\nKitayama, S., 201 ,  202 \\nKleck, B.E., 1 93 \\nKlein, M., 1 8  \\nKluckholn, c., 97 \\nKobuszewska, E., 52 \\nKohlberg, L., 46 \\nKohn, H., 1 14, 1 1 5 \\nKotkin, M., 1 56 \\nKrech, D., 1 12 \\nKruglanski, A.W., 101, 109 \\nKu Klux Klan, 98 \\nKuhn, M.H., 14 \\nLalonde, R.N., 171 \\nLane, R.E., 100 \\n\\n\\nIndex \\n259 \\nlanguage, 1 17-18 \\nLarsen, K., 83 \\nLassaline, M.E., 75 \\nLatane, B., 61, 180 \\nLa Voie, L., 129 \\nLea, M., 82 \\nleadership, 88, 178 \\nLeBon, G., 180 \\nL'Ecuyer, R., 14 \\nLee, J., 196 \\nLee, S., 82 \\nlegitimacy/illegitimacy, 167, 170-2, 175-6 \\nLegrenzi, P., 149 \\nLeippe, M.R., 141 \\nLemaine, G., 151, 168, 169, 171, 174, 176, \\n177 \\nLeung, K., 201 \\nLevine, J.M., 57, 125, 126 \\nLe Vine, R.A., 1 19 \\nLevy, S.B., 102 \\nLewin, K., 54 \\nLeyens, J.-P., 17, 129, 214, 215 \\nLindskold, S., 196 \\nLinville, P.W., 69 \\nLinz, J.J., 1 23 \\nLippitt, R., 54 \\nLlavata, E., 152 \\nLloyd, B., 13 \\nLodewijkx, H.G.M., 1 92 \\nL6pez-Saez, M., 1 99-210 \\nLord, R.G., 88, 90 \\nLorenzi-Cioldi, F., 9, 10, 1 3, 1 5, 1 9, 20-1 , 22, \\n26, 1 32, 214 \\nLott, A.J.and Lott, B.E, 83, 131 \\nLouis-Gueerin, c., 1 3  \\nLow, S.M., 1 17 \\nLubovitcher Hassidim, 102 \\nLucca, N., 201 ,  213 \\nLuhtanen, R., 212, 221 ,  222, 226 \\nLui, L., 80, 87 \\nLuria, A.R., 27 \\nMaass, A., 70, 142, 143 \\nMackie, D., 82, 143, 144, 146, 2 1 1  \\nmajorities, influence of, 142 \\nMakris-Botsaris, E., 1 9  \\nMann, J., 80, 196 \\nMaoris, 59 \\nMaras, P., 176 \\nMarch, J.G., 98 \\nMarcia, J.E., 14 \\nMariategui, J.c., 1 1 8  \\nMarkus, H., 1 8, 47, 48, 201 , 202 \\nMarques, J.M., 124-41, 1 50, 215, 220 \\nMartin, R., 143, 144 \\nMartinez-Taboda, M., 21 1-29 \\nMaslow, A.H., 54 \\nMazzara, B., 36 \\nMcAdams, D.P., 18, 19 \\nMcCallum, D.M., 181, 1 87 \\nMcCusker, C., 201 \\nMcDougall, W., 180 \\nMcGarty, C., 77, 79, 82, 86, 90, 91, 92, 127, \\n139 \\nMcGrath, J.E., 54 \\nMcGuire, W.J. and McGuire, C.V., 14, 23 \\nMcPartland, T.S., 14 \\nMcPhail, C., 180 \\nMead, G.H., 2, 47 \\nMedin, D.L., 75, 77, 88-9, 89-91, 92 \\nMentzer, S.J., 193 \\nMervis, C.B., 75 \\nMessick, D., 21 1 \\nmeta-contrast principle, 77-80, 91, 120-1, \\n127, 128, 1 36-7, 138 \\nMeyer, G., 55, 127 \\nMiller, N., 196 \\nMiller, R.L., 47 \\nminimal group paradigrn, 1 10, 1 82-3 \\nminorities, 142, 166 \\ninfluence of, 70-2, 73, 147-8 \\nMischel, W., 31-2, 41 \\nMlicki, P.P., 213 \\nMoghaddam, F.M., 178, 219 \\nMoise, L.C., 218 \\nMontei1, J.M., 35 \\nMontero, M., 218 \\nMoore, M., 156 \\nmorale, 178 \\nMorales, J.F., 164, 199-210, 215 \\nMoreland, R.L., 57, 104, 105, 125, 126 \\nMormons, 102 \\nMorris, W.N., 56 \\nMoscovici, S., 13, 70, 142, 149 \\nmotivation, 24-5, 28, 178 \\nMoya, M.C., 1 54-65 \\nMucchi-Faina, A., 142 \\nMugny, G., 70, 125, 142-53 \\nMullen, B., 214 \\nMultiple Alternative Matrices (MAMs) \\nexperiments, 183-6 \\nMurphy, G.L., 89 \\nMurrell, A., 80, 196 \\nMussolini, B., 100 \\n'myths of descent', 1 18 \\nNakbi, J.L., 1 5  \\nNascimento-Schulze, C.M., 215, 221 \\nnation-state, 1 14-15, 1 16, 1 1 8-19, 121 \\nnational identity, 1 14-23 \\n\\n\\n260 \\nSocial identity: international pers\\npectives \\nNational Rifle Association, 99 \\nNational Right to Life Committee, 99 \\nNavarro, E., 146 \\nneeds, satisfaction of, 104 \\nNeisser, D., 75, 92 \\nNemeth, c., 70, 72, 149 \\nNeuberg, S.L., 23, 76 \\nNezlek, J., 56 \\nT\\nhe Night Porter (film), 1 58 \\nNiznik, J., 1 22 \\n'normative fit' principle, 79, 91, 127-8 \\nnormative influence, 126, 1 34 \\nnorms, group, 96, 106, 1 33-4 \\ndenotative, 127-9, 1 34, 1 35, 1 36-7 \\nprescriptive, 129, 134, 1 35, 1 36-7, 138 \\nOakes, P.J., 31, 32, 33, 75-92, 1 19, 1 27, 1 28, \\n146, 162, 2 1 1  \\nOdbert, H.S., 16 \\nO'Dea, T.F., 102 \\nOldenquist, A., 55 \\nOnorato, R., 86 \\nOrtony, A., 90 \\nOsgood, C.E., 1 96 \\nOstell, A., 83 \\nOstrom, T.M., 8, 104 \\n'outgroup homogeneity effect', 8 \\nPaez, D., 1 24-41, 2 1 1 -29 \\nPages, R., 2 \\nPaicheler, H., 1 3  \\nPalmonari, A., 1 3, 1 5  \\nPapastamou, S., 143, 144 \\nparalysis, ingroup, 146, 148 \\nParelman, A., 1 56 \\nPark, B., 35, 69 \\nParsons, T., 97 \\nparticularization, 32, 35 \\nPavelchak, M.A., 23, 36 \\nPawluczuk, B., 5 1  \\nPeabody, D., 1 6  \\nPemberton, M.J., 196 \\nPeplau, L.A., 1 55, 156, 162 \\nPerez, J.A., 1 25, 142-53 \\nPerie, M., 85 \\npersecution, collective, 1 24-5 \\npersonal attraction see interpersonal \\nattraction \\npersonal identity, 24-5, 31, 47-9, 54-6, 76, \\n77, 160, 228 \\ncommon knowledge in, 14-17, 19-20 \\nand social identity, 1-1 1 , 213 \\nas a social representation, 1 3-23 \\npersonal relationships see interpersonal \\nrelationships \\nPersonnaz, B., 168 \\nPfost, K.S., 157 \\nPhinney, J.S., 222 \\nPiaget, J., 46 \\nPichevin, M.F., 21, 37-8 \\nPinkley, R.L., 181, 1 85, 187, 191, 193, 1 95 \\nPiolat, M., 37-8 \\nPizzigati, K.A.L., 16 \\nPodeschi, D.M., 32 \\npolarization, group, 81-3, 9 1  \\npolitical participation, 222, 229 \\nPombeni, M.L., 15 \\npositive self-identity/evaluation, 6, 24-5, \\n181 \\nsee also self-esteem, personal \\npositive social identity, 93, 121-2, 221 \\nand manifest influence, 148 \\nuse of new dimensions of intergroup \\ncomparison to attain, 166-79 \\nsee also self\\n-esteem, collective \\nPratkanis, A.R., 44 \\nPratto, F., 36 \\nprimus inter pares (PIP) effect, 27 \\nsee also superior conformity of the self \\nproductivity, group, 60, 63, 67, 68 \\nProshansky, H.M., 1 17 \\nprototypicality, 30-1, 32, 33-5, 37, 38, 41-2, \\n75-92, 127-9, 134 \\nPryor, J.B., 104 \\nQuattrone, G., 69 \\nQuestionnaire o\\nf\\n Social Perception, 48 \\nQuevedo, S., 212 \\nRabbie, J.M., 1 80, 1 82, 1 90 \\nRamirez, S., 123 \\nRauma, D., 124, 141 \\nRavaud, J.F., 13, 14-15 \\nRaviv, A., 109 \\nrealistic conflict theory, 1 88 \\nrecategorization, and discontinuity eff\\nect, \\n1 96-7 \\nreferent informational influence, 1 39 \\nregionalism, 1 16, 1 17, 121 \\nReicher, S., 22-3, 76, 1 19, 1 27, 146, 162, \\n21 1 \\nReis, H.R., 56, 1 96 \\nrelationships see intergroup relationships; \\ninterpersonal relationships \\nReykowski, J., 45, 46 \\nReynolds, Y., 92, 1 1 5  \\nRicci Bitti, P., 1 3  \\nRichard III Society, 99 \\nRiecken, H.W., 59 \\nRiley, S., 156 \\n\\n\\nIndex \\n261 \\nRips, L.J., 90 \\nRivera, A.N., 222 \\nRobalo, E.M ., 130 \\nRobbers' Cave experiments, 106-7 \\nRobinson, P.W., 19 \\nRocha, S.A., 130 \\nRodriguez, C.P., 214 \\nRogers, C., 54 \\nRokeach, M., 96 \\nRook, D.S., 56 \\nRosch, E., 30-1, 41, 75, 76, 77, 80, 87 \\nRosenberg, S., 164 \\nRoss, G.F., 168, 170, 171, 175 \\nRoth, E .M., 89 \\nRothbart, M., 35, 69 \\nRoux, P., 25 \\nRubin, Z., 1 56 \\nRucker, M.H., 156 \\nRush, M.C., 88 \\nRussell, CT., 107 \\nRussell, J.E.A., 88 \\nRust, M.C., 196 \\nRutherford, J.F., 107 \\nSachdev, I., 83, 105 \\nSalazar, J .M., 1 14-23, 219, 224 \\nSalisbury, R.H., 99 \\nSanchez-Mazas, M., 146, 148, 150 \\nSande, G.N., 16 \\nSanders, G.S., 81 \\nSarbin, T.R., 2 \\nSarchielli, G., 13 \\nSchachter, S., 54, 59, 1 26 \\nSchadron, G., 214 \\nSchei̖r, M., 226 \\nschematic traits, 47 \\nSchoben, E.J., 89 \\nSchopler, J., 180-97, 203, 209 \\nSears, D.O., 222 \\nSedikides, C., 8 \\nself \\nand the ego, 44 \\nfluidity of the, 3 \\nas object, 44 \\nand prosocial attachment, 44-6 \\nsocial, 44-6 \\nas subject, 44 \\nsuperior conformity of the, 9, 16, 27 \\nself-categorization, levels of, 14, 31, 36, \\n146 \\nself-categorization theory, 7-8, 30-3, 34, 35, \\n65, 93-4, 105 \\nand idea of depersonalized social \\nattraction, 84 \\nand group polarization, 8 1-3 \\nprototypicality in, 30-1, 33, 34, 76-81 ,  \\n85-6, 91, 127-9 \\nand social influence, 147-8 \\nself-consciousness, 212, 226-8, 229 \\nself-definition, 7-8 \\nself-esteem \\ncollective, 212, 220-3, 229 \\npersonal, 6, 19, 20, 21 1-12, 220-3, 229 \\nself-other comparison, 24-42 \\nsee also differentiation; similarity \\nrelationships \\nself-permanence, 3 \\nself-presentation, 177-8 \\nself-representation, duality of, 2 \\nself-stereotyping, 126-7, 1 28, 131 \\nself structure, 45 \\nself-we-others schemata, 44-52 \\nSemin, G.R., 17 \\nSerino, C., 24-42 \\nSerooio, R.G., 133 \\nSeton-Watson, H., 1 1 5  \\nSewell, D.R., 88 \\nShaw, M.E., 126 \\nSherif, CW., 54, 107, 196 \\nSherif, M., 5, 54, 69, 81, 83, 106-7, 155, 196, \\n197 \\nShils, E., 100 \\nSills, D.L., 98 \\nSilver, M., 83 \\nsimilarity relationships, 3, 4, 6-8, 14, 27-30, \\n33-5, 44-5, 47, 89-90 \\nasymmetries in, 37-41, 42 \\nand group beliefs, 95, 108 \\nintergroup, 14, 14-17, 20-2 \\nand the meta-contrast principle, 77-80 \\nmetrical approach to, 28-9 \\nand national identity, 120-1 \\nsee also covariation hypothesis; \\nprototypicality \\nSimmel, G., 125 \\nSimon, B., 132 \\nSimon, H.A., 98 \\nSinclair, A., 215 \\nSingh, A., 22 \\nSingleton, R., 81 \\nSmith, A.D., 1 1 5, 1 1 8  \\nSmith, C., 214 \\nSmith, J., 47 \\nSmith, P.B., 219, 225 \\nSmith, P.M., 83 \\nSmith, V.A., 181 \\nSnyder, CR., 45, 47, 193 \\nSobczyk, K., 52 \\nsocial attraction, and prototypicality, 83-5, \\n91 \\n\\n\\n262 \\nSocial identity: international pers\\npectives \\nsocial categorization, 4-8, 105 \\ncontext dependence of, 75-92 \\ndeductive effect of, 1 5  \\ngraded structure of, 75-92 \\nproductive character of, 31-2 \\nand social influence, 142-53 \\nsocial comparison see diff\\nerentiation; self­\\nother comparison; similarity \\nrelationships \\nsocial constructivism, and latent influence, \\n149 \\nsocial control, and intragroup differentiation, \\n1 24-41 \\nsocial creativity, 166-79 \\nsocial identity theory, 4-8, 55, 65, 93, 105, \\n1 19, 167-9, 181-6 \\nsocial influence see influence \\nsocial loafing, 61 \\nsocial representations, in personal identity, \\n1 3-23, 213 \\nsocial self, 44-6 \\nsociocentrism, 9 \\nSokolowska, G., 52 \\nSpears, R., 82, 132 \\nSrull, T.K., 28 \\nstability/instability, 167, 170, 171, 175-6 \\nStahelski, A., 1 87 \\nstate, 1 14-1 5, 1 16, 1 18-19, 1 20 \\nstateless nationalism, 1 16 \\nstatus, 167-8, 170-2, 175-6, 212-23, 228-9 \\nStaw, B.M., 70 \\nStephenson, G.M., 1 57 \\nstereotypes, 4, 7, 1 5, 2 14, 215 \\nsee also self-stereotyping \\nStewart\\\" 1 32 \\nStory, A.L., 85 \\nStotland, E., 103 \\nStrenta, A., 193 \\nStroup, H.H., 107 \\nsubjective group dynamics, 131-2, 1 39-41 \\nsuccess of groups, 64-5 \\nSuls, 1.M., 47 \\nsuperior conformity of the self, 9, 16, 27 \\nsupra-nationalism, 1 16, 1 18, 121, 122 \\nsymbolic individual mobility strategy, 219, \\n220 \\nSzuster-Zbrokewicz, A., 51, 52 \\nt-group movement, 57 \\nTajfel, H., xviii, 4, 5, 6, 26, 41, 55, 77, 105, \\n1 19, 1 25, 126, 127, 1 3 1 , 145, 1 55, 166, \\n168, 169, 181 \\non legitimacy/stability of \\ngroups, 167, 171, \\n175-6 \\nminimal group paradigm, 1 10, 1 82-3 \\non minority groups, 103 \\non national and ethnic identity, 1 14 \\non similarity/difference judgements, 24, 25, \\n29 \\non temporal aspects of social identity, 163 \\nTap, P., 1 3, 19, 46 \\nTaylor, D.M., 178, 219 \\nTaylor, L.A., 166-79 \\ntemporal aspects of social identity, 163, 176-7 \\nterritoriality, 1 17, 120-1 \\nThibaut, 1.W., 106, 1 54, 183, 192 \\nToch, H., 100, 109 \\nTorregrosa, 1.R., 123 \\nTougas, F., 222 \\nTrans-Species Unlimited, 98-9 \\ntranscendence, and national identity, 120, 121 \\nTriandis, H.C., 56, 201, 203, 213 \\nTriska, J.F., 100 \\nTruman, D.B., 99 \\nTuckman, B.W., 57 \\nTurner, J.e., xviii, 5, 14, 24, 34, 35, 36, 41, 55, \\n65, 75-92, 93, 105, 1 19, 1 26, 128, 1 32, \\n1 39, 146, 163, 166, 167, 169, 181, 1 84, \\n1 85, 203, 21 1 \\non comparative fit, 127 \\non group polarization, 81, 82, 83 \\non influence of groups, 147-8 \\non meta-contrast principle, 77, 79, 80, 1 27, \\n162 \\nand the minimal group paradigm, 9, 182 \\non normative fit, 1 27-8 \\non positive self-identity, 6 \\nand prototypicality theory, 30, 75-92, 127 \\non referent informational influence, 139 \\non self-definition, 7-8 \\nsee also self-categorization theory \\nTversky, A., 27-8, 29, 34, 41, 77 \\nValencia, J., 122 \\nvalues, 96-8, 212, 220-3, 225, 226 \\nVander Zanden, J.W., 98 \\nvan Knippenberg, A., 83, 168, 171, 174, 176, \\n177, 178, 215 \\nVega, L., 199-210 \\nVerkuyten, M., 1 9  \\nVillareal, M., 1 22, 201, 2 1 3  \\nVine, I., 1 1 5 \\nVitti, E., 142 \\nVolpato, C., 9, 10, 142, 144 \\nWattenmaker, W.O., 89 \\nWetherell, M.S., 76, 79, 81, 82, 83, 1 19, 127, \\n146, 162, 2 1 1  \\nWheeler, L., 56, 1 96 \\nWhite, B.J., 54, 107 \\n\\n\\nWhite, R., 54 \\nWicklund, R., 193 \\nWilder, D.A., 95, 108, 143, 144, 145-6 \\nWilke, H., 83, 178, 215 \\nWilkes, A.L., 25, 29, 127 \\nWilliams, J., 1 55, 172 \\nWilliams, K., 61, 180 \\nWilliams, R.M., 97 \\nWinter, D.G., 18 \\nWish, M., 17 \\nWisniewski, E.J., 75 \\nwitch-hunting, 124 \\nWittenbraker, J., 184 \\nWolfe, D., 103 \\nWood, G., 69 \\nWorchel, S., 53-73 \\nIndex \\nWorth, L.T., 143 \\nWright, S.c., 178, 219 \\nWyer, R.S., 32 \\nYamagishi, T., 124 \\nYzerbyt, V.Y., 129, 214 \\nZander, A., 98, 103, 125 \\nZavalloni, M., 13, 25, 26 \\nZebrowitz, L., 219 \\nZeegers, W., 13, 16 \\nZeigamik, B., 64 \\nZiller, R., 46 \\nZimbardo, P.G., 141, 180 \\nZuckier, H., 36 \\n263 \\n\\n\\nSocial Identity, \\nIntergroup Conflict, and \\nConflict Reduction,\\nVolume 3\\nRichard D. Ashmore\\nLee Jussim\\nDavid Wilder,\\nEditors\\nOXFORD UNIVERSITY PRESS\\n\\n\\nSOCIAL IDENTITY,\\nINTERGROUP CONFLICT, and\\nCONFLICT REDUCTION\\n\\n\\nRUTGERS SERIES ON SELF AND SOCIAL IDENTITY\\nSeries Editors\\nRichard D. Ashmore and Lee Jussim\\nRutgers, The State University of New Jersey\\nEditorial Board\\nKay Deaux, City University of New York\\nKenneth J. Gergen, Swarthmore College\\nDorothy C. Holland, University of North Carolina at\\nChapel Hill\\nDaniel M. Ogilvie, Rutgers, The State University of\\nNew Jersey\\nSuzanne Ouellette, City University of New York\\nSeymour Rosenberg, Rutgers, The State University of\\nNew Jersey\\nDiane N. Ruble, New York University\\nRonald T\\naylor, T\\nemple University\\nPeggy A. Thoits, Vanderbilt University\\nDavid Wilder, Rutgers, The State University of New\\nJersey\\nVolume 1\\nSelf and Identity: Fundamental Issues\\nVolume 2\\nSelf, Social Identity, and Physical Health: Interdisciplinary \\nExplorations\\nVolume 3\\nSocial Identity, Intergroup Conflict, and Conflict Reduction\\n\\n\\nRUTGERS SERIES ON SELF AND SOCIAL IDENTITY\\nVolume 3\\nSOCIAL IDENTITY,\\nINTERGROUP CONFLICT,\\nand CONFLICT REDUCTION\\nEdited by\\nRichard D. Ashmore\\nLee Jussim\\nDavid Wilder\\n1\\n2001\\n\\n\\n3\\nOxford\\nNew York\\nAthens\\nAuckland\\nBangkok\\nBogotá\\nBuenos Aires\\nCape T\\nown\\nChennai Dar es Salaam Delhi Florence Hong Kong Istanbul\\nKarachi\\nKolkata\\nKuala Lumpur Madrid Melbourne Mexico City Mumbai\\nNairobi\\nParis São Paulo Shanghai Singapore T\\naipei T\\nokyo T\\noronto Warsaw\\nand associated companies in\\nBerlin Ibadan\\nCopyright © 2001 by Oxford University Press\\nPublished by Oxford University Press, Inc.\\n198 Madison Avenue, New York, New York 10016\\nOxford is a registered trademark of Oxford University Press.\\nAll rights reserved. No part of this publication may be reproduced, \\nstored in a retrieval system, or transmitted, in any form or by any means, \\nelectronic, mechanical, photocopying, recording, or otherwise, \\nwithout the prior permission of Oxford University Press.\\nLibrary of Congress Cataloging-in-Publication Data\\nSocial identity, intergroup conflict, and conflict reduction / [edited by] Richard D. Ashmore,\\nLee Jussim, and David Wilder.\\np. cm.—(Rutgers series on self and social identity ; 3)\\nIncludes bibliographical references and index.\\nISBN 0-19-513742-6; 0-19-513743-4 (pbk.)\\n1. Group identity. 2. Conflict management. 3. Ethnic relations. I. Ashmore, Richard D. II.\\nJussim, Lee J. III. Wilder, David, 1949–\\nIV. Series\\nHM753.S6197\\n2001\\n302.4—dc21\\n00-053761\\n9 8 7 6 5 4 3 2 1\\nPrinted in the United States of America\\non acid-free paper\\n\\n\\nPreface to Volume 3\\nThis volume is important to me both professionally and personally: it is a\\nsignificant part of my evolving life story. At the same, my co-editors and I\\nbelieve that this book provides a start toward an important new perspective\\non intergroup violence, a problem that brings extensive material destruc-\\ntion and widespread human suffering around the globe.\\nIn the spring of 1964, while still an undeclared and uncertain history ma-\\njor, I read Gordon Allport’s The Nature of Prejudice for a course entitled,\\n“Medieval Anti-Semitism.” It was an epiphany—I would become a social\\npsychologist, and I would seek to understand the causes and cures of inter-\\ngroup conflict. That summer I enrolled in my first psychology classes, fin-\\nished the psychology major during my senior year, and went on to get my\\nPh.D. in social psychology with a specialization in intergroup relations.\\nWhen I read Allport and chose my occupation and specialty area, I did\\nnot know Erikson’s seminal work on identity development. As I look back,\\nthough, his ideas about adolescent identity search fit my experience very\\nwell. Becoming a social psychologist and seeking to understand prejudice\\nand discrimination allowed me to bring together my childhood identifica-\\ntions (with two parents who were simply good people, with a succession of\\noutstanding teachers [from Mrs. Cirino in the fourth grade to Professor\\nLewis Spitz at Stanford who, to this day, provides the model for how I teach\\na large undergraduate class]), my then current concerns and values (I be-\\nlieved in 1964 that conflict among groups was the greatest problem facing\\nhumans; I felt that we should and could do something about this), and my\\nfuture work (becoming a scientist and uncovering why people engage in\\ngroup-based harm-doing).\\n\\n\\nNow it is 2000. With many twists and turns, I today still see myself as an\\nintergroup relations researcher. And I continue to believe that intergroup\\nconflict is the major problem facing the world. What is different, one of\\nthose twists and turns, is that I now consider myself also an identity re-\\nsearcher. This volume allows me to bring together, then, my personal life\\nstory, especially my commitment to understanding intergroup conflict, with\\nmy evolving professional self-definition as a social scientist who studies self\\nand identity.\\nAlthough the world today is, in many ways, quite different from that of\\nalmost 40 years ago, group-based violence, aggression, and suppression are\\nno less in evidence. Although the global cold war has ended, we now ex-\\nperience a seemingly never-ending succession of small and local “wars.”\\nThese are global in a new way—group-based strife occurs on all continents\\nand every corner of the world and yet, because of powerful new communi-\\ncations technologies, the whole world often is involved if only as CNN\\nviewers.\\nSince I entered the field in the mid-1960s, the social and behavioral \\nsciences have made great strides in seeking to understand and ameliorate\\nintergroup conflict. At the same time, there are still many gaps in our\\nknowledge.\\nMy co-editors and I believe that social identity may provide a powerful\\nperspective for filling in some of these gaps. Social identity has certainly\\nbeen implicated in past work on group strife, but these efforts have been\\npiecemeal and isolated from one another across disciplinary boundaries.\\nOur goal in this volume was to bring together scientists from diverse social\\nand behavioral sciences to address the multiple ways that social identity and\\nintergroup conflict are interrelated. We believe that our contributors have\\ndone an excellent job, on the one hand, of highlighting major current theo-\\nries, concepts, and methods and, on the other, of identifying the crucial \\nissues that need to be addressed if we are to develop a more complete and\\ncoordinated social identity perspective on intergroup conflict. We hope that\\nthis volume contributes to the development of such a paradigm and that it,\\nalong with other approaches, facilitates social technologies that significantly\\nreduce the incidence and intensity of people harming others because of\\ngroup membership.\\nPiscataway, New Jersey\\nRichard D. Ashmore\\nMay 2000\\nvi\\nPreface to Volume 3\\n\\n\\nAcknowledgments\\nMany individuals and organizations helped us as we planned and carried out\\nthe Third Rutgers Symposium on Self and Social Identity and produced\\nthis volume. We thank you all. We received so much help from so many\\nsources that we are sure to have forgotten to mention some here, and for\\nthat we apologize.\\nWe first thank the symposium speakers/chapter authors. They were \\nchosen because they each have distinguished scientific careers, and that past\\nbehavior predicted their performance in this endeavor. They produced\\nhigh-quality symposium presentations and excellent chapters appearing in\\nthis book. In addition, they had to deal with fairly picky editors. We thank\\neach of you for sticking with us and producing the present polished contri-\\nbutions.\\nNext, we wish to thank our publisher, Oxford University Press, and our\\neditors there, Joan Bossert, Philip Laughlin, and Catharine A. Carlin. Joan\\nhas been an invaluable source of guidance and support since she first re-\\nceived the proposal for the Rutgers Series, and Philip and Catharine ably \\nassisted us as we moved the present volume from an idea, to a conference,\\nto a book.\\nThere are many people to thank for sharing their ideas about the topic of\\nsocial identity and intergroup conflict, for suggesting speakers/authors, and\\nfor providing comments on chapter drafts. We especially wish to thank the\\nmembers of our editorial board, who provided valuable feedback on initial\\nand revised proposals. In addition, the following people provided input\\nconcerning one or more facets of the conference/book: Myron Aronoff,\\nMahzarin R. Banaji, James Jones, Louis Kreisberg, B. Ann Bettencourt,\\n\\n\\nYael Zerubavel, David Sears, Daniel M. Ogilvie, Jessica Heppen, and Rick\\nLau.\\nThis volume, as with its two predecessors, deserves to be part of the Rut-\\ngers Series on Self and Social Identity. We have been helped at every step of\\nthe way by the faculty, staff, students, and administrators of Rutgers–The\\nState University of New Jersey. The following members of the Rutgers\\nUniversity community were particularly helpful: Steven Lione, Thelma\\nCollins, Deana Guden-Pagnozzi, and the staff of the University Inn and\\nConference Center; Stacey Hersh-Ehling of the Rutgers News Service;\\nJoanne Aguglia of the Douglass College Center; Rosemary Manero and her\\ncolleagues in the Rutgers Dining Service; Marc Beebe, Justine Cardone,\\nKathy Maloney, Rae Frank, Maria Brewer, Anne Sokolowski, and Alison\\nSmith, all of the Department of Psychology, who exhibited patience and\\ngood humor in their extremely efficient handling of day-to-day tasks and\\ndeadline pressures.\\nOther members of the Rutgers community deserve special thanks. First,\\nCharles Flaherty, chair of the Department of Psychology, has supported the\\nRutgers Series since its inception, and he made departmental staff and other\\nresources available for our use in putting together the third symposium and\\npresent associated book. Second, we express our gratitude to the faculty and\\ngraduate students of the doctoral program in social psychology. Members\\nof the Social Psychology Area served as hosts for speakers, helped with\\nsymposium registration, transported speakers to and from airports and train\\nstations, directed traffic between symposium sessions, and helped conduct\\nmyriad other essential tasks. Third, we thank the following members of the\\nRutgers community for their generous financial assistance: Joseph J. Seneca\\n(Vice President for Academic Affairs), and Richard F. Foley (Dean, Faculty\\nof Arts and Sciences).\\nWe are also grateful for grants from the National Institutes of Health\\n(R13MH60129-01; Carolyn Morf, Chief, Personality and Social Cognition\\nProgram) and from the National Science Foundation (SBR-9811485;\\nSteven J. Breckler, Program Director, Social Psychology) that supported\\nboth the Third Rutgers Symposium and the production of this volume.\\nviii\\nAcknowledgments\\n\\n\\nContents\\nContributors\\nxi\\n1\\nintroduction: Social Identity and Intergroup Conflict\\n3\\nLee Jussim, Richard D. Ashmore, and David Wilder\\nPART I / The Contribution of Individuals’ Identities and the Collective\\nIdentities of Social Groups to Intergroup Conflict\\n2\\nIngroup Identification and Intergroup Conflict: When Does Ingroup Love\\nBecome Outgroup Hate?\\n17\\nMarilynn B. Brewer\\n3\\nEthnic Identity, National Identity, and Intergroup Conflict: The \\nSignificance of Personal Experiences\\n42\\nThomas Hylland Eriksen\\nPART II / The Contribution of Ethnic and National Identities to \\nPolitical Conflict in the United States\\n4\\nThe Meaning of American National Identity: Patterns of Ethnic Conflict\\nand Consensus\\n71\\nJack Citrin, Cara Wong, and Brian Duff\\n\\n\\n5\\nCommunal and National Identity in a Multiethnic State: A Comparison of\\nThree Perspectives\\n101\\nJim Sidanius and John R. Petrocik\\nPART III / The Contribution of Social Identity to Violent Intergroup\\nConflict\\n6\\nSocial and Role Identities and Political Violence: Identity as a Window on\\nViolence in Northern Ireland\\n133\\nRobert W. White\\n7\\nIndividual and Group Identities in Genocide and Mass Killing\\n159\\nErvin Staub\\nPART IV / The Role of Social Identity in Reducing Intergroup Conflict\\n8\\nThe Role of National Identity in Conflict Resolution: Experiences from \\nIsraeli-Palestinian Problem-Solving Workshops\\n187\\nHerbert C. Kelman\\n9\\nconclusion: T\\noward a Social Identity Framework for \\nIntergroup Conflict\\n213\\nRichard D. Ashmore, Lee Jussim, David Wilder, and Jessica Heppen\\nIndex of Subjects\\n251\\nIndex of Names\\n265\\nx\\nContents\\n\\n\\nxi\\nContributors\\nRichard D. Ashmore, Department of Psychology, Tillett Hall, Rutgers, The State\\nUniversity of New Jersey, 53 Avenue E, Piscataway, NJ 08854-8040. E-mail: \\nashmore@rci.rutgers.edu.\\nMarilynn B. Brewer, Department of Psychology, Ohio State University, 1885 Neil\\nAvenue, Columbus, OH 43210. E-mail: brewer.64@osu.edu.\\nJack Citrin, Department of Political Science, 210 Barrows Hall #1950, University \\nof California, Berkeley, Berkeley, CA 94720-1950. E-mail: gojack@uclink.\\nberkeley.edu.\\nBrian Duff, Department of Political Science, 210 Barrows Hall #1950, University \\nof California, Berkeley, Berkeley, CA 94720-1950. E-mail: duff@socrates.\\nberkeley.edu.\\nThomas Hylland Eriksen, Department of Social Anthropology, University of Oslo,\\nMoltke Moesvei 31, Eilert Sundts hus, 6th floor, P. O. B. 1091 Blindern, N-0317\\nOSLO, Norway. E-mail: t.h.eriksen@ima.uio.no.\\nJessica Heppen, Department of Psychology, Tillett Hall, Rutgers, The State Uni-\\nversity of New Jersey, 53 Avenue E, Piscataway, NJ 08854-8040. E-mail: \\njheppen@rci.rutgers.edu.\\nLee Jussim, Department of Psychology, Tillett Hall, Rutgers, The State University\\nof New Jersey, 53 Avenue E, Piscataway, NJ 08854-8040. E-mail: jussim@rci.\\nrutgers.edu.\\nHerbert C. Kelman, Psychology Department, Harvard University, Cambridge, MA\\n02138. E-mail: hck@wjh.harvard.edu.\\nJohn R. Petrocik, 210 McAlester Hall, Department of Psychology, University of \\nMissouri-Columbia, Columbia, MO 65211. E-mail: PetrocikJ@missouri.edu.\\n\\n\\nJim Sidanius, Department of Psychology, 1285 Franz Hall, P. O. Box 951563, \\n405 Hilgard Avenue, Los Angeles, CA 90095-1563. E-mail: Sidanius@psych.\\nucla.edu.\\nErvin Staub, Department of Psychology, University of Massachusetts, Amherst, MA\\n01003. E-mail: estaub@psych.umass.edu.\\nRobert W. White, Department of Sociology, Indiana University at Indianapolis, \\nCavanaugh Hall 303, 425 University Boulevard, Indianapolis, IN 46202. E-mail:\\nrwwhite@iupui.edu.\\nDavid Wilder, Department of Psychology, Tillett Hall, Rutgers, The State Uni-\\nversity of New Jersey, 53 Avenue E, Piscataway, NJ 08854-8040. E-mail:\\nwilder@psych-b.rutgers.edu.\\nCara Wong, Department of Political Science, 210 Barrows Hall #1950, University \\nof California, Berkeley, Berkeley, CA 94720-1950. E-mail: cwong@bravo.\\nberkeley.edu.\\nxii\\nContributors\\n\\n\\nSOCIAL IDENTITY,\\nINTERGROUP CONFLICT, and\\nCONFLICT REDUCTION\\n\\n\\nThis page intentionally left blank \\n\\n\\nLee Jussim\\n1\\nRichard D. Ashmore\\nDavid Wilder\\nIntroduction\\nSocial Identity and Intergroup Conflict\\nIntense conflict between nations, ethnic groups, and political groups has\\ncreated some of the most enduring and lethal social problems humans have\\never faced. Slavery, conquest, exploitation, war, mass murder, and genocide\\nare among the most extreme manifestations of intergroup conflict. Such\\nconflicts remain as much a part of the human condition today as ever. \\nOver the last two decades, social and behavioral scientists have increas-\\ningly emphasized the role of self and identity in the causes and conse-\\nquences of intergroup hostility and harm-doing. This is, in part, because\\nissues of self and identity occur at multiple levels of analysis: individual\\nwithin a social context, groups within a society, and nations of the world.\\nAt the level of individual within a social context, ethnic identity may con-\\ntribute to both in-group bonds and hostility toward other groups (Jones,\\n1997). At the level of groups within a society, social identity theory and\\nself-categorization theory emphasize the potential for group-based identi-\\nties to foster support for the status quo among higher power and status\\ngroups, and to foster intergroup competition and movements for political\\nchange among lower power and status groups (T\\najfel, 1981; Thoits & Vir-\\nshup, 1997; Turner, 1987). Finally, at the level of modern nations, identity\\nis implicated in a variety of conflicts—state violence against internal mi-\\nnorities (e.g., Goldhagen 1996; Staub, 1989), armed uprisings by ethnic\\ngroups against central governments (e.g., Gurr, 1993), and international\\nconflicts (e.g., Greenfeld & Chirot, 1994; Keegan, 1993). Furthermore,\\naddressing issues of identity may be necessary to reduce certain conflicts\\nbetween groups (e.g., Kelman, 1997; Kriesberg, Northrup, & Thorson,\\n1989).\\n3\\n\\n\\nAlthough self and identity play a critical role in some (perhaps most) in-\\ntergroup conflicts, social scientific analyses of that role have thus far been\\nlargely isolated from each other. This is unfortunate because, just as the sci-\\nentific analysis of conflict has blossomed in recent years, so too has there\\nbeen a resurgence in work on self and identity (Ashmore & Jussim, 1997).\\nMultiple and diverse lines of contemporary social science research and\\ntheory demonstrate that self and identity are central to the understanding\\nof human thought, feeling, and action. They are also crucial in accounting\\nfor interrelationships between the individual and larger sociocultural insti-\\ntutions and systems (Holland, 1997; T\\najfel, 1981; Thoits & Virshup, 1997).\\nThus, there is much in the areas of self and social identity that might inform\\nour understanding of intergroup conflict.\\nThe main purposes of this book are to elucidate ways in which: \\n• Social identities create and exacerbate intergroup conflicts.\\n• Intergroup conflict influences social identity.\\n• Addressing social identity issues might help reduce some inter-\\ngroup conflicts. \\nT\\no achieve these goals, we invited a group of talented social and behav-\\nioral scientists to address relationships between intergroup conflict and\\nidentity. First, we asked them to address the locus of social identity. Are so-\\ncial identities primarily phenomena of individual psychology or are social\\nidentities primarily phenomena that reside in the larger social structure?\\nThat is, do social, group-based identities occur primarily in the minds of in-\\ndividual people? Or are they embedded in the culture, language, norms, and\\nsocial practices of a group of people? Or do social identities emerge from\\nsome combination of individual psychology and socio-cultural discourses,\\nbeliefs, and practices? \\nSecond, we sought contributors whose work addresses different types of\\nintergroup conflict. Although developing a taxonomy of the dimensions on\\nwhich intergroup conflicts may vary is beyond the scope of this book, it is\\nclear that social identities may be involved in a great variety of intergroup\\nconflicts. For example, whether it is “identity politics” in American elec-\\ntions, ethnic cleansing in Yugoslavia, or riots targeting Muslims in India,\\ngroup-based identities often play a critical role in maintaining, and some-\\ntimes exacerbating, intergroup conflicts. Although addressing all types of\\nintergroup conflicts is beyond the scope of any single book, we sought con-\\ntributions that would address a wide range of conflicts: conflicts played out\\nwithin normal legal structures; conflicts involving illegal acts of aggression;\\nconflicts over symbolic and ideological issues; and conflicts over concrete\\nresources and power.\\nAlthough each contributor did not address the role of identity in this full\\nrange of conflicts, the contributions addressed many different types of con-\\nflicts (rather than focus on a single type of conflict—e.g., genocide). There-\\n4\\nIntroduction\\n\\n\\nfore, the first pair of contributions (by Brewer and Eriksen) focuses on \\nidentifying broad principles or common themes underlying a range of in-\\ntergroup conflicts. The second pair (by Citrin, Wong, and Duff, and Sida-\\nnius and Petrocik) focuses on the extent to which ethnic identity in the\\nUnited States remains strong and predicts strength of attachment to U.S.\\npolitical institutions and culture. The third pair of contributions (by White\\nand Staub) addresses the role of social identity in lethal violence by and\\nagainst the state. The seventh chapter (by Kelman) addresses the role of \\nsocial identity in reducing intergroup conflict. \\nIn setting the stage for the chapters to follow, we address the following\\nquestions: What do we mean by social identity? What do we mean by con-\\nflict? What is the organizational scheme for this volume? What specific\\ntopics do the chapters address?\\nWhat Do We Mean by Social Identity?\\nSelf and identity are not simple concepts with widely agreed upon mean-\\nings. As we pointed out in the first volume in this series (Ashmore & Jussim,\\n1997), the terms have been used in a bewildering diversity of ways. Some-\\ntimes, the same word or phrase is used by different writers to refer to differ-\\nent concepts; sometimes different words or terms are used to refer to what\\nappears to be the same construct. T\\no make matters worse, self and identity\\nare both evolving concepts, which have meant different things in different\\nhistorical epochs (Baumeister, 1997; Danziger, 1997). Self and identity are\\nalso doubly contested concepts today. First, the nature and meaning of group-\\nbased identities, such as national, ethnic, or gender-based identities, are \\noften contested in the context of sociocultural discourses (for example, con-\\nsider the potentially strong reactions, both pro and con, that the statement\\n“My country right or wrong” sometimes evokes). Second, the meaning and\\nnature of self and identity as social scientific constructs are contested among\\nresearchers from different traditions (Danziger, 1997; Holland, 1997). So-\\ncial constructivists, for example, often argue against the possibility of even\\nstudying self and identity objectively (Holland, 1997), whereas many psy-\\nchologists and sociologists implicitly assume that self and identity can be\\nstudied objectively (Jussim & Ashmore, 1997). \\nThis volume, however, does not attempt to address the multiplicity of\\ncontroversies surrounding self and identity (though see Ashmore & Jussim,\\n1997, which does just this). This volume focuses on social identity (and \\nits role in intergroup conflict). Social identity is a narrower and more spe-\\ncific subset of the phenomena included under the general rubric of “self \\nand identity.” As such, it may be at least somewhat easier to introduce the\\nterm in a way that will meet with some agreement across disciplines and\\nperspectives.\\nSocial Identity and Intergroup Conflict\\n5\\n\\n\\nWithin psychology and sociology, one of the clearest and most common\\ndefinitions of social identity was provided by T\\najfel (1981, p. 255): “that part\\nof an individual’s self-concept which derives from his knowledge of his\\nmembership in a social group (or groups) together with the value and emo-\\ntional significance attached to that membership.” Thus, according to T\\najfel,\\nsocial identity has two components: belief that one belongs to a group (e.g.,\\n“I am an American”) and the importance of that group membership to one’s\\nself (e.g., “and I am damn proud to be a citizen of the greatest country on\\nearth”). \\nWe have not insisted that the contributors use this definition; they were\\nfree to define social identity in the manner they saw as most appropriate to\\ntheir work. Nonetheless, five of the seven contributors cite and use T\\najfel’s\\ndefinition. Limitations to this definition, however, were implicitly high-\\nlighted by the two authors (Eriksen, Kelman) who did not use it. T\\najfel’s \\ndefinition is almost purely individualistic, focusing exclusively on how the\\nindividual thinks and feels about group memberships. Both Eriksen and\\nKelman suggest that ethnic and national identities are not purely individu-\\nalistic—that such identities emerge within specific sociocultural contexts.\\nSociocultural discourses, national myths, and intergroup relations are all es-\\nsential to the creation of any particular ethnic or national identity, so that,\\naccording to Eriksen and Kelman, social identity resides at least partly\\nwithin one’s national or cultural community, rather than exclusively within\\nthe individual.\\nWhat Do We Mean by Intergroup Conflict?\\nConflict, of course, can occur between almost any type of groups (e.g.,\\nsports teams, corporations, university departments, nations, political par-\\nties, ethnic groups). The chapters in this book, however, primarily (but not\\nexclusively) emphasize conflict between ethnic, racial, religious, and na-\\ntional groups. This emphasis reflects some of the dominant concerns within\\nthe social sciences. Those concerns, we suspect, have largely been moti-\\nvated by a longstanding desire to understand some of humanity’s deepest\\nand most tragic social ills—war, genocide, mass murder, and institutional-\\nized oppression.\\nWithin social psychology, conflict is often defined as some sort of incom-\\npatibility of goals, beliefs, attitudes, and/or behavior (e.g., Myers, 1999).\\nThe nature of the conflicts addressed in this volume are many and varied.\\nSome address economic and political discrimination within a nation; others\\naddress how such discrimination can flare up into murderous intergroup \\nviolence. Five chapters focus on specific, real-world conflicts. Eriksen ad-\\ndresses conflicts between Serbs and Bosnians; Fijians and ethnic Indians in\\nFiji; Hindus and Muslims in India. Citrin et al. and Sidanius and Petrocik\\naddress conflicts between white Americans and ethnic minorities in\\n6\\nIntroduction\\n\\n\\nAmerica. White discusses the multi-sided conflict in Northern Ireland, and\\nKelman describes his own workshops focusing on reducing the conflict be-\\ntween Israelis and Palestinians. \\nThree chapters (Brewer, Eriksen, Staub) seek to provide broad, general\\ninsights into the intersection of social identity and intergroup conflict.\\nBrewer develops a theoretical perspective on the steps leading to progres-\\nsively greater intergroup hostility. Eriksen derives general insights into in-\\ntergroup hostility on the basis of his analysis of three real-world conflicts.\\nStaub focuses specifically on the origins of genocide and mass murder. \\nThe starting point for understanding intergroup conflict often varies\\nacross the different social science disciplines. Within social psychology,\\nconflict of some type or another is often addressed, implicitly or explicitly,\\nin theories and approaches to intergroup relations (e.g., Allport, 1954;\\nOakes, Haslam, & Turner, 1994; Sherif, 1958; T\\najfel, 1981). The goal of\\nsuch research is typically to provide broadly applicable general explanations\\nof phenomena such as bias, prejudice, and discrimination. Social psycho-\\nlogical theories customarily provide individual-level explanations for conflict\\nand the method of choice is, in general, laboratory experimentation. The\\nstarting point for many such perspectives, therefore, is the accumulated ex-\\nperimental evidence addressing issues such as stereotypes and prejudice.\\nThis is in stark contrast to many of the other social sciences, such as an-\\nthropology, political science, and sociology. Work in these disciplines often\\nfocuses on understanding the history, sources, and nature of particular in-\\ntergroup conflicts. When general principles are sought, however, the start-\\ning point is usually real-world conflicts between nations, ethnic groups, or\\nreligious groups. There is often more emphasis on organizational, institu-\\ntional, and sociocultural levels of explanation. The methods of choice tend\\nto be surveys, interviews, ethnographies, archives, and other sources for his-\\ntorical analysis. Consequently, the literatures addressing intergroup conflict\\nin the different social sciences have developed largely independently, and\\noften may appear to be talking past one another. One purpose of this book,\\ntherefore, was to begin to redress this tendency by drawing contributors\\nfrom several social sciences (anthropology, political science, psychology,\\nand sociology). \\nThemes Underlying This Book\\nSeveral themes provide the framework for the specific topics addressed by\\nthe volume: (1) identity issues are important to diverse intergroup conflicts;\\n(2) conflict influences social identity; (3) social identity can contribute to in-\\ntergroup conflict; (4) addressing issues of social identity may be useful or\\neven necessary for reducing some conflicts; and (5) to understand how iden-\\ntity figures in intergroup strife, it is necessary to consider multiple levels of\\nanalysis.\\nSocial Identity and Intergroup Conflict\\n7\\n\\n\\nIdentity Issues Are Important to Diverse Intergroup Conflicts\\nRegardless of whether intergroup conflicts are played out peacefully in cul-\\ntural and political arenas, or become violent and murderous, social identity\\nis usually involved. On the peaceful side, consider the various conflicts\\nwithin the United States over issues such as affirmative action, immigration,\\nand official use of the English language. These conflicts often seem to in-\\nvolve social identities—with policy preferences and attitudes often varying\\namong different ethnic groups in such a manner as to enhance the power,\\nstatus, or privileges of one’s own group.\\nOr consider the more brutal side: the various incidents of mass murder\\nand genocide that have occurred throughout human history. In many cases,\\nextreme nationalist movements link a deep commitment to a national or\\nethnic identity to hatred of a minority group—a pattern that can be found\\nin the Turkish genocide of the Armenians, the German genocide of the\\nJews, and in Serbian mass murder of Bosnians.\\nConflict May Influence Social Identity\\nPeople are usually affected, sometimes deeply, by intergroup conflicts. Cul-\\ntural stereotypes and prejudices can lead members of stigmatized groups to\\ndisidentify with and disengage from tasks in which they are most heavily\\nstigmatized (Steele, 1997). Some African-American youths, for example,\\ndisengage from academic achievement because negative intellectual stereo-\\ntypes about African Americans are too threatening and because high\\nachievement may be seen as “acting white” (Fordham & Ogbu, 1986;\\nSteele, 1997).\\nFurthermore, groups that have once been victimized by conquest, en-\\nslavement, or genocide may, in turn, undergo identity changes that render\\nthem more vulnerable to committing atrocities against other groups in the\\nfuture. At least sometimes, a “we better get them before they get us” atti-\\ntude becomes a critical component of a national or ethnic identity. When\\nsuch an attitude takes hold on a large scale, the society may become less\\nlikely to consider making peace with enemies, more likely to seek violent\\nsolutions to conflicts, and more likely to commit genocide (du Preez, 1994;\\nGoldhagen, 1996; Staub, 1989).\\nSocial Identity May Contribute to Intergroup Conflict\\nIntergroup violence can be fostered by social identities as well as stem from\\ngroup-based self-interest (the notion of realistic group conflict [Ashmore &\\nDel Boca, 1976]). For example, the recent Serbian-Kosovar conflict was\\nstarted over a material issue: who, Serbs or Kosovars, would have political\\nauthority over Kosovo? But one of the reasons this was important to the\\nSerbian government was the historical linkage of Kosovo with Serbian\\n8\\nIntroduction\\n\\n\\nidentity (it was the location of a 600-year-old stand by Serbs against invad-\\ning Turks).\\nAddressing Issues of Social Identity May Be Useful for Reducing Some Conflicts\\nAt the same time, social identity can assist in alleviating intergroup conflict.\\nFor example, Sherif’s (1958) classic work on “superordinate goals” demon-\\nstrated that members of two groups in conflict can work toward a shared\\ngoal. Doing so can involve a new sense of “us,” which likely contributes \\nto the lessened hostility that “superordinate goals” produce (cf. Gaertner,\\nDovidio, Anastasio, Bachman, & Rust, 1993).\\nTo Understand How Identity Figures in Intergroup Strife, It Is Necessary \\nto Consider Multiple Levels of Analysis\\nThe incorporation of the final organizing variable—level of analysis—\\nunderscores that the links of identity to intergroup conflict cannot be re-\\nduced either to the characteristics of individuals or to the qualities of\\ngroups. Fortunately, however, different social and behavioral sciences tend\\nto emphasize different levels of analysis. Thus, it was necessary that con-\\ntributors to this volume come from diverse academic disciplines that cover\\nthe full continuum—from those that focus primarily on the individual (as\\nmost psychologists do), to others that emphasize understanding the rela-\\ntionship of individuals to institutions and social structure within a society\\n(as do many sociologists and political scientists), to still others that investi-\\ngate multiple societies (as do many anthropologists). \\nThe Chapters in This Volume\\nThree sections, each including two chapters, address different aspects of re-\\nlations between identity and conflict: (1) general analyses of how social\\nidentities contribute to conflicts between groups and how conflict can influ-\\nence social identity; (2) analyses of the extent to which ethnic identities \\ninfluence degree of patriotism and attachment to mainstream American po-\\nlitical culture and institutions; and (3) perspectives on how social identity\\ncontributes to lethal intergroup violence. The fourth topic is conflict reso-\\nlution, which is addressed in a chapter that describes a specific conflict re-\\nduction intervention and the role of social identity in this intervention.\\nHow Social Identities Contribute to Intergroup Conflict\\nThe first section analyzes the role of identity, as both a group- and indi-\\nvidual-level variable, in contributing to intergroup conflict. Ethnic and na-\\ntional rivalries have a long history of leading to intergroup conflicts with\\nSocial Identity and Intergroup Conflict\\n9\\n\\n\\nhorrendous consequences. In the extreme, ethnic identity has led to mass\\nmurder (e.g., in the former Yugoslavia). National identities, especially in the\\nform of aggressive nationalism, have contributed to genocides and global\\nwar (e.g., the Nazi Holocaust and World War II). Ethnic identities have\\nalso contributed to slavery, discrimination, and exploitation. \\nIn chapter 2, Brewer reviews the psychological research on social \\nidentity and intergroup strife. Social psychologists have long studied how\\nindividuals’ social identities and group memberships influence their reac-\\ntions to, and perceptions of, others. Much work has addressed the role of\\nindividual-level stereotypes, prejudice, and discrimination in intergroup\\nconflict. In addition, two related theories—social identity theory and self-\\ncategorization theory (T\\najfel, 1981; Turner, 1987) have explicitly identified\\nways in which individuals’ group memberships and social identities may\\nlead to intergroup strife. Brewer integrates existing theoretical perspectives\\non social identity with social cognition research on stereotypes and preju-\\ndice and with her own work on optimal distinctiveness into a new perspec-\\ntive. Her analysis suggests that intergroup harm-doing is the end result of \\na series of progressively more hostile steps that include the psychological\\ncategorization of people into groups, ingroup trust and preference, per-\\nceived moral superiority of the ingroup over outgroups, and antagonism to-\\nward outgroups. She argues that although progression from each step to the\\nnext is not inevitable, the occurrence of each step increases the likelihood\\nthat the next will occur.\\nIn chapter 3, Eriksen addresses this topic by focusing on three case stud-\\nies of the role of identity in intergroup conflict: the Serbian-Bosnian con-\\nflict; the Hindutva movement in India; and relations between ethnic Indians\\nand Fijians on Fiji. In all three cases, Eriksen describes how conceptions of\\n“us” and “them” (social identities and intergroup relations) are influenced\\nby societal/cultural phenomena such as the existing power structure, na-\\ntionalistic leaders, national myths, and modernization. He then discusses\\nhow these socially and historically constructed views of “us” and “them” in\\nturn contribute to intergroup conflict. \\nToWhat Extent Do Ethnic Identities Influence Attachment \\nto the Mainstream Political Culture of the United States?\\nThe second section of the book focuses specifically on relations between\\nethnicity and attachment to mainstream American political culture. T\\no what\\nextent have different ethnic groups adopted mainstream American values\\nand become strongly attached to American political institutions? T\\no what\\nextent does a strong ethnic identity undermine attachment to American \\npolitical culture? T\\nwo chapters address these issues.\\nIn chapter 4, Citrin, Wong, and Duff assess ethnic similarities and dif-\\nferences regarding key, symbolic aspects of American identity. Levels of\\nchauvinism, patriotism, and pride in America are strikingly similar among\\n10\\nIntroduction\\n\\n\\nblacks, whites, Asians, and Hispanics. Personal commitment to an Ameri-\\ncan identity serves to bind Americans to the United States and to bring di-\\nverse groups together in providing a common identity, and, to some extent,\\na common set of interests. \\nHowever, alongside this common sense of attachment to America, many\\npeople, especially members of minority groups, express a strong sense of \\nattachment to their ethnic groups. Among whites, furthermore, a strong at-\\ntachment to a white racial identity seems to foster intergroup conflict.\\nSpecifically, such whites tend to be most resistant to policies aimed at im-\\nproving the social and political conditions facing Blacks. In contrast, strong\\nethnic attachment among minority groups did not seem to foster inter-\\ngroup hostility.\\nIn chapter 5, Sidanius and Petrocik compare three general perspectives—\\nmelting pot, pluralist, and social dominance—regarding relations between\\nethnic identity and national identity among African-American, white, La-\\ntino, Asian, and Jewish United States citizens, and among resident aliens liv-\\ning in the United States They find that: (1) ethnic identities remain strong\\nfor many people; (2) ethnocentrism exists among all groups they studied; \\n(3) differences in patriotism between whites and the non-black ethnic groups\\nis small or nonexistent; but (4) differences in patriotism between African\\nAmericans and whites are larger. Sidanius and Petrocik interpret these results\\nas most consistent with the social dominance model, which predicts that the\\nstrongest support for a society’s culture and institutions will be found among\\nthose groups that benefit most from that culture and those institutions.\\nSocial Identity and Lethal Intergroup Conflict\\nThe third section of the book is devoted to the role of national and ethnic\\nidentity in the instigation of deadly violence. Intergroup hostility, even\\nwhen intense, does not always erupt into violence. And even when violent,\\nconflicts do not always erupt into mass killing, ethnic cleansing, or geno-\\ncide. Nonetheless, systematic and widespread state-sponsored violence is\\nclearly one of the ultimate manifestations of conflict between groups. From\\nbiblical stories of conquest, to the Romans’ treatment of foes and rebellious\\npeoples (e.g., Carthage, Israel), the Crusades, the Turkish slaughter of Ar-\\nmenians, the German Holocaust, the Cambodian genocide, and widespread\\nkillings in Rwanda and Bosnia, mass murder and genocide have emerged as\\ngruesome recurrent themes of intergroup relations. Thus, two chapters ad-\\ndress the role of social identity in lethal intergroup conflicts.\\nIn chapter 6, White presents an analysis of the ongoing hostilities in\\nNorthern Ireland. In tracing the historical roots of the existing conflict, he\\nidentifies three primary perpetrators of violence: the Irish Republican para-\\nmilitaries, Protestant paramilitaries, and the state’s security forces. He ana-\\nlyzes the role of religious, political, and national identities in perpetuating\\nthe conflict and the violence. He concludes that: (1) social identities play a\\nSocial Identity and Intergroup Conflict\\n11\\n\\n\\ncrucial role in maintaining the conflict; and (2) very different types of iden-\\ntities underlie each group’s involvement. IRA members often view them-\\nselves as working for freedom and to overcome the oppression of the Irish\\nby the British; Protestant paramilitary members view themselves as fighting\\nfor the defense of their religious/ethnic group; and British political leaders\\nview themselves, at least according to their public statements, as unbiased\\npeacekeepers (a view that White argues is not completely consistent with\\ntheir actions there). \\nIn chapter 7, Staub presents a social psychological analysis of factors\\nleading to, and preventing, genocide. He describes the social and historical\\nconditions that often predispose a society toward genocidal violence. In so\\ndoing, he highlights the various ways in which self and social identity are in-\\nvolved in the processes of mass killing. For example, he concludes that “in-\\njured” social identities—the belief that one’s group has been unfairly vic-\\ntimized in the past and, therefore, is at greater risk of being victimized in\\nthe future—is one major precursor of genocide (“we have to get them be-\\nfore they get us”). Staub also discusses the types of interventions that seem\\nto avert genocide. \\nSocial Identity Approaches to Reducing Intergroup Conflict\\nIn chapter 8, Kelman describes his extensive experience running conflict-\\nresolution workshops for Palestinians and Israelis. Through these work-\\nshops, it has become clear that each side tends to have components of a na-\\ntional identity that seem to reduce the conflict to a zero-sum game. Indeed,\\neven acknowledging the identity and existence of the other group has often\\nbeen viewed as a threat to one’s own identity and existence as a nation.\\nHowever, Kelman also points out that, although the core identity of both\\npeoples may be non-negotiable, there are many peripheral aspects of those\\nidentities that have emerged from these workshops as negotiable. Kelman’s\\nchapter, therefore, describes the role of identity in both contributing to the\\nIsraeli-Palestinian conflict and in seeking solutions to that conflict.\\nConclusion\\nResearch on identity and on conflict have long histories within the social\\nand behavioral sciences. Until relatively recently, however, research on rela-\\ntions between identity and conflict has been less common. In the last 20\\nyears, researchers in anthropology, political science, psychology, and soci-\\nology have begun to study how group identifications contribute to conflict,\\nhow intergroup conflict can influence identity, and, most recently, how\\nidentities might be harnessed to reduce intergroup hostility. This volume\\ndoes not summarize all such existing work. The following chapters do pre-\\nsent a sophisticated and detailed analysis of some of the most fundamental\\n12\\nIntroduction\\n\\n\\nissues involved in understanding identity and conflict. Chapters address\\nconflict between ethnic, racial, religious, and national groups; they address\\nethnic differences in national attachment; political conflicts over territory;\\nand genocidal conflicts. As the chapters take stock of existing work on iden-\\ntity and conflict within several social scientific disciplines, we suspect that\\nboth their diversity and some similarities in underlying themes will stand\\nout.\\nReferences\\nAllport, G. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley.\\nAshmore, R. D., & Del Boca, F. K. (1976). Psychological approaches to under-\\nstanding intergroup conflict. In P. Katz (Ed.), Towards the elimination of racism\\n(pp. 73–123). Elmsford, NY: Pergamon Press.\\nAshmore, R. D., & Jussim, L. (Eds.) (1997). Self and identity: Fundamental issues\\n(Rutgers Series on Self and Social Identity, Volume 1). New York: Oxford Uni-\\nversity Press.\\nBaumeister, R. F. (1997). The self and society: Changes, problems, and opportuni-\\nties. In R. D. Ashmore, & L. Jussim, (Eds.), Self and identity: Fundamental issues\\n(Rutgers Series on Self and Social Identity, Volume 1; pp. 191–217). New York:\\nOxford University Press.\\nDanziger, K. (1997). The historical formation of selves. In R. D. Ashmore, & L. Jus-\\nsim, (Eds.), Self and identity: Fundamental issues (Rutgers Series on Self and Social\\nIdentity, Volume 1; pp. 137–159). New York: Oxford University Press.\\ndu Preez, P. (1994). Genocide: The psychology of mass murder. Southampton, UK: \\nBowerdean Publishing Co.\\nFordham, S., & Ogbu, J. U. (1986). Black students’ school success: Coping with the\\n“burden of acting White.” Urban Review, 18, 176–206.\\nGaertner, S. L., Dovidio, J. F., Anastasio, P. A., Bachman, B. A., & Rust, M. C.\\n(1993). The common ingroup identity model: Recategorization and the reduc-\\ntion of intergroup bias. In W. Stroebe & M. Hewstone (Eds.), European review of\\nsocial psychology, 4, pp. 1–26, London: Wiley.\\nGoldhagen, D. (1996). Hitler’s willing executioners: Ordinary Germans and the holo-\\ncaust. New York: Alfred A. Knopf.\\nGreenfeld, L., & Chirot, D. (1994). Nationalism and aggression. Theory and Society,\\n23, 79–130.\\nGurr, T. R. (1993). Why minorities rebel: A global analysis of communal mobiliza-\\ntion and conflict since 1945. International Political Science Review, 14, 161–201.\\nHolland, D. (1997). Selves as cultured: As told by an anthropologist who lacks a\\nsoul. In R. D. Ashmore, & L. Jussim, (Eds.), Self and identity: Fundamental Issues\\n(Rutgers Series on Self and Social Identity, Volume 1, pp. 160–190). New York:\\nOxford University Press.\\nJones, J. (1997). Prejudice and Racism (second edition). New York: McGraw-Hill.\\nJussim, L., & Ashmore, R. D. (1997). Fundamental issues in the study of self and\\nidentity—contrasts, contexts, and conflicts. In R. D. Ashmore, & L. Jussim,\\n(Eds.), Self and identity: Fundamental issues (Rutgers Series on Self and Social\\nIdentity, Volume 1, pp. 218–230). New York: Oxford University Press.\\nSocial Identity and Intergroup Conflict\\n13\\n\\n\\nKeegan, J. (1993). A history of warfare. New York: Alfred A. Knopf.\\nKelman, H. C. (1997). Some determinants of the Oslo breakthrough. Negotiaion\\nJournal, 13, 327–340.\\nKreisburg, L., Northrup, T. A. & Thorson, S. J. (1989). Intractable conflicts and\\ntheir transformation. Dynamics of Identity, 4, 55–82. \\nMyers, D. G. (1999). Social psychology (fifth edition). New York: McGraw-Hill.\\nOakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality.\\nCambridge, MA: Blackwell.\\nReich, W. A. (1997). An expanded role for identity in the analysis of intergroup con-\\nflict. Unpublished manuscript.\\nSherif, M. (1958). Superordinate goals in the reduction of intergroup conflict.\\nAmerican Journal of Sociology, 63, 349–363.\\nStaub, E. (1989). The roots of evil: The origins of genocide and other group violence. Cam-\\nbridge: Cambridge University Press.\\nSteele, C. (1997). A threat in the air: How stereotypes shape intellectual identity and\\nperformance. American Psychologist, 52, 613–629. \\nT\\najfel, H. (1981). Social identity and intergroup relations. London: Cambridge Univer-\\nsity Press.\\nThoits, P. A., & Virshup, L. K. (1997). Me’s and we’s: Forms and functions of social\\nidentities. In R. D. Ashmore, & L. Jussim, (Eds.), Self and identity: Fundamental\\nissues (Rutgers Series on Self and Social Identity, Volume 1; pp 106–133). New\\nYork: Oxford University Press.\\nTurner, J. C. (1987). Rediscovering the social group: A self-categorization theory. New\\nYork: Basil Blackwell.\\n14\\nIntroduction\\n\\n\\nPART I\\nTHE CONTRIBUTION OF \\nINDIVIDUALS’ IDENTITIES\\nAND THE COLLECTIVE \\nIDENTITIES OF SOCIAL\\nGROUPS TO INTERGROUP\\nCONFLICT\\n\\n\\nThis page intentionally left blank \\n\\n\\nMarilynn B. Brewer\\n2\\nIngroup Identification and \\nIntergroup Conflict\\nWhen Does Ingroup Love Become Outgroup Hate?\\nSocial psychologists’ interest in social identity and ingroup bias arose in the\\ncontext of research on intergroup relations (T\\najfel &Turner, 1986; Brewer &\\nBrown, 1998). As a consequence, there has been a coupling of the study of in-\\ngroup favoritism and outgroup prejudice, the implicit assumption being that\\nthe more we know about why individuals become attached to their ingroups,\\nthe more we will understand why they come to dislike or derogate outgroups.\\nThe present chapter seeks to challenge that assumption by examining more\\ncarefully the nature of the relationship between ingroup love and outgroup\\nhate. The basic thesis is that the formation of ingroups and ingroup identifi-\\ncation arises independently of attitudes toward outgroups. Attachment to in-\\ngroups is presumed here to be the primary process, fundamental to indi-\\nvidual survival and well-being. Ingroup identification thus precedes\\noutgroup hostility and intergroup conflict but is not by itself a sufficient ex-\\nplanation for such conflict. Instead, as this chapter will argue, we need to\\ntrace the steps that lead from ingroup formation to outgroup hostility, speci-\\nfying the conditions under which the two become entangled.\\nSumner’s Theory of Ethnocentrism\\nAn early theory of the interrelationship between ingroup attachment and\\nidentification and outgroup antagonism was implicit in the definition of\\n“ethnocentrism” when the term was introduced by Sumner (1906). Ethno-\\ncentrism was described by Sumner as a universal characteristic of human\\nsocial groups whereby\\n17\\n\\n\\na differentiation arises between ourselves, the we-group, or in-group, and\\neverybody else, or the others-group, out-groups. The insiders in a we-group\\nare in a relation of peace, order, law, government, and industry, to each other.\\nTheir relation to all outsiders, or others-groups, is one of war and plunder.\\n. . . Ethnocentrism is the technical name for this view of things in which\\none’s own group is the center of everything, and all others are scaled and rated\\nwith reference to it. . . . Each group nourishes its own pride and vanity,\\nboasts itself superior, exalts its own divinities, and looks with contempt on\\noutsiders. . . . (pp. 12–13)\\nThe features characteristic of Sumner’s ethnocentrism syndrome include\\nnegative attitudes toward outgroups as well as positive feelings and evalua-\\ntions of the ingroup. Further, an explicit negative correlation between in-\\ngroup and outgroup attitudes is postulated: the greater the attachment and\\nsolidarity within the ingroup, the greater the hostility and contempt di-\\nrected toward outgroups. This hypothesized relationship between ingroup\\nlove and outgroup hate derived from Sumner’s functional theory of the ori-\\ngins of social groups and intergroup conflict:\\nThe relation of comradeship and peace in the we-group and that of hostility\\nand war towards others-groups are correlative to each other. The exigencies\\nof war with outsiders are what make peace inside, lest internal discord should\\nweaken the we-group for war. . . . Thus war and peace have reacted on each\\nother and developed each other, one within the group, the other in the inter-\\ngroup situation. . . . Sentiments are produced to correspond. Loyalty to the\\ningroup, sacrifice for it, hatred and contempt for outsiders, brotherhood\\nwithin, warlikeness without—all group together, common products of the same\\nsituation. (1906, pp. 12–13, emphasis added)\\nThe last 20 years of social psychological research on ingroup bias and in-\\ntergroup relations have cast doubt on the validity of the functional relation-\\nship between ingroup and outgroup attitudes and behavior presumed by\\nSumner. Results of the laboratory experiments using the “minimal group\\nparadigm” designed by T\\najfel and his colleagues (T\\najfel, Billig, Bundy, & Fla-\\nment, 1971) explicitly challenged the assumption that differential treatment\\nof ingroup and outgroup members derives from cooperative interdepen-\\ndence within groups and competitive relations between groups. In the mini-\\nmal group experiments, mere categorization of persons into differentiated\\nsocial groupings has proved to be sufficient to give rise to preferential dis-\\ncrimination in favor of ingroup over outgroup members, in the absence of\\nany differential interdependence (T\\najfel et al., 1971; Turner, 1978; Brewer,\\n1979; Diehl, 1990). Field studies assessing the relationship between identifi-\\ncation and attachment to ingroups and discrimination against outgroups\\nhave also failed to find any straightforward negative correlation between in-\\ngroup positivity and intergroup attitudes (Hinkle & Brown, 1990). For ex-\\nample, in a study of the reciprocal attitudes among 30 ethnic groups in East\\nAfrica, Brewer and Campbell (1976) found that almost all of the groups ex-\\nhibited systematic differential positive evaluation of the ingroup over all out-\\n18\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ngroups on dimensions such as trustworthiness, obedience, friendliness, and\\nhonesty, but the correlation between this measure of ingroup positivity and\\ndistancing from outgroups was essentially .00 across the 30 groups (p. 85).\\nThus, the question of when and how ingroup identification is related to out-\\ngroup hostility and intergroup conflict requires further investigation.\\nDecomposing Ethnocentrism: Ingroup, Intergroup, and \\nOutgroup Orientations\\nSumner’s definition of ethnocentrism cited at the beginning of this chapter\\ncontains four separate propositions. The four distinguishable elements can\\nbe characterized as follows:\\n1. Human social groups are organized into discrete ingroup-out-\\ngroup categories (the social categorization principle).\\n2. Individuals value their ingroups positively and maintain positive, \\ncooperative relationships with members of the ingroup (the in-\\ngroup positivity principle).\\n3. Ingroup positivity is enhanced by social comparison with out-\\ngroups in which ingroup attributes and outcomes are evaluated as\\nbetter than or superior to those of outgroups (the intergroup compar-\\nison principle).\\n4. Relationships between ingroup and outgroups are characterized by \\nantagonism, conflict, and mutual contempt (the outgroup hostility\\nprinciple).\\nIn contrast to Sumner’s (1906) expectation that these four elements nec-\\nessarily cohere into a pattern that is universally characteristic of intergroup\\nrelations, a review of relevant social psychological research suggests that the\\ncomponents can be distinguished both empirically and conceptually. How-\\never, the components are not entirely independent features of ethnocentric\\nattitudes, either. Instead, I propose that the elements represent a systematic\\nprogression along a continuum of possible relationships between ingroup\\nformation and intergroup behavior in which each element in the progres-\\nsion provides a necessary but not sufficient condition for occurrence of the\\nsubsequent elements. Within this system, the first two elements are proba-\\nbly universal characteristics of human social groups (as Sumner postulated),\\nbut the third and fourth elements require additional social-structural and\\nmotivational conditions that are not inherent in the processes of group for-\\nmation itself.\\nSocial Categorization:The Basic Substrate\\nVirtually all current theories of social identity and intergroup relations rec-\\nognize social categorization as the basic cognitive process underlying all in-\\nIngroup Identification and Intergroup Conflict\\n19\\n\\n\\ntergroup phenomena (e.g, Allport, 1954; T\\najfel, 1969). Categorization (a)\\npartitions the multidimensional variability among human beings into dis-\\ncrete subsets, accompanied by (b) accentuation of perceived intracategory\\nsimilarities and intercategory differences (Doise, Deschamps, & Meyer,\\n1978; T\\najfel, 1969; T\\najfel & Wilkes, 1963). The basic processes of catego-\\nrization and category accentuation are presumed to be the same whether we\\nare talking about individuals’ partitioning of the world of physical objects\\nand events or of the social world. Individuals learn to classify objects as\\nfunctionally interchangeable and develop concepts that distinguish mem-\\nbers of one category from those of another as a fundamental tool for nego-\\ntiating the physical and social environment. \\nThe assimilation and contrast processes underlying category formation\\nand accentuation are independent of any self-involvement in the category\\nsystem. Just as one can acquire a concept of “apple,” which may include a\\nrepresentation of the prototypic member of the category, a theory of the\\n“essence” of appleness, and an understanding of the variability among ex-\\nemplars of the category, so one can acquire similar conceptualizations of so-\\ncial objects such as “librarian” or “African American.” Stereotypic represen-\\ntations (prototypes) of social categories are a natural consequence of\\ncategorization and concept formation processes, motivated by the need for\\ncognitive simplification and social structure. As with any other category\\nconcept, whether social category stereotypes are positive or negative (or\\nmixed) is a function of what has been learned about the content of the cate-\\ngory representation.\\nWhat makes social categorization different from nonsocial categoriza-\\ntion systems, however, is the potential for self-involvement in that social\\nperceivers are themselves members of some social categories (ingroups) and\\nnot others. Thus, an additional categorical distinction between ingroup and\\noutgroups may be superimposed onto particular social classifications. In the\\nabsence of the engagement of this ingroup-outgroup distinction, social \\ncategories are merely cognitive concepts that help individuals to make sense\\nof their social world and to direct their behavior toward other individuals.\\nWhen social categories are also recognized as social ingroups, however, ad-\\nditional processes associated with self-categorization (Turner, Hogg, Oakes,\\nReicher, & Wetherell, 1987) are engaged that lay the groundwork for fur-\\nther social differentiation. Categorization of the self as an ingroup member\\nentails assimilation of the self to the ingroup category prototype and en-\\nhanced similarity to other ingroup members. Self-categorization provides\\nthe cognitive substrate for attachment to ingroups and differentiation from\\noutgroups—the first step toward ingroup bias and discrimination. \\nIngroup Positivity:The Causes and Consequences of Social Identification\\nDifferentiation between ingroups and outgroups does not necessarily carry\\nevaluative implications, and acknowledging one’s ingroup membership is\\n20\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nnot equivalent to identification with that social group. This is clear from\\nT\\najfel’s (1981) definition of social identity as “that part of an individual’s\\nself-concept which derives from his knowledge of his membership of a so-\\ncial group . . . together with the value and emotional significance attached\\nto that membership” (p. 255; emphasis added). Social identification, then,\\nentails affective and evaluative processes that are above and beyond mere\\ncognitive classification of the self and others into a shared social category.\\nThe affective significance of social identification arises from the felt attach-\\nment between the self and the ingroup as a whole. \\nSocial identification represents the extent to which the ingroup has been\\nincorporated into the sense of self, and at the same time, that the self is ex-\\nperienced as an integral part of the ingroup. With high levels of social iden-\\ntification, the group’s outcomes and welfare become closely connected to\\none’s own sense of well-being (Brewer, 1991). According to social identity\\ntheory, it is this engagement of the self that accounts for the positive valua-\\ntion of the ingroup and positive orientations toward fellow ingroup mem-\\nbers. Self-worth is both projected onto and derived from positive ingroup\\nevaluation. The distinction between ingroup membership and ingroup\\nidentification implies that social identification is not an automatic by-\\nproduct of categorization. Identification implies some motivation to define\\noneself in terms of that group membership and to achieve and maintain \\ninclusion in the ingroup category. Since identification entails some sacrifice\\nof an autonomous self-concept, the question arises as to why individuals\\nwould attach some measure of their self-worth and well-being to the fate of\\na collective. \\nOne explanation that has been offered is that social identity and ingroup\\npositivity are engaged in the interests of enhanced self-esteem (Turner,\\n1975; T\\najfel, 1978). But group outcomes are not always positive or self-\\naggrandizing, and identification often entails assimilation of group failures\\nas well as successes (Brewer & Weber, 1994; Crocker & Major, 1989; Hirt,\\nZillmann, Erickson, & Kennedy, 1992; Turner, Hogg, Turner, & Smith,\\n1984). Overall, results of research on the causal role of self-esteem in social\\nidentification are far from convincing (Abrams & Hogg, 1988; Rubin &\\nHewstone, 1998). In fact, there is good reason to believe that attachment of\\nthe self to the group (or other objects) leads to positive valuation of the\\ngroup, rather than vice versa (Greenwald & Banaji, 1995). \\nIf strength of ingroup identification is largely independent of ingroup\\nstatus, then motives other than self-enhancement must be implicated in se-\\nlecting and maintaining social identities. An alternative motivational expla-\\nnation is provided by optimal distinctiveness theory (Brewer, 1991). The\\ntheory postulates that human beings have two powerful social motives: a\\nneed for inclusion that motivates assimilation of the self into large, imper-\\nsonal social collectives, and an opposing need for differentiation that is satis-\\nfied by distinguishing the self from others. As opposing motives, the two\\nneeds hold each other in check. When a person feels isolated or detached\\nIngroup Identification and Intergroup Conflict\\n21\\n\\n\\nfrom any larger social collective, the drive for inclusion is aroused; on the\\nother hand, immersion in an excessively large or undefined social collective\\nactivates the search for differentiation and distinctiveness. \\nIdentification with distinctive social groups emerges from the processes\\ninvolved in satisfying both needs simultaneously. Groups that are exclusive\\nrather than inclusive, groups that have clear categorical boundaries—satisfy\\nthe need for inclusion (intragroup assimilation) at the same time that they\\nprovide a basis for satisfying the need for differentiation (intergroup con-\\ntrast). Clear ingroup boundaries serve to secure both inclusion and exclu-\\nsion. According to optimal distinctiveness theory, ingroups that meet these\\nsimultaneous conditions become an integral part of the individual’s sense of\\nself and the basis for a secure and stable self-concept. Group identification\\nin this sense involves “transformations in the definition of self and the basis\\nfor self-evaluation. When the definition of self changes, the meaning of\\nself-interest and self-serving motivation also changes accordingly” (Brewer,\\n1991, p. 476). \\nOptimal distinctiveness theory proposes that social identification is the\\nproduct of the search for inclusion and differentiation, rather than a conse-\\nquence of the search for self-esteem. Nonetheless, although the desire for\\nself-enhancement may not be a primary cause of ingroup identification,\\nidentification may well lead to a motivation to view the ingroup in the most\\nfavorable possible terms. Drawing on Heider’s (1958) principles of cogni-\\ntive balance, the connection between self and ingroup constitutes a unit re-\\nlationship. Assuming that affect toward the self is primarily one of postive\\nregard, positive attitudes toward the group creates balance between unit\\nand sentiment relations. There is empirical evidence that positivity biases\\nassociated with ingroup identification arise automatically and without\\nawareness. Perdue, Dovidio, Gurtman, and Tyler (1990), for instance,\\ndemonstrated that ingroup signifiers (terms such as “we,” “us,” and “our”)\\nautomatically activate positive evaluative responses, as indicated by differ-\\nential reaction time to recognize subsequent words as “good” (vs. “bad”).\\nConsistency motives are also implicated in relationships between atti-\\ntudes and behavior; positive evaluations and expectations give rise to trust\\nand cooperative behavior that in turn justify positive feelings and future ex-\\npectancies. Thus, the process of assimilation between self and others that\\nunderlies ingroup formation instigates a “benign cycle” of positive affect\\nand behavior.\\nThe emphasis at this stage on positive affect and evaluations directed to-\\nward the ingroup is not devoid of implications for intergroup relations.\\nPositive attitudes and expectations about ingroup members, perceived simi-\\nlarity, and cooperative orientation are all limited to those who share mem-\\nbership in the common ingroup. The extension of positive self-regard to\\nothers stops at the boundary between ingroup and outgroups; attitudes to-\\nward those outside the boundary are, at best, characterized by indifference.\\nDifferential indifference is not without its consequences. When positive re-\\n22\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ngard, empathy, and cooperation are extended to some individuals but not\\nothers on the basis of category membership, an initial form of intergroup\\ndiscrimination is evident (Brewer, 1979, 1996). Compared to ingroupers,\\noutgroupers are less likely to be helped in ambiguous circumstances (Frey\\n& Gaertner, 1986; Gaertner & Dovidio, 1977; Gaertner, Dovidio, & John-\\nson, 1982), more likely to be seen as provoking aggression (Baron, 1979;\\nRogers & Prentice-Dunn, 1981), less likely to receive the benefit of the\\ndoubt in attributions for negative behaviors (Weber, 1994), and likely to be\\nseen as less deserving of public welfare (Pettigrew & Meertens, 1995).\\nStudies of ethnic and racial prejudice in the United States and Europe\\nconfirm that the essence of “subtle racism” is not the presence of strong\\nnegative attitudes toward minority outgroups but the absence of positive sen-\\ntiments toward those groups (e.g., Dovidio & Gaertner, 1993; Pettigrew &\\nMeertens, 1995; Stangor, Sullivan, & Ford, 1991, Study 1). As Gaertner,\\nDovidio, Anastasio, Bachmann, & Rust. (1993) put it, “Initially it was \\nassumed that in situations in which blacks were treated differently than\\nwhites, whites discriminated against blacks, and thus reflected subtle, indi-\\nrectly expressed negative racial attitudes. Alternatively, it is possible that, \\nat some fundamental level, aversive racism reflects a pro-white (i.e., pro-\\ningroup) rather than the solely anti-black bias that was originally pro-\\nposed” (p. 4).\\nIf social groups differ in access to resources and power, the benefits of in-\\ngroup favoritism accrue to members of some groups more than others and\\ncontribute to substantial differences in outcomes between the groups as a\\nwhole. Ultimately, many forms of discrimination may develop not because\\noutgroups are hated, but because positive affects such as admiration, sympa-\\nthy, and trust are reserved for the ingroup and withheld from outgroups\\n(Pettigrew & Meertens, 1995). Differential positivity toward ingroups over\\noutgroups also sets the stage for more pernicious forms of intergroup \\ndiscrimination.\\nIntergroup Comparison: Social Competition for Positive Distinctiveness\\nIngroup positivity implies a sort of self-absorption whereby group members\\nattend to the concerns and welfare of their own group with disregard for\\nthe status or outcomes of those outside the ingroup boundary. The orienta-\\ntion is simply “us” versus “not us,” with little need to differentiate among\\noutgroups or to take outgroup outcomes into account while pursuing in-\\ngroup goals. Satisfaction with the ingroup’s well-being is evaluated against\\ninternal standards and expectations, independent of the status of those out-\\nside the group.\\nThe idea that assessments of goodness and satisfaction can be made in an\\nabsolute sense against clear internal standards may hold true for some di-\\nmensions of evaluation or some conditions of group living. In studies of the\\nindividual self-concept, Neidenthal and Beike (1997) demonstrated that\\nIngroup Identification and Intergroup Conflict\\n23\\n\\n\\nsome self-evaluations are affected by comparative context, but others re-\\nmain stable and independent of comparative information. A parallel distinc-\\ntion may hold for ingroup self-evaluations. Groups that are physically or\\npsychologically isolated from outsiders, for instance, may measure success\\nin terms of basic sustenance or survival in the face of environmental hard-\\nships, or in terms of adherence to dogma or beliefs and practices that are\\nnot shared by those outside the group. \\nIn most cases, however, standards of evaluation are not so clear-cut, and\\nknowledge of the status of outgroups becomes relevant to assessing the state\\nof welfare of the ingroup. The conditions of social comparison at the inter-\\ngroup level are parallel to those for interpersonal comparison in the pursuit\\nof self-evaluation as outlined by Festinger (1954). The need for social com-\\nparison is aroused when there is uncertainty about one’s standing on some\\ndimension of self-evaluation, uncertainty that can be resolved by comparing\\none’s own position to that of relevant others. The relevance requirement is\\nparticularly interesting because it implies that the self and other (or ingroup\\nand outgroup) must be perceived as similar in some sense (i.e, as subject to\\nthe same values or sharing a common frame of reference for evaluation) in\\norder to be useful targets for comparative appraisal. Thus, when ingroup\\nevaluation is dependent upon intergroup comparison, it is necessary to\\nidentify specific outgroups for comparison and to distinguish between out-\\ngroups that are relevant to ingroup self-regard and those that are not. \\nComparison (whether interpersonal or intergroup) is not inherently\\ncompetitive. This depends on what motives social comparison is serving\\n(Wood, 1989). Comparison can be undertaken for purposes of objective\\nself-appraisal or to motivate self-correction and improvement. When as-\\nsessments are relative and evaluative, however, the better the other is judged\\nto be, the worse the self-evaluation. In this case, ingroup favoritism takes a\\nsubtle shift from motivated perceptions that “we are good” to perceptions\\nthat “we are better.” When all of the groups in a social milieu are jockeying\\nfor this type of ingroup advantage, ingroup-outgroup relations are charac-\\nterized by what Turner (1975) referred to as “social competition” for posi-\\ntive identity. Combined with motivations for self-enhancement, then, social\\ncomparison becomes social competition and the pursuit of positive self-\\nregard can be achieved only at the expense of the other. \\nIn this case, competition is not necessarily inherent in the availability of\\npositive outcomes but is created by the evaluation that is attached to out-\\ncomes for the ingroup and outgroup. If resources (economic and social \\nbenefits) are valued in an absolute sense, increases in ingroup welfare can be\\nachieved even if the outgroup is also benefitting. However, when valuation\\n(e.g., subjective states such as status) is relative, the welfare of the ingroup\\nand the outgroup is psychologically a zero-sum distribution. We cannot \\nimprove our position or sense of well-being unless the outgroup is doing \\nless well than we are. When resources are also attached to intergroup differ-\\nences in status and power, this zero-sum mentality is exacerbated. Ingroup\\n24\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nbias motivated by a desire to establish or maintain positive distinctiveness \\nof the ingroup relative to the outgroup is a second form of intergroup dis-\\ncrimination that can be differentiated, at least conceptually, from simple \\ningroup favoritism (Brewer, 1996). \\nInterestingly, the more similar groups are in their values and aspirations,\\nthe more acute the intergroup social competition. When the ingroup and\\noutgroup are arrayed along a single dimension of evaluation that all groups\\nregard as important and relevant to collective self-esteem, comparative ap-\\npraisals are inevitably competitive. In a context of self-enhancement mo-\\ntives, outgroup achievements are threatening to ingroup position and the\\ncloser the outgroup, the greater the threat (T\\nesser, 1988). Social competi-\\ntion and intergroup threat are diluted when the outgroup is very remote or\\ndistant from the ingroup on the dimension of evaluation. Ingroups that en-\\njoy very high status often display largess or magnanimity, rather than in-\\ngroup discrimination, toward outgroups that are much lower down on the\\nstatus ladder. Conversely, low-status groups often exhibit outgroup posi-\\ntivity in evaluations that are related to the status distinction, conceding out-\\ngroup superiority on dimensions where the status difference is clear and\\nlarge (Hinkle & Brown, 1990; Mullen, Brown, & Smith, 1992). \\nEither of these forms of outgroup favoritism, however, is qualified by\\nstrategies that serve ingroup positivity overall. Whereas low-status groups\\nexhibit outgroup positivity on status-relevant evaluations, they generally\\ndisplay greater ingroup bias (compared to their high-status counterparts)\\non dimensions unrelated to the status differential (Brewer & Campbell,\\n1976; Brewer, Manzi, & Shaw, 1993; Mullen, Brown, & Smith, 1992). Simi-\\nlarly, outgroup largess on the part of high-status groups quickly reverts to\\ningroup favoritism when the status gap between ingroup and outgroup is\\nperceived to be closing (Sachdev & Bourhis, 1991; Sidanius, 1993).\\nIn a pluralistic context where multiple dimensions of evaluation are possi-\\nble, social competition for positive distinctiveness may be more diffused and\\nless intense. In research on individual self-evaluation maintenance it has\\nbeen demonstrated that individuals feel threatened by a friend’s achieve-\\nments when the domain of performance is highly important and relevant to\\ntheir own self-evaluation, but a friend’s accomplishments are celebrated\\nwhen the domain is not self-important (T\\nesser, 1988). Similar effects have\\nbeen observed in research on ingroup bias and intergroup discrimination.\\nMummendey and Schreiber (1983) found that when individuals are given the\\nopportunity to choose dimensions of evaluation, they evaluate ingroup prod-\\nucts more favorably than outgroups on dimensions chosen by the ingroup\\nbut rate the outgroup positively on other dimensions. In general, in-group\\nfavoritism is highest on dimensions of evaluation that are important to the\\ningroup, but is mitigated for dimensions that are important to the outgroup\\nbut relatively unimportant to the ingroup (Mummendey & Schreiber, 1984;\\nMummendey & Simon, 1989). \\nIn interpreting these findings, Mummendey and Schreiber (1983) sug-\\nIngroup Identification and Intergroup Conflict\\n25\\n\\n\\ngest that social competition occurs only when there is no alternative for es-\\ntablishing one’s own positive identity than at the expense of the outgroup.\\nWhen positive identity is possible for both groups at the same time, judg-\\nments are more influenced by fairness than by ingroup bias. Thus, positive\\ndistinctiveness does not necessitate across-the-board favoritism of ingroup\\nover outgroups. Ingroup distinctiveness at the expense of outgroups obtains\\nonly when valuation is relative and on dimensions that are particularly im-\\nportant to the group’s collective self-worth.\\nOutgroup Antagonism and Intergroup Aggression\\nIntergroup comparison and social competition provide an initial connection\\nbetween valuing the ingroup and devaluing relevant outgroups. In the lit-\\nerature on social motives (McClintock, 1972), however, a clear distinction is\\nmade between competition—the motivation to seek relative gain for the in-\\ngroup over others—and aggression—the motivation to harm the other as an\\nend in itself. Following this distinction, we can distinguish between inter-\\ngroup discrimination that is based solely on ingroup favoritism, and dis-\\ncrimination or prejudice that entails an active component of outgroup dero-\\ngation and aggression (Levin & Sidanius, 1999; Struch & Schwartz, 1989).\\nAggression directed against an outgroup may originate in the service of\\nprotection or enhancement of the ingroup, but it may also become an end\\nin itself—outgroup hate that has become dissociated from ingroup en-\\nhancement or desire to benefit the ingroup.1\\nWhen outcomes are objectively (or perceived to be) zero-sum, it be-\\ncomes difficult to pin down the locus of bias (Brewer, 1979). If ingroup out-\\ncomes improve only as outgroup benefits decrease, it is not clear whether\\nthe motivation for discriminatory behavior is to benefit the ingroup (i.e.,\\ndiscrimination for us) or to disadvantage the outgroup (discrimination\\nagainst them). Still, the experimental literature on intergroup discrimina-\\ntion provides evidence that the primary motivation is to benefit the ingroup\\nrather than harm the outgroup. In studies where evaluations of the ingroup\\ncan be assessed independently of evaluations of the outgroup, conditions\\nthat increase ingroup identification and favoritism generally involve en-\\nhancement of ingroup evaluations rather than decreased evaluation or dero-\\ngation of outgroups (e.g., Bettencourt & Dorr, 1998; Brewer, 1979; Brewer,\\nManzi, & Shaw, 1993). \\nMore direct evidence is provided by research on the asymmetry of posi-\\ntive and negative ingroup bias (Mummendey & Otten, 1998). Although the\\ninitial experiments by T\\najfel et al. (1971) claimed that ingroup bias in the\\nminimal intergroup situation occurred for allocations of both positive and\\nnegative outcomes, subsequent research suggests that when the outcomes\\nto be distributed are negative or harmful, the usual intergroup discrimina-\\ntion may be lessened or may disappear all together. Hewstone, Fincham,\\nand Jaspars (1981) modified the normal paradigm by asking group members\\n26\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nto subtract money from ingroup and outgroup recipients. Although some\\ningroup bias was observed in this context, the levels were lower than those\\nobtained with the standard allocation matrices. Mummendey et al. (1992)\\nextended this finding by asking participants to distribute different durations\\nof an unpleasantly high-pitched tone to ingroup and outgroup members.\\nWith harm as an outcome, ingroup bias was generally eliminated and\\nstrategies to equalize or minimize the amount of aversive stimulation were\\nobserved.\\nMummendey and Otten (1998) have suggested that the positive-negative\\nasymmetry in intergroup discrimination is explained by normative con-\\nstaints that make it more difficult to justify differential allocations that harm\\nothers directly than just benefit some more than others. As long as everyone\\nobtains some benefit, it is relatively easy to find justification for why the in-\\ngroup might be entitled to benefit more than the outgroup. There is even\\nevidence that such ingroup benefiting is considered normative in its own\\nright (Blanz, Mummendey, & Otten, 1997; Platow, O’Connell, Shave, &\\nHanning, 1995). Thus, allocations that favor the ingroup over the outgroup\\nmay be motivated primarily by intentions to be a “good” ingroup member,\\nrather than by any sentiments against the outgroup. \\nInflicting harm, however, is not sanctioned so directly. It is more difficult\\nto justify protecting the ingroup if it comes at the expense of inflicting harm\\non others (Mummendey & Otten, 1998). Thus, even when the ingroup may\\nbe disadvantaged, intergroup discrimination is constrained when there \\nis no direct motivation to harm or derogate the outgroup. These are the\\nconstraints that are lifted when outgroups are viewed with hatred or con-\\ntempt—emotions that justify outgroup harm above and beyond ingroup\\nbenefit. Yet outside of the laboratory, the world is full of evidence of the \\nexistence of such self-justified harm toward outgroups, even at great costs\\nto the ingroup. A desire for positive distinctiveness can account for prefer-\\nential hiring of ingroup members over outgroup members and for selective\\ninvestment in projects that benefit one’s own group over others. But wars \\nof conquest, pograms, and ethnic cleansing require explanation that goes\\nbeyond that of achieving positive distinctiveness for the ingroup.\\nSocial identity theory provides an adequate framework for understand-\\ning ingroup bias and intergroup discrimination of the positive types de-\\nscribed earlier. Assimilation of the self to the ingroup accounts for ingroup\\npositivity, and from there, the extension to relative enhancement of the in-\\ngroup over outgroups under conditions of intergroup comparison is not a\\nbig motivational leap. The story becomes more complicated, however,\\nwhen one tries to use the same framework to account for more virulent out-\\ngroup hate and intergroup hostility. T\\no justify aggression against outgroups\\nin the interest of the ingroup, the very existence of the outgroup, or its\\ngoals and values, must be seen as a threat to the maintenance of the ingroup\\nand to one’s own social identity. Thus, understanding the relationship be-\\ntween ingroup identification and outgroup hostility requires understanding\\nIngroup Identification and Intergroup Conflict\\n27\\n\\n\\nhow the interests of the ingroup and those of the outgroup come to be \\nperceived as in conflict. For that purpose, we need to reconsider the funda-\\nmental motivations underlying ingroup attachment and the path from in-\\ngroup identity to outgroup threat.\\nFrom Ingroup Formation to Intergroup Conflict: Another Look\\nRealistic competition over scarce vital resources provides one explanation for\\nperceived conflict of interests. When resources for survival or flourishing are\\nscarce and groups are locked into negative interdependence (competing for\\nthe same resources), the link between ingroup survival and destruction of the\\noutgroup is direct. But analyses of warfare and intergroup violence call into\\nquestion the necessary role of realistic, objective conflicts of interest in many\\ncases of intergroup conflict (e.g., Sears & Funk, 1991). Perceived threat is of-\\nten subjective and symbolic, and overt conflict often appears to be the prod-\\nuct of prior intergroup antagonism rather than its cause.\\nThe thesis that is being developed here gives prior status to ingroup \\ncategorization and identification for understanding the origins of intergroup\\nbehavior. This view assumes that ingroups are formed for reasons indepen-\\ndent of intergroup conflict and competition over resources. The social and\\npsychological functions served by ingroup formation and identification pro-\\nvide the necessary conditions for collective action—either intragroup or \\nintergroup. Hence, ingroup attachment and loyalty lay the groundwork for\\nintergroup conflict, but outgroup antagonism is not a necessary extension of\\ningroup positivity and enhancement. Explicating the relationship between\\nsocial identity and intergroup conflict requires a fuller understanding, first,\\nof the functions of ingroup identification and, then, of the additional condi-\\ntions that lead to a connection between ingroup attachment and outgroup\\nhostility.\\nIngroup Formation: The Boundaries of Cooperation and Trust\\nIf ingroup-outgroup distinctions do not always involve intense (or even\\nmild) competition or conflict over scarce resources, there is need for a\\ntheory of the evolution of social groups that does not depend on intergroup\\nconflict per se. Such a theory starts from the recognition that group living\\nrepresents the fundamental survival strategy that characterizes the human\\nspecies. In the course of our evolutionary history, humans abandoned most\\nof the physical characteristics and instincts that make possible survival and\\nreproduction as isolated individuals or pairs of individuals in favor of other\\nadvantages that require cooperative interdependence with others in order\\nto survive in a broad range of physical environments. In other words, as a\\nspecies we have evolved cooperation rather than strength and social learn-\\ning rather than instinct as basic adaptations.\\n28\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nThe result is that, as a species, human beings are characterized by obliga-\\ntory interdependence (Brewer, 1997; Caporael, 1997). For long-term survival,\\nwe must be willing to rely on others for information, aid, and shared re-\\nsources, and we must be willing to give information and aid and to share \\nresources with others. At the individual level, the potential benefits (receiv-\\ning resources from others) and costs (giving resources to others) of mutual\\ncooperation go hand-in-hand and set natural limits on cooperative interde-\\npendence. The decision to cooperate (to expend resources to another’s \\nbenefit) is a dilemma of trust since the ultimate benefits depend on every-\\none else’s willingness to do the same (Brewer, 1981). A cooperative system\\nrequires that trust dominate over distrust. But indiscriminate trust (or indis-\\ncriminate altruism) is not an effective individual strategy; cooperation must\\nbe contingent on the probability that others will cooperate as well.\\nSocial categorization and clear group boundaries provide one mechanism\\nfor achieving the benefits of cooperative interdependence without the risk of\\nexcessive costs. Ingroup membership is a form of contingent altruism. By\\nlimiting aid to mutually acknowledged ingroup members, total costs and\\nrisks of nonreciprocation can be contained (see T\\nakagi, 1996, for a related ar-\\ngument). Thus, ingroups can be defined as bounded communities of mutual\\ntrust and obligation that delimit mutual interdependence and cooperation. \\nAn important aspect of this mutual trust is that it is depersonalized\\n(Brewer, 1981), extended to any member of the ingroup whether personally\\nrelated or not. Psychologically, expectations of cooperation and security\\npromote positive attraction toward other ingroup members and motivate\\nadherance to ingroup norms of appearance and behavior that ensure one\\nwill be recognized as a good or legitimate ingroup member. Symbols and\\nbehaviors that differentiate the ingroup from local outgroups become par-\\nticularly important here, to reduce the risk that ingroup benefits will be \\ninadvertently extended to outgroup members, and to assure that oneself \\nis recognized as a member of the ingroup and entitled to those benefits. As-\\nsimilation within and differentiation between groups is thus mutually rein-\\nforcing, along with ethnocentric preference for ingroup interactions and \\ninstitutions.\\nThe optimal distinctiveness model of social identity (Brewer, 1991) was\\nbased on this evolutionary perspective on the functions of ingroup for-\\nmation and differentiation. The theory assumes that psychological me-\\nchanisms at the individual level (opposing needs for differentiation and \\ninclusion) co-evolved with structural requirements at the group level for\\nsuccessful group functioning and coordination (size and boundedness). The\\nrelative strength of inclusion and differentiation needs at the individual\\nlevel interacts with group properties such as size and permeability to de-\\ntermine social identification and ingroup attachment and loyalty. The psy-\\nchology of assimilation and differentiation limits the extent to which strong\\nsocial identification can be indefinitely extended to highly inclusive, super-\\nordinate social groups or categories. \\nIngroup Identification and Intergroup Conflict\\n29\\n\\n\\nThus, one implication of optimal distinctiveness theory is that ingroup\\nloyalty, and its concomitant depersonalized trust and cooperation, is most\\neffectively engaged by relatively small, distinctive groups or social cate-\\ngories. Groups that satisfy these needs are those that provide the necessary\\nboundaries on mutual obligation and trust. Optimal group identities be-\\ncome essential to a secure self-concept. Once group identification has been\\nachieved, maintaining a secure sense of inclusion and boundedness becomes\\ntantamount to protecting one’s own existence. When optimal inclusion\\nwithin a distinctive ingroup is not achieved, the sense of self is threatened\\nand vulnerable.\\nIntergroup Differentiation and Outgroup Stereotypes\\nA consequence of ingroup identification and intergroup boundaries is that\\nindividuals modify their social behavior depending on whether they are in-\\nteracting with ingroup or outgroup members. Ingroup behavior is governed\\nby norms and sanctions that reinforce expectations of mutual cooperation\\nand trustworthiness. Depersonalized trust is supported by implicit under-\\nstandings that ingroup members will monitor the behavior and interactions\\nof other group members, sanctioning deviations from group expectations\\nabout appropriate ingroup attitudes and behavior. Thus, shared ingroup\\nmembership may be taken as prima facie evidence that other members of the\\ngroup will live by the codes of conduct that bind them as a group (Kramer,\\nBrewer, & Hanna, 1996).\\nJust as there is a realistic basis for ethnocentric trust of ingroups, differ-\\nences in norms and sanctions applied to ingroup behavior compared to be-\\nhavior in interactions with outgroup members provide a realistic basis for\\noutgroup distrust and negative stereotypes. At the same time that groups\\npromote trust and coopration within, they caution wariness and constraint\\nin intergroup interactions. Thus, even in the absence of overt conflict be-\\ntween groups, the differentiation between ingroup and outgroup behavior\\ncreates a kind of self-fulfilling prophecy in the realm of intergroup percep-\\ntions. As LeVine and Campbell (1972) put it, “if most or all groups are, in\\nfact, ethnocentric, then it becomes an ‘accurate’ stereotype to accuse an\\noutgroup of some aspect of ethnocentrism” (p. 173). \\nCombined with the accentuation principle that exaggerates perceived\\ndifferences between social categories, this leads to a set of “universal stereo-\\ntypes” to characterize ingroup-outgroup differences. Whereas “we” are\\ntrustworthy, peaceful, moral, loyal, and reliable, “they” are clannish, exclu-\\nsive, and potentially treacherous. What is particularly interesting about this\\npattern of stereotypes is that the same behaviors that are interpreted as rea-\\nsonable caution on the part of the ingroup in dealings with outgroup mem-\\nbers become interpreted as “clannishness” and indicators of potential\\ntreachery when exhibited by outgroupers toward the ingroup. \\nEvidence for this hypothesized universal difference in ingroup versus out-\\n30\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ngroup stereotypes was provided by data from our survey of intergroup per-\\nceptions in East Africa (Brewer & Campbell, 1976). Respondents in the \\nsurvey were presented with a lengthy list of character traits (both positive \\nand negative) and asked to indicate which groups were most likely to exhibit\\neach trait. When responses to this trait list were factor analyzed, the first \\nfactor that emerged was a bipolar evaluative factor interpretable as trustwor-\\nthiness versus untrustworthiness. Positive loadings on this factor were asso-\\nciated with traits such as “peaceful,” “obedient,” “honest,” “gentle,” and\\n“friendly.” Negative loadings were obtained for traits such as “quarrelsome,”\\n“disobedient,” “dishonest with others,” “cruel,” and “hot-tempered.” An \\nindex of positive regard was constructed by taking the frequency with which\\na group was mentioned in connection with positive traits on this factor and\\nsubtracting the frequency of mentions in connection with the negative traits.\\nComparing ingroup and outgroup ratings on this index revealed that all 30\\ngroups in the survey rated their own group more positively than the average\\nrating given to outgroups. Further, 27 of the groups rated the ingroup higher\\nthan any outgroup in the sample (Brewer & Campbell, 1976, p. 79). Although\\nthe assignment of negative traits such as “quarrelsome,” “dishonest,” and\\n“cruel” varied considerably across outgroups, attributions of positive traits\\nsuch as “peaceful,” “honest,” and “friendly” were almost universally reserved\\nfor the ingroup.\\nThe socialization experiences that lead to differential expectations and\\ncomfort in ingroup dealings compared to dealings with outgroups may be\\ngeneralized in the form of a generic schema that is brought to all intergroup\\ncontexts. Whenever ingroup-outgroup distinctions become salient, the in-\\ntergroup schema is activated. Just such a generic schema was suggested by\\nT\\najfel in his initial explanations for the minimal intergroup discrimination\\nfindings (T\\najfel et al., 1971). At that time, T\\najfel (1970) speculated about a\\n“generic norm” of intergroup behavior that prescribed outgroup competi-\\ntion. More recently, Insko and Schopler (1987) also postulated such a group\\nschema as one explanation for their robust finding that in interdepen-\\ndent situations, people are consistently more competitive in intergroup ex-\\nchanges than when facing the same situation in an interpersonal context—a\\nfinding that is known as the “individual-group discontinuity effect” (Insko\\net al., 1987, 1990). A primary component of the group schema, as devel-\\noped by Insko and his colleagues, is schema-based distrust—the learned be-\\nlief or expectation that intergroup relations are inherently competitive and\\ntherefore the outgroup cannot be trusted and the ingroup’s welfare must be\\nprotected (see also Fiske & Ruscher, 1993). In an experimental test of the\\nschema-based mistrust hypothesis, Insko et al. (1990) found that group dis-\\ncussion preceding an intergroup prisoner’s dilemma game was characterized\\nby frequent references to the potential untrustworthiness of the other\\ngroup, and there was a strong negative correlation between the amount of\\nrecorded distrust during discussion and later cooperation in the PDG play.\\nFurther, groups more than individuals chose to withdraw from play when\\nIngroup Identification and Intergroup Conflict\\n31\\n\\n\\nthat option was made available, where withdrawal serves as an indicator of\\nfear that the other player will prefer competition to cooperation.\\nThe Role of Emotions in Intergroup Contexts\\nWhen distrust leads groups to compete in an effort to protect themselves\\nfrom anticipated outgroup defection, a self-fulfilling cycle of competitive-\\nness is set in motion (Kelley & Stahelski, 1970). What is of importance here\\nis the idea that this cycle is the product of a misattribution process whereby\\nuniversal ethnocentric behavior (intragroup trust and lack of trust in inter-\\ngroup encounters) is used to infer malevolent intent on the part of out-\\ngroups. Similar misattribution processes may be at work in interpreting and\\nevaluating intergroup differences more generally. If individuals feel discom-\\nfort and unease in interactions with outgroups, those feelings may well be\\nattributed to intergroup differences in values, attitudes, and cultural prac-\\ntices. Hence, negative affect becomes associated with intergroup differences\\nand, through a process of misattribution of arousal, intergroup anxiety is\\ntransformed into more virulent intergroup emotions of fear, hatred, or dis-\\ngust (Smith, 1993; Stephan & Stephan, 2000). It is this emotional compo-\\nnent that is postulated here to be the critical ingredient that turns inter-\\ngroup comparison into intergroup antagonism.\\nOne perspective on the etiology of intergroup emotions can be derived\\nfrom the theory of optimal distinctiveness (Brewer, 1991) described earlier.\\nThe theory holds that strong ingroup identification is associated with meet-\\ning needs for secure inclusion (though similarity to the ingroup) and secure\\ndifferentiation (through intergroup distinctions). From this perspective,\\nthere are two fundamental sources of potential threat to a person’s social\\nidentity. One threat comes from feelings of marginalization from the \\ndesired ingroup—loss of a secure sense of inclusion and fit. One way of cop-\\ning with this type of threat is to reaffirm support for ingroup positivity \\nand moral superiority. However, hypervaluation of the ingroup alone may\\nactually increase the distance between the self and the ingroup ideal. Dero-\\ngation of outgroups, however, increases the contrast between ingroup \\nand outgroup and, indirectly, the relatively similarity between the self \\nand the ingroup by comparison. Thus, threats to inclusion are predicted \\nto heighten feelings of moral superiority, intolerance of difference, and con-\\ncomitant emotions of contempt and disgust toward relevant outgroups.\\nThe emotions associated with moral superiority may justify some nega-\\ntive discrimination against outgroups, but do not necessarily lead directly to\\nhostility or conflict. In various contexts, groups have managed to live in a\\nstate of mutual contempt over long periods without going to war over their\\ndifferences. The emotions of contempt and disgust are associated with\\navoidance rather than attack, so intergroup peace may be maintained\\nthrough segregation and mutual avoidance. However, social changes that\\ngive rise to the prospect of close contact, integration, or influence from the\\n32\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\noutgroup engage the second form of threat to social identity—the threat of\\nloss of distinctiveness, accompanied by feelings of invasion. The emotions\\nengendered by this form of threat are fear and anger—emotions that propel\\naction against the outgroup rather than mere avoidance. This analysis sug-\\ngests that the combined emotions of contempt (engendered by moral supe-\\nriority in the service of insecure identity) and anger (engendered by fear of\\ninvasion and loss of distinctiveness) provide the potent ingredients that are\\nsufficient to kindle hatred, expulsion, and even ethnic cleansing.\\nSome Moderating Conditions\\nThe preceding analysis indicates that there are a number of steps by which\\nthe effects of social categorization and ingroup-outgroup differentiation\\nmay provide a fertile ground for conflict and hate. As the steps progress, the\\nline is crossed between the absence of trust and the presence of active dis-\\ntrust, or between noncooperation and overt conflict. In this final section I\\nconsider a few of the conditions that may exacerbate or moderate the\\nprocess whereby maintaining ingroup integrity and loyalty paves the way to\\noutgroup hate and hostility.\\nCommon Goals.\\nThe presence of superordinate goals or common threat is\\nwidely believed to provide the conditions necessary for intergroup coopera-\\ntion and reduction of conflict (e.g., Sherif , 1966). This belief is an extrapo-\\nlation of the general finding that intragroup solidarity is increased in the face\\nof shared threat or common challenge. It may be true that loosely knit in-\\ngroups become more cohesive and less subject to internal factioning when\\nthey can be rallied to the demands of achieving a common goal. The dy-\\nnamics of interdependence are quite different, however, in the case of\\nhighly differentiated social groups. Among members of the same ingroup,\\nengaging the sense of trust necessary for cooperative collective action is es-\\nsentially nonproblematic. In an intergroup context, however, perceived in-\\nterdependence and the need for cooperative interaction make salient the\\nabsence of mutual trust. Without the mechanism of depersonalized trust\\nbased on common identity, the risk of exploited cooperation looms large\\nand distrust dominates over trust in the decision structure. It is for this rea-\\nson that I have argued elsewhere (Brewer, 2000) that the anticipation of posi-\\ntive interdependence with an outgroup, brought on by perceptions of com-\\nmon goals or common threat, actually promotes intergroup conflict and\\nhostility. When negative evaluations of the outgroup such as contempt are\\nalso already present, common threat in particular may promote scapegoat-\\ning and blame rather than mutual cooperation.\\nPerceived positive interdependence with the outgroup also threatens \\nintergroup differentiation. T\\no the extent that feelings of secure inclusion,\\ningroup loyalty, and optimal identity are dependent upon the clarity of in-\\ngroup boundaries and intergroup distinctions, shared experiences and co-\\nIngroup Identification and Intergroup Conflict\\n33\\n\\n\\noperation with the outgroup threaten the basis for social identification. Par-\\nticularly for individuals who are exclusively vested in a single group identity,\\nthe threat of lost distinctiveness may override the pursuit of superordinate\\ngoals and lead to resistance to cooperation (collaboration) even at the cost\\nof ingroup self-interest.\\nPower Politics. Moral superiority, distrust of outgroups, and social compari-\\nson are all processes that emerge from ingroup maintenance and favoritism\\nand can lead to hostility and conflict between groups even in the absence of\\nrealistic conflict over material resources or power. When groups are politi-\\ncal entities, however, these processes may be exacerbated through deliber-\\nate manipulation by group leaders in the interests of mobilizing collective\\naction to secure or maintain political power. Social category differentiation\\nprovides the fault lines in any social system that can be exploited for politi-\\ncal purposes. When trust is ingroup-based, it is easy to fear control by out-\\nsiders; perceived common threat from outgroups increases ingroup cohe-\\nsion and loyalty; appeals to ingroup interests have greater legitimacy than\\nappeals to personal self-interest. Thus politicization—an important mecha-\\nnism of social change—can be added to the factors that may contribute to a\\ncorrelation between ingroup love and outgroup hate.\\nCulture. In addition to psychological needs, there are social structural and\\ncultural factors that may promote the intensity of ingroup attachment and\\nloyalty and enhance the distance between ingroup and outgroups. When\\nthe salience and strength of intragroup interdependence and mutual obliga-\\ntion is increased, the importance of maintaining group boundaries is also\\nincreased. Hence we might expect that in collectivist societies ingroup-\\noutgroup distinctions and distrust of outgroups would be higher than in \\nindividualistic societies where social interdependence is less emphasized.\\nIndeed, findings from cross-cultural studies of ingroup bias in collectivist\\nand individualist societies supports this prediction (Triandis, 1995).\\nSocial Structural Complexity. A strong relationship between intense ingroup\\nfavoritism and outgroup antagonism might also be expected in highly seg-\\nmented societies that are differentiated along a single primary categoriza-\\ntion, such as ethnicity or religion. And this would be especially true if the\\ncategorization is dichotomous, dividing the society into two significant sub-\\ngroups. Such segmentation promotes social comparison and perceptions of\\nconflict of interest that give rise to negative attitudes toward outgroups and\\nhigh potential for conflict. By contrast, the potential for intergroup conflict\\nmay be reduced in societies that are more complex and differentiated along\\nmultiple dimensions that are not perfectly correlated. Such a complex social\\nstructure gives rise to cross-cutting category distinctions that mean that, at\\nthe individual level, a person may be attached to one ingroup by virtue of\\nethnic heritage, to another by religion, to yet another based on occupation,\\n34\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nor region of residence, and so forth. With this profusion of social identities\\nsome individuals will be fellow ingroup members on one category distinc-\\ntion but outgroupers on another. Such cross-cutting ingroup-outgroup dis-\\ntinctions reduce the intensity of the individual’s dependence on any particu-\\nlar ingroup for meeting psychological needs for inclusion, thereby reducing\\nthe potential for polarizing loyalties along any single cleavage or group dis-\\ntinction and perhaps increasing tolerance for outgroups in general.\\nSummary and Concluding Perspectives\\nThe main point of this chapter has been to argue that any relationship be-\\ntween ingroup identification and outgroup hostility is progressive and con-\\ntingent rather than necessary or inevitable. Ingroup formation arises from\\nprocesses of social categorization and the need for cooperative interdepen-\\ndence within groups. At the individual level, identification with ingroups \\nis motivated by the need to belong and to participate in bounded coopera-\\ntive social units. The bounded nature of ingroup attachment means that\\ntrust and cooperation are limited to ingroup relations and withheld from \\noutgroups.\\nOnce the self is attached to a distinctive ingroup, additional motives to\\nachieve positive valuation, maintain secure inclusion, and protect ingroup\\nboundaries are engaged. It is at this point that outgroups may become rele-\\nvant to the maintenance or enhancement of ingroup identity. Social compe-\\ntition for positive valuation shifts initial indifference about outgroup out-\\ncomes to concern for the relative position of the ingroup over outgroups.\\nUnder these conditions, benefiting the ingroup comes at the expense of\\noutgroup benefits, and positive outcomes for the outgroup may arouse re-\\nsentment and antagonism if the comparison threatens important bases of\\ningroup evaluation. This is the first step in connecting ingroup identity to\\nnegative outgroup attitudes and underlies many forms of social discrimina-\\ntion and intergroup competition for social status.\\nAt a more extreme point in the progression from ingroup identity to in-\\ntergroup conflict, perceived threats to ingroup inclusion or differentiation\\nfrom outgroups promote perceived conflict between ingroup identity and\\nthe very existence of the outgroup or its goals and values. This is the point\\nat which preservation of the ingroup justifies direct aggression against \\nthe outgroup and underlies conflicts over political control, dominance, and\\nwars of annihilation. Such intense outgroup hostility is not a necessary or\\ninevitable concomitant of strong ingroup identity, but the basic processes of\\nsocial categorization that give rise to differential treatment of ingroup and\\noutgroup lay the groundwork for distrust and vulnerability that make this\\nescalation possible.\\nThis analysis of the relationship between ingroup identity and outgroup\\nattitudes has a number of implications for research on intergroup relations\\nIngroup Identification and Intergroup Conflict\\n35\\n\\n\\nand the prevention of the more virulent forms of intergroup conflict. First,\\nit implies that the study of ingroup bias, prejudice, and discrimination\\nshould make a clearer distinction between discriminatory attitudes and be-\\nhavior that are based on ingroup preference and positivity and discrimina-\\ntion that is based on antagonism and hostility toward outgroups. This is not\\nto suggest that ingroup-based prejudices are entirely benign or acceptable.\\nIndeed, these are the prejudices that are most likely to be subtle, uncon-\\nscious, and pervasive and thus more difficult to detect and address than\\n“old-fashioned” outgroup hatred. But understanding the motives underly-\\ning discrimination based on group membership leads to different prescrip-\\ntions for preventing or reducing the discriminatory outcomes.\\nWhen differential cooperation, positivity, and trust toward others is\\nbased on shared ingroup membership, discrimination can best be reduced\\nby conditions that foster greater inclusiveness, extending the boundaries of\\nthe ingroup to include former outgroup members. This is the basic ratio-\\nnale for the common ingroup identity model for reducing intergroup bias\\nthrough recategorization of ingroup and outgroup into a superordinate \\ncategory identity (Gaertner et al., 1993). This is most likely to be successful\\nwhen symbols of superordinate category membership are highly salient and\\nindividual needs for differentiation are not aroused.\\nWhen intergroup attitudes and relations have moved into the realm of\\noutgroup hate or overt conflict, however, the prospect of superordinate\\ncommon group identity may constitute a threat rather than a solution\\n(Brewer, 2000). When outgroup attitudes are characterized simply by lack\\nof trust rather than active distrust, there is opportunity for capitalizing on\\ncommon interests and shared identities. But when intense distrust has al-\\nready developed, common group identities are likely to be seen as threats\\n(or opportunities) for domination and absorption. In this case, the prescrip-\\ntion for conflict reduction may first require protection of intergroup\\nboundaries and distinctive identities.\\nFrom this analysis there is an inevitable tension between promoting co-\\noperative interdependence and shared interests at high levels of inclusive-\\nness and the need to maintain and recognize intergroup differentiation and\\ndistinctiveness. The negative effects of divisiveness will be most virulent\\nwhen individuals have very limited options for meeting their needs for se-\\ncure, optimal group identification. When social cleavages are defined along\\na single dimension of differentiation, vulnerability to threats to inclusion\\nand differentiation is very high, and maintenance of ingroup identity is\\nmost likely to be achieved at the cost of high levels of distrust and antago-\\nnism toward outgroups. When societies are more complex, when individu-\\nals have multiple avenues for meeting needs for inclusion and distinctive-\\nness, intergroup comparisons along any particular division are likely to be\\nless competitive or critical to member identities. Under these conditions,\\nthe positive benefits of distinctive ingroup identities may be attained with-\\nout the costs of intergroup hostility and conflict.\\n36\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nNote\\n1. Note, we are dealing here with prejudice and hostility at the collective level.\\nThe hatred that a specific individual bears toward a particular type of people is not\\nat issue unless that hatred is grounded in norms and attitudes shared by the ingroup\\nas a whole.\\nReferences\\nAbrams, D., & Hogg, M. A. (1988). Comments on the motivational status of self-\\nesteem in social identity and intergroup discrimination. European Journal of Social\\nPsychology, 18, 317–334.\\nAllport, G. W. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley.\\nBaron, R. A. (1979). Effects of victim pain cues, victim’s race, and level of prior insti-\\ngations upon physical aggression. Journal of Applied Social Psychology, 9, 110–114.\\nBettencourt, B. A., & Dorr, N. (1998). Cooperative interaction and intergroup bias:\\nEffects of numerical representation and cross-cut role assignment. Personality and\\nSocial Psychology Bulletin, 24, 1276–1293.\\nBlanz, M., Mummendey, A., & Otten, S. (1997). Normative evaluations and fre-\\nquency expectations regarding positive and negative outcome alloctions between\\ngroups. European Journal of Social Psychology, 27, 165–176.\\nBrewer, M. B. (1972). (Series editor). HRAFLEX Ethnocentrism Interview Series.\\nNew Haven, CN: Human Relations Area Files.\\nBrewer, M. B. (1979). In-group bias in the minimal intergroup situation: A cognitive\\nmotivational analysis. Psychological Bulletin, 86, 307–324.\\nBrewer, M. B. (1981). Ethnocentrism and its role in intergroup trust. In M. Brewer\\n& B. Collins (Eds.), Scientific inquiry in the social sciences (pp. 214–231). San Fran-\\ncisco: Jossey-Bass.\\nBrewer, M. B. (1991). The social self: On being the same and different at the same\\ntime. Personality and Social Psychology Bulletin, 17, 475–482.\\nBrewer, M. B. (1996). In-group favoritism: The subtle side of intergroup discrimi-\\nnation. In D. Messick & A. T\\nenbrunsel (Eds.), Codes of conduct: Behavioral research\\ninto business ethics (pp. 160–170). New York: Russell Sage Foundation.\\nBrewer, M. B. (1997). On the social origins of human nature. In C. McGarty & S. A.\\nHaslam (Eds.), The message of social psychology (pp. 54–62). Oxford: Blackwell.\\nBrewer, M. B. (2000). Superordinate goals versus superordinate identity as bases of\\nintergroup cooperation. In D.Capozza & R. Brown (Eds.), Social identity processes:\\nTrends in theory and research (pp. 117–132). London. Sage.\\nBrewer, M. B., & Brown, R. (1998). Intergroup relations. In D. Gilbert, S. Fiske, \\n& D. Lindzey (Eds.), The handbook of social psychology. Fourth Edition (Vol. 1, \\npp. 99–142). Boston: McGraw-Hill.\\nBrewer, M. B., & Campbell, D. T. (1976). Ethnocentrism and intergroup attitudes: East\\nAfrican evidence. Beverly Hills, CA: Sage.\\nBrewer, M. B., Manzi, J., & Shaw, J. S. (1993). In-group identification as a function\\nof depersonalization, distinctiveness, and status. Psychological Science, 4, 88–92.\\nBrewer, M. B., & Weber, J. G. (1994). Self-evaluation effects of interpersonal versus\\nintergroup social comparison. Journal of Personality and Social Psychology, 66,\\n268–275.\\nIngroup Identification and Intergroup Conflict\\n37\\n\\n\\nCaporael, L. R. (1997). The evolution of truly social cognition: The core configura-\\ntions model. Personality and Social Psychology Review, 1, 276–298.\\nCrocker, J., & Major, B. (1989). Social stigma and self-esteem: The self-protective\\nproperties of stigma. Psychological Review, 96, 608–630.\\nDeschamps, J-C., & Brown, R. (1983). Superordinate goals and intergroup conflict.\\nBritish Journal of Social Psychology, 22, 189–195.\\nDiehl, M. (1990). The minimal group paradigm: Theoretical explanations and em-\\npirical findings. In W. Stroebe & M. Hewstone (Eds.), European review of social\\npsychology (Vol. 1, pp. 263–292). Chichester, England: Wiley.\\nDoise, W., Deschamps, J-C., & Meyer, G. (1978). The accentuation of intra-\\ncategory similarities. In H. T\\najfel (Ed.), Differentiation between social groups\\n(pp. 159–170). London: Academic Press.\\nDovidio, J. F., & Gaertner, S. L. (1993). Stereotypes and evaluative intergroup \\nbias. In D. Mackie & D. Hamilton (Eds.), Affect, cognition, and stereotyping\\n(pp. 167–193). San Diego, CA: Academic Press.\\nFestinger, L. (1954). A theory of social comparison proccesses. Human Relations, 7,\\n271–282.\\nFiske, S. T., & Ruscher, J. (1993). Negative interdependence and prejudcie: Whence\\nthe affect? In D. Mackie & D. Hamilton (Eds.), Affect, cognition, and stereotyping\\n(pp. 239–268). San Diego, CA: Academic Press.\\nFrey, D., & Gaertner, S. L. (1986). Helping and the avoidance of inappropriate in-\\nterracial behavior: A strategy than can perpetuate a non-prejudiced self-image.\\nJournal of Personality and Social Psychology, 50, 1083–1090.\\nGaertner, S. L., & Dovidio, J. F. (1977). The subtlety of white racism, arousal, and\\nhelping behavior. Journal of Personality and Social Psychology, 35, 691–707.\\nGaertner, S. L., Dovidio, J. F., & Johnson, G. (1982). Race of victim, non-respon-\\nsive bystanders, and helping behavior. Journal of Social Psychology, 117, 69–77.\\nGaertner, S. L. , Dovidio, J. F., Anastasio, P., Bachman, B., & Rust, M. (1993). The\\ncommon ingroup identity model: Recategorization and the reduction of inter-\\ngroup bias. In W. Stroebe & M. Hewstone (Eds.), European review of social psy-\\nchology (Vol. 4, pp. 1–26). Chichester, England: Wiley.\\nGreenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Attitudes, self-\\nesteem, and stereotypes. Psychological Review, 102, 4–27.\\nHeider, F. (1958). The psychology of interpersonal relations. New York: Wiley.\\nHewstone, M., Fincham, F., & Jaspars, J. (1981). Social categorization and similarity\\nin intergroup behaviour: A replication with “penalties.” European Journal of \\nSocial Psychology, 11, 101–107.\\nHinkle, S., & Brown, R. (1990). Intergroup comparisons and social identity: Some\\nlinks and lacunae. In D. Abrams & M. Hogg (Eds.), Social identity theory: Con-\\nstruction and critical advances (pp. 48–70). London: Harvester Wheatsheaf.\\nHirt, E. R., Zillmann, D., Erickson, G., & Kennedy, C. (1992). Costs and benefits of\\nallegiance: Changes in fans’ self-ascribed competencies after team victory versus\\ndefeat. Journal of Personality and Social Psychology, 63, 724–738.\\nInsko, C. A., & Schopler, J. (1987). Categorization, competition and collectivity. In\\nC. Hendrick (Ed.), Personality and Social Psychology Review. Vol. 8: Group Processes\\n(pp. 213–251). Beverly Hills, CA: Sage.\\nInsko, C. A., Pinkley, R., Hoyle, R., Dalton, B., Hong, G., Slim, R., Landry, P.,\\nHolton, B., Ruffin, P., & Thibaut, J. (1987). Individual versus group discon-\\n38\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ntinuity: The role of intergroup contact. Journal of Experimental Social Psychology,\\n23, 250–267.\\nInsko, C. A., Schopler, J., Hoyle, R., Dardis, G., & Graetz, K. (1990). Individual-\\ngroup discontinuity as a function of fear and greed. Journal of Personality and \\nSocial Psychology, 58, 68–79.\\nKelley, H. H., & Stahelski, A. J. (1970). The social interaction basis of cooperators’\\nand competitors’ beliefs about others. Journal of Personality and Social Psychology,\\n16, 66–91.\\nKramer, R. M., Brewer, M. B., & Hanna, B. A. (1996). Collective trust and collec-\\ntive action: The decision to trust as a social decision. In R. Kramer & T. Tyler\\n(Eds.), Trust in organizations (pp. 357–389). Thousand Oaks, CA: Sage.\\nLevin, S., & Sidanius, J. (1999). Social dominance and social identity in the United\\nStates and Israel: Ingroup favoritism or outgroup derogation? Political Psychology,\\n20, 99–126.\\nLeVine, R. A. & Campbell, D. T. (1972). Ethnocentrism: Theories of conflict, ethnic \\nattitudes and group behavior. New York: Wiley.\\nMcClintock, C. (1972). Social motivation: A set of propositions. Behavioral Science,\\n17, 438–454.\\nMullen, B., Brown, R., & Smith, C. (1992). Ingroup bias as a function of salience,\\nrelevance, and status: An integration. European Journal of Social Psychology, 22,\\n103–122.\\nMummendey, A., & Otten, S. (1998). Positive-negative asymmetry in social dis-\\ncrimination. In W. Stroebe & M. Hewstone (Eds.), European review of social psy-\\nchology (Vol. 9, pp. 107–143). Chichester, England: Wiley.\\nMummendey, A., & Schreiber, H. (1983). Better or just different? Positive social\\nidentity by discrimination against or by differentiation from outgroups. European\\nJournal of Social Psychology, 13, 389–397.\\nMummendey A., & Schrieber, H. (1984). ‘Different’ just means ‘better’: Some obvi-\\nous and some hidden pathways to in-group favoritism. British Journal of Social\\nPsychology, 23, 363–368.\\nMummendey, A., & Simon, B. (1989). Better or different? III: The impact of impor-\\ntance of comparison dimension and relative ingroup size upon intergroup dis-\\ncrimination. British Journal of Social Psychology, 28, 1–16.\\nMummendey, A., Simon, B., Dietze, C., Grunert., M., Haeger, G., Kessler, S.,\\nLettgen, S., & Schaferhoff, S. (1992). Categorization is not enough: Intergroup\\ndiscrimination in negative outcome allocations. Journal of Experimental Social \\nPsychology, 28, 125–144.\\nNeidenthal, P. M., & Beike, D. (1997). Interrelated and isolated self-concepts. Per-\\nsonality and Social Psychology Review, 1, 106–128.\\nPerdue, C. W., Dovidio, J. F., Gurtman, M. B., & Tyler, R. B. (1990). “Us” and\\n“Them”: Social categorization and the process of intergroup bias. Journal of Per-\\nsonality and Social Psychology, 59, 475–486.\\nPettigrew, T. F., & Meertens, R. W. (1995). Subtle and blatant prejudice in Western\\nEurope. European Journal of Social Psychology, 25, 57–75.\\nPlatow, M., O’Connell, A., Shave, R., & Hanning, P. (1995). Social evaluations of\\nfair and unfair allocations in interpersonal and intergroup situations. British Jour-\\nnal of Social Psychology, 34, 363–381.\\nRogers, R. W., & Prentice-Dunn, S. (1981). Deindividuation and anger-mediated\\nIngroup Identification and Intergroup Conflict\\n39\\n\\n\\ninterracial aggression: Unmasking regressive racism. Journal of Personality and \\nSocial Psychology, 41, 63–73.\\nRubin, M., & Hewstone, M. (1998). Social identity theory’s self-esteem hypothesis:\\nA review and some suggestions for clarification. Personality and Social Psychology\\nReview, 2, 40–62.\\nSachdev, I., & Bourhis, R. (1991). Power and status differentials in minority and ma-\\njority group relations. European Journal of Social Psychology, 21, 1–24.\\nSears, D. O., & Funk, C. L. (1991). The role of self-interest in social and political\\nattitudes. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, \\npp. 2–92). New York: Academic Press.\\nSherif, M. (1966). In common predicament: Social psychology of intergroup conflict and \\ncooperation. New York: Houghton Mifflin.\\nSidanius, J. (1993). The psychology of group conflict and the dynamics of op-\\npression: A social dominance perspective. In S. Iyengar & W. McGuire (Eds.),\\nExplorations in political psychology (pp. 183–219). Durham, NC: Duke University\\nPress.\\nSmith, E. R. (1993). Social identity and social emotions: T\\noward new conceptualiza-\\ntions of prejudice. In D. Mackie & D. Hamilton (Eds.), Affect, cognition, and\\nstereotyping (pp. 297–315). San Diego, CA: Academic Press.\\nStangor, C., Sullivan, & Ford, T. E. (1991). Affective and cognitive determinants of\\nprejudice. Social Cognition, 9, 359–380.\\nStephan, W., & Stephan, C. (2000). An integrated threat theory of prejudice. In \\nS. Oskamp (Ed.), Reducing prejudice and discrimination (pp. 23–45). Mahwah, NJ:\\nLawrence Erlbaum.\\nStruch, N., & Schwartz, S. H. (1989). Intergroup aggression: Its predictors and \\ndistinctness from in-group bias. Journal of Personality and Social Psychology, 56,\\n364–373.\\nSumner, W. G. (1906). Folkways. New York: Ginn.\\nT\\najfel, H. (1969). Cognitive aspects of prejudice. Journal of Social Issues, 25, 79–97.\\nT\\najfel, H. (1970). Experiments in intergroup discrimination. Scientific American,\\n223(2), 96–102.\\nT\\najfel, H. (1978). Differentiation between social groups: Studies in the social psychology of\\nintergroup relations. London: Academic Press.\\nT\\najfel, H. (1981). Human groups and social categories. Cambridge: Cambridge Univer-\\nsity Press.\\nT\\najfel, H., Billig, M., Bundy, R., & Flament, C. (1971). Social categorization and \\nintergroup behaviour. European Journal of Social Psychology, 1, 149–178.\\nT\\najfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior.\\nIn S. Worchel & W. Austin (Eds.), Psychology of intergroup relations (pp. 7-24).\\nChicago: Nelson-Hall.\\nT\\najfel, H., & Wilkes (1963). Classification and quantitative judgement. British Jour-\\nnal of Psychology, 54, 101–114.\\nT\\nakagi, E. (1996). The generalized exchange perspective on the evolution of altru-\\nism. In W. Liebrand & D. Messick (Eds.), Frontiers in social dilemma research\\n(pp. 311–336). Berlin: Springer-Verlag.\\nT\\nesser, A. (1988). T\\noward a self-evaluation maintenance model of social behavior. \\nIn L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 181–\\n227). San Diego, CA: Academic Press.\\nTriandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press.\\n40\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nTurner, J. C. (1975). Social comparison and social identity: Some prospects for \\nintergroup behaviour. European Journal of Social Psychology, 5, 5–34.\\nTurner, J. C. (1978). Social categorization and social discrimination in the minimal\\ngroup paradigm. In H. T\\najfel (Ed.), Differentiation between social groups (pp. 101–\\n140). London: Academic Press.\\nTurner, J. C. , Hogg, M., Turner, P\\n., & Smith, P\\n. (1984). Failure and defeat as deter-\\nminants of group cohesiveness. British Journal of Social Psychology, 23, 97–111.\\nTurner, J. C. , Hogg, M., Oakes, P., Reicher, S., & Wetherell, M. (1987). Rediscover-\\ning the social group: A self-categorization theory. Oxford: Blackwell.\\nWeber, J. G. (1994). The nature of ethnocentric attribution bias: Ingroup protection\\nor enhancement? Journal of Experimental Social Psychology, 30, 482–504.\\nWood, J. V. (1989). Theory and research concerning social comparisons of personal\\nattributes. Psychological Bulletin, 106, 231–248.\\nIngroup Identification and Intergroup Conflict\\n41\\n\\n\\nThomas Hylland Eriksen\\n3\\nEthnic Identity, National Identity,\\nand Intergroup Conflict\\nThe Significance of Personal Experiences\\nThere exists by now an enormous literature on contemporary ethnicity and\\nnationalism in its various forms. It deals with a wide range of phenomena—\\nNorth American multiculturalism and indigenous rights movements, post-\\nSoviet ethnonationalism in Central and Eastern Europe, urban minority\\ndilemmas and Islamic revivalism in Western Europe, indigenista movements\\nin Latin America, and processes of political fission and fusion in contempo-\\nrary Africa. The analytical focus of this chapter will be on the concept of\\nidentity. The naïve question to be asked at the outset is: what is it about\\nidentity politics that makes it such a formidable force in the contemporary\\nworld? Some important issues will have to be omitted—notably the rela-\\ntionships among identity politics, globalization, and reflexive modernity\\n(e.g., Bauman, 1993; Friedman, 1994; Giddens, 1994). Instead, what is of-\\nfered amounts to an anthropological perspective on the relationship be-\\ntween personal identity and political identity, using conflicts based on iden-\\ntity politics as empirical examples. \\nThis chapter sets out to do three things. First, a brief overview of the\\nstandard social anthropological perspective on the politics of identity is pro-\\nvided. Identity politics should be taken to mean political ideology, organiza-\\ntion, and action that openly represents the interests of designated groups\\nbased on “essential” characteristics such as ethnic origin or religion, and\\nwhose legitimacy lies in the support of important segments of such groups.\\nMembership in such groups is generally ascribed, unlike membership in\\nother political groups (socialists, liberals, trade unions, etc.). Second, by way\\nof examples from India, Fiji, and Yugoslavia, parallels and differences be-\\ntween some such conflicts are highlighted. The examples are chosen mainly\\n42\\n\\n\\nfor their differences: what they have in common is, apart from their tempo-\\nral location in the 1990s, apparently only that they are based on ideologies\\nof culture and identity. The Yugoslav conflict has been extremely brutal and\\ntragic; the Fijian one has involved only a few casualties, but has led to im-\\nportant constitutional changes; while the Indian conflict, although occa-\\nsionally bursting into violence, is largely contained within the framework of\\ninstitutional Indian politics. If it can persuasively be argued that these con-\\nflicts have important features in common, it is likely that those features will\\nalso be present in other settings where identity politics has a major impact.\\nThe third and final part of the chapter draws on classic political anthropol-\\nogy in order to suggest some general features of the relationship between\\npersonal identity and politics.\\nThe Anthropology of Identity\\nThe last decades of the 20th century saw a dramatic reconceptualization \\nof core concepts, including culture and society, within the social sciences.\\nUntil the 1960s, the close overlap between culture and ethnicity, or even\\nculture and nationhood, was generally taken for granted in the scholarly\\ncommunity. During the past 30 years, however, hardly a single serious con-\\ntribution to the field has failed to point out that there is no one-to-one rela-\\ntionship between culture and ethnicity (the seminal text here is Barth,\\n1969); that cultural differences cut across ethnic boundaries; and that ethnic\\nidentity is based on socially sanctioned notions of cultural differences, not\\n“real” ones. While ethnic identity should be taken to refer to a notion of\\nshared ancestry (a kind of fictive kinship), culture refers to shared represen-\\ntations, norms, and practices. One can have deep ethnic differences without\\ncorrespondingly important cultural differences (as in the Bosnian example\\nbelow); and one can have cultural variation without ethnic boundaries (as,\\nfor example, between the English middle class and the English working\\nclass).\\nSeveral recent debates in anthropology and neighboring disciplines pull\\nin the same direction: away from notions of integrated societies or cultures\\ntoward a vision of a more fragmented, paradoxical, and ambiguous world.\\nThe currently bustling academic industry around the notion of globaliza-\\ntion (see Featherstone, 1990, for an early, influential contribution) repre-\\nsents an empirically oriented take on these issues, focusing on the largely\\ntechnology-driven processes that contribute to increasing contact across\\nboundaries and diminished importance of space. This focus on unbounded\\nprocesses rather than isolated communities has contributed to a reconcep-\\ntualization of the social, which is radically opposed to that of classic Durk-\\nheimian sociology and anthropology; where flux, movement, and change\\nbecome the rule and not the exception in social life (Strathern, 1991; Han-\\nnerz, 1992; Lash & Urry, 1994). \\nEthnic Identity, National Identity, and Intergroup Conflict\\n43\\n\\n\\nParticularly in North America, the classic concept of culture has been\\nused for what it is worth in domestic identity politics, leading in some cases\\nto controversial policies of multiculturalism, where individuals have been\\nendowed with special rights in accordance with their ethnic origins. Critics\\nmight point out that multiculturalism in some of its versions resembles\\napartheid; also, that by positing a simple one-to-one relationship between\\nethnic origin and culture, it not only encouraged a “disuniting of America”\\n(Schlesinger, 1992) but also contributed to reifying misleading notions of\\nculture seen as the commonalities of a bounded set of individuals, like so\\nmany nationalisms writ small. \\nA further disruptive tendency has been the so-called postcolonial move-\\nment in literary studies, spilling into anthropology and other disciplines,\\nthat has raised the question of who has the right to identify whom; a stan-\\ndard text in this field of discourse is Edward Said’s Orientalism (1978), al-\\nthough Frantz Fanon developed similar insights two decades earlier. Said\\nand others argued, briefly, that ethnocentrism was deeply embedded in\\nWestern scholarship dealing with non-Western peoples. Postcolonial critics\\nalso tend to call attention to the multiplicity of voices (an academic cliché\\nby the late 1990s) present in any society and the general unwillingness of\\nacademic researchers to give all of them the attention they deserve. \\nT\\nwo related debates defined the field for many years. First, there was the\\ncontroversy over primordialism and instrumentalism. Was ethnic identity\\n“primordial”—that is, profoundly rooted in, and generative of, collective\\nexperiences; or did it arise as an ad hoc supplement to political strategies?\\nAn early, powerful defense of the instrumentalist view was represented in\\nAbner Cohen’s work on urban ethnicity in Africa (1969, 1974), showing the\\nconscious manipulation of kinship and cultural symbols by political entre-\\npreneurs seeking political gain. This perspective is still used with consider-\\nable success in studies of identity politics. Who, then, were (or are) the pri-\\nmordialists? Clifford Geertz is often associated with this view, arguing as \\nhe does along hermeneutic lines that cultural systems are more or less self-\\nsustaining and are thus not subject to the willful manipulation of individuals\\n(Geertz, 1973), a perspective he retained when writing about nationhood \\nin the Third World (Geertz, 1967). Typically, however, ethnicity studies\\nwere—and are—instrumentalist in their basic orientation (Rex, 1997). \\nThe second debate, usually framed as the opposition between construc-\\ntivism and essentialism, concerns the question of whether ethnic or national\\ncommunities are created more or less consciously, or whether they grow or-\\nganically, as it were, out of preexisting cultural communities. In nationalism\\nstudies the most highly profiled antagonists regarding this issue have been\\nthe late Ernest Gellner (1983, 1997) and Anthony D. Smith. Smith (1986,\\n1991) has developed an intermediate position in arguing the importance of\\npreexisting ethnies for the development of nationalism while acknowledging\\nits essential modernity. Gellner, on the other hand, champions the view that\\nnations are entirely modern creations—the progeny of industrialism and\\n44\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nthe state—that more or less fraudulently invent their past to gain a sem-\\nblance of antiquity and deep roots; his final statement on the issue is repro-\\nduced in the posthumously published Nationalism (Gellner, 1997) under the\\nheading “Do Nations Have Navels?”, a pun on the biblical enigma relating\\nto Adam’s navel. Regarding definitions, there are also important differences\\nbetween theorists. While Gellner holds nations to be ethnic groups who ei-\\nther control a state or who have leaders who wish to do so, Anderson (1983)\\nsees no necessary link between the abstract “imagined community” of the\\nnation and particular ethnic groups; indeed, several of the main examples in\\nhis famous book on “imagined communities,” including the Philippines and\\nIndonesia, are multiethnic countries. Yet others have distinguished between\\nethnic nations and “civic” ones (Smith, 1991). There is nonetheless general\\nagreement that nations are by definition linked with states, whether they\\nare based on a common ethnic identity or not.\\nIn anthropology at least, the recent shift toward the study of identities\\nrather than cultures has entailed an intense focus on conscious agency and\\nreflexivity; and for many anthropologists, essentialism and primordialism\\nappear as dated as pre-Darwinian biology. In addition, there seems to be\\ngood political sense in discarding the old, static view of culture, which is be-\\ning used for many political purposes that are difficult to endorse by acade-\\nmics committed to democratic values, ranging from the Balkan war to dis-\\ncrimination against ethnic minorities in Western Europe. Further, this is an\\nage when the informants talk back. It could, perhaps, be said that a main\\npurpose of an earlier anthropology consisted of identifying other cultures.\\nRepresentatives of these so-called other cultures are now perfectly able to\\nidentify themselves, which leaves the scholars either out of a job or with a\\nnew mission—that is, to identify their identifications—in other words, to\\nstudy reflexive identity politics. \\nNot so many years ago, anthropology was still a discipline fueled by a\\nprogrammatic love of cultural variation for its own sake, and anthropolo-\\ngists involved in advocacy tended to defend indigenous peoples’ or other\\nminorities’ traditional ways of life against the onslaught of modernity. A\\nmain tendency in recent years has, on the contrary, consisted of decon-\\nstructing instrumentalist uses of notions of authenticity and tradition, and\\nshowing not only that the internal variation within a group is much greater\\nthan one would expect but also that traditionalist ideologies are, paradoxi-\\ncally, direct results of modernization (e.g., Roosens, 1989). This theoretical\\nshift is a very significant one. It offers a method for investigating extremely\\nwell the strategic action, politics of symbols, and contemporary processes of\\nidentity politics within a uniform comparative framework. \\nIn other words, cracks in the edifice of mainstream social and cultural\\nanthropology, some of them directly inspired by events well beyond the\\nconfines of academia, have led to a widespread reconceptualization of soci-\\nety and culture. Reification and essentialism have become central terms of\\ndenunciation; multiple voices, situational identification, and cultural flows\\nEthnic Identity, National Identity, and Intergroup Conflict\\n45\\n\\n\\nare some of the key words delineating the current intellectual agenda. It has\\nbecome difficult if not impossible to talk of, say, Nuer culture, Hopi cul-\\nture, Dutch culture, and so on, since such terms immediately invite critical\\nquestions of whose Nuer culture, Hopi culture, and so on, intimating that\\nthere are an infinite number of versions of each culture, none of which is\\nmore “true” than the others (Holland, 1997). Ethnicity and nationalism,\\nthen, become the political reifications or constructions of a particular au-\\nthorized version of a culture, freezing that which naturally flows, erecting\\nartificial boundaries where they did not exist before, trimming and shaping\\nthe past to fit present needs, and inventing traditions where no organic tra-\\nditions exist, or are not adequate, to ensure a sense of continuity with the\\npast. \\nA new kind of political responsibility has entered academia in acute ways\\nduring the last decades. Academic or semiacademic statements about na-\\ntions, ethnic groups, or cultures may now immediately be picked up, or as-\\nsimilated more or less subconsciously, by ideologists and politicians wishing\\nto build their reputation on national chauvinism, ethnic antagonism, enemy\\nimages, and so on. The liberal academic establishment thus wags a warning\\nfinger at those who dare to talk of culture as the cause of conflicts, shaking\\ntheir heads sadly over those lost souls who have not yet heeded the words of\\nleading theorists such as Barth (1969) and Gellner (1983), criticizing those\\nwho do not realize that culture is chimerical and fleeting, and that reified\\nculture is a dangerous tool. It is, thus, not only intellectually correct but\\nalso politically correct to reject all forms of essentialism. \\nThe current scholarly orthodoxy on ethnicity and the politics of identity\\ncan be summed up as follows:\\n• Although ethnicity is widely believed to express cultural differ-\\nences, there is a variable and complex relationship between eth-\\nnicity and culture; and there is certainly no one-to-one relationship\\nbetween ethnic differences and cultural ones. \\n• Ethnicity is a property of a relationship between two or several\\ngroups, not a property of a group; it exists between and not within\\ngroups. \\n• Ethnicity is the enduring and systematic communication of cultural\\ndifferences between groups considering themselves to be distinct. It\\nappears whenever cultural differences are made relevant in social\\ninteraction, and it should thus be studied at the level of social life,\\nnot at the level of symbolic culture. \\n• Ethnicity is thus relational and also situational: the ethnic character\\nof a social encounter is contingent on the situation. It is not, in\\nother words, inherent.\\nThis instrumentalist framing of ethnicity, which may appear simply as a\\nset of methodological guidelines, has firm, although usually untheorized,\\nphilosophical foundations and is, as I have tried to show elsewhere (Eriksen,\\n46\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\n1998b), deeply embedded in empiricist thought. I shall argue that this ap-\\nproach, notwithstanding its strengths, is limited in overemphasising choice\\nand strategy (instrumental aspects) when analyzing identity politics. As a re-\\nsult, the self is taken for granted (A.P. Cohen, 1994), and it is therefore not\\nshown how it can be possible to mobilize particular aspects of personal\\nidentity for antagonistic identity politics. Yet, anthropology as a discipline is\\nin a privileged situation to study the dynamics of identity politics, precisely\\nbecause of its focus on the ongoing flow of social interaction.\\nAlthough an enormous amount of anthropological research has been\\ncarried out on ethnicity and nationalism since around 1970, surprisingly\\nfew studies have dealt with violent conflicts and conflict resolution (see,\\nhowever, T\\nambiah, 1994; Turton, 1997). The dominant approaches to eth-\\nnicity have been instrumentalist (with a focus on politics) or constructivist\\n(with a focus on ideology), and research questions have concentrated on the\\nestablishment and reproduction of ethnically incorporated groups, not on\\nthe circumstances under which ethnicity may become politically less impor-\\ntant. While my examples (below) and the ensuing discussion will indicate\\nthe fruitfulness of these approaches, it is also necessary to point out the\\nneed for a phenomenological understanding of social identity, which sees it\\nas emerging from experiences, not as a mere construct of ideology. In this, I\\nfollow scholars such as A. P. Cohen (1994) and Jenkins (1996), who have\\ncalled for an anthropology of identity that does not concentrate exclusively\\non its political and ideological aspects, but also strives to understand the\\nself. \\nAs noted by Holland (1997), anthropologists are generally associated\\nwith a culturalist view of the self, arguing its cultural specificity as against\\npsychologists, who have been more prone to a universalist view—the self as\\nsomething proper not to particular cultures but to humanity, with universal\\ncharacteristics lurking below a thin veneer of culture. For the purposes of\\nthe present argument, it is not necessary to take a stance on this contro-\\nversy, partly since it can be presupposed that modernity creates a particular\\nkind of selves with important shared characteristics everywhere (Giddens,\\n1991), but also because this chapter restricts its scope to the relationships\\namong personal experiences, ideology, and political mobilization. As the \\nexamples will, it is hoped, show, the similarities are more striking than the\\ndifferences here.\\nCulture and the Breakup of Yugoslavia\\nNo other recent ethnic conflict has been more intensely studied, discussed,\\nand moralized over than the breakup of Yugoslavia in 1991. This was \\nfollowed by three major wars and a number of smaller skirmishes, and the\\nsituation in many parts of ex-Yugoslavia remains unstable and tense. In \\nEurope, the outbreak of war in Yugoslavia has been interpreted by hun-\\nEthnic Identity, National Identity, and Intergroup Conflict\\n47\\n\\n\\ndreds of commentators. In the press, it was occasionally argued, along \\nsociobiological or Hobbesian lines, that humans are driven by aggressive\\ninstincts that emerge when the social fabric falls apart, in this case orga-\\nnized as factions ultimately based on kinship. Another view, popular in the\\nEuropean nationalist right, implied that ethnic conflict was inevitable when\\ndifferent groups are forcibly integrated into one state. Most scholars have,\\non the contrary, tended to focus on the cultural logic of feuding in Balkan\\nsociety, the deep economic crisis underlying the conflict, or the rise of \\nSerbian supremacy during the 1980s, looking for contradictions within Yu-\\ngoslav society rather than into human nature for an explanation of the con-\\nflict. A widespread view nevertheless sees the cultural differences between\\nthe constituent groups as a basic cause of the conflict (cf., the influential\\nanalysis by Ignatieff, 1994, or Huntington’s controversial model, 1996). \\nYugoslavia was a state that came into being twice: after the First World\\nWar and after the Second World War. The first Yugoslav republic (1918–\\n41) was for all practical purposes the Serb monarchy writ large, and was rid-\\ndled with continuous internal tension. Croats and Slovenes reluctantly \\nsupported King Alexander’s regime, seeing it as a possible defense against\\nItalian, Austrian, and Hungarian aggressors. It was a precarious state, peri-\\nodically dictatorial, that had been on the verge of collapse several times \\nbefore it fell apart as Germany invaded the country in 1941. While Serbs\\ngenerally resisted the Germans, Croats collaborated and saw the German\\nintervention as an opportunity to create their own state. \\nBefore the formation of the second Yugoslav republic in 1945, both of\\nthe largest constituent groups suffered large-scale massacres—Serbs at the\\nhands of Croatian fascists (Ustasa) in 1941 and Croats killed by Serbian\\ncommunists (Partisans) in 1945. The new Yugoslav state (1945–91) was a\\nnonaligned socialist federation led by a pro-Serbian Croat, Josip Broz Tito,\\nuntil his death in 1980. Ethnicity was officially declared a nonissue in so-\\ncialist Yugoslavia. This official blindness stems from the Marxist view that\\nclass is a more objective and more authentic vessel of social identity than\\nethnicity or nationality, which were officially seen as expressions of false\\nconsciousness. This does not, however, imply that Yugoslav policies were\\nparticularly repressive regarding expressions of cultural distinctiveness sym-\\nbolizing ethnic identity; on the contrary, in this area Yugoslavia was more\\nliberal than many Western European countries whose leadership feared\\nseparatism and social fission. Ethnic identity was seen as politically irrele-\\nvant, and partly for this reason, the use of various languages and the prac-\\ntice of different religions were tolerated in civil society. It is true that the\\nmerging of Serbian and Croatian into one language, Serbo-Croatian, in \\nthe 1950s signaled an attempt at building a unitary Yugoslav identity, but\\nthe two languages were so closely related that few, except Croat intellectu-\\nals reacting against the relegation of specific Croat variants as “dialect,”\\nseem to have taken offense (Schöpflin, 1993). Albanian remained an official\\nlanguage in Kosovo, as did Slovene in Slovenia. \\n48\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nThere were nearly twice as many Serbs as Croats in Yugoslavia, and\\ntwice as many Croats as Slovenes or Muslims. Regarding the territorial \\ndimension, Slovenes were—and are—largely confined to the nearly mo-\\nnoethnic republic of Slovenia. Both Croatia and Serbia had large minorities\\nof Serbs and Croats, respectively, as well as Gypsies, while Serbia also in-\\ncluded nearly a million Albanians (in Kosovo) and smaller numbers of Hun-\\ngarians and others. In Macedonia, most of the population were (and are)\\nSlav-speaking Macedonians, while the Montenegrins of Montenegro are\\nculturally close to the Serbs. Bosnia-Herzegovina was the most throroughly\\nmixed republic, with roughly equal numbers of Serbs, Croats, and Bosnian\\nMuslims, often living in mixed areas. As is well known, the conflicts of the\\n1990s have modified this picture somewhat, creating large monoethnic \\nterritories in formerly mixed areas.\\nAlthough the ruling Communist party seems to have believed that a\\ncommon Yugoslav identity would eventually supersede the national identi-\\nties based on ethnic membership, ethnic identity remained strong in most\\nparts of the country throughout the postwar era. There were nevertheless\\nimportant exceptions, particularly in cities such as Belgrade, Sarajevo, and\\nZagreb, where many people increasingly identified themselves primarily as\\nYugoslavs and where mixed marriages were common.\\nEthnic identities did, in other words, not disappear during the existence\\nof Yugoslavia. In some urban areas they were arguably weakened, but it\\ncould be—and has been—argued that the nonethnic character of Yugoslav\\npolitics actually led to its strengthening as a vehicle for the political opposi-\\ntion and made it possible for Serbs to gain control over the armed forces\\nand state bureaucracy: since political ethnicity officially did not exist (only\\ncultural ethnicity did), there were no institutionalized ways of preventing\\none group from dominating the public sector. \\nThe wars in the former Yugoslavia have bequeathed to the world the \\nneologism ethnic cleansing. It is nevertheless easy to show that the conflicts\\ninvolving Serbs, Croats, Bosnian Muslims, Slovenes, and Albanians were\\nnever conflicts over the right to assert one’s ethnic or cultural identity, but\\nwere based on competing claims to rights such as employment, welfare, and\\npolitical influence. What needs to be explained is the fact that the conflicts\\nover these resources were framed in ethnic terms rather than being seen as,\\nsay, regional, class-based, or even ideological. \\nThe relevant questions are therefore: what is the stuff of ethnic identity\\nin the former Yugoslavia; in which ways do the groups differ from one an-\\nother, and why did group allegiances turn out to be so strong? Bosnia-\\nHerzegovina may be considered as an example. There are three large ethnic\\ngroups inhabiting Bosnia: Serbs, Croats, and Muslims. The main difference\\nbetween the groups is religion—Serbs are Orthodox and Croats are\\nCatholics. (Ironically, religious fervor was not particularly widespread in\\nprewar Bosnia.) They all have common origins: Slav immigration into \\nIllyria from the north took place between c. a.d. 400 and 700, and the cul-\\nEthnic Identity, National Identity, and Intergroup Conflict\\n49\\n\\n\\ntural differences between Croats and Serbs are perhaps comparable to those\\nbetween Norwegians and Swedes. The “objective” difference between\\nBosnian Christians and Bosnian Muslims, further, has been compared to\\nthe difference between English Protestants and English Catholics (Cornell\\n& Hartmann, 1998). Unlike the impression sometimes given in Serbian and\\nCroatian propaganda, Bosnian Muslims are not the descendants of alien \\ninvaders, but of locally residing converts. Although each group has its nu-\\nmerical stronghold, many Bosnian regions and villages were mixed before\\nthe war. This implies, among other things, that they went to school to-\\ngether, worked together, and took part in various leisure activities together.\\nA Serbian villager in Bosnia had more in common, culturally speaking, with\\na Muslim co-villager than with a Serb from Belgrade. This would hold true\\nof both dialect spoken and way of life in general. However, since religion\\nturned out to be the central marker of collective identity in the Bosnian\\nconflict, the effective boundary was drawn not between villagers and city-\\ndwellers, but between religious categories.\\nThe boundaries between the groups may seem arbitrary. However, \\nthe large, “national” groups are clearly embedded in smaller, local networks\\nbased on kinship and informal interaction, as well as being culturally\\nfounded in religious schisms, collective myths or memories of treason and\\nresistance under Ottoman rule, massacres, deception, and humiliations. \\nAlthough it is tempting to argue that any so-called cultural trait can be \\nexploited in the formation of national or ethnic groups, it is obvious that\\nnot just anything will do. Nothing comes out of nothing, and strong collec-\\ntive identities—such as the ones revealed during the war in Bosnia—are \\nalways embedded in personal experiences. In one of the most detailed \\naccounts of ethnicity at the village level in prewar Bosnia, Bringa (1996)\\nshows that although cultural differences between the groups were perhaps\\nnegligible, and although relations among Serbs, Muslims, and Croats might\\nbe cordial at the local level, there were nevertheless important social prac-\\ntices of affiliation that created boundaries between them—not in the cos-\\nmopolitan Sarajevo middle class embraced by Western commentators, per-\\nhaps, but elsewhere. Intermarriage was rare, the close informal networks \\nof friends tended to be monoethnic, and the discrete groups maintained \\ndifferent, sometimes conflicting myths of origin. The intimate sphere, in\\nother words, seems to have been largely monoethnic and by this token,\\nBosnia was a plural society in the classic sense (Furnivall, 1948); the public\\narenas were shared, but the private ones were discrete. \\nOne may choose not to speak of such features of social reality and every-\\nday life in terms of “culture,” but they are no more “invented” than any\\nother social fact. People do not choose their relatives, they cannot choose to\\ndo away with their childhood and everything they learned at a tender age.\\nThese are aspects of identity that are not chosen, that are incorporated and\\nimplicit. People relate to them as reflexive agents, but they do so within\\nlimitations that are not chosen. Such limitations form the objective founda-\\n50\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ntions of social identification. When analysts such as Cornell and Hartmann\\n(1998) argue that ethnic identities before the war were weak, sometimes so-\\ncially irrelevant, and in many cases ambiguous (many had parents from dif-\\nferent groups), they refer to particular segments of society—a main example\\nis the Yugoslav basketball team—whose members were active participants in\\npan-Yugoslav contexts. \\nThe maintenance of ethnic boundaries in socialization and the private\\nsphere reveals a main cause of the failure of Yugoslav social engineering in\\ndoing away with ethnic identification. It does not explain the outbreak of\\nwar in the early 1990s, but it indicates why the groups that emerged were \\nso strong, and why they were based on ethnicity (seen as fictive kinship) \\ninstead of, say, class or region. Their foundation must be sought not in the\\nbiology of kinship, as some might want to argue, but in the phenomenology\\nof social experience, the raw material of personal identity. This argument\\nwill be elaborated after an examination of two very different examples of \\nintergroup conflict in polyethnic societies.\\nThe Fijian Coup-D’etat\\nThe Pacific island-state of Fiji, located at the ethnographic crossroads of\\nMelanesia and Polynesia, is perhaps less heterogenous than Yugoslavia, but\\nit is scarcely less ethnically divided. Its population of about 800,000 is\\nlargely composed of two ethnic categories: Fijians and Indians. The Fijians\\nare indigenous, largely Christian, speak a Polynesian language, and make\\nup slightly less than half the population. The Indians are uprooted “over-\\nseas Indians” whose ancestors were brought to Fiji during colonialism un-\\nder the British indentureship system described, probably a trifle too grimly,\\nas “a new system of slavery” by Tinker (1974). They are overwhelmingly\\nHindus (with a Muslim minority), speak a locally modified dialect of Hindi,\\nand were slightly more numerous than the indigenous Fijians until the po-\\nlitical changes in the late 1980s leading to mass migration of Indo-Fijians.\\nThe small minorities of Europeans and Chinese are politically insignificant,\\nbut the economically powerful Europeans, representing the former colonial\\nregime, have in no small measure shaped the Fijian public sphere, notably\\nthrough establishing English as the national lingua franca.\\nThe relatively brief history of democratic Fijian politics up to 1987 has\\nbeen described with the metaphors of balance and power sharing (Premdas,\\n1993; Kelly, 1998). While Indians in practice wielded disproportionate eco-\\nnomic power, it was tacitly agreed that Fijians should be politically para-\\nmount. However, there were indications that the “equilibrium” model was\\nunder severe stress in the early- to mid-1980s, and in the 1987 elections, a\\ncoalition supported by most Indians and only a few Fijians won. Although\\nthe new prime minister was a Fijian, many saw his government as a vehicle\\nfor Indian communal politics. In May and September 1987, the military\\nEthnic Identity, National Identity, and Intergroup Conflict\\n51\\n\\n\\nseized power through two successive coups-d’etat, explicitly doing so to\\nprotect “native” Fijian interests. In the period following the coups, thou-\\nsands of Indians emigrated. \\nThe tension between Fijians and Indians had been evident throughout\\nFijian history. Unlike the situation in Yugoslavia and especially Bosnia, \\nnobody would question the view that there are deep cultural differences \\nbetween Indians and Fijians. Their languages, cultural traditions, religions,\\nand gender relations differ markedly—indeed, one of their recent ethnogra-\\nphers has poetically described striking, and culturally potent, differences \\nin body language between Fijians and Indians (Williksen-Bakker, 1991).\\nThere seems to be little informal interaction between the groups, most\\nrural areas are dominated by one or the other, and intermarriage has always\\nbeen nearly nonexistent. The effective separation of Fijians from Indians\\nhas always been much deeper than that obtaining between the major groups\\n(or “nationalities”) in rural Bosnia after the Second World War. Until 1987,\\npolicies of compromise had nonetheless ensured political stability and had\\nmade Fiji a remarkably liberal and relaxed society.\\nInvoking the slogan “Fiji for Fijians” (which had been launched before\\nthe 1977 elections), the writers of the new constitution, promulgated in\\n1990, ensured continued indigenous Fijian dominance of the political\\nsphere, by according that group disproportionate representation in parlia-\\nment, ruling that only Fijians can become prime ministers, and giving \\nFijians preferential treatment in other areas as well, such as religion. In ad-\\ndition, Fijians are guaranteed control over most of the arable land, about 82\\npercent of which has been communally owned by Fijian kin groups since\\nthe beginning of colonialism in the 19th century. \\nThe conflict culminating in the military coups can, at one level of analy-\\nsis, be seen as a clear case of group competition. Indians have done better\\neconomically than Fijians. Ironically, this may partly be explained through\\nthe British colonial policy of indirect rule relating to Fijians, who were al-\\nlowed to retain important traditional institutions, such as chieftainship and\\nthe rudiments of a caste system, making them in consequence unprepared\\nto compete with Europeans and Indians in a capitalist economy later. The\\nIndians, by contrast, had shed important aspects of their traditional social\\norganization in the process of migration, and were accustomed to economic\\nindividualism through the indentureship system whereby they were made\\nto work on the European-owned plantations. \\nThe demographic growth rate among Indians has been higher than\\namong Fijians (just as the Muslims in Bosnia were more prolific than the\\nother groups), and there was generally a growing sentiment among Fijian\\nleaders that they were becoming a minority in the land of their ancestors.\\nThe inegalitarian measures introduced by the military regime, discriminat-\\ning between categories of citizens on ethnic grounds, were condemned by\\nthe international community, but less strongly than one might have ex-\\npected in a different setting. What is remarkable about the Fijian case is the\\n52\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nnativist quality of the supremacist rhetoric—how they brought “sons-of-\\nthe-soil” arguments to bear on a national legislation, creating, in effect, a\\ntwo-tier society where non-Fijians were relegated to the status of second-\\nclass citizens. They argued that their culture, like that of Maoris in New\\nZealand and Aborigines in Australia, was threatened with marginalization\\nfrom outside forces. In contrast, nobody in countries like Mauritius,\\nTrinidad and T\\nobago, and Guyana—which are in many ways similar to Fiji\\nbut lack substantial indigenous populations—would have been able to in-\\nvoke arguments of cultural authenticity and preservation of traditional cul-\\ntures in a bid to introduce differential treatment for different ethnic groups\\n(see Eriksen, 1992, 1998a, for details on Trinidad and Mauritius).\\nIt may well be asked whether contemporary Fijians, being Methodists and\\nproficient English speakers, are any less culturally uprooted than, say, Trini-\\ndadians of African descent or, for that matter, Fijians of Indian descent. The\\nquestion of cultural authenticity is outside the scope of this chapter; let me\\nnow make a few more pertinent points relating to this example.\\nFiji had developed an informal formula for interethnic accommodation\\nwhere the largest ethnic groups divided societal sectors between them. In\\naddition there were—and still are, at least to some extent—developing\\nfields of shared meaning and cross-cutting alignments, such as the common\\nuse of English as a national language and a shared educational system. Nev-\\nertheless the segregation between the groups in both social and cultural do-\\nmains is more striking to the outside observer than tendencies toward as-\\nsimilation. Conflict avoidance would thus have to rely chiefly on group\\ncompromise rather than the development of a hybridized, shared identity—\\nan option that has occasionally been proposed by politicians and intellectu-\\nals in other insular, postcolonial plantation societies such as Mauritius and\\nTrinidad. Finally, owing to historical circumstances and cultural differences,\\nIndians and Fijians have participated in different ways and have succeeded\\nto varying degrees in the modern sectors of politics and the economy. It\\ncould indeed be argued that processes of modernization in Fiji, far from \\nreducing cultural differences, have deepened them, at least at the socially\\noperational level. Unlike in Bosnia, it is possible to refer to differences in\\nlocal organization, cosmology, and traditional economical practices when\\naccounting for the ethnic conflict in Fiji, which is nevertheless much less \\nviolent than the Bosnian one. This is a reminder, against cultural determin-\\nists à la Huntington (1996), of the relative unimportance of cultural differ-\\nences for ethnic conflict—and it also indicates one of the main strengths of\\nthe constructivist-instrumentalist perspective on identity politics: cultural\\ndifferences do not in themselves lead to intergroup conflict, but are invoked\\nstrategically to mobilize support. At the same time, it must be conceded\\nthat the differences in life-worlds and personal identities in Fiji, as in\\nBosnia, explain why the political cleavages were given ethnic expression.\\nThe differences were already there before they were exploited for particular\\npolitical ends.\\nEthnic Identity, National Identity, and Intergroup Conflict\\n53\\n\\n\\nHindutva: An Apparent Anomaly in Contemporary Indian Society\\nMy third and final example differs from the two previous ones in significant\\nways. India is a tough case for any scholar trying to develop a general theory\\nof ethnicity or nationalism, and with few exceptions, it does not figure in\\ngeneral introductory texts on the field. India is hardly a state based on cul-\\ntural similarity or even equality in the Western sense; it is a country with\\ndeeply embedded hierarchies and a very considerable degree of internal cul-\\ntural variation. Its population of nearly a billion is divided by language, reli-\\ngion, caste, and culture, and it has often been argued that India is culturally\\nmore complex than continents such as sub-Saharan Africa or Europe. Al-\\nthough 80 percent of the population are Hindus in one meaning of the\\nword or another, India also has the second-largest Muslim population in the\\nworld (after Indonesia) and more Christians than all the Scandinavian\\ncountries put together. Since independence (and partition) in 1947, India\\nhas been defined in Gandhian-Nehruvian terms as a secular, federal country\\nusing English and Hindi as national languages, but with another dozen or\\nso official regional languages.\\nSince the early 1980s—but particularly forcefully during the 1990s—a\\nformerly marginal political movement has steadily increased its influence in\\nIndia, culminating in its victories in the successive general elections of 1998\\nand 1999. This is the movement often referred to as hindutva, meaning\\nroughly “Hindu-ness,” which rallies behind slogans to the effect that India\\nshould be redefined as a Hindu country. The hindutva movement, led by an\\norganization called the RSS (Rashtriya Swayamsevak Sangh) began mod-\\nestly in the interwar years, and its more recent parliamentary wing, the BJP\\n(Bharatiya Janata Party, “The Indian People’s Party”) is now in power not\\nonly federally, in New Delhi, but in several of the states as well.\\nThe rhetoric of hindutva is strongly reminiscent of European ethnic na-\\ntionalism. It invokes ancient myths of bitter defeats and noble sacrifices, re-\\nframing them to fit a contemporary political scene. It quotes liberally from\\n19th-century poets and sacred texts, and it redefines history to make the\\npast conform to a redefined present. It advocates a return to the roots, con-\\ndemns Westernization and its adverse moral effects on the young, praises\\nthe family as the key institution of society, and seeks to promote the vision\\nof India as a hindu rashtra—a Hindu nation. While the late Rajiv Gandhi al-\\nlowed himself to be photographed wearing a Lacoste shirt and khaki shorts,\\nBJP leaders always wear traditional Indian clothes. The main enemy image\\nis nevertheless not the West but Islam, which is depicted as a martial and\\ncruel religion alien to the subcontinent, and Indian Muslims (the descen-\\ndants of converts, like Bosnian Muslims) are represented partly as traitors to\\nHinduism, partly as foreign invaders. The demolition of a mosque in the\\nnorthern town of Ayodhya in December 1992, the ensuing riots in several\\nIndian cities, and the call for the rebuilding of a Hindu temple allegedly de-\\n54\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nstroyed by a Mughal ruler four centuries ago marked a climax of sorts in\\nthis respect (see van der Veer, 1994, for details).\\nThe phenomenal rise of this traditionalist movement is a result of several\\nconnected processes of sociocultural change or modernization. First, the\\nvery notion of hindutva, Hindu-ness, is a modern one. Hinduism is not a\\n“religion of the Book.” It is an noncentralized religion with scores of holy\\nscriptures, thousands of avatars (incarnations of divinities), and very many\\nways of worshiping them. The idea of the Hindu identity as an imagined\\ncommunity based on cultural similarity is alien to Hinduism as such, which\\nis a religion based on complementarity, difference, and hierarchy. Regard-\\ning political Hinduism, some Indian commentators actually speak of a\\nSemitization of Hinduism whereby it takes on structural characteristics from\\nthe great religions of West Asia. \\nSecond, the hindutva movement is explicitly modeled on European na-\\ntionalism—some early hindutva ideologists were even warm admirers of\\nHitler—which has been, for 150 years, an attempt to reconcile change and\\ncontinuity by talking of roots and traditions in a situation of industrializa-\\ntion and urbanization. This is obvious in hindutva practice, whereby issues\\nregarding national anthems, dress, and foreign foods are given prominence,\\nwhile profound social changes continue to affect everyday life as before.\\nThere is a clear connection between the rise of the BJP and the liberaliza-\\ntion of the Indian economy, the rise of a substantial new middle class with \\na strong consumerist orientation and the rapid spread of new mass media\\nincluding the Rupert Murdoch-controlled Star TV network. While liber-\\nalization of this kind stimulates consumerism (perceived as Westernization),\\nit also indirectly boosts traditionalism since the new patterns of consump-\\ntion and the new media scene may indicate that cherished traditions are \\nunder threat. \\nThird, the “contagious” influence from political Islam is obvious; hin-\\ndutva is the assertion of Hindu identity as opposed to Muslim identity both in\\nPakistan and in India itself. Doubly ironic, hindutva has double origins in\\nEuropean romanticism and West Asian political Islam. When its first ide-\\nologist, Dr. Veer Savarkar, wrote in the 1920s that “Hindutva is not the\\nsame thing as Hinduism,” he was therefore right, but not for the reasons he\\nbelieved. Savarkar saw hindutva as a wide-ranging social movement emanat-\\ning from Hindu faith and practices, while a more historically correct ac-\\ncount sees it as the result of cultural diffusion from Europe and West Asia.\\nFourth, and perhaps most important, the hindutva movement can be\\nseen as a reaction against a growing egalitarianism in Indian society. Al-\\nready in the 1950s, policies attempting to improve the conditions of the\\n“Untouchables,” the lowest castes, were introduced, and during the 1990s,\\nvery radical measures have been proposed to this effect—and in some cases\\ncarried out. About half of India’s population are now defined as being either\\nDalits (“Untouchables”), tribals, low-caste people, or “OBCs” (Other Back-\\nEthnic Identity, National Identity, and Intergroup Conflict\\n55\\n\\n\\nward Classes), and in theory, all of 49.5 percent of jobs in the public sectors\\nshould be reserved for these groups, following the recommendations of the\\ngovernment-appointed Mandal commission. Since the early 1990s, this\\nprinciple has been enforced in many areas. Naturally, many members of the\\n“twice-born,” upper castes feel their inherited privileges eroding away, and\\nhindutva is largely a movement representing the interests of the disen-\\nchanted upper castes. It is for the most part a reaction against the move-\\nment toward greater equality in Indian society. Although hindutva seems to\\npromote equality among Hindus, an implication of its traditionalist Hin-\\nduism is the reinvigoration of the caste system, which in effect benefits only\\nthe “twice-born” castes. \\nThis analysis of hindutva must by necessity be a superficial one (see\\nHansen, 1999, for a full treatment). It must be remarked, however, that in-\\nterreligious marriages (and, indeed, intercaste marriages) are rare outside\\ncertain elite groups. Casual interaction between Hindus and Muslims is far\\nfrom unusual, but as in the Bosnian and Fijian cases, the intimate (family)\\nspheres, as well as personal networks of close friends, rarely cross religious\\nboundaries. Social classification in India is nonetheless complex, and as will\\nbe indicated later, the Hindu-Muslim divide is only one of several possible\\nsocial dichotomies—unlike in Bosnia and Fiji, where religious or ethnic\\ncontrasts tend to be paramount.\\nSome Comparisons\\nPrevious sections have outlined three contemporary conflicts involving col-\\nlective identity as a political resource. The differences are obvious; the \\nfocus here will therefore be on the similarities.\\nThe conflicts have three important sociological features in common.\\nFirst, there is in all three cases competition over scarce resources. As Horowitz\\n(1985) and many others writing about group conflict in contemporary soci-\\neties have shown, such conflicts invariably involve perceptions of scarcity\\nand struggles to retain or attain hegemony or equality. Successful mobiliza-\\ntion on the basis of collective identities presupposes a widespread belief that\\nresources are unequally distributed along group lines. “Resources” should\\nbe interpreted in the widest sense possible, and could in principle be taken\\nto mean economic wealth or political power, recognition, or symbolic\\npower—although what is usually at stake is either economic or political re-\\nsources. This feature is easy to identify in all three examples described\\nabove: Fijians and Indians compete over relative political and economic\\npower; the constituent groups of Bosnia compete over political power\\nand/or sovereignty; hindutva is an attempt to defend the political and eco-\\nnomic interests of “Hindus” in secular India.\\nSecond, modernization actualizes differences and triggers conflict. With the\\nintegration of formerly discrete groups into shared economic and political\\n56\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nsystems, inequalities are made visible, as comparison between the groups\\nbecomes possible. In a certain sense, ethnicity can be described as the\\nprocess of making cultural differences comparable, and to that extent, it is a\\nmodern phenomenon. The Fijian example, where the increasing integra-\\ntion of Fijians into the modern sphere made it apparent that Indians were\\ndoing better economically, illustrates this point. In India, the rise of the\\nDalit movement struggling for recognition and equal rights on behalf \\nof “Untouchable” groups is an expression of the modern value of equality,\\nand the counterreaction from the Hindu right is an attempt to stop egali-\\ntarianism from spreading, as well as reflecting—almost with the accuracy of\\na mirror image—symbolic competition with Muslims within and (espe-\\ncially) outside India. The Bosnian example, admittedly, seems less straight-\\nforward, as socialist Yugoslavia was in many ways no less modern than its\\nsuccessor countries (some would indeed argue that at least at the level of\\nideology, it was infinitely more modern than them). What is clear, and\\nwhich also holds true for other Eastern and Central European countries, \\nis that the sudden introduction of liberal political rights and a capitalist\\neconomy around 1990—core characteristics of non-socialist moderniza-\\ntion—created a new dimension of comparison between individuals and new\\narenas of competition. \\nThird, the groups are largely self-recruiting. Intermarriage is rare in all three\\ncases (excepting urban Yugoslavia). Although biological self-reproduction is\\nby no means necessary for a strong collective identity to come about, it\\nshould be kept in mind that kinship remains an important organizing princi-\\nple for most societies in the world, and a lot of what passes for ethnicity at the\\nlocal level is really kinship. Kinship has an important social dimension in ad-\\ndition to its symbolic side, which is highlighted in ideologies of fictive or\\nmetaphoric kinship. Symbolic boundaries are never effective unless under-\\npinned by social organization. \\nFurther, there are several important ideological similarities. First, at the\\nlevel of ideology, cultural similarity overrules social equality. Ethnic national-\\nism in Yugoslavia, political Hinduism in India, and the “sons-of-the-soil”\\nrhetoric of Fiji all depict the ingroup as homogeneous, as people “of the\\nsame kind.” Internal differences are undercommunicated, and moreover,\\nin the wider political context, equality values are discarded for ostensible\\ncultural reasons. (Although it could be argued that hindutva is a Trojan\\nhorse concealing upper-caste interests with all-Hindu rhetoric, the point \\nis that it stresses the commonalities of all Hindus irrespective of caste or\\nlanguage.)\\nSecond, images of past suffering and injustice are invoked. Serbs bemoan the\\ndefeat at the hands of the Turks in Kosovo in 1389; Hindu leaders have\\ntaken great pains to depict Mughal (Muslim) rule in India from the 1500s as\\nbloody and authoritarian; and indigenous Fijian leaders compare their\\nplight to that of other indigenous peoples who have suffered foreign inva-\\nsions. Violence targeting the descendants of the invaders can therefore be\\nEthnic Identity, National Identity, and Intergroup Conflict\\n57\\n\\n\\nframed as legitimate revenge. Even hindutva leaders, who claim to represent\\n80 percent of India’s population, complain that Hinduism is under siege and\\nneeds to defend itself with all means available. \\nThird, the political symbolism and rhetoric evokes personal experiences. This is\\nperhaps the most important ideological feature of identity politics in gen-\\neral. Using myths, cultural symbols, and kinship terminology in addressing\\ntheir supporters, promoters of identity politics try to downplay the differ-\\nence between personal experiences and group history. In this way, it be-\\ncomes perfectly sensible for a Serb to talk about the legendary battle of\\nKosovo in the first person (“We lost in 1389”), and the logic of revenge is\\nextended to include metaphorical kin, in many cases millions of people.\\nThe intimate experiences associated with locality and family are thereby\\nprojected onto a national screen. This general feature of social integration\\nhas been noted by Handelman (1990), analyzing national rituals, and much\\nearlier in Turner’s (1967) studies of ritual among the Ndembu of Zambia. In\\nshowing that rituals have both an instrumental and an emotional (or sen-\\nsory) dimension—one socially integrating, the other metaphorical and per-\\nsonally meaningful—Turner actually made a point crucial to the present\\nanalysis—namely, that loyalty to a larger collectivity (such as a tribe or a na-\\ntion) is contingent on its imagery being personally meaningful. \\nFourth, first-comers are contrasted with invaders. Although this ideological\\nfeature is by no means universal in identity politics, it tends to be invoked\\nwhenever possible, and in the process, historical facts are frequently\\nstretched. In Fiji, the Fijian population—although genetically a Polynesian-\\nMelanesian mix—has a strong case here, although it is less obvious that\\nIndo-Fijians can be immigrants to a country in which they were born, and\\ntherefore legitimately deprived of equal rights. Regarding Bosnia and India,\\nas mentioned above, there is nothing to suggest that the ancestors of Mus-\\nlims in the respective countries were more recent arrivals than the ancestors\\nof Christians or Hindus, although Islam is a relatively recent import. What\\nis interesting here is how the varying depth of cultural genealogies (“roots”)\\nis used to justify differential treatment. The historical location of the self\\nalong the dimensions of descent and place is thereby invested with political\\nsignificance.\\nFifth and finally, the actual social complexity in society is reduced to a set of\\nsimple contrasts. As Adolf Hitler wrote in Mein Kampf, the truly national\\nleader concentrates the attention of his people on one enemy at the time.\\nSince cross-cutting ties reduce the chances of violent conflict, the collective\\nidentity must be based on relatively unambiguous criteria (such as place, \\nreligion, mother-tongue, kinship). Again, internal differences are under-\\ncommunicated in the act of delineating boundaries toward the demon-\\nized Other. This mechanism is familiar from a wide range of interethnic \\nsituations, from social classification in Zambian mining towns (Epstein,\\n1992) to Norwegian–Sami relations in sub-Arctic Scandinavia (Eidheim,\\n1971), the Sinhalese–T\\namil conflict (Kapferer, 1988) and Quebecois nation-\\n58\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nalism (Handler, 1988): the Other is reduced to a minimal set of “traits,” and\\nso is the collective Self.\\nThese similarities do not necessarily indicate that there are universal\\nmechanisms linking personal selves and larger collectivities, but they do\\nsuggest that there is a universal “grammar” common to contemporary iden-\\ntity politics everywhere. In the final sections of this chapter this argument\\nwill be pursued slightly further, and it will be suggested that universal con-\\nnections between the self and the collective exist, which must be under-\\nstood not only to account for traditional societies but to make sense of the\\npresent. Far from being an “atavistic” or “primitive” counterreaction to\\nglobalization or modernization, identity politics is a special case of some-\\nthing more general—namely, collective identity anchored in personal \\nexperiences.\\nWhere Is the Identity of Identity Politics?\\nSocial scientists have proposed many typologies of ethnic conflict, dividing\\nthe groups involved into categories such as majority, minority, irridentist,\\nand separatist, using variables such as division of labor, relative political\\npower, and historical intergroup relations as criteria of classification. In my\\nview, this kind of exercise can at best generate a limited understanding of\\nthe dynamics of group conflict. T\\no begin with, the very adjective “ethnic” is\\nhardly appropriate to describe all conflicts based on identity politics. Indian\\nHindus are not an ethnic group in any meaningful sense, and it is a matter\\nof definition whether Serbs, Croats, and Muslims in Bosnia should be con-\\nsidered ethnic groups (they have shared origins only a few centuries back).\\nMany contemporary conflicts displaying some or all of the features listed\\nabove cannot be seen as ethnic. I note a few African examples: The Su-\\ndanese civil war is partly fought over religion (northeners are Muslims try-\\ning to Islamicize the south), partly over culture and language—neither\\nnorthern nor southern Sudanese are ethnic groups. Hutus and Tutsis in\\nRwanda and Burundi, like the constituent groups of Bosnia, are culturally\\nvery close; they speak the same language and have the same religion. The\\nSomali civil war presents an even more puzzling case, as Somalia is one of\\nthe few sub-Saharan states that are truly ethnically homogeneous and so far\\nthe only one that seems to have relinquished the trappings of statehood\\ncompletely, having dissolved into warring clans (an intermediate level of so-\\ncial organization, between the family and the ethnic group) since the early\\n1990s. T\\no the northwest of Somalia, one of the great forgotten wars of\\nAfrica is being fought over a contested border area between Eritrea and\\nEthiopia. Eritrea, which seceded from Ethiopia in 1991, has never been\\nbased on religion or ethnic identity, but has a vague legitimacy as a histori-\\ncal nation in the brief period of Italian colonialism before the Second\\nWorld War. The current war is being fought between Tigrinya speakers on\\nEthnic Identity, National Identity, and Intergroup Conflict\\n59\\n\\n\\nboth sides of the borders, who are united through religion, language, cus-\\ntoms, history, and even kinship ties, but they are no less bitterly divided po-\\nlitically. This conflict in turn creates its strange bedfellows in the alliance\\nbetween Ethiopian Tigrinyas and Amhara speakers from the highlands,\\nwho are traditional enemies.\\nIn other words, the concept “ethnic conflict” is misleading, whether it is\\nused to classify phenomena or to explain hostilities. Several of the alterna-\\ntive terms one might consider are, however, no less misleading: “Cultural\\nconflict” will clearly not do, as it is obviously not what is usually thought of\\nas cultural differences that lie at the heart of the conflicts. At the village\\nlevel, even Hindus and Muslims in India hold many of the same beliefs and\\nworship in similar ways. The low-intensity conflict in Fiji involves groups\\nthat are by any criterion more culturally different than, say, the Bosnian\\ngroups. All the conflicts considered here are over resources perceived as\\nscarce: territory, political power, economic gain, employment, recogni-\\ntion—rights in a wide sense. What they have in common is their successful\\nappeal to collective identities perceived locally as imperative and primor-\\ndial, identities associated with a deep moral commitment, whether ethnic\\n(based on notions of kinship and descent), regional (based on place), or reli-\\ngious (based on beliefs and forms of worship). For these reasons, the term\\n“identity politics” is preferable as a generic term for all such political move-\\nments, whether nonviolent or violent.\\nThis final section will therefore amount to an attempt to unravel the\\nidentity of identity politics. What is it that makes it so powerful? What is\\nthe “identity” that such political movements can draw upon? \\nBenedict Anderson proposes an answer in the introduction to his seminal\\nImagined Communities (1983), where he points out that nationalism has\\nmore in common with phenomena such as religion and kinship than with\\nideologies like liberalism and socialism. He argues that nationalism (and,\\none might add, any form of identity politics) expropriates personal identity,\\ntransforming intimate experiences into the raw material of politics. I owe\\nmy existence to my parents, and by metonymical extension they represent\\nthe larger, abstract collective. I harbor tender feelings for my childhood,\\nwhich by extension becomes my group’s glorious and tragic history. I feel\\nattached to the place where I grew up, which was not just any arbitrary\\nplace but the nation (or, as the case might be, the sacred land of Hinduism,\\nthe traditional territory of the Fijians, the tormented country of the brave,\\nbut sadly misunderstood Serbs). Indeed, this argument can profitably be\\nseen as echoing Turner’s aforementioned argument on the instrumental and\\nemotional dimensions (or “poles”) of ritual. In both cases, the integrative\\nstrength of the imagined community (be it a tribe or a nation) depends on\\nits ability to mobilize emotions proper to the intimate sphere of kinship and\\npersonal experience. \\nThe conditions for this transformation to take place—the move from an\\ninterpersonally anchored identity to an abstract national, ethnic, or reli-\\n60\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ngious identity—are usually tantamount to certain general conditions of\\nmodernity (cf. Gellner, 1997). It is through school and mass media that\\npeople are taught to identify with an abstract, mythically rooted community\\nof people “of the same kind.” Through the replacement of traditional\\neconomies with an abstract labor market, they become participants in a\\nlarge-scale system of subsistence. Through the implementation of a bureau-\\ncratic system of political management, their allegiances are at least partly\\nmoved from the concrete to the abstract community. \\nIt is important to remember, as theorists of nationalism and ethnicity\\nhave pointed out time and again, that identification is relational, situational,\\nand flexible, and that each person carries a number of potential identities,\\nonly a few of which become socially significant, making a difference in\\neveryday life. Even fewer gain political importance, forming the basis of\\npower struggles and group competition. This is not, however, to say that\\ncollective identities can be created out of thin air. They have to be con-\\nnected, in credible ways, to people’s personal experiences. These experi-\\nences in turn are flexible—not only historians but everybody else as well \\nselects and interprets events to make a particular kind of sense of the past—\\nbut not indefinitely so. Regarding our main examples, in Fiji virtually no-\\nbody doubts whether he or she is Indian or Fijian, and politics—whether\\nbased on compromise or conflict—will have to take this into account for the\\nforeseeable future. In Yugoslavia after the breakup, cross-cutting ties and\\ncultural hybridity were undercommunicated. Cosmopolitanism was in-\\ncreasingly seen as a suspect, unpatriotic attitude, and people of mixed an-\\ncestry were forced to choose a bounded, unambiguous identity: they had \\nto select past experiences that made them either Serbian, Croatian, or \\nMuslim—more or less like the proverbial North African mule, who speaks\\nincessantly about his uncle, the horse, but never mentions his father, the\\ndonkey. In India, finally, some of the strongest scholarly arguments against\\nthe lasting influence of hindutva have actually pointed toward people’s per-\\nsonal experiences (Frøystad, 1999). Since Indian everyday life is still perme-\\nated by caste distinctions, and caste continues to define the very fabric of\\nsocial integration, these scholars argue that hindutva—the idea that all Hin-\\ndus have something profound in common—is so counterintuitive to most\\nIndians that it can unite Hindus only as long as the enemy image of Mus-\\nlims can be kept ablaze. \\nAll the basic components of political identity familiar from classic politi-\\ncal anthropology can be identified in contemporary identity politics: it is\\nbased on a sometimes ambiguous mix of kinship and locality; it has well de-\\nveloped myths of origin and myths of past suffering; and it distinguishes\\nclearly between “us” and “them.” The main difference between, say, the no-\\nmadic Nuer society studied by E. E. Evans-Pritchard (1940) in the 1930s\\nand Serbian (or Croatian) nationalism today is probably that of scale: while\\nthe Nuer rarely imagined themselves as members of larger groups than the\\nclan, a Serb in Vojvodina can readily identify him or herself with a Serb in\\nEthnic Identity, National Identity, and Intergroup Conflict\\n61\\n\\n\\nKosovo. The act of transformation from personal, concrete social experi-\\nences to the abstract community is naturally much more demanding in a\\nlarge-scale society than in a village-based one, hence the importance of\\nmodern institutions of communication, economic transactions, and political\\nrule for the growth of abstract communities.\\nSome Final Lessons from Political Anthropology\\nHaving long ago abandoned the early ambition of becoming “a natural sci-\\nence of society,” social and cultural anthropology has for decades been re-\\nluctant to formulate lawlike propositions about the functioning of society.\\nThe constructivist turn of recent years seems to confirm that contemporary\\nanthropology is less concerned with absolute truths than with the analysis\\nof local cultural constructions. This need not be so, and the study of current\\nidentity politics may illustrate the power of comparative anthropology in\\ngenerating general hypotheses.\\nEarly instrumentalist research on ethnic groups, particularly in Africa (as\\nin A. Cohen, 1974), searched for the logic of group cohesion, which they\\nassumed to be roughly the same everywhere. The related, actor-based per-\\nspective developed by Barth (1969) and his colleagues assumed the logic of\\naction to be quite universal—people act to maximize benefits. Later analy-\\nses of the constructedness of ethnic and national ideologies (the seminal\\ntext is Hobsbawm and Ranger, 1983; see Chapman, Malcolm, McDonald,\\n& T\\nonkin, 1989, for an overview) also emphasize universal characteristics of\\na particular kind of societal formation (the modern state) and its relation-\\nship to group identities based on notions of culture. The canonical texts on\\nnationalism (such as Gellner, 1983; Anderson, 1983; Smith, 1991) also have\\nclear universalist ambitions. As this chapter has made clear, these ap-\\nproaches have obvious strengths, but they need to be supplemented by de-\\ntailed research on the experiential world of the everyday—the Lebenswelt\\n(life-world) of the actors. A renewed focus on the informal, intimate, and\\noften noninstrumental dimensions of everyday life reveals that terms such\\nas “ethnicity” by themselves explain little. The parallels between a sup-\\nporter of the BJP and a supporter of Serbian supremacists should not be lo-\\ncated to their respective “ethnic” identities or “civilizational” membership,\\nbut to the fact that their everyday life, social networks, and personal obliga-\\ntions connect them to particular groups that may be exploited politically,\\ngiven the right circumstances. It should also be kept in mind that class poli-\\ntics can sometimes be a form of identity politics (Shore, 1993), which can\\nprofitably be understood along the same lines as ethnic or religious identity\\npolitics. The cause of group allegiance lies in the everyday, not in the over-\\narching ideology.\\nIn order to complete this analysis, it is necessary to go a few decades\\nback, to classic political anthropology, in order to see how the perspectives\\n62\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ndeveloped earlier in the chapter can be enriched by the work of previous\\ngenerations. \\nThe integration of persons into groups can be described as the work of\\nan inverted refrigerator: the function of a refrigerator is to generate inward\\ncoldness, but in order to do so it more or less inadvertently, as a side effect,\\ncreates outward warmth. Conversely, groups form to create warmth for\\ntheir members, but they necessarily create some outward coldness in order\\nto be able to do so. Under particular circumstances, the outward coldness is\\nmore readily perceptible than the inward warmth. A sociological principle\\noriginally formulated by Georg Simmel, known as “Simmel’s Rule,” simply\\nstates that the internal cohesion of a group is contingent on the strength of exter-\\nnal pressure. This principle may explain why group integration generally is\\nso much stronger in small groups, especially if they are oppressed, than in\\nlarge ones—why, for example, Scots seem to have fewer difficulties defining\\nwho they are than do the English. \\nAn interesting corollary of Simmel’s Rule is the fact that what kind of\\ngroup emerges depends on where the perceived pressure comes from. Both\\ngender-based and class-based social movements have periodically been suc-\\ncessful, given that the perceived threat was seen, not as alien religion or for-\\neign ethnic groups, but as male supremacy and ruling classes, respectively.\\nIn accordance with this, some inhabitants of Sarajevo during the war felt\\nthat the conflict was really an urban–rural one, since city-dwellers had a lot\\nin common, irrespective of religion, that they did not share with rural peo-\\nple. Strong opposition groups in India, similarly, argue against a view of \\nIndian politics as divided between Hindu communalists and liberals because\\nthey see the main problem of Indian society as one of poverty and distribu-\\ntion of resources, to which neither of the parties seems to give priority. In\\nFiji, finally, the immediate reason for the 1987 coup was the establishment\\nof a government of national unity promising to address issues of social wel-\\nfare and economic development rather than intergroup issues. In other\\nwords, redefinitions of societal cleavages are entirely possible insofar as they\\ndo not contradict people’s everyday experiences too obviously. \\nIn the course of this chapter, cross-cutting ties and conflicting loyalties\\nhave already been mentioned as mitigating forces in situations of inter-\\ngroup conflict. Phrased within the terminology used here, one might say\\nthat shared experiences across boundaries reduce the risk of conflict. In\\nMax Gluckman’s reinterpretation of Evans-Pritchard’s Nuer material from\\nthe 1930s (Gluckman, 1956–1982), this point was made forcefully. The\\nNuer were organized along kinship lines across villages, but they were also\\nlocally integrated in villages. The women married out of the village and the\\nlineage, so that everybody had affines (in-laws) in other villages. Further-\\nmore, men were tied to nonrelatives through trade, initiation rituals, and\\nfriendship. All of these factors led to a reduction in the incidence of vio-\\nlence among the Nuer. In contemporary identity politics, it can easily be\\nseen—and has been remarked above—how political leaders emphasize in-\\nEthnic Identity, National Identity, and Intergroup Conflict\\n63\\n\\n\\nward similarity and outward boundaries in order to reduce the potentially\\nmitigating impact of cross-cutting ties. They are true to people’s everyday\\nlife, but try to emphasize certain experiences at the expense of others (in-\\nward solidarity and similarity, outward conflict and difference).\\nAnother, even more time-honored principle from political anthropology\\nis the twin notions of fusion and fission in tribal societies. When the sole\\norganizing principle for a group lacking hierarchies and formal political of-\\nfice is kinship, there are limits to the group’s growth; at a certain point, it\\nsplits into two. Without such a fission, internal conflicts would soon be-\\ncome overwhelming given the simple social organization of such societies,\\nand the effects would be destructive. Fusion of discrete groups has also been\\nstudied extensively, but in many acephalous societies it is seasonal (nomadic\\ngroups fuse in the dry season or in winter) and fragile.\\nA more dynamic view of contraction and expansion of tribal groups \\nwas developed, especially by Africanists, from the 1940s (Fortes & Evans-\\nPritchard, 1940). In studies of feuding and political competition, they\\nshowed how two or several local groups that might be periodically involved\\nin mutual feuding united temporarily when faced with an external enemy.\\nThis form of organization, described as segmentary by Evans-Pritchard, fol-\\nlows the proverb often cited in recent years to explain the logic of the So-\\nmali civil war: “It’s me against my brother, my brother and I against our\\ncousins, and our cousins, my brother and I against everybody else.” A form\\nof segmentary logic is apparent in politics nearly everywhere; a distinguish-\\ning mark of modern nation-building has nevertheless been its attempt to\\nchannel loyalties away from various subnational levels of identity in order to\\nmonopolize the political loyalty of individual citizens. \\nThe segmentary logic creates a fluid, relational political organization\\nthat, in its pure form, is impracticable in modern state societies given their\\nrequirements for stability, centralized power, and reified systems of political\\nrepresentation. This does not, however, mean that segmentary identi-\\nfication does not continue to exist, and one of the causes of oppositional\\nidentity politics in modern nation-states is their not providing subnational\\nidentity groups appropriate political arenas, thereby encouraging counter-\\nreactions in the form of identity politics directed against the state. \\nThe formation of identity-based political groups generally entails both\\nan expansion and a contraction of the focus for identification. At the time of\\nthe breakout of conflict in Bosnia, the federal or even state level was in-\\ncreasingly seen as irrelevant—the process was one of fission. At the same\\ntime, internal conflicts and schisms within each constituent group were\\nminimized, and as a result each group became more coherent and united\\nthan before. In the cases of Fiji and India, this is even more obvious: among\\nFijians, rivalry between chiefs and clans has diminished in importance as \\nFijian politics has grown increasingly ethnic; similarly, rifts within the In-\\ndian population on the basis of regional origin, which could formerly lead\\nto Indian subgroups supporting Fijian-dominated governments, have be-\\n64\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\ncome much less important since the military coup. In the Indian case, the\\nvery idea of hindutva implies an enormous expansion of the ingroup for\\nHindus. Trying to bridge differences based on language, caste, region, and\\nculture, hindutva tries to create a morally committing all-Indian Hindu\\nidentity based on symbolic equality. This is, in the Indian context, a very\\nradical move. Simultaneously, Indian citizenship and (overarching, supra-\\nreligious) national identity become less important since the federal Indian\\nstate includes millions of non-Hindus, who are depicted as internal enemies\\nby hindutva spokespersons.\\nIn all the examples considered here, group segmentation at a higher\\nlevel, and the ensuing formation of imagined communities larger than the\\nlocality but smaller than the state, is immediately related to the need for a\\nfirm boundary in a situation of conflict within the state. The alternative\\nidentity of national citizenship, which encompasses the other as well, no\\nlonger functions. Interestingly, contemporary identity politics is very simi-\\nlar to nationalism—for example, in its appeal to mythical foundations, its\\nabstract postulation of similarity and equality, its rejection of segmentary\\nidentity formation, and its attempt to reduce a world of many small differ-\\nences to a world of only a few, major ones. In many cases, it is more success-\\nful than nationalism, particularly in postcolonial, multiethnic states. This \\nis not only because identity politicians promise its adherents that they will\\nwin zero-sum games against political competitors, but also because they \\nare able to represent themselves as natural extensions of people’s personal,\\nexperience-based identities. \\nA challenge for modern states, thus, consists in coping with the fact that\\npersonal identity can be exploited politically not only by the state itself but\\nby others as well, not because the self is infinitely multifaceted (it is not),\\nbut because the experiences and relationships that make up the self can be\\nexpanded symbolically in several, often conflicting ways. Processes of seg-\\nmentary fusion and fission, the formation of different kinds of groups in-\\nvolving overlapping personnel owing to the functioning of Simmel’s Rule,\\ngroup-based antagonism, and competition: these ways of expressing politi-\\ncal interests, underpinned by shared meaning within the ingroup, are no\\nmore eradicated by modernity than is personal identity. The challenge thus\\nconsists in laying the foundations for “a sense of belonging to a community\\nlarger than each of the particular groups in question” (Laclau, 1995, p. 105),\\nand this can be done only by first acknowledging both the richness and the\\nvariability of personal identities.\\nIn the face of violent identity politics, “ethnic cleansing,” and the strong\\nattention to “roots” and historically based identities (ungenerously de-\\nscribed as “the narcissism of small differences” by Michael Ignatieff) char-\\nacterizing many societies in recent years, it is not surprising that intellectu-\\nals have recently tried to think essentialism away, emphasising the endlessly\\nflexible and fluid character of human identification. A typical expression of\\nthis position can be found in a recent text by the influential sociologist \\nEthnic Identity, National Identity, and Intergroup Conflict\\n65\\n\\n\\nZygmunt Bauman, where he states: “If the modern ‘problem of identity’ is\\nhow to construct an identity and keep it solid and stable, the postmodern\\n‘problem of identity’ is primarily how to avoid fixation and keep the options\\nopen” (Bauman, 1996, p. 18).\\nIn today’s world it may be ethically imperative to endorse Bauman’s posi-\\ntion, but it is equally important to keep in mind that humans are not free-\\nfloating signifiers, and no amount of benevolent intentions will be able to\\nchange people’s life-worlds overnight. Rather than trying to think them\\naway, it is necessary to understand them and come to terms with their \\nenduring power. Notwithstanding globalization and the universalization \\nof modernity, cultural differences continue to exist, within and between\\nplaces, within and between nations and ethnic groups. It is also, however,\\ndoubtlessly true that carbon can be turned into graphite as well as dia-\\nmonds, and the ways in which cultural differences become socially relevant\\nvary importantly. But to pretend they do not exist outside ethnic and na-\\ntionalist ideologies would be intellectually indefensible; people’s personal\\nexperiences are the very raw material of such ideologies. Here lies an im-\\nportant limitation in constructivist models of identity. Collective identities\\nare constructed, consciously or not, but nothing comes out of nothing. In\\nlocating the universal not in the workings of identity politics (it changes\\nhistorically and varies geographically), nor in the eternal sovereignty of \\nthe state (the same objection applies), but rather in the social life-worlds in\\nthat individuals make sense of the world, we may have found a basis for\\ncomparison that will outlive academic fads and contemporary politics. \\nAcknowledgments\\nThe author would like to express his thanks to the editors and three anonymous\\nreaders for their critical readings and useful comments on an earlier version of the\\nchapter.\\nReferences\\nAnderson, B. (1983). Imagined communities: Reflections on the origin and spread of na-\\ntionalism. London: Verso.\\nBarth, F. (Ed). (1969). Ethnic groups and boundaries. Oslo: Scandinavian University\\nPress.\\nBauman, Z. (1993). Postmodern ethics. Cambridge: Polity.\\nBauman, Z. (1996) From pilgrim to tourist; or a short history of identity. In S. Hall\\n& P. Du Gay (Eds)., Questions of cultural identity (pp. 18–36). London: Sage.\\nBringa, T. (1996). Being Muslim the Bosnian way: Identity and community in a central\\nBosnian village. Princeton, NJ: Princeton University Press.\\nChapman, M., McDonald, M., & T\\nonkin, E. (1989). Introduction—history and so-\\ncial anthropology. In E. T\\nonkin, M. McDonald, & M. Chapman (Eds.), History\\nand ethnicity (pp. 1–21). London: Routledge.\\n66\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nCohen, A. (1969). Custom and politics in urban Africa: A study of Hausa migrants in a\\nYoruba town. London: Routledge & Kegan Paul.\\nCohen, A. (1974). Two-dimensional man: An essay on power and symbolism in complex \\nsociety. London: Routledge & Kegan Paul.\\nCohen, A. P. (1994). Self consciousness: An alternative anthropology of identity. London:\\nRoutledge.\\nCornell, S., & Hartmann, D. (1998). Ethnicity and race: Making identities in a chang-\\ning world. London: Pine Forge Press.\\nEidheim, H. (1971). Aspects of the Lappish minority situation. Oslo: Scandinavian Uni-\\nversity Press. \\nEpstein, A. L. (1992). Scenes from African urban life: Collected Copperbelt essays. Edin-\\nburgh: Edinburgh University Press.\\nEriksen, T. H. (1992). Us and them in modern societies: Ethnicity and nationalism in\\nTrinidad, Mauritius and beyond. Oslo: Scandinavian University Press.\\nEriksen, T. H. (1998a). Common denominators: Ethnicity, nationalism and compromise in\\nMauritius. Oxford: Berg.\\nEriksen, T. H. (1998b). Culture and ethnicity: A second look. Lecture at 6th SIEF\\nconference, Amsterdam, 21 April 1998.\\nEvans-Pritchard, E. E. (1940). The Nuer. Oxford: Clarendon.\\nFeatherstone, M. (Ed). (1990). Global culture. London: Sage.\\nFortes, M., & Evans-Pritchard, E. E. (Eds.). (1940). African political systems. London:\\nOxford University Press.\\nFriedman, J. (1994). Global culture and local process. London: Sage.\\nFrøystad, K. (1999). Personal communication.\\nFurnivall, J. S. (1948). Colonial policy and practice: A comparative study of Burma and\\nNetherlands India. New York: New York University Press.\\nGeertz, C. (Ed.). (1967). Old societies and new states: The quest for modernity in Africa\\nand Asia. New York: The Free Press.\\nGeertz, C. (1973). The interpretation of cultures. New York: Basic Books.\\nGellner, E. (1983). Nations and nationalism. Oxford: Blackwell.\\nGellner, E. (1997). Nationalism. London: Weidenfeld & Nicolson.\\nGiddens, A. (1991). Modernity and self-identity. Cambridge: Polity.\\nGiddens, A. (1994). Beyond left and right. Cambridge: Polity.\\nGluckman, M. (1956/1982). Custom and conflict in Africa. Oxford: Blackwell.\\nHandelman, D. (1990). Models and mirrors: Towards an anthropology of public events.\\nCambridge: Cambridge University Press.\\nHandler, R. (1988). Nationalism and the politics of culture in Quebec. Madison: Wiscon-\\nsin University Press.\\nHannerz, U. (1992). Cultural complexity: Studies in the social organization of meaning.\\nNew York: Columbia University Press.\\nHansen, T. B. (1999). The saffron wave: Democracy and Hindu nationalism in modern\\nIndia. Princeton, NJ: Princeton University Press.\\nHobsbawm, E., & Ranger, T. (Eds.) (1983). The invention of tradition. Cambridge:\\nCambridge University Press.\\nHolland, D. (1997). Selves as cultured: As told by an anthropologist who lacks \\na soul. In R. Ashmore & L. Jussim (Eds.), Self and identity: Fundamental issues\\n(pp. 160–190). Oxford: Oxford University Press.\\nHorowitz, D. L. (1985). Ethnic groups in conflict. Berkeley: University of California\\nPress.\\nEthnic Identity, National Identity, and Intergroup Conflict\\n67\\n\\n\\nHuntington, S. (1996). The clash of civilizations and the remaking of a world order. New\\nYork: Simon and Schuster.\\nIgnatieff, M. (1994). Blood and belonging: Journeys into the new nationalism. London:\\nVintage.\\nJenkins, R. (1996). Social identity. London: Routledge.\\nKapferer, B. (1988). Legends of people, myths of state. Baltimore, MD: Smithsonian \\nInstitution Press.\\nKelly, J. D. (1998). Aspiring to minority and other tactics against violence. In D. C.\\nGladney (Ed.), Making majorities: Constituting the nation in Japan, Korea, China,\\nMalaysia, Fiji, Turkey and the United States (pp. 173–197). Stanford, CA: Stanford\\nUniversity Press.\\nLaclau, E. (1995). Universalism, particularism and the question of identity. In \\nJ. Ratchmann (Ed.), The identity in question. London: Routledge.\\nLash, S., & Urry, J. (1994). Economies of signs and space. London: Sage.\\nPremdas, R. (1993). Balance and ethnic conflict in Fiji. In J. McGarry & B. O’Leary\\n(Eds.), The politics of ethnic conflict regulation (pp. 251–274). London: Routledge.\\nRex, J. (1997). The nature of ethnicity in the project of migration. In M. Guibernau\\n& J. Rex (Eds.), The ethnicity reader: Nationalism, multiculturalism and migration.\\nCambridge: Polity.\\nRoosens, E. E. (1989). Creating ethnicity. London: Sage.\\nSaid, E. W. (1978). Orientalism. New York: Pantheon.\\nSchlesinger, A. M. Jr. (1992). The disuniting of America. New York: Norton.\\nSchöpflin, G. (1993). The rise and fall of Yugoslavia. In J. McGarry & B. O’Leary\\n(Eds.), The politics of ethnic conflict regulation (pp. 172–203). London: Routledge.\\nShore, C. (1993). Ethnicity as revolutionary strategy: Communist identity construc-\\ntion in Italy. In S. Macdonald (Ed.), Inside European identities (pp. 27–53). Oxford:\\nBerg.\\nSmith, A. D. (1986). The ethnic origins of nations. Oxford: Blackwell.\\nSmith, A. (1991). National identity. Harmondsworth: Penguin.\\nStrathern, M. (1991). Partial connections. Savage, MD: Rowman and Littlefield.\\nT\\nambiah, S. J. (1989). The politics of ethnicity. American Ethnologist, 16, 335–349.\\nTinker, H. (1974). A new system of slavery: The export of Indian labour overseas 1830–\\n1920. Oxford: Oxford University Press.\\nTurner, V. (1967). The forest of symbols: Aspects of Ndembu ritual. Ithaca, NY: Cornell\\nUniversity Press.\\nTurton, D. (Ed.) (1997). War and ethnicity: Global connections and local violence. Wood-\\nbridge, Suffolk: University of Rochester Press.\\nvan der Veer, P. (1994). Religious nationalism: Hindus and Muslims in India. Berkeley:\\nUniversity of California Press.\\nWilliksen-Bakker, S. (1991). Fijians in business: A study in the transformation of\\nsymbols. Doctoral dissertation, University of Oslo.\\n68\\nThe Contribution of Identities to Intergroup Conflict\\n\\n\\nPART II\\nTHE CONTRIBUTION \\nOF ETHNIC AND \\nNATIONAL IDENTITIES \\nTO POLITICAL CONFLICT \\nIN THE UNITED STATES\\n\\n\\nThis page intentionally left blank \\n\\n\\nJack Citrin\\n4\\nCara Wong\\nBrian Duff\\nThe Meaning of American \\nNational Identity\\nPatterns of Ethnic Conflict and Consensus\\nNational identity is one among many often co-existing and overlapping so-\\ncial identities, including territorial, racial, religious, linguistic, and gender\\nidentities. The psychological task for individuals is to order and integrate\\ntheir different collective “selves” (Ashmore & Jussim, 1997). The political\\ntask for government is to balance the need for national unity with the com-\\npeting claims of other group identities. In modern society, emphasizing loy-\\nalty to the “nation” and making this the essence of the individual’s political\\nself-definition emerged as the dominant way to boost social solidarity\\n(Greenfield & Chirot, 1994). As an immigrant nation, the United States has\\nalways faced the problem of coping with ethnic diversity; the motto e\\npluribus unum expresses the desire for a strong sense of common American\\nidentity without indicating the proper balance between the national “one”\\nand the ethnic “many.”\\nThe contemporary meaning of American national identity is the focus of\\nthis chapter. Immigration and differences in fertility rates have changed the\\nethnic and religious composition of the United States (Farley, 1996;\\nWarner, 1993). Compared to 50 years ago, the country is more diverse, with\\nmany more people today from Latin America, Asia, and Africa. Moreover,\\nthis demographic change has occurred in the context of technological trans-\\nformations eroding national sovereignty in economic life and of the emer-\\ngence of intellectual trends legitimating the primacy of ethnicity in political\\nlife.\\nScholars disagree about whether globalization and multiculturalism have\\nweakened Americans’ sense of attachment to the “nation,” particularly\\namong minority groups (Reich, 1992; Barber, 1995; Hollinger, 1997;\\n71\\n\\n\\nWolfe, 1998; de la Garza, Falcon, & Garcia 1996). There also are conflict-\\ning views about the implications of a stronger sense of ethnic, as opposed to\\nnational, identity. Some scholars (Raz, 1994; Young, 1990) regard ethnic\\npride as a source of individual self-esteem among cultural minorities that\\ncontributes to the achievement of their group’s goals. Others (Schlesinger,\\n1992; Miller, 1995) claim that the emphasis on ethnic distinctiveness (in-\\ngroup favoritism) inevitably weakens common bonds and intensifies group\\nconflict (outgroup hostility), raising the specter of cultural and political\\nBalkanization. \\nThis chapter employs survey research to analyze the meaning and politi-\\ncal consequences of American national identity in the context of the demo-\\ngraphic, economic, and intellectual changes noted above. The conceptu-\\nalization of national identity formulated here draws on social identity\\ntheory (T\\najfel & Turner, 1986; Thoits & Virshup, 1997) to encompass both\\nself-categorization (or identification as) and affect (or identification with).\\nHowever, it goes beyond merely assessing the individual’s positive (or nega-\\ntive) feelings about the nation to determine the normative foundations of\\ntheir sense of American identity. \\nThe content of American identity is the set of ideas and sentiments that\\nform the conceptual framework of nationhood. At the cultural level, this\\nrefers to a particular collective representation of the nation indicating the\\nsubjective criteria for membership in the national community (Citrin, Haas,\\nMuste, & Reingold, 1994). A given conception of national identity thus in-\\ncludes normative assumptions about how members of different ethnic and\\ncultural groups should relate to one another. For example, must minority\\ngroups assimilate to a dominant language, religion, or political outlook in\\norder to achieve full acceptance as fellow-nationals, or can a nation accom-\\nmodate loyalty to a variety of cultural traditions? Historical experience sug-\\ngests that there is no single answer to this question, so it is important to\\nconceive of national identity in a way that recognizes the possibility of sev-\\neral conceptions of “Americanness” vying for popular support. \\nThe next section of the chapter provides a more detailed account of the\\ntripartite definition of national identity that distinguishes among the di-\\nmensions of self-categorization, affect, and normative content. The second\\nsection summarizes the content of competing ideological conceptions of\\nAmerican identity, with particular attention to their normative assumptions\\nregarding the articulation of strong ethnic identities. The analysis of survey\\ndata, presented in the third section, addresses three main questions: \\n1. What is the degree of national versus ethnic identification in\\nAmerican public opinion, as assessed by several different measures\\nof self-categorization and affective support? \\n2. What is the pattern of consensus and cleavage among white, black,\\nHispanic, and Asian respondents? Specifically, do minority ethnic\\ngroups have a weaker sense of national identity than whites, and\\n72\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nare there ethnic group differences in how people order their na-\\ntional and ethnic identities? \\n3. What is the impact of national and ethnic identities on policy pref-\\nerences? Drawing on symbolic politics theory (Sears, 1993), the \\nfinal section of the chapter tests hypotheses about the relative in-\\nfluence of these identifications on preferences regarding foreign\\npolicy, racial policies, and “multicultural” issues such as immigra-\\ntion and group representation. \\nThese analyses shed light on the impact of a superordinate “American”\\nidentity on feelings about “outgroups,” both domestic and foreign, and pro-\\nvide a basis for speculating about the implications of different patterns of\\nnational and ethnic identity for conflict and cooperation among America’s\\nracial and ethnic groups.\\nThe Dimensions of Social Identity\\nIdentity is a slippery concept. It is an assertion of both sameness and differ-\\nence. One answers the question “What is your identity?” by naming who\\none is like. This entails drawing boundaries: one is the same as some others\\nand different from everyone else. For this reason, a social identity both in-\\ntegrates and divides. \\nThe first step in measuring a social identity is to determine the basis of\\nself-categorization; one identifies the characteristic(s) shared by those with\\nwhom one psychologically belongs. We each possess multiple potential \\nsocial identities whose degree of overlap and whose relative significance \\nfor our self-concept and behavior may vary (Thoits & Virshup, 1997).\\nMoreover, while self-categorization may be the initial step in the formation\\nof a psychologically meaningful social identity, the range of one’s options is\\nfinite. Biology, the life cycle, social structure, the economy, and the govern-\\nment create the categories that demarcate social identities with potential\\npolitical consequences. Thus, some social identities are optional and others\\nimposed. Furthermore, the borders dividing identity groups are permeable,\\nbut not entirely open. It is easy to be a Bulls fan and then to “exit” psycho-\\nlogically for the Celtics when Michael Jordan retires. It is harder to shed\\none’s nationality or ethnicity and almost impossible to change one’s race or\\nsex. Finally, even when one can adopt a particular social identity by self-\\ncategorization (Turner, Hogg, Oakes, Reicker, & Blackwell, 1987), this\\nchoice frequently must be affirmed by the agreement of others in the group\\nthat you do indeed possess the criteria for membership. The daughter of\\nChinese immigrants may call herself an American, yet might be told by \\nfellow-citizens of European origin that she does not share their national\\nidentity.\\nSelf-categorization refers to “identification as,” connoting a perceived\\nThe Meaning of American National Identity\\n73\\n\\n\\nself-location in a group, and should be distinguished from “identification\\nwith,” which indicates positive affect toward others in the group and must be\\nmeasured separately. Social identity theory (T\\najfel, 1978) posits that mere\\nawareness of belonging to a group engenders positive feelings about the\\ngroup and a tendency to act on behalf of other group members, even if they\\nare unknown to one personally. In the same vein, Miller (1995) has argued\\nthat the moral value of a strong sense of national identity is that it fosters\\ndiffuse feelings of sympathy and obligation toward fellow citizens. \\nIn addition, there is a difference between membership in a group and psy-\\nchological attachment to it. One can identify with members of a minority\\ngroup, agreeing that they deserve more access to jobs or political power,\\nwithout categorizing oneself as a member of that group. This example un-\\nderscores the potential significance of identities based on shared values\\nrather than sociological similarities and the need to determine the condi-\\ntions under which “identification as” and “identification with” are closely\\nconnected. \\n“Identification with” entails learning about a group’s defining customs,\\nexpectations, and values, and making them one’s own. These common val-\\nues or ideas define the third component of a social (e.g., national) identity—\\nits content. The specific content of a group identity is socially constructed, in\\nthe sense that consensual decisions shape and reinforce the normative crite-\\nria for membership. Since these criteria are vulnerable to challenge and\\nsubject to change, it is important to assess the extent of their legitimacy. \\nStill, social identities are not automatically political. They are politicized\\nwhen feelings of identification with a group are combined with a belief in\\nadvancing its goals through collective action (Miller, Gurin, Gurin, &\\nMalanchuk, 1981). During the British Raj, Indians undoubtedly differenti-\\nated themselves cognitively from the English. The emergence of an ide-\\nology and organized movement demanding change in the cultural, eco-\\nnomic, and legal status of Indians infused this social identity with political\\ncontent. \\nThe Affective Dimension: Patriotism versus Chauvinism\\nA nation denotes a group of people seeking or possessing a common home-\\nland. Nationalism as a doctrine asserts that a group of people sharing char-\\nacteristics that differentiate them, in their own minds, from others should be\\npolitically autonomous. The affective dimension of national identity (iden-\\ntification with) refers to feelings of closeness to and pride in one’s country\\nand its symbols. Many scholars go further, however, and distinguish love of\\none’s own country from a sense of superiority to other countries (Schaar,\\n1981; Staub, 1997; Kosterman & Feshbach, 1989; Sullivan, Fried, & Dietz,\\n1992; Connor, 1993; T\\naylor 1995). They invoke the concept of patriotism\\nto refer to the positive emotion of love for one’s own people and homeland\\n74\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nand use “nationalism” pejoratively, not technically, to designate arrogance\\nand contempt toward other countries. Writing in this vein, for example,\\nGeorge Orwell described patriotism as defensive and nationalism as aggres-\\nsive. For Schaar (1981, p. 285), nationalism is a perversion of patriotism,\\n“its bloody brother.” Psychologists developing separate measures for these\\nconcepts tend to give the label “nationalism” or “blind patriotism” to name\\ndisdain for other countries and a drive to dominate them (Bar-T\\nal, 1993;\\nKosterman and Feshbach, 1989; Staub, 1997). \\nDrawing on social identity theory’s distinction between ingroup fa-\\nvoritism and outgroup hostility clearly is useful when studying national and\\nethnic identity. The emotive use of the term “nationalism,” however, is un-\\nnecessarily confusing. “Chauvinism,” not nationalism per se, is the term\\nused here to refer to an extreme and bounded loyalty, the belief in one’s\\ncountry’s superiority, whether it is right or wrong. Whether patriotism, an\\nideology of mutual affection among those with a common sense of national\\nconsciousness (Kelman, 1997, p. 170), and chauvinism empirically slide into\\neach other, such that caring for one’s own typically is associated with hos-\\ntility toward others, is a controversial issue to be explored below (see also\\nBrewer, this volume). \\nNormative Conceptions of American Identity\\nHistory shows that the attributes used to define the idea of a “people” or\\n“nation” are numerous and malleable (Greenfield, 1992). Historians distin-\\nguish “civic” nationalism, where criteria for belonging are identical to citi-\\nzenship and so, in principle at least, open to all, from “ethnic” nationalism,\\nwhere membership is based on a sense of shared blood (Connor, 1993).\\nThere is a modern tendency to endorse the principle of ethnicity as the le-\\ngitimate basis for political autonomy, but there are prominent counterex-\\namples of multiethnic states such as the United States (Greenfield, 1992).\\nAs Max Weber put it, nationality ultimately is based on the principle of ter-\\nritory, ethnicity on lineage. \\nWhatever the specific foundations of a sense of national identity, nation-\\nalism, in its neutral, technical meaning, implies that membership in the na-\\ntion is the citizen’s overriding group loyalty, taking precedence over other\\navailable foci of affiliation, including ethnicity, in circumstances where they\\nconflict. If a strong national identity means consciousness of special bonds\\nto one’s fellow-nationals, so that they become, at least symbolically, a large\\nextended family, then constructing a nation may involve psychological\\nlosses as well as gains. The development of a national identity may require\\ndiminishing or abandoning one’s attachment to other groups, including\\none’s ethnic group. \\nThe third dimension of our conception of national identity refers to its\\nnormative content. This means the particular set of ideas about what makes\\nThe Meaning of American National Identity\\n75\\n\\n\\nthe nation distinctive—ideas about its members, its core values and goals,\\nthe territory it ought to occupy, and its relations to other nations (Citrin et\\nal., 1994). In a multiethnic society, these ideas lay down the nature of le-\\ngitimate commitments to subnational communities of descent. In contem-\\nporary American politics, liberalism, nativism, and multiculturalism are \\nalternative political theories about how society should be organized and\\nhow national identity should be defined. Each proposes a different solution\\nfor how properly to balance national and ethnic identities, with conflicting\\nimplications for policymakers. The analysis below therefore assesses the \\ndegree of popular support for these competing ideas. \\nThe liberal conception of American identity is “civic,” not “ethnic.”\\nCommitment to the national “creed” of democracy and individualism\\n(Huntington, 1981; Gleason, 1980) is what makes one an American. Since\\nbelonging to the nation is equated not with shared blood but with common\\nbeliefs and customs (Gleason, 1980; Lind, 1995), anyone, regardless of an-\\ncestry, can become American through adherence to the dominant set of\\nideals, which include equality of opportunity and respect. The liberal image\\nof nationality is ethnically inclusive, in principle, if not always in practice\\n(Smith, 1997). \\nWhen it comes to ethnic diversity, the liberal conception of national\\nidentity is optimistic about the ability of contemporary American society to\\nassimilate newcomers. This means that ethnic Balkanization is not a serious\\nthreat, so there is no reason to extirpate the traditions of new immigrants as\\nattempted by the Americanization program in the 1920s. Over time, the\\nlargely Asian and Hispanic newcomers will blend into the cultural main-\\nstream, just like their European predecessors. Thus, people are free to\\nhonor their ethnic heritage if they wish, without undermining America’s\\nability to create one people out of many. Still, according to the liberal con-\\nception of national identity, when the claims of nationality and ethnicity\\nconflict, the former should take precedence. In psychological terms, liberal-\\nism rejects the notion of primordial identities, the idea that differences\\namong communities of descent are fundamental and enduring, exercising an\\ninherent dominion over the individual’s political conduct. \\nRogers Smith (1997) documents the persistent gap between the liberal\\ntheory of national identity and more nativist political practices. From the\\nbeginning of the United States, the official definition of “Americanness” ex-\\ncluded blacks and Native Americans and consigned women to a lesser role.\\nThen, in the 19th and 20th centuries, restrictive immigration and natural-\\nization laws discriminated on a racialist basis. Asians were denied citizen-\\nship, Mexicans were repatriated, and in 1924, immigration was limited to a\\nsmall number of people of Northern European origin. \\nThe nativist response to ethnic diversity (Higham, 1988) is to insist on\\ncultural conformity. If there is to be immigration, the nation’s policy should\\nfavor the admission of people who already are familiar with American po-\\nlitical culture (Brimelow, 1995). Once here, newcomers should undergo a\\n76\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nprogram of indoctrination that cleanses them of their traditional loyalties\\nand imparts knowledge of cultural practices. Nativism thus gives an ethno-\\ncentric cast to American national identity. Here, nationality does not simply\\ntrump one’s ethnicity; rather, nationality fuses with a particular Anglo-\\nSaxon ethnicity. \\nScholars disagree about the extent and the voluntary nature of cultural\\nassimilation of immigrants to America (Gerstle, 1997; Hollinger, 1997;\\nGordon, 1964; Glazer & Moynihan, 1959). While the passage of the 1964\\nCivil Rights Act and the 1965 Immigration and Nationality Act marked the\\nascendancy of the liberal conception of citizenship, in previous decades na-\\ntivism and racism strongly influenced public policy. Gerstle (1997, p. 556)\\nargues that government support for nativism and the end of mass immigra-\\ntion suppressed the hyphenated American identities that had thrived earlier\\nand narrowed the range of acceptable cultural behavior. Even so, nativism\\nnever entirely succeeded (Glazer & Moynihan, 1959); every immigrant did\\nnot truly “melt” into a single, unchanging political culture. \\nMulticulturalism is an alternative ideological response to the presence of\\nethnic diversity in America. The “soft” or “liberal” version of this concep-\\ntion of national identity emphasizes the virtues of cultural pluralism and\\ncalls on the government to help ensure tolerance toward minorities (Miller,\\n1995; Appiah, 1994; Raz, 1994). “Hard” or “radical” multiculturalism goes\\nfurther, conceiving of the nation as a confederation of ethnic groups with\\nequal rights and construing ethnicity as the preferred basis of one’s political\\nidentity (Citrin, Sears, Muste, & Wong, forthcoming; Sears, Citrin, Chele-\\nden, & van Laar, 1999). \\nUnlike liberalism, then, multiculturalism assumes that one’s ethnic iden-\\ntity is essential to one’s personal dignity and self-realization (Raz, 1994).\\nThis conception of national identity does hold that differences among com-\\nmunities of descent are basic and persisting. Since no culture is superior to\\nany other, none should be privileged in a multiethnic polity (Young, 1990;\\nGoldberg, 1994). T\\no assure that minority cultures survive and flourish—\\nsomething that is viewed as indispensable for the well-being of their mem-\\nbers—“hard” multiculturalism therefore demands group rights and govern-\\nment efforts to preserve minority cultures (Okin, 1997). Without such\\nsupport, it is argued, the standing of minority ethnic groups inevitably will\\ndecline in the face of the cultural and economic power of the numerical \\nmajority, and members of minority ethnic groups will lose the core of their\\nsocial identity. \\nMulticulturalism emphasizes ethnic consciousness and pride, but pro-\\nvides no apparent basis for social solidarity among America’s ethnic groups\\nsave the principle of mutual tolerance (Raz, 1994). Because of this, critics\\n(Schlesinger, 1992; Kateb, 1994) warn that elevating the psychological\\nsalience of ethnic identity, as Yugoslavia’s tragic recent history illustrates,\\nwill increase group conflict or make it difficult to mobilize citizens on be-\\nhalf of national goals. In the vocabulary of social identity theory, they claim\\nThe Meaning of American National Identity\\n77\\n\\n\\nthat legitimating ingroup favoritism will lead inexorably to hostility and ag-\\ngression toward outgroups, including other ethnic groups within the\\nAmerican national community.\\nResearch Questions and Method\\nDiscussion about the consequences of changing the balance between na-\\ntional and ethnic identities often proceeds in an empirical vacuum, without\\nmuch reference to the prevailing attitudes and beliefs of the American pub-\\nlic. This chapter employs survey data to present evidence about the follow-\\ning questions: What is the level of patriotism in the United States, and does\\nit differ substantially across ethnic groups? Do white, black, Hispanic, and\\nAsian Americans differ in how they order their national and ethnic identi-\\nties or in how they conceive of membership in the national community?\\nAnd does one’s particular conception of national identity influence opinions\\nabout government policies?\\nPsychological theories regarding intergroup relations differ in their pre-\\ndictions about the preeminence of ethnicity in how people define themselves\\npolitically (Sidanius, Levin, Rabinowitz, & Frederico, 1999; Sears et al.,\\n1999). Our own expectations are guided by symbolic politics theory (Sears,\\n1993; Sears et al., 1999). This approach assumes that individuals possess sta-\\nble predispositions established in early socialization (e.g., through parents or\\nmass media) and reinforced by later experience (e.g., through education or\\nwork). Ethnicity has no inherent priority in identify formation. Ethnic iden-\\ntification is one potentially significant predisposition, national attachment \\nis another. The relative strength of these identities varies according to one’s\\nsocialization experiences, and their influence on current behavior depends\\nupon the nature of environmental cues. For example, feelings of national\\nidentity are likely to be evoked by international terrorist activities because\\nthese raise the salience of one’s status as an American, but not by proposals to\\nincrease the tax on tobacco.\\nBy emphasizing the role of social learning, the symbolic politics perspec-\\ntive is compatible with either ethnic antagonism or harmony in multiethnic\\nsocieties. In the American context, it appears that national pride and a sense\\nof the country’s exceptionalism are strongly socialized attitudes, and this \\nsocialization may mitigate ethnic conflict by integrating diverse groups in an\\noverarching identity (Horowitz, 1985). All the country’s ethnic groups are\\nexposed to patriotic norms and the country’s liberal political tradition (Hartz,\\n1955; Lipset, 1996) that transmits belief in individual rights, not group\\nrights. The potency of these cultural messages would predict only minor \\nethnic differences in national identity and support for multiculturalism. \\nSymbolic politics theory conceives of national and ethnic identities as\\nstructures of valenced beliefs stored in memory (Lau & Sears, 1986). Which\\nof a person’s political predispositions influences her political choices partly\\n78\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\ndepends on the nature of the stimuli she encounters, such as events, poli-\\ncies, and candidates’ messages (Zaller, 1992). When a particular attitude\\nsuch as national pride is evoked, the theoretically predicted response is to\\nact in a way that reinforces one’s preexisting orientation. Accordingly, sub-\\njective conceptions of national identity are hypothesized to govern opinions\\nabout specific issues, such as foreign policy, immigration, or teaching\\nAmerican history, which evoke symbols of nationhood, but not about “neu-\\ntral” issues, such as protecting endangered species.\\nThe survey method relies on self-reports as the basis for classifying social\\nidentities. The main body of evidence here comes from the 1996 General\\nSocial Survey (GSS) of the American public aged 18 and older conducted\\nby the National Opinion Research Center (Davis & Smith, 1997). This sur-\\nvey employed a split-sample design and the results are based on the 1,367\\nrespondents of the total sample of 2,094 who were asked questions about\\nnational identity. In part because the 1996 GSS national survey included\\nvery few Hispanic and Asian respondents, comparable evidence from the\\n1994 and 1995 Los Angeles County Social Surveys (LACSS), conducted by\\nUCLA’s Institute for Social Science Research (Sears et al., 1999) also is re-\\nported. The Los Angeles data have special interest since they delineate the\\ndegree of ethnic consensus and conflict in an extraordinarily diverse local\\ncommunity where the salience of multiculturalism is a political constant.\\nThe 1994 LACSS sample included 279 whites, 231 blacks, 264 Latinos, and\\n47 Asian respondents, and the corresponding numbers in the 1995 sample\\nwere 259, 71, 202, and 47.\\nThe 1996 GSS included a variety of items regarding beliefs and feelings\\nabout America. From this set the analyses in this chapter use items covering\\nthe different aspects of patriotism and chauvinism, as well as other ques-\\ntions assessing subjective beliefs about the normative criteria for American\\nnationality. Other questions about the respondent’s level of pride in distinct\\nfeatures of American society and history were used to provide convergent\\nevidence regarding national attachment. In addition, questions concerning\\none’s political self-categorization and one’s level of identification with one’s\\nethnic group provided evidence about the balance between national and\\nethnic identities.\\nPublic Attachment to America\\nThe present survey evidence shows clearly that most Americans retain a\\nstrong, positive sense of national identity. T\\nable 4.1 presents the responses\\nof the 1996 GSS national sample, broken down by ethnicity, to an array of\\nrelevant items, with the relatively few “don’t know” answers excluded. (For\\nthe full wording of the questions, see Davis & Smith, 1997). T\\nable 4.1\\ngroups these questions to reflect the tripartite definition of national identity\\noutlined above and to indicate the composition of multi-item attitude\\nThe Meaning of American National Identity\\n79\\n\\n\\ntable 4.1 Beliefs about American National Identity (1996 GSS)\\nT\\notal\\nWhites\\nBlacks\\nHispanics\\n(1367)\\n(1018)\\n(156)\\n(61)\\nPolitical Self-Categorization\\nThink of self mainly as “just American” in\\nsocial and political issues*\\n90%\\n96%\\n66%\\n79%\\nAffective Attachment to America\\nPatriotism scale items:**\\nHow important is being an American to you?\\n(% saying ‘most important thing in their life’)\\n46\\n47\\n49\\n26\\nHow close do you feel to America? (% saying \\n‘close’ or ‘very close’)\\n81\\n83\\n69\\n83\\nAgree ‘I would rather be a citizen of America \\nthan any other country’\\n91\\n92\\n89\\n79\\nPride in Aspects of America\\n% saying ‘very proud’ or ‘somewhat proud’\\nThe way democracy works\\n83\\n85\\n77\\n81\\nIts political influence in the world\\n80\\n81\\n77\\n75\\nIts history\\n88\\n92\\n69\\n81\\nIts fair and equal treatment of all groups in \\nsociety\\n57\\n60\\n42\\n51\\nChauvinism scale items:\\nAgree ‘World would be a better place if people \\nfrom other countries were more like Americans’\\n40\\n40\\n47\\n47\\nAgree ‘America is a better country than most \\nother countries’\\n81\\n83\\n74\\n73\\nAgree ‘America should follow its own interests, \\neven if this leads to conflicts with other nations’\\n44\\n44\\n50\\n39\\nAgree ‘People should support their country even \\nwhen it’s in the wrong’\\n32\\n30\\n41\\n33\\nNormative Conceptions of American Identity\\n% Saying very or fairly important for a True \\nAmerican:\\nNativism scale items:\\nT\\no have been born in America\\n69\\n67\\n82\\n42\\nT\\no be a Christian\\n54\\n50\\n77\\n50\\nT\\no have lived in America for most of one’s life\\n73\\n71\\n84\\n77\\nAssimilationism scale items:\\nT\\no be able to speak English\\n93\\n94\\n91\\n85\\nT\\no feel American\\n87\\n89\\n81\\n72\\nT\\no have American citizenship\\n93\\n92\\n93\\n93\\nMulticulturalism\\nAgree ‘It is impossible for people who do not \\nshare American customs and traditions to \\nbecome fully American’\\n35\\n34\\n47\\n34\\nAgree ‘Ethnic minorities should be given \\ngovt assistance to preserve their customs \\nand traditions’\\n17\\n11\\n44\\n45\\nThink it is better if groups adapt and blend into \\nlarger society rather than maintain their \\ndistinct customs and traditions\\n58\\n61\\n56\\n50\\n*This question was not asked in the 1996 GSS.  The data we report is from the 1994 GSS.\\n**The items listed under Patriotism, as well as those under the headings Chauvinism, Assimilationism, and\\nNativism, are the items that were used in constructing these scales for the analyses appearing in other tables.\\n\\n\\nscales. The Political Self-Categorization item asks respondents to consider\\nwhich identity is primary to them when they think about political and social\\nissues. The Patriotism and Chauvinism items assess the affective dimension\\nof national identity. These items are similar in content to other measures of\\npsychological attachment to the nation (Kosterman & Feshbach, 1989;\\nStaub, 1997). Another set of “pride” questions taps the respondent’s sense\\nof pride in regards to specific aspects of American society. The “true Ameri-\\ncan” questions capture normative conceptions of national identity and are\\nmodeled on an earlier Traditional Americanism scale (Citrin, Reingold, &\\nGreen, 1990). In addition, several items assess support for the maintenance\\nof diverse cultures within the political community, a core principle of the\\nmulticulturalist conception of American identity. \\nClearly, table 4.1 shows that patriotic sentiment remains pervasive in the\\nAmerican public. One poll is just a snapshot in time, of course, and the level\\nof positive affect one records will vary with the response categories used to\\nmeasure the underlying attitude. Nevertheless, the consistency in expres-\\nsions of pride and belonging is impressive. When asked to rate the impor-\\ntance of being an American on a 10-point scale, fully 46 percent of the 1996\\nGSS sample said it was “the most important thing in their life.” In response\\nto similar questions, 81 percent said they felt “close to America” and 91 per-\\ncent agreed that they would “rather be a citizen of America than any other\\ncountry.” Responses to the items in the “Pride” subheading in table 4.1\\nshow that this generalized feeling of attachment carries over to judgments\\nabout most specific features of national experience. For example, 83 percent\\nof the 1996 GSS sample expressed pride in “the way democracy works in\\nAmerica.” However, many fewer, 57 percent, said they were proud of the\\ncountry’s record in treating all groups in society equally. Clearly, one can\\nhave a strong sense of patriotism without believing that the nation is per-\\nfect. \\nNor does everyone extend pride in America to a conviction in the coun-\\ntry’s superiority. While 81 percent of the 1996 national sample did agree\\nthat America is a better country than most others, just 40 percent thought\\nthe world would be a better place if people from other nations were more\\nlike Americans. Only 32 percent agreed that one should support their coun-\\ntry ”even when it is wrong,” compared to 50 percent who disagreed. Chau-\\nvinism, in the sense of a latent hostility toward other nations, is present, but\\nthe dominant outlook seems to be just a preference for one’s own people\\nand place. \\nThe pattern of answers to the questions about what is important for \\nbeing “truly American” reinforces this conclusion. Each of the attributes\\nlisted was deemed either “very” or “fairly” important, suggesting that most\\nAmericans believe that there is a distinct national identity. The more inclu-\\nsive political criteria—American citizenship (87 percent) and simply feeling\\nAmerican (93 percent)—were chosen more often than the ascriptive quali-\\nties of being native born (69 percent) or Christian (54 percent). The ability\\nThe Meaning of American National Identity\\n81\\n\\n\\nto speak English (93 percent) might appear as an ethnocentric criterion, but\\nthis is an achieved rather than an ascribed trait and research indicates that\\nwithin two generations most immigrants do learn the country’s common\\nlanguage (Portes, 1996). In fact, language minorities and foreign-born resi-\\ndents are as likely as white or black Americans to emphasize the linguistic\\ncriterion for national identity (Citrin et al., 1990). \\nMulticulturalism asserts the need to maintain minority cultures within\\nAmerica and regards the symbols of assimilation with suspicion, if not out-\\nright hostility. The 1996 GSS asked respondents whether they believed it is\\n“better for a country if different racial or ethnic groups maintain their dis-\\ntinct customs and traditions” or if they “adapt and blend into the larger so-\\nciety.” Of those who expressed a definite opinion, 58 percent said that\\n“blending in” would be better. (In the case of this question, however, fully\\n27 percent of the respondents did not express a definite opinion, suggesting\\nsubstantial ambivalence about the choice as posed. When the respondents\\nwithout a definite opinion are included, 43 percent of the total sample are\\nin favor of “blending in,” while 31 percent favor the maintenance of cultural\\ndiversity). Even fewer, only 17 percent of the GSS sample, agreed with the\\nposition of many advocates of multiculturalism that ethnic minorities\\nshould receive government support to preserve their traditions.\\nAt the same time, there was considerable confidence about the possibility\\nof assimilation. More respondents (42 percent) disagreed than concurred\\n(35 percent) with the proposition that “it is impossible for people who do\\nnot share American customs and traditions to become fully American.” Per-\\nhaps because of this optimism, many Americans may seem to feel that as-\\nsimilation and maintaining connections to one’s ethnic heritage are mutu-\\nally compatible. Public opinion endorses the importance of a common\\nnational identity without insisting on a forced march to cultural conformity.\\nFor example, a 1994 GSS poll found that only 26 percent of the public said\\nthat ”ethnic history is getting too much attention in the public schools,” an-\\nother indication of the tendency to view the existence of a unifying Ameri-\\ncan culture as compatible with respect for pluralism (Merelman, Streich, &\\nMartin, 1998). \\nUsing exploratory and confirmatory factor analysis, we combined the\\nitems listed in table 4.1 to construct Patriotism (alpha = .64) and Chauvinism\\n(alpha = .66) indices and summary measures of Nativist (alpha = .79) and \\nAssimilationist (alpha = .64) conceptions of Americanism. Although there is\\nstatistical evidence for treating these dimensions as distinct, clearly they are\\ninterrelated. Nativism and Assimilationism correlate .58 (Pearson’s r) in the\\n1996 GSS sample, and Patriotism and Chauvinism also are associated \\n(r = .37). An interesting note is that Assimilationism was more strongly re-\\nlated to Patriotism (r = .47) than to Chauvinism (r = .38), whereas the obverse\\nwas true for Nativism (r = .29 and .45, respectively), a pattern pointing to a\\n“psycho-logical” bond between an exclusionary definition of American iden-\\ntity and a sense of superiority to other peoples. However, beliefs about the\\n82\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nvalue of cultural diversity are not strongly related to these variants of positive\\nattachment to the nation. Responses to the question about whether or not\\ngroups should blend into the mainstream culture correlated positively, but\\nweakly, with the Patriotism (.13), Chauvinism (.13), Nativism (.07), and As-\\nsimilationism (.13) scales. Admittedly, these correlations may be attenuated\\nowing to the use of just a single-item to measure support for assimilation into\\nthe mainstream culture and due to the low variance on the scales. \\nNormative Consensus or Ethnic Conflict? \\nAlthough strong feelings of national attachment are the norm, the possi-\\nbility of ethnic conflict in outlook remains. The nativist prediction that \\ndiversity erodes national unity assumes that minority groups are less com-\\nmitted to the idea of a common American identity. The multiculturalist\\nperspective yields a similar expectation. If ethnicity is the dominant crite-\\nrion of one’s social identity, then minority group members should be less\\nlikely than those in the dominant ethnic group to view the entire nation as\\n“theirs” and, therefore, to be less likely to express the usual forms of pride\\nand affection. \\nDespite frequent claims that there is a vast racial divide in American\\nopinion (Kinder & Sanders, 1996), our surveys indicate that ethnic differ-\\nences are slight when it comes to patriotic sentiment (table 4.1). There is a\\nstrong similarity in the outlook of blacks and whites regarding the personal\\nimportance of American citizenship. Blacks (69 percent) are somewhat less\\nlikely to say they feel “very close” or “close” to America than whites (83\\npercent) or Hispanics (83 percent), but, again, the overwhelming majority\\nof all three groups express a strong sense of attachment to the nation. In the\\n1994 LACSS, 79 percent of the whites and 73 percent of the black respon-\\ndents said they were “extremely” or “very” proud to be an American.\\nThe opinions of Hispanics point to a continuing connection to their\\ncountries of origin among recent immigrants. As table 4.1 shows, they are\\nless likely than either blacks or whites to say being an American is impor-\\ntant to them, more willing to say they would move to another country, and\\nless likely to say that being born in America or feeling American is an im-\\nportant characteristic of being a true American. The same general pattern\\nexists among the Hispanic and Asian respondents in the Los Angeles data\\n(reported in full in Sears et al., 1999). Only 44 percent of the Hispanic re-\\nspondents and 58 percent of the Asians in the Los Angeles study felt “ex-\\ntremely” or “very” proud to be an American, but these differences from\\nwhite opinion are largely because many of these respondents are recent ar-\\nrivals in the country and are not yet citizens. The findings of the Latino\\nNational Political Survey (de la Garza, Falcon, & Garcia, 1996, p. 346) also\\nindicate that after controlling for demographic characteristics, including\\nlength of time in the United States, Mexican Americans “express patriotism\\nThe Meaning of American National Identity\\n83\\n\\n\\nat levels equal to or higher than do Anglos.” Somewhat surprisingly, the\\nLos Angeles survey found virtually no ethnic differences on the question of\\nwhether “it is better if groups change so they blend into the larger society\\nrather than maintain their distinct cultures.” \\nThe pervasive patriotic and assimilative tendencies among all ethnic\\ngroups frays, however, when one moves from the realm of ideals to judging\\nhistorical experience. In the 1996 GSS survey, as reported in table 4.1,\\nblacks and, to a lesser extent, Hispanics were less likely than whites to ex-\\npress pride in America’s history, in the way its democracy works, and in the\\ntreatment of all social groups. Members of minority groups also are less\\nlikely than whites to agree that America has provided “people of my ethnic\\ngroup a fair opportunity to get ahead in life” (LACSS) and more likely to\\nagree that the government should give special assistance to ethnic minori-\\nties (GSS). Minority groups, then, clearly display a heightened sensitivity to\\ntheir collective status, largely agreeing with whites about the goal of e\\npluribus unum, but, as previous research also found (Hochschild, 1995), \\nasserting more often that discrimination and unequal opportunities for \\nminority groups are ongoing problems in the United States.\\nThe normative conceptions of what it means to be a “true American”\\nalso are very similar in all ethnic groups. In this regard, blacks are somewhat\\nmore likely to endorse what we have termed ascriptive or nativist criteria of\\nidentity. They are significantly more likely (by the chi-square test) than ei-\\nther white or Hispanic respondents in the 1996 GSS to agree that being\\nborn in America, living in America for most of one’s life, and being a Chris-\\ntian are “very important” qualities of national identity. In an important\\nsense, these opinions express the visceral quality of their American roots for\\nthe black respondents. African Americans constitute the oldest group of\\nnonwhite immigrants in the country, despite the forcible nature of their ar-\\nrival. Thus, the symbolic meaning of their invocation of nativist criteria for\\nnational identity might well be a positive statement about their own “true\\nAmerican-ness” rather than prejudice toward outsiders. \\nSymbolic politics theory emphasizes the role of socialization in the for-\\nmation of social identities and would predict that common exposure to the\\npatriotic themes so dominant in American life would engender high levels\\nof national attachment across all ethnic groups over time. Of course, so-\\ncialization is ongoing and multifaceted, and there generally are group dif-\\nferences in prior learning and current experience that have an impact on\\npolitical predispositions (Kinder & Sears, 1985). T\\nable 4.2 summarizes the\\nethnic group differences in feelings of national identity by comparing their\\nmean scores on the multi-item measures of Patriotism, Chauvinism, Na-\\ntivism, and Assimilationism Indices, with each variable scored from 0 (low)\\nto 1 (high). The results show that in all three ethnic groups, positive attach-\\nment to American nationality prevails. Nevertheless, ANOVA indicates that\\nseveral of the ethnic differences that do appear around the highly favorable\\nnorm are statistically significant. Blacks do have lower mean scores than\\n84\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nwhites on Patriotism, though higher scores on Nativism. Hispanics also\\nhave lower Patriotism scores than whites, as well as less support for the As-\\nsimilationist conception of national identity. Whether one should empha-\\nsize the element of consensus or these minor and spotty differences is a\\nmatter of debate (compare Sidanius, this volume). But whatever one makes\\nof the divergences among whites, blacks, and Hispanics, clearly the glass of\\npatriotism is much more than half full in every group. The real ethnic di-\\nvide in American politics emerges on specific policies bearing directly on\\nthe status of minority groups, not on conceptions of national identity. \\nThe results of a multiple regression analysis in which race, ethnicity,\\ngender, age, income, education, region, and immigrant generation were\\nemployed as predictors of the Patriotism, Chauvinism, Nativism, and As-\\nsimilationism scales, respectively, confirm this conclusion. (For space rea-\\nsons, the full results are not reported here, but will be provided on request.)\\nWith the imposition of statistical controls, there is no “race effect” on Pa-\\ntriotism or Chauvinism scores. Blacks remain significantly more likely than\\nwhites to endorse the nativist conception of American identity, largely be-\\ncause of their greater support for Christianity as a legitimate defining crite-\\nrion of nationality. This suggests that another social identity, religion, is im-\\nportant to understanding ethnicity, national identity, and intergroup\\nconflict in America. \\nMore generally, there is a striking similarity in the demographic under-\\npinnings of the four measures of national attachment: the elderly, those\\nwith lower levels of education, and those residing in the South were signifi-\\ncantly more likely to have high, “pro-America” scale scores. Finally, first-\\ngeneration immigrants were significantly less likely to have high scores on\\nthe Patriotism, Nativism, and Assimilationism indices. However, neither\\nsecond- nor third-generation Americans differ from those with an even\\nThe Meaning of American National Identity\\n85\\ntable 4.2 Ethnic Differences in National Identity1 (1996 GSS)\\nMean Score on a 0 to 1 Scale\\nT\\notal\\nWhites\\nBlacks\\nHispanics\\nAffective attachment to America\\nPatriotism\\n.80\\n.81\\n.76\\n.75\\nChauvinism\\n.59\\n.59\\n.62\\n.56\\nNormative conceptions of American identity\\nAssimilationism\\n.86\\n.86\\n.85\\n.79\\nNativism\\n.64\\n.62\\n.79\\n.60\\n1. The four attitude measures are additive indices created by summing responses to the several groups of\\nitems designated in table 4.1.  The response options for each item were first recoded to range from zero to\\none.  After summation, index scores were also recoded to range from zero to one with higher scores indicat-\\ning higher levels of patriotism, chauvinism, etc.  Reliabilities (Chronbach’s alpha) for the respective scales\\nare patriotism (.64), chauvinism (.66), assimilationism (.64), and nativism (.79).\\n\\n\\nlonger family history in the country, once again suggesting that socializa-\\ntion into a sense of national pride and belief in a distinctive American iden-\\ntity does not take long. \\nNational, Hyphenated, and Ethnic Identities\\nSince citizens can have multiple loyalties (or identities), it is important to\\nassess the intensity of ethnic and national identifications separately.\\nWhether these attachments are complementary or competitive is an em-\\npirical matter. The 1996 GSS asked respondents separate questions about\\nhow close they felt to the United States and to their own racial or ethnic\\ngroup. A cross-tabulation of answers to these items show that 63 percent of\\nthe white, 65 percent of the black, and 66 percent of the Hispanic respon-\\ndents expressed feelings of closeness to both the nation and their own ethnic\\ngroup. Among whites, only 11 percent felt close attachment to their own\\nethnic group, but not to the country as a whole, whereas 24 percent of the\\nblack and 14 percent of the Hispanic respondents fell into this category.\\nThis suggests a stronger salience of an ethnic identity among blacks, with\\nthe possible implication that they would be more likely than whites to \\nprivilege ethnicity over nationality when trade-offs must be made. Still, the\\ndominant result is that a large majority among both races expresses af-\\nfinity to both identity groups. Subjectively, national and ethnic identities in\\nthe United States tend to be experienced as complementary rather than\\ncompeting.\\nT\\nable 4.1 reports the responses to the political self-categorization ques-\\ntion: “When you think of social and political issues, do you think of yourself\\nmainly as a particular ethnic, racial, or nationality group, or do you think of\\nyourself mainly as just an American?” (The data come from the 1994, not\\nthe 1996 GSS poll, which did not include the item). This question wording\\ncalls for a ranking of identities and explicitly places the choice in a political\\ncontext. Of the white respondents, 96 percent answered “just an American,”\\ncompared to 66 percent of the black and 79 percent of the Hispanic respon-\\ndents (see table 4.1). The uniformity of opinion among whites is consistent\\nwith the view of the various groups of European immigrants having\\n“melted,” so that their most salient political identity is no longer cultural or\\nreligious, but simply a superordinate identity as Americans (Alba, 1990;\\nWaters, 1990; Roediger, 1994). \\nThe 1994 and 1995 LACSS studies asked a similar self-categorization\\nquestion with equivalent results. In the pooled samples, 93 percent of the\\nwhite respondents gave the overarching “just an American” self-designa-\\ntion. Blacks (69 percent), Hispanics (70 percent), and Asians (71 percent)\\nwere significantly less likely (by the chi-square test) than whites to say they\\nthought of themselves as just Americans. When forced to choose, however,\\na large majority in all three minority ethnic groups opted for a common na-\\n86\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\ntional identity, not ethnic particularity, as their first choice. Moreover,\\namong Hispanics and Asians, those who categorized themselves in terms of\\nethnicity tended to be first-generation immigrants, suggesting, once again,\\nthat the absorption of a sense of national identity is the normal outcome of\\nsocialization in this country. \\nThe 1995 LACSS allowed respondents to say whether they identified as\\njust Americans or as both Americans and as members of an ethnic, racial, or\\nnationality group. As reported in table 4.3, only 17 percent of the whites\\nopted for a dual or hyphenated identity. But, when given this opportunity to\\nchoose, a majority in all three minority groups preferred to categorize\\nthemselves as both Americans and members of an ethnic group and not as\\nexclusively “American.” Unfortunately, there are no national data to repli-\\ncate this result, but it is consistent with other small-scale studies testing so-\\ncial identity theory that suggest the relatively greater salience of an ethnic\\nidentity for members of smaller, easily demarcated, homogeneous, and \\ndisadvantaged groups (Prentice & Miller, 1999; Brewer, this volume). The\\nhistorically dominant group in a multiethnic state often fuses rather than\\nhyphenates their ethnic and national identities (Connor, 1993). It is beyond\\nthe scope of this chapter to explore the antecedents of these alternative self-\\ncategorizations by minority group members. Among immigrant groups, it\\nappears that length of residence in the United States and speaking English\\ndiminish the likelihood of identifying oneself primarily in ethnic terms \\n(de la Garza et al., 1996), but these factors could not account for the varied\\nresponses of blacks. \\nIdentities and Policy Preferences\\nThe political relevance of social identities rests partly on how they influence\\nattitudes and behavior toward one’s own and other groups. A strong sense\\nof group identity is expected to promote conduct that favors one’s fellow\\ngroup members, even in a “minimal group” (T\\najfel, 1978). But since there\\nare multiple dimensions of self-identification, how one defines the “I” as\\n“we” when a political decision arises is likely to be significant. Symbolic\\npolitics theory holds that broad attitudes, including feelings of national and\\nethnic identity, are more likely to determine preferences on specific issues\\nwhen those predispositions are central to an individual’s self-definition and\\nwhen they are cued by stimuli associated with the objects of these attitudes\\n(Sears, 1993; Zaller, 1992). For example, a strong sense of Jewish identity is\\nmore likely to govern opinions about American policy toward Israel or a\\nmandatory school prayer than the future of the spotted owl. In addition, as\\nshown above, ethnic identity may indeed be more significant for the con-\\nstruction of the social self among some groups than others. \\nNationalism is an integrating ideology that overrides the claims of less\\ncomprehensive group loyalties. In fact, liberal theorists (T\\namir, 1997;\\nThe Meaning of American National Identity\\n87\\n\\n\\ntable 4.3 Political Self-Categorization: American, Dual, or Ethnic Identity (1995 LACSS)\\nWhites\\nBlacks\\nHispanics\\nAsians\\n(n = 254)\\n(70)\\n(174)\\n(46)\\nJust an American Identity\\n78%\\n16%\\n11%\\n4%\\nBoth American and ethnic identity\\n17\\n54\\n59\\n59\\nEthnic identity\\n5\\n30\\n31\\n37\\nWhites\\nBlacks\\nHispanics\\nAsians\\nUS born \\nNot US born \\nUS born \\nNot US born\\nUS born \\nNot US born \\nUS born \\nNot US born\\n(n = 223) \\n(31)\\n(68)\\nna\\n(59)\\n(115) \\n(8)\\n(38)\\nJust an American identity \\n80%\\n65%\\n16%\\nna\\n24%\\n4% \\n13%\\n3%\\nBoth American & ethnic identity \\n17\\n16\\n54\\nna\\n53\\n62 \\n63\\n58\\nEthnic identity \\n3\\n19\\n29\\nna\\n24\\n34 \\n25\\n40\\n\\n\\nHollinger, 1997; Miller, 1995) justify holding on to a national identity on\\nthe grounds that this sentiment generates feelings of special obligation to\\nand responsibility for the other members of one’s national community. If\\nthis surmise is correct, then people should be less likely to feel a sense of\\nduty to members of groups who define themselves as somehow outside or\\ndifferent from the national “family.” Accordingly, since multiculturalism\\nvalidates the maintenance of group differences, the stronger one’s attach-\\nment to an overarching American national identity, the more likely one\\nshould be to oppose recognition and support for groups emphasizing their\\ncultural distinctiveness. Alternatively, if the majority group views ethnic mi-\\nnorities as disadvantaged members of a community with a shared identity,\\nfeelings of national attachment could accommodate, even facilitate, support\\nfor policies aimed at helping them. \\nIn considering the political implications of ethnic consciousness, Tyler,\\nLind, Ohbuchi, and Sugawara (1998) found that assimilated and “bicultur-\\nalist” members of minority groups (that is, those who emphasize their\\nAmerican identity and dual identity, respectively) were equally likely to ex-\\npress trust in government, to regard existing institutions as fair, and to com-\\nply with government policy. Both groups were significantly more likely\\nthan “separatist” respondents from the same ethnic group (those emphasiz-\\ning their ethnic identity) to express these attitudes. The present data show\\nthat the relative salience of an ethnic as opposed to national self-identifica-\\ntion affects how people evaluate multiculturalist principles (see table 4.4).\\nMinority group members for whom ethnic identification takes priority over\\nnational identity generally are less likely than those who say they view poli-\\ntics mainly as an American to express pride in America and to agree that\\ntheir group has had a fair chance to get ahead, and are more likely to believe\\nthat groups should maintain their own distinct cultures, to deny that politi-\\ncal organizations based on ethnicity promote separatism, and to think that\\npeople are best represented by leaders with the same ethnic background.\\n(See table 4.4 for results of chi-square tests for statistical significance.) \\nT\\nable 4.4 also suggests that the association between an ethnic self-identi-\\nfication and support for multiculturalism is stronger and more systematic\\namong blacks than among either Hispanics or Asians. For example, among\\nblacks, 61 percent of those who identified themselves as “just American” felt\\nit is better for minority groups to blend into the larger society rather than\\nmaintain their distinct cultures, compared to only 25 percent of those who\\ndefined themselves primarily as members of an ethnic group. Among His-\\npanics, the equivalent figures were 60 percent and 54 percent, a far smaller\\ndifference, and among Asians, they were 57 percent and 48 percent, respec-\\ntively. The future trend in how the different minority groups balance their\\nethnic and national identities thus has clear implications for the pattern of\\nintergroup conflict over issues of cultural recognition and representation.\\nOrdinary least-squared regression analysis was employed to delineate the\\nconnections between the affective and normative dimensions of national\\nThe Meaning of American National Identity\\n89\\n\\n\\ntable 4.4 The Impact of Political Self-Categorization on Beliefs about the American National Community (1994 LACSS)\\nBlacks\\nHispanics\\nAsians\\nJust\\nEthnic\\nJust\\nEthnic\\nJust\\nEthnic\\nAmerican\\nIdentity\\nAmerican\\nIdentity\\nAmerican\\nIdentity\\n(n = 156)\\n(68)\\n(182)\\n(66)\\n(36)\\n(8)\\nFeel extremely or very proud to be an American\\n81%\\n54%\\n53%\\n22%\\n61%\\n33%\\nAgree that American society has provided people of my ethnic\\n50\\n22*\\n70\\n63\\n69\\n63\\ngroup a fair opportunity to get ahead in life\\nThink it is better if groups change so that they blend into the larger\\n61\\n25\\n60\\n54\\n57\\n48#\\nsociety (as in the idea of a melting pot) rather than maintain their\\ndistinct cultures1\\nAgree that political organizations based on race or ethnicity\\n70\\n57#\\n71\\n52*\\n70\\n38#\\npromote separatism and make it hard for all of us to live together\\nLeader background doesn’t matter for representation\\n68\\n40*\\n67\\n50*\\n60\\n38\\nChi-square test of differences between “just American” and “ethnic identity” respondents within each racial/ethnic group:\\n*p<.05\\n#p<.10\\n1The percentages in this row represent pooled data from the 1994 and 1995 LACSS.\\n\\n\\ntable 4.5 The Relationship between Attitudes T\\noward National Identity  and Policy Preferences (1996 GSS)\\nAffective Attachment\\nNormative Conceptions\\n(3)\\n(6)\\nPatriotism\\nAssimilationism\\n(1)\\n(2)\\n& Chauvinism\\n(4)\\n(5)\\n& Nativism\\nPatriotism\\nChauvinism\\n(patrsm) (chauv)\\nAssimilationism\\nNativism\\n(assim) (nativ)\\nPolicy Preferences\\nMulticulturalism\\n(A) Need to share American customs\\n.05\\n.38**\\n–.08\\n.42**\\n.17**\\n.27**\\n–.07\\n.29**\\n(B) Support melting pot as ideal\\n.20*\\n.24**\\n.18\\n.17\\n.25**\\n.01\\n.33**\\n–.10\\n(C) Reduce immigration levels\\n.09\\n.18**\\n.06\\n.17**\\n.26**\\n.17**\\n.21**\\n.09*\\n(D) No cultural benefit from immigrants\\n.04\\n.20**\\n–.01\\n.21**\\n.11**\\n.15**\\n–.01\\n.16**\\nRacial Policies\\n(E) Oppose affirmative action\\n.02\\n–.01\\n.01\\n.00\\n–.05\\n–.03\\n–.05\\n–.01\\n(F) Oppose govt. help for blacks\\n.06\\n.14*\\n.03\\n.13\\n.12*\\n.06\\n.08\\n.04\\n(G) Oppose more spending on blacks\\n.17\\n.06\\n.14\\n.00\\n–.01\\n.06\\n.03\\n.05\\nDomestic spending\\n(H) Oppose social spending\\n.03\\n.03\\n–.03\\n.05\\n.03\\n.00\\n.05\\n–.02\\nForeign Policy\\n(I) Support more spending on defense\\n.23*\\n.31**\\n.12\\n.25*\\n.19*\\n.13*\\n.10\\n.10\\n(J) Support protectionism\\n.18**\\n.42**\\n.06\\n.40**\\n.29**\\n.23**\\n.16**\\n.16**\\nEntries are unstandardized regression coefficients for national identity variables in equations with controls for race, ethnicity, age, education, income, gender, region (South), generation, \\nideology, and party identification.\\n*p<.05\\n**p<.01\\n\\n\\nidentity on the one hand and policy preferences on the other. The statistical\\nmodel estimates the effects of the Patriotism, Chauvinism, Assimilation,\\nand Nativism Scales, (columns 1, 2, 4, and 5 in table 4.5), respectively, on \\nissues related to cultural diversity (rows A–D), racial issues (rows E–G), \\ndomestic spending (row H), and foreign policy (rows I–J). The equations\\nincluded race, ethnicity, age, education, income, gender, region, immigrant\\ngeneration, partisan affiliation, and liberal-conservative self-identification\\nas control variables. Owing to the small number of Hispanic respondents in\\nthe 1996 GSS sample, this analysis is confined to whites and blacks. For\\nsimplicity of presentation, the table omits the effects for the control vari-\\nables and reports only the unstandardized regression coefficients for the\\nmeasures of national identity (Patriotism, Chauvinism, Assimilationism,\\nand Nativism). All variables were recoded to vary between 0 and 1, and the\\ndependent variables all are scored in the “nationalist” or “conservative” \\ndirection. Accordingly, the positive coefficient of .24 for the association \\nbetween Chauvinism and the Melting Pot item (column 2, row B) indicates\\nthat believing that America is superior to other countries is strongly related\\nto the idea that ethnic minorities in the United States should blend into the\\nmainstream culture. \\nT\\nable 4.5 also reports the results of models that simultaneously included\\nboth Patriotism and Nationalism Scales (column 3) and the Assimilationism\\nand Nativism Scales (column 6) as predictors. These equations were de-\\nsigned to provide a tentative estimate of the relative degree of association\\nbetween each of these affective orientations and normative conceptions of\\nnational identity. The results (which are reviewed by comparing the coeffi-\\ncients across columns on a row-by-row basis) do provide some evidence of\\ntension between a strong sense of national identity and multiculturalism.\\nAlthough the respondent’s race has no relationship to these multiculturalist\\npolicy positions in the multivariate model, three of the four indicators \\nof national identity predict support for the belief that it is impossible for\\npeople who do not share American customs to become fully American (row\\nA). With regard to the question asking whether ethnic and racial groups\\nshould blend into mainstream society, Patriotism, Chauvinism, and Assimi-\\nlationism had statistically significant relationships with giving the melting-\\npot response (row B). Immigration is the major source of ethnic diversity\\nand new claimants for cultural recognition in America, and Chauvinism,\\nAssimilationism, and Nativism, but not Patriotism, had statistically signifi-\\ncant relationships with support for reducing the level of immigration (row\\nC) and disagreement that immigrants have a positive cultural impact on\\nAmerica (row D). \\nWhen Patriotism and Chauvinism are included as predictors simultane-\\nously, the more aggressive outlook tends to have the stronger association\\nwith opposition to increased cultural diversity in the United States. And\\nwhen the effects of Assimilationism and Nativism are compared, it is the\\nlatter more exclusionary attitude that generally appears to have the stronger\\n92\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nnegative relationship to support for multiculturalism. An ethnocentric con-\\nception of American national identity, therefore, may intensify internal po-\\nlitical conflict by hardening resistance to the demands of cultural minorities\\nfor recognition and support. \\nT\\nable 4.5 also reports on opinions about government efforts targeted at\\nhelping blacks (rows E–G) and about spending on domestic social programs\\n(measured by a Social Spending Index combining beliefs about whether the\\ngovernment should spend more on health, education, welfare, and the envi-\\nronment—row H). The results are a virtual mirror image of the earlier\\nfindings about the factors influencing opinions about cultural diversity and\\nimmigration. In the case of the domestic policy issues, race is strongly asso-\\nciated with preferences, with blacks consistently more favorable toward\\ngovernmental activism. But none of the four measures of national identity\\nhas a statistically significant relationship with scores on the Social Spending\\nIndex, opinions about affirmative action, or attitudes about whether the\\ngovernment should spend more to help blacks. The only coefficients in the\\narray that attain conventional levels of statistical significance are the weak\\nassociations of Chauvinism and Assimilationism with opposition to special\\nefforts to improve the living conditions of blacks. The results of these\\nanalyses indicate that generalized feelings of national attachment neither\\nboost support for assisting one’s fellow citizens through government spend-\\ning nor enhance opposition to policies, such as affirmative action, aimed at\\nassisting blacks or other ethnic minorities. \\nThis may not be surprising, since the controversies over domestic spend-\\ning and racial policy tend not to be framed in terms that engage attitudes\\ntoward the nation’s integrity or power. As hypothesized above, the impact\\nof social identities (and other predispositions) on specific policy preferences\\nshould vary across political domains. Feelings of national attachment\\nshould be strongly engaged only when the specific issues clearly touch on\\nthe values of American power, sovereignty, or purity. The data presented in\\ntable 4.5 consistently support this proposition. In addition to their effects\\non beliefs about cultural diversity and immigration, all four measures of na-\\ntional attachment had statistically significant effects in predicting opinions\\nabout whether military spending should be increased and on scores on an\\nAmerican Protectionism Index (alpha = .65), constructed by summing an-\\nswers to items about limiting foreign imports, prohibiting foreigners from\\nbuying land in America, and requiring television stations to give preference\\nto American programs. \\nA final regression analysis, reported in table 4.6, explored potential dif-\\nferences in the connections of ethnic as opposed to national identities on\\nracial policy preferences. Here, responses to both the question “How close\\ndo you feel to the United States?” and the question “How close do you feel\\ntoward your own racial or ethnic group?” were included as predictors of\\nopinions regarding affirmative action, support for government efforts to\\nhelp blacks, and government spending for blacks. In this analysis, the vari-\\nThe Meaning of American National Identity\\n93\\n\\n\\nable coding and controls are identical to those reported for table 4.5. How-\\never, to distinguish between the impact of ethnic identifications among\\nwhites and blacks, these equations were estimated separately for whites and\\nblacks. (Because of the paucity of Hispanic respondents, the analysis could\\nnot be replicated for that group.) \\nT\\no highlight the results, table 4.6 reports only the coefficients for na-\\ntional attachment and ethnic closeness. Confirming the results reported\\nabove, positive affect for the nation as a whole has no impact one way or the\\nother on the racial policies queried. There are obviously significant racial\\ndifferences in policy preferences, but among blacks, feeling close to one’s\\nown ethnic group had no great additional influence on support for policies\\naimed at improving their collective status. The fear of some political com-\\nmentators that black racial identification leads to a highly distinctive politi-\\ncal outlook seems overstated. \\nAmong whites, however, a stronger sense of ethnic identity does have a\\nsignificant effect across the three policies; white identification seems consis-\\ntently to increase opposition to government policies designed to help\\nblacks. Given our concern about the impact of ethnic identification on in-\\ntergroup conflict, these results are sobering. “Identity politics,” by which\\nwe mean the tendency to judge issues and events in terms of how they affect\\nthe standing of one’s own group, can juxtapose one’s ethnic and national\\n94\\nThe Contribution of Ethnic and National Identities to Political Conflict\\ntable 4.6 National Attachment and Ethnic Identity as Predictors of Opinion \\non Race-T\\nargeted Policies (1996 GSS)\\nOppose\\nOppose Special\\nOppose More\\nAffirmative\\nGovt. Help\\nSpending on\\nAction\\nfor Blacks\\nBlacks\\nAmong White Respondents\\nClose to own ethnic group (close =1)\\n.07**\\n.05*\\n.08*\\n(white identification)\\nClose to the US\\n–.02\\n–.02\\n–.03\\n(national attachment)\\nSample size (n)\\n543\\n546\\n347\\nAmong Black Respondents\\nClose to own ethnic group (close =1)\\n.12\\n–.05\\n.09\\n(black identification)\\nClose to the US\\n–.11\\n.16\\n.01\\n(national attachment)\\nSample size (n)\\n72\\n74\\n50\\nEntries are unstandardized regression coefficients for national and ethnic identity variables in equations with\\ncontrols for age, education, income, gender, region (South), generation, ideology, and party identification.\\n*p<.05\\n**p<.01\\n\\n\\n“selves.” While minorities may be more likely to express a strong ethnic\\nidentity, counterintuitively, it appears that the effects of ethnic identification\\nare more prominent among whites. In a disturbing way, the results in table\\n4.6 are a reminder that the evocation of white consciousness, whether\\nthrough nativist mobilization or through reaction to the assertion of ethnic\\npride by other groups, can lead to divisiveness and prejudice. \\nConclusion\\nA large body of survey evidence confirms that a positive sense of national\\nidentity is pervasive in the United States, as is the belief that there is a\\nunique American identity. In terms of affective attachment, patriotism—the\\nsense that America is “best for me”—was more widespread than chauvin-\\nism—the sense that this country is inherently superior to all others. Signifi-\\ncantly, too, we found no evidence of a deep ethnic divide in feelings about\\nAmerican nationality. The slightly lower level of patriotism among blacks,\\nas a group, than whites is less compelling than the fact that the large major-\\nity of all ethnic groups express pride in and closeness to their country and\\nits symbols. In addition, there is a broad consensus among all ethnic groups\\nregarding the main criteria for American identity. \\nThe strong sense of attachment to an American identity among ethnic\\nminorities should calm fears about the consequences of the nation’s in-\\ncreased ethnic diversity. Nearly as much as whites, minority respondents in\\nour surveys rejected the idea of organizing political life along ethnic lines.\\nFurthermore, most of them do not express their primary political identity\\nin purely ethnic terms. \\nThese results do not fit easily into an image of public attitudes as driven\\nprimarily by interest-based conflict between dominant and subordinate eth-\\nnic groups. The symbolic politics perspective outlined here presumes that\\nattitudes and values are acquired through a process of reinforcement, and\\nthat the role of purely instrumental motivation often is secondary (Sears &\\nFunk, 1990). The survey evidence suggests that members of all ethnic\\ngroups may be attracted to an American identity because of its traditional\\nliberal premises. In both California and national surveys (Citrin et al., 1990,\\n1994), the criteria of a “true American” chosen as important most often by\\nall four major ethnic groups were egalitarianism and tolerance (“treating\\npeople of all races equally” and “respecting people’s freedom to say what\\nthey want”).\\nThe staying power of this liberal conception of American identity may\\nhelp explain why many minority respondents viewed attachment to the\\ncountry and a sense of closeness to their own ethnic group as complemen-\\ntary rather than competitive. The results presented above show that mi-\\nnority group members frequently conceive of themselves in dual terms,\\ndefining themselves as both Americans and part of a distinctive racial or\\nThe Meaning of American National Identity\\n95\\n\\n\\nethnic group. Whites, by contrast, generally see themselves as “just Ameri-\\ncans”; for them, national and ethnic identities are merged. What is critical,\\nboth psychologically and politically, is how the dual identity—national and\\nethnic—is accommodated intrapersonally. That is, when does the con-\\nsciousness of being simultaneously American and black lead to a sense of\\nambivalence that pits the self against either the nation or the group, and\\nhow is this conflict among loyalties resolved? The present data reinforce\\nthe familiar point that blacks identify with American ideals, not American\\nrealities. As James Jones suggests (personal communication), their dual\\nidentity reflects recognition of the persistence of what Myrdal (1944) called\\nthe “American dilemma.”\\nIn the political domain, a strong sense of national identity consistently\\nwas related to opposition to liberal immigration policies and official support\\nfor maintaining minority cultures. There is, then, an ideological tension be-\\ntween multiculturalism and nationalism. Strong feelings of national attach-\\nment also were associated with approval for increasing America’s military\\nstrength, as well as a willingness to take quite drastic measures to reduce the\\npotential impact of foreign interests and influences. These issues, of course,\\nfocus on the boundaries between citizens and strangers, a distinction that is\\nless relevant for opinion formation when it comes to domestic spending or\\nracial policies. \\nThe strength of national attachment was unrelated to how whites felt\\nabout policies targeted at helping blacks, their fellow-citizens. However, a\\nstrong ethnic identification among whites was associated with greater an-\\ntagonism to affirmative action and more government spending to assist\\nblacks. In this instance, ingroup favoritism does seem to lead to greater \\ninter-group conflict. \\nThe portrait of normative consensus among the general public seems\\nstrangely discordant with the seeming prevalence of ethnic competition in\\nAmerican political life. The division of political offices, government con-\\ntracts, and other benefits among ethnic groups is closely watched by ac-\\ntivists. Racial issues deeply divide the country’s two political parties. Given\\nall this and more, what accounts for the failure of the surveys to detect in-\\ntense disagreements between the “majority” and “minority” groups when it\\ncomes to attitudes about national identity?\\nOne answer is that the existence of a gulf between the views of political\\nactivists and ordinary citizens is not uncommon. Another possibility is that\\nthe current approach to measuring national and ethnic identifications is in-\\nadequate. Surveys may pay too much attention to abstract norms and too\\nlittle to responses to distributive conflicts or actual social encounters among\\nmembers of different groups. In this regard, researchers should frame issues\\nin a way that asks respondents to make trade-offs among their multiple\\nidentities when it comes to allocating costs and benefits. This would enable\\na more comprehensive assessment of intragroup, as well as intergroup, con-\\nflict in how these choices are made. \\n96\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nAnother important issue concerns the enduring quality of the tendency\\nof recent immigrants to prefer a dual or hyphenated political identity. The\\npresent data indicate that a developing sense of patriotism and a sense of\\nidentity as an American are part of the process of acculturation that His-\\npanic and Asian groups are experiencing. More generally, today’s immi-\\ngrants, like their predecessors, become “desocialized from their native cus-\\ntoms over time, even if the emotional significance of attachment to the\\ngroup persists” (Glazer & Moynihan, 1975, p. 8). Because of this continu-\\ning affective tie, appeals to ethnicity still can be an effective strategy for mo-\\nbilizing political support for the pursuit of group or individual interests\\n(Horowitz, 1985). The utility of ethnicity in political combat, then, may \\nreinforce the tendency of political leaders to emphasize the salience of this\\naspect of identity and strive to sustain its psychological significance within\\ntheir ethnic group. \\nIn addressing the unresolved issue concerning the conditions under\\nwhich hostility toward outgroups accompanies ingroup pride, Gaertner,\\nDovidio, Nier, Ward, & Banker (1999) argue for the efficacy of a superordi-\\nnate common identity in mitigating intergroup conflict. Applied to the\\nlarger realm of American national politics, the question posed by social\\nidentity theory becomes whether an enhanced sense of ethnic identification\\namong minorities leads toward satisfactory conflict resolutions or toward\\nentrenched Balkanization. Clearly, the mobilization of group consciousness\\namong the disadvantaged can force movement toward less inequality and\\nmore social justice. On the other hand, the enhanced expression of ethnic\\nidentification in one group tends to be emulated, often provoking resistance\\nto change and the hardening of group boundaries. Whatever the direction\\nof future change in public attitudes, the evidence summarized here points to\\nthe advantages of a common sense of American identity founded on the \\nrealization of equal citizenship. \\nReferences\\nAlba, R. D. 1990. Ethnic identity: The transformation of white America. New Haven:\\nYale University Press.\\nAppiah, K. A. (1994). Identity against culture: Understandings of multiculturalism.\\nBerkeley, CA: Doreen B. T\\nownsend Center for the Humanities.\\nAshmore, R. D., & Jussim, L. (1997). Self and identity: Fundamental issues. New York:\\nOxford University Press. \\nBarber, B. (1995). Jihad vs. McWorld. New York: Times Books.\\nBar-tal, D. (1993). Patriotism as fundamental beliefs of group members. Politics and\\nthe Individual, 3, 45–62.\\nBrimelow, P. (1995). Alien nation. New York: Random House.\\nCitrin, J., Sears D., Muste C., & Wong, C. (Forthcoming). Multiculturalism in\\nAmerican public opinion. British Journal of Political Science.\\nCitrin J., Haas E. B., Muste C., & Reingold B. (1994). Is American nationalism\\nThe Meaning of American National Identity\\n97\\n\\n\\nchanging?: Implications for foreign policy. International Studies Quarterly, 38,\\n1–31.\\nCitrin J., Reingold B. & Green. D. P. (1990). American identity and the politics of\\nethnic change. Journal of Politics, 52, 1124–1154.\\nConnor, W. (1993). Beyond reason: The nature of the ethno-national bond. Ethnic\\nand Racial Studies, 163, 373–389.\\nDavis J., & Smith. T. (1997). The general social survey codebook. Chicago: National\\nOpinion Research Center.\\nde la Garza, R. O., Falcon A., Garcia C. (1996). Will the real Americans please stand\\nup: Anglo and Mexican-American support of core American political values.\\nAmerican Journal of Political Science, 40, (2), 335–350.\\nFarley, R. (1996). The new American reality : Who we are, how we got here, where we are\\ngoing. New York: Russell Sage Foundation.\\nGaertner, S. L., Dovidio J.F., Nier J. A., Ward C. M., & Banker, B. S. (1999). Across\\ncultural divides: The value of superordinate identity. In D. Prentice & D. Miller.\\n(Eds.), Cultural divides: The social psychology of cultural contact (pp. 173–212). New\\nYork: Russell Sage Foundation.\\nGerstle, G. (1997). Liberty, coercion, and the making of Americans.” Journal of\\nAmerican History, 84, 524–558.\\nGlazer, N. & Moynihan D. P. (1963). Beyond the melting pot. Cambridge: MIT Press.\\nGlazer, N., & Moynihan. D. P. (1975). Ethnicity: Theory and experience. Cambridge,\\nMA: Harvard University Press.\\nGleason, P. (1980). American identity and Americanization. In S. Thernstrom, A.\\nOrlov, & O. Handlin (Eds.), Harvard Encyclopedia of American Ethnic Groups\\n(pp 31–58). Cambridge, MA: Harvard University Press. \\nGoldberg, D. T. (1994). Introduction: Multicultural conditions. In D. T. Goldberg,\\n(Ed.), Multiculturalism: A critical reader (pp. 1–41). Cambridge, MA: Blackwell.\\nGordon, M. (1964). Assimilation in American life: The role of race, religion, and national\\norigins. New York: Oxford University Press.\\nGreenfeld, L. (1992). Nationalism : Five roads to modernity. Cambridge, MA: Harvard\\nUniversity Press.\\nGreenfield, L., & Chiriot, D. (1994). Nationalism and aggression. Theory and Soci-\\nety, 23, 79–130.\\nHartz, L. (1955). The liberal tradition in America: An interpretation of American political\\nthought since the revolution. New York: Harcourt, Brace.\\nHigham, J. (1988). Strangers in the land: Patterns of American nativism, 1860–1925.\\n2nd ed. New Brunswick, NJ: Rutgers University Press.\\nHochschild, J. L. (1995). Facing up to the American dream: Race, class, and the soul of the\\nnation. Princeton, NJ: Princeton University Press. \\nHollinger, D. (1995). Postethnic America. New York: Basic Books.\\nHollinger, D. A. (1997). National solidarity at the end of the twentieth century: \\nReflections on the United States and liberal nationalism. Journal of American \\nHistory, 84, 559–569.\\nHorowitz, D. (1985). Ethnic groups in conflict. Berkeley: University of California\\nPress.\\nHuntington, S. P. (1981). American politics: The promise of disharmony. Cambridge,\\nMA: Belknap Press.\\nKateb, G. (1994). Notes on pluralism. Social Research, 61, 511–537. \\n98\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nKelman, H. (1997). Nationalism, patriotism, and national identity: Social-psycho-\\nlogical dimensions. In D. Bar-tal & E. Staub (Eds.), Patriotism (pp. 165–189).\\nChicago: Nelson-Hall.\\nKinder, D. R., & Sanders, L. M. (1996). Divided by color: Racial politics and democratic\\nideals. Chicago: University of Chicago Press.\\nKinder, D., & Sears, D. O. (1985). Public opinion and political action. In \\nG. Lindzey & E. Aronson (Eds.), Handbook of Social Psychology. 3rd ed., vol 2., \\n(pp. 659-741) Reading, MA: Random House.\\nKosterman, R., & Feshbach, S. (1989). T\\nowards a measure of patriotic and national-\\nistic attitudes. Political Psychology, 10, 257–274.\\nLau, R., & Sears, D. (Eds.). (1986). Political cognition : The 19th Annual Carnegie\\nSymposium on Cognition.. Hillsdale, NJ: Lawrence Erlbaum Associates.\\nLind, M. (1995). The new nationalism and the fourth American revolution. New York:\\nThe Free Press.\\nLipset, S. M. (1996). American exceptionalism: A double-edged sword. New York: W. W.\\nNorton. \\nMerelman, R. M., Streich, G. & Martin, P. (1998). Unity and diversity in American\\npolitical culture: An exploratory study of the national conversation on American\\npluralism and identity. Political Psychology, 19, 781–808.\\nMiller, A. H., Gurin, P., Gurin G., & Malanchuk O. (1981). Group consciousness\\nand political participation. American Journal of Political Science, 25, 494–511.\\nMiller, D. (1995). On nationality. Oxford: Clarendon Press.\\nMyrdal, G. (1944). An American dilemma: The Negro problem and modern democracy.\\nNew York, London: Harper & Brothers.\\nOkin, S. M. (1997). Is multiculturalism bad for women? Boston Review, 22, 5.\\nPortes, A. (Ed.) (1996). The new second generation. New York: Russell Sage Foundation.\\nPrentice, D. A. & Miller, D. T. (Eds.) (1999). Cultural divides: Understanding and\\novercoming group conflict. New York: Russell Sage Foundation.\\nRaz, J. (1994). Multiculturalism: A liberal perspective. Dissent, 41, 67–79.\\nReich, R. B. (1992). The work of nations: Preparing ourselves for 21st century capitalism.\\nNew York: Vintage Books.\\nRoediger, D. R. (1994). Towards the abolition of whiteness: Essays on race, politics, and\\nworking class history. London: Verso.\\nSchlesinger, A. M. Jr. (1992). The disuniting of America: Reflections on a multicultural\\nsociety. New York: Norton.\\nSchaar, J. H. (1981). Legitimacy in the modern state. New Brunswick, NJ: Transaction\\nBooks.\\nSears, D. O. (1993). Symbolic politics. In S. Iyengar & W. J. McGuire, (Eds.), Explo-\\nrations in political Psychology (pp. 113–149). Durham: Duke University Press.\\nSears, D. O., Citrin J., Cheleden S., & Van Laar C. (1999). Cultural diversity and\\nmulticultural politics: Is ethnic Balkanization psychologically inevitable? In D.\\nPrentice & D. Miller, (Eds.), Cultural divides: The social psychology of cultural con-\\ntact. (pp. 35–79) New York: Russell Sage Foundation. \\nSears, D. O., & Funk, C. L. (1990). Self-interest in Americans’ political opinions. \\nIn J. J. Mansbridge, (Ed.), Beyond self-interest (pp. 147–170) (1999). Chicago:\\nUniversity of Chicago Press.\\nSidanius J., Levin S., Rabinowitz J. L., & Frederico C. M. (1990). Peering into the\\njaws of the beast: The integrative dynamics of social identity, symbolic racism\\nThe Meaning of American National Identity\\n99\\n\\n\\nand social dominiance. In D. A. Prentice & D. T. Miller, (eds.), Cultural divides:\\nUnderstanding and overcoming group conflict (pp. 80-132) New York: Russell Sage\\nFoundation. 1999.\\nSmith, R. M. (1997). Civic ideals: Conflicting visions of citizenship in U.S. history. New\\nHaven: Yale University Press.\\nStaub, E. (1997). Blind versus constructive patriotism: Moving from embeddedness\\nin the group to critical loyalty and action. In D. Bar-tal & E. Staub, (Eds.), Patri-\\notism (pp. 213–228).Chicago: Nelson Hall.\\nSullivan, J., Fried A., & Dietz M. (1992). Patriotism, politics, and the presidential\\nelection of 1988. American Journal of Political Science, 36, 200–234.\\nT\\najfel, H. (1978). Social categorization, social identity and social comparison. In H.\\nT\\najfel (Ed.), Differentiation between social groups (pp. 61–76). London: Academic\\nPress. \\nT\\najfel, H., & Turner J. (1986). The social identity theory of intergroup behavior. \\nIn S. Wrochel & W. Austin, (Eds.), Psychology of intergroup relations. Chicago:\\nNelson-Hall.\\nT\\naylor, C. (1995). Cross purposes: The liberal-communitarian debate. In Philosophi-\\ncal arguments. (pp. 181–203). Cambridge, MA: Harvard University Press.\\nThoits, P. A., & Virshup, L.K. (1997). Me’s and we’s: Forms and functions of social\\nidentities.” In R. D. Ashmore & L. Jussim, (Eds.), Self and identity: Fundamental\\nissues. New York: Oxford University Press. \\nTurner, J. C., Hogg, M.A. Oakes, P. J. Reicher, S. D. & Blackwell, M. S. (1987). \\nRediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.\\nT\\nyler, T., Lind E. A., Ohbuchi K., & Sugawara, I. (1998). Conflict with outsiders:\\nDisputing within and across cultural boundaries. Personality & Social Psychology\\nBulletin, 24, 137–147.\\nWarner, R. S. (1993). Work in progress: T\\noward a new paradigm for the sociological\\nstudy of religion in the U.S. American Journal of Sociology, 98, 1044–1093.\\nWaters, M. (1990). Ethnic options: Choosing identities in America. Berkeley: University\\nof California Press. \\nWolfe, A. (1998). One nation after all. New York: Viking.\\nYoung, I. M. (1990). Justice and the politics of difference. Princeton, NJ: Princeton\\nUniversity Press.\\nZaller, J. (1992). The nature and origins of mass opinion. New York: Cambridge Uni-\\nversity Press. \\n100\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nJim Sidanius\\n5\\nJohn R. Petrocik\\nCommunal and National Identity\\nin a Multiethnic State\\nA Comparison of Three Perspectives\\nAs recent episodes of “ethnic cleansing” in Rwanda, Bosnia, and Kosovo\\nmake painfully clear, some multiethnic states seem as inclined to interethnic\\nviolence at the dawn of the 21st century as they were at the start of the 20th.\\nIndeed, the national mobilization of a 50-year-long cold war may have tem-\\nporarily restrained hostilities that lost none of their underlying virulence.\\nGrievances that were set aside by a need to face a common enemy have now\\nbecome priority issues for organized interests, parties, and public policy.\\nFurthermore, ethnic conflict is not peculiar to nation-states mobilized into\\nthe stand-off between the cold war alliances, but appears quite generalized\\nacross the globe (see Gurr & Harff, 1994).\\nIn many cases, intergroup conflict within multiethnic states turns on the\\ncompatibility of subgroup identities and commitments to the nation-state.\\nThus, one wonders if Israel can rely only upon Jewish citizens? Can a\\nGypsy also be Hungarian? Are Finns of Swedish extraction loyal to Fin-\\nland? Can Francophones be as interested in the welfare of Canada as En-\\nglish speakers? Can a Turkish immigrant (and his children) be as German as\\nan ethnic German? Can Chinese, Filipino, and Mexican immigrants also be\\ngood Americans? The national loyalty question arises because “social ho-\\nmogeneity and political consensus are regarded as prerequisites for, or fac-\\ntors strongly conducive to, stable democracies” (Lijphart, 1977, p. 1). These\\nquestions are crucial because the modern nation-state is predominantly \\n(if not universally) organized around ethnic definitions of citizenship\\n(Horowitz, 1985). Yet few nations are homogeneous and, as migration in-\\ncreases, ethnic diversity creates more opportunities for conflict within them\\nalong ethnic, racial, religious, and linguistic divides (hereafter referred to\\n101\\n\\n\\ncollectively as “communal” differences). A better understanding of the\\nmanner in which people are able to negotiate loyalty to their ethnic particu-\\nlarisms while adopting or retaining the civic culture and loyalty demanded\\nby the nation-state seems as worthwhile today as it ever was in the past.\\nThis is especially true given that it is the institutions of the state that pro-\\nvide the principal framework within which communal differences are re-\\nsolved. This chapter examines the relationship between ethnic and national\\nidentities within the United States.\\nDiversity as a Source of Conflict\\nWhile we recognize a history of conflict among communal groups, the\\nethos of American society treats such behavior as aberrant, largely idiosyn-\\ncratic individual choices. Communal conflict is seen as a “breakdown” of a\\nnormal condition. We argue the contrary: communal conflict is a manifes-\\ntation of a systematic tendency for: (1) groups to form social identities\\naround communal differences, (2) there to be an inverse relationship be-\\ntween the strength of communal and common identities, and (3) both mate-\\nrial and symbolic differences between groups to become occasions for con-\\nflict. The thesis underlying the following analysis is that social diversity\\nshapes social identities that provide material and symbolic reasons for com-\\nmunal conflict in the United States. Further, the thesis asserts that conflict\\nis an unavoidable, systematic by-product of social diversity itself. Material\\ninterests that coincide with social identities (especially when they are redis-\\ntributive) will make communal conflict more common and severe. Institu-\\ntions can moderate its severity. A civic and social culture can be constructed\\nthat pronounces communal conflict illegitimate, thereby reducing it to a\\nminimum. But it will always be present because social identities will always\\nbe salient to some fraction of the society, and any material interest that co-\\nincides with a communal distinction will insert that communal difference\\ninto the conflict because of the sensitivity of people to communal distinc-\\ntions (Lijphart, 1977; Powell, 1982).\\nEthnic, Racial, and National Loyalty\\nThis chapter analyzes ethnic conflict by focusing on certain relationships\\nbetween communal identity and national loyalties. We have two “nation-\\nality” groups—Hispanic Americans and Asian Americans. The first is asso-\\nciated with the countries of Latin America (principally Mexico), even\\nthough many are the progeny of families that have been in the United\\nStates since the early 18th century. Although some Asian Americans also\\ncome from families that are long-term residents of the United States, most\\nare recent immigrants. Consequently both fit into the conditions that are\\n102\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\npresumed to create the “divided loyalty” often attributed to identifiable\\nsubnational communities. \\nAfrican Americans have a singular history. Even though many can trace\\ntheir time in the United States to the 17th century, slavery and segregation\\nhave made them strangers in the land. Africa may not exert a sense of iden-\\ntity for many, but the separation from the larger society that was forced on\\nthem for almost three centuries created a group experience that is, we be-\\nlieve, functionally equivalent to being placed on the margins of the society. \\nSpecifically, this chapter examines the relationship between national loy-\\nalty and a preference for one’s ethnic and “racial” group. Surprisingly little\\nempirical research has been directed at this issue. Within the American\\ncontext, most thinking and research directed at this question falls into two\\ncategories: the classical “melting pot” model of incorporation and an alter-\\nnative “pluralist” perspective, with the latter giving rise to multicultural \\nvisions of the social order. \\nThe rarely examined assumption of these images of American society is\\nthat there is no inherent conflict between subgroup and national identity. A\\nfew attempts to examine the relationship have usually reported a compati-\\nbility, sometimes an easy one, between these identities (de la Garza, Falcon,\\n& Garcia, 1996). However, this assumption and the empirical evidence in\\nfavor of it flies in the face of emerging work on social dominance, which\\nfinds incompatibility between subgroup and national social identities. In\\ngeneral, this work suggests that national identity is generally associated with\\ncommunal identity for one group and the exclusion of others (Sidanius,\\n1993; Sidanius, Feshbach, Levin, & Pratto, 1997; Sidanius & Pratto, 1999).\\nThe purpose of this chapter is to further examine the linkage between com-\\nmunal identity and national loyalties in the United States.\\nThe United States is a hard case because disaffection is limited and disaf-\\nfected groups are few in number. The extreme group cleavages that assure a\\nnegative association between commitment to common and subgroup iden-\\ntities are absent in the United States. Moreover, the prevailing civic culture\\nemphasizes the compatibility of subgroup and common identities and loy-\\nalty. Yet, as the following data show, such negative relationships actually ex-\\nist. The nuances of this relationship and their implications for multiethnic\\nstates are elaborated below.\\nA Thematic History of American Identity and Ethnic Diversity\\nThe earliest comparative studies of mass politics quickly established that\\nconflict in even the most well-ordered democratic societies closely followed\\ncommunal divisions. Religious, ethnic, racial, linguistic, and regional differ-\\nences shaped social and political debate more strongly than most other dis-\\ntinctions (class, for example; see Lipset & Rokkan, 1967; Horowitz, 1985;\\nLijphart, 1979; Petrocik, 1981, 1998). The history of the United States fol-\\nCommunal and National Identity in a Multiethnic State\\n103\\n\\n\\nlows this pattern (see, e.g., Kleppner, 1970; Kelley, 1979). Communal con-\\nflict was fueled by cycles of immigration that transformed the nation from a\\npopulation of largely English Protestants (with some dissenter religions) to\\none that included continental nationalities, Catholics, and Jews. Making\\neven the most modest assumption about fertility patterns, it seems likely\\nthat immigrants and first-generation Americans (who could be expected to\\nbe more like their immigrant parents than the children of “old stock”\\nAmericans) represented upwards of a third of the population of the country\\nby 1900.\\nAmerican politics was affected by this diversity from the earliest days of\\nthe republic. The Alien Act of 1798, for example, was passed by the declin-\\ning Federalists to limit the ability of immigrants to vote in the hope that the\\ngrowth of the Democratic-Republicans would be slowed. Their lack of suc-\\ncess did not forestall many further attempts to limit the impact of immigra-\\ntion and population diversity on an originally English, Protestant society.\\nThe Irish, and Catholics in general, were the first to redefine what America\\ncould represent; central and north Europeans came next; southern and east-\\nern Europeans were the major group for about 60 years in the late 1800s\\nand early 1900s; immigration from Asia during this same period was impor-\\ntant in the west. In different ways and at different times, this population di-\\nversification led to new political parties and legislation to limit the number\\nof immigrants and their influence on American life. \\nThe surge in immigration in the late 1980s—the immigration rate was\\n2.3 per thousand in 1980 but 6.1 per thousand in 1990—reawakened latent\\nsensitivity to the new “strangers in our midst.” This reawakening has been\\nespecially pronounced in California, where the number of new immigrants\\nwith a non-European “racial” and ethnic heritage is very large. The new\\nimmigrants, like their predecessors, are perceived by the native population\\nto be substantially different, raising familiar questions about how they fit\\ninto the society, where their loyalties and commitments lie, what changes\\nshould be made to accommodate them, and what costs they will impose on\\nthe society. School boards, zoning boards, city councils, utilities regulators,\\nlocal and state elections, and presidential politics have all been forums for\\nconflicts between Americans and the new immigrants. Some of the disputes\\nare material, turning on taxes and social welfare costs; others are symbolic\\n(e.g., Chinese residents of the San Gabriel Valley objected to a switch in the\\ntelephone area code from 818 to 626 because of the numerological signifi-\\ncance of the numbers for the Chinese). Unresolved problems between\\nwhite and black Americans have acquired renewed significance with the \\nrising tide of immigration.\\nAmericanization and the Melting Pot\\nSpecific disputes between natives and immigrants have always been con-\\nflicts over the identity of the national community: about what values, view-\\n104\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\npoints, and interests were compatible, and which could not be abided be-\\ncause they served and promoted loyalties that were hostile to a common\\nAmerican identity. It rested on the expectation that different groups em-\\nbraced significantly different values, beliefs, and perceptions. A de facto\\npolicy of “Americanization” was the first way in which the nation dealt with\\ndiversity. Everybody was expected to adopt the language, dress, customs,\\nbeliefs, and loyalties of the dominant English culture that defined the\\nUnited States. By the end of the 19th century, the polity replaced “Ameri-\\ncanization” with the “melting pot.” \\nThe melting pot replaced the expectation of submersion in the preexisting\\nEnglish Protestant culture with a preference (not necessarily universal) for a\\ncontinuous reformulation of what constituted American culture. Immi-\\ngrants, in the melting pot formulation, had valuable things to offer America,\\nand this value was to be recognized as American society incorporated the lan-\\nguage, cuisine, social sensibilities, and economic and political values of the\\nimmigrant cultures. This reformulation of American culture, coupled with\\nmarriage across the lines of national origin and religion, became the alterna-\\ntive resolution of ethnic diversity and the suspicions it produced. It resolved\\n(at least intellectually) the struggle over whether immigrants were real Amer-\\nicans by loosening (if not severing) the connection between WASP and\\nAmerican. The result was a continually changing American culture and a\\ncontinually changing American ethnicity still expressing the basic theme\\n“from many, one” (see e.g., Salins, 1997). Restrictions on European (and es-\\npecially German) nationality groups during World War I, immigration lim-\\nits after 1920, and “Americanization” campaigns (largely tied to learning En-\\nglish and shedding the most obvious habits and custom of the immigrant\\nculture) through World War II produced substantial assimilation and largely\\neliminated the salience of nationality and ethnic differences by the middle\\n1960s (Walch, 1994; Freedman, 1996).\\nThe Unmelted Pot of Pluralism\\nA more widespread recognition of the pluralism of American society is of-\\nten attributed to the reemergence of a white ethnic identity in response to\\nthe civil rights movement (Greeley, 1971, 1976; Novak, 1971), but, as Mil-\\nton Gordon (1964) demonstrated almost four decades ago, and others con-\\nfirmed (e.g., Wolfinger, 1965; Parenti, 1967), resistance to melting-pot\\npressures was strong and seems to be to the present day (Alba, 1990). While\\nthe notion of the melting pot has been the popular metaphor for most of\\nthis century, it co-existed with a “pluralist” image of “salad bowls,” “quilts,”\\nor “glorious mosaics” to explain how one could be an American while \\nretaining a distinct “racial” or ethnic identity.\\nWhile there is variation in the precise manner in which the term “plural-\\nism” is understood, the central idea is that subgroup identities are retained.\\nIn a pluralist society, communal segments defined by religion, language, na-\\nCommunal and National Identity in a Multiethnic State\\n105\\n\\n\\ntional origin, or “race” remain salient to individuals. Catholics recognize\\nProtestants; the progeny of Italian immigrants see “Yankees” and Polish\\nAmericans; blacks and whites are aware of a “racial” divide; Louisiana Ca-\\njuns recognize differences between themselves and their Protestant neigh-\\nbors. While the pluralist image expects social views and policy preferences\\nto vary by ethnic and “racial” segment, there is no notion that the differ-\\nences are unbridgeable because the pluralist political culture expects in\\ndividuals in the subgroups to feel an identity and loyalty to the national\\ncommunity, and accept a set of obligations to it. Some formulations of plu-\\nralism expect positive supra-communal effects from the communal identity.\\nRather than contributing to “disuniting” and fractionation (Schlesinger,\\n1992), these distinct ethnic loyalties are thought to co-exist and even pro-\\nmote a common national identity through the communal identity (Powell,\\n1995; de la Garza et al., 1996).\\nThe Melting Pot, Pluralism, and Conflict \\nA melting-pot, immigrant society expects virtually no communal conflict\\nbecause the melting pot society has virtually no communal divisions around\\nwhich material or symbolic grievances can be politicized. Pluralism, al-\\nthough it has a strong expectation of communal conflict and expects groups\\nto be sensitive to how social resources (e.g., income, education, employ-\\nment opportunities, and social standing) are distributed, assumes a political\\nframework within which affected groups have different advantages and are\\nable to bring them to bear in ways that ensure that their core interests are\\npromoted. (Dahl, 1961, offers the classic exposition; but also see Polsby,\\n1963, for a significant elaboration of the basic theory that guided virtually\\nall research on the topic.) Every turn to government may energize inter-\\ngroup conflicts because such collective action enforces contributions that\\nsome groups will not prefer to make, especially when the outcome is redis-\\ntributive. In the textbook model of political pluralism, all groups participate\\nequally and have a similar positive regard for the basic institutions that\\nstructure the political competition because no group is permanently ex-\\ncluded or permanent losers to any other groups.\\nDiversity, Social Dominance, and Communal Conflict\\nCentral to the pluralist argument is the notion that all ethnic groups be re-\\ngarded as co-equal partners in the pursuit of the American dream. However,\\nin contrast to this vision, there are a cluster of theories of intergroup rela-\\ntions that suggest that “co-equality” never describes relationships among\\nsalient groups within multiethnic societies. These theoretical models have\\nbeen referred to as group dominance models, where social dominance\\ntheory is the most explicit and recent statement of this general position (see\\n106\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nSidanius & Pratt. 1999; for related models see Blumer, 1960; Jackman,\\n1994). These group dominance models suggest that societies tend to be \\norganized as group-based hierarchies, with dominant groups enjoying dis-\\nproportionate power, prestige, and privilege. This group-based hierarchy\\napplies to both democratic and nondemocratic states, despite differing dis-\\ncourses concerning inclusivity or pluralism. \\nSocial dominance theory, as it has been developed by Sidanius and Pratto\\nand their colleagues (see, e.g., Pratto, Sidanius, Stallworth, & Malle, 1994;\\nSidanius, 1993; Sidanius & Pratto, 1999) shares with pluralism an expecta-\\ntion that communal identities will not be submerged into a larger national\\nconsciousness, but, contrary to pluralism, it expects some groups to be relatively\\nconsistent losers in the struggle. The foundation for this expectation is social\\ndominance theory’s assertion that social systems are structured as group-\\nbased social hierarchies with one or few groups at the top of the hierarchy,\\nwith all others—sometimes in a carefully ordered fashion—arrayed below\\nthem. The mechanisms and processes that establish and maintain the ex-\\npected social hierarchy can be informal—group-based associational pat-\\nterns that create and limit opportunities—or formal—political and eco-\\nnomic institutions that promote the interest of dominant segments (see\\nespecially Sidanius & Pratto, 1999). \\nIn time, institutions can become identified with specific groups rather\\nthan as common institutions of the society, and considerable political con-\\nflict can turn on perceptions of who is served by them. Examples of this in\\nthe United States are easy to observe. Social welfare agencies are popularly\\nidentified with minority ethnic groups. White Americans often resist finan-\\ncial pleas from urban school districts because the school systems in big cities\\nare identified with minority children. And while police forces are staffed\\nwith “officer friendlies” for whites, they are commonly regarded as hostile\\nand dangerous by African Americans and other minorities (see Sidanius &\\nPratto, 1999, chap. 8). \\nExclusionary Patriotism\\nNot only do such group dominance models have much to say about how\\npositive and negative social value will be distributed across the group-based\\nsocial hierarchy, but these models also have implications for the general in-\\nterface between national and group identity. Since public institutions are\\nclosely linked to dominant groups, evaluations of them should vary with the\\nsalience of an individual’s identification with the dominant as compared to\\nthe subordinate groups. Specific intergroup differences in evaluations of\\npolitical figures, the police, or particular government agencies are well\\nknown. The interest here is in demonstrating a more general relationship\\nbetween communal identity and a common national identity. This can have\\nseveral manifestations, but our data allow us to examine only one aspect: re-\\ngard for the nation or patriotism.\\nCommunal and National Identity in a Multiethnic State\\n107\\n\\n\\nWe expect the dynamics of communal conflict in an environment of \\nsocial dominance to create exclusionary patriotism: a situation in which \\ncommunal and national identities are supportive and positively correlated\\namong socially dominant groups, but in conflict among those in subordi-\\nnate communal groups. Specifically, we expect the following: \\n1. The prominence of the society’s public institutions in maintaining\\ncommunal hierarchies will cause dominant communal groups to be\\nmore identified with the nation’s symbols.\\n2. We expect a differential relationship between the intensity of com-\\nmunal identities and a common national identity across groups.\\nAmong dominate groups we expect communal and national identi-\\nties to be fused and positively correlated; among subordinate\\ngroups we expect them to be at odds and negatively correlated.\\n3. We expect the negative relationship to be moderated by the experi-\\nences of the subordinate group. That is, those further down the\\ncommunal hierarchy are likely to see more of a conflict between\\ntheir communal identity and national symbols that are associated\\nwith the communal identity of the dominant group. The subjective\\nsocial distance among the groups predicts some of the effect. Of\\ncourse, a real history of significant conflict between the dominant\\nand subordinate group would lead to even stronger effects. \\nThis relationship should be observable even though “exclusionary pa-\\ntriotism” is specifically contrary to the American civic culture. The key\\nvariable is not the avowed norms but the actual operation of the society. \\nBecause the society functions in a way that allows and even encourages a\\nperception that the political order is more closely identified with European-\\nAmericans than it is with other major racial and ethnic groups, the dynam-\\nics of exclusionary patriotism will shape the relationship between subgroup\\nidentities and attitudes toward American national identity.\\nEmpirical Implications of the Interface Between \\nNational and Ethnic Attachment\\nThe melting-pot, pluralist, and group dominance models of American soci-\\nety lead to a different set of expectations concerning the interface between\\nnational and communal identity. This chapter will explore three expecta-\\ntions for each model. \\n• The Salience of Ethnic Identity. The melting-pot model predicts that\\nethnic or “racial” identification will be a trivial social category for\\nmost Americans. Both the pluralist and group dominance perspec-\\ntives predict that these social identities will remain important and\\nsalient social identities. The salience of ethnicity should express it-\\n108\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nself not only in terms of social identity but also in terms of a gener-\\nalized ethnocentric bias against Americans from other ethnic\\ngroups.\\n• Patriotism and Ethnicity. Both the melting-pot and pluralist models\\npredict that patriotic attachment to the United States will be essen-\\ntially equal across ethnic groups. In contrast, and for reasons ex-\\nplained above, the group dominance model predicts that dominant\\nEuropean-Americans will have significantly higher levels of patrio-\\ntism than will members of subordinate groups (e.g., Asian, Latino,\\nor African Americans).\\n• The Interface Between National and Communal Identity. The melting-\\npot and pluralist models predict no correlation between patriotism\\nand communal identity. The melting-pot model predicts no corre-\\nlation because it expects essentially no ethnocentrism in the society.\\nThe variance in ethnocentrism is trivial and idiosyncratic. The plu-\\nralist model of the society predicts no correlation between commu-\\nnal identity and regard for national institutions because the wins\\nand losses of one’s ethnic group are not predetermined by an inher-\\nently subordinate position to dominant groups who use public in-\\nstitutions to enforce their preferences. Indeed, the results of de la\\nGarza et al. (1996) even suggest that ethnic identity can actually\\npromote a stronger sense of regard for the nation. However, social\\ndominance theory (Sidanius et al., 1997) expects the correlation \\nbetween national and communal identity to vary across the social\\nstatus continuum. It should be positive among dominants (e.g., \\nEuropean-Americans) and negative among subordinates (e.g.,\\nAfrican-Americans). T\\nable 5.1 summarizes the expected relation-\\nships between ethnic and national attachment for each model.\\nThus far, we know of only two studies that have attempted to explore the\\nrelative plausibilities of these conflicting models (but see also Citrin et al.,\\nthis volume). Using a nationally representative sample of Mexican Ameri-\\ncans, de la Garza and his colleagues (1996) found clear empirical support\\nfor the pluralist perspective. They found that Mexican Americans who were\\nstrongly attached to their ethnic heritage were no less patriotic than \\nMexican-Americans with weak ethnic attachments. Furthermore, their data\\nshowed virtually no relationship between the strength of this ethnic identity\\nand core American values such as individualism. Even more striking was\\ntheir finding that Mexican Americans appeared slightly more patriotic than\\nnative-born white Americans. \\nA later study by Sidanius and colleagues (1997) found only partial \\nsupport for the de la Garza findings. When comparing the relationships \\nbetween ethnic and national identity between European- and African-\\nAmericans, the results showed clear support for the group dominance per-\\nspective in two respects. First, European-Americans showed significantly\\nCommunal and National Identity in a Multiethnic State\\n109\\n\\n\\nhigher levels of patriotism than African-Americans. Second, there was clear\\nevidence of asymmetry in the correlation between patriotism and commu-\\nnal identity across dominant and subordinate ethnic groups. Among domi-\\nnants (i.e., whites), ethnocentrism was positively correlated with patriotism,\\nwhile among subordinates (i.e., blacks) ethnocentrism was negatively corre-\\nlated with patriotism, as the social dominance model expects. While the\\ndominant-subordinate asymmetry was quite consistent for the contrast be-\\ntween European- and African Americans; it was less consistent with respect\\nto the contrast between European and Latino Americans. Moreover, while\\nthe asymmetry held among a sample of European- and Latino-American\\ncollege students, it did not hold among a national probability sample of\\nwhite and Latino adults (see Sidanius et. al., 1997). In Sidanius et al.’s\\n(1997) analysis of an American national probability sample, the results were\\nessentially consistent with the de la Garza et al. (1996) findings.\\nHow might we explain the somewhat conflicting findings between the de\\nla Garza et al. (1996) and the Sidanius et al. (1997) studies? First and most\\nobviously, the de la Garza et al. study only compared whites and Latinos,\\nwhereas the Sidanius et al. study also included blacks. Second, even the\\nslightly different findings with respect to Latinos might have something to\\ndo with major differences in political context. In recent years, immigration\\nfrom Mexico and Central America has become a salient and contentious po-\\nlitical issue in California, where the Sidanius et al (1997) data were col-\\nlected. It is quite possible that Southern California’s highly racialized public\\ndiscourse concerning illegal Latino immigration (e.g., Proposition 187) in-\\nduced Latinos to feel under attack and as political outsiders as African\\nAmericans have felt for the bulk of American history. Finally, we think that\\ntheir analysis does not measure their key conceptual variable very well. The\\nacculturation variable, the centerpiece of their analysis, is only a weak ap-\\n110\\nThe Contribution of Ethnic and National Identities to Political Conflict\\ntable 5.1 Hypotheses Derivable from the Melting-Pot, Pluralism, and Group\\nDominance Models\\nTheoretical Model\\nHypotheses\\nMelting-pot\\nPluralism\\nGroup Dominance\\n1. Ethnicity remains strong \\nNo\\nYes\\nYes\\nsocial identity\\n2. Dominants will be more \\nNo\\nNo\\nYes\\npatriotic than subordinates\\n3. There will be a correlation \\nNo\\nNo, or positive \\nYes. Positive \\nbetween ethnocentrism and \\ncorrelation\\ncorrelation among \\npatriotic attachment to\\ndominant, negative \\nthe nation\\ncorrelation among \\nsubordinates\\n\\n\\nproximation of subgroup attachment. Their measure categorized individu-\\nals by place of birth and language facility: U.S.-born Mexican Americans\\nwho were English speakers were regarded as the most acculturated while\\nforeign-born Spanish speakers were classed as the least acculturated. Accul-\\nturation, however, is not the relevant variable. Although there is probably a\\nrelationship between the individual’s level of acculturation and the strength\\nof his or her preference for Mexicans over non-Mexican-Americans, there\\nis no reason to assume that it is strong. It may, in fact, be extremely weak.\\nAmong European Americans (the Irish, Italians, Poles, etc.), identity with\\nnational origin is poorly related to rates of endogamy, the observance of\\nethnic customs, and facility with the immigrant language (Alba, 1990). A\\npriori, there is no compelling reason to expect Mexican Americans to be \\ndifferent. \\nT\\no be sure, the null results also appear with the one approximate mea-\\nsure of identity used in the analysis. Respondents classed as Latinos were\\nthose who, when asked their “racial” identity, chose the option “something\\nelse” rather than white. This group, slightly less than half of their sample,\\nclearly thought of themselves in ethnic terms. In our experience, those who\\nself-describe as Latinos are more ethnically identified than those who do\\nnot use this appellation—and this subgroup was not less patriotic or less in-\\ndividualistic in their study. However, an unwillingness of the respondent to\\nchoose the category “white” as their ethnic or racial group is a weak mea-\\nsure of identity, and the null results probably should not be surprising.\\nConsider that it is fairly common for white Americans to respond with a na-\\ntional origin distinction when asked their nationality, but there is tremen-\\ndous variability in the strength of attachment to the national origin of\\nAmericans who call themselves Italian, Polish, or Lithuanian when report-\\ning their nationality. This variance almost certainly exists among Mexican\\nAmericans. It is at least reasonable to suspect that it does, and the weak rela-\\ntionship de la Garza and his colleagues report for the measure may only re-\\nflect the indeterminate property of the categorization.\\nThe finding that Mexican Americans are more patriotic and individualis-\\ntic than native-born whites is even less persuasive. As de la Garza and his\\ncolleagues note, the whites with whom they draw the contrast were avail-\\nable for the analysis only because they lived in areas with large Mexican-\\nAmerican populations. These whites are almost certainly not representative\\nof the white population as a whole, given the patterns of residential segre-\\ngation that exist in the United States. We would expect such whites to be\\npoorer and less educated. Their social class is likely to orient them toward a\\npreference for government action (the exact opposite of the “individualism”\\nexpected of the median American in the de la Garza et. al. formulation).1\\nThe social values that permit these whites to live comfortably in Mexican-\\nAmerican neighborhoods may also have negative implications for expres-\\nsions of patriotism, depressing, or even reversing, therefore, the expected\\nrelationship between it and nativity and ethnicity.\\nCommunal and National Identity in a Multiethnic State\\n111\\n\\n\\nMethod and Design\\nWe used five different data sets in these analyses and a total of 7,400 \\nrespondents. \\nData set 1: UCLA Student Sample\\nThis data set consisted of 725 randomly sampled UCLA undergraduates\\ncollected in the fall of 1993. The sampling frame stratified a list of all regis-\\ntered students into four “ethnic” strata (whites, blacks, Latinos, and Asians),\\nand randomly sampled from each stratum. The subjects were enticed to\\nparticipate by the offer of four $50 prizes. The ethnic breakdown of all of\\nthe samples is found in table 5.2. The students were given a questionnaire\\nprimarily assessing their attitudes toward patriotism and their affect toward\\nvarious American ethnic groups. However, unlike the Sidanius et. al. (1997)\\nstudy, in this chapter we will explore the responses of resident-aliens as well.\\nPatriotism Measures.\\nKosterman and Feshbach (1989) and Sidanius et al.\\n(1997) have shown that patriotism could be well measured by an eight-item\\nscale consisting of items such as: “I am proud to be an American,” and\\n“Every time I hear the national anthem, I feel strongly moved.” Not only\\ndid this scale have a high degree of face validity, but it also had a fairly high\\nlevel of reliability (with a Cronbach a = .91). Because the patriotism mea-\\nsures were assessed using different scales across the different studies, in or-\\nder to aid in cross-study comparisons, we transformed all of the patriotism\\nmeasures to the same 0.00 to 1.00 scale, where 0 represents the lowest pos-\\nsible score a respondent can get and 1.00 represents the highest possible\\nscore a respondent can get. \\nEthnocentrism. While communal identity can be regarded as expressing the\\ngeneral salience of one’s “racial,” ethnic, or religious identity, we will use\\nthe term “ethnocentrism” to refer to preference for communal ingroups over\\ncommunal outgroups. The ethnocentrism measures were operationalized\\nby use of thermometer rating scales. The respondents were asked to indi-\\ncate their degree of positive or negative feeling toward each of four ethnic\\ngroups using a seven-point rating scale (7–very positive, 4–neutral, 1–very\\nnegative). The groups were whites, African Americans, Latinos, and Asians.\\nT\\nwo different types of ethnocentrism scores were computed: (1) general-\\nized ethnocentrism and (2) hierarchical ethnocentrism. Generalized ethno-\\ncentrism (GEI) was used to measure the respondents’ degree of ethnocen-\\ntrism against the “generalized other,” regardless of whether this generalized\\nother had relatively high or low ethnic status. GEI was computed simply by\\ntaking the thermometer rating given to one’s ethnic ingroup minus the\\nthermometer rating given to the average ethnic outgroup. Hierarchical eth-\\nnocentrism (HEI), on the other hand, was designed to be sensitive to the \\n112\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nhierarchical nature of the status differences among the ethnic groups. Thus,\\nHEI was measured slightly differently, depending upon whether one was a\\nmember of a dominant or subordinate ethnic group. For whites, hierarchi-\\ncal ethnocentrism (HEI) was defined in precisely the same way as for gener-\\nalized ethnocentrism (GEI). However, consistent with the notion of sub-\\nordination to a dominant social group, HEI for the each of the three\\nsubordinates groups was defined as the positive rating given to one’s ethnic\\ningroup minus the rating given to higher-status whites. \\nData Set 2: Los Angeles County Social Survey (LACSS)\\nThese data are drawn from two years (1997 and 1998) of the Los Angeles\\nCounty Social Survey (LACSS). The LACSS is a countywide random-digit\\ndial telephone survey of adults living in households conducted by the Com-\\nputer Assisted T\\nelephone Interviewing unit of UCLA’s Survey Research\\nCenter. The LACSS employs a 12 call-back procedure, systematically vary-\\ning the day of the week and the time of day, before dropping any numbers\\nfrom the sample. A total of 1,232 respondents were interviewed (see table\\n5.2).\\nPatriotism Measures. T\\nwo items were used to operationalize patriotism: (1)\\n“I find the sight of the American flag very moving,” and (2) “I have great\\nlove for the United States.” The questions were answered on a four-point\\nscale from “4–strongly agree” to “1–strongly disagree.” The reliability of\\nthe scale was considered adequate (Cronbach a = .68).\\nEthnocentrism.\\nGeneralized and hierarchical ethnocentrism were defined\\nby use of affect differentials in the same manner as used in Data set 1, the\\nUCLA student sample. \\nCommunal and National Identity in a Multiethnic State\\n113\\ntable 5.2 Data Sets and Ethnic Groups\\nSamples\\nData Set 1\\nData Set 2\\nData Set 3\\nData Set 4\\nData Set 5\\nEthnic\\nUCLA Students\\nLACSS\\nNES\\nNES\\nGSS\\nGroup\\n1993\\n1997–98\\n1992\\n1996\\n1996\\nT\\notal\\nWhites\\n153\\n539\\n1,905\\n1,325\\n1,015\\n4,937\\nBlacks\\n113\\n139\\n312\\n191\\n170\\n925\\nLatinos\\n98\\n221\\n207\\n149\\n54\\n729\\nAsians\\n72\\n68\\n28\\n22\\n10\\n200\\nJews\\n32\\n—\\n—\\n—\\n—\\n32\\nImmigrants\\n257\\n265\\n—\\n—\\n55\\n577\\nT\\notal\\n725\\n1,232\\n2,452\\n1,687\\n1,304\\n7,400\\n\\n\\nEthnic Grievances. Another series of questions assessed respondents’ “sense\\nof ethnic grievance”—that is, the extent to which their ethnic or racial group\\nhas been disadvantaged by those in power. These questions assessed whether\\nrespondents: (1) believed that their group should have more influence in gov-\\nernment; (2) believed that their group was often held back by other people;\\n(3) believed that the government helps particular groups and does not pay\\nenough attention to people in general; (4) believed that public officials from\\ntheir own group would be more likely to do things respondents approved of;\\n(5) would not vote for someone who was not likely to help their group; (6)\\noppose tax increases that help groups other than their own; and (7) would\\nvote for a candidate from their own group, even if that candidate did not\\nagree with them on every issue. All questions except the first two used an\\nagree or disagree format (which also included options for don’t know and re-\\nfused, and neither agree or disagree if this was volunteered by the respon-\\ndent). The response options for the first question were should have more in-\\nfluence, less influence, or just about right (as well as don’t know and refused).\\nThe response options for the second question were often held back, some-\\ntimes held back, and treated as well as anybody else (as well as don’t know and\\nrefused). For this question, both held back and sometimes held back were\\ncoded as “agreement” (i.e., agreeing that their group has been held back).\\nData Set 3: National Election Study—1992 (NES)\\nThe third data set was a national probability sample of Americans collected\\nas part of the 1992 National Election Study (NES) conducted by the Cen-\\nter for Political Studies at the Institute for Social Research at the University\\nof Michigan. Altogether, the 1992 study interviewed 2,452 respondents and\\nthe analysis is limited to the four major ethnic groups that are the focus of\\nthis study (i.e., whites, African Americans, Latinos, and Asians; see table 5.2\\nfor details.\\nPatriotism. Patriotism was defined by two questions that had a content that\\nwas virtually identical to the patriotism questions used in data set 2: (1)\\n“When you see the American flag flying does it make you feel extremely\\ngood, very good, somewhat good, or not very good?” (2) “How strong is\\nyour love for your country? Extremely strong, very strong, somewhat\\nstrong or not very strong?” (Cronbach a = .78). \\nEthnocentrism. Ethnocentrism was defined in precisely the same way in data\\nsets 2 and 3.\\nData Set 4: National Election Study—1996 (NES)\\nThe fourth data set was the 1996 National Election Study (NES), also con-\\nducted by the Center for Political Studies at the Institute for Social Re-\\n114\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nsearch at the University of Michigan (N = 1,687; see table 5.2). This data\\nset did not contain measures of patriotism, but it did contain feeling ther-\\nmometers that were used to compute ethnocentrism in the same way as in\\ndata sets 2 and 3. The 1996 NES permitted a replication of the communal\\nidentity patterns observed in the 1992 study, and a check against a one-time\\npattern.\\nDataset 5: General Social Survey—1996 (GSS)\\nThis data set (N = 1,304 ) was used to replicate earlier findings and examine\\ndimensions of national identity beyond what was possible with the patrio-\\ntism index used in the 1992 NES study (data set 3) and the LACSS samples\\n(data set 2). While the dimension of patriotism, or love of the nation, is a\\ncentral dimension of national attachment, it is clearly only one of several\\ndifferent possible ways in which citizens can be psychologically attached to\\nthe nation. Previous research (Kosterman & Feshbach, 1989; Sidanius et\\nal., 1997; Hofstetter, Feierabend, & Klicperova-Baker, 1999) has shown\\nthat there are at least two important dimensions of national attachment: pa-\\ntriotism (love of country and its symbols) and nationalism (the desire that\\none’s country dominate other countries). Previous research also suggests\\nthat while dominants are significantly more patriotic than subordinates,\\nthey are not more nationalistic than subordinates (see Sidanius et al., 1997). \\nThe 1996 GSS permitted a further exploration (as well as replication) of\\nthe communal-national trade-off because it contained 34 questions con-\\ncerning Americans’ feelings of national attachment. While the 34 national\\nattachment items formed a reasonable coherent and homogeneous scale\\n(Cronbach a = .84), examination of the items revealed distinctly different\\ntypes of national attachment within this scale. T\\no further explore this issue\\nand before any further analyses were attempted, we first factor-analyzed\\nthese items using an alpha factor extraction approach. While there were as\\nmany as 10 factors with eigen values greater than 1.00, the scree test sug-\\ngested that only four of these factors were meaningful. Altogether these\\nfour factors accounted for approximately 37 percent of the total variance.\\nThe factors were interpreted as: \\n1. Patriotism or pride in one’s nation (51 percent of the common vari-\\nance) was primarily defined by nine items such as: “Proud of the\\nway democracy works in America,” “Proud of America’s economic\\nachievements,” “Proud of America’s political influence in the\\nworld,” Proud of America’s scientific and technical achievements.”\\n2. Nationalism (20 percent of common variance) essentially embraced\\na chauvinistic and belligerent attitude toward other nations and\\npeoples (see also Bar-tal & Staub, 1997). The items most strongly\\ndefining this dimension were: “TV should give preference to\\nAmerican films,” “Foreigners should not be allowed to buy land,”\\nCommunal and National Identity in a Multiethnic State\\n115\\n\\n\\n“America should follow its own interests,” “People should support\\ntheir country even when the country is wrong,” and “People would\\nbe better off if they were more like Americans.”\\n3. Unwillingness to leave America (16 percent of common variance) pri-\\nmarily expressed people’s unwillingness to move away from\\nAmerica (e.g., “Willing to move outside of America,” “Willing to\\nmove outside of North America”).\\n4. Importance of American identity (13 percent of common variance), as\\ndistinct from patriotism or pride in one’s nation, expressed people’s\\nsense of social identity as being American. The dimension was pri-\\nmarily defined by items such as: “It is important to have lived in\\nAmerica for life,” “It is important to have been born in America,”\\n“It is important to have American citizenship.”\\nFactor scores were computed for each respondent on each factor. Not\\nsurprisingly, these factor scores were positively correlated, with the\\nstrongest correlation found between “Patriotism” and “American identity”\\n(r = .42), and the weakest between “nationalism” and “unwilling to leave\\nAmerica.” ( r = .10).\\nThe 1996 GSS included one question that we used to assess the salience\\nof ethnic identification. This question read “Compared to people who are\\nnot (your ethnicity), would you say that you feel much closer to people who\\nare (your ethnicity), somewhat closer to them, or not closer at all?” The\\nterm “your ethnicity” did not actually appear—it was always replaced by the\\nrespondent’s own ethnicity (e.g., if they said they were Asian, it read\\n“Asian”; if they said they were “Latino” etc.).\\nResults\\nThe alternative models of American society posit ethnic differences (or in\\nthe case of the melting-pot model, no ethnic differences) in: the salience \\nof ethnic identity, group levels of patriotism, and the relationship between\\npatriotism and ethnocentrism.\\nIssue 1: The Continued Salience of Ethnic Identity\\nWithin the melting-pot model, ethnic identification should not be a salient\\nsocial identity. We explored the validity of this assumption by examining: \\n(1) the degree to which American citizens also think of themselves in\\n“racial”/ethnic terms; (2) ethnocentric bias against other American ethnic\\ngroups, and (3) people’s political grievances as a function of their own \\nethnic/racial ingroups. First T\\nable 5.3 shows ethnicity to be a salient social\\nidentity for most Americans. For example, some 76 percent of European-\\nAmericans in the nation as a whole feel close (34 percent are very close) to\\n116\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nother whites. Among, African, Latino and Asian Americans, this percentage\\nis even greater at 89 percent, 81 percent, and 84 percent, respectively. The\\nsame general pattern was found for Americans in Los Angeles County. How-\\never, not surprisingly and in both cases, the salience of ethnicity varied sig-\\nnificantly as a function of ethnic group membership. In both cases ethnicity\\nwas more salient for the ethnic minorities than for European Americans (na-\\ntional data set: X2(6) = 81.71, p < .001; LA dataset X2(6) = 61.75, p < .001).\\nSecond, using the Los Angeles and national probability samples, we ex-\\namined the generalized ethnocentrism scores of each ethnic group. If one’s\\nethnic/racial particularism is no longer a salient feature of people’s social\\nidentity, as a strict reading of the melting-pot model would expect, then\\nAmericans from one particular ethnic or “racial” group should not exhibit\\nethnocentric bias against other American “racial”/ethnic groups. However,\\nas the data in table 5.4 show, all four major American ethnic/racial groups\\ndisplay an ethnocentric bias against other groups. While European Ameri-\\ncans in Los Angeles appeared to be slightly less ethnocentric than members\\nof the other major ethnic groups (the thermometer difference was approxi-\\nmately 3 degrees compared to differences of 8, 10 , and 8 degrees among\\nblacks, Hispanics, and Asians, respectively), the level of general ethnocen-\\ntric bias among European Americans in the nation as a whole appeared to\\nbe no more or less extreme than among other major ethnic groups in the\\nnation as a whole (i.e., an 11 degree average difference compared to an av-\\nerage difference of 17 degrees for African Americans, 10 degrees for Lati-\\nnos, and 8 degrees for Asians). This very broad and generalized level of eth-\\nnocentric bias is consistent with the thrust of social identity theory (T\\najfel &\\nTurner, 1986).\\nFinally, ethnic group membership also has political relevance and politi-\\ncal overtones. Using the 1997 LACSS, we asked respondents from each of\\nthe major ethnic groups a series of questions concerning whether they felt\\nCommunal and National Identity in a Multiethnic State\\n117\\ntable 5.3 Feelings of Closeness to One’s Own Ethnic/Racial Group (Citizens\\nOnly)\\nWhites\\nBlacks\\nHispanics\\nAsians\\nUSA\\nLA\\nUSA\\nLA\\nUSA\\nLA\\nUSA\\nLA\\nRespondent feels:\\nVery close\\n34%\\n27%\\n70%\\n58%\\n47%\\n43%\\n46%\\n37%\\nSome closeness\\n42\\n52\\n19\\n39\\n34\\n40\\n38\\n41\\nNot close\\n24\\n21\\n11\\n10\\n19\\n14\\n17\\n22\\nT\\notal\\n100% 100%\\n100% 100%\\n100% 100%\\n100% 100%\\nNumber of cases\\n975\\n559\\n166\\n140\\n343\\n432\\n14\\n95\\nSource: The U.S. national data are drawn from the 1996 General Social Survey. The Los Angeles data are\\nfrom the 1997 Los Angeles County Social Survey.\\n\\n\\nthat members of their own ethnic group were being fairly treated by the po-\\nlitical system, whether public officials from their own ethnic group would\\ntreat them more fairly, and whether they would vote for a member of their\\nown group, even if they disagreed with the respondent (see table 5.5). For\\nexample, while members of the ethnic minorities are significantly more\\nlikely than whites to feel that their own ethnic group needs more influence\\nin the political system (X2(3) = 15.82, p < .01), a majority of all groups feel\\nthat their ethnic ingroups need more political influence. Similarly, a ma-\\njority of all ethnic groups believe that some groups get too much help and\\nnot enough attention is paid to what is good for all people. The only issue\\nwhere one’s ethnic identity does not seem to play a major role concerns re-\\nspondents’ willingness to vote for a member of one’s own ethnic group,\\neven when the respondent disagrees with the candidate. T\\nable 5.5 shows\\nthat a distinct minority of respondents is willing to vote strictly on the basis\\nof ethnicity, varying from 23 percent for blacks to 6 percent for Asians.\\nAltogether, ethnicity remains an important and salient social identity\\nwith clear political overtones. In general, the data give little support to the\\nmelting-pot image, at least to the extent that the melting pot assumes that\\nAmericans are blended and substantially unaware of their “racial” and eth-\\nnic differences. T\\nable 5.5 also makes clear that the pluralist notion of con-\\ntented collaboration among competing groups is at least contested. Ethnic\\nminorities are particularly likely to believe that they suffer at the hands of\\nothers because of their communal identity and that members of their group\\nare likely to be more sympathetic and protective.\\n118\\nThe Contribution of Ethnic and National Identities to Political Conflict\\ntable 5.4 Thermometer Differences Among Ethnic/Racial Groups (Citizens\\nOnly)\\nWhites\\nBlacks\\nHispanics\\nAsians\\nUSA\\nLA\\nUSA\\nLA\\nUSA\\nLA\\nUSA\\nLA\\nRating of:\\nWhites\\n71\\n81\\n73\\n79\\n71\\n75\\n67\\n74\\nBlacks\\n62\\n77\\n85\\n87\\n69\\n65\\n63\\n66\\nHispanics\\n60\\n78\\n70\\n79\\n78\\n79\\n59\\n66\\nAsians\\n58\\n79\\n61\\n78\\n63\\n66\\n71\\n77\\nOwn group\\n71\\n81\\n85\\n87\\n78\\n79\\n71\\n77\\nAverage of\\nother groups\\n60\\n78\\n68\\n79\\n68\\n69\\n63\\n69\\nDifference\\n+11\\n+3\\n+17\\n+8\\n+10\\n+10\\n+8\\n+8\\nNumber of cases\\n3151\\n559\\n495\\n140\\n343\\n432\\n61\\n95\\nSource: The U.S. national data are drawn from the 1992 and 1996 National Election Study surveys of those\\nyears. The Los Angeles data are from the 1997 and 1998 Los Angeles County Social Survey.\\n\\n\\nIssue 2: Ethnicity and Patriotism\\nOur second question concerns whether the degree of patriotic attachment\\nto the nation is the same across all major ethnic groups. While both the\\nmelting-pot and pluralist models would expect no differences among the\\ngroups, the group dominance model of American society expects consider-\\nable differences. Dominant European Americans should have significantly\\nhigher levels of patriotism than Hispanics, Asians, and African Americans.\\nIn addition, since African Americans have long been, and remain, the group\\nat the very bottom of America’s ethnic hierarchy, the group dominance ap-\\nproach would expect the mean level of patriotism among African Americans\\nto be relatively low and no greater than that found among resident-aliens,\\nor people barred from participation in the political system. T\\no explore these\\nquestions, we compared patriotism as a function of ethnicity across four of\\nthe five data sets (UCLA students, Los Angeles County residents, and two\\nrepresentative samples from the nation as a whole).\\nThe results in table 5.6 are inconsistent with both the melting-pot and\\nthe pluralism models. Contradicting the melting-pot and the pluralist per-\\nspectives on America, patriotic attachment is not uniform across the ethnic\\nstatus hierarchy. However, the differences are not substantially large. More-\\nover, the absolute scores indicate that all groups have relatively high patrio-\\ntism scores. The bulk of the intergroup difference, reflected in the correla-\\ntion ratios (varying between .14 and .35), is due to the relatively low\\nCommunal and National Identity in a Multiethnic State\\n119\\ntable 5.5 Sense of Grievance Among Those Identifying with an Ethnic/Racial\\nGroup (Citizens Only)\\nWhite\\nBlack\\nHispanic\\nAsian\\nPercent agreeing:\\nBelieve that ones group needs more \\ninfluence\\n58\\n86\\n71\\n67\\nBelieve that he/she has been “held back” \\nbecause of their ethnic/racial identity\\n27\\n77\\n88\\n67\\nBelieve that some groups are helped too \\nmuch and not enough attention is paid \\nto what is good for all people\\n54\\n55\\n59\\n67\\nBelieve that public officials from own \\ngroup are more likely to do preferred \\nthings\\n30\\n41\\n59\\n56\\nOppose programs that help some group \\nbut not mine\\n47\\n46\\n44\\n33\\nWould not vote for a candidate if he/she \\nwould not help people like respondent\\n55\\n73\\n61\\n28\\nWould vote for a candidate from my group \\neven if I disagreed about some things\\n21\\n23\\n17\\n6\\nNumber of cases\\n151\\n53\\n162\\n46\\nSource: The 1997 Los Angeles County Social Survey.\\n\\n\\npatriotism scores of the African Americans. But even their scores are not\\nparticularly low, except in the UCLA student data set. In an absolute sense,\\nthen, no group is substantially alienated from the American polity, but there\\nis a clear tendency for patriotism to be highest among European Americans\\nand unequivocally lower among America’s most subordinate group, African\\nAmericans.2 \\nThese differences survive our tests for the confounding effects of other\\nsocial attitudes. The test used demographic information such as gender, age\\nand socioeconomic status as proxies for confounding effects that might be\\ncorrelated to a general sense of disaffection resulting from material disad-\\nvantages tied to social status. The analysis was done with multiple regres-\\nsion that controlled for gender, age, and SES while testing for the partial \\neffect of a binary coded ethnic group membership variable. In all of the re-\\ngression analyses, European-American ethnicity was treated as the contrast\\ncategory. Use of this coding scheme had the distinct advantage of providing\\na direct test of the patriotism differences between the dominant ethnic\\n120\\nThe Contribution of Ethnic and National Identities to Political Conflict\\ntable 5.6 Patriotic Attachment to America as a Function of Ethnicity Across Four\\nSamples.\\nEthnic Group\\nResident Alien in US\\nMore\\nthan 5\\nLess than\\nWhites\\nAsians\\nLatinos\\nJews\\nBlacks\\nyrs\\n5 yrs\\nSample 1 (UCLA Students, 1993): F(6,703) = 16.64, p<.001, h = .35\\nMean patriotism\\n.693\\n.628\\n.617\\n.555\\n.432\\n.585\\n.477\\nN\\n153\\n72\\n98\\n32\\n106\\n219\\n30\\nSD\\n.23\\n.20\\n.23\\n.23\\n.24\\n.20\\n.22\\nSample 2 (L.A. County, 1997–98): F(4,1203) = 9.53, p<.001, h = .18\\nMean patriotism\\n.849\\n.806\\n.818\\n—\\n.757\\n.763*\\n—\\nN\\n530\\n66\\n217\\n—\\n138\\n257\\n—\\nSD\\n.20\\n.19\\n.22\\n—\\n.25\\n.23\\n—\\nSample 3 (NES-1992): F(3,2206) = 66.47, p<10–10, h = .29\\nMean patriotism\\n.794\\n.712\\n.789\\n—\\n.589\\n—\\n—\\nN\\n1,736\\n26\\n167\\n—\\n281\\n—\\n—\\nSD\\n.22\\n.22\\n.22\\n—\\n.29\\n—\\n—\\nSample 5 (GSS-1996): F(3,1204) = 8.72, p<.01, h = .14\\nMean patriotism\\n.531\\n—\\n.505\\n—\\n.488\\n.480*\\n—\\nN\\n973\\n—\\n48\\n—\\n38\\n49\\n—\\nSD\\n.11\\n—\\n.13\\n—\\n.13\\n.16\\n—\\n*Indicates that length of residency for this alien group is unknown.\\n\\n\\ngroup (i.e., whites) and each of the subordinate ethnic groups, net of the \\neffects of all other factors.\\nThe results of these analyses for the four different data sets are found in\\ntable 5.7. The table reports only the coefficients for the ethnic group con-\\ntrasts (i.e., white vs. ethnic minority). The coefficients for gender, age, and\\nvarious SES measures are not individually reported because they are only\\nproxies for unexamined control variables that have no theoretical impor-\\ntance.3 The tables show a moderate degree of consistency across data sets:\\nafter considering the effects of the major demographic factors, the patrio-\\ntism differences among ethnic groups in table 5.6 persist in table 5.7. This\\nwas especially true among the college students. Without exception, the\\ncontrasts between socially dominant whites and all of the subordinates\\ngroups showed whites to have significantly higher patriotism scores, with,\\nas before, the most powerful contrast being blacks versus whites (t = –8.60, p\\n< .001, b = –.39). The contrasts between whites on the one hand, and Asian\\nand Latino Americans on the other hand, tended not to be significant\\nwithin the other general population samples. \\nAfrican Americans, though exhibiting relatively high absolute levels of pa-\\ntriotic attachment and attaching relatively high importance to their Ameri-\\ncan identity (see figure 5.1), express less patriotic sentiment than any of the\\nother ethnic groups, and they are more likely to exhibit the exclusionary con-\\nflict between their communal identity and national sentiment. Indeed,\\nAfrican Americans tended to be no more patriotic than resident-aliens (see\\ntable 5.6). In the student data set, the patriotism scores among African Amer-\\nicans were significantly lower than those found among resident-aliens in the\\nUnited States for more than five years (Scheffé post-hoc, p < .001). The effect\\nis apparent even after controlling for demographic differences between\\ngroups: European Americans have a significantly greater level of patriotic \\nattachment than resident-aliens (data set 1: t = –4.79, p < .001; t = –4.54, \\np < .001; dataset 2: t = –2.02 , p < .05; dataset 5: t = –1.83, p < .05). African-\\nCommunal and National Identity in a Multiethnic State\\n121\\ntable 5.7 Patriotism and Ethnocentrism Among Dominant and Subordinate-\\nGroups\\nEuropean- vs. \\nEuropean- vs. \\nEuropean- vs. \\nAfrican-Americans\\nHispanic-Americans\\nAsian-Americans\\nB\\nb\\nB\\nb\\nB\\nb\\nUCLA\\n–.25**\\n–.39**\\n–.06*\\n–.10*\\n–.08*\\n–.10*\\nLACSS 1997/1998\\n–.07**\\n–.11**\\n.01\\n.01\\n–.01\\n–.01\\nNES 1992\\n–.22**\\n–.30**\\n–.00\\n–.00\\n–.00\\n–.02\\nGSS 1996\\n–.04**\\n–.10**\\n.00\\n.00\\nNA\\nNA\\nNote: Numbers under the columns labeled “B” are standardized coefficients from a regression that controls\\nfor gender, age, and socio-economic status. Numbers in the “b” columns are the unstandardized coefficients.\\n**p<.05; *p = .05\\n\\n\\nAmericans, controlling for demographic differences, are never significantly\\nmore patriotic than resident-aliens. Among university students, African\\nAmericans had significantly lower patriotism scores than aliens residing in\\nthe United States five years or longer (data set 1: –t = 5.20, p < .001).\\nIssue 3: Patriotism and Communal Identity\\nBoth the melting-pot and pluralist models of American society expect eth-\\nnocentrism and national attachment to be uncorrelated. This expectation is\\nnot met in these data. Feelings about the nation are related to communal\\nidentity. As one would infer from social dominance theory, this relationship\\nis asymmetrical across the social status hierarchy. Ethnocentrism and na-\\ntional assessments are positively correlated for whites but negatively related\\nfor others, especially African Americans.\\nThe following analysis regressed patriotism on the hierarchical ethnocen-\\ntrism (HEI) index within each major ethnic group within the data sets\\nwhere we had appropriate measures (data sets 1, 2, and 3). The results were\\nreasonably consistent across all three data sets, but were, again, most con-\\nsistent among the university students (see table 5.8). For example, the more\\nwhite students were ethnocentric against African Americans, the more pa-\\ntriotic they were (b = .37, p < .05; see table 5.8). The same general tendency\\n122\\nThe Contribution of Ethnic and National Identities to Political Conflict\\nfigure 5.1 Importance of American Identity as a Function of Ethnicity\\nWhites\\nLatinos\\nResident-aliens\\nBlacks\\nEthnicity\\nImportance of American Identity\\n\\n\\nheld with respect to whites’ HEI against both Latinos and Asians (i.e., b =\\n.44, p < .01, and b = .41, p < .05, respectively). Among subordinates, on the\\nother hand, the exact opposite trend was found. Without exception, ethno-\\ncentrism among subordinates was negatively related to patriotism. Thus,\\nwhile the regression of patriotism upon ethnocentrism (HEI) among whites\\nagainst Latinos was significantly positive (i.e., b = .44, p <. 01), the corre-\\nsponding regression of patriotism on HEI against whites among Latinos\\nwas negative (i.e., b = -.49, p <.01). T\\no test whether or not these asymmetries\\nin the relationships between ethnocentrism and patriotism were statistically\\nsignificant across the status continuum, we used simple slopes analyses to\\ntest for interaction between ethnic group and ethnocentrism (see Aiken &\\nWest, 1993). As can be seen for the student dataset in table 5.8, these slope\\ndifferences were significant in all cases (i.e., F = 19.63, p < .001, F = 20.80, p\\n< .001, F = 17.78, p < .001 for the white vs. black, white vs. Latino, and\\nwhite vs. Asian contrasts, respectively).\\nThe intergroup asymmetries were substantially weaker in the general\\npopulation but generally consistent with expectations. Only two results are\\npatently contrary to the exclusionary patriotism hypothesis: in the Los An-\\nCommunal and National Identity in a Multiethnic State\\n123\\ntable 5.8 Patriotism Regressed upon Hierarchical Ethnocentrism Within Domi-\\nnant and Subordinate Groups (Entries are unstandardized regression coefficients.)\\nEthnocentrism Index among \\nEthnocentrism Index among\\nInteraction\\nDominant Group (i.e., Whites)\\nSubordinate Groups\\nStatistics\\nAmong\\nAmong\\nAmong\\nAmong Whites\\nBlacks\\nLatinos\\nAsians\\nF-statistics\\nWhite-\\nWhite-\\nWhite-\\nBlack-\\nLatino-\\nAsian-\\nfor Slope\\nBlack\\nLatino\\nAsian\\nWhites\\nWhite\\nWhite\\nDifferences\\nUCLA Sample\\n.37*\\n–.47***\\n19.63**\\n.44**\\n–.49***\\n20.80**\\n.41*\\n–.51*\\n17.78**\\nLos Angeles Sample 1997–98\\n–.00n.s.\\n–.09n.s.\\n0.20n.s.\\n–.00n.s.\\n–.19*\\n2.05n.s.\\n–.20**\\n–.10n.s.\\n4.94*\\nNES 1992 Sample\\n.15**\\n–.21**\\n30.09***\\n.16**\\n–.12n.s.\\n8.64**\\n.16**\\n–.19n.s.\\n0.73n.s.\\n*p<.05, **p<.01 ***p<.001\\n\\n\\ngeles sample white ethnocentrism toward blacks and Hispanics is unrelated\\nto patriotism. Every other coefficient is properly signed. The insignificant\\ndifference between whites and Asians in the 1992 NES is a product of a\\nsmall number of cases; however, the numerical value of the differences is\\nsubstantial. Altogether, six of the nine interactions showed that the slope of\\nthe relation between ethnocentrism and patriotism differed between domi-\\nnants and subordinates. For whites, seven of nine correlations between eth-\\nnocentrism and patriotism were positive and significant. For the other eth-\\nnic groups, five of nine correlations between ethnocentrism and patriotism\\nwere negative and significant (see table 5.8).\\nSummary and Discussion\\nThis paper offered three basic ways of thinking about the linkage between\\ncommunal identity and national attachment in American society: (1) the\\ntraditional American melting-pot approach, (2) the pluralist model, and (3)\\nthe group dominance approach. In an effort to assess the relative validity \\nof these different ways of thinking about the intersection of national and\\nethnic identity, we explored three issues (see table 5.1). These issues con-\\ncerned: (1) the salience of ethnic identity and the presence of ethnocentrism\\nagainst other groups of Americans, (2) the relationship between patriotic at-\\ntachment to the nation and ethnic group membership, and (3) the nature of\\nthe interface between communal identity and patriotic attachment.\\nThe results are mixed, but on balance more supportive of expectations\\nproffered by social dominance theory. Ethnic subgroup identities remain a\\nsalient social identity for each of the major groups we examined and, consis-\\ntent with the expectations of social identity theory (T\\najfel & Turner, 1986),\\naverage Americans display clear ethnocentric bias against other ethnic\\ngroups of American citizens. Further, and inconsistent with both the melt-\\ning-pot and pluralism perspectives, patriotic attachment to the nation was\\ngenerally not uniform across ethnic groups. European Americans have\\nstronger patriotic attachment than those from subordinate communal\\ngroups. This tendency was particularly strong and consistent with respect\\nto the white/black contrast. There was not a single instance in which this\\nwhite/black contrast failed to be statistically reliable.\\nFinally, the pluralism image of the society was contradicted by the corre-\\nlation between ethnic and national attachment across ethnic groups. Within\\nthe pluralism model of American society, communal attachments should\\nhave no implication for national attachment. That is not the case, especially\\nfor black Americans. Here, the data were clearly more consistent with the\\ngroup dominance thesis than with the pluralism perspective. The clear\\nasymmetry in the relationship between communal identity and patriotism\\nacross the social status continuum can be regarded as a special case of ideo-\\n124\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nlogical asymmetry discussed by social dominance theory and demonstrated in\\na number of studies (Fang, Sidanius, & Pratto, 1998; Levin & Sidanius,\\n1999; Levin, Sidanius, Rabinowitz, & Federico, 1998; Sidanius, Pratto, &\\nRabinowitz, 1994; Sidanius & Pratto, 1999).4 The robustness of these basic\\nasymmetrical findings is supported not only by their consistency with work\\ngenerated by social dominance research but also by recent work outside of\\nthis tradition (Hofstetter et. al, 1999). \\nThe distinctive relationships observed for black Americans raises the\\nlikelihood that the degree to which communal and national identity are\\nconnected depends upon contextual and situational factors that cannot be\\nsystematically considered in this chapter. For example, the degree to which\\nthe group dominance perspective accurately describes intergroup status re-\\nlationships may well be the critical variable that links communal and extra-\\ncommunal identities. African Americans, almost uniquely, have a history of\\nsubordination that makes it easy and reasonable to associate the institutions\\nof the nation-state (and sentiments such as traditional patriotism) with\\nabuse and oppression. In such an environment it is simple to understand\\nwhy they, relative to Hispanics, Asians, or any other group of immigrants,\\nare inclined to see a conflict between their communal identity and their\\nsentiments about the nation. Many other groups, including European im-\\nmigrants in their first decades in this country, also suffered (Freedman,\\n1996), but the severity of the oppression of American blacks is relatively\\nunique. \\nThe issue that policymakers and interpreters of the American social or-\\nder need to address is whether the patterns characteristic of white/black\\ncommunal and national sentiments can be approximated symbolically—in\\nthe absence of the distinctive history of blacks. We suspect that it can. Peter\\nSkerry’s (1993) analysis of Mexican Americans in San Antonio and Los An-\\ngeles offers affirmative evidence on this point. He persuasively argues that\\nthe ethnic politics practiced by Mexican Americans in San Antonio replicate\\nthe pluralist experiences of European immigrants, with the result that San\\nAntonio Mexican Americans participate in and feel committed to the insti-\\ntutions of the society. In Los Angeles, by contrast, Skerry found an ethos\\nthat defined Mexican Americans as oppressed, a mind-set that created an\\ninsurgent political style, and a dramatically less supportive view of the soci-\\nety’s institutions. In brief, in objectively similar conditions, an ideologically\\nshaped political ethos in Los Angeles created among L.A. Mexican Ameri-\\ncans a subordinate group perspective that established the conditions for ex-\\nclusionary patriotism. We suspect, although the data to demonstrate it are\\nlacking, that the strong asymmetry in the UCLA student data set is a reflec-\\ntion of the “oppressed minority” ethos of Hispanic and Asian politics in the\\nregion. This speculation suggests that future research in this area should ex-\\nplore the material and perceptual situational and contextual factors associ-\\nated with the linkage of communal and national identity.\\nCommunal and National Identity in a Multiethnic State\\n125\\n\\n\\nNotes\\n1. We have doubts about the appropriateness of measuring individualism with\\nthe questions used in their study. They are, more appropriately, viewed as mea-\\nsures of some attitude about the government’s responsibility for the social welfare\\nof the population. This is almost certainly related to an individualism-communi-\\ntarian attitude, but it is a related orientation, not individualism itself. The failure\\nof the three questions to form a reliable scale (de la Garza, 1996, p. 342) is a fur-\\nther indicator of the weakness of the questions as indicators of some underlying\\nsocial orientation.\\n2. However, as mentioned before, unlike the other data sets examined, the 1996\\nGSS data set contained a much richer array of national attachment questions. These\\nadditional questions afforded us the opportunity to examine the connection between\\nethnic and national attachment across additional dimensions of national attachment.\\nThus, besides the dimension of patriotism, we were also able to identify three addi-\\ntional dimensions of national attachment: (1) nationalism—i.e., an aggressive and\\ndominance-oriented stance toward other nations; (2) unwillingness to leave America,\\nand (3) importance of American identity. Most interestingly, the ethnic differences with\\nregard to these three additional dimensions of national attachment were very differ-\\nent from those found for patriotism. First, there were no significant ethnic groups\\ndifferences with respect to nationalism. Second, while there was an overall difference\\nwith respect to Unwillingness to leave America (i.e., F(3,1249) = 10.68, p < .001, h =\\n.16), inspection of Scheffé post-hoc comparisons showed that this difference was es-\\nsentially due to the fact that resident-aliens were significantly less opposed to leav-\\ning the United States than all other American groups—not surprisingly.\\nThe one national attachment dimension where the group differences were quite\\ndifferent from previous results was importance of American identity. Not only did the\\ndifferent ethnic groups show significant overall differences here (F(3,1290) = 13.68, p\\n< .001; h = .18), but the nature of these differences were qualitatively different from\\nthose found with respect to patriotism (see figure 5.1). As seen in figure 5.1, while\\none should expect resident-aliens to have the lowest level of American identity,\\ngiven what we have already seen, it is not immediately obvious that African Ameri-\\ncans should have the highest level of American identity. Furthermore, controlling\\nfor demographic differences did not substantially change this picture. The analyses\\nshowed that while the resident-aliens had significantly lower American identity\\nscores than the dominant European Americans (t = –3.74, p < .001), African Ameri-\\ncans had unambiguously greater American identity scores than the dominant Euro-\\npean Americans (t = 3.50, p < .001). Rerunning these regression analyses and recod-\\ning ethnicity using blacks as the contrast group revealed that, everything else being\\nequal, blacks also had significantly greater American identity than all groups except\\nLatinos.\\n3. Gender had a significant relationship to patriotism in all the data sets; age was\\nsignificant in every data set except the UCLA data (where age differences were at-\\ntenuated). Patriotism was higher among males and older people. Income, education,\\nand similar class measures were usually of trivial importance.\\n4. The ideological asymmetry hypothesis posits that while the correlations be-\\ntween ingroup bias and certain system-justifying ideologies will be positive among\\ndominants, these correlations will be negative among subordinates.\\n126\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nReferences\\nAiken, L. S., & West, S. G. (1993). Multiple regression: Testing and interpreting interac-\\ntions. Newbury Park, CA: Sage Publications. \\nAlba, R. D. (1990). Ethnic identity: The transformation of white America. New Haven:\\nYale University Press.\\nAllen, J. P., & Turner, E. (1997). The ethnic quilt. Population diversity in Southern \\nCalifornia. The Center for Geographical Studies, Department of Geography,\\nCalifornia State University, Northridge.\\nBar-tal, D., & Staub, E. (1997). Patriotism in the lives of individuals and nations.\\nChicago: Nelson-Hall.\\nBlumer, H. (1960). Race prejudice as a sense of group position. Pacific Sociological \\nReview,1, 3-5.\\nDahl, R. A. (1961). Who governs? New Haven: Yale University Press.\\nde la Garza, R. O., Falcon, A., & Garcia, F. C. (1996). Will the real Americans\\nplease stand up: Anglo and Mexican-American support of core American politi-\\ncal values. American Journal of Political Science, 40, 335–351.\\nFang, C. Y., Sidanius, J., & Pratto, F. (1998). Romance across the social status con-\\ntinuum: Interracial marriage and the ideological asymmetry effect. Journal of\\nCross Cultural Psychology, 29, 290–305.\\nFredrickson, G. M. (1981). White supremacy: A comparative study in American and\\nSouth African history. Oxford: Oxford University Press.\\nFreedman, S. G. (1996). The inheritance: How three families and America moved from\\nRoosevelt to Reagan and beyond. New York: Simon and Schuster.\\nGordon, M. (1964). Assimilation in American life: The role of race, religion, and national\\norigins. New York: Oxford University Press.\\nGreeley, A. M. (1971). Why can’t they be like us. New York: Dutton. \\nGreeley, A. M. (1976). Ethnicity: A preliminary reconnaissance. New York: Wiley.\\nGurr, T. R., & Harff, B. (1994). Ethnic conflict in world politics. Boulder, CO: \\nWestview.\\nHofstetter, C. R., Feierabend, I. K., & Klicperova-Baker, M. (1999). Nationalism\\nand ethnicity: A community study. Paper delivered at the Annual Meeting of the\\nInternational Society for Political Psychology, Amsterdam, The Netherlands,\\nJuly 17–21, 1999.\\nHorowitz, D. L. (1985). Ethnic groups in conflict. Berkeley: University of California\\nPress.\\nHuckfeldt, R. & Kohfeld, C. W. (1989). Race and the decline of class in American poli-\\ntics. Urbana: University of Illinois Press.\\nJackman, M. (1994). The velvet glove: Paternalism and conflict in gender, class and race\\nrelations. Berkeley: University of California Press.\\nKelley, R. (1979). The cultural pattern in American politics: The first century. New York:\\nKnopf.\\nKleppner, P. J. (1970). The cross of culture. New York: Free Press.\\nKosterman, R., & Feshbach, S. (1989). T\\noward a measure of patriotic and national-\\nistic attitudes. Political Psychology, 10, 257–274. \\nLevin, S., & Sidanius, J. (1999). Social dominance and social identity in the United\\nStates and Israel: Ingroup favoritism or outgroup derogation? Political Psychology,\\n20, 99–126.\\nCommunal and National Identity in a Multiethnic State\\n127\\n\\n\\nLevin, S., Sidanius, J., Rabinowitz, J. L., & Federico, C. (1998). Ethnic identity, \\nlegitimizing ideologies and social status: A matter of ideological asymmetry. \\nPolitical Psychology, 19, 373–404.\\nLijphart, A. (1977). Democracy in plural societies: A comparative exploration. New\\nHaven: Yale University Press.\\nLijphart, A. (1979). Religious vs. linguistic vs. class voting: the “crucial experiment”\\nIn comparing Belgium, Canada, South Africa, and Switzerland. American Political\\nScience Review, 73 (June), 442–458.\\nLipset, S. M., & Rokkan, S. (1967). Cleavage structures, party systems, and voter\\nalignments: An introduction. In S. M. Lipset & S. Rokkan, (Eds.), Party systems\\nand voter alignments (pp. 1–64) New York: Free Press.\\nNovak, M. (1971). The rise of the unmeltable ethnics. New York: Macmillan.\\nParenti, M. (1967). Ethnic politics and the persistence of ethnic voting. American\\nPolitical Science Review, 61, 717–725.\\nPetrocik, J. R. (1981). Party coalitions: Realignments and the decline of the New Deal\\nparty system. Chicago: University of Chicago Press.\\nPetrocik, J. R. (1998). Reformulating the party coalitions: The “Christian Democra-\\ntic Republicans.” Paper delivered at the 1998 annual meeting of the American\\nPolitical Science Association. Sep 3-6. Boston Marriot Copley Place Hotel. \\nPolsby, N. W. (1963). Community power and political theory. New Haven, Yale Uni-\\nversity Press.\\nPowell, C. L. (1995). My American journey. New York: Random House.\\nPowell, G. B. Jr. (1982). Contemporary democracies: Participation, stability, and violence.\\nCambridge: Harvard University Press.\\nPratto, F., Sidanius, J., Stallworth, L. M., and Malle, B. F. (1994). Social dominance\\norientation: A personality variable predicting social and political attitudes. Jour-\\nnal of Personality and Social Psychology, 67, 741–763.\\nSalins, P. D. (1997). Assimilation, American style. New York: Basic Books.\\nSchlesinger, A. M., Jr. (1992). The disuniting of America. New York: Norton.\\nSidanius, J. (1993). The psychology of group conflict and the dynamics of oppres-\\nsion: A social dominance perspective. In S. Iyengar & W. McGuire (Eds.), Explo-\\nrations in political psychology (pp. 183–219). Durham, NC: Duke University Press. \\nSidanius, J., Feshbach, S., Levin, S., & Pratto, F. (1997). The interface between eth-\\nnic and national attachment: Ethnic pluralism or ethnic dominance? Public Opin-\\nion Quarterly, 61, 103–133.\\nSidanius, J., & Pratto, F. (1993). The inevitability of oppression and the dynamics of\\nsocial dominance. In P. Sniderman & P. T\\netlock (Eds.), Prejudice, Politics, and the\\nAmerican dilemma. (pp. 173–211). Stanford University Press. \\nSidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierar-\\nchy and oppression. New York: Cambridge University Press.\\nSidanius, J., Pratto, F., & Rabinowitz, J. (1994). Gender, ethnic status, ingroup at-\\ntachment and social dominance orientation. Journal of Cross-Cultural Psychology,\\n25, 194–216.\\nSkerry, P. (1993). Mexican-Americans: The ambivalent minority. New York: Free Press.\\nSmith, T. W. (1991). What do Americans think about Jews? Working papers on contempo-\\nrary anti-Semitism. New York: The American Jewish Committee, Institute of\\nHuman Relations.\\nT\\najfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior.\\n128\\nThe Contribution of Ethnic and National Identities to Political Conflict\\n\\n\\nIn S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7–24).\\nChicago: Nelson-Hall.\\nWalch, T. (1994). Immigrant America: European ethnicity in the United States. New\\nYork: Garland Publishing.\\nWolfinger, R. E. (1965). The development and persistence of ethnic voting. Ameri-\\ncan Political Science Review, 59: 896–908.\\nCommunal and National Identity in a Multiethnic State\\n129\\n\\n\\nThis page intentionally left blank \\n\\n\\nPART III\\nTHE CONTRIBUTION OF \\nSOCIAL IDENTITY TO \\nVIOLENT INTERGROUP \\nCONFLICT\\n\\n\\nThis page intentionally left blank \\n\\n\\nRobert W. White\\n6\\nSocial and Role Identities \\nand Political Violence\\nIdentity as a Window on Violence in Northern Ireland\\nMuch of the violence in the contemporary world appears, on the surface, to\\ninvolve two groups of people engaged in “them versus us” behavior. From\\nthis view, group members develop a social (group) identity in comparison\\nwith members of some other group. This comparative self-definition leads\\nto heightened differences across some social category, for example, across\\ntribes or ethnic groups, and, ultimately, violence. Examples of intergroup\\nviolence based on social identity might include Tutsi versus Hutu, Arab ver-\\nsus Israeli, and Serb versus Croat. I suspect, however, that much of this vio-\\nlence does not reduce to a straightforward “them versus us” situation.\\nIt has been argued that violence in Northern Ireland, as an example, in-\\nvolves individuals with competing social identities based on religion—of\\nIrish Protestants (the majority community) in conflict with Irish Catholics\\n(the minority community); for example, Whyte, 1990, p. 97; Cairns, 1982.\\nThis view is subject to question. The Irish Republican Army (IRA) has\\nbeen the most violent actor in Northern Ireland since 1969 (as measured\\nby deaths caused). Typically, the IRA is viewed as a Catholic defense orga-\\nnization. Yet, the origins of the IRA lie in the Irish Protestant community,\\nand there have been some Protestant members of the organization (see\\nMcKittrick, Kelters, Feeney, & Thornton, 1999, pp. 462–463; White,\\n1997a). \\nThe complexity of the political violence in Northern Ireland may be\\ndemonstrated by a relatively straightforward examination of who has killed\\nwhom since 1969.1 The examination shows that there are three broadly de-\\nfined violent actors: Irish Republican paramilitaries, Protestant (or “Loyal-\\nist”—loyal to the British crown) paramilitaries, and the security forces.2\\n133\\n\\n\\nT\\nable 6.1 presents the number of persons killed by each actor for each \\n“status” of victim in the Northern Ireland conflict (Sutton, 1994). This is\\none of the world’s most studied conflicts, and these data are consistent with\\nother attempts to enumerate the dead (see White, 1997a; McKittrick, et al.,\\n1999, pp. 15–19, 1473–1493; Whyte 1990, p. viii).\\nThe primary target of Protestant paramilitaries is Catholic civilians. Six\\nhundred and twelve (67 percent) of their 911 victims were Catholic civil-\\nians. Only 26 (of 911—3 percent) of their victims were Irish Republicans.\\nIrish Republicans killed the most people. And, Irish Republicans killed a\\nlarge number of Protestant civilians (20 percent, or 377/1,926). However,\\nthe single largest group of Irish Republican victims was members of the\\nBritish army. Adding in members of the Royal Ulster Constabulary (RUC),\\n134\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\ntable 6.1 Agency Causing Deaths and the Status of the Victims, Northern \\nIreland, 1969–1993\\nStatus of Agent\\nSecurity Forces\\nStatus of\\nIrish Republican\\nProtestant\\nBritish\\nRUC-\\nVictim\\nParamilitaries\\nParamilitaries\\nArmy \\nUDR/RIR1\\nOther\\nT\\notal\\nBritish Army\\n494\\n1\\n3\\n2\\n3\\n503\\nRUC\\n280\\n6\\n2\\n1\\n3\\n292\\nUDR/RIR\\n196\\n3\\n2\\n0\\n1\\n202\\nRepublican\\n154\\n26\\n123\\n18\\n5\\n326\\nParamilitaries\\nProtestant\\n27\\n61\\n11\\n2\\n1\\n102\\nParamilitaries\\nCatholic\\n181\\n612\\n134\\n30\\n30\\n987\\nCivilians\\nProtestant\\n377\\n111\\n17\\n6\\n27\\n538\\nCivilians\\nPolitical\\n25\\n31\\n1\\n2\\n1\\n60\\nActivists\\nPrison\\n23\\n2\\n0\\n0\\n0\\n25\\nOfficers\\nFormer\\n62\\n0\\n0\\n0\\n0\\n62\\nSecurity\\nOfficers\\nOther\\n107\\n58\\n3\\n1\\n7\\n176\\nT\\notals\\n1926\\n911\\n296\\n62\\n78\\n3273\\nSource: Sutton, 1994.\\n1. The Royal Ulster Constabulary and the Ulster Defence Regiment/Royal Irish Regiment. The RUC is\\nthe local police force in Northern Ireland. The UDR, a locally recruited regiment of the British Army, was\\nreplaced by the Royal Irish Regiment.\\n\\n\\nthe police force in Northern Ireland, and the Ulster Defence Regiment\\n(UDR; a locally recruited regiment of the British Army, later replaced by\\nthe Royal Irish Regiment; RIR) shows that about half (970 of 1926) of the\\nvictims of Republican paramilitaries were members of the security forces.\\nIrish Republicans do not exclusively, or even primarily, target the Protestant\\ncommunity in Northern Ireland (see also White, 1997a, 1997b; Bruce,\\n1997; Drake, 1998).\\nState agents have killed a relatively large number of people in Northern\\nIreland. The British Army killed 296 people. The Royal Ulster Constabu-\\nlary and the Ulster Defence Regiment/Royal Irish Regiment killed 62 peo-\\nple. The two communities have not suffered equally at the hands of the\\nstate. It is not surprising that state agents have killed a large number of Irish\\nRepublican paramilitaries, given that Republicans shoot and bomb them.\\nHowever, state agents killed 164 Catholic civilians compared to 23 Protes-\\ntant civilians (and 131 Republican paramilitaries). In Northern Ireland the\\nstate is not a neutral actor (see also White & White, 1995; White, 1999;\\nThatcher, 1993, pp. 384–385). T\\nable 6.1 suggests that between 1969 and\\n1993, Irish Republicans and the security forces confronted each other with\\nviolence, while Protestant paramilitaries directed their violence against\\nCatholic civilians (see also note 3).\\nMy objective is to draw on the concepts of social and role identity to pro-\\nvide insight into the complexity of violence in places like Northern Ireland.\\nThis violence does not reduce to a “them versus us” model, but rather in-\\nvolves a complex mix of identities. I suspect that several conflicts in the\\nworld today, upon in-depth examination, will be shown to be as complex as\\nthe Northern Ireland conflict. With respect to Northern Ireland, I argue\\nthat the violence of Protestant paramilitaries may be understood in “them\\nversus us,” or in Irish Protestant versus Irish Catholic, terms. Protestant\\nparamilitaries target their opponents because they are Catholic or Irish\\nCatholic. Irish Republican paramilitary activity is best understood as a com-\\nbination of behavior based on a social identity and behavior based on a role\\nidentity. However, for Republicans the social identity dimension of the con-\\nflict is not Irish Catholic versus Irish Protestant in Northern Ireland, but\\ninstead Irish Nationalist versus the British government and its agents, in\\nBritain and Ireland. Also, an Irish Republican role identity (described be-\\nlow) complements the Irish social identity. Finally, the British view the con-\\nflict as an Irish problem—of Irish Protestants versus Irish Catholics (them\\nvs. them). This has important implications for the behavior of British elites,\\nBritish soldiers, and Protestant and Republican paramilitaries, as demon-\\nstrated below.\\nLike many other conflicts in the world, the violence in Northern Ireland\\nis centuries in the making. Historical background is therefore presented\\nfirst. An in-depth examination of the various identities at work in Northern\\nIreland follows.\\nSocial and Role Identities and Political Violence\\n135\\n\\n\\nThe Historical Setting: Background to the Complexity\\nNorman/English soldiers invaded Ireland in the 12th century. From then\\nuntil the mid-sixteenth century, political violence in Ireland was primarily\\nmotivated by what today is referred to as national identity. The English\\nfound a people different from themselves in language, customs, and outlook\\non life. For the next four hundred and so years there were numerous violent\\nconflicts between native Irish and invading English (Cronin, 1980).\\nIn the 16th century, with the Reformation, religion became important.\\nThe Irish and Old English in Ireland remained Catholic. Most people in\\nGreat Britain embraced Protestantism. In 1601, Irish Chiefs rebelled and\\nlost. A few years later they fled to the European continent. Their lands—in\\nUlster in the North of Ireland—were confiscated and given to English and\\nScottish settlers. In the 1640s, the native Irish in Ulster rebelled. They were\\nsupported by the Old English in Ireland. Both groups were Catholic. From\\nthis point religion and national identity became intertwined in Anglo-Irish\\npolitics. When Oliver Cromwell invaded in 1649 he did so with particular,\\nProtestant religious zeal. His zeal was also a function of the fact that it was\\nIrish Catholics who were being pacified; Cromwell never treated Catholics\\nin England as harshly as he did Catholics in Ireland (Fraser, 1973/1997;\\nCronin, 1980, pp. 488–489). There was a further confiscation of Irish land\\nand property, which were placed in trustworthy, Protestant hands. In the\\nlate 1680s, William of Orange seized the throne in England. James II, a\\nCatholic, fled to Ireland. He was supported by his fellow Catholics in Ire-\\nland, who sought relief from Protestant domination. William’s army de-\\nfeated James’s army at the Battle of the Boyne, and for the third time in a\\ncentury the Irish suffered the results of a lost war (e.g., O’Leary & Mc-\\nGarry, 1993). \\nThese 17th century wars established a Protestant ascendancy in Ireland.\\nBy the end of the century roughly 10 percent of Irish land was in Catholic\\nhands, though perhaps 90 percent of the population was Catholic (i.e., “na-\\ntive Irish”; O’Leary & McGarry, 1993, pp. 66–69). The wars were about \\nreligion and political and social control of Ireland. The native Irish were\\nvirtually excluded from Irish social and political affairs. In the process, a\\nsiege mentality was generated among the Protestant elite. Placed in control\\nof a conquered people were a Protestant Ascendancy that was vastly out-\\nnumbered and constantly fearful.\\nIn the late 18th century the political philosophy of Republicanism was\\nintroduced into the mix. The United Irishmen, the first Irish Republicans,\\nwere founded by Presbyterian merchants and manufacturers in Belfast and\\nDublin. They were influenced by the French and American Revolutions,\\nand by a desire for economic and political freedom. They found allies in\\nIrish Catholics, who sought political and social relief. Republican ideology\\nis a political philosophy, not a religious or ethnic one. In the 1790s, it was a\\nradical philosophy. The United Irishmen sought such reforms as universal\\n136\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\n(male) suffrage and no property qualification (see Cronin, 1980). The\\nUnited Irishmen rebelled in 1798. Unlike the republicans in the United\\nStates and France, they failed.\\nIn response to the violence in Ireland, the Act of Union (effective Janu-\\nary 1, 1801) disbanded an independent Irish parliament and created the\\nUnited Kingdom of Great Britain and Ireland. This did not end political \\nviolence in Ireland. Irish Nationalists rejected the Union and aspired for an\\nindependent Irish nation. Irish Republicans sought to achieve this nation\\nthrough force of arms. There were Republican rebellions in 1803, 1848,\\nand 1867. The rebellions failed, but there remained in Ireland persons\\ncommitted to an independent Irish Republic. Throughout the 1870s and\\n1880s there was Republican-inspired violence. Throughout the 19th cen-\\ntury, the general Protestant community in Ireland—“Unionists”—sup-\\nported the Union. This was especially so in northeast Ireland, where the\\ndescendants of the 17th-century settlers feared that “Home Rule” would\\nlead to “Rome Rule.” In Belfast in the 1880s, for example, there was wide-\\nspread rioting in opposition to a Home Rule Bill for Ireland (see Cronin,\\n1980; Lyons, 1973).\\nNorthern Ireland dates from events in 1916–1922. In 1916, there was a\\nrebellion the day after Easter. The rebels lost, but they reorganized as the\\nIrish Republican Army (IRA) and the political party Sinn Féin (founded\\n1905). The two organizations worked to break the connection with the\\nUnited Kingdom. They established a revolutionary Irish government in\\nDublin and carried out military activities against the British presence in Ire-\\nland. The IRA and Sinn Féin were successful throughout most of Ireland.\\nIn the northeast, the IRA and Sinn Féin were opposed by both the state’s\\nsecurity agents—for example, the Royal Irish Constabulary—and by a\\nProtestant paramilitary organization, the Ulster Volunteer Force. In re-\\nsponse to the political unrest, the British passed the Government of Ireland\\nAct (1920), which partitioned Ireland. The Act created Northern Ireland,\\nwhich comprised six of the nine counties of the province of Ulster, and the\\nIrish Free State. The parliamentary headquarters for Northern Ireland was\\nestablished in Belfast. Dublin became the seat of power for the Free State.\\nAt its creation, Northern Ireland was approximately two-thirds Protestant\\nand one-third Catholic. The Irish Free State (declared a republic in 1949),\\nwas (and is) approximately 95 percent Catholic.\\nIn 1921, negotiations between Republicans and British representatives\\nled to a treaty. The IRA and Sinn Féin split over the Anglo-Irish Treaty\\n(1921), and a civil war ensued. Those Republicans opposed to the treaty\\nmaintained the IRA and Sinn Féin, and fought against former comrades\\nwho accepted the treaty and the new Irish Free State. Pro-treaty forces won\\nthe war, and, among other things, this cemented partition. Those who re-\\nmained in Sinn Féin and the IRA refused to recognize partition and the\\ngovernments in Dublin and Belfast. They also maintained their organiza-\\ntions. From 1939 to 1945 and from 1956 to 1962, the IRA engaged in un-\\nSocial and Role Identities and Political Violence\\n137\\n\\n\\nsuccessful paramilitary campaigns in England and Northern Ireland, in an\\nattempt to create a united Ireland (see Bell, 1979).\\nIn Northern Ireland, the Protestant community embraced the new\\nprovince. The Catholic community, both in Northern Ireland and in the\\nIrish Free State, largely rejected it. Thus, partition created a double minority\\nsituation. In Northern Ireland Protestants dominated a minority Catholic\\ncommunity who believed they had been gerrymandered into a Protestant\\nstate. Within Ireland as a whole the Protestant community was in the minor-\\nity. The government of the Free State/Republic rejected Northern Ireland\\nand claimed the entire island for itself.\\nIn the 1960s, there was a civil rights campaign in Northern Ireland. The\\ncampaign sought to redress grievances that Nationalists/Catholics had\\nfaced since the founding of the province. Civil rights protests led to coun-\\nterdemonstrations by Protestant extremists, which led to widespread riot-\\ning in Northern Ireland in August 1969. The British government sent in\\ntroops to restore order. It is out of events in the mid-to late 1960s that the\\nthree actors of table 6.1 are derived: Irish Republican paramilitaries, Protes-\\ntant paramilitaries, and the security forces (Bell, 1979, 1993).\\nOrganizationally, Irish Republicans can legitimately trace themselves to\\nthe rebellion in 1848. Veterans of 1848 were involved in rebellious activity\\nin the 1860s, and some of those involved in the 1860s (and 1870s and 1880s)\\ntook part in the 1916 Easter Rising (see Bell, 1979, 1987, p. 136). In 1969\\nand early 1970, the IRA and its political wing, Sinn Féin, split again and the\\nProvisional IRA (hereafter, IRA) and Provisional Sinn Féin (hereafter, Sinn\\nFéin) were created. The founders of the Provisionals were Republican vet-\\nerans, some with activist histories that dated to 1916. Contemporary\\nProtestant paramilitary organizations date from the mid-1960s and the cre-\\nation of the Ulster Volunteer Force (named for the UVF of the 1910s). In\\n1971 the Ulster Defence Association (UDA) was started as a coordinating\\nbody for vigilante groups. The Ulster Freedom Fighters (UFF) became the\\nparamilitary wing of the UDA (see Flackes & Elliott, 1989, pp. 272–277,\\n278).\\nPolitical violence in Northern Ireland is hundreds of years in the mak-\\ning. It is not simply about religion, but instead involves a complex mix of\\nnational identity, religion, and political aspirations. Social identities and\\nrole identities, as described below, provide an insight into this complexity.\\nIdentities: Social (Group) and Role\\nAccording to T\\najfel (1981, p. 255), social identity is that “part of an indi-\\nvidual’s self-concept that derives from his knowledge of his membership of\\na social group.” T\\najfel typically refers to groups based on race, ethnicity, or\\nreligious identification. For T\\najfel, the need to view oneself positively may\\nheighten differences between groups and lead to conflict. While T\\najfel fo-\\n138\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\ncuses on intergroup relations, Turner, Hogg, Oakes, Reicher, & Blackwell\\n(1987) extend the social identity perspective to emphasize intragroup be-\\nhavior. They define social identity as self-categorizations that “define the\\nindividual in terms of his or her shared similarities with members of certain\\nsocial categories in contrast to other social categories” (Turner, Oakes,\\nHaslam, & McGarty, 1994, p. 454). Thoits and Virshup (1997) summarize\\nthese two perspectives by noting that they focus on “we’s,” in the sense that\\nsocial identities are “collective identities.” The focus is on how the indi-\\nvidual and the group mesh together. While the social groups around which\\na collective “weness” develops are often described in terms of race, gender,\\nand ethnicity, a social identity may also develop around a political group-\\ning—for example, the Ulster Volunteer Force or the Irish Republican\\nArmy. My use of the term “social identity” is consistent with Thoits and\\nVirshup’s (1997) usage, in which they refer to a collective “weness” that ex-\\nists in comparison to a collective “other.”\\nRole identities focus on “me’s” (Thoits & Virshup, 1997). Role identities\\nact as “internalized positional designations” (Stryker, 1980, pp. 60–61; see\\nalso McCall and Simmons, 1978) that locate and define individuals within a\\ngroup. Role identities refer to individual-level self-conceptions of the re-\\nquirements attached to social positions. Individuals internalize these re-\\nquirements, which become identified as a part of the self. Role identities\\nhave “salience” for individuals (Stryker, 1980, 1968). In certain situations\\none role identity may be more salient for an individual than another role\\nidentity. In addition, there is the potential for significant role conflict and\\nrole strain—when the expectations associated with roles conflict or compete\\n(Stryker, 1980). My use of the term “role identity” follows the definition of\\nStryker and refers to self-conceptions of the requirements attached to a so-\\ncial position that develop from social interaction—for example, by learning\\nto be a Volunteer in the Irish Republican Army from other IRA Volunteers.\\nThe social or group (collective) identity is conceived of as a group-level\\nphenomenon, where the individual conceives of the self as one of “us”—\\noften in contrast to “them.” The focus is on cognitive processes, comparing\\none’s group to some other group. As Thoits and Virshup (1997, p. 115)\\nstate, “Group- or category-based identities are collective-level self-concep-\\ntions; they are identifications of the self with a collectivity.” In contrast, role\\nidentities focus on who the individual is. Instead of comparing “us” vs.\\n“them,” the individual considers “Who am I?” by drawing on social interac-\\ntion and incorporating the expectations of other people into their self-\\nconceptualization. Both kinds of identity are social, but in different ways\\n(see also Stryker, 2000; Klandermans and de Weerd, 2000). For those who\\nemploy the concept of social identity, the self is a product of comparison\\nand categorization processes in which “the perceiver appraises the self in re-\\nlation to others.” For those who employ the concept of role identity, the self\\nis a product of incorporating others’ expectations and appraisals, of apprais-\\ning the self from the perspective of others—for example, in role-taking \\nSocial and Role Identities and Political Violence\\n139\\n\\n\\nexercises (see Thoits & Virshup, 1997, pp. 120–122). In terms of violent \\npolitical activists in Northern Ireland, persons may kill on behalf of their\\ngroup, and/or they may kill because they have internalized such behavior as\\npart of the role of being a particular political activist. Indeed, activists may\\nkill because they are motivated simultaneously by both a social identity and\\na role identity.\\nIn the following, I argue that the violence of Protestant paramilitaries is\\nbest understood from a social identity perspective. Protestant paramilitaries\\nact in defense of their social group by killing members of the other (Catholic)\\nsocial group. In contrast, I argue that Irish Republican paramilitaries are mo-\\ntivated by both social and role identities. Irish Republicans target the security\\nforces in Ireland because these forces represent an oppressor of their “Irish-\\nness” (the British government) and because they have internalized the role\\nrequirements of being an Irish Republican. Finally, I argue that the British\\nperspective may be understood in social identity terms. However, instead of\\ntaking a “them versus us” perspective—for example, British forces versus\\nIrish Republican forces—the British take a “them versus them” perspective\\nand view the violence as an Irish problem, involving Irish Protestants versus\\nIrish Catholics.3\\nData Sources: Accounts from Activists\\nThe accounts below are from very different sources. None of the data were\\ncollected with this particular research topic in mind. The accounts from\\nIrish Republicans are from my ongoing research on the causes and conse-\\nquences of small-group political violence. These accounts are taken from\\ninterviews that centered on why the respondents became involved in the\\nIrish Republican movement. The Republican accounts represent my inter-\\npretation of how Irish Republicans may be understood in terms of social\\nand role identities. This interpretation is based on interaction with Irish\\nRepublicans over a 17-year period. I have formally interviewed perhaps 100\\nIrish Republicans, and have informally met with an unknown number of\\nothers. Many of the respondents were low-level Sinn Féin workers, such as\\nadvice center staffers. Included among the respondents are several promi-\\nnent Republicans and several veterans of the IRA. Many of the respondents\\nwere born and raised in families where a parent or other significant adult\\nwas active in Republican politics or had strong Republican sympathies (see\\nWhite, 1993; White & Fraser, 2000).\\nThe accounts from Protestant paramilitaries, British elites, and British\\nsoldiers are taken from secondary sources. These sources include interviews\\nwith journalists, memoirs, and published accounts of speeches. I did not\\ncollect these data. My interpretation of the accounts of Protestant paramili-\\ntaries, British elites, and British soldiers, therefore, is qualified. With re-\\nspect to Protestant paramilitaries, I present the accounts to illustrate the\\n140\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nconventional wisdom with respect to these people—that is, that the motive\\nof these paramilitaries is sectarian and is based on a strong identification\\nwith being Protestant. This conventional wisdom is based on solid social\\nscience research (e.g., Bruce, 1992). The statements from British elites and\\nBritish soldiers are consistent with survey data, noted below, which shows\\nthat Northern Ireland is not a high-priority item for the British public.\\nThis is one indicator that British people may interpret the conflict as one of\\n“them versus them.”\\nNorthern Ireland: The Actors and Their Identities\\nIn Northern Ireland, political actors represent two communities that reside\\nthere and one external interest group: the Protestant community, the Irish\\nNationalist community, and British representatives. Within the Protestant\\ncommunity, there are Unionists and Loyalists. Unionists want Northern\\nIreland to remain part of the United Kingdom. Loyalists are Unionists who\\ntake an extreme position on the Union and are willing to defend it with\\nparamilitary activity. Irish Nationalists, who are predominantly Catholic,\\nseek an end to the Union and a united Ireland. Among Irish Nationalists are\\nIrish Republicans, who are willing to use paramilitary activity to bring\\nabout a united Ireland. British interests are represented by political ap-\\npointees, for example, the secretary of state for Northern Ireland, and per-\\nsons involved with the security apparatus, such as the British army.\\nThe Protestant Community: Unionists.\\nIn many ways, the Unionist mentality is a siege mentality. In spite of\\ntheir political power in Ireland, Unionists have always been in the minority,\\nand they have, therefore, always perceived their position to be threatened.\\nWhen it was created, Northern Ireland offered Unionists a gerrymandered\\nmajority situation, but this was not guaranteed to last forever. Historically,\\nthe Protestant ascendancy in Ireland maintained their position only with\\nthe help of the English and British state. The British parliament created\\nNorthern Ireland as a haven for pro-British Protestants; the Protestant\\ncommunity in Northern Ireland did not do this on its own.\\nT\\noday, most Protestants in Northern Ireland are Unionists (there are\\nsome Catholic Unionists). In addition, many Protestants in Northern Ire-\\nland identify not with Ireland but with Britain. When asked in 1968, on the\\neve of the current violence, which term best describes themselves (Irish,\\nBritish, Ulster, Sometimes British-Sometimes Irish, Anglo-Irish), only 39\\npercent of Protestants chose “British.” An additional 32 percent chose “Ul-\\nster.” Perhaps surprising, the first choice of 20 percent of Protestants to this\\nquestion was “Irish” (see Rose, 1971, p. 485). In 1978, in the midst of a vio-\\nlent conflict, the question was repeated on a survey by Edward Moxon-\\nSocial and Role Identities and Political Violence\\n141\\n\\n\\nBrowne. Moxon-Browne found that the number of Protestants whose first\\nidentity was “British” increased significantly, to 67 percent. The number\\nwho chose “Ulster” fell to 20 percent, and only 8 percent chose “Irish”\\n(Moxon-Browne 1983, pp. 5–7).\\nThe Protestant Community: Loyalists\\nLoyalists (loyal to the British crown) tend to be working-class Unionists\\nwho view violent opposition to Irish nationalism as both justified and ne-\\ncessary; they support, or are, Protestant paramilitaries (see Bruce, 1992, \\npp. 14–15; Nelson, 1984). There is general agreement that Protestant para-\\nmilitaries are motivated by sectarianism, by deep-seated mistrust, and by\\nhatred of Catholics. Steve Bruce, an authority on Protestant paramilitaries,\\nstated in 1994:\\nAlthough leaders of the UDA [Ulster Defence Association] and UVF [Ulster\\nVolunteer Force] have periodically tried to confine the violence of their mem-\\nbers to some notion of “legitimate targets” (that is, republicans) or have\\nclaimed to be trying so to restrict the deaths, very many of their victims were\\nchosen simply because they were Catholics. Even in these more publicity-\\nconscious days there are loyalist paramilitary leaders who say that they like to\\nkill republicans but an ordinary Catholic will do. (1994, p. 125)\\nThe data in table 6.1 support this interpretation. The majority of victims\\nof Protestant paramilitaries are Catholic civilians. Because Catholics tend to\\nbe Irish Nationalists, they are by (Protestant paramilitary) definition trai-\\ntors and subject to violence. Protestant paramilitaries punish the general\\nCatholic community when they are threatened or want to hit back. Because\\nIrish Republicans are hard to target, the objective is to create enough fear in\\nthe Catholic community that they, in turn, will force the Republican para-\\nmilitaries to quit. At its most basic level, Loyalists view themselves as fight-\\ning for tribal survival (Bruce, 1992, p. 55). \\nIn identity terms, Protestant paramilitaries view the world from a\\nProtestant social identity, in opposition to Catholics. Consider the follow-\\ning from a former Protestant paramilitary, John White. White was a lead-\\ning member of the Ulster Freedom Fighters (UFF). He describes his up-\\nbringing:\\nI remember my mother, an uneducated working-class woman, saying to me,\\nSon, we don’t go into a united Ireland, and me saying, why not, Mammy, and\\nI remember her saying, because you’ll not get a job, and you’ll probably not\\nget a house, and they’ll murder us, you know. And my Da, he just had a con-\\nfirmed hatred for Roman Catholicism. I remember one time my brother\\nbrought a Roman Catholic to the house, a girlfriend, and my Da told her to\\nget out of the house and not come back again. So, you know, I thought there\\nmust be some genuine fears there, when even my Da was saying, no bringing\\nT\\naigs home to this house. (Stevenson, 1996, p. 62)4\\n142\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nIn 1976, White confessed to the 1973 murders of Paddy Wilson and\\nIrene Andrews. Wilson was a member of the Social, Democratic, and\\nLabour Party (SDLP), a moderate, middle-class political party associated\\nwith the Catholic community in Northern Ireland. The SDLP then, and\\nnow, advocated a united Ireland, but through peaceful means. Now out of\\nprison, John White commented on why Wilson, who was publicly known\\nto be unarmed, was a target:\\nMy analysis then was that he [Wilson] supported the concept of a united Ire-\\nland; the IRA were trying to pursue a united Ireland; therefore, by his\\nrhetoric, he was supporting the IRA. It may have been warped at that time,\\nbut that was the way I viewed it when I was twenty two years of age. (Steven-\\nson, 1996, pp. 59–61)\\nJohn White’s account is not unique among former Protestant paramili-\\ntaries. The following is from Alex Calderwood, a veteran of the Ulster De-\\nfence Association (UDA). He states, “I grew up hating Catholics.” \\nI admit that I was a sectarian bigot. . . . It was my background and the soci-\\nety that programmed me in a sense, but I was responsible for what I done. I\\nwas a bigot and I hated Catholics. (Stevenson, 1996, p. 65)\\nA central theme in the accounts of Protestant paramilitaries is a sense of\\n“us versus them,” Protestant versus Catholic. The account of another for-\\nmer Protestant paramilitary, William Smith, is consistent with this:\\nIn the seventies these communities were close-knit communities. . . . You\\nstill had the extended family situation, and word-of-mouth and rumor was far\\nmore valid than printed matter or television. But there’s no one thing, it was\\nthe whole situation at the that time. You had segregated schools, you had seg-\\nregated areas, you had a whole history dealt down to you by word-of-mouth\\nabout troubles, the IRA, and all that—that was all part of your dogma the\\nwhole way through your life. So whenever the bogeyman appeared in 1970,\\nthis was the bogeyman we were told about. The bogeyman appeared and the\\nrest became part of life—there was thousands and thousands, I mean, the vast\\nmajority of people were involved in it. So when your community felt threat-\\nened, you felt threatened. (Stevenson, 1996, p. 67)\\nBased on these accounts, it appears that persons who became Protestant\\nparamilitaries were motivated by an “us versus them” perspective. These\\naccounts suggest that the paramilitaries learned this perspective because\\nthey were encouraged, by family, friends, and the Protestant community\\naround them, to view things comparatively—Protestant versus Catholic.\\nThus, Catholics are people to be hated because “they’ll murder us.” \\nBeginning in the early 1970s, large numbers of people were recruited\\ninto Protestant paramilitary organizations like the Ulster Volunteer Force,\\nthe Ulster Defence Association, and the Ulster Freedom Fighters. It can be\\nexpected that members of these groups developed social identities that were\\nparticular to these groups—that is in addition to their Protestant social\\nSocial and Role Identities and Political Violence\\n143\\n\\n\\nidentity they probably also developed social/collective identities based on\\ntheir membership in a particular paramilitary group. As members of these\\ngroups they may also have developed role identities (see below). However,\\nthe key to understanding these groups and their members remains their\\nProtestant social identity. This social identity is common to most if not all\\nProtestant paramilitaries, and is probably the most important motivator of\\ntheir behavior.\\nThe Irish Community: Nationalists\\nAn Irish Nationalist is a person who wants all of Ireland to be ruled by Irish\\npeople (Irish people being defined as people living in Ireland, even if they\\nidentify themselves as British). Irish Nationalism predates the Reformation\\nand the Act of Union (effective January 1, 1801; Cronin, 1980). T\\noday, most\\nIrish Nationalists are Catholic. With respect to Northern Irish politics, the\\nterms “Catholic” and “Nationalist” are often used interchangeably, but they\\nare not synonymous. Catholics in Northern Ireland do not exclusively iden-\\ntify themselves as Irish. When Richard Rose (1971) surveyed the Northern\\nIrish public in 1968, he found that when asked which term “best describes\\nthe way you think of yourself” (British, Irish, Ulster, sometimes British-\\nsometimes Irish, Anglo-Irish), 76 percent of Northern Irish Catholics chose\\nthe “Irish” option. Only 15 percent of Catholics identified themselves as\\n“British,” and even fewer (5 percent) identified themselves as “Ulster.” T\\nen\\nyears later, in the midst of violence, when Edward Moxon-Browne included\\nthe question on his survey, the percentage had fallen; 69 percent of\\nCatholics chose “Irish” as their first option. Again, only 15 percent of\\nCatholics identified themselves as “British,” and few (6 percent in 1978)\\nidentified themselves as “Ulster.” Also in 1978, when asked “which of these\\nwould you least like to call yourself? British, Irish, Ulster, sometimes\\nBritish-sometimes Irish, Anglo-Irish, Other?”, less than 2 percent (1.7 per-\\ncent) of Catholics chose “Irish”—that is, very few Catholics rejected the\\nIrish national identity (see Moxon-Browne, 1983, pp. 6–9). In contrast to\\nthe Protestant community, the national identity of Catholics/Nationalists in\\nNorthern Ireland appears to be more stable (see also Moxon-Browne,\\n1983). \\nThe Irish Community: Irish Republicans\\nAt its most basic level, an Irish Republican is someone who “seeks to sepa-\\nrate Ireland from England by force” (Cronin, 1980, p. 23). At their core, all\\nRepublicans agree on two things: (1) the British government is the cause of\\npolitical strife in Ireland, and (2) a British withdrawal from Ireland can be\\ngained only through force of arms (see White & Fraser, 2000). These be-\\nliefs are evident in accounts from Republicans and may be interpreted in\\nboth social and role identity terms. \\n144\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nConsider the following from Respondent 1. He describes events that led\\nhim into the Irish Republican Army. Recognizing that he is Irish and that\\nthe British were killing “our people” was essential for his recruitment into\\nthe IRA. In the early 1970s he went to school in England:\\nRESPONDENT 1 [Hereafter, R]: I went to England and as I say everything\\nthen was at a distance, it wasn’t as immediate. . . . But then Bloody Sunday\\n[in which British troops shot dead 14 Nationalist civilian protesters, in 1972]\\nhappened and again this incredible sense of outrage. And then, just through-\\nout the year, watching the British, some of the British people—their attitudes\\nto us, the Irish. And it was just—fuck sake. The attitudes were just that—\\nyou’re “Paddy.” They automatically call you Paddy when you go over there.\\nI’d say, “My name’s Ruairí.” “Sure, you’re Paddy. You’re all called Paddy and\\nMick.” I said, “No, my name’s Ruairí.” I took it from my friends there because\\nit was usually a joke, there was no malevolence in it. But, there were people\\nwho used it as a derogatory term. And just throughout the year, watching\\nthings going at home. And following along Bloody Sunday I decided that\\nwhat I was doing, which I wasn’t happy with anyway, that I was wanting to go\\nhome more and more. And then when I left England to travel for a couple of\\nmonths, hitched around, and I came home. And as soon as I came home I got\\ninto Sinn Féin and it was a natural progression from there. Because my beliefs\\nwere fairly well formulated at the time. The British were killing our people, they\\nwere locking them up and they were nothing more than Stormont [the\\nNorthern Ireland parliament]. (emphasis added) \\nSimilar comments were offered by a Protestant Republican (see also Hynd-\\nman, 1996). He was born and raised in East Belfast. By his background he\\nwas an unlikely candidate for involvement with Irish Republicans. When he\\nmoved to England, he discovered that he was Irish.\\nINTERVIEWER [Hereafter, I]: You mentioned our people, you said they\\nwere killing our people or shooting our people. Do you consider yourself\\nIrish? \\nR: Oh yes, I am Irish you see. \\nI: But I mean, the typical—\\nR: (Laughs) I know it’s uh—I understand what you mean. Given my back-\\nground, right? For a long time I thought I was British. I didn’t really know\\nwhat I was, actually. When I was younger I went to England to work due to\\nthe high unemployment at home. And, then I was made very aware of my iden-\\ntity, and it was Irish. I was left in no doubt that I was a Paddy. And uh, you\\nknow the old stories about people going to try to find flats, you know, some-\\nwhere to live in London, where the signs are in the window, “No Irish,” “No\\nColoureds.” It was a good thing to experience because it did give me a per-\\nception, it did give me an education which I might not necessarily have gotten\\nif I had stayed in the North [of Ireland]. But certainly that made me very\\naware of my identity. Then moving from my Loyalist area, through other lo-\\ncations, finally to end up in a Republican area was a great experience. Living\\nthere which had always—had been a place which I had visited quite a lot, but\\nSocial and Role Identities and Political Violence\\n145\\n\\n\\nliving in it is a lot different from visiting. T\\no experience the continual daily\\nharassment, the constant harassment. It’s a great education as well. T\\no see the\\npoverty as well. [RESPONDENT 2, emphasis added]\\nThese two respondents acquired an Irish social identity by personally expe-\\nriencing discrimination and devaluation of their Irishness. The develop-\\nment of this social identity helped to make them potential recruits to Irish\\nRepublicanism. Their recruitment was also influenced by their acquiring a\\nRepublican role identity.5\\nRespondents 1 and 2 became Republicans when they were adults. Many\\nRepublicans, however, are born and raised in Republican families and \\nthey are influenced by Republicanism at an early age. One result of this in-\\nfluence is an Irish social identity. Respondent 3 was raised in a Republican\\nfamily: \\nR: the only solution to our problems, and it has been my opinion all down\\nthrough the years, is a united Ireland. It’s the only way forward for us. There\\nis no other way. The other way is a creeping, cringing, crawling spirit of the\\nIrish nation. We desire freedom, we seek freedom, and it’s not too much to\\ndemand. All we’re seeking is our own country. . . . I think it only makes\\ngood common sense that if you are the owner of a house, I always make this\\nanalogy, that if you’ve got a four-roomed house, and there’s one person occu-\\npying the house. And there’s quite a large family of you. And you make a de-\\nmocratic decision time after time for the person that’s occupying your house,\\nundemocratically, for to remove . . . to get out of the house. That democra-\\ntic decision is taken. But the man won’t move. What’s your alternative? Get\\nhim by the scruff of the neck and throw him out. It’s the only way. Our people\\nhas taken the democratic decision in this country here several times to ask Britain to\\nleave this country. So as we can apply our own national government. But she\\nhasn’t adhered to that democratic decision, and that’s why this struggle is still\\ngoing on. (Respondent 3, emphasis added)\\nRespondent 4 was also raised in a Republican family. He stated the following:\\nR: [When I joined, my goal for the Movement] was to see the complete free-\\ndom of this country. In other words, what the Republican goal always was, to\\nbreak the connection with England and to establish an Irish Republic. An Irish So-\\ncialist Republic. And that was my reason for joining the Movement and that is\\nwhy I got involved. (Respondent 4, emphasis added)\\nLike Protestant paramilitaries, Republicans—both from Republican fami-\\nlies and not—develop a social identity that places them in an “us versus them”\\nsituation. The “them,” however, is the British government, not the Protes-\\ntant community in Northern Ireland. \\nThere is evidence that Irish Republicans develop a role identity that\\ncomplements their Irish social identity, and that this role identity develops\\nthrough social interaction with family members and/or peers who are Re-\\npublicans. Consider the following from the three respondents raised in \\nRepublican families:\\n146\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nI: Why in general did you get involved in the Movement?\\nR: Well, Bob, I was involved with the Republican Movement at a very early\\nage because of my father being a member of the Republican Movement as\\nwell. From 1916 ‘til the Civil War my home was a place for anybody who\\ncame on the run, or to seek help. It was always given to them. So I was sort of\\nbrought up, you know what I mean, within Republican circles, my father be-\\ning a Republican. And uh, I suppose it rubbed off on me. I always had Repub-\\nlican ideals, you know? (Respondent 3)\\nR: Well, the reason why I decided to get involved in the Republican Move-\\nment was that I was born into the Republican tradition. Both on my father’s\\nand my mother’s side. And that is one of the reasons. And reading Irish his-\\ntory I realized and came to the conclusion that the Republican Movement\\nwere the true inheritance of the Republican faith. That is the reason why I \\nbecame [involved]. \\n. . .\\nR: [Reading the question] Did any particular person have a strong influence—\\neh, to get involved? Well, I wouldn’t just say to get involved, but, there was\\none woman, she was my grandmother. Incidentally my four grandparents\\nwere teachers. But this particular—my mother’s mother, she was an outstand-\\ning republican sort of thing, you know? When she died in 1969—She never\\nwavered. . . . She was strong in her Republican [beliefs] to the day she died.\\nAnd when she—her house, as a matter of fact, was the headquarters of the\\nIrish Republican Army [in her area]. . . . she was a very highly respected\\nwoman sort of thing there. . . . And she had a big influence on my life there\\nbecause as you may know you’re—well, in this country anyway your grand-\\nchildren go on holidays with their grandparents. . . . She never wavered, as\\nI said. And she had a big influence. And even up to this day, if any questions\\ncome up [I would refer to her Republicanism]. (Respondent 4)\\nR: Republicanism in our house was a—sort of what would be termed now as\\npure Republicanism, in that my father wouldn’t of been very sympathetic to-\\nward a socialist line of thinking. It would of been more of a Republican prin-\\ncipled directly. And, uh, seen the great danger in socialist tendencies within it.\\nAnd, to some extent that would of influenced my direction for a number of\\nyears. And even yet to some extent that you would fear a socialist—too big a\\nsocialist involvement within it. And uh, he was very much just a straight line\\nRepublican. \\nI: Is that why you would have gone with Provisional rather than Official\\n[when the Republican Movement split in 1969/70]?\\nR: That’s right. That would of been a big influence actually. And uh, not only\\nfrom hearing it in the past but actually hearing him talking about it at that\\nparticular time, eh, that the split come. My direction like would of been influ-\\nenced a big part by his sort of talking to me. (Respondent 5)\\nIt appears that these children of Republicans were raised such that they ac-\\nquired Republican principles and learned the role of being a Republican—\\nfor example, to help other Republicans “on the run” and to never “waver”\\nin your commitment. This acquisition and learning occurred through inter-\\nSocial and Role Identities and Political Violence\\n147\\n\\n\\naction with other Republicans. Accounts from Republicans not born and\\nraised in Republican families suggest that they also acquire a role identity\\nthrough interaction with both their family members (even though they are\\nnot Republican) and their peers who are Republicans. The following is\\nfrom Respondent 1. He is not from a Republican family. He joined the IRA\\nin the early 1970s. He was strongly influenced by his personal experiences\\nwith peers who were involved in the Irish Republican Movement. His in-\\nvolvement stems from a family background that stressed being a responsible\\nperson, and his social interaction with people who he believed were being\\nresponsible—that is, persons who were willing to risk their lives in a just\\nstruggle against a much stronger opponent (and his development of an Irish\\nsocial identity, as noted earlier).6\\nR: The main reason that I got involved with the Republican Movement is that\\nI was brought up to believe that you should treat people as equals and that no-\\nbody had the right to walk over you, nor did you, more importantly, did you\\nhave the right to walk over anybody else. And the Republican Movement, it\\nmay have historical background, yes, but essentially it got off the ground be-\\ncause it came out of the community and it was a community that accepted it\\nbecause it was its own people. And therefore I suppose you could say it was\\nmy contemporaries that, if anybody influenced, would have influenced me to\\njoin. But it was just a natural thing because it was the time of ferment, of po-\\nlitical ferment that was in it. I felt that I couldn’t close my eyes to it and that I\\nhad a responsibility to do something. I’m not sure, mind you, that would be\\neverybody’s reply—it sounds terribly altruistic [laughs]. But the point is that\\nthat’s just the way I felt about it. I just felt that I had had a pretty good life\\ncompared to most of the lads that would have been in the movement in Derry.\\nI come from a middle class background and I never really had to worry about\\nanything and I just felt there was a responsibility for me to do something. But\\nit wasn’t as clear cut as that. \\n. . .\\nR: I didn’t join because my friends were in the movement. That’s not the rea-\\nson I joined. I was motivated to join the Republican Movement because I\\nknew that the people who were in it were sincere. I mean, a 19-year-old kid or\\nan 18-year-old kid who makes a conscious decision to go out and fight a vastly\\nsuperior army on the streets, to take on a government and all its forms, and to\\nchange society—He may not think in terms of vast political thought about\\nhow society’s going to be changed by him. But if he makes that conscious de-\\ncision instead of happily waiting until he get his welfare check, and then goes\\noff and plays pool or drinks beer and goes to dances—I mean they’re de-\\nnounced as being mindless thugs but that is just not true. It could never for\\nme be a description of them. (Respondent 1)\\nAs it was for Protestant paramilitary organizations, in the early 1970s\\nIrish Republican organizations received large numbers of recruits. As with\\nProtestant paramilitaries, the new Republican paramilitary recruits were in-\\nfluenced by a social identity. Unlike Protestant paramilitaries, the Republi-\\n148\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\ncan recruits did not interpret their activity as Catholic versus Protestant,\\nbut rather as Irish versus British. In addition, and also unlike Protestant\\nparamilitaries, the Republican paramilitary recruits joined a movement with\\na lengthy history. Along with this history came established roles for mem-\\nbers of the Republican Movement—“she never wavered.” Out of interac-\\ntion with persons who were already Republican (be they family members \\nor peers) the Republicans acquired a role identity that complemented their\\nsocial identity.\\nThe British Perspective:“Them vs.Them”\\nThe British interpretation of the violence is that it is an Irish problem in\\nwhich members of two Irish communities, Protestant and Catholic, cannot\\nget along. This is implied in the accounts of Respondent 1 and Respondent\\n2, one Catholic and the other Protestant, but both “Paddy” in Britain. In ef-\\nfect, the British deny the social identity of Protestant paramilitaries, who\\nview themselves as British, and they deny the political aspirations of Irish\\nRepublican paramilitaries, who do not want to be citizens of the United\\nKingdom. \\nThis is not a new phenomenon. In 1921, on a visit to Belfast, King\\nGeorge V said:\\nI speak from a full heart when I pray that my coming to Ireland today may\\nprove to be the first step towards an end of strife amongst her peoples, what-\\never their race or creed. In that hope I appeal to all Irishmen to pause, to\\nstretch out the hand of forgiveness and conciliation, to forgive and forget and\\nto join in making for the land which they love a new era of peace, content-\\nment and goodwill. (see Lyons 1973, p. 426)\\nFifty years later, Reginald Maudling was the British Home Secretary re-\\nsponsible for Northern Ireland affairs at Westminster (from 1970-1972). \\nIn his memoirs, he described how the “Irish” problem perplexed him, an\\nEnglishman.\\nIt is very hard for an Englishman to understand the feelings of those who live\\nin Northern Ireland. The history of their struggles is a long one, and they\\ntend to cherish every moment of hatred in it. The deep divide between\\nCatholic and Protestant, which, incidentally, is more racial than religious, had\\nbeen handed down faithfully from generation to generation. (Maudling 1978,\\np. 180)\\nTraditionally, Ireland is not that important for British elites. This makes\\nthe violence that much more an “Irish problem.” After helping execute\\nCharles I, Oliver Cromwell spent nine and one-half months in Ireland,\\npacified the country, and then returned to the real business at hand—\\nrunning England (Fraser, 1973/1997). In the early part of this century,\\nthere were a series of debates and proposals associated with the Home Rule\\n[for Ireland] Bill in 1914. William Manchester, in his biography of Winston\\nSocial and Role Identities and Political Violence\\n149\\n\\n\\nChurchill, describes the Irish situation as being, “foremost in everybody’s\\nmind, so certain to burst into flames at any moment” (1983, p. 463). But, as\\na meeting concluded Sir Edward Grey read out a document from the For-\\neign Office. It was an ultimatum from Austria to Serbia, part of the begin-\\nning of what became World War I. The war developed, and the Irish issue\\nwas “deferred.” Churchill later recalled, \\nThe parishes of Fermanagh and Tyrone faded back into the mists and squalls\\nof Ireland, and a strange light began immediately, but by perceptible grada-\\ntions, to fall and grow upon the map of Europe. (Manchester, 1983, p. 463)\\nThe “Irish question” also has not had an impact on the voting British\\npublic. Northern Ireland constitutes less than 3 percent of the population\\nof the United Kingdom. Since 1983, the province has sent only 17 (of 635)\\nMPs to Westminster. In response to Gallup polls on important issues in the\\nUnited Kingdom, Northern Ireland traditionally has “fluctuated between\\nnil mention and coming at the bottom of a list of ten or a dozen issues” (see\\nO’Malley, 1990, p. 226). \\nBecause Northern Ireland has not been a high priority for the British\\npublic, British soldiers, recruited from that public, appear to be relatively\\nunaware of the politics of the situation (at least in the early years of the re-\\ncent conflict). Consider the following from two soldiers sent to Northern\\nIreland in 1969. \\nBRITISH SOLDIER 1: The boys knew where they were supposed to go in\\nterms of sealing off streets, so it was then a question of getting men down to\\nthe corners and road junctions and giving them cover in case there was shoot-\\ning. We hadn’t a clue what was going on really, but the feeling I got within half\\nan hour was one of relief, on both sides. The Irish are terribly curious and they\\ncame out of all sorts of corners, and I think both Protestants and Catholics\\nwere relieved that there might be some stability. . . . (emphasis added;\\nArthur, 1987, p. 7)\\nBRITISH SOLDIER 2: For the first week I slept underneath an awning by a\\nbaker’s shop. I’d chosen the baker’s shop because I knew we’d get fresh bread\\nand buns each day. This was the period of complete daze, immediately after\\nthe initial rioting. No one knew what was going on, so we had to start talking to the\\npeople to find out what might happen. We felt we were terribly welcome, particu-\\nlarly by the Protestants, but they kept on saying: ‘Why have you got your ri-\\nfles pointing at us, and your backs to the Catholics, when they’re the ones\\nwho started it?’ So immediately we came up against the sectarian problem:\\nwhich way do you face? Who do you defend? Which side do you protect? (emphasis\\nadded; Arthur, 1987, pp. 9–10)\\nThe perspective on Ireland of British elites, citizens, and soldiers may\\nbest be described as viewing the conflict as “them versus them.” The degree\\nto which Britishers identify with Britain and adopt potential role identi-\\nties—“As a British soldier my job here is to keep the peace”—may be less\\nimportant than the fact that they view what happens in Northern Ireland as\\n150\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nexternal to themselves, or at least not of their own making, that it is a con-\\nflict caused by “others” and it is not a reflection of British policies or \\nactions. \\nBy denying the identity of those Protestants in Ulster who identify\\nthemselves as British, British elites contribute to the fears of these people.\\nBy refusing to recognize the political claims of the Republicans, and by ac-\\ntively denying these claims, British elites leave Republicans with violence as\\na means of expressing their political desires. Perhaps most important, view-\\ning the conflict as external, or involving “them versus them,” contributes to\\npolicy mistakes on the part of the British government, as described below. \\nImplications\\nIn Ireland there are three violent political actors: Protestant paramilitaries,\\nIrish Republican paramilitaries, and state agents. Each perceives the vio-\\nlence in Northern Ireland from a different perspective. These perceptions\\nhave real consequences. This is demonstrated in the British policies of\\ncriminalization and Ulsterization, which were implemented in the mid-\\n1970s. Their impact is still felt today.\\nIn 1972, IRA prisoners led by Billy McKee, an IRA veteran of the 1930s\\nand 1940s, went on a hunger strike in order to gain political status. In the\\nmidst of negotiations that led to a truce between the British government\\nand the IRA in late June, political status was granted to both Republican\\nand Protestant paramilitary prisoners. Criminalization revoked this status.\\nAfter March 1, 1976, persons convicted of paramilitary crimes were treated\\nas ordinary criminals. The British government denied the political nature\\nof the IRA’s activity and equated IRA violence with the sectarian violence of\\nProtestant paramilitaries. Ulsterization dates from 1977 and is designed to\\nmake the Royal Ulster Constabulary the primary policing agent in North-\\nern Ireland (e.g., Flackes & Elliott, 1989, p. 392 ; see also Bell, 1993, \\np. 517). As a result, the violence became that much more an Irish problem\\n(of “them vs. them”).\\nThe implications of criminalization and Ulsterization were profound.\\nIrish Republican prisoners rejected criminalization. They refused to wear\\nprison clothes and wrapped themselves only in blankets. The protest esca-\\nlated, ultimately resulting in the hunger-strike of 1981. In an attempt to re-\\ngain political status, 10 Irish Republican prisoners fasted to death (see Bell,\\n1993). The behavior of the Republican prisoners on the strike may be un-\\nderstood in terms of a role identity and role behavior. Hunger-striking has a\\nlengthy history among Republicans; the hunger-strike weapon is part of the\\nRepublican repertoire. The first Republican to die on a hunger strike in this\\ncentury was Thomas Ashe, in 1917. There were deaths by Republicans on\\nhunger strike in the 1910s, 1920s, 1940s, and the 1970s. People like Ashe, it\\ncan be argued, were role models. One dimension of the role identity of be-\\nSocial and Role Identities and Political Violence\\n151\\n\\n\\ning an Irish Republican is to confront the state while in prison with the\\nweapon of the hunger strike. Billy McKee, the senior Republican hunger\\nstriker in Belfast in 1972, was active in the IRA in the 1940s when then \\nsenior Republicans were on hunger strike. Some of those on the hunger\\nstrike in 1981 were under the command of people like McKee (if not Mc-\\nKee himself) when they were new recruits to the Republican Movement in\\nthe early 1970s. \\nThe hunger strike in 1981 led to a mass mobilization of the general \\nNationalist community in Ireland. One of the strikers, Bobby Sands, was\\nelected as an abstentionist MP to Westminster. The chief beneficiary of the\\nhunger strike was the Republican movement. Both Sinn Féin and the IRA\\nattracted large numbers of new recruits (see Bell, 1993; White, 1993). For\\nthe British government, who held firm against the hunger strike, it was a\\ndisaster. Ulsterization probably saved the lives of British soldiers, but it, like\\ncriminalization, did not stop Irish Republicans, who claim that they ‘target\\nthe uniform’ rather than the religion of members of the security forces (see\\nFlackes & Elliott, 1989, p. 414, table 4). While killing a British soldier may\\nhave more of an impact, Republicans were quite willing to kill members of\\nthe RUC (see White, 1997a, 1997b; Bruce, 1997)). \\nCriminalization and Ulsterization had very different results for Protes-\\ntant paramilitaries. Criminalization did not lead to deaths from hunger\\nstrikes by Protestant paramilitary prisoners. Although there have been\\nProtestant paramilitary prisoners on hunger strike, none have maintained a\\nfast to the death. There are a number of reasons for this and two are impor-\\ntant here. First, it must be acknowledged that the conditions for Protestant\\nparamilitary prisoners have never been as severe as they have been for Irish\\nRepublican paramilitaries. Pro-British Protestant paramilitary prisoners are\\nless of a threat to the state than are anti-state Republican prisoners (see\\nWhite, 1999). Second, and very important for present purposes, there is no\\nhistory of hunger striking to the death by Protestant paramilitaries. The\\nhistoric and contemporary “role models” (martyrs) that Irish Republicans\\nhave—Thomas Ashe in 1917, Sean McCaughey in 1946, Michael Gaughan\\nin 1974, and Bobby Sands in 1981, for example—are not available to\\nProtestant paramilitaries. The role identity of a Protestant paramilitary\\nprisoner confronting the British government by dying on hunger strike\\ndoes not exist.\\nIndeed, at the beginning of the violence a Protestant paramilitary role\\nidentity could not exist because such people, and such organizations, had\\nnot yet formed. When the Ulster Volunteer Force and the Ulster Freedom\\nFighters were created in the mid-1960s and early 1970s, a previous genera-\\ntion of Protestant paramilitaries was not there to serve as paramilitary lead-\\ners (role models). The original UVF was largely recruited into the security\\nforces of Northern Ireland in the 1920s. That organization ceased to exist,\\nand its former members became police agents of Northern Ireland (see Far-\\nrell, 1980, pp. 35, 44–45). While some of the parents of the Protestant para-\\n152\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nmilitary recruits of the 1960s and 1970s were members of the RUC and the\\nUlster Defence Regiment (UDR; e.g., T\\naylor, 1999, p. 82), this is qualita-\\ntively different from being raised in a family with a tradition of involvement\\nin illegal paramilitary activity. A leader of a vigilante group described their\\nactivity at the beginning of the current violence:\\nWe never planned to go on the kill. There was no time that we sat down and said\\n‘That’s it. Stiff a T\\naig.’ Mind, we planned doing something to the Ardoyne af-\\nter the three Scots boys [three British soldiers] were killed [in 1971] but for\\nsome reason it never came off. No, it was ground up. One or two volunteers\\njust started doing it. (taken from Bruce, 1992, p. 54; emphasis added)\\nContrast this with the description of Billy McKee’s activities in the early\\n1970s. The interviewer, the journalist Peter T\\naylor, asked a Republican who\\nwas recruited back into the IRA by McKee to comment on his activities in\\nthe early 1970s:\\nWhat did he [McKee] tell you about the policy towards British soldiers?\\nHe didn’t indicate that there was going to be an immediate offensive\\nagainst the British army. He said, “These things take time. People have to be\\ntrained. People have to be motivated. People have to be equipped. All this won’t \\nhappen overnight.” But the intention was there and it sounded good to me.\\n(T\\naylor, 1998, p. 74; emphasis added)\\nFor Protestant paramilitaries, criminalization reinforced their interpre-\\ntation that Republicans are out to destroy their community. While Republi-\\ncans state that they only target uniforms, not the ethnic/religious belief of\\nthe occupant, Protestants (paramilitaries and in general) view attacks on the\\nRUC as sectarian because the force is 90 to 95 percent Protestant. In re-\\nsponse to the killing of members of the RUC, Protestant paramilitaries\\nkilled Catholic civilians (see White, 1997a, 1997b; Bruce, 1997).\\nThe accounts from Protestant paramilitaries suggested that they do not\\ninterpret their behavior in terms of a role identity. However, these accounts\\nare from secondary sources and social and role identities were not the focus\\nof those who collected the accounts. Also, the accounts are from activists\\nwho were not born and raised in Protestant paramilitary families. The vio-\\nlence in Northern Ireland has spanned a generation. Many of the Protes-\\ntant (and Republican) paramilitaries of the 1970s and early 1980s now have\\nadult children. It is very possible that with the continuity of the Protestant\\nstruggle there has developed among Protestant paramilitaries a role identity\\nthat has been passed to another generation in the same manner as Irish Re-\\npublicans pass on their role identity. T\\noday, the leader of the Ulster Demo-\\ncratic Party, the political wing of the Ulster Defence Association, is Gary\\nMcMichael. Gary McMichael is the son of John McMichael, a leading\\nProtestant paramilitary who was killed by the IRA in 1987 (Cusack & Mc-\\nDonald 1997, p. 267; Bruce, 1992, pp. 232–233). If a Protestant paramili-\\ntary role identity has developed, this identity should be consistent with the\\nSocial and Role Identities and Political Violence\\n153\\n\\n\\nProtestant social identity, and the basic social identity differences among\\nProtestant paramilitaries, Irish Republican paramilitaries, and the British\\nshould remain. \\nConclusion\\nMuch of the violence in the world today appears to be caused by heightened\\nsocial identities, by groups that define themselves as “us versus them.” On\\nthe surface, violence in Northern Ireland fits this pattern. The above case\\nstudy reveals, however, that there are three significant violent actors in\\nNorthern Ireland, not two. These actors are Protestant paramilitaries, Irish\\nRepublican paramilitaries, and the state’s security forces. Each of the actors\\ninterprets the violence in different social identity terms. Protestant para-\\nmilitaries see themselves as British and they define themselves as Protes-\\ntants in contrast to Irish Catholics. Further, they perceive that Irish Repub-\\nlicans act in a manner consistent with an Irish Catholic identity. Irish\\nRepublicans, in contrast, act in a manner consistent with an Irish social\\nidentity. Their conflict is not with Irish Protestants but with British parlia-\\nments that lay claim to Irish territory. They view themselves as being in a\\nwar of national liberation against the British government. Protestant para-\\nmilitaries interpret Republican activity differently. For Protestant paramili-\\ntaries, targeting the British in Northern Ireland means targeting Protes-\\ntants in Northern Ireland—they view themselves as the British in Northern\\nIreland. In the midst of these different interpretations, British people have\\ntraditionally viewed the violence in Ireland from an outsider’s perspective.\\nThe violence is between Protestant Irish and Catholic Irish, and the British,\\nas the lawful government, are caught in the middle. This adds to the com-\\nplexity of the situation, especially because British government actions are\\nsubject to different interpretations by Protestant paramilitaries and Irish\\nRepublican paramilitaries.\\nIn addition to their Irish social identity, the data suggest that there is an\\nIrish Republican role identity and that this influences Republicans’ behav-\\nior. This role identity develops through social interaction with other Re-\\npublicans. Although the data did not necessarily indicate the development\\nof a Protestant paramilitary role identity, there is no reason to believe that\\nsuch an identity has not developed over time. Similarly, British elites and\\nBritish soldiers probably have role identities. As with Republicans, these\\nrole identities probably develop through social interaction. \\nAt the moment, there is a peace process underway in Northern Ireland.\\nIt has been a long time coming, with origins that date from events in the\\nmid- to late-1980s. As of yet, peace has not been fully confirmed. Involved\\nin this peace process are the three general actors that are the subject of this\\npaper—Protestant paramilitaries, Irish Republican paramilitaries, and the\\nBritish government (also involved are several other actors, including non-\\n154\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nviolent political activists and the Irish Government in Dublin). The differ-\\nent perceptions that the actors have of themselves and their counterparts\\ncontribute to the difficulty of achieving a lasting peace in Ireland.\\nThe Northern Ireland conflict is hundreds of years in the making. The\\nconflict is not unique. Many of the conflicts that attract our attention today\\nare decades, if not centuries, old. Serb versus Croat and Hutu versus Tutsi\\nare centuries old conflicts. There are more than two violent actors in\\nNorthern Ireland. Other violent settings also involve more than two actors,\\nif not currently then historically. The British were key actors in the carving\\nup of Palestine. T\\noday, the United States plays a key role in that conflict, as\\nArab states and Palestinian paramilitaries are aware. Similarly, for the vic-\\ntims of the bombing of Kosovo (whether or not it was justified), the United\\nStates and its allies were significant actors. An in-depth investigation of\\nmany conflicts that appear to involve two actors engaged in “them versus\\nus” behavior may show them to be as complex as the Northern Ireland \\nconflict.\\nAcknowledgment\\nThis research was supported by grants from the Harry Frank Guggenheim Founda-\\ntion, the National Science Foundation (grant SES-8318161), and the IUPUI Office\\nof Faculty Development. I thank Stephen Heathorn, Shel Stryker, Peggy Thoits,\\nEric Wright, the participants in the Symposium, and the reviewers for their com-\\nments.\\nNotes\\n1. My use of the term “political violence” follows Tilly’s conception of the term\\n“any observable interaction in the course of which persons or objects are seized or\\nphysically damaged in spite of resistance” (Tilly, 1978, p. 176). Political violence in\\nIreland is not confined to Northern Ireland. The Irish Republican Army recruits\\nand carries out operations on an all-Ireland basis. Protestant paramilitaries have\\nkilled people in the Irish Republic, and the IRA has been active throughout Ireland,\\nEngland, and on the European continent. Included in the Sutton (1994) data in\\ntable 6.1 are persons who were killed outside of Northern Ireland, in events related\\nto the conflict there. By far the vast majority of persons killed because of the North-\\nern Ireland conflict were killed in Northern Ireland (see White, 1997a). \\n2. These actors are “broadly” defined because there are in fact several different\\nRepublican, Protestant, and security force organizations active in Northern Ireland.\\nThese include the Irish Republican Army, the Irish National Liberation Army, the\\nUlster Volunteer Force, the Ulster Freedom Fighters, the British army, and the\\nRoyal Ulster Constabulary. \\n3. This is too simple a description of a complex situation. While Irish people are\\nin general “thems” to Britishers (as described below), there are probably degrees of\\n“themness.” In particular, one group of “them”—Ulster Protestants—identifies with\\nBritain, is devoted to the British crown, proudly flies the Union Jack, etc., while an-\\nSocial and Role Identities and Political Violence\\n155\\n\\n\\nother group of “them”—Irish Nationalists—identifies with Ireland, rejects the\\nBritish crown, flies the Irish flag, etc. While Irish people may be “them,” some of\\n“them” are likely to be more tolerable than others. This helps to explain differences\\nin state repression, whereby Republican paramilitaries experience more repression\\nat the hands of the British than do Protestant paramilitaries (e.g., White, 1999). It\\nalso helps explain the much larger number of Catholic civilians versus the relatively\\nsmall number of Protestant civilians killed by the security forces. Catholic civilians\\ncome from the dis-loyal “them” community. They are subject to much more state\\nrepression (see White & White, 1995).\\n4. “T\\naig” is a slang term for a Catholic person. It is typically used by Protestants\\nin a derogatory fashion. It is a corruption of the Irish first-name, “T\\nadhg.”\\n5. In order to demonstrate the development of an Irish identity I present infor-\\nmation from two respondents who lived in England prior to their joining the IRA.\\nThe life stories of these two respondents are such that their understanding of their\\nIrishness was practically forced on them, which makes for a good presentation. The\\nother Republican respondents (Respondents 3, 4, and 5), developed an Irish social\\nidentity without first living in Britain.\\n6. Determining when a person acquires a role identity is difficult. Some people\\nmay acquire role identities only after joining an organization and formally moving\\ninto the role. Others may acquire role identities from interaction with persons in\\nroles prior to their actually joining a given organization. In a sense, they may find\\nthemselves in the role, and then join the organization. Imagine the potential recruit\\nwho goes out on an operation or two with his friends and quickly finds herself a \\nregular member of an IRA unit. Similarly, consider the child of a Republican who, at\\nthe age of 18, is interned under the assumption that he is a member of the IRA, but\\nis not. Interred without trial with other Republican prisoners, having been brought\\nup in Republican surroundings, this individual probably knows the role—or learns it\\nvery quickly—of being an IRA prisoner, adapts quickly to prison, and joins the IRA.\\nReferences\\nArthur, M. (1987). Northern Ireland: Soldiers talking. London: Sidgwick & Jackson.\\nBell, J. B. (1979). The secret army: The IRA 1916–. Dublin: Irish Academy Press.\\nBell, J. B. (1987). The gun in politics: An analysis of Irish political conflict, 1916–86. New\\nBrunswick, NJ: Transaction.\\nBell, J. B. (1993). The Irish troubles: A generation of violence. New York: St. Martin’s.\\nBruce, S. (1992). The red hand: Protestant paramilitaries in Northern Ireland. Oxford:\\nOxford University Press.\\nBruce, S. (1994). The edge of the Union: The Ulster Loyalist political vision. Oxford: \\nOxford University Press.\\nBruce, S. (1997). Victim selection in ethnic conflict: Motives and attitudes in Irish\\nRepublicanism. Terrorism and Political Violence, 9, 56–71.\\nCairnes, E. (1982). Intergroup conflict in Northern Ireland. In H. T\\najfel (Ed.), Social\\nidentity and intergroup relations (pp. 277–297) Cambridge: Cambridge University\\nPress.\\nCronin, S. (1980). Irish nationalism: Its roots and ideology. Dublin: Irish Academy\\nPress.\\nCusack, J. & McDonald, H. (1997). UVF. Dublin: Poolbeg Press.\\n156\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nDrake, C. J. M. (1998). The role of ideology in terrorists’ target selection. Terrorism\\nand Political Violence, 10 (Summer), 53–85.\\nFarrell, M. (1980). Northern Ireland: The orange state. London: Pluto Press.\\nFlackes, W. D., & Elliott, S. (1989). Northern Ireland: A political directory. Belfast:\\nBlackstaff Press.\\nFraser, A. (1973/1997). Cromwell: Our chief of men. London: Weidenfeld and Nicolson.\\nHyndman, M. (1996). Further afield: Journeys from a Protestant past. Belfast: Beyond\\nthe Pale Publications.\\nKlandermans, B., & de Weerd, M. (2000). Group identification and political\\nprotest. In S. Stryker, T. J. Owens, & R. W. White, Self, identity, and social move-\\nments. (pp. 68–90) Minneapolis: University of Minnesota Press.\\nLyons, F. S. L. (1973). Ireland since the famine. London: Fontana.\\nManchester, W. (1983). The last lion: Winston Churchill, 1874-1932, Visions of glory.\\nNew York: Little, Brown and Co.\\nMaudling, R. (1978). Memoirs. London: Sidgwick & Jackson.\\nMcCall, G., & J. L. Simmons. (1978). Identities and interactions. New York: Free\\nPress.\\nMcGarry, J., & O’Leary, B. (1995). Explaining Northern Ireland. Oxford: Blackwell\\nPublishers.\\nMcKittrick, D., Kelters, S. Feeney, B., & C. Thorton. (1999). Lost lives: The stories \\nof the men, women, and children who died as a result of the Northern Ireland troubles.\\nEdinburgh: Mainstream Publishing.\\nMoxon-Browne, E. (1983). Nation, class and creed in Northern Ireland. Aldershot:\\nGower Publishing.\\nNelson, S. (1984). Ulster’s uncertain defenders: Protestant political, paramilitary, and\\ncommunity groups and the Northern Ireland conflict. Belfast: Appletree Press; \\nSyracuse, NY: Syracuse University Press.\\nO’Leary, B., & McGarry, J. (1993). The politics of antagonism: Understanding Northern\\nIreland. London: Athlone Press.\\nO’Malley, P. (1990). The uncivil wars: Ireland today. Boston: Beacon.\\nRose, R. (1971). Governing without consensus: An Irish perspective. Boston: Beacon\\nPress.\\nStevenson, J. (1996). “We wrecked the place”: Contemplating an end to the Northern Irish\\ntroubles. London: The Free Press.\\nStryker, S. (1968). Identity salience and role performance: The relevance of \\nsymbolic interaction theory for family research. Journal of Marriage and the\\nFamily, 30, 558–564.\\nStryker, S. (1980). Symbolic interactionism: A social structural version. Menlo Park, CA:\\nBenjamin/Cummings.\\nStryker, S. (2000). Identity competition: Key to differential social movement partic-\\nipation? In S. Stryker, T. J. Owens, & R. W. White, Self, identity, and social move-\\nments (pp. 21–40). Minneapolis: University of Minnesota Press.\\nSutton, M. (1994). An index of deaths from the conflict in Northern Ireland, 1969–1993.\\nBelfast: Beyond the Pale Publications.\\nT\\najfel, H. (1981). Human groups and social categories: Studies in social psychology. Cam-\\nbridge: Cambridge University Press.\\nT\\naylor, P. (1999). Loyalists. London: Bloomsbury.\\nThatcher, M. (1993). The Downing Street years, 1979–1990. New York: Harper-\\nCollins\\nSocial and Role Identities and Political Violence\\n157\\n\\n\\nThoits, P. A., and Virshup, L. K. (1997). Me’s and we’s: Forms and functions of \\nsocial identities. In R. Ashmore & L. Jussim (Eds.), Self and identity: Fundamental\\nissues. New York: Oxford University Press.\\nTilly, C. (1978). From mobilization to revolution. Reading, MA: Addison-Wesley.\\nTurner, J. C., Oaks, P. J. Haslam, S. A. & McGarty. C. (1994). Self and collective:\\nCognition and social context. Personality and Social Psychology, Bulletin 20,\\n454–463.\\nTurner, J. C., with Hogg, M. A., Oakes, P. J., Reicher, S. D., & Blackwell, M. S.\\n(1987). Rediscovering the social group: A self-categorization theory. Oxford: Basil\\nBlackwell.\\nWhite, R. W. (1993). Provisional Irish Republicans: An oral and interpretive history.\\nWestport, CT: Greenwood Press\\nWhite, R. W. (1997a). The Irish Republican Army: An assessment of sectarianism.\\nTerrorism and Political Violence, 9 (Spring), 20–55.\\nWhite. R. W. (1997b). The Irish Republican Army and sectarianism: Moving \\nbeyond the anecdote. Terrorism and Political Violence, 9, 120–131.\\nWhite, R. W. (1999). Comparing state repression of pro-state vigilantes and anti-\\nstate insurgents: Northern Ireland, 1972-75. Mobilization, 4, 189–202.\\nWhite, R. W., & Fraser, M. R. (2000). Identities and long term movement \\nactivism: Republican Sinn Féin. In S. Stryker, T. J. Owens, & R. W. White \\n(Ed.), Self, identity, and social movements. (pp. 324–346). Minneapolis: University\\nof Minnesota Press.\\nWhite, R. W. & White, T. F. (1995). Repression and the liberal state: Northern \\nIreland, 1969-72. Journal of Conflict Resolution, 39, 330–352.\\nWhyte, J. (1990). Interpreting Northern Ireland. Oxford: Clarendon Press.\\n158\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nErvin Staub\\n7\\nIndividual and Group Identities in\\nGenocide and Mass Killing\\nMass killing and genocide, in which a group turns against another with sys-\\ntematic and widespread violence, has been a tragic aspect of life in the 20th\\ncentury. How does such violence, especially its most extreme form, geno-\\ncide, come about? What is the motivation for such violence and how does it\\nevolve? What are the social conditions, characteristics of cultures, and psy-\\nchological processes of individuals that make it more likely? Most impor-\\ntant for the present volume, what role do individual and group identities\\nplay in genocide and mass killing?\\nThe U.N. Genocide Convention has defined genocide as “acts commit-\\nted with intent to destroy, in whole or in part, a national, ethnic, racial or\\nreligious group.” I see the words “in part” in the definition above as prob-\\nlematic. When some people in a group are killed, for example, to intimidate\\nthe whole group, or when members of the population are killed in the belief\\nthat this will eliminate support for guerrillas, as in Guatemala, this may best\\nbe considered mass killing, not genocide. However, genocide and mass\\nkilling have similar origins and prevention requires similar actions. Because\\nsome nations objected to the inclusion of political groups, they have not\\nbeen included in the Genocide Convention. Some authors have called vio-\\nlence that aims to eliminate a political group, “politicide” (Harff & Gurr,\\n1990) while others, including myself, prefer the term “genocide”. Influ-\\nences leading to genocide against political and other kinds of groups and\\nthe preventive efforts required seem to be the same (Staub, 1989b).\\nI define “mass killing” as killing many members of a group without the\\nintention to eliminate the group (Staub, 1989b). Such killings can be part of\\nan evolution toward genocide. It is important for both understanding and\\n159\\n\\n\\nprediction, and therefore prevention, that similar influences lead to mass\\nkilling and genocide. While the likelihood of group or collective violence is\\npredictable, its exact nature, whether it will be mass killing or genocide,\\nmay not be.\\nIn this chapter I first outline a conception of the origins of genocide and\\nother collective violence that I have developed over the last 20 years and\\nhave applied to the analysis of a number of specific instances of genocides\\nand mass killing (see Staub, 1989b, 1996a, 1999b). I next examine elements\\nof this conception in detail as I extend it by exploring the role of identity in\\nthe generation of such violence. I will look at different types of and aspects\\nof identities: individual identity and varied forms of it; social identity, the\\nextent to which a person’s identity is based on membership in the group;\\ngroup self-concept, or how the members see and define their group. I will\\nalso discuss the psychological and social processes through which people\\nstrengthen their identities in difficult times, often in destructive ways.\\nOne of the important reasons for understanding the origins of genocide\\nand mass killing is to find ways to prevent them. Thus, in the last section \\nof the chapter I will consider prevention and the role of identities in \\nprevention.\\nThe Origins of Genocide and Mass Killing\\nOver the years I have developed a model of the origins of genocide and\\nmass killing (Staub 1989b, 1996b, 2000). This is depicted schematically in\\nfigure 7.1. Genocide or mass killing usually begins with some form of “in-\\nstigation.” Instigation may have little to do with the group that ends up as\\nthe victim. Difficult life conditions in a society are such instigators: severe\\neconomic problems, such as inflation, depression, and unemployment; po-\\nlitical disorganization or political chaos; and rapid social changes.\\nDifficult life conditions usually frustrate fundamental human needs—\\nwhat I have called “basic needs.” All of the following psychological needs\\nmay be frustrated by difficult life conditions: the need for security, for a\\npositive identity, for effectiveness and control, for connection to other peo-\\nple, and for a meaningful comprehension of reality and of one’s own place\\nin the scheme of things. These needs are viewed here as universal, even\\nthough culture may shape their strength, manifestation, and available av-\\nenues for fulfillment. Especially when these needs are reasonably fulfilled,\\nanother important need also enters—the higher need for spirituality or\\ntranscendence, the need to go beyond a focus on one’s own self, needs, and\\nconcerns (Staub, 1989b, 1996a, 1998a, 1999a, 1999b, 1999c).\\nConflict between groups is another instigator. One type of conflict has to\\ndo with vital needs, such as territory needed for living space. Conflicts can\\nbe intractable, often for psychological reasons, such as a group’s identity be-\\n160\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\ning tied to living in a certain territory. Another type of conflict occurs when\\na powerful group limits the opportunities and resources of a subordinate,\\npowerless group.\\nGroup conflict also frustrates basic needs. What needs it affects will de-\\npend on specific conditions, but it is likely to affect security and identity.\\nConflict between superordinate and subordinate groups, as well as the un-\\neven effects of difficult life conditions on different groups, may give rise to a\\nsense of injustice. While justice may not be a fundamental need in the same\\nway as the needs described to this point, it is nonetheless a powerful human\\nmotive.\\nThe instigation and the resulting frustration of basic needs leads to psy-\\nchological and social processes that turn a group against others. These in-\\nclude individuals turning to some group for identity; people elevating the\\ngroup by devaluing other groups and taking action against their members;\\nthe ingroup scapegoating outgroups; and ideologies or visions of social\\narrangements that promise a better life but also identify enemies.\\nOne example is Rwanda. All the instigators were present there. There\\nwere increasing economic problems, starting in the late 1980s. The inva-\\nsion of a Tutsi-led group of rebels in 1990, consisting of refugees from vio-\\nlence against Tutsis in previous decades or the children of refugees, began a\\ncivil war. Political demands by varied groups added to political chaos and\\ndisorganization. The Tutsis were scapegoated for all the problems. An in-\\ntense ideology of “Hutu power” was developed and propagated, with the\\nIndividual and Group Identities in Genocide and Mass Killing\\n161\\nfigure 7.1 Influences and Processes Contributing to Genocide and Mass Killing\\n\\n\\nTutsis defined as enemies to be destroyed (des Forges, 1999; Prunier, 1995;\\nStaub, 1999b).\\nThese psychological and social processes often lead to action against the\\ndevalued, scapegoated group, or a group that is seen as an ideological en-\\nemy, and thus is the starting point for an evolution toward possible mass\\nkilling or genocide. Individuals and groups change as they harm others.\\nThey justify their actions by devaluing their victims more and more. They\\nbecome desensitized to the suffering of their victims. The standards of \\nacceptable social behavior change, allowing and even encouraging violence.\\nInstitutions change or new ones are established that serve discrimina-\\ntion and violence. In the end, there may be a reversal of morality: killing \\nthe scapegoat or ideological enemy may become a “higher good” (Staub,\\n1989b).\\nUnless halted by bystanders, the violence evolves in intensity. Unfortu-\\nnately, both internal bystanders (members of the involved groups) and ex-\\nternal bystanders (other nations and the international community) tend to\\nremain passive. This affirms and encourages the perpetrators. At times \\noutsiders—groups and nations—support the perpetrators because of past\\nties or a desire to promote what they define as their own national interests.\\nWhen the kind of psychological and social processes here described do not\\nlead to genocide, usually bystanders are involved in a positive way—they in-\\ntervene in some way to alter the processes leading to collective violence\\n(Staub, 1989b, 1999b, 2000).\\nCertain characteristics of cultures make the genocide or mass killing\\nprocess more likely. A history of devaluation of a group of people is a pri-\\nmary characteristic. Very strong respect for authority and a monolithic soci-\\nety with a limited range of values, in contrast to a pluralistic society, are also\\nimportant. Unhealed wounds in a group owing to past victimization, which\\nlead members of the group to feel easily threatened and respond with what\\nthey see as defensive aggression, make it more likely that such a genocidal\\nprocess unfolds in response to instigation. The nature of individual identi-\\nties that the society has fostered and the nature of group self-concept are\\nalso important. \\nThe influences here described do not act separately and independently.\\nThey interrelate and join in meaningful ways. I will examine here some as-\\npects of this interrelation; others can be seen and deduced from the subse-\\nquent discussion (see figure 7.1 and Staub, 1989b). It is the frustration of\\nbasic needs by instigating conditions that leads group members, whose indi-\\nvidual identity is shaken, to turn to the group for identity, to focus more on\\ntheir social identity, or to “give themselves over” to an identity group. This\\nfrustration also leads to scapegoating and the creation of destructive ideolo-\\ngies (which identify enemies), that turns the group against another group.\\nThe kind of ideology that is adopted, and the scapegoat and ideological en-\\nemy that is selected, very much depend on characteristics of the culture. \\nWhether the group turns against another depends in part on whether \\n162\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nthe devaluation of some group is part of the culture, the nature of that de-\\nvaluation, and how intense it is (Staub, 1989b, 1999b). It depends on past\\nvictimization of the group and unhealed wounds, which make it more vul-\\nnerable to threat. It depends on the strength of “authority orientation,”\\nhow deeply and unconditionally people are accustomed to be led, to respect\\nand obey authorities and the extent they are accustomed to gain support\\nfrom their group. The stronger this orientation, the more they will be af-\\nfected by difficult life conditions and other instigators, and the more diffi-\\ncult it will be to stand on their own. Their need to turn to new authorities\\nwho might be able to guide and support them will be greater.\\nA history of devaluation of a group, which all members of a society tend\\nto absorb to some extent, a strong authority orientation, and a monolithic\\nrather than pluralistic-democratic culture, as well as other cultural charac-\\nteristics (Staub, 1989b), all make it less likely that people will oppose initial\\nactions against a victim group, or initial “steps along the continuum of de-\\nstruction.” The passivity of bystanders has the effect of allowing and even\\nencouraging the evolution of violence. \\nThe psychological processes that initiate and maintain the increasing \\nviolence are fostered both by intense instigators and the presence of cul-\\ntural characteristics. The social conditions that constitute the instigators af-\\nfect both perpetrators and internal bystanders. As the violence evolves, it\\nconstitutes culture change; further evolution happens in a changing and\\nchanged cultural context. \\nIdentity and the Origins of Mass Killing and Genocide\\nI now describe how individual and group identities are involved in the paths\\nto genocide or mass killing depicted in figure 7.1. T\\no begin, severe eco-\\nnomic problems deeply impact people. Unemployment, or the inability to\\nprovide for oneself and one’s family owing to severe inflation, makes people\\nfeel insecure and ineffective. Such conditions call into question people’s\\npersonal identity and self-worth: Who am I if I have no work and cannot\\ntake care of the basic material needs of myself and my family? At the same\\ntime that identity is questioned, ties to one’s group can also weaken because\\nthe problems originate in society and social conditions.\\nIntense political conflict and disorganization also affect people’s experi-\\nence of their group and perception of the world. Rapid social change, which\\noften occurs at times of economic and political crises, adds to social chaos\\nand disorganization. Since substantial social change requires adjustment,\\neven positive change places demands on people. \\nThe different elements of difficult life conditions, which may occur sepa-\\nrately but frequently are present in some combination, frustrate many of\\nthe basic needs. In difficult times, focusing on their own needs, people be-\\ncome disconnected from each other. At times when they need it most, peo-\\nIndividual and Group Identities in Genocide and Mass Killing\\n163\\n\\n\\nple lack connection and support. The social upheaval makes their old\\nworldview inadequate in providing a comprehension of reality: an under-\\nstanding of how people, society, and the world operate, and of their own\\nrole or place in the scheme of things. \\nAs a result, when a society experiences persistent, intense life problems,\\nits members desperately need a renewed basis for identity, security, a sense\\nof effectiveness, connection to other people, and comprehension of reality.\\nThey turn for identity and the fulfillment of other basic needs to a new\\ngroup, or to leaders who strengthen the group through scapegoating out-\\nsiders or offer it new identity, meaning, and hope through ideologies that\\npromise a better life and offer the fulfillment of basic needs.\\nIdentity, Scapegoating, and Ideology\\nThree types of “identities” are discussed in this chapter: first, personal identi-\\nties, or the ways in which individuals answer the question, who am I?; sec-\\nond, social identities, or the extent to which individual identity is rooted in,\\nor connected to, the group; third, group self-concepts, or the socially shared\\nway members perceive and experience their group. Relevant to all of them\\nis a group’s culture and the social and political institutions that both express\\nand shape that culture. The nature of individual identity is shaped by cul-\\nture; in turn, the identities that characterize individuals maintain the cul-\\nture. Social identity theory has focused on how people categorize them-\\nselves (and others) as members of groups, how they identify themselves\\nwith certain groups, and the tendency they have to enhance themselves by\\ncomparing their group favorably with other groups (T\\najfel, 1982; Turner,\\n1987; Myers, 1999). It emphasizes the psychological experience of rooted-\\nness in and connection to the group. \\nBut social identities also have content, as defined by the culture. “We”\\nare intelligent and hardworking, or easygoing and ready to enjoy life, or\\nbrave and ready to fight for our rights or for maintaining important values.\\nThe content of social identity is what I refer to as group self-concept.\\nWhile social identity theory stresses the tendency to favorably evaluate\\none’s group, group self-concept can be positive or negative (superior or\\nweak and vulnerable). As in the case of individual identity, consciously held\\nviews and evaluations of the group can exist side by side with feelings about\\nand valuations of the group that may not be consciously held, such as a\\nsense of vulnerability or weakness.\\nIndividual identity is, in part, a social identity rooted in groups to which\\none belongs ranging from family to nation (T\\najfel, 1978, 1982; Turner,\\n1987; Bar-T\\nal & Staub, 1997). It is not surprising, therefore, that in difficult\\ntimes people try to strengthen their identity through personal identification\\nwith some group. A good social identity—an identity based in a group that\\nis seen in a positive light, for example, as effective and powerful—can also\\n164\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nbe a good source of security, feelings of effectiveness and control, and con-\\nnection. The values and worldview of the group also provide a comprehen-\\nsion of reality. \\nScapegoating some other group can strengthen both individual and\\ngroup identity and also help fulfill other needs. If another group is viewed\\nas responsible for life problems, then neither I nor my group are to blame.\\nScapegoating at such times is usually a social process in that it is done to-\\ngether with like-minded others. This not only builds identity but strength-\\nens security and connection. \\nIdeologies are almost always a part of the genocidal process. I see ideolo-\\ngies as visions of social arrangements that tell people how to live life. They\\nare blueprints for the organization of societies and relations among indi-\\nviduals. The ideologies created or adopted in difficult times offer the satis-\\nfaction of basic needs. They promise people a better life. Sometimes their\\nstated purpose is to improve life for all human beings—what I have called\\n“better world ideologies.” Communism has been such an ideology. Nation-\\nalistic ideologies promise better life to people in a particular group by en-\\nhancing the group’s wealth, power, or purity (Staub, 1989b).\\nOften, ideologies have both better life and nationalistic components.\\nThis was true of the Nazi ideology. The Khmer Rouge ideology was osten-\\nsibly a better world ideology, focusing on total social equality. But it was\\ngroup based, looking to the group’s history as a simple agricultural society\\nfor ways to create this social equality. Without explicit elaboration it also\\nincluded a strongly nationalistic agenda. This expressed itself in the espe-\\ncially intense violence against the Vietnamese in Cambodia, in hostility to\\nVietnam, an ancient enemy, and in genocidal violence against other minori-\\nties in Cambodia. \\nLegitimizing ideologies (Levin, Sidanius, Rabinotitz, & Frederico, 1998;\\nStaub, 1989b, 1999b) offer justifications for the power and privilege of\\ndominant groups. Violent reactions by powerful groups to demands by sub-\\nordinate groups for rights and privilege have been one of the most frequent\\nsources of mass killing since World War II (Fein, 1993). In my view, this \\nresponse of dominant groups is not simply to a threat to their power and\\nprivilege. They also respond to a threat to their identity and worldview\\n(Staub, 1989b). They have come to see themselves as rightfully occupying\\npositions of power and influence, owing to their background, history, dili-\\ngence, intelligence, bravery, or other aspects of their presumed nature. The\\nworld is arranged the “right way” with them in power (see Sidanius &\\nPetrocik, this volume).\\nPeople usually hold or turn to and adopt ideologies as part of a group or\\na movement. At times some people create or put together ideas that form\\nthe ideology, as did Hitler (1925) or the leaders of the Khmer Rouge in\\nCambodia (Staub, 1989b). However, it is when shared among a group of\\npeople that an ideology comes to life. The ideology defines the group and\\ngives the group and its individual members a positive social identity. It also\\nIndividual and Group Identities in Genocide and Mass Killing\\n165\\n\\n\\nprovides a basis for connection. Participating in the movement can make\\npeople feel effective and in control. The goals of the movement become the\\nmotivating force for members. The ideology and the movement that arise\\nfrom it often become central sources of people’s identity and comprehen-\\nsion of reality.\\nWhile human beings need positive visions in difficult times, the ideolo-\\ngies that emerge usually identify another group as an enemy. This group is\\nseen either as the cause of problems, or as standing in the way of the ide-\\nology’s fulfillment, or both. They, therefore, become the objects of hostility.\\nHarmful actions begin to be directed at them. \\nThe Role of Individual Identities in Genocide\\nCertain individual identities that a culture promotes make the genocidal\\nprocess more likely. Research and theory have identified differences in the\\ntype of self-concept or individual identity that is shaped by individualist and\\ncollectivist cultures (Triandis, 1994). Differences in women’s and men’s\\nidentities in Western societies, including the United States, have also been\\nnoted (Sampson, 1988). It has been proposed, for example, that male iden-\\ntity is more autonomous, more like the identity shaped by individualistic\\ncultures, while female identity is more relational, like identities shaped by\\ncollectivist cultures.\\nConsidering conditions that give rise to genocide, both relational/collec-\\ntivist and autonomous/individualist identities may have vulnerabilities. Dif-\\nficult life conditions may especially frustrate autonomous persons. They\\nhave learned to stand on their own, but now they are unable to effectively\\nfend for themselves or their families. American soldiers who became pris-\\noners of war in the Korean War had more difficulty in resisting brainwash-\\ning than, for example, Turkish soldiers, seemingly because the former were\\ntrying to face it alone while the latter supported each other (Kincaid, 1959).\\nRelational identities may also create difficulties. People with such identi-\\nties may feel interconnected with other people in a way that makes it espe-\\ncially difficult to speak out against and oppose the direction the group is be-\\nginning to take. As their group or society begins to scapegoat, identify an\\nenemy, discriminate, or take other harmful actions against another group,\\nthey are more likely to remain silent and go along. This makes more proba-\\nble the evolution of increasing discrimination and violence. \\nI have suggested that relational/collectivist identities may be differenti-\\nated into embedded and connected (Staub, 1993a). The embedded identity,\\nwhich I just described, is characterized by a strong connection to other peo-\\nple that also embodies dependence on the group and an inability to separate\\nfrom it. In contrast, people who have developed connection to others as in-\\nherent to their identities, but who also have learned to stand on their own,\\nto be separate, have connected selves or identities. \\nThe nature of personal identity has implications for peoples’ relationship\\n166\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nto their group. People with connected selves may be less likely to give\\nthemselves up to the group in difficult times. They may be less likely to re-\\nlinquish their own identity and give themselves over to a social or group\\nidentity than people with embedded selves. Seemingly paradoxically, be-\\ncause they can use others for support more effectively, those with connected\\nself-concepts may be even less likely to do this than people with autono-\\nmous/individualist identities. As a result, people with connected selves may\\nbe less likely than those with either embedded or autonomous identities to\\nremain passive as the evolution of harm-doing begins. They may be more\\nlikely to speak out and oppose actions contrary to their own values, or to\\nthe values they believe the group ascribes to. This is extremely difficult \\nto do but is essential in halting the evolution of increasing violence. \\nI am suggesting that identity-related differences in individuals’ relation-\\nship to the group have implications for group violence. Related to this, I\\nhave proposed that some individuals are “constructively” patriotic while\\nothers are “blindly” patriotic (Staub, 1997). In research exploring this we\\nfind that constructive patriots express love for their group, but also a will-\\ningness to question and criticize the policies and practices of the group that\\nare contrary to caring values and concern for human welfare, or contrary to\\nwhat they see as important group values (Schatz & Staub, 1997; Schatz,\\nStaub, & Lavine, 1999). Although this has not been assessed in the research,\\nI expect that constructive patriots would have connected identities. In con-\\ntrast, blind patriots, who I expect to have either individualist or embedded\\nself-concepts, love their group but are unwilling to criticize it or to question\\nits policies or practices. Different identity types and associated relationships\\nto the group have clear implications for supporting or opposing destructive\\nviews and group practices. \\nOther classifications of individual identities that are relevant to genocide\\nare possible. Some people may have secure identities (Staub, 1993a). Pre-\\nsumably their basic need for security has been fulfilled in the course of their\\nlives, especially early lives. Perhaps they also have had experiences with tak-\\ning effective action to protect themselves. As a result they feel reasonably\\nsecure in the world. They would be less affected by instigating conditions,\\nsuch as difficult life conditions, and less moved to the psychological and so-\\ncial processes that were described as a consequence of such conditions.\\nOthers, perhaps due to the frustration of the need for security early in life,\\nincluding victimization or other traumas, may develop insecure identities,\\nmaking it more likely that these psychological processes leading to genocide\\nunfold in response to instigating conditions.\\nAuthority Orientation and Identity\\nObservation and research indicate that in many cases of genocide or mass\\nkilling the society has been characterized by strong respect for authority\\n(Gourevich, 1998; Kressel, 1996; Smith, 1998; Staub, 1989b,1999b). “Au-\\nIndividual and Group Identities in Genocide and Mass Killing\\n167\\n\\n\\nthority orientation” in the culture is a term I prefer to “authoritarian per-\\nsonality” because it focuses purely on respect for authority and hierarchy,\\nwithout implications of repression of sexual feelings and other characteris-\\ntics that have been described as aspects of the authoritarian personality.\\nAuthority orientation affects the nature of identity in a number of ways.\\nIndividuals who have embedded selves are connected to others through net-\\nworks of dependent relationships. People in an authority-oriented culture\\nalso seem deeply embedded in the group, but this is based on their orienta-\\ntion to authority. The experiences they have as this orientation is developed\\nin them, normally from early childhood on, diminishes the evolution of a\\nstrong independent self. \\nIn the face of difficult life conditions or group conflict, people who have\\nrelied on leaders for guidance and protection will find it more difficult to\\nbear the threat, anxiety, and frustration of basic needs. When policies and\\npractices are instituted in a group that harm others, they will be less willing\\nto speak out and to oppose the authorities and the rest of the group. Strong\\nrespect for authority also makes obedience to immoral orders by authorities\\nmore likely.\\nThe Nature of Group Self-concept\\nWeakness, Superiority, and a Positive Self-Concept. The nature of group self-\\nconcept (by which I mean the way members consensually see and experi-\\nence their group) is also important. How do members of a group, people\\nwho identify with a group, see their group? What is their conception and\\nexperience of the group? And, how do differences in group self-concept\\npredispose toward mass killing or genocide? I have hypothesized (Staub,\\n1989b) that both a sense of weakness and vulnerability, and feelings and be-\\nliefs of superiority, may make it more likely that groups will turn to and en-\\ngage in systematic violence against others. When a group feels weak and\\nvulnerable, life problems in the society or conflict with other groups will in-\\ntensify vulnerability and the need to defend oneself. Elevating the group,\\nscapegoating another group, creating “positive visions” that identify ene-\\nmies, and self-defense through aggression become more likely. \\nA sense of superiority, when frustrated, may also become a powerful\\nsource of psychological processes and actions that contribute to genocide.\\nFeeling superior makes individuals and groups vulnerable to disconfirma-\\ntion of their superiority and, as a result, likely to experience narcissistic\\nwounds (Baumeister, 1997). However, a self-concept of superiority is differ-\\nent from a high self-esteem. Feelings of superiority may be the result of a\\nhistory of success by a society in many realms combined with the usual eth-\\nnocentrism of groups that leads to interpretation of this success as evidence\\nof the group’s superiority. But more often, a belief in one’s superiority is ex-\\naggerated and may represent a compensatory identity. In the lives of most\\nindividuals and groups there are enough small and large experiences of inef-\\n168\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nfectiveness, failure, and loss that, under normal conditions, a more moder-\\nate sense of self would develop.\\nA positive identity in individuals is likely to arise from experiences of\\nnurturance and affirmation and from success in overcoming obstacles.\\nComparable experiences on the group level can give rise to positive group\\nself-concepts. The absence of these may lead to low self-concept. Their ab-\\nsence, or negative experiences, overlaid with experiences of success but\\nwithout incorporating the difficult, painful experiences into one’s identity,\\ncan lead to a feeling of superiority, which is what I regard as a compensatory\\nidentity. In children a defensive self-esteem or defensive egotism, which was\\nfound to be different from high self-esteem, was associated with more bul-\\nlying of others. Defensive self-esteem was ascribed to children who de-\\nscribed themselves and were described by others as having positive self-\\nesteem, but were also described by others as “always wants to be the center\\nof attention; thinks too much of himself or herself; can’t take criticism”\\n(Salmivalli, Kaukiasnen, Kaistaniemi, & Lagerspetz, 1999, p. 1271).\\nGermany may be a good example of having a superior group self-\\nconcept. The group self-concept of Germans early in this century was ex-\\ntremely positive, of a kind that may be described as a superior compen-\\nsatory identity. This was due to recent military successes, economic\\nstrength, a history of cultural achievement, effective civic institutions, and\\nan internal climate that propagated self-satisfaction with German culture\\nand society beyond what is “normal” ethnocentrism (Staub, 1989b). \\nThese positive experiences might account for what seems was an early\\n20th-century German sense of superiority. But Germans also had a history\\nof problems and suffering. Many Germans died in wars in previous cen-\\nturies. During the Thirty Years’ War (1618–1648), a third of Germany’s\\npopulation died (Craig, 1982). After struggles among many smaller units,\\nGermany became a single country only late in the 19th century. History\\nleaves its mark, and perhaps underlying the outward sense of group superi-\\nority there was also a deep-seated national vulnerability. Germany was then\\nfrustrated by a series of events beginning with the loss of World War I.\\nThese included a revolution that overthrew the monarchy, hyperinflation,\\neconomic depression, political chaos, and actions by other countries that\\nGermans felt were humiliating to them (e. g., the French occupying the\\nRuhr district to ensure the receipt of compensation Germany was forced to\\npay after World War I). Whether it was defensive egoism, or normal ethno-\\ncentrism, or even a “genuine” sense of superiority, the nature of their group\\nidentity was likely important in Germans’ reactions to the intense life prob-\\nlems following World War I. \\nA decade ago I wrote: “in groups and in individuals very high self-evalu-\\nation often masks self-doubt. Persistent life difficulties may contradict high\\nself-evaluation and bring self-doubt to the surface. Even when there is no\\nunderlying self-doubt, a very high self-evaluation may be associated with\\nlimited concern for others. Among individuals, a moderately positive self\\nIndividual and Group Identities in Genocide and Mass Killing\\n169\\n\\n\\nconcept is most strongly associated with sensitivity and responsiveness to\\nother people” (from Staub, 1989, p. 55; see also Jarymowitz, 1977; Reese,\\n1961). This still seems a reasonable conclusion. Both individuals and groups\\nhave to value themselves in order to value other people, but not value them-\\nselves so strongly that others do not matter, or so highly and unrealistically\\nthat setbacks are experienced as devastating blows to collective and personal\\nself-worth. \\nCultural Devaluation and Ideologies of Antagonism \\nA form of group identity that I have called “ideology of antagonism” is es-\\npecially important in genocide and collective violence. Inherent in group\\nidentity is a differentiation between “us” and “them.” In my view a devalua-\\ntion of “them” is not inherent in, but easily arises out of, this differentiation\\n(see Brewer, this volume). There are many factors that develop a persistent\\ndevaluation of one group by another, which then becomes part of the\\ngroup’s culture. They include justification of past mistreatment of the other\\ngroup, fear of the other’s difference, and the need to create identity when\\none group has split off the other. (The latter may be an important origin of\\nChristian anti-semitism [Staub, 1989b]). Devaluation embedded in a cul-\\nture is a crucial cultural source of a group turning against another.\\nAt times, two groups have had a long history of conflict with mutual vio-\\nlence. The evolution of increasing violence has already occurred, and as a\\nresult new violence easily flares up. In the course of this past history, not\\nonly the view of the other but the identity of each group as well has come to\\nbe defined around this enmity: the other is my enemy, and I am the enemy\\nof the other. In short, group self-concept comes to center on violent con-\\nflict with the other group (see Kelman, this volume).\\nThe Hutus and Tutsis in Rwanda seem to be an example of ideologies of\\nantagonism. Up to the 20th century, the Tutsis were dominant, but group\\nmembership was defined by a complex combination of ethnic origin and\\nclass, with rich Hutus sometimes acquiring a classification as Tutsis. Under\\ncolonial rule by the Belgians, Tutsi dominance was intentionally enhanced,\\nwith Hutus receiving increasingly bad treatment. They became an under-\\nclass with limited rights and privilege. In 1959, in a violent revolt, they\\nkilled about 50,000 Tutsis. The Hutus ruled the country after it gained in-\\ndependence in 1962. There were significant massacres of Tutsis in the\\n1960s and 1970s, with continued discrimination and more limited, sporadic\\nviolence against them after that.\\nIn the 1990s, a Tutsi-led group, consisting mainly of descendants of\\nrefugees over the past few decades, invaded Rwanda. In the course of the\\nfollowing few years, a complex political and military situation developed in\\nthe country. Dominant Hutus increasingly advocated “Hutu power,” with\\nmedia propaganda explicitly or implicitly promoting the destruction of \\nTutsis. The genocide of 1994 followed. It was stopped by the victory of the\\n170\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nTutsi-led army over the government (see des Forges, 1999; Prunier, 1995;\\nStaub, 1999b).\\nCultural devaluation is extremely persistent, and ideology of antagonism\\nis even more so. Neither tend to go away on their own, unattended, simply\\nwith the passage of time. They remain part of the culture, expressed in lit-\\nerature, in everyday relations, and in the life of institutions. They remain\\npart of the identity of individuals and groups. Active social processes are re-\\nquired to change them.\\nVictimization, Healing, and Identity\\nWhen a group has been victimized in the past, without significant healing,\\nit has an increased likelihood of becoming a perpetrator of mass violence\\n(see Eriksen, this volume). People who experience trauma are deeply af-\\nfected (Herman, 1992; Janoff-Bulman, 1992; McCann & Pearlman, 1990).\\nGroup victimization diminishes the different selves—the individual self, \\nsocial identity, and group self-concept. Who am I and who are we, if we are\\nso treated? Since individual identity is partly based on group identity, even\\nthose members of the group who are not present and have not been indi-\\nvidually in danger are deeply affected.\\nVictimized groups are more likely to respond to threat with violence. In\\ncase of conflict, their understandable self-focus makes it difficult for them\\nto consider the needs of others. Feeling insecure, their self vulnerable, and\\nperceiving the world as a dangerous place, they experience threat as more\\nintense than it is. They are more likely to engage, therefore, in aggression\\nwhich they see as defensive (Staub, 1998b). \\nHealing following victimization makes it less likely that a group turns\\nagainst another and perpetrates violence. Acknowledgment of the group’s\\nsuffering by others and the expression of caring and empathy contribute to\\nhealing. The group itself can also promote healing by directly engaging\\nwith its past experience, including certain kinds of memorials and rituals of\\nmourning and remembrance (Staub, 1998b). Healing from both individual\\ntrauma and group victimization involve repair of identity, as well as connec-\\ntions to other people (Pennebaker & Keough, 1999). The two are inher-\\nently related. The reestablishment of trust in people and the capacity to ex-\\nperience connection to them help repair identity.\\nPerpetrators also need to heal. At times, their woundedness has con-\\ntributed to their actions. Their own violence further wounds them. It seems\\nimpossible to inflict tremendous harm on others, to kill many people and\\nnot become wounded or damaged. When perpetrators and victims continue\\nto live together, as in Bosnia and Rwanda, for reconciliation to take place,\\nsome healing by both groups has to precede it. Reconciliation is required,\\nin turn, to stop a continuing cycle of violence.\\nIn the case of the Armenians who survived the genocide by the Turks in\\n1915–1916, healing was interfered with by the denial of Turkey that a geno-\\nIndividual and Group Identities in Genocide and Mass Killing\\n171\\n\\n\\ncide had taken place and by the complicity in this denial of many other\\ncountries. The relationships of the recently created Armenia to other na-\\ntions, especially Turkey and Azerbaijan populated by ethnic Turks, have\\nbeen affected by this history and the identity it has helped create. That\\nidentity has components of vulnerability, including a view of the world as\\ndangerous and, perhaps, an ideology of antagonism in relation to Turks. \\nIt is difficult to identify victim and perpetrator groups that have healed\\nand thus see the consequences of such healing. While there has been aware-\\nness of the need for individuals to heal following the mass killing in Bosnia\\nand the genocide in Rwanda, there has been little prior consciousness of the\\nneed for whole groups to heal. There have certainly been natural processes\\nthat have moved groups toward healing, but there have probably been no\\nintentional efforts specifically aimed at this by victimized groups themselves\\nor by the community of nations or by other bystanders. \\nA consciousness about the need for healing seemed to have emerged, at\\nleast in contemporary times, with respect to individual survivors of the\\nHolocaust, and later Vietnam veterans, as well as victims of individual vio-\\nlence such as sexual or physical abuse (Herman, 1992; McCann & Pearl-\\nman, 1990). The aim of such healing has been to improve individual lives,\\nnot reducing the likelihood of later group violence.\\nGermany might serve as an example of healing, even if far from com-\\nplete. Germany was wounded by the sequence of events that contributed to\\nthe Holocaust and by its own horrific actions during World War II and the\\nHolocaust. Afterwards, the Germans were led to hesitantly engage with\\ntheir past through the Nuremberg trials, other trials, and through the rest\\nof the world reminding Germans of their conduct. At the same time, in-\\nstead of acting in a generally punitive and harsh way, the world offered sup-\\nport, help, and connection to Germany. This ranged from the Marshall\\nplan, to the willingness by French leaders to make connections with Ger-\\nmany, to Germany’s becoming a member of the European Community, and\\nto other demonstrations of the world accepting Germany into the interna-\\ntional community. In different ways and for different reasons, this was true\\nof both West and East Germany. Reconnecting with others is an avenue to\\nhealing for both victims and perpetrators. \\nChanges in Identities as Harm Doing Evolves\\nAs noted above and depicted in figure 7.1, genocide and mass killing are\\nusually the outcome of an evolution of increasing harm doing by perpetra-\\ntors. There may be one victim group, although in the course of this evolu-\\ntion the boundaries expand and violence toward other groups also becomes\\nprobable. How do individual and group identities change along the way?\\nThe perpetrators come to devalue members of the victim group more\\nand more. They justify their actions, in part, by the nature of the victim,\\nwho is not only increasingly devalued but also increasingly excluded from\\n172\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nthe moral realm, the realm of people to whom moral values and moral con-\\nsiderations apply (Opotow, 1990; Staub, 1989b). Perpetrators also justify\\ntheir actions by the “higher” values of their ideology. Along the way the\\nperpetrators may come to see themselves as people who are willing to make\\nthe sacrifices required, including the killing of the victims, to fulfill impor-\\ntant goals. Thus, they may see themselves, as Himmler encouraged the SS\\nto see themselves, as bearing noble sacrifice for their higher cause (Hilberg,\\n1961). \\nIn addition to their identity changing in relation to the victims, perpetra-\\ntors are likely to become desensitized to human suffering in general. Deval-\\nuation tends to extend from the original victims to new groups, which \\nprepares the way for the extension of violence (Staub, 1989a, 1989b). In\\nGermany, for example, after the euthanasia program, the killing of physi-\\ncally handicapped, mentally ill, and other “inferior” Germans, after the\\ngenocide of the Jews and gypsies, and after violence against people in occu-\\npied countries, the authorities began to consider the killing of physically\\nugly German prison inmates (Lifton, 1986). \\nAs boundaries weaken and new groups are devalued and become poten-\\ntial or actual victims, the perpetrators’ identity as people who value human\\nbeings and human welfare is likely to change. As this process unfolds perpe-\\ntrators are likely to distance themselves from all human beings not a part of\\ntheir group, however that is defined. At the extreme, might it be the case\\nthat even the welfare of one’s ingroup and of friends and family become of\\nless concern?\\nThe identity of bystanders also changes. Internal bystanders—members\\nof the perpetrator group who are not themselves perpetrators—usually re-\\nmain passive. They are also affected by instigating conditions. They have\\ngrown up in a culture that imbues members with respect for authority and\\ndevalues the victim group. Thus, they share a “cultural tilt” with perpetra-\\ntors. This cultural background, often in combination with fear of a repres-\\nsive system, reduces the motivation to oppose the group as it moves along a\\ncontinuum of increasing violence.\\nPassivity changes bystanders. It is difficult to see others suffer and do\\nnothing. T\\no justify their passivity, reduce their guilt, and reduce empathy\\nwith victims that would make them suffer, passive bystanders tend to dis-\\ntance themselves from victims, in part by increasingly devaluing them. This\\nmakes it even more likely that they will remain passive. It also leads some\\nbystanders to join perpetrators.\\nFor example, a group of psychoanalysts in Berlin remained passive as\\ntheir fellow Jewish analysts were the object of harassment and persecution.\\nThey began to use racially appropriate Nazi terms to refer to Jewish col-\\nleagues, like “not pure Germans.” These served both as euphemisms and as\\nmeans to distance and objectify their fellow analysts. They accepted a rela-\\ntive of Herman Goring, untrained in psychoanalysis, as the head of their In-\\nstitute. They began to rewrite psychoanalytic theory to fit Nazi ideology\\nIndividual and Group Identities in Genocide and Mass Killing\\n173\\n\\n\\nand doctrine. Some of them became participants in the euthanasia program,\\nand in the end some got involved in the extermination of Jews (Staub,\\n1989a).\\nThe process in this specific instance may be an example of how by-\\nstanders distance themselves, even from familiar others and from close asso-\\nciates. Adopting at first an “objective,” disinterested stance, and introducing\\nideological thinking change both the view of and emotional relationship to\\nthe other. Through this process bystanders, as well as perpetrators, change\\nthe nature of their own self. Those changes in identity are intertwined with\\nchanges in society and culture. As violence evolves, the norms and standards\\nof society change, allowing and often promoting actions that were previ-\\nously inconceivable. For instance, in Rwanda, neighbors killed neighbors\\nand people gave up members of their families to the killers (des Forges,\\n1999; Gourevich, 1998; Staub & Pearlman, in press).\\nI have suggested that these changes in perpetrators and bystanders repre-\\nsent changes in identity. Although identity is a broad concept, we may see it\\nas including a person’s self-concept and orientation to and evaluation of the\\nself, as well as values relating to and orientation toward and evaluation of\\nother people. Thus, I am proposing here that a person’s view of the world is\\nalso part of personal identity (Epstein, 1980). Some or all of these elements\\nof identity will change in the course of a societal evolution toward genocide.\\nThe changes in a person’s self-concept and in the group’s collective sense of\\nself mutually reinforce each other.\\nHowever, the negative changes in the view of the other, and the changes\\nin the self, as individuals and a society move along a “continuum of destruc-\\ntion” (Staub, 1989a, 1989b), are not the whole story. Even in very simple\\nsocieties there is variation in personality and, presumably, self-concept.\\nOnly a small percentage of people have the “modal identity type” that may\\nbe seen as characteristic of the society (Konner, 1978). In genocidal soci-\\neties, in addition to perpetrators and passive bystanders, there are usually\\nsome people who attempt to help victims. The rescuers of Jews in Nazi Eu-\\nrope have been the primary group of such helpers who have been exten-\\nsively studied (Fogelman, 1994; Oliner & Oliner, 1988; T\\nec, 1986). But in\\nother societies as well, there have been heroes who have endangered them-\\nselves to save lives. \\nI have heard several such stories in Rwanda, both in conversation and in\\nformal interviews I conducted with people whose lives were saved. In one\\ncase, for example, a Hutu man who had worked for a woman sent another\\nman to help her after her husband was killed. The man who came, a\\nstranger carrying a Bible, repeatedly endangered himself as he faced off\\npeople who came to the house wanting to take her away. The woman telling\\nthis story was unclear about his motivation to help: she suggested his reli-\\ngion, his being a good man. Having some difficulty accepting such self-\\nsacrificial behavior, she wondered about possible selfish motives but could\\nnot identify any.\\n174\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nOnce people begin to help in some limited way, they are further shaped\\nby their own actions. Like violence, so helping evolves. People who have\\nagreed to help for a short time, or to a limited extent, change as a result of\\ntheir actions. They become more committed to the welfare of those they\\nhave helped and often to the welfare of human beings in general. They\\ncome to see themselves as people who are willing to endanger themselves to\\nhelp others. Their values and identity change (Staub, 1989b, 1993b). \\nThe Role of Identity in the Prevention of Genocide and Mass Killing\\nThe Role of Leaders \\nHow can genocide and mass killing be prevented and what is the role of dif-\\nferent kinds of identities in prevention? How can people gain positive iden-\\ntities in difficult times without turning to destructive groups or making\\ntheir groups destructive? One way is for leaders to offer positive visions that\\nare constructive, visions that offer hope but are realistically aimed at solving\\nlife problems without identifying enemies, visions that connect rather than\\nseparate people. The New Deal under Roosevelt was one example of this.\\nThis can be difficult and at times even dangerous for leaders when there is\\nalready strong enmity against a potential victim group.\\nHealing by Victimized Groups \\nAn important contributor to prevention is healing by victimized groups.\\nHealing can help repair damaged identities and worldviews (as well as fulfill\\nother frustrated needs). It can recreate some sense of security and trust in\\nthe world. As a result, it makes it less likely that a victim group becomes a\\nperpetrator. Acknowledgement of the group’s suffering, expressions of em-\\npathy and sorrow, and other forms of affirmation by the rest of the world\\ncan contribute to healing. Punishment of perpetrators can also help victims.\\nIt can communicate to victims that what was done to them is not regarded\\nas normal and acceptable by the world. That message, in turn, makes it\\nseem less likely that it will happen again (Staub, 1998b, 1999c). \\nEngaging with their painful experiences can also help survivors of group\\nviolence heal. Trauma victims frequently avoid such engagement even\\nthough they are often affected by what has happened to them in many ways:\\nthrough intrusive imagery, hyper arousal, negative feelings about the self,\\nmistrust and difficulty in relating to people (Herman, 1992; McCann &\\nPearlman, 1990). Engagement can lead to integration of the self and recon-\\nnection with people (Pennebaker & Keough, 1999). It can lead to finding\\nmeaning, perhaps in helping other people, working for the prevention of\\nsimilar violence against people, or living life in a different way. \\nHealing makes reconciliation possible. In places like Rwanda and Bosnia,\\nIndividual and Group Identities in Genocide and Mass Killing\\n175\\n\\n\\nwhere victims and perpetrators live together, or where mutual violence has\\noccurred, reconciliation is essential to prevent continued cycles of violence.\\nThe self is to a significant extent defined by relationships to others. Ideo-\\nlogies of antagonism, intense devaluation and hostility toward another, \\nor strong fear of a neighbor are central aspects of identity. Reconciliation,\\nwhich when genuine requires forgiveness and overcoming devaluation and\\nantagonism, changes identity in central ways. As a result of healing and rec-\\nonciliation, feelings of diminished self-worth and vulnerability can lessen.\\nMistrust in people and in the world can also be reduced. Both the desire for\\nrevenge as a form of justice and the readiness to use force to defend oneself\\nand one’s group can decrease. Trust in people, trust in the world as a livable\\nplace, and hope for the future can reemerge.\\nHow can healing, reconciliation, and positive connection to others in\\ngeneral be promoted in previously victimized groups? We have so far only a\\nlimited repertoire of skills and techniques for this. T\\no begin with, victimized\\ngroups have to start with at least some desire for change, which, in case of\\nintense antagonism, may be lacking. When there is some openness, con-\\ntexts for healing can be created by bringing members of different groups to-\\ngether. People can write about and talk to each other about what has hap-\\npened to them (Pennebaker & Keough, 1999). They can talk about their\\npain, sorrow, loss, and anger. As they listen to each other and respond with\\nempathy, pain, which has been a source of disconnection, can become a\\nsource of connection (Staub & Pearlman, 1996).\\nThis is one aspect of a project that we have been conducting in Rwanda\\non “healing, forgiveness, and reconciliation.” In addition to such experien-\\ntial aspects of promoting healing and reconciliation, the training we have\\nused also had psycho-educational aspects. Participants learned about and\\ndiscussed the impact of trauma on people in the hope that this would help\\nthem understand their own experience better. They learned about avenues\\nto healing. They heard about the fulfillment and frustration of basic needs\\nand they received information and engaged in discussion about the origins\\nof genocide (Staub, 2000; Staub & Pearlman, in press).\\nThis last element of the training has had seemingly dramatic effects.\\nLearning about a number of other genocides and about how genocide\\ncomes about seemed to reaffirm for participants their own humanity. If\\nthese terrible things have been done to others as well, and if there are un-\\nderstandable human processes that give rise to them, then the victims in\\nRwanda remain part of the human realm in spite of what happened to them.\\nIn addition, understanding the forces that led to genocide seemed to hu-\\nmanize the perpetrators, both in the survivors’ eyes and, perhaps, in their\\nown, thereby contributing to reconciliation. The type of training described,\\nwhich we conducted in Rwanda, can help not only victims but also mem-\\nbers of the perpetrator group. Genocide and mass killing, inflicting violence\\non others, deeply wounds the perpetrators. It also wounds members of the\\nperpetrator group who have remained passive bystanders. They also are in\\n176\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nneed of healing. Healing, in turn, can open the way to reconciliation. As\\nreconciliation enhances a feeling of security, it furthers the possibility of\\nhealing.\\nSome of the participants in our project used the approach provided in\\nthe training, integrated with their own traditional approach, with groups \\nin the community. Formal, experimental assessment showed that trauma\\nsymptoms decreased among community members who participated in\\nmeetings led by people we trained, both over time, and in comparison to\\ngroups led by different trainers who used their organizations’ traditional\\napproach, as well as in comparison to no treatment control groups. Orien-\\ntation to members of the other ethnic group became more positive among\\ncommunity members who participated in the meetings that used this \\nintegrated approach, both over time and in comparison to people in the\\nother groups (Staub, Pearlman, Gubin and Hagengimana, unpublished \\nresearch).\\nHealing by leaders is also important. Like the other members of a vic-\\ntimized group, so the people who emerge as leaders have been affected by\\npast victimization. In Bosnia, a number of the Serb leaders have had mem-\\nbers of their families killed by the Croat regime that killed many Serbs and\\nJews during World War II. Outsiders who engage with such leaders need to\\nbe aware of and attempt to address their wounds to individual and group\\nidentity, if they are to effectively influence policies and practices (Staub,\\n1999c).\\nHostile groups can also benefit from joint efforts in behalf of shared\\ngoals. Depending on the nature of hostility and past injuries, this may re-\\nquire prior dialogue, problem solving in small groups (see Kelman, in this\\nbook), or healing. But significant engagement by members of groups with\\neach other in the course of working together for shared goals can bring\\nbenefits in overcoming devaluation and antagonism (Allport, 1954; Petti-\\ngrew, 1997; Staub, 1989b, 1999c).\\nCulture Change\\nAnother group-level issue is culture change. Cultural devaluation and au-\\nthority orientation shape identities. Monolithic cultures, with a limited set\\nof values and limited access of at least some groups of people to the public\\ndomain, also make genocide more likely, while pluralistic cultures make\\ngenocide less likely. When respect for authorities is moderate and there are\\nvaried values and points of view that can be expressed in the public domain,\\npeople are more able to tolerate uncertainty and therefore social upheaval\\n(Soeters, 1996).\\nDemocracies, which are pluralistic, are unlikely to engage in genocide\\n(Rummel, 1994). This, however, seems especially true of “mature” democ-\\nracies. (Staub, 1996b) These are democracies in which laws are enforced, in\\nwhich all groups can participate in public life, and which have a “civil soci-\\nIndividual and Group Identities in Genocide and Mass Killing\\n177\\n\\n\\nety “ (i. e., well-established civic institutions). Germany was not this kind of\\ndemocracy at the time of the Weimar Republic, nor was Argentina at the\\ntime of “the disappearances” (Staub, 1989b), when elected governments\\nwere regularly replaced by military rule. Working to create democracy and\\na civil society is one way to bring about culture change that makes genocide\\nless likely. Such a culture is likely to form individuals whose relationship to\\nthe group is more akin to constructive, rather than blind, patriotism and\\nwho are less likely to be embedded in their group.\\nMultiple group identities seem important in preventing genocide. When\\npeople can gain identity and fulfill other basic needs through connection to\\na variety of groups, they are less likely to give themselves over to and lose\\nthemselves in a group that turns against others. Their ability to oppose de-\\nstructive ideologies and practices will increase. \\nHowever, these groups need to be reasonably independent of each other.\\nIn many cases of genocide, subgroups of society, which could have been the\\nsource of at least somewhat independent identities, have subordinated\\nthemselves to or have become embedded in the state or in an ideological\\nmovement. This was recently the case with the churches in Rwanda. Many\\nchurch leaders, both in the dominant Catholic Church and in the minority\\nProtestant churches, were long connected to the ruling elements of the\\ncountry. They became accomplices during the genocide, facilitating the\\nkilling of Tutsis (des Forges, 1999; Gourevich, 1998).\\nInclusive Caring, Moral Courage, and the Behavior of Bystanders\\nWhat type of socialization will lead children to speak out against harmful\\npractices within their groups? What is required for people to exert influ-\\nence within their group so that it does not remain a passive bystander in the\\nface of violence against people outside the group? Warmth, affection, effec-\\ntive but not punitive guidance, allowing children reasonable autonomy,\\nhelping them develop a sense of significance and a “voice,” and guiding\\nchildren to help and not harm others make it likely that they will develop\\ninclusive caring, connected identities and constructive patriotism, as well as\\nmoral courage (Staub, 1996a). Such characteristics would make genocide\\nless likely.\\nConstructive patriotism has a value basis in caring about human welfare.\\nHolding such a value motivates criticism and opposition to harmful policies\\nand practices of the group. Caring about human welfare is a value that has\\nto develop in children through experiential learning. Without love and af-\\nfection and positive guidance in their own lives, without feeling valued,\\nchildren are unlikely to learn to value others.\\nT\\no publicly oppose harmful practices and advocate benevolent ones of-\\nten requires moral courage. This is not necessarily physical courage, but the\\nability and willingness to tolerate conflict. The kind of individual identities\\npeople develop and the nature of social identities, what the group requires\\n178\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nof its members, are central to this. Even when there is no physical danger,\\npeople often won’t speak out against harmful and potentially violent prac-\\ntices out of concern for ostracism or exclusion from the group. Is it possible\\nto create cultures in which opposition for the sake of human welfare is \\nvalued? \\nIt is relevant in this context that, according to one researcher, about one-\\nthird of rescuers of Jews in Nazi Europe were marginal to their group (T\\nec,\\n1986). They had a different religion, or one foreign parent, or moved to\\ntheir community from some other place, or may have been simply person-\\nally different and therefore less taken in by or less invested in their commu-\\nnity. This marginality may have made them less identified with their group,\\nperhaps fearing less the possibility of ostracism and enabling them to have a\\nseparate, independent perspective.\\nInclusive caring means the development of caring for people beyond the\\ngroup. It has to do with the relationship of “us” and “them,” with the devel-\\nopment of an identity that makes individuals and groups not draw a line at\\nthe boundary of their group that diminishes concern for the welfare of\\nthose beyond the group. (Staub, 1996a). Many of the rescuers grew up in\\nfamilies where this line between their group and minorities was not as\\nsharply drawn as in the rest of the group. The parents and the family en-\\ngaged with people outside the “ingroup,” often including Jews (Oliner &\\nOliner, 1988).\\nIndividual identities based on inclusive caring would lead people to want\\ntheir countries to be active bystanders, who would respond when there is \\nviolence against people outside their borders. Self-concepts that embody\\ninclusive caring would be likely to move individuals to exert influence on\\ntheir governments not to remain passive. This is essential for the preven-\\ntion of mass killing and genocide. While this has rarely happened, the \\nlimited examples that exist strongly suggest that if the community of na-\\ntions responds early, as harm doing just begins, the evolution of increasing\\nviolence can usually be stopped. \\nEven groups of individuals can exert significant influence. After the revo-\\nlution in Iran the Bahai, for a long time a devalued and persecuted group,\\nwere immediately intensely persecuted. Over 200 Bahai were executed. The\\ninternational Bahai community immediately made representations to the\\nvarious governments where Bahai lived and to the international community.\\nThe United Nations, as well as individual nations, passed resolutions con-\\ndemning the killing and persecution. While the persecution did not stop,\\nthe killings did (Bigelow, 1993), with some limited resumption, unfortu-\\nnately, in the 1990s. \\nIn places where violence could have been much greater, as in Northern\\nIreland, or between Israelis and Palestinians, usually external bystanders\\nhad some role in limiting violence. Internal bystanders, members of differ-\\nent groups reaching out to each other, perhaps enabled by the behavior of\\nexternal bystanders, also seems to have had a role (Staub, 1999b). In North-\\nIndividual and Group Identities in Genocide and Mass Killing\\n179\\n\\n\\nern Ireland, the British military and political presence provided support to\\npolitical and social leaders who have worked for opportunities in jobs and\\nmore social equality for Catholics, as well as to community organizations\\nthat brought members of the two groups together for dialogue and other\\nforms of engagement with each other (Cairns & Darby, 1998; Staub, 1999b;\\nsee, however, White, this volume).\\nWith early, committed, and resolute actions by the community of na-\\ntions, the evolution of violence can probably be inhibited much of the time\\nwithout the use of military force (Staub, 1999b). But for this to happen re-\\nquires important changes in identities: in individuals’ views of themselves\\nand their nations and in nations’ views of themselves and of the interna-\\ntional community. It requires committed bystanders and the creation of an\\ninternational system that leads to action when action is required. \\nSummary and Conclusions\\nIdentities of individuals, the nature of the relationship to the group in\\nwhich they invest their identity, and group-self concept are all important in\\nthe generation of genocide, mass killing, and other harmful behavior by\\ngroups toward other groups. They are also important in prevention. How-\\never, identities must be considered in the context of other aspects of indi-\\nvidual and group psychology. Identity is a concept that refers to certain\\nstructures in individual personality with related processes. In reality, of\\ncourse, human beings (and cultures and societies), are unitary wholes with\\ninterrelated structures and processes.\\nImportant concepts related to identity that have relevance to genocide\\nand mass killing include values, the way the self and others are divided (“us”\\nand “them”), orientation to authority, the individual’s relationship to the\\ngroup, and basic human needs (figure 7.1). One may conceptualize all these\\nas part of identity, or separate them and see them as concepts with their\\nown structures and processes. I find it useful to separate these and then con-\\nsider the relationships among them. As I suggested in this chapter, the na-\\nture of a person’s identity and group self-concept will certainly affect orien-\\ntation to authority, divisions between “us” and “them,” and so on. \\nI have proposed a series of steps that can take members of a group from\\ninstigating factors (left side of figure 7.1) to full-fledged genocide or mass\\nmurder (right side of figure 7.1). Instigating factors include intergroup con-\\nflict and difficult life circumstances (e.g., poverty) that threaten basic needs\\n(e.g., need for security, sense of self-worth, connections to others). Behav-\\niors that elevate the ingroup and derogate outgroups (e.g., scapegoating)\\nprovide means of coping with these threats to the self. Derogation of the\\noutgroup can encourage initial acts of violence against the outgroup. Initial\\nviolence, in turn, sets in motion a cycle of further violence and polarization\\n180\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nthat becomes tied to self and group identity. This cycle may eventually esca-\\nlate to mass killing and genocide. The development of this chain of events is\\nmoderated by both cultural variables (top of figure 7.1) and the behavior of\\nbystanders (bottom of figure 7.1). For example, ideologies that legitimize\\nviolence and a group’s history of being a victim of violence are two cultural\\nfactors that can increase the likelihood that initial instigators will lead to\\nmass killing or genocide. On the other hand, involvement of bystanders\\n(whether ingroup members or outsiders, such as the international commu-\\nnity) can mitigate the march from instigators to eventual genocide.\\nOne of the important relationships suggested by this chapter is that be-\\ntween identity and basic needs (figure 7.1). Security, connection, and effec-\\ntiveness are important from birth on, and their fulfillment and frustration\\nbegin at birth. When we consider motivation for action that may have to do\\nwith identity, under many conditions a complex of motives and needs will\\nbe involved. When people engage in destructive ways of fulfilling their need\\nfor identity (for example, by joining an ideological movement that aims to\\ndestroy “enemies”), they are also usually seeking to satisfy other basic needs\\nthat I have proposed: security, connection, a sense of effectiveness and con-\\ntrol, and the need for comprehension of reality. The same is true when peo-\\nple act in positive ways. Reconciliation with others may arise from and has\\neffects on identity, but also fulfills varied basic needs and expresses and af-\\nfirms certain values.\\nIndividual identity is a storehouse of and is shaped by all the experiences\\na person has, at least all significant experiences. For example, it is shaped by\\na history of fulfillment and frustration of basic needs. Some may regard it,\\ntherefore, as superordinate to other aspects of personality and their related\\nprocesses. In my view, however, the most useful theory will maintain and\\ndevelop conceptual distinctions. The important task for basic theory build-\\ning is to identify the interrelations of concepts and related processes.\\nHow are these concepts, underlying structures, and processes interre-\\nlated in the tragic domain of human life this chapter addresses? Is identity a\\nmore basic need than others? When a basic need such as security or con-\\nnection is fulfilled, does that automatically strengthen identity? How can\\nbasic needs be fulfilled constructively for whole groups of people? Will \\nanswers to these and other questions help in halting and preventing group\\nviolence? How can we apply gains we make in theoretical understanding to\\nhelp us in the practical realm, in preventing violence?\\nThese questions point to an obligation of psychologists. The obligation\\nis even deeper than applying knowledge to the great practical issues of hu-\\nman life. It is to be purposeful in our efforts to gain knowledge, whether in\\nthe laboratory or through experience in real-life settings. The obligation is\\nto be guided by a vision of reducing human suffering and enhancing human\\nwelfare, especially when the knowledge we seek relates to such important\\nmatters as the origins and prevention of genocide and mass killing.\\nIndividual and Group Identities in Genocide and Mass Killing\\n181\\n\\n\\nReferences\\nAllport, G. W. (1954). The nature of predjudice. Reading, MA: Addison-Wesley.\\nBar-T\\nal, D., & Staub, E. (1997). Introduction: The nature and forms of patriotism.\\nIn D. Bar-T\\nal & E. Staub (Eds.), Patriotism in the lives of individuals and groups.\\nChicago: Nelson-Hall.\\nBaumeister, R. F. (1997). Evil: Inside human violence and cruelty. New York: W.H.\\nFreeman.\\nBigelow, K. R. (1993). A campaign to deter genocide: The Baha’i experience. In H.\\nFein, (Ed.), Genocide watch. New Haven: Yale University Press.\\nCairns, E., & Darby, J. (1998). The conflict in Northern Ireland. American Psycholo-\\ngist, 53, 754–760.\\nCraig, G. A. (1982). The Germans. New York: New American Library.\\ndes Forges, A. (1999). Leave none to tell the story: Genocide in Rwanda. New York:\\nHuman Rights Watch.\\nEpstein, S. (1980). The self-concept: A review and the proposal of an integrated\\ntheory of personality. In E. Staub (Ed.), Personality: Basic aspects and current re-\\nsearch. Englewood Cliffs, NJ: Prentice-Hall.\\nFein, H. (1993). Accounting for Genocide after 1945: Theories and some findings.\\nInternational Journal of Group Rights, 1, 79–106.\\nFogelman, E. (1994). Conscience and courage: Rescuers of Jews during the Holocaust.\\nNew York: Anchor Books.\\nHarff, B., & Gurr, T. R. (1990). Victims of the state genocides, politicides and group\\nrepression since 1945. International Review of Victimology, 1, 1–19.\\nGourevich, P. (1998). We wish to inform you that tomorrow we will be killed with our\\nfamilies. New York: Farrar, Straus and Giroux.\\nHerman, J. (1992). Trauma and recovery. New York: Basic Books.\\nHilberg, R. (1961). The Destruction of the European Jews. New York: Harper & Row.\\nHitler, A. (1925). Mein Kampf. Translated by Ralph Manheim. Boston, MA:\\nHoughton Mifflin Company.\\nJanoff-Bulman, R. (1992). Shattered assumptions. New York: The Tree Press.\\nJarymowitz, M. (1977). Modification of self-worth and increment of prosocial sensi-\\ntivity. Polish Psychological Bulletin, 8, 45–53.\\nKincaid, E. (1959). In every war but one. New York: Norton.\\nKonner, M. (1978). Social and personality development: An anthropological \\nperspective. In M. Lamb (Ed.), Social and personality development. New York:\\nHolt.\\nKressel, N. J. (1996). Mass hate: The global rise of genocide and terror. New York:\\nPlenum Press.\\nLevin, S., Sidanius, J., Rebinowitz, J. L., & Frederico, C. (1998). Ethnic identity, \\nlegitimizing ideologies and social status: A matter of ideological asymmetry.\\nPolitical Psychology, 19, 373–404.\\nLifton, R. J. (1986). The Nazi doctors: Medical killing and the psychology of genocide.\\nNew York: Basic Books.\\nMcCann, L. I., & Pearlman, L. A. (1990). Psychological trauma and the adult survivor:\\nTheory, therapy, and transformation. New York: Bruner/Mazel.\\nMiller, A. (1982). For your own good: Hidden cruelty in child-rearing and the roots of \\nviolence. New York: Farrar, Straus and Giroux.\\nMyers, D. (1999). Social psychology. New York: McGraw-Hill.\\n182\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nOliner, S. B., & Oliner, D. (1988). The altruistic personality: Rescuers of Jews in Nazi \\nEurope. New York: Free Press.\\nOpotow, S. (Ed.). (1990). Moral exclusion and injustice. Journal of Social Issues, 46 (1).\\nPennebaker, J. W., & Keough, K. A. (1999). Revealing, organizing and reorganizing\\nthe self in response to stress and emotion. In R. J. Contrada, & R. D. Ashmore\\n(Eds.), Self, social identity, and physical health. (pp. 101-121) London: Oxford. \\nPettigrew, T. F. (1997). Generalized intergroup contact effects on prejudice. Person-\\nality and Social Psychology Bulletin, 23, (2), 173–185.\\nPrunier, G. (1995). The Rwanda crisis: History of a genocide. New York: Columbia\\nUniversity Press.\\nReese, H. (1961). Relationships between self-acceptance and sociometric choices.\\nJournal of Abnormal and Social Psychology, 62.\\nRummel, R. J (1994). Death by government. New Brunswick, NJ: Transaction \\nPublishers.\\nSalmivalli, C., Kaukiainen, A., Kaistaniemi, L., & Lagerspetz, K. M. J. (1999). Self-\\nevaluated self-esteem, peer-evaluated self-esteem, and defensive egotism as pre-\\ndictors of adolescents’ participation in bullying situations. Personality and Social\\nPsychology Bulletin, 25, 1268–1278.\\nSampson, E. E. (1988). The debate on individualism. American Psychologist, 47,\\n15–22.\\nSchatz, R., & Staub, E. (1997). Manifestations of blind and constructive patriotism.\\nIn D. Bar-T\\nal & E. Staub (Eds.), Patriotism in the lives of individuals and groups.\\nChicago: Nelson-Hall.\\nSchatz, R., Staub, E. & Lavine, H. (1999). On the varieties of national attachment:\\nBlind versus constructive patriotism. Political Psychology, 20. 151–175.\\nSimpson, C. (1993). The splendid blond beast. New York: Grove Press.\\nSmith, N. S. (1998). The psychocultural roots of genocide. American Psychologist, 53,\\n743–753.\\nSoetters, J. L. (1996). Culture and conflict: An application of Hofstede’s theory to\\nthe conflict in the former Yugoslavia. Peace and Conflict: Journal of Peace Psychol-\\nogy, 2 (3), 233–244.\\nStaub, E. (1989a). Steps along the continuum of destruction: The evolution of by-\\nstanders, German psychoanalysts and lessons for today. Political Psychology, 10,\\n39–53.\\nStaub, E. (1989b). The roots of evil: The origins of genocide and other group violence. New\\nYork: Cambridge University Press.\\nStaub, E. (1993a). Individual and group selves, motivation and morality. In T. Wren,\\nand G. Noam, (Eds.), Morality and the self. (pp. 337–359) Cambridge: MIT Press.\\nStaub, E. (1993b). The psychology of bystanders, perpetrators and heroic helpers.\\nInternational Journal of Intercultural Relations, 17, 315–341.\\nStaub, E. (1996a). Altruism and aggression in children and youth: Origins and cures.\\nIn R. Feldman, (Ed.), The psychology of adversity. Amherst: University Massachu-\\nsetts Press.\\nStaub, E. (1996b). Cultural-societal roots of violence: The examples of genocidal \\nviolence and of contemporary youth violence in the United States. American \\nPsychologist, 51, 117–132.\\nStaub, E. (1997). Blind versus constructive patriotism: Moving from embeddedness\\nin the group to critical loyalty and action. In E. Staub & D. Bar-T\\nal (Eds.), Patri-\\notism. Chicago: Nelson-Hall.\\nIndividual and Group Identities in Genocide and Mass Killing\\n183\\n\\n\\nStaub, E. (1998a). Basic human needs and their role in altruism and aggression. Un-\\npublished manuscript. Department of Psychology. University of Massachusetts\\nat Amherst.\\nStaub, E. (1998b). Breaking the cycle of genocidal violence: Healing and reconcilia-\\ntion. In J. Harvey (Ed.), Perspectives on loss: A sourcebook. Philadelphia: T\\naylor and\\nFrancis.\\nStaub, E. (1999a). A brighter future: The evolution of caring and nonviolent chil-\\ndren. Book in preparation.\\nStaub, E. (1999b). The origins and prevention of genocide, mass killing, and other\\ncollective violence. Peace and Conflict: Journal of Peace Psychology, 5 (4), 303–336.\\nStaub, E. (1999c). The roots of evil: Social conditions, culture, personality and basic\\nhuman needs. Personality and Social Psychology Review, 3, 179–192.\\nStaub, E. (2000a). Genocide and mass killing: Origins, prevention, healing and \\nreconcilation. Political Psychology, 21, 367–382.\\nStaub, E. (2000b). Mass murder: Origins, prevention and U.S. involvement. Encyclo-\\npedia of violence in the United States. New York: Charles Scribner’s Sons, Reference\\nBooks.\\nStaub, E., & Pearlman, L. (1996). Trauma and the fulfillment of the human poten-\\ntial. Workshop presented of the meetings of the International Society for Trau-\\nmatic Stress Studies. November, San Francisco.\\nStaub, E., & Pearlman, L. A. (in press). Healing, reconciliation and forgiving after\\ngenocide and other collective violence. In Forgiveness and Reconciliation, Radnor,\\nPA. T\\nempleton Foundation Press. \\nStaub, E., Pearlman, L. A., Gubin, A., and Hagengimana, A. Healing, forgiveness\\nand reconciliation in Rwanda: Intervention and experimental evaluation. Manu-\\nscript in preperation. Department of Psychology, University of Massachusetts at\\nAmherst.\\nT\\najfel, H. (1978). Social categorization, social identity and social comparison. In H.\\nT\\najfel (Ed.), Differentiation between social groups (pp. 61–76). London: Academic\\nPress.\\nT\\najfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psy-\\nchology, 33, 1–39.\\nT\\nec, N. (1986). When light pierced the darkness: Christian rescue of Jews in Nazi-occupied\\nPoland. New York: Oxford University Press.\\nTriandis, H. C. (1994). Culture and social behavior. New York: McGraw-Hill.\\nTurner, J. C. (1987). Rediscovering the social group: A self-categorization theory. New\\nYork: Basic Blackwell.\\n184\\nThe Contribution of Social Identity to Violent Intergroup Conflict\\n\\n\\nPART IV\\nTHE ROLE OF SOCIAL \\nIDENTITY IN REDUCING \\nINTERGROUP CONFLICT\\n\\n\\nThis page intentionally left blank \\n\\n\\nHerbert C. Kelman\\n8\\nThe Role of National Identity in\\nConflict Resolution\\nExperiences from Israeli-Palestinian \\nProblem-Solving Workshops\\nMy colleagues’ and my work as scholar-practitioners has focused on analysis\\nand resolution of protracted, seemingly intractable conflicts between na-\\ntional, ethnic, or other kinds of identity groups, best exemplified by inter-\\ncommunal conflicts, such as those in Cyprus, Northern Ireland, Sri Lanka,\\nBosnia, and apartheid South Africa. My own most intensive and extensive\\nexperience, over the past quarter-century, has been with the Israeli-Pales-\\ntinian conflict, and my analysis draws primarily on that experience.1\\nUsing the Israeli-Palestinian conflict as a case in point, this chapter ex-\\namines the way in which issues of national identity can exacerbate an inter-\\nnational or intercommunal conflict and the way in which such issues can be\\naddressed in conflict-resolution efforts. The chapter starts out with a brief\\nhistory of the Israeli-Palestinian conflict, setting the stage for the identity\\nissues at the heart of the conflict. It then proceeds to describe the struggle\\nover national identity between the two peoples, which has led them to per-\\nceive their conflict in zero-sum terms, with respect to not only territory and\\nresources but also national identity and national existence. Next, it argues\\nthat long-term resolution of this and similar deep-rooted conflicts requires\\nchanges in the groups’ national identities, such that affirmation of one\\ngroup’s identity is no longer predicated on negation of the other’s identity.\\nSuch identity changes are possible as long as they leave the core of each\\ngroup’s national identity intact. Furthermore, the chapter proceeds to ar-\\ngue, such changes need to be and can be “negotiated” between the two\\ngroups. One venue for negotiating identity, described in the next section, is\\nprovided by the problem-solving workshops between Israeli and Palestinian\\nelites that my colleagues and I have convened for many years. Finally, the\\n187\\n\\n\\npaper concludes with an illustration of the possibilities and limits of the ne-\\ngotiation of identity, based on a joint Israeli-Palestinian exploration of the\\nproblem of Palestinian refugees.\\nThe Israeli-Palestinian Conflict\\nThe Israeli-Palestinian conflict is now more than a century old (see T\\nessler,\\n1994, for a comprehensive account of the history of the conflict, or\\nMendelsohn, 1989, and Gerner, 1991, for shorter accounts). Its origins go\\nback to the birth of political Zionism at the end of the 19th century (see\\nHalpern, 1969, and Hertzberg, 1973). The early decades of the 20th cen-\\ntury brought to Palestine waves of Jewish immigrants who purchased land,\\nbuilt settlements and social institutions, and clearly signaled their intention\\nto establish a Jewish homeland and ultimately a Jewish state in Palestine.\\nThe growing Jewish presence was soon perceived as a threat by the Arab\\npopulation of the land, which was itself influenced by the development of\\nArab nationalism and the construction of a specifically Palestinian identity\\n(see Muslih, 1988, and Khalidi, 1997). Violence first erupted in the 1920s\\nand has continued to mark the relationship between the two peoples ever\\nsince.\\nDuring the period of the British mandate, which was established after\\nWorld War I, various formulas for the political future of Palestine were ex-\\nplored, including partition and establishment of a federal state, but none\\nwas found to be acceptable to both the Arab and the Jewish populations (or\\nindeed to either one of them). In November 1947—in the wake of World\\nWar II and the decimation of European Jewry—the United Nations Gen-\\neral Assembly voted to end the British mandate over Palestine (on May 15,\\n1948) and to partition the land into a Jewish and an Arab state. The Zionist\\nleadership accepted the partition plan, with reservations. The Arab leader-\\nship, both within Palestine and in the neighboring states, rejected it. Fight-\\ning between the two sides broke out immediately after adoption of the UN\\nresolution and turned into all-out war after May 15, 1948, when the British\\nforces withdrew, the Jewish leadership in Palestine declared the indepen-\\ndent state of Israel, and regular armies from the neighboring Arab states\\njoined the fray. Fighting continued until early 1949.\\nIn July of 1949, Israel and the Arab states signed armistice agreements\\n(though the state of war continued). The armistice lines became the official\\nborders of the State of Israel. These borders included a larger portion of\\nPalestine than the UN partition plan had allotted to the Jewish state. The\\nArab state envisioned by the partition plan did not come into being. T\\nwo\\nparts of mandatory Palestine remained under Arab control: the West Bank,\\nwhich was eventually annexed by Jordan, and the Gaza Strip, which came\\nunder Egyptian administration. The establishment of Israel and the war of\\n1948–49 also created a massive refugee problem, with the flight or expul-\\n188\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nsion of hundreds of thousands of Palestinian Arabs from their homes in the\\npart of Palestine that became the State of Israel. I shall return to the refugee\\nproblem later in this chapter.\\nThe map changed radically as a result of the Arab-Israeli war of June\\n1967—and, along with it, the political atmosphere in the Middle East. By\\nthe end of the Six-Day War, as Israelis called it, Israel occupied the West\\nBank and the Gaza Strip, thus extending its control over the entire territory\\nof mandatory Palestine. It also occupied the Sinai Peninsula and the Golan\\nHeights—Egyptian and Syrian territories, respectively. The new geopoliti-\\ncal and strategic situation created by the 1967 War led to the Palestinianiza-\\ntion of the Arab-Israeli conflict, bringing it back to its origin as a conflict\\nbetween two peoples over—and increasingly within—the land they both\\nclaimed (Kelman, 1988).\\nThe Palestinianization (or re-Palestinianization) of the conflict has mani-\\nfested itself in the actions of the Arab states, of the Palestinian community it-\\nself, and of Israel. Israel’s neighboring Arab states gradually withdrew from\\nthe military struggle against Israel—though not before another major war in\\n1973—leaving it, essentially, to the Palestinians themselves. The disengage-\\nment of the Arab states became dramatically clear with the 1977 visit to\\nJerusalem of President Anwar Sadat of Egypt, the largest and most powerful\\nArab state, an initiative that led to the Camp David accord of 1978 and the\\nEgyptian-Israeli peace treaty of 1979. The Palestinians took repossession of\\ntheir struggle, which in the years between 1949 and 1967 had been mostly in\\nthe hands of the Arab states. Fatah, under the leadership of Yasser Arafat, and\\nother Palestinian guerrilla organizations grew in strength and eventually\\ntook over the Palestine Liberation Organization (PLO), which was originally\\na creature of the Arab League. Between the 1960s and the 1980s, the Pales-\\ntinian movement gradually shifted its emphasis from the liberation of all \\nof Palestine through armed struggle against Israel to the establishment of an\\nindependent Palestinian state in the West Bank and Gaza through largely \\npolitical means. The end of the occupation became the immediate goal of the\\nmovement and, with the onset of the intifada—the uprising in the West Bank\\nand Gaza—in December 1987, the occupied territories became the focal\\npoint of its struggle.\\nOn the Israeli side, the intifada further underlined the Palestinianization\\nof the conflict in the wake of the 1967 War. What had been largely an inter-\\nstate conflict between 1948 and 1967 had now been internalized by Israel—\\nthat is, transformed into a continuous confrontation with a resentful Pales-\\ntinian population, living under occupation within Israel’s post-1967\\nborders. Many Israelis were persuaded by the intifada that continuing occu-\\npation was not tenable and that the Palestinians were indeed a people,\\nwhose national movement had to find some political expression if there was\\nto be a peaceful accommodation between the two sides (Kelman, 1997c).\\nBy the end of the 1980s, there was a strong interest on all sides in finding\\na peaceful accommodation and an increasing recognition that some version\\nThe Role of National Identity in Conflict Resolution\\n189\\n\\n\\nof a two-state solution would provide the best formula for a broadly accept-\\nable historic compromise. The political obstacles to such a solution, how-\\never, remained severe. A number of strategic and micropolitical considera-\\ntions—traceable, in particular, to the end of the Cold War and the\\naftermath of the Gulf War—eventually brought the leaderships on both\\nsides to the negotiating table at the Madrid Conference in 1991 and the\\nsubsequent talks in Washington. These talks, however, never developed\\nmomentum. It was only after Prime Minister Yitzhak Rabin came into\\npower in Israel in 1992, at the head of a government led by the Labor Party,\\nand gradually (and reluctantly) concluded that Israel would have to deal di-\\nrectly with the PLO leadership in order to make progress in the negotia-\\ntions, that a breakthrough was finally achieved. This breakthrough oc-\\ncurred in the secret Oslo talks, which culminated in the exchange of letters\\nof mutual recognition between the PLO and the State of Israel (which, in\\nmy view, was the most significant achievement of the Oslo process) and the\\nDeclaration of Principles signed in Washington in September 1993 (see\\nKelman, 1997c, for further details).\\nThe Oslo breakthrough occurred because a number of long-term and\\nshort-term interests of the parties—to which I alluded briefly—persuaded\\nthem of the necessity of reaching an agreement. “But a significant factor\\ncontributing to the breakthrough was the conclusion, on both sides, that\\nnegotiations were not only necessary but also possible—that they could\\nyield an acceptable agreement without jeopardizing their national existence.\\nThis sense of possibility evolved out of interactions between the two sides\\nthat produced the individuals, the ideas, and the political atmosphere re-\\nquired for negotiations” (Kelman, 1997d, p. 213). A variety of unofficial\\ncontacts between the two sides, including the workshops to be described in\\nthis chapter, contributed to creating this sense of possibility and the climate\\nconducive to negotiations (see Kelman, 1995).\\nThough the Oslo agreement was a genuine breakthrough in the effort to\\nresolve the Israeli-Palestinian conflict, the process of negotiating an interim\\nagreement and initiating final-status negotiations has faced numerous ob-\\nstacles and experienced dangerous setbacks, all reminding us of the in-\\ntractability of conflicts between identity groups, such as the Israeli-Palestin-\\nian conflict. Significant changes in the conflict have taken place with the\\nestablishment of a Palestinian Authority (PA) in the West Bank and Gaza\\nStrip, which constitutes the scaffolding for a Palestinian state. But the path\\nto a negotiated agreement has been obstructed by such events as the He-\\nbron massacre of Palestinians at prayer, the assassination of Yitzhak Rabin,\\nthe suicide bombings against Israeli civilians, the continuation of the Israeli\\nsettlement process in the West Bank, the deteriorating economy and quality\\nof life in the West Bank and Gaza, the corrupt and autocratic leadership of\\nthe PA, and the three-year rule of Prime Minister Netanyahu in Israel.\\nThe election of a new Israeli government in May of 1999 brought new\\nhope that serious negotiations of the final-status issues will be pursued. But\\n190\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nthe negotiations failed and, in late September 2000, a new intifada broke\\nout, producing a violent Israeli response and a serious deterioration of the\\npeace process. Nevertheless I feel fairly confident that the negotiations will\\neventually succeed in establishing a Palestinian state. I am less confident\\nthat this new state will be truly independent, secure, and viable, and that its\\nestablishment will form the basis of a long-term peace and a cooperative,\\nmutually enhancing relationship with Israel, conducive to ultimate recon-\\nciliation between the two peoples. How the identity issues that divide the\\ntwo peoples are addressed in the negotiations and the public debates that\\nsurround them will play a major role in determining the quality and dura-\\nbility of the solution achieved at the negotiating table.\\nThe Struggle over National Identity\\nAlthough national identity is carried by the individual members of a na-\\ntional group and can thus be studied as a property of individuals, the pre-\\nsent chapter refers to it as a collective phenomenon—as a property of the\\ngroup. “Insofar as a group of people have come to see themselves as consti-\\ntuting a unique, identifiable entity, with a claim to continuity over time, to\\nunity across geographical distance, and to the right to various forms of self-\\nexpression, we can say that they have acquired a sense of national identity.\\nNational identity is the group’s definition of itself as a group—its concep-\\ntion of its enduring characteristics and basic values; its strengths and weak-\\nnesses; its hopes and fears; its reputation and conditions of existence; its in-\\nstitutions and traditions; and its past history, current purposes, and future\\nprospects” (Kelman, 1997a, p. 171; see also Kelman, 1998b).\\nIn the Israeli-Palestinian and other such conflicts, the threat to collective\\nidentity is a core issue in the conflict, which is integrally related to the\\nstruggle over territory and resources. Both peoples and their national\\nmovements claim the same territory, and each seeks ownership of that terri-\\ntory and control over its resources as the basis of an independent state that\\ngives political expression to its national identity. The integrity of this col-\\nlective identity is critical to each group for several reasons. First, the in-\\ntegrity of the national identity is an end in itself, in that the identity serves\\nas a source of distinctiveness, unity, and continuity for the group and of a\\nsense of belongingness for its members. Second, the national identity con-\\nstitutes the ultimate justification of the group’s claim to ownership of the\\nland and control of its resources. And third, the national identity provides a\\nfocus for developing and maintaining the group’s distinctive culture, reli-\\ngion, and way of life. The collective identity of each group is bolstered by a\\nnational narrative—an account of the group’s origins, its history, and its re-\\nlationship to the land—that explains and supports its sense of distinctive-\\nness, its positive self-image, and the justice of its claims and grievances.\\nIn conflicts such as that between Israelis and Palestinians, in which the\\nThe Role of National Identity in Conflict Resolution\\n191\\n\\n\\ntwo sides live in the same space and claim ownership of the same territory, it\\nis not only the actions of the other but also the identity and the very exis-\\ntence of the other that are a threat to the group’s own identity. The other’s\\nidentity and its associated narrative challenge the group’s claims to owner-\\nship—at least to exclusive ownership—of the land and its resources. The\\nother’s presence in the same space, particularly if it is accompanied by \\ndemands for a share of the power and for recognition of the other culture,\\nreligion, and/or language, is perceived as a threat to the integrity and cohe-\\nsiveness of the group’s society and its way of life.\\nThese dynamics lead to a view of the conflict as a zero-sum struggle, not\\nonly around territory but also around identity (Kelman, 1987). Acknowl-\\nedging the other’s identity becomes tantamount to jeopardizing the iden-\\ntity—and indeed the national existence—of one’s own group. Each side\\n“holds the view that only one can be a nation: Either we are or they are.\\nThey can acquire national identity and rights only at the expense of our\\nidentity and rights” (Kelman, 1987, p. 354). Thus, over the course of the \\nIsraeli-Palestinian conflict, there has been a systematic tendency on each\\nside to deny the other’s identity as a people, the authenticity of the other’s\\nlinks to the land, the legitimacy of the other’s claims to national rights, and\\nthe very existence of the other as a national group (Kelman, 1978, 1982).\\nNegation of the other’s identity and of the narrative in which it is embedded\\nbecomes so important to the conflict that it is incorporated in the identity\\nthat each group constructs for itself and in the narrative that it presents to\\nthe world (Kelman, 1999).\\nThe contrasting Israeli and Palestinian narratives about the creation of\\nthe State of Israel in 1948 both rely on the negation of the other to bolster\\nthe justice of their own cause. For Israelis, the creation of Israel represented\\na rightful return of the Jewish people to its ancestral homeland. Establish-\\nment of a Jewish state in Palestine did not, in their eyes, constitute an injus-\\ntice to the Arabs who resided there, because Palestinian Arabs were not a\\npeople, distinct from the Arab inhabitants of surrounding countries, and\\nnever exercised sovereignty in Palestine. Moreover, in the Israeli narrative,\\nthe responsibility for the Palestinian refugee problem and the suffering of\\nthe Palestinian Arab population rests with their own aggressive and incom-\\npetent leadership, which rejected all compromise and initiated violent at-\\ntacks in the effort to block the establishment of Israel. For Palestinians, by\\ncontrast, the creation of Israel represented an act of usurpation by Euro-\\npean settlers, who forcefully displaced the indigenous population and de-\\nstroyed their society, their property, and their way of life. In the Palestinian\\nnarrative, Jews are a religious group, not a nation entitled to its own state,\\nand Zionism is a form of settler colonialism that imposed itself on a region\\nin which it has no roots. Each identity gains some of its strength and legiti-\\nmacy from negating and delegitimizing the other.\\nIdentities that rest in part on negation of the other inevitably take on \\nan exclusivist and monolithic character (Kelman, 1997b). In the Israeli-\\n192\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nPalestinian conflict, such exclusivist and monolithic definitions of identity\\nhave begun to give way in recent years. For significant segments of the two\\npopulation, however—and in some respects even for large majorities—such\\ndefinitions still prevail. \\nIn the Israeli-Palestinian case, a defining element of each group’s identity\\nis its relationship to the land and its history. Insofar as this relationship is\\nexclusive—that is, insofar as the group’s identity rests on the view that the\\nland and its history belong to it alone and that the other’s claims on them as\\npart of its own identity are illegitimate and inauthentic—there is little room\\nfor conflict resolution. Conflict resolution becomes an option when the\\nparties accept the possibility that certain elements of identity may be shared\\nwith the other, acknowledging that the other also has a profound attach-\\nment to the land, anchored in authentic historical ties to it. Israelis and\\nPalestinians have been gradually moving toward acceptance of shared ele-\\nments of identity as they have been searching for a political formula for\\nsharing the land. It has proved more difficult for the two sides, so far, to ac-\\ncept Jerusalem as a shared element of the two identities and to develop a\\npolitical formula to reflect that view.\\nIdentities that rest on negation of the other also take on a monolithic\\ncharacter; that is, all dimensions of the group’s identity—such as ethnicity,\\nreligion, and language—tend to be viewed as highly correlated. The ide-\\nology calls for complete correspondence between ethnic boundaries, politi-\\ncal boundaries, boundaries of emotional attachment, and boundaries of in-\\ntensive interaction. Self and other are, in principle, completely separated\\nalong all of these lines. The Israeli-Palestinian conflict, as well as other pro-\\ntracted conflicts, particularly between identity groups living in close prox-\\nimity within a small space (such as Northern Ireland or Cyprus), might be\\nmore amenable to resolution if there were some degree of disaggregation of\\nthe monolithic identity, based on distinctions between different types of\\nboundaries. Such distinctions would allow for the development of a tran-\\nscendent identity—not in place of the particular ethnonational identities, but\\nalongside of them. In the Israeli-Palestinian case, a transcendent identity\\ncould be fostered by separating the concept of the state, as a sovereign po-\\nlitical entity, from that of the country, as a geographical entity. This distinc-\\ntion would allow the two communities to treat the entire country (Eretz\\nYisrael or Palestine) as an object of common sentimental attachment and as\\nthe framework for common instrumental pursuits (in such areas as develop-\\nment and use of water resources, environmental protection, public health,\\nand tourism), while living in and identifying with separate political states\\nwithin that country.\\nThe zero-sum view of identity and the mutual denial of the other’s iden-\\ntity that I have described create serious obstacles to conflict resolution. All\\nissues tend to become existential—matters of life and death for each side.\\nCompromise solutions that involve sharing of the land or agreeing on dif-\\nferent boundaries for different purposes are likely to threaten exclusivist\\nThe Role of National Identity in Conflict Resolution\\n193\\n\\n\\nand monolithic identities. The demonized other is not trusted to negotiate\\nin good faith and respect agreements. In short, when acceptance of the\\nother’s national rights and recognition of the other’s national identity are\\nseen as relinquishing the group’s own rights and jeopardizing its own iden-\\ntity, distributive solutions based on compromise are hard to achieve. Even if\\nthe parties agree to make certain compromises in response to reality de-\\nmands and external pressures, these compromises are unlikely to lead to\\ndurable changes in the relationship between the conflicting groups, con-\\nducive to stable peace, mutually enhancing interaction, and ultimate recon-\\nciliation. Lasting change requires mutual adjustments in collective identity.\\nIdentity Changes\\nThe stubborn resistance to change in collective identities is widely recog-\\nnized and taken for granted. Yet identities have to change, at least tacitly, if\\nprotracted identity conflicts are to be settled and, certainly, if they are to be\\nresolved in a way that transforms the relationship and opens the way to rec-\\nonciliation. South Africa provides perhaps the best illustration of an arena\\nof intense, protracted conflict in which fundamental identity changes paved\\nthe way to resolution and reconciliation, although it also illustrates the dif-\\nficulties in changing the worldviews and the structural realities that became\\nentrenched in the apartheid era.\\nDespite their undeniable rigidities, identities are potentially changeable\\n(and indeed negotiable) for two reasons: First, unlike territory and resources,\\nthey are not inherently zero-sum; though they are perceived and debated as\\nsuch in intense conflicts, it is in fact not the case that A’s identity can be rec-\\nognized and expressed only if B’s identity is denied and suppressed. \\nIf the two identities are to become compatible, however, they have to be\\nredefined. And this points to the second reason for the potential change-\\nability of group identities: they can be redefined because they are to a large\\nextent constructed. T\\no view national identity as a social construction does\\nnot imply that it is manufactured out of nothing. There may be cases in\\nwhich one can properly speak of an imagined past, invented to buttress a\\nnewly formed identity (cf. Anderson, 1983). Generally, however, the social\\nconstruction of an identity draws on a variety of authentic elements held in\\ncommon within a group: a common history, language, or religion; common\\ncustoms, cultural expressions, experiences, values, grievances, aspirations\\n(Kelman 1997b). Typically, the social construction of an identity involves a\\ndual process of discovery (or rediscovery) and creation of such common ele-\\nments (Kelman, 1997a). The social construction of the identity implies a\\ndegree of arbitrariness and flexibility in the way the identity is composed\\n(which elements are admitted into it and which omitted from it), and in\\nwhat its boundaries are (who is included and who is excluded). These\\nchoices depend on the opportunities and necessities perceived by the elites\\n194\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nthat are engaged in mobilizing ethnonational consciousness for their politi-\\ncal, economic, or religious purposes (Kelman, 1997b). Serbs and Croats, for\\nexample, share a common language and culture, but differ in religion and\\nhistorical experiences. Political leaders have at times focused on the simi-\\nlarities in the effort to shape them into a single nation; at other times they\\nhave magnified the differences to define them as separate—and mutually\\nantagonistic—nations.\\nThus, although national identities are generally constructed out of real\\nexperiences, these experiences can be ordered in different ways, resulting in\\ndifferent boundaries and priorities. As a consequence, they can be—and\\ntypically are—deconstructed and reconstructed. “In fact, the reconstruction\\nof identity is a regular, ongoing process in the life of any national group.\\nIdentities are commonly reconstructed, sometimes gradually and some-\\ntimes radically, as historical circumstances change, crises emerge, opportu-\\nnities present themselves, or new elites come to the fore” (Kelman, 1997b,\\np. 338). Clearly, therefore, there is room for maneuver in a group’s self-\\ndefinition, particularly with respect to the definition of group boundaries\\nand the priorities among different elements of the group’s identity.\\nChanges in identity over the course of a protracted conflict come about\\nthrough a combination of changed perceptions of the necessity and the possi-\\nbility of resolving a conflict that has become increasingly costly to the par-\\nties. The mounting costs and dwindling prospects of governing Algeria for\\nthe French, of maintaining apartheid for white South Africans, of the occu-\\npation of Palestinian territories for Israelis, and of the armed struggle for\\nthe Palestinians created the necessity for changes in identity: Algeria as an in-\\ntegral part of France, South Africa under exclusive white control, Israel\\nwithin the borders of Greater Israel, and Palestinian repossession of the en-\\ntire homeland were assigned lower priority in the national identities of\\nthese groups as it became clear to a majority that these aspirations could not\\nbe realized at an acceptable cost. \\nWhat made it possible to change these priorities was often the discovery\\nthat accommodation of the other’s identity need not destroy the core of the\\ngroup’s own identity, and that a compromise solution to the conflict was\\ntherefore negotiable. This kind of learning can take place in the course of\\nofficial or unofficial interactions between the groups or their members, in-\\ncluding the problem-solving workshops that my colleagues and I have con-\\nducted: In the course of Israeli-Palestinian workshops, for example, partici-\\npants have learned to differentiate their image of the enemy by discovering\\nthat there are potential negotiating partners on the other side, that there is\\na distinction between the other’s ideological dreams and operational pro-\\ngrams, and that the other has positive goals beyond destruction of their\\ngroup (Kelman, 1987). They were enabled to enter into the enemy’s per-\\nspective, thus discovering the historical sources of the other’s claims and\\ngrievances, the depth of the other’s fears, and the authenticity of the other’s\\nsense of peoplehood. They began to visualize a different future, discovering\\nThe Role of National Identity in Conflict Resolution\\n195\\n\\n\\npossibilities for mutually beneficial coexistence and cooperation. As such\\nexperiences multiply, and as the learnings produced by them are infused\\ninto the two political cultures, each group may gradually change its identity\\nby eliminating the negation of the other’s identity as an element of its own\\nidentity and perhaps even admitting the possibility of a partnership as a new\\nelement of its own identity.\\nNegotiating Identity\\nThe changes I have described are often the result of an explicit or implicit\\nprocess of negotiating identity. At its core, national identity is clearly non-\\nnegotiable; indeed, the very idea of negotiating identity sounds like an oxy-\\nmoron. National identity is a collective psychological conception, which\\ncannot be dictated or prescribed by outsiders. A group of people who define\\nthemselves as a nation cannot be told that they have no right to do so be-\\ncause their self-definition does not conform to some set of theoretical, \\njuridical, or historical criteria for doing so, or because their nationhood is\\ninconvenient to others. Nor does it make sense to tell them how to draw the\\nboundaries of the group: whom to include and whom to exclude. People are\\na nation if they perceive themselves as such and are prepared to invest \\nenergy and make sacrifices in terms of that perception (Kelman, 1978).\\nNeither Palestinians nor Israelis will give up the core of their identity: their\\nsense of peoplehood, their attachment to the land, their conviction about\\nthe historical authenticity of their links to that land, their commitment to\\ntheir national culture, language, and way of life. Nor will they give up the\\nnational narrative that substantiates the justice of their cause. \\nBut there are many elements that can be added to or subtracted from an\\nidentity without jeopardizing its core. In fact, changes in less central ele-\\nments of the identity are often advocated precisely in order to protect the\\ncore of the identity. It was on that basis that the majority of Israelis and\\nPalestinians came to accept territorial compromise—that is, a shrinking of\\nthe territorial dimension of their identity—as the best available option for\\nmaintaining their national identity. The Peace Now movement in Israel, for\\nexample, advocated withdrawal from the occupied territories largely on the\\ngrounds that this was the only way in which Israel could maintain its char-\\nacter as both a Jewish state and a democratic state. Yehoshafat Harkabi\\n(1986), a former chief of Israeli military intelligence and a prophetic voice\\nin the debate about Israeli-Palestinian peace, explicitly advocated a “smaller\\nIsrael”—a “Zionism of quality” rather than a “Zionism of acreage.” He \\nargued that Israel had to choose between withdrawing from the West \\nBank and making way for a Palestinian state there, or annexing the West\\nBank with the consequence that Israel would eventually become a Palestinian\\nstate. On the Palestinian side, the territorial dimension of the Palestinian\\nidentity has gradually changed as the movement reflected on its realistic \\n196\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\noptions. The thinking of the PLO evolved from advocacy of a Palestinian\\nArab state in the whole of Palestine, to a secular democratic state, and even-\\ntually to a Palestinian state alongside of Israel comprising the West Bank\\nand Gaza (cf. Muslih, 1990). Significant segments of both societies still re-\\nject territorial compromise on religious or ideological grounds and link\\ntheir national identity to possession of the land in its entirety. But the Pales-\\ntinian and Israeli mainstreams have by now come to terms with a national\\nidentity that finds its political expression in only part of the land, as evi-\\ndenced by the opinion polls that are now conducted on a regular basis in\\nboth societies.\\nSuch changes in elements of identity are a legitimate subject for “nego-\\ntiation” between groups whose identities clash, because the identity that\\none group chooses for itself has significant implications for the rights, in-\\nterests, and identity of the other. Whenever one group translates the self-\\ndefinition of its nationhood into action—“by making territorial claims, by\\ndemanding an independent state, by seeking to redraw borders, by declar-\\ning who is included in the national identity and who is excluded from it, or\\neven by selecting a name for itself” (Kelman, 1997b, p. 337)—the other is\\ninevitably affected. Each group, therefore, has a legitimate concern about\\nthe way the other defines itself, the way it formulates its national identity. It\\nis not surprising, then, that identity issues play an important role in the for-\\nmal and informal processes of pre-negotiation and negotiation—that is, in\\nthe efforts to open a way to the negotiating table and to reach agreement\\naround the table.\\nT\\no some extent, identity issues are part of the subject matter of the offi-\\ncial negotiations. I have already referred to the territorial dimension of\\nidentity: insofar as Israelis and Palestinians are negotiating on the basis of a\\n“land for peace” formula, they are accepting territorial limits to their na-\\ntional identities, which have, after all, been historically linked to the whole\\nof the land. Similarly, the mutual recognition between Israel and the PLO,\\nas expressed in the exchange of letters between Arafat and Rabin—which I\\nhave described above as the most important breakthrough of the Oslo\\nagreement —can be viewed as a product of the negotiation of identity: an\\nact of acceptance and legitimization of the other who in the past had been\\ndefined as the antithesis to the self.\\nAlthough redefined identities are thus promulgated around the official\\nnegotiating table, the “negotiation” of identity is primarily an informal, un-\\nofficial process in which members of the conflicting parties explore and \\ninvent ways of accommodating their group identities to one another. The\\npurpose of negotiation in this looser sense of the term is not to produce \\npolitical agreements, but to develop joint understandings and formulations\\nthat can help pave the road to political agreements at the official level. Im-\\nplicitly and explicitly, this kind of “negotiation” has been a central focus for\\nproblem-solving workshops between Israelis and Palestinians that my col-\\nleagues and I have conducted over the past quarter-century.\\nThe Role of National Identity in Conflict Resolution\\n197\\n\\n\\nProblem-Solving Workshops\\nProblem-solving workshops are the central instrument of interactive problem\\nsolving, an unofficial, third-party approach to the resolution of international\\nand intercommunal conflicts, derived from the pioneering work of John\\nBurton (1969, 1979, 1984) and anchored in social-psychological principles\\n(see Kelman, 1972, 1979, 1992, 1998a). A workshop is a specially con-\\nstructed, private space in which politically involved and often politically in-\\nfluential (but generally unofficial) members of conflicting communities can\\ninteract in a nonbinding, confidential way. The microprocess of the work-\\nshop provides them the opportunity to penetrate each other’s perspective;\\nto explore both sides’ needs, fears, priorities, and constraints; and to engage\\nin joint thinking about solutions to the conflict that would be responsive to\\nthe fundamental concerns of both sides.\\nLet me describe a typical “one-time” workshop for Israeli and Palestin-\\nian participants. By “one-time” I mean that this particular group of Israelis\\nand Palestinians convenes only for this one occasion. Some of the indi-\\nviduals may have participated in more than one such workshop, and the \\none-time workshops that we have held over the years have had a cumulative\\neffect within the two societies. But, until 1990, we made no effort to recon-\\nvene the same group of participants for another occasion. The workshops\\ntake place under academic auspices and are facilitated by a panel of social\\nscientists knowledgeable about international conflict, group process, and\\nthe Middle East. The workshops usually begin with two pre-workshop \\nsessions, about four hours in length, during which the third party meets\\nseparately with each of the two parties. The workshop itself typically lasts\\nabout two and a half days, often scheduled over an extended weekend. The\\nparticipants include three to six Israelis and an equal number of Palestin-\\nians, plus a third party of three or more members.\\nThe Israeli and Palestinian participants have included parliamentarians,\\nleaders and activists of political parties or political movements, journalists,\\neditors, directors of think tanks, and politically involved academicians—that\\nis, scholars who not only publish academic papers but who also write for\\nnewspapers and appear in the media, who serve as advisors to political lead-\\ners, and some of whom move back and forth between government and\\nacademia. Some of our participants have been former diplomats, officials,\\nor military officers, and many were later to become negotiators, ambas-\\nsadors, cabinet ministers, parliamentarians, and leading figures in the media\\nand research organizations. We look for people who are within the main-\\nstream of their societies and close to the center of the political spectrum. At\\nthe same time, they have to be people who are at least willing to explore the\\npossibility of a negotiated solution and to sit down as equals with members\\nof the other party.\\nThe central ground rule of problem-solving workshops is the principle\\nof privacy and confidentiality. In the early days of our work, confidentiality\\n198\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nwas particularly important for the protection of our participants, because\\nthe mere fact that they were meeting with the enemy was controversial and\\nexposed them to political and even physical risks. Confidentiality is equally\\nimportant, however, for the protection of the process that we are trying to\\npromote in workshops. We are trying to encourage the participants to talk\\nand listen to each other, rather than focus on their constituencies, on third\\nparties, or on the record. We want them to think out loud, to experiment\\nwith ideas, to explore different options, without having to worry about how\\nothers would react if their words in the group were quoted outside. We\\nwant them to engage in a type of interaction that is generally not feasible\\namong parties engaged in a bitter conflict—a type of interaction that, in-\\ndeed, deviates from the conflict norms that usually govern their behavior:\\nan interaction that is analytic rather than polemical, one in which the par-\\nties seek to explore each other’s perspective and gain insight into the causes\\nand dynamics of the conflict; an interaction that is problem-solving rather\\nthan adversarial, one in which the parties sidestep the usual attempt to allo-\\ncate blame and, instead, take the conflict as a shared problem that requires\\njoint effort to find a mutually satisfactory solution.\\nThe agenda of a problem-solving workshop is designed to allow this kind\\nof interaction to unfold. The core agenda of a one-time workshop has four\\ncomponents. First, each side is asked to discuss its central concerns in the\\nconflict—the fundamental needs that would have to be addressed and the\\nexistential fears that would have to be allayed if a solution is to be satisfac-\\ntory to it. The parties are asked not to debate the issues raised, although\\nthey may ask for clarification of what the other says. The purpose is for\\neach side to gain an adequate understanding of the other’s needs, fears, and\\nconcerns, from the perspective of the other. Once they have demonstrated\\nthat they understand the other’s needs to a significant degree, we move to\\nthe second phase of the agenda: joint thinking about possible solutions.\\nWhat participants are asked to do in this phase is to develop, through an in-\\nteractive process, ideas about the overall shape of a solution for the conflict\\nas a whole, or perhaps, a particular issue in the conflict, that would address\\nthe needs and fears of both sides. They are given the difficult assignment of\\nthinking of solutions that would meet not only their own side’s needs, but\\nthe needs of both sides.\\nOnce the participants have developed some common ground in this\\nprocess of joint thinking, we turn to the third phase of the workshop: dis-\\ncussion of the political and psychological constraints within the two soci-\\neties that would create barriers to carrying out the ideas for solution that\\nhave been developed in the group. We deliberately leave the discussion of\\nconstraints to the third phase, so that it does not hamper the creative\\nprocess of jointly generating new ideas. Finally, depending on how much\\nprogress has been made and how much time is left, we ask the parties to en-\\ngage in another round of joint thinking—this time about ways of overcom-\\ning the constraints that have been presented. The participants are asked to\\nThe Role of National Identity in Conflict Resolution\\n199\\n\\n\\ncome up with ideas about what their governments, their societies, and they\\nthemselves might do—separately or jointly—that would help to overcome\\nthe barriers to negotiating mutually satisfactory solutions to the conflict.\\nThe third party in our model enacts a strictly facilitative role. It does not\\npropose solutions, nor does it participate in the substantive discussions. Its\\ntask is to create the conditions that allow ideas for resolving the conflict to\\nemerge out of the interaction between the parties themselves. The facilita-\\ntion of the third party, however, is an important part of the process. The\\nthird party sets the ground rules and monitors adherence to them; it helps\\nto keep the discussion moving in constructive directions, tries to stimulate\\nmovement, and intervenes as relevant with questions, observations, and\\neven challenges. It also serves as a repository of trust for parties who, by de-\\nfinition, do not trust each other: They feel safe to come to the workshop\\nbecause they trust the third party to maintain confidentiality and to protect\\ntheir interests.\\nWorkshops have a dual purpose: to produce changes, in the form of new\\ninsights and ideas, in the individual participants; and to transfer these\\nchanges into the political process at the levels of both public opinion and\\ndecision making. These two purposes may at times create contradictory re-\\nquirements—most notably in the selection of participants (Kelman, 1992).\\nT\\no balance these contradictory requirements, we look for participants who\\nare not officials, but politically influential. They are thus more free to en-\\ngage in the process but, at the same time, because of their positions within\\ntheir societies, any new ideas that they develop in the course of a workshop\\ncan have an impact on the thinking of decision makers and the society at\\nlarge.\\nThe Israeli-Palestinian workshops we carried out until 1990 were all\\none-time events designed to create a climate conducive to movement to the\\nnegotiating table. In 1990, we organized our first continuing workshop, in\\nwhich a group of high-level Israelis and Palestinians met periodically over a\\nthree-year period (Rouhana & Kelman, 1994). In 1994, Nadim Rouhana\\nand I convened a Joint Working Group on Israeli-Palestinian Relations,\\nwhich is still in progress.2 For the first time in our program, this group has\\nbeen engaged in producing jointly authored concept papers on some of the\\nfinal-status issues in the Israeli-Palestinian negotiations and on the future\\nrelationship between the two societies. One of three papers that the group\\nhas published so far—a paper on the problem of Palestinian refugees and\\nthe right of return (Alpher & Shikaki, et al., 1999)—provides the starting\\npoint for my detailed illustration of the negotiation of identity, to be pre-\\nsented in the next section.\\nMuch of the discussion in our workshops, from the beginning and until\\nthis day, has focused, in effect, on a process of negotiating collective identi-\\nties. In our experience, such a process can be productive only if it is based\\non mutual respect for the core of the other’s identity and on the principle of\\nreciprocity. Each side must know that the other does not seek to undermine\\n200\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nits group identity, and each must take care not to undermine the other’s\\nidentity. And each must know that the risks it takes in acknowledging the\\nother’s claims, rights, and authenticity will be reciprocated by the other’s ac-\\nknowledgment of its claims, rights, and authenticity.\\nStarting from the understanding that neither side is prepared to negoti-\\nate the core of its identity—its peoplehood, its relationship to the land, the\\nbasic justice of its cause—or the general lines of its national narrative, there\\nremain various elements of each group’s identity that can be “negotiated” in\\nthe interest of mutual accommodation. Let me illustrate some of the possi-\\nble changes in identity that can be and have been discussed in problem-\\nsolving workshops and similar encounters and that have, over time, begun\\nto penetrate the Israeli and Palestinian political cultures.\\n1. Many members of both communities have become able to remove the\\nnegation of the other’s identity as an integral part of their own identity.\\nThough the other may still be seen as an obstacle to achieving one’s own\\nnational goals, the other is not as often seen as the antithesis of one’s own\\nidentity whose demise is a condition for one’s own survival. Thus, many \\nIsraelis have come to accept the reality of Palestinian peoplehood, particu-\\nlarly after observing Palestinians’ readiness to make sacrifices for their na-\\ntional cause during the intifada and Palestinians’ celebration of the signing\\nof the Oslo agreement in September 1993. Interestingly, Israelis saw paral-\\nlels between these events and their own struggle for statehood and celebra-\\ntion at attaining it—a significant degree of identification with the other\\nwhose existence had previously been denied. Many Palestinians, on their\\npart, now recognize the right of Israelis to their state, on the grounds that\\nthe state has existed for over half a century and that its dismantlement\\nwould create a new injustice to the generations that were born into it. Very\\nfew Palestinians, on the other hand, are prepared to acknowledge the his-\\ntorical links of Jews to the land, which might be seen as justification for the\\nestablishment of the Jewish state in the first place.\\n2. We have seen signs of softening of the exclusiveness of group identity,\\nwhich allows for the recognition that—despite the validity of one’s own\\nclaims—the other too has valid claims. The recognition of elements of\\nidentity shared with the other, such as a common identification with the\\nland, opens the way to political solutions based on sharing territory and re-\\nsources. In a recent workshop, for example, mainstream Israelis and Pales-\\ntinians were able to agree—much to everyone’s surprise, including their\\nown—on a formula for sharing Jerusalem: a united city containing the capi-\\ntals of both states. Public opinion polls on both sides also suggest greater\\nwillingness to share Jerusalem than had been widely assumed (Segal, 1999).\\n3. Workshop participants have experimented with disaggregating the\\nmonolithic nature of their identities, recognizing that there are different\\nboundaries of group identity (such as ethnic boundaries, political bound-\\naries, boundaries of sentimental attachment) that do not necessarily coin-\\ncide. This recognition opens the way to the development of transcendent\\nThe Role of National Identity in Conflict Resolution\\n201\\n\\n\\nidentities, which might allow the two peoples to maintain a common at-\\ntachment to the country while “owning” only part of that country as their\\npolitical state. The concept of a “united country with divided sovereignties”\\nwas discussed in one of our workshops in the early 1980s. In a more recent\\nworkshop, the idea of establishing different kinds of boundaries was ex-\\nplored in the attempt to find solutions to the problem of Israeli settlements\\nin the areas in which Palestinians hope to establish their state.\\n4. Workshop discussions can help to identify outdated elements of\\ngroup identity, which refer to maximalist goals and dreams of glory or self-\\naggrandizing images that have no current political relevance but poison the\\nclimate for conflict resolution. Examples here might be Palestinian refer-\\nences to the armed struggle as the way to eliminate the Zionist entity, or \\nIsraeli references to the Zionist project of making the desert bloom. Work-\\nshops have often sensitized participants to words and images that humiliate\\nand frighten the other and could be discarded with minimal cost to group\\nidentity.\\n5. In the course of workshop discussions, participants may decide to re-\\norder the priorities within their national identities, such that certain ele-\\nments (e.g., territorial ambitions), which may not have been given up but\\nhave become too costly to pursue, are relegated to low priority and thus be-\\ncome available for negotiated compromise. Thus, over time, Palestinians\\n(in our workshops and in the larger society) decided to give priority to end-\\ning occupation and establishing a Palestinian state over recovering the lost\\nland in its entirety. Israelis gave priority to maintaining the Jewish character\\nof Israel over controlling the whole of the land.\\n6. Finally, workshop participants may “negotiate” changes in national\\nnarratives that accommodate the other’s view of history as much as possible,\\nsuch as accepting a share of the responsibility for the course of the conflict.\\nThe above-mentioned concept paper of our current Joint Working Group\\non Israeli-Palestinian Relations on “The Palestinian refugee problem and\\nthe right of return” (Alpher & Shikaki, et al., 1999) provides a good illustra-\\ntion of such an effort to negotiate identity. I shall conclude this chapter with\\na more detailed discussion of this concept paper, trying to show what made\\nit possible to achieve common ground in this effort, what still remains to be\\ndone, and possible directions for achieving further progress.\\nThe Problem of Palestinian Refugees\\nFrom the point of view of the struggle over national identity, the problem\\nof Palestinian refugees is probably the most important and the most diffi-\\ncult issue to resolve in the Israeli-Palestinian conflict. It may ultimately be\\nmore important and more difficult than the issue of Jerusalem, because it\\nbears directly on the contrasting views of the justice or injustice of the es-\\ntablishment of Israel presented by the two national narratives.\\n202\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nThe problem of Palestinian refugees goes back to the 1948 war that \\nfollowed the declaration of an independent State of Israel. In the course \\nof that war, somewhere between 600,000 and 760,000 Palestinian Arabs—\\nabout half of the Arab population of mandatory Palestine—were displaced\\nfrom their homes in what became the State of Israel (Morris, 1987, \\npp. 297–298). They were turned into refugees and most of them and/or\\ntheir descendants remain refugees to this day. The precise number of Pales-\\ntinian refugees today is in dispute. The figure given by UNRWA—the\\nUnited Nations Relief and Works Agency, which is charged with the wel-\\nfare of Palestinian refugees—is almost three and a half million (as of Janu-\\nary 1998). This number includes Palestinians who registered as refugees\\nwith UNRWA after losing their homes in 1948, as well as their descen-\\ndants. The majority of this population lives in refugee camps in Gaza, the\\nWest Bank, Jordan, Lebanon, and Syria. Palestinian refugees in Lebanon\\n(some 350,000) and Syria (some 300,000) by and large remain stateless, as\\ndo many Palestinians living elsewhere in the Arab world. \\nPalestinians and Israelis have differed sharply in their account of the\\nPalestinian exodus in 1948. According to the Palestinian narrative, “the\\nArab refugees were forcibly expelled by Jewish forces or left in a panic flight\\nto escape massacre and . . . they were helped on their way by occasional\\nmassacres, committed by Jewish forces to keep them running” (Alpher &\\nShikaki, et al., 1999, p. 173). According to some Israeli accounts, the Arab\\npopulation was urged by its own leaders to leave their homes in anticipation\\nof an early return after the defeat of the Jewish forces by the Arab armies.\\nThe more recent writings of Israeli “revisionist historians” (e.g., Morris,\\n1987) support the claim that there were indeed some systematic efforts by\\nJewish forces to expel the Arab population at various times and places. No\\ndoubt, a combination of circumstances created refugees in a war that was\\nfought in the midst of the country’s towns and villages.\\nHow the refugee problem is ultimately resolved raises major practical\\nand symbolic issues for both sides that reach to the core of their respective\\nidentities. By practical issues, I refer to those relating to the implementation\\nof any agreement on the refugees—that is, to the concrete steps that will be\\ntaken to deal with the plight of the current refugee population. By symbolic\\nissues, I refer to those relating to the principle of the right of return of the\\nrefugees to their original homes.\\nAt the practical level, Palestinians are concerned with normalizing the\\nstatus of the refugees to the satisfaction of the individual Palestinians in-\\nvolved and of the entire Palestinian community. Normalization of the status\\nof individual refugees could take any one or combination of the following\\nforms: return to Israel (perhaps their original homes, where that is still fea-\\nsible), return to the Palestinian state, financial compensation, resettlement\\nin a country outside of Palestine, integration with citizenship in the state in\\nwhich they are now living, or Palestinian citizenship and permanent status\\nin the state in which they are now living. The options offered to the\\nThe Role of National Identity in Conflict Resolution\\n203\\n\\n\\nrefugees must be such as to satisfy the entire Palestinian community that\\nthe final peace agreement was not achieved at the expense of any part of\\nthat community.\\nAt the symbolic level, Palestinians are asking for an Israeli acknowledg-\\nment of the right of return as a matter of principle, even if they are prepared\\nto compromise on the implementation of that right. From the Palestinian\\nperspective, Israeli acknowledgment of the right of return presupposes ad-\\nmission of direct moral responsibility for the plight of the refugees. Such an\\nacknowledgment and admission would confirm the Palestinian narrative\\nabout the consequences for the Palestinian people of the Zionist enterprise,\\nof Jewish settlement in Palestine, and of the establishment of Israel. In \\nconfirming the Palestinian narrative, such an acknowledgment would also\\nrecognize the Palestinian national rights that are anchored in this narrative.\\nThere seem to be two interrelated, but separable, reasons for Palestinians’\\nemphasis on this symbolic issue. Psychologically, they need an Israeli ac-\\nknowledgment that Palestinians have a right to return to the homes from\\nwhich they were wrongfully displaced, in order to satisfy their sense of justice\\nand be able to let go of the conflict. Politically, they need this acknowledg-\\nment to support the validity of their claims (such as claims for compensation)\\nin the bargaining that lies ahead.\\nFor Israelis, too, major practical and symbolic issues are at stake in the\\nresolution of the refugee problem. At the practical level, Israelis are pri-\\nmarily concerned with the impact of a return of Palestinians in large num-\\nbers to Israel proper (i.e., to Israel within its pre-1967 borders). A large-\\nscale return of refugees (beyond a symbolic number of up to 100,000)\\nwould be disruptive to the society because it would raise many divisive\\nquestions about the ownership of homes, villages, and other properties that\\nhave changed hands or been destroyed after the 1948 events. T\\no what pre-\\ncise destination would refugees return whose former houses are now in-\\nhabited by Jewish families or whose homes or villages no longer exist? Fur-\\nthermore, and most centrally, Israelis are concerned about the impact of a\\nlarge-scale return of Palestinians to Israel on the demographic balance and\\nthe Jewish character of the state. In short, they see a large-scale return of\\nrefugees to Israel proper as creating an existential threat to Israel by under-\\nmining the Jewish majority and the viability and stability of the state. There\\nis also concern among Israelis about a mass return of refugees to the future\\nPalestinian state because of the impact on the stability and economy of that\\nstate, but this concern is shared by the Palestinians themselves, who are at\\nleast as mindful as Israelis about the absorptive capacity of the Palestinian\\nstate.\\nAt the symbolic level, Israelis have serious concerns about acknowledg-\\ning the principle of a Palestinian right of return, even if Palestinians agree\\nthat this right will not be implemented through a large-scale return of\\nrefugees to Israel proper. Acknowledgment of that principle is particularly\\ntroubling, from the Israeli perspective, if it entails admission of moral re-\\n204\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nsponsibility for the plight of the refugees. Such an acknowledgment and ad-\\nmission would undermine the Israeli narrative about the rightful return of\\nthe Jewish people to its ancestral homeland, the establishment of the State\\nof Israel, the Arab rejection of partition and attack on the new state, and the\\nresulting refugee problem. Again, there seem to be two reasons for Israelis’\\nstance on this symbolic issue and, in particular, for their reluctance to ac-\\ncept direct moral responsibility for the plight of the refugees. Psychologi-\\ncally, they are not willing to accept a national self-image that negates the\\nrighteousness of their cause and implies that Israel was “born in sin.” Po-\\nlitically, they are concerned that accepting the principle of the right of \\nreturn and moral responsibility for the refugee problem would keep the \\nissue open, even after a peace agreement has been signed, and make Israel\\nvulnerable to future claims. \\nOur Joint Working Group on Israeli-Palestinian Relations discussed, in\\nconsiderable detail, the divergent narratives of the two communities; the\\nfundamental needs, fears, and concerns of each community that would have\\nto be addressed if a solution of the problem of Palestinian refugees is to be\\nacceptable to each; and the general principles that must underlie such a so-\\nlution. The joint concept paper that finally emerged from the group’s dis-\\ncussion (Alpher & Shikaki, et al., 1999) starts with a review of the two nar-\\nratives, of the basic needs of each side, and of the principles that must\\ngovern a solution. It then proceeds to present four possible solutions to the\\nrefugee/right of return issue. Solutions 1 and 2 reflect the traditional posi-\\ntions—conceivably, the opening negotiating positions—of the Palestinian\\nand Israeli sides, respectively. Solutions 3 and 4 are compromise solutions,\\nfrom a Palestinian and an Israeli standpoint, respectively. These two solu-\\ntions represent serious efforts, on the part of each side—based on intensive\\ndiscussions within the group—to come up with a compromise that would\\naccommodate the concerns of the other side without jeopardizing its own\\nvital interests. The paper discusses the advantages and disadvantages of\\neach solution, highlights the areas of agreement between solutions 3 and 4\\nas well as the remaining gaps between the two sides, and points to possible\\nways in which these gaps might be bridged.\\nSolutions 3 and 4 represent a considerable narrowing of the gap between\\nthe two sides when compared to the opening positions of solutions 1 and 2.\\nStill, there remains a significant gap between the two positions, particularly\\nwith respect to the symbolic issues that divide the two sides. Solution 3 calls\\nfor an Israeli acknowledgment of the moral right of Palestinians to return\\nto their homes based on Israel’s acceptance of responsibility for creating the\\nrefugee problem (although it recognizes that the actual exercise of the right\\nof return would be limited). Solution 4, the Israeli compromise proposal,\\nacknowledges that Israel shares practical (but not moral) responsibility for\\nthe plight of the refugees, and accepts the right of return to the Palestinian\\nstate, but not to Israel proper (although it agrees to the return of a limited\\nnumber of refugees to Israel). \\nThe Role of National Identity in Conflict Resolution\\n205\\n\\n\\nThe third party tried very hard to encourage the participants to bridge\\nthe remaining gaps between solutions 3 and 4 and try to come up with a\\njoint solution 5—a compromise of the compromises, as it were. We did not\\nsucceed in this effort, however. In part, further movement toward a joint\\nposition on the symbolic issues was blocked by the great difficulty of the as-\\nsignment. We were dealing here with issues that reached to the core of each\\ngroup’s identity and associated narrative. For Israelis to acknowledge any-\\nthing more than shared practical responsibility for the refugee problem and\\nfor Palestinians to accept anything less than an Israeli acknowledgment of\\nmoral responsibility would undermine their respective narratives, and most\\nmembers of the group were not prepared to go that far. Furthermore, at\\nleast some of the participants may have preferred to leave the final bridging\\nof the gap between the two sides to the official negotiators. Since that final\\nbridging would inevitably entail large concessions on one or both sides,\\nthey may have felt that it could best be done in the context of an overall\\nagreement, where concessions on the refugee issue could be made with ap-\\npropriate trade-offs on other issues.\\nEven though the Joint Working Group did not come up with a joint so-\\nlution to the refugee problem, the concept paper that it produced repre-\\nsents a significant contribution to the final negotiation of this issue. The\\npublication of the paper is in itself significant, in that it is the joint product\\nof a group of politically mainstream and influential Israelis and Palestinians,\\nwho worked on it over the course of two years and agreed to make it public.\\nIt is also significant that the Israeli and Palestinian authors of the paper ex-\\nplicitly premise their discussion on the assumption of a two-state solution,\\nadvocating the establishment of a viable, independent Palestinian state. In\\nthe paper, each group recognizes the other’s narrative, without abandoning\\nits own. The different options for resolving the refugee problem are evalu-\\nated in terms of their success in addressing the fundamental needs and fears\\nof each side. What emerged from this effort is a considerable narrowing of\\nthe gap between the two sides, even though the paper does not entirely\\nbridge that gap. Solutions 3 and 4 represent substantial movement from the\\ninitial positions reflected in solutions 1 and 2—movement based on each\\nside’s sincere attempt to accommodate the basic concerns of the other.\\nWith respect to the practical issues—that is, the actual implementation\\nof the right of return—the gap between the two sides was substantially nar-\\nrowed in the Working Group’s paper. “Solutions 3 and 4 concur that, at the\\npractical level, the core of the solution involves four components: a ‘return,’\\nhowever defined, of a limited number of refugees to Israel proper; return of\\na larger number to the Palestinian state; permanent absorption in host\\ncountries (notably Jordan); and compensation. Both sides agree that a suc-\\ncessfully negotiated solution should close the file on the refugee issue”\\n(Alpher & Shikaki, et al., 1999, p. 184). The two solutions disagree about\\nsome important details. For example, the Israelis frame the limited return\\nof refugees to Israel proper as part of a program of family reunification,\\n206\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nwhereas the Palestinians frame it as part of the exercise of the right of re-\\nturn. With regard to compensation, solution 3 calls for both collective and\\nindividual compensation for the refugees’ losses, whereas solution 4 calls\\nfor collective compensation only. Furthermore, solution 4 stipulates a simi-\\nlar mechanism of collective compensation for Jewish refugees from Arab\\ncountries (albeit without an operational link to the compensation of Pales-\\ntinian refugees), whereas solution 3 eschews any such linkage. These differ-\\nences clearly require further negotiation. But what is important is that the\\ntwo solutions agree on the crucial practical issues of implementing the right\\nof return: Palestinians concede that there will be no large-scale return of\\nrefugees to Israel proper; and Israelis accept the refugees’ right of return \\nto the Palestinian state, agreeing to leave it up to Palestinian authorities to\\ncontrol the inflow of refugees in accord with the state’s absorptive capacity.\\nWith respect to the symbolic issues—those relating to the principle \\nof the right of return—the concept paper represents significant movement\\ntoward a mutually acceptable formulation, although there is still a wide gap\\nbetween the two positions. Solution 4, the Israeli compromise statement,\\nacknowledges that Israel “shares practical (but not moral) responsibility, to-\\ngether with the other parties to the process that culminated in the 1948 war,\\nfor the plight and suffering of the refugees, and that rectification of that\\nplight by all parties is a central goal of the Arab-Israeli peace process”\\n(Alpher & Shikaki, et al., 1999, p. 181). This statement comes closer than\\npast statements by mainstream Israelis to acknowledging the suffering of\\nthe Palestinian refugees and accepting a share of the responsibility for it,\\nand thus beginning to address the Palestinians’ sense of justice. The move-\\nment reflected in this statement was generated by an intensive process of\\nnegotiating identity within the Working Group. Though it represents sig-\\nnificant movement, this acknowledgment is not sufficient to meet the Pales-\\ntinians’ call for full Israeli acknowledgment of the principle of the right of\\nreturn, based on acceptance of direct moral responsibility for the plight of\\nthe refugees. Solutions 3 and 4, thus, still remain far apart at the symbolic\\nlevel. This remaining gap is understandable in that the two solutions,\\nthough they represent serious efforts at compromise, have their point of de-\\nparture in two sharply clashing national narratives. The question is whether\\nit might be possible to bridge that remaining gap through a further process\\nof negotiating identity.\\nClosing the gap at the symbolic level is clearly difficult because of the\\nsharp clash in national narratives that gives rise to it. Neither party can be\\nexpected to give up the core of its narrative in the course of the current ne-\\ngotiations; it is likely to take several generations before such changes can\\noccur. But it is possible, as part of the process of negotiation and reconcilia-\\ntion, for the parties to make changes in how they formulate their narratives,\\nwhere they place their emphasis, how they present the other within their\\nown narratives, and what they require by way of acknowledgment or valida-\\ntion of their narratives. Such changes at the edges of the narratives can po-\\nThe Role of National Identity in Conflict Resolution\\n207\\n\\n\\ntentially be “negotiated,” in the sense of being explored and jointly formu-\\nlated by the parties in a spirit of mutual accommodation.\\nThus, in the case of the right of return, we can assume that Palestinians\\nwill not abandon their conviction that they have a moral right to return to\\nthe homes from which they feel they were wrongfully displaced by the \\nIsraelis, and that Israelis will not admit guilt for the plight of the refugees.\\nThe question is whether it is possible to jointly formulate an Israeli ac-\\nknowledgment that would be substantially responsive to what Palestinians\\nneed to hear about the right of return without requiring either side to aban-\\ndon these core elements of their narratives. In the terms of my summary,\\ntoward the beginning of this section, of the symbolic issues that the refugee\\nproblem raises for the two sides, the question would be: Can the parties \\nformulate an Israeli acknowledgment that would provide significant\\n(though probably not total) satisfaction of Palestinians’ sense of justice and\\nvalidation of their claims for restitution, without requiring Israelis to negate\\nthe righteousness of their cause and making them vulnerable to future\\nclaims?\\nLet me offer three possible approaches to answering this question, which\\ncan be seen as three steps in the negotiation of identity that might further\\nnarrow the gap between the two sides manifested in solutions 3 and 4.\\nFirst, it may be possible to redefine the conception of responsibility in a way\\nthat would not imply guilt and thus be acceptable to the Israeli public, and\\nyet be viewed by Palestinians as confirming their narrative, validating their\\nclaims, and satisfying their sense of justice. One possibility here is for Israel\\nto accept a degree of causal responsibility without necessarily accepting\\nguilt. An Israeli acknowledgment might, for example, describe the refugee\\nproblem as a tragic and unjust outcome of a historical process that arose\\nwith the establishment of the State of Israel. It might go on to say that,\\nwhile there are disagreements about the precise historical events and the\\ndistribution of blame, Israel acknowledges that it was an active participant\\nin this historical process and, indeed, a beneficiary of it, and thus bears part\\nof the responsibility for these events. The statement might conclude with\\nan offer of compensation as an expression of that responsibility. A statement\\nof this kind clearly implies causal responsibility without explicitly acknowl-\\nedging guilt. It thus has the potential, if properly fine tuned, of an acknowl-\\nedgment consistent with both the Palestinian and the Israeli narratives.\\nSecond, it may be possible to find a mutually acceptable moral basis for the\\nright of return, which satisfies Palestinians’ need for an Israeli acknowledg-\\nment of the right of return as a moral principle without linking that moral\\nright to the refugees’ original homes in Israel. One approach here would be\\nto anchor the moral basis for the right of return in return to Palestine: \\nIsraelis might acknowledge the Palestinians’ moral right to return to Pales-\\ntine, based on the proposition that both peoples have legitimate claims to\\nthe land and that sharing the land (by establishing two states within it) \\nis, therefore, a just solution to the conflict. Although the two sides differ in\\n208\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\ntheir views of the justice and injustice of the events of 1948, the two-state\\nsolution allows them to develop a shared view of justice for the future. \\nIsraeli acknowledgment of the refugees’ moral right to return to Palestine,\\nwhile not fully addressing Palestinians’ sense of justice, would do so to a sig-\\nnificant degree. It would also strengthen the validity of Palestinian claims\\nby confirming (in a context of reciprocity) Palestinian national identity and\\nrootedness in the land. For Israelis, anchoring the moral basis for the\\nrefugees’ right of return in return to Palestine, rather than to their original\\nhomes, would greatly reduce their concerns about future claims, even\\nthough it may not entirely eliminate them.\\nThird, in conjunction with the two approaches just discussed and possi-\\nble other approaches to narrowing the gap between the two sides at the\\nsymbolic level, it may be possible for each side to find ways to affirm the\\nidentity of the other. Such affirmations, perhaps in the form of acknowledg-\\nments that the other has integral links to the land and national rights, could\\ncompensate each side for the concessions they make to resolve the refugee\\nissue: the Israelis for the risks they take in accepting causal responsibility for\\nthe plight of the refugees, and the Palestinians for accepting something less\\nthan the right of the refugees to return to their original homes. Affirmation\\nof the other’s identity can take place only in the context of a process of ne-\\ngotiating identity, based on the principle of reciprocity, in which each side\\nlearns to uncouple negation of the other from affirmation of the self. Each\\ncan affirm the other’s identity to the extent that it feels assured that it is not\\nthereby jeopardizing its own identity and its own claims.\\nThe three approaches to narrowing the remaining gap between the Is-\\nraeli and Palestinian positions in our Working Group’s paper on the prob-\\nlem of refugees all represent efforts to negotiate identity. As such, they\\nsearch for ways whereby each group can accommodate the narrative of the\\nother without undermining its own narrative. In such a process, each group\\nis encouraged to affirm or protect the identity of the other if, in turn, its\\nown identity is protected or affirmed. Each may be willing to grant some-\\nthing more to the other than it was originally prepared to offer, or receive\\nsomething less from the other than it originally demanded, in return for a\\nsubstantial concession by the other at the symbolic level.\\nConclusion\\nIn the Israeli-Palestinian conflict, as in other protracted ethnic conflicts, the\\never-present disputes over territory, resources, and political control are ex-\\nacerbated by perceived threats to national identity and national existence\\nthat underlie the actions and reactions of the opposing communities.\\nThreats to identity and existence create obstacles to the settlement of con-\\nflicts, even when both parties have concluded that a compromise agreement\\nis in their best interest. Moreover, even when specific issues in conflict are\\nThe Role of National Identity in Conflict Resolution\\n209\\n\\n\\nsettled and political agreements signed—often with the mediation of pow-\\nerful third parties—these agreements may not lead to stable peace, fruitful\\ncooperation, and ultimate reconciliation between the two parties, unless\\nthey have formed a new relationship based on mutual respect for their na-\\ntional identities.\\nThe experience of my colleagues and myself in problem-solving work-\\nshops on the conflicts in Israel/Palestine, Cyprus, Sri Lanka, and Northern\\nIreland, and experiences with similar efforts in other parts of the world,\\nsupport the central argument of this chapter about the role of national\\nidentity in intercommunal conflict and conflict resolution. A central feature\\nof such deep-rooted, protracted conflicts is the struggle over national iden-\\ntity between the two parties, which leads them to perceive their conflict in\\nzero-sum terms, not only with respect to territory and resources, but also\\nwith respect to national identity and national existence. Long-term resolu-\\ntion of such conflicts and reconciliation between the former enemies re-\\nquires changes in the groups’ national identities; in particular, they require\\na redefinition of each group’s identity so that affirmation of its own identity\\nis no longer predicated on negation of the identity of the other. Such iden-\\ntity changes are possible, provided they leave the core of each group’s iden-\\ntity and national narrative—its sense of peoplehood, its attachment to the\\nland, its commitment to the national language, culture, and way of life—in-\\ntact. Furthermore, such changes must and can be “negotiated” between the\\ntwo groups. It is possible for the groups to accommodate their national\\nidentities and the surrounding narratives to one another, as long as each \\nrespects and acknowledges the other’s peoplehood and thus reassures the\\nother that the core of its identity will not be jeopardized by changes in \\nthe periphery. Although some negotiation of identity takes place around the\\nnegotiating table, it is primarily a process that must engage the entire body\\npolitic on each side, at all levels and through different media. Problem-\\nsolving workshops provide a specially constructed arena for engaging in this\\nprocess and jointly developing new insights and formulations that can be\\ninjected into the public debate and can penetrate the political culture on\\nboth sides.\\nA central lesson from our experience is that national identity, though\\nvery much part of the problem in ethnic conflicts, can also become part of\\nthe solution. The way we talk about our identity affects the way we think\\nabout it and ultimately the way we act on it. In groups that are caught up in\\nprotracted conflict, identity depends on the conflict and is shaped by the\\nconflict: Many elements of identity are constructed as vehicles for pursuing\\nthe conflict. It should be possible, within limits, to reconstruct these ele-\\nments as vehicles for peace and reconciliation. What is needed is an invest-\\nment of identity in conflict resolution and in a new relationship with the\\nformer enemy. Development of such a new, transcendent identity confronts\\nmany obstacles. It cannot bypass the political process of negotiating a mu-\\ntually acceptable agreement, nor can it be allowed to threaten or undermine\\n210\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nthe particularistic identity of each group. But within these constraints, the\\npotential for reconstructing the national identities of former enemies in the\\nservice of peace and reconciliation exists and needs to be nurtured.\\nNotes\\n1. The work reported here is carried out under the auspices of the Program on\\nInternational Conflict Analysis and Resolution (PICAR), which I direct at the\\nWeatherhead Center for International Affairs, Harvard University. I am grateful to\\nthe Center for providing a home for PICAR, to the William and Flora Hewlett\\nFoundation for providing financial support for PICAR over a number of years; and\\nto Donna Hicks, Nadim Rouhana, and my other PICAR colleagues.\\n2. I am grateful to the organizations that have provided financial support for the\\nJoint Working Group over the years: the Nathan Cummings Foundation, the\\nCarnegie Corporation, the Ford Foundation, the Charles R. Bronfman Foundation,\\nthe U.S. Information Agency, and the William and Flora Hewlett Foundation, as\\nwell as the Renner Institut in Vienna and the Weatherhead Center for International\\nAffairs.\\nReferences\\nAlpher, J., & Shikaki, K., with the participation of the additional members of the\\nJoint Working Group on Israeli-Palestinian Relations (1999). Concept paper:\\nThe Palestinian refugee problem and the right of return. Middle East Policy, 6 (3),\\n167-189. (Originally published as Weatherhead Center for International Affairs\\nWorking Paper No. 98-7. Cambridge, MA: Harvard University, 1998.)\\nAnderson, B. (1983). Imagined communities: Reflections on the origins and spread of na-\\ntionalism. London: Verso. \\nBurton, J. W. (1969). Conflict and communication: The use of controlled communication in\\ninternational relations. London: Macmillan.\\nBurton, J. W. (1979). Deviance, terrorism and war: The process of solving unsolved social\\nand political problems. New York: St. Martin’s Press.\\nBurton, J. W. (1984). Global conflict: The domestic sources of international crisis.\\nBrighton, England: Wheatsheaf.\\nGerner, D. J. (1991). One land, two peoples: The conflict over Palestine. Boulder, CO:\\nWestview Press.\\nHalpern, B. (1969). The idea of a Jewish state (2nd ed.). Cambridge, MA: Harvard\\nUniversity Press.\\nHarkabi, Y. (1986). Israel’s fateful hour. New York: Harper & Row.\\nHertzberg, A. (1973). The Zionist idea. New York: Atheneum.\\nKelman, H. C. (1972). The problem-solving workshop in conflict resolution. In\\nR.L. Merritt (Ed.), Communication in international politics (pp. 168–204). Urbana:\\nUniversity of Illinois Press.\\nKelman, H. C. (1978). Israelis and Palestinians: Psychological prerequisites for mu-\\ntual acceptance. International Security, 3, 162–186. \\nKelman, H.C. (1979). An interactional approach to conflict resolution and its appli-\\ncation to Israeli-Palestinian relations. International Interactions, 6, 99–122.\\nThe Role of National Identity in Conflict Resolution\\n211\\n\\n\\nKelman, H. C. (1982). Creating the conditions for Israeli-Palestinian negotiations.\\nJournal of Conflict Resolution, 26, 39–75.\\nKelman, H. C. (1987). The political psychology of the Israeli-Palestinian conflict:\\nHow can we overcome the barriers to a negotiated solution? Political Psychology,\\n8, 347–363.\\nKelman, H. C. (1988, Spring). The Palestinianization of the Arab-Israeli conflict.\\nThe Jerusalem Quarterly, 46, 3–15.\\nKelman, H.C. (1992). Informal mediation by the scholar/practitioner. In J. Bercov-\\nitch & J. Rubin (Eds.), Mediation in international relations: Multiple approaches to\\nconflict management (pp. 64–96). New York: St. Martin’s Press.\\nKelman, H. C. (1995). Contributions of an unofficial conflict resolution effort to the\\nIsraeli-Palestinian breakthrough. Negotiation Journal, 11, 19–27.\\nKelman, H. C. (1997a). Nationalism, patriotism, and national identity: Social-\\npsychological dimensions. In D. Bar-T\\nal & E. Staub (Eds.), Patriotism in the lives\\nof individuals and nations (pp. 165–189). Chicago: Nelson-Hall. \\nKelman, H. C. (1997b). Negotiating national identity and self-determination in \\nethnic conflicts: The choice between pluralism and ethnic cleansing. Negotiation\\nJournal, 13, 327–340.\\nKelman, H. C. (1997c). Some determinants of the Oslo breakthrough. International\\nNegotiation, 2, 183–194.\\nKelman, H. C. (1997d). Group processes in the resolution of international conflicts:\\nExperiences from the Israeli-Palestinian case. American Psychologist, 52, 212–220.\\nKelman, H. C. (1998a). Social-psychological contributions to peacemaking and\\npeacebuilding in the Middle East. Applied Psychology: An International Review,\\n47(1), 5–28.\\nKelman, H. C. (1998b). The place of ethnic identity in the development of personal\\nidentity: A challenge for the Jewish family. In P. Y. Medding (Ed.), Coping with\\nlife and death: Jewish families in the twentieth century (Studies in Contemporary Jewry\\nXIV) (pp. 3–26). Oxford: Oxford University Press.\\nKelman, H. C. (1999). The interdependence of Israeli and Palestinian identities:\\nThe role of the other in existential conflicts. Journal of Social Issues, 55(3),\\n581–600.\\nKhalidi, R. (1997). Palestinian identity: The construction of a modern national conscious-\\nness. New York: Columbia University Press.\\nMendelsohn, E. (1989). A compassionate peace: A future for Palestine, Israel, and the\\nMiddle East (rev. ed.). New York: Farrar, Straus & Giroux.\\nMorris, B. (1987). The birth of the Palestinian refugee problem, 1947-1949. Cambridge:\\nCambridge University Press.\\nMuslih, M. (1988). The origins of Palestinian nationalism. New York: Columbia Uni-\\nversity Press.\\nMuslih, M. (1990). T\\nowards coexistence: An analysis of the resolutions of the Pales-\\ntine National Council. Journal of Palestine Studies, 19 (4), 3–29. \\nRouhana, N. N., & Kelman, H. C. (1994). Promoting joint thinking in international\\nconflicts: An Israeli-Palestinian continuing workshop. Journal of Social Issues, 50\\n(1), 157–178.\\nSegal, J. M. (1999). Defining Jerusalem. Middle East Insight, 14 (1), 27–28, 51–54.\\nT\\nessler, M. (1994). A history of the Israeli-Palestinian conflict. Bloomington and \\nIndianapolis: Indiana University Press.\\n212\\nThe Role of Social Identity in Reducing Intergroup Conflict\\n\\n\\nRichard D. Ashmore\\n9\\nLee Jussim\\nDavid Wilder\\nJessica Heppen\\nConclusion\\nToward a Social Identity Framework \\nfor Intergroup Conflict\\nWe began this volume with two goals: (1) to review and organize social sci-\\nentific knowledge, across several disciplines, about social identity and inter-\\ngroup conflict; (2) to set the stage for future work by highlighting the theo-\\nries, concepts, and methods with the most promise. These goals provide the\\nbasic structure for this concluding chapter. First, we take a look back: we\\nsummarize the chapters, distill the major points, and identify areas of agree-\\nment and disagreement. Second, we take a look forward: again drawing\\nfrom the preceding chapters, we present a framework that highlights the\\nideas that we feel should be the highest priority as we and other social and\\nbehavioral scientists attempt to construct a more comprehensive and coher-\\nent social identity perspective on intergroup conflict.\\nA Look Back \\nAs noted in the introductory chapter, the authors were given considerable\\nfreedom in how they carried out their task, and this shows in the types of\\ncontributions they have produced. Both chapters in Part I are conceptual\\nanalyses; however, Brewer presents a general theory, whereas Eriksen uses\\nconflicts in three societies to illustrate his key points. In Part II, the chap-\\nters by Citrin, Wong, and Duff and Sidanius and Petrocik report empirical\\ninvestigations, both based on survey data. In Part III, White presents a de-\\ntailed analysis of the group-based violence in Northern Ireland, while Staub\\ndescribes the place of self and identity within a general framework for un-\\nderstanding genocide. In the single chapter in Part IV, Kelman delineates\\n213\\n\\n\\nthe ways in which national identity is involved in problem-solving work-\\nshops designed to help ameliorate the Israeli-Palestinian conflict. We turn\\nnow to a more in depth review of these chapters.\\nPart I: The Contribution of Individuals’ Identities and the \\nCollective Identities of Social Groups to Intergroup Conflict\\nAlthough both Brewer and Eriksen present conceptual analyses, there are\\nclear differences between these contributions. Brewer details a very tight\\ntheory addressing a very specific question, When does ingroup love become\\noutgroup hate? This theory is stated in the most general and abstract form\\nand is similar to the grand hypothetico-deductive theories of learning in the\\n1940s (e.g., Hull, 1943) and to social psychological models of the 1950s\\n(e.g., Festinger’s theories of communication [1950], social comparison\\n[1954], and cognitive dissonance [1957]). And, as is true of this latter group\\nof conceptual frameworks, it is individualistic in orientation, focusing on\\nthe thoughts and feelings of individuals, though admittedly in a complex \\nsocial environment. Further, the model seeks support and testing primarily\\nin laboratory experiments. \\nBrewer begins with Sumner’s notion of ethnocentrism: social groups ex-\\nist in a state of war with one another, and, as a consequence, ingroup love\\nand outgroup hate are functionally interconnected sentiments necessary for\\ngroup survival. Brewer, however, suggests that Sumner’s theory can be de-\\ncomposed into four separate propositions: (1) social categorization—indi-\\nviduals and groups partition the social world into categories, and this cate-\\ngorization involves ingroup assimilation and outgroup contrast, which\\nmake social categories more distinct in individual and collective representa-\\ntions than in reality; (2) ingroup positivity—given that self is a member of\\nsome groups and that self is positively valued, there is a tendency to view\\none’s group (the ingroup) as good; (3) intergroup comparison—individuals\\ncompare their group to other groups, which can lead to judgments, not just\\nthat we are good but also that we are better than other groups; (4) outgroup\\nhostility—the preceding steps, together with certain additional conditions, lead\\npeople to hate an outgroup and actively work to harm it. Brewer suggests\\nthat these four propositions are not highly correlated, as Sumner argued,\\nbut instead constitute a progression of possible relations between ingroup\\norientations and feelings about, and behavior toward, the outgroup.\\nBrewer further proposes that social categorization and ingroup positivity\\nare likely universal. She also draws an important implication from the latter\\nprinciple. Much discrimination is driven by ingroup positivity and is, thus,\\nfavoritism for one’s own group, rather than discrimination against an out-\\ngroup. She presents laboratory experimental evidence in favor of this pro-\\nposal. She also notes that an important task for future research will be to\\ndevelop ways of disentangling pro-ingroup from anti-outgroup discrimina-\\ntion in real-world settings.\\n214\\nConclusion\\n\\n\\nThis raises the question that Brewer began with, When does ingroup\\nlove (and “discrimination for us”) turn to outgroup hate (and “discrimina-\\ntion against them”)? Brewer argues that it is not “realistic, objective con-\\nflict” that provides the significant impetus. Rather, she suggests that the an-\\nswer lies within her optimal distinctiveness theory of social identity, which\\nshe links to evolutionary theory. Her argument is that humans evolved, not\\nas isolated individuals or even in families, but instead as members of larger\\nsocial groups. T\\no allow the individual to learn from and cooperate with oth-\\ners, humans had to trust other people. However, to trust all others would be\\ndangerous. Thus, human beings developed the propensity to trust all others\\nwithin their group (“depersonalized trust”). When groups encounter one an-\\nother, each operating according to this same orientation gives rise to uni-\\nversal ingroup-outgroup stereotypes: “we” are close knit and cooperative;\\n“they” are clannish and treacherous. \\nThe final ingredient in the progression to outgroup hate is provided by\\nemotions. When members of different groups interact, they feel uncom-\\nfortable with one another and misattribute this negative emotional state to\\n“They are___________.” Under conditions of limited contact and little \\nsocial mixing, the dominant emotions are likely to be contempt (“They are\\ninferior and repugnant”), and this should motivate distancing (Karen \\nHorney’s “moving away from” [Ashmore, 1970]). However, under condi-\\ntions of close contact and possible integration, people will likely feel \\ninvaded, and this will give rise to anger and fear (“They are violent and dan-\\ngerous”). According to Brewer, it is these social conditions and associated\\nfeelings and attributions that turn ingroup love into outgroup hate and \\nproduce group-based harm-doing (Horney’s “moving against”). \\nIn contrast to Brewer’s development of a general theory of group identi-\\nfication and intergroup conflict with most evidence provided by laboratory\\nresearch, Eriksen is concerned with identity politics, especially how social,\\nespecially ethnic, identities are manipulated by political elites, and experi-\\nenced and enacted by everyday citizens, in real-world ethnic conflicts. Ac-\\ncordingly, his data are not from controlled experiments, but rather are case\\nstudies of three actual “ethnic conflicts,” though it is clear from his analysis\\nthat he would not accept this label. Eriksen builds his case, not by a step-by-\\nstep elaboration of a small set of ordered propositions as Brewer does, but\\ninstead by first describing the current position in cultural anthropology\\nconcerning ethnicity and its implications for intergroup conflict. This posi-\\ntion essentially argues that ethnicity is not primordial (a social category de-\\nfined by shared biological descent or ancestry and thus a grouping that is\\nfundamental and enduring for both the group itself and individual group\\nmembers) and not centrally about what a group of people share. Instead,\\nethnicity is best seen as a set of socially constructed group and interpersonal\\nperformances that communicate to self and others the boundaries between\\ngroups, which are not fixed entities but rather collectives in flux. T\\no capture\\nthis view, many anthropologists have turned from studying ethnicity per se\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n215\\n\\n\\nto seeking to understand social identities. One implication of this position is\\nthat group identity is conceptualized as both conscious and strategic—indi-\\nviduals are aware of their multiple social identities and deploy them situa-\\ntionally in order to achieve personal and group goals. \\nAlthough Eriksen generally accepts this basic position, he notes that it\\nleaves the person and his or her everyday experiences out of the ethnicity-\\nidentity-intergroup conflict equation. He proposes that, in addition to con-\\nceiving ethnicity as about social construction, boundaries, and strategic en-\\nactment, it is necessary to consider the personal identities of individuals and\\nhow these can become fused with social identities. More specifically with\\nregard to intergroup conflict, he suggests that, when the lived experience of\\nindividual group members is consonant with the key elements of a social\\nidentity currently proffered by authorities, it sets the stage for destructive\\nintergroup conflict fought in the name of that group identity. Thus, for ex-\\nample, although Bosnians and Serbs in Bosnia shared many cultural ele-\\nments and in multiple ways were quite similar to one another, they had\\nlived separate lives in the private sphere, with little intermarriage and few\\ncross-group close friendships. This situation of separate private spheres al-\\nlowed group myths and pejorative outgroup beliefs to be part of the every-\\nday experience of members of each group. Thus, when leaders espoused the\\nseparate and conflicting religiously based social identities of Bosnian and\\nSerb, the citizenry had personal selves and everyday experiences that made\\nthese appeals resonate. \\nAlthough they take two very different approaches and seem to offer in-\\ncommensurate, if not incompatible, frameworks for understanding inter-\\ngroup conflict, the Part I contributions are actually much closer than might\\nappear at first glance. Brewer starts with the individual and does a nice job\\nof identifying and ordering the cognitive and affective factors that set the\\nstage for ingroup attraction and, in turn, outgroup hostility. She concludes\\nby describing factors that moderate the model she has constructed. These\\nmoderators are all societal variables (e.g., social structural complexity), \\nand it is precisely such factors that Eriksen elaborates in his contribution.\\nEriksen draws the circle complete by his insistence that only when socially\\npropagated group identities fit with the lived experience of individuals and\\ntheir personal sense of self will the stage be set for ethnically based violence.\\nIn complementary ways, then, both Brewer and Eriksen note the impor-\\ntance of individual and social factors in the ways in which social identity is\\nimplicated in intergroup hostility.\\nThe foregoing should not be construed as meaning that Brewer and\\nEriksen are in total agreement about the personal and societal underpin-\\nnings of group-based harm doing. The clearest point of friction concerns\\nthe role of objective conflict between groups as a precipitator of hostile re-\\nlations. Although he never uses the term, Eriksen at several points presents\\nideas consistent with the realistic group conflict (RGC) view of intergroup\\nconflict (Ashmore & Del Boca, 1976; Boulding, 1962). According to this\\n216\\nConclusion\\n\\n\\nperspective, groups come into conflict over limited material resources—\\nland, energy sources such as oil, and the like. Whereas Eriksen identifies\\nobjective competition as crucial to conflict development and proliferation,\\nBrewer sees such “realistic” competition as much less important. In fact, she\\nmight be read as suggesting that tangible conflicts come only after the stage\\nis set by the elements offered in her model.\\nThe second major disagreement between Brewer and Eriksen is equally\\ninstructive. Brewer’s presentation is all about the here and now, about indi-\\nviduals’ thoughts and feelings about me, us, them, in the present moment.\\nEriksen, on the other hand, grounds his analysis in history. For each of his\\nthree examples, he describes what happened in the past, not just as stage\\nsetting but also as crucial to understanding why the groups are fighting\\nnow. This difference reflects, in part, two contrasting disciplinary para-\\ndigms. Brewer is a psychological social psychologist, and present-day prac-\\ntitioners of this social science continue to be much influenced by Kurt\\nLewin’s proscription against “historical” or “genetic” accounts: do not ex-\\nplain a person’s current behavior by recourse to what happened to that per-\\nson in her childhood; instead account for action on the basis of the contem-\\nporaneous field of forces acting on the person. Although Lewin’s ahistorical\\napproach referred to the life history of the individual, many of his students\\nwho, in turn, shaped today’s social psychologists, took Lewin’s approach\\nmore generally to mean that the important determinants of action are cur-\\nrent social factors. As a consequence, not only has the life history of the in-\\ndividual been left to developmental psychologists, but the past of social\\ngroups has been left for historians. In contrast, social and cultural anthro-\\npologists, viewing culture as, in part, something that one generation passes\\nto the next, have been much more inclined to see a group’s history as im-\\nportant to understanding its present organization and functioning—Thus,\\nEriksen’s descriptions of current conflicts feature historical information. \\nOne important next step will be to bring these perspectives together.\\nOne way, which will be elaborated below, is to extend the notion of life\\nstory and personal narrative into the analysis of social identity and group-\\nbased conflict. The basic notion is that, as individuals and groups, we make\\nsense of who we are now, in part, by looking back and creating stories—\\nabout self, about our ethnic group, about our nation. And it is these stories,\\nas part of personal and group identities, that may be crucial in setting the\\nstage for violence and oppression done in the name of us vs. them.1\\nPart II: The Contribution of Ethnic and National Identities to \\nPolitical Conflict in the United States\\nThe two chapters in the second part follow very well from Eriksen’s contri-\\nbution. Methodologically, he suggests that it is useful to look closely at spe-\\ncific intergroup conflicts; substantively, ethnic conflict within nation-states\\nis a major issue around the globe and is not limited to Third World or de-\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n217\\n\\n\\nveloping countries. In Part II, the focus is on patterns of ethnic divergence\\nand convergence in just one nation, the United States, arguably the most\\neconomically advanced country in the world. Further, only one type of con-\\nflict is considered: political contentions among ethnic groups.\\nCitrin, Wong, and Duff’s central concern is American national identity:\\nWhat is it? Are there ethnic group differences in beliefs about American\\nidentity? How do individuals prioritize ethnic and national identity? How\\ndo ethnic and national identity relate to policy preferences? Citrin and his\\nco-workers set the stage for their empirical research by carefully laying out\\ntheir notion of social identity, in this case national identity. Following T\\najfel\\n(1978) they conceive of national identity as involving both “identification\\nas,” or self-categorization (e.g., “I am an American”) and “identification\\nwith,” or affective attachment to the nation. Regarding the latter, they fur-\\nther distinguish patriotism, or love for one’s country (“Being American is\\nvery important to me”) from chauvinism, or feelings of national superiority\\nand degrading of other countries (“If only other countries were like\\nAmerica the world would be a much better place”). \\nCitrin et al. suggest that they go beyond T\\najfel (a point we return to be-\\nlow in considering the multiple dimensions of social identity) by adding a\\nthird component to national identity—normative content. The use of the\\nword “normative” is deliberate and does not mean average. Instead, norma-\\ntive conceptions are assertions about how society should be organized, how\\npeople should behave to make society work well, and how society should\\ntreat groups and individuals in the nation. As political scientists, Citrin et al.\\nbring an important theme to the volume—normative as pre- and pro-scrip-\\ntive, not just descriptive. This reflects the subfield of political theory and\\nthe issue of how political systems should be constructed. Given the empha-\\nsis on ethnicity and national identity, Citrin and his colleagues stress three\\ndifferent normative views of how ethnic and national commitments should\\nbe related in multiethnic states: liberal, in which national membership de-\\npends on sharing a set of ideals (e.g., respect for the individual, freedom)\\nand not shared ethnicity (hence, it is often termed civic, not ethnic, nation-\\nalism); nativism, in which American identity is equated with one ethnic\\ngroup, white Anglo Saxons, and thus, to be an American means accepting\\nthe values and other characteristics of this dominant ethnic group; multicul-\\nturalism, in which ethnicity is prized above national identity.\\nCitrin et al. also preface their data reporting by introducing symbolic\\npolitics theory (Sears, 1993), which stresses that political orientations are\\nsocialized in childhood and maintained by adult engagements (e.g., work).\\nAccording to the authors, national identity is widely socialized in the\\nUnited States, and, thus, there should be more ethnic consensus than con-\\nflict concerning national identity. \\nCitrin et al. address their questions about national identity and predic-\\ntions from symbolic politics theory using survey data collected from both a\\nnational sample and Los Angeles County respondents. Concerning their\\n218\\nConclusion\\n\\n\\nfirst question, What is American identity?, they found considerable agree-\\nment that, in regard to political issues, most Americans define self as “just\\nan American” (self-categorization) and feel considerable pride in and patri-\\notism for America, but not high levels of chauvinism. Also, there was con-\\nsiderable support for nativistic beliefs (though the pattern was uneven) and\\neven more support for assimilation notions. \\nIn answer to their second question, Are there ethnic group differences in\\nviews on American identity?, Citrin et al. concluded that there are few dif-\\nferences in defining American identity except on issues pertaining to ethnic\\ngroups per se (e.g., how the nation treats minority groups). A most intrigu-\\ning finding was that, on average, African Americans were more nativist than\\nwhites. This was especially true for endorsing being a Christian as a crite-\\nrion for being “a true American.” As the authors note, this suggests yet an-\\nother important social identity, religion. \\nTurning to the third question, How do individuals prioritize ethnic and\\nnational identity?, Citrin and colleagues found that, when given the option,\\nabout 50 percent of African Americans chose “both ______ [my ethnicity]\\nand American,” about one-third chose ethnicity as their primary identity,\\nand about one-sixth chose “just American.” Hispanics and Asians responded\\nsimilarly, but were somewhat less likely to select “just American” and some-\\nwhat more likely to choose “both _______ [my ethnicity] and American”\\nand ethnic identity. For whites, on the other hand, “just American” was far\\nand away the predominate response. \\nThe final question was, How do ethnic and national identity relate to\\npolicy preferences? As predicted by symbolic politics theory, race (given\\nsmall sample sizes, only whites and blacks were compared) did not predict\\nviews on multicultural policies or international affairs (military spending or\\nprotectionism), whereas national identity did. On race-related policies, race\\ndid predict policy preferences. However, when ethnic identification was\\nadded to the predictive equation, felt closeness to one’s ethnic group did not\\nenhance support for race-based policies for blacks, but for whites felt close-\\nness increased disapproval of programs that benefited minority groups.\\nThat it was among whites that identification with one’s group related to po-\\nlitical issues regarding race may seem paradoxical to some given that the\\npractice of identity politics is most often associated with minority groups.\\nThe Citrin et al. contribution highlights three points. First, social identity\\nis multidimensional. Their chapter considers self-categorization, affective\\nattachment, and normative content. Below we identify additional possible\\nfacets of social identification. Second, a particularly important aspect of ex-\\nperiencing ethnicity, though often neglected, is the “shoulds” attached to\\none’s group. These are related to the notion of normative content, but they\\nfocus, not on broad issues of what it means, for example, to be a true Ameri-\\ncan, but on the very specific group norms for how “one of us” should act—\\nthe food we should eat, the way we should talk, and the like—and the pres-\\nsures an individual feels from members of her or his own group to abide by\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n219\\n\\n\\nthese “rules.” T\\no address this issue, Contrada et al. (in press) have recently\\ndeveloped a self-report instrument to assess individual differences in own-\\ngroup pressure to behave in specific ways. Third, as Citrin and his colleagues\\nnote, social identities are not necessarily political identities. This raises the\\ncrucial question, How do people get from social definition to action? \\nIn the second chapter of Part II, Sidanius and Petrocik take up much the\\nsame general topic as Citrin and his collaborators; the relationships be-\\ntween ethnicity and national attachment in the United States. And like \\nCitrin et al., Sidanius and Petrocik identify three perspectives on this issue:\\nmelting pot (each ethnic group contributes something to an emergent \\nnational American identity, and, in the process, members of these ethnic\\ngroups lose their own sense of ethnicity and become identified with, and \\nattached to, the nation); pluralism (individuals and groups retain their\\nuniqueness, but all agree to be tolerant and to also identify with the nation);\\ngroup dominance model (in all societies, groups are arranged in a hierarchy\\nof prestige and power, and this means that groups not only maintain their\\ndistinctive identity, but also that, for those at the top, ethnocentrism and\\npatriotism are positively linked, but for those at the bottom this correlation\\nis negative). Again in parallel with the preceding chapter, Sidanius and\\nPetrocik assess with survey data the predictions they derive from each per-\\nspective. These data come from five different samples, from university stu-\\ndents, to Los Angeles residents, to a national sample. \\nSidanius and Petrocik address three specific issues: the salience of ethnic\\nidentity; the relation between ethnicity and patriotism; the relationship be-\\ntween patriotism (attachment to the nation) and ethnocentrism (by which\\nthey mean a negative attitude toward American ethnic groups relative to\\none’s own group). In regard to the first issue, Sidanius and Petrocik con-\\nclude that ethnic identity is salient for Americans and that this is especially\\ntrue of minority-group Americans. With respect to the second issue, Sida-\\nnius and Petrocik find that white Americans are highest in patriotism and\\nAfrican Americans lowest. Concerning the third issue, Sidanius and Petro-\\ncik report that, for white Americans, patriotism and ethnocentrism are of-\\nten positively correlated, but the correlation for minority Americans is gen-\\nerally negative. \\nOf the many excellent points in the Sidanius and Petrocik contribution,\\nwe underline just one, their group dominance model. Individuals develop\\ntheir intergroup attitudes and ethnic groups come into conflict within soci-\\nety, and all known societies contain group dominance structures: Some so-\\ncial categories are privileged and others are disadvantaged. Thus, the soci-\\netal context for intergroup relations is never neutral or even-handed, and\\nthis is true even in economically advanced nations with strong democratic\\ntraditions. This is a particularly important reminder at this time in the\\nUnited States, where we surmise that a large portion of the dominant white\\npopulace believes that the civil rights and women’s movements have some-\\nhow leveled the playing field for all Americans and now it is time to treat\\n220\\nConclusion\\n\\n\\neveryone as individuals and not as group members. Without denying the\\nsocial changes that have occurred over the past four decades, it is simply not\\ntrue that American society and culture treat all people and groups as equal. \\nThe chapters by Citrin et al. and Sidanius and Petrocik present a para-\\ndox: both address the same basic issue (the interplay between ethnic and na-\\ntional identity in the United States), they deal with two similar folk models\\n(though they use somewhat different labels—liberal [Citrin et al.] and plu-\\nralism [Sidanius & Petrocik]; nativist [Citrin et al.] and Americanization\\nand melting pot [Sidanius & Petrocik]; see also Fredrickson [1999] for a\\nhistorical review of models of American ethnic group relations), they both\\ndraw on similar previous studies (especially that by de la Garza, Falcon, &\\nGarcia, 1996), and they take the same basic methodological approach (sur-\\nvey data; drawing in one case from the same survey). But they reach quite\\ndifferent conclusions: Citrin and co-workers see a small ethnic divide in the\\nUnited States and little tension between ethnic and national identity,\\nwhereas Sidanius and Petrocik perceive a large ethnic divide and consider-\\nable conflict between ethnic and national attachment. \\nThere are three factors that make this paradox more apparent than real.\\nFirst, they use different items to assess ethnic/national self-categorization.\\nSecond, the two chapters are not really directed at precisely the same set of\\nphenomena. Citrin et al. predict policy preferences from ethnicity and eth-\\nnic identity, whereas the primary outcome variable of interest to Sidanius\\nand Petrocik is the correlation between national attraction (especially pa-\\ntriotism) and prejudice against U.S. ethnic groups other than one’s own\\ngroup. Third, and we believe most important, in reporting results, Citrin\\nand colleagues concentrate on absolute levels of attachment to the United\\nStates (which are generally high) and play down group differences (which,\\nfor some measures, are substantial), whereas Sidanius and Petrocik stress\\ngroup differences and deemphasize absolute levels of national attachment\\nby ethnic group members. \\nThe Citrin et al. and Sidanius and Petrocik chapters underline four im-\\nportant points. First, how one asks people to report their group self-catego-\\nrization is crucial. Some items implicitly pull for “just American,” whereas\\nothers pull for ethnicity and nationality self-codings either instead of, or in\\naddition to, labeling self as “American.” Our recommendation is that sur-\\nveys concerned with ethnicity and national identity should use the 1995\\nLACSS item (described by Citrin et al.) that offers the respondent the fol-\\nlowing choices: “just an American,” “_______ [whichever national or ethnic\\ngroup the person sees self as a member of],” “both American and ______\\n[self-categorized ethnic group].” This seems to us to provide the most neu-\\ntral way of getting at how respondents define and categorize self in terms of\\na superordinate national identity (“American”) and more particular ethnic\\nself-codings. Although the results reported by Citrin et al. are from a small\\nand local sample, they suggest that the modal white American sees self as\\n“just American,” while minority Americans define self in three very distinc-\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n221\\n\\n\\ntive ways: about half select a dual identity, about one in three categorizes\\nself in terms of ethnicity only, and about one in six regards self as just\\nAmerican.\\nSecond, the Citrin et al. and Sidanius and Petrocik chapters remind us\\nthat it is easy to slip from ethnic identity, or identification, to ethnicity, or\\nsimple ethnic group differences. Although both chapters report analyses\\ndealing with individual differences in group identification (especially self-\\nlabeling), they also report analyses in terms of ethnic group differences.\\nHowever, simple ethnic group differences do not necessarily mean differ-\\nences as a function of ethnic or national identity. In addition, such an ap-\\nproach obscures the wide variation within any social category in the content\\nand degree of identification with the group. For example, what other social\\nidentity, intergroup attitude, and public policy preference differences are\\nthere among minority group Americans who self-label as “just American,”\\n“American and ____,” and minority group label only?\\nThird, ethnic groups are not monolithic. Obviously there is variation by\\nage, socioeconomic status, and a variety of other variables that Citrin et al.\\nand Sidanius and Petrocik partial out in some analyses. At the same time,\\nsome of these partialed-out variables constitute important social identifica-\\ntions. For example, there is considerable variation in the African-American\\n(and white) community in terms of religious commitment, especially defin-\\ning self as a “fundamentalist Christian.” \\nFourth, just as ethnic and national identity are multifaceted, so too is \\nnational attachment. Both chapters note that there is a positive form of na-\\ntional attachment (patriotism) and a more virulent brand (chauvinism/\\nnationalism—basically “We are the best in the world . . . and others\\nshould emulate us”), as well as issues of willingness to leave the country and\\nbeliefs about what makes someone a true American. \\nThe foregoing summary suggests that the Citrin et al. and Sidanius and\\nPetrocik chapters have not resolved the question of how ethnicity and eth-\\nnic identity relate to national identity and attachment. They have, however,\\nprovided some important clues. In general, there is fairly high attachment\\nto the United States in all groups, but there are also substantial differences\\nbetween groups, especially on issues pertinent to ethnicity. More impor-\\ntant, the two chapters begin to point us toward the necessary next genera-\\ntion of research. This will need to use multiple measures of ethnic or na-\\ntional self-categorization, especially items that do not tilt toward or away\\nfrom ethnic (or national) identification, and multiple measures of national\\nattachment. Thus, future work will be directed at two important questions:\\nHow do the various facets of ethnic and national identity (self-categori-\\nzation, felt closeness, and the like) relate to the various components of \\nnational identity? How, in turn, do these sometimes competing and some-\\ntimes complementary forms of self-identification relate to variables perti-\\nnent to intergroup conflict (e.g., attitudes toward other groups, political\\npolicies and programs)? \\n222\\nConclusion\\n\\n\\nPart III: The Contribution of Social Identity to Violent Intergroup Conflict\\nWhite continues the focus of Part II to look intensively at one conflict, but\\nthe site of his analysis is Northern Ireland, not the United States, and the\\ntype of conflict is deadly violence, not political strife. White begins his\\nchapter in a very dramatic fashion by presenting a table indicating who has\\nkilled whom in the conflict. This table not only helps him make some very\\nimportant points about the conflict but also reminds the reader that the ul-\\ntimate outcome of identity-based disputes is that people may kill one an-\\nother. Of the many excellent points made by White, we highlight three: (1)\\nthe past importantly shapes the contemporary conflict; (2) the violence in\\nNorthern Ireland is three- not two-sided; (3) identity is involved in the con-\\nflict in multiple ways. \\nFirst, the current fighting is rooted in a long history, and that past not\\nonly shapes the present conflict but also sets parameters for how the hos-\\ntility might be reduced. White describes how the violence in Northern Ire-\\nland began with the Norman/English invasion of Ireland over 700 years\\nago, how religion was introduced into the conflict in the 16th century with\\nthe Reformation, how a Protestant ascendancy was created during the 17th\\ncentury, how Northern Ireland was created in the early part of this century,\\nand how the current protracted violent confrontation began in the 1960s. \\nThis history is important to the current conflict in three major ways.\\n1. It underlines the importance of losing for group definition. On multi-\\nple occasions over seven centuries, the Irish have lost battles and wars to the\\nEnglish. These repeated defeats have been incorporated into the group nar-\\nrative of the Republicans of Northern Ireland. This parallels the history\\nand national narrative of Serbs as described by Eriksen (this volume; see\\nalso Staub this volume). This 700-year record also suggests why the Irish\\nRepublican Army may be so reluctant to “decommission,” which, as this\\nchapter is being written (spring 2000), may derail the recently agreed to\\npeace accord. The Irish have lost many times to the English; they can not\\naccept another “defeat,” which is how many perceive decommissioning,\\neven if the current agreement is widely supported by Protestants and\\nCatholics.\\n2. The Protestants in Ireland were imported by the British and took\\nover land and status from defeated Irish. They have always been in a mi-\\nnority in Ireland as a whole (though a majority in Northern Ireland), and in\\nWhite’s words, have a “siege mentality.” It is likely that being a numerical\\nminority with political and economic majority status (as whites in Southern\\nAfrica and the Tutsi in Rwanda until recently) also has important implica-\\ntions for social identity. Groups that have superior power but are numeri-\\ncally in the minority (power majority, numerical minority) likely stereotype\\nthe other group (power minority, numerical majority) as, at the same time,\\nboth immature and childlike (to justify their inferior position) and poten-\\ntially violent in a primitive way (to make sense of past uprisings and fears of\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n223\\n\\n\\nwhat might happen if “they” won). This is much the same set of images that\\ncharacterized U.S. whites’ views of blacks during the slavery period (Boskin,\\n1970). In complementary fashion, the self-image of power majority yet nu-\\nmerical minority groups features traits such as grown up and mature (re-\\nsponsible, productive, and the like) and civilized.\\n3. As White notes, a long history of fighting an external oppressor also\\nsets the stage for role identities in this struggle, a point to which we return\\nbelow. Here we simply note that often roles are hard to give up, thereby\\nmaking intergroup conflict difficult to reduce.\\nSecond, what looks to an outsider as a simple two-sided religious con-\\nflict—Northern Ireland Catholics versus Protestants—is, upon more care-\\nful analysis, really a multisided political conflict with a religious gloss.\\nWhite shows that there are three key contestants—Protestant paramili-\\ntaries (who are just a portion of the Protestant community); Irish Republi-\\ncans (most IRA members are Catholic, but not all Catholics are affiliated\\nwith the IRA, and some IRA members are Protestant); and the state, espe-\\ncially British authorities. At the end of his analysis, White notes that there is\\nactually a fourth important party, the people and government of the Repub-\\nlic of Ireland, and, with the current efforts at peace, the role of that party\\nwill likely be increasing. The United States has also been involved in the re-\\ncent efforts to resolve the conflict, and thus there is a fifth “side.” The cru-\\ncial point for the present volume is that apparently “us against them” con-\\nflicts are seldom just two-sided; and the various parties are likely to have\\ndifferent “identity shares” in the conflict, which leads to the next major\\npoint we wish to underscore about White’s chapter.\\nThird, identity is implicated in multiple and diverse ways in the conflict.\\nT\\no uncover how identity enters into the political violence, White focuses on\\naccounts given by the various participants—quite simply how do various\\nparticipants describe the conflict and their role in it? White begins his\\nanalysis by using Thoits and Virshup’s (1997) distinction of “we” from “me”\\nsocial identities. The former are collective social identifications (member of\\nsome social category [e.g., nation, ethnic group]), whereas the latter are role\\nidentifications (occupant of a particular role [e.g., parental, occupational]).\\nWhite argues that for Republicans their involvement reflects both a collec-\\ntive social identity (“we”) and role identity (“me”). Importantly, the collec-\\ntive self-definition is not me as Irish Catholic versus Irish Protestants, but\\nme as Irish Nationalist versus the British. The role identity is a set of enact-\\nments that the person has learned from either growing up in a Republican\\nhome or as part of joining a Republican organization. White illustrates a\\nRepublican role identity by describing the hunger strikes of the early 1980s.\\nHe notes that hunger strikes have a long history in Republican efforts to get\\nthe British out of Ireland, and this means not only that these are part of the\\n“Republican prisoner” role but also that there are specific men who have\\ngone on hunger strikes (and often died) and who now serve as role models\\n(see also Aretxaga, 1993). \\n224\\nConclusion\\n\\n\\nWhite’s suggestion that role identities can be important parts of involve-\\nment in intergroup conflict is a significant addition to role identity models\\n(McCall & Simmons, 1966; Stryker,1987). Just as the normal everyday ac-\\ntivities of society require roles, so, too, do protracted conflicts. What might\\nthese conflict roles be? Obviously, there are the roles of active fighter (sol-\\ndier) and prisoner. But in addition, “armies” require that other functions be\\nfulfilled. One important function is intelligence, or the collecting of infor-\\nmation to assist in waging the fight. White describes a British soldier who\\nslept outside a bakery; it is possible that the owner of the bakery could have\\nprovided intelligence reports to either the Republican or Protestant para-\\nmilitaries and, if so, would have taken on a conflict-related role identity. \\nIn addition, when an army is an insurgent one, other roles are necessary,\\namong them, fund-raiser, politician, and civilian support. In one of the Re-\\npublican narratives reported by White, the respondent describes his grand-\\nmother, of whom he said, “She never wavered.” She may also have supplied\\na safe house to paramilitaries and fulfilled additional functions in support of\\nothers with active fighting roles, and, in so doing, her role of “supporter”\\nwould have been active and not passive. \\nAccording to White, the Protestant Loyalists and British actors in\\nNorthern Ireland construe their involvement solely in “we” social identity\\nterms. For the Protestant paramilitaries, the conflict is “us against them,”\\nwhere “We are British Protestants” and “They are Irish Catholics.” For the\\nEnglish, White argues that the conflict is “them versus them,” a religious\\nconflict among the Irish. \\nWhite acknowledges possible role identities for Protestants and British,\\nbut we believe that he underplays them. Although in the earliest years of the\\ncurrent “troubles,” the Loyalists may have been relatively unorganized with\\nlittle sense of history, certainly as the conflict evolved it is probable that\\nProtestant paramilitary groups became relatively well organized and devel-\\noped systems of roles. Role identities are even more probable for the British\\nor English first-line participants in Northern Ireland, the military person-\\nnel. It is likely that many English soldiers construed their involvement there\\nas “me as a soldier doing a soldier’s job,” and not as “we British.”\\nWhite’s primary data are verbal accounts, and gathering such narratives\\nin the midst of an armed conflict is very difficult. This difficulty raises some\\nquestions about White’s conclusions. As he notes, the verbalizations he re-\\nports are not factual histories or descriptions, but instead they are stories\\nthat a person has told to someone else. As with any story, the content of the\\naccount depends, in part, on who is talking to whom for what purpose. This\\nraises the most general question: Are the narratives given by the three\\ngroups of protagonists (Republicans, Protestant paramilitaries, English sol-\\ndiers and elites) comparable? This can be partitioned into two more specific\\nqueries: Who gave the accounts? T\\no whom were the accounts given and\\nhow did the respondent think the accounts would be used? Regarding the\\nfirst or “who?” question, White acknowledges that the three groups were\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n225\\n\\n\\nnot equated: many of the Republicans were Sinn Fein members; Protestant\\nparamilitary accounts were from former paramilitaries; the English voices\\nwere from soldiers at the start of the current conflict (late 1960s) and politi-\\ncal leaders. Regarding the “to whom and how used?” question, the Republi-\\ncans were all interviewed by White himself, and it is likely that they were\\nassured confidentiality and told that the conversations were for research\\npurposes. The Protestant paramilitary stories and statements by English ac-\\ntors were taken from secondary sources, most from memoirs and remem-\\nbrances or from official statements. The question of how the respondents\\nthought the accounts would be used is less certain, but at least some had\\nclear public relations and political implications. As a consequence, these\\nstatements may have been tilted toward a “we” social identity perspective\\nand away from explaining the individual’s roles in the conflict. \\nWhile White presents a detailed analysis of just one violent conflict,\\nStaub describes a general framework for understanding genocide that is\\nbased on, and applied to, several different murderous group conflicts. Basi-\\ncally, this model indicates that mass killing and genocide are the end result\\nof a set of personal and social processes that are set in motion by one of two\\ngeneral types of instigators: difficult life circumstances (e.g., hyperinflation)\\nand intergroup conflict. These instigators threaten basic needs, including\\nthose related to identity. This sets the stage for a spiral toward large-scale\\nsystematic violence as follows: the psychological elevation of the ingroup\\n(and generally a derogation of the outgroup), which facilitates initial harm-\\ndoing; this, in turn, produces altered societal institutions and individual\\nidentities. These lead to the final step in the model, mass killing and geno-\\ncide. T\\no this causal sequence, Staub adds two moderators. The first, culture,\\ninvolves a variety of societal-level factors (e.g., history of devaluation of \\nan outgroup); the second, bystanders, indicates individuals both within \\nand outside the group who are not (at least initially) involved in the harm-\\ndoing.\\nAmong the many insights provided by Staub’s model, we highlight five.\\nFirst, both group and individual identity play important roles in intergroup\\nviolence. Regarding the former, Staub suggests that ethnic groups and na-\\ntions vary in their “group self-concept,” and this shared notion of “us” can\\nlead some groups toward mass killing and genocide while directing others\\naway from such harm-doing. In particular, Staub highlights as a cultural\\nfactor predisposing toward systematic intergroup violence, “unhealed vic-\\ntimization,” the belief that “we” have been hurt coupled with no opportu-\\nnity to personally and socially come to terms with this wound. \\nSecond, and closely related to the first point, a personal and societal\\nsense of vulnerability (“They have killed our people before, and they could\\nattack us at any time”) can not only predispose a group toward violence but\\nalso make preemptive violence more likely. As Eriksen noted regarding\\nSerbs in the former Yugoslavia and White with regard to the Irish Republi-\\ncans, these groups had histories of losing, which justified, in the partici-\\n226\\nConclusion\\n\\n\\npants’ minds, acts of “getting them before they get us.” This type of mind-\\nset helps sustain these conflicts \\nThird, regarding the through-time set of factors leading to intergroup\\nviolence, most important for the present volume is that these involve\\nchanges in individual and group identities. Those in a society who engage\\nin initial violence are more receptive to self-definitions that make their\\nharm-doing not only acceptable but, in some cases, also morally right; those\\nwho at the outset are bystanders can come to see killing or support for\\nkilling as part of how “I ought to behave” and alter their self-conception \\naccordingly. \\nFourth, in terms of culture moderators, it is possible to identify societal-\\nlevel identity types that are more likely to participate in mass killing, and\\nthese are not neatly divided along Eastern versus Western modal self-\\nconcept configurations. In fact, as Staub notes, both the autonomous indi-\\nvidual self more common in Western industrialized societies and the more\\nsocial and connected self-configuration that is more common in Eastern\\nand less industrialized nations have their own vulnerabilities to genocide.\\nThe challenge for the next generation of researchers is to assess these types\\nand test Staub’s suggestions. (See Norenzayan et al. [1999] and Lyengar et\\nal. [1999] for additional ideas about different types of cultures and the im-\\nplications for intergroup strife.) \\nFifth, third parties play significant roles both in creating conditions ripe\\nfor large scale group violence and as a major resource in limiting or stop-\\nping genocide. Staub describes how Belgians built on existing Hutu-Tutsi\\ndominance systems and added the notion of “blood” to help make the soci-\\nety more manageable for them. Years later this made the violence more\\nlikely. \\nPart IV: The Role of Social Identity in Reducing Intergroup Conflict\\nStaub’s emphasis on the importance of bystanders, especially external by-\\nstanders, as crucial to preventing the escalation to mass killing and genocide\\nleads very nicely into the final part, which focuses on reducing conflict. The\\nchapter in this section, by Kelman, describes problem-solving workshops as\\na conflict-reduction method that features third parties. Kelman begins his\\nchapter by describing the history of the Israeli-Palestinian conflict. Just as\\nEriksen and White suggest that the past is important to understanding why\\nand how intergroup harm-doing develops, Kelman proposes that efforts to\\nreduce conflicts must take into consideration the historical foundations of\\ncurrent strife. Next, Kelman lays out his notion of national identity. He ar-\\ngues that, “Although national identity is carried by the individual members\\nof a national group and can thus be studied as a property of individuals, the\\npresent chapter refers to it as a collective phenomenon—as a property of\\nthe group.” In the simplest terms, national identity is the shared and con-\\nsensual definition of us as a group and nation; this includes our claims to\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n227\\n\\n\\nland and continuity over time, our beliefs about who we are, and what\\nmakes us distinctive. Kelman suggests that national identity is accompanied\\nby a “national narrative—an account of the group’s origins, its history, and\\nthe relationship to the land.” Thus, for Kelman, social identity as a nation\\ninvolves both a self-definition (national identity) and group story (national\\nnarrative). \\nIn seeking to understand how national identity might underlie conflict,\\nKelman argues that in protracted intergroup fights, especially in which\\nboth sides have claim to the same land, the groups’ national identities can\\nbecome exclusive (e.g., “This land belongs to us and can belong to no one\\nelse”) and monolithic (all the multiple dimensions of nationhood [religion,\\nethnicity, history, and the like] become seen as highly intercorrelated). Both\\nof these properties make conflict resolution more difficult, identity exclu-\\nsivity by making sharing impossible by definition, and identity unidimen-\\nsionality by making us all the same and reducing any possibility that in\\nsome ways they are like us. One final aspect of protracted conflicts such \\nas that between the Israelis and Palestinians is that the process of fighting\\ncan become part of the national identities—“We are the people who fight\\nthem.” This echoes Staub’s observation that engaging in intergroup vio-\\nlence changes individuals. Kelman proposes that intergroup conflict can\\nchange whole nations and become part of the shared collective self-defini-\\ntion and the consensual national narrative.\\nKelman then describes how identity change, though difficult, is still pos-\\nsible. Here he makes two very significant points. First, although people and\\nnations see their identities as fixed, they are, in fact, not. Identities, like all\\nbelief systems, are not immutable essences, but rather are socially con-\\nstructed and, as such, are modifiable. Second, unlike land and other physical\\nobjects, identities are not zero-sum; the definition of “us” need not include\\nthe defeat or devaluation of “them”.\\nHow might these possible changes in national identity be achieved? Kel-\\nman describes the problem-solving workshops that he and his colleagues\\nhave been conducting for many years in the Middle East. These feature\\nthird parties who serve both to moderate the discussions and to provide an\\nunbiased resource for both sides. Another key aspect of the workshops is\\nthat the participants are influential people in the Israeli and Palestinian\\ncommunities (e.g., business leaders) but not formal political leaders. Thus,\\nthey have “clout” but do not need to represent a political position. \\nThe workshops attempt to find ways to bridge the gaps between the two\\nsides. Kelman highlights how this involves altering national identities. For\\nexample, workshop participants have considered what aspects of their na-\\ntional identity might be outdated and thus might be psychologically dis-\\ncarded. Kelman concludes his contribution with a detailed description of\\nhow one workshop grappled with a crucial issue in the Israeli-Palestinian\\nconflict: what to do with, and for, the Palestinian refugees. A key feature of\\nhis analysis is that this question has both material (what Kelman terms\\n228\\nConclusion\\n\\n\\n“practical”) and symbolic facets. On the material or objective side, how can\\nIsrael and Palestine accommodate the large number of refugees in other\\ncountries? On the symbolic side, how can each side resolve the refugee is-\\nsue in such a way that it does not do violence to its national identity and\\ngroup narrative?\\nA Look Forward: A Social Identity Framework \\nfor Intergroup Conflict\\nAs the preceding “look back” indicates, our authors have done an excellent\\njob of addressing the first goal of this volume: to survey what the social and\\nbehavioral sciences have learned about the ways in which social identity and\\nintergroup harm-doing are interconnected. They have also succeeded in\\naddressing our second goal in putting together this book: to identify the\\nneeded next steps in moving toward a more complete account of social\\nidentity and intergroup conflict. In this section we pull together and orga-\\nnize the next-generation conceptual and methodological tools identified by\\nour contributors in the form of a social identity framework for intergroup\\nconflict that is schematically depicted in figure 9.1. It is important to em-\\nphasize that this is not a tight deductive model such as proposed by Brewer;\\ninstead, we offer a heuristic scheme that identifies important constructs and\\nsuggests how they might be interconnected. We present this framework as\\nfollows. First, we provide a brief overview of the scheme. Second, the ad-\\nvantages of phrasing the framework in terms of a general threat or coping\\nparadigm are described. Third, the details and implications of the frame-\\nwork are presented.\\nOverview of Framework \\nThe figure depicts three general sets of constructs—threat and appraisal;\\ndestructive coping: intergroup harm-doing; and social identity—and identi-\\nfies these as mutually influencing one another. The arrow from “threat and\\nappraisal” to “destructive coping: intergroup harm-doing” marks this as a\\ngeneral threat/coping framework as used in psychological analyses of stress\\n(Lazarus & Folkman, 1984). When members of a group perceive actual or\\npossible harm to themselves and other ingroup members (threat), they seek\\nto makes sense of this (appraisal), and if the other group is somehow judged\\nas responsible for the threat, they may cope with the appraised threat de-\\nstructively by harming that group and its members. \\nThe return arrow from “destructive coping” to “threat” indicates that in-\\ntergroup harm-doing can escalate threat and appraisal: when one group in-\\nflicts harm on another, it raises the possibility of retaliation (i.e., threat).\\nThe figure indicates that threat and appraisal and harm-doing are each con-\\nnected to social identity, and also that social identity moderates the link of\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n229\\n\\n\\nthreat and appraisal to intergroup violence. Social identity consists of both\\ngroup (socially shared) and individual (personal) representations about own\\nand other group. Group identities that feature own group as victim and\\nother group as oppressor or aggressor heighten objective threat and ap-\\npraisal of danger; the reverse path indicates that being threatened and see-\\ning another group as responsible is likely to influence individual and group\\ndefinition (e.g., living in a state of war, declared or not, with another group\\nwill help create social and personal identities congruent with this state). \\nSimilarly, the “intergroup harm-doing” to “social identity” arrow indi-\\ncates that groups that engage in outgroup violence will have their personal\\nand group self-definitions altered, again to bring them into a psychologi-\\ncally consistent relation (e.g., “we are the people who are making the world\\nsafe/pure/ . . . by subjugating/killing . . . them”). Finally, “social iden-\\ntity” moderates the link between “threat and appraisal” and destructive cop-\\ning.” Groups that have a socially shared sense of “us” as victim and “them”\\nas dangerous are more likely to cope with threat via intergroup violence\\nthan groups that do not have such a view of “us” and “them.”\\nThe figure is obviously simplified in several respects. First, the frame-\\nwork is from the perspective of only one group in relation to just one other\\ngroup. As White’s chapter in particular informs us, this is seldom the case in\\nreal-world conflicts, and the framework, if it is to prove useful, will need to\\nbe expanded to include additional groups beyond “us” and “them.” Second,\\nthere are many other instigators to intergroup violence than threat. Often\\none group aggresses against another for material gain because the group’s\\nleaders believe that they can get away with it. For example, the people of\\nWest Africa in the 17th through 19th centuries posed no threat to Ameri-\\ncans; slaves were introduced into the United States simply because Euro-\\n230\\nConclusion\\nfigure 9.1 Social Identity Framework for Intergroup Conflict\\n&\\n\\n\\npeans and other slave traders had more power than Africans, and American\\nplantation owners wanted cheap labor. Third, threat, even if it is linked\\nclearly to some outgroup, does not always lead to violence. Fourth, there\\nare many inputs into intergroup conflict in addition to threat and appraisal\\nand social identity. At the societal level, type of economic system and politi-\\ncal structures are important, and, at the individual level, personality and so-\\ncial attitudes are likely significant contributors to group-based harm-doing.\\nFifth, the figure suggests a static view of the links among the variables de-\\npicted. This is not the case. As Staub’s chapter makes very clear, there is a\\nthrough time path to mass killing and genocide. Thus, our figure stands for\\nmany causal paths that unfold through time. For example, a group might\\nexperience a large influx of immigrants to their country as an economic\\nthreat (“They will take our jobs”); this may, in turn, lead to discriminatory\\nlaws and outbreaks of violence against that group; if continued over time,\\nthese actions will likely come to be prominent parts of views of us and them;\\nwhether this would escalate to widespread and systematic harm-doing such\\nas genocide would depend on threat and appraisal and social identity ac-\\ncording to figure 9.1. \\nAdvantages of Threat/Coping Framework \\nThere are four major advantages to using a general threat/coping paradigm.\\nFirst, as has been demonstrated at the individual level, unpredictable and\\nuncontrollable threats are experienced as particularly stressful, and such\\nthreats often lead people to seek the comfort and security of the ingroup\\n(Worchel, 1999). For most people, the world today is characterized by un-\\npredictable and uncontrollable forces and events—modernization in many\\ndeveloping countries and globalization of the economy for all nations (Erik-\\nsen, this volume), rapid technological change, and often unplanned social\\nchange. Thus, the model suggests that intergroup strife will be more, not\\nless, likely as we enter the 21st century, even though much of the globe is\\nexperiencing considerable material progress. \\nSecond, as stress and health workers have demonstrated, it is fruitful to\\ndistinguish problem- from emotion-focused coping. In the former, the per-\\nson, upon encountering a situation that signals current or future harm, at-\\ntempts to deal directly with the source of threat (e.g., upon learning that she\\nhas high blood pressure, which it is explained to her increases the risk \\nof health problems, the person changes her diet to reduce salt and fat).\\nEmotion-focused coping, on the other hand, involves attempts to reduce\\nthe fear and anxiety that accompany threats. Thus, using the same example,\\nthe newly diagnosed hypertensive individual could drink alcohol to avoid\\nthinking about the mysterious and scary condition about which she has just\\nlearned. As applied to intergroup conflict, the 1929 Stock Market Crash\\nushered in a worldwide depression. In the United States, one response was\\nthe set of political measures termed the New Deal, which among other\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n231\\n\\n\\nthings sought to provide assistance for the economically afflicted. This\\nproblem-focused coping can be contrasted with the rise of Nazism in Ger-\\nmany at about the same time. Hitler’s programs sought more to alleviate\\nfeelings of despair and national impotence following the loss of World War\\nI and the economic and political turmoil of the Weimar years, and these led\\nto the “Final Solution” and the Holocaust.\\nThird, there are clear individual differences in styles of coping, and the\\npreceding chapters suggest that there are likely reliable societal and group\\ndifferences in styles of coping as well. Thus, for example, Staub notes that\\nsocieties differ in degree of authority orientation. It is likely that those high\\nin authority orientation will more likely accept violence-against-outgroup\\n“solutions” for threats offered by leaders even if they contradict societal\\nnorms about harm-doing.\\nFourth, and very important for the present analysis, there is evidence\\nthat self and identity are implicated in the threat-coping process (Contrada\\n& Ashmore, 1999). Thus, there is reason to believe that an identity-inter-\\ngroup conflict model based on a threat-coping framework may be fruitful.\\nFramework Details and Implications\\nDestructive Coping: Intergroup Harm-Doing.\\nAs depicted in figure 9.1, our\\nmodel seeks to account for intergroup harm-doing, by which we mean\\nharm by members of one group to those of another group based on group\\nmembership. As suggested in the figure, we see intergroup harm-doing as\\nmultidimensional. Intergroup violence can involve physical and/or psycho-\\nlogical hurting. For example, during the slavery and Jim Crow eras in the\\nUnited States, African Americans were subjected to both physical violence\\n(e.g., beatings, lynchings, poor housing) and psychological violence (e.g.,\\nsocietal stereotypes of innate inferiority and racial etiquette that made in-\\nterracial contacts demeaning for blacks). T\\noday the balance is much shifted\\ntoward psychological harm-doing (persisting cultural stereotypes), though\\nphysical violence against blacks has not been eliminated. \\nAnother important dimension is the degree to which harm-doing is sys-\\ntematized. At one extreme, members of one group engage in local and\\nspontaneous acts of violence against outgroup individuals. At the other ex-\\ntreme, there are elaborate societal systems for hurting the other group. In\\nNazi Germany, there was an organized set of institutions for carrying out\\nthe “Final Solution.” It is not just open violence that can be highly system-\\natized. The slavery system in the United States, the caste system in India,\\nand apartheid in South Africa all were undergirded by complex sets of \\nsocial institutions (legal, educational, and the like) that oppressed certain\\ngroups. \\nThere are two other important dimensions of intergroup harm-doing\\nnot depicted in the figure: (1) extent—how widely across the other group is\\nthe violence extended? and (2) duration—how long has the harm-doing\\n232\\nConclusion\\n\\n\\nbeen going on? As both increase so too does the strength of the links of in-\\ntergroup harm-doing to social identity and threat. \\nThreat and Appraisal. According to our model, a significant contributor to\\nintergroup harm-doing is threat and appraisal. Threat and appraisal are\\nhighly related but separable concepts. Just as work on social problems has\\nprofited from a distinction between actual conditions in the social environ-\\nment and the public “seeing” and labeling of these as problematic (Thomp-\\nson & Fine, 1999), we too feel it is useful to distinguish the objective threat\\nand the individual and collective recognition and interpretation (or ap-\\npraisal) of this condition (see also Northrup, 1989). Thus, for example, dur-\\ning the 1860s many Chinese immigrated to the United States to work on\\nthe transcontinental railroad. The societally projected image of Chinese\\nduring this period was relatively benign. However, when the railroad was\\ncompleted and many Chinese became competitors for jobs, unions and\\nother groups promoted public representations (in newspapers and the like)\\nof the Chinese as dangerous, and these came to be shared widely enough\\nthat local discrimination occurred and national immigration restrictions\\nwere implemented. Thus, it was not the sheer numbers of Chinese that\\nmattered or even the economic competition with whites, but the shared so-\\ncial representation of Chinese as dangerous. \\nThreat, like harm-doing, is viewed as multidimensional. Figure 9.1 iden-\\ntifies three important dimensions: material, symbolic, link to other group.\\nMaterial and symbolic threat are often treated as opposite ends of a single\\ncontinuum. We believe, however, that it is best to consider these as poten-\\ntially independent dimensions. As noted above, material threat is the core of\\nrealistic group conflict (RGC) approaches to intergroup conflict (Ashmore\\n& Del Boca, 1976). These are often contrasted with “symbolic politics” ap-\\nproaches (see Citrin et al., this volume; Sears et al., 1999). In symbolic poli-\\ntics (SP) models, groups come into “unrealistic” non-self-interested conflict\\nover symbolic issues. For example, Whose language should be spoken in\\nthis country? Whose history should be taught in our schools? Our view is\\nthat the conflict between the RGC and SP positions is more apparent than\\nreal. Material things, especially land, often acquire symbolic meaning. And\\nsymbols may have significant material implications. Continuing the same\\nexamples, deciding “whose language” will, in turn, partially shape which\\ngroup’s members can make the better living economically. Whose history is\\ntaught will legitimate and buttress the economic and political status of some\\ngroups and undermine that of others. Thus, even symbolic conflicts have\\nimportant material implications.\\nIn his contribution to this volume, Kelman stresses that the Israeli-Pales-\\ntinian conflict has both material, or “practical,” and symbolic facets, and\\nboth of these need to be addressed in attempts to reduce the conflict. The\\nother chapters in this book tend to emphasize either material or symbolic\\nthreat. As noted above, Brewer’s primary argument is in terms of symbolic\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n233\\n\\n\\nthreat, whereas Eriksen stresses material conflicts. Citrin and his colleagues\\nwere guided by symbolic politics theory; Sidanius and Petrocik’s group\\ndominance model is, at its core, a materialist framework—group dominance\\nin the service of material outcomes. \\nThe third dimension of threat depicted in the figure is link to the other\\ngroup. Some threats directly and clearly involve “them.” Some examples:\\nChinese competing with whites in the United States after the completion of\\nthe transcontinental railroad, Hutus and Tutsi attacking one another in\\nRwanda (Staub, this volume), and laws in India designed to set aside jobs for\\nthe lower castes (Eriksen, this volume). At the same time, intergroup harm-\\ndoing can be triggered by threats that do not objectively arise from the\\nother group. In Germany in the 1920s, the hyperinflation and other eco-\\nnomic problems did not arise because of the actions of Jews. Hitler identi-\\nfied this group as a cause, however, and this appraisal was widely accepted\\ngiven the longstanding societal representations of Jews as sneaky and dan-\\ngerous (Goldhagen, 1996). Examples of types of threats that do not directly\\ninvolve the outgroup that yet can start a cycle of intergroup harm-doing are\\neconomic instability and hardship (poverty, globalization that fractures \\ntraditional economic practices and procedures, inflation), political insta-\\nbility and corruption (changing governments, lack of societal agreement on\\npolitical processes, widespread graft by public officials), and rapid social\\nchange (large-scale movements from rural to urban areas, changes in family\\nroles, and other social changes associated with “modernization”). \\nIn sum, our conceptualization of threat underlines two major points.\\nFirst, threats can be about tangible things (material threat) and about intan-\\ngibles as represented in valued symbols (symbolic threat). Further, a given\\nthreat can have both a material and symbolic component. Second, conflict\\nbetween groups can begin with factors that are not objectively related to the\\ngroups at all. Quite simply, some nongroup-related threats can initiate the\\nset of processes that lead to intergroup violence. In these cases, leaders di-\\nrect appraisal to “them” as the cause and away from the economic, political,\\nand social factors, leading individuals to see themselves and their groups as\\nin danger. In such cases, groups cope with the emotions aroused by the\\nthreat rather than with the threat itself.\\nThere are two other important dimensions of threat that are not de-\\npicted in the figure, and these parallel those for intergroup harm-doing: (1)\\nextent—how widely across the ingroup is the threat experienced? and (2)\\nduration—how long has the threat been in existence? As both increase, so\\ntoo do the links of threat to intergroup harm-doing. \\nAppraisal is necessary for threat to lead to collective group-based vio-\\nlence. Although the figure suggests that this is a static phenomenon, it is\\nnot. Instead, appraisal unfolds over time. T\\no capture this idea we have iden-\\ntified both primary and secondary appraisal. Primary appraisal is an initial\\nmaking sense of the threat: What is the problem? What or who is responsi-\\nble? Thus, for example, when a society with many foreign workers (as is\\n234\\nConclusion\\n\\n\\ntrue of many nations) encounters an economic downturn, it might be ex-\\nplained relatively spontaneously as “them taking jobs from us,” and this\\ncould lead, without much thought or organization, to isolated attacks on\\nthe foreign workers.\\nHowever, for violence to be systematic and widespread requires mobiliz-\\ning a larger segment of society and providing more organized paths to ac-\\ntion. This involves secondary appraisal guided by both group and individual\\nsocial identities as indicated in the figure. Regarding the former, group\\nidentities must first be invoked to make sense of the threat. As Brewer,\\nEriksen, and Staub note, leaders are crucial in this process (see also\\nWorchel, 1999). At least initially, these leaders are not likely to be national\\npolitical leaders but instead proponents of “fringe” groups, labor leaders, or\\nlocal religious figures. And, in an early coping stage, the efforts they direct,\\nthough more organized than a spontaneous attack or a “race riot,” are still\\nrelatively reflexive.\\nBut what makes individual citizens ready to accept a group-based (“them\\nagainst us”) explanation for their troubles? Eriksen’s chapter points us in the\\nright direction when he notes that the shared group social identities prof-\\nfered by leaders must be consonant with the social and personal identities of\\nindividuals and also fit with their daily experience and enactment of these\\nselves. This melding of a group identity with the identities of individuals\\ncan result in mobilization of substantial numbers of individuals for action,\\nin this case a second stage of harm-doing.\\nAs actions against the outgroup develop and widen, the ingroup engages\\nin secondary appraisal where the “guilt” of the outgroup is solidified in so-\\ncietally projected images, and harm-doing is explained as a necessary or\\nperhaps even a noble activity.\\nIf mobilization is to be enduring, individuals will have to accept some\\npersonal costs. This raises the crucial question: Under what circumstances\\nwill individuals and groups be willing to endure personal hardship to either\\ninflict harm or to experience harm? Stated somewhat differently, what are\\nthe factors that determine whether threat leads to mobilizing individuals\\nand the group to undertake action? As figure 9.1 suggests, we believe that\\npart of the answer lies in the content and structure of both group and indi-\\nvidual social identity. With regard to the group narrative, a national or eth-\\nnic story that features being a victim will increase the likelihood of mobi-\\nlization, and this will be true if there is also the sense of group vulnerability\\n(Eriksen, White, Staub, this volume). At the individual level, we suggest\\nthat the relative lack of valued personal selves makes a society susceptible to\\ngroup-based mobilization. Thus, for example, in a nation in which men and\\nwomen are not able to successfully enact work and family roles because of\\npoverty, discrimination, and the like, they will be more likely to be suscepti-\\nble to social identities that involve outgroup violence or support others \\ndoing harm.\\nAn important implication of this line of reasoning is that protracted con-\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n235\\n\\n\\nflicts create social identities, at both the group and individual level, that\\nmake the progression to widespread and systematic violence more likely.\\nHarming others based on group membership or standing by while others\\ndo this changes people. Cognitive consistency models suggest that people\\nand groups would develop personal and shared reasons that justify this \\nbehavior. These have included belief systems specifying that they are not\\nhuman (and, thus, norms about hurting others do not apply), are inferior\\nhumans (and, thus, need to be taken care of as slaves), or are powerful and\\ndangerous (and, thus, must be killed before they destroy us). These ratio-\\nnalizations become part of transformed group and individual identities \\nas depicted by the arrow from intergroup harm-doing to social identity in\\nfigure 9.1. Given the focus of this volume on social identity, we turn now to\\na more detailed analysis of the group and individual identity portion of the\\nscheme.\\nSocial Identity. As indicated at the top of figure 9.1, and suggested at several\\nplaces above, social identity is here viewed as a significant input to the\\nthreat-intergroup harm-doing relationship. In the proposed framework, we\\ndistinguish group from individual social identity. \\n• Group Social Identity. Group social identities are group-level phe-\\nnomena. This means that both the referent and locus of the identity is \\nthe group. By referent we mean what the identity is about, the group—\\nAmerica/Americans; Serbia/Serbians, and so on. The content of the iden-\\ntity is the set of things that are thought to “define,”—for example, America\\nand Americans, including geography (e.g., 50 states, mostly between\\nCanada and Mexico), politics (e.g., democracy, two-party system), econom-\\nics (e. g., capitalism), morals and values (e.g., caring, freedom), history (e.g.,\\nrevolution, civil war), and the like. By locus of the identity, we mean that\\nthe group is also the level of representation of the identity. Such representa-\\ntion takes both psychological and socio-structural form. Regarding the for-\\nmer, labeled “Social Representations” in the figure, the group identity is a\\nset of beliefs that are widely accepted in the society, even though many indi-\\nviduals do not know or endorse all of the beliefs (Thompson & Fine, 1999).\\nThe socio-structural form of group identity, “Societally Projected Images”\\nin figure 9.1, is manifested in two primary and intertwined ways: (1) it is\\nwoven into the political, economic, and social structure of the society, often\\nin implicit ways (e.g., institutional racism in the United States [ Jones,\\n1997]); (2) it is conveyed by the societal mechanisms of socialization such as\\nschools and the media (e.g., cultural racism in the United States [ Jones,\\n1997, Worchel, 1999]). \\nIn addition to the rather static group social identity notion (“We are\\n_____”), we propose that there is a group-level analog to life story and per-\\nsonal narrative at the individual level (McAdams, 1997; Gergen & Gergen,\\n1988). That is, just as individuals develop stories that specify not only what\\n“I am like now” but also “how I got here” and “where I will likely go in the\\n236\\nConclusion\\n\\n\\nfuture,” we propose that groups develop narratives that tell the story of\\ntheir group. This proposed construct is explicit in only one of the foregoing\\nchapters; Kelman makes national narrative a central part of his analysis, but\\nhe draws no link to life story or self-narrative at the individual level. The\\nnotion of group narrative, however, is implicit in several other contribu-\\ntions. For each of the three case studies of intergroup conflict described by\\nEriksen, a group narrative was propagated by political leaders who identi-\\nfied “us” as noble and deserving and “them” as dangerous (e.g., in Bosnia\\nthe Serbian national narrative featured past defeats at the hands of powerful\\noutsiders, especially the Turks, and the local Muslim community was de-\\nfined as representing external invaders even though this definition was inac-\\ncurate). In describing the role identities of Republicans, White emphasized\\nthe importance of a family history of involvement in the cause and the fact\\nthat this history was often passed down from one generation to the next.\\nWe suggest that, as one generation of Republicans socialized the next, they\\nnot only told the story of “our family’s” involvement but also told the story\\nof “We Republicans,” the narrative of this group. \\nThere are four additional important points about group narrative:\\n1. Group stories can lie relatively dormant for periods of time. For\\nexample, in Yugoslavia during the Tito period, the divergent group\\nnarratives of the Serb, Croat, and Muslim communities were not in\\nthe foreground of societally projected images or social representa-\\ntions of the citizenry.\\n2. At certain times and under certain conditions, there are forces\\npushing for demonstrating and featuring a national narrative.\\nThreat, in the form of societal crises or outgroup actions, can\\nbring group stories to prominence and political leaders can use\\nsuch narratives (Brewer, Eriksen, this volume).\\n3. Group narratives can change. As Kelman noted, history is not\\nfixed, and, thus, groups can alter their story (Rothman, 1997).\\n4. Group narratives often contain folk theories of ethnicity, and these\\ngenerally do not match scientific understanding of this concept. As\\nEriksen notes, most current-day anthropologists reject the notion\\nthat ethnicity is primordial, and yet it is precisely this version of\\nethnicity that activists and politicians use in mobilizing people to\\naction on the basis of ethnicity.\\n• Individual Social Identity. At the individual level, identity components\\nare generally partitioned into social identities and personal identities (Ash-\\nmore & Jussim, 1997b), and we follow that convention here even though\\ndrawing the line between personal and social is not always easy (Deaux,\\n1992). Social identities are those self definitions that identify self in terms\\nof, or in relation to, other people and groups (e.g., “I am a teacher” which\\nspecifies a role that I and other teachers enact in relation to other people\\nwho fill the role of student; “I am an American,” which specifies that I am \\nT\\noward a Social Identity Framework for Intergroup Conflict\\n237\\n\\n\\na member of a social category along with others). Personal identities are\\nself-definitions that identify me as a unique individual (e.g., “I am obedi-\\nent”). Personal identity also includes self-narrative (an organized story\\nabout me, again as an individual). \\nGiven the thrust of this volume, we feature social identity here, though\\nwe briefly describe below how individual-level personal identity might be\\nimportant to intergroup conflict. We follow Thoits and Virshup (1997) in\\npartitioning individual-level social identities into “we’s” (i.e., self as a mem-\\nber of a collective group [e.g., “I am an African American”]) and “me’s” (i.e.,\\nself as fulfilling a role [e.g., “I am a mother”]). \\n1. Individual social identity: “we selves.” In the application of social identity\\nto intergroup conflict, “we selves” are obviously important and have been\\nthe most often studied. Thus, in this volume, all authors have been con-\\ncerned with how individuals define self in relation to an ethnic or national\\ngroup. And, they have done this in a variety of ways: self categorization in\\nterms of an ethnic label such as black or white, felt closeness to one’s group,\\nbeliefs about one’s group and another group, and the like. This diversity \\nof approaches raises an important question for the next generation of work\\non social identity and conflict: what are the components of “we” social \\nidentities? \\nMost researchers regard self-definition in terms of collective social\\ngroups as multidimensional, but there is far from consensus on what these\\ndimensions are. T\\no illustrate, table 9.1 lists the distinctions and components\\nthat three research groups have proposed for ethnic identity, racial identity,\\nand cultural identity, perhaps the most often studied types of “we” social\\nidentities. On the same line are listed dimensions that are the same or\\nhighly similar. It can be seen that there are five areas of partial agreement\\non what are the major dimensions of “we” social identity. T\\nwo of the three\\narticles included: (1) salience/centrality; (2) evaluation and feelings/private\\nregard; (3) values and attitudes/ideology; (4) interest and knowledge/famil-\\niarity; (5) behaviors and practices/behavior and language.\\nAt the same time, no dimension is mentioned by all three authors. And,\\nthere are three components listed by just one author: (1) self-identification,\\n(2) commitment, (3) public regard. De la Garza et al. and Sellers and his as-\\nsociates may have omitted self-identification because they were proposing\\nscales to measure discrete dimensions of a specific social identification, and,\\nin so doing, they may have taken for granted that the first step in social\\nidentification is placing self in a social category. At the same time, it is likely\\nthat a comprehensive assessment of “we” social identity will require atten-\\ntion to self-identification, including such issues as preferred ethnic label and\\nthe possibility of self-subtyping. Just as perceivers discern subtypes within\\nage (Brewer, Dull, & Lui, 1981), gender (Clifton, McGrath, & Wick, 1976),\\nand racial categories (Devine & Baker, 1991), it is likely that people may de-\\nfine self in terms of such subtypes (e.g., Eriksen noted that Yugoslavians liv-\\ning in Belgrade and other urban areas had very different views of being a\\n238\\nConclusion\\n\\n\\nSerb, Muslim, or Croat than those in rural areas, and this may involve sub-\\ntyping). Regarding commitment, most would agree that something akin to\\nthe Eriksen/Marcia notion of the developmental path that the person took\\nto her or his current sense of group identification is a potentially important\\ncomponent of racial/ethnic/cultural identity. Cross’s (1995) work is a re-\\nlated but different approach to social identity as a developmental construct.\\nFinally, even though only Sellers et al. explicitly include a public regard di-\\nmension as part of social identity, it is likely that others would regard this as\\na potentially important dimension.\\nT\\nable 9.1 illustrates one other complication in conceiving “we” social\\nidentities: to what extent must the “other” group be included within the\\nconcept of collective social self-definition? De la Garza and associates are\\nvery clear that Latino cultural identity includes familiarity with and lan-\\nguage skills relevant to both the ingroup and the “other” group. Should as-\\npects of identification in terms of “them” routinely be included in assess-\\nments of social identities?\\nAs is apparent, table 9.1 suggests a large number of possible components\\nof “we” social identities. As researchers grapple with which are necessary to\\ninclude, it is important to consider construct validity. In the simplest terms,\\nwhere does collective social identity end and related variables (causes, cor-\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n239\\ntable 9.1 Dimensions of “We” Social Identity Proposed by Three Research\\nGroups\\nAUTHOR(S)\\nPhinney (1995)\\nde la Garza et al. (1995)\\nSellers et al. (1998) \\nCONSTRUCT\\nEthnic identity\\nLatino(a) cultural identity\\nAfrican-American \\nracial identity\\nDIMENSIONS\\nSelf-identification\\n—\\n—\\nSalience\\n—\\nCentrality\\nEvaluation of\\n—\\nPrivate regard\\nand\\nfeelings about\\ngroup\\n—\\n—\\nPublic regard\\n—\\nValues and attitudes\\nIdeology\\nInterest in\\nFamiliarity with\\n—\\nand\\nLatino(a) culture\\nknowledge of\\nAmerican culture\\ngroup\\nEthnic behaviors \\nBehavior \\n—\\nand practices\\nand\\nlanguage\\nSpanish proficiency\\nEnglish proficiency\\nCommitment\\n—\\n—\\n\\n\\nrelates, and consequences) begin? For example, is “interest in and knowl-\\nedge about” one’s ethnic group part of one’s social identity, or is it best seen\\nas an outcome predictable from identification? This question becomes even\\nmore difficult regarding behavior: are ethnic behaviors part of social iden-\\ntity or an outcome that can be used to assess the construct validity of social\\nidentity measures? \\nOne of the most important messages of the present volume is that even\\nconflicts that appear to be simple “us against them” on the basis of one so-\\ncial identity distinction (e.g., religion in Northern Ireland) also involve\\nmultiple other social identities. These other social identifications are not\\njust regarding ethnic group and nation, which have been the focus of the\\nforegoing chapters. For example, in Part II of the volume addressing ethnic\\nand national identity in the United States, there was evidence that gender,\\nsocioeconomic status, and religion were associated with ethnic and national\\nidentifications and with the intergroup attitudes and policy positions under\\ninvestigation. Thus, as highly social creatures, people have multiple identi-\\nfications with collectives, and these interrelate with one another in complex\\nways to shape orientations toward other groups and toward political issues.\\n2. Individual social identity: “me selves.” As White’s chapter reminds us,\\nhow people relate to their own and other groups depends not just on collec-\\ntive self-definitions (“we’s”), but also on thinking of self in terms of social\\nroles (“me’s”). As noted above, role identities pertinent to group conflict are\\nespecially likely to arise and be important in protracted conflicts where\\nmany “supporting” roles are needed to carry on the fight. Also as noted\\nabove, there is a second major way in which role social identities are impor-\\ntant to understanding intergroup conflict: to the extent that societal condi-\\ntions allow individuals to have and successfully fulfill valued role identities\\n(e.g., husband and father), it will be more difficult to recruit them for doing\\nviolence against others. On the other hand, when poverty, oppression,\\nmodernization, and other social forces make it difficult for large numbers of\\npeople to fulfill social roles, it makes it easier for them to be drawn to social\\nidentities that provide feelings of positive evaluation and potency even if\\nthey also explicitly or implicitly involve doing harm to others. \\nIn addition, the relation between social role organization and group\\nboundaries is likely to have important implications for group-based vio-\\nlence. Eriksen’s observation that Bosnians and Serbs had separate private\\nspheres means that they did not have role identities (husband-wife, in law,\\nand the like) across social category lines. Social roles organized largely along\\nin-group lines may make it easier to recruit people to intergroup harm-do-\\ning. Further, it is likely that friendships, too, were predominantly within\\ngroup. We suggest that cross-group friendship is likely to reduce inter-\\ngroup violence (Ashmore, 1970; Pettigrew, 1998). If I see myself as Richard\\n(Bosnian) friend of Mel (Serb) and we have extensive first-hand experience\\nwith one another, it will be harder to mobilize either of us for “us against\\nthem” identity conflict.\\n240\\nConclusion\\n\\n\\nPersonal Identity.\\nAlthough social identities are most pertinent to inter-\\ngroup conflict, we suggest that personal identifications (self-construals that\\nidentify me as a unique individual) are also likely important. We offer two\\nillustrative examples. First, although Staub treated “authority orientation”\\nas a societal-level construct, it is likely to have an individual-level represen-\\ntation also, with substantial individual differences even within a society high\\n(or low) in this orientation. T\\no the extent that I define myself as obedient to\\nauthority, I will be a more likely recruit for group-based violence fomented\\nby political leaders. Second, as noted above, we believe that self narratives\\nare important to identity at both the group and individual level. White’s\\nchapter suggests that self stories are most likely to be conflict-related in so-\\ncieties characterized by protracted conflict. White notes that many current-\\nday Republicans come from families with a long commitment to the cause\\nof freeing Ireland from England. Further, in these families, that tradition is\\na central part of the stories that parents and grandparents tell to children,\\nand, in turn, this history of one’s family’s involvement becomes part of the\\nindividual’s personal story. Thus, just as Kelman noted that the group nar-\\nrative can feature “We are the people who fight them,” so, too, at the indi-\\nvidual level, a central part of an individual’s narrative can be that “My family\\nand I are people who fight them.”\\nWhose Social and Personal Identities? It is useful to distinguish three general\\ntypes of actors in intergroup conflict situations:\\n1. Elites—individuals who have political, economic, or information\\npower. As Brewer, Eriksen, Staub, and Kelman note, elites play a\\npivotal role in social strife, often in propagating “us versus them”\\nsocial identities to mobilize citizens (Worchel, 1999).\\n2. Agents of violence—people who actively and directly carry out the\\nharm-doing (e.g., in White’s chapter, the Protestant and Republi-\\ncan paramilitaries and English soldiers).\\n3. “Ordinary _____s” (as per Goldhagen’s [1996] “ordinary Ger-\\nmans”)—general citizens or average group members. Whatever\\nthe merits of Goldhagen’s arguments, he has highlighted a crucial\\npoint, How do ordinary citizens become participants in organized\\nviolence between groups? Staub suggests that bystanders who “do\\nnothing” while organized violence is done will be touched, possi-\\nbly even taking on the identity of the hurting party.\\nCaveats\\nWe reiterate that the proposed framework is not a tight deductive theory,\\nbut instead a heuristic scheme intended to highlight the possible roles of\\nidentity in intergroup conflict. Thus, it is necessary to note two significant\\ncaveats.\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n241\\n\\n\\nOne of the most important caveats about the model is what is left out.\\nMany factors that contribute to intergroup violence are not depicted. As\\nnoted above, these omissions are made to underscore the importance of so-\\ncial identity. At the same time, as with any framework, our scheme hides as\\nwell as highlights. Even some factors that we feel are quite important (e.g.,\\nthe role of political, economic, and religious leaders in promoting identity-\\nbased conflict) are only indirectly represented in the figure (e.g., as soci-\\netally projected images under group social identity).\\nA second caveat concerns context. As depicted in figure 9.1, we have es-\\nsentially frozen in time a conflict between two groups and also wrenched\\nthat conflict out of its regional and global context. As is clear in the chapters\\nby Eriksen, Sidanius and Petrocik, White, and Kelman, however, many eth-\\nnic and national conflicts unfold over a relatively long time period, and, as\\nnoted above, early events profoundly influence the current situation\\n(Fisher, 1990; Northrup, 1989; Rothman, 1997). \\nWhite’s chapter highlights the “space/place” context caveat for our\\nmodel: In the case of Northern Ireland, England/Britain is an “outside”\\n(not “them” or “us” if one focuses just on the two groups in Northern Ire-\\nland) yet potent factor (and if White is correct, the “them” party for Repub-\\nlicans). Further, White notes but does not emphasize other important out-\\nsiders—the Republic of Ireland and the United States. The differing\\nresponses of the world community to the conflict in Yugoslavia and Rwanda\\nsuggest the importance of regional context, together with associated strate-\\ngic interests and belief systems (e.g., racism). The fighting in Kosovo, on\\nthe edge of Europe, led to a military campaign by NATO, whereas the\\ngenocide in Rwanda resulted in action from the world community that was\\n“too little too late” (New York Times, December 16, 1999).\\nIdentity and Conflict Reduction\\nGiven that we anticipate a future volume of the Rutgers Series on Self and\\nSocial Identity completely devoted to reducing intergroup conflict, here we\\ndiscuss just two issues important to defusing intergroup conflict: social psy-\\nchological mechanisms and types of interventions.\\nRegarding social psychological processes, several of the preceding chap-\\nters noted that getting people to accept higher order, superordinate identi-\\nties might be useful in ameliorating intergroup strife (Eriksen, Staub, Kel-\\nman). Brewer, too, raised this possibility, but also noted that according to\\nher optimal distinctiveness model and the present extension and elabora-\\ntion of it, overarching identity units could create conflict under some con-\\nditions. The possible utility of higher order identities for conflict reduction\\nis the direct focus of Gaertner and his colleagues. In multiple experiments,\\nthey have demonstrated that intergroup bias is significantly reduced when\\n242\\nConclusion\\n\\n\\nmembers of different groups adopt a superordinate or common group iden-\\ntity (e.g., Gaertner, Dovidio, Nier, Ward, & Banker, 1998)\\nPerhaps the furthest extension of the concept of superordinate categories\\nis that of “global culture.” The goal of global culture is to unite all people\\non the basis of characteristics shared as humans, thereby changing group\\nboundaries so members of currently different groups see themselves as be-\\nlonging to the same group (Worchel, 1999). However, it is important to\\nnote that global culture violates the basic need of identity uniqueness as\\nspecified by Brewer’s optimal distinctiveness model (see Brewer, this vol-\\nume). It also requires that individuals deny all previous group identities\\n(e.g., national, ethnic) in favor of the global identity. Thus, although the\\nnotion of global culture builds on research on higher order identities, it\\nconflicts with evidence that the benefits of superordinate identities occur\\neven when subordinate identities are also strong. These findings suggest\\nthat intergroup conflict reduction does not require that people give up their\\nethnic identity or other relatively specific collective social identities (Gaert-\\nner et al., 1998; Citrin et al., this volume).\\nAlthough the notion of higher order identities is appealing and Gaertner\\nand his co-workers have provided evidence for such an approach, we would\\nnote that the power of transcendent identifications is not just inside the\\nhead. Rather, we believe that it is action based on shared goals that is im-\\nportant. This was Sherif’s (1958) basic insight when he introduced the idea\\nof “superordinate goals”—goals that members of both groups desired but\\nthat could not be achieved by either group on its own; as a consequence the\\ngroups had to work together. Ashmore’s (1970) shared coping model took\\nSherif’s logic a step further by proposing that shared goals induced coop-\\nerative action and that such behavior promoted cross-group positive affec-\\ntive ties, especially friendships, and these emotional ties promoted stereo-\\ntype destruction, dissonance reduction, and the like that, in turn, produced\\nprejudice reduction (see also Pettigrew, 1998). The importance of cross-\\ngroup friendships as a mediator of intergroup contact fits with Eriksen’s\\nemphasis on lived experience. Just as the separate private spheres of Bosni-\\nans and Serbs provided fertile ground for religiously based hostility, so, too,\\ncould changes that promote cross-group friendship help in reducing con-\\nflict and making citizens of Bosnia more resistant to identity politics appeals\\nfrom leaders.\\nThe second issue related to reducing intergroup strife to be addressed is\\nimplementation of conflict reduction strategies. Here we discuss two ap-\\nproaches: educational programs designed to promote learning about other\\ngroups, and third-party interventions. Many school-based programs have\\nbeen designed to foster intergroup learning (Aboud & Levy, 1999; Gurin,\\n1999). Bilingual and multicultural educational programs and materials have\\nbeen shown to have positive effects on intergroup relations among students\\n(Bigler, 1999; Genesee & Gandara, 1999; Stephan & Finlay, 1999). The\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n243\\n\\n\\ngoal of these programs is to create an environment that allows groups to\\nmaintain their cultural identity and, at the same time, accept other groups\\n(Fisher, 1990). Through these programs, individuals may see that they are\\nless different from members of other groups than they thought (Stephan &\\nFinlay, 1999). The danger, of course, is that individuals may find learning\\nabout their group highly involving and satisfying, while ignoring or deni-\\ngrating lessons about other groups. If so, such curricula might unintention-\\nally magnify group boundaries and increase intergroup friction.\\nAnother approach to conflict reduction, designed for groups engaged in\\nactive, often violent conflict, is third-party intervention (e.g., Worchel,\\n1999; Rothman, 1997). Although the positive role of third-parties in reduc-\\ning conflict is most obvious in Kelman’s chapter in the present volume, it is\\nalso clear in the contribution by Staub. Fisher (1990) suggests that it is im-\\nportant to distinguish different types of third-party interventions, and pre-\\nsents a taxonomy of four types of third-party interventions that differ in\\nterms of whether or not the third-party focuses on the subjective (including\\nperceptions and evaluations of the conflict) and objective (including con-\\nflicts over both scarce resources [material] and over values [symbolic]) as-\\npects of the conflict. Third-party consultation, the type performed by Kel-\\nman and his colleagues, is seen as the most desirable because it involves a\\nclear emphasis on both the objective and subjective aspects of the conflict\\n(Fisher, 1990). In contrast, there are numerous examples of conflict inter-\\nventions carried out by other nations or aggregations of nations that focus\\nonly on objective facets of the conflict. Worchel (1999) suggests that such\\nthird-party interventions can sometimes save ethnic groups from persecu-\\ntion and destruction, but that they rarely lead to long-term peace. It is pos-\\nsible that these interventions are often unsuccessful, in part, because they\\nfail to adequately address the identity issues of the conflict. \\nThe chapters by Kelman and Staub raise an important question regard-\\ning third-party intervention: When should third parties intervene? At the\\nroundtable discussion at the conference associated with this volume there\\nwas clear agreement that early intervention was desirable, though Citrin\\npointed out that there is little political incentive for leaders to intervene un-\\ntil major conflict has emerged with clear evidence of harm-doing. Further,\\nRothman (1997) notes that negotiating too early in a conflict may instead\\nescalate identity conflicts.\\nSummary and Suggestions\\nIn this concluding chapter we both summarized the preceding chapters and\\noffered a conceptual framework that might guide future research.\\nPart I of the volume was, in part, concerned with the issue of the locus of\\nsocial identity—does social identification reside in the individual or in soci-\\nety? Brewer and Eriksen did not resolve this starting question, and, in fact,\\n244\\nConclusion\\n\\n\\nthese authors answered that social identity is both “in” society and “in” the\\nindividual. Thus, social identity, as with self and identity more generally, is\\nboth personal and social (Ashmore & Jussim, 1997b). Brewer and Eriksen\\nidentified the individual and societal building blocks of social identity, espe-\\ncially as they are linked to conflict between groups.\\nIn Part II we asked the contributors to address one type of conflict (po-\\nlitical strife) in one society (the United States). Not only did they share this\\ncharge, the authors also used a similar method, survey research. Yet given\\nthis shared topic and empirical approach, Citrin et al. and Sidanius and\\nPetrocik reached contrasting conclusions, the former seeing generally low\\nlevels of ethnically based political conflict and the latter concluding that \\nthe ethnic divide is wide and deep. As noted above, the difference in con-\\nclusions is more apparent than real, reflecting, in part, methodological and\\ninterpretational differences. The Part II authors leave us with many im-\\nportant ideas, including the multifaceted nature of ethnic and national \\nidentification.\\nIn Part III, White and Staub addressed violent conflict, the first in an\\nanalysis of Northern Ireland and the latter with a general model of the fac-\\ntors contributing to mass killing and genocide. These two chapters dovetail\\nwell in highlighting the importance of history and action for understanding\\nthe links between identity and conflict. White detailed the history of the \\nviolence in Northern Ireland and showed how the distant past was influen-\\ntial in shaping the most recent protracted fighting beginning in the late\\n1960s and is still much alive as the contestants seek to build on, and extend,\\nthe 1998 peace accord. Staub underlined that identities are not just in \\nthe head but, instead, both influence action and are shaped by action (and\\ninaction). \\nKelman, in the single Part IV chapter, described the problem-solving\\nworkshop approach that he and his colleagues have developed to ameliorate\\nprotracted conflicts. In so doing he underlined the central role that national\\nidentity plays in maintaining such conflict, and, at the same time, he offered\\nthe very optimistic message that national identity as a social construction\\ncan be altered.\\nThe second part of this concluding chapter was devoted to a “look\\nahead.” We presented a social identity framework for intergroup strife.\\nThis scheme, depicted in figure 9.1, identifies three major constructs: threat\\nand appraisal, destructive coping: intergroup harm-doing, and social iden-\\ntity. It is proposed that these are mutually causally linked and further that\\nsocial identity moderates the threat and appraisal to intergroup harm-doing\\nlink.\\nWe offer three general suggestions for theorists and researchers under-\\ntaking the next generation of work on social identity and intergroup con-\\nflict. First, we believe it will yield dividends if intergroup relations re-\\nsearchers looked to the concepts and methods of the general field of “self\\nand social identity.” Except for White’s use of role identity models, all the\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n245\\n\\n\\nchapters in this book were guided by social identity perspectives specifically\\nintended for understanding intergroup relations (e.g., T\\najfel, 1981; Turner,\\nHogg, Oakes, Reicher, & Blackwell, 1987). There is, however, a broad in-\\nterdisciplinary field concerned more generally with the issues of self and so-\\ncial identity (Ashmore & Jussim, 1997a). We urge scientists interested in in-\\ntergroup conflict to look more systematically at this field. We illustrated\\nwhat we have in mind above in linking “national narrative” as described by\\nKelman to the individual-level concept “life story” (McAdams, 1997) and\\npersonal narrative (Gergen & Gergen, 1988).\\nSecond, we suggest that those interested in the interface between iden-\\ntity and group conflict consider a broader range of conflicts than here ad-\\ndressed. Because we anticipate a future volume directed at identity and gen-\\nder, we did not include female-male relations in this volume. Yet gender\\nand other forms of intergroup conflict are likely to share some features with\\nethnic and national conflict but also to have some unique facets.\\nThird and finally, we hope that the general framework we have proposed\\nwill generate both empirical work and conceptual debate. Our broader\\nhope is that the heuristic scheme, and the volume as a whole, will spur oth-\\ners to develop a more comprehensive and coherent social identity paradigm\\nfor intergroup conflict\\nNote\\n1. History is also important for knowing the “objective” ways in which present\\nday material realities and symbols are constructed. For example, the Rwandan geno-\\ncide was made more likely by the distant and recent past of Rwanda: Several cen-\\nturies ago, the indigenous Hutu were subdued by Tutsi invaders (Smith, 1998), who\\nwere physically different (e.g., taller); the Tutsis implemented a feudal system that\\nsubjugated the numerically superior Hutus and propagated a value system that ex-\\nalted Tutsi physical attributes. Early in the twentieth century, colonial Belgians used\\nthese economic and value systems to help maintain order and their control over\\nboth Hutus and Tutsis. In addition, they added the notion of “blood” to explain\\ngroup differences. These historical “facts” shape today’s material and symbolic \\nrealities that guide action.\\nReferences\\nAboud, F. E., & Levy, S. R. (1999). Are we ready to translate research into pro-\\ngrams? Journal of Social Issues, 55, 621–625.\\nAretxaga, B. (1993). Striking with hunger: Cultural meanings of political violence in\\nNorthern Ireland. In K. B. Warren (Ed.), The violence within: Cultural and political\\nopposition in divided nations (pp. 219–253). Boulder, CO: Westview Press.\\nAshmore, R. D. (1970). Prejudice: Causes and cures. In B. E. Collins (Ed.), Social\\npsychology: Social influence, attitude change, group processes, and prejudice. Reading,\\nMA: Addison-Wesley.\\n246\\nConclusion\\n\\n\\nAshmore, R. D., & Del Boca, F. K. (1976). Psychological approaches to understand-\\ning intergroup conflict. In P. A. Katz (Ed.), Towards the elimination of racism\\n(pp. 73–123). New York: Pergamon Press.\\nAshmore, R. D., & Jussim, L. (Eds.) (1997a). Self and identity: Fundamental issues\\n(Vol. 1). New York: Oxford University Press.\\nAshmore, R. D., & Jussim, L. (1997b). T\\noward a second century of the scientific\\nanalysis of self and identity. In R. D. Ashmore & L. Jussim (Eds.), Self and iden-\\ntity: Fundamental issues (Vol. 1, pp. 3-19). New York: Oxford University Press.\\nBigler, R. S. (1999). The use of multicultural curricula and materials to counter\\nracism. Journal of Social Issues, 55, 687–705.\\nBoskin, J. S. (1970). Sambo. In G. Nash and R. Weiss (Eds.), The great fear: Race in\\nthe mind of America. New York: Holt, Rinehart and Winston.\\nBoulding, K. E. (1962). Conflict and defense: A general theory. New York: Harper.\\nBrewer, M., Dull, V. & Lui, L. (1981). Perceptions of the elderly: Stereotypes as\\nprototypes. Journal of Personality and Social Psychology, 41, 656–670.\\nClifton, A. K., McGrath, D., & Wick, B. (1976). Stereotypes of woman: A single\\ncategory? Sex Roles, 2, 135–148.\\nContrada, R. J., & Ashmore, R. D. (Eds.) (1999). Self, social identity, and physical\\nhealth (Vol. 2, pp. 3–20). New York: Oxford University Press.\\nContrada, R. J., Ashmore, R. D., Gary, M. L., Coups, E., Egeth, J. D., Sewell, A.,\\nEwell, K., & Goyal, T. (in press). Measures of ethnicity-related stress: Psycho-\\nmetric properties, ethnic group differences, and associations with well-being.\\nJournal of Applied Social Psychology.\\nCross, W. E. (1995). The psychology of Nigresence: Revising the Cross Model. In \\nJ. G. Ponterotto, J. M. Casas, L. A. Suzuki, and C. M. Alexander (Eds.), \\nHandbook of multicultural counseling (pp. 93–122). Thousand Oaks, CA: Sage \\nPublications.\\nDeaux, K. (1992). Personalizing identity and socializing self. In G. Breakwell (Ed.),\\nSocial psychology of identity and the self-concept (pp. 9–33). London: Academic Press.\\nde la Garza, M. F-O., Newcomb, M. D., & Myers, H. F. (1995). A multidimensional\\nmeasure of cultural identity for Latino and Latina adolescents. In A. M. Padilla\\n(Ed.), Hispanic psychology: Critical issues in theory and research (pp. 26–42). Thou-\\nsand Oaks, CA: Sage Publications.\\nde la Garza, R. O., Falcon, A., & Garcia, C. (1996). Will the real Americans please\\nstand up: Anglo and Mexican an-American support of core American political\\nvalues. American Journal of Political Science, 40, 335–351.\\nDevine, P. G., & Baker, S. M. (1991). Measurement of racial stereotype subtyping.\\nPersonality and Social Psychology Bulletin, 17, 44–50\\nFestinger, L. (1950). Informal social communication. Psychological Review, 57, 271–\\n282.\\nFestinger, L. (1954). A theory of social comparison process. Human Relations, 7,\\n117–140.\\nFestinger, L. (1957). A theory of cognitive dissonance. Stanford: Stanford University\\nPress.\\nFisher, R. J. (1990). The social psychology of intergroup and international conflict resolu-\\ntion. New York: Springer-Verlag.\\nFredrickson, G. M. (1999). Models of American ethnic relations: A historical per-\\nspective. In D. A. Prentice & D. T. Miller (Eds.), Cultural divides: Understanding\\nand overcoming group conflict (pp. 23–34). New York: Russell Sage.\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n247\\n\\n\\nGaertner, S. L., Dovidio, J. F., Nier, J. A., Ward, C. M., & Banker, B. S. (1999).\\nAcross cultural divides: The value of superordinate identity. In D. A. Prentice &\\nD. T. Miller (Eds.), Cultural divides: Understanding and overcoming group conflict.\\nNew York: Russell Sage.\\nGenesee, F., & Gandara, P. (1999). Bilingual education programs: A cross-national\\nperspective. Journal of Social Issues, 55, 665–685.\\nGergen, K. J., & Gergen, M. M. (1988). Narrative and the self as relationship. In \\nL. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 17–56).\\nNew York: Academic Press.\\nGoldhagen, D. (1996). Hitler’s willing executioners: Ordinary Germans and the holo-\\ncaust. New York: Alfred A. Knopf.\\nGurin, P., Peng, T., Lopez, G., & Nagda, B. A. (1999). Context, identity, and inter-\\ngroup relations. In D. A. Prentice & D. T. Miller (Eds.), Cultural divides: Under-\\nstanding and overcoming group conflict (pp. 133–170). New York: Russell Sage.\\nHull, C. L. (1943). Principles of behavior. New York: Appleton-Century-Crofts.\\nIyengar, S. S., Lepper, M. R., & Ross, L. (1999). Independence from whom? Inter-\\ndependence with whom? Cultural perspectives on ingroups versus outgroups. In\\nD. A. Prentice & D. T. Miller (Eds.), Cultural divides: Understanding and overcom-\\ning group conflict (pp. 273–301). New York: Russell Sage.\\nJones, J. (1997). Prejudice and racism (second edition). New York: McGraw-Hill.\\nLazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York:\\nSpringer.\\nMcAdams, R. C. (1997). The case for unity in the (post)modern self: A modest pro-\\nposal. In R. D. Ashmore & L. Jussim (Eds.), Self and identity: Fundamental issues\\n(Vol. 1, pp. 46–78). New York: Oxford University Press.\\nMcCall, G. J., & Simmons, J. L. (1966). Identities and interactions. New York: Free\\nPress. \\nNew York Times, 16 December 1999.\\nNorenzayan, A., Choi, I., & Nisbett, R. E. (1999). Eastern and Western perceptions\\nof causality for social behavior: Lay theories about personalities and situations. In\\nD. A. Prentice & D. T. Miller (Eds.), Cultural divides: Understanding and overcom-\\ning group conflict (pp. 239–272). New York: Russell Sage.\\nNorthrup, T. A. (1989). The dynamic of identity in personal and social conflict. In\\nL. Kriesberg, T. A. Northrup, & S. J. Thorson (Eds.), Intractable conflicts and their\\ntransformation (pp. 55–82). Syracuse, NY: Syracuse University Press.\\nPettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49,\\n65–85. \\nPhinney, J. S. (1995). Ethnic identity and self-esteem: A review and integration. \\nIn A. M. Padilla (Ed.), Hispanic psychology: Critical issues in theory and research\\n(pp. 57–70). Thousand Oaks, CA: Sage Publications.\\nRothman, J. (1997). Resolving identity-based conflict in nations, organizations, and com-\\nmunities. San Francisco, CA: Jossey-Bass.\\nRosenberg, S. (1997). Multiplicity of selves. In R. D. Ashmore & L. Jussim (Eds.),\\nSelf and identity: Fundamental issues (Vol. 1, pp. 23–45). New York: Oxford Uni-\\nversity Press.\\nSears, D. O. (1993). Symbolic politics. In S. Iyengar and W. J. McGuire (Eds.), \\nExplorations in political psychology (pp. 113–149). Durham: Duke University Press.\\nSears, D. O., Citrin, J., Cheleden, S., & Van Laar, C. (1999). Cultural diversity and\\nmulticultural politics: Is ethnic Balkanization psychologically inevitable? In \\n248\\nConclusion\\n\\n\\nD. A. Prentice & D. T. Miller (Eds.), Cultural divides: Understanding and overcom-\\ning group conflict (pp. 35–79). New York: Russell Sage.\\nSellers, R. M., Smith, M. A., Shelton, J. N., Rowley, S. A. J., & Chavous, T. M.\\n(1998). Multidimensional model of racial identity: A reconceptualization of\\nAfrican American racial identity. Personality and Social Psychology Review, 2, 18–39.\\nSherif, M. (1958). Superordinate goals in the reduction of intergroup conflict.\\nAmerican Journal of Sociology, 63, 349-363\\nSmith, D. N. (1998). The psychocultural roots of genocide: Legitimacy and crisis in\\nRwanda. American Psychologist, 53, 743–753.\\nStephan, W.G., & Finlay, K. (1999). The role of empathy in improving intergroup\\nrelations. Journal of Social Issues, 55, 729–743.\\nStryker, S. (1987). Identity theory: Developments and extensions. In K. Yardley &\\nT. Honess (Eds.), Self and identity: Psychosocial perspectives (pp. 89–103). New York:\\nWiley. \\nT\\najfel, H. (1978). Differentiation between groups: Studies in the social psychology of inter-\\ngroup relations. London: Academic Press.\\nT\\najfel, H. (1981). Human groups and social categories: Studies in social psychology. Cam-\\nbridge: Cambridge University Press \\nThoits, P. A., & Virshup, L. K. (1997). Me’s and we’s: Forms and functions of social\\nidentities. In R. D. Ashmore & L. Jussim (Eds.), Self and identity: Fundamental \\nissues (Vol. 1, pp. 106–136). New York: Oxford University Press.\\nThompson, L., & Fine, G. A. (1999). Socially shared cognition, affect and behavior:\\nA review and integration. Personality and Social Psychology Review, 3, 278–302.\\nTurner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Blackwell, M. S. (1987).\\nRediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.\\nWorchel, S. (1999). Written in blood: Ethnic identity and the struggle for human har-\\nmony. New York: Worth Publishers.\\nT\\noward a Social Identity Framework for Intergroup Conflict\\n249\\n\\n\\nThis page intentionally left blank \\n\\n\\nAborigines, 53\\nabstract communities, 60–61, 62\\naccentuation principle, 30\\nacculturation, 110–11\\nAct of Union (1801; Great Britain;\\n1801), 137, 144\\naffect, 72, 74–75, 81\\naffirmative action, 8, 93, 96\\nAfrican Americans, 11, 76, 78, 107,\\n126n.2\\ndisengagement by, 8\\nethnic group identification, 86–87,\\n89, 96, 117, 219\\nimmigration and, 104\\noppression of, 103, 125\\npatriotism and, 11, 83, 84–85, 95,\\n110, 119–24, 220\\npolicy preference issues, 93–94, 219\\npride in America, 83, 84\\nracism toward, 23\\nreligion and, 85, 219, 222\\nslavery and, 224, 230–31, 232\\nagents of violence, intergroup conflict\\nrole, 241\\naggression\\ndefensive, 162, 171\\nintergroup, 26–28, 35\\nSee also harm-doing; violence\\nAlbanians (ethnic), in Yugoslavia, 49\\nAlgeria, 195\\nAlien Act (1798; U.S.), 104\\naliens, resident. See resident-aliens\\naltruism, 29\\nAmericanization, of immigrants, 76,\\n105, 221\\nancestry\\nAmerican perspectives, 76\\nshared, 43, 215\\nanger, 33, 215\\nAnglo-Irish Treaty (1921), 137\\nantagonism\\ngroup-based, 65\\nideologies of, 170–71, 172, 176\\ntoward outgroups, 10, 17, 26–28, 32,\\n34, 36\\nanthropology, 7, 12\\nof identity, 43–47, 215\\npolitical, 62–66\\nanti-Semitism, 170, 234\\napartheid, 44, 194, 195, 232\\nappraisal, 229–31, 233–36, 245\\nArab-Israeli conflict, 188–189\\nCamp David accord (1978), 189\\nPalestinianization of, 189–91\\n251\\nIndex of Subjects\\n\\n\\nArab-Israeli conflict (continued)\\nSee also Isreal-Palestinian conflict;\\nPalestinians\\nArab-Israeli war (1967), 189\\nArab League, 189\\nArab nationalism, 188, 188\\nArabs, 133, 155, 188–89, 192, 203. See\\nalso Palestinians\\nArgentina, 178\\nArmenia, 172\\nArmenians, 8, 11, 171–72\\nAsian Americans, 11, 78, 86–87\\nChinese nineteenth-century immi-\\ngrants, 233, 234\\nethnic group identification, 117, 219\\nmulticulturalism and, 89\\nnational loyalty and, 102–3\\npatriotism and, 97, 119–21, 123–24\\npride in America, 83\\nassimilation\\nAmerican perspectives, 72, 76–77,\\n82–85, 92, 93, 105, 219\\nintragroup, 22, 28, 29, 214\\nof self, 27\\nSee also melting pot\\natrocities, 8. See also ethnic cleansing;\\ngenocide; mass killing\\nattachment\\ningroup, 17, 18, 20, 28, 29, 34, 35\\npsychological, 74, 81, 219\\nauthority orientation, 232\\ngenocide, mass killing, and, 162, 163,\\n167–68, 173, 177, 180\\npersonal identity and, 241\\nautonomous identities, 166, 167, 227\\navoidance, 32–33\\nAzerbaijan, 172\\nBahai, 179\\nBalkanization, 72, 76, 97\\nbasic needs, 160–64, 168, 180, 181, 226\\nBelgium, 170, 227, 246n.1\\nbetter world ideologies, 165\\nBharatiya Janata Party (BJP), 54, 55\\nbias, 7\\ningroup, 17, 18, 20, 24–27, 34, 36\\nBJP\\n. See Bharatiya Janata Party\\nblack Americans. See African Americans\\nblame, 33\\nblind patriotism, 75, 167, 178\\nBosnia-Herzegovina, 52, 53, 56–59, 237\\nethnic cleansing, 101\\nintergroup conflict in, 11, 49–50, 64,\\n187, 216\\nneed for healing, 171, 172, 175–76,\\n177\\nBosnians (people), 6, 8, 10, 49–50, 216,\\n240, 243\\nBritish mandate (Palestine), 188\\nBurundi, 59\\nbystanders, to violence, 162, 163,\\n173–74, 179–80, 181, 226, 227,\\n241\\nCambodia, 11, 165\\nCamp David accord (1978), 189\\nCanada, 58–59\\ncaring, inclusive, 178, 179\\ncaste system, India, 55–56, 57, 61, 232,\\n234\\ncategorization. See self-categorization;\\nsocial categorization\\nCatholics. See Roman Catholics\\nchauvinism\\nAmerican views on, 81–85, 92, 93, 95,\\n219, 222\\npatriotism vs., 74–75, 218\\nchildren, values development in, 178\\nChinese immigrants, American nine-\\nteenth-century hostility toward,\\n233, 234\\nChristians\\nAfrican American, 85, 219, 222\\nanti-Semitism and, 170\\nin Fiji, 51\\nin India, 54\\nin Yugoslavia, 49–50\\ncitizenship, 77\\nethnic definitions of, 101\\nCivil Rights Act (1964; U.S.), 77\\ncivil rights movements\\nin Northern Ireland, 138\\nin United States, 105, 220\\ncivil society, 177–78\\nclass, 48, 62\\nclass-based social movements, 63\\ncognitive balance, 22\\ncognitive dissonance, 214\\n252\\nIndex of Subjects\\n\\n\\ncollective identities, 56–61, 66, 144\\nIsraeli-Palestinian conflict and, 191,\\n194, 200–201\\nself and, 59, 139\\nYugoslavia and, 50\\nSee also group identity; social identity\\ncollectivist cultures, 34, 166\\ncommunal differences, national identity\\nand, 101–25, 220–21\\ncommunication, theories of, 214\\ncommunism, 49, 165\\ncommunities\\nabstract, 60–61, 62\\nimagined, 45, 60, 65\\ncomparison\\nintergroup, 19, 23–26, 27, 32, 214\\nmodernization as trigger for, 57\\nsocial, 24, 34, 214\\ncompensatory identities, 168–69\\ncompetition, 31, 65\\nfor resources, 24, 28, 56, 60, 217\\nsocial, 24–26, 35\\nconfidentiality, workshops and, 198–99,\\n200\\nconflict, 7, 8, 12–13. See also intergroup\\nconflict\\nconflict resolution\\nnational identity and, 187–211\\nworkshops on, 7, 12, 187, 190,\\n195–202, 210, 214, 227, 228, 245\\nconflicts of interest, 28, 34\\nconnected identity, 166–67, 181, 227\\nconquest, 3, 8, 27\\nconstructive patriotism, 167, 178\\nconstructivism, 44, 47, 53, 62, 66, 215\\nconsumerism, in India, 55\\ncontempt, 18, 19, 32–33, 215\\ncontent of identity, 72, 74, 75–76, 164,\\n218, 219\\ncooperation, 23, 28–30, 33–34, 35, 36\\ncoping\\ndestructive, 229–30, 232–33, 245\\nproblem– vs. emotion–focused,\\n231–32\\nshared, 243\\nstyles of, 232\\ncorruption, intergroup harm-doing and,\\n234\\ncourage, moral, 178–79\\ncriminalization (British policy), 151–\\n53\\nCroatia, 49\\nCroats, 48–50, 59, 133, 155, 177, 195,\\n237\\nCrusades, 11\\ncultural flows, 45–46\\ncultural identity, 238–39, 244\\nculture, 66\\nethnicity and, 43–46, 215\\nFiji, 52\\nglobal, 243\\nas intergroup conflict factor, 34, 46\\nYugoslavian issues, 48–51\\nSee also ethnic groups; genocide; mass\\nkillings\\nCyprus, 187, 193, 210\\nDalits (Untouchables caste), 55, 57\\nDeclaration of Principles (PLO and Is-\\nrael; 1993), 190\\ndefensive aggression, 162, 171\\ndemocratic societies, genocide and,\\n177–78\\ndestructive coping, 229–30, 232–33,\\n245\\ndevaluation of outgroups, 161–63,\\n170–73, 176, 177, 226\\ndifferentiation, 20–22, 29–33, 35, 36,\\n170\\ndiscrimination, 7, 10, 84\\napartheid policy, 44, 194, 195, 232\\nagainst outgroups, 18, 20, 23, 25,\\n26–27, 35, 36, 214–15\\nSee also prejudice\\ndisgust, 32\\ndistinctiveness, 34, 36\\nethnic, 72\\nof Israeli-Palestinian conflict issues,\\n191\\noptimal, 10, 21–22, 29–30, 32–33,\\n215, 242, 243\\npositive, 25, 26, 27\\ndistrust, outgroup, 30–36\\ndiversity\\nAmerican identity and, 71, 76–77,\\n82–83, 103–6\\nas conflict source, 102, 106–7\\nimmigration as source of, 92\\nIndex of Subjects\\n253\\n\\n\\ndominance\\ngroup, 106–9, 119, 124, 125, 220–21,\\n234\\nsocial, 11, 103, 106–10, 122, 124, 125\\nEaster Rising (1916; Ireland), 137, 138\\neconomic instability, intergroup harm-\\ndoing and, 163, 234\\neducation, 8, 61\\nabout conflict reduction, 243–44\\negalitarianism\\nin India, 55–56, 57\\nin United States, 95\\nEgypt, 188, 189\\nEgyptian-Israeli peace treaty (1979;\\nCamp David accord), 189\\nelites, intergroup conflict role of, 241\\nembedded identity, 166–67, 168, 178\\nemotion-focused coping, 231\\nemotions, 32–33, 215. See also specific\\nemotions\\nempathy, 23, 171, 175, 176\\nenemies\\nidentification of, 161, 162, 170–71\\nimage differentiation, 195–96\\nengagement of experiences, 175\\nEngland. See Great Britain\\nequality, social, 57\\nEritrea, 59–60\\nessentialism, 44, 45, 46, 65\\nEthiopia, 59–60\\nethnic cleansing, 4, 11, 27, 33, 49, 65,\\n101\\nethnic conflict, 59–60. See also inter-\\ngroup conflict\\nethnic grievances, 114, 118\\nethnic groups, 3, 7, 45\\nminority group identification with,\\n11, 86–87, 89, 95–96, 117, 219,\\n221–22\\nmulticulturalism and, 77\\nreciprocal attitudes among, 18–19\\nin United States, 71, 72, 76, 77–78,\\n86, 95–96, 103–8, 218–22\\nSee also specific ethnic groups\\nethnic identity, 3, 5, 6, 8–11, 72\\ncomponents of, 238–39\\nintergroup conflict and, 42–66, 215\\nmulticulturalism and, 77\\nin United States, 5, 10–11, 78, 86–87,\\n89, 93–96, 108–9, 116–18, 124,\\n219–22\\nSee also social identity\\nethnicity, 42, 62, 75, 215–16\\nculture and, 43–46, 215\\ngroup narratives and, 237\\nMarxist view, 48\\nmonolithic identity and, 193\\nsymbolic politics theory and, 78, 219\\nU.S. perspectives on, 76, 77, 86, 109,\\n119–22, 218–22\\nethnic pride, 72, 77\\nethnocentrism, 11, 29, 30, 44\\nAmerican perspectives on, 109, 110,\\n117, 122–24, 220\\ngeneralized, 112, 113, 117\\ngroup orientations, 19–28\\nhierarchical, 112–13, 122–23\\nSumner’s theory of, 17–19, 214\\nsuperiority feelings and, 168, 169\\nEuropean Community, 172\\nexclusionary patriotism, 107–8, 123–24,\\n125\\nexclusivist identities, 192–94, 201, 228\\nexperience. See personal experience\\nexploitation, 3, 10\\nFatah, 189\\nfavoritism\\ningroup, 17, 23–26, 34, 72, 75, 78,\\n214\\noutgroup, 25\\nfear, 32–33, 215\\nFiji, 6, 10, 42–43, 51–53, 56–58, 60, 61,\\n63, 64\\nFijians, 6, 10, 51–53, 56–58, 64\\nfission of groups, 64, 65\\nFrance, 169, 172, 195\\nfriendship, cross-group, 240, 243\\nfunctional theory, 18\\nfusion of groups, 64, 65, 87\\nGaza Strip, 188, 189, 190, 197, 203\\nGEI. See generalized ethnocentrism\\ngender, intergroup conflict and, 246\\ngender-based identity, 5, 166\\ngender-based social movements, 63\\ngender relations, in Fiji, 52\\n254\\nIndex of Subjects\\n\\n\\ngeneralized ethnocentrism, 112, 113,\\n117\\nGeneral Social Survey (1996), 79–86,\\n92, 115–16, 126n.2\\ngenocide, 3, 6–8, 10, 11, 213, 226, 231,\\n245\\ndefinition of, 159\\nethnic cleansing as, 4, 11, 27, 33, 49,\\n65, 101\\nHolocaust as, 11, 172, 173–74, 232\\nindividual and group identities in, 12,\\n159–81\\norigins of, 160–63\\nprevention of, 175–80, 181\\nGenocide Convention (U.N.), 159\\nGermany, 8, 169, 234\\nHolocaust and, 11, 172, 173–74, \\n232\\nNazism and, 165, 173–74, 179, 232\\nWeimar Republic and, 178, 232\\nWorld War II and, 172, 173–74, 179\\nYugoslavia and, 48\\nglobal culture, 243\\nglobalization, 43, 66, 71, 231, 234\\ngoals\\ncommon, 33–34, 243\\nsuperordinate, 9, 33, 243\\nGovernment of Ireland Act (1920;\\nGreat Britain), 137\\nGreat Britain\\nconflict in Northern Ireland and, 12,\\n134–38, 140–41, 144–45, 149–52,\\n154, 155–56nn.1–3, 180, 223–26,\\n242\\nFiji and, 51, 52\\nmandate over Palestine, 155, 188\\ngroup-based hierarchies, 107–8, 220\\ngroup dominance model, 106–9, 119,\\n124, 125, 220–21, 234\\ngroup identity, 3–5, 22, 74, 216\\nchanges to, 194, 201–2, 210–11, 227,\\n228\\ngenocide, mass killing, and, 159–65,\\n170–71, 178, 180–81, 226\\nin social identity framework, 230,\\n235, 236–37\\nSee also ingroups; outgroups; social\\nidentity\\ngroup integration, 63, 65\\ngroup narratives, 210, 228, 235, 236–37,\\n241, 246\\nIsraeli and Palestinian, 191–92, 202–9\\nSee also myths, national\\ngroups\\nintegration into, 63, 65\\nostracism from, 179\\nself-concepts of, 164, 168–70, 171,\\n174, 180, 226\\nGSS. See General Social Survey\\nGuatemala, 159\\nGulf War, 190\\nGuyana, 53\\nGypsies, 49, 173\\nharm, infliction of, 26–27, 214, 215, \\n226\\nharm-doing, intergroup, 3, 10, 229–36,\\n245\\nhate, 17, 18, 25, 27, 32–34, 36, 37n.1,\\n214–15\\nhealing, victimization and, 162, 163,\\n171–72, 175–77\\nHEI. See hierarchical ethnocentrism\\nhierarchical ethnocentrism, 112–13,\\n122–23\\nhierarchies, group-based, 107–8, 220\\nHindus, 6\\nin Fiji, 51\\nin India, 10, 54–61, 65\\nhindutva movement, 10, 54–58, 61, 65\\nHispanic Americans, 11, 78, 125\\nethnic group identification, 86–87,\\n89, 117, 219\\nnational loyalty and, 102–3\\npatriotism and, 83–84, 85, 97, 109,\\n111, 119–21, 123–24\\npride in America, 83, 84\\nSee also Mexican Americans\\nhistory, role of, 217, 223, 227, 245,\\n246n.1\\nHolocaust, 11, 172, 173–74, 232\\nJewish Palestinian homeland and, 188\\nHome Rule, Ireland and, 137, 149\\nhostility, outgroup, 17–19, 27, 28,\\n33–36, 37n.1, 75, 78, 97, 214, 216\\nhuman welfare, caring about, 178–79\\nHungarians, in Yugoslavia, 49\\nhunger strikes, 151–52, 224\\nIndex of Subjects\\n255\\n\\n\\nHutus, 59, 133, 155, 161–62, 170–71,\\n174, 234, 246n.1\\nidentification, 238\\ningroup, 17–36, 214–15\\nwith vs. as, 72, 73–74, 218\\nidentity, 12–13, 245\\nanthropology of, 43–47, 215\\nautonomous, 166, 167, 227\\ncollective. See collective identity\\ncompensatory, 168–69\\nconnected, 166–67, 181, 227\\ndefinitions of, 5\\nembedded, 166–67, 168, 178\\nethnic. See ethnic identity\\nexclusivist, 192–94, 201, 228\\ngender-based, 5, 166\\ngenocide, mass killing, and, 160,\\n161–81, 213\\ngroup-based, 3, 4, 5, 139. See also\\ngroup identity\\nindividual. See self\\ninsecure, 167\\nas intergroup conflict factor, 3, 4, 7–9\\nmonolithic, 192–94, 201, 228\\nnational. See national identity\\nnegation of, 192–93, 196, 201, 210\\nnegotiation of, 196–97, 200–202,\\n207–10\\nracial, 11, 238–39\\nrole, 135, 138–40, 144, 146–48,\\n151–54, 156n.6, 224–25, 237, 240,\\n245\\nsecure, 167, 181\\nsocial. See social identity\\nsocial construction of, 194–95\\nsuperordinate, 242–43\\nin threat-coping process, 232\\ntranscendent, 193, 201–2, 210–11,\\n243\\nidentity politics, 42–47, 53, 58–66, 215,\\n243\\nin United States, 4, 94–95, 97, 219\\nideological asymmetry, 124–25, 126n.4\\nideologies\\nof antagonism, 170–71, 172, 176\\nof better world, 165\\ngenocide, mass killing, and, 161–62,\\n165–66, 170–71, 173–74, 178, 181\\nmonolithic identity and, 193\\nnationalistic, 165\\nimagined communities, 45, 60, 65\\nimmigration, 231\\nto Palestine, 188\\nUnited States and, 8, 71, 76–77, 92,\\n96, 97, 104–5, 110, 233\\nImmigration and Nationality Act (1965;\\nU.S.), 77\\ninclusion, need for, 21–22, 29–30, 32,\\n33, 35, 36\\ninclusive caring, 178, 179\\nIndia, 6, 42–43, 60, 63\\ncaste system, 55–56, 57, 61, 232, 234\\nhindutva movement, 10, 54–58, 61,\\n65\\nMuslim population, 4, 54–58, 60, 61\\nnationalist movement, 74\\nIndians (Fiji people), 6, 10, 51–53,\\n56–58, 64\\nindividual-group discontinuity effect,\\n31\\nindividual identity. See individual social\\nidentity; self\\nindividualism, U.S. perspectives, 109,\\n111, 126n.1\\nindividualist cultures, 34, 166\\nindividual social identity, 230, 235, 236,\\n237–40\\ningroup positivity, 18–19, 20–23, 32, 36,\\n214\\ningroups, 10, 17–36, 214–16\\ncomparison and competition issues,\\n23–26, 34\\nethnocentrism and, 17–19, 29, 214\\nfavoritism toward, 17, 23–26, 34, 72,\\n75, 78, 214\\nformation of, 17, 19, 28–35\\noptimal distinctiveness theory, 21–22,\\n29–30, 32–33, 215\\noutgroup antagonism by, 26–28, 36\\npositivity toward, 18–19, 20–23, 32,\\n36, 214\\nscapegoating by, 161, 162, 164, 165,\\n180\\nsocial categorization and, 20, 214\\ninjured social identities, 12\\ninjustice, 57–58, 161\\ninsecure identities, 167\\n256\\nIndex of Subjects\\n\\n\\ninstigators of violence, 160–63, 173,\\n180, 181, 226, 230\\nInstitute for Social Research (University\\nof Michigan), 114\\nInstitute for Social Science Research\\n(UCLA), 79\\ninstrumentalism, 44, 45, 46–47, 53, 62\\nintegration, group, 63, 65\\ninteractive problem solving, 198\\ninterdependence, 33, 34\\ncooperative, 35, 36\\nobligatory, 29\\nintergroup comparison, 19, 23–26, 27,\\n32, 214\\nintergroup conflict\\ndefinition of, 6–7\\ndiversity as source of, 102\\nethnic and national identity as issues\\nin, 42–66, 215, 222\\ngender and, 246\\nidentity’s role in, 3, 4, 7–9\\ningroup identification and, 17–36,\\n214–15\\nIsraeli-Palestinian, 12, 187–211,\\n227–28\\nin multiethnic states, 101–2\\nmultilevel analysis of, 7, 9\\nin Northern Ireland, 133–55, 223–26\\nrealistic group conflict view of, 8,\\n216–17, 233\\nreduction of, 5, 7, 9, 12, 36, 242–44\\nself ’s role in, 3, 4\\nsocial identity and, 4, 5, 7, 8–12,\\n213–46\\nsocial identity framework for, 213,\\n229–42, 245\\nas violence instigator, 160–61, 180,\\n226\\nintergroup contrast, 22, 214\\nintergroup harm-doing, social identity\\nframework and, 229–36, 245\\nintergroup relations, 6, 11, 17, 18,\\n35–36\\nconflict-reduction strategies, 243–44\\nsocial categorization’s role in, 19–20\\nsocial identification and, 22–23\\nintermarriage, 57\\nin Fiji, 52\\nin India, 56\\nin United States, 105\\nin Yugoslavia, 50\\ninternational conflicts, 3, 187, 198. See\\nalso war; specific conflicts and wars\\ninterventions, as conflict-reduction\\nstrategy, 243, 244\\nintifada, 189, 191, 201\\nintolerance, 32\\nintragroup assimilation, 22, 28, 29, \\n214\\ninvasion, feelings of, 33\\nIRA. See Irish Republican Army\\nIran, 179\\nIreland, 136–38, 149, 155–56nn.1, 3,\\n223, 224, 242\\nhome rule, 137–38\\nSee also Northern Ireland\\nIrish Free State, 137–38\\nIrish Nationalists, 135, 137, 138, 141,\\n142, 144, 152, 224\\nIrish National Liberation Army, 155n.2\\nIrish Republican Army, 12, 133,\\n137–38, 140, 145, 151, 152,\\n155nn.1, 2, 223, 224\\nIrish Republicans, 134–37, 140–42,\\n144–49, 151–53, 223–26, 237\\nparamilitaries, 11, 133–35, 137–38,\\n140, 148–49, 151, 154,\\n155–56nn.2, 3\\nself-narratives, 241\\nIslam. See Muslims\\nIsrael, 133, 179, 187–211, 227–29, 233\\nCamp David accord (1978), 189\\ncreation of, 188, 189\\nDeclaration of Principles (1993), \\n190\\nnational identity concerns, 191–97,\\n200–201, 207–9, 214\\noverview of Palestinian conflict,\\n188–91\\nproblem-solving workshop participa-\\ntion, 7, 12, 198–202, 210\\nIsraeli-Palestinian conflict, 188–91\\nArab states and, 188–89\\nPalestinian refugee issue, 188–89,\\n192, 202–9, 228–29\\nPalestinian uprising, 189, 191, 201\\npeace moves, 189–90\\nSee also Arab-Israeli conflict\\nIndex of Subjects\\n257\\n\\n\\nJews, 8, 177, 179\\nanti-Semitism and, 170, 234\\nHolocaust and, 11, 172, 173–74, 232\\nin United States, 11, 104\\nZionism and, 188, 192, 196, 202, 204\\nSee also Israel\\nJoint Working Group on Israeli-Pales-\\ntinian Relations, 200, 202, 205–7\\nJordan, 188, 203\\njustice, 161\\nPalestinian sense of, 204, 207, 208,\\n209\\nKhmer Rouge, 165\\nkinship\\nethnicity vs., 57\\nidentity politics and, 58, 60, 61, 64\\nYugoslavian issues, 48, 50, 51\\nKorean War, 166\\nKosovars, 8\\nKosovo, 8–9, 48, 57, 58, 101, 155, 242\\nLACSS. See Los Angeles County Social\\nSurveys\\nlanguage\\nconflict over, 8\\nin Fiji, 51, 52, 53\\nin India, 54\\nmonolithic identity and, 193\\nin United States, 72, 82\\nin Yugoslavia, 48\\nLatino National Political Survey, 83\\nLatinos. See Hispanic Americans; Mexi-\\ncan Americans\\nleaders, role of, 175, 177, 235, 237, 242\\nLebanon, 203\\nliberalism, 76, 77, 78, 95, 218, 221\\nliberalization, in India, 55\\nlife conditions, 160, 161, 163, 166, 168,\\n180, 226\\nliterary studies, postcolonial movement\\nand, 44\\nLos Angeles County Social Surveys\\n(1994, 1995), 79, 83, 86–87, 221\\nLos Angeles County Social Surveys\\n(1997, 1998), 113–14\\nlove\\nof country, 74–75\\ningroup, 17, 18, 34, 214–15\\nLoyalists, Irish, 141, 142–44, 225\\nloyalty\\nidentity politics and, 58, 64\\ningroup, 28, 29–30, 33, 34\\nnational, 71, 75, 101–3\\npluralism and, 106\\nMacedonia, 49\\nMadrid Conference (1991), 190\\nMandal commission (India), 56\\nMaoris, 53\\nmarriage. See intermarriage\\nMarshall plan, 172\\nMarxism, 48\\nmass killing, 3, 6–8, 10, 11, 231\\ndefinition of, 159\\nindividual and group identities in, 12,\\n159–81, 226, 245\\norigins of, 160–63\\nprevention of, 175–80, 181\\nSee also genocide\\nmaterial threat, 233–34\\nMauritius, 53\\nmedia, 55, 61, 236\\nmelting pot, America as, 11, 92, 103,\\n104–9, 116–19, 122, 124, 220, 221\\nMexican Americans, 83–84, 109–11,\\n125\\nminimal group paradigm, 18\\nminority groups, 8\\nethnic identifications by, 11, 86–87,\\n89, 95–96, 117, 219, 221–22\\nmulticulturalism and, 77, 82\\nstate-sanctioned violence against, 3,\\n11\\nsubtle racism toward, 23\\nin United States, 71, 72, 77, 82–87,\\n89, 95, 96, 107, 219–22\\nSee also specific minority groups\\nmisattribution, 32, 215\\nmodernity, 44–45, 47, 65, 66\\nmodernization, 45, 234, 240\\nas conflict factor, 10, 56–57, 231\\nin Fiji, 53\\nmonolithic identity, 192–94, 201, 228\\nmonolithic societies, 162, 163, 177\\nMontenegro, 49\\nmoral courage, 178–79\\nmoral superiority, 10, 32–33, 34\\n258\\nIndex of Subjects\\n\\n\\nMughals, 55, 57\\nmulticulturalism, 42, 44\\nUnited States and, 71, 76–78, 81–83,\\n89, 92–93, 96, 103, 218\\nmultiethnic states, 75, 76, 78, 87, 218\\ncommunal and national identity in,\\n101–25\\nmultiple voices, 44, 45–46\\nmurder. See genocide; mass killing\\nMuslims\\nin Fiji, 51\\nin India, 4, 6, 54–58, 60, 61\\nin Yugoslavia, 49, 50, 52, 58, 59, 237\\nSee also Arabs\\nmyths, national, 6, 10\\nidentity politics and, 58, 61\\nYugoslavia and, 50, 216, 237\\nSee also group narratives\\nnarratives. See group narratives; self-\\nnarratives\\nNational Election Study (1992; 1996),\\n114–15\\nnational identity, 3, 5, 6, 8–11, 245\\nAmerican, 71–97, 101–25, 218–22\\nchanges in, 187, 194–96, 201–2,\\n210–11, 228\\ndefinition of, 191\\nintergroup conflict and, 42–66\\nIsraeli-Palestinian conflict issues and,\\n12, 187–211, 214, 227–28\\nnegotiation of, 196–97, 200–202,\\n207–10\\nNorthern Ireland hostilities and, 136\\nSee also myths, national; social iden-\\ntity\\nnationalism, 10, 42, 44–46, 65, 115\\nAnderson’s (Benedict) view, 60\\nArab, 188\\ncivic vs. ethnic, 75\\nIndian hindutva movement and, 54,\\n55\\nas integrating ideology, 87, 89\\npatriotism vs., 74–75\\nQuebecois, 58–59\\nnationalistic ideologies, 165\\nnationalist movements, 8, 74\\nNationalists, Irish, 135, 137, 138, 141,\\n142, 144, 152, 224\\nnationality, 48, 75\\nU.S. perspectives on, 76, 77, 86, 95,\\n105\\nNational Opinion Research Center, 79\\nnations, 3, 7\\ndefinition of, 74\\nethnic groups and, 44–45\\nloyalty to, 71, 75, 101–3\\nminority groups in. See minority\\ngroups\\nSee also specific countries\\nNative Americans, 76\\nnativism, 76–77, 82–85, 92, 218, 219,\\n221\\nNATO (North Atlantic Treaty Organi-\\nzation), 242\\nNazism, 165, 173–74, 179, 232\\nNdembu, 58\\nneeds, basic, 160–64, 168, 180, 181, 226\\nnegation of identity, 192–93, 196, 201,\\n210\\nnegotiation of identity, 196–97,\\n200–202, 207–10\\nNES. See National Election Study\\nNew Deal, 175, 231–32\\nNorthern Ireland, 7, 133–55, 187, 193,\\n210, 213, 242, 245\\nactivist accounts concerning, 140–51,\\n225–26\\nbystanders to violence in, 179–80\\nhistorical background of hostilities in,\\n136–38\\nhunger strikes in, 151–52, 224\\npolitical violence in, 133–55, 155n.1,\\n223–26\\nsocial and role identities in, 11–12,\\n135, 138–40, 143–44, 146–48,\\n151–54, 156nn.5, 6, 223–26, 237\\nNorway, 58\\nNuer, 61, 63\\nNuremberg trials, 172\\nobligatory interdependence, 29\\none-time workshops, 198–200\\noppression\\nof African Americans, 103, 125\\ninstitutionalized, 6\\noptimal distinctiveness, 10, 21–22,\\n29–30, 32–33, 215, 242, 243\\nIndex of Subjects\\n259\\n\\n\\n“ordinary” people, intergroup conflict\\nrole of, 241\\nOslo agreement (1993), 190, 197, 201\\nostracism, 179\\noutgroups, 10, 17–36, 214–16\\nAmerican perspectives on, 73\\nantagonism toward, 10, 17, 26–28,\\n32, 34, 36\\ndevaluation of, 161–63, 170–73, 176,\\n177, 226\\ndistrust of, 30–36\\nethnocentrism and, 18, 19, 214\\nfavoritism toward, 25\\nhatred of, 17, 18, 25, 27, 32–33, 34,\\n36, 37n.1\\nhostility toward, 17–19, 27, 28,\\n33–36, 37n.1, 75, 78, 97, 214, 216\\nintergroup comparison and, 23–26,\\n27, 32\\nscapegoating of, 161, 162, 164, 165,\\n180\\nsocial categorization and, 20, 214\\nsubtle racism toward, 23\\nPA. See Palestinian Authority\\nPakistan, 55\\nPalestine, 155, 188–90, 192, 197, 203,\\n208–9\\nPalestine Liberation Organization, 189,\\n190, 197\\nPalestinian Authority, 190\\nPalestinians, 179, 187–211, 227–29, 233\\nDeclaration of Principles (1993), 190\\nnational identity concerns, 191–97,\\n200–201, 207–9, 214\\nOslo agreement (1993), 190, 197, 201\\noverview of Israeli conflict with,\\n188–91\\nproblem-solving workshop participa-\\ntion, 7, 12, 198–202, 210\\nrefugee issue, 188–89, 192, 200,\\n202–9, 228–29\\nSee also Arab-Israeli conflict; Israeli-\\nPalestinian conflict\\npassivity toward victims, 173–74, 179\\npatriotism, 115\\nblind, 75, 167, 178\\nchauvinism vs., 74–75, 218\\nconstructive, 167, 178\\nexclusionary, 107–8, 123–24, 125\\nin United States, 10–11, 78, 81,\\n82–85, 92, 95, 97, 109–11, 119–24,\\n126nn.2–4, 219–22\\nPeace Now movement (Israel), 196\\npersonal experience, 47\\nidentity politics and, 58–61, 65, 66\\npersonal identity. See self\\nPLO. See Palestine Liberation Organi-\\nzation\\npluralism\\nin American society, 11, 103, 105–9,\\n118–19, 122, 124, 220, 221\\ngenocide and, 177\\npogroms, 27\\npolicy issues, U.S., 73, 78, 87–96, 219,\\n221\\nimmigration as, 76–77, 96, 104, 105\\nimpact of racism on, 77\\npolitical anthropology, 62–66\\npolitical groups, 3, 72, 139, 159\\npolitical instability\\ngenocide and mass killing origins\\nand, 163\\nintergroup harm-doing and, 234\\npolitical science, 7, 12\\npolitical violence, in Northern Ireland,\\n133–55, 155n.1, 223–26\\npoliticide, 159\\npoliticization, 34\\npolitics\\nAmerican ethnic group concerns\\nabout, 117–18\\nof identity, 4, 42–47, 53, 58–66,\\n94–95, 97, 215, 219, 243\\nof power, 34\\npositive distinctiveness, 25, 26, 27\\npositive self-concept, 169\\npostcolonialism, 44\\npower politics, 34\\npower structure, 10\\nprejudice, 7, 8, 10, 221, 243\\nagainst outgroups, 17, 23, 25, 36,\\n37n.1\\nSee also discrimination\\npride\\nin America, 10, 78, 79, 81, 83, 84, 95,\\n219\\nethnic, 72, 77\\n260\\nIndex of Subjects\\n\\n\\nprimary appraisal, 234–35\\nprimordialism, 44, 45, 76, 215, 237\\nprisoners of war, 166\\nproblem-focused coping, 231–32\\nproblem-solving workshops, 7, 12, 187,\\n190, 195–202, 210, 214, 227, 228,\\n245\\nProtestants (in Northern Ireland),\\n133–38, 140, 141, 151, 153, 154,\\n155–56n.3, 223–26\\nLoyalists, 141, 142–44, 225\\nparamilitaries, 11, 12, 133–35, 137,\\n138, 140–44, 149, 151–54,\\n155–56nn.1–3, 224–26\\nUnionists, 137, 141–42\\nProtestants (in Rwanda), 178\\nProvisional Irish Republican Army. See\\nIrish Republican Army\\nProvisional Sinn Féin. See Sinn Féin\\npsychology, 12\\npunishment, 175\\nQuebecois nationalism, 58–59\\nrace, policy preferences and, 93–95, 96,\\n219\\nracial identity, 11, 238–39\\nracism, 23, 77\\nRashtriya Swayamsevak Sangh, 54\\nrealistic group conflict, 8, 216–17, 233\\nreconciliation\\nidentity change as factor in, 194,\\n210–11\\nneed for, 171, 172, 175–77, 181\\nrefugees, Palestinian, 188–89, 192, 200,\\n202–9, 228–29\\nreification, 45, 46\\nrelational identities, 166\\nreligious groups, 7\\nin Fiji, 51, 52\\nin India, 10, 54–56, 57, 60\\nmonolithic identity and, 193\\nNorthern Ireland hostilities and,\\n11–12, 133–38, 140–44, 149, 151,\\n153, 154, 155–56nn.3, 4, 223–25\\nin United States, 71, 72, 104, 219,\\n222\\nin Yugoslavia, 48, 49–50\\nSee also specific religious groups\\nRepublicanism, 136–37, 140. See also\\nIrish Republicans\\nresident-aliens, in United States, 11,\\n119–22\\nresources\\ncompetition over, 24, 28, 56, 60, 217\\nas Israeli-Palestinian conflict issues,\\n187, 191–92, 201\\nrevenge, 57–58, 176\\nRGC. See realistic group conflict\\nrhetoric, political, 58\\nright of return, Palestinian refugees\\nand, 203–9\\nRIR. See Royal Irish Regiment\\nritual, 58, 60\\nrole identity, 135, 138–40, 144, 146–48,\\n151–54, 156n.6, 224–25, 237, 240,\\n245\\nRoman Catholics\\nin Northern Ireland, 133, 135–38,\\n140–44, 153, 154, 155–56nn.3, 4,\\n180, 223, 224\\nin Rwanda, 178\\nin United States, 104\\nRomans, 11\\nRoyal Irish Regiment, 135\\nRoyal Ulster Constabulary, 134–35,\\n137, 151–53, 155n.2\\nRSS. See Rashtriya Swayamsevak Sangh\\nRUC. See Royal Ulster Constabulary\\nRwanda, 11, 59, 101, 174, 223, 234,\\n246n.1\\nchurch leaders’ actions in, 178\\nconflict instigators, 161–62\\nideologies of antagonism and, 170–71\\nneed for healing in, 171–72, 175–77\\nworld community and, 242\\nSee also Hutus; Tutsis\\nSami, 58\\nscapegoating, 33, 161, 162, 164, 165,\\n180\\nSDLP\\n. See Social, Democratic, and\\nLabour Party\\nsecondary appraisal, 234–35\\nsecure identities, 167, 181\\nsecurity forces (Northern Ireland),\\n133–35, 138, 140, 151, 152, 154,\\n155n.2\\nIndex of Subjects\\n261\\n\\n\\nsegmentary organization, 64, 65\\nself, 3, 4, 215–16, 245–46\\ncollective identity and, 59, 139\\ndefinitions of, 5\\ngenocide, mass killing, and, 12, 159,\\n160, 162–67, 169, 171, 174, 176,\\n179, 180–81, 213, 226\\nidentity politics and, 47, 65\\nintergroup comparisons and, 23–24\\nmonolithic identity and, 193\\nin optimal distinctiveness theory, 22,\\n30, 32\\nrole identities and, 139–40\\nsocial identification and, 21, 87, 238\\nin social identity framework, 230,\\n235, 236, 237–41\\nin threat-coping process, 232\\nself-categorization, 3, 10, 20, 222\\nAmericans and, 81, 86–87, 219, 221\\nsocial identity and, 72, 73–74, 139,\\n218, 238\\nself-concept, group, 164, 168–70, 171,\\n174, 180, 226\\nself-esteem, 21, 22, 25, 72, 169\\nself-evaluation, 24, 25\\nself-narratives, 238, 241, 246\\nself-worth, 21, 26, 176\\nSerbia, 49\\nSerbs, 6, 59, 61–62, 223, 226\\nBosnia and, 10, 177, 216, 243\\nCroats and, 133, 155, 195\\ngroup narrative by, 58, 216, 237\\nKosovo and, 8–9, 57, 58\\nrole identities of, 240\\nYugoslavia and, 48–50\\nshared coping, 243\\nSimmel’s Rule, 63, 65\\nSinhalese-T\\namil conflict, 58\\nSinn Féin, 137, 138, 140, 152, 226\\nsituational identification, 45–46\\nSix-Day War (1967), 189\\nslavery, 3, 8, 10\\nin Fiji, 51\\nin United States, 224, 230–31, 232\\nSlavs, 49–50\\nSlovenes, 48–49\\nSlovenia, 48, 49\\nsocial categorization, 10, 19–20, 29, 33,\\n35, 214\\nsocial change, intergroup harm-doing\\nand, 234\\nsocial class. See class\\nsocial comparison, 24, 34, 214\\nsocial competition, 24–26, 35\\nSocial, Democratic, and Labour Party,\\n143\\nsocial dominance theory, 11, 103,\\n106–10, 122, 124, 125\\nsocial equality, 57\\nsocial identification, ingroup positivity\\nand, 18–19, 20–23, 32, 36, 214\\nsocial identity, 3–13, 47, 213–46, 228\\nAmericans’ ethnicity and, 116–18,\\n124\\nclass and, 48\\ncommunal differences and, 102,\\n108–9\\ndefinition of, 5–6, 21, 138–39\\ndimensions of, 73–74, 218, 219, 238\\ngenocide, mass killing, and, 160, 162,\\n164–65, 171, 178–79\\ninjured, 12\\nintergroup conflict framework and,\\n213, 229–42, 245\\nintergroup conflict reduction and, 5,\\n7, 9, 12, 36, 242–44\\nintergroup conflict role of, 3, 4, 7–9\\nmultilevel analysis of, 7, 9\\nmultiple aspects of, 178, 240\\nNorthern Ireland hostilities and,\\n11–12, 138–40, 143–44, 146, 148,\\n154, 156n.5, 223–26\\noptimal distinctiveness and, 10,\\n21–22, 29–30, 32–33, 215, 242, \\n243\\nsocial categorization’s role in, 19–\\n20\\ntypes of, 71\\nas violence factor, 5, 11–12, 133, 135,\\n154, 160, 162\\nSee also group identity\\nsocial identity theory, 3, 10, 21, 27, 164\\nU.S. national identity and, 72–75,\\n77–78, 87, 97, 124\\nsocialization, 31, 78, 84, 86, 218, 236\\nsocial movements\\nclass-based, 63\\ngender-based, 63\\n262\\nIndex of Subjects\\n\\n\\nsocial psychology, 7, 10, 17\\nconflict-reduction mechanisms,\\n242–43\\nsocial representations, 236, 237\\nsocial structure, 34–35\\nsocial welfare. See human welfare\\nsocietally projected images, 236, 237\\nsociology, 7, 12\\nSomalia, 59\\nSouth Africa, 187, 223\\napartheid policy, 44, 194, 195, 232\\nSri Lanka, 187, 210\\nstatus of groups, 24, 25, 35\\nstereotypes, 7, 8, 10, 223–24, 232, 243\\ncategorization and, 20\\noutgroup, 30–31, 215\\nSudan, 59\\nsuffering\\nacknowledgment of, 171, 175\\ndesensitization to, 173–74\\nimages of past, 57–58, 61\\nsuperiority\\nbeliefs of, 168–69\\nmoral, 10, 32–33, 34\\nsuperordinate category identity, 36\\nsuperordinate goals, 9, 33, 243\\nsuperordinate identities, 242–43\\nsymbolic politics theory, 73, 78–79, 84,\\n87, 95, 218, 219, 233, 234\\nsymbolic threat, 233–34\\nsymbolism, political, 58\\nSyria, 189, 203\\nT\\naig (term), 142, 156n.4\\nT\\namils, 58\\nterritory, Israeli-Palestinian conflict\\nover, 187, 191–93, 196–97, 201,\\n202\\nthird-party interventions, 243, 244. See\\nalso problem-solving workshops\\nThirty Years’ War (1618–1648), 169\\nthreat, 28, 32–34, 35, 36\\nfeelings of, 162, 163, 165, 171, 209,\\n237\\nmaterial vs. symbolic, 233–34\\nin social identity framework, 229–36,\\n245\\nthreat-coping frameworks, 229, 231–32\\ntolerance, 77, 95, 177\\ntraditionalism, 55\\ntranscendent identity, 193, 201–2,\\n210–11, 243\\ntribal societies, 64\\nTrinidad and T\\nobago, 53\\ntrust, 31, 36, 171, 175, 176\\ningroup, 10, 28–30, 34, 35, 215\\noutgroup distrust, 30–36\\nTurkey, 171–72\\nTurks, 8, 9, 11, 57, 172\\nTutsis, 59, 133, 155, 161–62, 170–71,\\n178, 223, 234, 246n.1\\nUlster Defence Association (UDA),\\n138, 142, 143\\nUlster Defence Regiment (UDR), 135,\\n153\\nUlster Democratic Party (UDP), 153\\nUlster Freedom Fighters (UFF), 138,\\n142, 143, 152, 155n.2\\nUlsterization (British policy), 151, \\n152\\nUlster Volunteer Force (UVF), 137,\\n138, 142, 143, 152, 155n.2\\nU.N. See United Nations\\nUnionists, Irish, 137, 141–42\\nUnited Irishmen, 136–37\\nUnited Kingdom. See Great Britain\\nUnited Nations, 179\\nGenocide Convention, 159\\nPalestine partition plan, 188\\nUnited Nations Relief and Works\\nAgency, 203\\nUnited States, 5–8, 10–11, 155, 217–22,\\n240, 245\\nAmerican identity conceptions in,\\n75–78, 81, 84, 92, 218–19\\nAmericanization of immigrants in,\\n76, 105, 221\\nchauvinistic views in, 81–85, 92, 93,\\n95, 219, 222\\ncommunal and national identity in,\\n101–25, 220–21\\ncommunal conflict in, 102, 106–8\\nhistorical perspectives on ethnic di-\\nversity in, 103–6, 221\\nidentity politics in, 4, 94–95, 97, 219\\nimmigration issues in, 8, 71, 76–77,\\n92, 96, 97, 104–5, 110, 233\\nIndex of Subjects\\n263\\n\\n\\nUnited States (continued)\\nmulticulturalism in, 71, 76–78,\\n81–83, 89, 92–93, 96, 103, 218\\nnational identity’s meaning in, 71–97,\\n218–20\\nNorthern Ireland conflict and, 224,\\n242\\npatriotism in, 10–11, 78, 81, 82–85,\\n92, 95, 97, 109–11, 119–24,\\n126nn.2–4, 219–22\\npolicy issues in, 73, 77, 78, 87–96,\\n104, 105, 219, 221\\npride in country in, 10, 78, 79, 81, 83,\\n84, 95, 219\\nslavery in, 224, 230–31, 232\\nunit relationships, 22\\nuniversal stereotypes, 30–31, 215\\nUniversity of California, Los Angeles\\n(UCLA), Institute for Science \\nSocial Research, 79\\nUniversity of Michigan, Institute for\\nSocial Research, 114\\nUNRWA. See United Nations Relief\\nand Works Agency\\nUntouchables (Indian caste), 55, 57\\nUVF. See Ulster Volunteer Force\\nvalues, genocide, mass killing, and, 162,\\n177, 178–79, 180\\nvictimization, 162, 163, 171–72,\\n175–77, 226\\nVietnam, 165, 172\\nviolence, 8, 216\\nagents of, 241\\nbystanders to, 162, 163, 173–74,\\n179–80, 181, 226, 227, 241\\ninstigators of, 160–63, 173, 180, 181,\\n226, 230\\nin Northern Ireland, 133–55, 155n.1,\\n223–26, 245\\nas revenge for past, 57–58\\nsocial identity’s role in, 5, 11–12, 133,\\n135, 154, 160, 162\\nstate-sanctioned, 3, 11\\nSee also aggression; ethnic cleansing;\\ngenocide; harm-doing; mass\\nkilling; war\\nvulnerability, sense of, 163, 168, 172,\\n176, 226\\nwar, 3, 6, 10, 27\\nin Ireland, 136, 137, 223\\nin Middle East, 188–89\\nin Rwanda, 161\\nSumner’s theory of, 18\\nin Yugoslavia, 47–51\\nSee also international conflicts; specific\\nwars\\nweakness, sense of, 168\\nWeimar Republic, 178, 232\\nWest Bank (Middle East), 188, 189,\\n190, 196, 197, 203\\nwomen\\nidentity and, 166\\nin United States, 76\\nwomen’s movement, 220\\nworkshops, problem-solving, 7, 12, 187,\\n190, 195–202, 210, 214, 227, 228,\\n245\\nWorld War I, 169, 188\\nWorld War II, 10, 172, 173–74, 177,\\n179. See also Holocaust\\nYugoslavia, 42–43, 47–52, 57, 61, 63,\\n77, 237\\nethnic cleansing policy, 4, 49, 101\\nworld community and, 242\\nSee also Bosnia-Herzegovina; Kosovo;\\nSerbs\\nZambia, 58\\nzero-sum perspective, 194, 210\\ndiscriminatory behavior and, 26\\nin Israeli-Palestinian conflict, 12,\\n187, 192, 228\\nsocial competition and, 24–25\\nZionism, 188, 192, 196, 202, 204\\n264\\nIndex of Subjects\\n\\n\\nAboud, F. E., 243\\nAbrams, D., 21\\nAiken, L. S., 123\\nAlba, R. D., 86, 105, 111\\nAlexander, king of Yugoslavia, 48\\nAllport, G. W., 7, 20, 177\\nAlpher, J., 200, 202, 203, 205, 206, 207\\nAnastasio, P. A., 9, 23\\nAnderson, Benedict, 45, 60, 62, 194\\nAndrews, Irene, 143\\nAppiah, K. A., 77\\nArafat, Yasser, 189, 197\\nAretxaga, B., 224\\nArthur, M., 150\\nAshe, Thomas, 151, 152\\nAshmore, Richard D., 3–14, 71, 213–49\\nBachman, B. A., 9, 23\\nBaker, S. M., 238\\nBanaji, M. R., 21\\nBanker, B. S., 97, 243\\nBarber, B., 71\\nBaron, R. A., 23\\nBar-tal, D., 75, 115, 164\\nBarth, F., 43, 46, 62\\nBauman, Zygmunt, 42, 66\\nBaumeister, R. F., 5, 168\\nBeike, D., 23–24\\nBell, J. B., 138, 151, 152\\nBettencourt, B. A., 26\\nBigelow, K. R., 179\\nBigler, R. S., 243\\nBillig, M., 18\\nBlackwell, M. S., 73, 139, 246\\nBlanz, M., 27\\nBlumer, H., 107\\nBoskin, J. S., 224\\nBoulding, K. E., 216\\nBourhis, R., 25\\nBrewer, Marilynn B., 5, 7, 10, 17–41,\\n75, 87, 170, 213, 214–17, 229,\\n233–34, 235, 237, 238, 241, 242,\\n243, 244–45\\nBrimelow, P., 76\\nBringa, T., 50\\nBrown, R., 17, 18, 25\\nBruce, Steve, 135, 141, 142, 152, 153\\nBundy, R., 18\\nBurton, John, 198\\nCairns, E., 133, 180\\nCalderwood, Alex, 143\\nCampbell, D. T., 18, 25, 30, 31\\nCaporael, L. R., 29\\n265\\nIndex of Names\\n\\n\\nChapman, M., 62\\nCharles I, king of England, 149\\nCheleden, S., 77\\nChirot, D., 3, 71\\nChurchill, Winston, 149–50\\nCitrin, Jack, 5, 6–7, 10, 71–100, 109,\\n213, 218–22, 233, 234, 243, 244,\\n245\\nClifton, A. K., 238\\nCohen, Abner, 44, 62\\nCohen, A. P., 47\\nConnor, W., 74, 75, 87\\nContrada, R. J., 220, 232\\nCornell, S., 50, 51\\nCraig, G. A., 169\\nCrocker, J., 21\\nCromwell, Oliver, 136, 149\\nCronin, S., 136, 137, 144\\nCross, W. E., 239\\nCusack, J., 153\\nDahl, R. A., 106\\nDanziger, K., 5\\nDarby, J., 180\\nDavis, J., 79\\nDeaux, K., 237\\nde la Garza, R. O., 72, 83, 87, 103, 106,\\n109, 110, 111, 221, 238, 239\\nDel Boca, F. K., 8, 216, 233\\nDeschamps, J-C., 20\\ndes Forges, A., 162, 171, 174, 178\\nDevine, P. G., 238\\nde Weerd, M., 139\\nDiehl, M., 18\\nDietz, M., 74\\nDoise, W., 20\\nDorr, N., 26\\nDovidio, J. F., 9, 22, 23, 97, 243\\nDrake, C. J. M., 135\\nDuff, Brian, 5, 10, 71–100, 213, 218\\nDull, V., 238\\nEidheim, H., 58\\nElliott, S., 138, 151, 152\\nEpstein, A. L., 58\\nEpstein, S., 174\\nErickson, G., 21\\nEriksen, Thomas Hylland, 5, 6, 7, 10,\\n42–68, 171, 213, 214–17, 223, 226,\\n227, 231, 234, 235, 237, 238–39,\\n240, 241, 242, 243, 244–45\\nEvans-Pritchard, E. E., 61, 63, 64\\nFalcon, A., 72, 83, 103, 221\\nFang, C. Y., 125\\nFanon, Frantz, 44\\nFarley, R., 71\\nFarrell, M., 152\\nFeatherstone, M., 43\\nFederico, C., 125\\nFeeney, B., 133\\nFeierabend, I. K., 115\\nFein, H., 165\\nFeshbach, S., 74, 75, 81, 103, 112, 115\\nFestinger, L., 24, 214\\nFincham, F., 26–27\\nFine, G. A., 233, 236\\nFinlay, K., 243, 244\\nFisher, R. J., 242, 244\\nFiske, S. T., 31\\nFlackes, W. D., 138, 151, 152\\nFlament, C., 18\\nFogelman, E., 174\\nFolkman, S., 229\\nFordham, S., 8\\nFord, T. E., 23\\nFortes, M., 64\\nFraser, A., 136, 149\\nFraser, M. R., 140, 144\\nFrederico, C. M., 78, 165\\nFredrickson, G. M., 221\\nFreedman, S. G., 105\\nFrey, D., 23\\nFried, A., 74\\nFriedman, J., 42\\nFroystad, K., 61\\nFunk, C. L., 28, 95\\nFurnivall, J. S., 50\\nGaertner, S. L., 9, 23, 97, 242–43\\nGandara, P., 243\\nGandhi, Rajiv, 54\\nGarcia, C., 72, 83, 103, 221\\nGaughan, Michael, 152\\nGeertz, Clifford, 44\\nGellner, Ernest, 44–45, 46, 61, 62\\nGenesee, F., 243\\nGeorge V, king of England, 149\\n266\\nIndex of Names\\n\\n\\nGergen, K. J., 236, 246\\nGergen, M. M., 236, 246\\nGerner, D. J., 188\\nGerstle, G., 77\\nGiddens, A., 42, 47\\nGlazer, N., 77, 97\\nGleason, P., 76\\nGluckman, Max, 63\\nGoldberg, D. T., 77\\nGoldhagen, D., 3, 234, 241\\nGordon, Milton, 77, 105\\nGoring, Herman, 173\\nGourevich, P., 167, 174, 178\\nGreeley, A. M., 105\\nGreen, D. P., 81\\nGreenfeld, L., 3, 71, 75\\nGreenwald, A. G., 21\\nGrey, Edward, 150\\nGubin, A., 177\\nGurin, G., 74\\nGurin, P., 74, 243\\nGurr, T. R., 3, 101, 159\\nGurtman, M. B., 22\\nHaas, E. B., 72\\nHagengimana, A., 177\\nHalpern, B., 188\\nHandelman, D., 58\\nHandler, R., 59\\nHanna, B. A., 30\\nHannerz, U., 43\\nHanning, P., 27\\nHansen, T. B., 56\\nHarff, B., 101, 159\\nHarkabi, Yehoshafat, 196\\nHartmann, D., 50, 51\\nHartz, L., 78\\nHaslam, S. A., 7, 139\\nHeider, F., 22\\nHeppen, Jessica, 213–49\\nHerman, J., 171, 172, 175\\nHertzberg, A., 188\\nHewstone, M., 21, 26–27\\nHigham, J., 76\\nHilberg, R., 173\\nHimmler, Heinrich, 173\\nHinkle, S., 18, 25\\nHirt, E. R., 21\\nHitler, Adolf, 55, 58, 165, 232, 234\\nHobsbawm, E., 62\\nHochschild, J. L., 84\\nHofstetter, C. R., 115, 125\\nHogg, M. A., 20, 21, 73, 139, 246\\nHolland, D., 4, 5, 46, 47\\nHollinger, D. A., 71, 77, 89\\nHorney, Karen, 215\\nHorowitz, D. L., 56, 78, 97, 101\\nHull, C. L., 214\\nHuntington, S., 48, 53\\nHuntington, S. P., 76\\nHyndman, M., 145\\nIgnatieff, Michael, 48, 65\\nInsko, C. A., 31\\nJackman, M., 107\\nJames II, king of England, 136\\nJanoff-Bulman, R., 171\\nJarymowitz, M., 170\\nJaspars, J., 26–27\\nJenkins, R., 47\\nJohnson, G., 23\\nJones, J., 3, 236\\nJones, James, 96\\nJussim, Lee, 3–14, 71, 213–49\\nKaistaniemi, L., 169\\nKapferer, B., 58\\nKateb, G., 77\\nKaukiasnen, A., 169\\nKeegan, J., 3\\nKelley, H. H., 32\\nKelley, R., 104\\nKelly, J. D., 51\\nKelman, Herbert C., 3, 5, 6, 7, 12, 75,\\n170, 177, 187–212, 213–14,\\n227–29, 233, 237, 241, 242, 244,\\n245, 246\\nKelters, S., 133\\nKennedy, C., 21\\nKeough, K. A., 171, 175, 176\\nKhalidi, R., 188\\nKincaid, E., 166\\nKinder, D. R., 83, 84\\nKlandermans, B., 139\\nKleppner, P. J., 104\\nKlicperova-Baker, M., 115\\nKonner, M., 174\\nIndex of Names\\n267\\n\\n\\nKosterman, R., 74, 75, 81, 112, 115\\nKramer, R. M., 30\\nKreisberg, L., 3\\nKressel, N. J., 167\\nLaclau, E., 65\\nLagerspetz, K. M. J., 169\\nLash, S., 43\\nLau, R., 78\\nLavine, H., 167\\nLazarus, R. S., 229\\nLeVine, R. A., 30\\nLevin, S., 26, 78, 103, 125, 165\\nLevy, S. R., 243\\nLewin, Kurt, 217\\nLifton, R. J., 173\\nLijphart, A., 101, 102, 103\\nLind, E. A., 89\\nLind, M., 76\\nLipset, S. M., 78, 103\\nLui, L., 238\\nLyengar, S. S., 227\\nLyons, F. S. L., 137, 149\\nMajor, B., 21\\nMalanchuk, O., 74\\nMalle, B. F., 107\\nManchester, William, 149–50\\nManzi, J., 25, 26\\nMarcia, 239\\nMartin, P., 82\\nMaudling, Reginald, 149\\nMcAdams, R. C., 236, 246\\nMcCall, G. J., 139, 225\\nMcCann, L. I., 171, 172, 175\\nMcCaughey, Sean, 152\\nMcClintock, C., 26\\nMcDonald, H., 153\\nMcDonald, M., 62\\nMcGarry, J., 136\\nMcGarty, C., 139\\nMcGrath, D., 238\\nMcKee, Billy, 151, 152, 153\\nMcKittrick, D., 133, 134\\nMcMichael, Gary, 153\\nMcMichael, John, 153\\nMeertens, R. W., 23\\nMendelsohn, E., 188\\nMerelman, R. M., 82\\nMeyer, G., 20\\nMiller, A. H., 74\\nMiller, D., 72, 74, 77, 89\\nMiller, D. T., 87\\nMorris, B., 203\\nMoxon-Browne, Edward, 141–42, 144\\nMoynihan, D. P., 77, 97\\nMullen, B., 25\\nMummendey, A., 25–26, 27\\nMuslih, M., 188, 197\\nMuste, C., 72, 77\\nMyers, D. G., 6, 164\\nMyrdal, G., 96\\nNeidenthal, P. M., 23–24\\nNelson, S., 142\\nNetanyahu, Benjamin, 190\\nNier, J. A., 97, 243\\nNorenzayan, A., 227\\nNorthrup, T. A., 3, 233, 242\\nNovak, M., 105\\nOakes, P. J., 7, 20, 73, 139, 246\\nO’Connell, A., 27\\nOgbu, J. U., 8\\nOhbuchi, K., 89\\nOkin, S. M., 77\\nO’Leary, B., 136\\nOliner, D., 174, 179\\nOliner, S. B., 174, 179\\nO’Malley, P., 150\\nOpotow, S., 173\\nOrwell, George, 75\\nOtten, S., 26, 27\\nParenti, M., 105\\nPearlman, L. A., 171, 172, 174, 175,\\n176, 177\\nPennebaker, J. W., 171, 175, 176\\nPerdue, C. W., 22\\nPetrocik, John R., 5, 6–7, 11, 101–29,\\n165, 213, 220–22, 234, 242, 245\\nPettigrew, T. F., 23, 177, 240, 243\\nPlatow, M., 27\\nPolsby, N. W., 106\\nPortes, A., 82\\nPowell, C. L., 106\\nPowell, G. B., Jr., 102\\nPratto, F., 103, 107, 125\\n268\\nIndex of Names\\n\\n\\nPremdas, R., 51\\nPrentice, D. A., 87\\nPrentice-Dunn, S., 23\\nPrunier, G., 162, 171\\nRabin, Yitzhak, 190, 197\\nRabinowitz, J. L., 78, 125, 165\\nRanger, T., 62\\nRaz, J., 72, 77\\nReese, H., 170\\nReich, R. B., 71\\nReicher, S., 20, 73, 139, 246\\nReingold, B., 72, 81\\nRex, J., 44\\nRoediger, D. R., 86\\nRogers, R. W., 23\\nRokkan, S., 103\\nRoosens, E. E., 45\\nRoosevelt, Franklin, 175\\nRose, Richard, 141, 144\\nRothman, J., 237, 242, 244\\nRouhana, Nadim, 200\\nRubin, M., 21\\nRummel, R. J., 177\\nRuscher, J., 31\\nRust, M. C., 9, 23\\nSachdev, I., 25\\nSadat, Anwar, 189\\nSaid, Edward, 44\\nSalins, P. D., 105\\nSalmivalli, C., 169\\nSampson, E. E., 166\\nSanders, L. M., 83\\nSands, Bobby, 152\\nSavarkar, Veer, 55\\nSchaar, J. H., 74, 75\\nSchatz, R., 167\\nSchlesinger, A. M., Jr., 44, 72, 77, 106\\nSchöpflin, G., 48\\nSchopler, J., 31\\nSchreiber, H., 25–26\\nSchwartz, S. H., 26\\nSears, D. O., 28, 73, 77, 78, 79, 83, 84,\\n87, 95, 218, 233\\nSegal, J. M., 201\\nSellers, R. M., 238, 239\\nShave, R., 27\\nShaw, J. S., 25, 26\\nSherif, M., 7, 9, 33, 243\\nShikaki, K., 200, 202, 203, 205, 206,\\n207\\nShore, C., 62\\nSidanius, Jim, 5, 6–7, 11, 25, 26, 78, 85,\\n101–29, 165, 213, 220–22, 234,\\n242, 245\\nSimmel, Georg, 63\\nSimmons, J. L., 139, 225\\nSimon, B., 25\\nSkerry, Peter, 125\\nSmith, Anthony D., 44, 45, 62\\nSmith, C., 25\\nSmith, E. R., 32\\nSmith, N. S., 167\\nSmith, P., 21\\nSmith, Rogers, 76\\nSmith, T., 79\\nSmith, William, 143\\nSoeters, J. L., 177\\nStahelski, A. J., 32\\nStallworth, L. M., 107\\nStangor, C., 23\\nStaub, Ervin, 3, 5, 7, 12, 74, 75, 81, 115,\\n159–84, 213, 223, 226, 227, 228,\\n231, 232, 234, 235, 241, 242, 244,\\n245\\nStephan, C., 32\\nStephan, W. [G.], 32, 243, 244\\nStevenson, J., 142, 143\\nStrathern, M., 43\\nStreich, G., 82\\nStruch, N., 26\\nStryker, S., 139, 225\\nSugawara, I., 89\\nSullivan, L. A., 23\\nSullivan, J., 74\\nSumner, W. G., 17–19, 214\\nSutton, M., 134, 155n.1\\nT\\najfel, H., 3, 4, 5, 7, 10, 17, 18, 20, 21,\\n26, 31, 72, 74, 87, 117, 124, 164,\\n218, 246\\nT\\nakagi, E., 29\\nT\\nambiah, S. J., 47\\nT\\namir, 87\\nT\\naylor, C., 74\\nT\\naylor, Peter, 153\\nT\\nec, N., 174, 179\\nIndex of Names\\n269\\n\\n\\nT\\nesser, A., 25\\nT\\nessler, M., 188\\nThatcher, M., 135\\nThoits, P. A., 3, 4, 72, 73, 139, 140, 224,\\n238\\nThompson, L., 233, 236\\nThornton, C., 133\\nThorson, S. J., 3\\nTilly, C., 155n.1\\nTinker, H., 51\\nTito, Josip Broz, 48, 237\\nT\\nonkin, E., 62\\nTriandis, H. C., 34, 166\\nTurner, J. C., 3, 7, 10, 17, 18, 20, 21, 24,\\n72, 73, 117, 124, 139, 164, 246\\nTurner, P\\n., 21\\nTurner, V., 58, 60\\nTurton, D., 47\\nT\\nyler, R. B., 22\\nT\\nyler, T., 89\\nUrry, J., 43\\nvan der Veer, P., 55\\nvan Laar, C., 77\\nVirshup, L. K., 3, 4, 72, 73, 139, 140,\\n224, 238\\nWalch, T., 105\\nWard, C. M., 97, 243\\nWarner, R. S., 71\\nWaters, M., 86\\nWeber, J. G., 21, 23\\nWeber, Max, 75\\nWest, S. G., 123\\nWetherell, M., 20\\nWhite, John, 142–43\\nWhite, Robert W., 5, 7, 11–12, 133–\\n58, 180, 213, 223, 224–27, \\n230, 235, 237, 240, 241, 242, \\n245\\nWhite, T. F., 135, 155–56n.3\\nWhyte, J., 133, 134\\nWick, B., 238\\nWilder, David, 3–14, 213–46\\nWilkes, 20\\nWilliam of Orange, 136\\nWilliksen-Bakker, S., 52\\nWilson, Paddy, 143\\nWolfe, A., 72\\nWolfinger, R. E., 105\\nWong, Cara, 5, 10, 71–100, 213, 218\\nWood, J. V., 24\\nWorchel, S., 231, 235, 236, 241, 243,\\n244\\nYoung, I. M., 72, 77\\nZaller, J., 79, 87\\nZillman, D., 21\\n270\\nIndex of Names\",\"difficulty\":\"hard\",\"domain\":\"Multi-Document QA\",\"length\":\"long\",\"question\":\"Considering both ‘Social Identity and Intergroup Conflict and Conflict Reduction’ and ‘Social Identity: International Perspectives’, how do the authors conceptualize the role of “group narratives” in shaping intergroup conflict, and what does this reveal about the limitations of traditional conflict resolution strategies focused solely on dialogue and mediation?\",\"sub_domain\":\"Academic\"}","display_format":"text","language":"","answer_status":"published","assets":[],"source_url":"https://huggingface.co/datasets/zai-org/LongBench-v2","history":"initial import","indexing_mode":"noindex","subproblems":[],"grids":[]}