The Social Identity Theory of Intergroup Behavior • 277

1964). Thus, real conflicts of group interests not again, it is probable that pure forms of this ex­ only create a~tago?istic. inter?roup relati?~s but treme are found only infrequently in real social READING 16 also heighten IdentificatIOn WIth, and pOSItIVe at­ situations. Examples that might normally tend to tachment to, the in-group. be near the interpersonal extreme would be the This identification with the in-group, however, relations between wife and husband or between has been given relatively little prominence in RCT old friends. Examples that would normally ap­ as a theoretical problem in its own right. The de­ proach the intergroup extreme are the behavior of velopment of in-group identifications is seen in soldiers from opposing armies during a battle, or RCT almost as an epiphenomenon of intergroup the behavior at a negotiating table ofmembers rep­ The Social Identity Theory conflict. As treated by RCT, these identifications resenting two parties in an intense intergroup con­ are associated with certain patterns of intergroup flict. relations, but the theory does not focus either upon Some of the theoretical issues concerning this of Intergroup Behavior the processes underlying the development and continuum are discussed by Turner (1982, 1984), maintenance ofgroup identity nor upon the possi­ Brown & Turner (1981), and Stephenson (1981); • bly autonomous effects upon the in-group and in­ the main empirical questions concern the condi­ tergroup behavior of these "subjective" aspects of tions that determine the adoption of forms of so­ Henri Tajfel • Formerly of the University of Bristol, England group membership. It is our conte?tion th~t the cial behavior nearing one or the other extreme. The John C. Turner. Macquarie University, Australia relative neglect of these processes m RCT IS re­ first-and obvious-answer concerns intergroup sponsible for some inconsistencies between the conflict. It can be assumed, in accordance with empirical data and the theory in its "classical" our common experience, that the more intense is form. In this sense, the theoretical orientation to an intergroup conflict, the more likely it is that Introduction The alternative to these approaches has been be outlined here is intended not to replace RCT, the individuals who are members of the opposite represented by the work of Muzafer Sherif and but to supplement it in some respects that seem to groups will behave toward each other as a func­ he aim of this chapter is to present an outline his associates and has been referred to by D. 1. us essential for an adequate social psychology of tion of their respective group memberships, rather Tof a theory of intergroup conflict and some Campbell (1965) as the "realistic group conflict intergroup conflict-particularly as the under­ than in terms of their individual characteristics or preliminary data relating to the theory. First, how­ theory" (RCT). Its point of departure for the ex­ standing of the psychological aspects of social interindividual relationships. This was precisely ever, this approach to intergroup behavior and in­ planation of intergroup behavior is in what Sherif change cannot be achieved without an appropriate why Sherif (1967, for example) was able to abol­ tergroup conflict must be set in context, in rela­ (1967) has called the functional relations between analysis ofthe social psychology ofsocial conflict. ish so easily the interindividual friendships formed tion to other approaches to the same problem. social groups. Its central hypothesis-"real con­ in the preliminary stages ofsome ofhis field stud­ Much of the work on the social psychology of flic't of group interests causes intergroup con­ ies when, subsequently, the individuals who had intergroup relations has focused on patterns ofin­ flict"-is deceptively simple, intuitively convinc­ The Social Context become friends were assigned to opposing groups. dividual prejudices and discrimination and on the ing, and has received strong empirical support of Intergroup Behavior An institutionalized or explicit conflict of ob­ motivational sequences of interpersonal interac­ (including Avigdor, 1953; Bass & Dunteman, jective interests between groups, however, does not tion. Outstanding examples of these approaches 1963; Blake & Mouton, 1961, 1962; Diab, 1970; Our point of departure for the discussion to fol­ provide a fully adequate basis, either theoretically can be found, respectively, in the theory of authori­ Harvey, 1956; Johnson, 1967; Sherifet al., 1961; low will be an a priori distinction between two or empirically, to account for many situations in tarian personality (Adorno et al., 1950) and in the Sherif & Sherif, 1953). extremes ofsocial behavior, corresponding to what which the social behavior of individuals belong­ various versions and modifications of the theory RCT was pioneered in social psychology by the we shall call interpersonal versus intergroup be­ ing to distinct groups can be observed to approach of frustration, aggression, and displacement (such Sherifs, who provided both an etiology of inter­ havior. At one extreme (which most probably is the "group" extreme of our continuum. The con­ as Berkowitz, 1962, 1969, 1974). The common group hostility and a theory of competition as re­ found in its pure form only rarely in real life) is flict in Sherif's studies was "institutionalized" in denominator of most of this work has been the alistic and instrumental in character, motivated by the interaction between two or more individuals that it was officially arranged by the holiday camp stress on the intraindividual or interpersonal psy­ rewards which, in principle, are extrinsic to the that is fully determined by their interpersonal re­ authorities; it was "explicit" in that it dominated chological processes leading to prejudiced atti­ intergroup situation (see Deutsch, 1949; Julian, lationships and individual characteristics, and not the life of the groups; and it was "objective" in the tudes or discriminatory behavior. The complex 1968). Opposed group interests in obtaining scarce at all affected by various social groups or catego­ sense that, given the terms of the competition, one interweaving of individual or interpersonal behav­ resources promote competition, and positively in­ ries to which they respectively belong. The other of the groups had to be the winner and the other ior with the contextual social processes of inter­ terdependent (superordinate) goals facilitate co­ extreme consists of interactions between two or group the loser. And yet, there is evidence from group conflict and their psychological effects has operation. Conflicting interests develop, through more individuals (or groups of individuals) that Sherif's own studies and from other research that not been in the focus of the social psychologist's competition, into overt social conflict. It appears, are fully determined by their respective member­ the institutionalization, explicitness, and objectiv­ preoccupations (see Tajfel, 1981, pp. 13-56, and too, that intergroup competition enhances intra­ ships in various social groups or categories, and ity of an intergroup conflict are not necessary Turner & Giles, 1981, for more detailed discus­ group morale, cohesiveness, and cooperation not at all affected by the interindividual personal conditions for behavior in terms of the "group" sions). (Fiedler, 1967; Kalin & Marlowe, 1968; Vinacke, relationships between the people involved. Here extreme, although they will often prove to be suf-

276 278 • Political Psychology The Social Identity Theory of Intergroup Behavior • 279 ficient conditions. One clear example is provided as individuals, to.d~vest themsel~es of an unsatis_ flict of inter.ests, in which .it is extremely. d~f!icult members of the out-group. The second statement by our earlier experiments (Tajfel, 1970; Tajfel et factory, u~derpnvtleged,.or stIg~atized grou for an indi;l~ual to conceive of the pos~lbtllty of is closely related to the first: the nearer members "betraying' hIS or her own group by movmg to the of a group are to the "social change" and the "in­ al., 1971), which we shall discuss briefly below, membership. The economic or socIal realities 0/a in which it was found that intergroup discrimina­ society may be such (as, for example, in the cas opposing group. Although this does happen on tergroup" extremes, the more they will tend to treat tion existed in conditions of minimal in group af­ o.f the millions of unemploy~dduri~g the Depres~ occasion, sanctions for such a move are, on the members ofthe out-group as undifferentiated items filiation, anonymity ofgroup membership, absence slon of the 1930s) that the ImpOSSibility of "get­ hole, powerful, and the value systems (at least in a unified social category, rather than in terms of of conflicts of interest, and absence of previous ting out" on one's own, as an individual, become ~ our cultures) are in fl~grant opposi.tion to it. To their individual characteristics. The vast literature hostility between the groups. an everyday reality that determines many form~ use an example from social-psychological research, in social psychology on the functioning of group Other social and behavioral continua are asso­ of intergroup social behavior. But even this ex­ it seems har.dly possible that one ~f the boys in stereotypes in situations of intense intergroup ten­ ciated with the interpersonal-intergroup con­ ample is still relatively extreme. Many social in­ Sherif's holIday camps would deCide to change sions is no more than an example of this general tinuum. One of them may serve to summarize a tergroup situations that contain, for whatever rea­ sides, even though some of his previously con­ statement. quasi-ideological dimension of attitudes, values, sons, strong elements of stratification perceived tracted friendships overlapped group boundaries. Thus, this preliminary conceptualization repre­ and beliefs that may be plausibly hypothesized to as such may tend to move social behavior away The intensity of explicit intergroup conflicts of sents an approach to the social psychology of in­ play a causal role in relation to it. This dimension from the pole of interpersonal patterns toward the interests is closely related in our cultures to the tergroup relations that takes into account social will also be characterized by its two extremes, pole ofintergroup patterns. This is as true ofgroups deoree of opprobrium attached to the notion of realities as weB as their reflection in social behav­ which we shall refer to as "social mobility" and that are "superior" in a social system as of those "r;negade" or "traitor." This is why the belief sys­ ior through the mediation of socially shared sys­ "social change." These terms are not used here in that are "inferior" in it. The major characteristic of tems corresponding to the "social change" extreme tems of beliefs. This convergence occurs at both their sociological sense. They refer instead to in­ social behavior related to this belief is that, in the of our continuum are associated with intense inter­ ends of the sequence just discussed; at the begin­ dividuals' belief systems about the nature and the relevant intergroup situations, individuals will not oroup conflicts. These conflicts can be conceived, ning, because it can be assumed without much dif­ structure of the relations between social groups in interact as individuals, on the basis oftheir individ­ ~herefore, as creating a subclass or a subcategory ficulty that the "social change" belief system is their society. The belief system of "social mobil­ ual characteristics or interpersonal relationships, of the subjective intergroup dichotomization char­ likely to reflect either an existing and marked so­ ity" is based on the general assumption that the but as members of their groups standing in certain acteristic of that extreme of the belief continuum. cial stratification or an intense intergroup conflict society in which the individuals live is a flexible defined relationships to members of other groups. They share the basic feature ofthe "social change" of interests, or both; at the end, because the con­ and permeable one, so that if they are not satis­ Obviously, one must expect a marked correla­ system of beliefs, in the sense that the multigroup sequences of the systems of beliefs arising from fied, for whatever reason, with the conditions im­ tion between the degree of objective stratification structure is perceived as characterized by the ex­ the social situations just mentioned are likely to posed upon their lives by membership in social in a social system (however measured) and the treme difficulty or impossibility ofan individual's appear in the form of unified group actions-that groups or social categories to which they belong, social diffusion and intensity of the belief system moving from one group to another. is, in the form of social movements aiming either it is possible for them (be it through talent, hard of"social change." This, however, cannot be a one­ The continuum of systems of beliefs discussed to create social change or to preserve the status work, good luck, or whatever other means) to move to-one relationship for a number of reasons, some so far represents one conjecture as to one impor­ quo. We shall return later to an elaboration of the individually into another group that suits them of which will be discussed below, although we tant set of subjective conditions that may shift so­ kinds of hypotheses that can be put forward con­ better. A good example of this system of beliefs, cannot in this chapter go into the details of the cial behavior toward members of out-groups be­ cerning the creation of change versus the preser­ built into the explicit cultural and ideological tra­ many social-psychological conditions that may tween the poles of "interpersonal" and vation of status quo. But before this is done, the ditions of a society, is provided in the following determine the transition in certain social groups "intergroup" behavior within particular situations realistic group conflict theory must be considered passage from Hirschman (1970): from an acceptance of stratification to behavior and societies. To conclude this part of our prelimi­ against this general background. characteristic of the intergroup pole of our first nary discussion, we must characterize briefly two The implications of this conceptualization for The traditional American idea ofsuccess confirms continuum-that is, to the creation ofsocial move­ further and overlapping continua, which can be intergroup relations in stratified societies and in­ the hold which exit has had on the national imagi­ ments aiming to change (or to preserve) the status considered as encompassing the major conse­ stitutions are both evident and direct. Whenever nation. Success-or, what amounts to the same quo (see Tajfel, 1978a; Giles & 10hnson, 1981, quences of social behavior that approaches one or social stratification is based upon an unequal di­ thing, upward social mobility-has long been provide a thorough discussion of this issue in the the other end of the interpersonal-intergroup con­ vision of scarce resources-such as power, pres­ conceived in terms of evolutionary individualism. context of seeking to predict the conditions under tinuum. They both have to do with the variability tige, or wealth-and hence there is a real conflict The successful individual who starts out at a low or uniformity within a group of behavior and atti­ of interests between social groups, the social situ­ rung of the social ladder, necessarily leaves his which ethnic groups will accentuate their distinc­ tudes concerning the relevant out-groups. The first ation should be characterized by pervasive ethno­ own group as he rises; he "passes" into, or is "ac­ tive languages, dialects, or accents). cepted" by, the next higher group. He takes his It may be interesting, however, to point to the may be described as follows: The nearer members centrism and out-group antagonism between the immediate family along, but hardly anyone else. close relationship that exists between an explicit of a group are to the "social change" extreme of over- and underprivileged groups (Oberschall, (pp. 108-109) intergroup conflict of interests, on the one hand, the belief-systems continuum and the intergroup 1973, p. 33). However, decades of research into and the "social change" system of beliefs on the extreme of the behavioral continuum, the more ethnic-group relations suggest that ethnocentrism At the other extreme, the belief system of "so­ other. One ofthe main features ofthis beliefsystem uniformity they will show in their behavior toward among stratified groups is, or at least it has been, cial change" implies that the nature and structure is the perception by the individuals concerned that members of the relevant out-group; an approach very much a one-way street. Milner (1975,1981) of the relations between social groups in the soci­ it is impossible or extremely difficult to move in­ toward the opposite extremes of both these con­ and Giles and Powesland (1975) summarize a great ety is characterized by marked stratification, mak­ dividually from their own group to another group. tinua will be correspondingly associated with deal ofevidence that minority or subordinate group ing it impossible or very difficult for individuals, This is precisely the situation in an intense con- greater in-group variability of behavior toward members-such as the American Blacks, the 280 • Political Psychology The Social Identity Theory of Intergroup Behavior • 281

French Canadians, the New Zealand Maoris, or fails to account for social change (in the sen . ciently firm and per~asive t.o prevent the creation a?alogue o~real-world ethnocentrism is in-group the South African Bantus-have frequently tended changes in the mutual relations, behavior an~e of f counitive alternatives to It), the result has been bIas-that IS, the tendency to favor the in-group to derogate the in-group and display positive atti­ titudes of large-scale human groups that h~ve bat­ ~ess :nd not more et?nocent~ism in the different over the out-group in evaluations and behavior. Not een tudes toward the dominant out-group. In other distinctly different in status in the past) part· statUS groups. The pnce of thIS has often been the only are incompatible group interests not always , ICU- ubordinate group's self-esteem. On the other sufficient to generate conflict (as concluded in the words, deprived groups are not always ethnocen­ larly when the processes of change become v ~and, tric in the simple meaning of the term; they may, rapid. St~tus ~ifferences betv:een groups ofteneZ whenever a subordinate group begins, for last section), but there is a good deal of experi­ in fact, be positively oriented toward the depriv­ no~ remaIn uOllater~lly assOCiated with low levels whateVer reasons, to question or deny its presumed mental evidence that these conditions are not al­ ing out-group. Data of this kind are not consistent of IOtergroup confliCt. For example, the generali_ characteristics associated with its low status, this ways necessary for the development of competi­ with a simple application ofRCT. (Recent detailed zation made. above-th~t certain forms of politi­ seems to facilitate the reawakening of a previously tion and discrimination between groups (Brewer, reviews of other field and laboratory data relevant c~l, economIC, and s~cI~l subordination of a so­ dormant conflict over objective resources. At the 1979; Turner, 1981), although this does not mean, to assessing the validity of the theory are provided CIal grouP. tend ~o elImlOate or even reverse its same time, it is likely that one of the counter­ of course, that in-group bias is not influenced by by Brewer, 1979, Stephenson, 1981, and Turner, ethnocentnsm-Is already dated. Research con­ reactions from the dominant groups in such situa­ the goal relations between the groups. tions will be to work for the preservation of the All this evidence implies that in-group bias is a 1981.) ducted over the last two decades reveals a chan a _ Some writers (including Gregor & McPherson, ing pattern in intergroup relations. Americ: previous~y .existing "subjective" and "objective" remarkably omnipresent feature of intergroup re­ 1966; Milner, 1975, 1981; Morland, 1969) have Blacks (Brigham, 1971; Friedman, 1969; Harri~' differentIatIOns. lations. The phenomenon in its extreme form has argued that the status relations between dominant & Braun, 1971; Hraba & Grant, 1970), French Atentative hypothesis about intergroup conflict been investigated by Tajfel and his associates. and subordinate groups determine the latter's iden­ Canadians (Berry, Kalin & Taylor, 1977), New in stratified societies can now be offered: An un­ There have now been in addition to the original tity problems. (By social status we mean a rank­ Zealand Maoris (Vaughan, 1978) and the Welsh equal distribution ofobjective resources promotes studies (Tajfel, 1970; Tajfel et al., 1971) a large ing or hierarchy of perceived prestige.) Subordi­ (Bourhis, Giles & Tajfel, 1973; Giles & Powesland antagonism between dominant and subordinate number of other experiments employing a similar nate groups often seem to internalize a wider social 1975), for instance, now seem to be rejecting (0; groupS, provided that the latter group rejects its procedure (methodological and conceptual issues evaluation of themselves as "inferior" or "second have already rejected) their previously negative in­ previously accepted and consensually negative concerning the experimental paradigm are dis­ class," and this consensual inferiority is reproduced group evaluations and developing a positive eth­ self-image, and with it the status quo, and starts cussed by Aschenbrenner & Schaefer, 1980; as relative self-derogation on a number of indices nocentric group identity. (Milner, 1981, and Tajfel, working toward the development of a positive Bornstein et al., 1983a; Bornstein et al., 1983b; that have been used in the various studies. Con­ 1982b, argue that these new data are likely to be a group identity. The dominant group may react to Branthwaite, Doyle, & Lightbown, 1979; Brown, sensual status itself-where subjective and ac­ genuine reflection ofsocial change.) This construc­ these developments either by doing everything Tajfel, & Turner, 1980; Turner, 1980, 1983a, corded prestige are identical-is problematic for tion of positive in-group attitudes has often been possible to maintain and justify the status quo or 1983b; and the results of the relevant studies are RCT, which conceptualizes prestige as a scarce accompanied by a new militancy over political and by attempting to find and create new differentia­ summarized most recently by Turner, 1983a, and resource, like wealth or power. Status differences economic objectives (see Tomlinson, 1970). tions in its own favor, or both. A more detailed in a wider theoretical and empirical context by between groups, like other inequalities, should tend But these developments do not rescue RCT in specification of some of the strategies and "solu­ Brewer, 1979; Brown & Turner, 1981; Turner, to accentuate the intergroup conflict of interests. its original form. The very suddenness with which tions" that can be adopted in this situation can be 1981, 1982), all showing that the mere perception Therefore, according to RCT, the impact of low the scene has changed effectively rules out objec­ found in Tajfel (1978a); we shall return later to a of belonging to two distinct groups-that is, so­ status upon a subordinate group should be to in­ tive deprivation and therefore new conflicting discussion ofsome of them. For the present, it will cial categorization per se-is sufficient to trigger tensify its antagonism toward the high-status group group interests as sufficient conditions for the be sufficient to state that, whether valid or not, the intergroup discrimination favoring the in-group. (Thibaut, 1950). Yet, under some conditions at "subordinate" group ethnocentrism. On the con­ hypothesis raises some important theoretical prob­ In other words, the mere awareness of the pres­ least, low social status seems to be cOITelated with trary, there has often been less objective depriva­ lems that need to be considered. The first question ence of an out-group is sufficient to provoke in­ an enhancement, rather than a lessening, of posi­ tion than there was in the past. An active and new is: What social-psychological processes are in­ tergroup competitive or discriminatory responses tive out-group attitudes. search for a positive group identity seems to have volved in the development of positive group iden­ on the part of the in-group. It could be argued that only conflicts of interest been one of the critical factors responsible for the tity? The second question concerns the conditions In the basic paradigm the subjects (both chil­ perceived as such create hostility. This requires that reawakening of these groups' claims to scarce re­ under which the status differences between social dren and adults have acted as subjects in the vari­ groups must compare their respective situations. sources (Dizard, 1970). groups are likely to enhance or to reduce inter­ ous studies) are randomly classified as members And, according to some views, it is only relatively In summary, RCT states that opposing claims group conflict. In order to continue the discussion of two nonoverlapping groups-ostensibly on the similar groups that engage in mutual comparisons; to scarce resources, such as power, prestige, or of these questions, we must now abandon specu­ basis of some trivial performance criterion. They therefore, many forms of status differences will wealth, generate ethnocentrism and antagonism lation and consider some relevant data. then make "decisions," awarding amounts of reduce perceived similarity (see Festinger, 1954; between groups. Therefore, low status should tend money to pairs of other subjects (excluding self) Kidder & Stewart, 1975). It follows that status to intensify out-group hostility in groups that are in specially designed booklets. The recipients are systems may reduce social conflict by restricting politically, economically, or socially subordinate. Social Categorization and anonymous, except for their individual code num­ the range of meaningful comparisons available to The evidence suggests, however, that where so­ Intergroup Discrimination bers and their group membership (for example, any given group. This hypothesis may be a useful cial-structural differences in the distribution of member number 51 of the X group and member tool to account for some of the determinants of resources have been institutionalized, legitimized, The initial stimulus for the theorizing presented number 33 of the Y group). The subjects, who social stability; but if it is taken to its logical con- and justified through a consensually accepted sta­ here was provided by certain experimental inves­ know their own group membership, award the tigations of intergroup behavior. The laboratory amounts individually and anonymously. The re- o elusion, it can account for no more than that. It tus system (or at least a status system that is suffi- 282 • Political Psychology The Social Identity Theory of Intergroup Behavior • 283

sponse format of the booklets does not force the characteristics of the experimental situation_' t raroup behavior. In turn, if we assume this­ large-scale social categories, are that the individu­ subjects to act in terms of group membership. other words, whether explicit references to gr In aenct the assumption is by no mean.s unre~son­ als concerned define themselves and are defined " oup In this situation, there is neither a conflict of members hIp commUnIcate to the subjects that th able-we must also assu~e .that thiS ~art~cular by others as members of a group. y interests nor previously existing hostility between are expected to, or ought to, discriminate. The fire form of intergroup behavlOr IS one which is ca­ We can conceptualize a group, in this sense, as the "groups." No social interaction takes place . b d b h" st pable of being induced by the experimenters much a collection of individuals who perceive them­ POint to e ma e a out t IS IOterpretation of th more easily than other forms (such as coopera­ between the subjects, nor is there any rational ~ results is that explicit references to group me tion between the groups in extorting the maxi­ selves to be members of the same social category, link between economic self-interest and the strat­ bership are logically necessary for operationalizi~ mum total amount ofmoney from the experiment­ share some emotional involvement in this com­ egy of in-group favoritism. Thus, these groups are in these minimal situations the major independen~ ers, or a fair division of the spoils between the mon definition of themselves, and achieve some purely cognitive and can be referred to as vari~ble-thatis, social catego?zation per se. Thi~ aroups, or simply random responding). And this degree ofsocial consensus about the evaluation of "minimal." requITes not merely that the subjects perceive them­ iast assumption must be backed up in its turn by their group and of their membership in it. Follow­ The basic and highly reliable finding is that the selves as similar to or different from others as in­ another presupposition: namely, that for some ing from this, our definition of intergroup behav­ trivial, ad hoc intergroup categorization leads to divid~als, b~t that they ar~ members of discrete reasons (whatever they may be) competitive be­ ior is basically identical to that of Sherif (1967, p. in-group favoritism and discrimination against the and dlscontlOuo~s categone~-thatis, "groups." havior between groups, at least in our culture, is 62): any behavior displayed by one or more actors extraordinarily easy to trigger off-at which point out-group. Fairness is also an influential strategy. Second, a detailed analysIs of the subjects' toward one or more others that is based on the ac­ There is also a good deal of evidence that, within we are back where we started from. The problem tors' identification of themselves and the others as postsession reports (Billig, 1972; Turner, 1975a) then must be restated in terms of the need to the pattern of responding in terms of in-group fa­ belonging to different social categories. shows that they do not share any common con­ specify why a certain kind of intergroup behav­ voritism, maximum difference (MD) is more im­ ception of the "appropriate" or "obvious" way to ior can be elicited so much more easily than other Social categorizations are conceived here as portant to the subjects than maximum in-group behave, that only a tiny minority have some idea kinds; and this specification is certainly not made cognitive tools that segment, classify, and order profit (MIP). Thus, they seem to be competing with of the hypothesis, and that this minority does not if we rest content with the explanation that the the social environment, and thus enable the indi­ the out-group, rather than following a strategy of always conform to it. Thirdly, the relevant experi­ behavior occurred because it was very easy for vidual to undertake many forms of social action. simple economic gain for members ofthe in-group. mental data do not support this interpretation. St. the experimenters to make it occur. (Tajfel, 1978a, But they do not merely systematize the social Other data from several experiments also show that Claire and Turner (1982) exposed observer-sub_ pp. 35-36) world; they also provide a system of orientation the subjects' decisions were significantly nearer jects to exactly the same experimental cues as nor­ for self-reference: they create and define the to the maximum joint payoff (MlP) point when mal categorized subjects; the former were required Two points stand out: first, minimal intergroup individual's place in society. Social groups, un­ these decisions applied to the division of money to predict the responses of the latter in the stan­ discrimination is not based on incompatible group derstood in this sense, provide their members with between two anonymous members of the in-group dard decision booklets. The categorized subjects interests; second, the baseline conditions for in­ an identification of themselves in social terms. than when they applied to two members of the out­ did discriminate significantly, but the observers tergroup competition seem indeed so minimal as These identifications are to a very large extent re­ group; that is, relatively less was given to the out­ failed to predict it and in fact expected significantly to cause the suspicion that we are dealing here with lational and comparative: they define the individual group, even when giving more would not have af­ more fairness than was actually displayed. some factor or process inherent in the intergroup as similar to or different from, as "better" or fected the amounts for the in-group. Billig and . The more general theoretical problem has been situation itself. Our theoretical orientation was "worse" than, members of other groups. It is in a Tajfel (1973) have found the same results even referred to elsewhere by one of us as follows: developed initially in response to these clues from strictly limited sense, arising from these consider­ when the assignment to groups was made explic­ our earlier experiments. We shall not trace the his­ ations, that we use the term social identity. It con­ itly random. This eliminated the similarity on the Simply and briefly stated, the argument (e.g., tory ofits development, however, but shall describe sists, for the purposes of the present discussion, performance criterion within the in-group as an Gerard and Hoyt, 1974) amounts to the follow­ its present form. of those aspects of an individual's self-image that alternative explanation of the results. An explic­ ing: the subjects acted in terms of the intergroup derive from the social categories to which he per­ categorization provided or imposed by the experi­ itly random classification into groups proved in ceives himselfas belonging. With this limited con­ menters, not necessarily because this has been this study to be a more potent determinant of dis­ successful in inducing any genuine awareness of Social Identity and Social Comparison cept of social identity in mind, our argument is crimination than perceived interpersonal similari­ membership in separate and distinct groups, but based on the following general assumptions: ties and dissimilarities not associated with catego­ probably because they felt that this kind of be­ Many orthodox definitions of "social groups" are rization into groups. Billig (1973), Brewer and havior was expected of them by the experiment­ unduly restrictive when applied to the context of 1. Individuals strive to maintain or enhance their Silver (1978), Locksley, OItiz and Hepburn (1980), ers, and therefore they conformed to this expec­ intergroup relations. For example, when members self-esteem: they strive for a positive self-con­ and Turner, Sachder and Hogg (1983) have all rep­ tation. The first question to ask is why should the of two national or ethnic categories interact on the cept. licated this finding that even explicitly arbitrary subjects be expecting the experimenters to expect basis of their reciprocal beliefs about their respec­ 2. Social groups or categories and the member­ social categorizations are sufficient for discrimi­ of them this kind of behavior? The Gerard and tive categories and ofthe general relations between ship of them are associated with positive or nation, and Alien and Wilder (1975) have provided Hoyt answer to this is that the experimental situ­ them, this is clearly intergroup behavior in the negative value connotations. Hence, social ation was rigged to cause this kind of expecta­ additional evidence for the importance of group everyday sense ofthe term. The "groups" to which identity may be positive or negative according tion in the subjects. This answer retains its plau­ classification compared to similarities between sibility only if we assume that what was no more the interactants belong need not depend upon the to the evaluations (which tend to be socially people without such classification. than a hint from the experimenters about the no­ frequency of intermember interaction, systems of consensual, either within or across groups) of The question that arises is whether in-group bias tion of "groups" being relevant to the subjects' role relationships, or interdependent goals. From those groups that contribute to an individual's in these minimal situations is produced by some behavior had been sufficient to determine, pow­ the social-psychological perspective, the essential social identity. fotm of the experimenter effect or of the demand erfully and consistently, a particularform of in- criteria for group membership, as they apply to 3. The evaluation of one's own group is deter- 284 • Political Psychology The Social Identity Theory of Intergroup Behavior • 285

mined with reference to specific other groups 1981; Giles & Powesland, 1975). Third, in-grou acquiesce in the .superiority of the winni~g o.ut­ sion, two three-person groups discussed an iden­ s through social comparisons in terms of value­ do ~ot compare themselves with every cognitivel uroup. This ~cqUleSC~?Ce ~y ~ group consIden~g tical issue, supposedly to gain an assessment of laden attributes and characteristics. Positively available out-group: the out-group must be p y ~ elf as legitImately mfenor has been shown 111 their verbal intelligence, and then briefly compared . d 1 . er- discrepant comparisons between in-group and celv~ ~s a re ev~nt c?mpans~n group. Similarity, ~t~dies by Caddick (1980, 1982), Commins and their respective performance. The subjects were out-group produce high prestige; negatively prOXImIty, and sltuatlOnal salience are among th Lockwood, (1979) and Turner and Brown (1978). 144 male undergraduates. The criterion for inter­ discrepant comparisons between in-group and variables that determine out-group comparabilit e Several other studies.report ~ndi~gs that are in line group differentiation was the magnitude of in­ out-group result in low prestige. and pressures toward in-group distinctivene y, with this interpretatIOn: 10sll1g m-groups do not group bias shown in the ratings of the groups' . . ~ should 1I1crease as a functIOn of this comparabil_ always derogate, but. so~etimes upgrade, their work. Half the triads, composed of Arts students, From these assumptions, some related theoretical ity. It is important to state at this point that, in man evaluations of the w1l1nll1g out-g:oups (fo~ ex­ believed that verbal intelligence was important for principles can be derived: social situations, comparability reaches a muc~ ample, Bass & Dunteman, 1963; Wllson & MIller, them (High Importance); half, composed of Sci­ wider range than a simply conceived "similarity" 1961). ence students, did not (Low Importance). Half the 1. Individuals strive to achieve or to maintain posi­ between the groups. Retrospectively, at least, the social-identity/so- sessions involved two Arts or two Science groups tive social identity. The aim of differentiation is to maintain or cial-comparison theory is consistent with many of (Similar Out-group), and half involved one Arts 2. Positive social identity is based to a large ex­ achieve superiority over an out-group on some the studies mentioned in the preceding section of and one Science group (Dissimilar Out-group). tent on favorable comparisons that can be made dimensions. Any such act, therefore, is essentially this chapter. In particular, in the paradigm of the Finally, in the Stable Difference condition, sub­ between the in-group and some relevant out­ competitive. Fully reciprocal competition between minimal group experiments, the intergroup dis­ jects were instructed that Arts students were defi­ groups: the in-group must be perceived as posi­ groups requires a situation of mutual comparison crimination can be conceived as being due not to nitely superior and Science students definitely in­ tively differentiated or distinct from the relevant and differentiation on a shared value dimension. conflict over monetary gains, but to differentia­ ferior in verbal intelligence; in the Unstable out-groups. In these conditions, intergroup competition, which tions based on comparisons made in terms ofmon­ Difference condition, there was no explicit state­ 3. When social identity is unsatisfactory, individu­ may be unrelated to the objective goal relations etary rewards. Money functioned as a dimension ment that one category was better than the other. als will strive either to leave their existing group between the groups, can be predicted to occur. of comparison (the only one available within the These variables were manipulated in a 2 x 2 x 2 and join some more positively distinct group Turner (1975b) has distinguished between social experimental design), and the data suggest that factorial design. and/or to make their existing group more posi­ and instrumental or "realistic" competition. The larger absolute gains that did not establish a dif­ The results showed that the Arts (High Impor­ tively distinct. former is motivated by self-evaluation and takes ference in favor of the in-group were sacrificed tance) groups were more biased than the Science place through social comparison, whereas the lat­ for smaller comparative gains, when the two kinds (Low Importance) groups, that similar groups dif­ The basic hypothesis, then, is that pressures to ter is based on "realistic" self-interest and repre­ of gains were made to conflict. ferentiated more than dissimilar groups in the evaluate one's own group positively through in­ sents embryonic conflict. Incompatible group There is further evidence (Turner, 1978a) that Stable condition, but that they were no more bi­ group/out-group comparisons lead social groups goals are necessary for realistic competition, but the social-competitive pattern ofintergroup behav­ ased (and sometimes even less so) in the Unstable to attempt to differentiate themselves from each mutual intergroup comparisons are necessary, and ior holds even when it conflicts with obvious self­ condition; and that, on some ofthe measures, there other (Tajfel, 1978a; Turner, 1975b). There are at often sufficient, for social competition. The latter interest. In this study, the distribution of either was a significant main effect for out-group simi­ least three classes of variables that should influ­ point is consistent with the data from the minimal monetary rewards or "points" was made, within larity: in-group bias increased against a similar out­ ence intergroup differentiation in concrete social group experiments that mere awareness of an out­ the minimal intergroup paradigm, between selfand group. Although these data are relatively complex, situations. First, individuals must have internal­ group is sufficient to stimulate in-group favorit­ an anonymous other, who was either in the in­ they do support some of our theoretical expecta­ ized their group membership as an aspect of their ism, and the observations (Doise & Weinberger, group or in the out-group. As long as minimal con­ tions and provide an illustration that variations in self-concept: they must be subjectively identified 1973; Ferguson & Kelley, 1964; Rabbie & ditions existed for in-group identification, the sub­ in-group bias can be systematically predicted from with the relevant in-group. It is not enough that Wilkens, 1971) that the possibility of social com­ jects were prepared to give relatively less to the social-identity/social-comparison theory. the others define them as a group, although con­ parison generates "spontaneous" intergroup com­ themselves when the award (either in points or in We have argued that social and realistic compe­ sensual definitions by others can become, in the petition. money) was to be divided between self and an tition are conceptually distinct, although most of­ long run, one of the most powerful causal factors Social and realistic competition also differ in anonymous member of the in-group, as compared ten they are empirically associated in "real life." determining a group's self-definition. Second, the the predictions that can be made about the conse­ with dividing with an anonymous member of the In an experiment by Turner, Brown, and Tajfel social situation must be such as to allow for inter­ quences for subsequent intergroup behavior of out-group. These results seem particularly impor­ (1979) an attempt was made to isolate the effects group comparisons that enable the selection and winning or losing. After realistic competition, the tant, since the category of "self," which is by no on intergroup behavior of the postulated autono­ evaluation of the relevant relational attributes. Not losing groups should be hostile to the out-group means minimal or ad hoc, was set here against a mous processes attributed to a search for positive all between-group differences have evaluative sig­ victors, both because they have been deprived ofa truly minimal in-group category, identical to those social identity. Children were used as subjects, and nificance (Tajfel, 1959), and those that do vary reward and because their interaction has been ex­ used in the earlier experiments. Despite this stark the manipulations involved decisions by the sub­ from group to group. Skin color, for instance, is clusively conflictual. However, when winning and asymmetry, the minimal group affiliation affected jects about the distribution of payments for par­ apparently a more salient attribute in the United losing establish shared group evaluations concern­ the responses. ticipation in the experiment, to be shared equally States than in Hong Kong (Morland, 1969); ing comparative superiority and inferiority, then, The theoretical predictions were taken outside by the in-group, between the in-group and the out­ whereas language seems to be an especially sa­ so long as the terms of the competition are per­ of the minimal categorization paradigm in a fur­ groups that were made relevant or irrelevant to lient dimension of separate identity in French ceived as legitimate and the competition itself as ther study by Turner (1978b). He used face-to-face comparisons with the in-group's performance. Canada, Wales, and Belgium (Giles & Johnson, fair according to these terms, the losing group may groups working on a discussion task. In each ses- Monetary self-interest (of a magnitude previously 286 • Political Psychology The Social Identity Theory of Intergroup Behavior • 287 ascertained to be of genuine significance to the fied societies implied by the theory when it is ap­ · s a disidentification with the erstwhile in­ classic paper on the psychology of status sug­ pieI subjects) would have produced no difference in plied to some of the problems raised in the seCond gested that discontent among low-status-group grOUP' the distribution decisions involving the two kinds section. members is lessened to the degree that of out-group; it would also have led to decisions intraclass rather than intergroup comparisons tending toward maximum in-group profit (MIP) 2. SOCIAL CREATIVITY are made. More recently, Rosenberg and Simmons (1972) found that self-esteem was rather than toward maximum difference (MD). Status Hierarchies and Social Change The group members may seek p~sitive disti?ctive- MD was the most influential strategy in the ss for the in-group by redefimng or altenng the higher among blacks who made self-compari­ choices. Furthermore, when the subjects could The reconceptualization ofsocial status attempted n~ rnents of the comparative situation. This need sons with other blacks rather than whites. Other work also suggests (see Katz, 1964; Lefcourt choose in-group favoritism (MD + MIP) and/or a earlier needs now to be made more explicit. Sta­ ~:t involve any change in th~ gr?up's actual S?­ & Ladwig, 1965) that, in certain circumstances, fairness strategy, they were both more discrimina­ tus is not considered here as a scarce reSOurce Or ciaI position or access to O?Jectlve resources In tory and less fair toward the relevant than the ir­ commodity, such as power or wealth; it is the out­ elation to the out-group. It IS a group rather than black performance was adversely affected by relevant comparison group. Other measures of in­ come of intergroup comparison. It reflects a ~n individualistic strategy that may focus upon: the low self-esteem induced by the presence group favoritism produced an interaction between group's relative position on some evaluative di. of the members of the dominant out-group. It reward level and type of out-group: more discrimi­ mensions of comparison. Low subjective status (a) Comparing the in-group to the out-group on follows that self-esteem can be enhanced by nation against the relevant than the irrelevant group does not promote inter-group competition directly; some new dimension. Lemaine (1966) found, comparing with other lower-status groups with high rewards, and less with low rewards. its effects on inter-group behavior are mediated for example, that children's groups that could rather than with those of higher status. This is Whatever may be other explanations for this in­ by social identity processes. The lower is a group's not compare themselves favorably with others consistent with the fact that competition be­ teraction, we can at least conclude that when re­ subjective status position in relation to relevant in terms of constructing a hut-because they tween subordinate groups is sometimes more ward levels are more meaningful, in-group favor­ comparison groups, the less is the contribution it had been assigned poorer building materials intense than between subordinate and domi­ itism is enhanced against a more comparable can make to positive social identity. The variety than the out-group-tended to seek out other nant groups-hence, for example, lower-class out-group, independently of the group members' of reactions to negative or threatened social iden­ dimensions ofcomparison involving new con­ or "poor white" racism. economic interests. Indeed, insofar as the subjects tity to be discussed below are an elaboration of structions in the hut's surroundings. The prob­ used the MD strategy, they sacrificed "objective" the principles outlined earlier in this chapter. lems that obviously arise here are those of le­ 3. SOCIAL COMPETITION personal and group gain for the sake of positive gitimizing the value assigned to the new social The group members may seek positive distinctive­ in-group distinctiveness. 1. INDIVIDUAL MOBILITY products-first in the in-group and then in the A study by Oakes and Turner (1982) also de­ other groups involved. To the extent that this ness through direct competition with the out­ serves mention here since it seems to provide some Individuals may try to leave, or dissociate them­ legitimization may threaten the out-group's group. They may try to reverse the relative posi­ direct evidence for the social competition inter­ selves from, their erstwhile group. This is prob­ superior distinctiveness, an increase in inter­ tions of the in-group and the out-group on salient pretation of the minimal group experiments. They ably more likely the more they approach the "so­ group tension can be predicted. dimensions. To the degree that this may involve simply compared the self-esteem of subjects cat­ Cial mobility" pole of the continuum of comparisons related to the social structure, it im­ egorized as in Tajfel et al. (1971) but who were belief-systems described previously. This strategy (b) Changing the values assigned to the attributes plies changes in the groups' objective socialloca­ not asked to complete the decision booklets with usually implies attempts, on an individual basis, of the group, so that comparisons which were tions. We can hypothesize, therefore, following subjects who were categorized and also discrimi­ to achieve upward social mobility, to pass from a previously negative are now perceived as posi­ RCT, that this strategy will generate conflict and nated in the normal manner. The latter subjects lower- to a higher-status group. In a four-group tive. The classic example is "black is beauti­ antagonism between subordinate and dominant were found to have higher self-esteem than the hierarchy, Ross (1979) found a direct linear rela­ ful." The salient dimension-skin color-re­ groups insofar as it focuses on the distribution of former-in line with the idea that discrimination tionship between low status and the desire to pass mains the same, but the prevailing value system scarce resources. Data relevant to this strategy have serves to achieve a positive social identity. Need­ upward into another group. Many earlier studies concerning it is rejected and reversed. The been referred to earlier in this chapter. less to say, work is progressing to replicate and report the existence of strong forces for upward same process may underlie Peabody's (1968) explore this finding. social movement in status hierarchies. Tendencies finding that even when various groups agree Let us assume as an ideal case some stratifica­ On the whole, the above studies provide some to dissociate oneself psychologically from fellow about their respective characteristics, the trait tion of social groups in which the social hierarchy confirmation for the basic social-identity/social­ members of low-prestige categories are known to is evaluated more positively by the group that is reasonably correlated with an unequal division comparison hypothesis. Further studies testing the many of us from everyday experience: they have possesses it. of objective resources and a corresponding status theory in both field and laboratory settings and been noted more systematically by lahoda (1961) system (based on the outcomes of comparisons in discussions of its application to the analysis of and Klineberg and Zavalloni (1969), among oth­ (c) Changing the out-group (or selecting the out­ terms of those resources). Under what conditions specific social contexts (e.g., male-female rela­ ers, and indirectly by the whole literature on ra­ group) with which the in-group is compared­ will this not lead to intergroup conflict-or, more tions, linguistic conflict, Protestant-Catholic con­ cial identification and preference. The most im­ in particular, ceasing or avoiding to use the precisely, to the development of competitive eth­ flict in Northern Ireland, prejudice and black iden­ portant feature of individual mobility is that the high-status out-group as a comparative frame nocentrism on the part of the subordinate group? tity, etc.) are to be found or are reviewed in Tajfel low status of one's own group is not thereby of reference. Where comparisons are not made First, to the extent that the objective and the (1978b, 1982a, 1982b) and Turner and Oiles changed: it is an individualist approach designed, with the high-status out-group, the relevant subjective prohibitions to "passing" are weak (see (1981). We shall now attempt to outline in general at least in the short run, to achieve a personal, not inferiority should decrease in salience, and our earlier discussion of the "social mobility" sys­ terms the analysis of inter-group behavior in strati- a group, solution. Thus, individual mobility im- self-esteem should recover. Hyman's (1942) tem of beliefs), low status may tend, in conditions 288 • Political Psychology The Social Identity Theory of Intergroup Behavior • 289 of unsatisfactory social identity, to promote the we shall argue presently that there are many sta­ th'e alternatives to the actual outcome are avail­ On the other hand, when the dominant group or widespread adoption of individual mobility strat­ tus differences that do not reduce comparability. able-whether other outcomes are conceivable. sections of it perceive their superiority as legiti­ egies, or at least initial attempts to make use of For the moment, we can note that both indi­ Status di fferences between social groups in social mate, they will probably react in an intensely dis­ these strategies. Insofar as individual mobility vidual mobility and some forms of social creativ­ systems showing variou degrees of stratification criminatory fashion to any attempt by the subor­ implies disidentification, it will tend to loosen the ity can work to reduce intergroup connict over can be distinguished in the same way. Whece sta­ dinate group to change the intergroup situation. cohesiveness of the subordinate group. This weak­ scarce resources-though with different implica­ tu" relations are perceived as immutable. a part of Such perhaps was the postbellum situation in the ening of subjective attachment to the in-group tions. The former is destructive of subordinate­ the fixed order of things, then social identity is southern United States: the whites, threatened by among its members will tend: (a) to blur the per­ group solidarity and provides no antidote to nega­ secure. It becomes insecure when the existing state those \\'ho had been their slaves, rapidly abandoned ception of distinct group interests corresponding tive social identity at a group level. The latter may of affairs begins to be questioned. An important their paternalistic stereotypes of the blacks as to the distinct group identity: and (b) to create restore or create a positive self-image but, it can corollary to this argument is that the dominanl or "childlike" in favor of openly hostile and deroga­ obstacles to mobilizing group members for col­ be surmised, at the price either of a collective re­ high-status groups, too, can experience in.<;ccure lory ones (Van der Berghe, 1967). The reactions lective action over their common illlerests. Thus. pression of objective deprivation or, perhaps. of social identity. Any threat to the distincti\-cJy su­ of illegitimately superior groups arc more com­ the Iow morale that follows from negative social spurious rivalry with some other deprived group. perior position of a group implies a potential loss plex (Turner & Brown, 1978). It seems that con­ identity can set in motion disintegrative processes It is interesting in this context that the French Ca­ of posilive comparisons and possible negau\'C com­ n icts of values are reduced by greater discri mi na­ that, in the long run. may hinder a change in the nadians, having recently gained a more asserti\'e parisons, which must be guarded against. Such a lion when superiority is assured. but by less group status. identity, are now apparently more disparaging of threat may deri\'e from the activity of the Iow-sta­ discrimination when il is unstable. This calls to Second, assuming that the barricrs (objective, other minority groups than are the English Cana­ tus group or from a conflict between the high-sta­ mind ~ollle Prisoner Dilemma studies in which moral, and ideological prohihitions) to leaving dians (Berry et aI., 1977). tus group's OWI1 value system (for example. the white discrimination against black opponents in­ one's group are strong, unsatisfactory social iden­ By reversing the conditions under which social sociopolitical morality) and the actual foundations creased the more cooperative was the opponent, tity may stimulate social creativity that tends to stratification does not produce intergroup connict, of its superiority. Like low-status groups, the high­ but decreased the more competitive he was (Baxter, reduce the salience of the subordinate/dominant we can hypothesize that negative social identity status groups will react to insecure social identity 1473: Cederblom & Diers, 1970). Baxter sug­ group conflict of interest. Strategy 2(c) mentioned promotes subordinate-group competitiveness to­ by scarching for enhanced group distinctiveness. gcsted in the title of his article ("Prejudiced Lib­ above is likely to be crucial here since. in general, ward the dominant group to the degree that: (a) In brief. then. it is true that clear-cut status dif­ erals')") that a conflict of values may underlie his access to resources such as housing, jobs, income, subjective identification with the subordinate ferences may lead to a quiescent social system in data. Research on the different effects of secure or education is sufficiently central to the fate of group is maintained; and (b) the dominant group which neither the "inferior" nor the "superior" and insecure status differences is reported in Tajfel any group that the relevant comparisons are not continues or begins to be perceived as a relevanl groups will show much ethnocentrism. But this (197Xb, 1982a. 1982b; see also Caddick, 1980 and easily changed or devalued. Few underprivileged comparison group. As a great deal of work has "ideal type" situation must be considered in rela­ Skevington, 1980). groups would accept puverty as a \'il1ue. but it may been done in social psychology on the determi­ tion to the perceived stability and Iegitimal.)' of Many of the points and hypotheses we have appear more tolerable to the degree that compari­ nants ofcohesiveness and loyalty within groups-· the system. Perceived illegitimacy and/or instabil­ advanccd in this chaptcr are not, in themselves, sons are made with even poorer groups rather than Hogg (1983), Turner et al. (1983). and Turner. ity provide new dimensions of comparabilitv that nc\\' (sce, for inswnce, Sherif', 1967: Runcilllan, with those that are better ofT (see Runciman. 1966). Sachdev & Hogg (l983) have recently looked in arc directly relevant to the attitudes and behavior 1%6: Milner. 1975; Billig, 1976). What is new. As noted above, some writcrs (Festinger, 1954; particular at the problem of how groups that are of the social groups involved. whatever theirposi­ wc think. is the integration of the three processes Kidder & Stewart, InS) imply that strategy 2(c) associated with costs and deprivations (such as tion in the system. This is the social-psychologi­ of social categorization, self-evaluation through is a dominant response to status differences be­ ~ubordinate ones) are able to maintain their cohe­ cal counterpart to what is widely known today a~ social identity. and intergroup social comparison, tween groups. The assumption is that intergroup siveness-we shall concentrate on the second con­ "the revolution of rising expectations." Providing into a coherent and testable framework for con­ comparability decreases as a direct function of dition. that individual mobility is unavailable or undesir­ tributing to the explanation of various forms of perceived dissimilarity. If this were the whole story, Our hypothesis is that a status difference be­ able, consensual inferiority will be rejected most intergroup behavior, social conflict, and social then, somewhat paradoxically. the creation of a tween groups does not reduce the meaningfulness rapidly when the situation is perceived as both change. This framework contains possibilities of consensual status system would protect social iden­ of comparison between them providing that there unstable and illegitimate. This is (or was) prob­ further development, and to this extent, we hope tity from invidious comparisons. The causal se­ is a perception that it can be changed. For example, ably the set of conditions underlying the develop­ that it may stimulate theoretically directed research quence would be as follows: similar groups com­ consider two football (or any other) teams that at ment of ethnocentrism among black Americans, in areas that have not been considered here. pare with each other; the outcome determines their the end of their season may have come first and French Canadians, and New Zealand Maoris, for But some cautionary points should be made. The relative prestige; the perceived status difference second in their league respectively. There is no instance. Vaughan (1978) reports that the perceived equation ofsocial competition and intergroup con­ reduces their similarity and hence comparability; argument about which has the higher status, but feasibility of social change (probably including, flict made above rests on the assumptions concern­ intergroup comparisons cease to be made; subjec­ alternative comparative outcomes were and, in the in this instance, the perceived illegitimacy of the ing an "ideal type" ofsocial stratification in which tive superiority and inferiority decrease in salience; future, still will be possible. When the new season present situation) is an important predictor of the the salient dimensions of intergroup differentia­ correspondingly, the groups' respective self-es­ begins, the teams will be as comparable and com­ developing Maori ethnocentl;sm; Friedman (1969) tion are those involving scarce resources. In this teems return to their original point. There may be petitive as they had been before. This example il­ argues that what we may term the "cognitive al­ respect, we have simply borrowed the central te­ occasions when this social-psychological recipe lustrates Tajfel's (1978a) distinction between se­ ternative" of black nationalism in the developing net of RCT. There is no reason, in fact, to assume for the maintenance of the status quo can be ob­ cure and insecure intergroup comparisons. The countries was influential in enhancing black that intergroup differentiation is inherently served in something like its pure form. However, crucial factor in this distinction is whether cogni- American social identity. conflictual. Some experimental work already 290 • Political Psychology The Social Identity Theory of Intergroup Behavior • 291 points clearly to the conclusion that evaluative ting nearer to this level so that differentials are acceptance of their eXi~ten~e by t~e groups in­ larity in intergroup behaviour. Europeall Journal ofSocial derogation of an out-group is conceptually and eroded. But the difficulty with this example~Ot volved. The proof of theIr eXIstence IS to be found Psychology, Vo!. 3(1),27-52. Blake, R. R., & Mouton, 1. S. (1961). Competition, commu­ empirically distinct from out-group hostility with many other similar examples-is that, in th~ in the large number ~f st~dies .(and .also ev~r~day . f l' IS nication and conformity. In 1. A. Berg & B. M. Berg (Eds.), (Turner et al., 1979). On the other hand, social­ case, th e preservatIOn 0 sa ary dIfferentials' occurrences in real lIfe) m whIch dIfferentIatlO~s Conformity and deviation. New York: Harper. identity processes may provide a source of inter­ probably associated with all kinds of objecti~S of all kinds are made between groups by theIr Blake, R. R., & Mouton, 1. S. (1962). The intergroup dynam­ group conflict (in addition to the cases outlined advantages that cannot be defined in terms o~ members although, on the face of it, there are no ics of win-lose conflict and problem-solving collaboration above) to the degree that the groups develop con­ ~f in union-management relations. In M. Sherif (Ed.), Imer­ m?ney a.lone. In turn, some these advantages "reasons" for these differentiations to occur. Ex­ glVup relations and leadership. New York: Wiley. flicting interests with respect to the maintenance wIll agam make sense only m the comparative amples of this have been provided in several stud­ Bornstein, G., Crum, L., Wittenbraker, 1., Harring, K., Insko, of the comparative situation as a whole. It seems framework of intergroup competition. Despite this ies mentioned in this chapter in which the intro­ C. A., & Thibaut, 1. (1983a). On the measurement ofsocial plausible to hypothesize that, when a group's ac­ confusing network of mutual feedbacks and inter­ duction by the subjects of various intergroup orientation in the minimal group paradigm. European Jour­ tion for positive distinctiveness is frustrated, im­ actions, the distinctions made here are important differentiations directly decreased the objective nal ofSocial Psychology, 13,321-350. Bornstein, G., Crum, L., Wittenbraker, 1., HalTing, K., Insko, peded, or in any way actively prevented by an out­ because they help us to understand some aspects rewards that could otherwise have been gained by C. A., & Thibaut, 1. (l983b). Reply to Turner's comments. group, this will promote overt conflict and hostility of intergroup behavior that have often been ne­ the in-group, or even directly by the individual. European Journal ofSocial Psychology, 13, 360-381. between the groups. This prediction, like many glected in the past. Findings of this kind, which can be generalized Bourhis, R. Y., Giles, H., & Tajfel, H. (1973). Language as a others, still remains to be tested. A further distinction must be made between widely to many natural social situations, provide determinant of Welsh identity. European Journal ofSocial Psychology, 3,447-460. explicit and implicit conflicts-a distinction that a clear example of the need to introduce into the Brantwaite, A., Doyle, S., & Lightbown, N. (1979). The bal­ has to do with conflicts that are "objective" in a complex spiral of social causation the social-psy­ ance between fairness and discrimination. European Jour­ 'Objective' and 'Subjective' Conflicts different sense. A conflict may be "objective" de­ chological variables ofthe "relational" and "com­ nal ofSocial Psychology, 9, 149-163. spite the fact that the goals the groups are aiming parative" kind discussed in this chapter. Brewer, M. B. (1979). In-group bias in the minimal intergroup None of the arguments outlined in this chapter for have no value outside of the context of inter­ situation: A cognitive-motivational analysis. Psychologi­ must be understood as implying that the social­ cal Bulletin. 86(2), 307-324. group comparison in that it may be institutional. REFERENCES Brewer, M. B., & Silver, M. (1978). Ingroup bias as a func­ psychological or "subjective" type of conflict is ized and legitimized by rules and norms (of what­ tion of task characteristics. European Journal of Social Adorno. T. w.. Frenkel-Brunswik, E., Levinson, D. 1., & being considered here as having priority or a more 8(3), 393-400. ever origin) accepted by the groups themselves. Sanford, R. N. (1950). The aLtlhoritarian personality. New Psychology, Brigham, 1. C. (1971). Views of White and Black schoolchil­ important causal function in social reality than the This was the case in Sherif's studies in their phase York: Harper. 1950. dren conceming racial differences. Paper presented at the "objective" determinants of social conflict of of competition between the groups; and it also is Alien, V. L., & Wilder, D. A. (1975). Categorization, belief Midwestern Psychological Association, Detroit, Michigan. similarity, and intergroup discrimination. Joumal of Per­ which the basic analysis must be sought in the the case in any football match and in countless Brown, R.l., Tajfel, H., & Turner, 1. C. (1980). Minimal group sonality & Social Psychology, 32(6),971-977. social, economic, political, and historical structures situations and intergroup discrimination: Comments on the other social activities. The behavior toward out­ Aschenbrenner, K. M., & Schaefer, R. E. (1980). Minimal paper by Aschenbrenner and Schaefer. European Joumal of a society. The major aim of the present discus­ groups in this kind ofexplicit conflict can be clas­ intergroup situations: Comments on a mathematical model ofSocial Psychology, 10(4),399-414. sion has been to determine what are the points of sified, in turn, into two categories, one of which and on the research paradigm. European Joumal ofSocial Brown, R. J., & Turner, J. C. (1981). Interpersonal and inter­ insertion ofsocial-psychological variables into the Psychology, 10, 389-398. can be referred to as instrumental and the other as group behaviour. In J. C. Turner & H. Giles (Eds.), IlIter­ Bass, B. M., & Dunteman, G. (1963). Biases in the evalua­ causal spiral; and its argument has been that, just noninstrumental. The instrumental category con­ Oxford: Basil Blackwell. tion of one's own group, its allies, and opponents. Joul'llal group behaviour. as the effects of these variables are powerfully Caddick, B. (1980). Equity theory, social entity and inrergroup sists of all those actions that can be directly re­ ofConflict Resolwion, 2,67-77. relations. 1, determined by the previous social, economic, and lated to causing the group to win the competition. Baxter, G. W. (1973). Prejudiced liberals? Race and informa­ Review of Personaliry alld Social Psychology, 219-245. political processes, so they may also acquire, in The noninstrumental category, which could be re­ tion effects in a two-person game. Joumal ofConflict Reso­ Caddick, B. (1982). Perceived illegitimacy and intergroup lution, 17, 131-161. turn, an autonomous function that enables them ferred to as "gratuitous" discrimination against the relations. In H. Tajfel (Ed.), Social identity and illtergroup Berkowitz, L. (1962). Aggression: A social psychological to deflect in one direction or another the subse­ out-group, includes the creation of negative ste­ relatiol/s. Cambridge: Cambridge University Press. analysis. New York: McGraw-Hill. Campbell, D. T. (1965). Ethnocentric and other altruistic quent functioning of these processes. reotypes and all other aspects of the "irrelevant" Berko~itz, L. (1969). The frustration-aggression hypothesis motives. In D. Levine (Ed.), Nebraska Symposium Oil Mo­ It is nearly impossible in most natural social in-group/out-group differentiations so well de­ revisited. In L. Berkowitz (Ed.), Roots ofaggression: A re­ tivation (pp. 283-311). Lincoln: Nebraska: University of situations to distinguish between discriminatory examination ofthe frustration-aggression hypothesis. New scribed, for example, in Sherif's studies. The first Nebraska Press. York: Atherton Press. intergroup behavior based on real or perceived category ofactions is both commonsensically and Cederblom, D., & Diers, C.l. (1970). Effects of race and strat­ Berkowitz, L. (1974). Some determinants of impulsive ag­ conflict of objective interests between the groups egy in the prisoner's dilemma. theoretically accounted for by assuming nothing gression: Role of mediated associations with reinforcements Journal ofSocial Psychol­ 81, 275-276. and discrimination based on attempts to establish more than the group's desire to win the competi­ for aggression. Psychological Review, 81, 165-176. ogy, Commins, B., & Lockwood, 1. (1979). The effects of status a positively valued distinctiveness for one's own tion-although this poses all the theoretical "com­ Berry, 1. W., Kalin, R., & Taylor, D. M. (1976). differences, favoured treatment, and equity on intergroup Multiculturalism and ethnic aUill/des in Canada. Kingston, group. However, as we have argued, the two can parison" problems discussed in this chapter; the comparisons. Europeall Journal of Social Psychology, 9, be distinguished theoretically, since the goals of second category of actions can be directly and Ontario: Queen's University. Billig, M. (1972). Social categorization in illlergroup rela­ 281-290. Deutsch, M. (1949). A theory of cooperation and competi­ actions aimed at the achievement of positively parsimoniously accounted for in terms of the so­ tions. University of Bristol, Bristol. tion. Human Relations, 2, 129-151. valued in-group distinctiveness often retain no cial-comparison/social-identity/positive-in-group­ Billig, M. (1973). Normative communication in a minimal Diab, L. (1970). A study of intragroup and inrergroup rela­ value outside of the context of intergroup com­ intergroup situation. European Journal ofSocial Psychol­ distinctiveness sequence described here. tions among experimentally produced small groups. Ge­ ogy, 3(3), 339-343. parisons. An example would be a group that does The implicit conflicts are those that can be 82, 49-82. Billig, M. (1976). Social psychology GIld intergroup relatiolls netic Psychology Monographs, not necessarily wish to increase the level of its own Dizard, 1. E. (1970). Black identity, social class, and Black shown to exist despite the absence of explicit in­ (Vo!. 9). London: Academic Press. power. 33, 195-207. salaries but acts to prevent other groups from get- stitutionalization or even an informal normative Billig, M., & Tajfel, H. (1973). Social categorization and simi- Psychiatry, 292 • Political Psychology The Social Identity Theory of Intergroup Behavior • 293

Doise, W., & Weinberger, M. (1973). Representations Esquisse d'une theorie de l'originalite socialite. Bulle . a'fel H. (1974a). Intergroup behaviOl; social comparison behavior? A reply to Branthwaite, Doyle and Lightbown. masculines dans differentes situations de rencontres mixtes. de Psychologie, 252(20), 1-2,1-9. tin T ~Ild social change. Paper presented at the Katz-Newcomb European JOl/rHal ofSocial Psychology, 10(2), 131-147. Bulletin de Psychologie, 26,649-657. Locksley, A., Ortiz, v., .& ~epburn, C. (1980). Social Cate 0­ Lectures, University of Mlch~gan, Ann Arbor. . Turner. J. C. (1981). The experimental social psychology of Ferguson, C. K., & Kelley, H. H. (1964). Significant factors nzallon and dlscnmmatlon behaVIOur: ExtinguiShin gh Tajfel. H. (1974b). SOCIal Identity and Intergroup behavIOUr. intergroup behavior. In J. C. Turner & H. Giles (Eds.), In­ t in overevaluation of own group's products. Journal ofAb­ minimal intergroup discrimination effect. Joul'/lal oJR e social Science Informal/on, 1~(2), 65~93. . tergroup behavior. Oxford: Basil Blackwell. normal & Social Psychology, 69, 223-228. sonaliry and Social Psychology, 39, 773-783. er· T .~ I H. (1975). The exit of SOCIal mobIlIty and the vOice of Turner. J. C. (1982). Towards a cognitive redefinition of the Festinger, L. (1954). A theory ofsocial comparison processes. Miln.er, D. (1975). Children and race. HarmondsWOrth a~:Cial change: Notes on the social psychology of intergroup social group. In H. Tajfel (Ed.), Social identiry and inter­ Human Relations, 7, 117-140. Middlesex: Pengum. ' relations. Social Sciencelnfornll/tion. 14(2),.101-118.. group relarions (pp. 15-40). New York: Cambridge Uni­ Fiedler, F. E. (1967). The effect of inter-group competition Milner, D. (1981). Racial prejudice. In l. C. Turner & H. Giles Tajfel, H. (1978a). The achleve.ment of group dlffer.enuatlOn, versity Press. on group member adjustment. Personnel Psychology, 20( I), (Eds.), IntergroLlp behavior. Oxford: Basil Blackwell. DifferentiatIOn berween soclOl groups: SlUdles II! rhe so­ Turner, J. C. (1983a). A second reply to Bornstein, Crum, 33-44. Morland, 1. K. (I?69). Race awareness among American and cial psychology ofintergrollp relations. London: Academic Wittenbraker, Harring, Insko and Thibaut on the measure­ Friedman, . (1969). Africa and the Afro-Americans: The Hong Kong Chmese chIldren. American Journal ofSociol. Press. ment of social orientations. European JOl/mal of Social changing Negro identity. PsychiallY, 32(2), 127-136. ogy, 75(360-374). T 'fel, H. (1978b). The psychological structure of intergroup Psychology, 13(4),383-387. Giles, H., & Johnson, P. (1981). The role of language in eth­ Oakes, P. J., & Turner, J. C. (1982). Social categorization and a~elations. In H. Tajfel (Ed.), Differentiation berween social Turner, J. C. (1983b). Some comments on ... "the measure­ nic group relations. In J. C. Turner & H. Giles (Eds.), In­ intergroup behavio~r: Does minima~ ~ntergroup diScrimi_ gIVlIPS: Studies in the so~ial psychology of intergroup re­ ment ofsocial orientations in the minimal group paradigm." rergroup behavior. Oxford: Basil Blackwell. natIOn make SOCial Identity more pOSitive? European Jour­ lations. London: AcademiC Press. European JOl/mal ofSocial Psychology, 13(4),351-367. Giles, H., & Powesland, P. E. (1976). Speech style and social nal ofSocial Psychology, 10, 295-30 l. Tajfel, H. (1981). Human groups and social categories. Cam­ Turner, J. C. (1984). Social identification and psychological evaluarions. London: Academic Press, European Mono­ Oberschall, A. (1973). Social conflict and social movements bridge, UK: Cambridge University Press. group formation. In H. Tajfel (Ed.), The social dimension: graphs in Social Psychology. New York: Prentice-Hall. . Tajfel, H. (1982a). Social psychology of intergroup relations. European developments in social psychology. New York: Gregor, A. J., & McPherson, D. A. (1966). Racial preference Peabody, D. (1968). Group judgments in the Philippines: Allnl/al Re\'iew of Psychology, 33, 1-39. Cambridge University Press. and ego identity among White and Bantu children in the Evaluative and descriptive aspects. Journal of Personality Tajfel, H. (Ed.). (1982b). Social identiry and intergroup rela­ Turner, J. C., & Brown, R. (1978). Social status, cognitive Republic ofSouth Africa. Generic Psychology Monographs, and Social Psychology, 10,290-300. tions. Cambridge: Cambridge University Press. alternatives and intergroup relations. In H. Tajfel (Ed.), 73,217-254. Rabbie, J. M., & Wilkens, C. (1971). Intergroup competition Tajtel, H., & Billig, M. (1974). Familiarity and categoriza­ Differentiation between social groups: Swdies in the so­ Harris, S., & Braun, J. R. (1971). Self-esteem and racial pref­ and its effects on intra- and intergroup relations. European tion in intergroup behavlOr. Journal of Experunental So­ cial psychology ofintergroup relations (pp. 201-234). New erences in Black children. Proceedings ofrhe 79th Annual Journal ofSocial Psychology, 1, 215-234. cial Psychology, Vol. J0(2), 159-170. York: Academic Press. Convention ofthe American Psychological Association, 6. Rosenberg, M., & Simmons, R. G. (1972). Black and White Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Turner, J. c., & Brown, R. J. (1976). Social status, cognitive Harvey, O. J. (1956). An experimental investigation of nega­ self-esteem: The urban school child. Unpublished manu­ Social categorization and intergroup behaviour. European alternatives and intergroup relations. In H. Tajfel (Ed.), tive and positive relations between small groups through script. JOl/rnal ofSocial Psychology, Vol. 1(2), 149-178. Differentiation between social groups: Srudies in the social judgmental indices. Sociomerry, 19,201-209. Ross, G. F. (1979). Multiple group membership, social mo­ Thibaut, J. (1950). An experimental study of the cohesive­ psychology of intergroup relations. London: Academic Hirschman, A. O. (1970). Exit, voice and loyalry: Responses biliry and intergroup relations. Unpublished doctoral dis­ ness of underpriviledged groups. HI/man Relations, 3,251­ Press. to decline in firms, organizations and states. Cambridge, sertation, University of Bristol. 278. Turner, J. c., Brown, R. J., & Tajfel, H. (1979). Social com­ MA: Harvard University Press. Runciman, W. G. (1966). Relative deprivation and sodaljlls, Tomlinson, T. M. (1970). Contributing factors in Black poli­ parison and group interest in ingroup favouritism. Euro­ Hogg, M. A. (1983). The social psychology ofgroup-Jorma­ tice. London: Routledge and Keegan Paul. tics. Psychiatly, 33(2), 137-281. pean JOl/mal ofSocial Psychology, 9(2), 187-204. tion: A cognitive perspecrive. Unpublished doctoral disser­ Sherif, M. (1967). Social interaction: Process and prodllcts. Turner, J. C. (l975a). Social categori~ation of social com­ Turner, J. C., & Giles, H. (Eds.) (1981). Intergroup behavior. tation, University of Bristol. Sherif, M., Harvey, O. l., White, B. l., Hood, W. R., & Sherif, parison in inrergroup relations. Unpublished doctoral dis­ Oxford: Basil Blackwell. Hraba, J., & Grant, G. (1970). Black is beautiful: A reexami­ C. W. (1961).lntergroup cooperation and competition: The sertation, University of Bristol. Turner, J. c., Sachdev, I., & Hogg, M. A. (1983). Social cat­ nation of racial preference and identification. JOLlrnal of Robber's Cave experimelll. Norman, OK: University Book Turner. J. C. (1975b). Social comparison and social identity: egorization, interpersonal attraction and group formation. Personaliry & Social Psychology, 16(3),398-402. Exchange. Some prospects for intergroup behaviour. EI/ropean Jour­ British JourHal ofSocial Psychology, 22(3),227-239. Jahoda, G. (1961). White man. Oxford, England: Oxford Uni­ Sherif, M., & Sherif, C. W. (1953). Groups in harmony and Ilal ofSocial Psychology, 5(1), 5-34. Van Den Berghe, P. L. (1967). Race and racism. New York: vel'S ity Press. tension; an integration of studies of intergroup relations. Turner, J. C. (1978a). Social categorization and social dis­ Wiley. lohnson, D. W. (1967). Use of role reversal in intergroup com­ New York: Harper. crimination in the minimal group paradigm. In H. Tajfel Vaughan, G. M. (1978). Social change and intergroup prefer­ petition. Journal ofPersonality & Social Psychology, 7(2), Skevington, S. M. (1980). Intergroup relations and social (Ed.), Differelltiation between social grollps: Studies in the ences in New Zealand. European Jourtlal of Social Psy­ 135-141. change within a nursing context. British Journal ofSocial social psychology of intergroup relarions (pp. 101-140). chology, 8,297-314. Julian, J. W. (1968). The study of competition. In W. E. & Clinical Psychology, 19,201-213. London: Academic Press. Vinacke, W. E. (1964). Intra-group power differences, strat­ Vinacke (Ed.), Readings in general psychology. New York: St Claire, L., & Turner, J. C. (1982). The role of demand char­ Turner. J. C. (I 978b). Social comparison, similarity and egy, and decisions in inter-triad competition. Sociomerry, American Book Company. acteristics in the social categorization paradigm. Europeall ingroup favoritism. In H. Tajfel (Ed.), Differentiation be­ 27,27-40. Kalin, R., & Marlowe, D. (1968). The effects of intergroup Journal ofSocial Psychology, 12(3),307-314. rween social groups: Studies in the social psychology of Wilson, w., & Miller, N. (1961). Shifts in evaluations of par­ competition, personal drinking habits and frustration in Stephenson. G. M. (1981). Intergroup bargaining and nego­ intergroup relations (pp. 235-250). London: Academic ticipants following intergroup competition. Joumal ofAb­ intra-group cooperation. Proceedings ofthe American Psy­ tiation. In J. C. Turner & H. Giles (Eds.), Inrergrollp be­ Press. normal & Social PS.'ichology, 63,428-431. chological Association, 3, 405-406. ha viol'. Oxford: Basil Blackwell. Turner, J. C. (1980). Fairness or discrimination in intergroup Katz, 1. (1964). Review of evidence relating to effects of de­ Tajfel, H. (1959). A note on Lambert's "Evaluational reac­ segregation on the intellectual performance of Negroes. tions to spoken languages." Canadian Journal ofPsychol­ American Psychologist, 19(6), 381-399. ogy, 13, 86-92. Kidder. L. H., & Stewart, V. M. (1975). The psychology of Tajfel, H. (1970). Experiments in intergroup discrimination. intergroup relations. New York: McGraw-Hill. Scientific American. 223(5), 96-102. Klineberg, 0., & Zavalloni, M. (1969). Nationalism and tribal­ Tajfel, H. (1972a. Experiments in a vacuum. In J. Israel & H. ism among African students. The Hague and Paris: Mouton. Tajfel (Eds.), The context ofsocial psychology: A critical Lefcourt, H. M., & Ladwig, G. (1965). The effect of refer­ assessment. London: Academic Press, European Mono­ ence group upon Negroes' task persistence in a biracial graphs in Social Psychology. competitive game. Journal of Personality and Social Psy­ Tajfel. H. (1972b. La categorisation sociale. In M. S (Ed.), chology, I, 688-671. Introdllction iz la psychologie sociale (Vol. I). Paris: Lemaine, G. (1966). Inegalite, comparison et incomparabilite: Larousse.