Social Identity Theory and the Organization

7/28/2019 Social Identity Theory and the Organization

1/21

Social Identity Theory and the OrganizationAuthor(s): Blake E. Ashforth and Fred MaelSource: The Academy of Management Review, Vol. 14, No. 1 (Jan., 1989), pp. 20-39Published by: Academy of ManagementStable URL: http://www.jstor.org/stable/258189Accessed: 13/08/2009 11:53

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=aom.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the

scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that

promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].
http://www.jstor.org/stable/258189?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=aomhttp://www.jstor.org/action/showPublisher?publisherCode=aomhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/258189?origin=JSTOR-pdf


2/21

? Academy of Management Review, 1989, Vol. 14, No. 1, 20-39.

S o c i a l I d e n t i t y T h e o r ya n d t h e Organization

BLAKEE. ASHFORTHConcordia UniversityFREDMAEL

Wayne State UniversityIt is argued that (a) social identification is a perception of onenesswith a group of persons; (b) social identification stems from the cat-egorization of individuals, the distinctiveness and prestige of thegroup, the salience of outgroups, and the factors that traditionallyare associcr>ed with group formation; and (c) social identificationleads to activities that are congruent with the identity, support forinstitutions that embody the identity, stereotypical perceptions of selfand others, and outcomes that traditionally are associated withgroup formation, and it reinforces the antecedents of identification.This perspective is applied to organizational socialization, role con-flict, and intergroup relations.

Organizational identification has long beenrecognized as a critical construct in the literatureon organizational behavior, affecting both thesatisfaction of the individual and the effective-ness of the organization (Brown, 1969; Hall,Schneider, &Nygren, 1970;Lee, 1971;O'Reilly &Chatman, 1986; Patchen, 1970; Rotondi, 1975).However, as discussed below, theoretical andempirical work has often confused organiza-tional identification with related constructs suchas organizational commitment and internaliza-tion and with affect and behaviors, which aremore appropriately seen as antecedents and/orconsequences of identification.Social identity theory (SIT) can restore somecoherence to organizational identification, andit can suggest fruitful applications to organiza-tional behavior. SIToffers a social-psychologicalperspective, developed principally by HenriTajfel (1978, 1981;Tajfel &Turner, 1985)and JohnTurner (1975, 1982, 1984, 1985). Following a re-

view of the literature on SIT, the antecedentsand consequences of social identification in or-ganizations are discussed. This perspective isthen applied to three domains of organizationalbehavior: socialization, role conflict, and inter-group relations.

Social Identity TheoryAccording to SIT, people tend to classify them-selves and others into various social categories,such as organizational membership, religiousaffiliation, gender, and age cohort (Tajfel &Turner, 1985). As these examples suggest, peo-ple may be classified in various categories, anddifferent individuals may utilize different catego-rization schemas. Categories are defined byprototypical characteristics abstracted from themembers (Turner, 1985). Social classificationserves two functions. First, it cognitively seg-

20


3/21

ments and orders the social environment, pro-viding the individual with a systematic means ofdefining others. A person is assigned the proto-typical characteristics of the category to whichhe or she is classified. As suggested by the lit-erature on stereotypes, however, such assign-ments are not necessarily reliable (e.g., Hamil-ton, 1981).Second, social classification enables the indi-vidual to locate or define him- or herself in thesocial environment. According to SIT, the self-concept is comprised of a personal identity en-compassing idiosyncratic characteristics (e.g.,bodily attributes, abilities, psychological traits,interests) and a social identity encompassing sa-lient group classifications. Social identification,therefore, is the perception of oneness with orbelongingness to some human aggregate. Forexample, a woman may define herself in termsof the group(s) with which she classifies herself(I am a Canadian; I am a woman). She per-ceives herself as an actual or symbolic memberof the group(s), and she perceives the fate of thegroup(s) as her own. As such, social identifica-tion provides a partial answer to the question,Who am I?(Stryker & Serpe, 1982;Turner, 1982).Note that the definition of others and the selfare largely "relational and comparative" (Tajfel& Turner, 1985, p. 16); they define oneself rela-tive to individuals in other categories. The cate-gory of young is meaningful only in relation tothe category of old. It should be noted, however,that social identification is not an all-or-nonephenomenon. Although many social categoriesare indeed categorical (e.g., Canadian, female,a member of XYZ Co.), the extent to which theindividual identifies with each category isclearly a matter of degree. Further, such identi-ties tend to be viewed positively inasmuch as theindividual vests more of his or her self-conceptions in valued personas (Adler & Adler,1987; Schneider, Hall, & Nygren, 1971). Thus,Jackall (1978) found that people working at me-nial jobs in a bank often distanced themselvesfrom their implied identity (e.g., This is only a

stopgap job; I'mtrying to save enough to startmy own business).The major focus of both SITand the presentpaper is to understand the implications of thesecond function of classification, that of socialidentification.

Social Identification and Group IdentificationSocial identification appears to derive fromthe venerable concept of group identification(Tolman, 1943).(Indeed, we will use social and

group identificationinterchangeably.) The liter-ature on group identificationsuggests fourprin-ciples that are relevant to our discussion. First,identificationis viewed as a perceptual cogni-tive construct that is not necessarily associatedwith any specific behaviors or affective states.Toidentify,an individual need not expend efforttoward the group's goals; rather, an individualneed only perceive him- or herself as psycho-logically intertwined with the fate of the group.Behaviorand affect are viewed only as potentialantecedents or consequences (Foote, 1951;Gould, 1975).As noted below, this conceptual-ization distinguishes identificationfrom relatedconcepts such as effort on behalf of the group(behavior) and loyalty (affect). However, ourview does contrast with some literatureon SIT,which includes affective and evaluative dimen-sions in the conceptualization of identity (e.g.,Tajfel, 1978).Second, social/group identification is seen aspersonally experiencing the successes and fail-ures of the group (Foote, 1951;Tolman, 1943).Often, identificationis maintained in situationsinvolving great loss or suffering (Brown, 1986),missed potential benefits (Tajfel,1982), ask fail-ure (Turner, 1981), and even expected failure(Gammons, 1986).Third,although not clearly addressed in theliterature,social identification is distinguishablefrom internalization (Hogg & Turner, 1987)(cf.Kelman, 1961; O'Reilly & Chatman, 1986).Whereas identificationrefers to self in terms ofsocial categories (Iam), internalizationrefers to

21


4/21

the incorporation of values, attitudes, and soforth within the self as guiding principles (Ibelieve). Although certain values and attitudestypically are associated with members of agiven social category, acceptance of the cate-gory as a definition of self does not necessarilymean acceptance of those values and attitudes.An individual may define herself in terms of theorganization she works for, yet she can disagreewith the prevailing values, strategy, system ofauthority, and so on (cf. "young Turks," Mintz-berg, 1983, p. 210; "counterculture," Martin &Siehl, 1983, p. 52).Finally, identification with a group is similar toidentification with a person (e.g., one's father,football hero) or a reciprocal role relationship(e.g., husband-wife, doctor-patient) inasmuchas one partly defines oneself in terms of a socialreferent. To be sure, the various literaturesreach this conclusion from different directions.Whereas identification with a group is argued tobe predicated on the desire for self-definition,identification with an individual-referred to as"classical identification" (Kelman, 1961, p. 63)is argued to be predicated on the desire to ap-pease, emulate, or vicariously gain the qualitiesof the other (e.g., Bandura & Walters, 1963; Ketsde Vries & Miller, 1984). Kelman (1961), for ex-ample, argued that in classical identification theindividual "attempts to be like or actually to bethe other person" (p. 63). Nevertheless, the ele-ment of self-definition suggests that these formsof identification are complementary. Indeed, wewill suggest that organizations often seek to gen-eralize identification with an individual to iden-tification with the organization through the rou-tinization of charisma.Social Identification and the OrganizationThe individual's organization may provideone answer to the question, Who am I? Hence,we argue that organizational identification is aspecific form of social identification. This searchfor identity calls to mind a family of existentialmotives often alluded to in the literature on or-

ganizational behavior, including searches formeaning, connectedness, empowerment, andimmortality (e.g., Denhardt, 1987; Fox, 1980;Katz & Kahn, 1978). To the extent the organiza-tion, as a social category, is seen to embody oreven reify characteristics perceived to be proto-typical of its members, it may well fulfill suchmotives for the individual. At the very least, SITmaintains that the individual identifies with so-cial categories partly to enhance self-esteem(Hogg & Turner, 1985; Tajfel, 1978). This is un-derstandable in view of the relational and com-parative nature of social identities. Through so-cial identification and comparison, the individ-ual is argued to vicariously partake in thesuccesses and status of the group: Indeed, pos-itive and negative intergroup comparisons havebeen found to affect a member's self-esteem ac-cordingly (Oakes & Turner, 1980;Wagner, Lam-pen, & Syllwasschy, 1986).

The individual's social identity may be de-rived not only from the organization, but alsofrom his or her work group, department, union,lunch group, age cohort, fast-track group, andso on. Albert and Whetten (1985) distinguishedbetween holographic organizations in which in-dividuals across subunits share a common iden-tity (or identities) and ideographic organizationsin which individuals display subunit-specificidentities. General examples of the former in-clude Ouchi's (1981) Theory Z organization inwhich "management styles are blended to-gether and diffused evenly throughout the entireorganization" (Albert & Whetten, 1985, p. 271)and Mintzberg's (1983) missionary organizationin which members strongly subscribe to a com-mon set of values and beliefs. Given the com-parative rarity of such organizations, however,the notion of a single or blended organizationalidentification is problematic in most complex or-ganizations. Thus, as discussed below, the or-ganizationally situated social identity may, infact, be comprised of more or less disparate andloosely coupled identities. This parallels work invarious social domains which indicates that in-

22


5/21

dividuals often retain multiple identities (Allen,Wilder, & Atkinson, 1983; Hoetler, 1985; Thoits,1983).Unfortunately, despite the longevity of thesocial/group identification construct, little re-search has been conducted on identificationwith organizations, as defined here. Conven-tional research on organizational identificationhas not distinguished identification from inter-nalization or cognition from behavior and affect.For example, Hall et al. (1970) defined organiza-tional identification as "the process by which thegoals of the organization and those of the indi-vidual become increasingly integrated andcongruent" (pp. 176-177), and Patchen (1970)defined it as shared characteristics, loyalty, andsolidarity. The lone exception is a study byO'Reilly and Chatman (1986) that distinguishedamong compliance, identification, and internal-ization. However, following Kelman's (1961)lead, they defined identification as "involve-ment based on a desire for affiliation" (p. 493),rather than as perceived oneness with the orga-nization.A particular problem in this area is the fre-quent confusion between organizational identi-fication and organizational commitment. Sometheorists equate identification with commitment,while others view the former as a component ofthe latter (see Wiener, 1982). The authors of theOrganizational Commitment Questionnaire(OCQ) (Mowday, Steers, & Porter, 1979, p.226)-the most frequently used measure of com-mitment during the last decade (Reichers,1985)-defined organizational commitment as"the relative strength of an individual's identifica-tion with and involvement in a particularorganization." In their view, commitment is char-acterized by a person's (a) belief in and accep-tance of the organization's goals and values, (b)willingness to exert efforton behalf of the organi-zation, and (c) desire to maintain membership.This formulation includes internalization, behav-ioral intentions, and affect, but not identification aspresently defined. Further, although identification

is defined as organization-specific, internaliza-tion and commitment may not be. An organiza-tion's goals and values may be shared by otherorganizations. Commitment scales consistentlyfeature generalized usage of the terms goalsand values, as in the OCQ item, "Ifind that myvalues and the organization's values aresimilar" (Mowday et al., 1979, p. 228). Respon-dents are not asked to limit responses to valuesthat are specific to their organization, if indeedthey could. Thus, an individual can score highon commitment not because he or she perceivesa shared destiny with the organization but be-cause the organization is a convenient vehiclefor personal career goals. If another organiza-tion proved more convenient, such an individ-ual could transfer to it without sacrificing his orher goals. For the individual who identified withthe organization, however, leaving the organi-zation necessarily involves some psychic loss(e.g., Levinson, 1970).This argument is supported by Mael's (1988)study of employed business and psychology stu-dents. He constructed a 6-item measure of orga-nizational identification based on the presentformulation (e.g., "Thisorganization's successesare my successes," p. 52), and subjected it andthe 15-item OCQ to confirmatory factor analysis.The two-factor model produced a X2ldf ratio of2.03:1 (i.e., 328.13/188) and an adjusted good-ness-of-fit index of .825; the single-factor modelproduced a ratio of 2.46:1 (i.e., 465.14/189) andan index of .780. The superior fit of the two-factormodel suggests that the identification and com-mitment constructs are indeed differentiable.In summary, the SIT conception of organiza-tional identification as shared identity is new tothe organizational behavior literature. To date,the perception of identification has been con-fused with internalization of organizationalgoals and values, and with behavior and affect.This is most clearly evident in research on orga-nizational commitment. Unfortunately, this con-fusion has impeded application of the rich find-ings of SITto organizations.

23


6/21

Antecedents and Consequencesof Social Identificationin Organizations

AntecedentsSIT is contradictory to conventional views ofgroup relations because according to it in-groupfavoritism tends to occur even in the absence ofstrong leadership or member interdependence,interaction, or cohesion. Laboratory studies uti-

lizing SIT'sminimal group paradigm have dem-onstrated that simply assigning an individual toa group is sufficient to generate in-group favor-itism (Brewer, 1979; Tajfel, 1982). Favoritism isnot dependent on prior perceptions of interper-sonal similarity or liking, and it occurs evenwhen there is no interaction within or betweengroups, when group membership is anony-mous, and when there is no link between self-interest and group responses (Turner, 1984).Even explicitly random assignment of individu-als to groups has led to discrimination againstout-groups and increased intragroup coopera-tion and cohesion (e.g., Billig & Tajfel, 1973;Locksley, Ortiz, & Hepburn, 1980).This led Turner (1984, p. 530) to propose theexistence of a "psychological group," which hedefined as "a collection of people who share thesame social identification or define themselvesin terms of the same social category member-ship." A member of a psychological group doesnot need to interact with or like other members,or be liked and accepted by them. It is his or herperception of being, say, a loyal patriot or sportsfan that is the basis for incorporation of that sta-tus into his or her social identity. The individualseems to reify or credit the group with a psycho-logical reality apart from his or her relationshipswith its members (Turner, 1984).The SIT literature suggests several factors ofdirect relevance to organizations which mostlikely increase the tendency to identify withgroups. The first is the distinctiveness of thegroup's values and practices in relation to those

of comparable groups (Oakes & Turner, 1986;Tolman, 1943). Distinctiveness serves to sepa-rate "figure from ground," differentiating thegroup from others and providing a unique iden-tity. Mael (1988) sampled the alumni of a reli-gious college and found a positive associationbetween the perceived distinctiveness of the col-lege's values and practices and identificationwith the college. Distinctiveness partly explainsthe missionary zeal often displayed by membersof organizations that are new and innovative(e.g., Perkins, Nieva, & Lawler, 1983)or organi-zations that pursue unique goals (e.g., Hall etal.'s 1970 study of the U.S. Forest Service).Within the organization, distinctiveness ingroup values and practices needs to be quali-fied by the clarity and impermeability of groupdomains or boundaries. For example, althoughit is likely that the values and practices of twofunctionally based subunits are more differenti-ated than those of two market-based subunits,suggesting distinctiveness, the former are morelikely to be sequentially or reciprocally interde-pendent and physically contiguous, suggestinga blurring of distinctiveness. This indeterminatedistinctiveness may account for the mixed sup-port for SIT in several field studies (Brown, Con-dor, Mathews, Wade, & Williams, 1986; Brown& Williams, 1984; Oaker & Brown, 1986; Skev-ington, 1981).Interestingly, even negatively valued distinc-tions have been associated with identification.Negatively regarded groups often utilize suchdefense mechanisms as recasting a negativedistinction into a positive one (Black is beautiful),minimizing or bolstering a negative distinction(We're not popular because we avoid playingpolitics), or changing the out-group with whichthe in-group is compared (Lemaine, Kastersz-tein, & Personnaz, 1978;Skevington, 1981; Wag-ner et al., 1986) (cf. social creativity, Tajfel &Turner, 1985). And the stronger the threat to thegroup, the stronger the defensive bias (vanKnippenberg, 1984). Such machinations mightpartly explain a person's often fierce identifica-

24


7/21

tion with countercultures (e.g., Martin & Siehl,1983) or disaffected groups in organizations(e.g., Jackall, 1978).A second and related factor that increasesidentification is the prestige of the group (Chat-man, Bell, & Staw, 1986;March & Simon, 1958).This is based on the earlier argument that,through intergroup comparison, social identifi-cation affects self-esteem. Mael (1988)found thatperceived organizational prestige was related toorganizational identification among samples ofworking university students and religious col-lege alumni. Individuals often cognitively (if notpublicly) identify themselves with a winner. Thisaccounts in part for the bandwagon effect oftenwitnessed in organizations, where popular sup-port for an individual or idea suddenly gainsmomentum and escalates, thus creating a risingstar. Desires for positive identifications effec-tively create champions, converting "the slight-est sign of plurality into an overwhelmingmajority" (Schelling, 1957, p. 32).Third, identification is likely to be associatedwith the salience of the out-group(s) (Allen et al.,1983; Turner, 1981).Awareness of out-groups re-inforces awareness of one's in-group. Wilder(cited in Wilder, 1981)categorized one set of sub-jects into two groups (in-group/out-group condi-tion), allegedly on the basis of preference forcertain paintings, and a second set into onegroup (in-group-only condition). Subjects as-sumed greater homogeneity in the in-group whenan out-group was present (in-group/out-groupcondition) than when no specific out-group wassalient (in-group-only condition). Awareness ofthe out-group underscored the existence of aboundary and caused subjects to assume in-group homogeneity. Similarly, Kanter (1977)found that the presence of females in a male-dominated sales force induced the males to ex-aggerate perceived masculine traits and differ-ences between the sexes.The well-known effects of intergroup competi-tion on in-group identification (e.g., Friedkin &Simpson, 1985) are a special case of this princi-

ple. During competition, group lines are drawnmore sharply, values and norms are under-scored, and we/they differences are accentu-ated (Brown & Ross, 1982; van Knippenberg,1984) (cf. cognitive differentiation hypothesis,Dion, 1979). Skevington (1980), for example,found that when high-status nurses (where sta-tus was based on training) were led to believethey would be merged with low-status nurses,they increased their in-group favoritism, em-phasizing their distinctiveness and superiorityover the low-status group.Finally, the set of factors traditionally associ-ated with group formation (interpersonal inter-action, similarity, liking, proximity, shared goalsor threat, common history, and so forth) mayaffect the extent to which individuals identifywith a group, although SIT suggests that theyare not necessary for identification to occur. Itshould be noted, however, that although thesefactors facilitate group formation, they also maydirectly cue the psychological grouping of indi-viduals since they can be used as bases for cat-egorization (Hogg & Turner, 1985;Turner, 1984).In complex organizations, the pervasivenessof this set of antecedents-the categorization ofindividuals, group distinctiveness and prestige,out-group salience, and group formation fac-tors-suggests that group identification is likelyto be prevalent. Also, although the SITliteratureindicates that categorization is sufficient foridentification to occur, the pervasiveness of for-mal and informal groups in organizations sug-gests that categorization is seldom the only fac-tor in identification. Thus, the consequences ofidentification suggested by SIT, discussed be-low, may well be intensified in organizations.ConsequencesThe SIT literature suggests three general con-sequences of relevance to organizations. First,individuals tend to choose activities congruentwith salient aspects of their identities, and theysupport the institutions embodying those identi-ties. Stryker and Serpe (1982) found that individ-

25


8/21

uals for whom a religious role was salient re-ported spending more time in that role and de-riving satisfaction from it, and Mael (1988)foundthat the identification of alumni with their almamater predicted their donating to that institution,their recruiting of offspring and others, their at-tendance at functions, and their satisfaction withthe alma mater. Thus, it is likely that identifica-tion with an organization enhances support forand commitment to it.A second and related consequence is that so-cial identification affects the outcomes conven-tionally associated with group formation, in-cluding intragroup cohesion, cooperation, andaltruism, and positive evaluations of the group(Turner, 1982, 1984). It is also reasonable to ex-pect that identification would be associated withloyalty to, and pride in, the group and its activ-ities. However, it should be noted that, given ourdiscussion of psychological groups, this affinityneed not be interpersonal or based on interac-tion. Dion (1973)demonstrated that one may likeother group members, despite their negativepersonal attributes, simply by virtue of the com-mon membership (cf. personal vs. social attrac-tion, Hogg & Turner, 1985). In short, "one maylike people as group members at the same timeas one dislikes them as individual persons"(Turner, 1984, p. 525).

Identification also may engender internaliza-tion of, and adherence to, group values andnorms and homogeneity in attitudes and behav-ior. Just as the social classification of others en-genders stereotypical perceptions of them, sotoo does the classification of oneself and subse-quent identification engender the attribution ofprototypical characteristics to oneself (Turner,1984, 1985). This self-stereotyping amounts to de-personalization of the self (i.e., the individual isseen to exemplify the group), and it increasesthe perceived similarity with other group mem-bers and the likelihood of conformity to groupnorms.Finally, it is likely that social identification willreinforce the very antecedents of identification,including the distinctiveness of the group's val-

ues and practices, group prestige, salience ofand competition with out-groups, and the tradi-tional causes of group formation. As the individ-ual comes to identify with the group, the valuesand practices of the in-group become more sa-lient and perceived as unique and distinctive(e.g., Tajfel, 1969).Perhaps the greatest contribution that SITmakes to the literature on organizational behav-ior is the recognition that a psychological groupis far more than an extension of interpersonalrelationships (Turner, 1985):Identification with acollectivity can arise even in the absence of in-terpersonal cohesion, similarity, or interactionand yet have a powerful impact on affect andbehavior. As discussed below, in crediting acollectivity with a psychological reality beyondits membership, social identification enables theindividual to conceive of, and feel loyal to, anorganization or corporate culture. Indeed,Turner (1982) claimed that "social identity is thecognitive mechanism which makes group be-haviour possible" (p. 21).

Applying Social Identity Theoryto OrganizationsThe explanatory utility of SIT to organizationscan be illustrated by applications to organiza-tional socialization, role conflict, and intergrouprelations.

Organizational SocializationAccording to the literature on organizationalsocialization, organizational newcomers arehighly concerned with building a situationaldefinition (Katz, 1980). Newcomers, it is argued,are unsure of their roles and apprehensiveabout their status. Consequently, in order to un-derstand the organization and act within it, theymust learn its policies and logistics, the generalrole expectations and behavioral norms, thepower and status structures, and so forth (Ash-forth, 1985).However, organizational newcomers also

26


9/21

are often concerned with building a self-definition, of which the social identity (or identi-ties) is likely to comprise a large part. For manyyears, writers in the personological tradition ofpersonality theory have noted the link betweensocialization and the self-concept, suggestingthat the emergence of situational and self-definitions are intertwined (see Hogan, 1976). Adeveloping sense of who one is complements asense of where one is and what is expected. Incomplex organizations, the prevalence of socialcategories suggests that social identities arelikely to represent a significant component of in-dividuals' organizationally situated self-defini-tions, and, indeed, many studies document thisidea (see Fisher, 1986; Mortimer & Simmons,1978;and Van Maanen, 1976, for examples).Developing Social Identifications. Althoughthe SIT literature is relatively mute about howsocial identification occurs, the literature on or-ganizational socialization suggests that situa-tional definitions and self-definitions bothemerge through symbolic interactions (Ashforth,1985;Coe, 1965;Reichers, 1987). Symbolic inter-actionism holds that meaning is not a given butevolves from the verbal and nonverbal interac-tions of individuals. For our purposes, interac-tion is defined broadly to include any symbol-ic transmission, from product advertisementsto orientation sessions. (As the SIT literaturereviewed above makes clear, interaction neednot be interpersonal-though in organizations,of course, it often is.) Through symbolic inter-actions the newcomer begins to resolve ambi-guity, to impose an informational framework orschema on organizational experience.With regard to self-definitions in particular,Van Maanen (1979) argued that conceptions ofthe self are learned by interpreting the re-sponses of others in situated social interactions.Drawing on the works of Charles Horton Cooley,George Herbert Mead, and Herbert Blumer,among others, he maintained that through in-teractions individuals learned to ascribe sociallyconstructed labels such as ambitious, engineer,and upwardly mobile to themselves and others.

An example was provided by Becker andCarper (1956). They interviewed graduate stu-dents in physiology, most of whom initiallyviewed physiology as a stopgap pending accep-tance into medical school. Becker and Carperfound, however, that through immersion in thesocial milieu many students gradually assumedthe identity of physiologists. Frequent interac-tion and social comparison with fellow students,observation of professors, and tutelage and re-inforcement by professors slowly shaped stu-dents' interests, skills, self-conceptions, andtheir understanding of the paradigms, values,norms, and occupational choices in the field.This perspective on social identification in or-ganizations suggests at least three implications.First, consistent with our earlier discussion, itsuggests that the often-noted effect of organiza-tional socialization on the internalization of or-ganizational values and beliefs is comprised inpart of an indirect effect via identification; that is,socialization effects identification, which in turneffects internalization. As noted, through self-stereotyping the individual typically adoptsthose characteristics perceived as prototypicalof the groups with which he or she identifies.Albert and Whetten (1985) argued that an orga-nization has an identity to the extent there is ashared understanding of the central, distinctive,and enduring character or essence of the orga-nization among its members. This identity maybe reflected in shared values and beliefs, a mis-sion, the structures and processes, organiza-tional climate, and so on. The more salient, sta-ble, and internally consistent the character of anorganization (or in organizational terms, thestronger the culture), the greater this internaliza-tion (Ashforth, 1985).However, socialization also has a direct effecton internalization, as suggested by the argu-ment that one may internalize an organization'sculture without necessarily identifying with theorganization, and vice versa. The relative im-portance of the direct (socialization -> internal-ization) and indirect (socialization -> identifica-tion -> internalization) effects most likely vary

27


10/21

across organizations, subunits, and roles. VanMaanen (1978) distinguished between investi-ture processes that ratify the newcomer's incom-ing identity and divestiture processes that sup-plant the incoming identity with a new organi-zationally situated identity. Total and quasi-totalinstitutions such as prisons, military and reli-gious organizations, professional schools, andorganizational clans provide prime examples ofdivestiture. In order to reconstruct the newcom-er's social identity, such organizations often re-move symbols of newcomer's previous identi-ties; restrict or isolate newcomers from externalcontact; disparage newcomer's status, knowl-edge, and ability; impose new identificationsymbols; rigidly prescribe and proscribe behav-ior and punish infractions; and reward assump-tion of the new identity (Fisher, 1986; Goffman,1961; Van Maanen, 1976, 1978). In such casesinternalization of organizational values dependslargely on the extent of identification with theorganization, subunit, or role. Indeed, the morethe organization's identity, goals, values, andindividual role requirements deviate from thesocietal mainstream, the greater the need fororganizationally situated identification.A second implication of the social identifica-tion perspective stems from the notion of reifica-tion. The existing organizational behavior liter-ature does not adequately explain how an indi-vidual can identify with, or feel loyal and com-mitted to, an organization per se. The implicitassumption is that regard for individuals simplygeneralizes to the group, that interpersonal re-lationships somehow are cognitively aggregatedto create an individual-organization relationship(Turner, 1984). We reverse this logic and arguethat identification with a group can arise quiteseparately from interpersonal interaction and co-hesion. In perceiving the social category as psy-chologically real-as embodying characteristicsthought prototypical of its members-the individ-ual can identify with the category per se (Iam aMarine). Thus, identification provides a mecha-nism whereby an individual can continue to

believe in the integrity of his or her organizationdespite wrongdoing by senior management orcan feel loyal to his or her department despite acomplete changeover of personnel.Third, the social identification perspectivealso helps to explain the growing interest insymbolic management (Pfeffer, 1981) and char-ismatic or transformational leadership (Bass,1985). To the extent that social identification isrecognized by managers to relate to such criticalvariables as organizational commitment andsatisfaction, managers have a vested interest inmanaging symbolic interactions. Although thecoherence of a group's or organization's identityis problematic, we believe that symbolic man-agement is designed to impart this identity, orat least management's representation of it.Through the manipulation of symbols such astraditions, myths, metaphors, rituals, sagas, he-roes, and physical setting, management canmake the individual's membership salient andcan provide compelling images of what thegroup or organization represents (Pondy, Frost,Morgan, & Dandridge, 1983).Interestingly, Martin, Feldman, Hatch, andSitkin (1983) noted that organizational cultures"carry a claim to uniqueness-that one institu-tion is unlike any other" (p. 438). We contendthat it is precisely because identification isgroup-specific that organizations make suchclaims. It is tacitly understood by managers thata positive and distinctive organizational identityattracts the recognition, support, and loyalty ofnot only organizational members but other keyconstitutents (e.g., shareholders, customers, jobseekers), and it is this search for a distinctiveidentity that induces organizations to focus sointensely on advertising, names and logos, jar-gon, leaders and mascots, and so forth.This link between symbolism and identifica-tion sheds light on the widespread interest incharismatic leaders. Because charismatic lead-ers are particularly adept at manipulating sym-bols (Bass, 1985), they are likely to engender so-cial and/or classical identification, that is, iden-

28


11/21

tification with the organization, the leader, orboth. Where the identification is classical, it maybe generalized to the organization through theroutinization of charisma (Gerth & Mills, 1946).Trice and Beyer (1986) contrasted the develop-ment of two social movement organizationsfounded by charismatic individuals: AlcoholicsAnonymous (AA) and the National Council onAlcoholism (NCA). The charisma of the AA'sfounder was routinized through an administra-tive structure, rites and ceremonies, oral andwritten tradition, and so forth, whereas the cha-risma of the NCA's founder was poorly routin-ized. The result, concluded Trice and Beyer, isthat the NCA has experienced greater difficultymaintaining the support of its members and do-nors.Identification and the Subunit. It should benoted, however, that the newcomer's emergingsituational definitions and self-definitions areapt to be largely subunit-specific. First, task in-terdependencies and interpersonal proximityare greater in the individual's immediate workgroup, suggesting a greater need for, and easeof, interaction. Second, given that people preferto compare their emerging beliefs with similarothers (cf. social comparison theory, Festinger,1954) and that interpersonal and task differenti-ation are greater between, than within, subunits(Lawrence & Lorsch, 1967), it is likely that thenewcomer will look first to his or her workgrouppeers. Third, given interdependence, proximity,and similarity, the subunit may be viewed bymembers as a psychological group, thus facili-tating social influence. According to Turner(1985; Hogg & Turner, 1987), the self-stereo-typing occasioned by psychological groupingcauses one to expect attitudinal and perceptualagreement with group members, such that dis-agreement triggers doubt and, in turn, at-titudinal/perceptual change. Thus, the newcom-er's perceptions gravitate toward those of thegroup. Finally, given the importance of the sit-uational definition to job performance and thecentrality of the social identity to the self-

concept, it is likely that a normative structurewill emerge to regulate and maintain these con-ceptions. This is consistent with Sampson's(1978) proposition that people attempt to man-age their lives in order to establish a sense ofcontinuity in their identity (identity mastery). Theupshot is that immediate groups often are moresalient "than a more abstract, complex, second-ary organization" (Brown, 1969, p. 353).Organizational socialization, then, can beseen under the SITperspective as an attempt tosymbolically manage newcomers' self-, if not sit-uational, definitions by defining the organiza-tion or subunit in terms of distinctive and endur-ing central properties. Identification with the or-ganization provides (a) a mechanism wherebythe individual can reify the organization andfeel loyal and committed to it per se (i.e., apartfrom its members) and (b) an indirect paththrough which socialization may increase theinternalization of organizational values and be-liefs.Role Conflict

Given the number of groups to which an in-dividual might belong, his or her social identityis likely to consist of an amalgam of identities,identities that could impose inconsistent de-mands upon that person. Further, these de-mands also may conflict with those of the indi-vidual's personal identity (Cheek &Briggs, 1982;Leary, Wheeler, & Jenkins, 1986). Note that it isnot the identities per se that conflict, but the val-ues, beliefs, norms, and demands inherent inthe identities.In organizations, conflicts between work-group, departmental, divisional, and organiza-tional roles are somewhat constrained by thenested character of these roles; that is, each hi-erarchical level encompasses the former suchthat the roles are connected in a means-endchain (March & Simon, 1958). Accordingly, thevalues and behavioral prescriptions inherent inthe organizational role tend to be a more ab-

29


12/21

stract and generalized version of those inherentin the workgroup role. Nevertheless, evennested identities can be somewhat at odds withone another (Rotondi, 1975; Turner, 1985; VanMaanen, 1976). In the course of assuming agiven identity (e.g., department), the group be-comes more salient and both intragroup differ-ences and intergroup similarities are cognitivelyminimized, thus rendering both lower order(e.g., workgroup) and higher order (e.g., orga-nization) identifications less likely. Also, giventhe association between identification and inter-nalization, a lack of congruence between thegoals or expectations of nested groups may im-pede joint identification. Not surprisingly, then,Brown (1969) found that task interdependenciesand the cohesion of the individual's functionalunit were negatively related to organizationalidentification or internalization.We speculate that the inherent conflict be-tween organizationally situated identities typi-cally is not resolved by integrating the disparateidentities. First, given the breadth of possibleidentities, integration would most likely provecognitively taxing. Second, given the oftenunique and context-specific demands of anidentity, integration would be likely to compro-mise the utility of each identity to its particularsetting. Instead, it is maintained that conflict be-tween identities tends to be cognitively resolvedby ordering, separating, or buffering the identi-ties. Suggestions of such processes abound.First, the individual might define him- or herselfin terms of his or her most salient social identity(Iam a salesman) or personal attribute (I want toget ahead); he or she also might develop a hi-erarchy of prepotency so that conflicts are re-solved by deferring to the most subjectively im-portant or valued identity (Stryker &Serpe, 1982;Thoits, 1983). Adler and Adler (1987) describedhow varsity basketball players resolved the con-flict between their athletic and academic rolesby defining themselves as athletes first and stu-dents second and by reducing their involvementin academics accordingly. Second, the individ-

ual might defer to the identity that experiencesthe greatest environmental press and mightminimize, deny, or rationalize the conflict (If Ihadn't bribed the official, I would have lost thecontract). This is akin to Janis and Mann's (1977)notion of defensive avoidance. Third, the indi-vidual might cognitively decouple the identitiesso that conflicts simply are not perceived (cf.value separation, Steinbruner, 1974). Laurent(1978) discussed how managers often are reluc-tant to inform subordinates about critical mat-ters, yet as subordinates, they complain aboutthe failure of their own managers to informthem. Finally, the individual might comply se-quentially with conflicting identities so that theinconsistencies need not be resolved for anygiven action (cf. sequential attention, Cyert &March, 1963).An example is provided by MortonThiokol, the manufacturer of the faulty solidrocket booster that led to the 1986 crash of thespace shuttle Challenger. A senior engineer ofthe company helped reverse a decision not tolaunch the Challenger when he was asked to"take off his engineering hat and put on hismanagement hat" (Presidential Commission,cited in Vaughan, 1986, p. 23).Related to this idea, Thoits (1983) suggestedthat the benefits of holding multiple roles (roleaccumulation), including resource accumula-tion, justification for failure to meet certain roleexpectations, and support against role failure orloss, are more likely to accrue if identities re-main segregated: "The actor's resources will bevaluable to others who do not share those re-sources themselves, the legitimacy of excusescannot be checked, and the consequences ofrole failure or loss can be contained more withinone sphere of activities" (p. 184).

To the extent this argument is valid, it suggeststhat one's identity is an amalgam of loosely cou-pled identities and that "the popular notion ofthe self-concept as a unified, consistent, or per-ceptually 'whole' psychological structure is pos-sibly ill-conceived" (Gergen, 1968, p. 306). Thisis consistent with evidence from SIT that partic-

30


13/21

ular social identities are cued or activated byrelevant settings (Turner, 1982, 1985) (cf. situa-tional identity, Goffman, 1959;subidentity, Hall,1971;hard vs. soft identity, Van Maanen, 1976).Most individuals slide fairly easily from oneidentity to another. Conflict is perceived onlywhen the disparities are made salient (Greene,1978). Thus, in SFT1ole conflict is endemic to so-cial functioning, but for the most part remainslatent: Only when individuals are forced to si-multaneously don different hats does their facil-ity for cognitively managing conflict breakdown.The argument also suggests that when an in-dividual compartmentalizes identities, he or shemay fail to integrate the values, attitudes,norms, and lessons inherent in the various iden-tities. This in turn suggests the likelihood of (a)double-standards and apparent hypocrisy (illus-trated by Laurent's, 1978, observation) and (b)selective forgetting. For example, in assumingthe identity of foreman, one may eventually for-get the values that were appropriate to the prioridentity of worker that now contradict the de-mands of the new identity (e.g., Lieberman,1956);that is, one unlearns tendencies that inter-fere with the ability to embrace the new, valuedidentity. Perhaps, then, wisdom is little morethan the ability to remember the lessons of pre-vious identities, and integrity is the ability to in-tegrate and abide by them.IntergroupRelations

For pedagogical purposes, we assume anideographic organization, that is, one com-prised of subunits in which members of eachshare a social identity specific to their subunit.This assumption allows us to speak of a sharedsubunit or group identity, even though in com-plex organizations the degree and foci of con-sensus remains problematic.Given this assumption, SIT suggests thatmuch intergroup conflict stems from the veryfact that groups exist, thus providing a fairlypessimistic view of intergroup harmony (Tajfel,

1982). More specifically, in SIT it is argued that(a) given the relational and comparative natureof social identifications, social identities aremaintained primarily by intergroup compari-sons and (b) given the desire to enhance self-esteem, groups seek positive differences be-tween themselves and reference groups (Tajfel,1978, 1981;Smith, 1983). Experimental and fieldresearch do suggest that groups are willing tosacrifice large monetary gains that do not estab-lish a positive difference between groups forsmaller gains that do (Brewer & Silver, 1978;Brown, 1978;Turner, Brown, &Tajfel, 1979), thatin-group members adopt more extreme posi-tions after comparison with an out-group thanwith fellow in-group members (Reid, 1983), andthat members prefer and selectively recall infor-mation that suggests intergroup differencesrather than similarities (Wilder, cited in Wilder,1981; Wilder & Allen, 1978). This suggests thatgroups have a vested interest in perceiving oreven provoking greater differentiation than ex-ists and disparaging the reference group on thisbasis (cf. social vs. instrumental competition,Turner, 1975). Further, this tendency is exacer-bated by contingencies that make the in-groupper se salient (Turner, 1981; Wilder, 1981), suchas a threat to the group's domain or resources(Brown & Ross, 1982; Brown et al., 1986) or, inTajfel's (1978)terms, where the group's identity isinsecure.The tendency toward subunit identification inorganizations, discussed above, suggests thatsubunits tend to be the primary locus of inter-group conflict. This tendency is exacerbated bycompetition between subunits for scarce re-sources and by reward and communication sys-tems that typically focus on subunit functioningand performance (Friedkin & Simpson, 1985;March & Simon, 1958). As noted, however, fieldresearch regarding the relationship betweensubunit differentiation and identification hasbeen inconclusive because it has confoundedthe basis of subunit formation (functional vs.market) and extent of interdependence (pooled,

31


14/21

sequential, reciprocal). Further, Brown andWilliams (1984) suggested that individuals whoregard their group identity as synonymous withtheir organizational identity are unlikely to viewother groups negatively. Just as a strong groupidentity unifies group members, so too should astrong organizational identity unify organiza-tional members. This is consistent with experi-mental research (Kramer & Brewer, 1984) andthe earlier discussion of holographic organiza-tions.However, where the organizational identity isnot strong and groups are clearly differentiatedand bounded, the tendency toward biased in-tergroup comparisons suggests several effects.Effects of Biased Intergroup Comparisons.First, the in-group may develop negative stereo-types of the out-group and deindividuate anddepersonalize its members (Horwitz & Rabbie,1982;Wilder, 1981). Hewstone, Jaspars, and Lal-

ljee (1982) studied British schoolboys from pri-vate and state secondary schools because of thehistory of conflict between the two systems. Theyfound that the groups differed in their percep-tions of themselves and each other, and that out-group perceptions were generally negative.What's more, these perceptions included self-serving (or group-serving) implicit theories ofwhy the groups differed and attribution biasesthat rationalized the successes and failures ofeach group (cf. social attribution, Deschamps,1983).This suggests a second effect of in-group bias:It justifies maintaining social distance and sub-ordinating the out-group (Smith, 1983; Sunar,1978). The in-group is seen as deserving its suc-cesses and not its failures, while the oppositeobtains for the out-group. Thus, Perrow (1970)found that members of functional subunitsacross 12 industrial firms were less likely to crit-icize the performance of their own unit and morelikely to advocate that their unit receive addi-tional power than were members from any othersubunit in their particular organization. Simi-larly, Bates and White (1961) sampled board

members, administrators, doctors, and nursesfrom 13 hospitals and found that each group be-lieved it should have more authority than theother groups were willing to allow, and Brownet al. (1986) found that members of five depart-ments in an industrial organization tended torate their own department as contributing themost to the company.Third, given symbolic interactionism, the de-sire for positive group differentiation, and thestereotyping of self, in-group, and out-group,emerging biases may soon become a contagion(Turner, 1984) that can be easily mobilizedagainst the out-group. In-group members oftencome to share pejorative perceptions of the out-group and experience the real or imaginedslights against other members as their own.Thus, major conflicts often cause an organiza-tion to polarize into rival camps, where, if anindividual is not on one side, he or she is be-lieved to be on the other side (Mintzberg, 1983).In the above study of hospitals, Bates and White(1961) found that where two groups disagreedon which should have greater authority over aparticular issue, respondents from each grouprated the amount of authority their own groupshould have higher than for issues which werenot in dispute, and gave the lowest rating to thegroup with which they disagreed. The initial dis-agreement had polarized each group's percep-tion of the situation.Finally, such competition exacerbates theabove tendencies because it threatens thegroup and its identity. Thus, as Horwitz andRabbie (1982) noted, "Both experimental andnaturalistic observations suggest that hostilityerupts more readily between [groups] than be-tween individuals" (p. 269). In-group and out-group relations may be marked by competitionand hostility even in the absence of "objective"sources of conflict (e.g., scarce resources). In-deed, Turner (1978) found the more comparablethe out-group, the greater the in-group bias.Hence, organizational subunits may claim to bepositively differentiated precisely because they

32


15/21

are not. This contrasts sharply with the conven-tional view that group conflict reflects competi-tion over rewards external to the intergroup sit-uation (cf. realistic group conflict theory, Cam-pbell, 1965; Tajfel & Turner, 1985).Qualifications to Intergroup Comparisons.The dynamics of intergroup comparison, how-ever, need to be qualified by the relative statusof the groups. The identity of a low-status groupis implicitly threatened by a high-status group,hence the defensive biases in differentiationnoted earlier. A high-status group, however, isless likely to feel threatened and, thus, less inneed of positive affirmation (Tajfel, 1982; vanKnippenberg, 1984). Accordingly, while a low-status group (such as a noncritical staff functionor cadre of middle managers) may go to greatlengths to differentiate itself from a high-statuscomparison group (such as a critical line func-tion or senior management), the latter may berelatively unconcerned about such comparisonsand form no strong impression about the low-status group. This indifference of the high-statusgroup is, perhaps, the greatest threat to theidentity of the low-status group because the lat-ter's identity remains socially unvalidated.Although the previous discussion suggeststhat subunits engage endlessly in invidiouscomparisons, three streams of research on SITsuggest otherwise. First, just as individuals se-lect similar others for social comparison, groupsalso restrict their comparisons to similar, proxi-mal, or salient out-groups (Tajfel & Turner,1985). Thus, the purchasing department may berelatively unconcerned with the machinationsof, say, the shipping or human resources de-partments.Second, van Knippenberg (1984) maintainedthat individuals are capable of making socialcomparisons on multiple dimensions, and thatmutual appreciation is possible where individu-als are superior on complementary or differentdimensions. The individuals validate each oth-er's relative superiority. Analogously, a field ex-periment by Mummendey and Schreiber (1984)

involving political parties found that in-group fa-voritism was strong on dimensions regarded asimportant to the in-group, but that out-group fa-voritism existed on dimensions regarded as un-important to the in-group but important to theout-group. It is quite conceivable that differenti-ated subunits would acknowledge one another'sdifferential expertise without necessarily com-promising positive differentiation.Finally, research on experimental and ethnicgroups indicates that groups are less likely toevidence ethnocentrism and defensive biases ifdifferences in the distribution of scarce re-sources or the outcomes of social comparisonsare viewed by the subordinate group as legiti-mate or institutionalized (Caddick, 1982;Tajfel &Turner, 1985). Indeed, in such cases the groupmay internalize the wider social evaluation ofthemselves as inferior and less deserving. Byaccident or design, systems of authority and ex-pertise in organizations (Mintzberg, 1983) oftenserve precisely this legitimating function, sug-gesting some stability in intergroup relations.In summary, SITargues that in the absence ofa strong organizational identity, the desire forfavorable intergroup comparisons generatesmuch conflict between differentiated and clearlybounded subunits. This is especially so if agroup's status is low or insecure. However, thisconflict may be mitigated to the extent thatgroups compare themselves on different dimen-sions or view the outcomes of comparisons aslegitimate or institutionalized.

Implications for ResearchGiven the paucity of research on SIT in orga-nizations, a research agenda might focus on

three objectives. First, in view of the frequentconfusion of organizational identification withsuch related constructs as commitment, loyalty,and internalization, the discriminability of iden-tification should be established. Mael's (1988)confirmatory factor analysis of the Organiza-tional Commitment Questionnaire and his new

33


16/21

measure of organizational identification offer apromising start. However, given the argumentthat individuals often have multiple (and con-flicting) identities within the organization, re-search should focus on salient subgroups aswell as the organization per se. Indeed, insofaras identification facilitates commitment and thelike, researchers should consider investigatingcommitment itself at the subgroup level. Recentwork on dual and multiple commitments is in-structive in this regard (e.g., Reichers, 1986).A second focus of research might be the pro-posed antecedents and consequences of socialidentification. Although experimental and cross-sectional field research have substantiated thesocial-psychological premises of SIT, the dy-namics of identification have not been estab-lished. Accordingly, longitudinal field researchthat focuses on a variety of newly created sub-units or organizations or on organizational new-

comers is strongly recommended. Such a de-sign would help to explore (a) how the anteced-ents interact to influence identification, (b) whatantecedents (if any) are necessary or sufficient,(c) the sequencing and timing of effects, and (d)if threshold conditions exist.Finally, although the applications of SIT to or-ganizational behavior were not intended to beexhaustive, they do suggest several specific av-enues for field research. For one, the role of or-ganizational socialization can be assessed bystructured observation of the interplay amongsymbolic interactions, symbolic management,and the emergence of social identities. Of par-ticular interest are the posited effects that iden-tification has on a person's internalization of or-ganizational values and on his or her reificationof the organization. Also important are themechanisms by which identification with lead-ers becomes generalized to the organization.For another, the disjointed resolution of role con-flicts can be evaluated by verbal protocol anal-ysis of conflict-laden decisions made over time.Of interest here are the factors associated withselecting a means of resolution, the possibility of

stable styles of resolution, the effects of differentmeans, and, more generally, the degree towhich various identities are cognitively inte-grated, the relative salience and priority of var-ious identities across organizations, subunits, hi-erarchical levels, and individuals, and the inter-action among role change, identity change, andselective forgetting. Finally, the roles that socialidentification and comparison processes have inintergroup conflict can be gauged by analyzingrelevant within- and between-group interac-tions. Research is particularly scarce on the fac-tors that affect the perception of group insecurity(and, hence, the desire for positive differentia-tion), the selection of reference groups, the di-mensions for intergroup comparison, and theperceived legitimacy and institutionalization ofthe organizational status quo. From an organi-zational development perspective, researchshould focus on the fairly unique means, sug-gested by SIT, of reducing dysfunctional inter-group conflict, such as enhancing the salienceand value of the organizational identity, in-creasing group security or at least legitimatingnecessary intergroup differences, and individu-ating out-group members.

ConclusionAccording to social identity theory, the indi-vidual defines him- or herself partly in terms ofsalient group memberships. Identification is theperception of oneness with or belongingness toa group, involving direct or vicarious experi-ence of its successes and failures. Group identi-fication and favoritism tend to occur even in theabsence of strong leadership or member inter-dependency, interaction, or cohesion. Identi-

fication is associated with groups that are dis-tinctive, prestigious, and in competition with, orat least aware of, other groups, although it canbe fostered by even random assignment to agroup. Identification can persist tenaciouslyeven when group affiliation is personally pain-ful, other members are personally disliked, and

34


17/21

group failure is likely. The concept of identifica-tion, however, describes only the cognition ofoneness, not the behaviors and affect that mayserve as antecedents or consequences of thecognition. Identification induces the individualto engage in, and derive satisfaction from, ac-tivities congruent with the identity, to view him-or herself as an exemplar of the group, and toreinforce factors conventionally associated withgroup formation (e.g., cohesion, interaction).This perspective, applied to several domains oforganizational behavior, suggests that:1. Organizational socialization can be under-stood in part as an attemptto symbolicallymanage the newcomer's desire for an iden-tity by defining the organization or subunitin terms of distinctive and enduring centralcharacteristics. Identification enables thenewcomer to reify the organization and feelloyal and committedto it per se, and facili-tates the internalization of organizationalvalues and beliefs.

2. Individuals have multiple, loosely coupledidentities, and inherent conflicts betweentheirdemands are typically not resolved bycognitivelyintegrating the identities,but byordering, separating, or buffering them.Thiscompartmentalizationof identitiessug-gests the possibility of double standards,apparent hypocrisy, and selective forget-ting.

3. In ideographic organizations, the desire fora salutary social identity predisposes orga-nizational subunits to intergroup conflict oncharacteristics hat are mutuallycompared.Thus, intergroup conflict may arise even inthe absence of such objective causes asscarce resources.

In summary, the concept of identification hasbeen neglected in organizational research. Thereformulated conception of identification as per-ceived oneness with a group, suggested by so-cial identity theory, offers a fresh perspective ona number of critical organizational issues, onlya few of which have been explored here.

ReferencesAdler, P., & Adler, P. A. (1987)Role conflict and identitysalience: College athletics and the academic role. SocialScience Journal, 24, 443-455.Albert, S., &Whetten,D. A. (1985)Organizational dentity.In L. L. Cummings & B. M. Staw (Eds.),Researchin orga-

nizationalbehavior (Vol.7, pp. 263-295). Greenwich, CT:JAIPress.Allen, V. L., Wilder, D. A., &Atkinson,M. L.(1983)Multiplegroup membershipand social identity.In T. R. Sarbin&K. E. Scheibe (Eds.), Studies in social identity (pp. 92-115).New York:Praeger.Ashforth, B. E. (1985)Climate formation: ssues and exten-sions. Academy of Management Review, 10, 837-847.Bandura, A., & Walters, R. H. (1963)Social learning andpersonality development. New York:Holt, Rinehart & Win-ston.Bass, B. M. (1985) Leadership and performance beyond ex-pectations. New York:Free Press.Bates, F. L., &White, R. F. (1961)Differentialperceptions ofauthority in hospitals. Journal of Health and Human Be-havior, 2, 262-267.Becker, H. S., & Carper, J.W. (1956)The development of

identificationwith an occupation. AmericanJournalof So-ciology, 61, 289-298.Billig,M., &Tajfel,H. (1973)Social categorization and simi-larity in intergroup behavior. European Journal of SocialPsychology, 3, 27-52.Brewer, M. B. (1979) n-groupbias in the minimalintergroupsituation: A cognitive-motivationalanalysis. Psychologi-cal Bulletin, 86, 307-324.Brewer,M. B., &Silver, M.(1978) n-groupbias as a functionof task characteristics. European Journal of Social Psy-chology, 8, 393-400.Brown, M. E. (1969) dentification nd some conditions of or-ganizational involvement. AdministrativeScience Quar-terly, 14,346-355.Brown, R. (1978)Divided we fall: An analysis of relationsbetween sections of a factoryworkforce.In H. Tajfel(Ed.),Differentiation between social groups: Studies in the so-cial psychology of intergroup relations (pp. 395-429). Lon-don: Academic Press.Brown,R., Condor,S., Mathews,A., Wade, G., &Williams,J.(1986)Explainingintergroupdifferentiationn an indus-trial organization. Journal of Occupational Psychology,59, 273-286.

35


18/21

Brown,R. J., &Ross, G. F. (1982)The battle foracceptance:An investigation into the dynamics of intergroupbehav-ior. In H. Tajfel (Ed.), Social identity and intergroup rela-tions (pp. 155-178). Cambridge, England: CambridgeUniversityPress.

Brown, R., & Williams, J. (1984)Group identification:Thesame thing to all people? HumanRelations, 37, 547-564.Brown, R. W. (1986)Social psychology, the second edition.New York:Free Press.Caddick, B. (1982)Perceived illegitimacyand intergroupre-lations. In H. Tajfel(Ed.),Social identityand intergrouprelations(pp. 137-154).Cambridge,England:CambridgeUniversityPress.Campbell, D. T. (1965)Ethnocentric nd otheraltruisticmo-tives. InD. Levine(Ed.),Nebraskasymposiumon motiva-tion(Vol.13, pp. 283-311). Lincoln:UniversityofNebraskaPress.Chatman, J.A., Bell, N. E., & Staw, B. M. (1986)The man-aged thought:The role of self-justification nd impressionmanagement in organizationalsettings. In H. P. Sims, Jr.,D. A. Gioia &Associates (Eds.), The thinking organization:Dynamics of organizational social cognition (pp. 191-214).San Francisco:Jossey-Bass.Cheek, J.M., & Briggs, S. R. (1982)Self-consciousnessandaspects of identity. Journal of Research in Personality, 16,401-408.Coe, R. M. (1965)Self-conceptionand professionaltraining.Nursing Research, 14, 49-52.Cyert, R. M., &March, J. G. (1963)A behavioral theory of thefirm.Englewood Cliffs,NJ:Prentice-Hall.Denhardt,R. B. (1987) mages of death and slavery in orga-nizational life. Journal of Management, 13, 529-541.Deschamps, J.C. (1983)Socialattribution. nJ.Jaspars,F. D.Fincham, & M. Hewstone (Eds.), Attribution theory andresearch: Conceptual, developmental and social dimen-sions (pp. 223-240). London:Academic Press.Dion, K. L.(1973)Cohesiveness as a determinantof ingroup-outgroup bias. Journal of Personality and Social Psychol-ogy, 28, 163-171.Dion, K. L. (1979) ntergroupconflictand intra-groupcohe-siveness. InW. G. Austin & S. Worchel(Eds.),Thesocialpsychology of intergroup relations (pp. 211-224). Mon-

terey, CA:Brooks/Cole.Festinger,L. (1954)A theory of social comparisonprocesses.Human Relations, 7, 117-140.Fisher, C. D. (1986)Organizationalsocialization:An integra-tive review. In K. M. Rowland & G. K. Ferris (Eds.),Re-search in personnel and human resources management(Vol.4, pp. 101 145).Greenwich, CT:JAIPress.Foote, N. N. (1951) dentification s the basis for a theory ofmotivation.American Sociological Review, 16, 14-21.

Fox, A. (1980)Themeaning of work. InG. Esland &G. Sala-man (Eds.), The politics of work and occupations (pp. 139-191).Toronto:Universityof TorontoPress.Friedkin,N. E., &Simpson,M. J.(1985)Effectsofcompetitionon members' identificationwith their subunits. Adminis-trative Science Quarterly, 30, 377-394.Gammons, P. (1986,November3)Livingand dying with theWoe Sox. Sports Illustrated, pp. 22-23.Gergen, K.J. (1968)Personal consistency and the presenta-tion of self. In C. Gordon & K.J.Gergen (Eds.),Theself insocial interaction(Vol. 1, pp. 299-308). New York:Wiley.Gerth, H. H., &Mills, C. W. (Eds.)(1946)From Max Weber:

Essays in sociology. New York:OxfordUniversityPress.Goffman, E. (1959) The presentation of self in everyday life.Garden City, NY:Doubleday.Goffman, E. (1961)Asylums: Essays on the social situation ofmental patients and other inmates. Garden City, NY: Dou-bleday.Gould, S. B. (1975) Organizational identification and com-mitment in two environments. Unpublished doctoral dis-sertation,Michigan State University,Lansing.Greene, C. N. (1978)Identificationmodes of professionals:Relationship with formalization, role strain, and alien-ation. Academy of Management Journal, 21, 486-492.Hall, D. T. (1971)A theoreticalmodel of career subidentitydevelopment in organizational settings. OrganizationalBehavior and Human Performance, 6, 50-76.Hall, D. T., Schneider, B., & Nygren, H. T. (1970)Personalfactors n organizational dentification.AdministrativeSci-ence Quarterly, 15, 176-190.Hamilton, D. L. (Ed.) (1981) Cognitive processes in stereotyp-ing and intergroup behavior. Hillsdale, NJ:Erlbaum.Hewstone, M., Jaspars,J., &Lalljee, M. (1982)Social repre-sentations, social attribution nd social identity:The inter-group images of 'public'and 'comprehensive'schoolboys.European Journal of Social Psychology, 12, 241-269.Hoelter,J.W. (1985)A structural heory of personal consis-tency. Social Psychology Quarterly, 48, 118-129.Hogan, R. (1976) Personality theory: The personological tra-dition.Englewood Cliffs, NJ:Prentice-Hall.Hogg, M. A., &Turner,J.C. (1985) nterpersonalattraction,social identificationand psychological group formation.European Journal of Social Psychology, 15, 51-66.Hogg, M. A., &Turner,J.C. (1987)Social identityand con-formity:A theoryof referent nformationalnfluence. In W.Doise & S. Moscovici (Eds.), Current issues in Europeansocial psychology (Vol.2, pp. 139-182). Cambridge, En-gland: Cambridge UniversityPress.Horwitz,M., &Rabbie, J.M. (1982) ndividualityand mem-

36


19/21

bership in the intergroupsystem. In H. Tajfel Ed.),Socialidentity and intergroup relations (pp. 241-274). Cam-bridge, England:Cambridge UniversityPress.Jackall,R. (1978)Workersn a labyrinth: obsand survivalina bank bureaucracy. Montclair,NJ:Allanheld, Osmun.Janis, I. L., &Mann,L. (1977)Decisionmaking:A psycholog-ical analysis of conflict, choice, and commitment.NewYork:Free Press.Kanter,R. M.(1977)Some effectsofproportions n grouplife:Skewed sex ratiosand responses to token women. Amer-ican Journal of Sociology, 82, 965-990.Katz, D., &Kahn, R. L. (1978)The social psychology of or-

ganizations (2nded.). New York:Wiley.Katz,R. (1980)Time and work:Towardan integrativeper-spective. In B. M.Staw &L. L.Cummings(Eds.),Researchin organizational behavior (Vol. 2, pp. 81-127). Green-wich, CT:JAIPress.Kelman, H. C. (1961)Processes of opinion change. PublicOpinionQuarterly,25, 57-78.Ketsde Vries, M. F. R., &Miller,D. (1984)The neurotic or-ganization:Diagnosing and changing counterproductivestyles of management. San Francisco:Jossey-Bass.Kramer,R. M., &Brewer, M. B. (1984)Effectsof group iden-tity on resource use in a simulated commons dilemma.Journal of Personalityand Social Psychology, 46, 1044-1057.Laurent,A. (1978)Managerialsubordinacy:A neglected as-pect of organizationalhierarchies. Academy of Manage-ment Review, 3, 220-230.Lawrence, P. R., & Lorsch,J.W. (1967)Organization andenvironment:Managing differentiationand integration.

Boston:HarvardUniversityPress.Leary,M. R., Wheeler, D. S., &Jenkins,T. B. (1986)Aspectsof identityand behavioral preference:Studies of occupa-tional and recreationalchoice. Social Psychology Quar-terly,49, 11-18.Lee, S. M. (1971)An empirical analysis of organizationalidentification.Academy of ManagementJournal,14, 213-226.Lemaine, G., Kastersztein,J., &Personnaz, B. (1978)Socialdifferentiation. n H. Tajfel(Ed.),Differentiationbetweensocial groups: Studies in the social psychology of inter-group relations (pp. 269-300). London:Academic Press.Levinson, H. (1970)A psychologist diagnoses merger fail-ures. HarvardBusiness Review, 48(2),139-147.Lieberman, S. (1956)The effects of changes in roles on theattitudes of role occupants. HumanRelations,9, 385-402.Locksley,A., Oritz, V., &Hepburn,C. (1980)Social catego-rization and discriminatorybehavior: Extinguishing the

minimal intergroupdiscriminationeffect. Journalof Per-sonality and Social Psychology, 39, 773-783.Mael, F. (1988)Organizational identification: Constructre-definition and a field application with organizationalalumni. Unpublished doctoral dissertation, Wayne StateUniversity,Detroit.March, J.G., & Simon, H. A. (1958)Organizations. NewYork:Wiley.Martin,J.,Feldman, M. S., Hatch,M. J.,&Sitkin,S. B. (1983)The uniqueness paradoxin organizationalstories.Admin-istrativeScience Quarterly,28, 438-453.Martin,J., & Siehl, C. (1983) Organizational culture and

counterculture:An uneasy symbiosis. OrganizationalDy-namics, 12(2),52-64.Mintzberg,H. (1983)Power in and around organizations.Englewood Cliffs,NJ:Prentice-Hall.Mortimer, .T., &Simmons,R. G. (1978)Adultsocialization.In R. H. Turner,J. Coleman, & R. C. Fox (Eds.),Annualreview of sociology (Vol.4, pp. 421-454). Palo Alto, CA:Annual Reviews.Mowday, R. T., Steers, R. M., & Porter, L. W. (1979)Themeasurement of organizational commitment.Journal ofVocationalBehavior, 4, 224-247.Mummendey, A., & Schreiber, H. J. (1984) "Different"ustmeans "better":Some obvious and some hidden path-ways to in-groupfavouritism.British ournalof Social Psy-chology, 23, 363-368.Oaker, G., &Brown,R. (1986) ntergrouprelationsin a hos-pitalsetting:A further est of social identity heory. HumanRelations,39, 767-778.Oakes, P. J.,&Turner,J.C. (1980)Social categorizationand

intergroupbehavior: Does minimal ntergroupdiscrimina-tionmake social identitymore positive?EuropeanJournalof Social Psychology, 10,295-301.Oakes, P., &Tumer, J. C. (1986)Distinctivenesand the sa-lience of social category memberships:Is there an auto-maticperceptualbias towardsnovelty?EuropeanJournalof Social Psychology, 16, 325-344.O'Reilly,C., HI,& Chatman, J.(1986)Organizational com-mitment and psychological attachment: The effects ofcompliance, identification, and intemalizationon proso-cial behavior.JournalofAppliedPsychology, 71,492-499.Ouchi, W. G. (1981)Theory Z:How American business canmeet the Japanese challenge. New York:Avon.Patchen, M. (1970)Participation,achievement and involve-ment on thejob. Englewood Cliffs,NJ:Prentice-Hall.Perkins, D. N. T., Nieva, V. F., & Lawler, E. E., m. (1983)Managing creation:The challenge of building a new or-ganization. New York:Wiley.Perrow, C. (1970)Departmentalpower and perspectives in

37


20/21

industrialfirms. In M. N. Zald (Ed.),Power in organiza-tions: Proceedings of the first annual Vanderbilt SociologyConference(pp. 59-89). Nashville, TN:VanderbiltUniver-sity Press.Pfeffer,J. (1981)Management as symbolic action: The cre-ation and maintenance of organizationalparadigms. InL. L. Cummings &B. M. Staw (Eds.),Research in organi-zational behavior (Vol.3, pp. 1-52). Greenwich, CT:JAIPress.

Pondy, L. R., Frost, P. J., Morgan, G., & Dandridge, T. C.(Eds.) (1983) Organizational symbolism. Greenwich, CT:JAIPress.Reichers, A. E. (1985)A review and reconceptualizationof

organizational commitment. Academy of ManagementReview, 10,465-476.Reichers, A. E. (1986)Conflictand organizationalcommit-ments. Journal of Applied Psychology, 71, 508-514.Reichers, A. E. (1987)An interactionistperspective on new-comer socialization rates. Academy of Management Re-view, 12,278-287.Reid, F. J.M. (1983)Polarizingeffectsof intergroupcompar-isons. European Journal of Social Psychology, 13, 103-

106.Rotondi, T., Jr. (1975)Organizational identification:Issuesand implications. Organizational Behavior and HumanPerformance, 13, 95-109.Sampson, E. E. (1978)Personalityand the location of iden-tity. Journal of Personality, 46, 552-568.Schelling, T. C. (1957)Bargaining,communication,and lim-ited war. Journal of Conflict Resolution, 1, 19-36.Schneider, B., Hall, D. T., &Nygren, H. T. (1971)Self-imageand job characteristicsas correlatesof changing organi-zational identification.HumanRelations,24, 397-416.Skevington, S. M. (1980)Intergroup relations and socialchange within a nursing context. British Journal of Socialand Clinical Psychology, 19, 201-213.Skevington, S. M. (1981)Intergrouprelations and nursing.European Journal of Social Psychology, 11, 43-59.Smith, K. K. (1983) Social comparison processes and dy-namic conservatism n intergrouprelations. In L. L. Cum-mings &B. M. Staw (Eds.), Research in organizational be-

havior (Vol.5, pp. 199-233).Greenwich, CT:JAIPress.Steinbruner, J. D. (1974) The cybernetic theory of decision:New dimensions of political analysis. Princeton, NJ:PrincetonUniversityPress.Stryker,S., & Serpe, R. T. (1982)Commitment, dentity sa-lience, and role behavior:Theoryand research example.In W. Ickes&E. S. Knowles(Eds.),Personality,roles, andsocial behavior (pp. 199-218).New York:Springer-Verlag.

Sunar, D. G. (1978)Stereotypes of the powerless: A socialpsychological analysis. Psychological Reports, 43, 511-528.Tajfel, H. (1969)Cognitive aspects of prejudice. JournalofSocial Issues, 25(4), 79-97.Tajfel,H. (1978)Theachievement of group differentiation. nH. Tajfel (Ed.), Differentiation between social groups:Studies in the social psychology of intergroup relations(pp. 77-98). London:Academic Press.Tajfel, H. (1981) Human groups and social categories: Stud-ies in social psychology. Cambridge, England: Cam-bridge UniversityPress.Tajfel,H. (1982) nstrumentality,dentityand social compar-isons. In H. Tajfel (Ed.), Social identity and intergroup re-lations (pp. 483-507). Cambridge, England: CambridgeUniversityPress.Tajfel,H., &Turner,J.C. (1985)The social identitytheoryofintergroupbehavior. In S. Worchel & W. G. Austin(Eds.),Psychology of intergroup relations (2nd ed., pp. 7-24).Chicago: Nelson-Hall.Thoits, P. A. (1983)Multiple identities and psychologicalwell-being: A reformulation nd test of the social isolationhypothesis. American Sociological Review, 48, 174-187.Tolman,E. C. (1943) dentificationand the post-war world.Journal of Abnormal and Social Psychology, 38, 141-148.Trice,H. M., &Beyer, J.M. (1986)Charisma and its routini-zation in two social movement organizations. In B. M.Staw &L. L. Cummings (Eds.), Research in organizationalbehavior (Vol.8, pp. 113-164).Greenwich, CT:JAIPress.Turner,J.C. (1975)Social comparison and social identity:Some prospects forintergroupbehaviour. EuropeanJour-nal of Social Psychology, 5, 5-34.Turner,J.C. (1978)Social comparison, similarity and in-group favouritism. In H. Tajfel (Ed.), Differentiationbe-tween social groups: Studies in the social psychology ofintergroup relations (pp. 235-250). London: AcademicPress.Turner,J.C. (1981)The experimental social psychology ofintergroupbehavior. InJ.C. Turner& H. Giles (Eds.), In-tergroupbehaviour (pp. 66-101). Chicago: University ofChicago Press.Turner,J.C. (1982)Towards a cognitive redefinition of the

social group. In H. Tajfel (Ed.), Social identity and inter-group relations (pp. 15-40). Cambridge, England: Cam-bridge UniversityPress.Turner,J.C. (1984)Social identificationand psychologicalgroup formation.In H. Tajifel Ed.),The social dimension:European developments in social psychology (Vol. 2, pp.518-538). Cambridge, England: Cambridge UniversityPress.

38


21/21

Turner, J. C. (1985) Social categorization and the self-concept: A social cognitive theory of group behavior. InE. J. Lawler (Ed.), Advances in group processes (Vol. 2,pp. 77-122). Greenwich, CT:JAIPress.Turner,J.C., Brown, R. J.,&Tajfel, H. (1979)Social compar-ison and group interest in ingroupfavouritism.EuropeanJournal of Social Psychology, 9, 187-204.

van Knippenberg, A. F. M. (1984) ntergroupdifferences ingroupperceptions.In H. Tajfel Ed.),Thesocial dimension:European developments in social psychology (Vol. 2, pp.560-578). Cambridge, England: Cambridge UniversityPress.Van Maanen, J. (1976)Breaking n: Socializationto work. InR. Dubin (Ed.), Handbook of work, organization, and so-ciety (pp. 67-130). Chicago: Rand McNally.VanMaanen, J.(1978)People processing:Strategies of orga-nizational socialization. Organizational Dynamics, 7(1),19-36.VanMaanen, J.(1979)The self, the situation,and the rules of

interpersonal relations. In W. Bennis, J. Van Maanen,E. H. Schein, &F. I. Steele (Eds.),Essays in interpersonaldynamics (pp. 43-101). Homewood, EL:Dorsey Press.Vaughan, D. (1986)Structuralsecrecy and organizationalmisconduct:NASAand the space shuttleChallenger. Pa-per presented at the meeting of the Academy of Manage-ment, Chicago.Wagner, U., Lampen, L., & Syllwasschy, J. (1986) n-groupinferiority, ocial identity and out-groupdevaluation in amodified minimal group study. BritishJournal of SocialPsychology, 25, 15-23.Wiener, Y. (1982)Commitment n organizations:A norma-tive view. Academy of Management Review, 7, 418-428.Wilder,D. A. (1981)Perceiving persons as a group: Catego-rization and ingroup relations. In D. L. Hamilton(Ed.),Cognitive processes in stereotyping and intergroup be-havior (pp. 213-257). Hillsdale, NJ:Erlbaum.Wilder,D. A., &Allen, V. L. (1978)Groupmembership andpreference for informationabout others. PersonalityandSocial Psychology Bulletin,4, 106-110.

Blake Ashforth Ph.D., Universityof Toronto)s Assis-tant Professorof Management at Concordia Univer-sity. Correspondence regarding this article may besent to him at the Departmentof Management, Fac-ulty of Commerce and Administration, ConcordiaUniversity,1455de Maisonneuve Blvd. W.,Montreal,Quebec, Canada H3G 1M8.Fred Mael is a doctoralcandidate in IndustrialPsy-chology at Wayne State University,Detroit.The authors are indebted to Ron Humphrey andPeggy Padgett for theirhelpful comments on earlierdrafts of the paper. Portionsof the paper were pre-sented at the 1988MidwestAcademy of Managementmeeting in Toledo, Ohio.

39

Social Identity Theory and the Organization

Documents

Transcript of Social Identity Theory and the Organization