Three Faces of Identity - Majorsmatter.net

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.majority of work in sociology is on the “me” aspectof the self, which includes the self-conceptand the identities that are incorporated into it.Once the self emerges in the human organism,a nascent self-concept soon follows.Rosenberg (1979) defined self-concept as thetotality of a specific person’s thoughts andfeelings toward him- or herself as an object ofreflection. The self-concept (or the “me”) canbe thought to consist of three broad classesof individual attributes (Rosenberg 1979,pp. 15–17): self-referring dispositions, physicalcharacteristics, and identities. Self-referringdispositions denote the abstract categoriespeople develop over their life courses andthen use to shape their response tendencies,including attitudes such as liberalism, traitssuch as altruism, values such as patriotism,and abilities such as athletic skill. Thoits &Virshup (1997) refer to these elements as one’sindividual identity because they represent waysto differentiate oneself from others (as opposedto representing communalities with othersthrough roles or group membership).Physical characteristics include one’s externalattributes such as being obese, deaf, or tall.These physical characteristics become sociologicallyinteresting when they are also incorporatedinto a person’s self-image and thus havethe potential to shape one’s behavior or one’ssocial and psychological well-being. They also,of course, have an external character that influenceshow others respond to us, shaping theirinternalization into the self-concept.The third component of Rosenberg’sself-concept is the focus of our review here:identity. There are four key sources of identitycharacterizations: personal or individualidentity, role-based identity, category-basedidentity, and group membership–based identity.Personal identity is the most elementarytype of identity, defined here as the socialclassification of an individual into a categoryof one (Rosenberg 1979). It denotes a uniqueindividual with self-descriptions drawn fromone’s own biography and singular constellationof experiences. Examples include: I am RoySmith, Pat Smith’s spouse; I served in Patton’sThird Army during World War II; I was bornin Robeson County Hospital at 3:14 PM; mymother had a cousin named Edna; I drovethe school bus during my junior year in highschool. Although personal identity consists ofunique identifiers and an individual narrative,it is social and institutional in origin. As such,soldiers are identified and differentiated fromother soldiers by their names, ranks, and serialnumbers; academics by their names, ranks,departmental affiliations, and the institutionsfrom which their highest degree was awarded.These distinctions are created and organizedby the institutions within which they occur.As individualized as it is, one’s personal identityinformation is the basis for one’s otheridentities. “If an individual could not be recognizedfrom one occasion to another as the sameperson, no stable social relationships could beconstructed” and no other identities could beformed (McCall & Simmons 1966, p. 65). Wedistinguish between these individuated narrativesin personal identities and what Thoits& Virshup (1997) termed “individual identity.”They defined individual identity as selfideasabstracted from one’s biographical detailsand framed in terms of broader social categoriessuch as working class, Midwesterner, diabetic,snowbird, or progressive. When suchabstracted self-categorizations are internalized,they often correspond to group or categoricalidentities as described below.Identities based on role relationships arethe most central in the theories that stressinternalization of identity meaning into theself-structure. We define role-identity as a socialposition a person holds in a larger socialstructure, 1 considers self-descriptive, and enactsin a role relationship with at least oneother person (Thoits 1995). Because it is selfdescriptiveand internalized, it becomes partof one’s self-concept. Role-identities are predicatedon recurrent interactions between role1 Here we adopt Stryker’s (2008) definition of social structureas socially patterned interactions and relationships noted fortheir regularity, resistance to change, and capacity to reproducethemselves.www.annualreviews.org • Three Faces of Identity 479

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.before discussing the specific theories, webriefly address motivation as it generallyapplies to these internalization theories.Nearly six decades ago, Foote (1951) attemptedto clarify role theory by using a socialpsychology of motivation that both rejectedbiological determinism (“the person impelledfrom within”) and cultural determinism (“theperson driven from without”) (p. 21). Foote’stheory of situated motivation is based uponsymbolic interactionist notions of language andidentification. Specifically, language is centralto motivation because it helps shape behaviorby enabling individuals to meaningfully understandand label their past actions in order to formulatepresent and future outcomes. Throughidentification, people appropriate and committo particular identities. People have multipleidentities, and their identities give their behaviormeaning and purpose. However, becausepeople also perform roles that are attached toparticular identities, the role/identity nexus isalso an important ingredient in motivation andchoice. Indeed, the very notion of identity insymbolic interactionism connotes an intimatelinkage between self and role (Burke & Tully1977) such that human beings are confrontedregularly with choices between alternative commitmentsand actions.Two influential mid-range identity theorieswere developed in the 1960s to represent theways in which structural role positions and theirinternalizations guided choices that actors makein social interaction. Both McCall & Simmons(1966) and Stryker (1968) attempted to link thestructural-functional insights about role relationshipsand their functions in larger sociologythat dominated the mainstream sociology of theday with the dynamic, processual insights aboutself that dominated microsociological thinking.The result was a remarkably fruitful theoreticaltradition that grew out of these two similar perspectiveson identity.character that individuals devise for themselveswhen occupying specific social positions. Andas discussed earlier, they saw role-identities asstemming from the preferred perceptions thatone has of oneself as one occupies various socialpositions. In this case, role-identities influencepeople’s everyday lives by serving astheir primary source of personal action plans.The theory has a view of people capable of creativityand improvisation in the performance oftheir roles, yet within the overall requirementsand restrictions of their social position(s). Thiscommingling of individuality, idiosyncrasy, andimpulsiveness with behavior constrained by socialconvention occurs through a dialog betweenthe “I” and the “me” bounded by thebroad dictates of one’s role-identity.Because people have multiple and oftencompeting role-identities, which also come andgo during one’s life course, an important theoreticalproblem in McCall & Simmons’s theoryis to explain which role-identities people valuemost and will thus attempt to perform. Theyargue that a person’s various role-identitiesget organized into a hierarchy of prominence,where a role-identity’s prominence reflects therelative value it has for his or her overall conceptionof one’s ideal self. 2 In this way, one’s prominencehierarchy is equivalent to one’s ideal self(McCall & Simmons 1966, pp. 76–80, 264).Prominence itself is predicated on many factors.First, people must assess the degree of commitmentthey have to a particular role-identity and,by extension, how much their self-esteem isbound to its successful activation. That is, commitmentsignifies how deeply a person stakeswho he or she is by virtue of his or her roleidentityand its performance. This conceptualizationis directly relevant to James’s (1890)definition of self-esteem as the ratio of one’sperceived success in a particular role-identityto one’s desired level of success.Role-Identity TheoryMcCall & Simmons (1966) defined roleidentityin dramaturgical language as the2 Their concept of ideal self, which traces back to Horney(1945), is part wish and part obligation with regard to anindividual’s most personal aspirations and wants for him- orherself (e.g., to be a distinguished professor, a loving mother,a successful businessperson).www.annualreviews.org • Three Faces of Identity 481

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.Second, when considering their ideal selvesvis-à-vis particular role-identities, people evaluatehow their prior actions generally comportwith their role-identity performances and howmuch they are living up to their ideal selves.Third, as a sign of the importance of reflectedappraisals, people try to determine how theirsignificant others will evaluate and appraise therole when and if it is activated. Fourth, individualsassess the rewards they may or may nothave received from the prior activation of a roleidentity,a notion consistent with social learningtheory (e.g., Bandura 1977).Identity TheoryStryker’s Identity Theory (1968, 1980, 2008)has been the dominant perspective on self andidentity within structural symbolic interactionismfor the past four decades. Since Stryker(2008) recently offered a focused treatment ofhis theory in the Annual Review of Sociology, weonly review its basic form here (see also Owens2003). Identity Theory sees a multifaceted selfcomposed of multiple identities arranged hierarchicallyin an identity salience structure. Themore salient an identity, the higher is the probabilityof its being invoked in an interactionalsituation that allows some agency or choice.The salience itself is based on two dimensions ofone’s commitment to the identity: interactionaland affective. Interactional commitment is theextensiveness of the interactions a person hasin a social network through a particular identity(e.g., the number of persons one interactswith based on the identity). Affective commitmentis a person’s emotional investment in relationshipspremised on the identity (e.g., howemotionally close others in the role relationshipare to the individual). Note that Stryker’suse of the term “commitment” is more multidimensionaland less psychological than the use ofthe same concept label in McCall & Simmons’sRole-Identity Theory above.In an early empirical test of the theory,Serpe (1987) showed that college students’commitments to their student-related identitiesat Time 1 impacted those identities’ saliencesat Time 2, whereas Time 2 student-identitysalience impacted contemporaneous commitmentto that identity. This pattern suggestedthat commitment precedes salience. However,the degree of structural freedom students hadin choosing one role-identity over another(i.e., coursework versus dating) was key. Onlysituations involving choice showed the pattern.More recently, Owens & Serpe (2003) foundthat behavioral and affective commitment wereboth significantly related to family identitysalience for Hispanics, but only behavioralcommitment was predictive of family-identitysalience for Anglos and blacks. They arguedthat the significance of affective commitmentfor Hispanics was heightened by the family’sparticular importance for this group.Finally, Stryker et al. (2005) recentlydiscussed an ongoing problem in IdentityTheory—how multilevel social structures facilitateor constrain one’s opportunity for,and the strength of, commitment to particularrole-identities (i.e., family, work, and volunteerism).They show that intermediate-levelsocial structures (e.g., neighborhood or school)influenced commitment most by fostering ingroupidentity-based relationships. Proximallevelsocial structures, though important, hadless impact than intermediate structures. Theproximal-level represents social embeddednessin multiple networks of social relations (e.g.,number of workmates who are also one’s relatives).Finally, large-scale stratification systems(e.g., gender, race, socioeconomic status) hadthe least impact on commitment to a particularidentity.Identity Accumulation TheoryThoits (1983, 1986, 2003) drew heavily onRole-Identity Theory (McCall & Simmons1966) and Identity Theory (Stryker 1968, 1980,2008) in her formulation of Identity AccumulationTheory, her description of the importancethat multiple role-identities can have for a person’spsychological and emotional well-being.Her theory asserts that multiple role-identitiescan be psychological resources that help reduce482 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.emotional distress (depression) and fosterglobal self-esteem in complex selves. UnderpinningIdentity Accumulation Theory are twokey assumptions. The first, an essential aspectof symbolic interactionism, is that identitiesprovide individuals with meaning and purposeby answering the question: Who am I? Second,roles give individuals structure and organizationby answering the question: What shouldI do? By extension, multiple role-identitiesprovide the person with an orientation towardlife situations that help foster well-being. AsAhrens & Ryff (2006) point out, more rolesalso are associated with beneficial individualresources, such as social connections, power,and prestige. The more options one has toobtain these resources, the better the outcome.Numerous recent empirical studies supportthe hypothesis linking multiple role-identitiesto lower levels of psychological distress(Kikuzawa 2006, Sachs-Ericsson & Ciarlo2000, Wethington et al. 2000), fewer physicalhealth problems ( Janzen & Muhajarine 2003),or both (Barnett & Hyde 2001). Conversely,Brook et al. (2008) cited other studies indicatingthat multiple role-identities canactually increase or prolong depression whenrole-identity demands are incompatible withother behavioral expectations or when the roleidentityclaims too much of a person’s time anddrains her energy. Brook et al. (2008) claimedthat the key to understanding the differencebetween positive and negative benefits of multipleroles is the mediating effect of positive ornegative emotions on the interactions betweenthe number of identities a person holds, theirsubjective importance to the person, and howharmoniously they interact with each other.Jackson (1997) added an important statusdimension involving race and ethnicity toIdentity Accumulation Theory and showedthat the multiple role-identities were associatedwith lower depression and greater happinessamong non-Hispanic white and MexicanAmerican men and women but not AfricanAmericans. Kikuzawa (2006) added a person’slocation in the life course to race/ethnicity ina cross-cultural study of depression levels forAmericans and Japanese age 60 and over. Shefound that role accumulation (spouse, parent,and community volunteer) benefited olderAmericans’ mental health, but any role beyondbeing either a spouse or a parent had no impacton well-being for older Japanese. Theoreticalunderstanding of these varying patterns indifferent racial, ethnic, and age groups is stillneeded.A common thread running through manystudies of multiple identities, including thework above on Identity Accumulation Theory,is the importance of Stryker’s Identity Theoryin the framing of the empirical research. Few,however, actually employ Stryker’s conceptionsof role interactional and affective commitmentin their empirical analyses. Even among thosewho try to hew closely to Stryker’s theoreticalconceptualizations, differences in the waysalience and commitment are operationalizedhelp explain empirical inconsistencies in the relationshipbetween multiple role-identities andwell-being.Serpe (1987) and Thoits (2003) may offer animportant piece missing in Identity AccumulationTheory: assessing the degree of choice peoplehave in enacting their many role-identities.If multiple role-identities are a resource, thenthe most consequential for well-being shouldbe the identities a person has some freedomto choose. For example, the mother role andworker role are more constraining because theyoffer less choice in enacting than the freedomto choose the PTA role and church choir role.Identity Control TheoryWhereas Stryker’s Identity Theory focusedon identity choices, Burke developed anelaboration of the theory to specify howinternalized meanings guided action after anidentity was adopted within an institutionalcontext (Stryker & Burke 2000). Burke’s (1991,Burke & Reitzes 1991) theoretical developmentsare often known as Identity ControlTheory to distinguish them from Stryker’smore structural focus on commitment andidentity enactment. Burke built on the workwww.annualreviews.org • Three Faces of Identity 483

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.of Powers (1973) to represent the relationshipbetween internalized identity meanings andperceptions of an interactional situation as acontrol system or cybernetic feedback loop[Robinson (2007) provides a recent review ofcontrol theorizing in sociology]. Meaningsare typically measured using bipolar semanticdifferential scales (Osgood et al. 1957, 1975).The dimensions on which meanings vary areassessed in each institutional context and thenmeasured at the level of the individual actor.Consistent with the other internalizationidentity theories summarized above, IdentityControl Theory emphasizes the importance ofunderstanding identity not just as a state ortrait characteristic of the individual, but as acontinuous process of affirmation and reaffirmationin social situations. “An identity process,”according to Burke & Reitzes (1991), “isa continuously operating, self-adjusting, feedbackloop: individuals continually adjust behavior tokeep their reflected appraisals congruent withtheir identity standards or references” (p. 840,emphasis in the original). Furthermore, Burkeand associates follow mainstream symbolic interactionismby viewing an identity as a set ofmeanings applied to the self in a social role[although later presentations of the theory explicitlyexpand its scope to consider identitiesthat come from group or category membership(e.g., Burke 2004, Burke & Stets 2009)]. Thesemeaning-sets act as a standard or reference forunderstanding who one is in a given situationand what are the expectations in order to maintainthat identity in the eyes of self and others.When an identity is brought into play in aninteraction situation, a feedback process ensuesvia reflected appraisals. That is, a cyberneticcontrol process balances any identity discrepanciesthat may arise when a particular identitystandard (the internalized meanings of self inrole, group, or category) is compared to theidentity’s situated meaning via a comparator.It is hypothesized that people behave so as tocounter and reduce any discrepancy that arisesfrom interaction. Since actions are generatedto reduce discrepancies, they can vary inquality, depending on the disturbing events.This feature of the theory allows it to describehow actors respond creatively to circumstancesother than normal role performances.Identity Control Theory is thoroughlyreviewed in Burke & Stets (2009). Here wedescribe only a few studies as exemplars ofthis research tradition. In one of the initialstatements of the theory, Burke (1991) poseda challenge to prevailing views of social stressby positing that disruption of the otherwisecontinuous identity process and the inability toclose the gap between disapproving reflectedappraisals and an identity standard will resultin distress. Burke & Stets (1999) show howself-processes can influence social structure,while Stets & Tsushima (2001) have extendedthe theory to an examination of the moderatinginfluence of group-based identities and rolebasedidentities on how people experience andcope with the potent emotion of anger. Morerecently, Burke et al. (2007) combined IdentityControl Theory, Status Characteristics Theory(discussed briefly at the end of the next section),and Legitimacy Theory in an experiment onhow gender factors into leader and subordinateidentity verification. They studied taskorientedgroups of four people (two womenand two men) when leadership is conferred by ahigher, legitimate authority (the experimenter).Among their findings are that female leaders(who were legitimated by the experimenterwhen the group was formed) and males (whowere not legitimated) both had higher levelsof identity verification than others. This studyillustrates an interesting new theme in IdentityControl Theory—the relevance of resources tothe ability of an actor to verify his or her identity(see Burke & Stets 2009, pp. 79–82, 229–33).This emphasis on a larger structure brings thetheory closer to the situational/cultural emphasisof the theories reviewed in the next section.THEORIES EMPHASIZINGCULTURE AND SITUATIONALCONTEXTPresumably, a “definition of the situation” is almostalways to be found, but those who are in484 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.the situation ordinarily do not create this definition,even though their society can be said to do so;ordinarily, all they do is to assess correctly whatthe situation ought to be for them and then actaccordingly.—Erving Goffman (1974, pp. 1–2)The theories of internalized identity meaningsreviewed in the previous section build on thefoundational insight of Mead and Cooley thatwe incorporate the social positions that we occupyinto our cognitive image of ourselves aspeople. Here, we shift attention to theories thatgive priority to situational, social structural, andcultural elements that lie outside the individual.We caution that our categorization heredoes not imply competition or contradiction.Instead, theories (and the conceptualizationsembedded within them) necessarily simplify realityby focusing on some elements to the exclusionof others. So the theories described here donot reject the idea that social roles, group memberships,and category memberships are incorporatedinto the self-image. Rather, they emphasizehow the elements of situations in whichactors are involved shape their behavioral, cognitive,and emotional reactions, rather than focuson the intraindividual features of identity inthe self-structure that are carried from situationto situation.This tradition of emphasizing the importanceof situational context for social actionbegan to develop in its modern form duringthe 1950s and early 1960s. It was partially aresponse to the sense that symbolic interactionism,and sociological social psychologymore generally, had lost its connection to thestructural focus of the larger discipline (Stryker2008). Foote’s (1951) influential theory of situatedmotivation, discussed above, was publishedin this period, using the concept of identity tofocus attention on the actor-situation nexus.Goffman (1959, 1963) emphasized that actorspresented themselves to others in a mannerthat served to maintain certain images. HisPresentation of Self in Everyday Life was concernedmore with the interactional presentationof self and how situational context supported orundermined it than with the underlying characterof the “self” that was being presented [seeHochschild (1983, appendix A) for a relateddiscussion]. Similar intellectual movementswere afoot in psychology, as Heider’s (1946,1958) Balance Theory drew attention to configurationsof interaction and their implications.Situated Identity TheoryAn early attempt to develop a formal, empiricallytestable theory based on this situationalfocus was Alexander’s Situated Identity Theory(Alexander & Knight 1971; summarized inAlexander & Wiley 1981). Building on Heiderand Goffman, Alexander conceptualized situatedidentities not as properties possessed bypersons or as features located in some externalenvironmental structure. Instead, the situatedidentity defined the relationship betweenan individual and the environment (especiallythe other actors within it) at a given time.Like many situational theories that followed,Alexander focused initially on relatively simplesituations in which actors were assumed tohave similar perspectives; he restricted his experimentalwork to situated activities that metthat consensus criterion. In his most cited study(Alexander & Knight 1971), he replicated a classiccognitive dissonance experiment, showingthat the desirability of the situated identity associatedwith an outcome predicted whetherpeople chose that outcome, often in violationof cognitive dissonance predictions. Althoughthis research tradition is not currently very active,it was a groundbreaking effort to considersystematically the impact of situational cues onidentity occupancy and social conduct flowingfrom that identity. It also pioneered the useof experimental methods in this domain withinsociology.We now move to a review of currentlyactive theories that emphasize the impact of situationaland cultural features on social interaction.We discuss two traditions in sociology andpsychology—Affect Control Theory and SocialIdentity Theory—that conceptualize the socioculturalenvironment in very different ways.www.annualreviews.org • Three Faces of Identity 485

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.We then briefly discuss Status CharacteristicsTheory. Although this theoretical researchprogram is not based in the identity tradition,its findings have some important connectionsto the phenomena that we explore here.Affect Control TheoryTo make the connection between the internalizedand situational/cultural conceptions ofidentity clear, we begin our discussion of themore situational/cultural conception of identitywith Affect Control Theory (Heise 1979, 2007),which shares many elements with Burke’s IdentityControl Theory. Both theories drew directlyon Powers’s (1973) insight that much ofhuman processing of stimuli from the environmentwas described by a control system ratherthan a direct translation of inputs into outputs.But even given this shared control imagery, thesubtle differences between the theories are anexcellent demonstration of the distinction thatwe make here between personal and social theoriesof identity. They also illustrate the dangerof making this categorization seem too rigid.Affect Control Theory was developed byHeise (1979) by combining two psychologicalresearch programs—Osgood’s work on themeasurement of affective meaning (Osgoodet al. 1957, 1975) and Gollob’s (1968) workon impression formation—with Powers’s newideas about control systems. The resultingtheory had three elements. First, it usedOsgood’s work to create a measurementsystem in which meanings were conceptualizedon three dimensions of affectivemeaning—evaluation (good-bad), potency(powerful-powerless), and activity (livelyquiet).Measuring meaning on just these threedimensions missed some nuances but allowedall the elements of a social event (the actor, thesocial action, the object-person at whom theaction is directed, and later emotions, setting,nonverbal behaviors, status characteristics, etc.)to be mapped into the same three-dimensionalspace. The second element of the theory was anempirical framework for describing how eventschanged the meanings within a situation. Usinga paradigm developed by Gollob, Heise combinedidentity labels and social behaviors intosimple event sentences (e.g., the mother slapsthe child) to study how the initial meanings ofthe event elements (a mother, slapping someone,a child) would be transformed by theircombination in the event (what do you think—on the three dimensions of affective meaning—of a mother who has slapped a child?).The third element of the theory is thecontrol system, adapted from Powers (1973),which it shares with Identity Control Theory.When social interaction deflects meanings awayfrom their stable, culturally determined values,Affect Control Theory proposes that actors tryto create new events to bring the transitory,situated meanings back into line with the stable,cultural meanings. Deflection of meaningsis defined mathematically, by the summed absolutedifferences between transient, situatedmeanings for event elements (actor, behavior,and object-person) and the culturally given referencevalues for each of those elements. Thesearch for a new event that, when enacted, willminimize that deflection then becomes a searchfor a three-number profile: the evaluation, potency,and activity of a behavior (what could Xdo next to remedy the situation?), a new labelfor the actor (what kind of a person would dosuch a thing?), or a new label for the objectperson(what kind of a person would deserve tohave that happen to them?). Once the profile isfound, one can search a cultural “dictionary” ofaffective meanings to find qualitative labels thatfit the affectively appropriate response.The contrast between Affect Control Theoryand the more recently developed IdentityControl Theory is described extensivelyin Smith-Lovin & Robinson (2006). Thetheories differ in their approaches to measurement,formalization, and emotional response.Here we emphasize only the differences in personalversus situational focus that are key to ourdiscussion.The first contrast between the theories is thereference level that is being maintained by thecontrol system in each theory. Identity ControlTheory’s reference level is the internalized486 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.meanings that an actor incorporates into theself-structure. In Affect Control Theory, peopletry to maintain meanings associated withthe entire situation—their own identities, theidentities of others, actions, and behavior settings(Smith-Lovin 1979). A second differenceflows directly from the first. In Affect ControlTheory, the actions of others are explicitlyconsidered as possible remedies for meaningdeflectingevents. In personal identity theories,the assumption is that people operate to maintaintheir own identities. Affect Control Theoryexplicitly considers how a situation in whichmeanings have been disrupted can be repairedby any of the actors that are copresent in a situation,even those who were not involved in thedeflecting event.Finally, the theories differ somewhat in theirassumptions about the stability of meaning.Since Identity Control Theory focuses on personal,internalized identity meanings, much ofthis research program looks at how these meaningsshift as a result of interactions over aperiod of sustained interactions (e.g., Stets &Burke 2005). In Affect Control Theory, meaningsare assumed to be quite stable and acquiredthrough long-term socialization processes fromthe culture at large. Most extreme dislocationsof meaning that cannot be resolved behaviorallyare resolved through relabeling a person or action.Since both theories share the core controlimagery of Powers (1973), much research thatsupports Affect Control Theory could supporteither theory (e.g., Robinson & Smith-Lovin1992, 1999), as well as self-consistency theoriesfrom psychology. Work that is more distinctiveto this theoretical tradition looks at featuresother than the central actor. For example,one series of studies examines how emotion displays(which signal transient meanings and deflectionsfrom identity) lead to labeling after adeviant act (e.g., Robinson et al. 1994). Anotherlooks at how people work to manage the identitiesof others, even at some cost to their ownself-image (C.L. Rogalin, D.T. Robinson, L.Smith-Lovin, submitted manuscript).Since Affect Control Theory focuses on thesituation, it often deals with identities that arenot core to an actor’s self-structure. Smith-Lovin (2007) reported an experiential samplingstudy that found we actually spend most of ourtime in such noncore identities. In fact, AffectControl Theory had no real conceptualizationof a stable, organized self-structure during thefirst three decades of its development (Heise1979, Smith-Lovin & Heise 1988, MacKinnon1994); it was a theory about roles, identities, andsituated action.Recent developments, however, make clearthat the boundary between personal and socialidentity theories is indistinct and everchanging.MacKinnon & Heise (2010) developa control theory of the self that is based on AffectControl Theory. The theory has two elements.First, actors use a “cultural theory of people”—the collection of categories and logical implicationsamong them—as a menu of possibilitiesfor self-identification and understanding others.Second, when self-identity meanings aredeflected in one situation (and if those deflectionscannot easily be behaviorally resolved),then actors seek out other possible identitiesin situations that, when maintained, will resolvethe deflection at the level of the self. Therefore,although it uses the mathematical control systemfrom Affect Control Theory, this new identitytheory clearly moves back into the realm ofinternalized, personal identity theories by concentratingon how people maintain a stable selfimageover a series of situational encounters. Inthis case, they do so by agentically entering situationsthat place them in different identitiesand lead to the experience of self-relevant situatedmeanings. Notice the similarity to the basicquestion posed (and answered) by Stryker’sIdentity Theory: How do people choose whichidentities to enact, when they have a choice?MacKinnon and Heise give a more formal answerto this question than Stryker, but they aredefinitely building in his domain.Having discussed how two very similar theoriesdiffer in their emphasis on personal versussocial/cultural focus, we now turn to a perspectivethat focuses on how contextual elementsaffect self-meanings that are derived from categorymemberships. Social Identity Theorywww.annualreviews.org • Three Faces of Identity 487

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.deals with cognitions about categories of ingroupsand out-groups (Tajfel & Turner 1979).Therefore, it would seem to be the prototype ofan intrapersonal identity theory. However, themechanisms described by the theory actuallycenter on the impact of the social environmenton those cognitions, placing this theory in thesituational/cultural category for our treatmenthere. We emphasize those contextual elementsof the theory in our description of it. 3Social Identity TheorySocial Identity Theory has its roots in work byTajfel during the late 1950s and 1960s on thesocial factors that influence perception. It wasfurther developed with Turner in the 1970s and1980s (Tajfel & Turner 1979, Turner & Tajfel1982; see also Hogg 1992). Like all the theorieswe review here, Social Identity Theorysees social actors as having multiple identitiesthat get activated by different social contexts.Here, the emphasis is on category memberships(e.g., being a Muslim or an Australian). Whena category membership becomes salient (relevantto the social context), self-perception andconduct become in-group normative. Perceptionsof other groups become out-group stereotypical.Depending on the nature of the relationshipbetween the groups (e.g., whether it iscompetitive or status-ranked), self- and otherperceptionscan shift toward different types ofperceptual discrimination.Unlike the control theories of identityreviewed above, Social Identity Theoryassumes a self-enhancement motive. Afterself-categorization occurs in the context of asituation, actors are motivated to make comparisonsthat favor the in-group (and sometimesdisparage the out-group). The strength of thistendency varies, depending on the potential formobility—it is stronger for immutable ascribedcategorizations (e.g., race, gender, nationality)3 Other treatments (e.g., Stets & Burke 2000) have emphasizedpoints of overlap between Social Identity Theory inpsychology and Identity Theory in sociology.than it is for achieved categorizations (e.g., educationaldegree status, membership on sportsteams), especially if the higher status is withinreach. Therefore, it is the social environmentand, in particular, the relevant group contrastswithin that environment that determinewhat dimensions of self-perception, otherstereotyping,and intergroup competition/discrimination occur. If I think of myself as anAmerican and I am in a social context wherethe British are my out-group, I might thinkof my American self as egalitarian (as opposedto the class-oriented British). But if I am (still)an American and I am comparing myself toan Italian, I might think of myself as orderlyand institution-upholding (as opposed to theunsettled, fractious Italians). The hypothesesof Social Identity Theory typically focus onhow variations in the social situation (therelevant out-groups, the beliefs about thedegree of mobility and social change possible,etc.) influence the content of the operativemeanings for the in-group (i.e., situated selfmeanings) and the out-group (stereotypes anddiscrimination).Notice that in Social Identity Theory, theconcept of salience has a very different meaningfrom that in Identity Theory. In Stryker’s IdentityTheory, identity salience is a stable part ofthe self—a result of commitment (frequent andaffectively valued relationships to others). It issomething that actors carry from situation tosituation. In Social Identity Theory, salience isconceptualized as the impact of the situation onself-categorizations. I might be an American inParis, but at a Civil War reenactment, I am aSoutherner, and in the American SociologicalAssociation, I am a professor and a socialpsychologist. And in each of these venues,I might have different meanings for thosecategorical identities, depending on the salientout-group at the time. In comparing Tajfel’sSocial Identity Theory and Stryker’s IdentityTheory, Hogg et al. (1995, p. 263) notedthat Tajfel’s theory put more emphasis onsocial context for identity and their meanings,whereas the sociological Identity Theoryfocused on self-structure for motivating the488 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.enactment of identities and long-term socialexperience for the internalized meanings.To end our section on situational/culturaltheories of identity, we turn to a theory thatis not considered an identity theory at all—theStatus Characteristics Theory branch of the expectationstates theoretical research program.We include the theory in this review not tocover all of its rich contributions to microsociology,but to highlight the elements of thetheory that deal with the relationships amongidentity and interaction. The increasing use ofthe theory outside of its scope conditions byidentity researchers justifies its placement here.Status Characteristics TheoryThe intellectual roots of Status CharacteristicsTheory are not in symbolic interaction, but inthe exchange tradition (Correll & Ridgeway2003). When people work together on a taskand will be rewarded jointly for its successfulcompletion (the scope conditions of the theory),actors are motivated to assess who willcontribute most effectively to the task completion.The actors for whom task expectationsare the highest are given action opportunities,receive positive evaluations, and have other interactionaladvantages. Actors’ characteristics(e.g., gender, race) become important withina situation when those characteristics areevaluated by self and others, and are perceived(either implicitly or explicitly) to be relevantto expectations about task performance. Thetheory’s core mechanism is that group memberswill exchange deference within the groupinteraction for the high-expectation actors’contributions to the task (resulting in higherbenefits for all group members, including thelower-expectation ones).Note, however, the similarities betweenStatus Characteristics Theory and Affect ControlTheory (Ridgeway & Smith-Lovin 1994).Both theories posit that largely consensualcultural meanings are imported into a situationto organize action within that setting. Boththeories suggest that the identities (and, therefore,cultural meanings) that are relevant aredetermined by the situation, not by an internalself-structure. And both provide a generativeaccount of how action unfolds in the situation,given these meanings. The fact that StatusCharacteristics Theory is frequently nowapplied outside of its scope conditions of collaborative,jointly rewarded task orientation—which nullifies the exchange mechanism—makes it closer to a situationally based identitytheory than the status-exchange theory of itsorigins. Researchers in both the Affect ControlTheory and the Identity Control Theory traditionshave used the theory to discuss how statusmeanings organize action within social situations(Ridgeway & Smith-Lovin 1994, Rashotte& Smith-Lovin 1997, Stets 1997, Stets &Harrod 2004, Cast et al. 1999). For example,Stets & Harrod (2004) find that actors inhigher-status positions have interactionalresources that allow them to sustain theiridentity meanings within situations, avoidingnegative emotions that come with lack ofidentity maintenance.Having reviewed the systematically developedtheories of identity and action thatemphasize internalized meanings and situational/culturalelements, respectively, we nowturn to a body of literature on collective identitythat has a very different emphasis. Indeed,a review of this literature a decade ago(Cerulo 1997) barely mentions social psychologicaltheories. Here, we try to summarizethe differing emphasis of this substantial literatureon collective identity and to draw outconnections between its treatment of identityand the social psychological literature reviewedabove.COLLECTIVE IDENTITYThe solidarity that derives from similarities is at itsmaximum when the collective consciousness completelyenvelopes our total consciousness, coincidingwith it at every point.—Emile Durkheim (1893)The sociological literature on collective identityis less organized around distinct midrangewww.annualreviews.org • Three Faces of Identity 489

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.theories than are the literatures on personalidentity and situated identity. Rather, it is awide-ranging literature coming out of a numberof traditions, including cultural sociology,social movements theory, feminist sociology,and cognitive sociology. Consequently,this section of our review is organized aroundconcepts and process rather than around specifictheoretical traditions. We attempt to drawconnections between this vibrant literatureand the midrange microsociological theoriesreviewed above.The internal and situational identity processesreviewed in the section above can leadidentities to serve as an organizational force,binding us to those with commonalities of interestand providing a social glue that can serveas a foundation for mobilizing joint action. Collectiveidentities should not be seen as a thirdtype of identity as much as an attempt by thisliterature to highlight another set of identityrelateddynamics at the group level. Both internalizedidentities and situational identities canunderpin the sense of connection and shareddestiny requisite for collective identification.Indeed, Nagel (1995, p. 21) refers to the collectivevariant of ethnic identity as “a dialecticbetween internal identification and externalascription.”Most definitions of collective identityinclude a notion of identification with sharedfeatures along with a recognition of sharedopportunities and constraints afforded by thosefeatures (Melucci 1989). Taylor & Whittier(1999, p. 170) define collective identity as the“shared definition of a group that derives frommembers’ common interests, experiences, andsolidarity.” A sense of we-ness, or connectionto other members of the group/category, isan essential component of collective identity,but the concept goes far beyond that. Prenticeet al. (1994) specifically distinguish betweengroup identities based on common bonds(attachments to individual group members)and those based on common identities (attachmentsdirectly to the group or category).The latter form of attachment is necessary toproduce a collective identity.Psychologists working in the social identityand self-categorization theoretical traditionsrefer to the “collective self” as nearly interchangeablewith “social identity” as used inthose theories (Brewer & Gardner 1996) andthus focus largely on its consequences forself-definition and interpersonal judgment.Sociological use of the term collective identityfocuses more on its consequences for mobilizingjoint action. This is closer to what Heise(1998) referred to as empathic solidarity. Heise(1998, p. 197) defines empathic solidarity as“a reciprocated sense of merged consciousnessand alliance, with faith in others’ commitmentsto shared purposes.” Heise argues that whenpeople take on the same identity, experiencethe same reality, and observe one another’sparallel emotions and collateral behaviors,a sense of common destiny and empathicconnection arises. It is this phenomenon,operating at the group level, that makes thesociological literature on collective identitydistinctive from the microsociological identityliterature that we review above.Collective Identity and theNew Social MovementsA concern with understanding what mobilizesjoint action motivates much of sociology’s attentionto collective identity. Collective identityis a central organizing concept in the literatureon new social movements. New socialmovement theory (Laraña et al. 1994; Melucci1989, 1994; Offe 1985; Turner 1969) juxtaposesmodern (post-1960s) movements with earliersocial movements, arguing that contemporarymovements are less about Marxist-style conflictover material interests and more concernedwith identity meanings and other symbolic resources.For example, Turner (1969) relatedprotest participation to identity dynamics andreferred to the emerging number of “identityseeking” movements. The identity dynamicsdescribed by Turner focused on personal identitiesand the identification/disidentificationwith various groups. Later researchers refinedthe notion of collective identity as a group-level490 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.phenomenon (Melucci 1989, 1994) central togiving a group or category the coherence andenergy necessary to mobilize its constituentsinto collective action. Nearly a decade ago,Polletta & Jasper (2001) reviewed the literatureon collective identity and social movements, revealinghow identity processes are deeply relatedto all aspects of social movements, including(a) movement emergence, (b) recruitmentand participation, (c) movement strategy, and(d) interpretation of outcomes (see Stryker et al.2000).Boundary WorkTaylor & Whittier (1999) describe three factorsthat contribute to the development of a collectiveidentity capable of motivating group-levelaction: (a) the creation of social boundaries,(b) the development and recognition of socialcriteria that account for a group’s structural position,and (c) negotiation of intergroup and intragroupmeanings. Boundaries identify who isand is not a member of a collective, but it isgroup consciousness that gives significance to acollectivity (Taylor & Whittier 1999, p. 179).In the case of mobilizing collective consciousnessin response to a dominant group or oppressiveunderstanding, this significance takes theform of what Morris (1999) calls oppositionalconsciousness. Strategies and practices used bycollectives to create, maintain, and transformcultural categories are collectively referred toas boundary work (Nippert-Eng 2002).Gamson (1995) points out that fixed categoriesare the basis for both oppression andpolitical power. Consequently, boundary workregarding collective identities such as race,gender, and sexuality can sometimes vacillatebetween taking on the goal of deconstructingboundaries (e.g., Lorber 2006) and takingon the goal of fixing the boundaries tomobilize on the basis of them (e.g., Jenson1995).Within this theoretical tradition, Lamont(1992) offered an in-depth analysis of the symbolicboundaries people draw when categorizingself and others in her comparative study ofupper-middle-class culture in France and theUnited States. Her work suggests that boundarywork around collective identities operatesmultidimensionally and that the relative prominenceof those dimensions can vary betweencultural settings (echoing some themes in SocialIdentity Theory). Her analysis distinguishedbetween three types of symbolic boundaries:moral, socioeconomic, and cultural. Accordingto Lamont’s research, French social and culturalspecialists draw stronger cultural boundaries,while social and cultural specialists inthe United States attend more to economicboundaries. More specifically, Lamont foundthat boundary-drawing activities in the UnitedStates vary with the degree to which they areembedded in occupational structures that aredependent on profit-making. U.S. social andcultural specialists in the nonprofit sector feltmore similar to, and connected with, other nonprofitworkers than to other social and culturalspecialists who were dependent on profitmaking.In contrast, French social and culturalspecialists identified more with the intellectualand cultural boundaries around their vocationsand less with the economic aspects.The boundary work literature focusesprimarily on the strategies and practices thatgroups use to construct collective identitiesand to manage the symbolic boundaries aroundthose identities. However, theorists have alsoattended to how external structures createand constrain the emergence and salience ofcollective identities. Again, this work picksup some themes emphasized in the sociocultural/situationaltheories of identity inmicrosociology.Competition and ContactTwo opposing logics guide the theoretical literatureregarding the impact of external forceson the salience of collective identity boundariesand the strength of collective identity bonds.One is that direct competition strengthensgroup boundaries and mobilizes collectiveidentity-based conflict. Competition theory(Olzak 1994) argues that when economic anddemographic changes lead to a breakdown inlabor market segmentation, the increase inwww.annualreviews.org • Three Faces of Identity 491

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.intergroup interaction and the intensificationof competition for scarce resources strain intergrouprelations and, consequently, increasecollective identification—particularly ethnicidentification. Olzak and colleagues haveamassed considerable support for the idea thatincreased economic competition leads to surgesin ethnic protests and collective action (Olzaket al. 1994, 1996; Soule 1992). Nagel (1995),on the other hand, found that cultural renewalamong Native Americans has taken place underexactly the kinds of conditions that are thoughtto produce cultural decline. One argument isthat positive economic conditions set the stagefor this renewal. According to Nagel, structuralexternalities such as successful land claims,increases in federal spending, and minorityset aside programs have increased the symbolicand material value of Native Americanidentity.In contrast to the idea that increasedinteraction and competition foster collectiveidentification, the cultural division of labortheory proposes that group solidarity andcollective identity arise when groups aredistinctively positioned in an occupationalstructure on the basis of cultural markers(Hechter 1978, 2000). Occupational segregationincreases intragroup interaction relativeto intergroup interaction and consequentlyincreases commonality of interests and futures.Thus, when cultural and economic boundariesoverlap, collective identity will arise on the basisof cultural, rather than economic, similarity.Okamoto (2003) extended and synthesizedthese contrasting arguments into a theory aboutthe shifting, layered nature of ethnic identities.In a study of the relationship betweeneconomic competition, cultural division of labor,and pan-Asian identity, Okamoto foundthat the collective identity boundaries trackedthe patterns of occupational segregation: WhenAsian ethnic groups occupy a shared place inthe occupational market, pan-Asian identitiesare more likely to emerge and mobilize panethniccollective action. When separate Asianethnic groups are segregated into different occupations,the separate ethnic identities remainsalient and pan-ethnic collective action is suppressed.Nested IdentitiesAs Okamoto’s work highlights, boundary workby collectives is not simply a matter of sharpeningcontrasts between competing groups or emphasizingdistinctiveness from a larger oppressiveculture. At times collective identificationis a process of crystallizing subgroup boundariesthat fractionalize a larger whole, or meldingsubgroups into larger, cohesive collectives.Green (1999) describes four groups of sectarianevangelical Protestants whose efforts animatedthe first wave of a Christian Right movement,but whose intergroup conflicts ultimately contributedto the decline of the movement. Thesuccess of a second wave of the Christian Rightmovement, according to Green, was facilitatedby the development of a new collective identityfor these groups.Similarly, in her study of the emergence ofa pan-tribal Native American identity, Nagel(1995) described the process by which subgroupscan join to form a larger common identity.She referred to ethnic renewal as theprocess by which new ethnic identities arebuilt/rebuilt out of historical social and symbolicsystems. Among Native Americans, ashared history of discrimination and oppression,combined with some of the positive economicexternalities described above, created acommon fate among ethnic groups with otherwisedistinctive cultural histories. These developmentsfacilitated the collective processes(e.g., institution building via the establishmentof new organizations and religions) and culturalpractices (development of new shared ritualsand symbols) involved in ethnic renewal.Identity, Emotions, and MobilizationClose on the heels of renewed attention tothe role of identity in social movements hasbeen a revival of interest in how emotionsmobilize joint action. Scholars have recentlybecome more focused on how emotions create492 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.solidarity and energize collective bonds (Britt& Heise 2000, Collins 1990, Gould 2004,Heise 1998, Jasper 1998). Britt & Heise (2000)point out that shared emotion is not enoughto generate collective identification and action.Using arguments from Affect Control Theory,Britt & Heise (2000) argued that negative emotionsthat are low energy and low potency (e.g.,shame and depression) are not useful for activatingcollective bonds or motivating joint action.In contrast, negative emotions that are higherin energy and potency (e.g., anger) can motivateindividual participation in collective actions—from mild protests to large-scale conflict. Totruly energize a mobilized collective identity,however, requires both energy and a positivesense of connection within the collective. Suchmobilization is facilitated, according to Britt& Heise, by positive, powerful, and energeticemotions like pride. Britt & Heise argued thatcertain emotion transformations (e.g., shame topride) are difficult to make without transitioningbetween more affectively similar emotionsalong the way. They illustrated how aspectsof participation in the gay rights movementcould be seen as efforts to transform shame tofear to anger to pride in an effort to constructa solidary and energized collective identity.In a participant observation study of politicalavoidance in several volunteer organizations,Eliasoph (2002) examined the complement tothis process. Eliasoph studied how political disengagementis socially produced in interaction.Her analysis focused on how individuals balancedideas of citizenship with feelings of powerlessness.She described the process of “culturework” done by volunteers working to addresspolitical avoidance by transforming feelings ofpowerlessness into expansion of self-interest. Inthis way, Eliasoph showed how emotion transformationin an effort to protect personal identitycan also inhibit activation and mobilizationof a collective identity.Collective Identity and ModernityMuch has been written about the relationshipbetween modernity and contemporary loci ofcollective identity. The conventional argumentis that social and institutional complexity oflate modernity (Giddens 1991) or postmodernity(Gergen 1991) has fractionalized the contemporaryself. Wimmer (2002) speaks to thisidea in his study of nationalism and ethnicity,but argues for the opposite relationship. Accordingto Wimmer, contemporary notions ofidentity are responsible for the emergence ofmodernity. Wimmer distinguishes three positionson the issue of nationalism and ethnicity.The first is that nations and ethnic groups aretruly modern phenomena. Second, nations andethnic groups as we currently understand themare transitory—what he refers to as “birth painsof modernity.” Third, national and ethnic identitiesare perennial and basic to human socialorganization. Wimmer then presents what hedescribes as a radical modern argument: thatethnicity nationalism produced modernity aswe understand it. According to Wimmer, it wasthe fusion of three notions of peoplehood thatresulted in the modern politicization of ethnicity:(a) the notion of the people as a sovereignentity—with power by means of political (suchas democratic) procedures; (b) the notion ofpeople as citizens—with rights and responsibilities;and (c) the notion of people as an ethniccommunity bound by common political destinyand shared cultural features. The comminglingof these notions translated into three modernpolitical principles: democracy, citizenship,and national self-determination. Wimmer furtherargues that nations differ in the relativesalience of these principles. According toWimmer, the French and Swiss emphasizedemocracy and rely on the concept of democracyto derive the principles of citizenship andnationalism. In contrast, Wimmer argues thatGermany, Greece, and Israel stress nationalityand from it derive the principles of democracyand citizenship.While social movement theorists focuson both aspects of collective identity—common interests/fate and common symbolicmeanings—there is a strong thread in themodernity/postmodernity literature thatdocuments a shift over time in the relativewww.annualreviews.org • Three Faces of Identity 493

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.importance of these two aspects in shapingcategory-based action. Pre-1960s movementsare presumed to be motivated more by morestructural, ascriptive category membershipsand jointness of interest, whereas post-1960smovements are thought to be driven moreby meaning-based bonds (Laraña et al. 1994,Melucci 1994, Offe 1985). This argument aboutthe trend away from the importance of structurallydefined group identities toward morepersonally and culturally defined group identitiesparallels the findings in social psychologyabout the transition (over a similar historicalperiod) from more structural/institutional definitionsof self to more personal/dispositionaldefinitions of self (Turner 1976).Fractal Organizationof Identity DynamicsCollective identities are to groups what socialidentities are to individuals (Owens 2003,p. 227). Therefore, collective identity processescan sometimes mirror the processes describedby the theories about internal and situationalidentity dynamics reviewed above. The processof boundary work, often fueled by competitionor contact, can lead to the development of anidentity, making it available for members of acategory to see themselves as having a commonset of interests or fate. Sometimes new labelsare developed and added to a culture’s “culturaltheory of people” (MacKinnon & Heise 2010),making them available for individuals to adoptand incorporate into a self-structure. Further,intergroup dynamics can lead to situations thatallow groups and/or individuals to adopt differentidentity orientations at different momentsin a political/social process.While identity accumulation theory (Thoits1983, 2003) and self-complexity theory(Linville 1987) argue that having access tomultiple role-identities can provide the selfwith resources that help provide a bufferagainst stress, the social movement literaturereveals that collective identities with layersof nested identities can provide a movementwith a richer set of resources for mobilizingeffective collective action (Heckathorn 1993,Lichterman 1999, Oliver & Marwell 1988,Richards 2004). Moreover, just as multipleidentities within self-structures can sometimeslead to identity conflict, competing or layeredidentities within collectives can createopportunities for conflict and fractionalization.Richards (2004) illustrates this point withher examination of the delicate balancing actrequired to mobilize and maintain collectiveidentity in the ethnically and economically diversewomen’s movements in Chile. She studiedthe relationships between the poor, workingclasspobladoras, the indigenous Mapuche, andthe National Women’s Service (SERNAM).Nested class and ethnic identities set the stagefor potential fractures among these variousgroups involved in the women’s movement.Both the pobladoras and the Mapuche describebeing left out of the programs implemented bySERNAM. Pobladoras identified gender as importantin explaining their activism—yet initiallyorganized around roles as wives andmothers. Richards describes how the pobladoras’self-views became transformed through participationin the movement. Meanwhile, theMapuche claims of difference and disadvantageconflicted with SERNAM’s goal of equality/sameness,causing schisms in both discourseand agendas. Richards’s work showed howmembers of the women’s movement navigatethe danger of fragmentation by suppressing theseparate identities in the superordinate collective.This careful negotiation of a coherent,integrated collective identity echoes the findingsof Lichterman’s (1999) study of a sexualminority activist group as it attempted to createand maintain a solidary collective identityas a coalition-building network. Lichtermanfound that compared to some of its constituentgroups, the LGBT coalition-buildinggroup actively avoided reflective discourse andengaged in more unified talk carefully designednot to raise the specter of differenceor to activate potentially schismatic organizingprocesses.494 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.CONCLUSIONS ANDOPPORTUNITIESThis review, with its unusual structure and itscoverage of the relatively disconnected literatureson identity within the microsociology andsocial movements areas, is designed to allowreaders to see opportunities for enriching bothtraditions. We deal first with the opportunityfor connections within the microlevel traditionsand then with potential for enrichment acrossthe two distinct literatures.The relatively subtle distinction that wehave drawn between theories that emphasizehow self-structures and internalized meaningsorganize social life, on the one hand, and howwe import cultural meanings and shape themwithin the context of situated interaction, onthe other, points to a need for two types of newresearch. While much of the microsociologicalliterature has used either surveys (to studyself-structure and internalized meanings) orexperiments (to study situated meanings andhow they shift in context), we may need newstudy of naturally occurring situations to fullyexploit the linkages between internalized structureand situated action. Some insights fromStryker’s Identity Theory could be linked to thefast-growing literature on networks to developthe linkage between commitment and saliencemore fully and to explore how it impacts situatedaction. The new work by MacKinnon &Heise (2010; see also Moore & Robinson 2006)about how people use movement from onesituation to another to maintain fundamentalsentiments toward the self could be linkedto Thoit’s work on identity accumulation toshow how situated action accomplishes mentalhealth benefits (or, among some categories ofpeople, fails to do so). Exploring the few competingtheoretical predictions from IdentityControl Theory and Affect Control Theory(Smith-Lovin & Robinson 2006) might allowus to assess the relative value of emic, individualizedmeaning measurement (as used inBurke’s work) and mathematical formalization(as used by Heise). It might also reveal somescope conditions for when people privilegeinternalized, idiosyncratic meanings over adefinition of situation that is institutionallyimposed.Connections between the microsociologicalliterature and the group-level treatment of collectiveidentity are more challenging. The obviouspoint of departure is that the people who aremotivated by collective identities toward politicalor social action are undergoing the processesalready well described by the microsociologicaltheories. This fact alone might provide someleverage for determining when recruitment, reframingof a social situation, and mobilizationmight succeed or fail. More interesting are theways in which the collective identity literaturecan suggest new topics for microsociological research.Much of this literature focuses on thestructural conditions under which new optionsare created for a definition of the situation. Thedefinition of the situation is a central processin microsociological identity theories, but it israrely studied explicitly. The collective identityliterature’s use of Goffman’s framing concept,and its attention to the inter- and intragroupprocesses that successfully motivate actors toview a contested situation in a new way, couldpoint to a useful exploration of these key processesat the microlevel. A second area of potentialenrichment comes from the collectiveidentity focus on the group as a unit that generatessocial action. As we noted above, thereis a parallel between processes at the individualand group levels, both in defining situations andin generating action. Although theorists shouldavoid loose analogies, the processes by whichgroups manage multiple, nested, layered identitiesand use particular identities (and the emotionalresponses associated with them) to motivateorganized collective action may mirror theprocess of managing nested identity structureswithin individual selves. An exploration of thisprocess might require a refocusing of attentionon group-level studies in microsociology. Wehave excellent midlevel theories of group processin sociological social psychology, and theyare becoming increasingly connected intellectuallyto the issue of identity [see, for an excellentexample, Lawler et al.’s (2009) new bookwww.annualreviews.org • Three Faces of Identity 495

on affective commitments to groups]. We needto push these developments to see how groupsorganize actions vis-à-vis other outside grouplevelinterests.DISCLOSURE STATEMENTThe authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.ACKNOWLEDGMENTSWe thank Tiffani Everett, Steven Foy, Robert Freeland, Elizabeth Hordge-Freeman, ChristopherD. Moore, Victor Ray, Kim Rogers, Daniel B. Shank, and Allison Wisecup for helpful commentson an earlier draft of this review.Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.LITERATURE CITEDAhrens C, Ryff C. 2006. Multiple roles and well-being: sociodemographic and psychological moderators. SexRoles 55:801–15Alexander CN Jr, Knight GW. 1971. Situated identities and social psychological experimentation. Sociometry34:65–82Alexander CN Jr, Wiley MG. 1981. Situated activity and identity formation. In Social Psychology: SociologicalPerspectives, ed. M Rosenburg, RH Turner, pp. 269–82. New York: Basic BooksBandura A. 1977. Social Learning Theory. Englewood Cliffs, NJ: Prentice HallBarnett RC, Hyde JS. 2001. Women, men, work, and family. Am. Psychol. 56:781–96Brewer MB, Gardner WL. 1996. Who is this “we”? Levels of collective identity and self representations.J. Personal. Soc. Psychol. 71:83–93Britt L, Heise DR. 2000. From shame to pride in identity politics. See Stryker et al. 2000, pp. 252–71Brook AT, Garcia J, Fleming M. 2008. The effects of multiple identities on psychological well-being. Personal.Soc. Psychol. Bull. 34:1588–600Burke PJ. 1991. Identity processes and social stress. Am. Sociol. Rev. 56:836–49Burke PJ. 2004. Identities and social structure: the 2003 Cooley-Mead Award address. Soc. Psychol. Q. 67:5–15Burke PJ, Reitzes DC. 1991. An identity theory approach to commitment. Soc. Psychol. Q. 54:239–51Burke PJ, Stets JE. 1999. Trust and commitment through self-verification. Soc. Psychol. Q. 62:347–66Burke PJ, Stets JE. 2009. Identity Theory. New York: Oxford Univ. PressBurke PJ, Stets JE, Cerven C. 2007. Gender, legitimation, and identity verification in groups. Soc. Psychol. Q.70:27–42Burke PJ, Tully JT. 1977. The measurement of role/identity. Soc. Forces 55:880–97Cast AD, Stets JE, Burke PJ. 1999. Does the self conform to the views of others? Soc. Psychol. Q. 62:68–82Cerulo KA. 1997. Identity construction: new issues, new directions. Annu. Rev. Sociol. 23:385–409Collins R. 1990. Stratification, emotional energy, and the transient emotions. In Research Agendas in the Sociologyof Emotions, ed. TD Kemper, pp. 27–57. Albany: SUNY PressCooley CH. 1902. Human Nature and the Social Order. New York: ScribnerCorrell SJ, Ridgeway CL. 2003. Expectation states theory. In Handbook of Social Psychology, ed. J Delamater,pp. 29–51. New York: KluwerDurkheim E. 1893 (1984). The Division of Labor in Society. Transl. WD Halls. New York: Free PressEliasoph N. 2002. “Close to Home”: the work of avoiding politics. In Cultural Sociology, ed. L Spillman,pp. 130–40. Oxford: Oxford Univ. PressFoote NN. 1951. Identification as the basis for a theory of motivation. Am. Sociol. Rev. 16:14–21Gamson WA. 1995. Constructing social protest. In Social Movements and Culture, ed. H Johnston,B Klandermans, pp. 85–106. New York: RoutledgeGergen KJ. 1991. The Saturated Self: Dilemmas of Identity in Contemporary Life. NewYork:BasicBooks496 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.Giddens A. 1991. Modernity and Self-Identity: Self and Society in the Late Modern Age. Stanford, CA: StanfordUniv. PressGoffman E. 1959. The Presentation of Self in Everyday Life. Garden City, NJ: DoubledayGoffman E. 1963. Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs, NJ: Prentice-HallGoffman E. 1974. Frame Analysis: An Essay on the Organization of Experience. NewYork:Harper&RowGollob HF. 1968. Impression formation and word combination in sentences. J. Personal. Soc. Psychol. 10:341–53Gould DB. 2004. Passionate political processes: bringing emotions back into the study of social movements.In Rethinking Social Movements: Structure, Meaning and Emotion, ed. J Goodwin, JM Jasper, pp. 155–75.Lanham, MD: Roman & LittlefieldGreen JC. 1999. The spirit willing: collective identity and the development of the Christian right. In Wavesof Protest: Social Movements Since the Sixties, ed. J Freeman, V Johnson, pp. 153–68. Oxford: Oxford Univ.PressHechter M. 1978. Group formation and the cultural division of labor. Am. J. Sociol. 84:293–318Hechter M. 2000. Containing Nationalism. Oxford: Oxford Univ. PressHeckathorn DD. 1993. Collective action and group heterogeneity: voluntary provision versus selective incentives.Am. Sociol. Rev. 58:329–50Heider F. 1946. Attitudes and cognitive organization. J. Psychol. 21:107–12Heider F. 1958. The Psychology of Interpersonal Relations. New York: WileyHeise DR. 1979. Understanding Events: Affect and the Construction of Social Action. Cambridge, UK/New York:Cambridge Univ. Press. 197 pp.Heise DR. 1998. Conditions for empathic solidarity. In The Problem of Solidarity: Theories and Models, ed.P Doreian, T Fararo, pp. 197–212. Amsterdam: Gordon & BreachHeise DR. 2007. Expressive Order: Confirming Sentiments in Social Actions. New York: Springer. 163 pp.Hochschild AR. 1983. The Managed Heart: Commercialization of Human Feeling. Berkeley: Univ. Calif. Press.307 pp.Hogg MA. 1992. The Social Psychology of Group Cohesiveness: From Attraction to Social Identity. Hemel Hempstead,UK: Harvester WheatsheafHogg MA, Terry DJ, White KM. 1995. A tale of two theories: a critical comparison of identity theory withsocial identity theory. Soc. Psychol. Q. 58:255–69Horney K. 1945. Our Inner Conflicts: A Constructive Theory of Neurosis. New York: NortonJackson PB. 1997. Role occupancy and minority mental health. J. Health Soc. Behav. 38:237–55James W. 1890. The Principles of Psychology. New York: Henry HoltJanzen BL, Muhajarine N. 2003. Social role occupancy, gender, income adequacy, life stage and health: alongitudinal study of employed Canadian men and women. Soc. Sci. Med. 57:1491–503Jasper JM. 1998. The emotions of protest: affective and reactive emotions in and around social movements.Sociol. Forum 13:397–424Jenson J. 1995. What’s in a name? Nationalist movements and public discourse. In Social Movements andCulture, ed. HJB Klandermans, pp. 107–26. New York: RoutledgeKikuzawa S. 2006. Multiple roles and mental health in cross-cultural perspective: the elderly in the UnitedStates and Japan. J. Health Soc. Behav. 47:62–76Lamont M. 1992. Money, Morals, and Manners. Univ. Chicago PressLawler EJ, Thye SR, Yoon J. 2009. Social Commitments in a Depersonalized World. New York: Russell SageFound.Laraña E, Johnston H, Gusfield JR. 1994. New Social Movements: From Ideology to Identity. Philadelphia: TempleUniv. PressLichterman P. 1999. Talking identity in the public sphere: broad visions and small spaces in sexual identitypolitics. Theory Soc. 28:101–41Linville PW. 1987. Self-complexity as a cognitive buffer against stress-related illness and depression. J. Personal.Soc. Psychol. 52:663–76Lorber J. 2006. Shifting paradigms and challenging categories. Soc. Probl. 53:448–53MacKinnon NJ. 1994. Symbolic Interactionism as Affect Control. Albany: SUNY PressMacKinnon NJ, Heise DR. 2010. Self, Identity, and Social Institutions. New York: Palgrave MacMillanwww.annualreviews.org • Three Faces of Identity 497

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.McCall GJ, Simmons JL. 1966. Identities and Interactions. New York: Free PressMcPherson M, Smith-Lovin L, Cook J. 2001. Birds of a feather: homophily in social networks. Annu. Rev.Sociol. 27:415–44McPherson M, Smith-Lovin L, Brashears ME. 2006. Social isolation in America: changes in core discussionnetworks over two decades. Am. Sociol. Rev. 71:353–75Mead GH. 1934. Mind, Self, and Society from the Standpoint of a Social Behaviorist. Chicago: Univ. ChicagoPressMelucci A. 1989. Nomads of the Present. Philadelphia: Temple Univ. PressMelucci A. 1994. A strange kind of newness: What’s new in new social movements? See Laraña et al. 1994,pp. 101–32Moore CD, Robinson DT. 2006. Selective identity preferences: choosing from among alternative occupationalidentities. Adv. Group Process. 23:253–81Morris AD. 1999. A retrospective on the civil rights movement: political and intellectual landmarks. Annu.Rev. Sociol. 25:517–39Nagel J. 1995. American Indian Renewal: Red Power and the Resurgence of Identity and Culture. Oxford: OxfordUniv. PressNippert-Eng CE. 2002. Boundary work: sculpting work and home. In Cultural Sociology, ed. L Spillman,pp. 79–86. Oxford: Oxford Univ. PressOffe C. 1985. New social movements: challenging the boundaries of institutional politics. Soc. Res. 52:817–68Okamoto DG. 2003. Toward a theory of panethnicity: explaining Asian American collective action. Am. Sociol.Rev. 68:811–42Oliver PE, Marwell G. 1988. The paradox of group size in collective action: a theory of the critical mass. II.Am. Sociol. Rev. 53:1–8Olzak S. 1994. The Dynamics of Ethnic Competition and Conflict. Stanford, CA: Stanford Univ. PressOlzak S, Shanahan S, McEneaney EH. 1996. Poverty, segregation, and race riots: 1960 to 1993. Am. Sociol.Rev. 61:590–613Olzak S, Shanahan S, West E. 1994. School desegregation, interracial exposure, and antibusing activity incontemporary urban America. Am. J. Sociol. 100:196–241Osgood CE, May WH, Miron MS. 1975. Cross-Cultural Universals of Affective Meaning. Urbana: Univ. Ill.PressOsgood CE, Suci GJ, Tannenbaum PH. 1957. The Measurement of Meaning. Urbana: Univ. Ill. PressOwens TJ. 2003. Self and identity. In Handbook of Social Psychology, ed. JD Delamater, pp. 205–32. New York:KluwerOwens TJ, Aronson PJ. 2000. Self-concept as a force in social movement involvement. See Stryker et al. 2000,pp. 132–51Owens TJ, Serpe R. 2003. The role of self-esteem in family identity salience and commitment among African-Americans, Latinos, and whites. In Advances in Identity Theory and Research, ed. PJ Burke, TJ Owens,R Serpe, PA Thoits, pp. 85–104. New York: KluwerPolletta F, Jasper JM. 2001. Collective identity and social movements. Annu. Rev. Sociol. 27:283–305Powers WT. 1973. Behavior: The Control of Perception. Chicago: AldinePrentice DA, Miller DT, Lightdale JR. 1994. Asymmetries in attachments to groups and to their members:distinguishing between common-identity and common-bond groups. Personal. Soc. Psychol. Bull. 20:484–93Rashotte LS, Smith-Lovin L. 1997. Who benefits from being bold: the interactive effects of task cues andstatus characteristics. Adv. Group Process. 14:235–55Richards P. 2004. Pobladoras, Indigenas, and the State: Conflicts over Women’s Rights in Chile. New Brunswick,NJ: Rutgers Univ. PressRidgeway C, Smith-Lovin L. 1994. Structure, culture and interaction: comparing two generative theories.Adv. Group Process. 11:213–39Robinson DT. 2007. Control theories in sociology. Annu. Rev. Sociol. 33:157–74Robinson DT, Smith-Lovin L. 1992. Selective interaction as a strategy for identity maintenance: an affectcontrol model. Soc. Psychol. Q. 55:12–28498 Owens·Robinson·Smith-Lovin

Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.Robinson DT, Smith-Lovin L. 1999. Emotion display as a strategy for identity negotiation. Motiv. Emot.23:73–104Robinson DT, Smith-Lovin L, Tsoudis O. 1994. Heinous crime or unfortunate accident? The effects ofremorse on responses to mock criminal confessions. Soc. Forces 73:175–90Rosenberg M. 1979. Conceiving the Self. New York: Basic BooksSachs-Ericsson N, Ciarlo JA. 2000. Gender, social roles, and mental health: an epidemiological perspective.Sex Roles 43:605–28Serpe RT. 1987. Stability and change in self: a structural symbolic interactionist explanation. Soc. Psychol. Q.50:44–55Smith-Lovin L. 1979. Behavior settings and impressions formed from social scenarios. Soc. Psychol. Q. 42:31–43Smith-Lovin L. 2007. The strength of weak identities: social structural sources of self, situation and emotionalexperience. Soc. Psychol. Q. 70:106–24Smith-Lovin L, Heise DR, eds. 1988. Analyzing Social Events: Advances in Affect Control Theory. New York:Gordon & BreachSmith-Lovin L, Robinson DT. 2006. Control theories of identity, action and emotion: in search of testabledifferences between affect control theory and identity control theory. In Purpose, Meaning and Action:Control Systems Theories in Sociology, ed. K McClelland, T Fararo, pp. 163–88. New York: MacMillanSoule SA. 1992. Populism and black lynching in Georgia, 1890–1900. Soc. Forces 71:431–49Stets JE. 1997. Status and identity in marital interaction. Soc. Psychol. Q. 60:185–217Stets JE, Burke PJ. 2000. Identity theory and social identity theory. Soc. Psychol. Q. 63:224–37Stets JE, Burke PJ. 2005. Identity verification, control, and aggression in marriage. Soc. Psychol. Q. 68:160–78Stets JE, Harrod MM. 2004. Verification across multiple identities: the role of status. Soc. Psychol. Q. 67:155–71Stets JE, Tsushima TM. 2001. Negative emotion and coping responses within identity control theory.Soc. Psychol. Q. 64:283–95Stryker S. 1968. Identity theory and role performance. J. Marriage Fam. 30:558–64Stryker S. 1980. Symbolic Interactionism: A Social Structural Version. Menlo Park, CA: Benjamin CummingsStryker S. 2008. From Mead to a structural symbolic interactionism and beyond. Annu. Rev. Sociol. 34:15–31Stryker S, Burke PJ. 2000. The past, present, and future of an identity theory. Soc. Psychol. Q. 63:284–97Stryker S, Owens TJ, White RW, eds. 2000. Self, Identity, and Social Movements. Minneapolis: Univ. Minn.PressStryker S, Serpe RT, Hunt MO. 2005. Making good on a promise: the impact of larger social structures oncommitments. Adv. Group Process. 22:93–124Tajfel H, Turner JC. 1979. The Psychology of Intergroup Relations. Monterey, CA: Brooks/ColeTaylor V, Whittier NE. 1999. Collective identity in social movement communities: lesbian feminist mobilization.In Waves of Protest: Social Movements Since the Sixties, ed. J Freeman, V Johnson, pp. 169–94.Oxford: Oxford Univ. PressThoits PA. 1983. Multiple identities and psychological well-being: a reformulation and test of the socialisolation hypothesis. Am. Sociol. Rev. 48:174–87Thoits PA. 1986. Social support as coping assistance. J. Consult. Clin. Psychol. 54:416–23Thoits PA. 1995. Identity-relevant events and psychological symptoms: a cautionary tale. J. Health Soc. Behav.36:72–82Thoits PA. 2003. Personal agency in the accumulation of multiple role-identities. In Advances in Identity Theoryand Research, ed. PJ Burke, TJ Owens, RT Serpe, PA Thoits, pp. 179–94. New York: KluwerThoits PA, Virshup LK. 1997. Me’s and we’s: forms and functions of social identities. In Self and Identity:Fundamental Issues, ed. RD Ashmore, L Jussim, pp. 106–33. New York: Oxford Univ. PressTurner JC, Tajfel H. 1982. Social Identity and Intergroup Relations. Cambridge, UK: Cambridge Univ. PressTurner RH. 1969. The public perception of protest. Am. Sociol. Rev. 34:815–31Turner RH. 1976. The real self: from institution to impulse. Am. J. Sociol. 81:989–1016Weinstein EA, Wiley MG, DeVaughn W. 1966. Role and interpersonal style as components of social interaction.Soc. Forces 45:210–16Wethington E, Moen P, Glasgo N, Pillemer K. 2000. Multiple roles, social integration, and health. In SocialIntegration in the Second Half of Life, ed. K Pillemer, P Moen, E Wethington, N Glasgo, pp. 48–71.Baltimore, MD: Johns Hopkins Univ. PressWimmer A. 2002. Nationalist Exclusion and Ethnic Conflict. Cambridge, UK: Cambridge Univ. Presswww.annualreviews.org • Three Faces of Identity 499

ContentsAnnual Reviewof SociologyVolume 36, 2010Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.FrontispieceJohn W. Meyer ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣xivPrefatory ChapterWorld Society, Institutional Theories, and the ActorJohn W. Meyer ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣1Theory and MethodsCausal Inference in Sociological ResearchMarkus Gangl ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣21Causal Mechanisms in the Social SciencesPeter Hedström and Petri Ylikoski ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣49Social ProcessesA World of Standards but not a Standard World: Toward a Sociologyof Standards and StandardizationStefan Timmermans and Steven Epstein ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣69Dynamics of Dyads in Social Networks: Assortative, Relational,and Proximity MechanismsMark T. Rivera, Sara B. Soderstrom, and Brian Uzzi ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣91From the Sociology of Intellectuals to the Sociology of InterventionsGil Eyal and Larissa Buchholz ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣117Social Relationships and Health Behavior Across the Life CourseDebra Umberson, Robert Crosnoe, and Corinne Reczek ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣139Partiality of Memberships in Categories and AudiencesMichael T. Hannan ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣159v

Institutions and CultureWhat Is Sociological about Music?William G. Roy and Timothy J. Dowd ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣183Cultural Holes: Beyond Relationality in Social Networks and CultureMark A. Pachucki and Ronald L. Breiger ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣205Formal OrganizationsAnnu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.Organizational Approaches to Inequality: Inertia, Relative Power,and EnvironmentsKevin Stainback, Donald Tomaskovic-Devey, and Sheryl Skaggs ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣225Political and Economic SociologyThe Contentiousness of Markets: Politics, Social Movements,and Institutional Change in MarketsBrayden G King and Nicholas A. Pearce ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣249Conservative and Right-Wing MovementsKathleen M. Blee and Kimberly A. Creasap ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣269The Political Consequences of Social MovementsEdwin Amenta, Neal Caren, Elizabeth Chiarello, and Yang Su ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣287Comparative Analyses of Public Attitudes Toward Immigrantsand Immigration Using Multinational Survey Data: A Reviewof Theories and ResearchAlin M. Ceobanu and Xavier Escandell ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣309Differentiation and StratificationIncome Inequality: New Trends and Research DirectionsLeslie McCall and Christine Percheski ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣329Socioeconomic Disparities in Health BehaviorsFred C. Pampel, Patrick M. Krueger, and Justin T. Denney ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣349Gender and Health InequalityJen’nan Ghazal Read and Bridget K. Gorman♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣371Incarceration and StratificationSara Wakefield and Christopher Uggen ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣387Achievement Inequality and the Institutional Structure of EducationalSystems: A Comparative PerspectiveHerman G. Van de Werfhorst and Jonathan J.B. Mijs ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣407viContents

Historical Studies of Social Mobility and StratificationMarco H.D. van Leeuwen and Ineke Maas ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣429Individual and SocietyRace and TrustSandra Susan Smith ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣453Three Faces of IdentityTimothy J. Owens, Dawn T. Robinson, and Lynn Smith-Lovin ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣477Annu. Rev. Sociol. 2010.36:477-499. Downloaded from www.annualreviews.orgby Vanderbilt University on 01/03/12. For personal use only.PolicyThe New Homelessness RevisitedBarrett A. Lee, Kimberly A. Tyler, and James D. Wright ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣501The Decline of Cash Welfare and Implications for Social Policyand PovertySandra K. Danziger ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣523IndexesCumulative Index of Contributing Authors, Volumes 27–36 ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣547Cumulative Index of Chapter Titles, Volumes 27–36 ♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣♣551ErrataAn online log of corrections to Annual Review of Sociology articles may be found athttp://soc.annualreviews.org/errata.shtmlContentsvii