A Sociological Approach to Self and Identity*

Jan E. Stets and Peter J. Burke Department of Sociology Washington State University

* Chapter for Handbook of Self and Identity, edited by Mark Leary and June Tangney, Guilford Press, Forthcoming.

A Sociological Approach to Self and Identity

Thoughts on Social Structure
A sociological approach to self and identity begins with the assumption that there is a reciprocal relationship between the self and society (Stryker, 1980). The self influences society through the actions of individuals thereby creating groups, organizations, networks, and institutions. And, reciprocally, society influences the self through its shared language and meanings that enable a person to take the role of the other, engage in social interaction, and reflect upon oneself as an object. The latter process of reflexivity constitutes the core of selfhood (McCall & Simmons, 1978; Mead, 1934). Because the self emerges in and is reflective of society, the sociological approach to understanding the self and its parts (identities) means that we must also understand the society in which the self is acting, and keep in mind that the self is always acting in a social context in which other selves exist (Stryker, 1980). This chapter focuses primarily on the nature of self and identity from a sociological perspective, thus some discussion of society is warranted. The nature of the self and what individuals do depends to a large extent on the society within which they live. In general, sociologists are interested in understanding the nature of society or social structure: its forms and patterns, the ways in which it develops and is transformed. The traditional symbolic interactionist perspective known as the situational approach to self and society, sees society as always in the process of being created through the interpretations and definitions of actors in situations (Blumer, 1969). Actors identify the things that need to be taken into account for themselves, act on the basis of those identifications, and attempt to fit their lines of action with others in the situation to accomplish their goals. From this perspective, the inference is made that individuals are free to define the situation in any way they care to, with the consequence that society is always thought to be in a state of flux with no real organization or structure. As Stryker (2000, p. 27) recently remarked on this perspective: [It] tends to dissolve structure in a solvent of subjective definitions, to view definitions as unanchored, open to any possibility, failing to recognize that some possibilities are more probable than others. On the premise that

self reflects society, this view leads to seeing self as undifferentiated, unorganized, unstable, and ephemeral. Our view of self and society is rooted in the structural approach to the symbolic interactionist perspective (Stryker, 1980). Within this perspective, we do not see society as tentatively shaped. Instead, we assume that society is stable and durable as reflected in the patterned regularities that characterize most human action (Stryker, 1980, p. 65). Patterns of behavior within and between individuals have different levels of analysis, and this is key to understanding the link between self and society. At one level, we can look at the patterns of behavior of one individual over time and come to know that individual. By pooling several such patterns across similar individuals, we can come to know individuals of a certain type. At still another level, we can look at the patterns of behavior across individuals to see how these patterns fit with the patterns of others to create larger patterns of behavior. It is these larger, inter-individual patterns that constitute social structure. We provide an illustration. In this chapter, we will discuss how people act to verify their conceptions of who they are. A scientist, for example, may act in ways that make it clear to herself, as well as to others, that she is careful, analytical, logical, and experimentally inclined. She may engage in a variety of actions and interactions to convey these images. These are individual patterns of behavior and help us understand the individual scientist. These same patterns of behavior may be part of a larger social structure. We may find, for example, that scientists who are careful, analytical, logical, and experimentally inclined, and who do these things well, are elected to high positions in their scientific organizations. If we take a broader view, we may see that there is a flow of such persons into positions of prominence within their scientific societies, and into positions of eminence in policy and governmental circles. The result is that their pronouncements about being scientists and their activities as scientists help to maintain boundaries between themselves and non-scientists, as well as to keep resources flowing to the groups and organizations to which they belong. The flow of persons into positions of importance through the mechanism of elections and appointments is part of the social structure, as is the flows of resources they control, and the mechanisms that support and sustain these flows. 2

Individuals act, but those actions exist within the context of the full set of patterns of action, interaction, and resource transfers among all persons all of which constitute the structure of society. Social structures do emerge from individual actions, as those actions are patterned across individuals and over time, but individual actions also occur in the context of the social structure within which the individuals exist. In this way, social structure is a very abstract idea. It is not something we experience directly. We are not directly tuned to these patterns as they occur across persons and over time. Nevertheless, we can become aware of them and study them. Many of the patterns are well recognized, named, and attended to. They enter our everyday language as things like General Motors, the New York Yankees, the Brown family, Milwaukee. Some are recognized, but harder to point to, such as the working class or the country club set that do not have a legal status and do not maintain offices or locations. We can only point to individuals who may contribute to the patterns of behavior that constitute the structure. Some structures we tend not to see at all (without special effort or thought) such as the patterns of action that block access of African Americans to the education system or the patterns of actions that create the glass ceiling in organizations preventing qualified women from rising to positions of power and authority. Nevertheless, these too are parts of social structure and it is the job of sociologists to discover, attend to, and understand these patterns. The above implies that the basis for understanding social structure arises from the actions of individuals, keeping in mind that these agents (individuals) receive feedback from the structures they and others create to change themselves and the way they operate. In this chapter, we direct our attention to understanding selves that are producing actions, the patterns of which constitute social structure. However, as sociological social psychologists, we want the reader to keep in mind that persons are always embedded in the very social structure that is, at the same time, being created by those persons. It is this social context, or societal context, that is central in distinguishing sociological approaches to the study of the self.

Self and Identity in Sociology

Self The symbolic interactionist perspective in sociological social psychology sees the self as emerging out of the mind, the mind as arising and developing out of social interaction, and patterned social interaction as forming the basis of social structure (Mead, 1934). The mind is the thinking part of the self. It is covert action in which the organism points out meanings to itself and to others. The ability to point out meanings and to indicate them to others and to itself is made possible by language, which encapsulates meanings in the form of symbols. When ones self is encapsulated as a set of symbols to which one may respond to itself as an object, as it responds to any other symbol, the self has emerged. The hallmark of this process of selfhood is reflexivity. Humans have the ability to reflect back upon themselves, taking themselves as objects. They are able to regard and evaluate themselves, to take account of themselves and plan accordingly to bring about future states, to be self-aware or achieve consciousness with respect to their own existence. In this way, humans are a processual entity. They formulate and reflect, and this is ongoing. To be clear, the responses of the self as an object to itself come from the point of view of others to whom one interacts. By taking the role of the other and seeing ourselves for others perspectives, our responses come to be like others responses, and the meaning of the self becomes a shared meaning. Thus, paradoxically, as the self emerges as a distinct object, there is at the same time a merger of perspectives of the self and others, and a becoming as one with the others with whom one interacts. This becoming as one is possible through the shared meanings of the objects and symbols to which individuals respond in interaction. In using language, individuals communicate the same meanings to themselves as to others. The self is, thus, both individual and social in character. It works to control meanings to sustain itself, but many of those meanings, including the meanings of the self, are shared and form the basis of interaction with others and ultimately social structure. Self-Concept. Over time, as humans point out who they are to themselves and to others, they come to develop a concept/view of who they are. Here, humans are an entity that embodies content and a 4

structure. Sociologists have spent considerable time in understanding the content and structure of the self: ones self-concept. Early views of the self-concept were concerned only with self-evaluation. Selfconcept often meant self-esteem (ones evaluation of oneself in affective (negative or positive) terms)) (cf., Rosenberg, 1979). To broaden this view, Rosenberg (1979) suggested that there was more to the selfconcept than self-esteem. He defined the self-concept as the sum total of our thoughts, feelings, and imaginations as to who we are. Later conceptions elaborated and refined this view suggesting that the self-concept was made up of cognitive components (given the collection of identities) as well as affective components or self-feelings including self-esteem (both worth-based and efficacy-based self-esteem) (Franks & Marolla, 1976; Stryker, 1980). In general, the self-concept is the set of meanings we hold for ourselves when we look at ourselves. It is based on our observations of ourselves, our inferences about who we are, based on how others act toward us, our wishes and desires, and our evaluations of ourselves. The self-concept includes not only our idealized views of who we are that are relatively unchanging, but also our self-image or working copy of our self-views that we import into situations and that is subject to constant change and revision based on situational influences (Burke, 1980). It is this self-image that guides moment-to-moment interaction, is changed in situated negotiation, and may act back on the more fundamental self-views. For sociological social psychologists, the self-concept emerges out of the reflected appraisal process (Gecas & Burke, 1995). Although some of our self-views are gained by direct experience with our environment, most of what we know about ourselves is derived from others. According to the reflected appraisal process, which is based on the looking glass self (Cooley, 1902),1 significant others communicate their appraisals of us, and this influences the way we see ourselves. In a now classic review of studies on the reflected appraisal process, Shrauger and Schoeneman (1979) found that rather than our self-concepts resembling the way others actually see us, our self-concepts are filtered through our perceptions and resemble how we think others see us. Felson (1993) summarizes a program of research in which he has attempted to explain why individuals are not very accurate in judging what others think of them. Among the causes of the 5

discrepancy is the apprehension of others to reveal their views. At best they may reveal primarily favorable views rather than both favorable and unfavorable views. Consistent with other research (DePaulo, Kenny, Hoover, Webb, & Oliver, 1987; Kenney & Albright, 1987), Felson finds that individuals have a better idea of how groups see them than how specific individuals see them. Presumably, individuals learn the group standards and then apply those standards. In turn, when group members judge individuals, they use the same standards that individuals originally applied to themselves. Thus we find a correspondence in self-appraisals and others appraisals of the self. In our investigation of the reflected appraisal process with newly married couples, we find that social status derived from ones position in the social structure also influences the appraisal process. The spouse with the higher status (education, occupation, and income) in the marriage is more likely to not only influence their partners self-views, but also their partners views of them (Cast, Stets, & Burke, 1999). Spouses with a lower status in the marriage have less influence on the self-view of their higher status counterparts or on how their higher-status counterparts view them. Self-Evaluation. The aspect of the self-concept that has received a significant amount of attention in sociological social psychology is the evaluative part of the self-concept, better known as self-esteem (Rosenberg, 1979). Two dimensions of self-esteem have been identified: efficacy-based self-esteem (seeing oneself as competent and capable) and worth-based self-esteem (feeling that one is accepted and valued) (Gecas & Schwalbe, 1983). Others have labeled the distinction inner self-esteem (being effective) and outer self-esteem (acceptance by others) (Franks & Marolla, 1976). As Gecas and Burke (1995) point out, the significant interest in self-esteem is largely due to assuming that high self-esteem is associated with good outcomes such as personal success while low self-esteem is associated with bad outcomes such as deviance. While these associations are a bit misleading since research does not always show such consistency in these outcomes, part of the inconsistency may be rooted, among other things, in measuring self-esteem in global terms rather than more specific terms (Hoelter, 1986; Rosenberg, Schooler, Schoenbach, & Rosenberg, 1995).2 Nevertheless, self-esteem remains a high profile topic of investigation and has been examined from a variety of different viewpoints: as an outcome (Rosenberg, 6

1979), as a buffer against stress (Longmore & DeMaris, 1997), and as a motive that directs behavior (Kaplan, 1975; Tesser, 1988). Cast and Burke (1999) use identity theory as a theoretical framework for the integration of these different conceptualizations of self-esteem. They argue that self-esteem is intimately tied to the identity verification process. 3 They point out that: 1) high self-esteem has been found as an outcome of the identity verification process (Burke & Stets, 1999), 2) high self-esteem that is generated from the identity verification process can act as a buffer or resource when the verification process fails, and 3) the desire for self-esteem may be what motivates people to create and maintain situations or relationships that verify ones identity. They also argue that the two components of self-esteem (worth-based and efficacy-based) are each rooted primarily in the different bases of identities. They argue that verification of group-based identities has a stronger impact on worth-based self-esteem while verification of role-based identities has a stronger impact on efficacy-based self-esteem. Analyzing data from a sample of newly married couples, their results support the integration of the different viewpoints on self-esteem into identity theory. If (worth-based) self-esteem is a source of motivation, so too is self-efficacy (Bandura, 1982). Selfefficacy is seeing oneself as a causal agent in ones life. As Bandura (1995) points out, efficacy is a belief about ones causative capabilities. Whether one actually has control, objectively, is less relevant than what one perceives to be the case. Like self-esteem, positive outcomes have been associated with high self-efficacy such as effectively coping with lifes stresses and adopting good health habits (Bandura, 1995). Our own research finds that identity verification not only enhances feelings of self-worth as noted above, but also feelings of control over ones environment (Burke & Stets, 1999). Some have also recently linked self-esteem with efficacy by arguing that people with high self-esteem should also tend to perceive themselves as competent and, in turn, exhibit more involvement in social movements to try to effect social change (Owens & Aronson, 2000). Identity Because the self emerges in social interaction within the context of a complex, organized, differentiated society, it has been argued that the self must be complex, organized and differentiated as 7

well, reflecting the dictum that the self reflects society (Stryker, 1980). This idea is rooted in James (1890) notion that there are as many different selves as there are different positions that one holds in society and thus different groups who respond to the self. This is where identity enters into the overall self. The overall self is organized into multiple parts (identities), each of which is tied to aspects of the social structure. One has an identity, an internalized positional designation (Stryker, 1980, p. 60), for each of the different positions or role relationships the person holds in society. Thus, self as father is an identity, as is self as colleague, self as friend, and self as any of the other myriad of possibilities corresponding to the various roles one may play. The identities are the meanings one has as a group member, as a role-holder, or as a person. What does it mean to be a father, or a colleague, or a friend? These meanings are the content of the identities. Most interaction is between persons who occupy positions (statuses) in groups or organizations in society. Interaction is thus not between whole persons, but between aspects of persons having to do with their roles and memberships in particular groups or organizations: their identities. As a parent, we talk with our children. As a spouse, we talk to our partner. As a member of an organization, we talk to our employer. An assumption and implication of the above is that any identity is always related to a corresponding counter-identity (Burke, 1980). When one claims an identity in an interaction with others, there is an alternative identity claimed by another to which it is related. The husband identity is enacted as it relates to the wife identity, the teacher identity is played out in relation to the student identity and so forth. In each of these cases, there are things that are not talked about because they are not relevant to that identity, and there are things that are more likely to be talked about given the identity that is currently being claimed. There are various styles of interaction that are appropriate in each situation for each identity. We move into and out of these modalities very easily, and generally with very little thought. Often we operate in two or more identities at a time as in being both a friend and colleague. In examining the nature of interaction between identities of different persons, we can take two different perspectives: agency and social structure. In terms of social structure, we can focus on the external and talk about actors taking a role or playing a role. Here, the social structure in which the 8

identities are embedded is relatively fixed and people play out the roles that are given to them. Teachers do the things that teachers are supposed to do. Variations across persons taking on the same identities are viewed as relatively minor, except insofar as they impact the success (or failure) of a group or organization. Essentially, the social structure persists and develops according to its own principles; individuals are recruited into positions and individuals leave positions, but for the most part the positions remain. But there is also agency. As agents, individuals can make or create a role by making behavioral choices and decisions and engaging in negotiation and compromise as well as conflict. Research finds that making roles and accumulating role identities fosters greater psychological well-being (Thoits, 2001). Furthermore, Thoits finds that the reverse is also true: greater psychological well-being allows individuals to actively acquire multiple role identities over time, particularly voluntary role identities such as neighbor and churchgoer. When individuals feel good about themselves they take on more identities. In general, therefore, examining the nature of interaction between identities means addressing both social structure and agency. We must go back and forth and understand how social structure is the accomplishment of actors, but also how actors always act within the social structure they create. Identity Theory As Stryker (2000) points out, there are multiple views of identity within sociology. Some have a cultural or collective view of identity in which the concept represent the ideas, belief, and practices of a group or collective. This view of identity is often seen in work on ethnic identity, although identity is often not defined, thus obscuring what is gained by using the concept (e.g., Nagel, 1995; Scheff, 1994). This view lacks the ability to examine individual variability in behavior, motivation, and interaction. Another view, growing out of the work of Tajfel (1981) and others (e.g., Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) on social identity theory, sees identity as embedded in a social group or category. This view often collapses the group/category distinction and misses the importance of within group behavior such as role relationships among group members. A third view of identities grows out of the symbolic interactionist tradition, especially its structural variant (Stryker, 1980). This view takes into 9

account individual role relationships and identity variability, motivation, and differentiation. It is this work that we present and elaborate in this chapter. In addition, as we have argued elsewhere, social identity theory may be seen as a special case of this variant of identity theory (Stets & Burke, 2000). What the following views of identity theory have in common is a general set of principles that Stryker (1980) has enumerated as underlying the structural symbolic interaction perspective. These include: 1) that behavior is dependent upon a named or classified world and that these names carry meaning in the form of shared responses and behavioral expectations that grow out of social interaction. 2) That among the named classes are symbols that are used to designate positions in the social structure. 3) That persons who act in the context of social structure name one another in the sense of recognizing one another as occupants of positions and come to have expectations for those others. 4) That persons acting in the context of social structure also name themselves and create internalized meanings and expectations with regard to their own behavior. 5) That these expectations and meanings form the guiding basis for social behavior and along with the probing interchanges among actors shape and reshape the content of interaction, as well as the categories, names and meanings that are used. Negotiated meaning emerging from social interaction is the shared component in these views of identity theory Identity theory that grows out of structural symbolic interaction currently has two slightly different emphases (Stryker & Burke, 2000). In the work of Stryker and his colleagues (Serpe & Stryker, 1987; Stryker & Serpe, 1982, 1994), the focus is on how social structure influences one's identity, and in turn, behavior. In the work of Burke and his associates (Burke & Cast, 1997, 1999; Burke & Reitzes, 1981, 1991; Burke & Stets, 1999; Riley & Burke, 1995; Stets & Burke, 1996, 2000; Tsushima & Burke, 1999), the emphasis is on the internal dynamics within the self that influence behavior. Very similar to this version of identity theory is affect control theory, developed by Heise and his colleagues (Heise, 1979; MacKinnon, 1994; Smith-Lovin, 1987) that also focuses on the internal dynamics, but draws more heavily on the shared cultural meanings of identities as opposed to individual, subcultural, or group meanings. A third form of identity theory comes in the work of McCall and Simmons (1978). Though there hasnt been a clear program of research coming out of this version of identity theory, it does make 10

some important theoretical contributions to understanding identities that are important to review. We shall discuss all of these perspectives. To begin, we emphasize that the core of an identity is the categorization of the self as an occupant of a role, and incorporating, into the self, the meanings and expectations associated with the role and its performance (Stets & Burke, 2000). Sociological social psychologists see persons as always acting within the context of social structure in which others and themselves are labeled in that each recognizes the other as an occupant of positions or roles in society (Stryker, 1980). Thus, one assumes a role identity, thereby merging the role with the person (Turner, 1978). McCall and Simmons McCall and Simmons (1978, p. 65) define a role identity as the character and the role [our emphasis] that an individual devises for himself as an occupant of a particular social position. This follows the conception of Turner (1962) that criticizes the Lintonian (1936) role-theoretic character of social roles as too rigid and not allowing for the individual variability and negotiation that exists. McCall and Simmons (1978, p. 68) indicate that a role identity has a conventional dimension and an idiosyncratic dimension. The former is the role of role identity that relates to the expectations tied to social positions while the identity of role identity relates to the unique interpretations individuals bring to their roles. McCall and Simmons point out that the proportion of conventional versus idiosyncratic behavior tied to role identities varies across people and across identities for any one person. Since the self typically has multiple role identities, McCall and Simmons see the many different role identities as organized in a hierarchy of prominence. This organization reflects a persons ideal self (McCall & Simmons, 1978, p. 74). The prominence of an identity depends upon the degree to which one: 1) gets support from others for an identity, 2) is committed to the identity, and 3) receives extrinsic and intrinsic rewards from the role identity. The more prominent the role identity, the more likely it will be activated and performed in a situation. For successful enactment of a role in a situation, McCall and Simmons highlight the importance of negotiation with others in the situation. Enacting a role identity is always done in relation to a 11

corresponding counter-identity in the interaction, for example, husband to wife. Ones expectations associated with a role identity, whether they are conventional or personal, may differ from the expectations others associate for that role identity in the situation. Each party is trying to enact a role that meshes with the other; each has self-conceptions (their identity) as well as conceptions of the other (the others identity). This implies some degree of coordination and compromise between individuals so that smooth role performances can be achieved. Research on the leadership role identity evidences this negotiation. When individuals cannot negotiate leadership performances in a group that match their leadership identities, they become less satisfied with their role and are less inclined to remain in the group (Riley & Burke, 1995). Alternatively, when they can negotiate a leadership performance consistent with their identity, they become more satisfied and more inclined to remain in the group. Other work shows that when different but interrelated and complementary role behaviors are negotiated by role partners, a strong attachment to the group develops (Burke & Stets, 1999). Stryker Although developed independently, Stryker (1980) suggests a view of identities that is somewhat similar to that of McCall and Simmons (1978). He sees the many role identities that a person may have as organized in a hierarchy, but it is a salience hierarchy rather than a prominence hierarchy. A salient identity is an identity that is likely to be played out (activated) frequently across different situations. While the prominence hierarchy of McCall and Simmons addresses what an individual values, the salience hierarchy focuses on how an individual will likely behave in a situation. What one values may or may not be related to how one behaves in a situation although there is a significant relationship between the two (Stryker and Serpe 1994). However, there may be times when what one values may not be able to be expressed in a situation given situational constraints, so Stryker and Serpe argue that identity prominence and identity salience should be kept as distinct concepts (Stryker & Serpe, 1994). What importantly influences the salience of an identity is the degree of commitment one has to the identity. Commitment has two dimensions: a quantitative and qualitative aspect (Stryker & Serpe, 1982, 12

1994). In the former, reflecting the individuals ties to the social structure, commitment reflects the number of persons that one is tied to through an identity. The greater the number of persons to whom one is connected through having a particular identity, the greater is the commitment to that identity. With respect to the qualitative dimension of commitment, the stronger or the deeper the ties to others based on a particular identity, the higher the commitment to that identity.4 Stryker (1968, 1980) suggests that the greater the commitment to an identity, the higher will be the identity in the salience hierarchy. Once again, the relevance of social structure in understanding the self is made clear. Because people live their lives in social relationships, commitment takes these ties into account when explaining which identities persons are likely to invoke in a situation. For example, if a mans social network in terms of the number of others and the importance of those others is largely based on him occupying a particular role, such as father, then the father identity is likely to be invoked across various different situations. Research strongly supports the link between commitment, identity salience, and behavior consistent with salient identities. For example, Stryker and Serpe (1982) examine the religious role identity. Their six-item commitment scale measures the extensiveness and intensiveness of relations with others in life based on being in the religious role. For example, In thinking of the people who are important to you, how many would you lose contact with if you did not do the religious activities you do? (Extensiveness), or Of the people you know through your religious activities, how many are close friends? (Intensiveness). The salience of the religious identity is measured by asking respondents to rank the religious role in relation to other roles they may assume such as parent, spouse, and worker. Their measure of behavior is time in the religious role. Respondents are asked how many hours in an average week they spend doing things related to religious activities. Stryker and Serpe find that those committed to relationships based on religion have more salient religious identities that are associated with more time spent in religious activities. In another study, Callero (1985) examines the blood-donor role identity. In separate measures of the salience of the blood-donor role identity (in relation to other identities one might claim), commitment to the blood-donor identity (borrowing Stryker and Serpes 1982 commitment scale), and behavioral 13

measure of the identity (number of blood donations given in a six-month period), Callero reaches similar conclusions to that of Stryker and Serpe. The more one has relationships premised on the blood-donor identity, the higher the blood-donor role identity is in ones identity salience hierarchy, and the more this salient role identity is related to donating blood. Multiple Identities. The image of a hierarchy of identities, used by both McCall and Simmons and Stryker, highlights the fact that individuals have multiple role identities (which are ranked). And, this idea of multiple identities highlights the fact that individuals are always acting in the context of a complex social structure out of which these multiple identities emerge. Having multiple role identities may be good for the self. Indeed, self-complexity theory shows that more complex selves are better buffered from situational stresses (Linville 1985, 1987). Consistent with this, Thoits (1983, 1986) has shown that having multiple role identities is more beneficial than harmful to individuals because it gives their lives meaning and provides guides to behavior. Other studies have shown that the more one accumulates different role identities, the more positive its effects on mental health (see Thoits 2001 for a review). The positive effect of multiple identities on mental health may be contingent upon the kind of role identities being invoked. Thoits (1992) finds that obligatory role identities such as the parent identity, spouse identity, or worker identity are beneficial to mental health only when chronic strains in each role are low. Alternatively, voluntary role identities such as friend or neighbor significantly reduce psychological distress because they are less demanding physically and psychologically and they are easier to exit when their costs exceed their rewards. More attention is being given to understanding the development of multiple role identities and their outcomes for individual behavior. Smith-Lovin (2001) generates a series of predictions, derived mostly from social ecology and network theory, to explain why some individuals develop multiple identities and thus more complex selves than others. For example, she argues that the larger ones network of others, and the less homophilous (similar) they are, the more complex the self. Higher status actors will also have more complex selves than lower status actors because they are likely to have more diverse networks that range further through the social system. In general, Smith-Lovin draws our attention to the social 14

structures in which individuals are embedded because they influence the complexity of the self that is formed. Burke (2001) also examines how multiple identities are related to the social structure but in a slightly different fashion. He examines how the same two identities play themselves out differently for persons who are in two structurally different locations in a group. Specifically, he compares individuals who occupy the coordinator role in a group with those who do not occupy this position. He finds that those persons who hold the coordinator role in a group have task and social-emotional identity performances that are highly and positively correlated while those who do not hold the position of coordinator have task and social-emotional identity performances that are relatively independent. This suggests that the multiple identities that one holds may come to share meanings in response to the structural conditions in which the identities are played out again emphasizing the effect of structural location on identity processes. Burke If the work of Stryker focuses on the arrangement of identities and how they relate to social structure, Burkes work has focused more on the internal dynamics that operate for any one identity. In early work (Burke, 1980; Burke & Tully, 1977), it was argued that identity and behavior are linked through a common system of meaning. In order to predict how one behaves, we have to identify the meanings of the role identity for the individual. Drawing upon the conceptualization of meaning developed by Osgood, Suci, and Tannenbaum (1957), Burke and Tully (1977) present a method for the measurement of the selfmeanings of a role identity. Burke (1980) suggests that a person learns the meanings of a role identity in interaction with others in which others act toward the self as if the person had an identity appropriate to their role behavior. Thus, ones role identity acquires meaning through the reactions of others (Burke, 1980). This is not to say that persons do not import some of their own understandings into their role identities that may be different from others understandings. These differences are worked out through the negotiation process in interaction. Role identities generally contain a set of multiple meanings (Burke and Tully 1977). For example, the male role identity for John may mean being independent, competitive, and self-confident. 15

Additionally, different individuals may have different meanings for the same role identity. For example, a student may see him or herself as, and be seen by others as academic, if he or she regularly attends class, takes notes, passes exams, and finishes courses (Burke and Reitzes, 1981). Another student, however, may see themselves as sociable rather than academic if he or she finds opportunities to have fun with peers while at school: having friends over and going to parties. More generally, it is the meanings of identities have implications for how one behaves (with respect to the meanings of the behavior) and one's behavior confirms one's identity when they share meanings. More recent conceptions of identity expand on the notion of a correspondence of meaning between identity and behavior and incorporate the idea of a perceptual control system, a cybernetic model, based on the work of Powers (1973). This is where the internal dynamics of identities are most clearly seen (Burke, 1991, 1996; Burke & Reitzes, 1991; Riley & Burke, 1995). Since an identity is a set of meanings attached to the self in a social role, this set of meanings serves as a standard or reference for a person. When an identity is activated in a situation, a feedback loop is established. This loop has four components: 1) the standard (the self-meanings), 2) a perceptual input of self-relevant meanings from the situation including how one sees oneself (meaningful feedback in the form of reflected appraisals), 3) a process that compares the perceptual input with the standard (the comparator), and 4) output to the environment (meaningful behavior) that is a result of the comparison (difference) of perceptions of selfmeanings with actual self-meanings held in the standard. The system works by modifying outputs (behavior) to the social situation in attempts to change the input to match the internal standard. In this sense, the identity control system can be thought of as having a goal, that is, matching the situational inputs (perceptions) to the internal standards. What this system attempts to control is the perceptual input (to match the standard). What is important about the cybernetic model of the identity process is that instead of seeing behavior as strictly guided by the situation or strictly guided by internal self-meanings, behavior is seen to be the result of the relation between the two. It is goal-directed in that there is an attempt to change the situation in order to match perceived situational meanings with meanings held in the identity standard, that is, to 16

bring about in the situation the meanings that are held in the standard. Thus the model has the interesting implication of making different predictions about behavior from the same identity meanings depending upon the (perceived) situation. When self-meanings in the situation match self-meanings in the identity, the meanings of the behaviors correspond to these meanings. However, if the self-meanings perceived in the situation fail to match, behavior is altered to counteract the situational meanings and restore perceptions. Thus, for example, if one views herself as strong, and sees that others agree, she will continue to act as she has (strongly). But, if she sees that others appear to view her as weak, she will increase the strength of her performance in an effort to restore perceptions of herself as strong as reflected in the reflected appraisals.5 Identities are tied to social structure in the sense that the meaningful behavior outputs of an identity are role bahviors. Role behaviors are a means by which one strives to keep perceptions of self-relevant meanings in the situation in line with the meanings held in the identity standard (in other words, one strives for self-verification).6 Role behaviors are accomplished through interaction with others whose behavior is an output of their own identity processes that also strive for self-verification (Riley & Burke, 1995). All participants in the interactive setting mutually accomplish their respective self-verifications (if all goes well). Because each is motivated to match self-relevant meanings in the situation with selfmeanings in their respective identity standards, and because the actions of each change or disturb the meanings in the situation, self-verification is only accomplished by the cooperative and mutually agreed upon arrangement of role performances. However, this does not happen automatically. Performances are stretched, identity standards are altered,7 and negotiations are conducted as the participants seek ways to accomplish self-verification with at worst not disturbing the verification of others, and at best helping them to verify themselves. When congruity between reflected appraisals and the identity standard occurs, ties to role partners are strengthened. More recently, we have argued that commitment results from the self-verifying aspect of the identity process (Burke & Stets, 1999). In studying married couples, we show that the more the spousal roles of both partners are verified, the more it leads to the development of committed relationships, high levels of trust, and a perceived collective (group). 17

The cybernetic character of identity theory has led to a view of the nature of commitment that is slightly different from the view outlined earlier by Stryker. In this slightly different view, commitment to an identity is the sum total of the pressure to keep perceptions of self-relevant meanings in the situation in line with the self-meanings held in the identity standard (Burke & Reitzes, 1991). One is more committed to an identity when one strives harder to maintain a match between self-in-situation (perceptual input) meaning and the meaning held in the identity standard. Commitment thus moderates the link between identity and behavior making it stronger (high commitment) or weaker (low commitment). This does not negate the importance of the structural side shown in ties to role partners (Stryker & Serpe, 1982, 1994), but shows how those ties as well as other factors, such as rewards and praise one might receive for being in the role, bring about commitment as defined by Burke and Reitzes (1991) in terms of the strength of the self-verification response. The structural connection is maintained. For example, Burke and Reitzes (1991) show that those who are highly committed to a student identity (by having more ties to others, as well as receiving rewards for having the identity) have a stronger link between identity meanings (for example academic responsibility) and behavior meanings (for example, time in the student role, or grade point average) than those with lower levels of commitment. An extension that has been made to the identity model in identity theory concerns the nature of the meanings that are encapsulated in the identity standard. Originally thought of in terms of symbols in the tradition of symbolic interaction (cf. Stryker, 1980), Freese and Burke (1994) extend the notion of meaning to include not only symbols (shared meanings) but also signs, drawing from earlier work in symbolic interaction that had not been fully developed (Lindesmith & Strauss, 1956). Signs are signals (stimuli) that convey meanings through which individuals relate directly to their environment and all of the things in the situation insofar as they are used, transferred, or transformed: clothing, food, objects, air and so forth. There may be, in addition, symbolic meanings attached to the objects, for example a very expensive fountain pen may convey wealth and prestige, but as a writing implement that is manipulated to put ink on paper, it is also simply an object in the environment. In this way, Freese and Burke (1994) introduce the idea of the control of resources (though sign meanings), an idea that is essential to sociology 18

in understanding social structure. Thus, role performances are not just symbolic interactions, but also sign interactions. Persons manipulate signs (resources) and symbols in the situation to bring sign and symbolic meanings to match the sign and symbolic meanings held in their role-identities. By using this expanded model of identities, Burke (1997) was able to simulate the exchange of resources studied in network exchange theory to arrive at very close predictions to the final distribution of resources and power that were observed in laboratory experiments. The perceptual control system as it is applied to identities is a self-regulating system and is very similar to Carver and Sheiers (1981, 1998) theory of self-regulated behavior. In this way, sociologists and psychologists are thinking along similar lines with respect to understanding the self. We see that when the meanings of the self in the situation (based in part on feedback from others and in part on direct perception of the environment) are congruent with the meanings held in the identity standard, selfverification has occurred (Burke & Stets, 1999). This is also very similar to Swanns (1983, 1990) formulation of self-verification. We agree with Swann that people seek to verify their self-views in interaction, even if those self-views happen to be negative. Once again, sociologists and psychologists are thinking along similar lines in explaining the self. A Variant of Identity Theory: Affect Control Theory Affect control theory independently developed by Heise and his colleagues (Heise, 1979; MacKinnon, 1994; Smith-Lovin, 1987) is very similar to the cybernetic model of the identity process. Affect control theory also views identities as containing self-meanings, with a focus on the fundamental dimensions of meanings identified by Osgood, Suci, and Tannenbaum (1957) of evaluation, potency, and activity (EPA). These self-views are the fundamental sentiments persons hold about themselves in a social role (like the identity standard of Burke). When events in the situation disturb the perceived selfmeanings (called transients) so that they no longer match the fundamental sentiments, individuals act to create new events that restore the transients toward the fundamental sentiments. While these broad outlines show a high level of similarity between affect control theory and identity theory, there are some differences, and the two theories have each pursued slightly different questions. 19

One difference is that affect control theory uses culturally defined (shared) views of what an identity means, while identity theory has not confined itself to necessarily shared meanings, but has focused on the self-definition of self-meanings assessed along culturally chard dimensions (Burke and Tully, 1977). Identity theorists recognize that persons meanings tied to a role may, in part, be idiosyncratic to those persons. Another difference is that affect control theory has considered only the EPA dimensions of meaning for defining all identities (as well as behaviors), thus allowing direct comparisons among different identities. Identity theory, on the other hand, has tried to find the dimensions of meaning most relevant to the occupants of the positions, though these may vary from one role identity to another making direct comparisons more difficult. Emotion in Identity Theory In the identity cybernetic model, any discrepancy/incongruity between perceived self-in-situation meanings and identity standard meanings is signaled in the comparator. This discrepancy reflects a problem in verifying the self, and as a result of this the individual experiences negative emotional arousal such as depression and distress (Burke, 1991; Burke & Stets, 1999), anger (Bartels, 1997), and hostility (Cast & Burke, 1999). The absence of an error or discrepancy is self-verification and results in positive emotional arousal such as high self-esteem and mastery (Burke & Stets, 1999; Cast & Burke, 1999). The role of emotion in the identity control model is consistent with earlier arguments made by identity theorists on the relationship between identity and emotion. For example, McCall and Simmons (1978) argued that if a prominent identity has been threatened (by others not supporting ones role performance), an individual would experience a negative emotional response. Consistent with this, Ellestad and Stets (1998) report that when nurturing behavior is linked to fathering rather than mothering, women whose mother identity is prominent report the negative emotion of jealousy. Stryker (1987) discusses how emotion and identities are related given the salience hierarchy. He argues that identities that generate positive feelings should be played out more often and move up in the salience hierarchy while identities that repeatedly cause negative feelings should be less likely to be played out and move down in the salience hierarchy. He also argues that identities that are inadequately 20

played out should generate negative feelings because poor role performance results in others not supporting who ones identity claims. Therefore, identity theorists (McCall and Simmons, Stryker, and Burke) agree that negative emotion results from not meeting ones identity expectations and positive emotion results from meeting ones identity expectations. In a similar fashion, affect control theory (Heise, 1979; Smith-Lovin, 1995; Smith-Lovin & Heise, 1988) has posited emotional responses to the discrepancy between the self-meanings in the identity standard (fundamentals) and perceptions of self-relevant meanings in the situation (transients). Not only are particular emotions signaled in response to discrepancies of particular meanings, but the display of emotion is itself an event that changes the meanings in the situation that can move transients back toward the fundamental sentiments. For example, Heise (1989) suggests that the negative implications of a deviant act can be forestalled by the appropriate display of shame by the perpetrator. One difference between affect control theory and identity theory with respect to emotions is that affect control theorists argue that a discrepancy can generate positive emotion when one exceeds the expectations tied to identities in a situation (MacKinnon, 1994). Feelings vary in terms of their strength or intensity. Stryker (1987) argues that the strength of the emotional response to identity-related behaviors in situations should signal to individuals how important an identity is in their salience hierarchy with more important identities producing a stronger emotion. This parallels McCall and Simmons point as to the role of emotions in the prominence hierarchy. Burke (1991, 1996) hypothesizes that repeated interruptions in the self-regulating identity process cause more negative emotion than occasional or infrequent interruptions. Stets (2001) examines more closely the identity assumptions that: 1) a discrepancy leads to negative emotion, and 2) frequent interruptions in the identity process cause more intense or stronger negative emotion. These two assumptions are actually similar to Higgins (1989) self-discrepancy theory. For Higgins, negative emotion results from a discrepancy between ones actual state and ones ideal state. The negative emotion becomes more intense and frequent as the magnitude of the discrepancy increases. Stets examines the identity theory assumptions by studying the distributive justice process and individuals 21

emotional responses to injustice in a laboratory setting that simulates a work situation. If we translate the idea of disruption of the self-verification process into being over-rewarded or under-rewarded in a justice situation (in either case, ones standard is not being met), negative emotion should result, and more intense negative emotions should occur as the inequitable distributive process persists. Stets findings show that an identity discrepancy does not always lead to negative emotion. When one is over-rewarded, positive emotion results. Stets argues that when individuals receive rewards (goods) rather than punishments (bads), their standard quickly adjusts in a positive direction to the new level if the over-reward is relatively small. Making this adjustment has two consequences: the self is enhanced and any discrepancy between ones standard and ones perceptions (reward) is removed. The degree of exceeding the standard is important since an outcome that significantly exceeds ones standard in a positive direction may lead to negative emotion because the size of the discrepancy makes it too difficult to self-verify. This resolves the self-enhancement - self-verification debate in the literature in a different fashion than that offered by Swann and his colleagues (Swann 1990; Swann, Griffin, Predmore, & Gaines 1987). Swann and his associates argued that self-enhancement was dependent upon immediate affective reactions to social feedback while self-verification was dependent upon less immediate cognitive reactions to social feedback. Stets argues that what may be more important is the degree of disparity in meanings held in the input and identity standard. Self-enhancement may be more highly activated when a small discrepancy occurs in a positive direction. Self-verification may be more highly activated when a large discrepancy exists in a positive direction. Thus, rather than assuming that any discrepancy produces negative emotion because the information is not self-verifying as is assumed in identity theory, it may depend upon the size of the error registered in the comparator. Stets also finds that as the inequitable distributive process is repeatedly experienced by subjects, their emotions become less, not more, intense. This pattern again suggests that individuals are changing their standards, adjusting their standards to the level of rewards they received. Since a strong emotional response would signal a discrepancy between input meanings and standard meanings, a weaker emotional 22

response over time would suggest a closer correspondence between input and identity standard meanings. In general, these unexpected findings have implications for modifying assumptions in identity theory about the relationship between identity expectations, emotion, and the repetitiveness of disrupting the identity process. Another way in which the strength or intensity of an emotional response to identity disconfirmation has been examined is in the recent work of Stets and Tsushima (2001). Using a national probability sample, Stets and Tsushima find that more intense anger is associated with the lack of verification of group-based identities that are intimate such as the family identity that meet our need to feel valuable, worthy, and accepted. Less intense anger is associated with role-based identities that are less intimate such as the worker identity that fulfills our need to feel competent and effective. Burke (1991) has argued that greater distress will be felt by the individual when the self-verification process is interrupted by a significant other than a casual acquaintance. The fact that Stets and Tsushima find intense angry feelings in the family identity compared to the worker identity is consistent with this thesis. Group-based identities that are socio-emotional have strong ties that make others views about the self important. If the self is not verified, the emotional response can be powerful. According to identity theory, when negative emotion is felt, actors may either change what they are doing (the output end of the model), or they may think about the situation in a different way (the input side) in order to achieve greater congruence (Burke, 1991). In later work, Burke (1996) refers to these responses as different coping responses. One can modify the situation through some behavioral strategy or modify the meaning of the problem through some cognitive strategy.8 Ellestad and Stets (1998) reveal that the more salient the identity, the more likely it is that persons devise behavioral strategies to reassert their identity role, thereby maintaining who they are to themselves and significant others. More recent work finds that disruption of self-verification for group-based identities that are more intimate such as the family identity leads to coping strategies that are cognitive (activity on the input side of the identity model) while disruption of self-verification for role-based identities that are less intimate such as the

23

worker identity leads to behavioral strategies of coping (activity on the output side of the identity model) (Stets & Tsushima, 2001). The Hierarchy of Identity Control Systems and Identity Change in Identity Theory The identity control model has been further extended so that a particular identity standard is viewed as the output of a higher-level perceptual control process, thus embedding the identity control process within a hierarchical control structure (Burke, 1997; Burke & Cast, 1997; Powers 1973; Tsushima & Burke, 1999). For example, recent research has discussed the relationship between principle-level identity standards at a higher level of control and program-level identity standards at a lower level of control (Tsushima & Burke, 1999). Principle-level standards are abstract goal states such as values, beliefs, and ideals. Program-level standards are more concrete goals that are accomplished in situations. When the distinction of these two levels is applied to the parent identity, evidence reveals that some parents are more principle-oriented (e.g., desire their child to be a critical thinker, loving, autonomous) and some parents are more program-oriented (e.g., are concerned that their child makes her bed and gets to school). Those parents with more fully developed principle-level components of the parent identity are able to relate the principle-level to the program-level to alter programs so that the programs not only accomplish the mundane goals (getting the child to complete her homework) but also accomplish the higher-level goals (being independent). As a consequence, these parents experience higher efficacy and lower stress (Tsushima & Burke, 1999). If we extend our analysis beyond one identity and consider multiple identities, the hierarchical model is a useful way of understanding the relationship among multiple identities (Burke 2001). If we think of an identity as the set of all meanings held for oneself in terms of, for example, a particular role, then an identity standard might be thought of as a set or vector of meanings. Strictly, each meaning is part of a separate control system, but conceptually it is easier to think of the set or vector of meanings of an identity as part of a single control system. Thus, when multiple identities are enacted in a situation, separate control systems have been activated, each of which is acting to control self-in-situation meanings

24

to match an identity standard. Since control systems can be arranged hierarchically, higher, more abstract identities as well as lower, less abstract identities may be activated in a situation. This hierarchical model also helps us understand how identities change. Identity standards of lowerlevel control systems are the outputs of higher-level control systems. In other words, when a higher-level control system behaves, it provides the reference standard to the control systems just below it. When a higher-level system brings the higher-level perceptions into alignment with the higher-level standard, it does so by changing its outputs thereby changing lower-level standards. In this way the meanings contained in lower-level standards are altered (Burke & Cast, 1997). Further, because the overall perceptual control system is continuously operating to verify identity perceptions at all levels for identities that are activated, identity change is always going on, though at a much slower pace than behavior that alters the situation. Nevertheless, when actions cannot change the meanings in the situation to verify an identity, the identity standard itself will change toward the meanings in the situation.9 For example, Burke and Cast (1997) show that the birth of a child to a newly married couple provides a new set of meanings in the situation that is difficult to change. The consequence of this is that the gender identities of the husband and wife both change. Husbands become more masculine in their self-views while their wives become more feminine. Identity change has also been examined by Kiecolt (1994). She argues that a change occurs when a stressor such as chronic role strain or a life event disrupts valued role-identities, and among other things, people believe they can change, they see that the benefits of self-change outweigh the costs, and others provide support for their self-change. More recently, Kiecolt (2000) argues that involvement in social movements can result in change by changing ones salience hierarchy of identities. This can be done in three ways: 1) either adding or discarding an identity, 2) changing the importance of an identity without changing the ranking of the identity (for example, the activist identity can become more important as one becomes more involved in a social movement, but its importance relative to other identities does not change), or 2) changing the importance and ranking of an identity. One could also change the meanings of an identity. Consistent with the idea that higher levels of the perceptual control system change more 25

slowly, Kiecolt indicates that if social movement participation results in self-concept change, the change is gradual, not sudden. Most recently, Burke (forthcoming) has added an additional twist to identity change. Examining the spousal identities and behaviors of newly married couples over a three-year period, he shows how both the spouse identity and the performance of the spouse role change in response to marital interaction. The spousal identities were measured in terms of the extent to which persons thought they should do various activities associated with the spousal role such as cleaning the house, yard work, and maintaining contact with parents and in-laws. Role performance was measured by daily diary indication of the extent to which they reported they did these things over a four-week period, collected at a later point. Burke found that when the spouse identity was not verified (because their actual role performance was either greater than or less than would be expected given their self-views), the actual role performance shifted over time toward a closer correspondence with the standard contained in their spousal identity. At the same time, however, the spousal identity standard changed (though more slowly) toward the levels reflected in the actual performance in accordance with the theoretical expectations. Thus, we see that behavior adjusts to conform to the meanings contained in the identity standard, while the identity standard also slowly shifts over time to conform to the meanings of the behavior. Identity Theory as a Theory The predictive power of identity theory can be tested against alternative theories. In doing this, we are in a better position to identify the scope of identity theory, that is, where it does and does not apply, and the ways that the theory can be extended. For example, in a series of studies (Stets, 1997; Stets & Burke, 1996), we examine positive and negative emotion-based behavior among newly married couples. We compare predictions derived from expectation states theory (Ridgeway & Walker, 1995; Webster & Foschi, 1988) with those of identity theory. According to expectation states theory, those with more power should use more negative behavior in interaction as a way of maintaining the system of stratification (Ridgeway & Johnson, 1990). Higherstatus people such as men should use more negative behavior when they encounter challenges to their 26

position, particularly illegitimate opposition from lower-status people such as women. Alternatively, according to identity theory, those with a more masculine gender identity should be more likely to use negative behavior in interaction because the meaning of masculinity is related to dominance and competition and this is more consistent with a negative style than positive style of interaction. In studying newly married couples, we find that while the data are consistent with identity theory, they are inconsistent with expectation states theory (Stets & Burke, 1996). Wives rather than husbands use more negative behavior in conversation. While the expectation states predictions could be correct but not apply to intimate interactions, it might also be true that the predictions are wrong. The finding that women are more likely to use negative behavior in interaction could be a gender specific finding, or it could be part of a more general pattern of how powerless persons act perhaps in response to being discounted by others. To test this, Stets (1997) conducted a follow-up study to examine whether the effects of negative behavior in marriage was confined to gender. Stets investigates individuals in other powerless groups in our society such as the young, the less highly educated, and those having a low occupation to investigate whether their status produced the same negative behavior. Consistent with the results on gender, Stets finds that those with a low status on these other dimensions (age, education, and socioeconomic status) are also more likely to use negative behavior in marriage. Further, the identity effects (that is, that those with a more masculine identity are more negative) remain. These results identify how ones structural position and self-meanings combine to produce action that maintains both the status hierarchy and the self. In general, the above results together with the earlier results by Stets and Burke indicate that one's self-meaning tied to the identity may, in fact, be a better predictor of behavior than predictions laid out from expectation states theory. The above statement must be qualified. While expectation states theory may not predict emotionbased behavior, it may provide insight into cognitive-based behavior, particularly how individuals see themselves. In yet another follow-up on the above research, we examine who is influential in how we see ourselves by investigating spouses self-perceptions from their own viewpoint (self-views) as well as the viewpoint of their spouse (Cast, Stets, & Burke, 1999). We find support for the expectation states theory 27

prediction that the views of the spouse with higher status in the marriage about their lower status partner influences how the lower status partner views himself or herself. Further, the higher status spouse influences the lower status spouse's views of the higher status person. Thus, ones social structural position serves as a signal as to who is likely to have power in the interaction. In this way, one's position in the macro social order generalizes to the micro social order. This important process again reflects the social structurally contingent nature of identity processes that needs to be incorporated more fully into identity theory. Another way that the predictive power of identity theory has been tested against alternative theories is through an analysis of trust and commitment in marriage. Research has examined trust and commitment through the lens of exchange theory (Kollock, 1994; Lawler & Yoon, 1996). We argue that when a person's identity is repeatedly verified in interaction, several consequences will emerge including positive feelings, increased trust for the other, commitment to the other, and a perception that one is part of a group (Burke & Stets, 1999). Exchange theory posits many of these same outcomes, but for different reasons. In exchange theory, commitment is influenced not by repeated self-verification but by repeated exchange agreements (Lawler & Yoon, 1996). Such agreements generate an emotional buzz between actors in the form of satisfaction or excitement. These mild positive emotions lead to relational cohesion or the perception that one is part of a group, and cohesion influences commitment. On the one hand, there may be little difference between exchange theory and identity theory since repeated exchange agreements may be viewed as self-verification. What is gained through the exchange is a confirmation of the self as needing the thing gained. On the other hand, in exchange terms, value preferences guide one's behavior whereas in identity terms, the identity standard sets the value. In exchange terms, we seek out rewards and avoid punishment. In identity terms, we are motivated to seek self-verification. We see identity theory as applying to a wider range of situations and relationships than those examined by exchange theorists. For example, identity theory can apply to individuals who have a history of interaction. Indeed, most of our daily interactions are characterized by such interactions. In

28

general, we see that one of the ways of extending a theory is testing it with alternative theories. In doing this, we may arrive at a better sense of the strengths as well as limitations of our own theory. Integrating the Identity Theory Versions The program of research of Stryker and his colleagues, that of Heise, Smith-Lovin and their colleagues, as well as Burkes program of research offer important theoretical assumptions as to the nature of identities and how they operate. Strykers work highlights the fact that identities exist within and reflect social structure. Identities are constrained by social structure, but they also maintain and facilitate the further development of social structure. Burkes work along with that of Heise and SmithLovin, highlight the dynamic process that emerges when an individual claims an identity in a situation and what occurs when that claim is not verified. The ideas of the above researchers can be easily integrated (Stryker & Burke, 2000). For example, in thinking about the identity standard in the identity control model, we can conceptualize it as reflecting the role meanings of particular groups that one is committed to in society. The more strongly a person is committed to one or more of these groups, the greater the salience of that identity. The greater the commitment and salience to a particular identity, the more likely those meanings will be perceived to be personally relevant in a situation, and the greater the motivational force inherent in a discrepancy to bring the identity standard and self-perception in line. The greater the difficulty in aligning the identity standard with self-in-situation meanings, the more likely it is that the identity will decrease in salience and result in decreased commitment to the relationships in which the identity is premised upon. The integration of the above programs of research can be seen not only in the incorporation of committed and salient identities (from the social structure) into the identity control model, but also how functioning identities influence the social structure. For example, in some of our recent research, we find that when individuals verify each other identities, commitment to each other increases, and there is a shift in cognition such that the individuals come to see themselves as a collectivity or group, that is, a new social structure (Burke & Stets, 1999). On the other hand, when the individuals have problems verifying their identities, ties may be broken so the structure comes apart. Indeed, Cast and Burke (1999) evidence 29

that husbands and wives whose spousal identities are not mutually verified in their marriages are more likely to be divorced.

Integrating Sociological and Psychological Identity Theory: Social, Role, and Personal Identities
Social and Role Identities Identity theory in sociological social psychology has chiefly focused on role identities. However, individuals not only occupy roles in society, but they are also members of some groups (and not others) and therefore may take on particular social identities. Social identity theory in psychological social psychology has been instrumental in informing us as to the processes involved in social identity formation, activation, and motivation (Abrams & Hogg, 1990; Turner et al., 1987). Role identity theory and social identity theory have developed as disparate lines of research. Unlike Hogg and his colleagues (Hogg, Terry, & White, 1995), we see significant similarities between social identity theory and role identity theory. We recently called for a merger of the two theories that would yield a stronger social psychology, that is, a general theory of identity, since it would integrate the various bases by which individuals are tied to society (Stets & Burke, 2000). We have argued that the overlap between identity theory and social identity theory is striking. For example, the process of self-categorization into groups in social identity theory (Turner et al., 1987) is analogous to the process of identification into roles in identity theory (McCall & Simmons, 1978). In selfcategorization, people compare themselves to others, and those who are similar to the self are categorized with the self and are labeled the in-group while those who are different from the self are categorized as the out-group. In identification, persons identify themselves as an occupant of particular roles. Rather than seeing others as similar to oneself in interaction, individuals see themselves as set apart from others in the counter-roles others assume in the interaction. For example, sons and daughters are different from the corresponding counter-roles of mothers and fathers. Students are different from the corresponding counter-role of teachers. What theorists in both traditions share is the idea that when persons categorize

30

themselves as a member of a group or role, they do so by seeing themselves as an embodiment of a (group or role) prototype or standard. This prototype/standard contains the societal meanings and norms about the social category or role, serving to guide behavior. Broadly speaking, then, theorists in both traditions recognize that individuals view themselves in ways defined by the social structure. Therefore, persons are born into a particular society with social categories pre-existent to the individual (Hogg & Abrams, 1988; Stryker, 1980). While identity theorists and social identity theorists see somewhat different consequences when individuals take on an identity, these varying consequences are equally important in understanding the self. According to social identity theory, when individuals take on a group-based identity, there is uniformity of perception and action among group members (Haslam, Oakes, McGarty, Turner, Reynolds, & Eggins, 1996; Oakes, Haslam, & Turner, 1994). According to role identity theorists, taking on a rolebased identity results in different perceptions and action between individuals, as roles interact with counter-roles (Burke, 1980; Burke & Reitzes, 1981). As we point out, social identities and role identities can simultaneously exist in a situation with the result that there are both similarities (social identities) and differences (roles identities) with others (Stets & Burke, 2000). Within groups, individuals assume different roles (intragroup relations), but persons will also categorize themselves as members of one group (the in-group) and not another (the out-group) (intergroup relations). While inter-group relations activate a sense of belongingness and self-worth for individuals (focusing on who one is), intragroup relations activate a sense of self-efficacy (what one does). Both self-worth and self-efficacy are important dimensions of self-esteem, and both appear to be fostered through their ties to social and role identities, respectively (Cast & Burke, 1999). Personal Identities Social identity and identity theorists have both discussed personal identities, but they have remained peripheral in both theories. As another basis of the self, we think more attention should be given to personal identities, particularly as they relate to social and role identities. In social identity theory, the personal identity is the lowest level of self-categorization (Brewer, 1991; Hogg & Abrams, 1988). 31

Categorizing oneself in terms of the personal identity means seeing the self as distinct and different from others. The person is guided by her own goals rather than the groups goals. The activation of a social identity rather than a personal identity in a situation is a product of accessibility and fit (Oakes, 1987). This is the process of depersonalization, shifting the perception of the self from being unique toward the perception of the self as a member of a social category (Hogg et al., 1995). The Me becomes a We (Thoits & Virshup, 1997). The person sees herself as the embodiment of the in-group prototype rather than as a unique individual. Depersonalization does not mean a loss of ones personal identity but rather a change in focus from the personal to the group basis of an identity. Since social identity theorists see the personal and social identity as mutually exclusive bases of self-definition, one cannot be both at the same time. Deaux (1992) attempts to link the personal identity to the social identity. She indicates that while social identities are expressed along normative lines, there is an aspect of social identities that may be expressed along personal, idiosyncratic lines. Thus, personal identities may be linked to social identities by creating new ways of expressing ones membership in groups. Deaux also suggests that some personal identities may represent a general view of the self, and therefore may pervade all the membership groups to which one belongs. Our own work has addressed personal identities (Stets, 1995; Stets & Burke, 1994). We see personal identities as tied to an individual rather than being attached to a role in society. They operate across various roles and situations. In this sense, our conceptualization is very close to that of Deaux. In the same way that we regulate the meanings of our role identities and our social (group) identities, we also regulate the meanings of our personal identities. Thus, personal identities are not dispositions to act in a certain way, but rather, like role and group identities, a feedback control process maintains them. Like role and group identities, perceptions of ones personal identity in a situation are compared to ones meaning of the personal identity held in the standard. Any discrepancy between the two will register an error and either behavior, perception, or the identity standard will be modified to resolve the discrepancy.

32

Another way of looking at these three bases of identity group, role, and person is in terms of the resources that are controlled by each (cf. Freese & Burke, 1994). Again, the distinction is analytic, being difficult to sort out in any empirical situation, but one may think of group-based identities as controlling resources that support the group, qua group, sustaining it, its patterns of interaction, its boundaries and so forth. Role identities, on the other hand, control resources that sustain the role. Since most roles exist within groups, such resources may also work ultimately to sustain the group. Finally, person identities control the resources needed to sustain the individual as a biological being, maintaining food, clothing, shelter, love and so forth. In the same way that Deaux (1992) attempts to link personal identities to social identities, Stets (1995) attempts to link personal identities to role identities. Stets argues that personal and role identities may be related to each other through a common system of meanings. In other words, the meanings of role identities may overlap with the meanings of personal identities. For example, Stets (1995) finds that the masculine gender role identity is linked to the mastery personal identity through the shared meaning of control. Stets observes that when the meanings of role identities conflict with the meanings of personal identities, people may act without regard to their role identities in order to maintain their personal identities. In general, attempts are continuing to be made to integrate these different identity bases. Most recently, Deaux and Martin (2001) offered a model that links social and role identities. They proposed that each large-scale group identity is linked to an interpersonal network of others. These interpersonal networks consist of people who share, to varying degrees, the category membership to which a person is a member and who provide, also in varying degrees, support for the group identity. Support comes in assuming reciprocal roles for the identity one claims, thereby producing role and counter-role identities.

Future Research
As we think about future research, a number of issues emerge: integrating the various bases of identities (group, role, and person), understanding how the multiple identities a person has are

33

interrelated, and developing better measures of both the meanings of identities and the identity verification process. We briefly discuss each of these. Integrating the Social, Role, and Personal Identities We suggest that future research examine all three bases of identities: person, role, and social identities. This would lead to a more integrated and a stronger theory of identity. Within groups there are roles and persons playing out those roles. All are operating in a situation and we need to identify how they are related in the setting. For example, it would be important to investigate how much group and role identities are constrained by normative expectations. The less constrained the normative expectations, the more it may be that personal identities can influence not only behaviors, but also the content of the role or social (group) identities, creating unique patterns of interaction. Looked at another way, we could examine the extent to which certain kinds of personal identities influence particular role identities and group identities without disrupting smooth social relations and social order. Part of it may be due to how much ones role in the group carries with it greater power and thus increased freedom to express oneself according to ones personal identities. Alternatively, some groups may allow greater entre of personal identities into group interaction than other groups. Further, we might examine whether some identities are more malleable than others. For example, people may be more likely to adjust their personal identities to adapt to situations than to modify structurally constrained role or group identities. Another concern is the conditions under which group, role, and person identities compete in a situation, or alternatively, support one another. Are some personal identities an easier fit with some role and social identities and not others? How much are individuals aware of this fit or lack of fit among identities? Just as individuals have multiple role identities (Thoits, 1983), they also have multiple personal identities (Deaux, 1992). Finally, there is the issue of the effect of these identities among themselves over time. For example, personal identities may influence role and group identities when they are first taken on. Once a role or group identity becomes established, however, personal identities may have little impact (Stets, 1995).

34

Integrating the various bases of the self is challenging given that there are multiple personal identities, multiple role identities, and multiple social identities. How can we conceptualize this interrelationship? We might demarcate the self-concept as having distinct salience hierarchies or distinct identity control systems that refer to these three different kinds of identities. The movement within the hierarchies across situations is influenced by situational factors and/or commitment to those identities. At issue is how these hierarchies operate to produce particular combinations of identities in any one situation. Multiple Identities Future research needs to more fully address the occurrence of multiple identities in a situation. We expect that having two oppositional identities activated at the same time in a situation will result in much distress because the verification of one identity necessarily increases the discrepancy for the other. For example, a person may have the identity friend to a peer and daughter to her parents. The two groups may intersect when the peer visits in the persons home while her parents are present. We hypothesize that under such conditions there will occur some sort of change in the identity standards involved with the more important, salient, or more committed identity shifting the least. This means that identities higher in the hierarchy of importance or salience should take preference in the verification process in a situation over identities lower in the hierarchy. When two identities share common meanings, the situation is simpler. Controlling self-in-situation meanings to match the identity standard helps both identities. Verifying one of the identities will help verify the other.10 For example, consider a married person with children. If the spousal identity includes standards for providing material support for ones spouse, and if the parent identity includes standards for providing material support for ones children, then getting a well-paying job will help verify both identities. We hypothesize that identities with common meanings will tend to be activated together. Identities that are often activated together should develop similar levels of salience and commitment. If we try to understand how individuals manage their different identities in a situation, the matter is more complex when multiple personal, role, and social identities become activated not only for any one person in a situation, but for two or more persons in a situation. Furthermore, since this self is not a static 35

entity but an entity that is dynamic and can change, it is important to examine how these different identities change over time and come to shape a new self-concept. [the role of simulations??] Measurement There is also the measurement of identity that needs to be developed. As Stryker and Burke (2000) point out, if we consider an identity as a person schema, that is, how you see yourself as you move from one situation to the next, we might be able to measure identity salience by way of response latency measures. The idea is that greater responsiveness to cues related to an identity will increase the likelihood that the behavior related to the identity will be performed. The cues might be pictures that individuals can identify with. The quicker they identify the pictures perhaps the higher the identity in their salience hierarchy and the more likely it is to be played out across situations. Additionally, commitment may be indicated by the (computer measured) strength of response to bring disrupted identity descriptions back in line with self-conceptions. In this case, not the latency of response would be measured, but the strength and persistence in restoring the self-description. We need more direct measures of identity verification as well. In the past, self-verification has been assessed by the degree of agreement between how individuals thought they should behave and how others thought they should behave (Burke & Stets, 1999) or by how individuals thought they should behave and how they report they did behave (Burke, 2001). We need to ask individuals what they think other's perceptions are of them. To the extent that others perceptions are consistent with one's own perceptions, then verification is more directly captured. Further, if emotion is an outcome of the verification process, when need to identify whether verification (or the lack of it) occurs when a person is feeling positive or negative emotion. This requires moment-by-moment data that may be difficult to obtain. Conclusion The research programs on self and identity included under the rubric of identity theory are strong, active, and developing cumulative theory. However, there is still much work to be done.

36

References Abrams, D., & Hogg, M. A. (1990). Social identity theory: Constructive and critical advances. London: Harvester-Wheatsheaf. Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147 Bandura, A. (1995). Exercise of personal and collective efficacy in changing societies. In A. Bandura (Ed.), Self-efficacy in changing societies (pp. 1-45). New York: Cambridge University Press. Bartels, D. J. (1997). An examination of the primary emotions of anger and sadness in marriage within the context of identity theory. Unpublished MA, Washington State University, Pullman, WA. Becker, H. S. (1960). Notes on the concept of commitment. American Journal of Sociology, 66, 32-40. Blumer, H. (1969). Symbolic interactionism. Englewood Cliffs, NJ: Prentice-Hall. Brewer, M. B. (1991). The social self: On being the same and different at the same time. Personality and Social Psychology Bulletin, 17, 475-482. Burke, P. J. (1980). The self: Measurement implications from a symbolic interactionist perspective. Social Psychology Quarterly, 43, 18-29. Burke, P. J. (1991). Identity processes and social stress. American Sociological Review, 56, 836-849. Burke, P. J. (1996). Social identities and psychosocial stress. In H. B. Kaplan (Ed.), Psychosocial stress: Perspectives on structure, theory, life course, and methods (pp. 141-174). Orlando, FL: Academic Press. Burke, P. J. (1997). An identity model for network exchange. American Sociological Review, 62, 134150. Burke, P. J. (2001). Relationships among multiple identities. Bloomington, IN: The Future of Identity Theory and Research: A Guide for a New Century Conference. Burke, P. J. (Forthcoming). Marital socialization and identity change. Vancouver, BC: The Pacific Sociological Association. Burke, P. J., & Cast, A. D. (1997). Stability and change in the gender identities of newly married couples. Social Psychology Quarterly, 60, 277-290. 37

Burke, P. J., & Reitzes, D. C. (1981). The link between identity and role performance. Social Psychology Quarterly, 44, 83-92. Burke, P. J., & Reitzes, D. C. (1991). An identity theory approach to commitment. Social Psychology Quarterly, 54, 239-251. Burke, P. J., & Stets, J. E. (1999). Trust and commitment through self-verification. Social Psychology Quarterly, 62, 347-366. Burke, P. J., & Tully, J. (1977). The measurement of role/identity. Social Forces, 55, 881-897. Callero, P. L. (1985). Role-identity salience. Social Psychology Quarterly, 48, 203-214. Carver, C. S., & Scheier, M. F. (1981). Attention and self-regulation: A control-theory approach to human behavior. New York: Springer-Verlag. Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. Cambridge: Cambridge University Press. Cast, A. D. (2001). Identities and behavior. Bloomington, IN: The Future of Identity Theory and Research: A Guide for a New Century Conference. Cast, A. D., & Burke, P. J. (1999). Integrating self-esteem into identity theory. Portland, OR: Pacific Sociological Association. Cast, A. D., Stets, J. E., & Burke, P. J. (1999). Does the self conform to the views of others? Social Psychology Quarterly, 62, 68-82. Cooley, C. H. (1902). Human nature and social order. New York: Charles Scribner's Sons. Deaux, K. (1992). Personalizing identity and socializing self. In G. M. Blackwell (Ed.), Social psychology of identity and the self-concept (pp. 9-33). London: Surry University Press. Deaux, K., & Martin, D. (2001). Which context? Specifying levels of context in identity processes. Bloomington, IN: The Future of Identity Theory and Research: A Guide for a New Century Conference.

38

DePaulo, B.M., Kenney, D. A., Hoover, C. W., Webb, W., & Oliver, P. (1987). Accuracy of person perception: Do people know what kinds of impressions they convey? Journal of Personality and Social Psychology, 52, 303-315. Ellestad, J., & Stets, J. E. (1998). Jealousy and parenting: Predicting emotions from identity theory. Sociological Perspectives, 41, 639-668. Felson, R. B. (1993). The (somewhat) social self: How others affect self-appraisals. In J. M. Suls (Ed.), The self in social perspective (pp. 1-26). Hillsdale, NJ: Lawrence Erlbaum Associates. Franks, D. D., & Marolla, J. (1976). Efficacious action and social approval as interacting dimensions of self-esteem: A tentative formulation through construct validation. Sociometry, 39, 324-341. Freese, L., & Burke, P. J. (1994). Persons, identities, and social interaction. In B. Markovsky, K. Heimer, & J. O'Brien (Eds.), Advances in group processes (pp. 1-24). Greenwich, CT: JAI Press. Gecas, V., & Burke, P. J. (1995). Self and identity. In K. Cook, G. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology (pp. 41-67). Boston: Allyn and Bacon. Gecas, V., & Schwalbe, M. L. (1983). Beyond the looking-glass self: Social structure and efficacy-based self-esteem. Social Psychology Quarterly, 46, 77-88. Haslam, S. A., Oakes, P. J., McGarty, C., & Turner, J. C. (1996). Stereotyping and social influence: The mediation of stereotype applicability and sharedness by the views of in-group and out-group members. British Journal of Social Psychology, 35, 369-397. Heise, D. R. (1979). Understanding events: Affect and the construction of social action. Cambridge: Cambridge University Press. Heise, D. R. (1989). Effects of emotion displays on social identification. Social Psychology Quarterly, 52, 10-21. Higgins, E. T. (1989). Self-discrepancy theory: What patterns to self-beliefs cause people to suffer? In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 93-136). New York: Academic Press.

39

Hoelter, J. W. (1986). The relationship between specific and global evaluations of the self: A comparison of several models. Social Psychology Quarterly, 49, 129-141. Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of intergroup relations and group processes. London: Routledge. Hogg, M. A., Terry, D. J., & White, K. M. (1995). A tale of two theories: A critical comparison of identity theory with social identity theory. Social Psychology Quarterly, 58, 255-269. James, W. (1890). Principles of psychology. New York: Holt Rinehart and Winston. Kaplan, H. (1975). The self-esteem motive. In H. B. Kaplan (Ed.), Self-attitudes and deviant behavior (pp. 10-31). Pacific Palisades, CA: Goodyear. Kenny, D., & Albright, L. (1987). Accuracy in interpersonal perception: A social relations analysis. Psychological Bulletin, 102, 390-402. Kiecolt, K. J. (1994). Stress and the decision to change oneself: A theoretical model. Social Psychology Quarterly, 57, 49-63. Kiecolt, K. J. (2000). Self change in social movements. In S. Stryker, T. Owens, & R. White (Eds.), Identity, self, and social movements (pp. 110-131). Minneapolis: University of Minnesota Press. Kollock, P. (1994). The emergence of exchange structures: An experimental study of uncertainty, commitment, and trust. American Journal of Sociology, 100, 313-345. Lawler, E. J., & Yoon, J. (1996). Commitment in exchange relations: Test of a theory of relational cohesion. American Sociological Review, 61, 89-108. Lindesmith, A. R., & Strauss, A. L. (1956). Social psychology. New York: Holt Rinehart and Winston. Linton, R. (1936). The study of man. New York: Appleton-Century-Crofts. Linville, P. (1985). Self-complexity and affective extremity: Dont put all of your eggs in one cognitive basket. Social Cognition, 3, 94-120. Linville, P. (1987). Self-complexity as a cognitive buffer against stress-related illness and depression. Journal of Personality and Social Psychology, 52, 663-676.

40

Longmore, M. A., & DeMaris, A. (1997). Perceived inequity and depression in intimate relationships: The moderating effect of self-esteem. Social Psychology Quarterly, 60, 172-184. MacKinnon, N. J. (1994). Symbolic interaction as affect control. Albany: State University of New York Press. McCall, G. J., & Simmons, J. L. (1978). Identities and interactions. New York: Free Press. Mead, G. H. (1934). Mind, self, and society. Chicago: University of Chicago Press. Nagel, J. (1995). American indian ethnic renewal: Politics and the resurgence of identity. American Sociological Review, 60, 947-965. Oakes, P. (1987). The salience of social categories. In J. C. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher, & M. S. Wetherell, M. S. (Eds.), Rediscovering the social group: A self-categorization theory (pp. 117-141). New York: Basil Blackwell. Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality. Oxford: Basil Blackwell. Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. Urbana, IL: University of Illinois Press. Owens, T. J., & Aronson, P. J. (2000). Self-concept as a force in social movement involvement. In S. Stryker, T. J. Owens, & R. White (Eds.), Self, identity, and social movements (pp. 191-214). Minneapolis: University of Minnesota Press. Powers, W. T. (1973). Behavior: The control of perception. Chicago: Aldine Publishing. Ridgeway, C. L., & Johnson, C. (1990). What is the relationship between socioemotional behavior and status in task groups. American Journal of Sociology, 95, 1189-1212. Ridgeway, C. L., & Walker, H. A. (1995). Status structures. In K. S. Cook, G. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology. Boston: Allyn and Bacon. Riley, A., & Burke, P. J. (1995). Identities and self-verification in the small group. Social Psychology Quarterly, 58, 61-73. Rosenberg, M. (1979). Conceiving the self. New York: Basic Books. 41

Rosenberg, M., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global self-esteem and specific self-esteem: Different concepts, different outcomes. American Sociological Review, 60, 141-156. Rothbaum, F., Weisz, J. R., & Snyder, S. S. (1982). Changing the world and changing the self: A twoprocess model of perceived control. Journal of Personality and Social Psychology, 42, 5-37. Scheff, T. (1994). Emotions and identity: A theory of ethnic nationalism. In C. Calhoun (Ed.), Social Theory and the Politics of Identity (pp. 277-303). Philadelphia: Temple University Press. Serpe, R. T., & Stryker, S. (1987). The construction of self and reconstruction of social relationships. In E. Lawler & B. Markovsky (Eds.), Advances in group processes (pp. 41-66). Greenwich, CT: JAI Press. Shrauger, J. S., & Schoeneman, T. J. (1979). Symbolic interactionist view of self-concept: Through the looking glass darkly. Psychological Bulletin, 86, 549-573. Smith-Lovin, L. (1988). The affective control of events within settings. In L. Smith-Lovin & D. R. Heise (Eds.), Analyzing social interaction: Advances in affect control theory (pp. 71-101). New York: Gordon and Breach Science Publishers. Smith-Lovin, L. (1990). Emotion as the confirmation and disconfirmation of identity: An affect control model. In T. D. Kemper (Ed.), Research agendas in the sociology of emotions (pp. 238-270). Albany: State University of New York Press. Smith-Lovin, L. (1995). The sociology of affect and emotion. In K. S. Cook, G. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology (pp. 118-148). Boston: Allyn and Bacon. Smith-Lovin, L. (2001). Self, identity and interaction in an ecology of identities. Bloomington, IN: The Future of Identity Theory and Research: A Guide for a New Century Conference. Smith-Lovin, L., & Heise, D. R. (1988). Analyzing social interaction: Advances in affect control theory. New York: Gordon and Breach Science Publishers. Stets, J. E. (1995). Role identities and person identities: Gender identity, mastery identity, and controlling one's partner. Sociological Perspectives, 38, 129-150. Stets, J. E. (1997). Status and identity in marital interaction. Social Psychology Quarterly, 60, 185-217. 42

Stets, J. E. (2001). Justice, emotion, and identity theory. Bloomington, IN: The Future of Identity Theory and Research: A Guide for a New Century Conference. Stets, J. E., & Burke, P. J. (1994). Inconsistent self-views in the control identity model. Social Science Research, 23, 236-262. Stets, J. E., & Burke, P. J. (1996). Gender, control, and interaction. Social Psychology Quarterly, 59, 193220. Stets, J. E., & Burke, P. J. (2000). Identity theory and social identity theory. Social Psychology Quarterly, 63, 224-237. Stets, J. E, & Tsushima, T. M. (2001). Negative emotion and coping responses within identity control theory. Social Psychology Quarterly, 64. Stryker, S. (1968). Identity salience and role performance. Journal of Marriage and the Family, 4, 558564. Stryker, S. (1980). Symbolic interactionism: A social structural version. Menlo Park: Benjamin Cummings. Stryker, S. (1987). The interplay of affect and identity: Exploring the relationships of social structure, social interaction, self, and emotion. Chicago: American Sociological Association. Stryker, S. (2000). Identity competition: Key to differential social movement involvement. In S. Stryker, T. Owens, & R. White (Eds.), Identity, self, and social movements (pp. 21-40). Minneapolis: University of Minnesota Press. Stryker, S., & Burke, P. J. (2000). The past, present, and future of an identity theory. Social Psychology Quarterly, 63, 284-297. Stryker, S., & Serpe, R. T. (1982). Commitment, identity salience, and role behavior: A theory and research example. In W. Ickes & E. S. Knowles (Eds.), Personality, roles, and social behavior (pp. 199-218). New York: Springer-Verlag. Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: Equivalent, overlapping, or complementary concepts? Social Psychology Quarterly, 57, 16-35. 43

Swann, W. B., Jr. (1983). Self-verification: Bringing social reality into harmony with the self. In J. Suls & A. Greenwald (Eds.), Psychological perspectives on the self (pp. 33-66). Hillsdale, NJ: Erlbaum. Swann, W. B. Jr. (1990). To be adored or to be known? The interplay of self-enhancement and selfverification. In E. T. Higgins & R. M. Sorrentino (Eds.), Handbook of motivation and cognition (pp. 408-448). New York: The Guilford Press. Swann, W. B. Jr., Griffin, J. J., Jr., Predmore, S. C., & Gaines, B. (1987). The cognitive-affective crossfire: When self-consistency confronts self-enhancement. Journal of Personality and Social Psychology, 52, 881-889. Tajfel, H. (1981). Human groups and social categories: Studies in social psychology. Cambridge: Cambridge University Press. Tesser, A. (1988). Toward a self-evaluation maintenance model of social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp.181-228). New York: Academic Press. Thoits, P. A. (1983). Multiple identities and psychological well-being: A reformulation and test of the social isolation hypothesis. American Sociological Review, 49, 174-187. Thoits, P. A. (1986). Multiple identities: Examining gender and marital status differences in distress. American Sociological Review, 51, 259-272. Thoits, P. A. (1992). Identity structures and psychological well-being: Gender and marital status comparisons. Social Psychology Quarterly, 55, 236-256. Thoits, P. A. (2001). Personal agency in the accumulation of role-identities. Bloomington, IN: The Future of Identity Theory and Research: A Guide for a New Century Conference. Thoits, P. A., & Virshup, L. K. (1997). Me's and we's: Forms and functions of social identities. In R. D. Ashmore & L. J. Jussim (Eds.), Self and identity: Fundamental issues (pp. 106-133). New York: Oxford University Press. Tsushima, T., & Burke, P. J. (1999). Levels, agency, and control in the parent identity. Social Psychology Quarterly, 62, 173-189.

44

Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (Eds.), (1987). Rediscovering the social group: A self-categorization theory. New York: Basil Blackwell. Turner, R. H. (1962). Role-taking: Process versus conformity. In A. M. Rose (Ed.), Human behavior and social processes (pp. 20-40). Boston: Houghton Mifflin. Turner, R. H. (1978). The role and the person. American Journal of Sociology, 84, 1-23. Webster, M., Jr., & Foschi, M. (1988). Overview of status generalization. In M. Webster, Jr. & M. Foschi (Eds.), Status generalization: New theory and research (pp. 1-20). Stanford: Stanford University Press.

45

Endnotes

Cooleys (1902) classic reflected or looking glass self has three principal elements: the

imagination of our appearance to the other person, the imagination of his judgment of that appearance; and some sort of self-feeling, such as pride or mortification (p. 184).
2

Rosenberg et al. (1995) argue that specific self-esteem is most relevant to behavior while global

self-esteem is most relevant to psychological well-being. Thus, when one is examining a specific behavioral outcome, it is best predicted by specific self-esteem that is somehow connected to that behavior, whereas psychological well-being is best predicted by global self-esteem.
3

Identity verification involves the cognitive process of matching self-relevant meanings in the

situation to the meanings defined in the identity standard. A match signals self-verification or selfconfirmation.
4

Stryker (1980) speaks of commitment in terms of the costs of losing or giving up the identity,

reminiscent of the idea of commitment as side bets introduced by Becker (1960).

5

Cast (2001) has recently argued that we need to examine situations in which behavior influences the

self-meanings associated with an identity rather than the other way around. She argues that when we adopt new role identities, it is likely that our understanding of the role identity meanings is vague and loosely organized. Trying out different behaviors may help us crystallize the role identity standard meanings. Once we have settled in on those identity standard meanings, then those meanings will chiefly direct future role behavior.
6

Another means by which self-verification is achieved is in the input side of the identity model where

actors may modify what they perceive in the situation so that the resulting perceptions better match their identity standard.
7

We deal more explicitly with identity change below. Behavioral strategies and cognitive strategies are analogous to engaging in primary and secondary

control, respectively (Rothbaum, Weisz, & Snyder, 1982). In primary control, one attempts to influence 46

the situation. In identity theory, this occurs on the output side. Individuals act in order to match self-insituation meanings with the identity standard. If this action is ineffective, one may resort to cognitive strategies. This is making adjustment to the current situation, otherwise known as secondary control. For example, one may bias their perceptions in a direction that reduces any discrepancy between self-insituation meanings and identity standard meanings. If this is ineffective, one may change their identity standard as a last resort. This is another secondary control strategy.
9

Of course, another alternative is to leave the situation in order to deactivate the identity or to give up

the identity entirely, as in the case of divorce.

10

There are situations when two identities are unrelated to each other. An action of one identity leaves

the other identity unaffected.

47