The Development of Identity Theory
The Development of Identity Theory
Abstract and Keywords
This chapter summarizes identity theory in its more general formulation as it is viewed in the work of several modern researchers. It begins by identifying the symbolic interaction roots of identity theory, and shows the differences between structural symbolic interaction and traditional symbolic interaction. Next, it reviews all of the theorists who helped develop identity theory and emphasize the social structural version of symbolic interaction. The chapter also presents two theories that have a lot in common with identity theory.
The Symbolic Interaction Roots of Identity Theory
Historically, identity theory grew out of symbolic interaction (SI), particularly structural symbolic interaction (Stryker 1980 ) as we have already mentioned. Specifically, structural symbolic interaction is a version of symbolic interaction that stands in stark contrast to the traditional approach to symbolic interaction. We will summarize how these two versions of SI are different since this has important implications for understanding the basis of identity theory.
Both versions of SI have the same intellectual heritage by drawing on the seminal work of the pragmatic philosopher George Herbert Mead (1934) and earlier intellectuals such as William James (1890) and Charles Horton Cooley (1902) as we discussed in the last chapter. Herbert Blumer (1969) coined the term “symbolic interactionism,” and it was his ideas that led to the development of what we refer to as traditional symbolic interaction.
Although structural symbolic interaction and traditional symbolic interaction have many differences, which we shall review, we shall begin with one important commonality between the two. Most symbolic interactionists would agree that we can best understand social behavior by focusing on individuals’ definitions and interpretations of themselves, others, and their situations. By identifying the meanings that actors’ attributed to their surroundings, by getting “inside their head” and seeing the world from their perspective, we can understand why people do what they do (Meltzer, Petras, and Reynolds 1977). There is a strand in traditional SI that emphasizes the behavior of actors rather than their internal, subjective worlds (Couch, Saxton, and Katovich 1986). It takes seriously the pragmatic philosophers’ (p.34) theory of action. The focus is on how individuals’ construct actions and how these actions are coordinated with others to accomplish individual and collective goals in interaction. This strand of symbolic interaction does not deny the fact that humans act on the basis of the meanings that things have for them. They simply choose to focus their attention on the actions of humans. In general, however, symbolic interactionists share the premise that we need to “get at” social actors’ subjective world and understand their reality as they do. From this assumption, the traditional and structural versions of SI begin to diverge. The basis of this divergence is twofold. They differ as to the role of social structure in understanding self and social behavior, and they differ on the use of a prior theory and the development of theory to explain social psychological processes. We first turn to the role of social structure.
Substantively, the traditional version of SI has a tendency to neglect the relatively “fixed” nature of social structure in its analysis of social behavior. In the traditional version of SI, social structure (and society, more generally) is viewed as always in a state of flux, in the process of being created and recreated through the interpretations, definitions, and actions of individuals in situations. In situations, from the view of traditional SI, actors identify the things that need to be taken into account, they act on the basis of what it is they have identified, and they attempt to fit their lines of action with others in the situation to accomplish their goals. Individuals are free to define situations in any way they wish with the consequence that society is always thought to be in a state of flux with outcomes determined by negotiation, but with no overarching organization or structure from the view of the individual. If social structure exists at all, it is a temporary social order, which is assembled when actors greet one another and interact, disassembled upon actors parting, and assembled anew when actors meet again. As Stryker (2000, p. 27) has remarked on the traditional version of SI, “[It] tends to dissolve [social] 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.”
The structural version of SI examines the role of social structure in social life. Society is not continuously in a state of flux as the traditional version of SI would claim. Instead, it is viewed as stable and durable, as reflected in the patterned behavior within and between individuals. It is preexistent to the person; “in the beginning there is society” as Stryker (1997) titles one of his papers. We are born into a social world that is ongoing and organized, and we learn about this organization through socialization. Others (parents, educators, the media, and so forth) teach us what it is.
According to the structural version, we come to learn that within society there are an array of groups, networks, communities, and institutions that sometimes are distinct from one another and sometimes overlap in a (p.35) competitive or cooperative manner. We navigate in and around these various crosscutting groupings depending upon our tastes, and they influence who we become. As Stryker and Vryan (2003, p. 22) argue,
Social structures in general define boundaries, making it likely that those located within them will or will not have relations with particular kinds of others and interact with those others over particular kinds of issues with particular kinds of resources. Structures will also affect the likelihood that persons will or will not develop particular kinds of selves, learn particular kinds of motivations, and have particular symbolic resources for defining situations they enter.
Exposure to particular social structures helps shape individual goals. As Sewell (1992, p. 21) remarks: “Without a notion of heaven and hell a person cannot strive for admission into paradise; only in a modern capitalist economy can one attempt to make a killing on the futures market.”
Social structure provides both limits and possibilities for actors’ behavior. In chapter 1, we introduced the concepts of “agency” and “social structure” in order to set the foundation for understanding identities. We briefly revisit these concepts here to emphasize why we need to take into account the social structure, why we should not just give it “lip service.” Because social actors have agency, that is, their actions are oriented to individually held goals (from proximal goals such as accomplishing a particular task in a situation to distal goals as in achieving one’s moral objectives), individuals have the capacity to create social structures. Social structures emerge from individual actions as those individual actions are patterned across persons and over time. And, actors have the capacity to change social structures as well, thereby reorienting social behavior with the results that new patterns emerge. Social movements are a good example of a way that social action can mobilize social structural change, and movements such as the civil rights movement and the feminist movement have been effective in this change.
However, individual action also occurs in the context of social structures within which the individuals exist. On the one hand, social structures impose constraints on the agency of actors. For example, there is strong evidence on the intergenerational transmission of class position or the intergenerational transmission of aggression. On the other hand, social structures can provide resources and opportunity structures for actors such that they can overcome these constraints. For example, we take notice when the unexpected occurs, as when people become upwardly mobile actors in the class structure or when their experience of a violent upbringing becomes transformed into an adult life of nonviolence. Complexity is added to the picture when we see that although agency involves an individual accomplishing certain goals, if the goals are consistent with social structural arrangements, they are reinforcing not only for the person but also for the structures within which the interaction is embedded. If the individual’s goals are in opposition to social (p.36) structural arrangements, interaction may, on the one hand, become disruptive and destabilize the existing structures or, on the other hand, become squelched as the individual is prevented from obtaining his or her goals.
Considering the influence of the social structure makes us aware that individuals’ outcomes are not completely orchestrated by their own or even others’ actions as traditional symbolic interactionists would maintain. Structural arrangements persist according to their own principles and intrude into interaction, and they can constrain the actions of actors. Indeed, every situation has an implicit status hierarchy, a distribution of resources, a set of norms that shape and guide interaction and so forth, and this may constrain what actors can accomplish.
Structural symbolic interactionists grapple with how actors and social structure relate to one another. We think good sociological research is that which goes back and forth between the agency of individuals and the social structure in which their actions take place. Such work shows how structures are the accomplishments of actors, but it also demonstrates how actors are always acting within the structures they create and are thereby constrained. It is agents who are producing actions for their own goals, the patterns of which constitute social structure. But, social structural forces also act back on their creators guiding and limiting what individuals can do. Now we turn to the second difference between the traditional and structural versions of SI: the role of theory.
There are some traditional symbolic interactionists who agree with the Blumerian assumption that we cannot create theory or use a priori theory to explain social behavior, because we would be predicting social behavior and such prediction is impossible. They argue that there is no constancy or stability that exists, that can be studied, or from which we can develop theory. Within the traditional SI perspective, individuals are seen as actively and constantly constructing and reconstructing meanings and interpretations of themselves, others, and the interaction; coordinated lines of action are continuously being produced and reproduced, and social structure emerges in the situation rather than helping define and shape the situation. Because everything is in a state of flux, researchers are unable to obtain any reliable measures of their concepts or ideas. There is no stable reference point for measurement. If there is no reliable way by which to measure concepts, then a first step in building theory is not met. According to traditional SI, at best, what researchers can turn to are “sensitizing concepts” that provide them with some direction as to what and where to look when examining social action, recognizing that this direction may change as actors’ meanings and lines of action shift and change in the situation. Further, a priori theory cannot be imported into one’s study, because the local situation will be sufficiently unique (again, because of its constant construction and reconstruction by social actors), thereby rendering prior concepts from existing theories useless. All that social researchers can do is show, after the fact, what has developed in the situation. Essentially, the focus of traditional symbolic (p.37) interactionists is on describing and understanding rather than explaining and predicting.
Many structural symbolic interactionists are committed to developing and testing predictive explanations of social behavior as are we. This commitment is rooted in Manford Kuhn’s (1964) view of symbolic interaction, which stands in stark contrast to Blumer’s position. Unlike Blumer, Kuhn was interested in universal predictions of human conduct. His development of “self theory” was an effort to develop a set of generalizations about the self. Kuhn saw individuals as having a stable, core self. As a way of measuring this stable reference point—the self—he developed the Twenty Statements Test (TST), which measured people’s responses (allowing up to twenty answers) to the question, “Who am I?” This questionnaire was an important development by which symbolic interactionists could begin to examine internal processes in a quantifiable manner across people and situations.
Most structural symbolic interactionists maintain that there is sufficient constancy in social life to warrant the development of theoretical generalizations, which has been done with much success. Concepts can be measured; predictions about the self, action, and interaction can be developed and tested; theories can emerge from this testing; and further expansion and development of theories is possible with additional observations across place and time. Although there is change, there is enough stability to generate meaningful data that is sufficiently reliable and valid, which allows for generating and testing hypotheses, and which facilitates the growth of theory. In this way, the scientific procedure can be applied to an analysis of social life, and, as we will see, identity theory has been highly successful along these lines.
Introducing Identity Theory
Consistent with our focus, all of the theorists that we review below who have been instrumental in developing identity theory emphasize the social structural version rather than the traditional version of symbolic interaction. Sheldon Stryker emphasizes this approach the most, and Peter Burke probably emphasizes it the least. Nevertheless, all share the assumption that society is patterned and organized, and the self emerges within the context of a complex, organized society. If society is organized, so too must the self be organized. This reflects that dictum that the “self reflects society” (Stryker 1980 ). This idea follows from James’s (1890) notion that there are as many selves as there are different positions that one holds in society and as there are different groups who respond to the self.
In identity theory, different theorists focus on different aspects of one’s identity. For this reason, we can say that identity theory has slightly different emphases (Stryker and Burke 2000). In the work of Stryker and his (p.38) colleagues (Serpe 1987; Serpe and Stryker 1987; Stryker 1980 ; Stryker and Serpe 1982; 1994), the focus is on how the social structure influences one’s identity and behavior. Peggy Thoits’s (1983; 1991; 1995) research also has this emphasis. The work of Burke and his associates (Burke 1980; Burke 1991; 2004a; Burke and Reitzes 1981; 1991; Burke and Stets 1999; Cast, Stets, and Burke 1999; Stets and Burke 2000) emphasizes the internal dynamics within the self that influence behavior. A third emphasis is in the work of George McCall and J. L. Simmons (McCall 2003; McCall and Simmons1978). Here the focus is on how identities are maintained in face-to-face interaction. Though a clear program of research has not come out of McCall and Simmons’s work, they do make important theoretical contributions to understanding identities that are important to review. Below, we provide more detail on each of these orientations. Before reviewing these major emphases in identity theory, we’d like to briefly discuss the history of the idea of identity from some of the earliest theorists.
Erik Erikson (1950) was one of the first to bring this concept into the social sciences, though his usage of the term corresponded more to what we would call the self-concept, a more global representation of what Rosenberg (1979) described as all that one knows, thinks, and feels about oneself and who one is. Nelson Foote (1951) was perhaps the first to use the term “identity” close to the way it is used in identity theory. For Foote, the concept of identity gave motivation to the rather sterile concept of role. He recognized that roles prescribe relationships and behaviors, that there are expectations associated with the role positions. But the energy, motivation, and drive that make roles actually work require that individuals identify with, internalize, and become the role. The term “identity” thus describes this part of the individual who takes on and becomes the police officer or the quarterback or the mayor. It was this view of identity that McCall and Simmons (1978) used in their discussion of role identities. These were defined as “the character and the role that an individual devises for himself as an occupant of a particular social position” (p. 65). McCall and Simmons suggest that these are not idle musings, however, but serve as important influences on daily life and are a primary source of plans of action. There is motivational drive involved.
There is another side to identity that Greg Stone (1962) made clear. It is not simply the case that an identity involves an identification with and internalization of some position such as police officer by an individual. It is also the case that others identify the person as a police officer. Stone suggests that “a person’s identity is established when others place him as a social object by assigning him the same words of identity that he appropriates for himself” (p. 93). For Stone, “appearance” establishes and maintains one’s identity. By dressing a certain way, one announces to oneself and to the audience the identity that is being enacted such as one’s age, gender, occupation, and so forth. And as one’s position changes in and out of social relationships and in and out of social positions, one’s appearance and identity to which it refers also changes. Stone’s emphasis on appearance and (p.39) performance of one’s identity in interaction is a theme that continued to be discussed in McCall and Simmons, Stryker, and Burke’s work. We now turn to these contemporary theorists in more detail.
The Interactional Emphasis: George McCall and J. L. Simmons
McCall and Simmons (1978) are among the early originators of modern work in identity theory. In their interactional focus, we see the influence of Kuhn’s self theory (Kuhn and McPartland 1954) and George Homans’s (1974) exchange theory, among others. Unlike some who see the self as a study unto itself, McCall and Simmons are interested in how the self influences behavior. Further, since behavior emerges in interaction, they take seriously aspects of exchange such as negotiation and bargaining and rewards and costs that facilitate or impede action.
For McCall and Simmons, their central concept is a role identity. A role identity is one’s “imaginative view of himself as he likes to think of himself being and acting as an occupant”of a particular social position (McCall and Simmons 1978, p. 65). McCall and Simmons remark that role identities have a conventional dimension (the role of role identities), which includes the cultural expectations tied to social positions in the social structure that actors try to meet, and they have an idiosyncratic dimension (the identity of role identities), which involves the distinctive interpretations that individuals bring to their roles. For example, a professor identity may entail meanings of one as an educator and researcher. These are the conventional dimensions of the professor identity. Some may add to this an idiosyncratic dimension such as “friend to students” or “protector of students.” Either one of these is more distinctive meanings not typically found in the professor identity. Nevertheless, some may take on these meanings, and their behavior as professor will be slightly different than if they were guided completely by conventional meanings. McCall and Simmons point out that individuals can be at one extreme or the other on these dimensions. For example, they can rigidly adhere to the culturally defined behaviors attached to roles, or they can adopt unique behaviors such that they become unrecognizable to others, and others may perceive them as eccentric, even mentally disturbed. However, most individuals fall somewhere between the two extremes.
In their conceptualization of role identities, McCall and Simmons are more likely than other identity researchers in the structural SI tradition to discuss the idiosyncratic and idealized dimension of identities—to see identities as improvised and negotiated rather than as normative and conventional. In an effort to maintain an idealized conception of themselves—to legitimate their role identities—McCall and Simmons indicate that individuals enact role performances. This is the behavior (real or imagined) that they enact and that is guided by their role identity. Imagined behavior is what persons fantasize would maintain their role identity. In imagining, people provide (p.40) themselves with varying amounts of self-support, that is, self-confirming feedback that the behavior they would enact would fit their view of their identity. However, behavior that is portrayed in people’s imagination only partially legitimates an identity claim. An identity is also legitimated by behavior in the presence of others. When roles are performed in front of an audience, others evaluate persons’ identity performances and confirm or disconfirm their idiosyncratic imaginations of themselves as well as their conventional and normative views.
McCall and Simmons maintain that individuals typically claim more than one role identity. They conceptualize individuals’ multiple role identities as organized into a hierarchy within the self. They are primarily concerned with what they label a prominence hierarchy of identities. This hierarchy entails how individuals like to see themselves given their ideals, desires, or what is central or important to them. Perhaps what is central to them is being a good parent or professor or friend; thus they would claim the parent identity or professor identity or friend identity. The higher the identity in the prominence hierarchy, the more important it is. Where an identity appears in the prominence hierarchy depends upon several factors. First is how much individuals get support for the identity they are claiming in a situation. The more those individuals generate self-support and experience support from others for an identity they are claiming, the higher that identity in the prominence hierarchy.
Another factor influencing the placement of an identity in the prominence hierarchy is how committed individuals are to the role identity. If persons are invested in the identity that they are claiming, such that they derive a great deal of esteem or positive feelings when they live up to the view they have of the identity, then that identity is prominent to them. Still another factor influencing placement of an identity in the prominence hierarchy is the rewards individuals receive from the identity, both extrinsic and intrinsic. Extrinsic rewards include resources such as money, valued items, favors, and prestige that individuals obtain from others for an identity that they claim. Intrinsic rewards are the gratifications that individuals experience, internally, for the performance of a role. These may include a sense of efficacy or feeling of competence while enacting a particular behavior or following from the behavior. The more the extrinsic and intrinsic rewards associated with a particular identity, the higher that identity in the prominence hierarchy.
The above three factors, then—support, commitment, and rewards—influence the prominence hierarchy, or what McCall and Simmons sometimes label the ideal self. Which factor weighs more heavily into the ideal self is assumed to vary from person to person. For some it may be support from others, while for others it may be the internal and external gratification. Generally, the prominence hierarchy reflects persons’ priorities, which in turn serve to guide their actions across situations and over time. However, the prominence hierarchy is not the only determinant of behavior, because (p.41) prominent identities are not always activated in situations. Sometimes, less prominent identities get activated in a situation because of norms or pressures from others. For example, at a party, one might want to enact behavior associated with the professor identity such as instructing a student at the party about a way to collect his data for his dissertation, but the norm in the situation may be to behave along lines associated with the partygoer identity as in keeping discussions to topics that are not work related. Consequently, McCall and Simmons identify a second hierarchy of identities: a salience hierarchy.
The salience hierarchy of identities reflects the situational self rather than the ideal self. It is the self that responds to the expectations of the situation rather than to the desires of the self. It is the identity that is perceived as most advantageous to adopt in a situation in terms of getting support. Although the salience hierarchy is rather fluid as role identities become temporarily activated in different situations, the prominence hierarchy is more enduring and stable. Several factors influence the salience of an identity in a situation: prominence, support, rewards, and the perceived opportunity structure. McCall and Simmons maintain that the degree of prominence of the role identity is the most significant factor. The weight of the other factors in influencing the salience hierarchy varies from person to person.
The more prominent a role identity, the more likely it will be invoked in a situation. The second factor is how much individuals need support for a particular identity. According to McCall and Simmons, if support for an identity has recently been less than expected, particularly from an audience that is important to him or her, he or she will focus attention on another identity that has received prior support. The person will not seek to maintain a previously nonsupported identity, because he or she doesn’t want to risk the identity not being supported again. Thus, he or she will disinvest in a threatened identity and reinvest in an alternative, nonthreatened identity. For example, if the professor identity was not supported at a party but the parent identity was supported, the parent identity may be more likely to be activated in the future when a social event emerges. Looked at another way, an identity that has received prior support and in which the support is at expected levels or higher than expected levels will be the identity in which persons will continue to activate in a situation.
The third factor that determines the placement of an identity in the salience hierarchy is a person’s need for the kinds and amounts of intrinsic rewards (for example, esteem or physical gratification) and extrinsic rewards (for example, material possessions or power) gained through performance of an identity. Unlike the above in which expected or higher-than-expected levels of support for an identity increases the likelihood that individuals will continue to seek support for that identity, when persons’ receive less than expected or more than expected intrinsic or extrinsic rewards for an identity, they will be more likely to enact that identity in the future. In receiving less than the expected rewards for an identity, persons experience relative (p.42) deprivation; the more they are deprived, the more they desire the rewards that are tied to the identity. In receiving more than the expected rewards, individuals experience relative enhancement; and the more they obtain an excess of rewards, the more they will come to expect the surplus of rewards that are associated with a particular identity.
The final factor associated with the placement of an identity in the salience hierarchy is the perceived opportunity structure in the situation. Opportunity involves the amount of profit (rewards-costs) individuals will experience for playing out a particular identity in a situation. It is important to emphasize that reward/cost calculations are subjective; they are from the point of view of the actor in terms of what role identity is prominent, needs support, and is rewarding to the actor. A person may think that the playing-out of a particular identity by another actor in a situation is not profitable from the person’s perspective. However, the person needs to take the role of the other actor, for it is that actor’s role identity that is important, not the person’s role identity.
McCall and Simmons discuss the fact that individuals may not accurately assess the opportunity structures available. There may be a significant difference between what individuals perceive to be the costs and rewards for enacting a particular identity and what actually exists in a situation. For example, a person may not see the opportunity for profit in a situation as when a first-time mother sees no advantage to performing well in the worker identity at her job because she sorely misses her newborn and prefers spending the time with her child. So she may spend her time at work socializing with others and talking about her newborn, planning activities for her child, and so on. In this sense, she underestimates the rewards such as a salary advance for engaging in behavior associated with the worker identity. Alternatively, individuals may overestimate the rewards that may accrue for the enactment of a role identity. For instance, a person who invests in the worker identity may find himself the recipient of a low salary. Persons may also overestimate or underestimate the costs associated with performing a certain role identity. Indeed, a young lawyer may underestimate the costs associated with becoming a partner in a firm, or a new parent may underestimate the costs associated with having a child. Finally, persons may miscalculate the nature of the costs and rewards; the costs and rewards may be of a different kind from that which was anticipated. For example, rather than a behavior associated with an identity being costly in terms of physical energy, it may be costly in terms of emotional energy. Thus, a therapist may tire easily not because she is putting in long hours at the office but because her clients’ problems are emotionally taxing. And although one may expect material rewards such as money for enacting an identity, he or she may only receive social rewards such as praise and admiration.
For successful enactment of a role identity in a situation, McCall and Simmons highlight the importance of negotiation with others in the situation. Specifically, enacting an identity in an interaction is always done (p.43) in relation to a corresponding counteridentity of another as in the case of a husband supporting his wife, a mother instructing her child, or a professor teaching a student. However, one’s own expectations (whether conventional or idiosyncratic) as to how to act based on the content of a role identity (as its held for himself or herself) in a situation may differ from the expectations others associate with that role identity. For example, one may enact a role behavior that others have not seen before because the content of the role identity is more idiosyncratic. Another in the situation may rely on the conventional expectations associated with that role behavior because the content of the role identity (as he or she views it) is more conventional than idiosyncratic. For example, some professors may go out for a beer with their students after class while other professors may see this as unprofessional and inappropriate. A compromise is needed between disparate views.
Aside from the possible conflict in the meaning of identities and their corresponding behavior, each actor in an interaction has a view of his or her own identity as it relates to the identity of the other, and both actors need to enact a behavior that interrelates with the other. McCall and Simmons indicate that this requires a certain degree of coordination between individuals; and when conflict emerges, negotiation and compromises must follow so that there is support for each person’s behavior and the interaction can proceed smoothly. When interaction does run smoothly, it aids in the development of durable relationships, which helps stabilize persons’ prominence hierarchies (salience hierarchies in McCall and Simmons’s view are transitory and unstable, by nature). Essentially, people can come to settle on what is central or important to them if they have positive interactions and strong ties with others. This is not to say that individuals’ prominence hierarchy does not change as some priorities change over time and relationships end and others emerge.
Sometimes an identity is threatened in a situation as when others do not support one’s role performance (behavior). Here, the role identity is not being legitimated. Under these conditions, individuals will experience negative emotions, which may prompt them to use any one of the negative emotions in a series of what McCall and Simmons (1978, p. 92) label “mechanisms of legitimation” to cope with the negative feeling. One mechanism is short-term credit in which an identity that is currently not being supported is temporarily accepted by others because it was supported in the past. Essentially, actors draw upon a line of credit they have earned from prior identity support to “ride out” a current, unsuccessful role performance. Here we might have a physician who is having difficulty diagnosing a patient’s health problems. He might ask his patient to trust him until he figures out what might be the problem. Perhaps he might remind his patient that he has served her well in the past. If the physician still can’t make sense out of what the symptoms represent, he might send her to a specialist. He still hopes the patient will call upon him when she experiences additional ailments even though he has not found a diagnosis for her current ailment. (p.44)
Another mechanism is selective perception in which actors attend to cues that they think support an identity of theirs, and they do not attend to cues that do not support their identity. In the above example, a physician may focus on the patient’s positive nonverbal cues such as smiling, nodding her head, and showing a comfortable and relaxed posture, but he may ignore her verbal cues such as complaints that the medication he prescribed was ineffective in eliminating her symptoms or that he is unavailable for consultation when she needs it. Closely related to this is selective interpretation in which actors interpret cues as supportive of their identity when these cues are not supportive. So although the patient may voice her complaints to the physician, the physician may not interpret these as complaints. Instead, he may think that she is simply demonstrating how much she trusts him; otherwise, she would not be open and honest on her thoughts and feelings.
Other strategies to manage negative emotions when an identity is threatened include blaming others in the situation for not confirming an identity, criticizing them, and even sanctioning them for their lack of support. Here a physician may reprimand a patient for any criticism of him and inform the patient that she is no expert. One may also disavow an unsuccessful role performance as not what the person intended. So a physician may inform the patient that it was not his intention to prescribe the medication that was ineffective; rather his intention was to prescribe an alternative. Still yet, one might switch identities and in so doing, the alternative identity can be confirmed. Or the person can simply withdraw from the interaction. Essentially, these mechanisms help individuals avoid the pain associated with disconfirmed identities.
McCall and Simmons’s work has provided identity researchers with a theoretically rich and fruitful way of understanding the self and other in interaction. Their focus on how actors identify themselves in terms of taking on particular roles, and the implications of this identification for self-other interactions, has led to important contributions in identity theory. Recently, McCall (2003) has called for an expansion of self-identification by arguing that we need to investigate actors’ self-disidentifications, that is, who people claim they are not. Rather than answering the question, “Who am I?” following Manford Kuhn’s TST, he indicates that we should also answer the question, “Who am I Not?” McCall proposes that self-identification and self-disidentification can be regarded as the positive and negative poles of identity: the “Me” and “Not-Me.” In a pilot study examining these poles, he finds some interesting differences. For example, he finds that the “Me” is framed more in terms of roles and statuses, for example, “I am a student” or “I am a Catholic,” while the “Not-Me” is framed more in terms of characteristics and dispositions, for example, “I am not dishonest” or “I am not a wimp.” An important issue he raises for future research is how, over time, what is “Me” can become “Not-Me” and correspondingly, what is “Not-Me” can transform into “Me.” Although these identity changes can be brought about by expected role transitions throughout the life cycle, they may also occur (p.45) unexpectedly. Conceptualizing identity change as a movement between the positive and negative poles of identity serves as fertile ground for the theoretical development of identity theory.
The Structural Emphasis: Sheldon Stryker
Stryker is another early originator of contemporary work in identity theory. His structural emphasis in identity theory is rooted in bringing together ideas from role theory and Mead’s social psychology. The precursors of role theory such as Georg Simmel (Wolff 1950) and Ralph Linton (1936) brought to our attention the idea that society can be conceptualized as a structure of positions, and these positions carry with them expectations for behavior (roles). Viewing society as organized and patterned and merging these ideas with Mead’s (1934) ideas on the self and action serve as the foundation for a structural version of identity theory.
Like that of McCall and Simmons, Stryker’s key concept has been a role identity, but he has focused more on the normative, conventional aspect of role identities rather than the idiosyncratic aspect that McCall and Simmons address. In this way, he takes as a starting point the meanings of role identities that are largely shared among individuals. For example, the parent identity includes meanings of caregiver, teacher, and moral guider that most people would agree with, and parents would behave according to these meanings. Stryker is also less concerned than McCall and Simmons with how identities get negotiated in interaction. Instead, he is more concerned with how the social structure affects the self and one’s identity and, in turn, behavior. However, like McCall and Simmons, Stryker shares the idea that individuals typically claim more than one role identity, and given individuals’ multiple role identities, these identities can be conceptualized as organized into a hierarchy within the self.
As Stryker (Stryker and Burke 2000) has recently reiterated, society is made up of an enduring pattern of interactions and relationships that are differentiated yet organized. These interactions and relationships exist within groups, organizations, communities, and institutions. Persons live their lives in small networks of social relationships by playing out roles that support their membership in these networks. The probability of entering some networks (and not others) is influenced by the larger social structure within which the networks are embedded that makes available some networks (and not others) for individuals. Given the different positions persons may hold in different social networks and the corresponding roles and expectations associated with those roles, Stryker became interested in how persons choose one role behavior over another in a particular situation. This led him to look at the self for an answer.
For Stryker, people’s role choices are a function of the identities they claim in a situation. Identities are persons’ internalized role expectations in the sense that individuals take these expectations to be their own, as part of (p.46) who they are. For each role a person plays out in a social network, there is a corresponding identity attached to it. Thus, the role of father, colleague, or friend embeds individuals into the network of family, the academy, and friends, and there is a corresponding father identity, colleague identity, and friend identity attached to each role, respectively. The content of these various identities—these role expectations—becomes incorporated into individuals’ self-views.
If people play out different roles in different networks, Stryker reasoned that the corresponding role identities must be organized in some fashion such that some role identities are more likely to be invoked in a situation than other role identities. Thus, Stryker conceptualized persons’ role identities as organized in what he termed a salience hierarchy (Stryker 1980 ). A more salient identity (near the top of the hierarchy) is one that has a higher probability of being activated across different situations. If it has a greater chance of being activated across a variety of settings, then the behaviors associated with that identity that are in accord with the role expectations will be more likely to be enacted. Thus, the salience hierarchy identifies how social actors will likely behave in a situation.
Stryker’s approach to conceptualizing identities is somewhat different from McCall and Simmons’s approach. Recall that for McCall and Simmons, the salience hierarchy—the situational self—is “the person’s own preferences as to the subset of role identities he will enact in a given situation” (McCall and Simmons 1978, p. 84). Essentially, the salience hierarchy helps predict a person’s behavior in the short run. Longer-run predictions, according to McCall and Simmons, are determined by the prominence hierarchy—the ideal self—the relatively enduring aspect of the self that focuses on what is important to the self. More prominent identities influence which identities individuals prefer to enact in a specific situation. In this way, the prominence hierarchy influences the salience hierarchy.
For Stryker, the salience hierarchy—the readiness to act out an identity across situations—directly influences the choices people make among behavioral options. Thus, rather than the salience hierarchy predicting short-run behaviors as McCall and Simmons would argue, it predicts longer-run behaviors. In this sense, Stryker’s salience hierarchy has the same effect as McCall and Simmons’s prominence hierarchy; it captures the more enduring rather than fleeting source of behavior. However, McCall and Simmons’s prominence hierarchy is different from Stryker’s salience hierarchy in that the former assume that individuals are aware of their prominence hierarchy, that is, persons are self-aware of more important identities compared to less important identities. For Stryker, people may not be aware of how salient an identity is in their hierarchy, but their behavior would inform them as to its ranking in their hierarchy (Stryker and Serpe 1994). Additionally, each hierarchy carries different meanings (what is important, or the prominence hierarchy, compared to what one is ready to enact, or the salience hierarchy). Thus, identity prominence and identity salience should be kept (p.47) distinct. Future work will want to investigate whether it is fruitful in the development of identity theory to establish a causal ordering between identity prominence (importance) and identity salience (probability of enacting the identity).
Stryker (1980 ) maintains that one important factor that influences the salience of an identity is the degree of commitment one has to the identity. Broadly speaking, commitment is equated with the costs the person incurs for not playing out a role based on an identity (Stryker 1980 ). If the costs for giving up the identity are high, then commitment to the identity is high. Costs are examined along two dimensions: the number of ties and the strength of the ties to others in one’s social networks based on an identity.
The above two dimensions have been labeled the quantitative and qualitative aspects (Stryker and Serpe 1982; 1994) or interactional and affective components (Serpe 1987; Stryker 1987) of commitment, respectively. In the former, reflecting the individual’s ties to the social structure, commitment is the number of persons that one is related 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. Regarding the qualitative or affective dimension of commitment, the stronger or the deeper the ties to others based on a particular identity, the higher the commitment to that identity. Stryker (1968; 1980 ) suggests that the greater the commitment to an identity, the higher the identity in the salience hierarchy.
Once again, the relevance of social structure in understanding the self is made clear in Stryker’s conceptualization. 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 man’s 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. He’ll enact the father identity not only at home but also at work (by frequently talking about his children and displaying many pictures of them in his office), at social gatherings (again, discussing his children often), and so forth.
Empirical research strongly supports the link among commitment, identity salience, and the enactment of behavior that is 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, (p.48) 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 persons with many relationships based on religion (high commitment) 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 Serpe’s 1982 commitment scale), and behavioral 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 one’s identity salience hierarchy, and the more this salient role identity is related to donating blood.
In still another study, Nuttbrock and Freudiger (1991) investigate the salience of the mother identity among first-time mothers. They find that a more salient mother identity or the tendency for women to invoke the mother identity at school, at work, and with friends by talking about their child and showing pictures of their child influences behaviors consistent with the mother identity. Specifically, women with a more salient mother identity are more likely to accept the burdens of motherhood, that is, perform the parenting role without help from others such as their husband, and they are more likely to make sacrifices for their child including spending the necessary time and energy with the child.
Nuttbrock and Freudiger (1991), along with McCall and Simmons (1978), maintain that there is a causal ordering between identity prominence and identity salience, with an important identity influencing the salience of the identity. Some earlier work didn’t even maintain a distinction between identity prominence and identity salience and instead merged the two concepts both conceptually and empirically (see the discussion on this point by Stryker and Serpe 1994). Others have argued that we should not assume a causal ordering between identity importance and identity salience (Ellestad and Stets 1998; Stryker and Serpe 1994). Here, there is simply an association between the two processes. Not only is it possible that enacting an identity based on it being salient in one’s hierarchy of identities could reflect the importance of that identity to the person, but also the importance of the identity could guide the invocation of the identity in the situation. Since there is no empirical evidence, to date, that verifies a causal ordering, we prefer not to identify one.
The Perceptual Control Emphasis: Peter J. Burke
While Sheldon Stryker focuses on the hierarchical arrangement of identities and how identities are tied to the social structure, Peter Burke’s work focuses (p.49) more on the internal dynamics that operate for any one identity (Stryker and Burke 2000). In early work, Burke (1980; Burke and Reitzes 1981; Burke and Tully 1977) argued that identity and behavior are linked through a common system of meaning. When we try to understand a person’s behavior, the meaning that the behavior evokes should correspond to the meaning that is held in one’s identity. For example, if a person has a student identity that contains the meaning of being “academic,” then we should find that the person regularly attends class, take notes, passes exams, and finishes courses (Burke and Reitzes 1981; Reitzes and Burke 1980). However, if a person has a student identity that contains the meaning of being “social” rather than “academic,” then we should expect this person to spend much of his time socializing with his friends, attending parties and other social events, and so forth. Essentially, the meaning of one’s identity has implications for how one will behave, and one’s behavior confirms the meanings in one’s identity. For Burke, meaning is critical to understanding an identity. We discuss the role of meaning in identity theory in more detail in chapter 5. Here, we briefly review the major points.
Essentially, tied to each identity is a set of meanings that persons attribute to themselves when they are playing out or claim an identity. The meanings associated with the identity come to be known to the person through interaction with others in the situation in which others respond to the individual as if the person had these set of meanings. Thus, self-meanings develop from the reactions of others; and over time, a person responds to him or her self in the same way that others respond to the person, such that self-meanings become significant or shared by all (Burke 1980).
Drawing on the work of Osgood, Suci, and Tannenbaum (1957), Burke and Tully (1977) developed a method for the measurement of identity meanings that people claim for themselves. As we will discuss in more detail in chapter 5, to identify the self-meanings associated with an identity, a semantic differential framework is used in which respondents are given a set of bipolar adjectives. They respond to themselves as objects along the bipolar dimensions such as good and bad, powerful and powerless, or active and passive to help locate their identity meanings. Since any identity contains a set of multiple meanings (Burke and Tully 1977), multiple bipolar dimensions are provided for any one identity to which one can respond. For example, the gender identity of femininity includes being “noncompetitive” (competitive–not at all competitive dimension), “passive” (very active–very passive dimension) and “feelings easily hurt” (feelings not easily hurt–feelings easily hurt dimension) (Burke, Stets, and Pirog-Good 1988; (p.50) Stets and Burke 1996). The Burke-Tully procedure uses the meanings of the people in a particular subpopulation to formulate a particular identity rather than meanings from another source, such as the researcher’s own intuitive feeling or another population. The semantic differential method has been used to investigate the meanings associated with a variety of identities such as gender identity (Burke and Cast 1997; Burke, Stets, and Pirog-Good 1988; Stets and Burke 1996), the student identity (Reitzes and Burke 1980), the old age identity (Mutran and Burke 1979b), the environment identity (Stets and Biga 2003), and the moral identity (Stets and Carter 2006).
From the above measurement of meaning, we’ve come to understand some of the primary meanings of people’s identities within our culture. For example, we find that the student identity is comprised of multiple meanings including being academically responsible, intellectually curious, sociable, and personally assertive. The environment identity can be understood in terms of individuals holding either meanings of anthropocentrism (humans are independent and separate from the environment) or ecocentrism (humans are interdependent with the environment and should care for the physical world). The moral identity involves meanings of care and justice. What is important about the measurement of identities is that the meanings that individuals have for their identities affect how they will behave. For instance, those with a more feminine gender identity would be less likely to be a leader in a group, and those with an environment identity of ecocentrism would be more likely to recycle. Thus, when we identify the meanings of an identity for an individual, we can predict the meanings of the person’s behavior.
More recent conceptions of identity since Burke’s early work on meaning 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), which we discuss in more detail in the next chapter. This is where the internal dynamics of identities are most clearly seen (Burke 1991; 1996; Burke and Cast 1997; Riley and Burke 1995; Tsushima and Burke 1999). In the next chapter, we discuss in detail how identities operate within the perceptual control system. For now, we highlight the critical features of the cybernetic model of the identity process.
Since an identity is a set of meanings attached to the self, 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 identity standard (the self-meanings of an identity), (2) perceptual input of self-relevant meanings from the situation including how one sees oneself and the meaningful feedback that the self obtains from others (reflected appraisals), (3) a process that compares the perceptual input with the identity standard (the comparator), and (4) output to the environment (meaningful behavior) that is a function of the comparison (difference) of perceptions of self-meanings from the situation with actual self-meanings held in the identity 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 system can be thought of as having a goal of matching the situational inputs (perceptions) to the internal standards. What this system attempts to control is the perceptual input (to match the standard). When perceptions are congruent with the standard, identity-verification exists. (p.51)
In the cybernetic identity model, behavior is the result of the relation between perceived meanings of the self in a situation and internal self-meanings held in the identity standard. Behavior is goal-directed in that there is an attempt to change the situation in order to bring perceived situational self-meanings in line with the meanings held in the identity standard. When perceived self-in-situation meanings match self-meanings in the identity standard, the meanings of the behaviors correspond to these meanings and there is verification of the identity. However, if the perceived self-in-situation meanings fail to match, distress is felt and behavior is altered to counteract the situational meanings in an attempt to accomplish identity-verification. 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 experience distress, and she will increase the “strength” of her performance in an effort to restore perceptions of herself as strong as seen in the reflected appraisals.
The cybernetic nature of identities posited by Burke 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-in-situation meanings in line with the self-meanings held in the identity standard (Burke and Reitzes 1991). One is more committed to an identity when one strives harder to maintain a match between perceived self-in-situation 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 and Serpe 1982; 1994), but it 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 identity-verification response. The structural connection is maintained. For example, Burke and Reitzes show that those who are highly committed to a student identity (by having more ties to others as well as by 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.
More generally, research has been accumulating on understanding identities using the identity control model. For example, research has examined the emotional reactions to identity-nonverification. Even though we discuss the role of emotion more fully in identity theory in chapter 8, we briefly mention a few pertinent studies in this regard. Using a sample of newly married couples, Burke and Harrod (2005) find that, compared to partners who experience spousal identity-verification, those who experience spousal identity-nonverification report negative emotions. This is true whether the identity meanings of the spouse in the situation are overly positive or overly negative compared to their identity standard for such meanings. (p.52)
However, in a series of laboratory studies on the worker identity, Stets (2003; 2005) finds that identity-nonverification that is overly negative (workers are underevaluated relative to their worker identity) leads to negative emotions, while identity-nonverification that is overly positive (workers are overevaluated relative to their worker standard) leads to positive emotions. Stets argues that her results might be due to individuals in the laboratory responding automatically to positive feedback rather than processing the feedback more deeply, which may be more likely to occur in marriage.
Researchers have also examined who is more likely to experience identity-verification. For example, using a sample of Los Angeles County residents, Stets and Harrod (2004) find that individuals with higher status characteristics—for example whites (compared to nonwhites), males (compared to females), older individuals (compared to younger individuals), and the more educated (compared to the less educated)—are more likely to experience identity-verification across multiple identities. This helps us see one of the ways in which the external social structure (one’s position in terms of status) impinges on internal processes (the identity-verification process), thereby linking the structural emphasis (of Stryker) in identity theory with the perceptual control emphasis (of Burke).
Still other research takes into account interaction in dyads and groups and includes the idea that in many social settings there are multiple persons, each with their own identities, all seeking to have their own identities confirmed in the situation. As an illustration, Cast and her colleagues (Cast, Stets, and Burke 1999) find that in newly formed marriages, spouses with higher status (in terms of a higher education and occupational status) than their spouse are more likely than their lower-status spouses to influence their partner’s view of them (the higher-status spouses). The higher-status spouses also are more likely to influence their partner’s self-view. In this way, higher-status actors are more likely to have their identities verified than lower-status actors. Cast and her associates’ emphasis on the status of the actors and how each influences (or fails to influence) the other’s view shows us how the structural emphasis (of Stryker) in identity theory and the interactional emphasis (of McCall and Simmons) of identity theory can be merged into the perception control emphasis (of Burke).
Summary of Identity Theory
Figure 3.1 offers a “road map” on the development of identity theory. It highlights the important points regarding the roots of identity theory and the current emphases in identity theory that we have discussed. Notice the arrow going from the structural version of symbolic interaction to identity theory. The traditional version of symbolic interaction has contributed less to identity theory.
One can compare the interactional, structural, and perceptual emphases in identity theory by discussing how each explains identity performances or identity behavior. For McCall and Simmons, identity performances are a result of actors attempting to interrelate their identities with those of others in a situation. Every identity is played out in relation to a complementary identity; for example, the identity of doctor is played out in relation to the identity of patient, or the identity of salesclerk is performed vis-à-vis the identity of customer. When conflict between two identities emerges—as when a patient tells her physician that she is not listening to the patient’s symptoms or a customer complains that a salesclerk is not “helping” her—negotiation strategies and compromises are employed so that each actor’s identity claim can be confirmed and interaction can proceed smoothly. So the physician or salesclerk may apologize for her inattentiveness and work harder to help or listen.
For Stryker, identity performances or behavior is a function of how salient an identity is in one’s overall hierarchy of identities; a more salient identity is more likely to be invoked in a situation. For example, if the physician identity is salient to an individual, she will be more likely to “call it up” outside of work, as in talking about medicine and perhaps even “treating” others while at home or at a social event. One important factor that influences the salience of an identity is how committed one is to the identity. Greater commitment to an identity is a function of being tied to a larger social network that is premised on the identity and having stronger ties in that network. So if a physician can name most of her friends as physicians as well, and she has a close relationship with these friends, then the physician identity is likely to be salient for her.
For Burke, identity behavior is a function of the relationship between perceived meanings of the self in a situation and identity-standard meanings. When perceived self-in-situation meanings match identity-standard meanings, identity-verification exists, and the meanings of behavior are consistent with the meanings of the identity standard. When self-in-situation meanings do not match identity-standard meanings, behavior is modified to restore meanings of the self in the situation to correspond with identity-standard meanings, thereby moving the self from a state of identity-nonverification to identity-verification. For instance, if at a doctor’s appointment, a physician gets feedback from his patient that he is not a good plastic surgeon because the liposuction that he recently performed did not remove all of the cellulite on the patient’s legs, the physician is not being verified in his role identity. (p.55) The physician may respond in a variety of ways: he may disagree with the patient and show her how what she sees is not additional cellulite, he may agree with her and offer to perform another surgery to remove the remaining cellulite, or he may have some other response. Essentially, the responses are designed to move the physician from being in a nonverifying state to a verifying state.
The development of identity theory can be enhanced by merging Stryker’s ideas about identities at the social structural level with McCall and Simmons’s views at the interactive level and Burke’s conceptualization at the individual level. For example, in the above example of the woman responding in a manner in order to accomplish identity-verification, it would be valuable to identify the various ways in which individuals respond to identity-nonverification and the conditions under which this response occurs. McCall and Simmons provide a host of ways in which individuals can respond to disconfirmed identities such as using “short-term credit” to “ride out” an unsuccessful role performance, blaming others, or withdrawing from the interaction. These “mechanisms of legitimation”—as they label them—can be incorporated into the identity control model. However, we still have not identified the conditions under which they will occur. Relying on Stryker’s structural emphasis in identity theory, we might predict that those who have a more salient physician identity (that is, their network of social relationships involves others claiming the physician identity as well and their ties with these others are relatively strong) and who have greater commitment to that identity may be more likely to initially blame others for their nonverification. They might assume that others are not knowledgeable or skilled in their role identity and thus are not competent to evaluate them. The above illustrates how we might borrow ideas from the different emphases in identity theory in order to better predict social behavior.
By considering the macro, meso, and micro levels of identity theory simultaneously, we can expand identity theory beyond its current boundaries. By encouraging different theorists working within identity theory to look outside of their current conceptual borders and consider the influence of other pertinent sociological processes raised by identity theorists at other levels of analysis, we are in a position of developing a more general theory of identities. Essentially, there is still much to be done in identity theory, and one avenue for the future is working across the different emphases in identity theory, which, up to this point, have remained distinct in theory development and empirical testing.
Affinities to Identity Theory
Identity theory is not alone in dealing with these issues. We turn to two theories that have much in common with identity theory. One is in (p.56) sociology: affect control theory (Heise 1979; MacKinnon 1994; Smith-Lovin and Heise 1988); and the other is in psychology: self-verification theory (Swann, Rentfrow, and Guinn 2003). Affect control theory shares with identity theory its roots in structural symbolic interaction. Further, like Burke’s perceptual control emphasis, affect control theorists share the idea that the best way to model the identity process is through a cybernetic model. Self-verification theory shares with identity theory the assumption that people desire to confirm what they already believe about themselves and that persons will seek to maintain their self-views in the face of resistance. We discuss each theory in more detail below so that the reader may get a better sense of the similarities between the theories.
Affect Control Theory
David Heise, the originator of affect control theory, was initially influenced by the psycholinguistic measurement of meaning (Osgood, Suci, and Tannenbaum 1957) and the social psychology of impression formation, a largely cognitive activity. To begin, affect control theorists are interested in the definition of events and the affective reaction people have to these events. To arrive at definitions of events, they consider the actor in a situation, his or her behavior, the objects of one’s actions, and the setting. For example, an event might be, “A judge (actor) reprimands (behavior) the prosecutor (object) in the courtroom (setting).”
In order to identify people’s affective reaction to events, affect control theorists turn to Osgood and his colleagues who maintained that individuals universally respond to meaning along three primary dimensions: evaluation (the assessment of good/bad or nice/awful), potency (the level of something being powerful/powerless or big/little), and activity (how fast/slow, young/old, or noisy/quiet something is). These are called the evaluation/potency/activity (EPA) dimensions of meaning. Affect control theorists apply the EPA dimensions of meaning to each feature of an event. Thus, there would be EPA ratings for the actor, the behavior, the object, and the setting (ABOS), which are viewed as the primary elements of interaction.
In order to assess the average person’s affective reactions to each of the ABOS dimensions, Heise and his collaborators draw samples of individuals and ask them to provide an evaluation/potency/activity profile for concepts related to a variety of actors, behaviors, others, objects, and settings. For each concept, respondents use a scale that usually ranges from +4 to −4 in order to assess evaluation, potency, and activity. Zero is neutral on each of the dimensions—being neither good nor bad, strong nor weak, active nor passive. For example, in a sample of Canadian university females, the EPA profile of the actor, mother, is 2.7, 1.6, and 1.0; thus, on average, mothers are seen as very good, somewhat powerful, and slightly active (MacKinnon 1994). EPA profiles of actors, behaviors, objects, and situations comprise (p.57) fundamental sentiments in affect control theory. This is the culturally established EPA meanings of every element related to an event.
When different actors, behaviors, objects, and situations are combined into events, they create transient impressions or transient feelings. For example, “a mother screamed at her child” generates transient feelings that are negative for both the mother and the child. Ordinarily, mother and child might have EPA profiles that are, on the average, positive, but when the act of “screaming” occurs, it may change the sentiment of the actor (here, the mother) because she is doing a bad thing (screaming) as well as change the sentiment of the object (the child) who may have done a bad thing. Alternatively, “a mother hugging her child” generates transient feelings that are positive for both the mother and the child. The act of hugging is positive, and this behavior would enhance rather than diminish the positive sentiments attached to the mother (as the actor) and the child (as the object). Affect control theorists create impression formation equations that predict what transient feeling is likely to be experienced given different configurations of actors, behaviors, objects, and situations in events. The predictions take into account how behaviors and even settings can change existing fundamental sentiments. The reason affect control theorists use the term “transient impressions,” or “transient feelings,” is to recognize that once an event occurs, the feelings that it generates quickly can change as new events subsequently emerge.
Affect control theorists maintain that individuals attempt to experience events in which the transient feelings confirm fundamental sentiments. A discrepancy between fundamental sentiments and transient impressions is a deflection. When this occurs, something has to be done to bring the two into alignment. Either other behaviors are activated in order to restore fundamental sentiments or the event is reinterpreted. For example, in the above example, a restorative act for the event involving the mother screaming at her child may involve the mother apologizing to the child. Alternatively, the event could be reinterpreted if the modifier of “abusive” was included as a descriptive of the mother (thus changing her identity), or if, alternatively, rather than committing a bad act, the child committed a good act such as taking his first steps and the mother is “screaming” out of elation.
Identities take on significance for affect control theorists because individuals can take on roles and therefore claim role identities. Individuals can claim role identities as “observers” to events or as “actors” in events. The role identity meanings are understood along the EPA dimension as well. When affect control theorists focus on the identities of actors in events, the overlap with identity theory is most clearly seen (Smith-Lovin and Robinson 2006).
Theorists in both identity theory and affect control theory assume that in situations, actors try to maintain identity meanings, and that in situations, these meanings are often disturbed by the actions of others. In response, actors will emotionally react and be motivated to restore perceptions. This is (p.58) the basic control system approach which both traditions share (Smith-Lovin and Robinson 2006). However, while in identity theory, actors are trying to maintain their self-meanings held in their identity standard, in affect control theory, actors are trying to maintain meanings of all elements in the event: actor meanings, behavior meanings, object meanings, and situation meanings (ABOS meanings). In this way, while the control system in identity theory refers to the self, the control system in affect control theory refers to events.
Another difference between identity theory and control theory is in the nature of meaning (Smith-Lovin and Robinson 2006). While both Heise and Burke turned to Osgood and his collaborators as a guide for measuring meaning, Heise became interested in measuring meaning at the cultural level, and Burke became interested in measuring meaning at the individual level. In measuring meaning at the cultural level, Heise uses the EPA dimensions of meaning and applies them along the ABOS dimensions. Burke uses the semantic differential framework described above (and discussed in more detail in chapter 5) to measure individuals’ identity meanings. For Burke, other elements in the situation are left unmeasured, but they are nevertheless relevant in maintaining congruence between self-in-situation meanings and identity-standard meanings. We point out that although Heise’s approach allows us to understand cultural meanings, those meanings may not predict behavior for an individual in a specific situation. This is where Burke’s measurement procedure may garner more power. By drawing upon an actor’s own set of meanings, which are both cultural and idiosyncratic, in nature, we may be in a better position of predicting the actor’s behavior.
Self-verification theory has been developed and extensively tested by William Swann, Jr. and his students (Swann 1990; 2005; Swann, Griffin, Predmore, and Gaines 1987; Swann and Hill 1982; Swann, Rentfrow, and Guinn 2003). The theory does not specifically address identities; rather, it deals with the self more globally. However, in specific studies, Swann investigates individuals’ self-meanings when persons are in specific situations or roles. In this sense, it could be argued that he is studying identities.
In brief, self-verification theory assumes that people are motivated to verify or confirm currently held persistent self-views. People want to confirm their view of themselves as a means of bolstering the perception that their world is predictable and controllable. This desire for prediction and control is central to the theory. In people’s efforts to exert control over situations, they turn to their self-view as a guide. Indeed as identity theorists have argued, self-verification provides an emotional anchor that leaves one less vulnerable when encountering life’s events. When a person knows who (p.59) he or she is, others also will come to know and support the person, and this helps keep the person on an even keel (Cast and Burke 2002).
To facilitate self-verification, or people’s desires to confirm what they already believe about themselves, Swann suggested that individuals employ various strategies in interaction with others (Swann 1987). For example, individuals may engage in selective interaction, that is, choose to interact with others who confirm their identities and avoid those who do not (Swann, Pelham, and Krull 1989). Alternatively, they may display identity cues or lay claim to an identity by looking the part, for example, dressing a certain way or using a particular speech style so that others recognize their identity and behave appropriately, thereby confirming their identity. Individuals may also use interpersonal prompts, that is, interaction strategies that get others to behave toward them in a manner that is congruent with their identity (Swann 1987). If one receives disconfirming reactions from others, then interpersonal prompts may be used to counteract this disconfirmation. For example, Swann and Hill (1982) found that persons who thought of themselves as dominant reacted in an even more dominant fashion if they received feedback that they were submissive. And self-designated submissive persons acted in an even more submissive fashion when they received feedback suggesting they were dominant.
The above strategies imply that individuals create a verification context for themselves. This is an important assumption in identity theory as well (Burke and Stets 1999). People seek ways to establish and maintain social situations and relationships in which their identities are verified. According to identity theorists, these are self-verification contexts (Burke and Stets 1999). When one appears predictable in the eyes of others, this predictability, in turn, stabilizes the way that others respond to the self. And the stable way that others respond to the self further stabilizes one’s own self-views. Ultimately, individuals are dependent upon others to provide a steady supply of self-verifying feedback, and in so doing, a self-verifying environment develops.
What is interesting about self-verification theory is that it suggests that people prefer self-confirming feedback even when the self-view that is being confirmed is not a positive self-view. Although it seems counterintuitive that people will find negative feedback just as reinforcing as positive feedback, many empirical studies support this finding (Swann 2005; Swann, Rentfrow, and Guinn 2003 for reviews of this work). Individuals prefer others who verify not simply their favorable self-views but also their unfavorable self-views. This is especially true when individuals are given the time to access their self-views and compare the fit between their self-views and the feedback from others. If these mental resources are not made available to individuals, they will seek out others who enhance them rather than confirm their negative self-views.
In identity theory, the assumption is made that individuals desire to verify who they are even if that identity is negative (Burke and Harrod 2005). (p.60) There is nothing in identity theory that maintains that identity-verification only operates for positive identities. In the same way that people with positive self-views seek out positive feedback and positive interaction partners, people with negative self-views seeks negative feedback and negative interaction partners.
Research Given the Affinities
Researchers are beginning to look across identity theory, affect control theory, and self-verification theory and see where there may be some common ground. For example, in trying to understand individuals’ emotional reactions to verifying and nonverifying feedback in a laboratory setting, Stets (2005) turns to insights from both affect control theory and self-verification theory. As another example, although Swann has typically discussed self-verification strivings in terms of individuals’ self-views rather than identities per se, he recently argued that the self-verification motive operates for social identities (Pinel and Swann 2000). This opens the door to the identity-verification process operating for other bases of identities such as role identities and person identities, which we will discuss in greater detail in chapter 6. Perhaps a more striking example of looking across theories is recent work by affect control theorists Smith-Lovin and Robinson (2006), in which they compare both the similarities and differences between affect control theory and the perceptual control emphasis in identity theory. In comparing the two, they develop research questions on empirically testable differences that might prompt further research and advances in each theory.
In chapter 10, we discuss future directions in identity theory. One direction is linking identity theory with other theories in the discipline in order to broaden the scope of identity theory. This link is easier when theories share common ideas. Affect control theory and self-verification theory are two theories that are closer to identity theory than others that we shall discuss.