Jump to ContentJump to Main Navigation
Identity Theory$

Peter J. Burke and Jan E. Stets

Print publication date: 2009

Print ISBN-13: 9780195388275

Published to Oxford Scholarship Online: May 2012

DOI: 10.1093/acprof:oso/9780195388275.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 08 April 2020

Multiple Identities

Multiple Identities

(p.130) 7 Multiple Identities
Identity Theory

Peter J. Burke

Jan E. Stets

Oxford University Press

Abstract and Keywords

This chapter discusses the consideration of multiple identities that people have and that exist in situated interaction. It considers the case of multiple identities within a person, and then examines multiple identities within a situation. It also shows that the perceptual emphasis in identity theory conceptualizes identities as hierarchically arranged in a control system of identities instead of a prominence hierarchy.

Keywords:   multiple identities, situated interaction, perceptual emphasis, identity theory, control system of identities, prominence hierarchy

Much of our discussion about identities up to this point has been on what identities are and how they work. Moreover, although we have mentioned that people have multiple identities and that there are many identities from many people in operation in any situation of interaction, we have not considered the implications of this. Identities do not always operate in isolation, but they interact with other identities in particular situations. Thus, we must consider the importance of this observation for understanding more clearly some of its social implications. For this, we break the problem into two parts and first consider the situation in which different identities are held by the same person. Second, we consider the situation in which different identities are held by different persons interacting together in a social setting.

An example of the first is a situation in which an adolescent, who is both a friend and a daughter, is interacting in a situation that activates both of these identities at the same time, for example, when the daughter has a friend visit and her parents are present. In this situation, what it means to be a friend (perhaps by acting “sophisticated” with the friend) may be at odds with what it means to be a daughter (not acting “sophisticated”). An example of the different identities being held by different persons would be two persons, a professor and a student, interacting in a situation. One of the persons has a professor identity and the other person has a student identity. These identities may be at odds if what the student expects of the professor (give me a passing grade so that I can stay in school) differs with the identity of the professor (give grades according to merit and fairness with other students). In each of these cases, it may or may not be possible to verify all the identities involved at the same time. The consequences of these (p.131) possibilities are numerous. We consider first the case of many identities within a person.

Multiple Identities within a Person

As we mentioned earlier in the book, William James (1890) was among the first to note that we have as many “selves” as we have others with whom we interact. The idea of multiple selves has changed somewhat since James wrote, and we now talk about identities rather than selves, but the basic components of the concept have remained. We take on many identities over the course of a lifetime, and at any point in time we have many identities that could be activated. A person could be a student in one context, a friend in another, a mother, a daughter, a teacher, a blood donor, a homeowner, and so on. Each of these identities acts to control meanings/resources in a situation, such that relevant identities are verified.

When considering these multiple identities within a person, it helps to distinguish two approaches that might be taken. We can look at these identities from an internal framework or from an external framework. The internal focus attends to issues of how an individual’s multiple identities function together within the self and within the overall identity-verification process. How does the identity-verification process work for the multiple identities? The external focus addresses how the multiple identities that an individual has are tied into the complexities of the social structure in which the individual is embedded. Here we need to consider one’s commitment to the multiple identities and the way each identity ties one to a particular location in the social structure. We turn to each below.

The Internal Framework: Multiple Identities within the Perceptual Control System

When considering the relationship among identities within the hierarchical perceptual control system in which all identities reside, questions arise as to how the multiple identities relate to one another, how they are switched on or off, and, when they are on, how the person manages to verify each. Included in this last part is the question of what happens when one or more of the identities cannot be verified.

Prior work on the relationship among multiple identities within a person has focused on the way in which the multiple identities do or do not share meanings. For example, Linville’s (1985; 1987) self-complexity theory dealt with the idea that individuals with more complex selves were better buffered from situational stresses. The complexity of the self was defined as the number of “distinct” self-aspects one has. By distinct aspects we mean the number of roles, relationships, traits, or activities that do not share attributes or meanings. The suggestion was that if one self-aspect or identity (p.132) has problems, such problems will not spill over to other aspects, because they do not share meanings. Identities that share meanings, then, from their viewpoint, are potentially problematic because problems can spread from one identity to another along lines of common meaning.

Where do such complex selves come from? Stryker (1980 [2002]) suggests that the complexity of the self is a reflection of the complexity of society. As society becomes more differentiated in terms of groups, organizations, and roles available to persons, persons who take on more of these as identities become more complex themselves. The premodern self was generally simpler (with fewer affiliations and, therefore, identities) than the postmodern self. Further, because the networks of the premodern era were more embedded within one another, there was more sharing of cultural meaning and expectations (Pescosolido and Rubin 2000).

Consequently, the different identities in the postmodern era have less common or shared meanings compared with those in the premodern era. Thus, not only are there more identities available, but also they have less in common with one another. This means that the verification of these multiple and disparate identities calls for the manipulation of more and disparate meanings across more and disparate settings. Common meanings shared across identities would facilitate identity-verification, as there would be fewer different meanings to control. It would make sense, then, that people would select to occupy positions in the social structure that, as much as possible, shared meanings, thereby taking on identities that shared meanings.

Stets (1995) looked at the relationship between gender role identity and mastery, a person identity, and suggested that the two are related through a common dimension of meaning concerning the degree to which the person controls aspects of his or her environment. Because of this common meaning, she suggests, enactment of one identity has implications for the other, since each is controlling the same shared meaning. Deaux (1992; 1993) also proposed this idea of common characteristics shared among social and personal identities. She used the concept of common “traits,” though one could substitute the idea of common meanings with the same effect.

Shared among all of these researchers is the idea that identities that have common meanings are likely to be activated together whenever those meanings are present in the situation and that multiple identities might work together in the identity-verification process to control those meanings in the situation or might suffer together when there are problems dealing with the shared meanings. Additionally, Deaux suggests that identities that share many meanings are more prominent or important because of that sharing, and such identities may work together to control the meanings of identities lower in the prominence hierarchy. Identity theory further develops the idea of shared meaning and the extent to which one identity may control another, but it does so within the context of the hierarchical control system, and we turn to that now.

(p.133) Salience, Prominence, and Commitment

The salience of an identity is the likelihood that it will be activated. Identities that are more salient are more likely to be activated in any situation. By being activated, you recall, we mean that an identity is attempting to verify itself. This means that perceptions of relevant meanings are being made, the comparator is assessing the degree to which the perceived meanings match those in the identity standard, and the comparator is sending an appropriate signal to the output system. Depending upon whether or not the identity was being verified, this output would continue behavior unchanged (no error) or modify the social behavior. Behavior would be modified to counteract any disturbances and alter the situational meanings in order to reduce the error or difference between their perceptions and the identity standard. Although a person may have multiple identities, if only one of them is activated at any point in time, then, for that period, they may as well have only the one identity.

If more than one identity is activated in a situation, we expect that the identity with the higher level of prominence, or the identity with the higher level of commitment, will guide behavior more than an identity with a lower level of prominence or commitment. As McCall and Simmons (1978) suggest, performances strongly suggested by more prominent (important) identities are more likely to be carried out than are those suggested by less prominent identities. If one identity is more important than another is, then verification of that identity is more important than verification of another. In a similar fashion, an identity that has more commitment than another (that is, more other people depend upon that identity than the other), that identity is more likely to be verified, thus fulfilling our commitments to the many rather than to the few. Thus, prominence and commitment not only influence the level of salience of an identity but also help sort out the question of what to do next when multiple identities are activated. If one identity or another has to wait for verification because we cannot work on all of them at once, those that are less prominent and those that have lower commitment are the ones to wait. In this way, identities can be distinguished and compared in terms of their prominence and level of commitment. However, identities are also related more directly in terms of the hierarchy of control in which they are embedded.

The Hierarchical Control System

To understand this idea of the hierarchy of control of meanings that is central to identity control theory, we need to understand the hierarchical nature of the overall perceptual control system in which identities are located. The overall perceptual control system, as we noted, is composed of an interlocking set of individual control systems at multiple levels (Tsushima and Burke, 1999) such as that depicted in figure 7.1. To facilitate the presentation, we (p.134)

                   Multiple Identities

Figure 7.1. Model for Three Identities within a Person

will first discuss the relationships among identities located at the same level in the overall control system, such as those labeled B and C in the figure. We will then discuss the relationship among identities at different levels in the system such as those labeled A and B in the figure.

At the lowest level of the hierarchy, we consider the two identities labeled B and C in the figure. Each of the identities has its own standard and its own perceptions of meanings, and each modifies its own output (behaviors in the situation) to verify itself by keeping the perceived situational meanings in agreement with the meanings in its standard in the manner we discussed in chapter 4. Although we consider only two identities for the person, the ideas can be generalized to a larger number of identities that might be labeled D and E and so on. Because all of these identities exist within one person, we note that the output of all of these identities must combine to control the social behavior of that individual. The individual’s behavior controls situational meanings to make them congruent with the meanings held in all of the identity standards. That is, each identity is controlling meaning by adjusting the behavior of the same individual. Thus, although there are possibly many identities, there is only one behavioral output stream because there is only one person to act. This implies that the behavior of an individual must “satisfy” several individual identities simultaneously by altering the situation in ways that change all of the self-relevant meanings perceived by all of the different identities. If a person has the identities of professor and spouse and both are activated, the person must adjust perceived meanings to confirm or verify (p.135) both the professor identity and the spousal identity.

This process occurs for all activated identities in a situation, so that perceptions of all of the self-relevant meanings of all of the activated identities are simultaneously controlled. For this to happen, all of the meanings must be either unrelated or aligned. They cannot remain in opposition. Were, for example, two self-relevant meanings perceived in the situation to be in opposition with each other, as one was brought into alignment, the other would be moved out of alignment. One cannot both be good and bad, for example, or both strong and weak. When different identity standards require oppositional meanings, as in our above example of the daughter/friend trying to be “sophisticated” to her friend, but not so sophisticated to her parents, the system is put into an impossible situation in which one or both identity standards cannot be verified. To the extent this happens, the identity standards themselves must shift as people’s identities change to remove the conflict. People re-identify themselves, changing the self-meanings held in their identity standards. To understand this idea of identity change and how it works, as we will in chapter 9, however, we must understand the hierarchical structuring of identities. We will discuss that in the next section.

When the meanings held in different identity standards are unrelated to one another, an action that changes meanings in the situation to verify one identity will leave the other unaffected. For example, getting an ovation for an excellent scientific talk may verify one’s scientist identity but may be irrelevant for one’s spousal identity.

When two identities share common meanings, the situation is much simpler. Control of the situation to change self-relevant perceptions on the shared dimension of meaning helps both identities. Verifying one of the identities will help verify the other and the two identities can coordinate their outputs to verify both. For example, consider a married person with children. If the spousal identity includes standards for providing material support for one’s spouse, and if the parent identity includes standards for providing material support for one’s children, then getting a well-paying job will help verify both identities.

Levels of Control

We move now to identities that are above those at the lowest level in the hierarchy. In doing so, there are additional ways in which identities relate to one another. At levels other than the “lowest,” the outputs of each individual control system provide the standards for identities that are at lower levels in the control system. Figure 7.1 shows two identities within a person that are hierarchically arranged. Note that the output of “higher” identity (A) is the standard of the “lower” identity (B). Although the lower identity follows the model we have been discussing, with its perceptions, standard, comparator, and output of social behavior that matches the meanings of perceptions and (p.136) standard, the higher identity differs in that it does not control social behavior directly. It has its perceptions, comparator, and standard, but its action is to control the standard or goals of the “lower” identity. The lower identity acts to match the meanings in the situation to those meanings held in its standard (its goals), but those meanings are set by the higher identity. Thus, what the lower identity actually does depends upon the goals set by the higher identity. Since the higher identity in some sense controls the lower identity, they cannot be in conflict—the lower identity is the servant of the higher identity. The higher identity does not tell the lower identity how to verify itself; it only tells the lower identity what meanings need to be verified.

Consider, for example, the identity professor as the lower identity. In addition, consider one’s gender identity as the higher identity. How one verifies the professor identity would depend upon the setting of one’s gender identity. Being male provides a set of meanings for the professor identity that are different from the meanings for the professor identity that are provided if one is female. The professor identity standard, or what it means to be a professor, varies by the kind of gender identity one has.

There are many possible arrangements for the identity at the higher level. The output of one higher-level identity may serve as the standard for several lower-level identities. In this way, all the lower-level identities can be seen as in the “service” of the higher-level identity, and the lower-level identities are thus coordinated in their endeavors. For example, the way in which I am a professor and the way I am a parent are both influenced by a higher-level gender identity. Indeed, as Burke (2001) has shown, the coordination needed by different identities, each seeking their own verification, can only be provided by having a common higher-level identity.

It is also possible that several higher-level identities contribute their outputs to a single standard for a lower-level identity. This was the model Burke (1997) used in creating an identity model for network exchange. Several higher-level identities, each controlling a different set of meanings in the situation, had their outputs combined into a single standard for a lower-level identity that controlled negotiation behavior. Finally, it is possible that these arrangements can be combined in various ways with multiple higher identities controlling multiple lower identities in the overall control system.

Levels of Perception

We discussed the hierarchy of control in which higher-level identities control lower-level identities by slowly adjusting the standards of the lower-level identities. This is what one might call the output side of the system. On the input side, perceptions are also hierarchical, and, since it is the perceptions that are controlled, perhaps we should have discussed this side of the control hierarchy first. What makes some control systems “higher” than others? Is it possible that the higher and lower control systems in the hierarchy could (p.137) switch places? How do the levels differ? The answer to these questions lies in the nature of the perceptions that are controlled. Perceptions that are higher in the control hierarchy are more complex in that they consist of patterns of perceptions at lower levels. For example, perceptions of lines and angles are at a lower level, while perceptions of squares and triangles are at a higher level, consisting of patterns of angles and lines. Controlling perceptions of squares and triangles necessarily involves controlling perceptions of their lower-level components, angles, and lines. It is this relationship between the “parts” and the “whole” that distinguishes levels in the control system. Control of perceptions of “wholes” occurs at a higher level than control of perceptions of “parts.” Control of perceptions of wholes is achieved by controlling the lower-level perceptions of parts. The levels cannot be reversed.

Following this logic, identities at a higher level are more general than identities at a lower level. Tsushima and Burke (1999) examined parent identity standards that exist at two different levels, which they called a principle level (higher) and a program level (lower). They indicated that both the perceptions and the standards that exist higher in the hierarchy are more abstract and more general. These higher perceptions and standards organize perceptions and standards that are lower in the hierarchy and that are more concrete and situated. Principle-level standards are conceptualizations of abstract goal states such as values, beliefs, and ideals. Program-level standards of the parent identity are more concrete goals accomplished in situated activity, such as going to the store or making sure the children get off to school. In their study, Tsushima and Burke (1999) interviewed fifteen married and fifteen single mothers with children in elementary school. Questions for the interview were organized in three areas concerning standards for education, for discipline, and for the parent identity in general. Coders were trained to examine the interviews and classify the responses discussing each standard in terms of whether the standard was at a principle or program level. Tsushima and Burke discovered that parents were primarily oriented either toward program- or toward principle-level standards for parenting, and that it was possible to classify parents as to being mainly oriented toward principle-level or program-level goals.

Tsushima and Burke provide examples of goals or standards at each of the two levels to help illustrate this rather abstract discussion. With respect to education, some parents spoke of begging, forcing, or harping on their children to do homework. Begging, forcing, and harping are programs of activity that these parents used with the goal of getting the children to do their homework. With respect to discipline, programs including confrontation, setting timers, and grounding the children were mentioned, each with respect to some situational disturbance created by the child. Each of the different programs of activity with respect to education and discipline were used to accomplish the lower-level goals of getting the children to act in particular ways within particular situations. They were immediate, short-term ways of dealing with immediate problems.

(p.138) In contrast, other parents had parent identity standards that provided higher-level meanings/goals for education and discipline. These higher-level standards were higher in that they helped organize and relate various lower-level programs. For example, one parent spoke of her educational goals being to challenge her children, to be a good teacher. With respect to discipline, another parent said that it was important to listen to her children, to give them love, and to be consistent. These higher-level standards do not prescribe particular activities or programs of activity, but they do provide ways to evaluate programs of activity in terms of their ability to achieve the higher-level goals. These are longer-range goals that are achieved over time by selecting more immediate situational programs of activity that are consistent with the long-range goals.

These examples illustrate different levels at which identities and identity standards exist in the control hierarchy. Perceptions at the higher levels are composed of patterns of perceptions at the lower levels. Control of perceptions at the higher levels involves controlling the patterns of perceptions at the lower levels. The accomplishment of programs of activity by the parents involves controlling perceptions of activities and meanings such that the programs are carried out. For some parents who had not developed parental identity standards at higher levels, this was all they could attempt. Managing perceptions at the higher level of general principles and values involves seeing patterns at the program level and controlling those patterns. The higher-level standards are met when perceptions indicate that the lower-level programs are functioning in such a way to match the patterns indicated at the higher level.

Although this discussion has been in terms of lower- and higher-level standards for the parent identity, the same principles hold for identities in general. For some parents, the parent identity consisted of particular programs of activity; but for other parents, the parent identity consisted of general principles that guided all of the programs in which they might engage. In a similar fashion, one’s moral identity, which is a higher-level identity, would have standards that are met when perceptions of patterns in the programs of activity in any of the many role or social identities match those high-level standards (Stets, Carter, Harrod, Cerven, and Abrutyn 2008). Perceptions of the meanings of actions as a spouse, as a worker, as a parent, or as Rotary club member all contribute to one’s perceptions of oneself as a moral person. On the output side, being a moral person is realized by one’s actions in all of these roles. On the input side, the pattern of perceptions of accomplished meanings in all these roles contributes to the perceptions of oneself as a moral person.

The hierarchical organization of identities in these various ways within the perceptual control system helps to maintain self-relevant perceptions close to their identity standards at all levels simultaneously. This general principle holds for all identities at whatever level. All of the comparators shown in figure 7.1 act to measure the difference between perceived (p.139) meanings and those meanings held in the identity standards. The outputs of these comparators (each a function of the difference between perceptions and standards) work either to change behaviors in the situation for the lowest-level identities, or, for higher levels, change the identity standards for identities at lower levels. This is the identity-verification process, which is a dynamic, ongoing, continuous process of counteracting disturbances that occur in the situation. Such disturbances may be the result of others’ behaviors in the situation, one’s own behavior in the situation, or ongoing physical processes in the situation.

Through the identity-verification process, actions (output behaviors acting on the meanings and resources flows in the environment) are taken that alter the situation and hence the self-relevant meanings in that situation to bring them into congruence with the standards held in the identity. To the extent that identity-verification succeeds or fails, and perceived self-relevant meanings are or are not brought into congruence with their standards, three concurrent outcomes result. Emotional reactions that are positive occur when the discrepancy is decreasing or nonexistent, and negative emotional reactions occur when the discrepancy exists or increases. We talk more about this in the next chapter. At the same time of this ongoing emotional response, the output of lower-level identities brings about behavior that changes the meanings in the situation. This moves one’s perception toward the standard, and higher-level identities change their outputs (changing meanings held in the identity standards) to move the standard of the lower-level identity toward the meanings perceived at that level (though at a much slower rate). In the longer run, the identity system moves toward congruence between perceptions and the identity standards at all levels through the operation of both mechanisms.1

The External Framework: Multiple Identities within the Social Structure

This last idea of the nexus of identities within a person leads naturally to our other perspective of looking at the multiple identities as multiple ties to the social structure. Sociology has long conceptualized persons as occupying multiple positions or roles within the organized matrix of social relations we call the social structure (Linton 1936; Merton 1957; Parsons 1949; Turner 1978). That these multiple positions may come into conflict with one another within the individual was a logical next step that has been explored in a number of ways using the ideas of role conflict (Gross, Ward, and McEachern 1958), role strain (Secord and Backman 1974), and status inconsistency (Jackson 1962; Jackson and Burke 1965; Lenski 1954).2

Although this work was conceptualized in terms of the multiple positions that people hold and the multiple sets of expectations held for them rather than the multiple identities of those individuals, a translation to the latter perspective is straightforward. Indeed many of the effects of role conflict or (p.140) status inconsistency (for example, to create distress of one sort or another) only make sense when the individual cares about the conflicting expectations. They would care if they have internalized them as standards for themselves. This idea is very close to our current understanding of the concept of identity and the conflicts that can arise between them when the different identities attempt to control the same meanings to different contrasting set points held by the different identity standards as discussed above.

The viewpoint that identities are tied to social structural positions (i.e., that individuals’ memberships and roles in the groups, organizations, and networks to which they belong form the basis of many of their identities) grows out of the ideas of structural symbolic interaction theory (Stryker 1980 [2002]). This perspective suggests a number of ways in which the identities may relate to oneoanther in terms of the way in which the positions are connected within the social structure. From this perspective, we note three different conditions: (1) persons may have multiple role identities within a single group, (2) persons may have the same role identities but in different groups, and (3) persons may have different role identities within intersecting groups.

For each of these three conditions, we are talking about multiple identities that are simultaneously activated. If an identity is not activated, it has no effect, since no identity-verification is taking place and no behavior is being used to control perceptions relevant to that identity. For this reason, the abstract concept of multiple identities makes sense when there may be multiple identities for people to call up (activate) on different occasions. The overall number of identities held by a person may influence the likelihood that activated identities may conflict with one another. It is likely that, aside from sleep or otherwise being inattentive, one or more identities is always activated, providing guidance for our perceptions and behaviors. The question then is what are the conditions for multiple identity activation?

Multiple Identities within a Single Group

There are two forms of this condition. One may have several roles within a group (e.g., husband, father, son, and brother within an extended family reunion, or task leader and social emotional leader in the same task group). In this case, these different identities will often be concurrently activated, and this will have both internal and external consequences. Internally, we hypothesize that the concurrent activation of these identities will lead, over time, to the identities developing similar levels of salience and commitment for the individual, as well as developing shared meanings. We suggest that the similar levels of salience will develop because the identities are often activated together, and with similar levels of activation, they should develop similar levels of salience or the probability of activation. Those that are activated together frequently will become more salient; those that are activated together less frequently will become less salient.

(p.141) There are also external consequences. For example, with respect to commitment or the number and depth of ties to others in terms of the identity, when identities are activated together, they are activated in the presence of a common set of others present in the situation. Thus, these others will come to know the person in terms of those identities and develop ties with the person in terms of those identities. The more (or fewer) persons that one knows in terms of these commonly activated identities, the greater (or lesser) is the level of common commitment to the identities.

With the identities being concurrently activated, we also expect that they will develop common, shared, or overlapping meanings. Others in the group in which the multiple identities are commonly activated are likely over time to develop expectations about the way one engages in behavior relevant to each of the identities, such that the meanings defining the identities are consistent, shared, and mutually reinforcing. In this way, each of the identities becomes tied to the overall structure and functioning of the group in ways that make the verification process for all of the identities much more coordinated and shared. For example, if one person has the role of secretary and the role of treasurer in a group, and these two identities though initially separate are activated frequently together, the two will come to share meanings and perhaps merge into a single role. However, if there are times when the two roles cannot be activated together, or if there are restrictions on their being performed concurrently, such a merger is less likely and the sharing of meaning will be less.

Indeed, this was found by Burke (2003) in a study of task leadership identities and social emotional leadership (maintaining interpersonal relations) identities in small groups. Each group member was conceived to have some level of both of these identities. Some persons saw themselves as generally high on task leadership; others saw themselves as generally low or somewhere in the middle on task leadership. Similarly for self-views (identities) on social emotional leadership, some saw themselves as being high, and others saw themselves as low. These two identities were independent across persons, so that knowing how persons viewed themselves in task leadership terms did not predict how they saw themselves on social emotional leadership terms. Burke found that these identity meanings were predictive of the kind of social behavior group members engaged in along the same dimensions of meaning. Those who generally saw themselves as low on task leadership did not perform much task leadership activity, while those who saw themselves as high on task leadership did so. Similarly, those who generally saw themselves as high on social emotional leadership did perform more social emotional leadership activity, while those who saw themselves as low did not do so. Further, there was evidence that when identity and performance did not match, performance was adjusted upward or downward over time to bring about a closer match. For example, those who saw themselves as high on task leadership, if initially performing at low levels of task leadership (a mismatch), over time increased their level of task (p.142) leadership performance (bringing about a closer match between identity meanings and performance meanings).

Because the two identities were independent across group members, one would expect that the activities would be independent across persons. And that is what Burke (2003) found for persons who were not legitimated by external authority as coordinators for their groups. However, for persons who were designated as coordinators for their groups, the two types of activity were positively correlated. Persons who were high on task leadership activity were also high on social emotional leadership activity. Being in the coordinator role apparently changed performance expectations compared to expectations for persons not in the coordinator role. It was expected that the coordinator would engage in both types of leadership somewhat independent of the level of their self-view on each of the dimensions, while such expectations were apparently not held for persons who were not in the coordinator role. Coordinators therefore tended to do both types of leadership or neither type.

In the short run, this change in leadership performances for coordinators may have had little effect. However, if such coordinated task and social emotional leadership performance persisted over longer periods, we would expect that the self-views of the coordinators would change to become more consistent with their performances as constrained by expectations of the group members. They would come to see themselves as high on both dimensions or low on both dimensions or somewhere in the middle. There would come to be a blurring of the separation of the two types of leadership for coordinators as their identities changed.

A second form of multiple identities within a single group exists when a person has an activated identity in one group and something in the situation activates an identity that the person has in another group. For example, I am in a board meeting as the director of a company division and something calls up my golfer identity or my spouse identity. This is the perspective that Stryker (2000) seems to suggest in his analysis of competing identities within the context of social movements. He raises the question of why all members of a social movement are not equally committed and active within the movement—a question that is not often raised by persons studying social movements—but a question that is certainly relevant to an activist within the movement. By taking an identity perspective, Stryker points out that persons have multiple identities, which need to be verified, and people have different levels of commitment and salience for each of the identities, with the result that people spend more or less time in each identity. People may have, for example, spousal identities or worker identities that preclude the full participation in a social movement. With two or more identities activated, the person engages in behavior that attempts to verify, to whatever extent possible, all of the activated identities, with the result that each is influenced by the requirements of the other.3

As another example, consider Harriett, who is a professor, at a faculty meeting, and her faculty identity is activated. At that meeting, someone (p.143) mentions having seen and enjoyed a movie playing in town. This information is relevant to Harriett’s spousal identity, as she and her husband had been contemplating seeing the movie. As a result, Harriet’s spousal identity becomes activated, and she stores away the information about the movie to convey to her spouse later. However, while her spousal identity is activated, she is engaging in identity-verification processes with respect to that identity as well as the faculty member identity, with the result that she may attempt to speed up the faculty meeting in an attempt to accompany her husband to the movie that evening. In this way, each of the activated identities may influence the salience and commitment of the other, with the likelihood that the more salient identity will have greater influence on the less salient identity.

Multiple Identities Based on a Common Role within Multiple Groups

Here we consider the case in which the multiple identities are multiple in a sequential sense and not necessarily activated at the same time. For example, a person may have the identity “friend” in separate nonoverlapping groups or “treasurer” in several nonoverlapping voluntary associations. Each of these role identities resides in the same individual, and many of the meanings of these identities are shared, having arisen in a common culture. That is, what it means to be a friend of A is similar to what it means to be a friend of B. As a result, we would expect that any differences in the different friend identities a person has would diminish over time. Parsimony would argue that where the friend identities did not have to be different, they would become alike, reducing the information load for carrying around different expectations and meanings. Additionally, we would expect that such identities would become highly salient because of the extended network of connections to others through the identities (commitment). That is, as a number of studies have shown, increased commitment to an identity leads to the increased salience of the identity (e.g., Stryker and Serpe 1982).

Multiple Identities in Intersecting Groups

Here we consider the case in which the different identities that a person has in different groups become simultaneously activated if and when the different groups come into contact or overlap in some way. As in our earlier 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 person’s home while her parents are present. Within this situationally aggregated set of persons, both identities will be activated, and sets of meanings and expectations from both identities will be relevant. This is often the situation when role conflict is present. The meanings and expectations for each identity come into conflict when both identities are activated. Under such (p.144) conditions, we would expect the identity standards involved to shift meanings, with the more salient or more committed identity shifting the least. However, to the extent that the identities share meanings, these meanings and expectations should reinforce each other rather than be in conflict. The increased number of ties in the larger network of others (parents and peers) will increase the level of commitment to the shared meanings and hence the identities that share those meanings.

The Person As a Container for Multiple Identities

Finally, it should be noted that having multiple identities also creates a nexus of those identities that are affected by the fact that a single individual holds them. Events and conditions that affect the individual have the capacity to affect all the identities held by that individual. For example, the individual may become overwhelmed by events with respect to one identity and suffer performance degradations with respect to otherwise unrelated identities, as when work suffers while an individual is going through a divorce or an individual who is sick may have trouble verifying any of their identities.

There is another way in which the common nexus of identities within a person becomes important. These identities may relate to one another through the transfer of information or other resources from one identity to which they are available to another identity held by the person. A simple example is that as a professor, I receive a salary. Some of this salary may be used by the professor identity to maintain that identity, but more of it is used by other identities, such as spouse or parent or homeowner to verify them. Because these identities are housed in one person, they can easily communicate with one another to make available these resources. This transfer of resources is sanctioned by society. However, other resources I gain as a professor (for example a laptop computer or a video camera) are not to be made available to other identities I have. Society has developed elaborate rules that indicate appropriate and inappropriate transfer of resources among identities held by one person. And the development of these rules is brought about because the multiple identities held by individuals do transfer resources between themselves. Misappropriation of funds, embezzlement, insider trading, and making personal long-distance calls at work are other examples in which the multiple identities held by persons inappropriately transfer resources among themselves.

Not all interidentity transfers are inappropriate. As we mentioned, using a salary in one identity that has been earned in another identity is appropriate. Using information one learns from a friend to plan an outing with one’s spouse and children is also appropriate. Using the skills one learns as a student while on a job is not only appropriate but also expected. The person is the common nexus of multiple identities, and through the person, those identities have opportunities to communicate, plan, share resources, and otherwise facilitate their mutual verification.

(p.145) Because people are the carriers of identities, identities meet and interact when people meet and interact. One can meet others because of a shared group membership (for example, belonging to the same union) or role relations (for example, doctors meet patients, nurses, and drug salesmen). Because individuals hold many identities, when one individual with a certain identity meets another because of the context of that identity, the identities of the two persons other than the ones that brought them together may become relevant and activated in unexpected ways.4

Consider, for example, a person who has a daughter in kindergarten. She may meet the mother of her daughter’s kindergarten friend while picking up her daughter at school. At first, the two adults may know each other as the mother of her daughter’s friend. Over time, however, other identities may become activated and known to each other. One may know, through her work identity, of a job opportunity that the other, with her work identity, could fulfill. Through this mechanism, identities become interconnected in somewhat random ways, both within and between people. This random character results from the structural arrangements of society, the connection of individuals to those structures, as well as the multiple identities that are housed within any given individual. Highly salient identities become activated and known to others who may then find additional ways to relate to the person through activating other identities of their own. Thus, the network of relations expands as identities find new ways of verifying themselves by activating relevant identities in other individuals. Lovers may meet at church, skiers at work, friends in voluntary association meetings, and coworkers at fraternal gatherings.

An Ecology of Identities

Smith-Lovin (2003) points out that people live in an ecology of identities that are made available by the culture we live in. In preliterate or premodern societies, there may be only a few identities that are available for individuals; for example, wife, mother, sister, and gatherer may encompass most of the identities available to a woman in such a society. In societies that are more complex, there are thousands of different identities available as roles, groups, categories, and divisions proliferate. The likelihood of having multiple identities activated would certainly depend upon the number of identities one has, and that depends, in part, on how many are available to us. So, clearly, having multiple activated identities is more likely in modern or postmodern societies than in premodern societies. However, Smith-Lovin (2003) also points out that the ecology of identities is not spread equally throughout complex societies. Children, for example, have fewer identities available to them than adults have, and adults who are higher up in the status structure have more identities available to them than adults lower in the status structure have. Persons who have more resources available may use those resources to access more identities. We know, for example, that (p.146) persons with higher education (both a resource and a source of status) are more likely to join one or more voluntary associations, thus adding to the number of identities they have.

Identities Provide Meaning

One of the early views of identity that grew out of the symbolic interaction framework, with its emphasis on symbols and meaning, is that identities provide “meaning” for individuals’ lives. Though it was never clear what “meaning” was in this context, it was good to have meaning and not good not to have meaning. A life without meaning is a life that is full of anomie; it has no purpose, no structure, and no framework. Without meaning, people have low self-esteem. Thoits (1983; 1986) and a number of other researchers suggested that identities provide a sense of purpose and meaning in life, integrating us with the actions and expectations of others. Identities do this because they define who we are as well as how and why we are to behave in normatively specified ways. Identities thus increase self-esteem and reduce depression and anxiety (Thoits 1983). From this view that identities provide positive “meaning,” it was a natural step to suggest that if one has more identities, one would have greater meaning, with the consequence of increased self-esteem and reduced depression and anxiety. More identities are, from this point of view, better.

Yet, as Thoits (2003) points out, this view runs counter to much earlier work on role conflict that more roles (identities) lead to greater chance of conflict between and among the roles (Bailey and Yost 2000; Goode 1960; Merton 1957). The competing demands of multiple roles, they argued, increased the distress of individuals. This idea was the core of what was known as stress theory. And, indeed, not all the research was fully supportive of the idea persons with more identities had higher self-esteem or lower distress. Often persons with more role identities had higher distress. At the same time, identity accumulation, or taking on more identities, did seem to benefit persons when considering some identities but not others.

To better understand this set of conflicting findings, Thoits (2003) suggested that it is important to distinguish between two different types of roles or identities: voluntary and obligatory. Voluntary identities, she argues, are those that are freely chosen by individuals and are easily exited. Obligatory identities are more compelling and, when taken on, are difficult to exit. Membership in a voluntary organization would be an example of a voluntary identity, while spouse, parent, and worker are examples of obligatory identities. By making this distinction, Thoits was trying to separate identities that could lead to distress from those that may be more beneficial. The idea, in part, was one of agency or motivation. By freely choosing some identities (voluntary), presumably a person would take on only those for which there was some personal benefit. And, for these identities, if problems began to ensue, the person could exit the identity and not suffer the consequences (p.147) of increased distress. The net result, of course, is increased benefit for those voluntary identities we maintain.

Another aspect of voluntary identities is suggested by Thoits (2003). People who are able to choose their identities are persons who have control over their lives compared to others who are induced by life’s circumstances into particular identities. Being in control of one’s life is, of course, a form of efficacy and is one basis for having high self-esteem. Thus, the cards are stacked for people to have high self-esteem by having identities that they choose, and among these identities, perhaps more is better.

This is what Thoits (2003) found, considering voluntary identities in contrast to obligatory identities: persons with more voluntary identities did seem to have greater self-esteem, greater mastery, and lower distress. Further, examining the data over time suggested that persons who gained voluntary identities had their self-esteem increase. Correspondingly, those who lost voluntary identities had their self-esteem decrease.

Identity theory has an alternative interpretation for both these results as well as for the issue of the effects of identity accumulation on self-esteem and depression. As Cast and Burke (2002) point out in their theory of self-esteem as we discussed in chapter 5, from the point of view of identity theory, it is the verification of identities that makes people feel good in general and feel good about themselves especially. Conversely, if identities are not verified, people feel bad, distressed, and angry (Burke 1991; 2004b; Burke and Harrod 2005). It does not matter whether the identity is a voluntary identity or an obligatory identity, whether it is a role identity, a person identity, or a social identity. The general principle is that identity-verification leads to good feelings and problems with verification leads to bad feelings. This general principle has been confirmed many times. Further, it makes sense that the more identities that are confirmed, the better one feels. Thus, accumulating multiple identities should have benefits, but only if the accumulated identities are all verified. If the accumulated identities are not verified, then there should be increasing distress, depression, and lower self-esteem.

We suggest that the voluntary identities a person holds are more likely to be verified, but not necessarily because of any intrinsic merit of them as opposed to obligatory identities. Rather, if an identity is being verified, and that makes us feel good about ourselves, we will want to continue in that identity. Whether it is more or less difficult to exit is not relevant, because we want to stay in it. We are successful at it. However, if we are not able to verify an identity, or if we are having a great deal of trouble verifying an identity, we will want to leave.

Cast and Burke (2002) showed, for example, that when couples verify their spousal identities, their feelings of self-esteem and self-efficacy are increased and their distress and depression decreased. Further, they found that people who have verification difficulties that persist over time are more likely to become separated or divorced. This, of course, results in giving up the spousal identity. Thus, one way to deal with identities that are not (p.148) verified is to leave them. However, if the identity is difficult to exit (like marriage), then fewer will leave the identity when there are verification problems. The result will be more problem (unverified) identities in the obligatory identity group than in the voluntary identity group and as a result more people with lower self-esteem in the obligatory identity group than in the voluntary identity group, which is consistent with what Thoits finds, but not for the same reason. The accumulation of multiple identities would appear, then, to have benefits only if they are verified. However, accumulation of verified identities would be good whether they were obligatory or voluntary identities.

Indeed, the accumulation of verified obligatory identities may be better than the accumulation of verified voluntary identities. The reason is that obligatory identities have another characteristic in society. These identities are difficult to leave because society makes it so. Moreover, society makes it so because it is important to society that there be continuity with respect to these identities. One has to think twice before taking on one of these identities because it is difficult to leave. This adds a level of importance to these identities in culture, and for individuals there is a built-in bias that such identities are taken as important, that is, have high prominence. We suggest that the verification (and nonverification) of prominent identities has more impact on emotions and feeling-states than the verification (or nonverification) of less prominent identities. Verifying an important identity should result in stronger feelings of self-worth or efficacy than the verification of a less important identity. Similarly, the failure to verify a prominent identity should have stronger negative consequences. This would lead us to expect that verification of obligatory identities should produce higher self-esteem than the verification of voluntary identities.

Multiple Identities Across Persons

We turn now to an examination of multiple identities within a situation rather than within a person. For now, we will assume that each person has only one identity; but with several persons interacting in a situation, each trying to verify his or her identity, we must be concerned with their effects on one another. In chapter 4, we saw how one identity functions to verify itself in the face of disturbances. Our concern now is that those disturbances are often other identities in the situation. Examples would include friends playing bridge, husbands and wives interacting, and members of a task group accomplishing their goal. Each of the persons has an activated identity and is engaging in behavior that both portrays the meanings of the identity and defends against discrepant meanings that are indicated or implied by others in the situation.

To facilitate this discussion, in figure 7.2 we adapt the figure of the identity model (figure 4.1) to a situation in which there are two persons, each (p.149) with an active identity, like a professor and a student. Again, each of the identities has all the components that we have talked about, including the identity standard, perceptions, a comparator, and output of meaningful behavior to the situation. Since the multiple identities are in multiple persons, each has its own output of meaningful behavior in the situation, unlike the case of multiple identities within one person. Each identity is trying to control its own perceptions of situational meanings to match the meanings held in the identity standard of that particular person/identity. The situation that one identity is controlling is the same situation that the other identity is controlling. So although there are two sets of meaningful behaviors, there is only one set of meanings in the situation, and that set is being produced by both persons and is being perceived by both persons.

Three possible scenarios can arise. Each identity may be trying to control the same meanings in the situation to be the same level that the other identity is trying to control. In this case, the identity standards of the two identities agree with each other as to what the meanings in the situation should be. As a result, the two identities can be seen to support each other; what one does to the situation is exactly what the other wants. In effect, they are on the same team. Perhaps, for example, two persons want the situation to be romantic in a similar way, so they cooperate to make it so.

The second scenario is that the two identities are trying to control the same meanings in the situation to be different levels. For example, one identity wants the situation to be romantic, while the other wants the situation to be businesslike and definitely not romantic. This is a case of true conflict because as one person moves the meanings in the situation to be in accord with their identity standard, the meanings in the situation are moving further

                   Multiple Identities

Figure 7.2. Identity Models for Two Interacting Persons

(p.150) out of accord with the meanings in the identity standard of the other person. With those identities and identity standards, the situation cannot verify both. Indeed, no situation could. As the situation moves toward a more businesslike atmosphere, the romantic identity is not verified; as the situation moves toward a more romantic atmosphere, the business identity is not verified. In this situation, each identity is a disturbance to the other identity.

There is a third (perhaps less interesting) scenario. In this, the meanings that one identity is trying to control are not relevant to the meanings that the other identity is trying to control. Two children are playing on the same beach, one building a sandcastle, the other digging a hole that is deep enough to be below the water line and fill up with water. Now, as long as the castle does not go over the hole, or the hole does not go under the sandcastle, what each is doing neither helps nor hinders what the other is doing. Each is controlling perceptions of meanings in the situation that are irrelevant to the other.

Each of these scenarios is, of course, a pure type. In most situations, the interaction is more complex in the sense that some meanings may be controlled in conflict, others in agreement, and others still that are irrelevant to one or the other of the identities involved—a mixture of types. Married couples provide a good example. Each spouse seeks to verify his or her own spousal identity in interaction with the partner. But, as any spouse knows, there are times when it works, and there are times when it does not work, or does not work as well. It works very well when what each does verifies not only the self but the other as well. It does not work well when what one does to verify the self “drives the other crazy” (that is, does not verify the identity of the other).

In our study of newly married couples (Burke and Stets 1999), we found exactly this result. Using procedures we discussed earlier, we measured the spousal role identity of about three hundred first married couples with no children who were interviewed right after they got married and then again a year later and a third time two years later. We measured the spousal identity of each partner. We also measured what each partner thought the spousal identity of the other ought to be. The spousal identity—that is, what it means to be a spouse for each person—was based on eleven items indicating various tasks or activities that are understood to be part of the spousal role. Each person was asked to what extent he or she felt he or she should do each of the activities, such as “work around the yard” or “shopping for the family.” The extent to which each thought he or she should engage in these activities defines, in part, for each individual, what it means to be a spouse. This, of course, would vary from individual to individual.

On the one hand, if the partner’s view of one’s spousal identity is consistent with the self-view, then the partner is likely to engage in activities toward the self that are consistent with the spousal self-meanings, thus verifying the self’s spousal identity. On the other hand, if the partner’s view of one’s spousal identity is inconsistent with the self-view, then the partner’s behavior is not likely to verify the self’s spousal identity. We thus have a (p.151) situation in which each partner is behaving in a common context, the household, trying to verify his or her own spousal identity, and this requires some coordination with their partner.

An analogy would be each person trying to keep the temperature of the home at a comfortable level. If one likes the temperature at 68 and the other likes the temperature at 72, then as one adjusts the thermostat to make the temperature comfortable, that setting makes the temperature uncomfortable for the other. In order not to argue about the temperature setting, the two will have to come to some agreement about the appropriate level for the temperature. This is also true about the spousal identities. If one person feels he or she should do all of the cooking and the partner would like to do half of the cooking, the behavior of one trying to confirm his or her spousal identity will result in the other partner not being able to confirm her or her identity. However, if one partner feels that his or her spousal role calls for him or her to cook most of the time, and the other partner feels his or her spousal role calls for him or her to cook some of the time, the two identities will support and reinforce each other.

We found there was quite a bit of variability across households in the degree to which the two spousal identities worked together in a complementary fashion. Although only about three and a half percent of the couples had perfect agreement across all eleven different activities, almost all couples agreed with each other within a point across each of the eleven activities. Of course, being newlyweds, these are people who are in love with each other, have known and dated each other for some time, and found themselves compatible enough to get married. We would therefore expect some compatibility in their self-views. Nevertheless, an average disagreement of one point on spousal activities has consequences, as we found (Burke and Stets 1999).

We examined a number of consequences that occurred as the degree of discrepancy increased from the zero of the few couples in complete agreement at one end of the spectrum to a maximum that averaged a little over one point across the eleven items. Two of the consequences that were examined were the degree to which each partner felt positive and negative self-feelings when the identity was or was not supported. We found that as the amount of discrepancy between the self-views of being a spouse and the views expressed by the partner for them increased, they felt an increasing amount of distress and depression. At the same time, as the discrepancy was lower, they felt an increasing amount of self-esteem and self-efficacy (Burke and Stets 1999). Being with a partner who confirms your self-view as a spouse makes you feel better about yourself. It also makes you feel competent and able; being with a partner who acts in ways that create situational meanings that do not support your self-view (because he or she has a different view of the way you should be in your spousal role) makes you feel depressed and distressed. Many identities interacting in the same situation, each trying to verify itself, works when they are controlling meanings in the same way and producing good feelings for the individuals involved. (p.152) When there are discrepancies or differences, the situation deteriorates for the individuals.

In addition to the self-feelings, we also examined the degree to which the partners trusted each other. It turned out that to the extent that persons’ spousal identities were confirmed by their partner in the household situation, they also developed a deeper trust for their partner. Their partner, after all, has behaved in ways that help them maintain their own spousal identities and feel good about themselves. As this happened for each of the partners, each became more trusting of the other—a situation of mutual trust developed. Along with this increasing mutual trust, the husband and wife developed an increased feeling of “we-ness” or couple orientation, and each was much more likely to speak of “we” rather than “I” in conversations with the other. The increased trust among these mutually verifying couples had its own consequences. The increased trust led to increased levels of mutual commitment and emotional attachment or love, which in turn helped keep the couples interacting even when times were more difficult.

These results led us to conceptualize something we called a “mutual verification context,” that is, a situation in which all the parties involved (the two spouses in our study) work together to create a context in which they can verify each other’s identities and have their own identity verified by the other’s identity. Such a mutual verification context is something that people desire and will maintain when they have it. A group of friends who all get along with one another might be an example. A family in which each member finds it possible to be himself or herself and each member likes and appreciates the others in the family might be another example. We further hypothesized that when such contexts cannot be developed or maintained, people would prefer to leave the situation in favor of a situation in which such a context is possible. Over time, this would result in an increase in the number of such mutual verification contexts, as ones that developed were maintained and ones that did not dissolved. Of course, this assumes that people are free to enter and leave such sets of relationships, and that is not always the case.

More Than Two Persons

A study by Riley and Burke (1995) examined a context in which there were multiple persons each trying to verify an identity in the same situation. This situation was a little more complex than with the couples because there were four persons in the situation trying to maintain an identity, each with respect to three others. Counting each pair of persons, a total of six relationships needs to be maintained. The identity that Riley and Burke investigated was a leader identity in the context of a small group. They measured the extent to which each member of a four-person group generally thought of himself or herself as a group leader. This was not a question of yes or no, but of degree. Some people thought of themselves as very much dominating and leading groups; some saw themselves as persons who avoided that role, (p.153) doing it very little, while others saw themselves at various points along the whole range in between.

Because people try to act in ways that verify their identities—that is, produce self-relevant meanings in the situation that match the meanings held in their identity standard—we expect that the meanings of the behaviors will ordinarily match those in the standard. Thus, people with strong leadership identities will act more like leaders, and those with weak or low leadership identities will not act like leaders or will do so much less. In the group, however, each person is not free to do what each wants to do. Each is constrained by what others are doing. For example, if two persons want to act as strong leaders in the group, they may end up competing with each other, neither being happy with his or her performance, because it does not match his or her strong leadership identity. Each may feel some distress because his or her leadership identity was not verified.

An additional feature of interaction in groups is that the members must share their interpretations of behaviors (i.e., share meanings) in order for them to coordinate their activities and accomplish the task. Thus, the behavior in which person A engages to display the meanings of his or her leadership identity must be similarly interpreted by person B. If what A does to show leadership is not what B takes to be leadership (because they don’t share meanings), then B will not respond in ways that verify A’s leadership identity. For the system to work, B must perceive A’s behavior in the same way that A perceives A’s behavior—taking the role of the other.

The groups Riley and Burke studied were laboratory groups composed of four randomly chosen student volunteers who were put together to solve a series of four human relations problems. Each problem was discussed in a separate discussion, and measures of perceived performance of each of the members were taken after each of the four discussions. They measured the leadership identity meanings of each of the participants prior to any of the discussions, they measured each person’s perceived meanings of their own leadership behavior after each discussion, and they measured the meanings of each person’s leadership behavior as perceived by others in the group after each discussion. From this they showed that each person’s leadership identity did predict his or her leadership behavior as perceived by himself or herself and by others across all the discussions (though the predictions for own perceptions were better than for other’s perceptions, as would be expected if people controlled their behavior to match their own perceptions of it). Riley and Burke also found that the perceptions of self and other were very highly correlated across the four discussions, thus indicating that self and others interpreted the meanings of the self’s behavior in very similar ways; they shared meanings.

Finally, Riley and Burke showed that the degree to which a person’s leadership behavior was not predicted by his or her leadership identity influenced members’ satisfaction with his or her role performance in the group (but not their overall satisfaction with the discussion). Group members who performed more or less leadership activity than was consistent with their (p.154) leadership identity were less happy than persons who performed at the expected level. And the degree of discrepancy or difference between actual and expected performance predicted the degree of dissatisfaction with their role in the group, confirming one of identity theory’s major points that people are distressed and upset by a failure to verify their identities.

In a further extension of this research, Burke (2006b) showed that when a group member performs more (or less) leadership in the group than would be expected on the basis of his or her leadership identity during one of the four discussions, that person changed his or her behavior in the next discussion to engage in less (or more) leadership activity. This was done, apparently, in an effort to counteract the disturbances in the system that caused him or her to perform too much (or too little). Thus, not only did the members feel bad when they could not perform at the level of their identity, they also attempted to do something about it when they had a chance. Persons with a strong leadership identity who were pushed out of that role in one discussion worked harder in the next discussion to (re)gain the role that was consistent with their identity. Similarly, persons who found themselves, for one reason or another, performing more leadership than was consistent with their identity backed off in the next discussion in an attempt to verify their identity. With more people in the situation, each trying to verify an identity, others in the situation may find it difficult to verify their own identity unless and until all the persons in the situation can establish a mutual verification context so that all identities can be verified simultaneously.


In this chapter, we examined the complexities that can arise when we consider more than one identity at a time operating in a situation to verify itself. Because each identity is a control system that operates to maintain consistency between perceived meanings and meanings in the identity standard, there are problems when two or more identities cannot agree on what those meanings should be. Whether the different meanings are held by different identities within a single person (being a student and a friend and a family member, for example), or the different meanings are held by different persons with different identities in a situation (a professor interacting with a student, for example), the meanings must be consistent or at least independent. They cannot be in disagreement without serious problems resulting either in identities changing or in people leaving the situation. In either case, the emotional reactions of the persons involved are quite negative. If the identities change so that they mutually verify one another in the sense that when one is verified other identities are also verified, people will feel good and will work to keep that situation as it is—what we called a mutual verification context. Such contexts are inherently stable and provide stability to the social structure of which they are a part.


(1.) It is, of course, possible that an identity will be dropped and the individual will no longer consider himself or herself to have that identity. For example, Cast and Burke (2002) show that spouses who have trouble verifying their spousal identity are more likely to become divorced.

(2.) Stryker and Statham (1985) and Stryker and Macke (1978) have nicely summarized much of this work.

(3.) Levels of commitment to each identity, as well as situational demands, would influence which identity is dominant in the situation even though both are activated.

(4.) Additionally, identities that one does not want known may also become known.