Attitudes As Actions: Social Identity Theory and Consumer Research

Barry R. Schlenker, University of Florida
ABSTRACT - A social identity approach is described which treats attitude expressions as interpersonal communications that present a particular image of the respondent to self and others. The approach predicts when people will associate themselves with particular symbols, and when attitude change will follow pro and counter-attitudinal behaviors. These predictions often differ from those advanced by dissonance and self-perception theories.
[ to cite ]:
Barry R. Schlenker (1978) ,"Attitudes As Actions: Social Identity Theory and Consumer Research", in NA - Advances in Consumer Research Volume 05, eds. Kent Hunt, Ann Abor, MI : Association for Consumer Research, Pages: 352-359.

Advances in Consumer Research Volume 5, 1978      Pages 352-359

ATTITUDES AS ACTIONS: SOCIAL IDENTITY THEORY AND CONSUMER RESEARCH

Barry R. Schlenker, University of Florida

[The present paper was facilitated by National Science Foundation grant BNS77-08182.]

ABSTRACT -

A social identity approach is described which treats attitude expressions as interpersonal communications that present a particular image of the respondent to self and others. The approach predicts when people will associate themselves with particular symbols, and when attitude change will follow pro and counter-attitudinal behaviors. These predictions often differ from those advanced by dissonance and self-perception theories.

The concept of attitude has occupied a central and almost hallowed place in the explanatory systems of theorists interested in human behavior. The field of social psychology was once defined as the study of attitudes (Thomas & Znaniecki, 1918), and the utility of the concept for the field of consumer behavior has hardly gone unrecognized: "No other single psychological construct has permeated consumer research as has the construct of attitudes" (Jacoby, 1976, p. 337).

Attitudes and Social Identity

Attitude is an intrapersonal construct designed to account for different individuals' consistent favorable or unfavorable response patterns vis-a-vis particular social objects and/or events. Attitude is usually defined as the amount of positive or negative affect held toward a particular object and/or event. However, all expressions of attitudes, whether given through gross overt behaviors, verbal statements, or questionnaire responses, are interpersonal communications. Like other actions, attitude expressions present a particular image of the respondent to himself and to others. Attitude responses carry diagnostic value, since they can: associate the person with particular groups of people (e.g., liberals vs. conservatives), reveal aspects of the person's personality (e.g., simple vs. complex; stable vs. unstable; optimistic vs. pessimistic), reveal the person's positive or negative orientation toward the social interaction and social setting (e.g., friendly vs. hostile), reveal the person's desired role in the interaction (e.g., dominant vs. submissive), and can be used to gain approval and avoid disapproval (e.g., conformity vs. nonconformity). In short, attitude expressions communicate aspects of the person's social identity and world view and set the tone for interactions. Even "private" attitudes, expressed only to oneself, occur in the context of imaginary social settings and serve the same functions in our private worlds. Every response used to infer our own and others' attitudes (except for physiological measures, which are notoriously unreliable) exists in a matrix of real or imagined social interaction, and can be used to influence the opinions and actions of others.

This focus on the interpersonal nature of attitudes is similar in some respects to the functional approach of the Michigan school (Katz, 1960), which examined attitudes according to their instrumental, ego-defensive, knowledge, and value expressive uses. Ultimately, these functions can be compressed to two: an informational facet, which involves cognitively organizing information about ourselves, others, and the world in a way that can guide actions; and an instrumental, or hedonic facet, which involves acting to maximize desirable outcomes and minimize undesirable outcomes. Expectancy-value or subjective expected utility models (e.g., Anderson, 1971; Fishbein & Ajzen, 1975; Jones & Gerard, 1967; Rosenberg, 1956) come closest to capturing these facets as they pertain to the conceptualization and measurement of attitudes. However, researchers who have worked with these positions have generally not devoted sufficient attention to the symbolic, interactional, and social identity aspects of attitudes, except indirectly through the assessment of the probability or utility components. Focusing on these aspects, though, generates hypotheses which might otherwise remain hidden and provides for the integration of seemingly disparate phenomena (e.g., Freudian defense mechanisms, reactions to inequity, dissonance) into a common framework. Thus, while accepting the basic usefulness of such models, it becomes worthwhile to explore avenues opened by a focus on social identity and attitudes.

Social Identity and Self-Presentation

The presentation of self is the activity of constructing and maintaining particular identities in social life. Identity is the composite of the social attributes associated with an individual. [It is worthwhile to distinguish between the self-concept and social identity. The self-concept is defined as the theory an individual has constructed about himself or herself (Epstein, 1973), and consists of perceptions of personal attributes and relationships between them. Social identity refers to "the individual's major role and social-type categorizations" (Gordon, 1968, p. 118). It describes what, who, and where the person is in social terms, and is not simply substitutable for the word self due to the more extensive social evaluation implied.] Each attribute creates a particular image of the person. Everyone--business person, politician, housewife, student, professor--develops an identity through and for social interaction and projects various aspects of that identity to particular audiences. Through impression management strategies people control how they see themselves and how they expect to be seen and treated by others. People respond to others on the basis of the identities those others create, so everyone finds it to his or her advantage to control the image of self presented for public and private consumption (Goffman, 1959; Jones & Wortman, 1973; Schlenker, 1975a).

The social identity approach derives in part from symbolic interactionism, a model of people which views them as symbol users and interpreters. According to the position, personality develops from and is inseparable from real or imagined social interactions (Carson, 1969). The focus is on the integral relationship between the person, others, society, and social symbols (cf. Brittan, 1973). Goffman's (1959, 1967, 1971) dramaturgical approach to social behavior developed from symbolic interactionism and views people as actors who perform roles available in society, managing the impressions they give to others in order to lay claim to specific social attributes. Although Goffman's work tends to minimize psychological factors and focuses upon interaction rituals, there is no reason to exclude internal, personal factors when studying self-presentation.

Constructing Identities: Attitudes and Identification

It is assumed that people manage the impressions they create to project a particular identity, both constructing it and maintaining it against actual or potential threats. Attitude expressions partially define identity and can arouse identity-maintenance concerns. Naturally, attitude topics vary in the degree to which they are relevant to identity concerns; the social judgment theory concept of ego-involvement is one way of conceptualizing such a continuum (Sherif & Hovland, 1961; Sherif et al., 1965).

Products similarly vary in the degree to which they have identity relevance. Some products can capitalize on their identity value; e.g., a Cadillac is a symbol of affluence and status, a Porsche is a symbol of a sporty life style. Other products, which initially might seem to suffer a handicap in this regard for one or another reason, can have symbolic value created for them; e.g., a Volkswagon "bug" seems at first glance to be ugly, but became a symbol of a pragmatic, environmentally-concerned life style. Even the use of products which might not appear to have any relevance for one's identity can be surprisingly revealing, at least to particular people. For example, a mother might never forgive her daughter-in-law for serving her son frozen or canned vegetables rather than fresh ones. To this particular mother, such behavior connotes laziness and indifference to her son's nutritional needs, and is an expression of the daughter-in-law's flawed personal qualities.

It often seems to be the goal of advertisers to increase the positive identity relevance of particular product lines or brands as a means of producing attitude change and ultimately affecting purchasing decisions. Desires for new products can be created by focusing on identity concerns (e.g., unless deodorants are used, doubts are raised about one's viability as a social participant) or previously neutral product brands are associated with existing relevant attributes (e.g., use of "Brut" deodorant identifies a "macho" individual). The classic techniques of testimonials and getting on the band wagon can be seen as attempts to create bonds between the consumer, some desirable image (e.g., attributes of the "star" or qualities of the symbol), and the product. "Join the Pepsi generation" (and have all the fun it entails for those who do); be like Farrah Fawcett-Majors or Joe Namath; smoke the cigarette of the rough-and-ready cowboy. Numerous studies have been devoted to specifying images associated with products (Jacoby, 1976).

From a social identity perspective, such attitude change is exemplary of a more general hypothesis: People act to increase their personal association with desirable attributes and outcomes, and act to decrease their personal association with undesirable attributes and outcomes. Personal association is defined as the probability that a particular attribute describes the individual; or in the case of outcomes, the degree to which the person is seen as responsible for producing the outcome. Desirability is defined as the worth of a particular attribute or outcome to the individual in a particular situation; it is defined idiosyncratically rather than normatively.

Research from a variety of areas provides strong support for the association hypothesis. In attributing the "causes" of particular actions, most people retrospectively report greater responsibility for successful task performances than for failing ones (e.g., Mynatt & Sherman, 1975; Schlenker, 1975b; Schlenker & Miller, 1977; Schlenker et al., 1976; Snyder et al., 1976; Wolosin et al., 1973; Wortman et al., 1973), and this is true whether the person worked on the task alone or with a group of others. People also more readily accept the implications of favorable rather than unfavorable personality information (e.g., Eagly, 1967; Jones, 1973; Steiner, 1968). Even when a person is objectively non-responsible for producing a desirable outcome, the same proclivities are evidenced. Cialdini et al. (1976) found that more students wore school-identifying products (e.g., school sweatshirts) after their school's football team had been victorious rather than nonvictorious; and that they used the pronoun "we" more when talking about a school victory rather than a defeat. The latter effect was strongest when the students had previously suffered a personal failure that threatened their self-images. When important self-images are threatened or otherwise in doubt, people display a greater association/dissociation effect.

As the above suggests, individuals often prefer to associate themselves with attributes that are evaluated by most people as desirable, e.g., competence, attractiveness. Such findings are frequently explained as being the result of self-esteem needs, social approval needs, or cognitive consistency needs. But it seems somewhat simplistic to propose that people's identity-relevant actions are taken merely to maximize self-esteem and/or to maximize approval from others. It would then be difficult to explain why some people present themselves as incompetent, hostile, unfriendly, unworthy of responsibility for success, or even out of touch with reality and mentally ill (cf. Braginsky et al., 1969; Carson, 1969). Braginsky et al. (1969), for example, found that schizophrenics' self-presentational behaviors were affected by the identity they wished to project to particular audiences. Mental patients who were old-timers in an institution and who enjoyed open ward privileges were asked to complete mental tests or talk with a psychiatrist. These patients were assumed to prefer the freedom, with attention from the staff, provided by the open wards, and would want to avoid being seen as either healthy enough to be sent home or ill enough to be confined to restrictive closed wards. When the patients believed that the test/interview was for the purpose of diagnosing their mental health, with high scores indicating they should be sent home, they presented themselves as mentally ill. They gave frequent reports of hallucinations and showed other signs of abnormal, incompetent functioning. But when they believed that the test/interview was for the purpose of diagnosing their mental illness, with high scores indicating they should be placed in undesirable closed wards, they presented themselves as mentally healthy. They gave few reports of hallucinations and showed signs of normal, competent functioning. Thus, schizophrenics, often defined as being out of touch with reality and incapable of normal functioning, were quite adept at modifying their self-presentations to fulfill their interaction goals.

Similarly, identity-relevant actions don't seem motivated by "needs" for cognitive consistency. If they were, it would be difficult to explain the prevalence of inconsistent belief systems or belief-behavior relationships (cf. Tedeschi et al., 1971). Schlenker (1975a), for example, found that people who view themselves as "failures" will present themselves in a self-effacing manner when the failures are publicly known but will present themselves quite positively when the failures are unknown to the audience.

Such data falls into place when viewed from a social identity perspective. The desirability of an attribute or outcome can only be assessed with reference to the identity the individual is attempting to project. As Braginsky et al., (1969) demonstrated, different attributes acquire different values depending upon one's goals in a situation. Further, certain attributes can only be claimed by certain people. Most people might like to be seen as intelligent, competent, witty, attractive, respected, and so forth. But only certain individuals can successfully lay claim to such attributes; reality does impose constraints on projected identities. Inaccurate claims to social attributes produce disruption of interactions, anxiety, guilt, and the possible censure and punishment of the offender (Goffman, 1959, 1967). Not only does an illicit identity (i.e., one the individual has no "right" to claim) produce embarrassment and anxiety for the discovered offender, it creates similar states for those who are duped into believing the misrepresentation and then realize their mistake. Schlenker (1975a) supported the hypothesis that people must believe they can successfully (without irrefutable challenge) act in accord with the identities they project before they will present themselves in a normatively positive fashion to an audience.

Claimed images or attributes can create obligations for the person. An obligation is defined as the belief that certain physical characteristics or behaviors should accompany claims to a particular image or attribute. A person who claims intellectual prowess should be able to demonstrate it; a person who claims physical beauty should possess some such recognizable features. Attributes (and products) vary in the degree to which they create obligations, and people vary in the degree to which they perceive obligations. A man might avoid a "Joe Namath" cologne if he believes wearing it would make him the object of jests about his baldness, pot belly, lack of coordination, and failures with women. A second man with a similar set of characteristics might not believe that any obligations derive from the use of the product (and thus not expect such chiding), and revel in his fantasies of himself as playboy extraordinaire.

Images or attributes also vary in the degree to which claiming them produces public (vs. only private) concerns. The personal desirability of a normatively desirable attribute will be Low when the identity it implies creates public obligations which the person believes he or she cannot satisfy. But if a normatively desirable attribute only creates private obligations (e. g., those of a Walter Mitty fantasy world) and no public ones, its personal desirability will be high. The use of some products (e.g., "personal" items meant to be concealed), even if it would create obligations if publicly known, can be hidden away from public scrutiny.

In the realm of consumer behavior, this translates into the prescription that one must be careful of the type of image associated with particular products. The individual must attach a high personal desirability to the image. But this will only occur when he or she favorably evaluates the particular image in others, and either (a) believes that their are no public obligations associated with the image, or (b) believes that he or she can fulfill whatever public obligations are perceived, or (c) believes that although there might be public obligations that can't be fulfilled, he or she can "hide" public use of the product and not have to fulfill them. If public obligations exist which cannot be hidden and cannot be personally fulfilled, normatively desirable images will be rejected.

This would seem to be a particularly fruitful area for consumer research. Although worthwhile data do exist on product and brand images, to my knowledge nothing has been done with respect to (a) obligations that derive from those images, (b) consumers' abilities to fulfill whatever obligations exist, and (c) the public vs. privateness of whatever obligations may exist. Knowledge of the interrelationship between these factors relative to particular products could provide a more adequate predictor of purchasing decisions and behaviors.

Maintaining Identities: Attitudes, Accounting, and Counterattitudinal Actions

One of the more salient areas of attitude research has been on the relationship between behaviors and subsequently expressed attitudes, particularly with respect to counterattitudinal behavior. Attitude change following counterattitudinal behavior has typically been explained via dissonance theory (Aronson, 1968; Festinger, 1957) or self-perception theory (Bem, 1967, 1972). The social identity approach offers an alternative conceptualization of the area which focuses on the interpersonal concerns raised by such actions (Riess & Schlenker, 1977; Schlenker, 1973, 1975c; Schlenker & Forsyth, 1977; Schlenker & Schlenker, 1975).

The counterattitudinal advocacy paradigms employed by psychologists typically induce subjects to lie, cheat, harm others, refrain from doing what they want to do, or otherwise make themselves appear to be immoral, unattractive, incompetent, or irrational. In other words, they place subjects in a predicament that could threaten their projected identities. Following actions which create a predicament (i.e., ones which are embarrassing, undesirable, illegal, anti-normative, or disruptive) people try to account for their behaviors. Accounting is an attempt to describe (interpret) the action and its consequences in personally desirable terms. Typically, this translates as interpretations which are socially acceptable (Lyman & Scott, 1968). People are "scientists'' in the sense that they try to understand and explain the causes of events through attributions. But accounting can go beyond "logical" attribution in that a search is made to "explain" one's otherwise undesirable actions in desirable ways. Accounting involves the use of attributions and attitude expressions in the service of maintaining a particular identity.

Although there are numerous accounting tactics, two general classes of responses predominate--excuses and justifications. Excuses are attempts to remove from oneself personal responsibility for the predicament. Excuses include tactics such as (a) denying one did it, (b) stating it was an accident or the consequences were unforeseeable, (c) citing external coercive pressures (e.g., "He made me do it"), and (d) citing internal coercive pressures (e.g., "I couldn't help it, I was [starving] [drunk] [mentally ill] [possessed by demons]''). Justifications allow a person to admit responsibility for the predicament but alter the perceived negativity of the consequences of the predicament. For example, one might justify harming another by noting why the other person deserved it (and hence the act was not "bad") or by minimizing the amount of harm done ("It was only a scratch"). Thus, successful excuses allow one to escape from the predicament by minimizing personal responsibility, while successful justifications allow one to escape by minimizing the aversive consequences.

It is hypothesized that the particular type of account which is used in any situation will depend upon (a) its subjective probability of extricating the person from the predicament or minimizing the potential punishment, and (b) the utility of the perceived outcomes it generates (the degree to which it eliminates or minimizes the potential negative sanctions from observers and from self guilt). People will use that accounting tactic with the most positive (or least negative) subjective expected utility. The characteristics of the actor, situation, and audience affect these probabilities and utilities, since they affect the standards involved in judging the action and the degree to which a particular account could be challenged or refuted by the "facts." As this implies, accounts which can be obviously refuted by known information are least likely to be used because they are easiest to challenge, making the person not only still accountable for the original action but also sanctionable for the attempted deceit. Space does not permit a full exploration of accounting here, but two points are worth noting. First, predictions can be made about when particular types of accounting tactics will be used. Second, the accounting tactics mentioned above bear a striking resemblance to phenomena previously discussed under the headings of (a) modes of dissonance reduction, (b) equity restoration techniques, and (c) defense mechanisms in psychopathology.

People frequently internalize the accounts they deliver and come to believe their own interpretation of the situation, producing relatively enduring attitude change or restructuring. It is hypothesized that people will internalize their accounting tactics when they negatively evaluate the action that created the predicament. If a person does not negatively evaluate the act, or feels the act is much less negative than does an audience, an account might be used to placate the audience but will be recognized by the actor as a "necessary lie" (cf. Schlenker & Forsyth, 1977, for an elaboration of this hypothesis). When "necessary lies" are told, it can complicate the person's existence; he or she must then keep multiple sets of books to recall which accounts were delivered to which audiences. Thus, an account which is accepted by both self and others facilitates one's effective functioning.

The amount of accounting necessary following a predicament is directly related to both personal responsibility for the predicament and the magnitude of its negative consequences. Responsibility is defined as the attribution that an individual freely and purposefully behaved to produce the primary consequences associated with the action, and did not engage in the behavior because of coercive pressures, accident, or to achieve other associated consequences. The more negative the action and the greater the responsibility of the actor, the more imperative it is to eliminate, or at least minimize, the potential guilt and punishment.

Consistent with the accounting hypothesis, it has been found that significant attitude change occurs following counterattitudinal actions only when a person appears to be personally responsible for engaging in the behavior and the behavior produces aversive consequences (Calder et al., 1973; Collins & Hoyt, 1972; Hoyt et al., 1972). Envision the typical paradigm. A subject is given either high or low choice (high vs. low responsibility, respectively) about making a speech which persuades or does not persuade (high vs. low harm, respectively) an audience. If no choice exists, the person has a ready-made visible excuse for the behavior and does not have to further account for it. If no harm is done, there is no need to account for the action. Only when responsibility and aversive consequences coexist is accounting necessary. But how can a high choice, high harm subject account for the behavior? The action can't be excused, since it was emphasized that the choice to perform it was solely his. It can be justified, though, by showing why little or no harm occurred. If the subject expresses attitudes which are roughly congruous with the position taken in the speech, then no real harm has occurred--after all, the subject personally endorses the beliefs the audience adopted. The counterattitudinal behavior becomes neither a lie nor harmful to others.

Research examining social identity theory indicates that responsibility and consequences are important not because they impact on cognitive consistency, but because they define the nature of the social identity dilemma (e.g., Forsyth et al., 1977; Riess & Schlenker, 1977; Schlenker, 1975a, 1975c; Schlenker & Forsyth, 1977; Schlenker & Schlenker, 1975). In many (though not all) counterattitudinal behavior situations, social identity theory makes the same predictions as does Aronson's (1968) revision of dissonance theory, which stresses that dissonance is produced when the implication of a behavior is inconsistent with one's self-concept. However, the two approaches differ in the hypothesized mediating mechanisms behind attitude change. Dissonance theory focuses on cognitive consistency-inconsistency, while social identity theory focuses on accounting following social predicaments. The difference is best illustrated by the fact that most counterattitudinal behaviors do two things: (1) produce a predicament for the person since he or she has, by definition, lied and perhaps harmed others, thus being subject to sanctions and (2) produce cognitive inconsistency. This natural covariation has perhaps caused dissonance theorists to focus on the wrong variable. According to dissonance theory, motivation arises from cognitive inconsistency; according to social identity theory, motivation arises from the nature of the identity predicament. Although these variables covary in most situations, they can be separated such that a "lie" is made to produce positive rather than aversive consequences. When this is done, data indicate that inconsistency per se is particularly non-troublesome, and people behave according to predictions derived from social identity theory rather than dissonance or self-perception theory (Schlenker & Schlenker, 1975).

It is necessary to elaborate some of the implications of the above. The above does not imply that people do not organize attitudes and beliefs in a way that frequently gives the appearance of logical (or psychological) consistency. They do. All models of information integration (e.g., Anderson, 1971; Osgood & Tannenbaum, 1955; Wyer, 1974) are in some sense "consistency" models, since they describe how information is processed to achieve a "consistent" behavioral orientation. However, information integration models do not automatically lead to the conclusion that inconsistency has major motivational by-products that steer social behaviors in a particular direction. Only the so-called cognitive consistency theories (balance theory, dissonance theory) posit such drive-reduction needs arising from cognitive inconsistency. It is one thing to discuss the process of information integration, another to assume that cognitive inconsistency per se produces psychological needs which must be satisfied. The fact that people do weight and integrate information in order to make sense out of their environment and to function effectively in it is certainly compatible with (indeed, assumed by) social identity theory. Similarly, "incentive theory" (Elms, 1967; Janis & King, 1954; King & Janis, 1956) ideas about the effects of cognitive contact and biased scanning on attitude change following role-playing are quite compatible with information processing models. However, most research attention has been devoted to people's attitudinal reactions to predicaments. In such situations, motivation does not appear to arise from intra-psychic, cognitive consistency pressures, but from the nature of the predicament for social identity.

Social identity research suggests that many of the dissonance-type ideas about counterattitudinal behavior, effort justification, and so forth, might not have the marketing implications previously thought. For example, getting a consumer to reach to a higher shelf to get a product (exerting maximal effort) might not automatically lead to more favorable attitudes toward the product. It might do so only to the extent that the behavior created a predicament for the individual. Say, he was asked by another shopper why he strained to get the product when an equivalent one was in easy reach, believed so doing was somewhat irrational, believed it would harm his identity if he admitted this was irrational, and believed that an account based on product-liking would be most effective. It would be rare for a person to encounter this specific sequence of events. Or, instead of using "free giveaways" of new products, dissonance theorists might advocate making consumers pay some small amount for it or making them mail a postcard (exerting effort) rather than having it left on their doorstep. Once again, unless a predicament is generated, the strategy does not seem likely to have the most positive effects. Hedonic considerations suggest fewer people will be exposed to the products, and those that are will probably not develop any more favorable attitudes than otherwise. Marketing strategies based on "insufficient justification" do not appear to have strong data to recommend them.

The only marketing area that does seem to yield data consistent with the "insufficient justification" idea involves the use of pricing techniques. Some dissonance researchers have recommended that products be sold for the largest amount possible (that will still get purchases) so that dissonance is aroused (Doob et al., 1969). Yet when a direct relationship is found between pricing and liking for a product, it is probably best to conceptualize it as due to economic evaluation processes; the relationship between price and perceived product quality has generated more laboratory research during the past eight years than any other topic in consumer behavior (Jacoby, 1976). Many products are simply perceived as more valuable the higher their price, at least up to some upper limit (Gabor & Granger, 1966). This is true irrespective of whether a purchase has been made. [This is not to suggest that postdecisional distortions of the values of accepted and rejected alternatives do not occur. However, this well documented phenomenon does not follow exclusively from dissonance theory. Virtually identical predictions can be derived from conflict models, the idea of unequivocal behavior orientation (Jones & Gerard, 1967), social identity theory (based on derivations of the association hypothesis), and self-perception theory. Though the mediating processes may be debated, the end results relevant for consumer behavior seem to be comparable for each approach.]

Maintaining Identities: Attitudes, Acclaiming, and Proattitudinal Actions

The social identity approach is built around the self-presentational nature of attitude expressions; it focuses upon the potential consequences of attitude expressions and the ways the consequences and the situation affect attributions which can be made about the person by himself and others. As was seen, many counterattitudinal actions produce consequences which are identity-threatening and which require accounting tactics. A complimentary process should occur following proattitudinal actions. Proattitudinal actions are often identity-supporting, and, depending upon the nature of the action, might produce consequences which are highly desirable. Hence, the association hypothesis indicates that people should want to appear personally responsible for relevant proattitudinal behaviors. For example, a person who contributed to a charity would prefer the attribution that the donation was made because he believed in the cause, and would want to avoid attributions such as "He did it because it is expected of those in his societal position," or "He did it because he was pressured into it by the canvasser for the charity.'' Responsibility implies that the person is entitled to the appropriate rewards, credit, respect, and feelings of self-worth that derive from the action. As the desirability of the consequences increases, so should the value of the accompanying image and personal association.

If an individual already appears associated with the behavior and its obviously desirable consequences, no further actions are necessary; he or she can reap the appropriate rewards. However, if the actor does not appear to be responsible for the action or if the consequences appear to be less desirable than the actor believes they should appear, acclaiming tactics will occur. Acclaiming is an attempt to describe (interpret) an action and its consequences in ways which reflect favorably on the actor. The two most salient acclaiming tactics are entitlings, which involve attempts to increase one's responsibility for an action and its consequences, and enhancements, which involve attempts to increase the apparent positivity of the consequences of the action. As with accounting, predictions can be derived about which acclaiming tactics will be used in particular situations based on the tactics' subjective expected utilities.

It is hypothesized that when environmental variables decrease the amount of responsibility that can be attributed to a person for proattitudinal actions, the more desirable the actions and their consequences, the more the person will use acclaiming tactics. Responsibility can be demonstrated by increasing the extremity of one's position on the topic or by appearing highly involved with the topic, such as by stressing its personal importance, its value to society, and one's behavioral commitment to it. Extremity and involvement are typically conceptualized as dimensions of one's attitude toward an issue (Scott, 1968). When a person already appears responsible for a desirable action, little or no additional acclaiming is required; responsibility and recognition are already theirs. But when environmental pressures decrease responsibility and dissociate the person from the positive consequences, he or she should use acclaiming tactics to increase the likelihood that responsibility is attributed.

As an example of the application of this hypothesis in role-playing situations, consider a politician who delivers a speech which coincidentally favors the popular side of some local issue. Naturally, the politician would prefer the audience to make the attribution that the speech reflected his personal attitudes (personal attribution) rather than that the speech merely reflected the desire for reelection (role attribution). To insure that the origin of the speech is attributed to personal attitudes, the politician might demonstrate increased gesticulation, hand-pounding, and voice inflection during the speech (as compared to a situation in which responsibility was not in doubt). In addition, he should evidence more favorable attitudes toward the issue at the conclusion of the speech than were held by him the previous day, at least up to some maximal point. This show of involvement with the issue would serve as a means of convincing the audience, and perhaps himself, that the behavior was desirable and was due to personal feelings rather than external role constraints.

Consistent with the above, research has found that when important positive consequences are produced by behaviors (e.g., delivering a proattitudinal speech), subjects subsequently express more favorable attitudes toward the topic when they are given no choice rather than high choice about engaging in the behavior; the effect has been found following both proattitudinal (Schlenker & Riess, 1977) and counterattitudinal(Schlenker & Schlenker, 1975) actions. These results fail to support predictions of the opposite effect (greater attitude change following high rather than low decision freedom) derived from self-perception theory.

These findings indicate that when the consequences of an action are identity-relevant and important, the type of "insufficient justification effect" which occurs will depend upon whether the consequences are desirable or undesirable. When the consequences are undesirable, insufficient Justification (i.e., high responsibility) for the behavior is directly related to subsequent attitude change; as choice increases, so does attitude change. When the consequences are desirable, though, insufficient justification is inversely related to subsequent attitude change; as choice increases, attitude change decreases.

These results do not imply that self-perception principles are majorly incorrect. Self-perception theory, in conjunction with even more general ideas about attribution, has been quite successful in predicting attitudes in situations where (a) the consequences of the action are minimal and social identity concerns are not aroused (e.g., Taylor, 1975), and (b) subjects are given false or misleading information about their internal states and must make sense out of the information available to them (e.g., Nisbett & Valins, 1972). As the desirability of an action increases, though, social identity concerns become engaged and people appear to follow a pattern of rationalization rather than "rational'' information processing.

Conclusions

The social identity approach highlighted here has potential for integrating a wide variety of phenomena under a common rubric. The approach weaves together symbolic interaction (people are symbol users and interpreters), attribution theory (behaviors have social meanings which reflect on the identity of the actor with certain probabilities and consequences), hedonic principles (people act to maximize subjective expected utility), and social influence (people use impression management strategies for accounting and acclaiming purposes, which influence self and others).

In the realm of attitudes, the approach is not competitive with information processing models or with functional approaches to attitudes. Rather, it complements these by focusing on the communicative, interpersonal, social, and symbolic uses of attitude expressions. It emphasizes both rationalization and rational information processing. Rationalization is not conceptualized as an aspect of Freudian intrapsychic conflicts determined sometime during the first five years of life, or as due to needs for cognitive consistency. Instead, it is viewed as an identity protecting and enhancing tactic that arises from the nature of the social interaction process. People are "action centers" (Heider, 1958) who intend to produce certain effects and who are held accountable for what they do. The rewards and punishments (both from others and from pride or guilt) associated with accountability provide a powerful hedonic impetus for rationalization.

At one level, it has been argued that the basic motivational processes described by dissonance theory may well be in error--needs for cognitive consistency do not appear to have the omnipotent driving force that has been attributed to them. At another level, the kinds of predictions typically made by dissonance theory can be reconceptualized as a "special case" of the more general social identity approach. Such a possibility has been indirectly suggested by Zimbardo (1969, p. 15), who believes that Festinger's theories of social communication, social comparison, and dissonance, "may be subservient to more basic phenomena which characterizes this approach as a 'face-saving theory,' in which the individual is motivated to modify and distort both internal and external reality in order to make them appear consistent with having made the 'correct' decision."

Although the attribution process is salient in the social identity approach, attributions serve as the foundation for information about one's identity. These attributions can be biased by one's self-interests in order to control identity; attributions are acted on and manipulated through accounting and acclaiming tactics. From this perspective, attribution does not play the passive role it does in self-perception theory, where attributions about our attitudes are the end points of the attitude change process. There is rarely, if ever, only one attribution that can be made about an action. People exercise the opportunity to sort through the attributional possibilities until one or more can be found which best serve self-interests and establish the preferred type of social identity.

Finally, the social identity approach is most directly applicable to symbols that are (or can be made) at least somewhat identity relevant. The boundaries of identity relevance are cloudy at the moment, since relatively little empirical work has been directed at specifying when people evaluate a symbol as relevant versus irrelevant to their identity. But it seems reasonable to suggest that many products have some identity relevance, or at least that many products can take on such relevance through particular types of advertising campaigns. Campbell's soups, for example, once stressed their relevance to the image of the stay-at-home housewife by emphasizing attributes associated with mother's good home cooking. Today, changing times have dictated another strategy; the value of the product is stressed for the image of liberated women who want to provide proper nourishment for their families while escaping the house quickly for an afternoon game of tennis. The degree of identity relevance of particular products or product brands is an empirical question, but it would seem that a large number of products already have the requisite relevance to make social identity hypotheses applicable.

REFERENCES

E. Aronson, "Dissonance Theory: Progress and Problems," In R. Abelson, E. Aronson, W. McGuire, T. Newcomb, M. Rosenberg and P. Tannenbaum (Eds.), Theories of Cognitive Consistency: A Sourcebook, (Chicago: Rand McNally, 1968)

N. H. Anderson, "Integration Theory and Attitude Change," Psychological Review, 78 (1971), 171-206.

D. J. Bem, "Self-perception: An Alternative Interpretation of Cognitive Dissonance Phenomena," Psychological Review, 74 (1967), 183-200.

D. J. Bem, "Self-perception Theory," In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (New York: Academic Press, 1972)

B. M. Braginsky, D. D. Braginsky and K. Ring, Methods of Madness: The Mental Hospital as a Last Resort (New York: Holt, Rinehart and Winston, 1969)

A. Brittan, Meanings and Situations (Boston: Routledge and Kegan Paul, 1973)

B. J. Calder, J. Ross and C. A. Insko, "Attitude Change and Attitude Attribution: Effects of Incentive, Choice, and Consequences," Journal of Personality and Social Psychology, 25 (1973), 84-89.

R. C. Carson, Interaction Concepts of Personality (Chicago: Aldine, 1969)

R. B. Cialdini, R. J. Borden, A. Thorne, M. R. Walker, S. Freeman and L. R. Sloan, "Basking in Reflected Glory: Three (Football) Field Studies," Journal of Personality and Social Psychology, 34 (1976), 366-375.

B. E. Collins and M. F. Hoyt, "Personal Responsibility- For-Consequences: An Integration and Extension of the 'Forced Compliance' Literature," Journal of Experimental Social Psychology, 8 (1972), 558-593.

A. N. Doob, J. M. Carlsmith, J. L. Freedman, T. K. Landauer and S. Tom, Jr., "Effect of Initial Selling Price on Subsequent Sales," Journal of Personality and Social Psychology, 11(1969), 345-350.

A. H. Eagly, "Involvement as a Determinant of Responses to Favorable and Unfavorable Information," Journal of Personality and Social Psychology, 7(1967), Whole No. 643.

A. C. Elms, "Role-Playing, Incentive and Dissonance," Psychological Bulletin, 68(1967), 132-148.

S. Epstein, "The Self-Concept Revisited: Or a Theory of a Theory," American Psychologist, 28(1973), 404-416.

L. Festinger, A Theory of Cognitive Dissonance (Evanston, Ill.: Row, Peterson, 1957)

M. Fishbein and I. Ajzen, Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research (Reading, Mass.: Addison-Wesley, 1975)

D. R. Forsyth, M. Riess and B. R. Schlenker, "Impression Management Concerns Governing Reactions to a Faculty Decision," Representative Research in Social Psychology (1977), in press.

A. Gabor and C. W. J. Granger, "Price as an Indicator of Quality: Report on an Enquiry," Economica, 33(1966), 43-47.

E. Goffman, The Presentation of Self in Everyday Life (Garden City, N.Y.: Doubleday Anchor, 1959)

E. Goffman, Interaction Ritual: Essays on Face-to-Face Behavior (Garden City, N.Y.: Doubleday Anchor, 1967)

E. Goffman, Relations in Public (New York: Basic Books, 1971)

C. Gordon, "Self-conceptions: Configurations of Content," In C. Gordon and K. J. Gergen (Eds.), The Self in Social Interaction (New York: Wiley, 1968)

F. Heider, The Psychology of Interpersonal Relations (New York: Wiley, 1958)

M. F. Hoyt, M. D. Henley and B. E. Collins, "Studies in Forced Compliance: The Confluence of Choice and Consequences on Attitude Change," Journal of Personality and Social Psychology, 22(1972), 1-7.

J. Jacoby, "Consumer Psychology: An Octennium," In M. R. Rosenzweig and L. W. Porter (Eds.), Annual Review of Psychology, 27(Palo Alto, Ca.: Annual Reviews, Inc., 1976)

I. L. Janis and B. T. King, "The Influence of Role-Playing on Opinion Change," Journal of Abnormal and Social Psychology, 49(1954), 211-218.

E. E. Jones and H. B. Gerard, Foundations of Social Psychology (New York: Wiley, 1967)

E. E. Jones and C. Wortman, Ingratiation: An Attributional Approach (Morristown, NJ: General Learning Press, 1973)

S. C. Jones, "Self and Interpersonal Evaluations: Esteem Theories Versus Consistency Theories," Psychological Bulletin, 79(1973), 185-199.

D. Katz, "The Functional Approach to the Study of Attitudes,'' Public Opinion Quarterly, 24(1960), 163-204.

B. T. King and I. L. Janis, "Comparison of the Effectiveness of Improvised Versus Non-Improvised Role-Playing in Producing Opinion Changes," Human Relations, 9 (1956), 177-186.

S. M. Lyman and M. B. Scott, A Sociology of the Absurd (New York: Appleton Century-Crofts, 1970)

C. Mynatt and S. J. Sherman, "Responsibility Attribution in Groups and Individuals: A Direct Test of the Diffusion of Responsibility Hypothesis," Journal of Personality and Social Psychology, 32(1975), 1111-1118.

R. E. Nisbett and S. Valins, "Perceiving the Causes of One's Own Behavior," In E. E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins and B. Weiner (Eds.), Attribution: Perceiving the Causes of Behavior (Morristown, NJ: General Learning Press, 1972)

C. E. Osgood and P. H. Tannenbaum, "The Principle of Congruity in the Prediction of Attitude Change," Psychological Review, 62(1955), 42-55.

M. Riess and B. R. Schlenker, "Attitude Change and Responsibility Avoidance as Modes of Dilemma Resolution in Forced Compliance Settings," Journal of Personality and Social Psychology, 35(1977), 21-30.

T. S. Robertson, Consumer Behavior (Glenview, Ill: Scott, Foresman, 1970)

M. J. Rosenberg, "Cognitive Structure and Attitudinal Affect," Journal of Abnormal and Social Psychology, 53 (1956), 367-372.

B. R. Schlenker, "Self-Image Maintenance and Enhancement: Attitude Change Following Counterattitudinal Advocacy," Proceedings of the 81st Annual Convention of the American Psychological Association (Montreal, 1973)

B. R. Schlenker, "Self-Presentation: Managing the Impression of Consistency When Reality Interferes With Self-Enhancement," Journal of Personality and Social Psychology, 32(1975), 1030-1037. (a)

B. R. Schlenker, "Group Members' Attributions of Responsibility For Prior Group Performance," Representative Research in Social Psychology, 6(1975), 96-108. (b)

B. R. Schlenker, "Liking for a Group Following an Initiation: Impression Management or Dissonance Reduction?" Sociometry, 38(1975), 99-118. (c)

B. R. Schlenker and D. R. Forsyth, "Social Identity Theory: The Effects of Incentives on Attitude Change Following Counterattitudinal Behavior," Manuscript submitted for publication (Available from the authors, University of Florida, 1977)

B. R. Schlenker and R. S. Miller, "Egocentrism in Groups: Self-Serving Biases or Logical Information Processing?" Journal of Personality and Social Psychology (in press, 1977)

B. R. Schlenker and M. Riess, "Self-Presentations of Attitudes Following Commitment to Proattitudinal Behavior," Manuscript submitted for publication (Available from the authors, University of Florida, 1977)

B. R. Schlenker and P. A. Schlenker, "Reactions Following Counterattitudinal Behavior Which Produces Positive Consequences." Journal of Personality and Social Psychology, 31(1975), 962-971.

B. R. Schlenker, S. Soraci Jr. and B. McCarthy, "Self-Esteem and Group Performance as Determinants of Egocentric Perceptions in Cooperative Groups," Human Relations, 29(1976), 1163-1176.

W. A. Scott, "Attitude Measurement," In G. Lindzey and E. Aronson (Eds.), Handbook of Social Psychology, 2 (Reading, Mass.: Addison-Wesley, 1968)

M. Sherif and C. I. Hovland, Social Judgment: Assimilation and Contrast Effects in Communication and Attitude Change (New Haven, CT.: Yale University Press, 1961)

C. W. Sherif, M. Sherif and R. E. Nebergall, Attitude and Attitude Change: The Social Judgment-Involvement Approach (Philadelphia, Pa.: W. B. Saunders, 1965)

M. L. Snyder, W. G. Stephan and D. Rosenfield, "Egotism and Attribution," Journal of Personality and Social Psych, 33(1976), 435-441.

I. D. Steiner, "Reactions to Adverse and Favorable Evaluations of Oneself," Journal of Personality, 36(1968), 553-564.

S. E. Taylor, "On Inferring One's Attitudes From One's Behavior: Some Delimiting Conditions," Journal of Personality and Social Psychology, 31(1975), 126-131.

J. T. Tedeschi, B. R. Schlenker and T. V. Bonoma, "Cognitive Dissonance: Private Ratiocination or Public Spectacle?" American Psychologist, 26(1971), 685-695.

W. I. Thomas and F. Znaniecki, The Polish Peasant in Europe and America, 1(Boston: Badger, 1918)

R. J. Wolosin, S. J. Sherman and A. Till, "Effects of Cooperation and Competition on Responsibility Attribution After Success and Failure," Journal of Experimental Social Psychology, 9(1973), 220-235.

C. B. Wortman, P. R. Costanzo and T. R. Witt, "Effects of Anticipated Performance on the Attributions of Causality to Self and Others," Journal of Personality and Social Psychology, 27(1973), 372-381.

R. S. Wyer Jr., Cognitive Organization and Change: An Information Processing Approach (Potomac, Md.: Lawrence Erlbaum Associates, 1974)

P. G. Zimbardo, The Cognitive Control of Motivation (Glenview, Ill.: Scott, Foresman, 1969)

----------------------------------------