Stereotypes and Behavioral Confirmation: From Interpersonal to Intergroup Perspectives

"The organization of society is the framework inside of which social action takes place ... . Such organization and changes in it are the product of the activity of acting units and not of 'forces' which leave such units out of account."-Herbert Blumer (1962, p. 189). The longstanding interest aroused by research on behavioral confirmation in social interaction has been motivated, in part, by the implications of this phenomenon for interactions at an intergroup level. For if members of disadvantaged or stigmatized groups can be led to conform to the derogatory stereotypes held by a dominant group, the latter can maintain their positions of privilege and power. However, most studies on behavioral confirmation have concerned interactions between individual perceivers and individual targets. Moreover, these interactions have generally not been defined as involving intergroup relations, as the expectations induced in the perceiver have generally concerned personality traits attributed to the target as an individual rather than to the target's group membership per se. Hence, although it is quite possible that behavioral confirmation may take place during intergroup contact, the relevance of existing studies of interpersonal interactions to these settings needs to be established, both theoretically and empirically. Accordingly, our primary goal in this chapter has been to address this transition from the interpersonal to the intergroup levels of interaction by trying to answer the following question: How does stigmatization of one of the group members affect the confirmation of stereotypical expectations (a process that we labeled as "stereotype confirmation") during contact between members of these groups? To address this question, we began by articulating two processes believed to underlie stereotype confirmation in interpersonal settings: (1) the reciprocation of the perceiver to the target's anticipated behavior, and the target's own reciprocation of the perceiver's overtures; and (2) the pursuit of a confirmatory strategy by the perceiver, coupled with deference on the target's part. Then, we drew on the literature on interactions between nonstigmatized individuals (as perceivers) and stigmatized individuals (as targets) to suggest two behavioral styles likely to be adopted by the former-avoidance and dominance-in their dealings with the latter. Specifically, we reviewed relevant empirical data suggesting that avoidance coupled with reciprocation is particularly likely to lead to the confirmation of stereotypes regarding the sociability of the target, whereas the dominant style will facilitate the implementation of a confirmatory strategy and the consequent confirmation of expectations regarding the stigmatized target's competence to the extent that the target adopts a complementary submissive style. Based on these distinctions, we next identified which aspects of the interactions between members of stigmatized and nonstigmatized groups may facilitate stereotype confirmation and disconfirmation, respectively. Among these aspects, we first considered those that are related to the perceiver. We hypothesized that in unstructured interactions, category salience, intergroup anxiety, and prejudice each should facilitate the use of the avoidant style by the perceiver, whereas in structured interactions (involving the performance of a joint task), category salience should facilitate the use of a dominant style by the perceiver. A review of the relevant literature on the interactions between members of various stigmatized and nonstigmatized groups proved to be generally consistent with this prediction. Second, we considered aspects related to the stigmatized target, and how the target's response to the perceiver's behavior could affect stereotype confirmation. We proposed that this response may depend on the perceived permeability of group boundaries. When these boundaries are perceived as permeable, we proposed that the awareness that the perceiver may stereotype oneself ("stigma consciousness") is the crucial factor determining the occurrence of behavioral confirmation or disconfirmation. When not stigma conscious, targets are especially vulnerable to stereotype confirmation. When they are stigma conscious, we draw on social identity theory to suggest that they may try to dissociate themselves from their group in order to show that the stereotype does not apply to themselves. These actions, we have argued, may lead to perceptual disconfirmation. By contrast, when intergroup boundaries are perceived as impermeable, we suggested that targets may respond differently as a function of whether they perceive cognitive alternatives to the existing status differences. If there are none, targets may simply endorse the negative stereotypes and enact them. If there are such alternatives, targets may engage in a collective strategy of stereotype change. They are then likely to enact more positive stereotypes of their group and elicit behavioral disconfirmation. Third, we examined how the target's membership in a stigmatized group can affect the very structure of the interaction with a nonstigmatized perceiver. Specifically, we reviewed empirical evidence showing that members of stigmatized groups tended to occupy positions of lesser interpersonal status and power when interacting with nonstigmatized group members and that this inferior position makes them particularly vulnerable to behavioral confirmation. Based on evidence drawn from expectation states theory, we showed that this power imbalance concerns both formal and informal aspects of status and power. Finally, after having considered the impact of stigmatization on stereotype confirmation during interpersonal and intergroup interactions, we addressed the question of how stereotype confirmation and disconfirmation during these contacts can actually affect stereotype persistence and change. Based on theories of intergroup contact, we proposed that such an influence of the interpersonal on the intergroup level will hold only to the extent that the perceiver construes the contact as intergroup rather than interpersonal. This, we suggested, will be facilitated to the extent that the target pursues a collective strategy. This analysis has been consistent with the idea that stereotype confirmation processes are heavily influenced by factors defined at an intergroup level (such as permeability of group boundaries, group differences in power and status, as well as the perceived legitimacy and stability of these differences). Further, we have also suggested that these processes could play a role in the maintenance and change of intergroup relations. Nevertheless, in accord with the quotation that opens this section, we have argued that to understand this influence of the "micro" on the "macro" level, one must take into account the interactions between the sociostructural level of analysis as they affect the contexts in which dyadic interactions take place and the processes occurring at the level of these interactions themselves. This interactive standpoint has at least two major implications with respect to traditional social psychological research and theory on behavioral confirmation. First, it implies that the study of behavioral confirmation processes can, at least partially, inform an understanding of social processes occurring at superordinate levels of analysis and, hence, serve to address global social problems. Second, it implies that conceptualizations of behavioral confirmation processes should take into account the intergroup contexts in which the confirmation of stereotypes may take place. For such intergroup contexts can and do constrain the settings in which interactions between individuals take place as well as the expectations held by the interacting partners. Obviously, these two implications are complementary. It is only to the extent that it is informed by an understanding of the societal context in which interaction takes place that behavioral confirmation research can contribute to an understanding of these wider social problems. For example, an issue that may benefit from such an interactive analysis is soccer violence, a phenomenon that is seen as extremely prevalent among English supporters. Typically, the mass media portray supporters of the English soccer team as dangerous "hooligans" (see, for example, Murphy, Dunning, & Williams, 1991). In turn, these stereotypes may become the basis of the actions performed by non-English people when they interact with their targets out of their country. For example, Stott, Hutchinson, and Drury (2001) have shown that during the World Cup held in France in 1998, local youth, but also the police, tended to behave aggressively and indiscriminately toward these English fans, probably as a result of these stereotypes. This observation is consistent with a laboratory study showing that perceivers expecting a target to be hostile tended to behave more aggressively toward this target than perceivers interacting with a target who had not been described as such (Snyder & Swann, 1978). In this study, targets who had been randomly labeled as hostile did indeed behave more aggressively toward the perceivers who had so labeled them. Just like these targets, the supporters of the English team, who until their interaction with the local youth had been peaceful, felt unjustly treated and retaliated (Stott et al., 2001). Their behavior, which was publicized in the media, therefore served to reinforce the stereotype of the English supporter as a "hooligan". copy; 2003 Elsevier Science (USA). All rights reserved.