Patricia G. Devine, PhD, is Professor of Psychology at the University of Wisconsin, Madison. Before becoming Professor, she was a Visiting Fellow at Yale University and an Associate Professor at Wisconsin.

Dr. Devine received the Gordon Allport Intergroup Relations Prize from the Society for the Psychological Study of Social Issues in 1990 and the APA Distinguished Scientific Award for Early Career Contribution to Psychology in 1994. She is the author or co-author of several journal articles and is the coeditor of Social Cognition: Impact on Social Psychology (Academic Press, 1994). Her research interests include prejudice and intergroup relations, stereotyping, dissonance, and resistance to persuasion. Dr. Devine received her PhD in Social Psychology from Ohio State University in 1986.

Legal scholars, politicians, legislators, social scientists, and lay people alike have puzzled over the paradox of racism in a nation founded on the fundamental principle of human equality. Legislators responded with landmark legal decisions (e.g., Supreme Court ruling on school desegregation and the Civil Rights laws) that made overt discrimination based on race illegal. In the wake of the legislative changes, social scientists examined the extent to which shifts in whites' attitudes kept pace with the legal changes. The literature, however, reveals conflicting findings. Whereas overt expressions of prejudice on surveys declined (i.e., verbal reports), more subtle indicators (i.e., nonverbal measures) continue to reveal prejudice even among those who say they renounced prejudice. A central challenge presented to contemporary prejudice researchers is to explain the disparity between verbal reports and the more subtle measures.

Some reject the optimistic conclusion suggested by survey research and argue that prejudice in America is not declining; it is only changing form--becoming more subtle and disguised. By this argument, most (if not all) Americans are assumed to be racist, with only the type of racism differing between people. Such conclusions are based on the belief that any response that results in differential treatment between groups is taken as evidence of prejudice. However, this definition fails to consider intent or motive and is based on the assumption that nonthoughtful (e.g., nonverbal) responses are, by definition, more trustworthy than thoughtful responses. Indeed, nonverbal measures are assumed to be good indicators of prejudice precisely because they do not typically involve careful thought and people do not control them in the same way that they can control their verbally reported attitudes.

Rather than dismiss either response as necessarily untrustworthy, my colleagues and I have tried to understand the origin of both thoughtful and nonthoughtful responses. By directly addressing the disparity between thoughtful and nonthoughtful responses, our approach offers a more optimistic analysis regarding prospects for prejudice reduction than the extant formulations. To foreshadow, our program of research has been devoted to understanding (a) how and why those who truly renounce prejudice may continue to experience prejudice&endash;like thoughts and feelings and (b) the nature of the rather formidable challenges and obstacles that must be overcome before one can succeed in reducing the disparity between thoughtful and nonthoughtful responses.

Automatic and Controlled Processes in Prejudice

The distinction between automatic and controlled cognitive processes has been central to our analysis in prejudice reduction. Automatic processes occur unintentionally, spontaneously, and unconsciously. We have evidence that both low&endash; and high&endash;prejudiced people are vulnerable to automatic stereotype activation. Once the stereotype is well learned, its influence is hard to avoid because it so easily comes to mind. Controlled processes, in contrast, are under the intentional control of the individual. An important aspect of such processes is that their initiation and use requires time and sufficient cognitive capacity. Nonprejudiced responses require inhibiting the spontaneously activated stereotypes and deliberately activating personal beliefs to serve as the basis for responses. Without sufficient time or cognitive capacity, responses may well be stereotype-based and, therefore, appear prejudiced.

The important implication of the automatic/controlled process distinction is that if one looks only at nonthoughtful, automatic responses, one may well conclude that all white Americans are prejudiced. We have found important differences between low&endash; and high&endash;prejudiced people based on the personal beliefs that each hold, despite similar knowledge of and vulnerability to the activation of cultural stereotypes. Furthermore, low&endash;prejudiced people have established and internalized nonprejudiced personal standards for how to treat members of stereotyped groups. When given sufficient time, low&endash;prejudiced people censor responses based on the stereotype and, instead, respond based on their beliefs. High&endash;prejudiced people, in contrast, do not reject the stereotype and are not personally motivated to overcome its effect on their behavior.

A strength of this approach is that it delineates the role of both thoughtful and nonthoughtful processes in response to stereotyped group members. Eliminating prejudice requires overcoming a lifetime of socialization experiences, which, unfortunately, promote prejudice. We have likened reducing prejudice to the breaking of a habit in that people must first make a decision to eliminate the habit and then learn to inhibit the habitual (prejudiced) responses. Thus, the change from being prejudiced to nonprejudiced is not viewed as an all or none event, but as a process during which the low&endash;prejudiced person is especially vulnerable to conflict between his or her enduring negative responses and endorsed nonprejudiced beliefs. For those who renounce prejudice, overcoming the "prejudice habit" presents a formidable task that is likely to entail a great deal of internal conflict over a protracted period of time.

Prejudice With and Without Compunction

In subsequent work, we examined the nature and consequences of the internal conflict associated with prejudice reduction. Specifically, we have focused on the challenges faced by those individuals who have internalized nonprejudiced personal standards and are trying to control their prejudiced responses, but sometimes fail. We have shown that people high and low in prejudice (as assessed by a self&endash;report technique) have qualitatively different affective reactions to the conflict between their verbal reports concerning how they should respond in situations involving contact with members of stereotyped groups and how they say they actually would respond. Low&endash;prejudiced people, for example, believe that they should not feel uncomfortable sitting next to an African American on a bus. High&endash;prejudiced people disagree, indicating that it's acceptable to feel uncomfortable in this situation. When actual responses violate personal standards, low&endash;prejudiced people experience guilt or "prejudice with compunction," but high&endash;prejudiced individuals do not. For low&endash;prejudiced people, the coexistence of such conflicting reactions threatens their nonprejudiced self&endash;concepts. Moreover, these guilt feelings play a functional role in helping people to "break the prejudice habit." That is, violations combined with guilt have been shown to help low&endash;prejudiced people to use controlled processes to inhibit the prejudiced responses and to replace them with responses that are based on their personal beliefs.

Interpersonal Dynamics of Intergroup Contact

Until recently, our research has focused rather exclusively on the nature of internal conflict associated with prejudice reduction efforts. However, many of the challenges associated with prejudice reduction are played out in the interpersonal arena, and we believe it's important to explore the relevance of our work to issues of intergroup tension. Thus, one of our current lines of research is devoted to exploring the nature of the challenges created by the intergroup contact when people's standards are "put on the line."

In interpersonal intergroup contact situations, we have found that although low&endash;prejudiced people are highly motivated to respond without prejudice, there are few guidelines for "how to do the intergroup thing well." As a result, many experience doubt and uncertainty about how to express their nonprejudiced attitudes in intergroup situations. Thus, for low&endash;prejudiced people, their high motivation to respond without prejudice may actually interfere with their efforts to convey accurately their nonprejudiced intentions. Under these circumstances, they become socially anxious; this anxiety disrupts the typically smooth and coordinated aspects of social interaction. Their interaction styles become awkward and strained resulting in nonverbal behaviors such as decreased eye contact and awkward speech patterns. These are exactly the types of subtle responses that have typically been interpreted as signs of prejudice or antipathy. Indeed, it is not possible to distinguish between the type of tension that arises out of antipathy toward the group or social anxiety based on these signs alone.

We argue that it may be important to acknowledge that there are qualitatively distinct forms of intergroup tension experienced by majority group members, which are systematically related to their self&endash;reported level of prejudice. For some, the tension can arise out of antipathy, as was always thought in the prejudice literature, but for others, the tension arises out of anxiety over trying to do the intergroup thing well. Functionally then, we have different starting points for trying to reduce intergroup tension. Strategies for attempting to reduce intergroup tension differ when the problem is conceived as one of improving skills rather than one of changing negative attitudes.

Conclusion

To sum up, although it is not easy and clearly requires effort, time, and practice, prejudice appears to be a habit that can be broken. In contrast to the prevailing, pessimistic opinion that little progress is being made toward the alleviation of prejudice, our program of research suggests that many people appear to be embroiled in the difficult or arduous process of overcoming their prejudices. During this process, low&endash;prejudiced people are confronted with rather formidable challenges from within, as people battle their spontaneous reactions, and from the interpersonal settings in which people's standards are put on the line. We are sanguine that by developing a realistic analysis of the practical challenges faced by those who renounce prejudice, we may be able to identify strategies that may facilitate their prejudice reduction efforts.

It is important to recognize that we are not claiming to have solved the problem of intergroup prejudice, nor are we suggesting that prejudice has disappeared. The past several years have witnessed a disturbing increase in the incidence of hate crimes against minorities. And a sizable proportion of white Americans continue to embrace old&endash;fashioned forms of bigotry. Nevertheless, we hope that by developing an understanding of the challenges associated with breaking the prejudice habit, we may gain insight into the reasons low&endash;prejudiced people establish and internalize nonprejudiced standards. Armed with this knowledge, we may be able to encourage high&endash;prejudiced people to renounce prejudice. And when they do, we will be in a better position to understand their challenges and, perhaps, to assist them in their efforts.

Published in Psychological Science Agenda.

Back to Top