Although racism has characterized humankind throughout history, its content and expression has changed with changing times. Prior to landmark court decisions and the civil rights movement in the second half of the 20th century, overt racist attitudes were not only tolerated but institutionalized in the United States. The Plessy vs. Ferguson decision by the U.S. Supreme Court in 1896 had declared in its separate but equal doctrine that segregation by race was an acceptable and lawful practice as long as each group was afforded equal facilities. Segregation was entrenched in American society by this decision and by a proliferation of Jim Crow laws that explicitly legislated separate facilities for whites and non-white persons, even in such situations as that of Plessy in which the person's background was multiracial with primarily white ancestry.
It was only after the school desegregation case of Brown vs. Board of Education of Topeka in 1954, the Civil Rights Act of 1964, the National Voting Rights Act of 1965, and other subsequent statutes that institutionalized racist policies were ruled to be illegal. and were no longer viewed as acceptable practice in the nation. In the Brown case it was recognized that "separate educational facilities are inherently unequal." And in the acts of legislation that followed it was declared that segregation of public accommodations, barriers to voting and equal representation, and discrimination in hiring practices, employment, and housing were unlawful practices punishable under the law.
With changes in the country's laws also came changes in its norms. The direct expression of prejudicial views was no longer tolerated in many regions of the nation, and racism became more implicit, tempered by legally mandated standards of behavior. Irwin Katz and his colleagues (Katz, Wackenhut & Hass, 1986) noted that "the new type of racist avoids expressing opinions that are blatantly antiblack or segregationist, preferring ones that are relatively ambiguous and amenable to being defended on nonracial grounds" (p. 37).
One consequence of the change in societal norms and racial attitudes is that these attitudes have become more complex. In some cases, the individual may only be committed to racial equality in a superficial manner for the purposes of complying to strong social norms and presenting the self in a favorable light. In other cases, there may be sincere feelings of sympathy towards disadvantaged groups in society and a need to appear unprejudiced in his or her own eyes. In either case, interracial interactions are likely to be uncomfortable if accompanied by feeling of intergroup anxiety, salient negative stereotypes about the other's race, or the perception of realistic or symbolic threats among racial groups (Stephan & Stephan, 2000). The presence of conflicting attitudes, and less direct expression of negative racial beliefs, feelings or intentions, contrast with the explicit forms of institutionalized racism of America's past.
Sociologists have noted that American society has traditionally supported two different values that may conflict--egalitarianism and individualism. Rokeach (1968) has similarly emphasized the importance of equality and freedom as core values defining national political movements and their leaders. Conflict between these values underlie much of our national political debates. However, most of the views regarding modern forms of racism in this country stress that there is real conflict not just at a societal level but within the individual as well.
Katz et al. (1986) argue that many white Americans are ambivalent in their interracial attitudes. On the one hand, they are inclined to be sympathetic towards African Americans based on egalitarian values, a recognition of past discrimination, and a belief in social justice. But there is also a strong belief in the Protestant ethic ideals of individual responsibility, hard work, and the personal freedom to choose, and be rewarded for, more virtuous behavior. Higher rates of crime, unemployment, poor academic achievement, and other negative outcomes, to the extent they are viewed as the result of African Americans' own actions, may reinforce negative stereotypes and attitudes toward them. Thus, there exists a duality of values that may underlie the conflicting interracial attitudes of many white Americans.
One demonstration of the conflict underlying interracial attitudes was provided in a study by Weitz (1972). She examined both nonverbal (e.g., voice quality and behavioral measures) and verbal measures (trait ratings) of acceptance towards white and black "interaction partners" by male undergraduates. No actual interaction took place but measures were obtained prior to participants' expected meeting with their partners. It was found that both the verbal and nonverbal measures were positively related when participants thought their partners were white; however, when their partners were supposedly black the verbal measures were inversely related to the nonverbal ones. In other words, with a black partner positive expressions of verbal acceptance were paired with negative displays of nonverbal behavior, suggesting conflicted behavior by white participants and the transmission of mixed signals to black partners. This outcome most likely would be uncomfortable for both white and black individuals and discourage further interaction between them.
Donnerstein et al.
The disassociation between direct and indirect expressions of affect have been shown in many other studies. For example, Donnerstein, Donnerstein, Simon and Ditrichs (1972) used a contrived learning paradigm in which participants thought they were helping to condition other participants on a verbal learning task. Their role was to administer shocks to the other person, the "learner," as a means of providing feedback following errors. In actuality, no shocks were actually given. The white participants were told in some cases that the learner was white and in other cases that he was black. There were also manipulations of other factors, including anonymity of the white participant and the opportunity for the learner in the first round to retaliate later when the two individuals switched roles in a second round. One of the primary measures was the amount of direct aggression towards the learner as reflected in the intensity of shocks administered to the learner. In addition, an indirect measure of aggression was obtained by recording the duration of the shocks.
Donnerstein et al. (1972) found that when the participants' identities remained unknown to the learner, or when there was no expectation of later role reversal, they administered more intense shocks to a black learner than a white one. On the other hand, when they were not anonymous, or when they expected potential retaliation, they displayed less direct, but more indirect, aggression towards the black learner. In other words, open and continued interaction with the black learner caused their negative attitudes to be expressed in a disguised manner rather than directly. Again, the picture emerges of prejudicial attitudes being suppressed in one channel of communication but leaking out through another, undoubtedly less closely monitored, one. The Donnersteins have replicated these findings and extended their work in numerous other studies.
Other researchers have echoed this idea that prejudice still very much exists in America, but is simply expressed less directly as a response to changing laws and social norms. Gaertner and Dovidio's (1986) theory of aversive racism is one such viewpoint. This "perspective assumes that given the historically racist American culture and human cognitive mechanisms for processing categorical information, racist feelings and beliefs among white Americans are generally the rule rather than the exception" (p. 61). Aversive racists are those that adhere to egalitarian principles, support policies that promote racial equality, and who view themselves as unprejudiced, but who nevertheless harbor negative feelings and beliefs towards African Americans. These researches differentiate their view from that of Katz by not assuming "the widespread existence of genuinely problack, favorable components of whites' racial attitudes that are independent of egalitarian values" (p. 62). Thus, Gaertner and Dovidio's notion of aversive racism appears to offer a less optimistic outlook for the amelioration of existing prejudice.
One key prediction from this theory is that racial discrimination will most evident among aversive racists when there are no salient norms in the setting dictating nondiscriminatory behavior or where there are other salient aspects of the setting that render the person's behavior ambiguous or difficult to interpret. In other words, in those contexts where there are alternative explanations for one's behavior other than prejudicial attitudes and beliefs, overt discrimination is most likely to be expressed. In contrast, where norms of fairness and equal treatment are salient, behaving contrary to these norms is especially suggestive of prejudice and would be expected to confront one's egalitarian self-image.
This prediction also seems to follow from the application of attribution theory. Social norms provide a situational factor that may explain behavior and reduce the inference of dispositional qualities in the actor (Jones & Davis, 1965; Kelley, 1971). To the extent that other potential explanations exist for one's behavior, it is possible to avoid attributions of racist attitudes to the self and all of the self-incriminations that they would imply. Moreover, going one step further, it might be argued that through this process of self-perception an individual could anticipate the attribution process in others and gauge whether certain behaviors would be diagnostic of prejudice in the setting. Such diagnostic information could provide a basis for self-presentation strategies on the part of the individual. To what extent people actually in engage in this level of analysis and consciously suppress direct expressions of prejudice, as opposed to having internalized goals that regulate behavior at a preconscious level (Moskowitz, Gollwitzer and Schaal, 1999), remains unclear. It seems reasonable to assume that either process may operate, depending on characteristics of the individual and external factors (e.g., the degree experience in interracial settings).
Frey & Gaertner
Despite these open questions about the internal processes underlying such behavior, Gaertner and Dovidio have obtained strong support for their prediction. For example, in one study (Frey & Gaertner, 1986) white female undergraduates worked on a Scrabble-like task within a group setting in which the opportunity arose for them to help either a white or black coworker. In some cases, the coworker required help due to difficulty of the task (external cause) whereas in other cases this person's own lack of motivation (internal cause) necessitated the need for assistance. Additionally, the request for help came either from the coworker herself or from another person who requested aid on her behalf.
It was expected that when the coworker caused her own difficulties with the task, she would be relatively undeserving of assistance, and there would be an excuse to avoid helping her. However, if the request for aid came from a third-party this would heighten normative beliefs about the appropriateness of helping. Thus, only in the condition where the black coworker asked for help directly, and her difficulty was caused by internal factors, would discriminatory behavior be possible without implying prejudice on the part of the participants. In the other cases, there would be a more obvious violation of normative demands that would signal racist attitudes.
The results strongly supported this reasoning. On a number of different measures, offers to help and the quality of help offered were all lower for the black coworker who was responsible for her own need of assistance and directly asked for help. In all other conditions treatment was equivalent for black and white recipients. Moreover, the factors that justified not helping the black coworker did not appear to significantly undermine the offering of assistance to a white coworker, suggesting that these conditions facilitated the expression of racist attitudes rather than simply causing reduced help to all persons in need.
This research portrays prejudice in a different light than that which operated in the institutionalized racist culture of the United States prior to the civil rights movement and legislation of the 20th century. With the prevalence of egalitarian values, and new laws prohibiting discrimination based on race, ethnicity, religion, or other social categories, persistent prejudicial attitudes and stereotypes can no longer be expressed directly. Instead, they appear to find expression through channels and in contexts which render such behavior less detectable or more ambiguous in its meaning.