516 Prejudice, Racism, and Discrimination
Pettigrew and Meertens (1995) created separate multi-item
scales for blatant and subtle prejudice toward immigrants
and administered them to survey respondents from four
European countries with regard to several different target
groups. Across countries, confirmatory factor analyses sug-
gested that two-factor models surpassed a one-factor model,
but that a correlated two-factor model and a hierarchical model
in which blatant and subtle prejudice were first-order factors
subsumed under a general second-order factor were equally
viable models to account for the pattern of scale scores.
An advantage of using both subtle and blatant prejudice
scales is that a threefold typology emerged that yielded dif-
ferent patterns of responses to immigrants in Pettigrew and
Meertens’s (1995) research. Respondents who scored low on
both blatant and subtle prejudice scales were called “equali-
tarians,” a group who were most in favor of maintaining and
enhancing immigrants’ rights in their countries and who
presumably have internalized most strongly contemporary
norms of tolerance in their societies. Respondents scoring
high on both scales comprised “bigots,” who were most in
favor of returning immigrants to their home countries and
restricting immigrants’ rights and were assumed to have
rejected current norms against blatant prejudice. “Subtles”
were respondents scoring low on blatant prejudice but high
on subtle prejudice and were assumed to have only partially
and incompletely internalized norms against blatant preju-
dice. On immigration issues, “subtles” adopted a middling,
nonprejudicial stance between bigots and equalitarians and
required justification for restricting immigrants’ rights. The
“subtles” category, of course, is the analogue to symbolic,
modern, and aversive racism in that these people strive to
appear nonprejudiced and are assumed to express their biases
against immigrants in ways that do not violate current norms
against blatant prejudice.
Automatic and Controlled Processing
The Dissociation Model
As noted earlier in discussing ambivalence approaches to
prejudice, some prejudice researchers (e.g., Crosby et al.,
1980; Dovidio & Gaertner, 1986) have suggested that most
White Americans are prejudiced toward Black people and
that subtle behaviors that individuals can less readily monitor
and censor (e.g., helping, nonverbal behavior, reaction times
to briefly presented stimuli) are better gauges of White
Americans’ true racial attitudes. In an influential contribution
to the prejudice literature, Devine (1989) strongly challenged
and countered this view. She claimed that it implied that
prejudiced beliefs and attitudes were unamenable to change,
as well as that prejudice is an inevitable, unavoidable product
of normal cognitive processes.
As an alternative, Devine (1989) proposed a dissociation
model that emphasizes the importance of distinguishing
between automatic versus controlled cognitive processing
and the differentiation of stereotype activation versus per-
sonal beliefs. The automatic versus controlled processing dis-
tinction emerged in cognitive psychology during the 1970s
and subsequently has become an increasingly important con-
struct in social and personality psychology (see Bargh, 1989).
Automatic processing refers to unintentional, nonconscious
cognitive processing that occurs without effort or intention
and is unlimited by cognitive capacity. By contrast, con-
trolled cognitive processing refers to intentional, effortful,
and goal-directed processing of information that is assumed
to be under the person’s awareness and control but subject to
limitation by cognitive capacity (e.g., attentional limits). Ap-
plying this distinction to the relationship between stereotyp-
ing and prejudice, Devine (1989) suggested that stereotype
activationwas an automatic process that did not require in-
tention, attention, or cognitive capacity on the part of a per-
ceiver. Instead, whenever an appropriate cue is present, such
as the appearance of a Black person or a symbolic represen-
tation of one, a White U.S. perceiver’s stereotype of Black
people should be activated automatically.
Devine (1989) proposed that common socialization expe-
riences in late-20th-century America have led White people
in the U.S. to become equally knowledgeable about the
prevalent and generally negative stereotype of Black people,
regardless of their personal levels of prejudice. As a conse-
quence of this common knowledge, her dissociation model
predicted that automatic activation of the stereotype would
be equally strong and unavoidable for White U.S. perceivers,
regardless of the extent of their personal prejudice toward
Blacks.
Prejudiced and nonprejudiced White persons, however,
were expected in the dissociation model to differ in their per-
sonal beliefsconcerning Black people, and this difference in
personal beliefs regarding Blacks should be manifested on
cognitive tasks involving deliberate, controlled cognitive
processing. Specifically, on such a task, nonprejudiced White
persons should inhibit and override their negative cultural
stereotype of Blacks because it conflicts with their egalitarian
values and their personal beliefs and to replace the pejorative,
Black stereotype with more positive perceptions and attribu-
tions of Black persons. On this latter point, Devine’s (1989)
analysis of nonprejudiced perceivers agrees with aversive
racism and modern-symbolic racism theories in positing a