The linguistic intergroup bias as an implicit indicator of prejudice

The Linguistic Intergroup Bias (LIE) is the tendency to describe stereotypic events in more abstract terms than counterstereotypic events. We examined whether a paper-and-pencil measure based on the LIE could be used as an implicit indicator of prejudice. Experiment 1 demonstrated that a measure of implicit racial prejudice based on the LIE predicted whether subjects evaluated an African-American or a Caucasian as threatening. Experiment 2 extended this finding with indirect measures of threat. In both of these experiments, an explicit measure of prejudice failed to predict subjects' evaluations. Experiment 3 demonstrated that a measure of implicit gender prejudice based on the LIE was correlated with an implicit prejudice measure based on biased attributional processing. Across these three experiments, implicit and explicit measures of prejudice were largely uncorrelated. (C) 1997 Academic Press.