Two studies investigated the pervasiveness of race as a social categorization and whether the organization of information around racial categories is sensitive to contextual factors. Both studies measured accentuation eflects (more intra-than inter-race errors) and own-group bias (fewer confusions between own-than other-group members) in person memory, using the paradigm developed by Taylor, Fiske, Etcoff and Ruderman (1978 and Neuberg (1990) and Brewer (1988) argue that category-based processes have priority over attribute-based processes. Given our limited cognitive resources, it is both simpler (requiring less effort) and more efficient (requiring less time) for a perceiver to use categorical information to process information about individuals belonging to a group, than it is to analyse each person on an individual basis. According to Fiske and Neuberg (1990), the stage of 'initial categorization' occurs immediately as a perceiver encounters information (whether in the form of skin colour, clothing, speech style etc) sufficient to cue a meaningful social category. This initial stage indicates the pervasiveness of social categorization, with basic types of categories (e.g. race) thought to be accessed for virtually any target. The research reported below explores the ubiquity of social categorization and whether the organization of information around social categories is sensitive to contextual factors such as the relevance of a topic to a social category, and perceivers' processing goals.The paradigm we use to explore these issues has its origins in the pioneering studies of Henri Tajfel(1959), who proposed a theory of social judgment that provided a purely cognitive basis for the accentuation of intergroup differences. According to Tajfel, the categorization of objects (including social groups) is based on similarities within and differences between categories. Assimilation and contrast effects refer, respectively, to the accentuation of these basic processes, whereby perceivers minimize within-group differences and, especially, exaggerate between-group differences. Although the former effect was not significant in the original study (Tajfel and Wilkes, 1963), it has been found in later studies and the accentuation effect has been generalized from objects to social groups (Lilli, 1970;Marchand, 1970;Tajfel, Sheikh and Gardner, 1964). The magnitude of such accentuation effects can be seen as an index of the extent to which information is organized around social categories.Taylor, Fiske, Etcoff and Rudennan (1978, Experiment 1) extended this paradigm by having white subjects listen to a tape recording of a small-group discussion and simultaneously observe slides of each speaker (three black men and three white men) as he talked. After the 15-minute discussion, subjects were given a list of all the suggestions that had been made, and a set of pictures of all the participants. Subjects were asked to indicate which speaker had made each suggestion. The idea behind the study is simple: if race is used as...