ABSTRACT
Einstein’s reflections on social classification are instructive for students of intergroup perception. While the most obvious lesson is that his status as an ingroup member depends upon the value of his theory, there are two other ideas implicit in this brief quote that deserve attention, one more obvious than the other. The first simply notes that the same individual may be accurately described by a number of different category labels. Whatever the status of his theories, Einstein can fairly be characterized as a “citizen of the world,” and a “German,” and a “Jew.” Indeed, there is an inclusive or at least partly hierarchical relation among the categories. Citizen of the world subsumes German, and being a Jew overlaps with being German. The assumption that a person belongs to multiple categories is obvious, but the tone of the above quote invokes a second assumption, that membership in one category precludes membership in another. Einstein indicates that Germans will view him either as a German or as a Jew, and the French will view him either as a citizen of the world or as a German. Whereas the realities of the categorical relations involve inclusivity and overlap, the psychological representation of these categories implies exclusivity. It is the psychological representation of categorical relations-and not the external reality of those relations-that constitutes the focus of this paper. The above quote illustrates once again the importance of the stimulus error in psychology, which refers to the assumption that the perception of an object is determined solely by the characteristics of the stimulusthat a stimulus can be perceived in one and only one way. As applied to the study of social categories, the
stimulus error appears to have two corollaries. The first is that a stimulus person’s category membership is necessarily included in the psychological representation of that individual. Put less abstractly, a woman is represented as a woman, a black as a black, a Jew as a Jew, etc. Indeed, there is now good evidence that this assumption is not tenable. Early work by Asch (1940) on the meaning of the category “politician,” and more recent work by Lord and Lepper (1999) on the malleable relation between category and category exemplars has nicely shown that judgments about the nature of politicians, for example, depends strongly on the nature of the exemplars that are momentarily activated in memory, and that the same exemplar can be represented by a number of different category names. It could be argued that the fuzziness of such categories as “politician” or “statesman” allows fluidity in representation, but not so for categories with clearly established boundaries, such as gender or race. If we take as evidence for categorical representation the ability of a category label to activate or “call to mind” the relevant category member, then there is good evidence that women or blacks, for example, who are poor fits to their respective stereotypes are functionally excluded from membership in these categories (Rothbart & John, 1985; Rothbart & Lewis, 1988; Rothbart, Sriram & Davis-Stitt, 1996). Thus, even “obvious” group members may not necessarily be encoded as category members if the overall fit to the stereotype is poor. The important point is that it is not the clarity of the group boundaries (for example, with gender) or the societal importance of the category features (for example, with race) that necessarily determines categorical representation, but the degree of match between the individual and the group stereotype.
