The Influence of Learning to Distinguish Categories on Graded Structure

Research on the internal structure of categories has shown that central tendency, frequency of instantiation, and ideals are determinants of graded structure (Barsalou, 1985). Vast differences in the relative roles of these factors have been observed depending on the type of category, the immediate context of evaluation, and the domain knowledge or experiential history of the individual. In two experiments, using the simplest possible domain of novel artificial categories (lines of varying lengths), we tested whether the way in which a category is learned can shape its internal organization. Specifically, the impact of co-learning multiple categories within a classification scheme was assessed using typicality ratings. We found that the internal structure of a category took qualitatively different forms (i.e., graded structure based on ideals or central tendency; or even no graded structure at all) depending on the nature and number of the contrast categories during learning.