Rules and similarity in adult concept learning

This chapter reviews studies of rule-based category learning within the human adult literature, and contrasts these results with evidence for similarity-based accounts of category learning (e.g. exemplar and prototype models). Specifically, the chapter considers the contrast between rule-based versus similarity-based learning within research on unsupervised (spontaneous) categorization and supervised categorization. In the closing sections, it also presents and critically evaluates hybrid models of human adult categorization that are composed of both a rule and a similarity component.