View article

Similarity is a central construct in perceptual and cognitive science, with implications for everything from basic color and pattern perception, to object and phoneme recognition, memory retrieval, analogical reasoning and problem solving. Previous studies have examined people’s overt similarity judgments to understand the roles of similarity in cognition, and several model have been proposed to account for the data. This paper presents our simulations of these data with a model originally designed to simulate analogical reasoning (Hummel & Holyoak’s 1997, 2003a, LISA model), not similarity judgments. We show that the same mechanisms that LISA uses to simulate analogy also provide a natural account of disparate data on similarity judgments. These successes speak to the utility of LISA’s symbolic-connectionist knowledge representations as an account of human mental representations.