Cogsci Colloq: Robert Nosofsky (Psychological and Brain Sciences, Indiana University)

Title: Exemplar Model Accounts of Relations and Dissociations Between Categorization and Recognition

Abstract:

In this talk I provide an overview of an exemplar-based evidence-accumulation model of categorization and recognition decision making and its applications to empirical data, including its accounts of intriguing dissociations that have been demonstrated between categorization and recognition memory. According to Nosofsky and Palmeri’s (1997, 2015) exemplar-based random-walk model, people represent categories by storing individual exemplars in memory. Presentation of a test item causes the exemplars to be retrieved. The retrieved exemplars drive an evidence-accumulation process that leads to both categorization and recognition decisions. In the first part of my talk, I illustrate applications of the model to predicting the time course of decision making in tasks of both categorization and recognition.

Whereas the exemplar model presumes that categorization and recognition are mediated by the same memory representations and decision processes, this idea has been strongly challenged by results from the cognitive neuroscience literature. For example, behavioral studies have demonstrated intriguing dissociations in which amnesic subjects, despite having severely impaired old-new recognition memory, perform at normal levels in categorization tasks involving the same kinds of stimuli. In addition, neuroimaging studies reveal that different parts of the brain are activated when observers engage in categorization versus recognition tasks. These findings suggest, in contrast to the fundamental assumption of the exemplar model, that separate memory systems may mediate categorization and recognition. In the second part of my talk, I describe how I have attempted to address these challenges.