Running Title: Distinguishing GCM from PRT

Exemplar and prototype accounts of categorization phenomena differ primarily in terms of the manner in which category information is represented. However, the models also make auxiliary assumptions that are instantiated as model pa-rameters. Smith and Minda (1998) argue that the response scaling parameter γ in the exemplar-based Generalized Context Model (GCM) makes the model unnecessarily complex, and allows it to mimic prototype representations. We estimate the complexity of GCM with and without the γ parameter, as well as that of a prototype model. We then go on to assess the extent to which the models mimic each other’s behavior. The γ parameter does increase the complexity of the model, but this complexity only allows partial mimicry. Fur-thermore, if we adopt Minimum Description Length as the measure of model performance, the models become highly discriminable. Distinguishing GCM from PRT 2 An important question in categorization research is whether category informatio