Exemplar models and category specific effects

In recent years, there have been numerous reports of patients with brain damage who show selective identification or recognition deficits for objects from specific categories (see Forde, in press; Humphreys & Forde, 2000, for reviews). The most common deficit appears to be a selective impairment in the identification of living things, accompanied by relatively unimpaired recognition or identification of artificial or non-living objects. However, despite the large number of reported cases with category-specific processing deficits, there is still no agreement on the mechanisms that produce these deficits. It is not even clear whether all such cases can be understood in terms of a single process or mechanism, or whether categoryspecific deficits can be caused by a variety of different factors. In this chapter, we explore category-specific deficits from a theoretical viewpoint that evolved from recent research on perceptual categorization and identification. Although some efforts have been made to model category-specific deficits with connectionist models (e.g., Farah & McClelland, 1991; Humphreys, Lamote, & Lloyd-Jones, 1995), we are not aware of any attempts to apply classical models of categorization and identification1 to the neuropsychological data on category-specificity (with the exception of a study by Dixon, Bub, & Arguin, 1997, which will be discussed in detail later).