Testing the interactive two-step model of lexical access: Part I. picture naming

Abstract

The weight-decay (WD) and semantic-phonological (SP) models offer competing accounts of how brain damage disrupts processing within an interactive 2-step lexical-access framework. A few studies have compared these deficit models for how well they fit individual naming profiles on the Philadelphia Naming Test (PNT) (Dell, Lawler, Harris, & Gordon, 2004; Foygel & Dell, 2000; Ruml, Caramazza, Shelton, & Chialant, 2000). The results, good for both, offer little basis for choosing among them. Moreover, the numbers of patients whose profiles were fit in these studies was small. The largest had n =2 1 (Foygel & Dell, 2000) and was potentially biased by exclusion of nonfluent aphasics and those with many responses outside the standard error categories. Here, the WD and SP models are compared with data from a new sample that is larger and unbiased in those respects.