Abstract

This article criticizes the approach to language underlying the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders . Concepts from the philosophy of language illuminate taxonomic problems that vex users of the DSM nosology: lack of coverage, comorbidity, and within-category heterogeneity. Exception is taken to the operationism that results in a highly artificial DSM nomenclature, raising the specter of non-referential criterion sets. A dimensional approach is recommended because it would better correspond to an objectively seamless reality