Abstract
Background: The compilation of DSM-5 presented a substantial opportunity to develop a coherent, evidence-based classification of personality disorder. The irremediable problems with DSM-IV are widely recognized, the field seemed ready for change, and the data and methods for constructing a scientific classification are readily available. Rather than seize this opportunity, DSM-5 advanced an incoherent proposal lacking in evidential support and too poorly organized for clinical use. Methods: This article examines the problems with the proposal based on a consideration of the basic requirements of a satisfactory classification. It is suggested that an adequate system should have an explicit and coherent conceptual structure, be based on the best available scientific evidence, possess clinical utility, and be as parsimonious as possible. Results: The DSM-5 proposal fails to meet these criteria. Problems with the product and process suggest the need for a radical reconsideration of how personality disorders are classified and how classifications are compiled. Conclusions: The article proposes that greater emphasis be placed on developing a classification that provides the diagnostic information clinicians need to treat personality disorder and that future classifications should be constructed through an explicit process that is open to public scrutiny.