Abstract

The third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III) and its descriptive psychiatry-based intellectual antecedents imagined psychiatric disorders as discontinuous categories, presumably natural kinds, that would be empirically validated based on future scientific studies. Validation would emerge from a predicted convergence of clinical descriptions (symptom clusters that could be shown to be stable over the lifespan), laboratory results, and family studies. That future science is now arriving, but rather than validating the categorical DSM approach, large-scale genetics along with modern neurobiology and epidemiology have emphatically undercut it. Clinical description, laboratory studies, and family (now genetic) studies do not converge at all on distinct categories. Rather, modern studies are consistent with psychiatric disorders as heterogeneous quantitative deviations from health. The characteristics of these disorders have proven to be discoverable rather than invented and thus are grounded in nature. However, scientific results demonstrate that psychiatric disorders cannot reasonably be understood as discrete categories-and certainly not as natural kinds.