The reliability and validity of psychiatric diagnoses have always been a major concern. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) reliability field trials yielded ambiguous results, with some diagnostic categories scoring well below par. We argue that the emphasis on the reliability of psychiatric diagnoses, which has dominated psychiatric nosology and guided the endeavor of improving the DSM in its consecutive editions, is misguided and lacks in structural validity. In this article, we defend a pragmatic view on psychiatric disease as the most fruitful approach to an understanding of what the categorical distinctions in the DSM (can) represent. Disorders in the DSM are descriptions of clinical pictures and do not necessarily correspond to an identified pathological substrate. Although this is a logical result of the nature of psychiatric disease, it bears important consequences. The various DSM disease categories are not uniform but should be regarded as representing different kinds of disorders, ranging from a separation from normal behavior based on practical grounds to the discrete kind of disorders envisioned by proponents of a strong realistic view. We argue that the explication of kinds of disorders outlined in this article provides interesting perspectives on the problems of reliability and validity that the DSM faces.
Keywords: DSM-5; nosology; pragmatism; reliability; structural validity.
© 2016 John Wiley & Sons, Ltd.