Until the end of the nineteenth century, psychiatry generally used the relationship of symptoms to their provoking causes as an essential part of definitions of melancholic disorder. Although some kinds of cases, such as psychotic depressions, almost always displayed symptoms that implied disorder, diagnosticians understood that they had to consider context, because depressive disorder could often be symptomatically indistinguishable from profound normal sadness. In the late 1800s, the traditional contextual approach to diagnosis of depressive disorder began to divide into two distinct schools. On one side, Sigmund Freud and his followers emphasized the psychological etiology of all mental disorders, including depression, and their continuity with normal functioning. Adherents of this school studied and interpreted the patient+U0027s reported thoughts to surmise the existence of underlying unconscious pathogenic meanings and wishes. On the other side, Emil Kraepelin applied a classical medical model that examined the symptoms, course, and prognosis of depression and other disorders to define distinct physical pathologies. Kraepelin+U0027s approach inspired a cadre of researchers to translate it into a research program that often used statistical techniques to infer discrete disorders from manifest symptoms.
Many psychiatrists viewed the publication of the DSM-III in 1980 as finally resolving the struggle between the Freudian and Kraepelinian schools for the domination of psychiatric nosology largely in favor of Kraepelin+U0027s approach.1 We will see, however, that such a judgment is overly simplistic in many ways. Specifically with respect to depressive disorder, the DSM-III criteria in fact represented a rejection of key assumptions underlying both Freud+U0027s and Kraepelin+U0027s systems and an af rmation of a quite different research tradition that ignored the prior emphasis on contextual criteria.