More from my forthcoming book with Malte Ebach. Last post for the year, folks.
The Diagnostic and Statistical Manual of Mental Disorders
This text, known by its acronym the DSM (-I, -II-, III or -IV, and -V due in 2013), is the main standard classification of mental illness and disorders used across the world, although it is primarily the production of the American Psychiatric Association (APA). It is not the sole classification, there being also International Statistical Classification of Diseases and Related Health Problems (ICD), produced by the World Health Organization. Both systems are widely used in the context of drug prescription, medical insurance and administration, government statistics, and medical education.
The original system that the DSM-I codified was based upon psychoanalysis categories. DSM-I evolved out of the pre-war “Statistical Manual for the Use of Institutions for the Insane” after the second world war, in 1952, relying strongly upon military terminology and practice during the war, especially the Navy’s. Over the years it was revised extensively, and by DSM-III the APA abandoned the goal of earlier editors and authors to find an etiology for the diseases classified. Since very little in the way of etiologies for the diseases had been uncovered, it was seen that the role of the DSM was to provide practitioners with a way to efficiently and effectively diagnose conditions, and prescribe drug treatments and other treatments.
This meant that the DSM was not a nosology the way classifications of diseases were in medicine. Although medical science might not know what the etiologies of diseases were, the aim and project of medical research and classification was to move from phenomenology to etiology, and when a disease was finally explained in a way that might break it up into several distinct or more general conditions, medical science had little trouble doing so. Psychiatry, on the other hand seemed to move in the opposite direction. Instead of explaining and revising categories based on a knowledge of causal substrate, psychiatry revised based on “general concepts” of mental illness, some of which were in fact lay notions, and upon the availability of drugs to prescribe. What etiological research there was tended to be done by neuro-psychologists and neurologists instead.
The DSM is a case of a classification that is moving away from Theory rather than to it, largely because it is not an attempt at a natural system, but one of convention and operational use. However, the majority of those who employ it seem to think it is a natural scheme. This may retard the progress of psychiatry, as Dom Murphy thinks:
… classification can draw on causal discrimination in the absence of causal understanding. And it can use causal discrimination as a source of hypotheses. If we have good reason to believe [through this classification] that two syndromes depend on different pathologies, then we can orient research around finding out what they are. … The system of classification in the DSM is incoherent, heterogeneous, and provincial. It is incoherent in that it rests on a theory about the taxa of interest that requires symptoms to be expressions of underlying causes whilst at the same time it prohibits mention of these underlying causes in the taxonomy. It is heterogeneous in that it does not classify like with like at appropriate levels of explanation. And it is provincial in that it is cut off from much relevant inquiry. These complaints … reflect worries about the current state of biological psychiatry as a whole, since DSM-IV-TR is its flagship. 
With the approval of the DSM-V in 2012, critics noted in review that it was an amalgam of convenience and in some cases putative special interests. 
Apart from a connection with the available treatments, not all of which are clinically or epidemiologically tested, there is a minimal change of emphasis upon etiology in refining the categories of the manual. Nevertheless, the mere having of a systematic classification generates substantial research programs, and the results of neurobiological research has input much of what etiology there is in psychiatry. 
1. This is based upon the comprehensive research and discussion in Murphy 2006. Although Murphy’s theoretical discussion of classification in chapters 9 and 10 is consonant with ours, we arrived at similar ideas independently and in distinct domains.
2. 2006: 323.
4. Kupfer and Regier 2011, Regier, et al. 2009.
Kupfer, David J., and Darrel A. Regier. 2011. Neuroscience, Clinical Evidence, and the Future of Psychiatric Classification in DSM-5. American Journal of Psychiatry 168 (7):672-674.
Murphy, Dominic. 2006. Psychiatry in the scientific image. Cambridge, MA; London: MIT Press.
Regier, Darrel A., William E. Narrow, Emily A. Kuhl, and David J. Kupfer. 2009. The Conceptual Development of DSM-V. Am J Psychiatry 166 (6):645-650.
Thanks to Dom for providing me with a copy of his wonderful book.