The debate between Parker and Anderson & Haddad (Reference Parker, Anderson and Haddad2003) neatly summarised contemporary thinking on the question of antidepressant effect. It was a pity, though, that they provided no discussion of any historical perspective. The wonderfully clear account provided by David Healy (Reference Healy2002), for instance, shows how the marketing tail of psychopharmaceuticals now often wags the entire dog. The process by which this came about has been gathering momentum since the early 1960s. Healy explores its various causes and corollaries in detail. It is not, he argues, due to any uniquely pernicious qualities of drug companies since similar trends can be seen in relation to some other types of therapy.

If this additional, temporal dimension had been taken into account, one suspects that Gordon Parker might have placed greater emphasis on one of the factors that he identified as contributing to the current situation: namely that ‘“depression” is currently modelled as a single entity, varying only in severity’ (p. 102). The term ‘depression’ is thus semantically equivalent nowadays to ‘abdominal pain’, not to ‘appendicitis’ or ‘peptic ulcer’. If trials of an antacid, say, were undertaken on patients selected for ‘abdominal pain’ the results obtained would sometimes be favourable, sometimes not. Debate over ant-acid usefulness would exactly parallel that over the effectiveness of antidepressants.

How did we get into this situation? It seems likely that a lot of the blame can be laid at the door of DSM–III (American Psychiatric Association, 1980), which explicitly aimed for reliability of diagnosis. Unhappily, there was an implicit downside. The state of the art in psychiatry, when DSM–III was under development, was such that reliability could be attained only at the expense of validity. Partly as a consequence of choices that were made then, this problem still remains. It is no good blaming the failings of clinical trials, the machinations of drug companies, the uselessness of antidepressants or reporting bias, for our predicament. The main fault lies in the consequences of a bad choice of diagnostic system, made by our predecessors for what seemed, at the time, good reasons. The remedy must lie primarily in seeing DSM for the hindrance that it is, and one day replacing it with a system that separates the ‘peptic ulcers’ from the ‘appendicitises’.