At a meeting of several DSM-V workgroups in Washington, DC in April 2008, Gavin Andrews approached me with the proposal that the set of papers he and his colleagues were working on – proposing and defending a new meta-structure for DSM-V – be published in Psychological Medicine. I expressed interest and in consultation with my co-editor, Robin Murray, we agreed on a basic set of guidelines for this venture. First, all the individual papers would be peer- reviewed as for any other review submitted to Psychological Medicine. No promises were made about the eventual outcome of this review process. Second, we would commission a series of independent commentaries on this set of articles. Third, we would make every effort to publish all the accepted papers and commentaries in a single issue, and if possible in 2009.

We were able to stick by all elements of this agreement. All the submitted papers were eventually accepted although several had to go through more than one round of reviews as authors and reviewers worked out their differences. Many senior individuals worked hard on these reviews and I want to express my appreciation of their anonymous efforts as well as the excellent coordination of all of this by Jill Opalesky and Barbara Herrmann. We were pleased that all three senior individuals that we approached to write commentaries (Michael First, Assen Jablensky and Uli Wittchen) – chosen to represent a diversity of perspectives and experience – accepted the task.

Working on this series was a fascinating editorial experience. Many diverse issues were raised in the 40 reviews performed as part of this effort. Of the many issues raised, perhaps the most common, and the one that generated the most discussion, was the evidential standard to which the reviews should be held. Were these to be scholarly ‘full-bore’ literature reviews covering the relevant literature in a thorough fashion trying to marshal all available evidence for and against the proposals? Or, were they to be ‘thought-pieces’ or ‘expert opinion’ that had the goal of stimulating the field with the help of selected empirical evidence? Some (but not all) reviewers took the former position while the authors consistently took the latter. As editor, it was my task to try to adjudicate between these two perspectives.

These reviews together represent a substantial body of conceptual and empirical work. The authors propose a set of validating criteria and attempt to judge the success of their revision by those criteria. Many issues were raised. A selected short list would include: (a) what should be the guiding principle for a ‘meta-structure’ for a psychiatric nosologic system? A major problem is that the two leading candidates – ‘clinical similarity’ and ‘etiology’ – need not always result in the same meta-structure. A third possible approach – clinical utility (which might, for example, suggest putting all the child onset disorders together) – would yield yet a different pattern. (b) What level of evidence, distributed as it will be across varying classes of validators, is needed to change the meta-structure? (c) To what extent should the conceptual basis of the overarching categories be the same across all categories or differ depending on the kind of disorders included? (d) If we do use etiology as an organizing principle, what level of causal factors (e.g. genes, neurobiology, neuropsychology, personality, etc.) do we want to utilize? (e) What are the relative merits of a ‘splitting’ versus a ‘lumping’ approach to a diagnostic meta-structure?

I hope readers will find these papers stimulating. No one will claim that the outcome of this effort is definitive. But, we hope it has moved the discussion forward, clarifying some of the weighty issues involved. Psychiatric nosology is so challenging because it inevitably involves the meeting of empirical and conceptual/philosophical issues. Neither can be avoided. Furthermore, unlike in research, we do not have the luxury of waiting until some future date when all might be clear. Revisions need to be completed. Deadlines need to be met.

The authors of these papers have been invited to respond to the three commentaries in a future issue and have accepted that invitation. At least one more chapter in this discourse is yet to come.