This book brings together well-known experts from different professional backgrounds to consider the many conceptual problems that arise in the creation of systems of psychiatric classification. It addresses core themes such as the nature of psychiatric disorders, the historical foundations of diagnosis, the benefits and harms of classification and the complex concept of validation.

It would be difficult to underestimate the importance of the themes discussed in this book. The publication of DSM-III in 1980 marked a step change in how psychiatric diagnoses are made. Since that time DSM-III and its successor volumes have defined the range, boundaries, language and discourses of psychiatry. They have defined how the domain of psychiatric disorders is subdivided. They have stipulated which symptoms and signs are criteria for the presence of one psychiatric condition or another. In so doing, they have relegated non-criterial symptoms and signs to relative desuetude. The fact that we are on the threshold of DSM-5 makes the publication of this book very timely.

DSM-III was explicitly atheoretical and operational in its definitions of psychiatric disorders. It eschewed causal theorising partly as a way of countering what was seen as the undue influence of psychoanalysis. It was operational in order to meet an urgent problem of the poor reliability of psychiatric diagnoses.

Improvement in diagnostic reliability was an essential precursor of scientific progress. Adequate reliability is required to compare the results of research between one centre and others and to allow replication of research studies. Research has to start with ‘something’ to investigate. A system of classification, however tentative, is a necessary first step in the process. A diagnostic system also provides a language in which scientists can communicate with each other.

There are also downsides to what is described in this book as the ‘reliability first, validity second’ approach. One of these is that diagnostic entities become ‘self perpetuating feedback loops’. Although they may only be rough approximations to ‘real’ conditions (if such things can be said to exist), they become the phenotypes that are used in research. A genetic study of schizophrenia will be a study of DSM-defined schizophrenia. An antidepressant drug trial will be carried out on patients with DSM-defined depression. DSM categories determine what questions can be asked and therefore risk becoming a system that impedes rather than advances scientific progress.

The task we face is to retain the benefits of a system of classification and at the same time ensure that it does not create a conceptual tunnel vision. In one chapter, Paul McHugh sets out ideas on psychiatric diagnosis that he and his colleagues have developed over a number of years. He points out that the atheoretical approach has not yielded the gains in understanding that were anticipated more than 30 years ago. He argues that to make sense of psychiatric disorders we need to return to basic medical diagnostic principles such as localisation and process and to combine these with other concepts such as emergence. This model will have considerable intuitive appeal to clinicians and especially to those of us who can remember life before DSM-III.

These and other important themes are covered in this book. The material is complex but is made accessible by a number of factors. A main strength here is that the contributors are all pre-eminent experts in psychiatric nosology. The quality of writing and argument is uniformly excellent and the style is engaging. Each of the fifteen chapters is preceded by a short introduction from one of the editors and followed by an expert commentary.

Clinical care rests on the foundation of diagnosis. This book sets out to offer a comprehensive analysis of the concepts that we use to validate our diagnostic categories and succeeds admirably in this aim. It is important that clinicians are aware of the strengths and weaknesses of our diagnostic concepts. This book should be essential reading for all psychiatrists.