Peter Morrall, a senior lecturer in health and sociology, claims to have written “a polemic against the unified voice of conservatism and progressive viewpoints within the mental health industry” concerning homicides by people with mental illness. To support this claim Morrall offers us the following propositions: mental illness is a real entity; patients are at greater risk of committing suicide than homicide; the repercussions of psychiatric homicides are profound; killings are not caused by labelling theory or by moral panic; and both patients and public need protection. Few readers will find anything polemical in any of that.

Madness and Murder is a book of disconnected parts that fails to deliver the polemic the author promises. Chapters on mental illness, deviance, crime and homicide have the feel of an undergraduate text. The long-running debate between individualist and societal theories of crime is given an airing, while Dadd, M'Naghten, Foucault and Szasz duly make appearances. But how all this affects today's psychiatric homicides is not easy to see.

Morrall reserves the final chapter (entitled ‘The terror’) for his main point. He claims that psychiatrists caused the media panic about psychiatric homicides in the 1990s by their defensive attitude. He studied newspaper reports — he calls them a “catalogue of killings” — between 1994 and 1999. In 13 pages he lists 94 killings, not all of them in the UK, and a further 27 near-killings. Morrall acknowledges that newspaper reporting of these cases is “sloppy, careless and injudicious”, but he emphasises their frequent allusion to current or previous contact by the perpetrator with mental health services. He concludes that it is because psychiatrists interpret this “reporting of their professional gaffs as media orchestrated panics” that the panic took hold. And that claim is the nearest we get to a polemic.

It seems to me that Morrall has missed an opportunity. Psychiatry must articulate the role it and other agencies have in the prevention of psychiatric homicides. I had hoped Morrall would offer some suggestions on what the profession (or, if he prefers, ‘the industry’) should do. Is further restriction of those with mental illness feasible? Is it justified? Will it be effective in reducing risk? On these questions Morrall's polemic is deafeningly silent.