Tossed in an acronym salad

I am a great admirer of Ian Hacking. His history of probability, The Taming of Chance, is a model of lucidity. Philosopher he may be, but his handling of both the mathematical ideas and the social framing of statistical theory made that book valuable reading far outside the normal academic frame. Now he has turned his attention to an even more philosophically, clinically and socially troubling phenomenon: the dramatic growth over the past decades, especially in the United States, of individuals claiming to have "multiple personalities" and of the therapists who interpret, support, encourage and theorise them.

In attempting this task, Hacking is nothing if not ambitious. His text is at least three books in one. The first is relatively straightforward, a contemporary history of the rise of the "multiple personality movement", of and in which he would appear from his footnotes to have been a participant observer. The second is more historical and distinctly less obvious. Its thesis is that in the modern world the idea of memory has replaced that of the soul and that the birth of the "sciences of memory" can be traced to France in the ten-year period beginning in 1874. The third, most obviously "philosophical" theme, yet the one to which Hacking devotes the least attention, is the implication of claims of multiple personality for concepts of personhood and identity.

The current growth industries engaged in the pathologising of human behaviour have diagnosed an alarming range of previously unknown syndromes: hyperkinesis, minimum brain dysfunction, attention deficit disorder, Munchhausen syndrome by proxy, recovered memory, multiple personality disorder (MPD), dissociative identity disorder (DID)Ithe list seems endless, its protean forms each given brief authority by the US diagnostician's bible, the Diagnostic and Statistical Manual, now in its fourth incarnation. Some are the province of biological psychiatry and the drug companies, others of the burgeoning army of therapists and their rival theories of both origins and treatment. Recovered memory and multiple personality -or DID, to give it its currently accepted name - have been in the eye of the storm both because of the dramatic nature of their claims and because of their disturbing social implications. These diagnoses have been particularly attractive to therapists committed to the view that both syndromes originate in efforts by the abused child - most frequently a girl abused by an older man, a father, relative or friend - to find the psychic means to survive the experience. At their wilder reaches recovered memories extend to allegations about networks of satanic cults and child murderers. In opposition the parents who stand so accused in turn pathologise their accuser, said to be suffering from false memory syndrome.

The case for and against the "reality" of each of these conditions has been argued through the clinical and scientific press, in books of advocacy and increasingly in court cases and judicial inquiries. Hacking professes neutrality, though sometimes with a less than entirely straight face. His concern is different, for what intrigues him is the meaning of the extraordinary growth in the numbers and complexities of the reported cases of DID, especially over the past two decades.

He tracks their origin to occasional 19th-century accounts of individuals suffering from "dual personalities". The typical case would for most of the time appear orderly and rather repressed, but would on occasion fall into a prolonged trance-like sleep to awaken with no memory of her earlier state and with a quite different repertoire of behaviour, often outgoing and sexually freer. Once having "switched", the person would live for some time with the new character and then almost arbitrarily switch back again with no apparent memory of the intervening state. This was the period of Broca, of Ferrier, of debates about localisation of motor and speech activity, and, above all, with the limited technological and theoretical tools available, of the attempt to infer brain function from dysfunction. Thus these rare conditions naturally intrigued the protoneuroscientists of the period, who tied them to then-fashionable theories of cerebral liberalisation. For Hacking, though I find his case unconvincing, it is the time of the death of the soul and the birth of memory science.

But dual personalities had a relatively brief run for their money. Freud and his successors largely abolished the diagnosis, along with hysteria. New categories such as schizophrenia emerged. Despite its ostensible translation as split brain, schizophrenia, Hacking insists, is "real" and not to be confounded with the phenomenon of dual personality. MPD and DID were to remerge only in the transformed climate of late 20th-century America, in the context of intensive lobbying by therapists and therapies alike.

What was a rare condition in 19th-century France reached 6,000 in the US by 1986, and today there are, Hacking says, hundreds of such cases in every sizeable American town; it is an epidemic. Not only have the numbers of cases grown dramatically, but so has their complexity. No longer do sufferers have just two personalities, but many dozens, often only "revealed" during therapy. How can this be? The sceptical answer is that the "personalities", like false memories, are somehow implanted in suggestible clients. The therapists' response is that in the past only the very tip of iceberg of child abuse was revealed against massive denial; it took the new therapies and new feminist consciousness to expose it.

For Hacking the issues are less crudely polarised. He argues that the condition is indeed created, in that distressed people learn how to express their suffering in ways channelled by the therapeutic notions of the time; possession in the middle ages, hysteria in the 19th century, and acronym salad today. Such learned ways of being raise deep questions about the nature of identity which it becomes the proper province of the philosopher to explore.

As fascinating and significant as the phenomena Hacking describes are and as clearly as he writes, the agenda of his book is unclear and its theses concerning the soul and memory are unconvincing. There are brilliant passages, such as the superb account of the probability of false positives in screening procedures, that deserve extracting and circulating through every therapeutic practice in the country. But there are too many disingenuous disclaimers, too obtrusive a first person singular without the author ever coming quite clean about his own personal motivations in becoming so heavily involved in the DID movement. Here Hacking's own multiple personalities, as critic, philosopher, historian and participant, add up to less than a whole.

Steven Rose is professor of biology and director, Brain and Behaviour Research Group, Open University.