Brian Blackwell got a disorder and got away with it, so why shouldn't you? He battered his parents to death with a claw hammer but, instead of imprisoning him for life, Mr Justice Royce allowed him to plead manslaughter on the grounds of diminished responsibility. He will be eligible for parole in five years and seven months, which, for a defendant who might never have been released, is a result.

Blackwell's lawyers argued that he was suffering from acute narcissistic personality disorder. They weren't simply saying that he was selfish - most killers are that - but that he had a verifiable illness which meant he was no more fully responsible for his crimes than a blind man would be fully responsible for bumping into you in the street.

According to the textbooks, a sufferer - or should that be a victim? - has five or more of the following traits: an extravagant sense of self-importance; fantasies of unlimited success; the belief that they are special and can be understood only by their equals; an unreasonable sense of entitlement; a yearning for admiration; a willingness to exploit others; lack of empathy; envy of others and the belief that others envy him; and a visible arrogance or hauteur.

It sounds frightfully specific, but look at that list again. How many people do you know with five or more of its grandiose vices? My diffident self aside, I would say it encompasses most managers of my acquaintance, all actors, many politicians and the entire senior editorial staff of the New Statesman.

After the Blackwell case, Oliver James, the most garrulous of the Fleet Street shrinks, boldly declared that 80 per cent of all convicted prisoners had a personality disorder, and 74 per cent had more than one. James didn't follow his argument that they were bad because they were mad to its conclusion, and say that 80 per cent of convicts should go to the Court of Appeal and demand that their sentences be cut because of diminished responsibility. But then his trade is rarely consistent.

It is not only criminals who should be able to cop a plea. The late 20th century brought an explosion of psychiatric diagnoses. The bible of therapy is the Diagnostic and Statistical Manual of Mental Disorders, written by the American Psychiatric Association. The first edition (DSM-I) was published in 1952. It was a pamphlet which listed a mere 60 disorders. At 134 pages, the 1968 second edition might have been mistaken for a novella. The third, in its revised version of 1987, had 567 pages and was longer than most novels. DSM-IV, the current dictionary of delusion, was published in 1994 and would be easier to handle if it had appeared in two volumes. It has 886 pages and even in paperback weighs 3lb 4oz. DSM-V will be out in 2011. No one is expecting a haiku.

Herb Kutchins and Stuart A Kirk, two sceptical American academics, argue that psychiatry has expanded by pathologising everyday life. To give a sense of how its empire has inflated, they took ordinary human troubles and showed how they could be, and were being, reclassified as disorders:

An ambitious friend is striving for promotion and says he's not sleeping well. (Not sleeping equals major depressive disorder.)

A former classmate says she has never forgotten how you humiliated her in the playground. (Bearing grudges equals paranoid personality disorder.)

A colleague won't go for a drink after work and spends his weekends at home by himself. (A desire for solitude equals schizoid personality disorder.)

A wife says that after ten years of marriage she is no longer aroused by the sight of her husband. (Absence of sexual interest equals hypoactive sexual desire disorder.)

Your sister cannot stop thinking about a boyfriend who has left her. (Infatuation equals obsessive-compulsive person-ality disorder.)

Your child won't listen to his teacher. (Lack of respect for authority equals oppositional defiance disorder.)

A husband says his wife worries all the time. (Nervousness about the future equals generalised anxiety disorder.)

The list could go on, and indeed does go on in DSM-IV for 886 pages. Whether you are happy or sad, neat or messy, chaste or promiscuous, bumptious or withdrawn, fat or thin, drunk or sober, you have the symptoms of a mental disorder. It's a mixed blessing. On the one hand, you can always be ill. On the other, if you are accused of being criminally ill, you always have an alibi.

The cynical thing is to say that the spread of psychiatry is very profitable for therapists - conservative estimates say there are 200,000 in Britain indulging in 300 varieties of therapy - and equally profitable for the makers of Prozac, Xanax and beta-blockers, who were particularly heartened by the announcement in DSM-IV that there were 12 million sufferers from generalised anxiety disorder in the United States, most of them women and all of them potential drug-buyers.

As important as following the money is to question the medical status of the definitions of disorder. You don't have to trail behind Foucault into the postmodern wilderness and pretend that all mental illnesses are social constructs to wonder what a therapist is saying when he says you have a disorder. There are chronic illnesses such as schizophrenia and severe depression, which presumably Foucault never encountered, but what about the ocean of ordinary human vicissitude, in which most shrinks paddle and nearly all disorders lie?

If one psychiatrist says you have narcissistic personality disorder, would every competent psychiatrist who examined you say the same? You can prove that you don't have a tumour by having a scan, but how can you prove that you are not a narcissist? If you can't, wouldn't psychiatry fail Popper's falsification test and be a pseudo-science? Above all, what does saying you are a narcissist predict about your behaviour? A doctor can warn a heavy smoker that he has a one-in-three chance of contracting lung cancer if he does not stop. Could a psychiatrist say to a sufferer from narcissistic personality disorder that the chances of his taking a claw hammer to his mum and dad are one in three, one in ten, or a thousand, or a million?

There is a long history of empirical scepticism about psychiatry which goes back to Freud's original critics. A proud place in it belongs to Thomas Szasz, a Hungarian refugee from Nazism who found sanctuary in America. In 1980, he testified for the prosecution in the case of Darlin Cromer, which was not so different from the Blackwell trial. As with Blackwell, no one disputed that Cromer had killed her victim, a five-year-old black boy. Witnesses described how she had talked of "killing niggers" and of it being "the duty of every white woman to kill a nigger child". None the less, the defence argued that she was a victim of "schizophrenic paranoia", and should receive a reduced punishment.

Szasz would have none of it. He wondered how a psychiatrist who interviewed a murderer days or weeks after the crime could say what had been in Cromer's mind at the time of the killing. Asked why she had murdered, all he could say was: "She was a bad student . . . She was a bad wife. She was a bad mother. She was a bad employee in so far as she was employable. Then she started to engage in illegal drugs, then she escalated to illegal assault, and finally she committed this murder . . . Life is a task. You either cope with or it gets you. If you don't know how to build, you can always destroy." (The jury convicted.)

Szasz's words sound harsh, but then it is equally harsh to section someone under the Mental Health Act. It's not a question of whether psychiatry is liberal or conservative, but whether the hundreds of disorders it claims to have discovered are real or not.

Reviewing a recent collection of essays on Szasz in the current issue of the American journal Reason, Jacob Sullum said that the most heretical point he made about his colleagues was that they were in a losing fight against advances in neurology, which would in the end explain real and terrible mental diseases by examining physical causes.

We are a way from there yet, but it's moving on. My guess is that, when our understanding of the brain improves, all those disorders will return to where they were before Sigmund Freud: sufferers will talk to friends or relatives or look within themselves. They may pay money for a psychiatrist's time, but they will have ever fewer reasons for believing they are talking to a scientist. I hope and half-believe that today's psychiatric mania will one day be seen as an example of imperial overreach: the last spectacular expansion of a doomed intellectual system.

"Just think," people will say, "in the old days you could get away with murder."