Much of the published reaction to the government's Reforming the Mental Health Act (Department of Health, 2000) has been ethical in tone, focusing upon whether it might be justifiable to detain people with personality disorders who have yet to commit a criminal offence. An implication has been that it is perhaps unreasonable to ask psychiatrists to treat the behaviours associated with personality disorders prior to conviction. However, an alternative view might be that, in these proposals, psychiatry is being hoist with its own petard.

At both the psychodynamic and biological ends of the speciality's spectrum there are research findings which suggest that personality disorder is a legitimate concern of psychiatry. Psychotherapists (particularly in the context of a therapeutic milieu) claim to be able to treat personality disorders. Neuroscientists report organic correlates (e.g. implicating prefrontal cortex). Hence, if one takes the evidence at face value, personality disorders are brain disorders amenable to psychotherapeutic intervention. Why should not a democratically elected government, concerned for the safety of its citizens, ask psychiatrists to do what we say we can do: treat mental disorder? After all, we detain other patients without them having to commit an offence.

Of course, the problem is that commentators from within the psychiatric profession either do not believe such findings or they do not wish to bear their consequences. Both the psychodynamic and biological accounts of personality disorder, if indiscriminately applied, appear to diminish personal responsibility. If personality disorder justifies mitigation in the forensic setting, then large numbers of people in society are walking about with a trump card, to be played should they ever go to court. This is not fanciful: there are many sophisticated patients who can effectively use this card to their short-term advantage in their dealings with members of community mental health teams. These individuals have carte blanche to commit immoral acts, an excuse, a reason (i.e. their personality disorder), and if they should murder or maim, it is the health professional who will be held to account. It has never been more important for the discipline of psychiatry to establish a coherent and consistently applied approach to agency and responsibility in the context of personality disorder.

Parenthetically, it is worth noting that there are severe limitations at both ends of the spectrum referred to above. The evidence that psychotherapeutic interventions treat personality disorder seems often to emerge from institutions with a vested interest in demonstrating success. While I do not question the integrity of the researchers concerned, perhaps their papers should include ‘declarations of interest’. Also, the applicability of their findings to the real world seems limited: violent subjects with comorbid substance misuse are rarely accepted for psychotherapy lest they act out. On the biological side, despite the demonstration of correlates with psychopathy, it has to be admitted that these findings have yet to differentiate cause from effect. We have good evidence that when children learn a musical instrument, then they change the structure and function of motor regions in their brain. Might not discovering, learning and practising immoral conduct affect other brain regions similarly? Aristotle may have pre-empted us: