Bullmore et al ^{Reference Bullmore, Fletcher and Jones1} mount a defence of neuroscience in psychiatry, invoking history, a dawning golden age, Reil and Freud. Although ensuring that the curriculum for undergraduates and trainees should accurately reflect what is valuable for doctors wishing to understand and treat mental disorders, they do not fulfil the ambitions of their manifesto. Their argument against neuroscepticism is weak. Specifically, physical models for mental disorder imply a particular position on psychology which is known as analytical behaviourism, and which effectively denies the existence of mind as a reasonable concept. They may wish to advance this view but either do not realise it or do not say so.

Neuroscience is a materialist enterprise that generates and examines hypotheses about brain function, which may inform new ways of looking at mental life: but psychiatry cannot be ‘based’ on neuroscience without becoming neurology. If psychiatrists cease to occupy the no man's land of unknowability, others will. The point about reductionism is a parallel problem. The kind of conversation that psychiatrists engage in with patients could well be better informed by neuroscience, but the reason for contemporary ‘vague talk about neurotransmitters’ is that the innumerable diagnostic categories invented in psychiatry bear no relation to discrete pathognomonic anomalies: nobody would base a diagnosis of schizophrenia on a brain image whether functional or structural. Patients do not need to see their brains light up to know that they are experiencing voices. In psychiatry there is an underrated crisis of validity, which many get around by claiming that psychiatry is where the rest of medicine was before the discovery of microbes and so on. An alternative view would be that schizophrenia, for example, is indeed a ‘functional’ disorder: an illness but not a disease, an illness that is culturally plastic and to a great extent subjective in its essence.

Finally, the authors claim to refute the allegation that neuroscience is relatively bereft of therapeutic achievement. They fail to provide a single example of a ‘neuroscientific’ novelty since the 1960s that has transformed any really notable aspect of outcome in psychiatry. The one really big change, de-institutionalisation, could have occurred without any input from neuroscience at all; in fact, it was in large part a reaction against biomedicalism. It is doubtful that it would be deemed necessary to place yet another prominent polemical article in the Journal in defence of narrowly conceived neuroscientific hegemony within psychiatry, were this not the case.