Psychoanalysis: great art, bad science

Darian Leader, Last of the Psychoanalysts, is back banging his drum for the ideas of Sigmund Freud and against CBT, with his new book about melancholy, which is reviewed by Hanif Kureshi here.

Leader is a rabid critic of CBT, which he has compared to the brain-washing techniques of the Cultural Revolution. If only we all had access to psychoanalysts like him, he mourns, then we could accept that the roots of our mental illnesses are the sex and violence in our unconcious, and return to whatever Freudians call health.

Kureshi loves psychoanalysis, apparently. So do loads of other artists - like David Crane, creator of the Sopranos, who made a psychoanalyst one of the heroes of his TV series.

Artists like psychoanalysis because its method is very like their method - it relies on flashes of insights into the deep darkness of our motivation. But, like art, it refuses to test its findings. They must be self-evidently true about the human condition, like a novel or a film.

CBT, by contrast, has the guts to test its findings in clinical trials, something psychoanalysis has never done. Instead, psychoanalysis, like the novel, relies on anecdotal evidence and the haphazard observation of the author / analyst.

This is a great way to write a book, but a catastrophic way to formulate a medical theory.

And psychoanalysts are, in fact, often great writers - Freud is a far better writer than just about everyone working in CBT. But that doesn't mean his wild claims are right, anymore than the equally wild and untested claims of, say, DH Lawrence or Norman Mailer are right.

You have to test your findings, otherwise your insights are 'just' art, and not science.