Sir

While I am in sympathy with the thrust of Allan Snyder's thoughts on ‘autistic genius’ (Nature 428, 470–471; 2004), and the book by Michael Fitzgerald, Autism and Creativity, which he reviews, I think there is a danger of going overboard on the subject of pathology and creativity. Snyder cites me as saying that the philosopher Ludwig Wittgenstein had autistic traits, but I have never thought or suggested this. Snyder's misapprehension arose from a conversation that I had with the autistic scientist Temple Grandin — reported in my book An Anthropologist on Mars (Knopf, New York, 1995) — about another researcher's suggestion that Wittgenstein may have been autistic. Grandin erroneously attributed this opinion to me in her otherwise excellent book, Thinking in Pictures (Doubleday, New York, 1995).

I think that pathologizing genius, and pathologizing historical figures, has become an obsession with us; and also that the concepts of ‘autism’ (and ‘Tourette's syndrome’, etc.) have become so distended that they are vastly overused. It seems to me extremely unlikely, from the evidence we have, that Wittgenstein or Einstein or Newton were significantly autistic.

On the other hand, there is strong historical evidence to suggest that the eighteenth-century chemist Henry Cavendish was autistic (see O. W. Sacks, Neurology 57, 1347; 2001). Unlike most other supposed ‘autistic geniuses’, he showed a near-total incomprehension of common human behaviours, social relationships, states of mind, and money, as well as an almost obsessed attention to detail — which led to the great generalizations he was later to erect. But it is important to sift through the evidence with great care before diagnosing or pathologizing.