É«ºÐÖ±²¥

Hans Eysenck seems better at creating an ad hominem argument than at recognising one (É«ºÐÖ±²¥S, March 3).

There was nothing ad hominem in my description of the reductionist paradigm within which research seeking to assign genetic causes to violent crime in the United States is framed. I neither know nor care how those who conduct such research vote, and if Professor Eysenck, who says he does, tells me they are kind to their dogs I should not be in the least surprised.

In fact a considerable proportion of Professor Eysenck's letter is taken up in repeating points that I had myself made, that humans are biosocial animals, and that full explanations of human behaviour demand that we understand both domains. His little lecture on biochemistry does the psychometrician credit, but it merely summarises my points made in my own earlier Nature article.

Everyone pays lip-service to this principle, just as drivers do to the Highway Code, however bad their driving. But good science demands finding the appropriate level of analysis for effective explanation of the phenomena we are studying. The complex processes embraced within the term violence and their social and historical distribution simply do not permit naive genetic interpretation.

Professor Eysenck apparently believes that the effective explanation of the prevalence of violent deaths among poor young men in the United States lies in a lack of vitamin supplements rather than in the social context of a society riven by poverty, racism and unemployment and a powerful gun culture. Rather than reject this hypothesis outright, I would suggest that he try the experiment of handing out vitamins (double-blind of course) among the drive-by gangs of Los Angeles. He and I could make alternative predictions as to the consequences and we can settle the debate in true Popperian manner.

Steven Rose

Director

Brain and Behaviour Research Group

The Open University