The Nurture Assumption by Judith Rich Harris

Bloomsbury, £9.99, pp462

Judith Rich Harris has capsized traditional psychological thinking. Her determined hand does more than rock the cradle, it overturns our received idea of childhood. She argues that children are not significantly influenced by their parents. They have inherited their genes but are unlikely to have profited or suffered greatly from their example. Nurture can go hang - it is peers, not parents that have the power to change children.

This big, enthusiastic book excited extreme reactions in the US. Harris is adroit at wrong-footing other people's research (she brazenly invents research topics such as 'the effects of eating broccoli' to demonstrate how research findings can be manipulated). She is like a clever, unconventional child at the back of the class who keeps asking the strange but pertinent questions that are sure to embarrass teacher.

But there is a story within The Nurture Assumption that is alluded to but never fully told: about Harris's own family. She has two daughters, one of whom was adopted (she does not, tantalisingly, say which). The eldest was a model child and left her free to be an easy-going parent; the younger was a liability. She and her husband worried themselves sick about how best to manage her; they tried everything - and then 'gave up'. The youngest is now, in her mother's discouraging words, 'quite nice'.

Harris has been too ill to work outside the home; she sees it as an advantage that she has done all her thinking independently, not yoked to any single academic body. She has a healthy - though also slightly defensive - suspicion of academics and their preconceived ideas.

Until recently, she wrote textbooks on child development, a painful and ironic occupation. How much, having suffered anguish over a daughter who behaved badly, does she have invested in her own arguments? One can only guess at the extent to which her family history may have been a distorting drive.

Is she right? I have no research but my own life to pitch against her. Yet, instinctively, I feel that there is something compromising about accepting the balm she offers parents. Guilt may be a destructive accessory but Harris's arguments, taken to their logical conclusion, undermine parental responsibility and fill me with consternation. I flinch a little, too, from her easy moral definitions. She refers flatly to 'bad' children, which seems to me an extraordinary over-simplification.

There is nothing novel about the idea that children are influenced by their peers - they have always been so; it only becomes radical when hitched to the notion of parental inconsequence. We are all influenced by our peers, our genes and the world as it comes our way. And I cannot shed the idea that most of us have been influenced by our parents, too.