How much can parents shape their child's personality? According to most psychologists, your personality is largely determined by the way your parents treat you when you are a child. If you turn into a happy, well-adjusted adult, it's because your parents did their job well. If you have psychological problems, it's all your parents' fault.

This idea is what Judith Rich Harris calls 'the nurture assumption'. In her book of the same name, she argues that it is fundamentally mistaken. If Harris is right, her book will shake the very foundations of developmental psychology. Since Freud, almost every psychologist who has written about childrearing and personality has subscribed to the view that Harris attacks.

Thousands of studies by developmental psychologists purport to show the power that parents have to shape their children's personalities. Parents who care for their children in loving, supportive ways tend to have babies who are securely attached and who develop into self-confident adults. Parents who are too harsh have children who are aggressive or anxious, or both. And so on and so on.

Harris is aware of the difficulty of challenging such a consensus, but she sets about her monumental task with gusto and a lightness of touch that make her book a pleasure to read. As a writer of textbooks on child development, she has a command of the material that is probably superior to that of most university professors. She probes the methodology of the studies of personality development - and she finds it to be flawed.

Harris' first criticism is that none of these studies controls for genetic factors, and so fail to tease apart the effects of nature and nurture. There is, however, a parallel set of studies by behavioural geneticists, and these have shown that around fifty per cent of the variation in most personality traits can be accounted for by genetic differences. Putting these two sets of data together, Harris shows that the correlations that the personality development literature finds between parenting style and child personality are small enough to be explained almost entirely in terms of the shared genes.

But Harris is no genetic determinist. Far from it; she acknowledges that genes account for only half of the variation in personality traits, and accepts that we must look to environmental differences to explain the rest. What Harris denies is that the bit of the environment that most affects personality is provided by the parents. It is, she claims, a mistake to equate the 'genetic/environmental' distinction with 'nature/nurture', because the 'environment' that determines one's personality is not the home environment.

If the home environment has almost no effect on personality, what part of the environment is important? Harris argues that the most influential environment is provided by our peer group. It is children that socialise children, not parents. This is most clearly evidenced in immigrant families, whose children quickly pick up the language and accent of their peers. Along with language come all the other aspects of culture, from dress to attitudes. If there is ever a conflict between how your parents want you to behave, and the way your peers think you should behave, the peers win, at least during childhood (which is when it matters for personality development).

Harris is keen that her readers should not leap to the wrong conclusions. She denies that her 'group socialisation theory' means that it doesn't matter how you treat your children. Such an inference would be valid, of course, if the only reason for treating your children well was to ensure that they became well-adjusted adults. But this is not the only reason why one should be nice to one's kids. People should treat their children well, first and foremost, because their children are human beings, and, like all human beings, they deserve to be treated with respect. If you need a more prosaic reason to be good to your children, there is always the fact that treating them well will make it more likely that they will be nice to you when they are grown up and you are old. The one thing it won't do, however, is help them to be well-adjusted adults. Personality, Harris stresses, is highly context-dependent. Good parenting may help children to be happy at home, but the influence of parents stops at the front door. This is clear enough to anyone who has been to a school open-evening. There are always some parents who, after chatting with the teacher about their child, are amazed by how little the teacher seems to know their child. It never crosses their minds that little Johnny might behave completely differently in the classroom.

One of the virtues of Harris' theory is that it is consistent with evolutionary theory. This is something notably lacking from the theories that she criticises. From an evolutionary point of view, the nurture assumption immediately stands out as highly implausible. Nature would not have designed children to be putty in their parents' hands. Indeed, the Darwinian theory of parent-offspring conflict, first put forward by Robert Trivers in 1974, makes to the opposite prediction; we should expect children to be highly selective in the kinds of information that they accept from their parents. It is equally unlikely that the style of a baby's attachment to its mother should set the pattern for its future adult relationships; such an epigenetic rule would convey no selective advantage, and would quite probably be selected against.

As Steven Pinker points out in his Foreword, 'the attachment hypothesis owes its popularity to a tired notion bequeathed to us by Freud and the behaviourists: the baby's mind as a small blank slate that will retain forever the first few inscriptions written upon it'. In view of the immense amount of evidence that has accumulated against this view in the past few decades, it is amazing how many psychologists still believe in it. The persistence of this dogma provides yet more evidence (if any were needed) of the depressing human tendency to cling to cherished ideas even when they have been proved wrong. In view of this tendency, it is perhaps being too optimistic to state that Harris' book will succeed in persuading developmental psychologists to abandon the nurture assumption. Too many careers have been built on this foundation for it to be easily given up. Nevertheless, those who are more intellectually honest, or who do not have so much invested in the errors of the past, will recognise this book for what it is; a watershed in the history of psychology.