Why do men generally marry younger women? Why are children more likely to die at the hands of stepfathers than their natural fathers? For an increasingly influential band of biologists, psychologists and social scientists the answers to these and many other questions about our states of mind lie deep in our evolutionary history. Evolutionary psychology has emerged as one of the hottest scientific disciplines.

In the nineteenth century social Darwinists used Darwinism to justify class and racial hierarchies. But whereas in the past many liberals and radicals denounced social Darwinism as a reactionary creed, today they have embraced evolutionary psychology.

American writer Robert Wright, a leading advocate of the new creed, is at pains to emphasise that it is not a justification for old conservative prejudices. Indeed he takes pleasure in arguing that the 1950s, the conservatives' favourite decade of recent times, was based on profoundly 'unnatural' relationships between men and women. Our human nature, he argues, is attuned to the life we lived in the stone age and before. In this period, women played a full role in the activity of society - they were not stuck in the home bringing up kids. Wright also keys into liberal concerns about the dark side of family values, when he highlights the work of Martin Daly and Margo Wilson on child abuse and the murder of children by stepfathers.

Wright endorses the framework of evolutionary psychology as a way of explaining contemporary alienation. He has some sympathy for the American Unabomber, who wrote that: '[I] attribute the social and psychological problems of modern society to the fact that society requires people to live under conditions radically different from those under which the human race evolved.' (Time, 28 August 1995)

The idea is that human behaviour is the product of a long process of natural selection, through which certain behavioural patterns have been selected. But although every behaviour pattern is an evolutionary adaptation, none is adapted to the environment in which we live now. Many of the problems of modern society - the 'diseases of civilisation' as they are called - are said to originate in this mismatch. Specifically, argues Wright, contemporary society is too individualised; we need to rediscover a more communal form of living as practised by our ancestors. And if that means higher taxes and less economic growth, so be it.

In Britain last autumn, the Blairite think-tank Demos published Matters of Life and Death as No10 of their Demos Quarterly series, a report which set out the agenda of evolutionary psychology and considered its policy implications. The introduction to the report, 'An ism for our times', outlined the appeal of the idea. The fact that evolutionary psychology can be taken to back up a notion of community was an obvious attraction to the Demos writers. But in our age of relativism and uncertainty, they also clearly liked its tentativeness; proponents claim that the theory is widely applicable, but unlike the old social Darwinists they eschew rigid determinism and 'easy solutions'.

There is little doubt that the appeal of evolutionary psychology is linked to the spirit of the times. But is it a science? And does it tell us anything about the human condition?

The general failing of evolutionary psychology can be summarised simply: the theory asks too much of evolution, and understands too little about humanity.

Evolutionary psychology expects evolution to generate something it cannot - purposive, goal-driven human behaviour, and all the unique, complex faculties that make this possible, such as language.

As Stephen Jay Gould (something of a hate figure for evolutionary psychologists) points out in his most recent book, Life's Grandeur (1996), human culture has no equivalent in the natural world because there is progress in culture and society, whereas there is no such thing in the animal kingdom.

This capacity for progress in society arises from the fact that humans learn from each other and from previous generations. We pass on knowledge, ideas, technology, capital. Animals do not. Indeed, the nature of Darwinian evolution explicitly rules this out. What marks humanity out from the animal kingdom is not a unique genetic make-up. All genetically framed behaviour is limited, because all that genetic framing can do is sum up the evolutionary history of the species. Anything outside that experience cannot be adequately adapted to. There is no foresight in evolution, and it cannot generate reasoning and planning.

To emphasise the uniqueness of culture and society need not mean neglecting the role of natural evolution in understanding human development. It is the relationship between the two that must be grasped. But within this relationship, culture and society represent the driving force behind the development of human behaviour. It is here that evolutionary psychology understands too little about humanity.

Humanity does have a set of instincts that are genetically influenced. And these instincts are more complex than those of animals precisely because we continued to evolve genetically, up until about 50,000 years ago, while we were also beginning to develop the uniquely human capacities for cultural learning and societal planning. Insofar as we have biological attributes that facilitate sociality it was culture that drove the evolution of these features, not evolution that created culture.

The evolutionary psychologists' other major mistake is to forget the experience of the past 50,000 years. During this period, human society has transformed itself beyond recognition, and expanded the horizons of life in undreamt of ways. Yet the genetic make-up of human beings has remained much the same through that revolutionary journey.

Two of evolutionary psychology's favourite authorities are Leda Cosmides and John Tooby. They argue that we need to look at our modern selves as a kind of bottom-up creation of our evolved selves of 50,000 years ago. Our more complex capacities now, they say, are just worked up versions of the older simple ones: 'In our view, instead of culture manufacturing the psychology of social exchange de novo, content-specific, evolved psychologies constitute the building blocks out of which cultures themselves are manufactured.' (The Adapted Mind, 1992, p207)

In fact the opposite is the case. Insofar as we possess primitive instincts, these are transformed and reworked, in particular during child development, by the top-down influences of complex contemporary societies. We are not born as blank slates, but we are fundamentally changed by the higher, modern, mental states and social capacities made possible by the society in which we now live.

The limitations of the science of evolutionary psychology suggest that its popularity rests upon the way it chimes with the spirit of the times. One important element is a sense of alienation and a search for community. Underlying this search is the view that humanity's problems are caused by our own hubris, our neglect of the natural limits imposed by our environment. The kind of community that is sought is a modest, cautionary one.

At the launch of Matters of Life and Death, Demos director Geoff Mulgan suggested a link between the rise of ecology and of evolutionary psychology: 'Just as ecological understanding has shown us that there are all sorts of external limits to what humans can do, so evolutionary psychology shows parallel internal limits, which we transgress at a high cost.' He spelt out what such costs might be: 'Full freedom of consumption has led to obesity, drug abuse, and so on. This entails costs for both the individual and for society.'

For Mulgan, then, we need to limit human desires, because we are not designed to cope with unrestricted freedom. There is, however, little in evolutionary psychology to suggest that there are limits to human behaviour - largely because evolutionary psychology tells us very little at all about human behaviour. But for those, like Mulgan, who are predisposed to believe that there are constraints on human activity, evolutionary psychology seems to provide a useful way of grounding their arguments in science.

The claims of evolutionary psychology reflect today's prevailing mood of low expectations. Increasing disillusionment with social explanations for human behaviour, a sense that traditional social theory has proved an inadequate guide, and a pessimistic attitude towards the human potential have all opened up a new space for Darwinian arguments. In his book The Moral Animal, Robert Wright argues that the new discipline is the natural philosophy for the 1990s and that 'one reasonable reaction to evolutionary psychology is a self-consciousness so acute, and a cynicism so deep, that ironic detachment from the whole human enterprise may provide the only relief' (1994, p326). Unfortunately this is one of the few things about evolutionary psychology that is probably true.