Having been brought up in this environment, it is perhaps not surprising that much of what Leda Cosmides has to say about her own work in evolutionary psychology, a controversial field that she has helped to pioneer, resonates powerfully with the boundary-less science envisioned by her father.

As the 1970s ended, Cosmides and a small group of like-minded academics helped to lay the foundations of evolutionary psychology, a rapidly growing field that some commentators describe as a 'scientific revolution'. Briefly put, evolutionary psychologists hold that the human mind was sculpted by evolution. They believe our behaviour now is largely determined by mental mechanisms that served us equally well (perhaps even better) when humans were hunter-gatherers struggling to survive and reproduce in a very different environment.

Evolutionary psychologists believe that there is a universal human nature; that human characteristics are coded for by the genes all humans share and that the reason we have these particular characteristics rather than others is that natural selection - the environment favouring the reproduction of some genes over others - has systematically led to them.

Cosmides, 38, is associate professor of psychology and co-director of the Centre for Evolutionary Psychology at the University of California, Santa Barbara. She recalls the excitement at Harvard and elsewhere during the field's formative years: 'It was a strange kind of euphoria that is hard to duplicate. Because certain theories have such great explanatory power. . . and when you see the world resolving itself into a understandable framework ... well when that happens you know it.'

Cosmides never found abnormal behaviour terribly interesting, even at high school. She thought there were an infinite number of bizarre ways in which people could behave at anyone moment and yet did not. 'It seemed strange, ' she says, laughing, 'that almost everybody behaved so predictably that they could be swinging from the chandeliers but weren't.'

Human nature at its most fundamental level, she thought, must be relevant to providing answers for this orderly behaviour. At the age of 14 Cosmides became obsessed with Walden Two, B. F. Skinner's notion of a utopian community run by the judicious use of what he called 'behavioural technology' a phrase that still makes 'a chill of horror' run down her spine. Cosmides thought Skinner's work fascinating but completely wrong. The implication of his work - that the learning mechanisms of different species are essentially the same and that all human relationships are arbitrary social products seemed implausible to Cosmides. Skinner's work implied, for example, that a man could be made to feel delighted to find his wife in bed with another man; that the bond between mother and child could be easily erased.

The conviction that there was something terribly wrong with such attempts to explain human nature was one she took with her to Harvard in 1975, where she led a 'double life' between the biology and psychology departments, graduating in biology and staying on for a PhD in cognitive psychology.

While at Harvard she met her husband, John Tooby. The two have worked together ever since. At that time researchers were trying to connect evolutionary theory directly to observable human behaviour, paying scant attention to the psychological mechanisms that might underpin action. That was a bad idea, Cosmides says. She argues that the key reason we have an information-processing system, a brain, is to ensure our behaviour is contingent upon information from the environment. This relationship is facilitated by very subtle neurological programmes designed to create varieties of behaviour in response to different kinds of information. And since different people are exposed to information from different environmental stimuli, one should not expect manifest, observable behaviour to be uniform. In fact, such behaviour should be variable in a complex manner among people within a culture and between cultures.

In the hunt for a universal description of human nature, Cosmides has been influenced by the conceptual importance attached in other areas of science, particularly engineering, to identifying the level of 'invariance', that is, the level at which uniformity is exhibited by systems. The difficulties associated with scientifically studying observable behaviour variables have persuaded Cosmides that it is a big mistake to look for invariance at this level. 'That is why we have a brain, so that behaviour will not be invariant, so that it will be complexly associated with information from the environment. The level of invariance that we are interested in should therefore be in the development of neurological programmes which govern behaviour and knowledge acquisition. It is at the level of psychological mechanisms that you would expect invariance.'

When Cosmides and Tooby began publishing work that focused on the brain as an information-processing system, there was considerable opposition from the social science community. 'There was this long tradition of seeing the mind as a blank slate - as a general-purpose computer,' says Cosmides. 'But what people do not realise is that general purpose computers are not powerful - there is very little they can do.' To explain the argument evolutionary psychologists have chosen artefacts as metaphors. Cosmides likes to employ the Swiss army knife to illustrate the power of specialised designs that are bundled together in one neat package - rather like the specialised programmes developed by the brain.

One of the aims of evolutionary psychology, which has been called reductionist by its critics, is making psychology a part of the natural sciences. When there are disagreements in the natural sciences, scientists work across disciplines to focus on errors or on the possible need to revise standard models. Cosmides cites the efforts by physicists and chemists to duplicate results and get to the bottom of the scientific claims made for cold fusion. 'And that is because in the natural sciences you cannot get away with just saying 'oh, well, I don't care about it anyway'. But in the social sciences people do this all the time. Anthropologists, for example, will say they don't care about psychology, that it has nothing to do with them. A big part of evolutionary psychology is about the causal connection between disciplines - trying to make the discipline completely a part of the natural sciences. And that is not a reductionist but a naturalising enterprise.'

Other criticisms of evolutionary psychology are usually political or relate to preconceived ideas of how the brain works. There is a long-standing distaste among many academics for the notion that there is a universal human nature that consists of something more than a blank slate and a capacity to learn. Some quarters of the science community are worried by what they see as yet another attempt to define human nature in terms that would be exploited for ideological machinations: the horror of eugenics and Nazism loom large in their fears. Cosmides says that people who have such concerns cannot have read any evolutionary psychology because there is nothing in it that could encourage latterday Nazis. In any case, past notions of human nature have not exactly been helpful. 'Mao's China, Pol Pot's Cambodia, Stalin's Russia - these holocausts were based on the environmentalist assumption that human nature can be easily moulded any which way. In the past 80 years, more than 50 million people have been executed in the name of philosophies that rest on this view of human nature. Yet there are few senior scientists who are reluctant to embrace the notion that the mind is a general-purpose machine on the grounds that it has led to genocide in the past. People think they are being more ethical by being hyper-sceptical about modular, domain-specific views of the mind but they are not.'

Cosmides thinks such 'misconceptions' are born of a fear that the picture being drawn by evolutionary psychology places an enormous limit on human potential and that this limit would not exist if the mind were equipped with neurologica1 mechanisms designed to carry out general rather than specific tasks, a view Cosmides believes to be deeply flawed: 'If the mind was general purpose you would not be learning language at all. A mind with general-purpose mechanisms which are independent of content would not include any clues about what the structure of the world really is and would take an infinite amount of time to learn anything. Computationally such mechanisms would be very weak.'

It is also not the case that evolutionary psychology takes the 'mystery' out of life, she says. Seeing, falling in love, finding someone beautiful, enjoying the taste of chocolate are all things we do quickly, automatically, effortlessly and unconsciously. 'These things seem so simple and inevitable that we assume, wrongly, that the machinery that causes them is simple. We are blind to the elegance and complexity of our own mental machinery, our cognitive instincts.' Evolutionary biology provides 'lenses' that correct for this instinct blindness - 'it lets you see arabesques wbere your common sense had seen nothing but a blank slate.'

Recent work by Cosmides and Tooby has focused on co-operation and aggressive conflict. They believe people have complex mechanisms for engaging in individual co-operation that, in some contexts, operates against other groups. Soccer hooliganism is the kind of phenomenon that comes under this heading. 'In tribal societies you have conflicts between bands and so on,' Cosmides says 'We think there is an interesting set of programmes which become activated in these kinds of contexts and make people think differently. Certain kinds of situations activate these us-them conflicts.' Cosmides and Tooby believe that while these programmes are not usually active, certain cues in the environment will trigger them.

Cosmides says that if the kind of cues needed to activate this kind of psychology were known, then in principle society could be arranged to minimise the risk of activation. 'This could allow you to minimise ethnic conflicts and so on. If you don't study these things then you've got no chance of ever effectively dealing with these situations.'

In one research project Cosmides tested her theories about co-operation by using established tests of reasoning powers. Those taking part in the tests were far more successful at detecting cheats than detecting violations of logically identical rules having nothing do with co-operation (a 75 per cent success rate compared with 25 per cent). Results like these, says Cosmides, support her argument about the evolution of a neurological mechanism developed so  that humans could, for example, spot  cheating in a social system dependent on everyone co-operating and playing fair. So is it the case, then, that there is now no stopping the rise of evolutionary psychology? Cosmides laughs and says she is 'extremely optimistic' but could not, hand on heart, say there was absolutely no way the work will fall prey to its critics. 'Certainly I'd like to think that by the time I retire there will be no need for the words evolutionary psychology. It will just be known as psychology - a discipline continuous with the natural sciences.'

Reproduced by kind permission of the THES. The author has given his permission for use of his work free of charge.