The idea of a genetic blueprint that makes humans a simple matter of biological cause and social effect makes Richard Lewontin see red.

To be called "the most brilliant scientist I know" by none other than Stephen Jay Gould is no small achievement. Neither is it a small accomplishment to become probably the best known biologist-troublemaker in the United States. For a man who was once "rusticated" from Harvard (but then re-admitted) for failing too many courses, it is little short of miraculous.

Except that the last thing that Richard Lewontin, Alexander Agassiz professor of zoology at Harvard, is likely to believe in is miracles. Richard Lewontin is "a red, an old-fashioned Marxist. I'm a radical, in the sense that, wherever possible, I go back to square one, both in biology and in politics." In his book The doctrine of DNA, he says that he wants to encourage a "reasonable scepticism", and that has tended to be his position in the debates which have marked his academic career.

It was a combination of biological understanding and political principle that led him in the 1970s and 1980s into the struggles for which he is perhaps best known: against both sociobiology and its handmaiden, biological and genetic reductionism. His 1984 book, Not in our genes, co-written with Stephen Rose and Leon Kamin, remains the best refutation of genetic determinist claims that IQ, psychiatric disorders and behaviour generally are written in the genes. "Genetic reductionism, and all those claims about violence and IQ and temperament, DNA as the master molecule, are summed up in Sid Brenner's claim that if you gave him the complete DNA sequence of an organism and a large enough computer, he could compute the organism. Now that's just rubbish, it's not true and it's bad biology," he says.

³ÉÈËVRÊÓƵ

Lewontin sees the reductionist view that we are simply "readouts" of our genetic "blueprints" as the triumph of the pre-formationist view of the 18th century, in which the sperm contained a minute, perfectly formed human, which simply develops in a predetermined way. The basic error of the reductionists, he says, is to talk about genes "for" behaviour. He does not deny that they may find genes that influence behaviour, but "what they have found is a gene with some effect on the total physiology and morphology of the organism, which in turn mediates, in some way, behaviour. The interesting question is the mediation, the story - they should forget about the gene."

He entered into the struggle against genetic determinism, "not just because it's bad biology, there's plenty of bad biology, but because it legitimated certain social actions". In the US, where everyone has the message of the US constitution, "all men are created equal", drilled into them as a child, genetic determinism has played a particularly crucial role as an explanation of why this is not reflected by society.

³ÉÈËVRÊÓƵ

Sociobiology, the idea that human behaviour and social arrangements are biologically determined, plays a slightly different role, he points out. "Sociobiology doesn't speak to the question of differences, but to similarities. It says there is a human nature, which we inherited from our non-human ancestors, and which was enforced on us by natural selection." In particular, it says that human societies are "naturally" hierarchical. Attempting to create social equality, even while acknowledging differences in abilities, is a waste of time. The US capitalist system is therefore the best of all possible worlds, an argument recently made explicit by Kevin Kelly, the editor of Wired magazine in his book Out of Control, the New Biology of Machines. Sociobiology has now become enshrined in school textbooks in the US, which worries Lewontin deeply.

Apart from a disbelief that social phenomena can be reduced to biology, his objection to sociobiology is based on sociobiologists' obsession with explaining everything in terms of a single factor, natural selection. "What that ignores is the massive contribution of evolutionary and population genetics, which has produced a vast amount of theory and experimental evidence to show that natural selection is not the only thing that's operating. There's a lot of random drift, there's developmental noise, there's the history of the population. In fact, nowadays, if you're publishing a paper in evolutionary genetics, your null hypothesis has to be that it's all random, and you have to show that your data can't be accommodated within that. The sociobiologists ignore all that. For them, every pimple on the body has to be explained by natural selection."

He cites as an example of sociobiological "armchair theorising" an Israeli sociobiologist who says that blood is selected not merely to carry oxygen, but to be red (although in fact lobsters have green blood), so that if you are attacked by a predator, the red colour will scare the predator away. "You can tell those stories for ever, but any decent biologist should be revolted by such a thing."

In the late 1960s he resigned from the National Academy of Sciences, over its secret research to help the Vietnam war effort (he remains the only person to have resigned from the NAS, which is more exclusive than Britain's Royal Society). "Although the immediate issue was their war research, I realised that I had been seduced into something I was ideologically opposed to - all those medals and prizes."

More recently, he has turned his energies to the debate over the validity of DNA fingerprinting evidence in court cases. Aside from the practical laboratory problems with the technique, he has been particularly concerned to debunk the excessive claims of certainty made for it, often presented as claims of odds of billions to one against an accidental match between samples. For him this is not only a question of flawed population genetics, but an opposition to what he sees as underlying ideological assumptions about DNA as "master molecule", which leads juries to believe that evidence based on DNA must be certain. He has testified as an expert witness in two court cases, and was asked to testify in the O. J. Simpson trial but refused, because, he says, if he cannot possibly testify on behalf of the many indigent defendants in the US courts, he would certainly not testify for O. J. Simpson.

³ÉÈËVRÊÓƵ

The DNA fingerprinting experience still makes him angry. In the course of the debate, he wrote a paper for Science with a colleague, Daniel Hartl. First, Hartl was called by the FBI and asked to withdraw the paper. Next, at the very final stage before publication, after it had been reviewed and accepted, Daniel Koshland, then editor of Science called Lewontin and asked for changes in the paper. "So I told him, 'if you don't publish the paper the way it is, I'm going to make the biggest stink you ever heard, every newspaper in the US will hear about'. He went out and commissioned a rebuttal by people who had been opposing us in the courts."

After graduating from Harvard, Lewontin did his doctorate in the laboratory of Theodosius Dobzhansky, the population geneticist, at Columbia University. Population genetics describes how the frequency of different variants of a gene changes from generation to generation. At that time the key debate in population genetics was about how genetically different individual members of a species were. (Cheetahs, for example, are extremely genetically similar to each other.) While Dobzhansky maintained there was a lot of variation, Herman Muller, a developmental biologist, said there was very little.

"So I developed a formal experiment to answer the question, but there was no technology to do it." He discovered that Jack Hubby, at the University of Chicago, had a technique of gel electrophoresis, which involves separating proteins (the products of genes) according to their size. "So we tried out the technique, and there was the answer, spread out on a gel." After spells at North Carolina State University and the University of Rochester he moved to Chicago to collaborate with Hubby using the gel electrophoresis technique. His results showed that Dobzhansky was basically right, "and led to 20 years of everyone in the field grinding up organisms and putting them on a gel".

³ÉÈËVRÊÓƵ

For the past 23 years he has been at Harvard. He has made many contributions to population genetics. Latterly his work has focused on very fine details of mutations in the DNA code of genes. Sometimes such mutations do not affect the structure of the protein coded for by the gene. It might be expected that such mutations would not affect the fitness of the organism. In that case, all possible mutations would occur, and none would be favoured by the process of natural selection, in which unfit organisms are weeded out. In fact this does not happen: some mutations are found much more commonly than others. Lewontin has found evidence that one reason for this may be that different mutations affect the speed at which proteins are made, which is crucial to allow them to assume the right shape.

He describes himself as a philosopher and a methodologist, whose career has been devoted not to finding out things but to developing ways to find out things. One of his major themes is that there are no simple causes and effects, either in biology or in life. "I don't think if I had different views about biology I could be a Marxist, because my views about biology have to do with the understanding that things are both causes and effects . . . things interpenetrate each other. Part of that I get from Engels, but part of that makes me want to read Engels, because I come to that from biology."

For Lewontin there are no absolute laws in biology. "There is no law of increasing complexity in evolution. There are no general rules, or principles in biology, partly because living organisms are medium-sized objects, which are not governed by simple rules of physics. They are the nexus of a large number of weak forces. When the organism stops being the interaction of a large number of weak forces and instead becomes dominated by a single force, we say the organism is sick." He is scathing about much of contemporary biology. Many biologists, he says, suffer from "physics envy". "They are looking for simple causes and general phenomena, because that's how you get rich and famous. You don't get rich by saying 'everything depends'."

He says that what first attracted him to population genetics was its formal, mathematical aspects, which seemed far removed from practical, social and political concerns. This is somewhat ironic, given that genetic change in populations is what is meant by evolution, and debates about evolution have always been heavily entangled in politics.

³ÉÈËVRÊÓƵ

I ask him to predict the future. "I think genetic reductionism and biological determinism will continue to be dominant in biology. After all, in historical perspective, the environmental determinism of the two decades after the war was only a brief interlude in a century of biological determinism. The promises of medical research, like gene therapy, will continue, but very few of them will be fulfilled. I was lucky, I lived in the high time of science. I think funding is going to decline, but I can't see beyond the next capitalist economic crash."

David King is editor of GenEthics News.