Can testosterone really be held responsible for male aggression? Jerome Burne reports on a new book from the States which explodes the myths about this misunderstood hormone
"When a woman becomes a scholar," sneered the German philosopher Nietzsche, "there is usually something wrong with her sexual organs." Until recently, the idea that women were enslaved by their sexual hormones was common misogynistic currency. Over-emotional and prey to mood swings, they couldn't do all sorts of masculine jobs like run big companies, let alone the country, because of those hormones.

Now, of course, the roles are reversed. Women run companies, the country and jeer at males for being the puppets of their own imperious sexual organs, especially the testosterone they pump out. When there are reports of violence by gangs of young males, whether in Northern Ireland, Bosnia or just on the football terraces, commentators reach for phrases like "testosterone-fuelled rampage".

Here, for example, is a leading woman's magazine writing recently on hormones and sex: "The key to hard, fast, aggressive sex is testosterone. It makes you want to pursue sex, initiate, dominate. It is also your "warmone", triggering aggression, competitiveness, even violence." But does testosterone equal violence?

At first sight, it's an absurd question. We've always known that the way to turn a snorting, aggressive bull into a docile beast is to chop off his testosterone-producing cojones. Right-wing pundits regularly make common cause with militant feminists and call for castration of sex offenders to tame his sex-drive, too. But Robert Sapolsky, professor of biology and neurosciences at Stamford University in California, thinks we have got it all wrong. In a new book just published in the US, The Trouble With Testosterone, he shows how we've got it wrong about this misunderstood hormone.

He doesn't deny that testosterone sculpts all those distinctive male features - deeper voice, facial hair, bulkier muscles - but the equation isn't simple. "At first sight, the links are impressive," he says. "Men are generally more aggressive than women and they have more testosterone in circulation. What's more, they are at their most aggressive at precisely the time when they are brimming with the stuff - in adolescence." Women are thought to produce between a tenth and a fifth of the testosterone that men have.

But there's an elementary error of reasoning here. Just because we find a) high levels of testosterone with b) high levels of aggression, it doesn't mean that a) "causes" b), that b) could cause a), or that they could come together but be unrelated. "In fact," says Sapolsky, "there are many studies that show that if you put males together and measure their testosterone levels, the results don't tell you who is going to be the most aggressive. Also, if you start measuring a young guy's testosterone over a period, there is no link between maximum macho posturing and peak levels."

Let's look at the castration scenario again. It's true that if the source of testosterone is removed, aggression levels plummet, and if you then boost testosterone levels again with synthetic testosterone injections, aggression returns. But what if you only put back 20 per cent of the original amount? "Amazingly aggression rises to same level it was before the op," says Sapolsky. "What's more, aggression doesn't go any higher, even if you boost testosterone to twice its old level."

So, the more complex picture that is emerging is that we need some testosterone for aggression but the connection is vague and prey to other influences. It's not like many drugs - "the more you have, the bigger the effect". In fact, the way testosterone works is a valuable antidote to the fear that once we crack the genetic code we'll all be predictable automatons.

We know quite a lot about the links between genes and testosterone. Genes control the production of testosterone, how much and how fast. They also regulate the rate the liver recycles it from the blood and how many testosterone receptors - the cells that respond to the stuff - are built in the brain. "This should be an archetypal case for studying how genes can control behaviour," says Sapolsky. " Instead, it's an archetypal case for learning how little genes actually do."

Just how tenuous this testosterone-aggression link is can be seen in experiments done a number of times with monkeys. Put a group of strange males together and in the first few days they sort out their dominance hierarchy - who beats up who and who defers to who. Now, take monkey number three in the hierarchy and give him a huge shot of testosterone. What happens? He immediately becomes far more aggressive. So, testosterone causes aggression? No, a closer analysis reveals he's still showing loads of respect to numbers one and two but the rest are getting it in the neck more regularly. The point is that testosterone exaggerates the aggression that is already present.

You can even see this process at work in the brain. Two brain areas are crucial to aggression - the amygdala, which lights up when we get emotionally aroused, and the hypothalamus, which controls appetites such as sex and hunger. A rise in messages from the amygdala to the hypothalamus are a sure sign that we are getting pissed off. Now, suppose you inject testosterone into the area - there are testosterone receptors in the brain - is there a sudden flurry of messages? No, except when the wires, or axons, are already humming. Then there is a big surge. Testosterone exaggerates what's already there.

The more we uncover about the links between brains and hormones and behaviour, the more complicated they become. Old style distinctions between genes or environment no longer make any sense. Some castrated individuals do lose their aggression but some don't, and the ones who keep it are those who had the most experiences of being aggressive before the snip. Chop- their-balls-off lobby take note: biology does not tell us which kids will make trouble in inner-city hell-holes and testosterone doesn't explain gang warfare.

Humans are exquisitely sensitive to group influences and that is where aggressive or civilised attitudes are shaped. It's time testosterone was consigned to scrap heap of explanations, along with the wandering womb.

The Trouble With Testosterone is published by Scribner, an imprint of Simon & Schuster.