What happens when boy-girl meets girl-boy? If they're slugs living north ofManchester, not a lot: they just mate with themselves. Steve Jones explains the origins of sex

Look at the fashion pages, the news reports, even the book reviews. It's obvious: sex is about love, fantasy, jealousy and passion.

All of them are, in fact, beside the point. Sex means just one thing - having two parents rather than one. And when it comes to the meaning of women, that's even simpler. Women are the ones with large sex cells, men those with small. That definition is the only one that can cover Adam and Eve, apples and serpents, blue whales and blue-green algae.

There are, of course, a lot of tedious details. They fascinate those whose understanding of science is limited to metaphor: genetic homosexuality in fruit flies; rape in dolphins as two males gang up on a single female; and an infinity of other examples so often used to explain human behaviour in animal terms.

³ÉÈËVRÊÓƵ

All this is infantile, but some sexual minutiae are even interesting. Being a man, it transpires, is simply a matter of one tiny protein that binds on to a single site in the DNA and bends it. By so doing, it causes the early embryo to grow more quickly and shifts it irrevocably into the path of maleness.

What is most fascinating about sex is not its mechanics but its origin. Why bother with this first and least original of all sins? That is the oldest question in biology. Why do females tolerate males?

³ÉÈËVRÊÓƵ

Part of the answer lies in the fact that sex, not death, is the great leveller, because it mixes the biological capital of two people and disperses it among their children. Sex may also hold the key to the biggest mystery of all: why are babies born young? How can two old and decayed pieces of protoplasm produce, in a single more or less protracted spasm, a brand new model of what went before?

All men do is force women to copy their genes, without taking the trouble to have children themselves. Getting rid of males seems an obvious device, and many creatures, from lizards to aphids, have tried it. Just one parent, passing on all her genes to all her children - it sounds unbeatable. It is, though, always beaten in the long run. With the exception of one group of tiny freshwater creatures, the bdelloid rotifers, no purely asexual lineage has lasted long.

Men, however, do one useful job; they link pedigrees together. They tie family lines into a web of kinship and shared descent. Sex is what defines what a species actually is. Two creatures belong to the same species if they can mate and produce offspring. Genetic engineering is alarming because it breaches that taboo. It breaks the sex barrier by allowing DNA from one species to be moved promiscuously to any other with some simple (and notably unerotic) chemical tricks.

Sex has another implication, first pointed out by Francis Galton, Darwin's cousin. He noticed, in alpine villages in Italian Switzerland, an odd pattern in one asexually inherited character, the surname (which passes only down the male line). In one village everyone had one surname; in the next, a different one. Why, asked Galton, was this? Was it better, for some unknown biological reason, to have one particular name in each individual hamlet?

He saw that it was not. As soon as a man has no sons (no children at all, or only daughters) his name disappears. In time, inevitably and for reasons of simple arithmetic, a single surname takes over. Which one wins is a matter of chance.

The aristocracy shows how. The inheritance of most titles avoids sex as they pass only through males. That process is exclusive but, sooner or later, means extinction. Should an aristocrat fail to have sons, his title fades into history. The Knight of Glin, Desmond John Villiers FitzGerald of Glin Castle, County Limerick, is the 29th successor to that noble name. He is also, alas, the last knight, as all three of his children are daughters. His name will die with him.

Without constant refreshment, the number of noble names is bound to decrease each generation. The average persistence of a hereditary title in the middle ages was three generations, and all the 5,000 feudal knighthoods recorded in the Domesday Book have gone. Inheritance down a single line of descent opens all titles - and all genes - to the threat of instant death.

Two sets of human genes expose its dangers. The Y chromosome passes - like a title - from fathers to sons. Other genes are held on a special structure, the mitochondrion, that is transferred through the egg, through females. It is easy to follow their fate. How many men alive at, say, the time of William the Conqueror have a direct line through sons to a living man; and how many women a matching link through daughters to a woman alive today?

³ÉÈËVRÊÓƵ

There are around 300 British families who can trace direct descent from William - but only one, the Bassetts, can make the link through males alone. All the others need sex, men to women and back again, to follow a path back to the Conqueror. A third of the British population could do exactly the same thing, if only they could find the records. Sex is a remarkably democratic process.

In the past few weeks there has been a connection of names and genes. The Kohanim - the Cohens - are a Jewish lineage who are particularly proud of their history. They descend, they say, from the Priests: the religious aristocracy of early Judaism who disappeared with the destruction of Solomon's Temple more than two millennia ago. Only the priestly name persisted; as did the pattern of inheritance of priesthood, from father to son.

Now it seems that today's Cohens are indeed genetically distinct from other Jews - but only on the Y chromosome; inherited, like their name, asexually. One set of variants on the Y is more or less confined to that family, although they share their other genes with their co-religionists. A cultural tradition has been confirmed by biology.

³ÉÈËVRÊÓƵ

Without sex there will be genetic monopoly; one set of genes will - just like a surname - take over. Unfortunately, monopolies (even privatised ones) are inefficient. As they do not have to bother with competitors there is less need for innovation - for novel combinations of ideas. That is fine as long as there are no new challenges, but as soon as there are there is danger of ignominious collapse.

British slugs, like Swiss villagers, show what is going on. In the languid south, slugs - hermaphrodites though they are - indulge in sex. Boy-girl meets girl-boy, and matters proceed more or less as normal. For a hermaphrodite, though, there is always one temptation; what Woody Allen called sex with someone you really love. Any creature containing both male and female sexual organs can choose to mate with itself.

There are patterns in celibacy. For British slugs, sex stops at Manchester. North of there, they begin to fall into reproductive narcissism, mating with the partner under the skin. After many generations of so doing, all begin to share the same genes; they are, in effect, identical twins. Admittedly, for slugs things are changing. In the past decade, a wave of sex has spread from the south. Perhaps it is due to global warming, perhaps to imports that bring in more of their degenerate European cousins. The general picture, though, is clear. Slugs, like most creatures, find it easier to relinquish males in high mountains or cold climates than in the tropics.

This is because asexuality works best when the opponent - bad weather, or poor soil - is predictable. A single strategy, unimaginative though it is, may succeed. In the same way, huge fortunes accumulate mainly in places where there is political stability.

Giving up sex produces one set of genes that takes over. It may do well but it is in constant danger of obsolescence. Sex starts a system of competition in which new mixtures are steadily produced. These are the raw material upon which evolution acts to adapt the population to changing conditions. Most fail, but some are better than what went before. Adam Smith would have been delighted. Once sex gets started it is self-sustaining.

Enthusiasts for the free market see it as the most efficient economic system, just because it can never afford to stay still. The inbuilt contradiction shared by both sex and the market is the simple expense of keeping them going, of producing constant innovation against an irritating challenger. Lots of people suffer, genetically or financially, in each; but in the end both seem unbeatable.

So that is what sex is about - competition versus monopoly; the need to come up with new ideas or to sink into genetic oblivion.

It seems, I have to say, a deeply depressing view of life: sex as Thatcherism is a vision notably lacking in romance.

There are, though, new and exciting ideas about sexual reproduction that would stir the heart even of New Labour. All this competitive business is about what happens once the pastime gets going: it is an endless whirligig no creature can afford to jump off. But what started that intrinsically improbable machine in the first place? How did evolution come up with the curious idea of joining two creatures to make one?

The answer links the twin obsessions of sex and death. DNA is a complicated and unstable chemical. Each time it divides, it is damaged. My own genes are, I know, full of mistakes that were not there on their arrival on this earth, half a century ago. My body does its best: it has enzymes that repair them. Slowly, though, these are defeated and my cells (dividing asexually as they do) fall into decay. In time, they will drag me to the grave.

Those repair enzymes are now at the cutting edge of biological research, not least because their failure can give rise to cancer. New work shows that they go into a great frenzy of activity in one particular cell line - that giving rise to sperm and egg. At great metabolic expense, the errors are corrected, in time for the new generation to be born young and genetically renewed.

And that is how sex may have started. To proofread a manuscript an editor needs a perfect reference copy. Perhaps, in the very early days of life, the genetic material was a single short chain of nucleic acid bases. There were, no doubt, many mistakes: probably more than today because the earth was flooded with ultraviolet light, a potent agent of genetic damage. There were, no doubt, simple mechanisms for repair; but how to know what the correct sequence was? There was only one way, to float over to a friend and to line up your genes against hers in the hope of finding what had gone wrong. Sex may have started as a method of mutual cooperation in the face of physical decline; and, expensive though it is and unwelcome though its implications may be, that is what has kept it going.

It is a gratifyingly socialist view of evolution. That, then, is what sex really means; death and decay for those who indulge in it, but eternal life for their genes.

³ÉÈËVRÊÓƵ

Steve Jones is professor of genetics at University College, London. This article is based both on his lecture to the Royal Society this week and on his book, In The Blood, to be released in paperback by HarperCollins in March.