Warfare poses a problem for evolutionary biologists, especially those who want to explain human behaviour. Darwin's theory of evolution reveals life as an unending struggle, in which nature unhesitatingly rewards traits, tactics and behaviours that promote the survival and reproductive success of their bearers, often at the expense of others.

Set against this is the human male, whom history records as all too easily persuaded to risk his life for the good of a larger group, usually his countrymen, to whom he is at best only distantly related. Such things are not supposed to happen in a world of undiluted self-interest. This apparent contradiction forms one of the social scientist's favourite techniques for baiting evolutionists.

Some answers do suggest themselves. Promises of salvation or cash may work on desperadoes and the gullible, and sufficiently harsh punishment can persuade even the most reluctant of warriors. Another answer more directly implicates the theory: perhaps by fighting, a man increases the chances his non-combatant relatives may survive, prosper and have more children - the Darwinian currency rears its head. When humans roved the savannah in small, closely related bands this may have been the best strategy to promote one's genes, even if in contemporary societies such proclivities often go unrewarded.

Cooperation in a larger enterprise is a theme that swirls around the most fundamental aspects of biological and cultural evolution. Starting with the modest origin of life 3,500 million years ago, John Maynard Smith and Eors Szathmary advance the notion that the spectacular increase in complexity of the biotic world since then has been achieved by a series of significant evolutionary transitions.

ÁñÁ«ÊÓƵ

Here is what vexes the committed Darwinist: the defining feature of each of these transitions is that some entity voluntarily surrenders part of its own freedom to that of a larger group. In the language of genetics, an entity capable of replicating or reproducing on its own before the transition can only replicate as part of the larger more complex whole afterwards. But as the authors ask, why does competition among entities at the lower level not disrupt Nature's attempts to build complexity at a higher level? Whither self-interest?

The origin of life itself provides a view of the problem. Sometime in the past, independent self-replicating entities that we might now call genes joined up in long strings to form chromosomes. Chromosomes contain the molecules of deoxyribonucleic acid (DNA) arranged as sequences of genes that provide instructions for an organism's assembly. Rather than replicating on their own, the genes in the chromosome must now all replicate in concert. Why these solitary genes should ever join up with others - thereby handing over to the chromosome the "decision" about when it is best for all of them to replicate - is not obvious. Something must prevent parasitic genes from tearing down the edifice.

ÁñÁ«ÊÓƵ

The transition to chromosomes is not an isolated case. Our cells each contain a nucleus and separate entities known as organelles. Mitochondria are organelles that used to be free-living and self-replicating. At some stage they relinquished their independence to live in cells and perform useful tasks. At a later stage, individual cells capable of replicating on their own formed into colonies. Eventually, multicellular animals such as ourselves arose, in which only a few of the cells - the sperm and eggs - have the privilege of being the inheritors of the next generation: all of the remaining cells that make up our bodies are slaves to these few gametes, and eventually die with the organism.

Explaining slavery is not easy. Why, for example, do we not shed gametes from all over our bodies? Trees do. Beyond mere bodies, sometimes collections of bodies work together as societies.

So technical are many of the problems that the authors could have limited themselves to asking questions about any one of these major transitions. But their insight is that a small set of principles may unify transitions ranging from the origin of life right through to complex primate societies. This is an audacious claim even for Maynard Smith, a leading evolutionary theorist, and Szathmary, a theoretical chemist with several important contributions to questions about the origins of biological complexity.

The authors try to explain in terms of the immediate benefits conferred on the cooperators, the losses of sovereignty that characterise each of the major transitions. The details are specific to each particular transition, but in every case a higher level of complexity emerges from the harnessing of self-interest. Sometimes two genes can together accomplish more than their sum individually: in the authors' words "two men, each with one oar, can propel a boat, but one man with one oar will row in circles". Specialisation and division of labour, such as with bees, can also prove efficient.

Several general processes emerge that maintain a transition once it is in place. The "genetic bottleneck" is one. Multicellular organisms typically develop from one cell. This ensures all the cells in the body are, save for rare genetic mutations, identical. A high degree of relatedness makes it more likely that cells will behave altruistically towards one another, and provides one answer to the potentially embarrassing problem of shedding gametes. "Contingent irreversibility" is another process and refers to the notion that an entity that has replicated as part of a larger whole for long enough may lose the ability to do so on its own: our mitochondria would now perish if emancipated back into the wild.

"Central control" enables organisms to gang up on unruly elements within them that fail to cooperate. Genes in large complicated organisms have been selected to act in concert; there is, in one commentator's phrase, "a parliament of genes", although regular watchers of the mother of parliaments may wonder if this is the best metaphor. A "selfish" mutation to some gene that is part of this parliament makes that gene act for its own benefit, at the expense of the rest of the genes. The trouble is, for that selfish gene, there is so much genetic firepower available in the remainder of the organism, all of it arrayed against the one defector, that the defector is usually quickly subdued (readers may detect a distinct analogue to the behaviour of whips).

The central authority can act with great alacrity and nimbleness in a crisis. One form of selfish gene is known as a "driving element". If found on a chromosome responsible for determining the sex of the organism, a driving element ensures only chromosomes of one sex are produced. Genetic engineers have even used such drivers in attempts to control pests.

ÁñÁ«ÊÓƵ

Most of these attempts fail outside of the laboratory because suppressor mutations in natural populations rapidly evolve to turn off the driver. The male Y chromosome is particularly likely to give rise to driving elements. As a result this font of masculinity is very likely a graveyard of failed rogues, buried there by a strong central authority.

ÁñÁ«ÊÓƵ

The authors' rewarding efforts reveal an evolutionary landscape on which life plays out according to an enduring set of classical themes: "evil" in the form of selfish interest breathes life into "good" cooperative societies of genes and individuals, but constantly threatens to maul them. Kinship, the impossibility of going back home, and the collective will build structures but can only do so much against a force woven into the fabric of the society. Organisms are never the exclusively happy homes of cooperating genes all pulling together like ancient Olympian oarsmen.

Liars and cheats in the form of genetic renegades lurk around every corner. Indeed, as much as 90 per cent of the human genome (the total amount of chromosomal material in our cells) may consist of "junk" DNA, mute sequences of DNA that are never read by the cellular machinery. It is there because many genes will simply make copies of themselves until stopped almost literally by their bulk.

One cannot help but wonder if the human genome mapping industry realises that perhaps 90 per cent of its output never gets read. Often the renegade dog really bites. One heritable and deadly disorder known as myotonic dystrophy, has the property of "anticipation": the disease gets worse as it is passed from generation to generation.

One interpretation of this is that the gene responsible makes one or more additional copies of itself each generation at the time gametes are made inside the body. This gene is merely acting out its blind Darwinian ambition to increase its number of copies, but with sometimes devastating consequences for the rest of the parliament.

The cooperative cartels that have given us the biotic world are flimsy, depending as they do on a membership with their own interests at heart. Analogies to human societies beckon throughout, and one can almost hear Adam Smith applauding in the wings when it all works.

Modern societies themselves increasingly flirt with surrendering their sovereignty to ever larger units. The central authority - such as it is - of these larger units unfailingly trumpets the cause of mutual benefit. But as they do selfish voices chirp apocalyptically. Perhaps we should not be surprised.

Mark Pagel is research fellow, department of zoology, University of Oxford.