Science writer Matt Ridley once described the human mating system as "monogamy plagued by adultery", which sounds a little judgmental.
Perhaps we should just agree that we are an imperfectly pair-bonding species. Quite imperfectly - I am on my second marriage, and so is my wife - but the point is that we do form pairs: 89 per cent of the world's people get married before the age of 49.
Elsewhere in the animal world, monogamy is definitely a minority taste. Only 3 per cent of mammals are monogamous. Even among our closest relatives, the primates, only a quarter of the species form pair bonds. Moreover, the very fragility of the pair bond in human beings suggests that it is a behaviour we only adopted fairly recently in our evolutionary history. So when did we acquire it, and why?
There is a new explanation on the table. In a study published this week in the Proceedings of the National Academy of Sciences, scientists argue that the main reason why human beings - more precisely, male human beings - became monogamous was to keep their babies from being killed by other men. There are many species where an incoming male will kill a female's offspring by a previous male in order to make room for his own. It's especially common in mammals, where a female remains infertile while she is still producing milk for an existing baby. The new male is in a hurry to get on with fathering the bearers of his own genes, and if he kills her existing offspring she will become fertile again.
This may have been a particularly big problem in our own species, because human females may nurse a child for as long as two or three years. Infanticide is ugly, but unfortunately it makes sense as a male reproductive strategy. So it also makes sense for the father of the existing children to stick around and protect them from that fate. This was the hypothesis of the scientists from University College London and Oxford, Manchester and Auckland Universities who published the article in "Proceedings". "You do not get monogamy unless you already have infanticide, and you do not get a switch to paternal care [by males] if you don't already have monogamy," wrote Dr Christopher Opie of UCL.
The researchers decided to test the rival hypotheses: did males commit to monogamy to prevent infanticide, or to ease the burden on females and thus improve the children's survival chances? The cynics among you will already know the answer to this, but scientists actually have to prove things. They took a family tree of 230 mammals, including most of the primate species, and put in the details of their mating behaviour, rates of infanticide, and amount of paternal care (ie monogamy). Then they simulated the evolution of those species over a period of 75 million years, running the programme millions of times to see how monogamy rose or fell for each species under different circumstances.
The conclusion was clear: among primates, monogamy was always preceded by one thing and one thing only - infanticide by males. Once you have monogamy, there is usually a rise in the male commitment to caring for the offspring as well, but infanticide has to come first. Gwynne Dyer is an independent journalist whose articles are published in 45 countries.