‘Me last’ society?

Review of Unto Others :
The Evolution and Psychology of Unselfish Behaviour,
by Elliott Sober and David Sloan Wilson, Harvard University Press, 1998
The Psychologist, 13, p 145

This is the version originally submitted. It was edited prior to publication.


A caretaker from Essex made international news by donating one of his kidneys to a young girl he had never met, in an (as it turned out) vain attempt to save her life. Why did he do it? Why do people ever go to extreme lengths to help each other when there appears to be nothing in it for them? Or should I say – when there is nothing in it for their genes?

Evolutionary accounts of altruism have recently become popular, with such readable books as Ridley’s The Origins of Virtue and Wright’s The Moral Animal. They have in common an infectious enthusiasm for the idea that the ultimate causes of human motivation can be found in the selfish gene. With theories of kin selection, reciprocal altruism and game theory, apparent altruism can be explained as the kind of behaviour which – if our ancestors had carried it out – would have benefited the replication of their genes. In this way an extraordinarily wide range of human cooperative behaviours can make sense – but there is no real altruism.

Unto Others is not like this. Sober and Wilson take apart the usual assumptions in a systematic onslaught with which they claim to be “tearing down the edifice of individualism” (p 330). First they clearly distinguish between the concepts of evolutionary and psychological altruism. The former is concerned with behaviours that affect survival, reproduction and fitness regardless of how an organism thinks or feels about what it is doing; the latter is concerned far more with motivations and beliefs. Having made sure the reader understands why these are logically distinct – Sober and Wilson begin their task of building the case for true altruism of both kinds.

Their evolutionary argument hangs entirely on the status of “group selection” – among the most troublesome and controversial notions in evolutionary biology. Most biologists have learned, almost as an article of faith, that group selection does not occur. In other words, there is no evolution “for the good of the species”, only competition between selfish genes. The original reason for this dogma was sound enough. In the 1950s and 1960s biologists often invoked “the good of the group” as an evolutionary argument when there was no conceivable mechanism to make it work. In 1966 G.C. Williams exposed all the confusion, and group selection was out. But, say Sober and Wilson, group selection can be an important force in evolution, not just in sex ratios and parasites, but in human groups too. This is because what happens at the group level can be quite different from what happens at the level of the individual.

A weird example that helped me think about this is an illustration of “Simpson’s paradox”. In the 1970s the University of California was suspected of discrimination against women because the percentage of successful women applicants was significantly lower than that for men. However, every single department operated their admissions fairly and did not discriminate. The paradox was resolved by the discovery that women tended to apply far more often to the departments that were hardest to get into. So overall fewer of them succeeded.

To understand this example you have to think at the level of the group, and this is true in evolution too – whenever there is selection between groups. So when one faction goes to war and kills another, or when an agricultural society sweeps away a nomadic culture, there is selection between groups, and this means that features of those groups – like how much they help each other – can be a powerful force. No wonder human societies, as Sober and Wilson illustrate with a random survey of ethnographies, are characterised by strongly enforced social norms and behaviours that benefit the group.

They then turn to the question of psychological altruism. They overview the concepts of hedonism, egoism and altruism, and explore the importance of understanding motives as proximate mechanisms for producing adaptive behaviours. Unfortunately the main example they use is that of parental care, which seems the least peculiar kind of human caring, and the one most easily explained by the usual evolutionary arguments. In a last look at human uniqueness they conclude that cultural variation has produced adaptations that have nothing to do with genes, and their final conclusion is a pluralist view of human motivation. With group selection reinstated, they argue, humans can sometimes be selfish and egotistical, but sometimes truly altruistic.

Sober and Wilson have seriously infuriated biologists their with views on group selection. I doubt they will have as much impact on psychology, but they have certainly broken the usual mould of evolutionary psychology. Maybe altruism does not always have to be explained away as selfishness in disguise.