The Science of Discworld II

about points of fine detail. And there are some quite substantial areas of disagreement, such as the recommended treatment of women, or to what extent basic rights should be extended to the infidel. On the whole, however, there is a strong consensus in such teachings, for example an almost universal condemnation of theft and murder. Virtually all religions reinforce a very similar consensus of what constitutes ‘good’ behaviour, perhaps because it is this consensus that has survived the test of time. In terms of the barbarian/tribal distinction, it is a tribal consensus, reinforced by tribal methods such as ritual, but none the worse for that.
    Many people find inspiration in their religion, and it helps instil a sense of belonging. It enhances their feeling of what an awesome place the universe is. It helps them cope with disasters. With exceptions, mainly related to specific circumstances such as war, most religions preach that love is good and hatred is bad. And throughout history, ordinary people have made huge sacrifices, often of their own lives, on that basis.
    This kind of behaviour, generally referred to as altruism, has caused evolutionary biologists a great deal of head-scratching. First, we’ll summarise how they have thought about the problem and what kinds of conclusion they have reached. Then we’ll consider an alternative approach, originally motivated by religious considerations, which looks to us to be far more promising.
    At first sight, altruism is not a problem. If two organisms cooperate, by which in this context we here mean that each is willing to risk its life to help the other, 1 then both stand to gain. Natural selection favours such an advantage, and reinforces it. What more explanation is needed?
    Quite a lot, unfortunately. A standard reflex in evolutionary biology is to ask whether such a situation is stable – whether it will persist if some organisms adopt other strategies. What happens, for example, if most organisms cooperate, but a few decide to cheat? If the cheats prosper, then it is better to become a cheat than to cooperate, and the strategy of cooperation is unstable and will die out. Using the methods of mid-twentieth-century genetics, the approach pioneered by Ronald Aylmer Fisher, you can do the sums and work out the circumstances in which altruism is an evolutionarily stable strategy. The answer is that it all depends upon whom you cooperate with, whose life you risk your own to save. The closer kin they are to you, the more genes they share with you, so the more worthwhile it is for you to risk your own safety. This analysis leads to conclusions like ‘It is worth jumping into a lake to save your sister, but not to save your aunt.’ And certainly not to save a stranger.
    That’s the genetic orthodoxy, and like most orthodoxies, it is believed by the orthodox. On the other hand, though: if someone has fallen intoa lake, people do not ask ‘Excuse me, sir, but how closely related are you to me? Are you, by any chance, a close relative?’ before diving in to rescue them. If they are the sort of people who dive in, they do so whoever has fallen into the lake. If not, they don’t. Mostly. A clear exception arises when a child falls in; even if they can’t swim its parent is then very likely indeed to plunge in to the rescue, but probably would not do so for someone else’s child, and even less so for an adult. So the genetic orthodoxy does have a certain amount going for it.
    Not much, though. Fisher’s mathematics is rather old-fashioned, and it rests on a big – and very shaky – modelling simplification. 2 It represents a species by its gene-pool, where all that matters is the proportion of organisms that possess a given gene. Instead of comparing different strategies that might be adopted by an organism, it works out what strategy is best ‘on average’. And inasmuch as individual organisms are represented within its framework at all, which they are only as contributors to the gene-pool, it views competition between organisms as a direct ‘me versus thee’ choice. A bird that eats seeds is up against a bird that eats worms in a head-to-head struggle for survival, like two tennis-players … and may the best bird win.
    This is a bean-counting analysis performed with a bean-counting mentality. The bird with the most beans (energy from seeds or worms, say) survives; the other does not.
    From a