Why not Allah's omnipotent power? Or Lord Brahma's? Or even Yahweh's?

Why, I can't help wondering, is God thought to need such ferocious defence? One might have supposed him amply capable of looking after himself. Bear in mind, through all this, that the Editor being abused and threatened so viciously is a gentle and charming young woman.

Perhaps because I don't live in America, most of my hate mail is not quite in the same league, but nor does it display to advantage the charity for which the founder of Christianity was notable. The following, dated May 2005, from a British medical doctor, while it is certainly hateful, strikes me as more tormented than nasty, and reveals how the whole issue of morality is a deep wellspring of hostility towards atheism. After some preliminary paragraphs excoriating evolution (and sarcastically asking whether a 'Negro' is 'still in the process of evolving'), insulting Darwin personally, misquoting Huxley as an anti-evolutionist, and encouraging me to read a book (I have read it) which argues that the world is only eight thousand years old (can he really be a doctor?) he concludes:

The sentiment of this letter, if not its tone, is typical of many. Darwinism, this person believes, is inherently nihilistic, teaching that we evolved by blind chance (for the umpteenth time, natural selection is the very opposite of a chance process) and are annihilated when we die. As a direct consequence of such alleged negativity, all manner of evils follow. Presumably he didn't really mean to suggest that widowhood could follow directly from my Darwinism, but his letter, by this point, had reached that level of frenzied malevolence which I repeatedly recognize among my Christian correspondents. I have devoted a whole book (Unweaving the Rainbow) to ultimate meaning, to the poetry of science, and to rebutting, specifically and at length, the charge of nihilistic negativity, so I shall restrain myself here. This chapter is about evil, and its opposite, good; about morality: where it comes from, why we should embrace it, and whether we need religion to do so.

Several books, including Robert Hinde's Why Good is Good, Michael Shermer's The Science of Good and Evil, Robert Buckman's Can We Be Good Without God? and Marc Hauser's Moral Minds, have argued that our sense of right and wrong can be derived from our Darwinian past. This section is my own version of the argument.

On the face of it, the Darwinian idea that evolution is driven by natural selection seems ill-suited to explain such goodness as we possess, or our feelings of morality, decency, empathy and pity. Natural selection can easily explain hunger, fear and sexual lust, all of which straightforwardly contribute to our survival or the preservation of our genes. But what about the wrenching compassion we feel when we see an orphaned child weeping, an old widow in despair from loneliness, or an animal whimpering in pain? What gives us the powerful urge to send an anonymous gift of money or clothes to tsunami victims on the other side of the world whom we shall never meet, and who are highly unlikely to return the favour? Where does the Good Samaritan in us come from? Isn't goodness incompatible with the theory of the 'selfish gene'? No. This is a common misunderstanding of the theory — a distressing (and, with hindsight, foreseeable) misunderstanding.*) It is necessary to put the stress on the right word. The selfish gene is the correct emphasis, for it makes the contrast with the selfish organism, say, or the selfish species. Let me explain.

The logic of Darwinism concludes that the unit in the hierarchy of life which survives and passes through the filter of natural selection will tend to be selfish. The units that survive in the world will be the ones that succeeded in surviving at the expense of their rivals at their own level in the hierarchy. That, precisely, is what selfish means in this context. The question is, what is the level of the action? The whole idea of the selfish gene, with the stress properly applied to the last word, is that the unit of natural selection (i.e. the unit of self-interest) is not the selfish organism, nor the selfish group or selfish species or selfish ecosystem, but the selfish gene. It is the gene that, in the form of information, either survives for many generations or does not. Unlike the gene (and arguably the meme), the organism, the group and the species are not the right kind of entity to serve as a unit in this sense, because they do not make exact copies of themselves, and do not compete in a pool of such self-replicating entities. That is precisely what genes do, and that is the — essentially logical — justification for singling the gene out as the unit of 'selfishness' in the special Darwinian sense of selfish.

The most obvious way in which genes ensure their own 'selfish' survival relative to other genes is by programming individual organisms to be selfish. There are indeed many circumstances in which survival of the individual organism will favour the survival of the genes that ride inside it. But different circumstances favour different tactics. There are circumstances — not particularly rare — in which genes ensure their own selfish survival by influencing organisms to behave altruistically. Those circumstances are now fairly well understood and they fall into two main categories. A gene that programs individual organisms to favour their genetic kin is statistically likely to benefit copies of itself. Such a gene's frequency can increase in the gene pool to the point where kin altruism becomes the norm. Being good to one's own children is the obvious example, but it is not the only one. Bees, wasps, ants, termites and, to a lesser extent, certain vertebrates such as naked mole rats, meerkats and acorn woodpeckers, have evolved societies in which elder siblings care for younger siblings (with whom they are likely to share the genes for doing the caring). In general, as my late colleague W. D. Hamilton showed, animals tend to care for, defend, share resources with, warn of danger, or otherwise show altruism towards close kin because of the statistical likelihood that kin will share copies of the same genes.

The other main type of altruism for which we have a well-worked-out Darwinian rationale is reciprocal altruism (You scratch my back and I'll scratch yours'). This theory, first introduced to evolutionary biology by Robert Trivers and often expressed in the mathematical language of game theory, does not depend upon shared genes. Indeed, it works just as well, probably even better, between members of widely different species, when it is often called symbiosis. The principle is the basis of all trade and barter in humans too. The hunter needs a spear and the smith wants meat. The asymmetry brokers a deal. The bee needs nectar and the flower needs pollinating. Flowers can't fly so they pay bees, in the currency of nectar, for the hire of their wings. Birds called honeyguides can find bees' nests but can't break into them. Honey badgers (ratels) can break into bees' nests, but lack wings with which to search for them. Honeyguides lead ratels (and sometimes men) to honey by a special enticing flight, used for no other purpose. Both sides benefit from the transaction. A crock of gold may lie under a large stone, too heavy for its discoverer to move. He enlists the help of others even though he then has to share the gold, because without their help he would get none. The living kingdoms are rich in such mutualistic relationships: buffaloes and oxpeckers, red tubular flowers and hummingbirds, groupers and cleaner wrasses, cows and their gut micro-organisms. Reciprocal altruism works because of asymmetries in needs and in capacities to meet them. That is why it works especially well between different species: the asymmetries are greater.