James MacMillan, one of the conductors of the BBC Philharmonic orchestra, claimed in a speech last night that the "ignorance-fuelled" hostility to faith shown by "metropolitan arts, cultural and media elites" risks making society bland and uniform.

However, the composer, who is Roman Catholic, claimed that atheists have not succeeded in "beating religion into a pulp".

"The campaigning atheists, as opposed to the live-and-let-live variety, are raising their voices because they recognise that they are losing; the project to establish a narrow secular orthodoxy is failing." He added that the religious must carry on expressing their beliefs in the face of growing opposition.

"A smug ignorance, a gross oversimplification and caricature that serves as an analytical understanding of religion, is the common intellectual currency. The bridge has to be built by Christians and others being firm in resisting increasingly aggressive attempts to still their voices."

He concluded by saying that our lives will become meaningless unless the "mists of contemporary banality" are penetrated and the idea of the sacred is restored.

"I believe it is God's divine spark which kindles the musical imagination now, as it has always done, and reminds us, in an increasingly dehumanised world, of what it means to be human."

Genes that predispose a significant minority of men to homosexuality raise a Darwinian puzzle. If homosexual men rarely father children, homosexual genes should dwindle to the low frequency expected from recurrent random mutation, a frequency below one in a million. Even if Kinsey's estimate of one in ten is high, there can be no doubt that the abundance of homosexual men is too great to have stemmed from recurrent mutation alone.

As long as the (always implausible) social science orthodoxy was maintained that homosexual inclinations were entirely made, not born, there was little problem. The recent demonstration that, not for the first time, the politically correct is factually incorrect, changes all that. Moreover, contrary to two Letters to the Editor of this newspaper, the evidence that the 'gay' gene lies on the X chromosome (which a man receives only from his mother, and cannot pass to his sons) provides no let-out. A man passes his X chromosome to all his daughters and, on average, a quarter of his grandsons. Any gene that reduces a man's daughters is subject to strong negative selection. It should, other things being equal, disappear.

When Darwinians are challenged by some seemingly un-Darwinian fact of human life, they often invoke the distortions of civilization. Why have we a taste for sugar when it rots our teeth? Because civilization blunts the cutting edge of natural selection, and in our ancestral past sugar was too scarce to do anything but good. Darwinians have framed similar theories about homosexuality: forget the ephemera of modern life, how might homosexual genes have fared during all those millennia on the African savannah?

Some of these theories note that genes have different effects in different contexts. Genes that promote homosexuality in, say, bottle-fed individuals might foster some advantageous trait in breast-fed individuals. Before the teated bottle was invented, the gene would not have surfaced as a gene 'for' homosexuality at all. It would have been a gene 'for' something quite different, perhaps resistance to a virus. Obviously I name 'bottle' and 'virus' only for the sake of argument. The general point is that the effects of a gene may depend upon context. As a special case, they may depend upon which other genes are present in the body. Homosexuality may therefore manifest itself in some individuals, as a spinoff from a gene's positive selection because of its desirable effect in other individuals. A particular version of this theory postulates a gene that causes homosexuality in males but a completely different, beneficial, effect in females.

Another theory, the 'sterile worker,' starts from the well-understood observation that worker bees, ants, wasps, termites and naked mole-rats divert their energy and time away from reproduction and towards the welfare of their young collateral relatives. Perhaps Pleistocene children, while their macho fathers were away hunting, were left under the protection of a gay uncle? The uncle's genes, including those promoting homosexuality, would have a good chance of being reproduced by the children whom he protected as surrogate father.

Incidentally the newly discovered 'gay gene', being on the X chromosome, could be shared by a maternal uncle's nephews (and nieces) but not by a paternal uncle's nephews. It is tantalising to recall the anthropological finding that, in those many societies where uncle replaces father as economic and protective guardian of a child, it is universally the mother's brother not the father's brother. Admittedly, this "mother's brother effect" already has an alternative Darwinian explanation.

In any case, the sterile worker theory doesn't explain why the uncles, in addition to refraining from normal masculine activities, should enjoy making love to men. Indeed one might think that, left in camp with the women, there is another obvious way in which they could benefit their genes, over and above caring for their nephews and nieces. This brings me to my own favourite, the 'sneaky male' theory.

In harem-based species, like some seals and deer, a minority of males monopolises the females, leaving a surplus of bachelors. Those supernumerary males that have no hope of displacing a harem-master sometimes specialise in an alternative, 'best of a bad job,' strategy: sneaking quick copulations with females while his back is turned. Genes promoting sneaking skills are passed on, in parallel with genes promoting the dominant male skill of bashing up other males.

You can tell harem species by their sexual dimorphism - males larger than females. Humans are less dimorphic than elephant seals (a dominant bull typically outweighs 14 females) but dimorphic enough to suggest at least some legacy of harem-based history. Clandestine matings with females may have provided the only route for surplus bachelors to pass on their genes. Their skills may have included lulling harem masters into a false sense of security, and now here is the point. A genuine preference for other males might well carry more conviction than a simulated indifference to females. By analogy, women frequently remark that they feel 'secure' in the company of homosexual men, and monarchs have staffed their harems with eunuchs. Incidentally, experts doubt the widely-promulgated story that the Ottoman Sultan Ibrahim was so jealous of a rumoured liaison between a eunuch and an unidentified odalisque that he drowned his entire 280-strong harem in the Bosporus. In any case homosexual men are not eunuchs and they can fertilise women. According to the sneaky male theory, their homosexual orientation gained them privileged access to women and a minority stream of homosexual genes prospered.

Explanations buried in Pleistocene history are always less convincing where reproduction, rather than survival, is at stake. Early death may have been largely abolished nowadays, but genes still vary in their ability to get themselves reproduced. If a homosexuality gene lowers its own probability of being reproduced today, and yet still abounds in the population, that is a problem for commonsense as much as for Darwin's theory of evolution. And, intriguing as several of these theories may be, I have to conclude that it remains a problem.

A particular version of this theory postulates a gene that causes homosexuality in males but a completely different, beneficial, effect in females.