The in-crowd

New Scientist

vol 181 issue 2430 - 17 January 2004, page 36

Same-sex relationships are not a biological dead end. They are a glue that helps hold many animal societies together, and a fatal flaw in one of Darwin's central ideas, argues evolutionary ecologist Joan Roughgarden.

IN June 1997 I marched in my first gay pride parade. I walked up Market Street in San Francisco, from the Civic Center to the Ferry Building. The parade was one of the biggest I had ever seen, and the sidewalks on both sides were packed six deep. I had heard that 1 in 10 people is gay or lesbian, but had always felt this number exaggerated. At this parade, though, I began to realise for the first time that the number of gays may indeed plausibly reach that figure.

This number of gay and lesbian people posed a problem to me as a biologist. My discipline teaches that homosexuality is some sort of unexplained anomaly. If the purpose of sexual contact is reproduction, as the standard explanation has it, how can all these gay people exist? One might argue they are somehow defective, that some developmental error or environmental influence has misdirected their sexual fantasies. If so, gay and lesbian people are here for a brief time during our species' evolution, awaiting removal when natural selection prunes those with lower Darwinian fitness.

Hmm. I began to wonder about the evolutionary puzzle of homosexuality. If a theory says something is wrong with so many people, then perhaps the theory is wrong, not the people. But I feared I would have to leave the puzzle unsolved. In a few months I was to come out as a transgendered woman. I didn't know whether I would be fired from my professorship at Stanford University, California, and find myself working as a waitress in a transgender bar. In the event, I wasn't fired - although I was removed from all administrative responsibilities - and I wound up with more time to investigate how evolution has led to diverse manifestations of gender and sexuality.

I found that evolutionary theory had followed a wrong path that leads inexorably back to Darwin - specifically to his theory of sexual selection, which I have concluded should be declared not only false but unfixable. Although I believe many biologists acknowledge that recent findings about gender and sexuality are problematic, few go as far as me in recommending that Darwin's theory of sexual selection be tossed out completely. So let me sketch the steps that have led me to this rather drastic and provocative conclusion - and to a better understanding of the biology of homosexuality and gender.

There are two glaring flaws in Darwin's thinking. In 1871 he wrote, "Females choose mates who are "more attractive...vigorous and well-armed" just as "man can give beauty...to his male poultry" by selective breeding. Hence the peacock's tail, Darwin's frequent example, is supposed to reflect peahen taste in male fashion, and antlers a preference for strong warrior stags. "Males of almost all animals have stronger passions than females," he wrote, and, "The female...with the rarest of exceptions is less eager than the male... she is coy." In Darwin's view, males and females almost universally conform to their preordained roles of horny handsome warriors and discreetly discerning damsels. But the real world is far more diverse than that. In many species, including ours, females are not necessarily less eager than males, nor do females all yearn for Arnold Schwarzenegger. Females often solicit males, and males often decline. Moreover, in many species the supposed sex roles reverse. Even Darwin acknowledged species of birds, like the jacana, in which the females are highly ornamented and the males dull and drab, reversing the peafowl story.

Many animals, indeed, do not even sort neatly into two sexes at all. If you go snorkelling on a coral reef, about one-third of the fish you see make both eggs and sperm at either the same time or different times during their lives. These are called simultaneous or sequential hermaphrodites respectively, and are said to "change sex" when they switch from making eggs to making sperm or vice versa. In fact the most common body plan among multicellular organisms, including plants, is for a single individual to make both male and female gametes at some time during their life. So the condition whereby an individual can be unambiguously classified as either male or female should not be considered the norm.

Species may also feature more than one type of male and female. The multiple morphs of males in such species all produce sperm, but otherwise differ in body size, colour, morphology, behaviour and life history so much that an inexperienced naturalist might be tempted to classify them as different species. The same is true for multiple kinds of females that have nothing in common except that they all make eggs, such as yellow-throated and orange-throated side-blotched lizards, which lay eggs of different sizes. I have termed these distinct morphs as "genders", and this terminology allows one to say there are more genders than sexes. The bluegill sunfish of the north-eastern US and Canada, for example, has three male genders that I term controllers, cooperators, and endrunners. The large, orange-breasted controllers and medium-sized cooperators, whose dark, barred colour pattern resembles that of females, court females jointly. The controller fertilises most of the eggs, but allows the cooperator a limited role as well. The small, pale endrunner males lurk in the weeds waiting to dash in while a female is laying her eggs and deposit some sperm of their own.

The second problem with Darwin's notion of sexual selection is that in relatively social species, such as most birds and mammals, sexual contact - mating - is not necessarily, indeed not even often, about the transfer of sperm. Mating is mostly directed at forming and managing relationships that may ultimately result in the successful production and rearing of offspring. A simple count of how many times mating takes place relative to the number of young born illustrates the point. In humans, for example, suppose Ozzie and Harriet have two children, have been married for 50 years, and make love regularly each week - say, Thursday night. After 50 years they will have mated over 2500 times, and produced two children, thus mating 1250 times per offspring produced. Sounds inefficient? Not if we suppose that regular mating allows the couple to stay together to successfully rear their two children. Similarly, in birds, primates, indeed everywhere, lots of mating occurs at times and places that cannot possibly result in immediate offspring production.

By this stage of my research I was beginning to suspect that Darwin might be all wrong about sex. It seemed to me that social organisation in animals revolves around the control of access to reproductive opportunity, which includes all the things that animals need to reproduce: food and nest sites, for instance, as well as mates. Animals make direct use of the resources they control, but may also use them as bargaining chips to buy the help of others. Furthermore, the dynamics of animal societies also involve decisions about where to allocate friendship and cooperation among animals of both the same sex and the other sex. Different arrangements of cooperative effort lead to the emergence of different structures for families and small groups.

This theory, which I call social selection theory, provides a better explanation for much of the diversity we see in sexual practices. In bluegill sunfish, for example, social life does not, as traditional sexual selection theory requires, consist of females looking for males with great genes, or of males trying to fool females into thinking their genes are better than their neighbours'. Instead, it is about the ebb and flow of power to control access to reproductive opportunity. I suggest that controller males pay cooperator males for "marriage broker" services by allowing the cooperator to fertilise some of the eggs in his territory. In return, the cooperator male assists in courting females. Controller males without a cooperator male do not fare as well at attracting females. The feminine coloration pattern of the cooperator male may somehow promote this function, perhaps by allowing the cooperator male to develop a relationship with the females while the controller male is setting up and defending his territory.

The aspects of the relationships between animals that are managed by mating depend on the species' social system. Anthropologist Sarah Hrdy of the University of California at Davis has shown that female monkeys in India mate with multiple males so that each will refrain from harming the young because he might be the father. In addition to managing male power, mating helps the pair bond and ensures that males deliver on their promise of parental investment, preventing them from becoming dead-beat dads.

Social selection theory also explains a puzzle that goes back all the way to Aristotle: the "penises" of female spotted hyenas. The female's clitoris is enlarged to the size of a male penis, and fat deposits in a nearby skin pouch resemble a scrotal sac. Females erect their penis many times during the day in interactions with other females. Sexual selection theory has no explanation for such an unusual characteristic that is not used in mate choice. I suggest, though, that a female spotted hyena that did not have a penis would be excluded from the female groups that control access to reproduction. This is an instance of what I call a social inclusionary trait: a trait that gains an individual admission to a social group, whether or not it has any other use. The human brain, with all its powerful capacity for conversation, art and music, may be another such trait.

Same-sex sexuality in female bonobos is another social inclusionary trait. I conjecture that females that do not participate regularly in mutual face-to-face genital rubbing do not form the bonds needed to participate in the groups that control access to food, or enjoy the protection necessary to raise young successfully.

Taken to its logical conclusion, this argument might imply that even classic sexual ornaments such as the peacock's tail or a stag's antlers are not there to attract females by advertising the bearer's virility. Instead, these traits may be intended for members of the same sex more than for the opposite sex. They may be badges of admission to membership in power-holding cliques. I am not aware of any experiments to test whether secondary sexual characters are really badges or ornaments. Some experiments have shown how modifying traits such as feather colours affects mate choice. I feel such experiments should also investigate how these modifications affect same-sex relationships, including membership in power-holding cliques.

This new perspective on animal social behaviour, and its rejection of Darwin's theory of sexual selection, undercuts the subject of evolutionary psychology. Many biologists are becoming increasingly uneasy at how psychologists have retooled sexual selection theory into a theory of human personality, complete with evolutionary rationales for everything from beauty to rape. Being upfront about how problematic sexual selection theory is may help curtail this misuse of biology.

I have now come full circle to the question I started with, the puzzle of homosexuality and gender, and the difficulty it poses for Darwinian sexual selection theory. Author Bruce Bagemihl, in his book Biological Exuberance: Animal homosexuality and natural diversity, has catalogued over 300 vertebrate species in which same-sex genital contact regularly occurs. In some species, homosexuality is not very common - around 1 to 10 per cent of all mating. In others, such as bonobos, homosexual mating occurs as often as heterosexual mating. In some species only males participate, in others only females, in still others both sexes. Sometimes homosexuality is associated with pair bonds that last for years, and in others with short-term consortships. This broad occurrence of homosexuality among vertebrates raises the possibility that if it has a genetic basis at all, it has some broad adaptive significance, and is not an aberrant condition just a few species happen to be stuck with.

In humans, moreover, homosexuality is much too common for it to be considered a genetic aberration. Real genetic diseases are really rare, and their frequency inevitably depends on their severity. A disease that is uniformly lethal must arise anew each generation, so its frequency is equal to the mutation rate, say one in 1 million. A disease that causes only a 10 per cent drop in offspring production (fitness) is 10 times more common than a lethal disease - about one in 100,000. Similarly, a mere 1 per cent drop in fitness leads to a frequency of one in 10,000. If homosexuality has a frequency of 1 in 10, the fitness loss could be no more than 0.001 per cent, which is completely undetectable. A "common genetic disease" is a contradiction in terms, and homosexuality is three to four orders of magnitude more common than true genetic diseases such as Huntington's disease.

Indeed, I challenge the presumption that homosexuality leads to any reduction in fitness whatever. Throughout history and across cultures, homoerotic attraction has not precluded heteroerotic attraction. And there is little evidence that people who feel homoerotic attraction have, as a group, any less Darwinian fitness than those who don't. After all, many exclusively heterosexual people do not have offspring either. Even if those with homoerotic attraction did have marginally fewer children, they might make up for it by a better chance of survival - during wars, for example, when homoerotic bonds might lead soldiers to protect one another more vigorously. So what then, is the adaptive significance of homosexuality? Homosexuality has many uses, much as the ability to speak does. Homosexual contact is a way to communicate pleasure. And I suggest that homosexuality is a social inclusionary trait - that is, it provides animals, including perhaps humans at times, with admission to social groups. It evolves, I suggest, whenever same-sex cooperation helps achieve an evolutionarily successful life: to survive, find mates and protect one's young from harm. This plays out in different ways in different sexes and species. Sometimes, as with bonobos, same-sex cooperation provides group security and access to food that females need to successfully rear their young. For others, such as male Savanna baboons and probably some whales, it provides the allies they need to survive conflicts so that they may later mate. But the unifying principle is the same - homosexuality cements relationships that are crucial for a successful life.

Joan Roughgarden

Joan Roughgarden is a professor of biology at Stanford University in California.