The Science of Sexual Orientation
This is a chapter-by-chapter summary of my book, Gay, Straight, and the Reason Why: The Science of Sexual Orientation (Oxford University Press, 2011). Please see the book itself for a much more detailed account of the science, along with illustrations and over 600 references to the scientific literature.
Chapter 1: What is Sexual Orientation?
Sexual orientation has to do with the sex of persons that we find sexually attractive. In the most common usage, it’s defined as the trait that predisposes us to experience sexual attraction to people of the same sex as ourselves (homosexual, gay, or lesbian), to persons of the other sex (heterosexual, straight), or to both sexes (bisexual). Sexual orientation is often represented in terms of a graduated scale, such as the 7-point Kinsey scale, which ranges from 0 (exclusively attracted to persons of the opposite sex) to 6 (exclusively attracted to persons of the same sex), or a similar 5-point scale. This kind of classification emphasizes the commonality between heterosexual men and heterosexual women (because both are attracted to persons of the opposite sex) and between gay men and lesbians (because both are attracted to same-sex partners).
An alternative classification uses the terms gynephilic (sexually attracted to females) and androphilic (sexually attracted to males). This nomenclature emphasizes the commonality between heterosexual men and lesbians (because both are attracted to females), and between heterosexual women and gay men (because both are attracted to males). Bisexual people occupy the same central region in either classification.
In both sexes, the great majority of individuals are exclusively heterosexual, but the distribution of non-heterosexual people is different between the sexes. In men, exclusively homosexual (gay) individuals (group 5 in the 5-point classification) are more common than bisexual individuals, so the overall distribution is two-peaked (bimodal). In women, exclusively homosexual (lesbian) individuals are less common than bisexual individuals, so the overall distribution has only one peak.
It’s possible that surveys such as Laumann’s have underestimated the true numbers of gay and bisexual people, given that interviewees may be reluctant to admit to a stigmatized sexual orientation. However, more recent studies that have paid a great deal of attention to confidentiality and privacy have not come up with larger percentages.
In laboratory studies that have measured subjects’ genital arousal while viewing erotic images or videos, most men show arousal only to images of one sex (either female or male), whereas most women show arousal to images of both sexes. At the level of genital arousal, then, sexual orientation is usually more category-specific in men than in women.
Not many studies have examined how stable a person’s sexual orientation is over the life course. One large study conducted in New Zealand found that nearly all heterosexual men and women, and non-heterosexual (bisexual/gay) men, remained in those same groups over a 5-year time span, but significant numbers of non-heterosexual women changed their sexual orientation. Lisa Diamond of the University of Utah followed non-heterosexual women over a period of 10 years. Many of these women changed the way they described themselves over the course of the study; to what extent this represented an actual change in their sexual attractions is unclear, but it does seem that sexual orientation may be more fluid in women than in men.
Efforts to change people’s sexual orientation through psychotherapy or other interventions have been largely unsuccessful, and they may have traumatizing effects of those who undergo them. For that reason, and because being gay, lesbian, or bisexual is not a mental disorder, professional organizations such as the American Psychological Association discourage conversion attempts.
Chapter 2: Why We Need Biology
One reason for being interested in biological theories of sexual orientation is the evident weakness of other, more traditional approaches. This chapter reviews some of the non-biological theories.
Psychoanalytic theory. Sigmund Freud (1856-1939) proposed that the sex drive (libido) goes through several important transformations during infancy and childhood. He spelled out his ideas most clearly for boys, who were said to go through a period of sexual fixation on their mothers (oedipal phase) between roughly the ages of 3 and 5. Boys might become homosexual adults because they remained stuck in a homosexual phase that supposedly preceded the oedipal phase (pre-oedipal homosexuality) or because they failed to emerge from the oedipal phase (oedipal homosexuality). Freud believed that a close-binding, seductive mother, or distant or hostile mother, was the likely cause of such developmental arrests. In girls, events such as the birth of a younger brother, who might evoke envy on account of their possession of a penis, would be the precipitating factor driving the emergence of lesbian sexuality.
It is true that, on average, gay men report their relationships with their mothers as closer, and their relationships with their fathers as more distant, as compared with straight men. However, this likely reflects the fact that many boys who later become gay adults already have distinct personalities in childhood – personalities that may be attractive to mothers but aversive to fathers. In other words, Freud may have noted an actual correlation but misunderstood the direction of cause and effect. Freud’s ideas with regard to female homosexuality are too bizarre and inconsistent to be taken seriously.
Learning theory. Many ideas about sexual orientation see it as the product of some form of learning. For example, sexologist John Money (1921-2006) proposed that the mind of a newborn child is a blank slate with regard to gender characteristics, including sexual orientation, and that these characteristics are inculcated by rewards, punishments, and role modeling provided by parents, peers, teachers, etc. Most famously, Money documented the history of a child, born male, who lost his penis in a botched circumcision operation and was subsequently raised as a girl. This child, Money claimed, was successfully raised with a female identity and would eventually become a heterosexual woman, but this did not happen. In reality, the child reclaimed his male identity and insisted on becoming a heterosexual man. Other studies of children with unusual genital conditions support the idea that, however they are raised, their biological nature as male or female has a strong influence on their ultimate gender identity and sexual orientation.
Sexual orientation as a choice. Not many researchers think that conscious choice plays much of a role in the development of sexual orientation. Nevertheless, polls show that about one-third of Americans believe that being homosexual is a “lifestyle choice.” (No poll that I’m aware of has asked whether being straight is a choice.) If their sexual orientation was a choice, gay people should remember having made it, but according to surveys the great majority say that they did not. Most probably, the choice/no choice controversy arises from differences in the way people conceptualize sexual orientation. For people who think of it simply as a pattern of behavior, choice does seem relevant, because after all we do have conscious control over how we behave. At the level of sexual desire and sexual attraction, however, there is no evidence that conscious choice comes into the equation, whether for gay, straight, or bisexual men and women.
Evidence for the biological alternative. Biological psychologists working in diverse fields have made observations relevant to our understanding of sexual orientation:
· Homosexual behavior is common among non-human animals, and in at least one species (sheep) individual animals have a durable preference for same-sex partners.
· Gay and straight people differ in a wide variety of gendered traits, during both childhood and adulthood.
· There is evidence that sex hormones influence the development of these gendered traits.
· There is evidence that genes influence sexual orientation and gender.
· Differences in brain organization exist between gay and straight people.
· Differences exist in the structure and function of the bodies of gay and straight people.
· Birth order influences sexual orientation in men, apparently through biological processes.
The following chapters lay out the evidence for these statements and attempt to tie them together into a coherent biological theory of sexual orientation.
Chapter 3: The outline of a theory
Experimental studies in animals find that the levels of sex hormones to which an animal’s brain is exposed during a “critical period” of development strongly influences that animal’s tendency to approach, court, mate with, and partner with opposite-sex or same-sex partners. In mammals the key hormone is testosterone. In rats, for example, this critical period extends from a few days before birth to a few days after birth. High levels of testosterone, such as are typical seen in male fetuses (but which may be artificially induced in females), promote the development of brain circuitry responsible for male-typical traits, including sexual interactions with females. Low levels of testosterone, such as are typically seen in female fetuses (but which may be artificially induced in males), promote the development of brain circuitry mediating female-typical traits, including sexual interactions with males. Manipulating sex hormone levels later in life, in contrast, has relatively little effect on the animal’s sexual and gendered behaviors.
Several researchers have suggested that a similar mechanism influences the development of sexual orientation in humans. In this mechanism, the levels of testosterone (T) during a sensitive period of brain development, which in humans is likely to be well before birth, influences the sexual differentiation of the brain—that is, the formation of brain structures and patterns of connectivity that are more typical of males or females. These gender-differentiated brain systems in turn influence sexual orientation and other gendered traits during postnatal life.
According to this model, genetic variation between individuals can influence testosterone levels during the sensitive period, and also influence how the brain responds to testosterone. Through these pathways genes influence a person’s ultimate sexual orientation. There may also be random biological variations during development, not controlled by genes, that impinge on the same pathway. Finally, external events may affect the pathway, either during the sensitive period or later, during postnatal life. There is no direct evidence for such external influences in humans, but they have been documented in some animal studies.
This model didn’t originate with me: It can be traced back to the ideas of Magnus Hirschfeld in the early 20th century. More recently, researchers such as Guenther Doerner in Germany, Lee Ellis in the United States, and Qazi Rahman in Britain have promoted similar ideas. The following chapters of this book discuss the present status of the model.
Chapter 4: Childhood
There are differences between the behavioral and personality traits typically shown by girls and boys. For example, boys engage in more rough-and-tumble play than girls do, and this sex difference exists in other primate species too. It’s believed that this and other juvenile sex differences in monkeys result from differences in prenatal testosterone levels: increasing testosterone levels artificially in female monkey fetuses causes them to engage in more rough-and-tumble play when they are juveniles. Although similar experiments can’t be done in humans, nature has done the experiment for us: girls with the genetic condition called congenital adrenal hyperplasia (CAH), who are exposed to unusually high testosterone levels prenatally, engage in more rough-and-tumble play than other girls, and are gender-atypical in a variety of other (but not all) childhood traits. This suggests that prenatal sex hormone levels influence gendered juvenile characteristics in humans just as they do in animals.
Children who later become gay or lesbian adults (“pre-gay children”) differ, on average, from their same-sex peers in a variety of gendered traits. For example, boys who become gay men are (on average) less into rough-and-tumble play than boys who become straight men, while for girls who become lesbian adults it’s the reverse. This is known, first, on the basis of retrospective studies (studies in which gay and straight adults are asked about their childhood traits). It is also known from prospective studies: children who are markedly gender-nonconformist in childhood, when followed through to adulthood, are more likely to become gay, lesbian, or bisexual than children who are more typical in their gender characteristics. Finally, it is known from recent studies in which gay and straight adults provided video clips of themselves as children: Viewers who didn’t know the children’s ultimate sexual orientation judged the “pre-gay boys” to be more feminine, and the “pre-lesbian girls” to be more masculine, than the children who became heterosexual men and women.
In the simplest interpretation of these findings, gay people are (on average) gender-atypical in a variety of gendered traits during childhood because the biological processes of prenatal brain development have gone forward differently. It could be, for example, that testosterone levels were lower in “pre-gay” male fetuses, so that a variety of brain systems contributing to gendered traits, including childhood behaviors and adult sexual orientation, were less strongly masculinized than in other males. Similarly, it could be that testosterone levels were higher in “pre-lesbian” fetuses, causes childhood behaviors and adult sexual orientation to be more masculinized than in other females. Alternatively, there might be differences, not in the hormone levels themselves, but in the sensitivity of the brain to these hormones.
Chapter 5: Characteristics of gay and straight adults
Adult men and women differ, on average, in many cognitive and personality traits. In the cognitive realm, women perform better than men on some memory tasks, while men do better on visuospatial tasks. In the personality realm, men score higher in instrumentality (assertiveness, aggressiveness, independence, etc), while women score higher in expressiveness and sociability. Men prefer more thing-oriented activities, women prefer people-orientated activities. Not surprisingly, men rate themselves more masculine and women rate themselves as more feminine. There is evidence that these differences result in significant part from differences in the levels of testosterone and similar sex hormones before birth: For example, women who were exposed to unusually high levels of testosterone prenatally (due to a genetic condition called congenital adrenal hyperplasia or CAH) are shifted in the male direction in a number of these gendered traits.
In many but not all of these gendered traits, gay people score in a fashion that is shifted toward the other sex. Gay and bisexual men, on average, rate themselves more feminine than straight men do, while lesbians and bisexual women rate themselves more masculine than straight women do. Comparable gender shifts have been reported for gay people in visuospatial skills, memory skills, verbal fluency, occupational preferences, instrumentality, and empathy. Some of these shifts have been replicated in a number of different cultures.
Thus it appears that homosexuality (and probably bisexuality too) is not an isolated trait but part of a “package” of gender-variant traits that are evident both in childhood and adulthood. This package varies considerably from individual to individual. Nevertheless, the fact that this package exists, in a statistical sense at least, suggests that the factors that influence sexual orientation are (at least in significant part) those that influence gendered traits in general. Because prenatal hormones and genes are known to influence gendered traits, the question arises as to whether there is evidence that they influence sexual orientation too. The evidence bearing on this is discussed in the following chapters.
Chapter 6: The Role of Sex Hormones
Experiments in non-human animals have shown that testosterone levels during a “critical period” of brain development influence sex behavior, including preference for male or female partners, in adulthood.
Many studies of women with CAH, who were exposed to unusually high levels of testosterone and similar masculinizing hormones before birth, have found that these women are unusually likely to experience same-sex desire and same-sex relationships, compare with control groups such as their unaffected sisters. Data of this kind support the notion that prenatal testosterone levels influence sexual orientation—but equally, they suggest that these hormone levels may not be the entire story, as many affected women are still heterosexual.
CAH is an unusual condition—so what about the majority of women and men who don’t have any obvious endocrinological disturbance? Researchers have studied anatomical and physiological traits in gay and straight adults that are thought to serve as “historical markers”—providing information about hormone levels in those individuals when they were fetuses. Examples discussed in the book include the relative lengths of the 2nd and 4th fingers (the 2D:4D ratio), limb length, and certain physiological properties of the auditory system. These studies do suggest that lesbians and bisexual women did have higher-than-usual testosterone levels during fetal life. The evidence regarding gay and straight men is less clear.
Sex hormones affect brain development by interacting with a complex assembly of receptors, genes, and other molecules that control neuronal survival, growth, and the formation of synaptic connections. Differences between individuals in gendered traits such as sexual orientation could arise either because of differences in the levels of prenatal hormones themselves, or because of differences in the brain mechanisms that respond to hormones.
Chapter 7: The Role of Genes
Pioneering studies led by Boston University psychiatrist Richard Pillard found that homosexuality runs in families: finding one gay person in a family increases the chances of finding others. Pillard’s work related mainly to siblings, but other studies have found examples of multiple gay people cropping up at more distant locations on family trees.
Data such as these suggest that the clustering in families is not due to how parents raise their children, since the gay people among these more distant relatives were not raised by the same parents. Could the clustering be due to “gay genes” running in certain families? To answer this question researchers have studied twins. The idea is that, if sexual orientation is influenced by genes, “identical” twins (who share the same genes) should be more likely to share the same sexual orientation than “fraternal” twins (who share about half their genes, like non-twin siblings). Numerous studies have verified that this is the case. The studies have come up with quantitative estimates of the heritability of homosexuality—simply put, the fraction of the total causation of homosexuality that is genetic. These estimates have been quite diverse but generally range from about 25% to 50%, and are usually higher in men than in women. In other words, genes are a substantial part of the story of what makes people gay, but other factors must also play a role. These other factors could be biological factors that are not genetic, or factors that are not biological at all.
There have been several attempts to actually identify “gay genes” or at least to pin down the approximate location of such genes in the human genome. So far, these studies have only come up with suggestive findings. A large study currently under way at Northwestern University may produce more definitive results.
Gay genes and evolution. The existence of genes that predispose their owners to homosexuality is counter-intuitive: Given that gay sex doesn’t produce babies, and that gay people have fewer children than straight people, one would imagine that gay genes would be quickly eliminated from the gene pool by natural selection. Several hypotheses have been put forward to explain how such genes might persist in the population:
· In one idea, gay genes persist because of kin selection: that is, although the genes make a person more likely to be gay, they also induce other traits that help the reproduction of that person’s relatives. For example, they might cause a gay person to act in a benevolent fashion to his or her siblings, helping them have more children that they otherwise would. As these children would have some likelihood of inheriting the same genes, this action would help the genes persist. Mathematically, it is difficult to see how this effect could be strong enough to compensate for the gay person’s own reduced reproductive success, but one study conducted in American Samoa found that gay people there do help their relatives reproduce. Whether the amount of help is sufficient to satisfy kin selection theory isn’t known.
· In another class of ideas, gay genes persist because not all the people who possess them are gay, and in those who are not gay the genes actually promote reproductive success. An example is the fertile female hypothesis, which proposes that the female relatives of gay men (who have some chance of inheriting the same genes that made the gay men gay) are more fertile—perhaps because they are more strongly sexually attracted to males than other women and therefore engage in more heterosexual sex.
· Another idea is that certain feminizing genes in men improve their reproductive success by making them more attractive mates for women. When a man inherits one or two of these genes he remains heterosexual, but if inherits several of such genes he is tipped all the way to homosexuality. In this model, it is the heterosexual male carriers of the genes whose reproductive success is promoted by the gay genes.
All these ideas are quite speculative but there is some evidence to support them: for example, some studies report that women who have gay male relatives do have more children, on average, that women who have no such relatives. Still, it’s unlikely that the reason for the persistence of gay genes will be fully understood until these genes have been identified and their mode of action elucidated.
Chapter 8: The Brain
The human brain contains several regions that differ in size, internal structure, connections, or function between men and women. One region of particular interest is the hypothalamus, particularly its front (anterior) region. Damage to this region in experimental animals impairs sexual performance or, in some cases, changes the animal’s preferred mating partners from one sex to the other. In most mammals there is at least one cell group (or “nucleus”) in this region that is larger, on average, in males than in females, and this size difference between the sexes results from the higher testosterone levels in males during a critical period of brain development. In humans, this nucleus is called INAH3 (3rd interstitial nucleus of the anterior hypothalamus). In an autopsy study published in 1991, I reported that INAH3 was smaller, on average, in gay men than in straight men. More recently, similar observations have been made in sheep, a species in which a small proportion of males are naturally homosexual.
In addition, several more recent studies have reported on differences between gay and straight men, and between lesbian and straight women, in other brain regions including the cerebral cortex. In most cases, these differences represent gender shifts: i.e., the size of a structure in gay men is more typical of that seen in women, and vice versa for lesbians. Although these other brain regions may not be involved in sexual behavior, they may be involved in other traits that are gender-shifted in gay people, as discussed in chapters 4 and 5.
Functional differences between the brains of gay and straight people have also been reported. For example, a Swedish group used PET scanning technology to visualize brain activity in the hypothalamus and other brain regions while subjects were breathing air spiked with substances suspected of being human pheromones. In these studies, activity patterns in the anterior hypothalamus of gay men was similar to that seen in heterosexual women, while the patterns seen in lesbians was more similar to that seen in heterosexual men.
In general, the brain findings support the idea that the sexual differentiation of the brain goes forward differently in fetuses that ultimately become gay men or women, as compared with heterosexual individuals of the same sex.
Chapter 9: The Body
As already mentioned in previous chapters, there are subtle differences in bodily anatomy between gay and straight people—in such measures as finger-length ratios and arm-to-height ratios. Other anatomical differences have been reported in various aspects of bodily symmetry, although the interpretation of these findings is difficult.
Perhaps more interesting are the collection of bodily cues that provide the basis for gaydar—the ability to distinguish gay and straight people on the basis of subtle aspects of appearance and behavior. In controlled studies, judges (who may be straight or gay men or women) are able to pick out gay people at well above chance levels, using still photographs, voice recordings, videos, or moving stick figures that show only walking style (gait). When judges make errors in these studies, it is usually in the direction of judging some gay people to be straight, not in judging straight people to be gay. Thus, it appears that not all gay people are recognizable by gaydar, but nearly all people picked out by gaydar are in fact gay.
In most studies of gaydar, the cues that are used to recognize gay people are gender-atypical—for example, a man who walks with a more female-typical gait, which incorporates greater hip sway—is judged to be gay. In general, then, the existence of gaydar is consistent with a general model in which gay people, on average, are shifted toward the other sex in a variety of physical and behavioral traits. As previously discussed in Chapter 4, children who later become gay adults are also recognizably gender-atypical in their unconscious behaviors, as was demonstrated by the study of childhood videos supplied by gay and straight men and women. Thus this general sex-atypicality seems to exist throughout the lifetime of gay people, consistent with the hypothesis that some broad organizing factor, such as the level of prenatal sex hormones, influences a broad package of gendered traits including sexual orientation.
Chapter 10: The Older-Brother Effect
A group of Canadian researchers (Ray Blanchard and a number of associates) have conducted numerous studies of birth order and its effect on sexual orientation. In women there doesn’t seem to be any particular connection. In men, on the other hand, the Canadians find that being later-born in a family increases the chances of being gay. This is largely or entirely caused by older brothers: the more older brothers a boy has, according to Blanchard’s group, the more likely that he will become a gay man. However, it would take an improbably large number of older brothers to give a boy even a 50:50 chance of being gay by the older-brother effect alone. Thus the older-brother effect probably works in concert with other predisposing factors. Other research groups have verified the Canadians’ findings to a fair extent, though the various groups’ findings are not completely consistent.
It would be easy to imagine that the older-brother effect is caused by rearing conditions: For example, parents might make efforts to guide their oldest son toward heterosexuality but allow their younger sons free rein to explore other options. This seems not to be the reason, however, based on one study of non-standard (e.g. adoptive) families. It seems that it is the biological fact of having been a woman’s second or later son that is the responsible factor, not the childhood or adolescent experience of having older brothers.
Blanchard’s group has proposed a biological model to explain the older-brother effect. In their model, women pregnant with a first male fetus may develop antibodies against some male-specific antigens possessed by that fetus. During a second or later pregnancy with a male fetus, those antibodies bind to those antigens and interfere with their function, thus blocking some aspect of male-typical development and causing that fetus to have an increased likelihood of homosexuality in adulthood. This idea is plausible but as yet there is no direct evidence to support it.
It should be borne in mind that most gay men in contemporary society don’t have any older brothers, so the older-brother effect is certainly not the main cause of homosexuality. Even so, understanding the basis of the older-brother effect could give us important insights into the biological mechanisms that contribute to the development of everyone’s sexual orientation.
Chapter 11: Conclusions
The biological studies reviewed in this book support the idea that sexual orientation is an aspect of gender that emerges from the prenatal sexual differentiation of the brain. Whether a person ends up gay or straight depends in large part on how this process of biological differentiation goes forward, with the lead actors being genes, sex hormones, and the brain systems that are influenced by them.
In the simplest model based on these ideas, high levels of circulating testosterone during a prenatal critical period drive brain development in such a way that the person becomes typically masculine in a certain constellation of gendered traits, including sexual attraction to females. If testosterone levels are low during the critical period, the person becomes typically feminine in those same traits, including sexual attraction to males. As for what might cause differences in testosterone levels among fetuses of the same sex, there are several possibilities: they could be due to genetic variations between individuals, to some unknown external factors, or to sheer random variability.
In a related, alternative hypothesis, there is no difference between “pre-gay” and “pre-straight” fetuses in the circulating levels of testosterone, but their brains respond to testosterone in different ways, due to genetic differences in the receptors or other molecules that translate the hormonal signals into neuronal architecture. Some evidence suggests that the first mechanism is more important in females and the second mechanism is more important in males.
The findings described in this book don’t exclude the possibility that environmental factors (before or after birth) might influence sexual orientation, but positive evidence for such factors is lacking at present.
Another unresolved issue is whether there are different kinds of homosexuality that have different developmental origins. There are some hints in this direction: for example, it’s been reported that lesbians who identify as “butch” have more sex-atypical finger-length ratios that those who identify as “femme”—which one might interpret to mean that there is a class of more masculine lesbians whose sexual orientation can be traced back to the fetal hormonal mechanism, and another class of more conventionally feminine lesbians who owe their sexual orientation to some other mechanism. Nevertheless, the limited evidence available at present doesn’t support a simple dichotomy between conventionally gendered gay people and gender-atypical gay people. Rather, most gay people seem to have some gender-atypical traits. Much more research needs to be done in this area, in order to find out what correlations there may be between different kinds of gender-atypicality in gay people.
It’s surprising that men’s sexual orientation is so categorical—either gay or straight, with relatively few men in between. Surprising, because the developmental factors that are thought to influence sexual orientation, such as prenatal hormone levels, probably don’t neatly divide into two groups (low levels and high levels of testosterone, for example). It therefore seems likely that there is some “channeling” or “canalizing” process that forces individual developmental trajectories into one of two alternative pathways, at least in males. One can speculate that this process is a reciprocal inhibitory circuit that is operative in the brain during development, and possibly into postnatal life also. This circuit could set up an unstable, winner-take all interaction that leads to an either-or outcome even when the input signals are distributed continuously. Inhibitory links within the hypothalamus are candidates to participate in such a circuit.
Most people link biological ideas about sexual orientation to the notion that the same fixed proportion of gay people have existed at all times and places. This is not necessarily the case, however. For example, if older brothers increase the likelihood that a boy will become gay (Chapter 10), then changes in the prevalence of older brothers will change the prevalence of male homosexuality. Two hundred years ago, most men had older brothers, but now (with the radical decrease in family size) most men do not, so there should have been a decrease in the prevalence of male homosexuality caused by the older-brother effect.
There are a number of other ways in family size and other factors could impinge on the biological processes of sexual development, causing differences in the prevalence and possibly the nature of homosexuality, and heterosexuality too. For example, it’s been speculated that genes that “feminize” men have become more prevalent since the agrarian revolution of 10,000 years ago, because such genes might make men more effective co-operators or more caring fathers; if so, it could be that gay men have increased in number as a by-product of selection for these feminizing genes. I point out these examples, not because they are actually known to have operated, but because they suggest a more nuanced view of how biological mechanisms may impact human lives over time.