57 Comments

Embedded into this is the assumption that there are genes for being gay similar to sickle cell genes. I think the fact that some people end up gay is likely a consequence of the fact that there's a very complicated dance of sexualizing hormones that have to trigger at just the right times in order to successfully influence these behaviors and preferences during fetal development as well as later in life.

It's easy to imagine that there's a real tradeoff: in order to have robust sexual drive in most of the population, there's a chance that things go "awry" (from the perspective of the genes, not morally) and you end up with a percentage of the population attracted to the opposite sex. You could of course imagine some completely new system for generating sexual drive that wouldn't have this problem, but evolution doesn't allow you to make big jumps like that. Small adjustments to the system for sexual drive that would move away from a population of gay people will cause overall lower fertility.

In fact this absolutely must be true since we see homosexual attraction in many species across the whole animal kingdom. Those strong selection pressures have been there all along. If this were the kind of thing that could be optimized away, it would already have been.

A good analogy here might be why men form useless nipples and breast tissue. Here the small cost of making those nipples (including a chance for cancer) is dwarfed by the advantage provided by having robust breast and nipple-making genes to pass on to their female offspring.

Expand full comment

I've read :

"Occasional same-sex sexual behavior in animals is no model of exclusive homosexual orientation, which is uniquely human. https://rd.springer.com/article/10.1007/s10508-020-01708-9"

https://x.com/DegenRolf/status/1254050714134986752?s=20

Expand full comment

Also, it depends hugely on exactly how these traits are heritable. Let's take a model where there are 20 genes each of which increases your homosexuality a bit but anyone with less than 18 of them is straight. However, having more of those genes but less than 18 gives you a greater ability to understand what the opposite sex is likely to desire (I'm not saying this is the most likely explanation just a toy model to prove a point) and thus a reproductive advantage.

Since each of those genes undergoes positive selection in something like 95% of the population and negative selection in 5% of the population that could easily work out to net selective pressure for more such genes.

You can't just waive away the details of how the condition is heritable.

Expand full comment

You need to consider that people had more kids in the past. So if there is a gene that makes your kids more fecund, but increases the chances that one of them is gay, if you have 10 kids and 1 is gay you'll make up the difference, if you have 3 and one is gay, you won't. So the environment which now includes things like the pill and condoms will mediate this effect. If you had a full sample this wouldn't matter. But your sample doesn't include families who have the gay genes, but didn't have a gay child, you are excluding them from the set, so you the effect looks negative. In the past most people would have had enough kids to get a gay one and the math might have worked out.

Either way, most case of homosexuality, are not due to genes, but hormone exposure inutero. In the future with ectogenesis, that will be regulated so no one has to have a child with congentially low fertiality, which is clearly a birth defect.

Expand full comment

The number of children that survived to adulthood has been very close to 2 for nearly all of human history.

Expand full comment
Nov 7, 2023·edited Nov 7, 2023

It's not that heritable: most homosexual males with an identical twin are discordant. And you left out the rival explanation from Greg Cochran, the only one I find plausible:

https://en.wikipedia.org/wiki/Gregory_Cochran#Pathogenic_hypothesis_of_homosexuality

Remember that even small difference in fitness will add up over time. The expected frequency of an allele that reliably causes lower fitness should be at the rate of de novo mutations.

Expand full comment

From the article you linked:

"*Pathogenic hypothesis of homosexuality*

Cochran has argued that male homosexuality may be considered a disease because it generally reduces or eliminates reproductive output, and he and Ewald have speculated that homosexuality might be caused by infection with an unknown virus.[10][11][12] However, he does not suggest that the infectious agent that causes homosexuality is spread by homosexuals. Cochran's hypothesis is based on the argument that homosexuality is unlikely to be genetic because it does not follow simple Mendelian inheritance patterns and because natural selection should have largely eliminated genes that cause homosexuality.[10] *Cochran says that there is no positive evidence for the gay germ hypothesis.[11]* In 1999, journalist Caleb Crain published an article in the gay magazine Out in which he spoke with several sexual orientation researchers about the hypothesis.[13] Geneticist Dean Hamer called it an "interesting idea" which would need to be tested by experimentation, but that *he was skeptical as homosexuality doesn't appear in clusters.* J. Michael Bailey disagreed with the use of the term "disease", but gave Cochran the "benefit of the doubt". Elaine F. Walker, who researched a pathogenic cause of schizophrenia during pregnancy, *did not find it plausible.[13]*

*The dominant hypothesis in the scientific literature is that male homosexuality may be a result of incomplete masculinisation of the brain during fetal development, under the influence of sex hormones.[14]* Maternal immune responses have also been implicated.[15]"

How do you answer the criticism that a pathogenic cause would tend to cause the trait to appear in clusters, which apparently homosexuality does not? (I think this objection assumes that the "disease" must be communicable between adults who have it, but if it's only effective prenatally, like many teratogens, then I suppose we wouldn't expect that.)

In any case, "the dominant hypothesis in the scientific literature is that male homosexuality may be a result of incomplete masculinisation of the brain during fetal development, under the influence of sex hormones" strikes me as a hilariously inadequate "theory" of homosexuality. Homosexuality exists because fetal brains are not caused to be not gay ("are not completely masculinized")? No shit? How is that an explanation?

Expand full comment

The polio virus spreads via the digestive tract, and can't spread if it gets in the brain (and thereby causes paralysis). So if we just look at the people with paralysis, we won't see all the other people who were spreading the virus.

While some people claim that the distribution of homosexuality is the same everywhere, that's not true. Within the US, the GSS shows it to be more common among males raised in urban vs rural environments, and its absent from hunter-gatherers like these https://www.theatlantic.com/health/archive/2012/12/where-masturbation-and-homosexuality-do-not-exist/265849/

Expand full comment

The rise in gay individuals getting surrogates or IVF could plausibly have the opposite effect and don't forget that immigrants from conservative cultures and religious groups contribute an increasing percent of total reproduction (people often leave those groups but they constitute a genetic pool that keeps refreshing the overall gene pool).

Overall, it seems plausible we see a substantial decrease in the rate of homosexuality (tho if it happens whether that's a 2 generation thing or 200 generation thing depends hugely in mechanism) but it's not at all certain and it's far more complex than you suggest here.

I think that's really a shamr because as much as I appreciate not having to live up to the standards gay men set for each other I think it's generally easier and more pleasant to find a mate of the same sex for many people (don't get me wrong happily married now but just an observation).

Expand full comment

Well don't you worry, I'll find a way to have lots of kids and infuse the next generation with just a bit more (respectable) gayness. :)

Expand full comment

I think there has to be *some* benefit to "gay genes" as a whole, because while I accept your view that being gay probably has a much larger negative effect on fertility now (certainly for women, anyway) than it would have done in the EEA, it seems that it must always have had *some* negative effect on reproductive success (if it didn't one would wonder why we have sex drives at all).

It is very likely than lesbianism wasn't that detrimental to reproductive success in the EEA, since women probably didn't get a lot of choice in whether to mate. It is less obvious that this is also true for male homosexuality.

Expand full comment

There is also the issue of kin selection. Many women/couples make choices about whether to reproduce based on availability of resources, help with the children etc etc. The existence of a wealthy childless gay brother is exactly the kind of thing which might make someone decide to have more children and make those children more successful.

How that works out in the long run depends greatly on what the distribution of child numbers look like. If even a small fraction of men with reasonable success behave like Johnson or Musk then you can trade some immediate increase in offspring for greater chance to become a successful super reproducer.

It's alot harder to figure out these effects than you suggest. A great deal depends on indirect effects.

Expand full comment
Nov 7, 2023·edited Nov 8, 2023

One paper on sexual orientation and fertility intention among young people in a non-US country. Similar big gaps, also similar very big LGB share.

https://link.springer.com/article/10.1007/s11113-023-09773-3

Expand full comment

If this were true, homosexuality would've died out long ago. Even if historically social pressure led to some degree of heterosexual reproduction by gay people, it'd clearly still be a very disadvantageous trait to be averse from heterosexual sex.

So there are two other explanations you miss: 1) sexual orientation is polygenic, and having a few 'gay genes' - but not too many - as advantageous. In this scenario, each individual locus associated with homosexuality may on average be neutral relative to the average population fitness, and thus be in equilibrium, but only in a small fraction of the population would 'too many' 'gay alleles' occur by chance, causing deleterious aversion to heterosexual sex. I think this is similar to how some individual genes associated with autism - when in conjunction with enough other autism-related genes (autism is also a polygenic trait) - are associated with intelligence on their own.

The second explanation is that there are environmental-genetic interaction effects; the 'gay genes' don't make everyone with them gay, just a significant minority. One theory for why this would be advantageous in a social species is as a sort of population control mechanism, where it may be beneficial under some circumstances to produce non-reproducing members who can contribute to their genes' proliferation by helping their relatives's children survive. Obviously in extreme cases, like many insects, producing many non-reproducing offspring is the norm.

Expand full comment

Selection still applies to polygenic traits. The one caveat is that the rate of de novo mutations would apply to each allele that can potentially cause a trait.

> The second explanation is that there are environmental-genetic interaction effects; the 'gay genes' don't make everyone with them gay, just a significant minority.

We know for a fact based on identical twins that genes don't reliably make them gay.

Expand full comment

But the fitness curve for a polygenic trait isn’t necessarily monotonic. E.g., it may generally be good to be tall, but not too tall, so each tall allele is individually under positive selection, but occasionally two people with too many tall alleles between them mate and produce disadvantageously tall offspring. For polygenic traits it’s much more likely to get recurring disadvantageous phenotypes than with simple Mendelian genetics.

Expand full comment

Height is normally distributed. Sexual orientation is not. Similarly, it's not like there's a normal distribution in preferences for eating food vs gravel.

A normal distribution happens when you have a large number of genes of small effect, and we know that genes tend to evolve that way specifically to avoid big potentially harmful swings in phenotype. But, again, sexuality isn't like that.

Expand full comment

The distribution doesn't have to be normal for this kind of selection pattern to happen. Many poylgenic genetic diseases (see autism for example) are caused largely by mutations at loci that also occur among healthy people. So, in the autism example, variation at many associated-sites correspond to normal variation in other traits (like intelligence), but in combination, some set of mutations (or many different permutations of some as-of-yet-undiscovered set of mutations) combined causes an extreme phenotype that is deleterious, like autism. There's a Scott Alexander post on the autism/intelligence example (https://slatestarcodex.com/2019/11/13/autism-and-intelligence-much-more-than-you-wanted-to-know/).

Expand full comment

Given we have observed homosexual behaviors in non-human wild animals, where human-intermediated breeding can't account for continued propagation, it seems unlikely that non-standard sexuality will entirely disappear.

Aside from that, one argument I've encountered is that heterosexuality is a relatively recent evolutionary phenomenon (compared to sexual reproduction itself), which if you think about it makes sense - you have to have sexual reproduction before sexual dimorphism can evolve. The proper default, then, might be regarded as bisexuality, as opposed to heterosexuality; heterosexuality requires a lot of additional genetic machinery to achieve.

You may notice that homosexuality, then, is likely a side-effect of heterosexuality - strong heterosexual tendencies requiring both sexual dimorphisms, and complex attractiveness-filters for those dimorphisms - these same filters, expressed differently, result in homosexuality. In this view, insofar as heterosexuality is selected for, homosexuality may also be selected for, at the expense of bisexuality. This is likely the place you'd need to look for evolutionary advantage in homosexuality - if it is a byproduct of heterosexuality, the advantage heterosexuality offers is over -bisexuality-, rather than homosexuality.

(There may be a caveat here - homosexuality may have offered some kind of high-risk, high-reward payoff for men specifically in early human society, through warfare. Not just in the obvious way, either; there are examples in nature of homosexual pairings of animals in which they form an effective partnership to defend territories and offspring; black swans I think are the standard example there. This may better be regarded as an adaptive form of bisexuality, however.)

From this perspective, heterosexuality is an informationally-expensive trait, rather than the default; selection is not just the process of improving a species, it is also the process of maintaining information against entropic decay, using a finite amount of informational energy. (Each "selection event" adds one bit of information to the gene pool, in a certain sense; entropy, in various forms, is constantly "deleting" bits.) In that case, it may be too informationally expensive to maintain perfect heterosexuality; there's a lot of information needed to be encoded in order to achieve heterosexuality.

And in humans, the amount of information you'd need to encode to achieve heterosexuality is enormous - not just layers of filters for sexual dimorphism, but machinery to account for social behaviors that evolve more rapidly than genes themselves. And there are more important things to apply selective pressure towards - it would be more effective to ramp up "desire to have children".

Expand full comment

The real puzzle is obligate homosexuality, which does not seem common in other species of animals. The one species with a comparable frequency is sheep.

Expand full comment

Domestic sheep, or wild sheep?

How common in animals is obligate heterosexuality?

If homosexuality is modeled as a side-effect of heterosexuality, as opposed to a distinct phenomenon, then ramping up heterosexuality could also ramp up homosexuality; if the behavior is seen only in domesticated sheep, or wild populations descended from domesticated populations, I'd hazard a guess that human selection pressure against homosexual behavior (killing rams that engaged in homosexual behavior) would result in a selection bias against -bisexual- rams, paradoxically increasing obligate homosexuality.

Expand full comment

I don't know about wild sheep, I didn't think there were many still wild.

Humans have domesticated multiple species of animal. Sheep are unusual in the high frequency of obligate homosexuality, so there seems to be something about them rather than something about domestication.

Expand full comment

What exactly does obligate homosexuality mean? In earlier ages when homosexuals couldn't live as a couple and adopt many people who are homosexual and very much not bisexual would lay back and think of England (or more likely someone of their gender). A huge fraction of gay men I know engaged in some kind of sexual contact with a woman before figuring shit out.

So even in humans there is very little obligate homosexuality in a strong sense. I'd guess there are a lot of gay men in religious communities who are married with kids.

Expand full comment

These are male sheep that won't mate with any female sheep, but instead try to mount other males.

A woman is supposed to "lay back", a man is supposed to take an active role. Males that don't pursue heterosexual intercourse are going to be outbred by males who do.

Among women, bisexuals outnumber lesbians. Among men, gays outnumber bisexuals. Distribution of sex partners is bimodal, with one peak of course being much larger than the other, but the middle not being highest like in a normal distribution.

Expand full comment

I recall reading some sort of study like this, where researchers found that bisexuality had an advantage over heterosexuality in some animals because the risk of false negatives exceeded that of false positives. In other words, a bisexual animal has the risk of having sex with another animal of the same sex, possibly missing out on a heterosexual mating opportunity, but a heterosexual animal also has a risk of missing out, it might be repulsed by an androgynous-looking member of the opposite sex and miss out on a chance to mate with them. Overall, for many animals, heterosexuality might make them too picky and end up being selected against.

Expand full comment

I'm thinking that reproduction is going to change during this century. It already has to some extent, as LGB people use sperm banks, surrogates, etc. But we will soon not require a womb to create a child, and IVF takes care of the other part. No idea what social norms will develop around all of that or how that will impact fertility rates.

Expand full comment

I suspect that some of homosexualitys historical ways of increasing fertility was alloparrenting, which was a lot more relevant when we were hunter gatherers and before we became farmers. Farmer society was increadibly pressuring on having lots of kids and people knowing their place, and women specialized in being the caretakers of the family, leaving little use of alloparrenting outside of grandparents. Farmer society and especially industrial society is pretty new historiclly speaking, so it shouldnt be enough time to affect the genes heavilly.

In LGBTQ historic fiction, “married officially with a partner of opposite sex to fullfill societal expectations and having a kids, but i have my actuall gay love interest outside of the marriage and im deppressed/a wreck” is a pretty reoccuring trope

There also seems to be some amount of “im probably just unhappy or failing in some way” without considering the possibility that the person is gay.

Not everyone just knows that they ar X sexuality, and need real life experiences to realize what they want.

Pre LGBTQ acceptance society, theres likely some amount of people that never even realized they were not cis or straight

Expand full comment

Is there any data on whether political affiliation affects how many kids homosexuals have?

Otherwise, the “immense social pressure” theory seems pretty accurate, as you usually only need to have sex once for each kid to be born.

If you had between 3-6 kids historically and you were a closeted gay, then thats maybe 9 times of sex for that.

Im gay but only really realized it 2 years ago, but i had sex with a woman 2-3 times (with condoms).

I found the whole affair extremely uncomfortable but just thought it was normal cause it was new, and she also enjoyed it a great deal, so i just pushed through.

Pre-condoms, that could have been 2 babies. And thats just from personal ignorance, and super mild social pressures.

Having society persecute me or my family talk constantly about having babies could probably add a few more.

I cant give source for this, but there somes to be some ammount of marriages that are just formal ways of alliances, where the man and woman has no interest at all in sex and view it as a tedious affair.

So thats a mild evidence point for me

In case anyone wonders, i would like to get married and have 2-3 kids. Might adopt them or have some form of biological procedure for it.

Expand full comment

Why can’t there be new mutations? That’s how achondroplasia works. It’s genetic, autosomal dominant even, but people with the condition rarely reproduce. The reason we keep seeing it is new mutations.

Expand full comment

Gay genes have already been purged. Concordance for homosexuality in identical twins is low, much lower than you would expect to find for a random trait.

You lose a lot of information if the only thing you're willing to know is "sexual orientation is highly heritable". There's no particular reason to believe that "sexual orientation" is one thing about which it makes sense to make general statements.

For the classic example, "number of arms" is highly heritable, but while people who have two arms mostly have that many arms for genetic reasons, people who have one arm mostly do not have that many arms for genetic reasons. The problem comes into your reasoning when you decide that "number of arms" is one variable.

Expand full comment