I personally think the same thing is probably true of sexual orientation (which really is sexual attraction to gender not necessarily the same sex).

I would add that many anthropologists don't think people are born with a gender.

This doesn't sound plausible.

[tl;dr: If evolution didn't build in a powerful urge to make our gender behavior match our reproductive sex, then it made a huge error and missed a very easy and effective optimization.]

It may be the case that gender could be very loosely tied to biology in a philosophical or theoretical sense, but in the world that we're in right now, there are very strong adaptive reasons that gender expression and reproductive sex tend to stay close (statistically, of course) in any sexually dimorphic species, which I'd call a very strong "tie to biology".

Evolution 101-wise, gender can only be allowed to diverge from reproductive sex to a limited enough extent that it's more or less irrelevant to reproductive success. Evolution will make sure of that on a long enough time scale (at least up until the modern era, where we can to some extent decouple reproduction from sex).

If the two diverged commonly enough that animals were often foregoing sex with reproductively compatible partners in favor of incompatible ones that nevertheless matched the gender role they were interested in, then an adaptation that better facilitated reproductive matchings would easily emerge and dominate the population.

Note that such a compensating adaptation might even emerge as some form of social behavior, even if the impetus to that behavior was genetically driven; IMO, this doesn't make it any less tied to biology.

As an example, one suggestion [1] to solve the "gay problem" in evolution (why has full homosexuality, where a person is not at all attracted to members of the opposite sex, not been eradicated from the gene pool, since it should be so devastating to reproduction rates?) is that getting rid of the "gay gene" (or genes, or whatever) is actually a very difficult task for evolution to carry out (I'm anthropomorphizing evolution here for ease of speaking, not because I don't realize why that's wrong) for some reason. Difficult enough so that accepting the ~10% homosexuality rate was a better option, though obviously not ideal. So instead of getting rid of homosexuality, evolution tried to mitigate the "damage" that such behavior causes by enhancing an inclination for people to disapprove of it, which meant that even when people/animals were fully homosexual, they still tended to mate with members of the other sex due to social pressure. Thus the seemingly fitness-devastating 10% homosexuality rate was bumped down to a more ignorable number via social effects, and the presence of the gene was a net win. This is not to say that there's a "homophobia gene"; if this theory is correct, I'd guess that evolution more likely leveraged existing social behaviors (like wanting to fit in, or hating people that act differently) and turned them up to a slightly higher level.

In the case of gender identity, I suspect that there is a heavy dosage of social conformity involved in training people to signal their reproductive sex through gendered behavior. But I think it's biologically driven, or at least that it would be extremely surprising if it wasn't, since it's such low-hanging fruit. I'm sure that these biological imperatives are somewhat flexible, and that if pink was considered a boy color then boys would flock to the pink section of Toys R Us, rather than these things being hard-coded into the genome (though certain behaviors are definitely going to be hard-coded, since sex signaling had to take place before higher-level thought centers could be leveraged). But the inclination to figure out what these socially derived sex-signaling behaviors are is not a social construct - that's an evolutionary imperative, so while we may be able to change the particular expressions of gender that we see in the world, it's probably going to be rather difficult to prevent people from seeking them out and conforming to them.

This is why I'm always uncomfortable with nature vs. nurture questions - the environment that evolution optimizes any particular genome to succeed in includes the entire existing social structure, which was also influenced by previous rounds of evolution. So picking apart what is a "social construct" and what is "biological" is really a fool's errand, when it comes down to it - there's a delicate interplay between the two, and they always play off of each other.

[1] I should mention, there are other theories as well, the simplest being that even with the "gay gene", a person is only sometimes fully homosexual (twin studies have shown that homosexuality is definitely not 100% determined by DNA, though it's not 0%, either), so they do rather limited damage, and if tied to useful adaptations, there would be no particular evolutionary imperative to get rid of such a gene; the point, though, is that such arguments only hold up to a point, and if a large percentage of the population was gay, there would be much more selection pressure against that behavior, tamping down the ratio rather quickly to a lower level.

That's not how evolution works! Evolution works at the level of genes, and not individuals. It's easy to construct a model where a 10% gay population ends up being overall better for a population. Consider this made-up hypothesis: gay people are better at caregiving than non-gay people, so a population with gay people ends up with healthier adults who are able to have more, and healthier, children. For this scenario, gayness won't be "optimized" away because that leads to worse reproductive success for the population of genes involved. Nor is the presence of gay individuals "damage", because the result is an evolutionarily better population than one without gay individuals.

As another example, why does Down's syndrome exist? By your logic, shouldn't evolution have optimized that case away? That it hasn't means that changing how the 21st chromosome works is much harder than the impact of having a 1:733 failure rate. Why do you assume that any genetic component to being gay would be easy to change, without having negative consequences elsewhere in the population?

So your error is the belief that evolution emphasizes the reproductive success of individuals, when it deals instead with the reproductive success of genes. Some individuals don't need to reproduce so long as the overall gene population reproduces itself.

BTW, 100 years ago, pink was a boy's color, and young boys wore dresses too. Quoting from http://www.smithsonianmag.com/arts-culture/When-Did-Girls-St... "yet social convention of 1884, when FDR was photographed at age 2 1/2, dictated that boys wore dresses until age 6 or 7, also the time of their first haircut. Franklin’s outfit was considered gender-neutral."

I think this is actually a pretty reasonable statement, as long as it includes the caveat "on average" or "most of the time". Just as Down's syndrome isn't a huge problem for a population -- as long as it stays relatively uncommon.

A population comprised largely of people with Down's syndrome would likely be poorly adapted, and that's probably the case with a population comprised largely of gay people or transgender people as well. (Obviously, this is complete speculation, so I could be utterly mistaken.)

But yes, a population with a certain percentage of gay people could be better adapted just for having them, or alternatively, it could be better adapted because the same genetic diversity that leads a percentage of the population to be gay could be desirable in other ways.

Certainly. In the field of evolution, "on average", "statistically", or "most of the time" should be assumed to attach itself to almost every sentence (including this one).

All of this can be made much more precise, by the way, I just didn't mention it above because I already put up a huge wall of text. When it comes to deleterious mutations, there's a rule of thumb in evolution, which is to some extent mathematically provable: one mutation, one death. Statistically, what that means is that a single bad mutation will kill (where by "kill" I really mean "cause to not pass on one's genes to the next generation"), on average, one creature, no matter how bad the mutation is. If it's critical, then it will kill the first carrier before it's born; if it's not so critical, something like poor eyesight, then it will spread much further throughout the population before it kills (on average) one being.

This applies even in the face of mitigating factors. Taking the eyesight example, the fact that we have eyeglasses, and can correct poor vision, means that because poor eyesight kills less often than it did before eyeglasses the genes that cause it will spread much further throughout the population. The presence of the mitigating factor (eyeglasses) allows a potentially deadly gene to spread much further, so that on average it still kills one person per mutation.

So the fact that homosexuality has spread relatively far throughout the population either indicates that a) it is not a deleterious mutation overall (there's some significant benefit to the gene(s) that outweighs the lack of reproductive drive), b) that the mutation happens fairly often, so there are a lot of deaths due to it (this is the case with Down's syndrome), or c) that some damage-control mechanism exists so that the "death" rate is fairly low compared to the incidence of the gene.

In reality, it's probably some combination of all three possibilities; like I mentioned above, everything in evolution is statistical, so it never helps to look for single right answers.

Is the mutation which causes sickle-cell anaemia a "good mutation" or a "bad mutation"? It increases reproductive fitness in places where malaria is or was common, so it must be good, in an evolutionary sense.

How many deaths has it caused once the population of people carrying the haemoglobin gene mutation migrated to a location without malaria? Is that mutation now "good" or "bad"? How do you incorporate those numbers into your statistics?

Is the loss of eyesight a deleterious mutation? Definitely for a bird of prey, but not so for cave-dwelling creatures living in absolute darkness. For that matter, some people are attracted to people who wear glasses (and wearing zero-prescription glasses is such a turn-off!), so it might increase reproductive fitness.

Evolution doesn't know the future. If a population loses genetic resistance to a disease that's seemingly extinct, is that a "good" or "bad" mutation? How long does it take to judge that? After 1,000 years, should some thawed carcass reintroduce it and the species become extinct, does that count finally as a bad mutation and a single death?

For a real world example, consider the birds of New Zealand. They filled ecological niches which elsewhere were filled by mammals. Were these good mutations or bad ones? And when rats and weasels and cats and more were introduced to New Zealand, helping make many of those species extinct, then did those mutations retrospectively become deleterious?

If a genetic madness affects the leader of the US Strategic Air Command to issue orders which end up nuking a dozen Soviet cities, then what are the other cases which make that average out to one? If the nuking didn't occur, then what would the average have been?

What of a mutation which causes a speciation event? Is that a good mutation or a bad one? It's better for one environment and worse for the other.

There's a 10^-9 chance (1-in-a-million) that a "bad" mutation will mutate again back to the "good" form. With nearly 7 billion people in the world, that almost certainly happens a few thousand times every generation. In a generation we may be able to cure some genetic diseases through genetic engineering, so a "bad" mutation can be fixed.

With all those in mind, I can't figure out a way to get the numbers to come out "1" unless the definition of deleterious is defined to make it come out that way.

It's rather simple to prove in the simplified case, it's just a typical steady state assumption. If a population is in an equilibrium state, then the rate at which any mutation is introduced has to be equal to the rate at which it is removed from the population. So if one mutation has a 1% chance to kill its owner each generation, then to maintain equilibrium (in other words, to make sure the prevalence of the mutated gene in the population is stable), every time the mutation shows up anew, it must spread to 100 people, killing one of them. One mutation, one death.

Yes, that's super simplified, it neglects the possibility of multiple mutations, positive or neutral ones, back-mutation, interactions between members of the population, non-equilibrium states, etc. These will change the details of the math, sometimes quite substantially.

But the basic idea, that the worse a mutation is the less prevalent it will be, should hold.

The motivation for ignoring beneficial mutations (and back-mutations to beneficial states) is that they're extremely rare as compared to deleterious ones - most selection pressure in nature is aimed at merely preserving the functionality in the genome, weeding out new deleterious mutations rather than supporting new beneficial ones (though that is a critical role in the very long term, of course).

In my example earlier regarding extinct bird species on New Zealand, were there ever any beneficial mutations? After all, the genes no longer reproduce.

My point is that there is no stable state, so it's better to have a shorter-term definition of "beneficial" and "deleterious" based on relative reproduction fitness compared to others in the species population over a short time frame.

Additionally, beneficial and neutral mutations do not always spread to 100% of the species population. A Y-linked trait won't spread without a male lineage.

Yup, that's a pretty good description of the assumption/definition, for all the good and bad things it brings with it; I mostly agree with the rest of what you've said.

FWIW, I've mainly seen the one-mutation-one-death rule applied to arguments that attempted to place informational speed or capacity limits on the process of evolution, and in most cases it has turned out that these arguments fail when applied to the real world because of precisely the types of arguments that you've made against this rule (sexual recombination and the fact that evolution is never in a steady state tend to be the biggest problems).

An aside on the topic of defining beneficial/deleterious mutations: the problem is a very difficult one, because in reality the effect of a mutation is at best distributional and not measurable along a single axis of goodness/badness. I don't know much about the state of the art here, but if I was going to sit down and try to figure out a way to try to estimate "benificial-ness" of a mutation (even as a distribution), I'd say you've put your finger on exactly the right question to focus on, that there's a tension between what might be immediately beneficial and what would be beneficial in the longer term. For instance, a gene that made someone have babies like crazy by always sacrificing one's grandchildren for nourishment would be fantastic for 1-generation-out fitness, but terrible for 2-generations-out. Similarly, in an environment where pre-reproductive death was very common, an adaptation that decreased the likelihood to have children by a small factor but significantly increased the ability to keep them alive would be very beneficial at 2-generations but deleterious at 1. So pure-local effects measured 1 generation down the line won't necessarily tell us enough (though in the vast majority of cases, it's probably good enough).

On the other hand, a purely-global view is not right either, because as you've mentioned, environmental changes or genetic shifts within the species can turn previously helpful traits into "bad" ones. So measuring regret after the fact will not suffice, and further, it misses the fact that at the moment the mutation happens, there is some distribution that describes the likely outcomes of that mutation given the current state of the world, using no information from the future (even if we don't know it or can't feasibly calculate it). The extinct bird species definitely had beneficial mutations even though some freak event wiped them out later, and we need to account for that. So something more subtle is required.

In order to do better we'd probably have to make some assumptions to eventually cut off the familial dependencies, like perhaps that the presence or absence of a gene in an N-th grandparent can have no direct causal effect on the survival of the animal in question (i.e. direct nurture effects are limited in time - even this assumption is questionable in the face of things like family wealth). We'd also have to assume that we could quantify the expected changes in the environment into some sort of distribution, including distributional assumptions for expected changes in the rest of the population (luckily these changes should be rather slow, when measured in generations). Then we could in theory compare the expected size of a family tree branching from a member that possesses a new mutation after that N-th generation, to the base case, the family tree without that mutation. Of course, the "expected size" is a trivialization of the real distribution, which would better describe the possible effects of a mutation (and the details of that distribution might strongly effect the population dynamics).

It quickly gets messy, and requires a lot of assumptions. And even after all that, we wouldn't have a very clean number to work with, i.e. we couldn't easily continuous-ize the situation and write down an ODE that took the "beneficial-ness" of a mutation and showed us what would happen, because other details of the distribution would be important, as well.

Yes, of course, though in many cases genes achieve their own survival by boosting the survival and reproduction rates of their hosts.

It's easy to construct a model where a 10% gay population ends up being overall better for a population. Consider this made-up hypothesis: gay people are better at caregiving than non-gay people, so a population with gay people ends up with healthier adults who are able to have more, and healthier, children.

You're invoking group selection here, which is exactly what The Selfish Gene debunked in great detail; given your comment above, I'm surprised that you would make this argument.

From the point of view of the gene, in a society that contained a 10% gay population who were better at caregiving, a gene that selfishly reduced the probability of its host's homosexuality would thrive, because not only would its carriers benefit from the caregiving boost thanks to the other members of society without that gene, they would not suffer from the reduced reproductive potential. Only in the long term, as the gene spread throughout the population, would the caregiving benefits start to fade, and that's not a present-enough change in fitness to apply any evolutionary pressure against the gene (more precisely, it can't apply evolutionary pressure because it depends on the prevalence of the gene in other members of the population; it's a classic prisoner's dilemma situation, and if you're going to take one lesson from Dawkins, it's that evolution always chooses to defect).

As another example, why does Down's syndrome exist? By your logic, shouldn't evolution have optimized that case away? That it hasn't means that changing how the 21st chromosome works is much harder than the impact of having a 1:733 failure rate.

Down's syndrome would be exceedingly difficult to optimize away, because it falls into the category of commonly-reproduced-mutation; it is not the result of code that specifically causes Down's syndrome, it's the result of our genetic material being evolutionarily close to a state that results in Down's syndrome, so whenever something goes wrong, the maladaptive trait is rediscovered over and over. Same thing with most other chromosomal disorders (most of which end up filtered out very quickly, well before birth).

FWIW, that's another common theory about how homosexuality has survived, that normal people are "one mutation away" from being gay (or rather, of having the mutation that makes them potentially gay). Both of these cases still presume, however, that the negative consequences of the trait, when combined with the probability of the trait manifesting, are negligible enough compared to the genetic changes that would be required to move us more than "one mutation away".

Why do you assume that any genetic component to being gay would be easy to change, without having negative consequences elsewhere in the population?

I quite explicitly assumed exactly the opposite. My whole comment on that matter was predicated on the assumption that it is not easy to change susceptibility to homosexuality, and that social mitigation was a workaround.

The main reason I brought up homosexuality at all was that it is often pointed to as a counterexample to the idea that reproductively negative traits are weeded out of the gene pool; I wanted to make the point that evolution doesn't necessarily need to weed out such traits directly as long as it can find some way to control their side effects.

BTW, 100 years ago, pink was a boy's color, and young boys wore dresses too.

Yup, that doesn't surprise me. I absolutely believe that much, if not most, of what signals male/female in today's society is arbitrary. However, I think that the existence of some set of traits that each sex uses to signal reproductive class is very much innate.

It's kin selection, not group selection. Consider Dawkins' "Twelve Misunderstandings of Kin Selection" wherein he writes:

"To stick my neck out a little, it seems to me that, far from genes for altruistic behaviour being implausible, it may even be that a majority of behavioural mutations will turn out to be properly describable as either altruistic or selfish." ... "A gene for altruism, then, is any gene that, compared with its alleles, causes individuals to benefit other individuals at a cost to themselves." ... "But the kind of mutation that could lead to such altruistic restraint could be ludicrously simple. A genetic propensity to bad teeth might slow down the rate at which an individual could chew at the meat. The gene for bad teeth would be, in the full sense of the technical term, a gene for altruism, and it might indeed be favoured by kin selection."

The example I gave seems perfectly aligned with this definition of altruism and kin selection. Indeed, it's a weaker but analogous form of what leads to eusociality. You say "it can't apply evolutionary pressure because it depends on the prevalence of the gene in other members of the population", .... and I think I understand why we disagree. I wrote "population" but sometimes meant "species population" and at other times meant "gene population."

In an extreme hypothetical case, suppose that having a gay sibling help to raise a family meant a 5% improved chance that each child would live to adulthood and children in turn. Suppose also that having two gay siblings meant a -1% improved chance (perhaps because the person consumes more food, which could otherwise go to the children). Then there's strong kin selection here to have some, but not all, gay children. The descendants then become a larger part of the species population.

In this case, I don't see how homosexuality would be a "reproductively negative trait" for the gene, only for some of the individuals carrying the gene.

Personally, my suspicion is that homosexuality is more directly linked to a positive physical trait in the individual, though I don't have much to really back that up other than a vague sense that kin selection effects in evolution are rarely as strong as direct expressed ones. But yes, the "gay uncle" effect could explain it, too, and it's definitely an interesting enough phenomenon to be worth keeping in mind.

[1] In fact, I probably shouldn't react as negatively as I do against most invocations of the group selection argument, because oftentimes the points would be valid if expressed as kin selection arguments instead.

In theory, agreements should be more common than they are (sadly) in practice: http://www.scottaaronson.com/papers/agree-econ.pdf

What I've always wondered is which particular assumption of Aumann's agreement theorem is usually lacking on the Internet: honesty, rationality, common priors, or simply the willingness to continue the conversation long enough to resolve the disagreement.

BTW, in this thread I learned that I need to be more careful about how I use the term "population." :)

The problem here is that culture determines gender. Some cultures have two, some three, some even have more. Many cultures treat children as genderless and initiate them into genders in rites of passage. In some cases gender is contextual, so among the Norse and among the Greeks, there was a specific gender-based stigma attached to being the penetrated partner in male-male sex. Male-male sex was not stigmatized, only crossing the gender line and being penetrated as a woman.

I think it is a grave error to look at one's own culture and assume that it is biologically determined.

As for the "gay problem" we have to recognize at some point that every culture addresses human sexuality differently, and human sexuality is remarkably malleable. For example there are tribes in Papua which make young boys give oral sex to tribal elders as a part of a rite of passage as a way of them literally ingesting manliness in order to become men.

Once we look at our own sexual taboos involving who and what we are forbidden to have sexual relations with, and we recognize that these are socially contextual, not innate taboos, things change a great deal.

Absolutely. I never claimed otherwise.

All I think is that some sort of robust signaling mechanism that displays a person's (or animal's) reproductive "team" should be expected to exist in any sexually dimorphic species. In many animals, this is hard coded, but I suspect that in humans that was generalized to a high level imperative, "figure out what sex you are, and clearly display the appropriate characteristics so that mates can find you".

Once we look at our own sexual taboos involving who and what we are forbidden to have sexual relations with, and we recognize that these are socially contextual, not innate taboos, things change a great deal.

But this is exactly my point: the fact that these social sexual taboos so often go against behaviors that reduce evolutionary fitness suggests that supporting those social behaviors may, in fact, be precisely the way that evolution ended up most easily controlling those behaviors.

To be very clear about this: the fact that behavior is influenced socially rather than genetically does not necessarily mean that it's an accident of history. It very well could be a direct evolutionary adaptation that leaned on social behavior to implement itself. Nature tunes nurture, and nurture tunes nature, so arguing for one to the exclusion of the other is usually wrong.

That doesn't, of course, mean that we shouldn't try to overcome such evolutionary imperatives. But we should be aware of the fact that in such cases, the social behaviors are not completely arbitrary, and that we have an uphill battle to fight.

With humans though the linking of physical sex and gender is not as simple as you suggest. As I have said, some cultures (like ours) have two genders. Some have three, with children being genderless, and some have more genders than three. To pretend that gender is only about display of sex-based characteristics is to gloss over the fact that in most cultures it doesn't really work that way, nor does it really even in our own.

Gender is instead a social category and a social position. It affects division of labor and all sorts of other things. Different genders often have different taboos and these are often aimed at preventing gender-crossing, as well as maintaining a symbolic order between genders.

This is a very broad category of anthropology, and it's dangerous to assume that everyone structures their society around two genders fairly closely tied to biological sex, since this is not really the case.