As anyone who reads Andrew Sullivan knows, John Derbyshire is probably the most (what Andrew (and other gays) call) "homophobic" writer at National Review.

Apparently, in response to a post yesterday about the genetic origins of homosexuality, he got an email from a (vociferously non-conservative) supposed expert in the field, who wrote:

...if it were a genetic disease defense, it would have a certain very simple and identifiable inheritance pattern, and it certainly does not have that pattern. Identical twins would both have it, but the chance that a homosexual man's identical twin is also homosexual is only about 20%.

I have a theory about the genetic basis of human sexual orientation, that I never hear anyone discuss, but to me makes perfect sense, and fits the facts (including the one quoted above, assuming that it is indeed a fact). I discuss it here, and in comments to this post.

Simply put, some are born homosexual, some (probably more) are born bisexual, and most are born heterosexual. For the first and third groups, their sexuality is indeed thrust upon them. Errrr...so to speak.

If that's true, then the twin studies might actually provide some insight into the relative genetic component, assuming they're separated twins. If two separated twins (the cite above doesn't indicate whether they were separated or not) turn out to both be homosexual, that to me is a good indicator of homosexuality with either a genetic basis, or uterine environmental basis, or both. Of course, the only way to truly determine the genetic basis may be to allow human cloning...

In any event, it seems to me that my theory would dictate that the cases in which one twin is homosexual, and the other not, are cases in which both were born bisexual, and for various post partum environmental reasons, made different choices as to partners. It also explains why some homosexuals can be "cured," and others can't. The ones who can be were never homosexual in the first place--they always had a choice and simply decided to start choosing differently.

Long story short, this "expert's" provision of this "fact" (and I'm pretty sure I've seen different numbers) has little relevance to the debate, unless you really do have the simplistic viewpoint that all people are either purely homosexual, or heterosexual, with nothing in between.

The book Sperm Wars has some discussion of the genetics of homosexuality, including how homosexual behavior can provide greater reproductive success. The gist is that the extra experience provides an edge in seducing/deceiving hetero partners.

I can see that making sense to explain bisexuality, but how would a homosexual ever have success with a hetero partner, regardless of how much "extra experience" he/she had?

I think that there are probably genetic explanations for homosexuality that are completely independent of the sexual orientation of the phenotype (for instance, it might be a result of the combination of two or more genes which, when occuring separately, convey an advantage but that when combined express themselves as homosexuality).

The original assertion is questionable at best. Bailey and Pillard, comparing homosexual males to their siblings found that 52% of maternal (identical) twins were both homosexual, 22% of fraternal twins were both homosexual, 11% of adoptive brothers were homosexual, but, oddly, only 9.2% of non-twin biologic siblings were homosexual. In a further study on lesbians they found that 48% of maternal twins, 16% of fraternal twins, 6% of adoptive sisters, and 10% of non-twin biologic sisters were both homosexual.

Myria

Thanks, Myria, I've put your URLs into links (HTML is allowed here, as long as people play nice--I guess I should point that out in my comment template). Long URLs mess up my page formatting.

In John's case -- and I consider John a friend, albeit a slightly loopy one -- it goes farther, as he appears to consider (on the basis of etymology) that all gay males are inherently child-abusers.

(His argument: "pederasty", which is rooted in the Greek "paed-", "young", is also used as a synonym for anal sex. Ergo, anyone who has anal sex is a potential child molester.

Honestly, he has many good and worthy qualities... but he's a little eccentric.)

The genetic disadvantage of inherited homosexuality is enormous; it's like being
born with a fatal disease. We have a lot of experience with inherited fatal
diseases because there are a lot of them out there, but with a few interesting
exceptions they are all extremely rare.

An exception for example would be the gene for sickle cell anemia which in
some populations is quite common. It's fatal for a carrier if the person
gets two copies and it's not particularly good even if a person has only
one. Unless malaria is endemic, in which case the sickle cell anemia gene
carrier has an enormous advantage.

And that's the point, in every case where something as disadvantageous as
homosexuality is not extremely rare there's some common situation where
it turns out to be extremely advantageous. There's nothing subtle about
the advantage. It couldn't possibly be subtle, it has to be extreme to
overcome its extraordinary disadvantage.

Something like "The gist is that the extra experience provides an edge in
seducing/deceiving hetero partners," for example, seems way to subtle to
explain homosexuality. Given that it has to be a pretty dramatic advantage
it's hard to understand if there is such a thing why we haven't spotted
it.

The simplest explanation is that there is no advantage because except in
extremely rare circumstances there is no such thing as inherited homosexuality.

But if that's the case then why is their homosexuality (over and beyond the
extremely rare)? The only explanation I've been able to figure is that
the gene or genes favoring homosexuality are not single purpose. That is
they don't just code for homosexuality, and in fact that's just a side-effect
so in truth they code for something else. It's the something else that's
extremely advantageous. This is in other words a variation on the theme
already stated except this time we don't know what the linked trait is
and therefore are missing the obvious advantage.

But if this is true shouldn't we be able to use statistics to find the linked
trait given it's so advantageous? Well maybe. It's possible no one's
really done a thorough job of looking. It's also possible that the linked
trait hasn't been recognized as an entity. If it's not a 'something'
then there's no way for statistics to spot it.