Autonemesis

Easy: not all members of the species are homosexual. Most still breed. Therefore its impact on the species' "evolutionary strategy" is of little or no consequence to the species as a whole.

It doesn't. But like homosexuality, there are strong biases in society against this trait. Yet people are still turn out left-handed or gay.

Natural selection is blind to traits belonging to specific individuals. Whether a particular individual breeds or is left-handed is missing the point. The statistical effect of some individuals having the homosexuality trait on the survival of the whole species is - in the case of humans - insignificant (so far). Homosexuality simply does not play a decisive role, and so if it is genetic in origin, those genes can "piggyback" on the genome of those who do breed.

Doubting Didymus

Woah there, cowboy! The evolutionary impact of a trait is measured by its effect on the reproductive benifets of that trait. The impact on the species overall strategy does not enter the equasion. IF homosexuality were genetic, and IF it were always true that the trait neccessarily lowers the individual reproductive success, then we should see the trat become less and less common, and we would need to account for its persistance.

Colander of Truth

Homosexuality was discussed, if summarily, in the book <Genome>. It seems that there is a correlation with birth order and homosexuality, and although this hardly explains it, it offered an interesting look at what can go on.

Apparently, successive male fetuses can be allergenic to the mother. The resulting chemical warfare can affect the "masculinization" of the brain. Is this genetic? In part, indirectly, yes. It is physiological change as a result of the Y chromosone.

A good read, Genome, indeed.

thebeave

I used to think it was pretty far fetched myself, but consider that a person could be made into thinking they were a chicken if you mentally tortured and stressed them enough (i.e. brainwashing). Granted, that's pretty extreme, but if there was a tendency or openness to homosexuality to begin with, it would be that much easier to go from being largely straight to largely gay. Prison is perhaps an example where men (and women) undergo tremendous, unnatural social and mental stress. They are largely confined to interact with only their own gender for years on end. Engaging in homosexual behavior may be one manifestation of that rather unusual social situation.

MrDarwin

I think you're misunderstanding the nature of "social and cultural factors". If you consider "society" and "culture" to include the immediate family, then Freud (cough *quack* cough) had lots of ideas about how "social and cultural factors could engender such a strong and irreversible trait."

While I think Freud had it all wrong, I think it's a no-brainer that the influence of parents and siblings, as well as those who associate closely with a child during his upbringing--friends, classmates, neighbors, teachers and other authority figures, etc.--have a strong influence on that child's adult personality traits, up to and including sexual orientation. Most people don't "choose" to be shy, or talkative, or impulsive, or self-centered, either. The problem is that there are so many different factors, both genetic and non-genetic, that go into a person's personality and that interact in a complex way that differs for each and ever child, that it's probably impossible to dissect out one or another as being primarily responsible for a particular character trait.

[ December 12, 2002: Message edited by: MrDarwin ]</p>

pz

Yes, I agree that it is a problem in popular culture. However, it is a problem that is fueled by the scientific culture. We see it all the time: papers come out that claim to have found the "gene" for schizophrenia or obesity or alcoholism or whatever, and they are taken seriously enough that they get published in Science or Nature. I could scarcely believe it when a gene was described and named as a "breast cancer gene"; somebody really needed to be slapped for that one. I can't blame the lay press...they're just echoing the crap that gets published by the academic press. (Note that I do not consider the work or the data to necessarily be crap, and often I think it is quite useful and interesting stuff...but the interpretation of causality, which is usually all that makes it to the popular press, is garbage.)
I've heard a similar line before. "The stars incline, they do not compel." It's a very common rationalization used by astrologers. A similar phrase is disturbingly common in the literature of genetic determinism, especially of human genetics.

I'm sure you're aware of this, but for the sake of others reading this, there's something I want to emphasize. Genes do not cause disease. Defective alleles may lead to disease, sometimes with certainty, sometimes with an incidence dependent on environment. In this case, we are discussing homosexuality, which is not a disease at all. Disease models do not apply. I would also say that Mendel and Morgan do not apply. Part of the problem here is that people have been trying hard to fit disease and relatively simple genetic models to patterns of behavior for which they are simply not appropriate.

And I have been trying to stress that even that is not enough -- in these cases we also lack evidence for a genetic predisposition. Usually, when you read the papers that make these claims, you'll find that the 'evidence' boils down to "I can't think of any other way to account for this effect, so it must be genetic". Genes are the default explanation, not the explanation for which there is actual evidence.
Since I haven't reached that conclusion, I certainly can't claim that anyone else has published such a thing. I'm not saying that genes have no consequences, I'm saying that they are inadequate to explain behavior, and that the differences in behavior we see between hetero- and homosexual individuals are unlikely to have a genetic basis.

You are also asking for a paper that reaches a negative conclusion. Have you ever heard of the file drawer effect? It's extremely difficult to publish such a thing. I was privy to a very long-running series of knock-out experiments in Drosophila neurogenetics, and it was rather enlightening and mostly unpublishable. Gene after gene that had specific and interesting patterns of expression in the developing and adult nervous system was mutated, and mutant gene after mutant gene failed to have any perceptible effect. Only when a gene knockout did something interesting was any attempt made to publish it.
You mean, for example, <a href="http://flybase.bio.indiana.edu/.bin/fbidq.html?FBgn0004652" target="_blank">fruitless</a>, or <a href="http://flybase.bio.indiana.edu/.bin/golink?FB:FBgn0015374" target="_blank">courtless</a> or <a href="http://flybase.bio.indiana.edu/.bin/golink?FB:FBgn0003714" target="_blank">technical knockout</a>? These are, respectively, an RNA Pol II transcription factor, an ubiquitin conjugating enzyme, and a part of the mitochondrial small ribosomal subunit. Does this help you understand the behavior? It doesn't help me in the slightest, and I have a pretty good background in genetics, development, and neurobiology. Like I said, if I'm trying to understand architecture, it's not much help to show me that buildings fall down if you put sand in a lot of the bricks. It's an inappropriate level of explanation.

You are also treading on dangerous ground. These are 'broken' fruit flies -- they have gross pleiotropic defects that produce severely aberrant behavior in the affected individuals. I would say that there is another reason that they do not and cannot explain human homosexuality: they are invalid models. Gay people are not damaged goods, they merely express a behavior that is nearer one end of the range we see in all of us. The pattern that anatomists like Simon Le Vay have been claiming to see (which I find a bit shaky, too) is that male homosexual brains have some organizational features which are normally sexually dimorphic, but that resemble most those found in women. Any genetic explanation is going to have to somehow encompass the fact that these same behaviors are common in 50% of the population, and are clearly not the product of gross mutations of the kind seen in flies.

So yeah, the bottom line is that these mutations in flies have absolutely no relationship to the causes of homosexuality in humans. They might represent a dim beginning to studying the influences of a genetic contribution to behavior in general, but they have little explanatory power now.


There's another conceptual reason why these attempts are misleading. They presuppose a relatively simple one-to-one or small-number-to-one mapping of genes to traits, when it is really a many-to-many or even all-to-all mapping. The analogy I like to make to my classes is that of a Fourier transform of an image. The fft power spectrum is a rendering of the image in the spatial frequency domain, and you can easily go back and forth between the two. What you can't do is point to a spot in the frequency domain and assign it to a spot in the spatial domain -- it doesn't work that way. A point in the power spectrum represents that magnitude of a particular frequency across the whole of the spatial image. Blot out that point, and when you transform back you will sometimes see only a very subtle effect, and sometimes particular features (which use that spatial frequency component heavily) are strongly modified -- but it is still an error in understanding to say that one point in one domain has a causal relationship to one point in the other.

The analogy breaks down in one regard that makes my argument even stronger, though. The fourier transform uses a fixed function to generate the image from a power spectrum. The analogous process for the gene is called development, and it is more variable and contingent on a changing external state. That makes it even harder to translate the genomic state to the phenotypic one in any simple way.

Anyway, it's an analogy that seems to help a lot of math & engineering & comp sci majors grasp the idea -- think of the genome as a power spectrum, development as a kind of sloppy fourier function, and the organism as the resultant image. Most of the biologists go "huh?", but that just tells you that we ought to require more math of our biologists.
You've already brought up that important distinction between necessity and sufficiency. Genes are necessary. They are an indispensable component of the organism. At this point, that is so obvious to everyone that it feels faintly ridiculous to even have to mention it. The simple point I'm trying to make, that seems so obvious to me that I'm also a bit surprised that I have to bring it up, is that genes are by no means sufficient. We cannot explain differences in ordinary behavior, even differences that seem as extreme to us as mate preference, by recourse to genetics.

It's sad that people still try. It's appealing to be able to reduce complex differences to simple genetic causes: "oh, he's that way because he has Xq28". Really, I see it as very similar to the way people like to reduce complex events to "god willed it". Comprehensibility, even false comprehensibility, is comforting. But we should at least try to be more skeptical about it.

ps418

It is obvious -- no one with any understanding of the subject matter believes that genes alone are suffucient to explain interindividual differences in behavior. For most behavioral traits so far studied, genetic variance explains half or less of the behavioral variance.

The point of contention is not whether or not we can explain interindividual differences in behavior on the basis of genes alone -- everyone here agrees that you cannot-- the point of contention is whether we can explain interindividual differences in behavior without any recourse to genetics, e.g. solely in terms of environmental influences such as parenting.

Yes, this is a fine example of the type of genetic determinist explanation that no one with any familiarity with the subject holds.

[ December 12, 2002: Message edited by: ps418 ]</p>

pz

If I didn't despise smilies so much, I'd use a couple of those "head pounding against the wall" ones right here.

What you've just said is pretty much meaningless.
The answer is "yes". We have to. Since the genetic differences between people are relatively tiny, while the socially and culturally induced differences are comparatively huge, there's no other way around it.

ps418

A clear nonsequiter. Small genetic differences can and do often lead to large phentotypic differences. The relative genetic homogeneity of H. sapiens does not entail in any way that all interindividual differences in behavior have to be explained in solely in terms of "society" and "culture," a position abandoned long ago by most mainstream personality researchers.

You say that we have to explain differences in behavior without any recourse to genetic differences. Thank you for a clear statement of what you believe. Now, do you apply this rule to humans only? And do you apply it to all human behaviors? To take some examples, can we explain interindividual differences in obsessive/compulsive behavior solely in terms of "society" and "culture"? What about paranoid schizophrenia, or bulimia, or anorexia, or hyperactivity? Can we explain interindividual variability for these traits solely in terms of "society" and "culture"? And what about cognitive ability? Can we also explain interindividual difference in cognitive ability solely in terms of "society" and "culture"?

Maybe we can test the explanatory sufficiency of "society" and "culture" using some concrete examples?

[ December 12, 2002: Message edited by: ps418 ]

[ December 12, 2002: Message edited by: ps418 ]</p>

ps418

What a nutty comparison. First you try to paint behavior genetics as "genetic determinism," in defiance of fact and logic. Then you compare claims of genetic-influence-but-not-determination of behavior to the "rationalizations" of astrologers. Sometimes you just cant win. Oh yeah, there are no such things as "susceptibility loci" or quantitative traits. Either genes completely determine, or they play no role at all.

[ December 12, 2002: Message edited by: ps418 ]</p>