As a writer, there are few nicer things than reading a lucid and thought provoking response to an article you've written. PZ Myers, in responding to my article on the controversial theories of Joan Roughgarden, has written a gem of a blog post. Much of his post is devoted to scrutinizing and criticizing Roughgarden's theories. I'm not an evolutionary biologist, but I think Myers makes some excellent points. (He is a skeptical voice in my article as well.) Although he discusses several possible explanations for the evolutionary origins of homosexuality, he settles on a favorite:
Homosexuality is a byproduct. This is my favorite explanation, because ultimately it's about development. Why do men have nipples? Because women need them. Both men and women have the same set of genes (more or less), and follow very similar developmental pathways, and the nipple represents a developmental constraint or byproduct: mutations that knock out the male nipple might also knock out the female nipple, so the structure is retained in both sexes. Male nipples are a byproduct of a function needed by the other sex. We might also ask, why do some men love other men? The answer: because women need to love men. (We could also propose the complement, that lesbians exist because men need to love women.) If there are pathways that can predispose an individual to find males sexually attractive, the base structure is present in both men and women, and what we have are additional mechanisms to modulate the expression of the trait in men vs. women.
Debates like this are why I find the question of homosexuality and evolution so interesting.
My own guess is that there is no single adaptive explanation for homosexuality, and that this is why it's so difficult to pin down. I imagine Roughgarden is right about some animals (homosexuality acts as a social glue), and PZ Myers is right about others (gayness is like male nipples). The more I think about this answer, the more I like it. Wouldn't it be ironic if homosexuality was an example of convergent evolution, in which different species settled on the same answer at different times and for different reasons? Instead of being a deleterious mutation, homosexuality is a beneficial adaptation with multiple benefits . Take that, homophobes.
I agree with you that several of the possible explanations listed by PZ are going to be correct, some more correct than the other in one species, and a different combo in another.
The two you mention (social-glue and by-product hypotheses) appear the strongest, and they can also go hand-in-hand: a byproduct trait getting exapted for the social glue purposes.
I have linked to my old review of her book in this post.
homosexuality is a beneficial adaptation with multiple benefits
well, it better be mighty in its byproducts since it seems like it drives individual fitness down a lot (e.g., the group selective effects or inclusive fitness should be really compelling to correct for that). also, i believe the latest concordance between MZ twins shows around ~25%, for what it's worth....
Yes, as always in biology, there is no one simple answer, casuation is very complex etc...