In mathematics we often prove that some proposition is true by showing that the alternative is false. The principle can sometimes work in other disciplines, but it’s tricky. You have to have a very good understanding to know that some things are impossible (or close enough to impossible). You can do it fairly often in physics, less often in biology.
Sometimes you can make progress on a problem just by realizing that other people are calling something impossible, but don’t understand it well enough. For example, there were people who were sure that Neanderthals and anatomically modern humans couldn’t have interbred successfully because paleontologists had (often) said that Neanderthals were a different species. The problem was that the paleontologists meant that Neanderthals were a morphological species, a group whose skeletons could be distinguished pretty easily from those of AMH, but those talking about the nonexistence of gene flow took it to mean that they were a different biological species, one with which gene flow was impossible. There was no evidence for that, and it was bloody unlikely, considering that the two groups were only separated for a half a million years or so. Others, of course, said that humans would never have done the nasty with anything so repulsive. I’m not supposed to call those people stupid – so I guess it’s thesaurus time: dense, dim, doltish, dopey, dull, dumb. Take your pick.
Back in Lord Kelvin’s day, paleontologists were pretty sure, from sedimentation rates and the geological record, that the world was at least hundreds of millions of years old. Kelvin knew they were wrong, since he was much smarter than them (almost certainly true) and he couldn’t imagine how gravitational contraction could keep the sun going more than ten million years or so. This was before radioactivity, mass-energy, fusion, and all that. Kelvin didn’t have a good enough understanding of the Sun’s energy production to call impossible. He made a mistake – but he wasn’t stupid. Arrogant, maybe.
Peter Duesberg made a fool of himself by saying that retroviruses must be harmless to survive , since they need viable cells to replicate. Only viruses that kill cells in the process of replication should cause disease. However, as it turns out , the universe was not limited by his imagination. HIV does kill cells, but infected cells are also killed by lymphocytes.
There are many other viral diseases with non-Duesbergian mechanisms. LCM [lymphocytic choriomeningitis] only seems to cause severe effects, such as paralysis, in mice that bother to fight it (except for long-term kidney damage). You can successfully treat those mice with immunosuppressive therapy, although I can’t see why anyone would bother, other than Hugh Laurie and Geena Davis. You can get high levels of immune complexes (an antibody bound to an antigen) from some viral infections [such as Hepatitis C] , and those complexes can precipitate in many organs and cause damage, most notably causing kidney failure. Although with any luck you would already died of liver cancer, another non-Duesbergian pathological mechanism. The cancer doesn’t really help the virus, obviously, but it is an accidental long-term effect of actions that do aid the virus in the short term.
For that matter, there are a bunch of retroviruses related to HIV (lentiviruses) that trash various species, usually long after infection: feline immunodeficiency virus, Visna virus in sheep, etc.
So you have to keep this in mind when you’re trying to show that genetic mechanisms, at least as the main cause of homosexuality, don’t work – biology is complicated. Quite likely you’d have to put more than five minutes in on it: I did. You would wonder about heterozygote advantage, something like sickle-cell. But you would learn that you need a big advantage in order to generate a high frequency of a trait that reduced fitness so much – and big advantages are hard to find. The only known cases of such strong heterozygote advantage in humans are all defenses against falciparum malaria, which is a hell of a disease. And those defenses are limited to populations that spent a long time exposed to falciparum malaria – tropical and subtropical regions of the Old World. Moreover, any such gene would have an easily recognizable Mendelian family pattern , even if it had partial penetrance for some reason – and homosexuality does not. You might consider sexually antagonistic genes, or alleles that cheat at meiosis like the t-allele in mice (although they are rare) – but recent GWAS studies rule them out. Indeed, there is no gene that explains more than a tiny smidgen of the story: you know that, if you keep track of the literature. That fact pretty much rules out all the genetic scenarios. Which is not to say that genetics does not have some influence, but hell, that’s true of everything. I’m sure that there were heritable factors that influenced the probability that you’d die on Omaha Beach (ones that affected height, for example) but the true causal explanation is what we usually call the German theory. You can land there today and, without Germans, your chance of having your head removed by an 88 millimeter shell is practically zero.
You might note the obvious similarity between gay men and worker bees, and think about kin selection, except that there is no such similarity. Gay men are not super-avuncular. Talking about all their special help to kin is exactly like talking about the many fitness benefits of male lactation in humans – a fascinating story, except for the fact that it never happened. What would you call a biologist who publicly advocated a theory whose key mechanism was based on guy milk?
If you knew the MZ twin concordance, which is around 25%, you’d already know that some environmental factor had to be the main cause. If you knew history, let alone microbiology, you’d wonder about pathogens, because they have caused most of the common fitness-reducing syndromes. As for the notion that hosts would surely evolve resistance – microorganisms typically evolve a bit faster than we do. A bacterium can go through thousands of generations in a year.
You need to do a fair amount of spadework to make the case against a genetic cause. Even then, you can’t perfectly sure. It’s not like math. There’s always the possibility that there’s an undiscovered genetic phenomena that doesn’t have a name yet, not yet even a twinkle in someone’s mind’s eye. As Haldane would have said, things may be queerer than we can suppose. But that’s not the way to bet.