Since I'm going to be beating up on Thornhill and Palmer a lot, unless I get sick of shooting fish in a barrel, I thought I should include a post about something most of their critics seem to get wrong. Not that this shows T&P are right about the point at issue, only that the evidence being presented against them is, in this case, quite weak.
The strong thesis of Thornhill is that there is some adaptation humans possess which encourages rape under some circumstances because under some circumstances rape increases reproductive success. Many critics point out that rapes relatively rarely produce offspring, and carry high costs, so they are skeptical that any inclination to rape could ever be adaptive. That's fair enough; it's an empirical question, and Thornhill doesn't provide anything like conclusive evidence on his own side of the question (Palmer doesn't even think this particular thesis is true, so it isn't just T&Ps critics who complain about it).
However, some critics claim that the frequency of rapes which involve definitely non-fertile targets, or which don't involve vaginal penetration and ejaculation, constitute conclusive proof that there's no rape adaptation. This is obvious nonsense. If there were a genetic adaptation for rape, as Thornhill insists, there's no reason to think it would be some perfect mechanism which always operated precisely to increase reproduction. No adaptations are like that. All Thornhill needs is that there is some mechanism which exists because it sometimes causes reproductively successful rapes, and is selective enough that the increase in reproductive success due to those rapes outweighs the reproductive disadvantages produced by the mechanism. If the mechanism is not specific enough to reliably lead to reproductively successful rapes, that makes Thornhill's situation harder, as reproductively unsuccessful rapes have mostly the same reproductive disadvantages while providing no compensating reproductive benefits, but it doesn't show he can't be right. Probably most mechanisms for bringing about reproduction fail more often than they succeed, for various reasons. Thornhill's critics should focus on the enormous empirical hurdles his thesis faces, not try to imply that his thesis is in principle impossible.
Still, I can't let Thornhill completely off the hook even on this. Thornhill's own frequent emphasis on the high degree of specificity he expects adaptations to show encourages the reader to think he is assuming a very tightly focused mechanism, perhaps because that would make his task easier. Thus, it is partly his fault that his opponents don't consider whether a less specific mechanism could be operative.