lambdaphagy: there’s a lot of content in your latest response and I may not be able to give a proper response today.  Here are a few quick notes, though, on things that may have just been misunderstandings (I was not writing very clearly):

When I talk about people studying American college students, my point is that this is especially a problem for people who want to generalize their results cross-culturally.  A sense one gets from people like Pinker is that the universality of the psych features they’re describing is a result of EP, as though we might have initially thought that the mind was more culturally variable than it is, but EPists just checked empirically and it turns out it isn’t.  The tendency to generalize results in this way is a problem for all of psychology, and the WEIRD stuff is a great stab at pointing out and characterizing this problem.  But it’s a bigger problem if your claim is “this is human nature” rather than “this is what we found in a psych study of American college students” (in which case your reader can potentially say things like, well, I’m an American, so even if it doesn’t generalize along that dimension, it’s relevant to me … )

My memories of Pinker seem very different than yours.  To be clear, what I’m complaining about here is the gap between actually existing EP and the kinds of “we have discovered human nature, behold!” claims that I remember Pinker (and at times Tooby and Cosmides? … but these memories are all from many years ago, I could just be full of shit) making.

Yes, Pinker is entitled to his politics, but I remember him being very forceful about the notion that if we accept the scientific evidence, we have to accept something like the politics.  Saying “Pinker is entitled to his politics” feels to me a bit like saying “Rawls and Nozick are entitled to their politics,” which they are, but they also claim they’ve derived (parts of) their politics from first principles, and challenge you to reject the deduction, reject the principles, or accept the politics.  Pinker isn’t that direct, of course, but he’s more direct that I’d like one to be on the basis of any academic psychology in its current state (WEIRD problems, replication problems, etc.)  Note that much of this may just be that as a teenager I had never read anyone with this kind of forcefulness, and so it struck me as a model of forcefulness, perhaps more than it was.

My mention of Sperber was really loose; I don’t meant that his theory already existed when the EP theory appeared, I just mean that there are other explanations, so we don’t have to take the EP theory now for lack of anything else.  Sperber’s theory certainly existed when Pinker wrote about this stuff in How The Mind Works and The Blank Slate.

My criticism of the cheater detection stuff isn’t that it’s necessarily wrong (indeed, from what I’ve read about the issue, it’s still controversial?), but that it was a postdiction.  Our ability to reason better when things are framed in certain ways is interesting, but my point is that I don’t see Tooby and Cosmides as having a model which predicts this – that is, I can’t step inside their theory, forget about the data for a moment, and say, “ah, yes, evolving a cheater detection module makes sense in this context, given these constraints.”  Without sneaking in the data, I don’t know a priori why a cheater detection module should be the right answer to this engineering problem.  (This is what I meant in talking about “general purpose logic modules”: the question is not “do we have a general purpose logic module?” – we don’t – but “do we have a clear picture of the constrained optimization problem that evolution solved, such that we can see why a general purpose logic module is not a feasible solution?”)

The thing about evolutionary reasoning is that there was this constrained optimization problem in the distant past, and we have very incomplete evidence about it.  Actual, substantial theorizing about the situation has to involve specific claims about the constrained optimization problem which can drive various predictions: things like “there was a constraint in place that made features of this type infeasible.“  One can always go back, take an actual datum, and say, “I suppose this must have been optimal, given the constraints!”, but that isn’t predictive.  (”Predictive” in the usual scientific sense of anticipating information you haven’t seen – you can “predict” facts from the past, as long as no one has seen them yet.)

I’m not sure if that makes sense; I think there is probably a clearer, briefer way of saying what I’m saying, but I can’t seem to find it.