Dickins, cosmides, reasoning, modularity, and Wason's task

Abstract

Cosmides, and other "evolutionary psychologists," take Wason selection task results as paradigmatic evidence that to model human reasoning as systematic formal semantic reasoning is mistaken (Cosmides 1985, 1989). Indeed, these results are also supposed to exhibit domain and content specific modules selected for over our Pleistocene hunter-gatherer tribal existence. Such selectionist evolutionary explanations are to provide the true and proper foundation for all the cognitive/social sciences. To the contrary, twenty years before Cosmides took note of it in 1985, Professor Wason and other researchers identified this task as a most remarkable deviation from what they have established as normal human reasoning competence and its more typical performances, and they have a plausible formal explanation of what causes the problem in this formally unusual case (in effect, dealing with double negation). Cosmides's proposed social exchange or "cheater-detector" module is not a module nor is it content-specific in the relevant sense. Evolutionary psychologists have merely taken up a perspective characteristic of social anthropology, which has been nativist and Darwinian for most of its history as a discipline. But at its best anthropology, at least as a speculative historical discipline bears something like the same relationship to cognitive science as cosmology bears to physics and chemistry. © The Executive Management Committee/Blackwell Publishing Ltd. 2005.