ABSTRACT
From an evolutionary point of view, what is most overwhelmingly in need of explanation is the generality of cognition in humans, as compared to that of our immediate ancestors. Humans sing operas, build diesel engines and agonise about domain dependence. Above all they can sustain the kinds of societies required for each of these activities. What part a domain specialist apparatus plays in this new-found encyclopaedic cognition is a good question. But we believe that the wholesale current appeal to domain speci®city has led to the neglect of what should be the guiding observation ± the hugely more general capacities for reasoning in human beings. The rejection of homogeneous universal learning mechanisms, or of monolithic universally interpreted languages and logics, as simplistic bases for explaining the generality of human cognition, has led to the ``massive modularity'' view of reasoning (Sperber, 1994, 2002), but we will argue that this is a remarkably poor basis for explaining what is overwhelmingly in need of explanation.
