Distinctively human thinking: modular precursors and components

To what extent is it possible to see the human mind as built out of modular components? The question is interesting in its own right, but also because evolutionary psychologists have put forward a variety of arguments for massive modularity, and against domaingeneral models of cognition. This gives rise to a puzzle. For how, then, are we to explain the flexibility which is so distinctive of human thought and behavior? Part of the answer may be (a) that many of the conceptual / central modules are best described as modular learning mechanisms, and (b) that amongst these modules may be systems for generating and acquiring cultural products (especially cultural norms). This then makes possible cultural evolution, with proliferation and diversification of social structures. But there is a deeper puzzle which arises when we focus on distinctive human thinking, rather than the diversity of human behaviors and cultures. This is that human beings are manifestly capable of conjoining information freely across domains. We have no difficulty in thinking thoughts which link together information across modular barriers. How is this possible, if the arguments against domain-general cognitive systems are sound? For example, how is it that our thought-processes are computationally tractable, if our thinking can be so completely holistic, drawing contents from any domain? Fodor (1983, 2000) has famously argued that since human thought is plainly holistic, it is plainly non-modular, and so is plainly not computationally tractable. His conclusion is that human thinking must remain a mystery for the foreseeable future. A preferable response is surely to investigate how far we can get in building non -domain specific thinking out of modular components. I shall sketch an account which utilizes (a) a variety of different central / conceptual modules, of the sort postulated by evolutionary psychologists and others, (b) the re-use of input perceptual modules, in imagination, and (c) the use of a modular language faculty to integrate information across different domains. This account is consistent with recent cognitive interpretations of the archeological record, and is also consistent with recent experimental evidence collected by Liz Spelke and her colleagues (1996, 1999, 2001, forthcoming). References