Empiricists AND rationalists

Historical empiricists, such as Locke and Hume, and historical rationalists, such as Descartes, would all find comfort in what 21st-century cognitive science has to tell us about concept acquisition. As the rationalists insisted, there are innate input analyzers that compute perceptual representations — veridical representations of the distal world. Contrary to empiricist theories, these input analyzers are devices that do not have to be constructed by learning processes that operate over sensory representations. The example I alluded to in chapter 2 was depth perception.

Still, as the empiricists believed, an informational semantics (see chapter 13) seems appropriate for both sensory and perceptual representations. The input analyzers that create representations of color, of depth, and so on evolved to work as they do, and evolution is a process that is responsive to veridicality. That is, we have evolution to thank for guaranteeing that our representations of depth have the content they do and can fulfill the computational role required of them. Also, as the empiricists believed, garden-variety learning processes create previously unrepresented concepts from this initial stock of representational resources. What the empiricists did not envision were learning processes that create new primitives. But that is a long story, the story of this book. (Susan Carey, The Origin of Concepts, 2009, p. 448-449)


Mensagens populares deste blogue

The Demon Theory of Friction

A kilobyte rewritable atomic memory