Alison Gopnik: UCB

One of the most striking generalizations in evolutionary psychology is the correlation between sophisticated and flexible learning-based knowledge in adult organisms and a long protected period of immaturity in the young. In terms of the classic altricial vs. precocial, or r vs. k distinction in evolutionary biology, the altricial species are more likely to rely on learned information as adults. Human beings, of course, have a particularly extended childhood, and are particularly marked by their reliance on learning. This fact suggests that there may be a kind of cognitive division of labor between children and adults. Children are, as it were, the research and development division of the human species, while adults are production and marketing. Children are designed to be particularly good at learning and exploration and adults are designed to exploit the information they have learned as children for purposes of planning. I will review a wide range of developmental and neuroscientific evidence that supports this hypothesis, including some very recent studies in our lab that empirically show that children may be better causal learners than adults. In particular, I will discuss this idea in light of recent work on the computational bases of cognitive development using ideas about probabilistic models and Bayesian learning.