Primary Science Education Conference

By the time children are five years-old, they have solved all the hard problems of cognitive science: problems of face recognition, natural scene understanding, motor planning, navigation, number understanding, causal reasoning, language acquisition, understanding others’ thoughts, desires, and feelings, and understanding right and wrong.  The rapidity, robustness, and flexibility of children’s learning exceeds anything mastered by our most powerful artificial intelligence systems – and yet children achieve this remarkable learning while they appear to be busy climbing on jungle gyms, playing with blocks, and talking to stuffed animals.  Here I will talk about children’s remarkable ability to draw rich, abstract inferences from sparse data – and the puzzling gap between our formal theories of learning and the actual behavior of our most powerful learners.  I will point to the ways that children’s play has provided insight into the ways that they learn (and the many mysteries that remain), and I will show experiments looking at the relationship between uncertainty and children’s exploratory play, and trade-offs between exploration and instruction.  In the second part of the talk, I will discuss how children deploy some of their remarkable learning abilities to reason about their own and others’ goals, abilities, and emotions.  I will also introduce a new, open source, online developmental laboratory (Lookit!), which has the potential to expand both the questions we ask and the populations we reach.  I’ll end with some thoughts about what cognitive science has to offer education – and more importantly, the ways that educators might help transform the practice of cognitive science.  Finally, time allowing, I’ll show a few cartoons.