People show remarkable sophistication in their implicit reasoning about complicated physical systems: they can play tennis, shoot pool, hit a tetherball, etc. Here I go over recent evidence that such reasoning is supported by an intuitive, probabilistic model of Newtonian physics that supports forward-simulation of world dynamics to make predictions and inferences about objects. The intuitive physics that people use can account for performance in multiple object tracking in adults and reasoning about objects in infancy and childhood. Adults' kinematic predictions reveal that human intuitive physics include noisy dynamics, and that task-specific biases can arise from the same physical model given different task relevant uncertainty. Finally, online physical prediction reveals that people not only employ a noisy Newtonian forward simulation system, but also have a rapid (perhaps topological) qualitative reasoning system. Together, these results begin to characterize the intuitive physical system and its uses across domains of reasoning.
Ohio State navigation bar
