Abstract
Observing others is a powerful way to learn about goal-directed actions. But observational learning entails more than noticing the movement trajectory. It provides information about the actor’s intended goal and the means to achieve it. Prior developmental work suggests that long before children can achieve goals themselves, they can detect others’ goals. However, the question is still open as to whether young children notice efficient means to achieve a goal. Here, we addressed this question using a novel combination of methods (eye tracking, pupillary responses, EEG, and machine learning). Preschoolers (N=22) and adults (N=22) watched actors use efficient, adult-like and inefficient, child-like means of grasping a hammer to pound a peg. Our displays were designed to test whether children differentiate the efficiency of means to achieve a goal in a task where they do not use efficient means themselves. Eye tracking showed that participants at both ages looked equally long at the goal (peg), but differed in looking at the means: Adults looked longer at the hammer and hand and performed more gaze switches between hammer and hand and between hammer and peg; deep learning analysis distinguished efficient from inefficient grasps for adults, but not for children. Moreover, only adults showed differential physiological responses to efficient versus inefficient grasps with increased pupil dilation and differential patterns of action-related neural activity (EEG). Taken together, our findings show that children can actively direct their gaze to look at goal-directed actions without seeing whether the means are efficient or not. Moreover, findings suggest that the development of action perception is built from children’s own motor experiences.