Neurocognitive mechanisms of action sequence processing

Abstract

Imagine an ordinary action sequence, such as making a peanut butter sandwich. You first reach for two bread slices, grasp your peanut butter jar, open the jar, reach for a knife, insert it into your jar… and so forth. Even the simplest actions contain a complex stream of information from movements, objects, and other people and their goals. How do young infants, who have limited amounts of motor experiences or world knowledge, process the action sequences they observe around them? How do we as humans learn novel action sequences from mere observation? This thesis investigated the neurocognitive mechanisms that underlie the ability to learn and predict action sequences. Specifically, I focused on statistical learning—the ability to extract statistical regularities from the environment—as a candidate mechanism for learning novel action sequences. This thesis presents a series of experiments with infants, toddlers, and adults using a combination of behavioral, eye-tracking, and neuroimaging (EEG) methods. Findings show that, across development, observers correctly predict upcoming actions and their effects by successfully extracting the statistical regularities in action sequences. Further, findings show that our own motor system plays an important role in this process. The experiments described in this thesis collectively contribute to a better understanding of how statistical learning abilities contribute to action prediction across development.