From Errors to Expertise: How Humans Acquire, Adapt, and Refine Motor Skills

Abstract

Motor learning enables humans to acquire, adapt, and refine skilled movements through practice. While the learning mechanisms underlying motor adaptation and skill acquisition have been widely studied, less is known about how these motor learning types differ in their retention and generalization patterns, their underlying neural mechanisms, and whether similar mechanisms support continued improvements in motor execution during extended practice. This dissertation investigates these questions by comparing motor adaptation and de novo learning, and by examining how practice improves motor execution. Using visuomotor rotation and mirror reversal tasks to probe adaptation and de novo learning, respectively, we identify distinct behavioral signatures across the two learning types. Compared to adaptation, de novo learning shows slower reaction times, no aftereffects, robust retention across days, and broader generalization across spatial contexts and effectors. To examine the neural dynamics underlying these processes, electroencephalography (EEG) was used to measure neural activity during movement preparation and feedback processing. Established EEG markers, including the Readiness Potential, beta attenuation, alpha synchronization, and the P3 response, show stronger modulation across training during visuomotor rotation than during mirror reversal learning. Finally, improvements in motor execution during skill-based practice show robust retention and broad generalization, suggesting that continued improvements in motor acuity may share learning mechanisms with skill acquisition. Together, these findings show how both distinct and shared mechanisms contribute to the acquisition, retention, and generalization of motor skills.