Author's School

School of Engineering & Applied Science

Author's Department/Program

Biomedical Engineering


English (en)

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)

Chair and Committee

Kurt Thoroughman


On a daily basis, humans capably and effortlessly interact with their surrounding environment through the performance of accurate movements. Movements are often perturbed through the physical influence of the surrounding environment, interaction with objects, or injury, yet adaptation is both rapid and flexible. When adapting, humans are informed by their direct trial-and-error movement experience, incrementally updating predictive control on the next movement; how the brain processes sensed errors into adaptation is not fully known. We may also learn to move through the observation of the movements of others, as visually observed movement information may be transformed into motor memories that influence subsequent motor command; the neural computations underlying such a learning process are not well understood. In this thesis, I aimed to further understand how people incrementally update their predictive motor control in novel haptic environments as a function of sensation during action and during observation. Theories of motor learning suggest that adaptation scales with the size of experienced error. Previous studies have indicated the relationship between adaptation and sensed error can be modulated by statistics of the perturbing environment. In Chapter 2, we considered how the duration of experienced force perturbations might modulate adaptive strategy and found that people becomes increasingly sensitive to kinematic and dynamic sensory signals when experiencing perturbations of decreasing duration. We further found that subjects experiencing pulsatile forces adapted their steady-state feedforward prediction of dynamics with a persistently mismatched breadth when compared to the duration of experienced forces, but learned to closely match the experienced duration of full-movement forces. In the next two sections, we considered the learning effects of movement observation. In Chapter 3, we newly designed and implemented an experimental paradigm in which movement and observation were interleaved, varying the strength of perturbations and associated kinematics from trial-to-trial. Based on previous descriptions of long-term learning by observing, we hypothesized that incremental adaptation would be corrective with respect to observed errors but more modest in magnitude than gains from physical practice. Instead, we found that the incremental adaptive response of movement observation generally countered the direction of experienced forces and was similar in magnitude to the response following action, but was not error-corrective with respect to real-valued signals. Previous research had established an initial advantage when adapting to novel dynamics following observation but the learning processes influencing this effect were unknown. In Chapter 4, we newly demonstrated that the long-term movement observation resulted in adaptive changes in feedforward predictive dynamics. We found that observation generated a small, but significant, compensatory change in reach dynamics that could be characterized by a learned scaling of perturbation-appropriate kinematic signals, suggesting a transformation of visual inputs into a neural representation of environment dynamics.


Permanent URL: