UC Berkeley

During our daily lives, we make thousands of movements. When we stop and consider that doing something as ordinary as reaching for a glass of juice involves the precise sequential contraction of dozens of muscles simply to move our hand, we appreciate the immense problem that our brains are solving. If we then recognize that both the world and the body are constantly changing, the accuracy with which we move becomes quite staggering. Moving with such proficiency requires the motor system to be continuously learning and adapting. A host of neural structures are important for this behavior. One remarkable part of this system is the cerebellum, or "little brain": a phylogenetically ancient neural structure, containing over half of the neurons in the human central nervous system. Damage to this structure results in a loss of coordination, with marked impairments in the control of eye movements, the timing of simple rhythmic movements, and most intriguingly the ability to adjust well-learned motor skills.

The aim of this dissertation is to explore the processes of motor control and learning, with a special emphasis on the functional contribution of the cerebellum. Following a short introduction (Chapter 1), empirical evidence is provided from two classes of behavior. Chapter 2 deals with the production of rhythmic movements in a population of patients with cerebellar pathology. Chapters 3 through 5 involve the production of goal-directed reaching movements, carefully investigating the representation and correction of errors through the combined use of psychophysics, brain imaging, and patient studies.

In Chapter 2, patients with cerebellar pathology are observed to be impaired when producing rhythmic movements, particularly when the movements contain a distinct event that can be used to determine the performance error. In Chapter 3, we observe that by reshaping a target region, we can predictably impact the correction of movement errors during reaching movements toward that target. In Chapter 4, we provide physiological evidence of the representation of movement errors within the cerebellum, an effect only observed when appropriate measures are taken to factor

out the effects of changes in heart rate. In Chapter 5, we show that patients with cerebellar pathology are impaired in adjusting their movements to counteract a visual perturbation, and furthermore suggest that this impairment is equivalent whether the perturbation is applied suddenly or gradually.

Taken together, this work demonstrates that we learn to make better movements by rapidly evaluating our movements with respect to our goals, and correcting any mistakes with the help of the cerebellum.