A model of the role of the cerebellum in tracking and controlling movements

Abstract

The cerebellum contributes to motor control but also appears to track movements that it is not involved in producing. This is incongruent with motor theories of cerebellum. However, cerebellar motor function can be explained as a consequence of a tracking function. Accurate motor control requires state information about the motor system and about movement goals. I propose that the cerebellum is a neural analog of a state estimator. In a control system, state estimator malfunction leads to deficits which resemble symptoms of cerebellar dysfunction. State estimators are essential components in tracking and navigation systems. The cerebellum may have analogous roles. The hypothesis that cerebellar function is state estimation, not motor control, is consistent with established facts, avoids anomalies associated with motor theories, is testable, and suggests new ways to diagnose and treat cerebellar dysfunction.