The long term goal of this research is to establish a theoretical basis for understanding sensorimotor control in posture, including behavior and its underlying physiology. A focus in the present application will be the relationship of neural activity in the cerebellum to postural control. The methods will be entirely theoretical, and a major objective of the research will be to develop a mathematical structure describing the postural control system. The work will be accomplished with the collaboration of experimental physiologists. To achieve this goal, components of strategies that reduce the number of sensorimotor control parameters will be identified, given a set of task requirements such as maintaining stability. Then the components of strategies will be represented as relations between classes of stimuli and motor commands. Since classes of stimuli can also be equivalent for a population of neurons, populations of neurons will be identified whose activity expresses the same equivalence of classes of stimuli and motor commands. Specifically, patterns of neural activity in the cerebellum will be constructed using receptive field data, in order ultimately to determine to what extent populations of active cells in the cerebellum are directly controlling posture. A mathematical structure will be developed to describe postural control at both the behavioral and the neural-activity levels. Results of this research will clarify the unified nature of postural behavior and its underlying physiology, and will provide a more rational basis for diagnosis and treatment of balance disorders.