ABSTRACT
No issue in motor control has created as much controversy as the role of afferent input in movement. From 1950 to 1970, the intensive experimental efforts to resolve this issue primarily involved acute preparations in anesthetized animals, sensory alteration procedures, or reaction time paradigms in waking humans. Although these methodologies all have serious shortcomings for addressing the natural character of sensorimotor influences, certain relatively extreme positions were nevertheless formulated. In the last two decades, single neural unit recording in waking animals, unanticipated perturbation techniques in humans, and like procedures have yielded a moderation of those extremes, with increasing development of a common ground. Specifically, advocates of central nervous system pre-specification of muscle contraction and movement in the form of motor programs have altered their constructs and definitions to accommodate the ubiquitous evidence that almost all motor output parameters are modifiable during execution. The limited and almost strawman-like conceptualization of sensory influences being implemented exclusively via closed-loop, negative feedback circuits also has been rejected. Likewise, even zealous advocates of sensory influences have recognized that the particular contributions of a given afferent input vary considerably with the motor system under study, task goals, experience, and yet other variables. In this latter respect, the challenge for the next decade is to discover principles of motorsensory control that generalize across motor systems and tasks, despite the tremendous variation in both the degree and kind of obviously critical sensory contributions.
