ABSTRACT
A fundamental problem in action research is to determine how the motor system selects among alternative methods for achieving specific goals. We suggest that a useful approach to this problem is to identify optimization constraints. One constraint, introduced here, is to minimize times in joint angles far from resting positions. This constraint is shown to account for data from tasks in which adult human subjects pick up a cylindrical bar and transport it to various locations. The data take the form of overhand versus underhand grip choices. As assessed by awkwardness ratings of these two grip postures, grip posture choices are best predicted by assuming that subjects attempt to minimize awkwardness at the end of the first bar transport move they perform. We show that this outcome is consistent with the Extreme Joint Angle Constraint and is optimal in the sense that awkward joint angles, when they are adopted, allow for passive rotation of the arm into desired final positions.
