Postural invariance in three-dimensional reaching and grasping movements.

The question of whether the final arm posture to be reached is determined in advance during prehension movements remains widely debated. To address this issue, we designed a psychophysical experiment in which human subjects were instructed to reach and grasp, with their right arm, a small sphere presented at various locations. In some trials the sphere remained stationary, while in others (the perturbed trials) it suddenly jumped, at movement onset, to a new unpredictable position. Our data indicate that the final configuration of the upper limb is highly predictable for a given location of the sphere. For movements directed at stationary objects, the variability of the final arm posture was very small in relation to the variability allowed by joint redundancy. For movements directed at "jumping" objects, the initial motor response was quickly amended, allowing an accurate grasp. The final arm posture reached at the end of the perturbed trials was neither different from nor more variable than the final arm posture reached at the end of the corresponding stationary trials (i.e. the trials sharing the same final object location). This latter result is not trivial, considering both joint redundancy and the motor reorganization imposed by the change in sphere location. In contrast to earlier observations, our data cannot be accounted for by biomechanical or functional factors. Indeed, the spherical object used in the present study did not constrain the final arm configuration or the hand trajectory. When considered together, our data support the idea that the final posture to be reached is planned in advance and used as a control variable by the central nervous system.