Proprioceptive Bimanual Test in Intrinsic and Extrinsic Coordinates

Is there any difference between matching the position of the hands by asking the subjects to move them to the same spatial location or to mirror-symmetric locations with respect to the body midline? If the motion of the hands were planned in the extrinsic space, the mirror-symmetric task would imply an additional challenge, because we would need to flip the coordinates of the target on the other side of the workspace. Conversely, if the planning were done in intrinsic coordinates, in order to move both hands to the same spot in the workspace, we should compute different joint angles for each arm. Even if both representations were available to the subjects, the two tasks might lead to different results, providing some cue on the organization of the “body schema”. In order to answer such questions, the middle fingertip of the non-dominant hand of a population of healthy subjects was passively moved by a manipulandum to 20 different target locations. Subjects matched these positions with the middle fingertip of their dominant hand. For most subjects, the matching accuracy was higher in the extrinsic modality both in terms of systematic error and variability, even for the target locations in which the configuration of the arms was the same for both modalities. This suggests that the matching performance of the subjects could be determined not only by proprioceptive information but also by the cognitive representation of the task: expressing the goal as reaching for the physical location of the hand in space is apparently more effective than requiring to match the proprioceptive representation of joint angles.