A Donders' Like Law for Arm Movements: The Signal not the Noise.

Experiments were done to determine whether the starting position of the arm influences its final configuration (posture) when pointing to, or grasping, targets located within the common workspace of the arm. Subjects were asked to point to, or grasp, each of six targets from five, or seven, widely spaced starting positions. We found that the variability (standard deviation) of the arm's configuration, measured as the angle of inclination of the plane delimited by the arm and forearm, averaged about 4° for comfortable speed pointing movements and was only slightly higher for fast pointing movements. Comfortable speed reaches to grasp the targets were associated with slightly lower variability (3.5°) in final arm configuration. The average variability of repeated movements to a given target from a single start position (3.5°) was comparable to that of movements from different start positions to the same target (4.2°). A small difference in final arm inclination angle, averaged across all subjects and targets, of 3° was found between two pairs of starting positions. This small and possibly idiosyncratic effect is within the "noise" of final arm orientation variability for repeated movements (i.e., 3.5°). Thus, the variability of final posture is not for the most part due to different start positions, it is inherent to movement per se. Our results reconcile conflicting previous studies and are consistent with past works suggesting that a Donders' like law is indeed largely upheld for unconstrained visually guided arm movements. In summary, considering movements within a typical work space, when the hand is moved voluntarily to a given spatial location the posture of the arm is nearly the same regardless of its starting position. Importantly, variability is inherent to the rule.