MOTOR ADAPTATION TO CORIOLIS-FORCE PERTURBATIONS OF REACHING MOVEMENTS - END-POINT BUT NOT TRAJECTORY ADAPTATION TRANSFERS TO THE NONEXPOSED ARM

1. Reaching movements made in a rotating room generate Coriolis forces that are directly proportional to the cross product of the room's ang ular velocity and the arm's linear velocity. Such Coriolis forces are inertial forces not involving mechanical contact with the arm. 2. We m easured the trajectories of arm movements made in darkness to a visual target that was extinguished at the onset of each reach. Prerotation subjects pointed with both the right and left arms in alternating sets of eight movements. During rotation at 10 rpm, the subjects reached o nly with the right arm. Postrotation, the subjects pointed with the le ft and right arms, starting with the left, in alternating sets of eigh t movements. 3. The initial perrotary reaching movements of the right arm were highly deviated both in movement path and endpoint relative t o the prerotation reaches of the right arm. With additional movements, subjects rapidly regained straight movement paths and accurate endpoi nts despite the absence of visual or tactile feedback about reaching a ccuracy. The initial postrotation reaches of the left arm followed str aight paths to the wrong endpoint. The initial postrotation reaches of the right arm had paths with mirror image curvature to the initial pe rrotation reaches of the right arm but went to the correct endpoint. 4 . These observations are inconsistent with current equilibrium point m odels of movement control. Such theories predict accurate reaches unde r our experimental conditions. Our observations further show independe nt implementation of movement and posture, as evidenced by transfer of endpoint adaptation to the nonexposed arm without transfer of path ad aptation. Endpoint control may occur at a relatively central stage tha t represents general constraints such as gravitoinertial force backgro und or egocentric direction relative to both arms, and control of path may occur at a more peripheral stage that represents moments of inert ia and muscle dynamics unique to each limb. 5. Endpoint and path adapt ation occur despite the absence both of mechanical contact cues about the perturbing force and visual or tactile cues about movement accurac y. These findings point to the importance of muscle spindle signals, m onitoring of motor commands, and possibly joint and tendon receptors i n a detailed trajectory monitoring process. Muscle spindle primary and secondary afferent signals may differentially influence adaptation of movement shape and endpoint, respectively.