Sequence, time, or state representation: how does the motor control system adapt to variable environments?

Abstract

Does the observation of well-timed movements imply the existence of some internal representation of time, such as a hypothetical neural clock? Here we report the results of experiments designed to investigate whether subjects form a correct adaptive representation of mechanical environments that change in a very predictable manner. In these experiments, subjects were asked to execute arm movements over a twodimensional workspace while experiencing time-dependent disturbing forces. We provide a formal definition for time representation and conclude that our subjects didn't use time representation for motor adaptation under the tested conditions. Subjects performed arm-reaching movements in the following experiments: (1) six experiments in a sinusoidal time-varying force field; (2) six experiments in a simple sequence of alternating viscous force fields, in which the number of targets allowed for the approximation of the force by a complex state-dependent force field; and (3) six experiments in the same simple sequence of alternating viscous force fields, in which no state-dependent force field approximation was possible. We found that the subjects did not adapt to the timevarying force field and were unable to form an adequate representation of the simple sequence of force fields. In the latter case, whenever possible, they adapted to a single state-dependent field that produced forces similar to the two alternating fields. This state-dependent field produced the same forces as the applied sequence of fields only over the trajectories that subjects executed during the training phase. However, the state-dependent field was inadequate to produce the correct forces generated by the field sequence over a new set of trajectories.These results are not consistent with the hypothesis that subjects would develop a correct representation of time-dependent forces, at least under the tested circumstances. We speculate that the system responsible for adaptation of movements to external forces may be unable to employ temporal representation. While it is possible that such a representation may emerge in a more prolonged and/or intense training, our findings indicate a preference by the adaptive system to generalize based on representing dependence of external forces upon state rather than upon time.

Consecutive learning of opposing unimanual motor tasks using the right arm followed by the left arm causes intermanual interference

“…As initially mentioned, several previous studies demonstrated motor memory interference when two different tasks were learned using the same arm [17,18,21–25,32]. In some cases, this interference was shown to be reduced when associating the competing tasks with visual or proprioceptive context cues [33–37].…”

Augmented Dynamics and Motor Exploration as Training for Stroke

“…For motor recovery, training on a variety of tasks provides better improvement in overall function than repetitions of the same task [7], [8]. Rather than rote memorization of motor commands, the nervous system appears to learn associations between forces and movement states [18]. Allowing free movement presumably would provide the richest experiences of these dynamic relationships.…”

“…Researchers have shown that the motor system more successfully generalizes learned motor plans between movements spanning similar positions and velocities [16], and exhibits a preference for such generalization in a joint-based coordinate frame [17]. Evidence suggests that the motor system plans according to associations between expected forces and movement states [18]. Furthermore, the motor system has been shown to transfer skills between environments with overlapping characteristics [19]–[21].…”

Evaluation of negative viscosity as upper extremity training for stroke survivors

“…The motor system can generalize learned strategies to various movement trajectories spanning similar positions and velocities [29]. Rather than rote memorization of motor commands, the nervous system appears to learn associations between forces and movement states[30]. From a systems modeling perspective, broader excitation of movement states informs more accurate system identification [31].…”