| Abstract | [Introduction]
Perhaps the most desirable attributes of the human motor system are that it be capable of locating the limbs accurately in space using a variety of movement trajectories and that localization be accomplished relatively independent of changes in the initial conditions of limbs. Although it is well documented that these features are characteristic of the behavioral repertoire of both animals and humans, less clear is the nature of the underlying control mechanism(s). Neither of the currently popular closed-loop, feedback (2), or open-loop programming accounts (37) seem completely adequate. For example, although a closed-loop model could accommodate the fact that achievement of final position is possible in spite of a) changes in limb position prior to movement (39) or b) the introduction of abrupt changes in load during movement execution (25,35), it is a at loss when the same findings can be demonstrated under deafferentation conditions (8,26,35,40). Similarly, central motor feedback monitoring may handle deafferentation findings, but go awry when confronted with unforeseen changes in the movement context. Indeed, even a hybrid model that incorporates internal, central feedback loops (14,28) has great difficulty with finding that normal accuracy may result when monkeys are deafferented and, consequently, subjected to unpredictable movement perturbations (35).
In the present experiments, we set out to determine-on an a priori basis-whether any of the observed kinematic characteristics that arise in localization, violate the mass-spring model. Specifically, our tack was to introduce sudden and unexpected torque loads-which acted to drive the limb (in this case the index finger) in the opposite direction-and observe consequent effects on localization. Unlike numerous other studies, we were not particularly concerned in evaluating the various reflex responses to changed loading conditions. Rather, we wished to elucidate the effects of changing dynamic parameters and consequent kinematic variation on the attainment of a specified equilibrium position. In experiment 1,we show that the equilibrium position is accurately attained despite on-line perturbations. In experiment 2 we rule out possible alternative accounts by replicating this result in functionally deaferented individuals. |
|---|
