THE AMENDMENT OF LARGE-MAGNITUDE AIMING-MOVEMENT ERRORS

The amendment of large-magnitude errors is discussed with particular reference to the correction of directional errors. It is argued that the correction of a large-magnitude error may be understood as a double-step tracking situation in which the second step reflects an artificially induced error signal. At issue is how the motor system may respond to such stimuli in rapid succession. The temporal integration of error stimuli predicts an initial response which reflects the weighted average of the step positions. This approach, however, does not explain the marked increase in peak velocity for the corrective response, compared to an equivalent single-step response. Alternatively, it has been argued that the initial response is initiated as if it were a single-step response to the initial step position and is subsequently amended. The superposition hypothesis argued that the two responses are planned in parallel and overlap in time, to be superimposed one on the other. This hypothesis does not explain changes in the direction of a double-step response at its initiation. The braking hypothesis argues that the initial response is halted and a corrective response initiated as rapidly as possible. This approach cannot explain changes in the slope of an ATF as a function of the second target-step amplitude. A model of double-step tracking is proposed which integrates the temporal integration and braking hypotheses. Since the braking of the initial response would involve the application of large forces, it is argued that braking is facilitated by the temporal integration of step stimuli. The corrective response is then implemented as rapidly as possible. The implications of these findings to the understanding of directional errors is discussed.