Effects of speeds and force fields on submovements during circular manual tracking in humans.

Complex limb movements exhibit segmentation into submovements characterized as bell-shaped speed pulses. Submovements have been implicated in both feedback and feedforward control, reflecting an intermittent error-correction process. This study examines submovements occurring during a circular manual tracking task in humans, focusing on the amplitude-duration scaling of submovements and the properties of this scaling across changes in movement speed and external force load. The task consisted of intercepting and tracking a circularly moving target using a two-jointed, robotic arm that allowed external force fields to be imposed during tracking. Different speed levels and different levels of three types of force field were examined. Submovements were defined as fluctuations in the speed profile. The properties of the amplitude-duration ratio of the speed pulses were examined in relation to target speed and external force fields. The results show that the amplitude and duration of the submovements scale linearly in human manual tracking. The slope of the scaling was independently influenced by both target speed and external force fields. A common element in the increase in the scaling slope was increased tracking errors. Control experiments using passive movements and power spectral analysis showed that the submovements were not artifacts of the mechanical/acquisition system or the imposed force field. These results are consistent with the concept of stereotypy in which movements are constructed of scaled versions of a single prototype. Furthermore, the results support the hypothesis that submovements are integral to an error detection and correction control process.