On-line control of rapid aiming movements: Unexpected target perturbations and movement kinematics.

Abstract

Abstract Two experiments were conducted to assess the use of on-line visual information during rapid goal-directed aiming movements. In both experiments, participants were required to complete a discrete aiming movement to a single target on a graphics tablet. In Experiment 1, during 76% of the experimental trials, the target width remained constant throughout the movement. The remaining 24% of the trials were evenly divided between two target perturbations in which the width of the target unexpectedly increased or decreased in size upon movement initiation. During Experiment 2, the spatial location of the single target was perturbed to a new location closer to, or further away from the original target position. The proportion of perturbation trials remained constant across experiments. The results indicated that peak velocity was determined prior to movement initiation in order to meet the speed-accuracy demands of the original target width or movement amplitude. In contrast, during deceleration, participants modified their movement trajectories to account for a perturbation in target width or movement amplitude. These data suggest that on-line monitoring of visual information can be used to modify the latter half of a movement trajectory. The issue of how goal-directed aiming movements are controlled has been controversial since the seminal studies of Woodworth (1899) almost a century ago. A major concern has been whether the of aiming movements is a continual process as feedback-based models posit (i.e., Adams,1971), whether they are planned wholly in advance of execution (i.e., planned control; Plamondon, 1995a, 1995b; Schmidt, Zelaznik, Hawkins, Frank, & Quinn,1979), or if the process is some combination of the two (Beggs & Howarth, 1970; Crossman & Goodeve, 1963/1983). Woodworth (1899) originally coined the terms and control to describe simple aiming movements as a combination of both central and feedbackbased processes. Specifically, an initial impulse is programmed before the movement begins to bring the end-effector close to the target. Current is evoked to correct for error in the initial movement trajectory if response-produced feedback indicates that the limb will fall outside of the target boundaries. Thus, this second phase of the movement is dependent on visual and proprioceptive information about the position of the end-effector in relation to the target. According to Woodworth, it is the current phase that is responsible for the speed-accuracy relation observed during rapid goaldirected aiming. Although this original formulation by Woodworth has undergone numerous transformations, the important components of the model still feature prominently in current explanations of limb (i.e., Abrams, Meyer, & Kornblum, 1990; Beggs & Howarth, 1970; Carlton, 1979, 1981, 1992; Chua & Elliott, 1993; Meyer, Abrams, Kornblum, Wright, & Smith, 1988). A good model of limb should not only explain the well-known relation between speed and accuracy (Fitts, 1954; Fitts & Peterson,1964), but also how the characteristics of the aiming trajectories will change with the accuracy demands of the movement and the availability of task-relevant information. An examination of movement kinematics indicates that increases in movement time with increasing accuracy demands are primarily due to an increase in the time spent following peak velocity, when the limb is decelerating and executing any adjustments required to hit the target (i.e., Langolf, Chaffin, & Foulke, 1976; MacKenzie, Marteniuk, Dugas, Liske, & Eickmeier, 1987). Presumably, this extra time is required to process and use visual and proprioceptive feedback about the relative positions of the limb and target in order to reduce any error associated with the initial movement impulse (Chua & Elliott, 1993). This explanation of the dynamics of movement trajectories is based partly on findings which indicate that when vision is occluded upon movement initiation, less of the total movement time is spent after peak velocity, compared to situations in which vision is available over the entire course of the movement (Chua & Elliott, 1993; Elliott, Carson, Goodman, & Chua, 1991). …