Competition between movement plans increases motor variability: evidence of a shared resource for movement planning

Abstract

Do movement plans, like representations in working memory, share a limited pool of resources? If so, the precision with which each individual movement plan is specified should decrease as the total number of movement plans increases. To explore this, human participants made speeded reaching movements toward visual targets. We examined if preparing one movement resulted in less variability than preparing two movements. The number of planned movements was manipulated in a delayed response cueing procedure that limited planning to a single target (experiment 1) or hand (experiment 2) or required planning of movements toward two targets (or with two hands). For both experiments, initial movement direction variability was higher in the two-plan condition than in the one-plan condition, demonstrating a cost associated with planning multiple movements, consistent with the limited resource hypothesis. In experiment 3, we showed that the advantage in initial variability of preparing a single movement was present only when the trajectory could be fully specified. This indicates that the difference in variability between one and two plans reflects the specification of full motor plans, not a general preparedness to move. The precision cost related to concurrent plans represents a novel constraint on motor preparation, indicating that multiple movements cannot be planned independently, even if they involve different limbs.

Keywords: action; motor control; movement planning; parallel encoding; reaching.

Fig. 1.

Experimental configurations. A and B: start point(s) and possible target locations (30° arcs) in experiment 1 (A), in which auditory cues indicated which of two targets to reach to, and experiments 2 and 3 (B), in which cues indicated the response hand. C: timing for the two cueing conditions. In these examples, the cue indicates a movement to the left target or with the left hand, respectively. In the early cue condition, this information is provided at the time of the first tone (and coincides with the target onset in experiments 1 and 2 and the target presented early condition in experiment 3). In the late cue condition, this information is only provided at the onset of the imperative. In the target presented late condition in experiment 3, the target was presented at the same time as the imperative.

Fig. 2.

Experiment 1: effect of number of movement plans (within hand). A and B: movement variability (A) and response times (B) for target cued early or late. C and D: movement variability as a function of response time (C) and peak speed (D) for the different conditions. Error bars indicate ±1 SE.

Fig. 3.

Probability density distributions for direction error in experiment 1 for early cue no-jump (A), late cue no-jump (B), early cue jump (C), and late cue jump (D) conditions. Direction errors to the right target are in dark gray and to the left target in light gray.

Fig. 4.

Experiment 2: effect of number of movement plans between hands. A and B: movement variability (A) and response times (B) for hand cued early or late. C and D: movement variability as a function of response time (C) and peak speed (D) for the different conditions.

Fig. 5.

Probability density distributions for direction error in experiment 2 for early cue no-jump (A), late cue no-jump (B), early cue jump (C) and late cue jump (D) conditions. Direction errors with the right hand are in dark gray and with the left hand in light gray.

Fig. 6.

Experiment 3: effect of general preparedness. A and B: movement variability (A) and response time (B) for experiment 3A, in which there was no hand uncertainty because the hand was cued early (purple), and experiment 3B, in which movements with both hands had to be prepared because the hand was cued late (orange). Solid bars indicate conditions in which there was no target uncertainty because the target was presented at the first tone (early), as in the no-jump condition of experiment 2. Hatched bars indicate conditions in which the target was presented at the time of the imperative (late).