Movement duration does not affect automatic online control

Abstract

Pisella et al. (2000) have shown that fast aiming movements are automatically modified on-line in response to a change in target position. Specifically, when a movement is less than 300 ms in duration the reach is completed to a target’s new location even when one never intended to respond to the target jump. In contrast, when movements are slower, the reach is completed according to instructions. At present, it is unclear if it is possible for one’s intentions to guide the initial stages of these slow movements. To determine if the intentional control mechanism can guide the initial stages of a slow aiming movement, participants aimed to targets that could jump at movement onset, with a slow and very slow movement time goal. In particular, participants were to point towards (“pro-point”) or away from (“anti-point”) the target jump, with a movement time goal of 500 or 1200 ms. Results showed that in the anti-point condition, movement trajectories first deviated in the same direction as the target jump, followed by a response in the intended (opposite) direction. This suggests that while movement outcome is controlled by the intentional system, even in these slow aiming movements the automatic system is engaged at movement onset.

Introduction

Rapid aiming movements are modified on-line in response to a change in target position in order to bring the hand to the target’s new location (e.g., Bard et al., 1999, Chua and Enns, 2005, Cressman et al., 2006, Day and Lyon, 2000, Desmurget and Prablanc, 1997, Desmurget et al., 1999, Fecteau et al., 2001, Goodale et al., 1986, Gritsenko et al., 2009, Komilis et al., 1993, Magescas et al., 2009, Paulignan et al., 1991, Pisella et al., 2000, Prablanc and Martin, 1992, Prablanc et al., 1986, Pélisson et al., 1986, Soechting and Lacquaniti, 1983).¹ It has been suggested that these adjustments in limb trajectory occur automatically (i.e., independent of one’s intentions). Specifically, Pisella et al. (2000) have argued that fast aiming movements are driven by an “automatic pilot”, which escapes intentional control. This proposal is based on their findings that for movements completed between 125 and 350 ms, the hand is adjusted and the movement completed to the new target position even when participants are instructed to interrupt their movements in response to a target jump. Similar adjustments are not observed for movements completed in less than 125 ms as, presumably, there is not enough time for the central nervous system to detect errors in the limb’s trajectory in relation to the new target location and correct the unfolding movement (Blouin, Teasdale, Bard, & Fleury, 1995).

In support of Pisella’s claim of automatic visuomotor guidance, Day and Lyon (2000; see also Johnson, van Beers, & Haggard, 2002) have also shown that fast aiming movements are rapidly modified in the direction of a target jump, independent of instructions. In their paradigm, participants were required to respond as quickly as possible to a target jump by reaching in either the same (pro-point, PP) or opposite (anti-point, AP) direction. They found that all reaches, both PP and AP, were initially modified in the direction of the target jump, even though this was an incorrect response on the AP trials. Furthermore, these changes in trajectory occurred at a similar latency (∼125 ms into the movement) for both PP and AP trials. Accordingly, and in agreement with Pisella et al. (2000), Day and Lyon inferred that these adjustments were under “automatic” control.

By examining movement trajectories, Day and Lyon (2000) were able to observe that in their AP task, participants modified their trajectories twice, first in the direction of the target jump (an incorrect modification) and then in the opposite direction (a correct modification). The second modification ensured that participants completed the movement according to their intentions and is similar to the control observed by Pisella and colleagues for slow aiming movements. Specifically, Pisella found that when participants completed slower aiming movements (>300 ms), they did not touch down at a new target position but stopped their movements in response to a target jump as instructed.

Taken together, the results of Day and Lyon, 2000, Pisella et al., 2000 demonstrate that while the automatic pilot controls fast aiming movements initially, one’s intentions can control the latter portion of a movement if enough time is provided. This observation raises the question: Can intention guide a goal-directed aiming movement for its entirety (i.e., from start to finish), or is the automatic pilot always engaged at movement onset? Given (1) the differences in movement outcome observed by Pisella et al. (2000) for fast vs. slow movements (the latter never landed at the perturbed location) and (2) the fact that intentional control took over Day and Lyon’s trajectories only after sufficient time had elapsed, it is possible that if we allow participants to make slow movements, the automatic pilot will never be engaged, a proposal that has been put forth in the literature (see Pisella et al., 2000, Rosetti and Pisella, 2002) but never examined directly. This proposal implies that slow movements are prepared differently from fast ones, such that they do not deviate until a top-down process allows them to do so. This would be in contrast to fast movements, which, as we have described above, respond immediately and automatically to target perturbations.

In the current study we determined if intention controls the initial stages of a slow aiming movement by examining trajectories to double-step targets. Specifically, we examined the trajectories of slow aiming movements (500 and 1200 ms) completed to double-step targets when participants were instructed to PP or AP in response to a jumping target. In our task, target jumps occurred at movement onset and were visible to participants. An AP movement was adopted as opposed to a stop instruction, as anti-pointing has the benefit of providing (in the form of a trajectory deviation) a clear indication of intentional control onset. Furthermore, given that some of our movements were completed with a very long movement time (MT) goal, we were concerned that if participants were instructed to stop their movements in response to a target jump, the hand may drift to the target’s new position over time, even though the participant’s first response was a correct interruption of their movement. If we do not find deviations toward the jumping target in the anti-point condition in the slow aiming movements of the present study, it would indicate that intentional control is engaged from movement onset and overrides the automatic pilot. In contrast, if the hand initially deviates towards the target before being intentionally countermanded, it would suggest that the automatic pilot is engaged during aiming, regardless of movement duration, and controls the early portions of the movement until intentional processes take over.