You reach for the life-saving glass of water handed to you from the judge’s bench, with a bit of assistance from your co-defendant.

You take a sip and the rest is movie history.

What happens to your attention during that sequence of events? When we plan a movement, for example to reach for a glass of water or a USB stick, how and when do we deploy our attention? What, if anything, do we notice while we exercise targeted movements?

It has long been known that people shift their attention to a new location before reaching for it, or before moving their eyes towards it. This attention shift can be detected by measuring attentional performance at the location of the target of a planned movement. For example, visual discrimination between two confusable stimuli (for example, the letter E and its vertical mirror image ∃) is facilitated when the discrimination target is the same object to which people are about to shift their eye gaze. By contrast, performance in surrounding locations is close to chance. Moreover, it is impossible for people to direct attention to the target while saccading to another, spatially close location. Attention deployment and movement planning thus appear to be tightly coupled.

But how does the path of a planned movement or saccade affect the dispersion of attention? If I reach for a glass to my left, is my attention spread out along the left-right axis more than if I reach for the screen in front of me?

A recent article in the Psychonomic Society’s journal Attention, Perception, and Psychophysics addressed this question. Researchers Emma Stewart and Anna Ma-Wyatt were interested in determining whether the direction of a reach determines the way in which attention facilitates perceptual performance at locations surrounding the reach target.

Their experiment included three conditions: In one condition, participants had to reach for a cued location while keeping their eyes fixed on a central location (“reach-only”). In a second condition, the reach was accompanied by a saccade to the same target (“reach-and-saccade”), whereas in the third condition there was no reach and participants just saccaded to the target (“saccade only”). In all conditions, the emphasis was on a perceptual discrimination task involving locations close to the target.

The figure below describes the procedure on a trial that involved reaching with the hand. Participants pointed at a central resting point on the screen in front of them. When a cue appeared that signalled which of the two circles to reach for, participants started moving their hand to the target. Somewhere along the way the attentional probe appeared (in this case to the west of the target) whose orientation participants had to report at the end of the trial. The timing of the probe varied between trials and was used to map out the time course of attention deployment.

Participants were encouraged to reach very quickly, with negative auditory feedback being provided if a reach took longer than 600 ms.

The location of the attentional probe varied from trial to trial but was always chosen from the ensemble of locations in the small figure below:

The letters P and O designate probe locations that were either parallel to, or orthogonal to the direction of reach, respectively. That is, to move from the center to either target one’s hand would traverse through (or near) the points labeled P1 and P2. The locations O1 and O2, by contrast, were at right angles to either direction of reach.

The procedure in the other two conditions involving a saccade (either in addition to the reach or without a reach) was nearly identical, although of course it involved monitoring of people’s eye movements. In the saccade-only condition, participants rested their finger on a key on the keyboard instead of on the screen in anticipation of a trial.

The results of the study are nuanced and complex, with a few major insights that deserve to be highlighted. The figure below shows performance (i.e., correct report of the orientation of the attentional probe) for all locations as a function of when the probe appeared relative to the onset of movement. (Because these data are relative to the onset of reach, the saccade-only condition is omitted.)

The figure shows that when a saccade is involved (blue lines), performance on the parallel probes (P1 and P2; along path of movement) is relatively modest and time invariant. Performance on the orthogonal probes (O1 and O2), by contrast, increases strikingly once the reach has commenced and becomes highly accurate. The reverse pattern is observed when no saccade is involved (purple line): performance at all locations starts out high until the movement commences, and then declines with additional reach time.

Intriguingly, when performance is expressed as a function of the delay between the cue and the probe, thus permitting analysis of the saccade-only condition, the pattern of the saccade-only condition (not shown here) was found to overlap remarkably with the reach-and-saccade condition (blue line in the above figure). That is, attentional deployment increased over time and in particular for the orthogonal probes.

Taken together, the results suggest that the planning of the saccade drives the observed attentional effects. When a saccade is present, irrespective of whether people also have to reach, the location of the probe matters and orthogonal probes attract good performance later in the trial. When no saccade is present (reach-only condition) the pattern of attentional deployment is quite different and the location of the probe matters less.

In a nutshell, whenever a saccade is involved, attention was found to peak toward the end of the movement, whereas when a reach alone was executed, attention was found to decrease across the time course of the movement. The latter finding is consonant with earlier data reported by the same authors, which also found an attentional degradation throughout a reach. It is also consonant with other recent research that we reported on this blog, which showed that people’s movement trajectories are initially affected by conflicting information that competes for their attention. Clearly, attention is required to set up and commence an accurate movement.

But even with all the attention in the world, sometimes things just go wrong.

Psychonomics article focused on in this post:

Stewart, E. E., & Ma-Wyatt, A. (2017). The profile of attention differs between locations orthogonal to and in line with reach direction. Attention, Perception, & Psychophysics, 79, 2412-2423.