A common neural mechanism for preventing and terminating the allocation of attention

Abstract

Much is known about the mechanisms by which attention is focused to facilitate perception, but little is known about what happens to attention after perception of the attended object is complete. One possibility is that the focus of attention passively fades. A second possibility is that attention is actively terminated after the completion of perception so that the brain can be prepared for the next target. The present study investigated this issue with event-related potentials in humans, focusing on the N2pc component (a neural measure of attentional deployment) and the Pd component (a neural measure of attentional suppression). We found that active suppression occurred both to prevent the allocation of attention to known distractors and to terminate attention after the perception of an attended object was complete. In addition, the neural measure of active suppression was correlated with a behavioral measure of trial-to-trial variations in the allocation of attention. Active suppression therefore appears to be a general-purpose mechanism that both prevents and terminates the allocation of attention.

Figure 1.

Example stimulus displays from the three experiments. A, Experiment 1. The target was a circle of a specific color (red in this example) that could appear at either of the two lateral locations, and participants were asked to report whether the target had a notch on its top or bottom. B, Experiment 2. Participants were asked to direct attention to the location indicated by the vertical (or horizontal) cue bar and make a button-press response to indicate whether the circle with the target color was present or absent at this location. In this example, the target color was red and the participant was instructed to attend to the location indicated by the vertical bar. C, Experiment 3. The target was a central circle with the target color (red in this example). The target-color distractor was a circle with target color at a lateral location. Participants were asked to report whether the central circle was the target color or not, ignoring the lateral circles.

Figure 2.

Illustration of permutation test. A, The positive area from the actual grand average. B, The positive area from the permutation grand average for each iteration, which should contain only noise. C, Estimated null distribution from 500 permutations of the data. If the positive area from the observed grand average (red line) falls within the top 5% of values from the null distribution (indicated by the yellow area), the observed positive area is considered to be significant.

Figure 3.

Electrophysiological results from Experiment 1. A, Grand average waveforms for targets at contralateral versus ipsilateral electrode sites (averaged over PO7 and PO8). B, Topographic maps for the target during the N2pc period (left; 180–230 ms) and the Pd period (right; 290–340 ms). Rather than plotting contralateral minus ipsilateral, which forces the values to be zero or undefined on the midline, these maps show the voltage for right-target trials minus left-target trials. C, Permutation tests of the negative and positive areas from 100 to 400 ms. The blue bars indicate the estimated null distribution from 500 permutations. The red lines represent the observed values of the negative and positive areas (N2pc and Pd) from the grand average waveforms. The yellow areas indicate the top 5% of the permutation distribution. Because the red lines fall within the yellow regions, the observed values are significantly greater than would be expected by chance.

Figure 4.

Electrophysiological results from the discrimination array in Experiment 2. A, Grand average waveforms at contralateral versus ipsilateral electrode sites relative to the cued side (averaged over PO7 and PO8), shown with a short time scale to emphasize the early P1 component. B, The same data as in A, shown on a longer time scale to show the N2pc and Pd components. C, Grand average difference waveforms obtained by subtracting the ipsilateral waveforms from the contralateral waveforms (average of PO7 and PO8).

Figure 5.

Permutation tests of negative and positive areas from the contralateral-minus-ipsilateral difference waves (relative to the cued side) for the discrimination array in Experiment 2. The blue bars indicate the estimated null distribution from 500 permutations. The red lines represent the observed values of the negative and positive areas (N2pc and Pd). The yellow areas indicate the top 5% of the permutation distribution.

Figure 6.

Electrophysiological results from the cue array in Experiment 2. A, Grand average cue-elicited waveforms at contralateral versus ipsilateral electrode sites (averaged over PO7 and PO8). B, Permutation tests of the negative and positive areas from 100 to 400 ms. The blue bars indicate the estimated null distribution from 500 permutations. The red lines represent the observed values of the negative and positive areas (N2pc and Pd) from the grand average waveforms. The yellow areas indicate the top 5% of the permutation distribution. C, The same data as in A, shown on a longer time scale and without a highpass filter to show the late sustained component. D, Permutation tests of the negative and positive areas from 400 to 1600 ms.

Figure 7.

Electrophysiological results from Experiment 3. A, Grand average waveforms for the target-color distractor at contralateral versus ipsilateral electrode sites (averaged over PO7 and PO8). The ERPs were averaged separately for trials with fast RTs (left; presumably reflecting suppression of the salient distractor) and trials with slow RTs (right; presumably reflecting capture of attention by the salient distractor). B, Permutation tests of negative and positive areas for the target-color distractor in the short-RT trials (top) and the long-RT trials (bottom). The blue bars indicate the estimated null distribution from 500 permutations. The red lines represent the observed values of the negative and positive areas (N2pc and Pd). The yellow areas indicate the top 5% of the permutation distribution.

Figure 8.

Frequency distribution of RT for the target-color distractor trials, aggregated across all participants in Experiment 3. The red lines represent the boundaries of the short-RT trials (left) and the long-RT trials (right).