Feature salience and presentation timing dynamically shape location-based surround suppression

Human vision feels effortless, yet beneath that ease the brain runs a continuous negotiation over which locations deserve neural resources and which must be silenced. One of the strangest outcomes of that negotiation is location-based surround suppression: when attention locks onto a target, processing is actively suppressed at nearby locations and recovers only at more distant ones, as if the spotlight of attention cast a shadow around itself. Why would a system that constantly fights for limited neural capacity deliberately mute information right beside the thing it cares about? A new study published on 27 May 2026 in the journal Attention, Perception, & Psychophysics offers a detailed answer. Across six experiments, Rui Shi of Shenyang Normal University and Liaoning Normal University and Heming Gao of Liaoning Normal University show that this inhibitory shadow is not a permanent fixture of the attentional system but a dynamic state that can be sharpened, strengthened, or erased entirely, depending on how visually salient the stimuli are and how long they remain in view.

The suppressive surround first entered the scientific record through behavioral studies. In 2000, James Mounts reported that when a color singleton draws attention, items near it suffer reduced processing, forming an inhibitory ring around the focus of attention. Three years later, Fotios Cutzu and John Tsotsos provided psychophysical evidence for a suppressive annulus around an attended object and argued that such a ring is a computational necessity: once a winning location is selected, its neighbors must be dampened so that competing signals cannot contaminate the selected one. Neuroimaging then gave the idea anatomical flesh, revealing a center–surround profile of activation in early visual cortex — a so-called Mexican hat, with enhanced response at the attended location, reduced response just outside it, and normal response farther away. Later, direct neurophysiological recordings in humans confirmed that location-based suppression surrounds the focus of attention, and follow-up studies traced the profile to recurrent, feedback-driven processing within the visual cortical hierarchy. The brain, it seemed, does not merely boost the attended location; it actively digs a trench around it.

What much of this earlier work shared was an implicit assumption that the suppressive surround is a stable property of attention — switch the spotlight on, and its ring of inhibition comes with it. Shi and Gao suspected the picture was incomplete, and they concentrated on two forces known to shape visual selection. The first is salience, the degree to which a stimulus stands out from its surroundings. Vivid, high-contrast features capture attention automatically from the bottom up, whereas voluntary, top-down deployment depends on what the observer is instructed to attend. The second is time. Sharpening the neural representation of an attended item relies on feedback signals that cascade from higher to lower visual areas, and those recurrent loops need time to act; a stimulus flashed too briefly may register in early visual cortex yet never receive the top-down refinement that full attentional selectivity requires. If surround suppression depends on either ingredient, the classical picture of a constant inhibitory ring would need revision. The researchers therefore asked whether the suppressive surround is modulated by the color salience of the stimuli and by how long they are presented.

To find out, the team built six experiments around a single elegant task. Participants kept their eyes fixed at the center of a display while colored items appeared on an imaginary circle. One item was cued, pulling attention to its location, and participants then judged whether two targets — the cued item and a second item — were identical or different. The critical variable was the distance between the two targets along the circle. If the attended location suppresses its immediate surroundings, accuracy for the second target should be poorest when it sits right beside the cue and should improve steadily as the separation grows. Experiments 1 through 3 established a baseline with low-salience colors shown briefly, then separately extended the presentation time and increased the salience of the cue. Experiments 4 through 6 added a condition with high-salience cue-targets shown for a long time, a no-cue control in which no location was cued, and — most tellingly — high-salience targets shown briefly. Accuracy and reaction times across distances served as the spatial map of attention’s allocation.

Under the baseline of faint colors and brief displays, the classical signature emerged. Discrimination accuracy for the uncued target rose gradually and systematically as its distance from the cued location increased, indicating that nearby locations fell within the cued item’s suppressive surround while distant ones escaped it. Yet under these austere conditions the suppression was incomplete: it surfaced in accuracy alone. When the researchers either prolonged stimulus presentation or upgraded the cue to a high-salience color, the surround grew stronger. In the long-time condition and in the high-salience cue condition, evidence for surround suppression appeared in reaction times as well as accuracy, with performance at near locations suffering on both measures. Because degraded performance at locations flanking the attended one is the behavioral fingerprint of active neural inhibition, the authors concluded that surround suppression becomes more pronounced when attention has more processing time at its disposal or is anchored by a more distinctive cue. The suppressive surround, in other words, is not an all-or-none filter but a graded mechanism whose expression depends on how attention is deployed.

Then came the twist. When the targets themselves were rendered highly salient in Experiments 4 through 6, the suppressive surround vanished — and it did so even under the long presentation time that had produced the most pronounced suppression in the earlier experiments. Bottom-up competition from vivid, conspicuous targets apparently overwhelmed the inhibitory mechanism that top-down guidance and extended viewing had constructed. The no-cue control condition showed no distance-dependent pattern at all, confirming that the graded accuracy effects genuinely stemmed from the allocation of attention to the cued location rather than from low-level properties of the display. The asymmetry is striking: a salient cue, which binds attention through top-down guidance, deepens the suppression around its location, whereas salient targets resist being suppressed even inside a surround that has just been strengthened. The same physical property — visual distinctiveness — pushes in opposite directions depending on whether it decorates the attended item or the one being tested, and generous processing time cannot rescue the inhibition once strong bottom-up competition takes over.

Shi and Gao fold these results into a three-factor account. Location-based surround suppression, they argue, is dynamically modulated by the interaction of top-down control, indexed by the salience of the cue; bottom-up competition, indexed by the salience of the targets; and the availability of feedback-processing resources, indexed by stimulus presentation time. When attention is guided to a location by a distinctive cue, neural gain rises there and inhibitory mechanisms silence its immediate neighborhood, a process that requires recurrent feedback to reach full strength. Strong bottom-up signals at surrounding locations, however, counteract that inhibition, because visually striking items generate responses robust enough to survive, or override, the suppressive field. The framing dovetails with the biased-competition theory of visual attention, in which objects in a cluttered scene compete for neural representation and attention biases that contest, and with imaging evidence that attending to a feature sharpens neural population tuning through feedback processing across the human visual cortex. On this view, the inhibitory surround is not painted permanently around the spotlight; it is renegotiated from moment to moment among cue, target, and time.

The framework may also tidy a stubbornly mixed literature. Studies of the attentional surround have yielded effects ranging from subtle accuracy costs to robust inhibition, and reconciling them has proven difficult. The new results suggest that stimulus salience and viewing duration — parameters that varied quietly across earlier studies — could be the hidden moderators. Brief, low-salience displays may capture only a whisper of suppression visible in accuracy alone; longer and higher-contrast conditions reveal the full inhibitory profile in both accuracy and reaction time; and designs that render the tested items themselves salient may abolish the surround altogether. The findings also resonate with evidence that the center–surround profile of attention arises from recurrent processing in visual cortex and that attentional effects propagate backward through the ventral visual stream as processing unfolds. In such accounts, timing is not a nuisance variable to be averaged away but part of the mechanism that gives the attentional field its characteristic shape.

The implications stretch well beyond the laboratory. Any task that requires monitoring objects clustered around a point of interest — a driver watching a pedestrian while a second hazard stands beside them, a radiologist screening scans in which abnormalities crowd together, a controller tracking aircraft in tight formation — involves exactly the geometry the study probed. Whether the neighbors of an attended object are momentarily suppressed or survive depends on their visual prominence and on how long they remain visible, and designers of dashboards, warning systems and displays could exploit that rule by placing critical secondary information either far from the primary focus or making it distinct enough to escape the surround. Related work has documented the same suppressive architecture in visual working memory, hinting at a shared inhibitory mechanism across perception and short-term storage. In keeping with open-science practice, the authors have made all subject-level data from all six experiments publicly available, and the project received support from the National Natural Science Foundation of China and university research funds.

Where the work goes next is nearly as intriguing as what it found. Because suppression emerges only under particular combinations of salience and time, researchers can now chart its neural timetable directly, testing whether the delayed expression of reaction-time effects corresponds to the slower buildup of feedback signals from higher visual areas. The three-factor framework also invites developmental and clinical questions: if the inhibitory surround is dynamically tuned rather than hard-wired, atypical attention — in conditions such as ADHD or autism — might partly reflect altered salience weighting or slower feedback processing rather than a broken spotlight, and training regimes might eventually teach observers to deploy suppression strategically during visual search. For now, the study delivers a conceptual upgrade to one of psychology’s oldest metaphors. The spotlight of attention still shines wherever we point it, but the shadow it casts is not painted onto the walls of the visual system. It is redrawn continuously by what we are looking for, how loud the visual world is, and how much time the brain is given to finish the job.