Why suppressing two distractor features is harder than suppressing one
Think of distractor rejection as a filter tuned to a specific property, such as a particular color or shape. When only one distractor feature needs to be filtered out, that tuning can be sharp and effective. The 2016 study by Kerzel and Barras specifically set out to find the limits of distractor rejection after many encounters with distractors, and reported that rejection breaks down when more than a single distractor feature must be suppressed [1]. In other words, the filter works well for one feature but loses its grip when it has to cover several.
That fits a broader picture in which rejection is not one single switch but can depend on distinct mechanisms. One 2026 study describes a feature-specific form of distractor rejection that is sensitive to how often a stimulus appears and is relatively short-lived, alongside a separate feature-general form [4]. If rejection is partly tied to specific features, then asking it to handle multiple features at once could strain the system in a way that a single feature does not.
What rejection looks like when it works
The clearest demonstrations come from experiments where a distractor is made predictable. In one set of experiments, participants searched for a target shape while displays contained two colors; when a consistent distractor color was present, people found the target faster than when it was absent, showing that learned rejection of that color was operating [3]. Notably, this happened even though attention was being guided by shape, a different feature dimension, which suggests learned rejection and target guidance can run on separate tracks [3].
Rejection can also be driven by knowledge rather than repeated exposure. When participants were told to suppress one of two object categories (teddy bears or butterflies), responses on trials containing an item from the to-be-suppressed category were about 119 milliseconds faster than on neutral trials, and eye movements lingered less on the suppressed distractors [2]. That is a large, measurable benefit, and it shows that top-down categorical knowledge can support rapid rejection of a non-target [2].
When the limit shows up: location versus feature, and one versus many
Not every kind of suppression survives multiple distractors. In a task where a target appeared at a location previously occupied by a distractor, responses were delayed, reflecting lingering inhibition of that spot. That effect was significant only when a single distractor had been present; with multiple distractors, the effect was not reliably observed, and follow-up experiments explored possible reasons [5]. This is a spatial, location-based form of inhibition rather than suppression of a feature like color, so it does not by itself settle the feature question, but it is a concrete case where going from one distractor to several changed the outcome.
Put together, the evidence suggests the breakdown is not simply about the number of objects on screen. Rejection held up when the demand was to suppress one learned color while searching by shape [3], and when the demand was to suppress one of two categories [2]. The reported failure came when more than one distractor feature had to be suppressed [1]. The distinction that matters, then, is how many distinct features the rejection system must represent, not merely how cluttered the display is.
What this means for ignoring clutter in everyday life
A practical reading is that the brain is good at learning to ignore one consistent nuisance, such as a single recurring color or a familiar category of object, and can do so quickly and even when attention is focused on something else [2][3]. The difficulty rises when several different nuisance features compete for suppression at the same time [1]. That is consistent with the idea that some rejection is feature-specific and short-lived while other rejection is more general [4], though the exact division of labor between those mechanisms is still being worked out.
One caveat belongs right next to these claims: much of this evidence comes from controlled search tasks with briefly presented displays, and the strongest statement about breakdown concerns the specific conditions tested in the 2016 study [1]. The category experiment used only two categories and a single suppression instruction [2], and the color experiment used one consistent distractor color [3]. So the safe conclusion is that multi-feature suppression is where rejection is most likely to falter, not that it always fails.
Sources used in this answer
Distractor rejection in visual search breaks down with more than a single distractor feature.
Distractor rejection, studied after many encounters with distractors, breaks down when more than a single distractor feature must be suppressed.
Distractor suppression: Can top-down knowledge of non-target categorical information elicit rapid rejection?
Participants told to suppress one of two object categories rejected that category rapidly, with response times about 119 ms faster than neutral trials and shorter dwell times on suppressed distractors.
Learned distractor rejection persists across target search in a different dimension
Learned rejection of a consistent distractor color sped target search even when attention was guided by a shape-based target template in a different feature dimension.
Distinct underlying mechanisms contribute to learning-dependent distractor rejection
Learning-dependent distractor rejection appears to involve distinct mechanisms, including a feature-specific form sensitive to stimulus frequency and relatively short-lived, alongside a feature-general form.
The inhibitory effect of a recent distractor: singleton vs. multiple distractors
Inhibition of a recently occupied distractor location was significant with a single distractor but not reliably observed when multiple distractors were presented.