Dynamic Inhibitory Control Prevents Salience-Driven Capture of Visual Attention

The salience-driven selection theory is comprised of three main tenets: (a) the most salient stimulus within a monitored region of the visual field captures attention, (b) the only way to prevent salience-driven distraction is by narrowly focusing attention elsewhere, and (c) all other goal-driven processes are possible only after the most salient item has been attended. Evidence for and against this theory has been provided from two experimental paradigms. Here, event-related potentials (ERPs) recorded in a novel Go/No-Go paradigm disconfirmed all three of tenets of the theory. Participants were instructed to search cyan-item displays for a salient orientation singleton (Go trials) and to ignore randomly intermixed yellow-item displays that could also contain an orientation singleton (No-Go trials). ERP components associated with attentional orienting (posterior contralateral N2; N2pc), distractor suppression (distractor positivity; P-D), and stimulus relevance (P2a) were isolated to test predictions stemming from the salience-driven selection theory. On No-Go trials, the salient oddball elicited a P-D rather than an N2pc, indicating that it was suppressed, not attended. Moreover, a P2a emerged before the N2pc on Go trials, demonstrating that observers first evaluated the global color of each display and then decided to search for the oddball (Go trials) or to ignore it (No-Go trials). We conclude that goal-driven processes can lead to the prevention of salience-driven attention capture by salient visual objects within the attentional window. Public Significance Statement It is important to understand how humans mitigate distraction to prevent injury and to optimize performance and productivity. Some researchers believe it is possible to ignore potentially distracting visual stimuli, whereas others believe that salient distractors invariably capture attention. This debate has continued because most evidence for or against salience-driven distraction is open to multiple interpretations. We resolve the debate by isolating electrical brain activity associated with attentional orienting, stimulus relevance, and proactive suppression in a dynamic search task that required participants to withhold responses to a salient stimulus on half of the trials. Our participants were able to decide on the fly to attend to salient visual stimuli or to ignore them so that they did not divert attention. We conclude that salience does not determine the order of attentional selection in visual tasks.