Perceptual salience captures the eyes on a surprise trial

A number of characteristics of the visual system and of the visual stimulus are invoked to explain involuntary control of attention, including goals, novelty, and perceptual salience. The present experiment tested perceptual salience on a surprise trial, that is, on its unannounced first presentation following trials lacking any salient items, thus eliminating possible confounds by current goals. Moreover, the salient item's location was not singled out by a novel feature, thus eliminating a possible confound by novelty in directing attention. Eye tracking was used to measure involuntary attention. Results show a prioritization of the salient item. However, contrary to predictions of prominent neuro-computational and psychological salience models, prioritization was not fast-acting. Rather the observers' gaze was attracted only as the second fixation on average or later (depending on condition) and with a latency of more than 500 ms on average. These results support the general proposition that salience can control attention. However, contrary to most salience models, the present results indicate that salience changes attentional priority only in novel environments.