Capture of attention to threatening stimuli without perceptual awareness

Visual images that convey threatening information can automatically capture attention. One example is an object looming in the direction of the observer-presumably because such a stimulus signals an impending collision. A critical question for understanding the relationship between attention and conscious awareness is whether awareness is required for this type of prioritized attentional selection. Although it has been suggested that visual spatial attention can only be affected by consciously perceived events, we show that automatic allocation of attention can occur even without conscious awareness of impending threat. We used a visual search task to show that a looming stimulus on a collision path with an observer captures attention but a looming stimulus on a near-miss path does not. Critically, observers were unaware of any difference between collision and near-miss stimuli even when explicitly asked to discriminate between them in separate experiments. These results counter traditional salience-based models of attentional capture, demonstrating that in the absence of perceptual awareness, the visual system can extract behaviorally relevant details from a visual scene and automatically categorize threatening versus nonthreatening images at a level of precision beyond our conscious perceptual capabilities.