Abstract
Perhaps the most foundational assumption in perception research is that our experience of the world goes beyond the light reaching our eyes. This principle is familiar from visual illusions, perceptual constancies, and amodal completion — as when we perceive the continuity of an object behind an occluding surface. What is the limit to such phenomena? In particular, could we ever experience objects that are occluded completely? In fact, we frequently do: When a cloth or other fabric is draped over an object, we can often appreciate which properties the draped object might have, even when no part of the object is directly visible (Yildirim et al., 2016) — e.g., when a car is placed under a weather-protective covering, or when a child wears a bedsheet to dress up as a ghost. What is the nature of this experience? Do we only infer such hidden properties through thought and reflection? Or might we see such properties as part of automatic visual processing? Here, we investigate this latter possibility using visual priming. On each trial, one of two volumetric shapes (e.g., a cube or sphere) appeared suddenly, and subjects indicated which shape was shown. Crucially, each shape was preceded by a brief ‘cue’: A rendering of one of the shapes with a cloth draped over it. There were thus two types of trials: congruent-cued and incongruent-cued. Even though this cue was completely invalid and task-irrelevant, it facilitated subjects’ responses: They were faster to report the identity of the displayed shape when its draped cue was congruent vs. when it was incongruent. We suggest that the mind automatically represents objects even when they reflect no light whatsoever onto our eyes, such that visual processing computes the properties of completely occluded objects.