Visibility Is Not Performance: Feature-Specific Dissociations in Visual Processing

Poster Presentation 43.309: Monday, May 18, 2026, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Object recognition: Categories

Marjan Persuh1, Coryanne Coryanne Mulvey1; 1Manhattan University

A longstanding debate concerns whether visual experience is genuinely rich or whether we consciously perceive only a few detailed objects at a time. To address this question, we used a repetition-detection paradigm with very small set sizes, deliberately staying within the commonly assumed limit of three to four consciously perceived objects. The idea was that if all items should be consciously available under such minimal load, then performance limitations would reveal something deeper about the relationship between awareness and the ability to act on visual information. In Experiment 1, participants viewed arrays containing two or three items drawn from different feature domains, color, orientation, or shape, and judged whether any item repeated. In Experiment 2, the same task was used but participants additionally rated the visibility of the display. Across both experiments, performance depended strongly on feature type: color was easiest to process, whereas orientation posed the greatest difficulty. Crucially, visibility did not decrease in parallel with performance. Even when detection became harder at larger set sizes or for more demanding features, participants continued to report similar levels of visual clarity. Most importantly, visibility and task performance were not interchangeable indicators of conscious perception. In fact, the two measures sometimes moved in opposite directions. For example, at the larger set size, performance was lowest for orientation, yet subjective visibility was actually higher for orientation than for shape. This feature-specific reversal represents a dissociation: the feature that participants saw more clearly was not the one they could use most effectively for the detection task. Together, the results provide new evidence that what we consciously experience and what we can reliably report are not the same, supporting a view of visual perception that is phenomenologically rich even when behavioral performance is limited.