Abstract
The brain is thought to process different stimulus properties in parallel. The separately processed properties have to be bound together in order for us to experience and react to unified percepts. Here, I performed an experimental manipulation of the percept durations of surfaces, and examined the effects on percept durations of corresponding shape percepts. I did this by combining binocular rivalry with the bistable perception of the Necker cube. I overlaid orthogonal diagonal plaids on the front and back planes of the Necker cube, and presented only one of these plaids to each eye. This ensured binocular rivalry in the middle square of the Necker cube. The Necker cube was presented to both eyes. I examined the effect of increasing plaid contrast – which is known to decrease percept durations for the rivaling plaid in the other eye – on percept durations of the corresponding Necker cube percepts (where the surface partially occluded through rivalry is in the back). I predicted that the Necker cube percepts would at least to some extent follow the perceived plaid percepts. Results confirmed my prediction. Surprisingly, the effects of enhanced plaid contrast on Necker cube percept durations were even stronger than those on plaid percept durations. In a control condition where plaid contrast was also enhanced, but both plaids were presented to both eyes, I found effects that resembled those obtained in the rivalrous case. In an additional manipulation of the Necker cube percept durations –by introducing converging Necker cube diagonals, clearly favouring one of both depth percepts– we found no evidence for the opposite effect, of surface-to-shape binding. We conclude that the higher contrast surface is interpreted as nearer, irrespective of whether the other plaid is suppressed and occluded or not. These findings indicate that shape binds to surface, and surface not to shape.
Meeting abstract presented at VSS 2013