Abstract
Prolonged viewing of ambiguous stimuli (eg. the Necker Cube) causes spontaneous change in perception. Previous studies of electrophysiological correlates used subjects' response as time reference for averaging, entailing marked latency jitter. We introduced a new paradigm as follows: We attempted to entrain spontaneous reversals to specific time instances by onset/offset presentation and further added a comparison condition with depth-shaded, non-ambiguous stimuli which reversed externally. Finally, in an attempt to involve a larger neural population we combined 9 Necker cubes into a “Necker lattice”.
Methods. 16 visually normal subjects viewed a “Necker lattice” as ambiguous stimulus for spontaneous reversal (Exp. 2) and a non-ambiguous version for externally induced reversal (Exp. 1). Stimuli were presented for 800 ms followed by a blank interval of 400 ms during which the subjects indicated whether they had perceived a reversal compared to the preceding orientation. We recorded evoked potentials from 11 scalp locations.
Results. (1) Externally induced reversal: 95% of these reversals were correctly detected. In the event related potential (ERP), the difference trace (reversing vs. non-reversing condition) showed a series of 3 components beginning with a negativity at the occiput 200 ms after onset. (2) Spontaneous reversal: Subjects reported reversals only at stimulus onset, never during the presentation interval. Again, 3 ERP components were found. They had the same succession, polarities and scalp locations as those in (1) but appeared about 60 ms later.
Conclusions. Onset/offset presentation successfully entrained reversal, allowing synchronous averaging to an endogenous event. Perceptual disambiguation seems to be accomplished by the same structures that represent objects per se, and to occur early in the visual stream. This suggests that low level mechanisms play a crucial role in resolving perceptual ambiguity.