Abstract
We used event-related potentials (ERPs) to study the neural electrical activity associated with conscious perception of targets in a meta-contrast masking paradigm in which the detection of a target is inhibited by a subsequent mask with adjacent but non-overlapping contours. It has been proposed that early visual activity elicited by these stimuli is fed forward to higher brain areas, and that in order for the perception to reach consciousness, these higher brain areas must then feed this information back to the early visual areas. This theory further proposes that when these feedback signals from higher areas encounter the new activity elicited by the mask instead of the original target activity, the conscious perception of the target is inhibited. In the current experiment, after a fixation cue, target and mask pairs were presented using an SOA chosen so that observers reported perceiving roughly half of the masked targets. ERPs for trials where observers reported seeing the target were compared with those for unseen targets. We examined not only the differential activity elicited by detected and undetected targets, but also differential preparatory activity prior to the target onset that might predict subsequent target detection. For target locked ERPs, there was a difference in visual ERP components between detected and undetected targets starting with the N2 component, but no difference was evident in the earlier P1 and N1 components measured at occipital sites, consistent with the reentrant theory. Interestingly, preliminary analyses suggest that the ERPs elicited by the fixation cross preceding the targets were different for detected and undetected targets, implying that brain states preceding the target appearance influenced the likelihood of its detection. These results further support the assertion that attention modulation and feedback activity play a role in the efficacy of meta-contrast masking.