Is consciousness necessary for conflict detection and conflict resolution?

Is conflict control dependent on consciousness? To answer this question, we used high temporal resolution event-related potentials (ERPs) to separate conflict detection from conflict resolution in a masked prime Stroop task. Although behavioral interference effect was present in both the masked and unmasked conditions, the electrophysiological findings revealed more complex patterns. ERP analyses showed that N450 was greater for incongruent trials than for congruent trials and that it was located in the ACC and nearby motor cortex, regardless of whether the primes were masked or unmasked; however, the effects were smaller for the masked than unmasked condition. These results suggest that consciousness of conflict information may not be necessary for detecting conflict, but that it may modulate conflict detection. The analysis of slow potential (SP) amplitude showed that it distinguished incongruent trials from congruent trials, and that this modulation effects was reduced to a greater extent for the masked condition than for the unmasked condition. Moreover, the prefrontal-parietal control network was activated under the unmasked but not under the masked condition. These results suggest that the consciousness of conflict information may be a necessary boundary condition for the subsequent initiation of control operations in the more extended PFC-parietal control network. However, considering that the conflict interference effect was significantly reduced in the masked condition, it may be that, with larger unconscious conflict effects, more extensive cognitive control networks would have been activated. These findings have important implications for theories on the relationship between consciousness and cognitive control.