Can Non-Conscious Representations Support High-Level Cognition?
I will present a series of studies suggesting that higher-order cognitive functions e.g., complex sequence learning, episodic retrieval, aspects of working memory, and semantic categorization are not necessarily linked to states of conscious visual awareness. Recent research suggests that non-conscious items can be maintained ‘online’ for a few seconds, and be later retrieved to guide memory-based discrimination and decision confidence. Brain activity patterns associated with these non-conscious effects are not local. Large-scale parietofrontal areas seem to support the retention of non-conscious memoranda, a network involving the hippocampus and early visual cortex mediates the recognition of non-conscious items, and canonical brain areas of the semantic network can encode the meaning of non-conscious words. On the face of it, these studies challenge the long-held theoretical link between higher-order memory functions and conscious awareness and have ramifications for neurocognitive models of memory retrieval and consciousness, such as the global neuronal workspace model. However, methodological hurdles remain to make progress in this line of research, and I will propose a potential way to move forward.