ABSTRACT STRUCTURE, NOT SENSORY EXPERIENCE, DRIVES LEARNING AND NARRATIVE FORMATION

Abstract knowledge derived from past experiences guides the brain's ability to process new information and creatively reshape it, e.g., into novel narratives. Neural replay during rest has been identified as a mechanism for consolidating and reprocessing this information. Recent data suggest that neural replay can be detected with magnetoencephalography in humans. In humans, as in rodents, replay occurs in accelerated sequences, reorganized based on abstract knowledge rather than simple replication of sensory experiences. This mechanism likely supports generalization of learned frameworks to novel contexts, enabling flexibility and creativity. To establish behavioral effects of structural learning prior to our MEG study, we ran a behavioral study to characterize how learned abstract knowledge not only restructures memory but also supports narrative construction. Specifically, we quantified the contribution of abstract structural learning to participants’ accuracy in extracting ordered sequences, as well as its effect on how they subsequently organize objects into self-generated narratives. Participants first learned a sequential rule defining a target order of images that needed to be extracted (with the help of cues) from a scrambled presentation order. Next, they constructed their own narratives by organizing objects into self-generated event sequences. Consistent with our hypothesis, abstract structural knowledge acquired during learning facilitated participants’ performance when they were later presented with entirely new stimuli. Moreover, the narratives participants generated were systematically shaped by the learned structure, reflecting an internalized relational framework rather than a simple reproduction of sensory experiences. Next, our MEG study will examine whether abstract knowledge reshapes resting neural replay.