Cross-Modal Scaffolding: Music Enhances Hippocampal Binding and Separation for Visual Sequential Memory

Abstract

The human brain continuously segments experience into meaningful episodes while also encoding temporal relationships between events, yet the mechanisms that optimize this dual challenge remain poorly understood. Here we tested a theoretical framework in which structured temporal context from one modality (music) can organize such memory computations in another (visual) through coordinated modulation of the hippocampus. Using fMRI and a sequence learning paradigm, we show that musical accompaniment enhanced both boundary detection and sequential organization of visual event memory. Mechanistically, musical context accelerated development of neural responses to boundaries in hippocampus and prefrontal cortex while simultaneously optimizing hippocampal representational patterns for learning: strengthening pattern similarity for within-sequence items while reducing computational demands for discriminating representations of different sequences. Critically, musical context created conditions where contextual similarity became a stronger predictor of memory success than before, transforming similarity from an interference signal into a beneficial learning mechanism. Multivariate analysis further revealed that musical scaffolding enhanced hippocampal encoding of the sequential position of visual stimuli, demonstrating cross-modal transfer of temporal structure to visual sequence learning. Finally, we demonstrate functional specificity across hippocampal subfields, revealing how temporal structure cues can coordinate distinct computational processes within the memory circuit. These findings establish a framework for understanding how structured context signals like music can simultaneously optimize multiple aspects of memory organization, and provide mechanistic insight for educational and clinical interventions that have leveraged cross-modal temporal enhancement to improve human cognitive function.