Resonant cortical dynamics of speech perception

Abstract

What is the neural representation of a speech code as it evolves in time? How do listeners integrate temporally distributed phonemic information into coherent representations of syllables and words? How does the brain extract invariant properties of variable-rate speech? This talk describes a neural model that suggests answers to these questions, while quantitatively simulating speech and word recognition data. The conscious speech and word recognition code is suggested to be a resonant wave, and a percept of silence a temporal discontinuity in the rate that resonance evolves. A resonant wave emerges when sequential activation and storage of phonemic items in working memory provides bottom-up input to list chunks that group together sequences of items of variable length. The list chunks compete and winning chunks activate top-down expectations that amplify and focus attention on consistent working memory items, while suppressing inconsistent ones. The ensuing resonance boosts activation levels of selected items and chunks. Because resonance occurs after working memory activation, it can incorporate information presented after intervening silence intervals, so future sounds can influence how we hear past sounds. The model suggests that resonant dynamics enable the brain to learn quickly without suffering catastrophic forgetting, as described within Adaptive Resonance Theory.