Recall of overlearned speech in noise: Developmental ERP findings and an episodic buffer frame for culturally situated brain activity

Abstract

How does extensively practiced speech memory remain dependent on momentary sensory input, and how does this interaction shape brain activity during development? This perspective integrates exploratory behavioral and event-related potential (ERP) findings from a culturally situated spoken-sequence recognition paradigm with a working memory episodic-buffer account. We reanalyzed data from children and young adults who had memorized Quranic passages despite limited Arabic comprehension. Participants heard incomplete passages with or without static noise and selected the missing final word from phonologically plausible alternatives. Noise reduced recognition accuracy, and children generally responded more slowly. Descriptive ERP patterns revealed larger, more spatially distributed responses in children, spanning early auditory processing, mid-latency positivity, and later negative activity. We propose that recognition requires temporary binding of acoustic input, serial and rhythmic expectations, visual response cues, and consolidated sequence knowledge within an episodic buffer under executive control. Noise may impair perceptual fidelity and matching between transient input and stored representations. Developmental ERP differences may reflect representational coordination rather than neural efficiency. These preliminary findings motivate testable predictions concerning cross-modal congruence, familiarity, comprehension, and independently measured expertise. Culturally embedded learning practices provide informative contexts for examining interactions among working memory, long-term knowledge, and developing neural systems supporting cognition.

The authors' abstract, as published at the source. Brain and Cognition, 2026 · DOI ↗