Age and language experience modulate predictive processing in the visual modality

Abstract

Human language processing relies on the brain’s ability to predict structured input rapidly in time. While predictive coding has been widely studied in auditory language users, less is known about how visual language experience influences these mechanisms across their lifetime. We used electroencephalography (EEG) to examine how age and lifelong use of a signed language modulate neural prediction in the visual modality. We recorded EEG from Deaf signers of Austrian Sign Language (ÖGS) while they viewed videos of linguistic and non-linguistic biological motion. Signal complexity and neural coherence to optical flow in the stimuli were computed, and multivariate feature selection was applied to identify age-related patterns. Our results show that neural responses to linguistic motion are characterized by age-dependent increases in coherence to stimuli and changes in signal entropy, suggesting that visual linguistic experience refines predictive mechanisms over time. These effects were strongest in fronto-central and parietal regions, consistent with hierarchical sensory integration for language processing. This study demonstrates that predictive coding in the brain is shaped by modality-specific language experience and continues to evolve with age. The findings offer a new perspective on how sensory and linguistic systems interact to support inference in real-time visual processing.