id_688. NEONATAL LEARNING OF SPEECH SOUND PATTERNS
András Ambrus1,2, Silvia Polver1,3,4, Brigitta Tóth1, Gábor Háden1
1 HUN-REN Institute of Cognitive Neuroscience and Psychology, Research Centre for Natural Sciences, Magyar tudósok körútja 2, Budapest, 1117, Hungary
2 Department of Cognitive Science, Faculty of Natural Sciences, Budapest University of Technology and Economics, Budapest, Hungary
3 GRAMFC, Inserm UMR1105, Université de Picardie Jules Verne, Amiens, France
4 Department of Developmental Psychology and Socialisation, University of Padua, Via Venezia 8, Padua, 35131, Italy
INTRODUCTION: The automatic capacity to extract, encode and utilize statistical properties of the ever changing sensory environment, known as statistical learning, is a fundamental ability already present from birth. Detecting longer recurring auditory patterns consisting of as many as 10 pure tones is proved to be present in newborn babies. Whereas pure tones only provide information about pitch transitions, speech includes a wealth of acoustic information.
AIM(S): This study investigated whether sleeping newborns detect regularities in speech-like sound sequences.
METHOD(S): Using time-locked EEG measures, we compared 35 newborn babies' (up to 2 days of age) neural responses to regularly recurring and random pseudo-syllable sequences.
RESULTS: We found that neonates show distinct electrophysiological responses to regular versus random patterns, indicating that the neonatal auditory system is sensitive to structured syllable sequences.
CONCLUSIONS: These findings suggest that even at a very early age, infants can exploit redundancies in speech input, an ability underpinning later language acquisition. The present results provide an important step toward identifying early neural signatures of pattern learning in speech. Such responses may serve as potential biomarkers of language learning capacities in early infancy, offering a window into the developing mechanisms that support speech segmentation and grammar acquisition. Beyond their theoretical relevance, these results underscore the methodological value of the current paradigm: it provides a robust, non-invasive, and replicable tool for probing the infant brain’s capacity to extract structure from continuous auditory input. This paradigm can be extended to assess individual differences and atypical trajectories, contributing to the early identification of infants at risk for language disorders.
FINANCIAL SUPPORT: BIO-PREPA (2024-1.2.2-ERA_NET-2025-00023) Biomarkers and Behavioral Probes for Preclinical Perinatal Asphyxia Funding: ERA-NET NEURON Cofund2, Horizon 2020 Grant Agreement No. 964215 Principal Investigator and Project Coordinator: Brigitta Tóth