Simulating Early Phonetic and Word Learning Without Linguistic Categories
Developmental Science · 2025-03 · 28 (2)
DOI: 10.1111/desc.13606 ↗Kurz gefasst
Fachartikel in der Zeitschrift Developmental Science, online erschienen am 06.01.2025. Verfasst von Marvin Lavechin und 6 weiteren Autor:innen. Beteiligte Einrichtung: GIPSA-lab, Université Grenoble Alpes, Grenoble, France. Hauptthemen laut MEDLINE: Phonetics, Language Development. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 79 Literaturangaben. Gefördert von: MIAI, Agence de l'innovation de Défense, Grand Équipement National De Calcul Intensif.
Abstract (Originalsprache)
Before they even talk, infants become sensitive to the speech sounds of their native language and recognize the auditory form of an increasing number of words. Traditionally, these early perceptual changes are attributed to an emerging knowledge of linguistic categories such as phonemes or words. However, there is growing skepticism surrounding this interpretation due to limited evidence of category knowledge in infants. Previous modeling work has shown that a distributional learning algorithm could reproduce perceptual changes in infants' early phonetic learning without acquiring phonetic categories. Taking this inquiry further, we propose that linguistic categories may not be needed for early word learning. We introduce STELA, a predictive coding algorithm designed to extract statistical patterns from continuous raw speech data. Our findings demonstrate that STELA can reproduce some developmental patterns of phonetic and word form learning without relying on linguistic categories such as phonemes or words nor requiring explicit word segmentation. Through an analysis of the learned representations, we show evidence that linguistic categories may emerge as an end product of learning rather than being prerequisites during early language acquisition.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- GIPSA-lab, Université Grenoble Alpes, Grenoble, France.
- Laboratoire de Sciences Cognitives et de Psycholinguistique, Département d'Études Cognitives, ENS, EHESS, CNRS, PSL University, Paris, France.
- Laboratoire de Linguistique Formelle, Université Paris Cité, CNRS, Paris, France.
- IRIT, Université de Toulouse, CNRS, Toulouse, France.
- Meta AI Research, Paris, France.
ORCID-Profile: Maureen de Seyssel, Alejandrina Cristia
Förderung
- MIAI
- Agence de l'innovation de Défense
- Grand Équipement National De Calcul Intensif
- Canadian Institute for Advanced Research
- European Research Council
- J. S. Mc-Donnell Foundation
- Agence Nationale de la Recherche
Zitieren
APA
Lavechin, M., de Seyssel, M., Titeux, H., Wisniewski, G., Bredin, H., Cristia, A., & Dupoux, E. (2025). Simulating Early Phonetic and Word Learning Without Linguistic Categories. Developmental Science, 28(2). https://doi.org/10.1111/desc.13606
Vancouver
Lavechin M, de Seyssel M, Titeux H, Wisniewski G, Bredin H, Cristia A, et al. Simulating Early Phonetic and Word Learning Without Linguistic Categories. Developmental Science. 2025;28(2). doi:10.1111/desc.13606
BibTeX
@article{Lavechin2025,
author = {Lavechin, Marvin and de Seyssel, Maureen and Titeux, Hadrien and Wisniewski, Guillaume and Bredin, Hervé and Cristia, Alejandrina and Dupoux, Emmanuel},
title = {Simulating Early Phonetic and Word Learning Without Linguistic Categories},
journal = {Developmental Science},
year = {2025},
volume = {28},
number = {2},
doi = {10.1111/desc.13606}
}
EUVIMED