From Child Speech to Computational Insights: Exploring the Syntax of Bilingual Language Acquisition

Abstract

The question of how to computationally model bilingual language acquisition has long been hotly debated in the fields of computer science, psychology, and linguistics. With fundamental theories still dividing the field into opposite camps, very little consensus has been reached. We explore the intricate dynamics of bilingual language acquisition in children, emphasizing syntactic patterns, errors, and code-switching behaviors between English and Spanish. Utilizing a comprehensive dataset from the Child Language Data Exchange System (CHILDES), we analyze speech samples from young bilingual children to investigate the implications of multilingual upbringing on language development. Through detailed computational analysis, we identify distinct patterns and correlations for intra-sentential and inter-sentential code-switching, providing insights into how bilingual children navigate and integrate the structural constraints of both languages. Additionally, the examination of syntactic errors reveals how bilingual children adapt linguistic structures from one language to another, highlighting potential “carry-over” effects across languages. This work contributes to the broader discourse on language acquisition by offering a novel computational perspective to better understand the complex interplay of cognitive, social, and linguistic factors influencing bilingual children. Our findings advocate for refined computational models that more accurately reflect the nuanced reality of bilingual language acquisition, suggesting pathways for future research in both theoretical linguistics and applied computational work. This path forward must better attend to sociocultural factors and incorporate analysis through an interdisciplinary lens.