Asymmetrical Effects of Language Mixing in Bilingual Children's Word Learning

Language mixing is a common feature of bilingual environments, yet its effects on bilingual children's word learning remain inconsistent across studies. The present study examined how language mixing affects novel word learning in 3- to 4-year-old Korean-English bilinguals, extending previous work to an understudied language pair with a greater language distance than those studied previously. Following an established experimental paradigm, children were taught novel words embedded in either single-language or mixed-language sentences during a looking-while-listening task. For half the children, carrier sentences introduced the novel words in their dominant language; for the other half, they were introduced in their non-dominant language. Overall, children successfully mapped the novel word to referents in the single-language context but showed difficulty doing so in the intra-sentence code-switched language context. However, learning patterns diverged by the dominance of the carrier language; when carrier phrases were presented in a non-dominant language, language mixing significantly affected children's novel word learning, supporting learning only in a single-language context. No difference in language mixing was observed when the carrier language was dominant. These findings partially replicate and extend prior work on the effects of language mixing on young bilingual learners' word learning, suggesting that the costs of language mixing vary across learning contexts, depending on the specific switching directionality between two distant languages.