Edited by Fangfang Li, Karen E. Pollock and Robbin Gibb
[Bilingual Processing and Acquisition 10] 2020
► pp. 65–86
Bilingual children process distinct phonotactics of two languages in a mixed linguistic environment. The Unified Competition Model (MacWhinney, 2005), which posits language processing via weighing informational cues, would predict that bilingual children rely more on phonotactic probability than linguistic environment when organizing new words into languages. This chapter presents a study testing this hypothesis. Using nonwords of differing phonotactic probabilities, bilingual children were tested in predominantly Spanish or English contexts and asked to classify nonwords as English or Spanish. Data analyzed using a mixed effects generalized linear model indicated that bilingual children rely on phonotactic probabilities, not linguistic environment, to encode language membership supporting predictions of the Unified Competition Model and extending its explanatory function to bilingual language membership encoding.