Segmenting words from fluent speech during infancy – challenges and opportunities in a bilingual context

Abstract

Previous research shows that word segmentation is a language-specific skill. Here, we tested segmentation of bi-syllabic words in two languages (French; English) within the same infants in a single test session. In Experiment 1, monolingual 8-month-olds (French; English) segmented bi-syllabic words in their native language, but not in an unfamiliar and rhythmically different language. In Experiment 2, bilingual infants acquiring French and English demonstrated successful segmentation for French when it was tested first, but not for English and not for either language when tested second. There were no effects of language exposure on this pattern of findings. In Experiment 3, bilingual infants segmented the same English materials used in Experiment 2 when they were tested using the standard segmentation procedure, which provided more exposure to the test stimuli. These findings show that segmenting words in both their native languages in the dual-language task poses a distinct challenge for bilingual 8-month-olds acquiring French and English. Further research exploring early word segmentation will advance our understanding of bilingual acquisition and expand our fundamental knowledge of language and cognitive development.

Monolingual and bilingual infants’ word segmentation abilities in an inter‐mixed dual‐language task

“…How do we reconcile our current results with Polka et al. (), which showed that bilingual infants have difficulty segmenting in both of their native languages? There are several methodological differences between these two studies which may explain the discrepancy of results.…”

“…Our current study shows the same type of processing with a dual-language stream of English and French, indicating that this skill holds validity in a natural language context. How do we reconcile our current results with Polka et al (2017), which showed that bilingual infants have difficulty segmenting in both of their native languages? There are several methodological differences between these two studies which may explain the discrepancy of results.…”

“…It is noteworthy to add that this pattern held, even in the face of a bilingual stream. Regardless of whether monolingual infants were tested in two language using a blocked (as in Polka et al, 2017), or a mixed language task (as in this study), the bilingual context did not appear to affect this general languagespecific pattern. Similarly, bilingual infants showed language-specific patterns of word segmentation in that they are able to segment words in both of their native languages by 8 months of age.…”

“…There are several methodological differences between these two studies which may explain the discrepancy of results. First, Polka et al (2017) tested the languages in sequential blocks, while the study here tested them in a an inter-mixed task. One possibility is that infants are better able to process speech in the way that they experience it at home (Mattock, Polka, Rvachew, & Krehm, 2010).…”

“…For this reason, much of the initial work showing language-specific effects was focused on bi-syllabic segmentation (e.g., Jusczyk et al, 1999). Polka et al (2017) tested 8-month-old infants on a dual-language segmentation task where English and French were tested in succession within a single test. Under this dual-language protocol, results confirmed that 8-month-old, English and French-monolingual infants could segment words in their native language, but not in a non-native language, regardless of which language was tested first.…”