Poster #113 - The Development of Translation Equivalents in Bilingual Toddlers’ Vocabulary

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Language, Communication

Integrative Statement

Bilingual children are unique in that they must learn a word in each of their languages for each concept – called translation equivalents (TEs, Pearson et al., 1993). TEs are important as they provide evidence of the independence in which each vocabulary develops in bilinguals (Pearson et al., 1995; De Houwer et al., 2006), and impact bilinguals’ use of mutual exclusivity (Byers-Heinlein et al., 2013). Thus, examining the proportion of children’s vocabularies that have a TE – and the variables that predict this – is important to better understand the development and composition of bilinguals’ early lexicons.

Previous estimates of how many TEs bilingual children know have varied: 18% at 13 months (N = 31; De Houwer et al., 2006), 46% at 18 months (N = 20; Byers-Heinlein & Werker, 2013), and 37% at 24 months (N = 25; Poulin-Dubois et al., 2012). Another study (N = 13) reported an increase in TEs from 0% to 80% from 16 to 30 months (David & Wei, 2008). There are several potential explanations for these divergent findings: studies differed in whether comprehension or production was studied, they had relatively small samples, and differed in participants’ age, language profiles (e.g. more vs. less balanced exposure), and language pairs being learned.

To gain traction on these issues, the current study examined the development of TEs in the production of a large sample (N = 279) of bilingual toddlers learning English and French, aged 18-33 months, which allowed us to investigate the effects of age and language dominance in a linguistically uniform sample. Infants were exposed to each language 15% to 85% of the time according to a structured parent interview (Bosch & Sebastian-Galles, 2001; Byers-Heinlein et al., 2018). Vocabulary was measured using the Words and Sentences version of the MacArthur-Bates Communicative Development Inventories in American English (Fenson et al., 2007) and Canadian French (Trudeau, Frank, & Poulin-Dubois, 1999). TEs were calculated by comparing the words produced that referred to the same concept but were lexicalized in two languages (e.g., dog in English and chien in French are TEs). Following Byers-Heinlein and Werker (2013), percentage of TEs was calculated by dividing the total number of words with a translation equivalent by total vocabulary size (L1 + L2).

Across the entire age range, children showed a remarkably stable proportion of TEs in production, 33% on average, with substantial individual variability (range: 0% - 82%). Regression analyses showed no evidence of linear, quadratic, or cubic effects of age (ps > .05). Children with highly unbalanced exposure to their languages (80-20 group; 75-85% exposure to the dominant, and 15-25% exposure to the non-dominant) knew a smaller proportion of translation equivalents than children with more balanced exposure, F(3, 271) = 17.605, p < .001. Findings from this large sample reveal that the proportion of TEs does not appear to vary with age in this developmental period. Furthermore, stark differences in TEs knowledge of unbalanced as compared to balanced bilinguals suggest a qualitatively different path of lexical development.

Authors

• Ana Maria Gonzalez-Barrero, Concordia University
  Presenting Author

• Esther Schott, Concordia University
  Non-Presenting Author

• Krista Byers-Heinlein, Concordia University
  Non-Presenting Author