Bilingual morphological acquisition

Project title: Cross-linguistic transfer and morphological salience on English morpheme acquisition by Chinese-English and Spanish-English bilingual children

Typically developing bilingual children’s acquisition of English morphemes is attributed to child-internal factors such as cross-linguistic influences of the first (L1) or heritage language (e.g., Murakami & Alexopoulou, 2016; Paradis, 2011) and child-external factors such as language exposure (e.g., Babatsouli & Nicoladis, 2019; Bedore et al., 2018). In addition to these, the properties of the morphemes (e.g., morphological salience) also affect the morpheme acquisition. To illustrate, morphological salience factors include perceptual salience, semantic complexity, morphophonological regularity, syntactic category, and frequency, which were shown to account for 71% variance in L2 learners’ morphological accuracy (Goldschneider & DeKeyser, 2001). However, its effect on bilingual children’s morpheme acquisition has not been extensively studied. The present study thus extends this line of research by examining the effect of cross-linguistic influences of L1 (Chinese vs Spanish) and morphological salience on bilingual children’s accurate and error productions of six developmental English morphemes: present progressive -ing, plural -s, third-person singular -s, possessive -’s, articles, and regular past tense -ed.

Cross-sectional data were obtained from CHILDES (MacWhinney, 2000) and includes 48 transcripts of naturalistic conversations from two groups of bilingual children (n = 26 Chinese-English (CE), n = 22 Spanish-English (SE)). These children are age-matched and gender-balanced with a mean age of 4 years and 3 months (range = 3;0 - 5;11). Children’s morpheme production was measured using the Target-Like Use (TLU) for each child’s first 100 intelligible utterances. Their error productions were coded either as an omission error when there is no production of a morpheme in an obligatory context, or as a commission error such as double marking.To evaluate the effect of crosslinguistic influences, Multivariate Analyses of Covariances (MANCOVAs) were conducted with Language Group as a fixed factor and Age as the covariate. The analyses revealed a significant main effect of Language Group (F (2, 46) = 3.611, p = .006, partial η² = .351) and Age (F (2, 46) = 2.917, p = .019, partial η² = .304) on the TLU scores. Specifically, SE bilinguals were more accurate in third person singular -s, past tense -ed, and plural -s than CE bilinguals. Besides, CE bilinguals produced more omission errors than the SE bilinguals (F (1, 47) = 4.60, p = .037, partial η² = .09). As regards the effect of morphological salience, the Pearson correlation analyses showed that morphophonological regularity is correlated with the omission errors for both groups (r = . 89 for CE; r = . 81 for SE; p ≤.05). Moreover, the regression analyses showed that syntactic category is a significant predictor accounting for 79.6% of variances of TLU scores by SE bilinguals and 84.6% by CE bilinguals.

These results disclose the important roles of crosslinguistic influences from the L1 and morphological salience, particularly syntactic category in L2 morpheme acquisition by typically developing bilingual children.

• This project is being conducted at the DILL Lab by Xuan Wang in collabration with Yan Shi and Dr. McMillen.