Development of Mandarin by English-Mandarin bilingual children in a Chinese childcare centre : the role of input

Abstract

This study investigates the development of Mandarin in English-Mandarin bilingual children compared to their monolingual peers in a childcare centre in mainland China. It also examines the qualitative and quantitative aspects of input of the bilingual children's Mandarin development. Many previous studies examined bilingual children acquiring two languages in one-parent-one-language input conditions (e.g. De Houwer, 1990; Döpke, 2000; Meisel, 1990a; Paradis & Genesee, 1996; Ronjat, 1913; Silva-Corvalán, 2014; Thordardottir, 2014; Unsworth, 2014; Volterra & Taeschner, 1978), and most focused on children developing two Indo-European languages (cf. W Li, 2010; Yip & Matthews, 2007). However, the most common situation of children growing up bilingually takes place in immigration contexts where they are exposed to one language one environment mode (Qi, Di Biase, & Campbell, 2006). To date, this type of bilingual development has not received much attention. In recent years a growing number of native English-speaking people came to work or study in mainland China and their children became bilingual in context-bound one-language-one-environment situations, similar to most other children growing up in immigrant families. This means these children acquire English at home and Mandarin elsewhere e.g., at childcare centres. The effect of teachers' input at childcare centres on the mainstream language development of bilingual children has rarely been studied. Research questions thus follow: a) How do these bilingual children develop their Mandarin in the childcare centres? b) What role does input from teachers play in these children's Mandarin development? In order to address these questions, I carried out a multiple case study on seven three-to-five-year-old English-Mandarin bilingual children in a Chinese childcare centre. The main data includes a corpus of recordings of speech produced by the bilingual children, their monolingual peers and their teachers in the same childcare centre in the course of various activities over a four-month school term. I also collected supplementary data supplied by parents and teachers of the bilingual children, including questionnaires and interviews with them. In addition, some elicited comprehension and production tasks were proposed to bilingual children and their monolingual peers to complement the corpus data. Two linguistic domains, namely, Mandarin noun classifier and prepositional phrase (PP) with zà i 'at', are targeted for investigation as they manifest significant typological differences from the bilingual children's other language -English. For each domain of investigation, three types of comparison between bilingual children and their monolingual peers are focused on: (a) whether there exists different or similar patterns of acquisition, (b) whether different input conditions result in different patterns of acquisition, and (c) whether the input from teachers influences acquisition pathways in these two domains. Results reveal that bilingual children show a pattern similar to their monolingual peers in the acquisition of Mandarin noun classifiers. Both bilinguals and monolinguals predominantly use (and overuse) the general classifier gè while use of specific classifiers is rare. Children in either group rarely omit an obligatory classifier. Bilingual children's patterns of classifier acquisition are not as variable as their input, measured by the cumulative length of Mandarin exposure (CLME). Moreover, the pattern of teachers' use of classifiers appears to significantly influence both bilingual and monolingual children's acquisition within this domain. Firstly, that specific classifiers are quite rarely used by children is reflected in the input from teachers. Secondly, teachers never omit a classifier when it is obligatory, which may help children to know the obligatory use in practice. Thirdly, cases of children's overuse of the general classifier gè have also been found in teachers' productions, although the rate of overusing is much lower than that of the children. However, in the domain of Mandarin locative PP headed by zà i 'at', it is found that bilingual children follow a different pattern compared to their monolingual peers. Bilinguals show a strong preference for postverbal locative PP with zà i, but preverbal and postverbal zà i-PPs are equally divided in monolinguals. Moreover, about half of the postverbal zà i-PP utterances produced by bilinguals are non-target. In sharp contrast, non-target postverbal zà i-PP sentences were not observed in the monolingual children's productions. Comparison among the bilinguals found that although children with a larger amount of Mandarin exposure generally develop to a more advanced stage than others, those who use Mandarin more often do not lag behind even when their exposure time is less than others. Results from the analysis of teachers' input show that teachers' frequency of use of postverbal zà i-PP may influence the children's production in this domain. However, teachers' use of zai-PPs has consistently shown to be target-like. The results show that there is a possibility that bilingual children's non-target placement of zà i-PPs reflects cross-linguistic influence from the structure of the English prepositional phrase. Findings from this research will offer new insights about language contact and interaction in bilingual development. They will also shed light on the nature of input in the challenging aspects of bilingual children's linguistic development.

Date of Award	2018
Original language	English