The acquisition of motion event expressions by Uyghur-Chinese early successive bilinguals

This thesis explores the implications of Talmy’s (1986; 2000a; 2000b) motion event typology and Slobin’s (1996; 2003; 2006) elaborations on it in the context of early successive bilingualism (ESB). Its main objective is to bring the study of ESB acquisition of motion expressions to bear on some of the central issues of spatial language acquisition across childhood and of bilingual language development that have thus far been predominantly concerned with the domain of morphosyntax. As an attempt to redress the Eurocentricity characterising the current landscape of bilingualism research, the study focuses on a pair of non-Indo-European languages that has interesting typological/structural overlaps and differences (Uyghur vs. Chinese). It is guided by the following two general questions: 1) what is the relative role of universal cognitive factors versus language-specific properties in the early successive bilingual’s acquisition of motion expressions and 2) how language-specific factors modulate cross-linguistic influence (if at all) in the acquisition process. Four groups of Uyghur-Chinese early successive bilingual children (AO3; aged 4;6, 6;5, 8;4 and 10;6) were invited to narrate short cartoons depicting motion situations of differential levels of complexity (voluntary vs. caused motion). Each participant performed the same task once in Uyghur and once in Chinese with an interval of about 2 weeks. Monolingual Uyghur and Chinese speakers’ motion verbalisations served as the baseline for developmental and cross-linguistic comparisons. Speakers’ descriptions were analysed in terms of 1) the linguistic devices used to encode event components (information locus), 2) the frequency with which the components are expressed (information focus), 3) the number of components speakers typically encode in the utterance (utterance density) and 4) the syntactic complexity or compactness of motion utterances (utterance architecture). Our analyses showed that bilingual children followed the lexicalisation patterns of their L1 Uyghur from the earliest stage tested. Irrespective of language and motion situations, all children produce increasingly denser utterances. Specifically, for both voluntary and caused motion situations, they used the verb-framing pattern of encoding the framing event (Path) in the main verb and the co-event (Manner/cause) (if expressed) in a converb. Bilinguals were also found to be sensitive to the patterns of lexicalisation in their L2 Chinese from very early on. Specifically, they either encoded all event components in a resultative verb compound (RVC) (equipollently-framing) or the framing event in the main verb and co-event in a gerundive (verb-framing). However, the productive use of such patterns developed over time and, to an extent, depended on the motion situations involved. In terms of voluntary motion, even 10-year-old bilinguals did not reach the adult level of utterance density and architecture in their Uyghur. Unlike Chinese children and adults, bilinguals persistently employed the verb-framing pattern until age 8 when they fully converged with target equipollent-framing pattern. With respect to caused motion in Uyghur, bilinguals matched the adult pattern of utterance density and architecture from 8 onwards. In their Chinese, they started off with the equipollent-framing pattern until the verb-framing pattern became fully established at age 10. Bilingual children’s early reliance on verb-framing patterns in their Chinese verbalisations seems to indicate a strong L1 to L2 influence whereby they seek maximal structural convergence between the two linguistic systems during online production. However, the absence of such an L1 to L2 influence in expressing caused motion suggests that cross-linguistic influence is less likely to occur when the structural overlap between the two languages concerns a syntactically more complex construction (bi-clausal) while a less complex but L2-specific structure (mono-clausal) is available. In summary, our findings show the simultaneous but differential impact of both language-specific properties and universal cognitive factors in ESB acquisition of motion expressions. The observed differences in the developmental trajectories of the two languages, especially as evidenced by children’s consistently higher utterance density in Chinese, points to the weightier role of language-specific constraints. However, the increase in utterance density, i.e. children’s ability to focus on and retain more semantic components for expression over time regardless of language indicates the contribution of their developing general cognitive abilities. The occurrence and directionality of cross-linguistic influence seems to be determined by structural/typological overlap on the one hand and the syntactic complexity of the structures involved on the other