Acquisition Of Ba-construction In Mandarin-speaking Children With Specific Language Impairment

This thesis attempts to explore the acquisition of Ba-construction inMandarin-speaking children with specific language impairment (SLI). Specifically,both controlled experiment and spontaneous speech are adopted to examine theproduction and comprehension of Ba-construction, especially with three types ofcomplements, namely, resultative complement (RC), tendency complement (TC) andlocation complement (LC), in comparison to typically developing children. Also, theirmastery of thematic roles ‘agent’ and ‘theme’ in Ba-construction is investigated toshow whether they exhibit the same pattern as typically developing children do. Thenthe nature of specific language impairment will be discussed and explained from theperspective of Mandarin Chinese.Thirty-six children from Changsha, Guangzhou, Hefei and Xinxiang, who rangedin age from3;11to6;01(years; months), were involved in the present study. Theywere divided into three groups:12SLI subjects,12TDA (typically developingage-matched) subjects and12TDY (typically developing MLU-matched younger)subjects. All of them were selected by adopting a set of linguistic and nonlinguistictests. In the controlled experiments, given that research on Ba-construction has shownthat its typical form ‘NP1-BA-NP2-Verb-Complement’ appears around two years oldin child language, this structure was chosen as the test item. In the collection ofnaturalistic data, the process was scheduled within around ten days and for fiveone-hour sessions per child due to the rapid development of child language. Thus, onthe basis of the analysis of the data, the major findings are summarized as follows:The results from the controlled experimental data showed that SLI childrendiffered significantly from TDA children in the production, comprehension andimitation of Ba-construction, but little significant difference existed between SLIchildren and TDY children. In terms of performance on three different complementsin Ba-construction, SLI children were not as good as TDA children. Also, for SLIchildren, there was a significant difference between three different complements inthe production and imitation tasks, but not in the comprehension. The most frequentlyused complement in Ba-construction by SLI children was resultative complement. Inthe mastery of agent and theme in Ba-construction, while SLI children were similar toTDY children, they performed worse than TDA group, indicating that they were notas sensitive to the thematic roles in this structure as TDA children. Besides, instead ofthe expected responses, both SLI children and TDY children uttered more non-ba sentences than TDA children. These results were in accordance with the predictions ofFunctional Category Deficit Account and Representation Deficit for DependentRelations Hypothesis.The findings from the naturalistic data revealed that TDA children producedmore than SLI children and TDY children in the total number of Ba-construction. Andthe structure with resultative complement occupied more than that with tendencycomplement and location complement among three groups. In the production ofBa-construction by three groups, there were two major error types: one involved thosesentences where certain essential components were omitted while the other includedthose where some constrains were violated. These results indicated that in theacquisition of Ba-construction there were both the similarities and differencesbetween SLI children and typically developing children. In a word, both the analysisof controlled experimental and naturalistic data proved that SLI children haddifficulties in the acquisition of Ba-construction.