Discourse-grammatical features in L2 speech
: a corpus-based contrastive study of Chinese advanced learners and native speakers of English

  • Xiao CHEN

Student thesis: Doctoral Thesis

Abstract

It has been noted that even advanced EFL (English as a Foreign Language) learners who have sufficient grasp of the rules of grammar may use grammatical features in non-nativelike ways that cannot be simply categorized as grammatical errors or inappropriate English at the level of the clause or the sentence. Deviations emerge in the form of distributional patterns in extended discourse. They often involve features at the syntax-discourse interface, referred to as discourse-grammatical features in this thesis. Problems with these features are harder to detect than grammatical errors and easier to overlook in pedagogical grammar. The objective of this study, therefore, is to take a systematic, empirical look at how discourse-grammatical features in L2 speech differ from those in native speaker speech. The study is a corpus-based contrastive analysis of task-homogenous speech by Chinese advanced learners and native speakers of English selected respectively from the Chinese component of LINDSEI (The Louvain International Database of Spoken English Interlanguage) and LOCNEC (The Louvain Corpus of Native English Conversation). The investigation focused on Theme, person reference, and tense-aspect, which are three of the most problematic areas for Chinese learners identified through analytic induction and keyword analysis of the research corpus. The core of the investigation was a corpus inquiry into the frequency and distribution of different grammatical forms for Theme, person reference, and tense-aspect in learner and native speaker speech. Task type and proficiency level were also analyzed to determine if some of the marked usage is influenced by these variables while others are constant features in the L2 speech as a whole. The corpus analysis was conducted through multi-level annotation and statistical comparison of the research data. Quantitative findings were complemented by qualitative text analysis. Possible L1 influence at the levels of syntax, discourse and genre schema was carefully considered on the basis of previous research on Chinese as well as evidence from the additional post-task Chinese and English interviews. A further, pedagogical level of analysis was provided through the examination of how salient features indentified through corpus investigation are treated in a variety of widely used EFL textbooks in mainland China. The study demonstrates a theoretical and methodological shift from a sentence-based, categorical approach to a discourse-based, probabilistic approach to grammar in L2, while pushing forward the empirically-based understanding of the actual challenges Chinese learners are confronted with when using grammar in spontaneous discourse. The study develops the theoretical and empirical bases for teaching discourse grammar, showing how this can be a means of keeping grammar learning alive even at the advanced level. It also contributes to our knowledge of the effect of schema and task on linguistic forms and presents important pedagogical and methodological implications.
Date of Award4 Oct 2010
Original languageEnglish
Awarding Institution
  • City University of Hong Kong
SupervisorJohn FLOWERDEW (Supervisor) & Yong Wey David LEE (Co-supervisor)

Keywords

  • Second language acquisition
  • Corpora (Linguistics)
  • Grammar, Comparative and general
  • Syntax
  • Discourse analysis

Cite this

'