A Corpus-based Study Of Lexical Bundles In Academic Discourse

Lexical bundles function as basic units in English language and play an important role in academic discourse. They appear frequently in discourse realizing various discourse functions and are considered as the foundation of discourse construction. Previous studies at home and abroad mostly focus on the use of lexical bundles by native learners, or on the differences of lexical bundles used by native experts and native learners, or on those between native learners and second language learners. Few concern the differences between native experts and second language experts.This study aims to explore lexical bundles-four-word lexical bundles in particular used by Chinese scholars and native scholars in the field of applied linguistics. In view of this, the present study compiles two corpora, that is, native scholar writing corpus (NSC) and nonnative scholar writing corpus (NNSC), addressing the following research questions:(1) How are lexical bundles distributed in NNSC and NSC?(2) What are the most frequently used lexical bundles in NNSC and NSC respectively? Are there any differences between them?(3) What are the characteristics of identified target lexical bundles in terms of structure and what functions do they perform?The present study is a corpus-based study. The corpora adopted in the study are self-compiled, comprising800thousand words. Computer software AntConc is used to extract and identify the target lexical bundles, including frequency and number of all bundles counted. Then the structural patterns and functional roles are categorized and analyzed, presenting a quantitative and qualitative analysis as regards the similarities and differences of the four-word lexical bundles used by native scholars and Chinese scholars.The major findings are:1) Nonnative scholars use fewer lexical bundles than native scholars in academic discourse.2) The most frequently used top20four-word lexical bundles in NNSC and NSC differ both in frequency and type and it is notable from log-likelihood ratio result that Chinese scholars overuse eleven of the fifteen shared bundles when compared with native scholars.3) In terms of structure, most of the four-word lexical bundles in the two corpora are noun phrase fragments and prepositional phrase fragments and there are more types of phrasal fragments than clausal fragments, but the percentages of each of the phrasal bundles and clausal bundles are different. In terms of function, in NNSC corpus, research-oriented bundles occupy the highest percentage, while in NSC corpus the number of text-oriented bundles is the largest, but the percentages of each major category and sub-category are also different.On the basis of the findings, the present study extends lexical bundle researches to academic discourse, which can provide guidance and reference for EAP (English for academic purposes) teaching and possible new directions for future lexical bundle research are suggested.