High-frequency Lexical Bundles In Different Sections Of Applied Linguistics Research Articles

Lexical bundles function as basic units in English and play an important role in academic discourse. They appear frequently in discourse realizing a variety of discourse functions, and are considered the cornerstone of spoken and written discourse. Corpus-driven study argues that lexical bundles are register-(or genre-) sensitive (O’Keeffe et al.,2007:61):there are both frequency differences as well as structural and functional variations between lexical bundles used in different registers. However, most of the previous studies reported on variation of lexical bundles used in different texts, and few studies probed into the intra-textual differences of lexical bundles.In the present study, a corpus-based study is adopted to investigate the use of lexical bundles in each section of Applied Linguistics research articles (RAs). A Corpus of Academic Research English in Applied Linguistics (CAREAL) has been developed by collecting100published research papers written by expert writers in English during2008to2010. The file segmentation and bundle identification were completed by the corpus analyzing software, WordSmith Tools5.0(Scott,2008). Then, analyses were conducted to describe the structural and functional features of the four-word clusters in each section of the research articles.The major findings are summarized as follows:(1) While half of the lexical bundles identified in Applied Linguistics RAs are frequently used in at least two sections of the selected RAs, the other half of them are identified as frequent lexical bundles in only one certain section. These section-bound lexical bundles can be an indicator of intra-textual stylistic variation.(2) With regard to the structural features of lexical bundles, it is found that although RAs in Applied Linguistics rely heavily on preposition-and noun-based lexical bundles, there are different preferences in bundle structures across different parts of RAs:While the Literature Review sections, Methodology sections and the Discussion sections make considerable use of prepositional phrases, the Results sections employ much more noun-based bundles. Furthermore, some structures only appear in certain parts of a research article. For example, the Literature Review and the Discussion sections prefer anticipatory it+verb/adj fragments and (verb phrase+) that-clause fragments; pronoun/noun phrase+be (+...) fragments, adverbial clause fragments, passive verb+prepositional phrase fragments and copular be+noun phrase/adjective phrase fragments only frequently appear in the Results sections.(3) In terms of discourse function, overall, more than half of the lexical bundles in RAs belong to research-oriented bundles, and the next ranks the text-oriented bundles with participant bundles being far less frequent. Further analysis reveals that variations in function distribution of lexical bundles also exist across different parts of RAs:While research-oriented bundles exceed text-oriented bundles in Methodology, Results and Discussion sections, there is a roughly even split between research-and text-oriented bundles in the Literature Review sections. In addition, participant-oriented bundles are adopted more frequently in the Results and the Discussion sections. To be more specific, lexical bundles in the Results sections are mainly reader-focused engagement bundles while most of the lexical bundles in the Discussion sections are writer-focused stance bundles.The present study suggests that the register-sensitive nature of lexical bundles exist not only across different texts but also within different parts of a text. In order to facilitate learners’ mastery of such variation within the disciplinary discourse, the corpus-based approach is recommended to both course designers and classroom language teachers.