A Corpus-based Contrastive Study Of Metadiscourse In Research Article Abstracts

Research article has traditionally been acknowledged to be accurate and impersonal, withlittle subjective opinions. In fact, the fundamental purpose of research articles is to reportresearch findings to readers, in order to persuade the reader with their theories. Metadiscourse isused to guide readers to interpret the text, to express the writer’s attitudes and to build aninteractive relationship with the reader. It is an important strategy for persuasion. Abstract, as theminiature of a research article, serves as a time-saving device for readers’ quickly capturing themain points of the full text and as an important tool to retrieve information for researchers.Recently, more and more scholars have studied metadiscourse employment in research articles.However, studies about research article abstracts are mostly from the point of genre analysis orlinguistic features such as voice, tense, etc., metadiscourse is scarcely covered.The present study selects100research article abstracts from international key journals,50by English native writers and50by Chinese native writers, covering natural science and socialscience disciplines. It aims to make two comparisons: one is the contrastive study ofmetadiscourse distribution in research article abstracts by English and Chinese native writers; theother is the contrastive study of metadiscourse distribution in research article abstracts in naturaland social science disciplines. Furthermore, this study explores the underlying reason for thecultural and disciplinary variations.In the process of manual annotation, the author takes Hyland’s taxonomy for metadiscouseas the basis, dividing metadiscourse into two types: interactional and interactive metadiscourse.The former includes transitional markers, frame markers, evidentials, code glosses andendophoric markers. And the latter consists of hedges, boosters, engagement markers, attitudemarkers, and self mentions. Self mention is then classified into direct and indirect ones.The research findings show that:1. Metadiscourse accounts for an important part inresearch article abstracts. On the whole, the number of interactional metadiscourse is larger thaninteractive metadiscourse. Transitional markers, hedges and code glosses are the most frequentlyused devices. Endophoric markers, evidentials and engagement markers seldom occur.2. Thecultural comparison shows that English native speakers adopt a lot more metadiscourse devicesthan their Chinese colleagues, especially in code glosses, self mentions and evidentials. Theunderlying reason for the differences is that English culture is more writer-responsible, orreader-oriented, relying more on the writer to guide the text and to elaborate attitudes, thus withmore writer participation, while Chinese culture is on the contrary.3. The disciplinarycomparison shows that social sciences outnumber natural sciences in overall metadiscourse employment. The subcategories of transitional markers, frame markers, attitude markers, hedgesand self mentions are especially obvious. However, hard discipline writers use a lot more codeglosses than the soft discipline writers. The differences are mainly because studies in socialdisciplines generally have seldom uniform format or standards, so writers have to depend moreon rhetorical strategies and methods.This paper sheds some new lights on understanding the nature and significance ofmetadiscourse. By comparing cultural and disciplinary differences in employing metadiscourse,this paper provides certain help to Chinese scholars who are devoted to publish papers ininternational key journals, and at the same time, it is beneficial for writers to have a basicknowledge of the linguistic and rhetorical patterns within their disciplinary communities.