Comparative Analysis Of The Metadiscourse Devices In Sino-American College Students' Argumentative Compositions

Language is a tool for human communication. In any form of language communication, two levels of discourse are involved: the primary discourse level which consists of propositional meaning and referential meaning, and the metadiscourse level which consists of propositional attitudes, textual meaning and interpersonal meaning. On the primary discourse level, we supply information about the subject of our text and expand propositional contents; and on the metadiscourse level, we project ourselves into text guiding and directing our readers to organize, interpret, evaluate and react to our propositional contents.Nowadays, metadiscourse is attracting the attention of more and more linguists. And studies on metadiscourse show that metadiscourse plays an essential role in second language learning, especially in second language writing.Although many scholars have done sufficient researches on how to improve students' writing, they only pay attention to how to teach students the right grammar, how to use vocabulary in the right way, how to write a complete sentence. Besides, although some researchers have converted to teaching students to write in the discourse level, which focuses on the generation of contents, they are only working on an ideational dimension of writing.All these methods seem not to be so effective because most of scholars confine their researches to some specific language skills and foundation, which aim at a writer only and neglect the other important aspect of composition—a reader, because interaction is inherently a two-way process. Ignoring any respect, either the reader or the writer, may make the interaction less smooth, or even break the interaction.The present paper investigates the use of metadiscourse devices in the compositions written by Chinese college students and the ENL (English as native language) ones written by American college students. Metadiscourse devices are analyzed considering their number and function by using quantitative methods. The objective of this study is to find similarities and differences in the use of metadiscourse devices by Chinese and English-speaking writers of college level. With this comparison the author hopes to know if either of the two groups use metadiscourse devices more frequently, and to know the types of metadiscourse devices Chinese college students use so that better advice is given to guide and improve classroom instruction in the teaching of metadiscourse devices in China EFL context.The present study intends to address: 1) What are the overall features of the use of metadiscourse in the two types of English essays? 2) Are there any significant differences in the number and tokens of metadiscourse devices between the two types of English essays? If any, what are the possible reasons for these differences?The result shows metadiscourse devices are used frequently in the two kinds of compositions, and very similar frequency is found, with 63.17 and 65.13 totally in American and Chinese students' compositions respectively. It also indicates that the metadiscourse features used most often in the two kinds of compositions is textual markers, with a percentage of 49% and 37% in Chinese and American students' compositions respectively. However, significant differences exist in the use of metadiscourse features between the two corpuses. First of all, as far as the use of textual metadiscourse is concerned, there is a significant difference between the two types of compositions (Sig.=. 015). More specifically, the Chinese authors of the English compositions use much more textual metadiscourse devices than their American counterparts, with the mean frequencies of 2.524 and 1.774 per 100 words respectively. The significance is actually caused by textual markers (sig.=. 041), and there is no significant difference in the use of interpretative markers (sig.= .132). Possible causes of the overuse of textual metadiscourse features in Chinese students' compositions are as follows: L2 learners' wrong inner criteria about good writing, poor language level and the learners' exposure to the limited target language input. Secondly, as for the use of interpersonal metadiscourse features, significance exists between the two corpuses (sig.= .044), and the Chinese students use much less interpersonal metadiscourse features. The result suggests that the Chinese authors use much less hedges and attitude markers in their compositions, which indicates that they do not have sufficient awareness of establishing appropriate writer-reader relationship and that they have less knowledge of audience needs.In the end, this paper suggests the pedagogical implications, the limitations and future research dimensions.