论文总字数：47714字

摘 要

本文通过个案研究，探究了淮阴师范学院英语专业四级习作文本中的元话语使用效果。以Hyland的元话语人际模式为理论基础，选取142篇写作文本，对常用的元话语类型进行探究，结果显示，学生使用交际元话语比互动元话语更多，他们使用自称语，读者介入标记语，过渡语最多，使用内指标记语和言据语最少。本研究中作文分数的依据是专业四级评分标准，本文还对高分组和低分组的元话语使用进行了对比分析，我们发现：高分组和低分组的元话语使用频率和比重都不同，即互动元话语更多使用在高分组，而交际元话语更多使用在低分组；高分组的同学使用元话语呈现多样性，尤其是，他们频繁得使用了框架标记语；低分组的同学更多使用模糊语，自称语和读者介入标记语。

因此本文提出，在英语写作教学中教师注重培养学生的元话语使用意识，提高学生元话语使用质量。

关键词：元话语；交际类元话语；互动类元话语；习作；专四

Contents

1. Introduction 1

1.1Background of this study 1

1.2 Significance of this study 2

1.3 The overall structure of this study 2

2. Literature Review 3

2.1 Definition of metadiscourse 3

2.2 Classifications of metadiscourse 5

2.3 Relevant studies at home and abroad 9

3. Methodology 10

3.1 Research questions 10

3.2 Participants 11

3.3 Instruments 11

3.4 Data collection 12

4. Findings and Discussion 12

4.1 Types of metadiscourse used in writings of TEM 4 12

4.2 Frequency and proportion of metadiscourse 14

4.3 Statistic analysis of interactive resources 15

4.4 Statistic analysis of interactional resources 15

4.5 Discussion 16

5. Conclusions and Implications 19

5.1 Conclusions 19

5.2 Implications 20

Works Cited 21

1. Introduction

As we know, learns’ writing is close related to the appropriate use of metadiscourse. And to know of the situation of English majors’ metadiscourse in their compositions is necessary for teachers and learners so as to improve the quality of English writing. Generally speaking, a complete discourse is supposed to be divided into two levels, primary discourse and metadiscourse. Primary discourse provides relevant information about discourse topic and expands propositional meanings. Metadiscourse was first put forward by Zelling Harris in 1959 and referred to a way of understanding language use or indicating the way of authors or speakers to guide audience to understand discourse (Harris, 1959). It guides readers to organize, interpret, evaluate, and react to the propositional information. Its patterns of manifestation are mainly including words, phrases, short sentences, clauses, punctuation, charts and even paragraphs. It is widely used in different literary forms and plays an indispensible role in writing. Besides, the achievements of metadiscourse in applied areas have increased. This chapter first describes the background of this study. Following this, this chapter goes on to the significance of this study. At last, this chapter introduces the overall structure of this study.

1.1Background of this study

TEM 4, abbreviated for Test for English Major-4, is the exam for nationwide college English majors and the content of this exam covers listening, reading, writing and translating to test the comprehensive competence of English majors. Thus, writing is an important content of examination of TEM 4. The researches on L2 writing have prospered in recent years. This study mainly focuses on analysis of metadiscourse in TEM 4 compositions.

A growing number of linguists focus their attention on the research of metadiscourse from various perspectives. In academic discourse, most linguists will discuss the impacts of textual function of metadiscourse on understanding of academic discourse, including written and spoken language. Moreover，more and more scholars have paid more attention to metadiscourse in applied discourse, and they have done a great deal of research on metadiscourse in applied genres, such as news commentaries, English writings, speeches and so on.

1.2 Significance of this study

Although linguists have made great achievements in applied areas of metadiscourse, but the studies of metadiscourse using in writing of English examination are not many now. TEM 4 is a very important exam for English majors and their grades of TEM 4 are relevant to their graduation from college and future employment. In TEM 4, writing is a vital test content in the examination because the score of composition has a direct effect on the total points. However at the same time, writing is the weakness of the most examinees. During the test, examinees are pressed for time and have many questions to answer. Number of words in writing requires 200.When they come to writing, it is the last content in the test and little time is left. Many students spend plenty of time on practicing writing but sometimes the results have not lived up to their expectations. Many of them have difficulty in writing and make every effort to improve it to pass the exam.

According to research findings of scholars, the frequency and effectiveness of metadiscourse used in writing have a great influence on the composition, especially that of TEM 4. This paper will make an analysis on the use of metadiscourse in the writing of examinees based on Hyland’s interpersonal model of metadiscourse to explore the relationship between the using of metadiscourse and the quality of compositions. The study means to be of significance to college English writing teaching. On one hand, it means to help students understand and master the knowledge of metadiscoure by learning the use of metadicourse and apply it into writing to improve their writing competence. On the other hand, it makes teachers pay attention to teaching the knowledge of metadicourse and how to use metadicourse properly in different styles of writing.

1.3 The overall structure of this study

This study is composed of five chapters. The introduction part presents the study of metadiscourse and development of theories of metadiscourse. It also shows the background and importance of this study. The second chapter introduces various definitions of metadiscourse and linguists’ different opinions on its definitions. And the various classifications of metadiscourse are presented. All of these will be the theory basis of this study. The third chapter focuses on methodology of this study, including the research questions, participants, instruments and data collection. The fourth chapter is the main and essential part of this study. It shows the analysis of interactional and interactive resources of the data. Then discussion will be made about the analysis. The last part is about the conclusions of this study and implications.

2. Literature Review

2.1 Definition of metadiscourse

“Metadiscourse” was first put forward by Zelling Harries in 1959(Harris, 1959). After that, however, the definition of metadiscourse is controversial. Nash thinks it is difficult to establish the boundaries of metadiscourse. He argues “the word ‘metadiscourse’ may have a reassuringly objective, ‘scientific’ ring, but its usage suggests boundaries of definition no more firmly drawn than those of, say, ‘rhetoric’ or ‘style’. One reader may perceive a clear stylistic intention in something which another reader dismisses as a commonplace, ‘automatized’ use of language” (Nash, 100).

Differently, Crismore et al. (1989:2) refers to it as “the authors intrusion into the discourse, either explicitly or non-explicitly, to direct rather than inform, showing readers how to understand what is said and meant in the primary discourse and how to “take” the author”. He defines metadiscourse again in an influential paper in 1993 which states that metadiscourse is “linguistic material in texts, written or spoken, which does not add anything to the propositional content but that is intended to help the listener or reader organize, interpret and evaluate the information given”(Crismore et al. 1993:40).

Other scholars have slightly different opinions. Schiffrin believes it is the authors’ manifestation in a text to “bracket the discourse organization and the expressive implications of what is being said” (Schiffrin, 1980:231). Fairclough believes that metadiscourse is a kind of obvious intertextuality and authors can communicate with readers through it in discourse (Fairclough 52).

Metadiscourse has been characterized as simply “discourse about discourse” or “talk about talk” by Williams. He points out that the author communicates with readers on two levels. One of them is the primary discourse and the other metadiscourse. The primary discourse is the information about the discourse topic and it will expand the propositional information. According to Williams (1981: 226), metadiscourse is “whatever does not refer to the participant matter being addressed.” It leads readers to organize, interpret, evaluate and react to the propositional information.

Metadiscourse is defined by Vande Kopple (1985) as “the linguistic material which does not add propositional information but which signals the presence of an author”. He redefines metadiscourse as “discourse that people use not to expand referential material, but to help their readers connect, organize, interpret, evaluate and develop attitudes towards that material.”(1997: 2) In other words, metadiscourse refers to facets which make the text explicit and accessible to the reader, and engage the reader in the interaction.

Both Williams (1981) and Vande Kopple (2002) understand metadiscourse from propositional and non-propositional perspectives as a kind of language material that does not add propositional information but indicates the existence of speakers to help audience to organize, understand and evaluate information. Also, Beauvais regards metadiscourse as the explicit markers that help readers to identify how a writer’s arguments are to be understood (Yang, 2005).

However, Hyland argues that the above-mentioned definitions are biased because metadiscourse can not be recognized on the basis of propositional information and usually author’s opinions are often drawn by non-explicit markers in a discourse. Hyland defined it as “the cover term for the self-reflective expressions used to negotiate interactional meanings in a text, assisting the writer (or speaker) to express a viewpoint and engage with readers as members of a particular community” (Hyland, 2008：37).

Hyland proposed the interpersonal model of metadiscourse, dividing it into two kinds, interactive metadiscourse and interactional metadiscourse. Abdi emphasized interpersonal metadiscourse as “an indicator of the attempts made by writers to create an interaction with their reader, reach their audience and express their own truth-value judgments about the ongoing proposition” (Abdi, 2002:196).

In China, it was first introduced by Cheng (1997) and it has been studied increasingly by Chinese linguists. Cheng proposed that speakers in any kinds of communicative contexts must utilize metadiscourse to organize the words to make it convenient for audience to understand as easily as possible. Meadiscourse has nothing to do with propositional information and it is only used to strengthen ideational and interpersonal meanings. Li (2001) generalized metadiscourse as the words, phrases and sentences that can mark the structure of discourse and reflect the intention of speaker. They can appear in different positions and will not add any new information to the discourse. They exist by attaching to discourse, so they are not the primary discourse and they serve as words that regulate the primary information.

Based on the definitions of the previous scholars, this paper defines metadiscourse as all the discourse-linking terms that used to self-reflect and negotiate interactional meanings in a text, which have two kinds of functions that include assisting the writer (or speaker) to express a viewpoint and interacting with readers.

2.2 Classifications of metadiscourse

According to the linguistic theory of Halliday (1976), language possesses three metafunctions: ideational function, interpersonal function and textual function. Ideational function refers to that we talk about our experiences about the world and describe things and situations happened around us with language. Interpersonal function refers to that we communicate with others and establish and keep interpersonal relations with language. We affect others by language and express our ideas about this world and even change this world with language. Textual function refers to how to organize our words appropriately to make it indicate the relationship with other information and realize the delivery of information and the relationship with interactive context of speaker at the same time.

Different classifications have been proposed by various linguists. The most typical classifications of metadiscourse are proposed by Vanda Kopple, Crismore et al. (1993) and the interpersonal model of Hyland (2008) on the basis of systemic functional linguistics of Halliday. Vanda Kopple (1985) divides metadiscourse into two kinds. They are textual metadiscourse and interpersonal metadiscourse. Textual metadiscourse consists of text connectives, code glosses, validity markers and narrators. Interpersonal metadisourse includes illocution markers, attitude markers and commentaries. Their specific definitions are summarized in Table 1.

Table1: Vanda Kopple’ classification system for metadiscourse

(Hyland, 2008:32)

Crismore et al. divides metadiscourse into textual metadiscourse and interpersonal metadiscourse too, but it still differs from that of Vanda Kopple. The classification is presented in Table 2.Table2: Metadiscourse categorization of Crismore et al.

(Crismore et al. 1993:47-54)

Based on the theory of Halliday, Hyland’s interpersonal model analyses the compositions of these students. It is based on a functional approach which regards metadiscourse as the ways writers refer to the text, the writer or the reader. It acknowledges the contextual specificity of metadiscourse and, at a finer degree of delicacy, employs Thetela’s (1997) distinction between interactive and interactional resources to acknowledge the organizational and evaluative features of interaction. However, it takes a wider focus by including both stance and engagement features and by building on earlier models of metadiscouse.

Table3: Hyland’s interpersonal model of metadiscourse

(Hyland, 2008:49)

Hyland’s interpersonal model of metadiscourse has solid theoretical basic and strengthens that all resources possess interpersonal interactional function. This kind of classification is the newest, relatively impeccable and reasonable classification. It presents the interpersonal function, ideational function and textual function of metadiscourse in various discourses and realizes the interactivity of ideational meaning, interpersonal meaning and textual meaning. It emphasizes the interpersonal interactive potential of metadisourse. This type of classification is adopted in this study.

2.3 Relevant studies at home and abroad

More and more linguists devoted themselves into the research of metadiscourse in applied linguistics. They make a deep research on metadiscourse in different kinds of discourses, such as English argumentative writing of college students, news commentaries, and political texts.

Some scholars devoted to the relevance between the use of metadiscourse and the quality of writing from an empirical point of view to explore the existed cultural difference between students from various cultural backgrounds (Vande Kopple 1985; Crismore et al. 1993; Intaraprawat and Steffensen 1995; Cheng 1997). Some scholars have studied the function of metadiscourse in writing, such as teaching of composition (Yuan Liying, 2012), the writing of English majors (MaYue, 2013), reading and writing of non-English majors (Yuan Liying, 2013), the quality of writing.

There has been much research about analysis of metadiscourse in college students’ writing, but a little is made about the TEM 4 writing. So my focus is on the use and frequency of metadiscourse and how it affects the quality of writing.

ZhuYushan (2012) has studied metasidourse used in different styles of writing and he finds that metadiscourse used varies in different types of writing. Metadiscourse are used the least in narration than that used in argumentation and exposition. The main difference of metadiscourse using lies in the use of transitions, frame markers, hedges, attitude markers, self mentions and engagement markers. Argumentations use the six subcategories more frequently. Expositions use more transitions，frame markers and hedges than narrations do while self mentions and engagement markers are used less.

To sum up, the previous researches have great contribution to metadiscourse studies and English writing. But study gap still exists in the detail description of the relationship between metadiscourse and quality of writing of English majors. This study focuses on the relationship between the effect of metadiscourse in argumentation in composition lessons of TEM 4 and quality of writing.

3. Methodology

According to the aim of the present study, a case study is used to explore the use of metadiscourse of TEM 4 writing through collecting and analyzing English majors’ compositions. The research design is therefore described in this chapter, including research questions, participants and data collection.

3.1 Research questions

The purpose of this study is to find the relationship between the frequency of metadiscourse and the quality of writing. Based on Hyland’s interpersonal model of metadiscourse, the study is motivated to answer three questions. The research questions are addressed in the following:

1. What types of metadiscourse is used in the writings of TEM 4?

2. What is the frequency and proportion of metadiscourse in the writings of TEM 4？

3. What is the relationship between the frequency and proportion of metadiscourse and the marks of TEM 4 writings?

3.2 Participants

To investigate learners’ use of metadiscourse, sixty students with their five pieces of compositions are selected as participants of the present study. The students are English majors of second-year who are from the School of Foreign languages at Huaiyin Normal University. Almost every student has past CET 4 and about 70% of them have passed CET 6. In their writing class, they are required to practice TEM 4 writing. The writing part of TEM 4 includes two sections, compositions and notes. The study focuses on compositions. Words of every composition are acquired to reach 200 and the time is limited to 40 minutes in each of writing. The styles of compositions include argumentation, exposition and narration. This study investigates their argumentations, which require the students to express their view on a certain social phenomenon or problem and display the reasons to argue their arguments. In the writing class, the teacher explains some writing skills and ideas and provides them with some sentence patterns. The five topics investigated in this study are all assigned by the teacher and finished in limited time in class.

3.3 Instruments

To answer the research questions, quantitative analysis is employed. Besides, interview is used to explore the evaluation of writings. The two teachers, Mrs. Wang and Mrs. Song are the natural interviewees for they are the teachers who give the marks.

3.4 Data collection

Data of TEM 4 writing is collected from writing lessons, which is composed of 142 compositions selected from writings of sophomores at random. The selected compositions are all argumentations, whose topics are “The internet—A Blessing or Curse?”, “How to Improve Students’ Mental Health”, “Is Civil Servant an Ideal Job”, “The Main Difference between My College Life and My Middle School Life” and “One Way to Solve the Problem”.

After data collecting, statistical work on data has been done. The total number of words in all the compositions is 24627 and the average words of each piece of writing are 173. And the whole frequency of metadiscourse has been counted according to the classification of Hyland’s interpersonal model of metadiscourse, which mainly includes two types of metadiscourse resources, interactive and interactional dimension. Interactive dimension consists of transitions, frame markers, endophoric markers, evidentials and code glosses. Interactional dimension is composed of hedges, boosters, attitude markers, self mentions and engagement markers. At last, these different types of metadiscourse in writing are analysed in quantitative and qualitative perspectives.

4. Findings and Discussion

This chapter firstly describes the types of metadiscourse used in the compositions of TEM 4, and it secondly shows the findings of frequency and proportion of two types of metadiscourse used in participants’ compositions, interactive resources and interactional resources, then the chapter discusses the relationship between the use of metadiscourse and teacher’s marks of the compositions.

4.1 Types of metadiscourse used in writings of TEM 4

After skimming compositions, all the metadiscourse are underlined and classified into two types, interactive resources and interactional resources, according to the classification of Hyland. The different types of metadiscourse are shown in Table 4.

Table4: Types of metadiscourse

Students in this study obviously use some of metadiscourse in their writings to keep coherence of discourse. However, statistic working shows that the frequency and proportion of metadiscourse are different as in the following table.

Table 5: Frequency and proportion of interactive and interactional resources

The total number of interactive resources and interactive resources is 2264. The number of matadiscourse in every piece of writing is about 16 on average and almost one metadiscourse exists in every 7 or 8 words. So it is seen that metadiscourse plays a significant role in writing of TEM 4 and almost all students pay attention to the employment of metadiscourse in their writing. But the proportion of types of metadiscourse is different as shown in the above table that interactional metadiscourse (64. 27) is higher than that of interactive metadiscourse (35.73).

To answer the second and third study questions, analysis on frequency and proportion of metadiscourse is in the following sections.

4.2 Frequency and proportion of metadiscourse

According to statistic analysis, the frequency and proportion of different types of metadiscourse are showed and discussed in this chapter.

Table6：Frequency and Proportion of Sub-categories of Metadiscourse

According to Table 6, among the ten categories of metadiscourse, engagement markers, which accounts for 33.35% of the total number of metadiscourse, self mentions which accounts for 19.74%, and transitions, which accounts for 19.83%, have high frequency of occurrence. Another metadiscourse used most is frame marker. Hierarchy of structure is very important in argumentation and teachers often teach students to pay attention to hierarchy. Frame markers like “firstly”, “secondly” are used the most. Code glosses are words to elaborate propositional meaning. Students usually use “for example” and “such as” in TEM 4 writing.

In addition, Endophoric markers and Attitude markers have low frequency of occurrence, all of them accounting for only 1.81%. Endophoric markers and evidentials are used the least. It indicates that students are not good at quoting and citing.

4.3 Statistic analysis of interactive resources

The marks of 142 writings are collected and are divided into two groups, group of high marks and group of low marks. Here high marks refer to the score of writings over 13 and low marks refer to the score no higher than 9 (The highest mark is 15). The number of writings over 13 is 34 and the number of those under 9 is 24. Due to the different numbers, 20 writings are selected respectively from the two groups and proportion of metadiscourse is calculated and comparatively analyzed.

Table 7: Proportion of interactive resources in group of high marks and low marks

According to Table 7, it is concluded that interactive resources play a significant role in writing of TEM 4. They are used frequently in high marks writing and the number of them decline in the writings of students of low marks. Among them, frame markers are used more frequently in the group of high marks than that of group of low marks. But there is no significant difference in using other interactive resources, including transitions, endophoric markers, evidentials and code glosses.

4.4 Statistic analysis of interactional resources

Table 8: Proportion of interactional resources in group of high marks and low marks

As Table 8 shows, interactional resources are used more frequently in writings of group of low marks than that in group of high marks. Among them, boosters and engagement markers are used more frequently in group of high marks than that of low marks. Besides, hedges, attitude markers and self mentions are used more frequently in group of low marks than that of group of high marks. Although the number of interactional resources used does not increase with the scores, the higher they score, the more various means of interactive resources they use.

4.5 Discussion

As shown in Table 5 and Table 6, the usual types of metadiscourse are used by the sophomores of English majors in the writings of TEM 4 though metadiscourse is not intentionally taught by teachers in writing class. However, the frequency and proportion are used differently in their writings. Interactional resources are used much more frequently than interactive resources. Engagement markers, self mentions and transitions, have high frequency of occurrence. The most frequently used resource of metadiscourse is Engagement markers, because most of the topics of TEM 4 writing are about things which are closely related to our daily life, such as social phenomenona, social problems and campus life. Therefore, students tend to use words like “we, our, you, your”, to resonate readers. In the argumentation of TEM 4, writers have to express their opinions in the first paragraph and demonstrate their views in the following paragraphs. Thus, self mention and attitude markers are used frequently. Attitude marker is generally used in the beginning and at the end so it is used less than self mention. Transitions account for 19.83% and it is quite high. The most used transition markers are “and” and “because” and the two words show frequently in the compositions of low marks.

As shown in Table 7, first, frame markers are used more frequently in group of high marks than that of group of low marks. The possible reasons are as following. The writings of students of high marks have clearer structure. Students of low marks tend to use ‘firstly’, ‘secondly’ and ‘thirdly’ and such relatively easy words mostly, while students of high marks tend to use more complicated markers, such as “to begin with”, “in the first place”, “first of all”. It indicates that students of high marks use more advanced vocabulary when they are using frame markers. And using more frame markers makes argumentations seem more logical. Second, there is no significant difference in using other interactive resources, including transitions, endophoric markers, evidentials and code glosses. The only interactive resource that students of high marks use less than students of low marks do is transitions. However, students of low marks present low quality of inner coherence in writing. According to the survey, students of low marks present more severe colloquial tendency in the TEM 4 writing. And that is why it is hard for them to get high marks in TEM 4. It also shows that students of low marks tended to use “and” at the beginning and they will use “in addition” and “besides” and such progressive words. The students of high marks in writing, who get higher marks in TEM 4, will use more adversative conjunctions, such as “thus”, “because”, which used to explore the reasons and readers may think the argumentations more logical when they see these words. It is shown in Table 7 that students use endophoric markers and evidentials the least and that indicates students just express their own points of view rather than cite authoritative and objective facts to verify their standpoint. Writings like this are of low objectivity, which will not directly make the writings persuasive enough. Students of high marks use endophoric markers more frequently, and it shows that compositions of high marks are more objective. In addition, students of high marks use more code glosses than students of low marks. These code glosses help readers understand authors’ points of view and some abstract expressions. It indicates that compositions of high marks are more explanatory. Both students of high marks and students of low marks use code glosses like “for example” and “such as”. The variation of code glosses differs slightly.

According to Table 8, boosters and engagement markers are used more frequently in group of high marks than that of low marks. Attitude markers and self mentions are used more frequently in group of low marks than that of group of high marks. Too many hedges will make argumentations somewhat untrustworthy so that we can see from Table 8 that Hedges have the lowest frequency of use. Students of low marks have higher frequency of using hedges. They usually use “may”, “almost” and so on. It shows that students of low marks use more hedges and make their compositions be of less certainty. That is why their marks are lower. According to Hyland (2008), boosters are words that allow writers to close down alternatives, head off conflicting views and express their certainty in what they say. In other words, the more boosters writers use, the more certainty compositions show. According to Table 7, we find that students of the two groups tend to use boosters for emphasis, which makes the writings more cumbersome instead. For instance, they like to use “obviously” and ‘actually’. Students of high marks use more boosters than students of low marks. That means their compositions have more affirmative expressions. Attitude markers express attitudes and opinions of writers. Attitude markers do not have great influence on the quality of writing. That is because the attitude of author can be understood by readers between the lines. As table 7 shows, attitude markers are largely used and the amount of Attitude markers differs slightly between the two groups. Whether students of high marks or of low marks, they tend to use “I think” and “as far as I am concerned”. Only a few students use attitude markers such as “undoubtedly”, “definitely” and “absolutely”. Self mentions take up quite high proportion. However, the higher marks are scored, the less self mentions the students use. Students of low marks have more self mention expressions in their writings because most of them comment from their own perspective in the argumentation, especially the overuse of “I”. However, the students with high marks in writing have fast decrease in the using of “I”. They have the intention to reduce the use of ‘I’ to avoid subjective tendency in writing. Besides，the use of self mention is related to topics of writings. For example, when it occurs to the topic “the main difference between my college life and my middle school life”, writers use a large amount of “I” and “we”. Authors can establish contact with readers effectively by using engagement markers. Table 7 shows that engagement markers are used the most in TEM 4 writing. The proportion of group of high marks and group of low marks differs slightly. In their compositions, inclusive “we”, “you” and “our” are largely used. In argumentations, such types of metadiscourse like hedges, self mentions and engagement markers can not help the readers get a clear and logical image and evaluation.

According to interview of the evaluating teachers, the marks of compositions partly depend on the use of metadiscourse. Mrs. Wang said, “In argumentation, you are likely to give a high mark if the metadicourse are used correctly. Some metadiscourse can make the discourse coherent.” Mrs. Song said, “The goal of writing is closely related to many factors, such as the structure, the main idea, grammar and spelling. And metadiscourse is obviously helpful to organize the discourse.”

According to the discussion above, interactive resources are used more frequently in group of high marks and they decline in the writings of low marks except transitions. It indicates that students of high marks lay more emphasis on the structure and logic of compositions. Students of high marks use less interactional resources than students of low marks. The former uses more boosters such as “obviously” and “clearly” while the latter uses more hedges such as “may”, “might” and “possible”. This indicates that students of high marks use more affirmative expressions than students of low marks. The frequency of attitude markers such as “I think” and “I argue that” and self mentions such as “I” and “we” are used because of the topics, which are close related to the experience of writers’ life. The use of Engagement markers differ slightly between the two groups.

Generally speaking, English majors are inclined to utilize more interactional resources in TEM 4 writing. Students of high marks use more interactive resources while students of low marks use more interactional resources. Students of high marks use metadiscourse more properly and appropriately. That means the effectiveness of metadiscourse using, rather than the number, is proportional to the quality of writing. In fact，the quality of writing usually depends on learning and logic. And the use of metadiscourse is relevant to the logic relationship of writing. Therefore, the quality of metadiscourse using is an important factor to evaluate writing quality.

5. Conclusions and Implications

5.1 Conclusions

Although this paper is intent to be well organized but limitation still exists. For example, it just studies 142 pieces of writing of TEM 4 and it is not enough to represent the overall condition of metadiscourse in TEM 4 writing. Also, students of different areas may have different conditions of metadiscourse using. Moreover, the topic of writings varies and that may affect the occurrence of metadiscourse. All of these should be paid attention and be solved.

Through investigating 142 compositions of 60 English majors, this paper studies the use of metadiscourse in the argumentations of TEM 4 and the relationship between the number and quality of metadiscourse using and the quality of writing. The metadiscourse in the writings of TEM 4 is divided into two kinds, interactive resource and interactional resource based on Hyland’s interpersonal model of metadiscourse. In addition, the writings of TEM 4 are divided into two groups of low marks and high marks. The relationship between the use of metadiscourse and marks of writing is studied in this paper.

The research results show that students use more interactional resources than interactive resources in writing. They use self mentions, engagement markers and transitions the most and endophoric markers and evidentials the least. Also, the frequency of metadiscourse is different in the writings of low marks and high marks. Students of high marks use more interactive resources while students of low marks use more interactional resources. And student of high marks use metadiscourse in a more various way. The participants get higher marks partly because they use frame markers more frequently. The participants get lower marks partly because they use interactional resources such as hedges, self mentions and engagement markers.

5.2 Implications

English writing teachers can draw some suggestions from this study. Teachers should improve students’ inner standard of writing and teach more knowledge of how to use metadiscourse properly. Knowledge of metadiscourse can help students recognize the relationship between readers and authors, discourse and context. The teachers should teach the students to use more interactive resources, especially frame markers in the writings of argumentations. By doing this, learners can consider more about the readers, how to write and how to organize the discourse. In that case, students can learn more about the knowledge of metadiscourse and the role they play in the discourse and form the audience awareness. Only in this way can students improve their writing from sentence level to discourse level and improve the ability of writing.

Works Cited

Abdi, R. “Interpersonal metadiscourse: An indicator of interaction and identity”. Discourse Studies, 2002(2):139-145.

Cheng, X. and M. Steffensen. “Metadiscourse: A technique for improving student writing”. Research in the Teaching of English, 1996(2): 149-181.

Crismore, A. Talking with Readers: Metadiscourse as Rhetorical Act. New York: Peter Lang, 1989.

剩余内容已隐藏，请支付后下载全文，论文总字数：47714字