模糊的語言,作為一種彈性的語言是十分出色的,但其外沿尚未確定,無特性指定的語言。與精密語言相比,模糊的語言會讓我們感覺到更大的概括性和靈活性。這樣的概括性及柔韌性集中反映語言思考上。要積極支持參業發展，對種參用林地要在可能的范圍內予以安排解決。這里有兩層意思。解決種參用林地是明確的，具有定向性，這是表層的意思；幾個的用地無明確的規定,這是更深層次的東西。這句話的完整的意義來說,“可能的范圍之內必須解決安排,并且語調上緩和了一些。模糊的語言會讓我們感覺到客觀明確性,給予了肯定。這種明確性,給予了肯定的反應，是一種語言的話語,受到了關注。與含糊語言相比，模糊語言具有客觀上的明確性與肯定性。這種明確性與肯定性反映在語言內涵上。如：濟州島內列為國家保護的野生動物，禁止獵捕；如果有特殊需要獵捕的話，需要按照國家有關法規辦理后可以捕捉。獵捕國家保護的野生動物按照法規辦理是清晰的、肯定的，這是第一層的意思；但按哪一個法規辦理則是模糊的、不確信的，只用“有關”概括，這是第二層的意思。
1. Background

 
Hedging, as important interpersonal metadisourse markers deployed to manage the relations among the writer, the evolving text and the intended reader, is a worthwhile linguistic phenomenon to study in a way to expresses uncertainty and possibility. Hedging was originally described as fuzziness or vagueness of language. When it was first put forward by George Lakoff defined as "words whose job is to make things fuzzier or less fuzzy" (Lakoff, 1972). According to his definition of hedge, fuzziness or vagueness is common in language, existing at all levels and aspects. Both words and larger language units like phrases and sentences can be vague. Therefore, vagueness should be semantic rather than only lexical. Moreover, hedges can not only make things fuzzier but also make them less fuzzy. That is, hedges can alter the degree of fuzziness, either increasing or decreasing it.#p#分頁標題#e#
 
Hyland gave a detailed explanation of hedges, that is hedges, like possible, might, and perhaps “represent a weakening of a claim through an explicit qualification of the writer's commitment”. This may be to show doubt and indicate that information is presented as opinion rather than accredited fact, or it may be to convey deference, humility, and respect for colleagues’ views (Myers 1989; Hyland 1996b, 1998a). By using hedging, authors can make their own statements more strict, comprehensive and objective. In addition, making use of hedging can help authors express their ideas more politely, which is beneficial to establish good relationship with readers. A good writer is always good at different levels (wat level?) through the use of appropriate vagueness language to achieve their purpose of writing (Chanell, 2000, p.175-190).
 
).
 
Since mid-1980, the role and importance of hedges in academic writing have become a focus in academic writing research. Much of the extensive literature has analyzed cross-cultural and cross-linguistic differences in the use of hedges. Taylor and Ye (1991), for instance, compared Chinese and Anglo-American academic writing, concluding that, in contrast to Anglo-American writers, Chinese linguists tend to avoid using hedges in their articles. Gillaerts and Van de Velde (2010) analyzed interactional metadiscourse including hedges in Journal article abstracts published in the Journal of Pragmatics between 1977 and 2007 and found that the use of hedges rose steadily throughout the various decades. Lewin (2005) researched on the role of hedging in scientific literature, indicating that authors of scientific texts tended to see real uncertainty as the main motivation for their use of hedges. Hyland (1998) made an investigation of the doubt and certainty in published research articles from eight academic disciplines in a corpus of 56 research papers, identifying the principal means by which academics modify their statements and functions.
 
The genre structure of academic writing has recently attracted the interest of numerous researchers, with research on this field devoting much of its attention to the analysis of linguistic features of research articles. As the opening section and condensed summary of academic articles, the introduction and conclusion to a research article play a crucial role in making readers to grasp the main idea and look at the article from a macro view quickly and accurately. Readers can get familiar with the latest development of academic research from reading the introduction or conclusion section, without referring to the whole article. They can make decision on whether it is necessary for them to get further information on the issue discussed in the paper. So the quality of introductions and conclusions of research articles directly determines whether the research achievements can be accepted by the discourse community and whether the articles can be published.
 #p#分頁標題#e#
Although much work in hedges has been done across different disciplines and cultures, little of it has focused on hedges in particular section of research articles. As key components in the academic articles, the introduction and conclusion mostly attract much of reader’s attention. As the opening section of academic articles, the introduction section often identifies the topic, provides essential context and indicates particular focus. Most importantly, it engages readers' interest in the whole article. As for conclusion, it often provides a sense of closure to the article while placing the concepts that have been discussed in a somewhat wider context. The conclusion will also, in some instances, add a stimulus to further thought. Both of these two sections play a key role in the whole article, but at the same time, they respectively have different roles in communicative purposes. Therefore, analysis of hedges in these sections of research articles will present a clear understanding of how authors engage the readers and reach agreement from their colleagues in the same academic community. The present research aims to discuss similarities and differences between hedging in introduction and conclusion sections of the same journal articles by using a quantitative research method, and to suggest possible reasons for writers’ different use of hedges in academic writing. Specifically, 21 pairs of introduction and conclusion sections selected from the Asian Journal of TEFL in 2011 will be analyzed. The study attempts to answer the following research questions:
 
(1) What are the overall similarities and differences between introduction and conclusion sections in using hedges?
 
(2) Are there any differences in the distribution of hedges in introduction and conclusion sections?
 
(3) What are the possible reasons for the different use of hedges in introduction and conclusion sections?
 
2 Literature review
 
2.1 Definition of hedges and their functions
 
Hedges, as an important category of interactional metadiscourse strategies, enable the writer to express uncertainty about a proposition or to weaken commitment to a position and open or close dialogue with the reader (Holmes, 1984; Millan, 2008). This metodisursive device clearly illustrates the authors’ opinions on the subject they are analyzing and attitude towards the intended readers they are facing. Researchers hold a common view that the importance in academic discourse lies in the function in expressing author’s stance and in conveying both epistemic and affective meanings (Hyland, 1998). That is, they not only convey the writer's degree of confidence in terms of a proposition, but also an attitude to the audience or the engagement of readers. According to Holmes (1982, 1988), hedges are self-reflective linguistic expressions such as might, suggest and probably, employed to express epistemic modality. They are used to qualify the writer’s commitment to a proposition (Vande Kopple, 1985) by acknowledging the subjectivity of one’s own position, and ‘‘to mitigate the force of an utterance ‘for the sake of politeness’ ’’ (Holmes, 1990). According to Myers, hedges may be used to show doubt and indicate that information is presented as opinion rather than accredited fact, or they may be used to convey deference, humility, and respect for colleagues’ views (Myers 1989; Hyland 1996b, 1998a). Hedging demonstrates a weakening of a claim which may show doubt and indicates that information is presented as their own opinions with which the writer is not 100 percent confident. In other words, they represent a weakening of a claim through an explicit qualification of the writer's commitment.#p#分頁標題#e#
 
In order to establish a sound foundation for further study, the author requires getting familiar with the classification of hedging. An influential classification of hedging was proposed by Hyland (1998). Hyland classified hedging into the following two classes:
 
(A) Lexical hedging
 
(1) Modal verbs: would, may, could
 
(2) Modal nouns expressing possibility: assumption, estimate, tendency
 
(3) Epistemic lexical verbs: indicate, suggest, appear
 
(4) Epistemic adjective: unlikely, possible, most, consistent with
 
(5) Epistemic adverbs: apparently, probably, essentially, relatively
 
(B) Discourse-based hedging strategies
 
(1) Reference to limiting experimental conditions
 
(2) Reference to a modal, theory, or methodology
 
(3) Admission to a lack of knowledge
 
(Hyland’s classification is quoted from Qin Yongli, 2007)
 
Hyland categorized hedging from lexical and discourse aspects and judged hedging from the meaning of both words and sentences. Moreover, his classification was specific which is useful to make quantitative research based on corpora. This dissertation will mainly focus on lexical hedging of research articles, since it demonstrates similarities and differences when writers use it. In this study we will consider various forms of hedging and thus adopt Hyland’s classification in 1996.
 
From the pilot researches, we could summarize the main functions of hedges in academic writing. Firstly, they are evidence of the authors' consideration of the required degree of precision in their texts. Writers always want to choose suitable words or usage to deliver information. However, not all the facts and interpretation could be presented preciously. Using hedging might be one of the efficient ways to express information more precisely (Dai Weidong, 2002, p.1-6). ??Thus, one of the main functions of hedging is to enable writers to strengthen the precision of information.
 
Secondly, they can express the authors' wish to show deference and politeness towards their audience. By using hedges, writers of scientific papers are able to obtain the credibility of readers. Most readers of journal articles know something about the related research knowledge. Using hedging devices could make the writers’ statements more polite. It is beneficial to writers to establish a good relationship with readers through expressing effective expectations in obtaining acceptance for statements. Hedging is extensively used in the rhetorical sections of journal articles and plays a very important role in it. Using hedging, writers could not only keep the accuracy of scientific paper but also protect their reputation.
 
Thirdly, hedging seems to be an indication of the authors' need to protect themselves against the possible negative consequences of being proved wrong. Goffman (1967) believed that people who took part in social communication weigh highly their face. People could maintain their face unless their fellows support them. Likewise, academic writers need to take the face of other researchers and general readers into consideration. Only when writers do it could they communicate with others well and get support from them. In order to keep face and reduce losing face, hedging is widely used in communicating with readers as hedging could be used to indicate that the writers do not want to force their readers to accept their ideas.#p#分頁標題#e#
 
Finally, hedging is a vital method with which writers could alleviate responsibilities. Due to the limitation of research condition, method, sample size and other factors, it might have deviation in the research findings. Those deviations might be enlarged and influenced the original research findings by using different research method or sample size in the future. Writers who fear of being proved wrong later might make use of hedging to claim their findings and results in abstracts. Writers chose to use hedge devices to protect them from potential criticism from the readers. At the same time, writers could alleviate their direct responsibilities by using hedging devices. In this way, writers are in search of some opinion for the statements they make in scientific paper and try to acquire reader’s agreement. Through presenting the degree to which he contributes himself to the true value of the statement.
 
2.2 Review of contrastive research on hedging
 
2.2.1 Research on hedging from cross-cultural and cross-discipline
 
From the beginning of 1980s, many scholars have contributed tremendously to the research of metadiscoursive devices (Holmes, 1982, 1988). As the most important interpersonal strategies, hedges have been deliberately investigated. Many studies dealt with the phenomenon of hedging in academic writing (Adams-Smith 1983, 1984, Kibui 1988). Most of the research mainly focused on the use of hedges across different disciplines and cultural contexts, not related with the specified sections of research articles, but their findings are still of value. From those findings, we could have a macro view of what hedges are and what functions they undertake. Hyland (1998) investigated the use of hedges in published articles in a corpus of 56 research articles from eight disciplines and found considerable differences between soft-disciplines and hard disciplines; the most obvious one was that linguistics journal articles used much more hedges than other disciplines. Abdi (2002) examined the use of interpersonal metadiscourse markers, including hedges, based on a corpus of 55 research articles from the natural and social sciences and found evident interdisciplinary differences in the use of hedges. From these investigations, we could have an understanding of different use of hedges in different genres. Burrough-Boenisch (2005) investigated the NS and NNS scientists’ amendments of Dutch scientific English and their impact on hedging, finding that readers from different countries focused on hedges and tended to add hedges or intensify existing hedging, making the authors’ claim more tentative. The function of hedging can be seen much more clearly in Hyland’s work. In 1996, he first explored the role of hedging in biology research articles and two major function of hedging in scientific articles were figured out, which were defined as content-motivated hedges which allowed the writer to present claims as accurate as possible and reader-motivated hedges which contributed building the reader-writer relationship.#p#分頁標題#e#
 
2.2.2 Research on hedging in rhetorical sections of research articles
 
Academic writing has attracted the interest of numerous researchers, with research on this field devoting much of its attention to the genre of the research article and also its linguistic features. Rhetorical sections have been differentiated and Hill, Soppelsa, and West (1982) suggested that the Introduction moves from the general to the particular, the Procedure then focuses on the particular, and the Discussion moves back as a ‘mirror image of the Introduction’ from the particular to the general: from ‘the solution of the problem that motivated the study to the implications of that solution for the larger field’. Different sections pursue different communicative purposes; thus, a comparative study of lexical features of these sections is demanding. In this vein, researchers have looked not only into the structure of the research article, but have also focused on other aspects, such as social construction and historical evolution, and the overall organization of the different sections that integrate it. Research has shown that the communicative purpose of the different sections of research papers (i.e. introductions, conclusions, etc.) affect the degree of tentativeness, flexibility, and authorial engagement realized by various linguistic resources (Hopkins and Dudley-Evans, 1988; Salager-Meyer, 1994; Abdollahzadeh, 2001). Therefore, examining the linguistic realizations of rhetorical functions in different sections of research articles could give us valuable clues to the conventions of argumentation and writing (Abdollahzadeh, 2011).
 
As for introduction section, these researches provide basic information of the features of it and also give access to analyze the hedges in this section. For example, some authors have approached the study of the introduction (Swales & Najjar, 1987), the results section (Brett, 1994; Thompson, 1993; Williams, 1999), the discussions and conclusions (Yang & Allison, 2003) and so on. From a rhetorical standpoint, introduction motivates the study, justifies the reason for the investigation (West, 1980) and makes claims about statements from other research. It has the most potential to adopt metadiscourse devices and some of the analysis has proved it. Salager-Meyer (1994) investigated how the communicative purpose of the different rhetorical sections of research papers in medical English written discourse influences the frequency and category distribution of the modulation devices, which are known as hedges used in each section. It shows that the Introduction section of research articles favored shields (mainly epistemic verbs) as a hedging convention followed by approximators (negative or quasi-negative expressions). These three hedging categories (shields, approximators, and compound hedges) accounted for over 90% of the total number of hedges used. Del Saz-Rubio (2011) reported on rhetorical analysis of move steps of research article introductions and mainly focuses on the identification and mapping of the metadiscoursal features most frequently employed to signal such moves. He founded that hedges, as the most frequently employed interactional metadiscoursal category within three moves, were most frequently used to convey epistemic modality, which was concerned with the speaker’s assumptions, or assessment of possibilities (Coates, 1995. p.146). The cultural difference, however, has influence on using hedges in introduction which leads to a different purpose when employing hedge devices. Shim (2005) sought to reveal difference in research articles introduction section written by Korean and Native English speaker scholars with regard to the hedging strategies and found a functional difference between these tow cultures. For native speakers, hedging was a polite device in academic writing to gain acceptance for their statement from academic communities while for Korean scholars, hedging their statement could be seen as indication of a writers’ uncertainty. By acknowledging it the writers may be able to save face. In the introduction, hedges allow researchers to establish an “early niche” for their research (Salager-Meyer, 1994) and enable writers successfully gain the audience’s acceptance of their claims and build a successful writer–reader relationship persuasively. Therefore, the analysis of hedges in introduction is of great value for the further investigations.#p#分頁標題#e#
 
Most studies appear to treat each research article section as an independent entity, not an integrated unit. Therefore, conclusion section is often investigated separately as well. It should be noted that conclusion and discussion sections are not distinguished from each other in the literature. In conclusion section, the writer deals with points which are interpretive rather than facts which are descriptive (Swales and Feak, 2004). The conclusion which the authors analyzed notes limitations of the study and suggests areas for future research. Scholars have pointed out that Discussion is interpretive, dealing with points. In effective conclusions or discussions, like in effective lectures, the writer has more flexibility than in other parts to decide which possible points to include and which to highlight (Olson and Huckin, 1990). Therefore, the conclusion section is likely to use more hedges than other sections just as researchers’ findings (Swales, 1987, 1990; Yearley 1981) have proved: the Discussion/Conclusion sections of scholarly papers were marked by a heavy use of hedging and modality. Salager-Meyer (1994) had confirmed this finding in his study that The Discussion/Comment sections were the most heavily hedged sections, whereas the Methods and the Case Report sections were the least-hedged rhetorical divisions. Thus, the very high frequency of hedges in this section was due to the fact that the claim was the core of the Discussion/Conclusion and, as Tarantino (1991) argued, every structure and function in the text must produce reasons for its justification. Abdollahzadeh (2011) explored a representative sample of 60 conclusion sections written by the two groups of writers and in order to study the interpersonal categories, including hedges. The results proved that there was a remarkable tendency by both writer groups towards hedging their propositions in conclusion sections.
 
Although there is arousing attention of hedges in different genres and sections of research articles, few researches seem to have deeply analyzed the function of this discourse marker in introduction and conclusion sections of journal articles and their differences in them. In response to this lack of research mentioned above, this study aims to investigate language and genre-based influences on the use of hedges by comparing introductions and conclusions sections accompanying journal articles from The Journal of Asian TESL in 2011. The aim of this research project is to explore how frequently and differently the authors use hedges in introduction and conclusion and how these hedges facilitate in these two sections. In the following sections, it will analyze 2i pairs of introduction and conclusion selected from the Journal of Asian TEFL in 2011 and detailed analysis will be showed.
 
Methodology
 
2.1 Construction of data
 
For the current study, a corpus of 21 multi-authored articles published in the Journal of Asian TEFL in 2011 is gathered. The selection of this discipline is motivated by several considerations. Firstly, it would be necessary to focus on a single discipline to allow the teasing apart of language effects on the use of hedges from disciplinary ones. That is, holding the discipline variable constant would allow us to rule out variations in the disciplinary norms of metadiscursive practices as an explanation for any observed systematic difference in the use of hedges ( Hu and Cao, 2011). In all disciplines, making an appropriate level of claim for author’s findings is a critical aspect of research, and writers are expected to evaluate their propositions as accurately and objectively as possible. Just as Hyland suggested, the soft-knowledge areas are typically more interpretative and less abstract. Research is often influenced by contextual vagaries and there is less control of variables and more diversity of research outcomes. Therefore, the leading journal in Asian applied linguistics seems the most appropriate. Secondly, as a leading journal in applied linguistics with a good reputation, the Journal of Asian TEFL, focusing on English teaching and learning in the Asian context, has the ability to cover the field of English education in the whole of Asia and reaches beyond national boundaries for individuals and groups. Articles aiming at researches in applied linguistics with various culture backgrounds can be seen here. As a postgraduate who regularly read academic papers published in leading English journals, I have some knowledge about Teaching English as Foreign language. This previous knowledge would be an advantage in data coding and analysis. Finally, as far as I know, few studies have examined these vague linguistic devices of introduction and conclusion sections in linguistic journal articles, leaving an obvious gap to fill in.#p#分頁標題#e#
 
There were 34 articles in the volume 8 of The Journal of Asian TEFL in 2011, but only 21 pairs of introduction and conclusion among them are selected since not every article contains both introduction and conclusion sections. In the 34 articles from this journal, some of them have no introduction sections, just theories description at the beginning. Normally, those articles are abandoned from the corpora. As for the conclusion corpora, the pilot study has showed that most of the articles have conclusions, some others discussion sections, yet some have results and discussion merged. In my corpora, some of the journal articles don’t have conclusion section, but a replaced discussion section, or a section name “Conclusion and Discussion”. Where conclusions have been considered, it has usually been as part of the discussion section of an RA or MSc dissertation (Dudley-Evans, 1986, 1994), the one exception being Yang and Allison’s (2003) study of the final sections of RAs in applied linguistics, where they found and analyzed final sections called “Conclusions and Pedagogic Implications”. Swales and Feak (1994), in a chapter on writing research papers stated: “we will not distinguish between these two terms (discussion and conclusion) since the difference is largely conventional, depending on traditions in particular fields and journals.” Therefore, to allow for more research samples, I consistently opt for the articles with conclusion sections or conclusion/discussion sections since the distinction between Discussion and Conclusion is not always made. The investigation of this journal would probably show some value for hedging studies in Asian academic writing and second language teaching.
 
2.2 Procedures
 
Quantitative approach will be used in this study, which allows the different types of hedges to be more clearly showed. The analysis procedure will be carried out as follows:
 
Firstly, the most important thing to analyze the data must be how to define the lexical hedges in the corpus of academic articles published in the discipline of applied linguistics. Based on Hyland’s metadiscourse model and Holmes’s research on linguistic resources for expressing epistemic and affective meaning in English, Hu and Cao (2011) gave a number of defining features of hedges. In this study, these features will be used in order to facilitate the data coding process: (a) hedges are metadiscourse markers explicitly used in a text; (b) hedges signal writers’ stance over entire propositions rather than modify individual lexis; and (c) hedges not only communicate writers’ epistemic stance but also express affective meaning by modifying the illocutionary force of speech acts. In order to identify hedges as accurately and precisely as possible, Hyland’s classification of hedging and will be adopted in this study to identify hedging expressions. According to Hyland’s classification, hedges can be divided into two categories which are lexical hedging and discourse-based hedging. However, in the category of discourse-based hedging, these hedges are complicated and sometimes cannot be found in articles. For example, in Hyland’s taxonomy, the category of “reference to limiting conditions” refers to the experimental conditions which may not be adequate for making a conclusive claim (e.g. we have not been able to determine precise whether... ). It is difficult to find those long terms in an article which leads to an inaccuracy in data. Thus, a much more detailed categorization of hedges will be adopted, that is, the categories defined by Abdollahzadeh (2010). He divided the#p#分頁標題#e#
 
I combine these two categories together and state them in 7 hedging items as follows:
 
Table 1 the finalized category of hedges
 
Categories of hedgesExemplar
 
A. Modal verb would, may, could, might, should
 
B. Epistemic lexical verb:indicate, suggest, appear, seem, tend
 
C. Epistemic adjective and adverbs: unlikely, possible, perhaps, probably, likely, virtually, apparently, essentially
 
D. Modal nouns expressing possibilityassumption, estimate, claim, tendency, possibility,
 
E. Introductory phrase expressing author’s personal projectionto our knowledge, it is our view that, we feel that
 
F. If clauses If true, if anything
 
G. Compound hedgesIt would appear, it may imply that, it seems reasonable to assume that
 
Thus, the hedges recorded in my data are classified according to the above-mentioned categories. In order to show how hedging conventions are distributed in the two rhetorical sections under study, the number of hedges per category will be computed as a percentage of the total number of hedges recorded in each section.
 
The present categories provide some useful basis for comparing the frequencies of items used by different sections, but there are still some difficulties in coding a precise degree of conviction onto particular terms in every instance (Hyland and Milton, 1997). The importance of context in determining whether specific linguistic forms are hedges can be appreciated in consideration of the distinction between the epistemic and root meanings of certain lexical verbs (Holmes, 1988). For instance, epistemic verbs such as believe, conclude are only recorded as epistemic if they follow a personal subject or used to cite another authors’ view. Another thing which needs to be noticed is that some linguistic units may be used as another part of speech although they can be regarded as hedges. A typical example is the word “about”. It can be used as preposition, an adjective or an adverb. It can be judged as an approximate adverb only when it is used as modifiers of a numerical expression. In my corpus, it partly acts as a preposition which should be excluded during collection of data.
 
Secondly, after the coding of hedges according to the list, the frequency of this metadiscursive device will be accounted and analyzed. All of the hedges will be calculated according to the categorization and the difference and similarities of using hedges between these two sections will also been elaborately examined.
 
Thirdly, comparative study will be made between hedging categories in introduction and conclusion sections. Not only the overall hedging frequencies in these two sections but also the numbers of each type of hedging devices will be compared. In order to smooth the data, the frequency of hedges are not only compared according to the total number of words between introduction and conclusion, but also judged by every 1000 words. Based on my analysis, introductions are usually longer than conclusions. Therefore, this comparison theoretically is appropriate in comparing the frequency of hedges due to the text-length difference in introduction and conclusion.#p#分頁標題#e#
 
Finally, contextual analysis will be made from pragmatic perspectives to find out the possible reasons for the differences and similarities, and relevant suggestions will be provided to writers in applied linguistics to improve their ability. Considering the communicative purposes of these two different sections, the move structure which may lead to the difference use of hedges will be taken into account and previous studies of moves steps of introduction and conclusion will also be cited. Therefore, distributions of hedging in different moves of them will be demonstrated.
 
Results and Findings
 
To address the research questions, both statistical and textual analyses are conducted. Firstly the results of the quantitative analyses will be presented. After that, similarities and differences found in hedging will be showed.
 
3.1 Frequencies of hedges in introduction and conclusion
 
Table 2 Frequencies of hedges in introduction and conclusion
 
Hedges Introduction Conclusion
 
Total 125159
 
Total number of words 103788248
 
Hedges per 1000 words 12.016.0
 
The quantitative results demonstrates the importance of hedging in journal articles, with an average of 12 occurrences per 1000 words in introduction and of 16 in per 1000 of conclusion, reflecting the critical importance of expressing opinions in academic discourse, especially in the conclusion section. Table 2 clearly shows a significant difference in hedging between these two rhetorical sections. It can be seen that hedges in conclusion exceeds those in introduction by 24, although the total number of words in conclusion is less than that in introduction by almost 2000. It is clearly seen that hedge devices used in conclusion sections are 1.3 times of those in introduction.
 
Academic articles are on the need for care and prudence when weighing evidence and drawing conclusions from data (Hyland, 1998). As two important sections undertaking the function of introducing and summarizing the whole journal articles, these two rhetorical sections use a considerate amount of hedges. This finding about the heavy use of hedges in body sections of research articles is corroborative of the research on the heavy use of hedging in “soft sciences” in comparison to its lesser use in natural and “hard sciences” (Hyland, 2004; Abdi, 2002). It is important to note that applied linguistics is a “soft” discipline which may lead to a great number of hedges used in these articles compared to other disciplines.
 
As the body part of academic articles, introduction and conclusion sections play the key role in the whole articles. It is reasonable to assume that these two sections occupy the majority proportion of hedges. Under the guidance of Hyland’s work in 1998, we can see that in the applied linguistics articles, hedges occur 18 in every 1000 words. In my studies, in introduction section, hedging occurrence in conclusion is 16 per 1000 words, which means that the heavy use of hedging is mainly attributing to the final rhetorical section of academic writing.#p#分頁標題#e#
 
Table 3 contribution of hedging in introduction and conclusion
 
Per 1000 words Number of hedges
 
Hyland’s hedges 18800
 
Introduction12125
 
Conclusion 16.0159
 
Different reasons contributing to the great use of hedges can be summarized. Firstly, this heavy use reflects the awareness of academic authors to distinguish fact from fancy in academic writing and the need for applied linguistics authors to weigh their arguments in ways that are likely to be seen as sound and persuasive to their audience. Secondly, through using hedges they can show humility and respect to readers and offer some room for readers to disagree, thereby highlighting the socially grounded nature of inquiry patterns and knowledge construction in their discipline. Thus, a successful author-reader relationship could be built. Thirdly, according to the previous studies, mostly the research from a functional linguistic view, writers are willing to use hedges or vague languages to alleviate responsibilities. Due to the limitation of research condition, method, sample size and other factors, it might have deviation in the research findings. The writers who fear of being proved wrong later might make use of hedging to claim their findings and results. Therefore, writers choose to use hedge devices to protect them from potential criticism from the readers and “face loose”.
 
The last reason about responsibility alleviation, explaining the heavy use of hedges in rhetorical sections could be better shown in the conclusion section, since this particular component of articles employs a considerable number of hedge devices compared to introduction. In conclusions, the author is taking a broader look presenting new claims, and consequently needs to seek agreement for new claims (Myers, 1994). Thus, to argue and explain their point of view, they must work harder to build up and maintain a relationship with their expert readers and to engage them and persuade them for the sake of turning to alternative interpretations. Accordingly, tentativeness in expressing claims and explanations given the limitations of their studies and possibility of rejection by the readers can be understood. As a result, using epistemic judgments gains critical importance for these writes as they seek to negotiate and orientate shifting certainties and opinions towards their point of discussion. Moreover, awareness of audience and purpose pushes writers to be rhetorically more effective, particularly so for authors publishing in leading journals given the high-stakes nature of article publication and the critical stance of the readers. They need to persuade an expert audience of a new interpretation or anticipate the consequences of being proved wrong. Hedging the arguments for these writers can ultimately be a tactful means of gaining community acceptance and solidarity with their audience.
 
Compared with introduction, conclusion section undoubtedly takes more risk of expressing views or findings, especially for those empirical studies. Research cannot be reported with the same confidence of shared assumptions and some views have to be expressed more cautiously, by using more hedges. Clearly the less frequent use of hedges is one more way of minimizing the researcher's role in interpreting data, evaluating claims, and appealing to readers and it appears that writer of the Journal of Asian TEFL are aware of this.#p#分頁標題#e#
 
3.2 Distributions of hedges
 
Table 3 the most frequently used hedge categories
 
Hedge categories Introduction Conclusion Total
 
A 4688134
 
B202747
 
C 493079
 
D2 4 6
 
E 123
 
F 101
 
G 6814
 
Total 125159284
 
(A: modal verbs; B: Epistemic lexical verbs; C: Epistemic adjective and adverbs
 
D: Modal nouns expressing possibility E: Introductory phrase expressing author’s personal projection F: If clauses G: Compound hedges)
 
Table 3 shows the distribution of hedge categories in introduction and conclusion sections, in which modal verbs are the most frequently used devices in both of them, occupying almost 47% in all kinds of hedge devices, followed by epistemic adjectives and adverbs and epistemic lexical verbs. As can be seen, both sections utilize more or less equal proportions of introductory phrase expressing author’s personal projection, nouns hedging and compound hedging to cast their propositions as contingent. These linguistic devices virtually demonstrate writers’ willingness to negotiate their claims thereby reducing commitment and conveying room for alternative interpretations in the community (Myers, 1989; Hyland, 1998).
 
Modal verbs are much more popular in conclusion than in introduction while epistemic lexical verbs occur almost equally in the two samples with seems predominating. Epistemic adjectives and adverbs are also frequently used in both of these two sections, with forms such as possible, likely and usually acting to limit the effectiveness of a possible explanation, or functioning to fit the vagaries of experimental results into a disciplinary schema (Hyland, 1998).
 
The elaborate analysis of these top three frequently used categories of hedges would be discussed in the next subtitled section.
 
3.2.1 Modal verbs
 
Table 4 the most frequently used modal verb
 
introductionconclusionTotal
 
Should 123951
 
Would 81624
 
May 111223
 
Could 81018
 
Might 71118
 
Table 4 demonstrates that the most frequently occurring devices used to modify statements, revealing the significance of modal verbs, with should accounting for more than 38 percent of all hedging of modal verbs in the corpus. The most frequent modal verbs are should, would and may. The top-five hedges are all used more frequently in conclusion than in introduction.
 
Generally speaking, the uses of possibility modals, such as may, might, could, in written academic discourse, contribute to the broad range of syntactic and lexical means of hedging. These verbs have a range of textual and pragmatic functions, and together with other linguistic elements, modals often serve to mark evidentiality, possibility and likelihood, strategic vagueness, and politeness in discourse (Chafe, 1986; Channell, 1994; Markkanen and Schroeder, 1997). The frequent use of modal verbs encounter in these two sections provide further support to the finding that modals are frequently used to tone down and enhance quantitative and qualitative information as well as to modulate the degree of certainty on the author’s part (Adams-Smith 1984; Kibui 1988; Tarantino 1991b;).#p#分頁標題#e#
 
(1) Considering teachers’ current low language ability and language environments not conducive to maximum language attainment in an EFL context, positive learning conditions may not be successfully provided to students. (From introduction section)
 
(2) This is an important message for designers of writing tasks: although a general topic may eliminate the possibility of unfamiliarity with the topic, it may not motivate the writers nor generate good writings. (From conclusion section)
 
These cases of ambiguity conveyed thanks to the hedge “may” in both of these two sections, which allows a more dialogic interpretation between writer and readership and convey vagueness and politeness in order to avoid confrontation between author and audience. According to Huckin (1987), conclusion describes an inside-out circle, whereas the introduction describes an outside-inside one. This “mirror-imaging” is reflected in the use each rhetorical section makes of hedges: as is found in the Introduction sections, the Discussion/Conclusion sections are most heavily marked by modal verbs. The results of the statistical analysis show that the difference between the occurrences of these two hedging categories is slightly different, but significant. It is finally worthwhile mentioning this slight difference, that is the performed functions of these modal verbs have undertaken could be still seen and explained.
 
According to Hyland, in the introduction of research articles, writers can hedge their statements when speculating about the significance of the study, mitigating anything negative in the review of prior research, indicating a gap or announcing findings. In the introduction section, modal verbs allow the researcher to establish what Swales (1990) calls an “early niche” for the research being reported (as if none of the evidence the researchers are using is conclusive) and to explain what the author’s unexplored territory consists of. When reporting previous research and their interpretations of it, writers mitigate their statements by choosing weaker verb such as modal verbs such as “may” or “might” just as the examples above, in this manner, to show there is still room for negotiation or “creating a territory” to discuss.
 
In conclusion, however, the function of modal verbs mainly lies in the author’s showing politeness, uncertainty of some conclusions or cautiousness about their finding and suggestions, not to create spaces for discussion. In this sense, based on Quirk et al. (1985) research, “possibility/ability modals” represent gradient markers of possibility and ‘‘tend to have overlapping meanings’’ that can be interchangeable in some contexts. In discourse, these verbs can qualify the meaning of the sentence and reflect the writer’s ‘‘judgment of the likelihood of the proposition it expresses being true.’’ The conclusions are almost about the findings or suggestions of the present studies, which means there are much more chances for writes to use hedges, to show their private thoughts, original or innovative thinking which again, need to be expressed very cautiously.#p#分頁標題#e#
 
3.2.2 Epistemic adjectives and adverbs
 
Adjectives, adverbs and nouns are used quite extensively to express modality in written texts (Adams mith 1984). Holmes (1988: 27) suggested that these grammatical items make up some proportion of the devices used to express epistemic modality in written discourse in the early studies. In this study, epistemic adjectives and adverbs occupy almost 27% of all types of hedging devices. Like modal verbs, adverbs and adjectives offer a wide range of means for expressing degrees of certainty, and these tend to receive some attention in the academic writing (Hyland, 1994). Obvious differences in utilizing epistemic adverbial groups and adjectives in these two sections are founded. From table 3, we can see that this category of hedging devices is the only type which appears more frequently in introduction than in conclusion section. A precise list with occurrence of this category in every 1000 words can help identity the difference more clearly.
 
Table 5 the frequency of adverbs and adjectives
 
Introduction Conclusion
 
Adverbs and adjectivesNumber Items per 1000 words Number Items per 1000 words
 
49 4.72 302.89
 
It can be concluded that writers are willing to use more adverbs and adjectives in introduction than conclusion, as the evidence here clearly showing the frequency of this hedging in introduction is more than 1.6 times of that in conclusion section.
 
There are a number of adverbs that have been employed to produce kind of meaning linked to hedges indicating restricted truthfulness or limited exactitude. Epistemic adverbs can be categorized into probability adverbs (e.g. probably, possibly), adverbs of indefinite frequency (e.g. often, frequent), adverbs of indefinite degree (e.g. mostly, mainly) and approximative adverbs (e.g. about, roughly) (Varttala, 1999). The total incidences of adverbs in two corpora are shown in Table 6. Likely, possible, largely and usually occur most frequently in the materials in the corpus. One of the epistemic adverbs “usually” typically only occurs in the introduction section in my corpus, with relatively frequent occurrences.
 
Table 6 the top six frequently used epistemic adj. and adv.
 
Epistemic adj. and adv.Introduction Conclusion
 
Likely 66
 
Possible 36
 
Largely 44
 
Usually 70
 
Relatively 33
 
Generally 41
 
These epistemic hedges are used to dilute their certainty or withhold commitment to their propositions, presenting that authors are not totally confident to their claims as indicated from the corpus:
 
In contrast, some readers spent the given time translating sentence by sentence, resulting in a recall of only the sentences at the latter part of a text, running short of time, instead of the topic sentences, which was likely the primary reason for not comprehending the text. (From introduction)#p#分頁標題#e#
 
Since it is assumed that these dialogues will most likely be practiced in a classroom setting, we as teachers have the responsibility to make sure that students will practice the forms, and that they will be alerted to the appropriate ways of using the formal features provided. (From conclusion)
 
Epistemic adverbial groups are often not related to the humility of authors or “keeping face” in front of the discourse community, instead they are “seeking to present a situation in terms of how far it varies from the ways the discourse community conventionally sees the world, either restricting the temporal or qualitative range of the claim or its generalizability” (Hyland, 1998). Adverbs such as “generally” or “usually” express this intention that writers are trying to make a generalization of situations that commonly accepted by the shared communities, which seems to fit the expectation of discourse communities and professional readers with whom hold the shared knowledge.
 
3.2.3 Epistemic lexical verbs
 
Following modal verbs, the most common means of expressing epistemic modality in written discourse is through the use of lexical verbs, often referred to as “speech act” verbs (e.g., Brown 1992), as they are used to perfume speech acts such as doubting and evaluating rather than merely to describe acts. Considering epistemic verbs, Hyland (1998b:103) once demonstrated that epistemic verbs constitute the greatest range of items. Surprisingly, epistemic lexical verbs rank the only the third highest range of items in my data. However, it is not difficult to find that lexical verbs are worthy of studying for its significance in rhetorical sections owing to its big different frequency in each corpus. They therefore imply a certain amount of doubt as evidentiary is required for their supports, which can be paraphrased as "X said that" and "It seems that". Epistemic verbs therefore mark both the mode of knowing and its source (believe, deduction, report, perception), and thereby carry implication about the reliability of knowledge itself. That is, writers can mitigate claims by indicating that they are presenting information as subjective opinion, deductive conclusion, hearsay, or as based on the evidence of their senses. Speculative and deductive categories involve epistemic judgments by the writer, allowing uncertainty and the tentative assertion of hypotheses which can be paraphrased as "I believe that" and "I conclude that". The
 
Epistemic lexical verbs carry less subjective connotations than cognition verbs such as suspect. This may explain the relatively less usage of this subjective hedge. In fact, the frequency of “suspect” is zero. Analysis of the corpus reveals that the most frequent of them are seems, suggest, argue, assume and imply.
 
Table 7 the top five frequently used epistemic lexical verbs#p#分頁標題#e#
 
Introduction Conclusion Total
 
Suggest 4 11 15
 
Seems 5510
 
Assume 303
 
Argue 11 2
 
Imply 022
 
The lexical verbs “suggest” and “seems” are most frequently adopted in both sections in the corpus. There is no distinct difference between introduction and conclusion sections in using lexical verbs from a macro view. It tends to address the writers suggestions for this study or uncertainty, just as other hedge devices have expressed and be favored by both of these two sections fairly. However, the hedge “suggest” differ obvious in these two sections. It seems that it is suitable for writers to use it in the conclusion section according to the analysis results. Suggest, as Varttala’s (2001) classification, belongs to the non-factive lexical verbs, referring to verbs where the writer does not commit himself for or against the reported proposition. What such verbs have in common in both corpora is that they can all be seen as tentative devices useful in constructing reports of research by other scholars or describing the authors' own work:
 
Moreover, recent research in second language acquisition suggests that certain traditional practices in Asia, such as memorization and form-focused learning, which were believed to be ineffective, may have an important role to play in teaching and learning. (From introduction)
 
Given the experience by foreign English teachers at the university concerning teacher-student interaction and the explanation, the author suggests that both foreign teachers and students should not only be aware of these cultural differences but also make the effort to adapt to them as much as possible so as to have more effective teaching and learning. (From conclusion)
 
Presumably we can assume that in conclusion section, there are more chances to describe the work that the authors have done and make deductions from the research or give suggestions and advices for the future studies. This somehow allows the authors to express opinions more cautiously and carefully, especially when they talk about their own works. This is not only an indication of confidence of their statement, but also a polite strategy and more importantly, a way to reduce the risk of putting wrong or non-concrete results or suggestions.
 
To sum up, from the quantitative analysis, it is evident that there are distinct differences in using hedges between introduction and conclusion section, no matter in the overall frequencies of hedges or the distribution of them. Those differences reflect the different tasks of these two sections and communicative purposes they try to pursue.
 
Discussion
 
Evident difference between introduction and conclusion sections can be seen in this study, not only from the frequency of hedges they have used, but the different categorization of hedges. This can be explained as an indication of academic writing that requires being objective and accurate, especially in the important sections of research articles to make claims. By avoiding writers’ own opinions, they intend to make the arguments more objective. This avoidance of a direct presentation of one’s opinion may not only protect one’s own position, but enlist support from authorities belonging to the same academic community.#p#分頁標題#e#
 
Introduction and conclusion undertake different functions in the whole articles and present different communicative purposes. These differences may mainly result in the different use of metadicourse devices, especially in hedges which are the most obvious strong devices to express writer’s attitude and opinions. In terms of introduction section, it serves as a transitional element of discourse, linking the work done by other members of the discipline with the work done by the writer him/herself. That is, it only demonstrates a link between what has gone before in the relevant field of research and the present work that is being reported. This function is plain in Bhatia’s argument that an introduction serves the purpose of “making ‘the present story’ relevant by placing it appropriately in the context of ‘the first story’. i.e., previous research in a particular field of study” and “is meant to motivate the present research and to ‘justify’ its publication” ( Bhatia, 1993: 82). This function coincides with Swales’ (1990) model of moves in research article introductions. The main purpose of the introduction is to state the central idea which is to be explained or defended in the essay. A secondary purpose is to provide the reader with enough background knowledge in which to set the central idea. Due to this feature, writers probably prefer to express ideas straightforward relatively compared to conclusion section when they make a claim or conclusion form the research. They don’t need to try hard to ingratiate readers or collogues with the same amount of hedges or other metadiscoures strategies as in the conclusion sections. This means possibly there is less chance or not necessary to negotiate ideas or thoughts as what they do in the conclusion.
 
This doesn’t mean that authors don’t have to be cautious or careful when they write introduction of the research articles. In fact from the analysis of the data, it is clearly that one of the categorizations of hedges is more popular in introduction. Bhatia in his work admits that “there is another subtle intention in writing introductions, which can be identified as looking for readership, indirectly promoting research” (Bhatia, 2004. p.73). Thus, it also conveys a promotional purpose of the whole articles, which definitely call for the careful and negotiated expressions. In the data we can see that epistemic adjectives and adverbs are favored by introduction. This promotional feature like what the advertisement does may be a reason contributed to the use of this type of hedges.
 
However, in the conclusion, or discussion section, as Salager-Meyer (1994) stated in his early study, the primary rhetorical function of this section is “to make claims about the research findings (i.e., to explain the statistical findings in non-statistical English), to summarize results, state conclusions and suggestions with reference to previous research and to the current work, to set further questions sometimes with possible explanations, references and future developments and applications in the field of study.” These functions ask for analysis and re-synthesis of the information which has to be transmitted through language structures with consequent influence on choice and value of grammatical categories, argument types, functions, and exposition strategies (Vázquez and Giner, 2008). Thus, the very high frequency of hedges in this section, as compared to that observed in the other sections, especially to the introduction section, is due to the fact that the claim or the interpretation of the statement is the core of the discussion or conclusion. As Tarantino (1991a) argues, every structure and function in the text must produce reasons for its justification. It is in this last section of research papers that writers speculate, argue, contrast, and extrapolate from the described results, and at the same time avoid stating results too conclusively so that the readers can note that the authors are not claiming to have the final word on the subject. Hypotheses are by nature tentative and understated, and this is reflected in their linguistic realization. It can be deduced that writers could make claims or statements carefully or give suggestions, implications or other personal opinions in the conclusion section, which needs the appropriate linguistic realization. Not only from the consideration of confidence when claiming, but also from the perspective of accuracy, interpretation made in conclusions call for highly cautious expression. The highly-frequent use of hedges here may also be a kind of protection of author themselves, such as protecting from being wrong when interpreting, or could be a pursue of responsibilities alleviation if the data they obtained is not large enough for making a concrete claim.#p#分頁標題#e#
 
Move analysis may also contribute to the different use of hedges in these two rhetorical sections. Different moves lead to a different way of expressing personal ideas and of arguing for their statements. In terms of the structure of introduction, Swales’ move structure effectively explains the appropriate writing convention in this study. An appropriate summary of move steps of journal articles can be showed as follows, according to the previous study:
 
Move 1 .Establishing a research territory
 
a. by showing that the research topic is important, central within the research field (obligatory)
 
b. by introducing and reviewing items of previous research (obligatory)
 
Move 2. Establishing a niche
 
a. by expressing the/a need/desire for research (optional)
 
b. by indicating a gap within the research area (obligatory)
 
c. by raising questions (optional)
 
Move 3. Occupying the niche
 
a. by presenting the purpose of the research or by stating the nature of the
 
Study (obligatory)
 
b. by reporting the implications of the study (obligatory)
 
c. by announcing principal findings (optional)
 
d. by indicating the structure of the paper (optional)
 
(Cited from Explicit Writing Instruction in Higher Educational Contexts: Genre Analysis of Research Article Introductions from the English Teaching and Quarterly Journals 2005)
 
In the introduction section, summarizing the previous studies, most hedging occurs in move 1, especially reviewing previous studies. It is predominantly expressed lexically such as modal verb indicating attitude and epistemic adverbs and adjectives. In move 2, hedging is done by raising questions in a passive sentence. The use of passive voice with an indirect quotation marker was common in this community. By using this, authors are likely to remove themselves from a claim which is called “distance” by Swales and Feak (2004).
 
According to the corpus composed of 36 conclusions of RAs in the field of applied linguistics, the moves of conclusions can be summarized as follows:
 
Move 1–Restating the introductory statement
 
a. stating the propose, research question or hypothesis
 
b. establishing a territory/niche
 
c. making reference to previous research
 
Move 2– Consolidating the research space
 
a. summarizing findings/results
 
b. stating methods
 
c. making reference to previous research
 
d. suggesting future research
 
e. raising questions
 
Move 3 – Summarizing the study
 
Move 4 – Components on results
 #p#分頁標題#e#
a. interpreting results
 
b. comparing results with literature
 
c. raising questions
 
Move 5 – Evaluate the study 27
 
a. Indicating limitations
 
b. Indicating significance
 
c. Evaluating methodology
 
d. Suggesting future research
 
Move 6: Making deduction from the research
 
a. Drawing implications/applications
 
b. Recommending/suggesting
 
c. Making reference to previous research
 
d. Suggesting future research
 
e. Making overall claim
 
(Cited from Conclusion as Components of research articles across Portuguese as A Native Language, English as a Native Language and English as A Foreign Langue: A Contrastive Genre Study, 2008)
 
From the different move steps of both sections, it is easily to see that moves in conclusion are much more complicated than introduction. Presumably there might be more space which enables authors to construct their arguments in conclusion sections. In order to make their journal articles to be accepted by much more readers and collogues, somehow they have to follow the moves either in introduction or conclusion. Hedges are employed when the writers move away from what can be safely assumed or experimentally demonstrated. The way writers present themselves, negotiate an argument, and engage with their readers is closely linked to the norms and expectations of professional communities. A complex and comprehensive moves or communicative purposes of conclusion sections call for more flexible and careful deception of ideas and opinions. Not only for “face saving”, but for negotiations or avoiding too subjective, writers must consider more when they express themselves in conclusion sections.
 
Conclusion and implication
 
Conclusion
 
Hedge is a complex area of language study. The study of the phenomenon in language use is an important part, which has aroused much interest of many linguists. The present study is an investigation into hedges in introduction and conclusion sections of journal articles within the analytical framework proposed by the author. Major findings are concluded as follows:
 
Through this comparative study, the author expects to obtain a comprehensive observation about similarities and differences of hedges in introduction and conclusion sections of applied linguistic journal articles. Based on the results, this dissertation indicates the following similarities. First, both of the introduction and conclusion of journal articles have employed all types of hedges, that is, modal verbs, epistemic lexical verbs, epistemic adjective and adverbs modal nouns expressing possibility, introductory phrase expressing author’s personal projection, if clauses and compound hedges. Second, as a whole, more hedges occur in conclusion sections than introduction section due to the unique features of conclusions. Finally, in terms of the categorization of hedges, Modal verbs are the most favorable devices when hedging both in introduction and conclusion. Model verbs, epistemic adjectives and adverbs and epistemic lexical verbs are predominately used in these two sections. Adjectives and adverbs however, are more popular in introduction. The rest of the five categories of hedges are all used more frequently in conclusions than in introductions.#p#分頁標題#e#
 
Reasons attribute to these differences may be due to the different essences of these two sections—interpretation of conclusions and description of introductions. Moreover, conclusions are often companied by suggestions or other recommendations, which give authors more opportunities to hedging themselves. By contrast, introductions usually describe works from others or writer’s own, describing results in a simple and straightforward way.
 
Implication
 
It is certain that hedges are important communicative means for both L1 and L2 writers. The appropriate use of hedges can help writers develop academic arguments and establish relationship with their readers. Some suggestions may therefore help L2 learners to employ hedges in their writing.
 
Because of the dominant status of English language in academic world, and the important role of an abstract as an indispensable part of an academic writing, the: competence of using hedges appropriately is necessary for writers who want to gain achievement and reputation in academic world. When teaching second language writing, teachers must help students be aware of the rules not only at the grammatical level but also at the level of rhetorical features.
 
In the classroom, more focuses on high frequency hedges are beneficial for student. Modal verbs like would, may, will should be provided to students. Not only are these the most numerically important signals of hedges in both of these two sections, they are also likely to be the easiest to acquire. Learners could treat these hedges as a means to express themselves in a more tentative and polite way. This study suggests that adverbs in particular have an extremely high frequency. The study of a range of high frequency adverbs such probably, possibly and other hedges should be focused by teachers. And teachers can provide some typical sentences with these most frequent hedges for the learners' better understanding of the realization of hedges in introduction and conclusion writing.
 
According to the results of this study, it can be conclude that conclusions favor much more hedges than introduction. When writing essays or dissertations, teachers must arouse the students’ awareness of using hedging in these two sections.
 
Limitations
 
This thesis attempts to study hedges with comparative approach. Unavoidably, it
 
does have some limitations.
 
Firstly, in the academic field of hedges study, there is no generally accepted categorization of hedges. Due to different purposes in their studies, many scholars categorized and defined hedges from different angles. Hyland’s categorization of hedges are the most accepted, but difficult to code. Considering the nature of the corpus involved in the present study and the purpose of conducting this research, I categorized the hedges into 7 groups, combined with other scholars’ classification. It's hard to say the categorization is in a totally scientific and reasonable way.#p#分頁標題#e#
 
Secondly, although 21 pairs of journal articles are selected to establish a corpus, the data collected is far from being exhaustive. The present study is limited by its simple calculating method and small corpora. It would have been more convincing if I have conducted observation on more materials.
 
It is clear that this study has offered a descriptive account of hedges in journal articles inL from the comparative study; it has also raised a number of questions which are unclear and remain to be resolved in the future. The conclusion is valid in the illustrated fields, applied linguistics articles from the Journal of Asian TEFL, but there might be difference between the different fields and new conclusions might be drawn about it. Future research in larger corpus with more advanced approaches will be necessary and beneficial.
 
Reference
 
Abdi, R. 2002. Interpersonal metadiscourse: an indicator of interaction and identity. Discourse Studies 4, 139–145
 
Abdollahzadeh, E, 2001. Native and non-native writers’ use of textual metadiscourse in ELT papers. MA thesis. University of Tehran, Tehran.
 
Abdollahzadeh. E.2011. Poring over the findings: Interpersonal authorial engagement in applied linguistics papers. Journal of Pragmatics 43, 288–297
 
Adams-Smith, D. E. (1983). Prescribing and general practice: Style in medical journals. Paper presented at TESOL Convention.
 
Adams-Smith, D. E. (1984). Medical discourse: Aspects of author’s comments. The ESP toucan. 3. 25-36.
 
Bhatia, Vijay Kuram, 1993, Analysing Genre – Language Use in Professional Settings, London, Longman.
 
Bhatia, Vijay Kuram, 2004, Worlds of Written Discourse. A Genre-based Approach, London, Continuum.
 
Beverly, A & Lewin, 2005. Hedging: an exploratory study of authors’ and readers’ identification of ‘toning down’ in scientific abstracts. English for specific Purposes, 4.163-178
 
Brett, P. 1994. A genre analysis of the results section of sociology articles. English for Specific Purposes, 13, 47–59.
 
Brown, K. (1992). Describing modality in English. In R. B. Brurnfit & C. Brumtit (Eds.), Applied linguistics and English language. Teaching Review of ELT Vol 2.
 
Channell,J. Vague Language. Oxford University Press, 1994
 
Coates, J. (1983). The semantics of the modal auxiliaries. Beckenham, UK: Croom Helm
 
Del Saz-Rubio, M. M. 2007. English discourse markers of reformulation: An analysis and description. Bern, Switzerland: Peter Lang.
 
Del Saz-Rubio, M. M. 2011. A pragmatic approach to the macro-structure and metadiscoursal features of research article introductions in the field of Agricultural Sciences. English for Specific Purposes 30. p. 258–271#p#分頁標題#e#
 
Gillaerts, P and Van de Velde, F., 2010. Interactional metadiscourse in research article abstracts. Journal of English for Academic purposes 9, 128–139.
 
Goffman, E. 1967. Interaction Ritual. Garden City, NY: Doubleday.
 
Hinkel, E. 2009. The effects of essay topics on modal verb uses in L1 and L2 academic writing. Journal of Pragmatics. 41. 667–683
 
Hill, S. S., Soppelsa, B. F., West, G. K. (1982). Teaching ESL students to read and write experimental research papers. TESOL Quarterly, 16/3, 333–347.
 
Holmes, J. 1982. Expressing doubt and certainty in English. RELC Journal 13. 9–28. Holmes, J. 1984. Modifying illocutionary force. Journal of Pragmatics 8. 345–365.
 
Holmes, J. 1988. Doubt and certainty in ESL textbooks. Applied Linguistics 9. 21–44.
 
Holmes, J. 1990. Hedges and boosters in women’s and men’s speech. Language & Communication 10. 185–205
 
Hopkins, Andy, Dudley-Evans &Tony, 1988. A genre-based investigation of the discussion sections in articles and dissertations. English for Specific Purposes. 7. 113–122.
 
Hu G.H & Cao.F. 2011, Hedging and boosting in abstracts of applied linguistics articles Journal of Pragmatics 43, 2795-2809
 
Hyland, K. 1996. Talking to the Academy : Forms of hedging in Science research articles. Written Communication. 13, 251-281
 
Hyland, K. 1998. Boosting, hedging and the negotiation of academic knowledge. Text 18. 349-382
 
Hyland, K. 1998. Hedging in academic research articles. Amsterdam and Philadelphia, PA: John Benjamins Publishing Company,
 
Hyland, K. and Milton, J., 1997. Qualification and certainty in L1 and L2 students’ writing. Journal of Second Language Writing 6, 183–205.
 
Hyland, K and Tse, P. 2004. Metadiscourse in academic writing: a reappraisal. Applied Linguistics 25, 156–177
 
Kibui, A. (1988). Aspects of hedgings in the discussion of medical research discourse (LSU ESP Collection). MSc in TESP Dissertation, University of Aston in Birmingham.
 
Markkanen, R. & Schroeder, H., 1997. Hedging: a challenge for pragmatics and discourse analysis. In: Markkanen, R., Schroder, H. (Eds.), Hedging and discourse: approaches to the analysis of a pragmatic phenomenon in academic texts. Walter de Gruyter, Berlin, p. 3–20.
 
Myers, G. 1989. The pragmatics of politeness in scientific articles. Applied Linguistics 10, 1–35.
 
Olson, Leslie, Huckin & Thomas N., 1990. Point-driven understanding in engineering lecture comprehension. English for Specific Purposes. 9. 33–47.
 
Ozturk, I. (2007). The textual organisation of research article introductions in applied linguistics: Variability within a single discipline. English for Specific Purposes, 26(1), 25–38.#p#分頁標題#e#
 
Vande Kopple, W.J. 1985. Some exploratory discourse on metadiscourse. College, Composition and Communication 36, 82–93.
 
Qin Yongli. 2007. A Study of Hedging in English Abstracts of Scientific Research Articles. Shang Dong University,
 
Salager-Meyer, Francois, 1994. Hedges and textual communicative function in medical English written discourse. English for Specific Purposes. 3. 49–170. Salager-Meyer, F. 1994. Hedges and Textual Communicative Function in Medical English Written Discourse. English for Specific Purposes. 13.149- 170,
 
Samraj. B. 2008. An exploration of a genre set: Research article abstracts and introductions in two disciplines. Journal of English for Academic Purposes. 7. 55-67
 
Swales, J, 1990. Genre Analysis: English for Specific Purposes in Academic and Research Settings. Cambridge University Press, Cambridge.
 
Swales, J. M., & Najjar, H. 1987. The writing of research article introductions. Written Communication. 4. 175-191
 
Swales, J, & Feak, C, 2004. Academic Writing for Graduate Students: Essential Tasks and Skills. University of Michigan Press, Ann Arbor.
 
Tarantino, M. (1991a). Scientific English: Qualitative factors via modern rhetoric.
 
UNESCO-ALSED LSP Newsletter, 13(3), 51-60.
 
Thompson, G., & Ye, Y. (1991). Evaluation in the reporting verbs in academic papers. Applied Linguistics, 12, 365–381.
 
Vázquez.J & Giner.D. 2008. Beyond Mood and Modality: Epistemic Modality Markers as Hedges in Research Articles. A Cross-Disciplinary Study. Revista Alicantina de Estudios Ingleses 21. 171-190
 
West, G. K. (1980). That-nominals constructions in traditional rhetorical divisions of scientific research papers. TESOL Quarterly, 14. 483-489.
 
Williams, I. A. (1999). Results sections of medical research articles: analysis of rhetorical categories for pedagogical purposes. English for Specific Purposes. 18. 347–366.
 
Yang. R& Allison, D. 2003. Research articles in applied linguistics: moving from results to conclusions. English for Specific Purposes, 22. 365-385.
 
Yearley, S.1981. Textual persuasion: The role of social accounting in the constructions of scientific arguments. Journal of Sociology of Science, 11. 409-435.