引導性子句立場表達: 以語料庫為本的英語母語者及非母語者學術英文寫作分析

全文

(1)國立臺灣師範大學英語學系碩士論. 文. Master’s Thesis Department of English National Taiwan Normal University. 引導性子句立場表達: 以語料庫為本的英語母語者及非母語者學術英文寫作分析 Stance Expressed in Evaluative that-Clause: A Corpus-based Analysis on Academic Writing Written by Native and Nonnative Speakers of English. 指導教授：陳浩然博士 Advisor: Dr. Hao-Jan Chen 研究生：許嘉文 Student: Jia-Wen Hsu. 中華民國一○三年六月 June, 2014.

(2) 中文摘要本研究探討母語人士與非母語人士在學術英文寫作中以引導性子句所呈現的立場標記語的情況，並給予教學建議，幫助第二語言學習者適當地表達立場。對比研究文獻表示，近年來立場表達的使用情況在學術文章中已為廣泛，但第二語言學習者卻很少在這方面得到指導。這使本研究（1）探討母語者在學術文章中所呈現的立場表達程度，（2）對比臺灣學者和學生的立場建構上，與母語者有何相異特點。本研究的三個對比語料庫中的語料是由應用語言學的學術文章所構建而成的，蒐集了 50 篇由母語人士所撰寫的期刊文章，50 篇由臺灣學者所撰寫的期刊文章，以及 25 篇臺灣研究生的學術論文。完成語料的匯集後，從而進行定量和定性的研究探討。透過語料的詳細比較分析，也可找出臺灣研究生過度使用與未充分利用的立場標記語。（1）引導性子句的立場表達被廣泛地應用於學術寫作中，而研究結果顯示：與專業期刊學者相比，臺灣研究生所使用的立場標記頻率最為頻繁和廣泛。（2）（3）母兩組專業學者與研究生都傾向於使用動詞來做為引導性子句的主要述語。語人士使用較多具有模糊、可能性的標記語來表達試探性的立場，而臺灣研究生（4）普遍則過度使用肯定性與情態性的標記語，使文章較有果斷、說服的效果。（5）臺灣來說，臺灣學者及研究生傾向於使用虛擬主詞來避免直接的評論責任。研究生大量地使用肯定性的立場標記使文章充分表達自信，卻少量地利用試探性標記來呈現客觀。綜上所述，本文認為立場表達在學術寫作中是不可或缺的，但在沒有任何教學幫助下，我們不該期望學生能夠適當地、無誤地使用立場標記語。有限的立場表達方式和不夠充分的語言學習，使母語人士與非母語人士在表達立場上有顯著的差異。本文提供了學術寫作的教學建議，從而使學生增進立場表達的認識，並了解其在學術寫作中的涵義，進而對此類研究作出小小的貢獻。關鍵字: 語料庫分析、立場標記、引導性子句 i.

(3) ABSTRACT The aim of this thesis is to investigate and identify the characteristics of stance usage through the realization of that-complement clauses in the writings of native and non-native writers, and to make recommendations for writing instructors to assist L2 learners in controlling stance expressions in research reports. Findings of comparative studies are presented to show that there is a growing demand for stance conveyance in academic writing, but at present L2 students receive very little guidance in this matter. Such shortage arouse current study (1) to empirically examine the extent that native speakers of English display stance and show viewpoint in academic writings, (2) to discover whether Taiwanese writers, either expert or student, convey stance in ways different from native speakers. Three corpora of academic writing written in the field of applied linguistics were built by collecting 50 journal articles written by native speakers of English, 50 journal articles by local researchers in Taiwan, and 25 master theses by Taiwan graduate students. After the compilation of the corpora, quantitative analysis was conducted to compare and contrast the use of stance-that markers among writers. Through detailed comparison of multiple corpora, L2 students’ overused and underused stance-that markers can also be identified. Several discoveries were yielded: (1) evaluative that-clause was widely applied in academic writing for stance marking, with L2 students conveyed stance most frequently and extensively as compared to the journal writers. (2) All three groups inclined to use verbal predicates most frequently to control that-clauses. (3) While native expert writers projected greater extent of tentativeness through likelihood stance, L2 students exposed more assertiveness and forcefulness with excessive factive and attitudinal markers. (4) Taiwanese writers generally made considerable use of concealed subjects as an attempt to avoid taking direct responsibility of the ii.

(4) evaluation. (5) L2 students were found to have heavy reliance on factive markers to project assurance, and low preference on tentative stance to present objectivity. In conclusion, the thesis argues that stance marking in academic writing is essential but that students should not be expected to be able to appropriately employ stance without assistance. Limited repertoire of stance expressions and insufficient language input may have contributed to the mismatch exposed among the three groups of writers. This thesis hopes to offer educational implications for enhancing L2 writers’ writing quality, and thus make a small contribution to understanding the significance of stance in academic writing. Possible directions for future research are also provided.. Keywords: corpus analysis, stance taking, evaluative that-clause. iii.

(5) ACKNOWLEDGMENT This work would not have been possible without the help and support of many people. Frist and foremost, I would like to express my sincere gratitude to my advisor, Prof. Hao-Jan Chen for the continuous support of my research project, not to mention his patience, encouragement, and unsurpassed knowledge of corpus linguistics. Without his great guidance, I would not be able to put this thesis together. I am very grateful for the corpus computer software tools that Prof. Hao-Jan Chen provided for the research study. It was particular kind of him to show me the efficient ways to collect corpus data. I could not have imagined having a better mentor during my graduate years, for he has taught me more than I could ever give him credit for here. I would also like to thank my thesis committee members—Dr. Pei Chin Chang and Dr. Hsiao-hui Yang. Dr. Chang and Dr. Yang put great effort in proofreading my thesis and provided many insightful comments. Their advice and encouragement have been invaluable, for which I am extremely grateful. My gratitude also goes to my fellow classmates and friends, for the stimulating discussions and for all the fun we had during the past three years. Their kind support had given me strength during the thesis writing process. Last but not least, I would like to thank my parents, whose unconditional love and support are with me in whatever I pursue. Their patience and understanding during my graduate study years are greatly appreciated. I am indebted to my dear sister, who has always been my ultimate role model throughout my life. She has always been there for me whenever I need her care and support. This research would not have been possible without their great help and encouragement.. iv.

(6) TABLE OF CONTENT ABSTRACT (Chinese)………………………………………………...……………….. i ABSTRACT (English)…………………………………………………..……………... ii ACKNOWLEDGMENT………………………………………………..……………… iv TABLE OF CONTENT…………………………………………………...…..………... v LIST OF TABLES………………………………………………………....….………... viii LIST OF FIGURES………………………………………………………….……….… x CHAPTER ONE—INTRODUCTION………...……………………….….……..…...... 1 1.1 Background………………………………………………………...………….. 1 1.2 Purpose of the Study……………………………………………..…….……… 5 1.3 Significance of the Study………………………………………...……….…… 7 1.4 Definition of Key Term………………………………………..……….……… 9 1.5 Structure of the Thesis……………………………………...……………..…… 9 CHAPTER TWO—LITERATURE REVIEW…………………………………...…….. 11 2.1 Stance in English Academic Writing……………..…………………..……..… 11 2.1.1 Categorizing Stance………………………..……………….................. 13 2.1.2 Grammatical Realization of Stance………….….…………………..….. 21 Use of Modals………………………..……….……………….…….… 22 Use of Stance Adverbs……………………….……………….…..…… 25 Use of Extraposed Clauses…………………….……….………....…... 28 2.2 Evaluative That-Clause for Stance Marking…………….………………...……30 2.2.1 Theoretical Framework of Evaluative that………..…………..…..…… 2.2.2 Native Writers’ Preference on That–Clause Uses…..……..….…….…. 2.2.3 Learners’ That-Clause Usage Pattern………………..…..…….….…… CHAPTER THREE—METHODOLOGY…………………………...…..……..………. 30 34 36 41. 3.1 Corpus Data…………………………………………………….…….….…….. 41 3.1.1 The Reference Corpus………………………….……..…...…….….…. 42 3.1.2 Non-native Journal Corpus……………………………..…..…….…..... 43 3.1.3 The Learner Corpus………………..……………………..…..……..…. 44 3.2 Instruments………………………………………………………..…..….….... 45 3.3 Identification of Stance Markers……………………..……………..…..…...… 46 3.4 Analytical Methods…………………………………………………..…....….. 47 v.

(7) CHAPTER FOUR—RESULTS AND DISCUSSION…………….…………..…..…… 49 4.1 The Identified Stance Markers in the Three Corpora……………………......... 49 4.1.1 Grammatical Categorization of Stance-that Markers………....……….. 53 4.1.2 Exclusive Stance-that Markers in Each Corpus……………..….....…... 56 4.1.3 Top Twenty Stance-that Markers……………………………..….…..... 59 Suggest-that…………………………………………………..…..….. 63 Argue-that………………………..…………………………..………. 63 Reveal-that……………………….……………………………...…… 64 Point out-that……………………..………………………………..… 65 Claim-that…………………………..…………………………..……. 66 Fact-that……………………………..…………………………..…… 67 4.2 The Functional Categorization of Stance-that Markers………..…………..…. 68 4.3 Evaluative Source………………………………………………..…………..... 73 4.3.1 Evaluative Source Distribution of Likelihood Verbs……..………...…. 75 4.3.2 Evaluative Source Distribution of Factive Verbs…………..……...…… 76 4.4 Overused and Underused Stance-that Markers……………………..…..…...... 77 4.4.1 Stance Markers Overused by NNS-S when Compared with NS-J….... 78 Agree-that……………………………………………………..….….. 80 Disagree-that……………………………………………………...…. 82 Think-that…………………………………………………..……..…. 83 Believe-that…………………………………………………..……..... 83 4.4.2 Stance Markers Underused by NNS-S when Compared with NS-J…... 84 4.4.3 Stance Markers Overused by NNS-S when compared with NNS-J..…. 86 State-that…………………………………………………...……….... 87 Agree-that………………………………………………………….… 88 4.4.4 Stance Markers Underused by NNS-S when compared with NNS-J…. 89 4.5 Discussion of the Use of Stance-that Markers……………………………...… 89 4.5.1 Discussion of Structural Distribution of Stance Markers in 3 Corpora 90 4.5.2 Discussion of Functional Distribution of Stance Markers in 3 Corpora 92 4.5.3 Discussion of Evaluative Source Attribution…………………………... 93 4.5.4 Discussion of Overused and Underused Stance markers……………… 95 CHAPTER FIVE—CONCLUSION…………..……………………………………….. 98 5.1 Major Findings………………………………………………………………... 98 5.2 Pedagogical Implications……………………………………………..…..…… 100 5.3 Limitations of the Study………………………………………………….…….101 vi.

(8) 5.4 Suggestions for Future Research………………………………………..…… 102 REFERENCE…………………………………………………………………………. 103 APPENDIX………………………………………………………………………….... 109. vii.

(9) LIST OF TABLES Table 2.1 Biber’s Taxonomy of Stance Markers…………………………………...... 20 Table 2.2 Hyland and Tse’s Classification of Sentences Containing Evaluative that.. 33 Table 2.3 Synthesized Theoretical Framework of Stance-that Clauses with Example Sentences…………………………………………………………………. 40 Table 3.1 Overview of the Three Academic Corpora………………………….……. 42 Table 3.2 Journals Included in the NS-E Corpus……………………………..……... 43 Table 3.3 Journals Included in the NNS-E Corpus…………………………..……… 44 Table 3.4 MA Theses Included in the NNS-S Corpus…………………….………… 44 Table 4.1 Distribution of Grammatical Category of Stance-that in 3 Corpora……… 50 Table 4.2 Stance Coverage in 3 Corpora by Frequency of Occurrences……………. 50 Table 4.3 With Cut-off Criteria, Distribution of Stance-that in 3 Corpora………….. 51 Table 4.4 With Cut-off Criteria, Stance Coverage in 3 Corpora by Frequency of Occurrences………………………………………………………………. 51 Table 4.5 26 Mutual Stance-that Markers Across Three Corpora……………………. 55 Table 4.6 14 Exclusive Stance-that Markers in NS-J……………………………….. 57 Table 4.7 3 Exclusive Stance-that Markers in NNS-J………………………………. 58 Table 4.8 16 Exclusive Stance-that Markers in NNS-S……………………………... 58 Table 4.9 Top 20 Frequent Stance-that Markers in NS-J (size= 336,923 words)…… 60 Table 4.10 Top 20 Frequent Stance-that Markers in NNS-J (size= 316,946 words)… 61 Table 4.11 Top 20 Frequent Stance-that Markers in NNS-S (size= 371,064 words)... 62 Table 4.12a The Frequency of suggest-that…………………………………………. 63 Table 4.12b The Frequency of argue-that…………………………………………… 64 Table 4.12c The Frequency of reveal-that………………………………………………. 65 Table 4.12d Comparison of Reveal-that and Other Research-Act Verbs in NS-J and. viii.

(10) NNS-S………………………………………………………………… 65 Table 4.12e The Frequency of point out-that……………………………………….. 66 Table 4.12f The Frequency of claim-that……………………………………………. 67 Table 4.12g The Frequency of fact-that…………..………………………………… 68 Table 4.13 Distribution of Functions of Stance-that in 3 Corpora…………………... 69 Table 4.14a The Shared Stance-that Markers between NS-J and NNS-J…………… 72 Table 4.14b The Shared Stance-that Markers between NS-J and NNS-J…………… 72 Table 4.15 Overview of the Distribution of Evaluative Source…………………….. 74 Table 4.16 Evaluative Source Distribution of Likelihood Verbs……………………. 75 Table 4.17 Evaluative Source Distribution of Factive Verbs………………………... 76 Table 4.18 NNS-S’ Overused Stance-that Markers as Compared with NS-J……….. 78 Table 4.19 The Frequency of believe-that…………………………………………… 84 Table 4.20 NNS-S’ Underused Stance-that Markers as Compared with NS-J……… 84 Table 4.21 NNS-S’ Overused Stance-that Markers as Compared with NNS-J…….. 87 Table 4.22 NNS-S’ Underused Stance-that Markers as Compared with NNS-J……. 89 Table 4.23 Overview of NNS-S’ Most Overused/Underused Stance-that Marker Types……………………………………………………………………. 96. ix.

(11) LIST OF FIGURES Figure 2.1 Key Resources of Academic Interaction……………………………...… 16 Figure 2.2 Evaluative that Sentence Structure……………………………………... 32 Figure 4.1 Venn Diagram of Stance-that Marker Distribution…………….………. 54 Figure 4.2 Functional Distributions of Stance Markers in Three Corpora………… 69 Figure 4.3 Venn Diagram of Stance Marker Distribution between NS-J and NNS-S………………………………………………………………..…. 78 Figure 4.4 Venn Diagram of Stance Marker Distribution between NNS-J and NNS-S…………………………………………………………………... 86. x.

(12) CHAPTER ONE Introduction. 1.1 Background Stance-taking has become one of the most fundamental and multifaceted activities human accomplished through language. Recently, numerous linguists have been conducting research studies investigating the mechanisms used by writers to express personal feelings and assessments. Investigating how writers position themselves either subjectively or objectively in academic writing has been a popular area of research over the years. Academic writing classes are often taught with the idea of delivering content with a humble voice, that writers “must leave their personalities at the door and subordinate their views, actions and personality to its rigid conventions of anonymity” (Hyland, 2002b, p.351). It has been believed that subjective tone in academic writing may weaken the argument and easily lay oneself open to disagreement or criticism (Morley, 2007). Self-effacement has been treated as the synonym for illustrating social decorum in the academic writing circles (Morley, 2007). This style of writing creates an illusion of objectivity and displays a deliberate tone. Nevertheless, an illusion of impersonality is expressed as the writer is absenting themselves from what they compose. Until quite recently, adding personality into scientific writing was frown upon by EAP teachers and academic writing guidebooks, as the overall tone would be seen as unscholarly and vulgar. Several studies have documented that in fact there are times when personal language may be more appropriate, or at times unavoidable, as it adds flavor to the writing, revealing legible personal opinion, rather than someone else’s, in the concluding remarks (Hyland, 2002a; Hyland & Guinda, 2012). Thus, how to strike a balance between the two. 1.

(13) writing styles becomes the issue that every writer should pay attention to. Academic writing is no longer merely an exercise of delivering information and conveying content; rhetorically convincing arguments often persuade reader audience with the expression of writer’s attitude, view, or evaluation. It has been stated that the final aim of conducting academic research is to make one’s written work become a prominent literature resource in the target research field and also professional discourse community (Fløttum, 2012). Not seeing academic discourse as neutral and objective as established impression puts it, Fløttum (2012) proposes that “scientific observations and findings cannot speak by themselves, without any ‘help’ from or any traces of the author(s) behind them” (p.221). This states the importance of delivering traces of voice and stance in the academic contribution, when writing becomes a product of scientific and political discourse. As Hyland (2002b) states, “academic writing is not a single undifferentiated mass, but a variety of subject-specific literacies” (p. 352). It is not surprising to find that L1 English writers indicate stance differently across different disciplines (Charles, 2006; 2007; Hyland, 2002a; 2005; McGrath & Kuteeva, 2012; Silver, 2012; Wharton, 2012), and registers (Biber, 2004; 2006; Hewings, 2012; Vázquez Orta, 2010). Over the years, scholars have also been exploring how the L1 writers display stance through different grammatical structural features such as modals, intensifiers, and adverbials (Biber, 2004; 2006; Biber et al., 1999; Charles, 2007; Hyland, 2002a; Paradis, 2003) and semantic functional realizations as in the form of hedges, booster, directives, and evidentials (Biber, 2004; Hyland, 1998; 2002a; 2005; Martin & White, 2005; McCready & Ogata, 2007). It seems that writers select word choices according to their own disciplinary literacies. In other words, writers write in a way that makes the most sense to their readers in their community of practice (Hyland, 2002b). EAP researchers have extended this line of research not only by comparing and 2.

(14) contrasting the differing ways native-English (L1) expert writers and L1 student writers present argument (Biber, 2006), but also by drawing attention to how second language (L2) student writers manifest their viewpoint in academic prose (Hyland, 2002a; 2002b; 2005). Learning to write professionally and convincingly entails the correct practice of conveying one’s stance in academic writing. This writing skill may be difficult for L2 writers to master, not only should they break the language proficiency barriers, but also acquaint themselves with the accepted writing conventions in varying academic fields. Thus, we may postulate that L2 English speakers could encounter some difficulties when they are either expressing or interpreting stance uses. Even if one can comprehend every vocabulary in its sentence structure, the hidden meaning behind the words may not fully be inferred correctly. With flawed interpretation of others’ argument, the misunderstanding will cause one to lose the opportunity to consolidate knowledge framework and impede critical thinking ability. As English has been accepted as the lingua franca in the world today, the fact that it has privileged position in academic circles “[leads] to analogous discoursal behaviours” (Fløttum, 2012, p.222). Nonetheless, writing English academic papers might be a difficult task for non-native English speakers as they are expected to produce native-like writing, in which entails unfamiliar rhetorical conventions (Casanave, 2003). As Hood noticed, when published texts and L2 student texts are being compared in regards to structure patterning, published texts show more instances of encoding attitude by projecting stance signals to their readers with consistent position on the argument, whereas L2 student texts often forego opportunities to communicate such discourse function (Hood, 2004). We may argue that the contributing reasons of such difference may lie in between the area of both writing proficiency and academia status. Unfamiliarity with the target writing convention may 3.

(15) influence how well a non-native English writer constructs, expresses, and interacts with the reader and the wider community. In other words, despite the influence of language proficiency on writing skills, the real problem at heart perhaps involves one’s failure on meeting reader expectations. To increase the readability of one’s written work, one must adopt rhetorical practices, norms, and ideologies of the discipline and community that he/she wants to participate in. In this sense, writing becomes a process of social engagement (Zarei & Mansoori, 2011). Mismatch expectations between readers and writers is often the underlying reason that causes the writing as illogical in the eyes of the readers. For that reason, current study is motivated to investigate if Taiwan scholars follow the same pattern of stance usage as native English writers do, by examining academic writings composed by native and non-native writers. The present study extends previous research in two ways. First, it cross compares and contrasts the stance usage pattern in three dimensions: 1) L1 expert writers and L2 expert writers, 2) L2 expert writers and L2 student writers, and also 3) L1 expert writers and L2 student writers. By comparing three groups, major patterns of variation will be analyzed across differing academia status and language proficiency. Second, many studies have focused on stance expression in the realization of modals and semi-modals (Biber, 2006; Hyland, 2002a; Vázquez Orta, 2010), evaluative adjectives (also termed as intensifiers and grading adverbs) (Henderson & Barr, 2010), and stance adverbials (Ai, 2012; Biber, 2006; Biber & Finegan, 1988; Biber et al.,1999; Tseronis, 2009); few has examined clausal marking of stance through the forms of that-clause that are paired with stance verb, adjectives, and nouns. Thus, the present research aims to address the relative paucity of empirical studies of stance through clausal expressions in the form of that-complement clause.. 4.

(16) 1.2 Purpose of the Study As academic writing style over the years has evolved from delivering merely factual information into sprinkling authorial stance with a pinch of writer’s opinion and evaluation across textual passages, researchers are particularly interested in knowing the distribution ratio of stance markings in academic prose. To unveil such phenomenon, the first purpose of current study is to examine to what extent the native speakers of English display stance and show viewpoint in academic research journal articles. L1 writing in this case is treated as the reference data, and L2 writers can use it as a resource of expert practice to enhance and refine their own. The second purpose is to explore the usage pattern of authorial stance in academic writings written by L2 expert and L2 student writers, and further discover the discrepancies in how authorial stance is conveyed between writers in L1 and L2 contexts. Both Biber (2006) and Hyland (2002a; 2005) have confirmed that there are considerable differences among disciplines and registers. Hyland has also pointed out that hard sciences generally favor impersonal style by downplaying personal role to highlight the research study, whereas soft sciences such as humanities and social sciences tend to express evaluation and stance to “establish a credible scholarly identity and underline what they have to say” (Hyland, 2002b, p.353). Nevertheless, it may not be possible to expand the scope of investigation to uncover stance pattern in various study fields, as not every discipline is delivered through the English language in Taiwan. In order not to make English language and writing proficiency as the confounding factors, current study intends to only collect writings in applied linguistics or English language teaching, in which writers carry certain advanced level of standard English. Stance devices are searched to determine the different usage pattern in academic writing among expert native English writers, professional 5.

(17) non-native English writers in Taiwan, and graduate student writers in Taiwan, in the field of applied linguistics and TESOL (teachers of English to speakers of other languages). By classifying the stance markers into structural and functional categories, it is easier to discover the general usage pattern and preferences. Through preliminary statistical analysis, whether stance is conveyed more or less by whichever groups of writers would be known. The study may also indicate the overall percentage of stance device coverage in academic writings. It is hypothesized that stance would be conveyed more by native speakers comparatively, as they are more accustomed to incorporate personal viewpoints in written productions. Moreover, at this point of postulation, it is plausible to assume the underlying factor that causes differences to exist between professional and student writers may ascribe to the proficiency and academic status; whereas for native and non-native professional researchers, the discrepancy, if there is any, may due to L1 rhetorical convention that interfere with or transfer on the writing in L2, as demonstrated in the tension of politeness convention in different national or disciplinary ethos (Connor, 2002; Fløttum, 2012; Kuntjara, 2004; Rass, 2011; Xing et al., 2008). With the hope of making some contribution to the linguistic corpus field of research, current study is aimed to provide insights for L2 writers by illustrating how native speakers express viewpoint in academic writings; we can also further incorporate the corpus data to refine L2 writers’ writing style. Based on the objectives of the current study, the research questions formulated for this research are as follows: (1) Which grammatical category of stance markers (verb, adjective, or noun) in the form of that-complement clauses is used most frequently in the academic writing in the discipline of applied linguistics among the three groups: 6.

(18) 1) published journal articles written by native speakers of English 2) published journal articles written by non-native speakers of English 3) MA theses written by Taiwan graduate students (2) Among the stance markers in that-clausal structure, which functional category (e.g. likelihood, factive, nonfactive, attitudinal) is used most frequently in the academic writing in the discipline of applied linguistics among the three groups: 1) published journal articles written by native speakers of English 2) published journal articles written by non-native speakers of English 3) MA theses written by Taiwan graduate students (3) Are there any similarities or differences in the use of evaluative source (e.g. human, inanimate entity, concealed) among the three groups of writers? 1) published journal articles written by native speakers of English 2) published journal articles written by non-native speakers of English 3) MA theses written by Taiwan graduate students (4) Do the native speakers use more stance markers in academic writings than the non-native speakers overall? If so, to what extent? Is there any preference for native speakers to convey their standpoint?. 1.3 Significance of the Study This paper aims to conduct an investigation of stance based on a lexico-grammatical and discoursal semantic analysis of a wide class of expressions in English academic writings. The goal of this current corpus-based study is to uncover major patterns of stance usage variation among two groups of expert writers, native and non-native, and one group of student writers in Taiwan. Through quantitative. 7.

(19) corpus-based method, it is feasible to trigger and calculate the overall stance expressed lexically in a text. With concordancing function in the corpus tool, data is also being examined qualitatively by analyzing how the expressions are relatively at the sentential level contributing to the overall pragmatic purpose that is intended by writers. To reveal the similarities or differences of writing styles between native speakers and non-native speakers, previous studies have mainly focused their research on how L2 learners convey argument or carry opinion differently from the L1 professionals (Hyland, 2002a; Hyland & Tse, 2005b; Vázquez Orta, 2010). This way, it is easy to reveal the area or features that non-native speakers should polish in their written work. Novice writers may not be as confident to present their position by proclaiming or disclaiming; in other words, the tone of their prose tends to be more general. Nonetheless, the discrepancies of the writing style may attribute to writers’ English proficiency level and their status position in the academia. It may be more feasible to also compare and contrast the written work of the writers with similar academic status and English proficiency level. Therefore, instead of only contrasting the work between L1 experts and L2 learners as the previous studies, academic writings of L2 professionals are counted in to form a more completed picture for stance investigation. Present research also differs from the previous studies in the investigation of clausal marking of stance, which has not received much research attention in the literature. Moreover, many studies probing stance expression in written context are done in Anglo-medium academic circles. To date, few has explored such topic in the context of Taiwan.. 8.

(20) 1.4 Definition of Key Term Stance Stance, as defined by Hyland, refers to “the ways academic annotate their texts to comment on the possible accuracy or creditability of a claim, the extent they want to commit themselves to it, or the attitude they want to convey to an entity, a proposition or the reader” (Hyland, 2005, p.178). The ultimate goal for writer is to enable the composed work piece makes sense to potential readers and contributes to the wider community. Therefore, it is essential for writers to take readers into considerations. As Vázquez Orta states (2010, p.81):. Academic writers intrude and make comments on the information they convey through their texts; they convey judgements, align themselves with readers and express solidarity by anticipating objections and responding to an imagined dialogue with others, thus constructing the text with their readers.. Stance is often associated with the concept of voice, which is however another area that involves a broad range of meanings, aspects, and dimensions. In contemporary academic writing, while stance refers to expressions of personal opinions, voice is regarded as the idea of authorial visibility, signature, and personal stamp (Hyland & Guinda, 2012). Therefore, it is important to take note of how stance and voice are considered as two different concepts, which cannot be referred interchangeably. Thorough review on the notion of stance-taking is discussed in the following chapter.. 1.5 Structure of the Thesis As for the structure of the thesis, it is as follows: Chapter 2 is devoted to present a general introduction to stance, referring to the works of other researchers and offering new ideas and applications. It focuses on how stance is categorized 9.

(21) differently by numerous scholars throughout the years, and discusses how stance is delivered in English through grammatical realizations by different writers, with specific focus on evaluative that-clause constructions. Chapter 3 provides the details of the corpora data to be analyzed with word size shown in tables, and the underlying rationales for choosing those materials are given. Analytical methodology and the problems encountered are also explained. Chapter 4 is aimed to discuss the data results and findings, and reach a general conclusion on how different groups of writers express viewpoint in academic written texts by sorting out the patterns of stance usage. To form a complete picture of analysis, both quantitative statistical frequency data and concordance example sentences shown qualitatively are provided. Finally, Chapter 5 summarizes the findings of Chapter 4, and offers pedagogical implications and suggestions for further study in the area. Relevant appendix of all the stance markers searched in the current research is included in the end.. 10.

(22) CHAPTER TWO Literature Review. This chapter is aimed to examine the concept of stance that thus far has been investigated and conceptualized by various researchers. It is divided into two main sections; section 2.1 not only explores how stance has been researched and categorized by different scholars over the years, but also discusses how stance is indicated in linguistic realizations through uses of modals, adverbials, and complement clauses. Section 2.2 extends the study focus to evaluative that-clause and reviews the discrepancy of such usage found between writers of different English proficiency levels and discipline fields. Finally, an analysis framework for current research is adopted and synthesized after the thorough review.. 2.1 Stance in English Academic Writing Language classes or academic writing classes often focus on illustrating effective word choices that affect textual meaning, rather than projecting writer image and recognizing the intricate dynamic between writer and reader. Other than brushing up on the abilities of summarizing, synthesizing, and handling citations in advanced academic writing courses, learning to justify claims and display personal standpoint are also essential for academic success (Hyland, 2002b). It has been widely believed that content clarity and cohesive development of the written work are the important criteria in delivering persuasive message between two parties (Xing et al., 2008). It has also been proposed that the persuasive strength of one’s argument may be weakened when the writer avoids to show authorial attitudes and sureness and is “lessen[ing] the responsibility of proposing that statement as his/her own idea” (Kuntjara, 2004, p.18). 11.

(23) According to Fløttum, [A]uthors of English articles are overtly present in their texts; they are reader-friendly (guiding the readers through the text by explicit indications about what is or what will be done) and relatively polemical…and manifest a more collective voice. (Fløttum, 2012, p.227).. Academic texts which are written by native English writers are concise, direct, linear, and straight to the point, with author’s own opinions and perception presented. It is possible to notice numerous linguistic markers that serve to display authorial stance and interpersonal value judgment in the writing of native English speakers (Hyland, 2002a). EFL writers on the other hand may find the act of stating tings directly too aggressive as it violates the politeness and face system in Oriental culture. In addition, they may find themselves not yet have reached the academia status to exhibit rigid authorship when making claims. As a result, they tend to be indirect and often keep a distance away from the truth value of propositions by being in a neutral and objective position (Kuntjara, 2004). This act of indirectness may be demonstrated by their frequent usage of stance markers that illustrate the condition of being likely and probable, in both lexico-grammatical and discoursal level. Chinese writers have also been observed to exhibit ambiguous instances on the practice of citing quotation. While English writers tend to display rigid attribution of authorship, it may not be “obvious which portions of the text are attributed to whom” in the work of Chinese writers (Connor, 2002). This ambiguous feature may owe to the low profile attitude immerse in the Chinese traditional cultural values by concealing oneself behind oblique or indirect choice of words. Different expectation and consideration therefore may attribute to the discrepancies on how native English speakers and non-native English speakers display viewpoints, judgment, and evaluation through linguistic markers to indicate stance in academic discourse. Thus, it is clear that the differing rhetorical conventions in cultures, and the way a writer refer to their own act of 12.

(24) thinking all have an impact on how people compose and comprehend a written product. Investigations of personal expression have been carried out under different rubrics, including evaluation, evidentiality, intensity, affect, hedging, and stance (Biber, 2004). Stance refers to linguistic and non-linguistic forms and strategies that show a writer's commitment to the status of the information that he or she is providing; it also acts as markers of their personal feelings, attitudes, value judgments, assessments, affect, and emotional involvement (Gray & Biber, 2012; Precht, 2008). There have been considerable interests on how people position themselves with respect to a particular issue through stance, and both detailed analyses and large-scale investigations of stance usage pattern have been conducted over the years. Researchers who are interested in stance have attempted to classify them in different taxonomies; the following section gives an overview of how stance is being conceptualized throughout the years.. 2.1.1 Categorizing Stance Since the 1980, researchers have approached stance marking under multiple terms, such as evidentiality (Biber & Finegan, 1988; Gray & Biber, 2012; McCready & Ogata, 2007), affect (Biber & Finegan, 1988; Ochs & Schieffelin, 1989), hedges and boosters (Hyland, 1998; 2005; Precht, 2003), modality (Biber, 2004), (inter)subjectivity (Hyland, 2005), and appraisal (Martin & White, 2005). Despite the fact that numerous terminologies are used to annotate the concept of stance, researchers over the years have been putting great efforts in probing the nature in which writers position themselves in respect to other readers and writers. Chafe in 1986 firstly discussed how the linguistic markings of evidentiality were. 13.

(25) for writers to realize the purpose of indicating attitudes towards knowledge in proposition (Gray & Biber, 2012). The grammatical elements that are used to indicate knowledge source, reliability, and evaluation are either direct or indirect evidentials. In most cases, a direct evidential marker is used when the language user is certain about the stated claim; on the other hand, an indirect evidential is served to denote the state of uncertainty or to report a hearsay information without having to take responsibility for its truth. Whether they serve the purpose of inferencing, reporting, marking knowledge source, or simply denoting sensation, these markers all show language user’s attitude to, or belief in, the statement. As McCready and Ogata put it, “evidentials need not involve only assertion, but can also include an element relating to the speaker’s judgement of probabilities and expectation—in other words, a modal element" (McCready & Ogata, 2007, p.151). That is to say, when people are reporting information with evidentials, the statement may also entail their own opinion about the propositions they write. Ochs and Schieffelin in 1989 further investigated the notion of mood and affect that are to display utterance intensity which entails personal emotions, attitude and feelings such as surprise, anger, sadness, or fear in their proposition (Ochs & Schieffelin, 1989). Around the same time, Biber and Finegan brought these two terms, affect and evidentiality, together into a broader term—stance (Biber & Finegan, 1988). By examining linguistic markers according to the affective and evidential distinctions in a corpus-based study, Biber and Finegan discovered that affective meanings could either be positive (e.g. happily, fortunately) or negative (e.g. embarrassing, sadly), whereas evidential indicators hold a level of certainty (e.g. impossible, without doubt), commitment (e.g. will, demonstrate) or doubt (e.g. uncertain, perhaps, maybe) toward the status of knowledge. Precht (2003) further expanded the categorization of stance with quantifiers as 14.

(26) its third element. Quantifiers, which serve to modify writers’ statement, in English include: all, a lack of, many, most, few, a lot, a little, several, no, and none of the. In mathematics, quantification is the act of measuring, whereas in logic and semantic sense, it refers to the language that we use to specify the extent of validity of our claims. In other words, applying quantifiers in sentences somehow involves the act of demonstrating the usage of emphatics and hedges. While emphatic markers, such as must, will, prove, obvious, certain, essentially, and actually, carry a strengthening tone to stress a point and support claims with solid evidence and sound argument, hedges on the other hand present the message more tentatively with the following markers: may, could, seem, possible, likely, perhaps, and apparently (Precht, 2003). Hyland has also concentrated his extensive research on the prominence of hedging and boosting in academic writing (Hyland, 1998; 2002a; 2004). Hedges and boosters, being similar to what Precht has discussed as hedges and emphatics, are used to either reduce or increase the statement force; they “not only carry the writer’s degree of confidence in the truth of a proposition, but also an attitude to the audience” (Hyland, 1998, p.350). In views of Hyland, while writers use hedges by some means to limit their commitment to a proposition, boosters are applied to display high degree of certainty and involvement with the topic (Hyland, 1998). In 2005, he extended his work to the writer-reader textual interaction feature that one may find in academic discourse. Particularly interested in how interpersonal metadiscourse helps to develop writer-reader relationship, a model called “Model of Interaction in Academic Discourse” is presented. According to Hyland (2005), the two major constituents framing this model are stance and engagement.. 15.

(27) Figure 2.1 Key Resources of Academic Interaction (from Hyland, 2005, p.177) As the above figure displays, writers project their viewpoint through hedges, boosters, attitude markers, and self-mention; they additionally involve and engage readers by applying reader pronouns, personal asides, appeals to shared knowledge, directives, and questions in the writings (Hyland, 2005). While hedges imply that a statement is with plausible reasoning rather than knowledge, boosters express writers’ certainty by presenting work with assurance. It is of great importance for one to balance both ends of the spectrum, as both hedges and boosters “balance objective information, subjective evaluation and interpersonal negotiation, and this can be a powerful factor in gaining acceptance for claims” (Hyland, 2005, p.180). Writers are not only delivering information, they are in fact communicating ideas with persuasive argument and effecting solidarity with readers. This act can also be seen by the usage of attitudinal markers and self-mentions in academic prose. As the name implies, attitudinal markers demonstrate writers’ affective state, value, and reactions to convey surprise, frustration, agreement, doubt, and so on. When writers signal propositions with attitudinal markers, they “both express a position and pull readers into a conspiracy of agreement so that it can often be difficult to dispute these judgements” (Hyland, 2005, p.180). Several linguistic indicators are served to express such function, such as attitude verbs (e.g. admit, agree, anticipate, concede, ensure, expect, foresee, prefer, reflect), adverbs (e.g. hopefully, fortunately, 16.

(28) importantly, ironically, surprisingly, unbelievably), and adjectives (e.g. acceptable, appropriate, concerned, critical, crucial, noticeable, unacceptable, vital). As for self-mention, it is usually done by using first person pronouns to present writers’ interpersonal information. Inclusive we often performs as reader pronouns that bind writer and reader together to signal mutual understanding and as participants with similar goals. In this sense, by engaging and leading their readers to a scientific discovery, writers are recognizing readers’ presence and involving them as discourse participants for interpretation guidance (Hyland, 2005). Hyland emphasized that the essence of academic writing is for writers to construct and negotiate social relations with the reader audience. In other words, academic writing, as opposed to the traditional impression that it is merely presenting facts mechanically, is in fact to be dialogical, as defined by Hyland: Stance: They express a textual ‘voice’ or community recognized personality which, following others, I shall call stance. This can be seen as an attitudinal dimension and includes features which refer to the ways writers present themselves and convey their judgements, opinions, and commitments. It is the ways that writers intrude to stamp their personal authority onto their arguments or step back and disguise their involvement. Engagement: Writers relate to their readers with respect to the positions advanced in the text, which I called engagement (Hyland, 2001). This is an alignment dimension where writers acknowledge and connect to others, recognizing the presence of their readers, pulling them along with their argument, focusing their attention, acknowledging their uncertainties, including them as discourse participants, and guiding them to interpretations. (Hyland, 2005, p.176). For Hyland, stance encompasses evidentiality (in terms of hedging, boosting, and commitment toward a proposition), affect/attitude markers, and presence (through the use of self-mention in the text).The act of applying these linguistic markers in writing signals an interaction between the reader and writer, by supplying readers with tentative analysis and allowing rooms for alternative interpretation. This shows how 17.

(29) bearing audience in mind is what it makes an effective academic writing. Another approach to stance is that of the appraisal theory proposed by Martin and White in 2005. It is a system in which defines emotions and sentiment arouse during the interactive evaluation process between one’s goals, beliefs, and the environment. Appraisal theory encompasses engagement, attitude, and graduation as its three sub-divisions. Different from Hyland’s engagement concept, Martin and White viewed it as “means available for bringing voices other than the writer’s into a text, either explicitly or implicitly” (Hood, 2012; Martin & White, 2005). In other words, rather than putting the interpersonal negotiation between the reader and writer as the focus, engagement, in this sense, is looking at how writer present themselves while they acknowledge and engage with prior scholars’ established belief or values. This system produces a dialogic network of voices, either single-voiced (monogloss) or multi-voiced (heterogloss), through the expressions of projection (e.g. X suggests that…), modality (e.g. it might be that…), negation (e.g. it is not that…), and counter-expectancy (e.g. However, …). This perspective helps readers apprehend if the writer is taking a stand with, a stand against, or neutral with respect to other researchers’ value position (Martin & White, 2005). As for the second dimension, which examines expressions associated with writers’ opinions on a positive (e.g. good) or negative (e.g. bad) gradable scalar value, attitude is further expanded into three sets, namely affect, appreciation, and judgement. While affect mainly deals with the expressions of emotion and feelings (e.g. concerned, disappointed, satisfied, unaware), appreciation denotes aesthetic values of entity (e.g. important, complex, critical), and judgement describes moral assessments of people and their corresponding behaviours (e.g. discreet, honest, truthful). The last realm in appraisal system is graduation, in which deals with gradability of evaluations through force and focus. Intensifying the quality or quantity of an entity is done through the 18.

(30) usage of scaling force, such as important, very important, and little importance. Martin and White have also reminded us that these scaled expressions often exist in the context of academic research studies as well. For instance, proximity indicators (e.g. in the present study, a more recent study), quantification on study size (e.g. large-scale study, a number of participants), scope of procedural usage (e.g. a wide range of tools, widely applied) and the number of supporting literatures (e.g. many researchers, most studies show that…) can all be often found in academic writings (Hood, 2012). Focus, on the other hand, holds the effect of modifying the strength of boundaries between categories, whether it is softening through vague language usage or sharpening as the intensifiers and amplifiers would function. That is to say, the focus concept is somewhat indeed similar to that of hedging and boosting mentioned earlier. Building upon the earlier work, Biber and his colleagues (1999) have also provided another classification of stance, namely epistemic, affect, and style. Epistemic stance, or as other scholars call it—evidentiality, mainly deals with remarks that pronounce certainty, doubt, actuality, source of knowledge, imprecision, and viewpoint. It also implies some level or degree of certainty or evidence toward what is being presented in a clause. As style refers to how writers comment on information (e.g. honestly, in fact, to be honest), affect concerns with states, evaluations, personal emotions, and attitudes. He further proposed a framework that sorts numerous stance markers into both lexico-grammatical structural and semantic functional categorization. The three main structural groupings include: 1) modal and semi-modal verbs, 2) stance adverb(ial)s, and 3) extraposed complement clauses (Biber, 2006). Under each main category, devices are then allocated into four main sub-classes based on their semantic functions: likelihood, factive, non-factive, and attitudinal. Likelihood stance is similar to hedges in the sense that they both state writers’ condition of being likely, probable, and less 19.

(31) certain about the source or extent of knowledge. Attitudinal stance, which is also interchangeable as affect, expresses writers’ emotion, judgment, and appreciation. Factive stance, which bears a resemblance to booster’s function, is to present fact and make. assertion,. while. stance,. non-factive. often. being. projected. with. communication-related lexical items, is to avoid responsibility for proposition’s truth-value. Due to the fact that the functional sorting under to-clause is not straight-cut, and by noticing the subtle semantic differences between each stance marking, Biber labeled the categorization with much precision. The following table displays Biber’s taxonomy of stance under grammatical and functional classifications. Table 2.1 Biber’s Taxonomy of Stance Markers Structural Categories. Semantic Functions. Examples. a) Modals & semi-modals possibility/permission/ability. b) Stance adverbs. can, could, may, might. necessity/obligation. must, should, have to, got to, ought to. prediction/volition. will, would, shall, be going to. likelihood. apparently, evidently, perhaps,. factive. actually, always, certainly, indeed. non-factive. accordingly, generally, mainly. attitudinal. hopefully, importantly surprisingly. likelihood. appear, consider, estimate, guess. factive. certify, conclude, confirm, discover. non-factive. advise, argue, claim, convey, imply. attitudinal. agree, complain, ensure, prefer. c) Extraposed clauses Verb + That-clause. Adjective +That-clause likelihood. possible, probable, certain, clear, true. attitudinal. amazed, careful, necessary, obligatory. likelihood. assumption, claim, expectation, notion. factive. assertion, conclusion, discover, fact. non-factive. comment, proposal, remark, report. attitudinal. fear, hope, reason, view, thought. likelihood(probability). appear, happen, seem, tend. Noun + That-clause. Verb + To-clause. likelihood(cognition/perception) assume, consider, estimate, learn, see desire/ intention/ decision. aim, agree, choose, decide, intend. effort/ facilitation. afford, arrange, drive, enable, manage 20.

(32) Adjective + To-clause. speech act/ communication. ask, advise, claim, convince, report. likelihood/ certainty. certain, likely, unlikely, sure. attitude/ emotion/affect. concerned, content, pleased, surprised. evaluation. appropriate, essential, important, vital. ability/ willingness. able, competent, eager, insufficient. ease/ difficulty. difficult, impossible, pleasant, tough. Noun + To-clause. agreement, confidence, intention, plan. Despite the fact that there are many different terminologies used for the concept of stance by numerous scholars and researchers, they in fact have the same objective, to investigate the ways in which writers project their assessment, judgement, and viewpoint in their written production. From the early work in 1980s to the literature found recently, the different approaches to stance all contribute in generating a thorough paradigm in explaining and exploring what stance entails. How these linguistic markers of stance are realized in English grammatical structures is thoroughly discussed in the following sub-section.. 2.1.2 Grammatical Realization of Stance Numerous scholars have long been interested in the means by which writers encode personal attitudes and the status of knowledge in written pieces. While they have proposed framework to explore how interaction is realized in academic argument, there are studies that have investigated different linguistic features of stance. Unlike some languages that mark information source or its degree of certainty with inflectional affixes, English has no obligatory linguistic element to convey stance (Gray & Biber, 2012). Instead, there are numerous grammatical features for users to fulfill such functional purpose. Stance can either be delivered through single word or clausal structures in English and these include: modal verbs, stance adverbials, and complement clause constructions (Biber et al., 1999: 966-86). 21.

(33) Use of Modals A modal auxiliary verb modifies the meaning of the main verb in a sentence. While modals generally express speakers’ attitudes, modals can also convey the intensity and strength of those attitudes by the distinction of present and past-tense forms (e.g. may vs. might). In Longman Grammar of Spoken and Written English (Biber et al., 1999, p. 485), modality is a term used in syntactic and semantic analysis to refer to meanings connected with scaling degrees of possibility/ permission/ability, necessity/obligation, and prediction/volition. Having a wide variety of communicative functions, modals serve to express a matter of belief by making inferences or deduction, and a matter of action by giving permission, command, and obligation. Biber and his colleagues further classified modal auxiliaries into two sorts, extrinsic epistemic and intrinsic deontic: Each modal can have two different types of meaning, which can be labeled intrinsic and extrinsic (also referred to as ‘deontic’ and ‘epistemic’ meanings). Intrinsic modailty refers to actions and events that humans (or other agents) directly control: meanings relating to permission, obligation, and volition (or intention). Extrinsic modailty refers to the logical status of events or states, usually relating to assessments of likelihood: possibility, necessity, or prediction. (Biber et al., 1999, p.485).. In 2006, Biber investigated a 2.7 million words corpus called TOEFL 2000 Spoken and Written Academic Language, to represent the range of both spoken and written registers in which university students in US might encounter. The four focused registers for in-depth analysis included classroom teaching, class talk, textbooks, and written syllibi. Not only did this study compared stance usage across speech and writing, it also distinugished two communicative purposes: classroom management and instructional situations. It was found that stance devices were generally used more in spoken situations than written registers; yet there was still apprearances of stance expressions in written texts. Most stance expressions indicated were related to factive. 22.

(34) representation which added certainty to the propositional information that readers needed to know. The most frequently used stance markers in university settings turned out to be modal verbs. It was found that volition modals (e.g. will, would) were the most common modal class, followed by possibility modals (e.g. can, could, may), and finally necessity modals (e.g. must, should) as the least used category. Hyland in 2002 also did a related study, but instead of probing modals’ epistemic function, he investigated modals as markers of directives in a total of 2.5 million word corpus of published research articles, textbooks, and L2 student essays (Hyland, 2002a). The included journal articles were randomly selected from ten leading journals in eight disciplines. Textbook chapter readings that were used in first year classes at an English medium University in Hong Kong were chosen as the second corpus dataset. The third corpus encompassed project reports written by final year Hong Kong undergraduates. To ensure variability, textbook readings and student reports were also randomly picked from the eight fields as of journal article selection. The results showed a major distinction between hard (natural) and soft (social) sciences, in the extent to which precision and conciseness were more valued in hard sciences, through the use of directive modals to fulfill the economy goal in article writings. The following examples are extracted from Hyland’s study: [1]…the above definitions for the B2 index should be multiplied by a factor of 0.83. (Mechanical Engineering- Research Article) [2]…attention must be paid to the standard membrane microstrip and for low free-space radiation loss.. (Electronic Engineering-Textbooks). [3] Nonetheless, a radioiodinated probe has to be utilized in the test at the same time. (Biology-Project Reports) [4] Firstly, some configuration work should be done. The PC-Lab card should be switch on SW1 for setting 10 MHz internal pulse. (Mechanical Engineering-Project Reports) (Hyland, 2002a, p.236).. It was noted by Hyland that the tradition of precision and the impersonal writing 23.

(35) convention established in hard sciences influenced how the writers presented themselves and interacted with their readers. In contrast, writings found in soft sciences such as humanities often needed more elaboration, and directives were not as necessary to keep the writing concise. As Hyland puts it, some fields permit greater authorial presence than others and the use of directives emphasize that social relationships with discourse communities exercise strong constraints on a writer’s representations of self and others. (Hyland, 2002a, p.236).. In the same study, Hyland also discovered that, when being compared with student writers, expert writers overall used more obligation modals (e.g. must, ought) as cognitive acts to “instruct the reader to perform an action or to [reason] things in a way determined by the writer” (Hyland, 2005, p.184). Since student writers were sensitive to the concept of power relation and were also reluctant to establish control over readers, they only used half of what the expert writers would express to serve the cognitive directive function in writings. Instead, students were more comfortable in using textual-acts modals to guide readers to another part of the text. As students’ number of use in textual directives outnumbered that of expert writers, it seemed to the students that guiding or instructing readers to another textual passage carried less authority over the readers than telling them to interpret certain matter in a certain angle. The results proved that though the social beliefs of disciplines and genres are deeply embedded in our writing practices, “individual factors such as experience, confidence, or professional rank affect the choices made by particular writers in their moment-by-moment composing” (Hyland, 2002a, p.236). Another interesting survey of modality is done by Vázquez Orta in 2010, comparing how native English writers and non-native Spanish writers differ in the usage pattern of modal verbs in academic research articles. Focusing specifically on introduction and discussion sections, Vázquez Orta (2010) found that Spanish writers 24.

(36) tended to overuse the modal can and will to denote certainty, while would, should, and may were used significantly less to express probability and possibility. Native English speakers, on the contrary, showed more uses of would, should, and might as mitigators to tone down abruptness, or to show modesty and cogency in their writing. As Vázquez Orta reasoned (2010), the discrepancies lied in the different writing conventions of national culture.. Use of Stance Adverbs As Biber and his colleagues in Longman Grammar of Spoken and Written English (1999) noted, adverbs are categorized into three sorts: circumstance adverbs (e.g. eagerly, slowly, weekly), stance adverbs (e.g. apparently, clearly, perhaps, technically), and linking adverbs (e.g. additionally, moreover, nevertheless). Circumstance adverbs form the core propositional content and thus cannot be omitted without changing the sentential meaning; in contrast, stance adverbs are detachable in the sense that they are used to qualify a standpoint and add interpersonal meaning in discourse. In general, stance adverbs have “a rather loose connection with the rest of the elements in a sentence”, since they may be detached syntactically and semantically from the utterance in which they appear (Tseronis, 2009, p.44). It should be noted that there exist a slight difference in stance adverbial and stance adverb; stance adverbial encompasses sentential adverbs in the form of finite clauses (e.g. as X notes, as might be expected) and prepositional phrases (e.g. from our perspective, in my opinion, in brief, without doubt) that function as adverb, whereas stance adverb refers to part of speech that modifies verb, adjective or another adverb. As Biber remarks, while single-word stance adverbs are relatively common in academic writings, they generally appear far more frequently in conversation: Speakers use stance adverbs to convey their judgements and attitudes, to claim the. 25.

(37) factural nature of what they are saying, and to mark exactly how they mean their utterances to be understood. (Biber et al., 1999, p.766-777).. Single-word adverb, rather than sentential adverbial, has aroused more research interests among scholars over time, since it has been noted by Biber and his colleagues (1999) that prepositional phrases and finite clauses only account for 15% of stance adverbials. Additionally, single adverbs (e.g. frankly, obviously) are by far the most common category of stance adverbial in both spoken and written registers (Biber et al., 1999). In the investigation on university textbooks and syllabi, Biber (2006) discovered that certainty adverbs (e.g. actually, certainly) were most common in acadeimc written register, followed by style adverbs (e.g. generally) which were used to indicate usual case rather than invariable fact. Likehihood adverbs (e.g. probably, maybe) held similar frequency of use as the style adverbs, whereas attitude adverbs (e.g. astonishingly, ironically, surprisingly) were the least common. Another closely-related examination was done by Adams and Quintana-Toledo (2013), in which they inspected frequency distribution of stance adverbs in the sections of introduction and conclusion of research papers. Following the functional classification on stance framed by Biber et al. (1999), the markers were manually tagged as epistemic, attitudinal, and stylistic in the corpus. A number of meanings under the category of epistemic adverbs were further broken down into six subclasses: doubt and certainty, actuality and reality, source of knowledge, limitation, viewpoint or perspective, and finally imprecision. The researchers also noticed that semantic sorting was not an easy task as functional dividing was rather fuzzy; thus attitudinal and stylistic adverbials were not classified into additional subclasses. The result showed that epistemic adverbials were significantly favored and predominated in both introduction and conclusion sections, followed by attitudinal 26.