連繫功能詞與主題推進教學對增進EFL低成就者的寫作連貫性之研究 - 政大學術集成

全文

(1)國立政治大學英國語文學系碩士在職專班碩士論文. 指導教授：余明忠博士 Advisor：Dr. Ming-chung Yu. 連繫功能詞與主題推進教學對增進EFL低成就者的寫作連. 立. 政治大貫性之研究. ‧ 國. 學. The Improvement of Coherence in EFL Low Achievers‘ Writing. ‧. through the Instruction of Cohesive Devices and Thematic. n. al. y er. io. sit. Nat. Progression. Ch. engchi. i n U. 研究生：林舒悠撰 Name：Su-yu Lin 中華民國一百年十二月 December, 2011. v.

(2) The Improvement of Coherence in EFL Low Achievers‘ Writing through the Instruction of Cohesive Devices and Thematic Progression. A Master Thesis Presented to. 立. 政治大 Department of English,. ‧ 國. 學. National Chengchi University. ‧. n. er. io. sit. y. Nat. al. In Partial Fulfillmenti v. n U engchi of the Requirements for the Degree of. Ch. Master of Arts. by Su-yu Lin December, 2011.

(3) Acknowledgements. I would like to show my appreciation toward those who inspired and encouraged me in the process of doing the study. First of all, lots of gratitude goes to my advisor, Dr. Ming-chung Yu, who gave me many inspiring and perceptive suggestions. Without his consideration and patience, it would have been impossible for me to finish my thesis before my second baby was born. I also want to thank Dr. Chi-yee Lin and Dr. Chia-yi Lee, for being. 政治大 extended to all the teachers in the ETMA program for their professional knowledge 立. my committee members and giving helpful advice on my thesis. My many thanks are. and useful instruction. Last but not least, I am extremely grateful to Dr. Chien-ching. ‧ 國. 學. Mo, whose enthusiasm about teaching and expertise in etymology and reading. ‧. inspired me to do the research and to teach my students.. sit. y. Nat. Next, I would like to thank my colleagues who shared my work during my. io. er. further study in National Chengchi University. Besides, I really appreciate the support from the other ETMA researchers, Chih-yuan Yang, Lisa Tsai, Jessie Chen, Wei-ting. al. n. v i n Cworries Lu, and so on, for sharing my encouraging me all the time. Most U h e nand i h gc. important of all, my gratitude is also extended to my family members who always stand by me and support me. My parents‘ tolerance and my husband‘s considerate cooperation made it possible for me to complete the task.. iii.

(4) Table of Contents. Acknowledgements ................................................................................................................... iii Chinese Abstract ...................................................................................................................... xii English Abstract ...................................................................................................................... xiv Chapter One: Introduction ..................................................................................................... 1 Background ............................................................................................................................ 1. 治政大 Purpose and Significance of the Study .................................................................................. 5 立 Chapter Two: Literature Review ........................................................................................... 6 Motivation .............................................................................................................................. 3. ‧ 國. 學. Writing Process ...................................................................................................................... 6. ‧. Revision in Writing ................................................................................................................ 6 The Characteristics of Revision ......................................................................................... 7. y. Nat. sit. What to be Revised ............................................................................................................ 8. n. al. er. io. Low Achievers‘ Revision ................................................................................................ 10. i n U. v. Coherence in Writing ........................................................................................................... 13. Ch. engchi. Coherence Types .............................................................................................................. 14 Cohesion ..................................................................................................................... 14 Propositional Coherence ............................................................................................. 18 Coherence Analysis: Thematic Progression (TP) ............................................................ 20 Simple Linear TP ........................................................................................................ 20 TP with a Continuous/Constant Theme ...................................................................... 21 TP with a Hypertheme ................................................................................................ 21 Exposition of Split Rheme .......................................................................................... 22 TP with a Thematic Jump ........................................................................................... 22. iv.

(5) Other Two Rhematic Progression Types ..................................................................... 22 Previous Studies on TP .................................................................................................... 23 The Integration of CD and TP.......................................................................................... 25 Research Questions .......................................................................................................... 28 Chapter Three: Method ........................................................................................................ 29 Participants........................................................................................................................... 29 Selection........................................................................................................................... 30 Raters ............................................................................................................................... 31 GEPT Raters ............................................................................................................... 31. 政治大. Coherence Researchers ............................................................................................... 32. 立. Data Collection and Procedure ............................................................................................ 33. ‧ 國. 學. Product Data..................................................................................................................... 33 The Participants‘ Written Drafts and the Raters‘ Comments ...................................... 33. ‧. The Procedure of Collecting the Product Data ........................................................... 34. sit. y. Nat. The Two Participants‘ Written Works ......................................................................... 37. n. al. er. io. Process Data ..................................................................................................................... 38. i n U. v. Why Writing Conferences? ......................................................................................... 38. Ch. engchi. The Procedure of the Writing Conferences ................................................................. 39 Perception Data ................................................................................................................ 40 Journals ....................................................................................................................... 40 Questionnaires ............................................................................................................ 41 Interviews.................................................................................................................... 41 Data Analysis ....................................................................................................................... 42 The Analysis of Product Data .......................................................................................... 42 The Procedure of Analyzing CD and TP Types in the Written Essays ....................... 42 The Scale to Score CD and TP Type ........................................................................... 44. v.

(6) The Analysis of Process Data .......................................................................................... 46 The Analysis of Perception Data ..................................................................................... 47 Journals ....................................................................................................................... 48 Questionnaires ............................................................................................................ 48 Interviews.................................................................................................................... 49 Chapter Four: Results ........................................................................................................... 50 The Effect of Learning CD and TP on Writing.................................................................... 51 Effect 1: Progress in Writing Performance ...................................................................... 51 The Awareness of the Improvement in Writing Performance ..................................... 51. 政治大. The Changes Leading to the Progress in Writing Performance .................................. 53. 立. The Progress Shown in the Product Data ................................................................... 57. ‧ 國. 學. Effect 2: Improvement in Writing Coherence.................................................................. 59 The Awareness of the Improvement in Writing Performance ..................................... 59. ‧. Different Aspects of Perceiving the Progress in Writing Coherence .......................... 60. y. Nat. sit. The Progress in Writing Coherence Shown in the Product Data ................................ 63. n. al. er. io. Effect 3: the Frequency of Certain Categories in CD and TP .......................................... 65. i n U. v. The Analysis of CD..................................................................................................... 66. Ch. engchi. The Analysis of TP...................................................................................................... 69 The Raters‘ Comments on the Writing Coherence...................................................... 72 The Development of the Participants‘ Writing .................................................................... 73 The Development of Writing Coherence ......................................................................... 74 The Participants‘ Familiarity with CD and TP ............................................................ 74 The Development of Writing Skills ................................................................................. 77 The Preference for CD ................................................................................................ 77 More Focus on Content Level..................................................................................... 82 The Enhancement of Revision Stage .......................................................................... 84. vi.

(7) The Participants‘ Attitudes toward Writing Coherence ....................................................... 88 The Cultivation of Confidence in Writing Coherence ..................................................... 88 The Attitudes toward the Learning of CD and TP ........................................................... 90 Chapter Five: Discussion....................................................................................................... 97 The Effect of Instructing CD and TP on the Students‘ Writing ........................................... 97 Low Achievers‘ Progress in Writing and Writing Coherence ......................................... 97 The CD Categories and TP Types Adopted by the Underachievers ................................ 99 The Application of Certain CD Categories ................................................................. 99 The Application of Certain TP Types ........................................................................ 101. 政治大. The Development of Writing Coherence in Low Achievers‘ Writing ............................... 103. 立. The Possibility to Make Low Achievers Familiar with CD and TP .............................. 104. ‧ 國. 學. The Preference for CD ................................................................................................... 105 More Focus on Content Level ........................................................................................ 106. ‧. The Enhancement of Making Revision .......................................................................... 107. sit. y. Nat. The Participants‘ Attitudes toward Writing Coherence ..................................................... 108. io. er. Low Achievers‘ Confidence in Writing ......................................................................... 109. al. The Participants‘ Attitudes toward CD and TP ............................................................. 110. n. v i n Ch Chapter Six: Conclusion...................................................................................................... 112 engchi U Summary of the Major Findings ........................................................................................ 112 Implications........................................................................................................................ 114 Limitations ......................................................................................................................... 115 Suggestions for Future Studies .......................................................................................... 116 References .............................................................................................................................. 118 Appendices............................................................................................................................. 134 Appendix A—Classification of Halliday and Hasan‘s Cohesive Devices ......................... 135 Appendix B—The Thematic System ................................................................................. 136 Appendix C—Daneš‘s TP Types ....................................................................................... 137 vii.

(8) Appendix D—Consent Form ............................................................................................. 138 Appendix E—The Participants‘ Typed Works .................................................................. 139 Appendix F—CEEC Rating Scale for Compositions ........................................................ 140 Appendix G—Process of the Coherence Learning ............................................................ 141 Appendix H—The Chart of CD and TP............................................................................. 142 Appendix I—Reading Handout with CD and TP .............................................................. 145 Appendix J—The Test of CD and TP ................................................................................ 146 Appendix K—Peter‘s Writing Task ................................................................................... 147 Appendix L—Bob‘s Journal of Draft 6 ............................................................................. 148. 政治大. Appendix M—Questionnaire ............................................................................................. 149. 立. Appendix N—Interview Questions .................................................................................... 152. ‧ 國. 學. Appendix O—The Coding Scheme of Conference Talks .................................................. 153 Appendix P—The Coding Scheme of Journals ................................................................. 154. ‧. Appendix Q—The Interview Log ...................................................................................... 155. sit. y. Nat. Appendix R—The Coding Scheme of Interviews ............................................................. 156. io. er. Appendix S—GEPT Raters‘ Scores of the Participants‘ Essays and Coherence............... 157 Appendix T—Bob‘s and the Researchers‘ Analyses of CD and TP in Fight .................... 158. al. n. v i n C hScores of the Participants‘ Appendix U—The Researchers‘ e n g c h i U Writing Coherence .............. 159. Appendix V—The Excerpt of Attitudes toward CD and TP in Bob‘s Final Interview ..... 160 Appendix W—The Excerpt of Attitudes toward CD and TP in Peter‘s Final Interview ... 161. viii.

(9) List of Tables. Table 2.1 The Distinct Revising Preferences of Skilled and Unskilled Writers ...................................... 11 Table 3.1 The Reliability of the Raters .................................................................................................... 32 Table 3.2 The Reliability of the Researchers ........................................................................................... 33 Table 3.3 Scaling Chart of CD and TP .................................................................................................... 44 Table 4.1. 政治大. The Data Examined to Generate the Results ........................................................................... 50. 立. Table 4.2. The Questionnaire Items Related to Writing Performance ...................................................... 52. ‧ 國. 學. Table 4.3. The Questionnaire Items Related to Writing Coherence ......................................................... 59. ‧. Table 4.4. The Frequency of Cohesive Words in the Participants‘ Drafts................................................ 66. y. Nat. Table 4.5. al. er. io. Table 4.6. sit. The Thematic Patterning in the Participants‘ Writing ............................................................. 69. n. The Raters‘ Comments on Coherence ..................................................................................... 72 Table 4.7. Ch. engchi. i n U. v. The Confidence in Writing ...................................................................................................... 88 Table 5.1 The Most-frequently Used Cohesive Devices by Skilled Writers ......................................... 100 Table 5.2 The More Frequently-used Coherence Progression Types Found in Other Studies .............. 101. ix.

(10) List of Figures. Figure 2.1 Carter-Thomas‘s Three Types of Coherence ........................................................................... 14 Figure 3.1 The Selection of Participants ................................................................................................... 30 Figure 3.2 The Production of the Final Drafts .......................................................................................... 37 Figure 3.3 Daneš ‘s Simple Linear TP ...................................................................................................... 44 Figure 4.1. 政治大. The Distribution of Bobo‘s Writing ........................................................................................ 58. 立. Figure 4.2. The Distribution of Peter‘s Writing ......................................................................................... 58. ‧ 國. 學. Figure 4.3. Bob‘s Distribution of the Raters‘ Rating of Coherence ........................................................... 63. ‧. Figure 4.4. Peter‘s Distribution of the Raters‘ Rating of Coherence ......................................................... 63. y. Nat. Figure 4.5. al. er. io. Figure 4.6. sit. Bob‘s Distribution of the Researchers‘ Evaluation of Writing Coherence.............................. 65. v. n. Peter‘s Distribution of the Researchers‘ Evaluation of Writing Coherence ............................ 65 Figure 4.7. Ch. engchi. i n U. Bob‘s Distribution of Lexical Reiteration ............................................................................... 67 Figure 4.8 Peter‘s Distribution of Lexical Reiteration .............................................................................. 67 Figure 4.9 Bob‘s Distribution of Reference .............................................................................................. 68 Figure 4.10 Peter‘s Distribution of Reference ............................................................................................ 68 Figure 4.11 Bob‘s Distribution of Conjunction .......................................................................................... 68 Figure 4.12 Peter‘s Distribution of Conjunction ......................................................................................... 68 Figure 4.13 The Distribution of Bob‘s Use of Simple Linear TP and TP with a Constant Theme............. 70. x.

(11) Figure 4.14 The Distribution of Peter‘s Use of Simple Linear TP and TP with a Constant Theme ........... 70 Figure 4.15 The Distribution of Bob‘s Use of the Other TP Types ............................................................ 71 Figure 4.16 The Distribution of Peter‘s Use of the Other TP Types........................................................... 71. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xi. i n U. v.

(12) 國立政治大學英國語文學系碩士在職專班碩士論文提要. 論文名稱：主題推進與連繫功能詞教學對增進 EFL 低成就者的寫作連貫性之研究. 指導教授：余明忠博士. 研究生：林舒悠. 立. 論文提要內容：. 政治大. ‧ 國. 學. 英文作文教學在英語學習上是不可或缺的一環，而實際上在台灣，英文作文教學仍是透過分析文法和句型來進行的，但這樣的方式忽略了英文寫作能夠成功. ‧. 的主要關鍵——連貫性。由於連貫性的晦澀難解，教師與學生皆會認為在教授和. y. Nat. sit. 學習寫出有連貫性的文章是很複雜的。本文藉由教導學生如何於寫作時運用連繫. n. al. er. io. 功能詞(cohesive devices)和主題推進(thematic progression)兩種策略來探究作文中. i n U. v. 的連貫性，以研究其增進高中生英文寫作表現的可能性。本研究在不影響正常教. Ch. engchi. 學進度的前提下進行，研究者的 39 位臺北市高二學生首先接受如何分析課本文章連貫性的指導，並接著應用和檢視連繫功能詞以及主題推進於他們的寫作中。其中有兩位低成就者進一步地被挑選出來，藉由寫作會談(writing conference)以瞭解受試者在寫作時，如何利用以上兩種連貫性的策略於其英文寫作中；同時也透過訪談和日記撰寫的方式，來探討學生的學習連貫性策略的歷程。由相關的量化和質性資料可看出，本研究的結果顯示受試者在作文整體表現與作文連貫性上的分數明顯偏高；他們也被證明能夠和其他研究中的高成就者一般，運用相同的連繫功能詞(指稱、連接與重述)和主題推進(線性推進與主題連續推進)種類。在這歷時五個月的研究中，受試者也在連繫功能詞與主題推進的協 xii.

(13) 助下，培養出更注意作文內容以及進行適當修改的能力。此外，也由於受試者對於學習連繫功能詞與主題推進抱持著正面的態度，本研究建議應結合連繫功能詞和主題推進，並將其融入台灣現存的正規作文教學之中，藉著分析作文中的連繫功能詞與主題推進，協助學生理解抽象的連貫性進而使作文表現更進步。. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xiii. i n U. v.

(14) Abstract. As an indispensable element in English learning, writing instruction in Taiwan actually has been given through the analysis of grammar and sentence patterns. This however ignores another primary key to successful writing—coherence. Due to its obscurity, teachers and students both find it complicated to teach and acquire the ability to organize a written text logically and coherently. The current study. 政治大 them how to apply the coherence strategies to their writing—cohesive devices (CD) 立. investigated the possibility to improve the high school students‘ writing by instructing. and thematic progression (TP). In this study, with the teaching schedule unaffected,. ‧ 國. 學. the researcher‘s 39 second-graders in one Taipei senior high school were first taught. ‧. how to analyze the coherence of the reading passages in the textbooks. Then they. sit. y. Nat. were required to apply and examine CD and TP when writing. Two low achievers of. io. er. these students were further selected to investigate their writing development by examining their written texts with CD and TP applied in writing conferences, having. al. n. v i n C and interviews with the researcher journals between classes and writing h ekeeping ngchi U conferences.. Concluded from the quantitative and qualitative data collected, the results revealed that the two low-proficiency students were able to get high grades in the holistic writing performance and coherence of writing. They were meanwhile found to be able to apply the same categories of CD (reference, conjunction and reiteration) and types of TP (simple linear TP type and TP with a constant theme) as the ones used by the high-proficiency learners in previous studies. During the five-month study, the participants also cultivated the ability to focus more on content level and to revise properly in their writing with the help of CD and TP. Besides, since the participants xiv.

(15) held positive attitudes toward the learning of CD and TP, the researcher recommends that the combination of CD and TP should be integrated into the writing instruction in current normal English writing class in Taiwan to help students comprehend the abstract coherence and to improve their writing by analyzing CD and TP in the written texts.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xv. i n U. v.

(16) CHAPTER ONE INTRODUCTION. Background The focus of interest in writing instruction has been shifted from the product-oriented paradigm to the process-oriented paradigm since the past decades (Cumming, 1998; Emig, 1971; Flower, 1989; Hillocks, 1986; Krapels, 1990; Silva, 1993; Zamel, 1987). Adopted to investigate the writing process, Flower and Hayes‘s. 政治大 accepted and applied in the relevant research into the field of teaching writing 立. (1981a) three stages of writing—planning, writing, revising—has been widely. (Cumming, 1989; Fathman & Whalley, 1990; Ferris, 2003; Goldstein & Conrad, 1999;. ‧ 國. 學. Zamel, 1985). These three steps may take place recursively and simultaneously in the. ‧. writing process, so the ideas which are previously generated can be organized or. sit. y. Nat. corrected again after revising, and vice versa (Flower & Hayes, 1981; Murray; 1980;. io. er. Smith, 1982; Zamel, 1983a).. Among these three steps, revision is generally considered the most important. al. n. v i n C writing because this step promises the through the emphasis on changes in U h e nquality i h gc surface and meaning levels of the written texts (Bereiter & Scardamalia, 1987;. Fitzgerald, 1987; Villamil & de Guerrero, 1996). The surface-level changes, also called form-level or mechanical/grammatical revisions, are particularly preferred by both teachers (Ashwell, 2000; Cohen, 1990; Cumming, 1985; Fathman & Whalley, 1990; Ferris, 1995a, 1999, 2003, 2004; Zamel, 1985) and students with different ages and proficiency levels (Crowley, 1977; Faigley & Witte, 1981a; Fitzgerald & Markham, 1987; Sommers, 1980). The benefit of surface correction in writing instruction is therefore confirmed (Ellis, 1998; Ferris, 1995b, 1999, 2004, 2006; Reid, 1997). 1.

(17) However, the application of too many grammatical corrections may make writers ignore how to organize logical and coherent texts (Mohan & Lo, 1985; Perl, 1980). Truscott (1996, 1999, 2007) even proposed that grammatical corrections in teaching writing should not be emphasized too much so that students would not lose interest in writing. Therefore, besides grammatical revision, content-level/ discourse-level revision is indispensable in the writing process. With the content revised appropriately, the discourse becomes comprehensible, which is the key to good writing quality (Cumming, 1989; Goldstein & Conrad, 1999; Saito, 1994; Zamel,. 政治大 the contrary, the less successful writing may be described as ―fragmentary‖ or 立. 1985). Good-quality writing is commonly rated as ―well organized‖ and coherent; on. disconnected because its communicative intention is unlikely to be comprehended. ‧ 國. 學. (TEEP Attribute Writing Scales, cited in Watson Todd, Thienpermpool &. ‧. Keyuravong, 2004). These comments are all related to coherence, which is usually. sit. y. Nat. considered implicit due to readers‘ different interpretation (Hoey, 1991).. io. er. To analyze abstract coherence, two primary ways are developed: Theme-Rheme/Thematic progression types (hereafter TP) (Daneš, 1974) and Topical. al. n. v i n Structure Analysis (hereafterCTSA), which is adapted h e n g c h i U from the former (Lautamatti, 1987; Simpson, 2000). The application of TP or TSA is to find out the themes or topics between sentences and further to help identify propositional coherence,. whether it is explicitly displayed in a text or not. In fact, a theme or topic includes one or more units, so what actually connects the next theme or topic sometimes is hard to determine (Halliday & Hasan, 1976). Therefore, some specific linguistic connectors are needed. ―Manifested at the surface level of a text‖ (Watson Todd, Khongput & Darasawang, 2007, p. 12), these text-forming connectors are mainly based on Halliday and Hasan‘s (1976) cohesive devices (hereafter CD), which in some ways are similar to Faigley and Witte‘s (1981a) classification of surface-level revision. Due 2.

(18) to its easily-identified quality, CD is highly recommended to help recognize cohesion in a text (Mackay, 1987; Wang, 1998; Weissberg, 1984). Despite the advantage of CD, a paragraph, with the overuse or misuse of CD, may be seemingly full of cohesion but may be made up of independent and irrelevant sentences—no common proposition. This means there is no coherence in the text at all (Carell, 1982; Connor, 1984). Without propositional coherence, the text is doubtlessly beyond comprehension. Therefore, TP or TSA, combined with CD, is applied in some research to examine the coherence and cohesion in experts‘ and. 政治大 However, the integration of TP and CD seems to be more feasible when they are 立. students‘ writing (Lan, 2008; Lee, 2002; Liu, 2009).. taught as revision strategies to analyze coherence due to some problems of TSA.. ‧ 國. 學. Though TSA has also been applied as a revision strategy to examine student writers‘. ‧. writing coherence (Chiu, 2004; Connor & Farmer, 1990; Fan, 2008; Lee, 202; Liu,. sit. y. Nat. 2009; Shen, 2004), the process of searching for topics and the identification of topical. io. er. development is vague (Knoch, 2007; Lautamatti, 1987). Unlike TSA, TP, with more complete topical progressions, has been conducted to help improve students‘. al. n. v i n C h(Alonso & McCabe,U1998b, 2003). In the meantime, if coherence in previous research engchi students have the chance to learn how to analyze writing coherence on their own, content-level revision might be possibly achieved. Motivation In fact, research on revision and coherence analysis primarily lays lots of emphasis on the comparison of the writing performance between high- and low-proficiency students. On one hand, the former are found to revise more in the content level or more frequently when writing than the latter whose focus is always on the surface level (Fitzgerald, 1987; Matsumura, Patthey-Chavez, Valdés & Garnier, 2002; Perez, 2000). Besides, these high achievers are also found to revise both the 3.

(19) surface and content levels at any time in the writing process; therefore, the high-proficiency learners‘ writing quality is quite remarkable (Beach, 1976; Bridwill, 1980; Faigley, Cherry, Jolliffe & Skinner, 1985; Faigley & Witte, 1981a; Hall, 1990; Heuring, 1984; Monahan, 1984; Sommers, 1980; Stallard, 1974; Zamel, 1983a). On the other hand, as far as the content level is concerned, certain CD categories and TP types are also found to happen more frequently in skilled writers‘ writing than the unskilled ones‘ even though there was no instruction of CD and TP in the research on CD and on TP (to name but a few: Alonso & McCable, 1998a, 1998b, 2003; Chao,. 政治大 2007). In other words, the above-mentioned studies keep reinforcing the self-evident 立. 2002; Faigley & Witte, 1981b; Jalilifar, 2010; Lee, 2002; Li, 2009; Mustaffa & Aman,. fact—the high achievers inherently possess the ability to investigate more beyond the. ‧ 國. 學. form level in their writing process. Compared with the writing quality of their. ‧. counterparts, the low achievers‘ writing performance is always found to be inferior. io. er. unsolved (Porte, 1996, 1997; Vann & Abraham, 1990).. sit. y. Nat. since they tend to ignore the content level and their form-level problems remain. With the high achievers‘ writing skills identified in the research into revision. al. n. v i n C h(1990), Rubin (1975) and coherence analysis, Oxford and Wenden (1991) suggested engchi U that this finding should be feasible to improve the writing ability of the poor-skilled. writers. That is, these poor-skilled writers could enhance their writing by replicating the entire writing process that their counterparts go through. Nonetheless, little research follows the relevant suggestion to further examine the specific ways that teachers could use to help these poor writers. What is worse, it is also implied that these low-proficiency writers could not further apply certain CD categories and TP types as well as the high-proficiency ones (Chen, 2002; Faigley & Witte, 1981b; Ferris, 1994; Tseng, 2008). Based on the above-mentioned, more studies need to be conducted to see if the instruction of coherence analysis could also save 4.

(20) underachievers from being trapped in surface-level problems and help them to examine cohesion and coherence in the discourse level of their own writing as high achievers (Chao, 2002; Crompton, 2004; Faigley & Witte, 1981b; McCarthy, 1991; Reid, 1993; Scott, 1996). As a result, it is necessary for the present study to explore whether low achievers can make their writing better by learning to apply CD and TP to make more content-level revision than surface-level one. Purpose and Significance of the Study The purpose of this study is to guide low-proficiency students to focus more on. 政治大 revision strategies to help them examine their writing coherence. Furthermore, 立. the discourse level in the writing process by instructing these students CD and TP as. whether it is possible for the instruction of CD and TP to facilitate these low. ‧ 國. 學. achievers‘ writing performance is also another concern. Therefore, this current study. ‧. picked the senior high underachievers as its participants to examine: 1. the effect of. sit. y. Nat. learning CD and TP on writing; 2. the students‘ development of writing coherence by. io. er. examining CD and TP in their own writing; 3. their views of learning CD and TP. What this study investigated is of great importance because teachers often focus. al. n. v i n on such form-level issues asC word substitution, sentence h e n g c h i U structure drills, grammatical. mechanics, and so forth, instead of focusing on content-level problems (Silva, 1990). Besides, according to the College Entrance Examination Center (CEEC) in Taiwan, over half of all the examinees scored less than ten points and over ten percent of them got zero in writing in recent years‘ exams (Lin, 2009). Some Taiwanese senior high students apparently have difficulty composing coherent passages. The findings of the study can not only help teachers develop different ways to teach writing beyond surface-level focus but offer a promising solution to students‘ problems of writing coherence as well.. 5.

(21) CHAPTER TWO LITERATURE REVIEW. This chapter begins with the importance of revision in the writing process. Then the literature reviewed presents two main concerns of revision—what should be revised and what difference is in the revising patterns between skilled and unskilled writers. It is found why the skilled writers‘ writing quality is better lies in their focus on coherence. Then the subsequent sections center on the coherence instruction:. 政治大 research is presented with the research questions raised to verify the possibility of 立 cohesive devices and thematic progression. In the end, the requisite for doing the. Writing Process. 學. ‧ 國. enhancing low achievers‘ writing coherence.. ‧. To understand the complete process of how students produce texts, the attention. sit. y. Nat. has shifted from the product-oriented paradigm to the process-oriented one. Focusing. io. er. on the development of how a written text is formulated rather than on the analysis of the one-draft written product, the ―processing‖ (Myers, 1983, p. 19) is comprised of. al. n. v i n two primary features. First, C the writing process thatUa writer goes through is definitely hengchi. beyond the practice of imitating grammar and sentence patterns due to the complexity of the composing process (Faigley & Witte, 1981b). The next concern is the recursive progression among these stages—prewriting, writing and revision (Connor, 1987; Flower & Hayes, 1986; Myers, 1983). Considered linear in the product approach, the three stages have been verified to be mutually and recursively intertwined between and within paragraphs at the time of writing. In other words, writers move back and forth in the writing process (Flower & Hayes, 1986; Hyland, 2003; Zamel, 1982). Revision in Writing Among the writing stages, revision has long been viewed as the most crucial 6.

(22) part (Bereiter & Scardamalia, 1987; Elbow, 1981; Lowenthal, 1980; Sommers, 1980; Zamel, 1983a) because of its potential for improving the writing quality (Ash, 1983; Bamberg, 1978; Bracewell, Scardamalia & Bereiter, 1978; Bridwell, 1980). It is thus a must to first grasp what revision is and how it happens in writing. Then what to be revised is displayed and the previous research on revision is also discussed. The Characteristics of Revision The prospect of improving writing quality through revision lies in the following features. First, conceived as both goal- and problem-oriented processes, revision. 政治大 approach the meaning as much as possible (Bereiter & Scardamalia, 1986; Bridwell, 立 enables writers to rethink the written texts, to alter words or phrases and then to. 1980; Nold, 1981; Sommers, 1980, 1992). Second, revising may be initiated by any. ‧ 國. 學. rough ideas at both the end and start of sentences and the whole draft; therefore,. ‧. revisions made are likely to be trivial or crucial (Bereiter & Scardamalia, 1986;. sit. y. Nat. Bridwell, 1980; Cohen, 1990; Faigley & Witte, 1981a; Flower & Hayes, 1981; Nold,. io. er. 1981; Williams, 2004).. In addition, the way revision is conducted goes in a certain sequence. To begin. al. n. v i n with, writers, after finishingC the first draft, read through h e n g c h i U the whole text to identify the. inconsistency between the assigned topic and their own intentions to write down what they want to express. It is then essential for writers to choose the words, phrases or sentences for a change to make up for the discrepancy. Lastly, to improve their writing, writers further need to consider replacing the inappropriate words or phrases with preferred and feasible ones. The first two steps may simultaneously occur once in a while. In the last stage, writers manage to revise by substituting the alternatives for the problematic units in the text (Fitzgerald, 1987; Williams, 2004). In general, revising enables writers to move backwards and meanwhile forwards to check what may be missed so that the writing quality can be reassured. 7.

(23) What to be Revised Revision occurs repetitively at the time of composing texts while writers are unlikely to alter anything arbitrarily. Otherwise, the corrections in different stages may cause the disconnection throughout the writing process. A series of systematic entries used to modify anything inappropriate in the writing is thus required. One set of the revision entries formulated by Faigley and Witte (1981a, 1984) has been implemented to assess writing in plenty of research. In the revision measurement are two principal categories: changes in surface/form and content/meaning. The. 政治大 information and new one. In other words, even if some surface changes are made, 立. difference between the two changes is the existence of adjustments made for the given. such as spelling, tense, punctuation, and so on, any information related to the. ‧ 國. 學. conveyance of meaning remains still. On the other hand, the modification of any idea. ‧. and opinion is called content change. The alternation consists of two more. sit. y. Nat. subcategories: micro- and macro- changes. The former includes any discrete. io. er. perception, and the latter symbolizes the main information of the whole text. In fact, plenty of studies corroborate that surface revisions predominate the. al. n. v i n content ones when writers atC different ages and with h e n g c h i Udifferent language proficiency go through the writing process (Bridwell, 1980; Crowley, 1977; Dennett, 1990; Faigley & Witte 1981a; Fitzgerald & Markham, 1987; Hall, 1990; Stallard, 1974). Besides, most teachers, considering themselves as ―language teachers not writing teachers‖ (Zamel, 1985, p. 86), are still inclined to stress surface-level accuracy (Ashwell, 2000; Cohen, 1990; Cumming, 1985; Fathman & Whalley, 1990; Ferris, 1999, 2003, 2004; Ferris & Roberts, 2001; Zamel, 1985). Ferris (1997) found that large amounts of students‘ revision can be elicited and thus their errors will be decreased through teachers‘ particular feedback on grammatical issues (1999, 2003, 2004, 2006). The weight of surface correction may result from the long-predominant notion that writing 8.

(24) is rated as a product regardless of its implicit developing process and the need to train students to identify errors (Ferris, 2003; Lee, 2008b). Besides, most students have also been found to favor grammatical comments from their teachers (Leki, 1991; Oladejo, 1993; Saito, 1994). This may stem from their previous experiences of learning second language and the need to be assessed as academic performance and to pass the entrance exams (Hyland, 2003). However, with the process approach to teaching compositions proposed, some researchers (Cumming, 1989; Goldstein & Conrad, 1999; Saito, 1994; Zamel, 1985). 政治大 as coherence or organization instead of surface-level problems. Moreover, Truscott 立. have suggested the teacher feedback should be given on such content-level concerns. (1996, 1999, 2007) even contended that grammatical changes might cause the. ‧ 國. 學. hindrance to the writing improvement by presenting some convincing evidence, so he. ‧. proposed the form concern should not exist in writing instruction. In addition, it is. sit. y. Nat. suggested that teachers provide their students with the feedback on the surface and. io. er. content problems at the same time instead of paying separate attention to either of the two feedback types (Fathman & Whalley, 1990; Ferris, 2003; Krashen, 1984; Taylor,. al. n. v i n C hcontent commentsUare under operation has been also 1981). Then, how the form and engchi under discussion. Fathman and Whalley (1990) deduced that there is no difference between responding to form and content concurrently and responding to form or content respectively. Then Ashwell (2000) investigated which order is effective—form then content or the reverse—and he found the simultaneous application of form and content is the most appropriate way. To sum up, having been controversial all the time, the surface-level issue is undoubtedly essential in writing as long as it does not predominate over the content. issue in the whole revision process. For one reason, the approaches learners adopt to get hold of linguistic knowledge and that they use to comprehend diverse language 9.

(25) correction (Schwartz, 1993; Truscott, 1999) are different. For the other, the unskilled writers tend to make grammatical mistakes continuously (Raimes, 1985) since their language proficiency is not as good as the skilled writers. They are unlikely to solve the form-level problems in a short time. Therefore, it might be plausible for teachers to teach low-proficiency learners to learn how to solve their content-level problems. Low Achievers‟ Revision The low-proficiency writers, especially EFL learners, are found to correct grammatical problems in their early written texts (Heuring, 1984; Nold ,1981;. 政治大 process further than mechanical alternatives, so they are unlikely to make 立. Perl ,1980; Sommers ,1980). They are considered unable to stretch the revising. considerable progress in their writing quality (Fitzgerald, 1987; Matsumura, et al.,. ‧ 國. 學. 2002; Perez, 2000). In contrast, skilled writers have been proved to revise more in. ‧. discourse and surface changes than the less skilled ones (Ash, 1983; Faigley & Witte. sit. y. Nat. 1981a; Monahan, 1984; Nold ,1981; Sommers, 1980; Stallard, 1974; Zamel, 1983a).. io. er. Besides, these skilled writers never start to correct linguistic mechanics until they finish investigating and confirming the previously generated ideas (Victori,1999;. n. al. Zamel, 1983a).. Ch. engchi. i n U. v. The recognition of high achievers‘ correction mainly implies the possibility of helping low achievers improve their writing by comparing their writing performance with the former‘s (Oxford, 1990; Rubin, 1975; Wenden, 1991). In fact, this series of research just strengthens the self-evident actuality—successful writers‘ strategies and inclinations to compose good quality writing. This keeps being stressed by the comparison with their mediocre partners. These unsuccessful writers are thus recommended to imitate how and what their counterparts do when composing (Perl, 1979; Zamel, 1983a). The above-mentioned research on the revision preferences is illustrated in Table 2.1. 10.

(26) Table 2.1 The Distinct Revising Preferences of Skilled and Unskilled Writers Revision. Types. Skilled/Experienced writers (High. Unskilled/Inexperienced writers (Low. achievers). achievers). Discourse and sentence level changes. Sentence level changes (Fitzgerald,. (Ash, 1983; Faigley & Witte 1981a;. 1987; Heuring, 1984; Matsumura, et al.,. Monahan, 1984; Sommers, 1980;. 2002; Nold, 1981; Perez, 2000; Perl,. Stallard, 1974; Victori; 1999; Zamel,. 1980; Sommers, 1980). 1983a) Timing. Surface changes are mostly initiated at. Surface changes keep interrupted from. the later stage of writing (Zamel,. the start of writing (Zamel, 1983a). 1983a). 政治大. However, the attempt in the relevant research to deal with low achievers‘. 立. writing by just comparing their poor writing quality with their counterparts‘ is. ‧ 國. 學. doubted (Horowitz, 1986) for some reasons. First, little connection between language. ‧. proficiency and writing ability has been found to demonstrate that low achievers perform more poorly than high achievers because of their less competence in. y. Nat. io. sit. linguistic units (Jones, 1985; Raimes, 1987). In fact, it is students‘ writing strategies. n. al. er. instead of language proficiency that are closely related to the quality of their written. Ch. i n U. v. texts (Cumming, Rebuffot & Ledwell, 1989; Jones & Tetroe, 1987). Second, low. engchi. achievers are unlikely to reach the high achievers‘ good writing quality because they take the only correction of linguistic mechanics for granted instead of the discourse level in texts (Fitzgerald, 1987; Heuring, 1984; Matsumura, et al., 2002; Nold, 1981; Perez, 2000; Perl ,1980; Sommers ,1980). Besides, it is found that even most young adult writers with low English proficiency, whether they are natives or second language (SL) learners, generally have difficulty dealing with the organization of discourse (Brown, 1981; Nold, 1981; Zamel, 1983a). Third, the comparison with the advanced writers‘ writing cannot actually help underachievers to make any essential progress in writing; nevertheless, this could have a negative effect on the latter‘s 11.

(27) writing performance since their writing problems remain unresolved (Porte, 1996, 1997; Vann & Abraham, 1990). Most important of all, less-competent language learners have been proved to be capable of making some contributions to their own writing (Porte, 1995; Vann & Abraham, 1990). As a result, what these unskilled writers need is the specific way—in addition to grammatical correction—to improve their writing quality, which is ignored in most research on this filed. Therefore, instead of the constant emphasis on the differences of writing performance between less and more competent writers, more research is needed to. 政治大 especially the EFL learners, who spend more time and efforts dealing with 立. investigate the concrete and feasible ways to improve low achievers‘ writing,. grammatical structure (Porte, 1997). Though it is generally believed that teachers. ‧ 國. 學. have difficulty teaching the low-proficiency EFL learners to write the complete. ‧. passages (Cumming, 1989), research into improving these low achievers‘ writing by. sit. y. Nat. the instruction of both surface level and content level in writing is truly required.. io. er. Since these unskilled writers may leave the linguistic mistakes repetitively occurring (Raimes, 1985), one of the aims of the current study is to prevent the low achievers. al. n. v i n C h of the high achievers from being the negative counterparts again by enabling the engchi U former to examine the content level and to further improve their writing quality.. In general, content-level revision has been found to take place more frequently in high-proficiency writers‘ written texts, and these skilled writers tend to make content-level revisions prior to the surface-level ones (Ash, 1983; Faigley & Witte 1981a; Monahan, 1984; Nold ,1981; Sommers, 1980; Stallard, 1974; Victori,1999; Zamel, 1983a). Therefore, it can be concluded that high-quality writing is supposed to consist of good organization, reasonable sequence, and well-linked development of ideas to construct coherent discourse according to the analytical writing assessments (Hughes, 2003; Knoch, 2007; Watson Todd, et al., 2007; Weigle, 2002; Weir, 1990). 12.

(28) On the contrary, the poor-quality writing is disconnected due to the lack of appropriate connectives, which makes the writing proceed disparately and makes readers unable to understand the writers‘ intentions (Knoch, 2007; Watson Todd, et al., 2007). The above description of writing quality is mainly based on Faigley and Witte‘s revision taxonomy (1981a), and its indicators related to content are much more concrete and detailed; besides, though the specific indicators in the revision types and writing assessments are different, their focus primarily concentrates on whether the coherence exists in the written texts. Coherence in Writing 政治大 Coherence has been defined in a variety of ways. The most renowned is 立. Halliday and Hasan‘s (1976) definition that coherence is the connection of the surface. ‧ 國. 學. text testified by cohesive devices (CD). Their notion of coherence is composed of two. ‧. features, cohesion and register, which utterly concentrate on CD within and between. sit. y. Nat. sentences, though. Owing to their failure to describe coherence thoroughly, many. io. er. studies have been conducted to tackle with other possibilities of explaining the essence of coherence. Beyond the superficial explanation, coherence is actually found. al. n. v i n Cthat to be an informative structure readers to grasp the texts by the h eallows ngchi U. development of the theme-rheme construct and topics (Connor & Farmer, 1990; Daneš, 1974; Firbas, 1986; Lautamatti, 1987; Schneider & Connor, 1990). Concerning the content level of discourse, the second definition of coherence is that it can help writers/readers identify the connection in the content, and the connection is clarified by the relations between propositions which promote organization of the whole discourse (de Beaugrand & Dresseler, 1981; van Dijk & Kintsh, 1978; Meyer, 1975; Yule, 1996). Based on the above definition, different types of coherence have been developed to analyze the consistency and unity in writing.. 13.

(29) Coherence Types The coherence types are displayed systematically by Carter-Thomas (2008), whose diagram, as shown in Figure 2.1, clearly illustrates the main domains of the textual coherence. The coherence structure starts with the cohesion that primarily focuses on the combination of sentences to form paragraphs (Enkvist, 1978). Then the propositional coherence suggests how writers express their ideas to make readers understand the texts (Yule, 1996). As to the highest-level coherence, it deals with the more implicit communication acts within the discourse (Watson Todd, et al., 2007).. 政治大 Cooper, 1988; Watson Todd, et al., 2007), the last coherence type is ruled out in this 立. However, due to its complexity and uncertainty (Carter-Thomas, 2008; Connor, 1987;. study.. ‧ 國. 學. Figure 2.1 Carter-Thomas‘s Three Types of Coherence (2008, p. 2). ‧. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. Cohesion In this section, the definition of CD is presented first. The classification of CD categories is subsequently described in detail. Then the previous research into CD is also discussed to pinpoint the need of teaching CD to student writers and of combining the other dimension of discourse coherence. Cohesion, classified as a sentence-based approach in linguistic text analysis (Connor, 1987), is a rather specific way to detect the textual connection. The surface-level cohesion links the proposition in or between sentences with a series of 14.

(30) linguistic signals which are straightforwardly identified at the surface of a text (Bell, 1991; de Beaugrand & Dresseler, 1981; van Dijk & Kintsh, 1983; Hoey, 1983; Watson Todd, et al., 2007; Widdowson, 1978). With these signals, the sentences in a text are dependent on each other to keep the textual cohesion. Contrarily, without the cohesion linking the propositions in a text, the text will fail to be comprehensible in that the sequences within sentences disappear and the text splits into discrete pieces (Biesenbach-Lucas, Meloni & Weasenforth, 2000; Carreon, 2006; Hinkel, 2001; Zhang, 2000). Besides, since cohesion can be found within a sentence or sentences in. 政治大 a text and the implication of the text meaning. These connections are called cohesive 立. a text, the semantic connections expressed through it can be viewed as the meaning of. devices or text-forming ties included in a coherent text (Halliday & Hasan, 1976).. ‧ 國. 學. These cohesive devices, or cohesive ties (Hoey, 1991; Kieras, 1981; Weissberg,. ‧. 1984), are words or phrases that make it accessible for readers to decode the texts by. sit. y. Nat. identifying the relation across sentences and to connect new and old information.. io. er. These devices are: 1. grammatical cohesion, including reference, substitution, ellipsis, conjunction; 2. lexical cohesion, such as reiteration and collocation (see Appendix A).. al. n. v i n C hand ellipsis mostlyUappear in conversation, which does However, because substitution engchi. not correspond with the aim of the study, they are left out.. Reference. Reference, a semantic relation, can immediately pinpoint what is being mentioned in the previous context. The reference items can function within a context anaphorically and cataphorically—the same reference as a word or words previously and subsequently appearing in one discourse. Halliday and Hasan (1976) divided the referential cohesion into three sub-types: personal references (e. g., he and they), demonstratives (e. g., this) and determiners (e. g., the). Conjunction. Conjunction indicates how the clauses or sentences should be connected to the previous or subsequent ones. According to Halliday and Hasan 15.

(31) (1976), a five-category scheme connecting sentences is based on the specific meaning: additive (e. g., besides), adversative (e. g., however), causal (e. g., because), temporal (e. g., when) and continuative conjunction (e. g., after all). Reiteration and collocation. The two categories of lexical cohesion deal with the connection based on the words used. Reiteration is relatively complex because its five types required a teacher to teach step by step and to help learners to recognize them carefully. They are: 1. repetition, often including reference; 2. synonym, or near-synonym, also including reference; 3. antonym (e. g., ―white‖ vs. ―black‖); 4.. 政治大 whole), and 6. general words. Then collocation refers to the regular co-occurrence of 立 hyponym (e. g., the relation of ―flower‖ and ―tulip‖); 5. metonym (e. g., part vs.. certain words. For example, the following pairings, evening/dark, doctor/hospital,. ‧ 國. 學. garden/dig and boat/row, tend ―to appear in similar context when they occur in. ‧. adjacent sentences‖ (Halliday & Hasan, 1976, p. 286).. sit. y. Nat. Previous studies on CD. With these cohesive ties binding dispersed sentences. io. er. and paragraphs together to make the whole text a coherent unity (Halliday & Hasan, 1976; Zamel, 1983b), the relationship between cohesive devices and writing quality. al. n. v i n has been confirmed (FaigleyC & Witte, 1981; Hasan, h e n g c h i U1984; Pappas, 1985). Research on the application of CD in writing has been conducted predominantly to examine the. amount and frequency of different cohesive ties used in students‘ writing (Tierney & Mosenthal, 1983; Watson Todd, et al., 2007; Weissberg, 1984; Zhou, 2007) and to compare different devices applied to the written products between skilled and unskilled writers (Chen, 2002; Chiang, 1999; Faigley & Witte, 1981, 1984; Ferris, 1994; Tseng, 2008). Therefore, Reid (1993), McCarthy (1991) and Scott (1996) proposed it should be of importance for teachers to instruct CD explicitly since SL or EFL learners have more trouble dealing with the textual cohesion appropriately than native ones. However, few researchers, except for Tsao (2004) and Zhou (2007), 16.

(32) explicate the specific procedures for teaching cohesive devices. In view of this, the present study adapts the ways provided by Lee (1998, 2002), Liu (2000) and Zamel (1983b) to teach cohesion through some explicit exercises to be done in and between sentences or paragraphs written by students themselves. Despite the contributions of CD, some researchers have questioned the definite connection between the cohesion markers and coherence. For instance, Faigley and Witte (1981b) asserted that there are other elements leading to the successful writing, including ―a pragmatic unity, a unity of a text and the world of the reader (p. 201).‖. 政治大 namely, cohesion is not identical to coherence. From the theoretical and empirical 立. Connor (1984) concluded that a cohesively bound text is not necessarily coherent;. aspects, Carrell (1982) demonstrated that the cohesion does not always bring about. ‧ 國. 學. coherence as the former happens to be included in the latter. Tierney and Mosenthal. ‧. (1983) found the commonness of CD in text ―severely diminishes the usefulness of. sit. y. Nat. the cohesion concept as an index of coherence at a global or local level‖ (p. 228).. io. er. Therefore, the identification of CD in a text cannot be viewed as coherence. Take the following short passage as an example:. n. al. Ch. i n U. v. (2.1) The quarterback threw the ball toward the tight end. (2.2) Balls are used in many sports. (2.3) Most balls are spheres, but a football is an ellipsoid. (2.4) The tight end leaped to catch the ball. (Faigley & Witte, 1981b, p.201). engchi. The ball in (2.1) is the metonym of balls in (2.2), which is the hyponym of most balls in the next sentence. The ball in (2.4) can be referred to a football in (2.3) as the referential definite articles. These sentences are cohesively connected at the surface level because of the above-mentioned CD categories, while they are actually independent to each other because of lacking a general proposition. Moreover, Zamel (1983b) attributed the misuse of CD to the only focus on each respective cohesive device in teaching writing. Therefore, Kroll (1990) reported there is no definite connection between the content fluency and surface correctness. Among his 17.

(33) participants with different language backgrounds, some students could write fluently regardless of their poor English proficiency; on the other hand, others with good language proficiency might write incoherently. Due to these problems, the only identification of CD is not complete enough to comprehend the text meaning. To avoid failing to build textual coherence, propositional coherence also need to be taken into consideration. Propositional Coherence To solve some problems caused by CD, the main focus on coherence analysis in. 政治大 propositional coherence in writing is derived from the Functional Sentence 立. this section is propositional coherence. According to Connor (1987), the study of. Perspective (FSP). In view of this, the fundamental concepts of FSP are described. ‧ 國. 學. first. Then the method of teaching coherence adopted in the present study is presented. ‧. to explicate its combination with CD. Furthermore, the previous studies on teaching. y. sit. io. er. study.. Nat. the coherence analysis are conferred to initiate the research questions of the current. Functional sentence perspective (FSP). Developed by Czechoslovak linguists,. al. n. v i n Csentence FSP is defined as ―contextual (utterance)U organization, theme-rheme or heng i h c. topic-comment structure‖ (Firbas, 1974, p. 11). It has been perceived to be able to make discourse connection easily comprehended. Besides, the theory has been demonstrated to be closely related to the coherence examination (Alonso & McCabe, 2003; Connor & Farmer, 1990; Goodin & Perkins, 1982; Kopple, 1991; Lautamatti, 1987) because it consists of communicative dynamism, topic and comment, and theme and rheme (Kopple, 1982). Communicative dynamism is carried by each word with different degrees. In Firbas‘s (1974) view, every word in a sentence is ―… carrying some degree of [communicative dynamisms]‖ (p. 19), and each sentence can be divided into different 18.

(34) parts—topic and comment. The former is frequently anaphoric in discourse as it results from the introduced entity in the prior sentence (Bardovi-Harlig, 1990). The latter, on the contrary, expresses new, irrecoverable, or relatively less accessible information of the preceding sentence (Kopple, 1982). The topic-comment notion carries the same connotation as ―theme‖ and ―rheme‖ (Daneš, 1984; Fibras , 1974; Hajicova, 1994; Kopple, 1982; Schneider & Connor, 1990), the most common terms used to describe the sentence structures. Generally speaking, the first element of a sentence always holds the initial. 政治大 (Halliday, 1985, p. 39) and regarded as ―the point of departure of the message‖ 立. position and the rest are categorized as rheme. Theme is ―what the clause is about‖. (Halliday & Matthiessen, 2004, p. 64). To clarify Halliday‘s definition of theme,. ‧ 國. 學. Thompson (2004) separated themes into many subcategories to clarify different. ‧. themes (see Appendix B). As the given information, a theme should be placed in the. sit. y. Nat. beginning of a sentence or clause to be related to the preceding sentence, and a rheme,. io. er. as the new information concerning the theme, will then move the reader forward to the subsequent context. Therefore, through the balance and movement between theme. al. n. v i n C h role in the unityUand coherence of a text. and rheme, they play an important engchi. Theme-Rheme Patterning. Derived from the theme-rheme theory, some. approaches to examining coherence have been exploited. The first one is Theme-Rheme progression types (Daneš, 1974): the progression of themes and rhemes that are also named given and new information in texts. The second is Topical Structure Analysis (Lautamatti, 1987; Simpson, 2000): the exploration of a whole text by linking the discourse topic and sub-topics. The third one is Wastson Todd‘s (1998) Topic-based Analysis: the presentation of logical relations in texts by identifying the key concepts. Although the Topic-based Analysis is adapted and developed from TSA, its 19.

(35) complicated diagrams may not only confuse the assessing raters but hinder students‘ identification of abstract key concepts as well (Kroll, 2007). Besides, TSA originates from Theme-Rheme progression types, so their frameworks of topical development actually resemble each other (Albertini, 1990; Lan, 2008; Shen, 2004; Todd, 2007; Weisberg, 1974). Most of the studies on TSA focus on defining the three progression types as soon as the topics are identified, which is not definite enough to understand the topical development (Knoch, 2007). That is to say, how the sub-topics in a text are recognized remains obscure. Therefore, with the insufficient semantic elements. 政治大 implementation of TSA perplexing and challenging (Lautamatti, 1987). Therefore, the 立 detected, the process of searching for the topics and their progression would make the. present study adopts TP as its way to teach low achievers coherence analysis.. ‧ 國. 學. Coherence Analysis: Thematic Progression (TP). ‧. TP indicates the patterns how the flow of information exchanges between the. sit. y. Nat. consecutive occurrence of theme and rheme pairings in a text (Eggins, 1994). Besides,. io. er. it is convinced that the theme selection for all clauses and sentences is not random but patterning (Daneš, 1974). TP helps the reader/writer to build up the coherence in a. al. n. v i n text. There are five TP typesC developed by Daneš (1974): h e n g c h i U 1. Simple linear TP or TP. with linear thematization of rhemes; 2. TP with a continuous (constant) theme; 3. TP with a hypertheme; 4. Exposition of split rheme; 5. TP with a thematic jump. To describe a T-R nexus clearly, Daneš uses a horizontal arrow () and also a vertical arrow () to express the contextual link in and between sentences (see Appendix C). The five types of TP are illustrated below. Simple Linear TP In this type, the new information about the theme in one sentence becomes the old of the next sentence. Themes and rhemes in this type are connected continuously by repeating identical or relevant words. The following is a passage with the simple 20.

(36) linear TP type. (2.5) The stomach (T1) produces gastric juice, which contains dilute hydrochloric acid (R1). The acid (T2) kills most of the bacteria in the food (R2). The partly digested food (T3) passes next into the duodenum (R3). (Li & Liu, 2005, p. 56). In the example 2.5, the stomach is Theme 1 and the rest is Rheme 1. Then the acid in the second sentence summarizing Rheme 1 becomes Theme 2 in the next sentence. In the third sentence, kills most of the bacteria in the food is narrowed down to the focus only on the partly digested food as Theme 3. The given information keeps advancing to bring out more new information, which formulates the linear progression.. 政治大 In the second type, 立more than one rheme deal with the same theme. With the. TP with a Continuous/Constant Theme. ‧ 國. 學. same theme repeated, readers‘ attention is surely drawn to what the rhemes mention. The following passage is the example of this type.. ‧. (2.6) The brain (T1) contains 10 billion nerve cells, making thousands of millions of connections with each other (R1). It (T1) is the most powerful data processor we know (R2). And it (T1) is under intensive investigations by scientists nowadays (R3). (Li & Liu, 2005, p. 56). y. Nat. io. sit. In Example 2.6, Theme 1 the brain is substituted as it in the following two themes.. n. al. er. Therefore, on seeing the subjects of each sentence, readers can easily know the. i n U. v. passage is about human brain. The same theme is presented by three different and. Ch. engchi. new rhemes to make the passage understood easily. TP with a Hypertheme. In this type, each theme is derived from a hypertheme so that every sentence can be linked semantically or logically. This type is given in Example 2.7. (2.7) New Jersy (hypertheme) is flat along the coast and southern portion: the north-western region is mountainous. The coastal climate (T1) is mild, but there is considerable cold in the mountain areas during winter months. Summers are fairly hot. The leading industrial production (T2) includes chemicals, processed food, coal, petroleum, metals and electrical equipment. The most important cities (T3) are Newark, Jersy City, Paterson, Trenton, Camden. Vacation districts include Asbury Park, Lake Wood, Cape May and others. (Daneš, 1974, p. 120). In (2.7), Theme 1, 2 and 3 are all inferred from ―New Jersy‖, the hyperhteme. They can be viewed as the sub-categories of the hypertheme.. 21.

(37) Exposition of Split Rheme In this pattern, each sentence will share a common rheme which is then split into two or more partitions. Each of the divided rhemes will become the theme of the following sentence. See Example 2.8: (2.8) All substances (T1) can be divided into two classes (R1): elementary substances and compounds. An elementary substance (T2) is a substance which consists of atoms of only one kind (R2). A compound (T2) is a substance which consists of atoms of two or more different kinds (R2). (Daneš, 1974, p. 121). An elementary substance and a compound are both parts of two classes, Rheme 1 in the first sentence. And they have their own new information to describe them respectively.. 政治大 Though Daneš viewed 立 the last type as the modification of the simple linear TP. TP with a Thematic Jump. ‧ 國. 學. type, it is apparent to find one of the themes was skipped as a result of its redundancy or explicitness. The progression can be found in Dr. Martin Luther King‘s speech as. ‧. follows:. er. io. sit. y. Nat. (2.9) It (T1) is obvious today that America has defaulted on the promissory note in so far as her citizens of color are concerned (R1). Instead of honoring this sacred obligation (T2), American has given the Negro people a bad check, a check which has come back marked ―insufficient funds.‖ (R2) But we (T3) refuse to believe that the bank of justice is bankrupt (R3).. al. This passage can be illustrated by the simple linear TP type, with its Rheme 1, the. n. v i n CTheme promissory note, turning into obligation. However, what can be h e n2,gthisc sacred hi U substituted by we in the third sentence is the Negro people in the second sentence. Therefore, there must be a sentence, whose theme or rheme refers to Theme 2 and then moves to Theme 3, and it is omitted between these two sentences. In fact, because the theme carried in the deleted sentence appears both in the previous and later sentences, readers are still capable of indentifying we as Negro people. Other Two Rhematic Progression Types Despite Daneš‘s (1974) concern about the thematic content, other possible progression types, the connection between a rheme and the preceding theme or rheme,. 22.

(38) are not taken into consideration. To make the above possible links enhance the coherence progression, some researchers (Cloran, 1995; Crompton, 2004; Hawes & Thomas, 1996; Mauranen, 1996) proposed two more types whose focus is shifted to the rheme parts—rhematic progression (hereafter RP). The following dialogue is the example Cloran offers: (2.10) Mom: It (T1) is too cold for passion fruit now (R1). (2.11) Mom: They (T2) don‘t like the cold weather (R2). (2.12) Mom: Do you (T3) think we should plant a passion fruit vine at our new house (R3)? (2.13) Stephen: Yes…yes. (2.14) Mom: It (T4) usually takes a couple of years before you get many passion fruit on your vine (R4).. (Cloran, 1995, p. 390) The TP between Example 2.10 and 2.11 is Daneš‘s (1974) simple linear TP type as. 治政大what is happening between Rheme 1, passion fruit, becomes Theme 2, they. Then 立 Example 2.11 and 2.12 is the first type proposed by Cloran (1995), that is, Theme— ‧ 國. 學. Rheme progression (see Appendix C). The previous theme directly refers to the next. ‧. rheme. This can be defined as TP with a subsequent rheme. Then the progression in Examples 2.12, 2.13 and 2.14 put all their emphasis on rhemes. Therefore, Rheme 3. y. Nat. er. io. sit. and Rheme 4 are exactly the same thing. This type can be viewed as TP with interrelated rhemes (see Appendix C).. n. al. Ch. i n U. v. To sum up, five TP and two RP types are defined to make the coherence. engchi. progression more complete. Therefore, the present study will include these types to examine the students‘ written works. Previous Studies on TP Then this section presents the studies previously conducted on TP analysis. First of all, TP has been used to analyze the professional texts and to specify the progression types in them. The first kind of texts is the editorials in the newspapers written by native writers, which were compared with the ones written by nonnative writers (Hawes & Thomas, 1996). No significant difference has been found in these writers‘ essays. Sade (2007) then applied TP to explore the religious tracts, and 23.