美籍學習者作文偏誤的篇章分析與教學應用

全文

(1)國立台灣師範大學華語文教學研究所碩士論文 . 美籍學習者作文偏誤的篇章分析與教學應用 A Discourse Analysis in American Learner's Writing Errors with Pedagogical Implications . 指導教授：陳俊光博士研究生：范雅婷撰中華民國一零一年六月.

(2) ACKNOWLEDGEMENTS This thesis would not have been possible without the kind support and help of many individuals who contributed their valuable assistance to the preparation and completion of this research. First and foremost, I would like to express the deepest appreciation to my advisor, Professor Fred Jyun-gwang Chen, for his excellent guidance, great patience, and vital encouragement. Dr. Fred Jyun-gwang Chen has been my inspiration as I hurdle all the obstacles in the completion of this study. In addition, my sincere thanks go to Professor Feng-fu Tsao and Professor Chia-ling Hsieh for taking the time to serve on my thesis committee and providing me with invaluable suggestions. Moreover, I am heartily thankful to all the professors and staff at the Graduate Institute of Teaching Chinese as a Second Language, National Taiwan Normal University. Their efforts have enabled me to develop a better understanding of teaching Chinese as a second language. My thanks and appreciations also go to my classmates in developing the topic and people who have willingly helped me out with their zeal and abilities. Furthermore, I would like to thank Professor Cecilia Chang and Professor Li Yu, who provided me with informative resources and valuable insights in Chinese language pedagogy, which bear great relevance to this study, during my one-year’s overseas teaching practicum at William College. Last but not least, I would like to express my love and gratitude to my beloved families; for their understanding and endless love, through the duration of my studies.. i . .

(3) ABSTRACT Writing, as one of the four skills for acquiring a language, is an important indicator of successful acquisition of that language. It is also the most difficult skill for learners to acquire, especially at the discourse level. Many discourse errors are found in the production of L2 Chinese learners at a high proficiency level. However, previous research on Chinese learners’ writing has focused on the categorizations of the errors. Sources of errors and pedagogical applications are relatively less discussed. Moreover, the comparisons of errors at different L2 Chinese proficiency levels are rarely examined in literature. Halliday (1961) identifies three discourse functions: i.e. ideational, interpersonal, and textual. According to him, the ideational function refers to the function for composing human. experience.. The. interpersonal. function is concerned. with the discourse aspects of tenor or interactivity. The textual function is related to the internal structures and communicative functions. He further indicates that the textual function is to combine sentences into a discourse because effective interaction depends on extended discourse rather than words or sentences alone. Moreover, three types of textual function are identified: cohesion, information structure, and thematic structure. Based on Halliday’s discourse framework on cohesion, information structure and, thematic structure, this study seeks to address the following two research questions: 1) In English learners’ L2 composition, what types of errors do they commit and what are the causes of these errors? 2) Do these errors vary according to the learners’ L2 Chinese proficiency? A total of 89 L2 Chinese compositions from Williams College were collected for this study. The compositions were further categorized into three proficiency levels. In conjunction with the results of interviews conducted by the researcher, this study ii . .

(4) investigated the learners’ errors from the perspectives of both forms and functions. There are three major findings: first, linguistic typology has a great influence on the errors. This paper examined not only interlingual errors resulting from different typologies (i.e. isolating and inflecting languages, branching directions, pro-drop parameter, and subject- and topic-prominent languages), but also intralingual errors (i.e. overgeneralization and simplification). Second, in terms of cohesion errors, learners made the most errors in the areas of reference and conjunction. Within the reference errors, the error percentage was mitigated by the English learners’ L2 Chinese proficiency. As for conjunction errors, the errors made by the learners at the intermediate level were more than those at the elementary level, thus following a U-shaped learning curve. Finally, with respect to the discourse structure, the learners’ lack of recognition of the old-new information order and the concept of discourse resulted in their non-target-like ordering in sequencing information structure and thematic structure. Based on the results of the present study, pedagogical implications and suggestions for the further study are provided. Key words: error analysis, cohesion, information structure, thematic structure, linguistic typology. iii . .

(5) 中文摘要寫作為語言能力中的聽說讀寫四大項能力之一，也為語言課堂中的重要訓練項目。然而，寫作往往為四大能力中最難以掌握的一項，其中篇章層面的偏誤比例尤其甚高，高水平能力的學生作文仍有許多偏誤出現．然而，以往研究學生的篇章偏誤主要關注於各項偏誤的表面現象分類，偏誤來源以及教學應用較少深入探討之。本文以 Halliday（1961）的三種篇章功能為基礎，從銜接、信息結構、及主位結構來探討學習者的表現，研究目的為二：1）學習者進行漢語寫作時的各方面偏誤及原因為何？2) 這樣的偏誤現象，是否會隨著學習者漢語水平的提高而有所改變？本研究以英語母語者為研究對象，蒐集美國威廉大學（Williams College）二、三、四年級學生的作文共 89 篇（二年級 33 篇，三年級 29 篇，四年級 27 篇），配合面談，分別從形式和功能的角度來分析偏誤現象。研究結果主要有三：第一，偏誤與語言類型有很大的關係，本文除了從孤立語、屈折語、主要分枝方向、代詞脫落參數、主語顯著、主題顯著等面向探討語際偏誤來源之外，也經由面談討論其他偏誤來源，諸如過度類化與簡化等等．第二，銜接方面，學生在連接詞及指稱方面的偏誤率是最高的，指稱方面下的各類小項偏誤，隨著學習者漢語水平的增加而有明顯的改善。然而，在連接詞的部分，卻是中級組偏誤率高於初級組，產生了 U-型學習曲線的現象。第三，信息結構與主位結構方面，主要與語言類型中的新舊訊息順序有關，且與學生的篇章意識不足所致．本文也依據偏誤的類型與比例，提出關於漢語篇章銜接方面的教學應用。關鍵字：偏誤分析、銜接、信息結構、主位結構、語言類型. iv . .

(6) TABLE OF CONTENTS ACKNOWLEDGEMENTS .............................................................................................. i ABSTRACT ..................................................................................................................... ii 中文摘要 ......................................................................................................................... iv TABLE OF CONTENTS.................................................................................................. v LIST OF TABLES .......................................................................................................... vii LIST OF FIGURES ....................................................................................................... viii CHAPTER 1 ..................................................................................................................... 1 INTRODUCTION ............................................................................................................ 1 1.1 Significance of This Study ................................................................................. 1 1.2 Statement of the Problem ................................................................................... 1 1.3 Objectives of the Study ...................................................................................... 2 1.4 Description of This Study ................................................................................... 4 1.5 Organization of the Thesis .................................................................................. 5 CHAPTER 2 ..................................................................................................................... 7 LITERATURE REVIEW ................................................................................................. 7 2.1 Error Analysis ..................................................................................................... 7 2.1.1 Definition of Errors ................................................................................. 8 2.1.2 Error and Mistake .................................................................................... 9 2.1.3 Sources of Errors ................................................................................... 10 2.1.4 The Surface Structure Taxonomy .......................................................... 19 2.1.5 Five Steps of Error Analysis .................................................................. 26 2.2 Discourse Analysis ........................................................................................... 29 2.2.1 Definitions of Text and Discorse ........................................................... 29 2.2.2 Introduction of Discourse Analysis ....................................................... 30 2.2.3 Systematic Functional Garmmar ........................................................... 31 2.3 Lingustic Typology ........................................................................................... 44 2.3.1 Isolating Language and Inflecting Language ........................................ 45 2.3.2 Pro-drop Parameter ................................................................................ 46 2.3.3 Topic-prominent and Subject-prominent Language .............................. 50 2.3.4 Principal Branching Direction ............................................................... 51 2.4 Research of English Learners’ Chinese Discourse Errors ................................ 52 CHAPTER 3 ................................................................................................................... 55 METHOD ....................................................................................................................... 55 3.1 Purpose of the Study ......................................................................................... 55 v . .

(7) 3.2 Research Design ............................................................................................... 55 3.3 Participants ....................................................................................................... 56 3.4 Instruments ....................................................................................................... 57 3.5 Theoretical Foundations of Error Analysis ....................................................... 58 CHAPTER 4 ................................................................................................................... 59 RESULTS AND DISCUSSION ..................................................................................... 59 4.1 The Overall Performance.................................................................................. 59 4.2 Research Question One .................................................................................... 61 4.2.1 Cohesion ................................................................................................ 61 4.2.2 Information Structure and Thematic Structure ...................................... 71 4.3 Research Question Two .................................................................................... 77 4.3.1 Reference ............................................................................................... 77 4.3.2 Ellipsis ................................................................................................... 80 4.3.3 Conjunction ........................................................................................... 81 4.3.4 Lexical Cohesion ................................................................................... 82 CHAPTER 5 ................................................................................................................... 85 PEDAGOGICAL IMPLICATION ................................................................................. 85 5.1 Examination of Chinese Textbooks .................................................................. 85 5.1.1 Cohesion ................................................................................................ 87 5.1.2 Information Structure and Thematic Structure ...................................... 96 5.2 Pedagogical Application ................................................................................. 102 5.2.1 Cohension ............................................................................................ 102 5.2.2 Information and Thematic Structures .................................................. 125 CHAPTER 6 ................................................................................................................. 131 SUGGESTIONS FOR FUTURE RESEARCH............................................................ 131 REFERENCES ............................................................................................................. 133 APPENDIX A ............................................................................................................... 141 APPENDIX B ............................................................................................................... 145 . vi . .

(8) LIST OF TABLES Table 1 Detailed Information of Participants................................................................. 58 Table 2 The Overall Performance of Errors in Cohesive Device ................................... 60 Table 3 The Integrated Chart of Types of Errors and Their Sources ............................. 76 Table 4 Errors of Reference at Different Chinese Proficiency Levels ............................ 78 Table 5 Errors of Ellipsis at Different Chinese Proficiency Levels ................................ 80 Table 6 Errors of Conjunction at Different Chinese Proficiency Levels ........................ 81 Table 7 Errors of Lexical Cohesion at Different Chinese Proficiency Levels ................ 83 Table 8 Textbooks Used in America’s Universities (Overseas Compatriot Affairs Commission, R.O.C., 2000) .................................................................................... 85 Table 9 Textbooks Used in America’s Universities (Luo, 2010:73-74) .......................... 86 Table 10 Textbooks Examined in This Study................................................................... 87 . vii . .

(9) LIST OF FIGURES Figure 1 Types of Reference (Halliday & Hasan 1994: 33) ........................................... 34 Figure 2 The Overall Performance of Errors in Cohesive Device ................................. 61 Figure 3 Errors of Reference at Different Chinese Proficiency Level ........................... 79 Figure 4 Errors of Ellipsis at Different Chinese Proficiency Levels.............................. 80 Figure 5 Errors of Conjunction at Different Chinese Proficiency Levels ...................... 81 Figure 6 Errors of Lexical Cohesion at Different Chinese Proficiency Levels.............. 83 . viii . .

(10) CHAPTER 1 INTRODUCTION 1.1 Significance of This Study The significance of this study lies in two main areas. Theoretically, it is intended as a contribution to understanding the types of discourse errors learners exhibit, the sources from which the errors originate, and the effects the learners’ L2 proficiency makes on their learning. Specifically, this study investigates English-speaking learners’ Chinese compositions in order to analyze their discourse errors in terms of cohesion, information structure, and thematic structure. Form and function are both considered in the study. The sources of errors for each type are discussed from the interlingual and intralingual perspectives. Additionally, this study examines the relationship between discourse errors and the learners’ L2 Chinese proficiency. Pedagogically, by categorizing discourse errors, identifying their sources and examining the errors across different contexts and proficiency levels, this research seeks to provide useful information for Chinese language teachers. Specifically, it is hoped that by the results of this research, teacher of L2 Chinese writing will be able to design effective methods and materials for teaching writing to English-speaking learners.. 1.2 Statement of the Problem Writing is the most difficult skill for Chinese learners, compared to listening, speaking, and reading (Kao, Li, & Kuo, 1993), since it is rooted in the successive acqusition of the other three skills. Writing is concerned not only with the problems of syntax at the sentence level, but also with problems of textual organization at the 1. .

(11) discourse level. In addition, writing involves organizing and transmitting information. The writer is in the process of interacting back and forth between unceasingly complex knowledge while continuously developing text (Bereiter & Scardamalia, 1987). Chinese language teachers are concerned with the role of writing in their instructional planning because learners experience language by listening, speaking, reading and writing and must learn the skills interdependently. Thus, learners are often required to write compositions in their language courses. However, many errors are found in learners’ writing since a composition is a discourse or text task which involves more advanced skills than those at the sentence level. The well constructed text encompasses various linguistic components, such as appropriate use of cohesive devices, correct sequencing of information structure and thematic structure. Any inappropriate addition, omission, misordering, mis-selection or substitution can lead to a disintegration of cohesion and coherence in discourse. Discourse errors occur in learner’s compositions as a result of their lack of sufficient discourse competence. Although advanced L2 Chinese learners can speak Chinese with perfect fluency and few errors, many errors can, nevertheless, be found in their compositions, especially at the discourse level. Given the fact that L2 learners often lack discourse competence and produce discourse errors, it is necessary to examine the errors from a discourse perspective. Hence, this study uses the compositions produced by English-speaking learners of Chinese as the data base for error analysis at the discourse level. 1.3 Objectives of the Study As noted above, observation of learners’ errors at the discourse level is essential for developing learners’ ability of writing. Traditionally, language teaching progresses from 2. .

(12) vocabulary to phrases, phrases to simple sentences and simple sentences to complex sentences. The concept of discourse was largely ignored until scholars argued that the research on languages should go beyond the sentence level to the discourse level in 1950s (Harris, 1952; Weinrich, 1967; van Dijk, 1972). In 1970s, Chinese scholars introduced western theories of discourse, and began to apply these theories to analyze Chinese discourse. With the development of Chinese teaching, discourse analysis and error analysis targeting on Chinese writing have become more and more important, and resulted in more attention by researchers in this field. With respect to learners’ discourse errors in Chinese, previous research examined the errors. However, a majority of them concentrated on the presentation and classification of errors. The sources of errors were largely ignored, and pedagogical applications were rarely provided. In addition, the relationship between the discourse errors and learners’ Chinese proficiency was seldom discussed. Thus, this research attempts to go beyond current knowledge and understanding of language errors by providing a more comprehensive analysis and discussion. Therefore, the research objectives formulated in the present study are indicated below: (1) To identify, describe, classify discourse errors and sources of errors made by English-speaking learners of Chinese. (2) To compare and contrast categories and frequency of discourse errors for each level of language proficiency. (3) To identify the order of relative frequency in the categories of errors for each level of language proficiency in order to establish a hierarchy of frequency, which may reveal a hierarchy of learning difficulty for the categories at the discourse level. 3. .

(13) Based on the above objectives, the following two research questions are constructed. (1) In English-speaking learners’ compositions, what types of errors do learners commit at the discourse level? What are the causes of these errors? (2) Do the errors vary according to the learners’ L2 Chinese proficiency?. 1.4 Description of This Study In this study, an error analysis was conducted to analyze learners’ error types and the relationship between the errors and L2 Chinese proficiency. In the error analysis, the output produced by language learners is analyzed for errors through which the sources of such errors are identified and the frequency of error is considered proportional to the degree of learning difficulty. Corder (1976) points out that errors are used as an effective tool by learners to verify their hypotheses about the new system. Moreover, by classifying the errors that learners make as well as inferring the strategies that learners are using, an abundance of knowledge would be obtained by researchers interested in the process of second language acquisition. Learners would also benefit from their errors and, in fact, errors are considered indispensable in the process of learning according to Selinker (1992). Richards, Platt, and Platt (1992) believe that Error Analysis could be used to confirm learners’ command of a language as well as tool to examine the their learning difficulties. However, the theory was criticized due to its limiatons. For instance, Ellis (1994) suggests that a process of combining qualitative and quantative research practices appropriately to the interpretation of errors would contribute more to the research on second language acquisition. Therefore, interviews were also conducted in this research in order to provide a more comprehensive discussion of the learner errors.. 4. .

(14) 1.5 Organization of the Thesis Six chapters are included in this study, which. is organized as follows. Chapter 1 provides an overview of the thesis, including its significance, objectives, descriptions, and organization. Chapter 2 is the literature review, which is presented based on the central concepts of this research. Previous research such as error analysis, discourse analysis, linguistic typology, and other research on English-speaking learners’ Chinese discourse errors are reviewed. Chapter 3 delineates research methodology and design. Chapter 4 presents details of results and discussions. Chapter 5 provides pedagogical implications and applications. Finally, Chapter 6 is the conclusions and suggestions for further research.. 5. .

(15) 6. .

(16) CHAPTER 2 LITERATURE REVIEW This chapter presents an overview of previous work on related topics that provide the necessary background for the purpose of this research. The sections of this chapter that follow will include (1) error analysis, (2) discourse analysis, (3) linguistic typology, and (4) research on English-speaking learners’ Chinese discourse errors.. 2.1 Error Analysis In the literature of second language learning, two major linguistic methods were used to investigate learners’ target language productions: one was Contrastive Analysis, the other was Error Analysis. Contrastive Analysis was the main technique in the study of L2 learners’ errors in the 1950s and 1960s. It provided predictions of the potential difficulties L2 learners might have by systematically identifying the similarities and differences between native languages and target languages. By the late 1960s, doubts rose over the reliability of CA theory because very often errors predicted by CA did not really occur in learner's performance and errors which did arise were not always predicted by CA. Consequently, Error Analysis that focused on the errors learners made became an alternative. Corder (1967) first advocated in English language teaching/applied linguistics community the significant role errors play in a language learning process. Corder proposed that the process of language acquisition and the strategies of language learners were essentially the same for both first and second language learning. In addition, in L1 acquisition, children's incorrect utterances provided important evidence of language acquisition and language development. By classifying the errors learners made, 7. .

(17) researchers could learn how the language was acquired by making conclusions of the strategies learners used. Therefore, Corder emphasized the importance of errors in three aspects. First, teachers knew what needed to be taught; second, researchers realized how learning progressed; third, learners tested what they hypothesized about the L2. Other researchers also showed great interest in errors. Gass and Selinker (1983) defined errors as “red flags” that provided evidence of the learner’s knowledge of the second language. Researchers believed errors contain precious information regarding the learning strategies of language learners (Richards, 1974; Taylor, 1975; Dulay & Burt, 1972). Moreover, teachers could use them to assess learners’ learning and determine priorities for future teaching (Richards & Sampson, 1974). The sections that follow will review several key parts that constitute the Error Analysis theory: (1) definition of errors, (2) error and mistake, (3) sources of errors, (4) the surface structure Taxonomy, and (5) the five steps of error analysis.. 2.1.1 Definition of Errors Lennon (1991) indicated that in the process of investigating the errors of any L2 corpus, a definition of error should be established with reference to which potential errors could be assessed. Several researchers give definitions of errors as follows: (1) Errors are those flawed parts of learners’ conversation or composition that deviate from selected norm of mature language performance (Dulay et al., 1982: 138). (2) Generally. speaking,. errors. are. the. learner’s. incorrect. production. or. misunderstanding of the target language. It could be grammatical or pragmatic (Hu et al., 1989: 329). 8. .

(18) (3) An error is a linguistic form or a combination of structures that would not be produced by L2 native speakers under similar conditions or in the same context (Lennon, 1991: 182).. 2.1.2 Error and Mistake In order to analyze the learner’s language in an appropriate perspective, distinguishing “error” and “mistake” is essential. Corder (1967) held that it was of great importance to distinguish competence and performance. He argued mistakes were incorrect forms of “performance”: e.g. “unsystematic” slips of the tongue. However, errors were “systematic”, resulting from inadequate knowledge of the system of language. Corder (1979) also noted “An error is lack of competence and mistake is performance deviant.” Ellis (1997) pointed out that mistakes reflected occasional lapses in performance and they occurred because the learners failed to perform what they already knew. On the other hand, errors occurred because learners did not have enough knowledge to produce the correct forms, hence they reflected gaps in learners’ knowledge. Brown (2007) argued a mistake referred to a performance error in which learners failed to perform the understood knowledge correctly, such as a “slip” or “typo” while an error referred to a competence error, reflecting a gap in people’s knowledge of the language. It was an obvious deviation from the grammar used by adult native speakers. Richards and Schmidt (2002) stated that mistakes arose from distraction, exhaustion, inadvertence whereas errors occurred by imperfect learning. To summarize, an error is caused when the deviation arises as a result of lack of knowledge while a mistake or slip is made when the incorrect forms occurred as a consequence of incomplete performance of learners’ competence in the target language. 9. .

(19) 2.1.3 Sources of Errors Sources of errors are discussed and investigated by many researchers. Selinker (1972) identified five sources of errors: language transfer, transfer of training, strategies of. second. language. acquisition,. strategies. of. second. language. interaction,. and overgeneralization of linguistic structures. In addition, Corder (1974) noted three sources of errors: language transfer, overgeneralization or parallelism, and teaching-elicited error. Moreover, Richards and Sampson (1974) provided seven sources of errors: language transfer, intralingual obstruction, sociolinguistic context, modality, age, successions of approximate forms, universal hierarchy of difficulty. Brown (1980) proposed four sources of errors: interlingual errors, intralingual errors, context of learning, and communication strategies. James (1998) also indicated three sources of error: interlingual, intralingual, and induced errors. Chen (2007) concluded that sources of errors could be categorized into two types: interlingual errors and intralingual errors, and the latter could be further divided into overgeneralization and simplification. These are discussed in the following.. 2.1.3.1 Interlingual Errors Interlingual errors are the errors attributed to the native language. Therefore, they often refer to mother-tongue interference or L1 negative transfer (Lado, 1964). Extensive studies have already been done in the area of native language interference by contrastive analysts. However, error analysis holds a different concept towards L1 negative transfer from the contrastive analysis. Error analysis does not consider errors to be the persistence of old habits, but rather as a significant source that learners are acquiring and exploring the system of the second language. 10. .

(20) 2.1.3.2 Intralingual Errors Intralingual errors are considered as errors independent of learners’ native language. Richards (1974) held that learners tried to derive the rules based on what they have learned, but produced incorrect forms that resemble neither the native language nor the target language due to wrong hypotheses. Further, Richards and Sampson (1974) proposed that intralingual errors displayed generalizations resulting from fractional exposure to the target language rather than the structures or forms of the native language. Richards also categorized intralingual errors into four types: (1) overgeneralization or transfer of learned structures in the L2, (2) ignorance of rule limitations in L2 structures, (3) incomplete utilization of L2 rules, and (4) incorrect hypotheses caused by semantically or grammatically misunderstanding the rules in the target language. In addition, Ellis (1997) noted some types of errors were found to be prevalent because errors were the reflections of learners’ attempts to make language learning and application simpler. As for the differences between interlingual errors and intralingual errors, Brown (1994) argued that interlingual errors occurred more at the early stages of language learning; however, generalization within the target language appeared more when learners were gradually acquiring the second language. Intralingual errors could be further classified into overgeneralization and simplification.. Overgeneralization Overgeneralization often occurs due to the inappropriate application of target language rules. Richards (1974) maintained overgeneralization happened where the learner produced a deviant form based on his knowledge of other structures of the same 11. .

(21) linguistic domain in the target language. After having learned sentences such as “He reached the house at 10:00 o’clock,” an English learner may produce another sentence “*He leaved the house at 10:00 o’clock” (*leaved->left). The following passage illustrates the example of overgeneralization. (1) *一次，我的電腦軟體出了問題，我打通了服務熱線。接電話的是個小夥子，他說能聽懂英語，我就用英語告訴他我的電腦的問題。但他不明白我說的問題。以後我用漢語告訴他我的軟體的問題，他聽懂了，告訴了我解決辦法。 *Yícì, wǒde diànnǎo ruǎntǐ chūle wèntí, wǒ dǎtōng le fúwù rèxiàn. Jiē diànhuà de shì ge xiǎohuǒzi, tā shuō néng tīngdǒng yīngyǔ, wǒ jiù yòng yīngyǔ gàosù tā wǒde diànnǎo de wèntí. Dàn tā bù míngbái wǒ shuōde wèntí. Yǐhòu wǒ yòng hànyǔ gàosù tā wǒde ruǎntǐ de wèntí, tā tīngdǒng le, gàosù le wǒ jiějué bànfǎ. One time, my computer has a software problem. I called the service center and one guy picked up the phone. He said he understood English, so I told him my problem in English. But he does not understand me. Then I spoke Chinese to explain my software problem. He understood me and told me how to solve the problem. （Chao, 2005: 75）. In (1), the student was faced with a software problem in his computer. In the beginning of the example, he used English to express his problem, but the receptionist did not understand him. So, he tried to explain the situation in Chinese. However, yǐhòu (after) was chosen incorrectly in his attempt to connect the first part of the passage and the second part, since a reference point should be added before yǐhòu to express the time sequence, meaning “after” (the reference point) while ránhòu (“then”) could be used alone to connect the two sections. As a result, the learner mistakenly used yǐhòu alone to mark the time sequence of two things according to the rule of ránhòu. These types of overgeneralization errors were made due to an unfamiliarity of the second language rules and extension of the rules in inappropriate contexts.. 12. .

(22) Simplification Richards (1974) claimed that simplification was a learning strategy learners adopted to make learning easier. It occurred where a linguistic form in L2 was omitted by learners (George, 1972). Learners chose an easier and communication-driven method from a variety of hypotheses when they were learning or using a new language (Ellis, 1985). For instance, when it comes to using conjunctions to link words or phrases, learners prefer to choose the easiest one hé instead of bìng, bìngqiě, jí, yǔ, and tóng (All correspond to “and” in English). Even though there are a great many options in Chinese, hé is usually used to join nouns or noun phrases. However, the conjunction has constraints when connecting adjectives and verbs. Consequently, learners make errors as the sentences below: (2) *辦公室的環境比較有意思和年輕人在他們的工作上比較出色。 *Bàngōngshì de huánjìng bǐjiào yǒu yìsī hé niánqīng rén zài tāmen de gōngzuò shàng bǐjiào chūsè. The office environment is more interesting and young people have better performance at work. (3) *我們的教室很大和很乾淨。 *Wǒmen de jiàoshì hěn dà hé hěn gānjìng. Our classroom is big and clean. （Chao, 2005: 77）. Thus, simplification errors result from producing simpler linguistic forms.. 2.1.3.3 Other Sources of Errors In addition to interlingual and intralingual errors, researchers also discussed other 13. .

(23) sources of errors, including communication strategy based errors and induced errors. Brown (1980) believed communication strategy referred to learners’ conscious employment of verbal systems to deliver an idea when they had difficulties using linguistic forms correctly. Communication strategy was not discussed in this study because it was not related to learning process. As for induced errors, they were often initiated by the context of learning while interlingual and intralingual errors were usually elicited by learners themselves (Stenson, 1983). Brown specified that the context of learning consisted of teachers, teaching materials, and places where learning took place, such as a classroom or a social situation. Brown (1980) indicated that learners made errors because of a misleading explanation from the teacher, improper examples in a textbook, or a structure practiced in a drill but not appropriately contextualized. James (1998) also noted that induced errors took place when students were misled by teachers’ definitions, examples, explanations and arrangements of practice. James further classified induced errors into the following subcategories: material-induced errors, teacher-talk induced errors, exercise-based induced errors, pedagogical priority induced errors, and look-up errors. The sections that follow will present these induced errors.. Material-induced Errors James (1998) stated that material-induced errors were caused by inappropriate examples or explanations in the textbook. He also provided the following example to elaborate that students might learn incorrect forms from their textbook in which sentence (b) with present progressive tense was used to respond sentence (a). The correct response should be sentence (c).. 14. .

(24) (4) (a) “What’s the weather like in autumn?” (b) “*Many people are wearing raincoats and carrying umbrellas.” (c) “Many people wear raincoats and carry umbrellas.” (James, 1998: 191). Chen (2007) noted instances of inappropriate explanations in the textbook. The pattern “shì…de” was used to describe something that happened in the past, as shown in example (5). However, it could also be used in the future context, as shown in example (6) and (7). (5) 他是昨天來的． Tā shì zuótiān lái de. It is yesterday that he came. (6) 他是明天才開學的，所以今天不會來了． Tā shì míngtiān cái kāixué de, suǒyǐ jīntiān búhuì lái le. It is tomorrow that the school starts, so he will not come today. (7) 死了這條心吧！她是不會嫁給你的． Sǐ le zhètiáo xīn ba! Tā shì búhuì jià gěi nǐ de. Give it up! She will not marry you. （Chen, 2007: 23）. Teacher-talk induced Errors Teacher-talk induced errors result from teachers’ imprecise explanation which elicits its erroneous use of the second language (James, 1998). Stenson (1983) mentioned the methods teachers used to present materials and define lexical words might cause learners to commit errors. For instance, the word “worship” was often explained as “pray” by teachers. Thus, students produced 15. .

(25) “worshipping to God” because they overgeneralized the rule based on the old knowledge that ‘to” was the preposition attached to “pray”. In addition, grammatical errors may be caused due to insufficient or defective explanations by teachers. For example, Stenson found out that a teacher defined “as if” more or less synonymous with “like” into sentences with “as if”, and then required students to transform sentences with “like” into sentences with “as if”. Hence “He climbs like a monkey” should be transformed into “He climbs as if he were a monkey.” However, “*She cries as if the baby cries” was produced by a student for the original sentence “She cries like a baby” because the student thought “as if” was synonymous with “like” (Stenson, 1983). Liu (2005) provided the following sentences to explain the incorrect use of “dōu” caused by teachers’ inappropriate explanation. (8) *我們班都有13個人。 *Wǒmen bān dōu yǒu 13 ge rén. *Our class has 13 people. （Liu, 2005: 123） (9) *我們學校都有6個食堂。 *Wǒmen xuéxiào dōu yǒu 6 ge shítáng. *Our school has 6 dining halls. （Liu, 2005: 123）. In the two examples, “dōu” should be deleted. Liu (2005) pointed out the errors were elicited by the way teachers explained its meaning. Teachers tended to tell students “dōu”, meant “all”; consequently, “dōu” was chosen to deliver the concept of the aggregate number.. 16. .

(26) Exercise-based induced Errors Exercise-based induced errors refer to errors that result from specific exercises that require manipulation of language patterns (James, 1998). Students make errors when doing exercises of sentence combining, in which they are required to combine two simple sentences. Conditionals linked by “if” or “unless” are examples: (10) (a) I can’t afford a new car. (b) I shall win the lottery. (c) I can’t afford a new car unless I win the lottery. (d) *Unless I can afford a new car I shall win the lottery. (Anchalee & Somchoen, 2007: 177). In (10), sentence (a) and (b) should be combined to form sentence (c); however, the sentence (d) was produced by students, especially when they were told that “unless” was equivalent to “if…not”, which would cause them to replace the negative element in “can’t” with “unless”. Chen (2007) used the following examples to illustrate the problem: (11) 那隻狗把你怎麼了？ Nà zhī gǒu bǎ nǐ zěnme le? What did the dog do to you? (12) 我被那隻狗咬傷了． Wǒ bèi nà zhī gǒu yǎoshāng le. I was bitten by the dog. (13) 那隻狗把我咬傷了． Nà zhī gǒu bǎ wǒ yǎoshāng le. The dog bit me.. 17. .

(27) （Chen, 2007: 23）. In transformational drills, students were required to use bèi structure to answer (11). Even though (12) was correct, but the two sentences were not coherent semantically. The normal response would be sentence (13), “Nà zhī gǒu bǎ wǒ yǎoshāng le.” (the dog was bitten by me) “Wǒ bèi nà zhī gǒu yǎoshāng le” (I was bitten by the dog) would be the appropriate response to “Nǐ zěnme le?” (what happened to you?) because people want to know what happened to the person; therefore, the answer “being bitten by the dog” would be the appropriate response. As for the original sentence “What did the dog do to you”, “bit me” should be the answer to response. Hence, the actual function of a language is communication; thus, teachers should devise teaching activities which focus on authentic interaction rather than were structural drills.. Errors Induced by Pedagogical Priorities Teachers’ expectations have a great influence on learners’ achievement. Some teachers choose to prioritize one of the following: accuracy, fluency or the idiomatic in teaching communication, hence if fluency is regarded as superior, accuracy would have a lower priority or vice versa (James, 1998).. Look-up Errors Look-up errors are errors resulting from using a dictionary or grammar book then using the new words inaccurately (James, 1998). For example, a student wanted to speak “when” in “when I went to China, I went to Shanghai to buy things”, and she looked up the dictionary for the word. The dictionary used “shémeshíhòu” to explain the word. As a result, she misused the reference aid, producing “shémeshíhòu wǒqù Zhōngguó, wǒ qù Shànghǎi mǎi dōngxī” 18. .

(28) 2.1.4 The Surface Structure Taxonomy The surface structure taxonomy is based on “the way surface structures are altered”, and errors can occur because of change in surface structure in specific and systematic ways (Dulay, Burt & Krashen, 1982: 150). Errors are classified into five patterns: omission, addition, misordering, mis-selection, and blends. Each type will be illustrated as follows.. 2.1.4.1 Omission Omission refers to the absence of an element, which results in incorrect forms of the language. Learners in the early stages of the second language acquisition process tend to omit function words rather than content words (James, 1998). James (1998) distinguished ellipsis and omission, as in the following sentences: (14) (a) He’ll pass his exam, but I won’t (pass my exam). (ellipsis) (b) *He’ll pass his exam, and I’ll (pass my exam) too. (omission). James (1998: 106) Ellipsis refers to an intentional omission of words or phrases to make the passage coherent, while omission refers to an error which occurs as a result of words omitted. The phrase “pass my exam＂ in (a) is grammatical to be omitted. However, it is ungrammatical to be omitted in (b). Chen (2007) used the following Chinese examples to explain the difference between ellipsis and omission.. 19. .

(29) (15) 他昨天看了那部電影，我也看了（那部電影）．(ellipsis) Tā zuótiān kànle nà bù diànyǐng, wǒ yě kànle (nà bù diànyǐng). He watched that movie yesterday, so did I.. （Chen, 2007: 24）. (16) *這個計畫繁榮了國家經濟，那個計畫也繁榮了（國家經濟）．(omission) *Zhè ge jìhuà fánróng le guójiā jīngjì, nàge jìhuà yě fánróng le (guójiā jīngjì). *This plan prospered the national economy, and that plan also prospered (national economy). （Chen, 2007: 25）. It is grammatical to omit “nà bù diànyǐng” in (15) while it is ungrammatical to omit “guójiā jīngjì” in (16). Chen (2007) also cited several examples of the omission of functional words. (17) *笛子的聲音雖然很小，可是各地（都）能聽到． *Dízi de shēngyīn suīrán hěn xiǎo, kěshì gèdì (dōu) néng tīng dào. Although the sound of a flute is not loud, people hear it everywhere. （Li, 1996: 105, 106） (18) *他有兩（個）孩子． *Tā yǒu liǎng (ge) háizi. He has two children. （Tong, 1986: 20） (19) *只要我們努力學習，就一定（能）提高漢語水平． *Zhǐyào wǒmen nǔlì xuéxí, jiù yídìng (néng) tígāo hànyǔ shuǐpíng. *As long as we study hard, we (can) definitely enhance our Chinese proficiency. （Tong, 1986: 50）. Moreover, the omission example at the discourse level is shown below:. 20. .

(30) (20) a我們應該採取辦法來解決這些問題，比如，讓政府和學校解決貧困學生的學費問題，（因為）b所有的人都渴望得到教育和知識。 aWǒmen yīnggāi cǎiqǔ bànfǎ lái jiějué zhèxiē wèntí, bǐrú, rang zhèngfǔ hé xuéxiào jiějué pínkùn xuéshēng de xuéfèi wèntí, （yīnwèi） b suǒyǒu de rén dōu kěwàng dédào jiàoyù hé zhīshì. a We should adopt some measures to solve these problems. For example, devising some ways to push the government and the school to deal with the problem of tuition for students who have difficulties paying them （because） b all people desire to receive education and knowledge. （Yuan, 2009: 30）. In (20), part b is the cause, and part a is the effect. Thus, “yīnwèi” should be added to part B to present the cause-effect relationship of the two clauses. (21) a求婚的程序很簡單：b首先男的會偷偷地去問女孩子的父親是否同意將女兒嫁給自己，c若得到肯定的回答，就會去買一個戒指，d（然後）才會去向女孩子求婚。 aQiúhūn de chéngxù hěn jiǎndān: b shǒuxiān nán de huì tōutōudì qù wèn nǚháizi de fùqīn shìfǒu tóngyì jiāng nǚér jià gěi zìjǐ, c ruò dédào kěndìng de huídá, jiù huì qù mǎi yíge jièzhǐ, d（ránhòu）cái huì qù xiàng nǚháizi qiúhūn. a The process of a proposal is very simple: b first, the boy will ask the girl’s father to see if he agrees the marriage. c If the answer is positive, he will buy a ring, d and then propose to the girl. （Yuan, 2009: 30）. In (21), the paragraph describes the steps of a proposal. “Shǒuxiān” is used in b to mark the first step, and part c is the second step in which the man would purchase a ring if the father agrees to the marriage. Part d is the last step; therefore, “ránhòu” should be added to signal the coherence of the passage.. 21. .

(31) 2.1.4.2 Addition This type of error, which is the “result of all-too-faithful use of certain rules”, arises when any unnecessary additions of characters or items occur in students’ production (Dulay, Burt & Krashen, 1982: 156). Dulay, Burt and Krashen (1982) subcategorized this error into two types: regularization and double marking. They argued that regularization was caused by overlooking exceptions and applying incorrectly the old knowledge to other domains. For example, students’ production of the incorrect “goed” for “went”. Chen (2007) cited errors of the Chinese verb-copying construction to explain the addition errors in Chinese, as the following example shows: (22) *他發現這件事發現兩年了． *Tā fāxiàn zhèjiàn shì fāxiàn liǎngnián le. *He has discovered this for two years. （Chang, 1991a: 7） (23) *我畢業畢了三年了． *Wǒ bìyè bì le sān nián le. *I have graduated for three years. （Chang, 1991b: 410-415） (24) *我吃完飯吃了三個鐘頭了． *Wǒ chī wán fàn chī le sān ge zhōngtou le. *I have eaten for three hours. （Chang, 1991b: 410-415）. In (22), (23), and (24), the Chinese learners used the verb-copying construction and misapplied the rule to those sentences without considering the semantic constraints of 22. .

(32) the verbs. In addition to regularization, double marking is another type of addition error. It is defined as a failure to delete certain items which are required in some linguistic constructions but not in other structures, as in the example of “He doesn’t knows me.” One of the Chinese examples is given below: (25) *是張文幫我修改了作文的． *Shì Zhāngwén bāng wǒ xiūgǎi le zuòwén de. It is Zhangwen who revised my composition. （Tong, 1986: 40） (26) *這三個地方的風景都是很美． *Zhè sānge dìfāng de fēngjǐng dōu shì hěn měi. The three places have beautiful landscapes. （Tong, 1986: 126）. Besides the errors of addition mentioned above, in the discourse domain, extensive research also showed that English native speakers tended to overuse pronouns which should be deleted in a discourse (Xie, 1992; Gao, 1996). The following example illustrates the student’s overuse of “wǒmen” (we) in the paragraph. (27) 大家祝我們快樂，還送我們很多禮物。我們吃生日蛋糕。我們唱生日快樂歌。我們喝啤酒的時候，我們唱得非常難聽。 Dàjiā zhù wǒmen kuàilè, hái sòng wǒmen hěnduō lǐwù. Wǒmen chī shēngrì dàngāo. Wǒmen chàng shēngrì kuàilè gē. Wǒmen hē píjiǔ de shíhòu, wǒmen chàng de fēicháng nántīng. All the people wished us a happy birthday, and they gave us a lot of presents. We had cakes. We sang happy birthday songs. We did not sing well when we drank beer. （Yang, 2004: 63） 23. .

(33) 2.1.4.3 Misordering Misordering refers to deviations associated with the order of the constituents of the sentence. In other words, the errors are characterized by the incorrect placement of a morpheme or a group of morphemes in an utterance. Errors of this kind are usually interlingual errors which result from translating the target language according to the native language (Dulay, Burt & Krashen, 1982: 163), as seen in the following examples: (28) *He every time come late home. （Huang, 2010: 7）. In (28), the learner wrote the English sentence according to the left-branching order of Chinese. Thus, an ungrammatical sentence was created.. (29) (a) *我去辦公室時候，她在談話和同事。 *Wǒ qù bàngōngshì shíhòu, tā zài tánhuà hé tóngshì. When I went to the office, she was talking with her collegue. (b) 我去辦公室時候，她和同事在談話。 Wǒ qù bàngōngshì shíhòu, tā hé tóngshì zài tánhuà. (c) *他在蘇州住跟一個韓國朋友。 *Tā zài Sūzhōu zhù gēn yíge Hánguó péngyǒu. He lives in Suzhou with a Korean friend. (d) 他在蘇州跟一個韓國朋友住。 Tā zài Sūzhōu gēn yíge Hánguó péngyǒu zhù. （Huang, 2010: 7）. In (29), the learner produced the Chinese sentences ungrammaticlly because he/she wrote the sentence based on the right-branching order of English. It is clear that the 24. .

(34) word order of the incorrect Chinese sentences (a) and (c) corresponds to that of the English sentences.. 2.1.4.4 Mis-selection Mis-selection refers to the fact that learners choose the incorrect forms instead of the correct ones in the same language domain (Dulay, Burt & Krashen, 1982: 160), as shown in (30). (30) *He hurt herself. *I read that books.. －＞He hurt himself. －＞I read that book. (James, 1998: 108). In (31), “néng gǎibiàn chuántǒng sīxiǎng” (can change the traditional ideas) is the cause, not the goal. However, the learner chose the incorrect word “wèile” (in order to) to deliver the meaning of “yīnwèi” (because) in the sentence.. (31) a 歷史上的女英雄就是（為了）［因為］能改變傳統思想，b 才被大家稱讚。 a Lìshǐ shàng de nǚ yīngxióng jiù shì (wèile), [yīnwèi] néng gǎibiàn chuántǒng sīxiǎng, b cái bèi dàjiā chēngzàn. b Heroines in the history were acclaimed by people a (in order to) [because] they changed traditional ideas. （Yuan, 2009: 30）. 2.1.4.5 Blends Blending arises when two alternative grammatical forms are combined to produce an ungrammatical blend. The examples are as follows:. 25. .

(35) (32) *According to Eric’s opinion －＞According to Erica + In Erica’s opinion * The question is easy to be answered－＞The question is easy to answer + The question is easily answered (James, 1998: 111). Chen (2007) also provided the following examples to illustrate the errors of blending in Chinese. (33) *他很高高興興地回去了－＞他很高興地回去了＋他高高興興地回去了 Tā hěn gāogāoxìngxìng dì huíqù le -> tā hěn gāoxìng dì huíqù le+ tā gāogāoxìngxìng dì huíqù le. (Tong, 1986: 28) (34) *現在我們的漢語說得越來越比較好－＞現在我們的漢語說得越來越好＋現在我們的漢語說得比較好 Xiànzài wǒmen de Hànyǔ shuō de yuèlái yuè bǐjiào hǎo-> xiànzài wǒmen de Hànyǔ shuō de yuèlái yuèhǎo+xiànzài wǒmen de Hànyǔ shuō de bǐjiào hǎo. （Tong, 1986: 285）. Based on the five types of form errors, this study examines students’ compositions to categorize their discourse errors.. 2.1.5 Five Steps of Error Analysis Corder (1967, 1974) pointed out five procedures of error analysis, including collection of error, identification of errors, descriptions of errors, explanations of errors, and evaluations of errors.. 2.1.5.1 Collection of Errors The main objective of the first step is selecting a proper collection system. 26. .

(36) 2.1.5.2 Identification of Errors Once a corpus of learner language has been collected, the errors have to be identified. Four steps are included in this stage. First, set up the target language used for the evaluation. Second, distinguish between mistakes, lapses and errors. Mistakes occur because of some factors such as fatigue, carelessness, and stress rather than lack of competence. However, errors occur because learners do not have the competence to perform the correct forms or items of the target language. Third, make correct interpretation of the learners’ intentions. There are two types of interpretations: one is overtly erroneous, the other is covertly erroneous. The former takes place in the structures that are superficially deviant (“She goed to the park”) and the latter occurs in utterances that are superficially well formed but do not deliver the meaning the learner intends to mean (“Where do you go?”). Fourth, focus on deviations, including correctness and appropriateness. Correctness is an incorrect language usage (“I did went to the museum”) and appropriateness is an incorrect language use (“She can to do whatever she wants”).. 2.1.5.3 Description of Errors The third stage is describing the errors. A comparison of the learner's idiosyncratic utterances with a reconstruction of those utterances in the target language is necessary in this stage. Linguistic categories and surface strategy are two descriptive taxonomies of errors included in the process. The first can be chosen according to the rules in the grammatical books or textbooks. The other refers to the description of the superficial errors, such as omissions, additions, misorderings, and etc. Corder (1974) argued that researchers needed to focus on errors that occurred 27. .

(37) repeatedly. He also pointed out that the goal of error analysis was to explain learners’errors linguistically and psychologically. Therefore, the pattern the learners might use can be predicted. Once the errors are described properly, explanation of errors can begin.. 2.1.5.4 Explanation of Errors Two main sources of errors are proposed in foreign language learning. One regards errors as the influence of the mother tongue. The other holds that second language learners would commit errors as children do when they learn their first language because they consider the processes used in acquiring a first and a foreign language are identical. However, Corder (1967) proposed that errors would not necessarily be related to either the mother tongue or the target language since learners developed their own inter-language grammars, idiosyncratic dialects or approximate systems.. 2.1.5.5 Evaluation of Errors In error evaluation studies, judgments were based on three basic categories: comprehensibility, seriousness and naturalness of the grammar and the lexis. Brown (1994: 207-211) and Ellis (1995: 51-52) elaborated the model Corder (1967, 1974) proposed. Ellis (1997: 15-20) and Hubbard et al. (1996: 135-141) provided advice and examples on identifying and analyzing learners’ errors. In addition, Gass and Seniker (1994: 67) indicated 6 stages of error analysis: collection of data, identifications of errors, classifications of errors, quantifications of errors, analyses of sources of errors, and remediation of errors. In this study, in an attempt to investigate the problematic areas of English natives’ Chinese writing, error analysis was used as a major tool to capture and demonstrate 28. .

(38) college students’ problems at the discourse level.. 2.2 Discourse Analysis 2.2.1 Definitions of Text and Discorse The terms “text” and “discourse” are used in different ways by linguists and researchers in different fields. Halliday (1978) believed that a continuous process of semantic choice was provided by a text. De Beaugrande and Dressler (1981) defined text as a communicative event that needed to include the following seven criteria: cohesion, coherence, intentionality, acceptability, informativity, situationality, and intertextuality. However, Tischer et al. (2000) argued that cohesion and coherence were text-internal while the other five criteria are text-external. Schiffrin (1994) used the term “text” to differentiate linguistic materials from the environment in which “sayings” occurred (context) (p. 363). In other words, the text consists of the text-internal components, whereas the context consists of the text-external elements. In addition, Schiffrin (1994) indicated that all approaches within discourse analysis regarded text and context as the two essences that contributed to the communicative content of an utterance. Even though some researchers viewed text linguistics and discourse analysis in different ways, some definitions of the two were found the same. Text linguistics was defined as the investigation of real language use (De Beaugrande, 2002), and discourse analysis was also defined as the analysis of any aspect of language in use (Schiffrin, 1994; Fasold, 1990; Brown & Yule, 1983; Candlin, 1997). Another important characteristic of discourse studies is that they are essentially 29. .

(39) multidisciplinary. Therefore, they can cross the linguistics border into different and various domains, such as linguistics, poetics, history, psychology, sociology, anthropology, semiotics, and communication research, and etc. (van Dijk, 2002). Thus, the study adopted the general definition of discourse analysis as the study of language in use, and regarded both text and context as parts of discourse rather than considering them to be the same. In addition, a broader view was employed, that is, to use text and discourse to refer to the students’ compositions.. 2.2.2 Introduction of Discourse Analysis As for the origins of modern discourse analysis, van Dijk (1985) indicated three main points: first, the early research focus was mainly on a descriptive and structuralist undertaking which was in the domains of linguistics and anthropology. Second, researchers dedicated to investigating native or popular discourse genres, including folktales, myths, stories, and some ritual interaction forms. Third, the functional analysis of sentence and discourse structure as well as the first interests toward text linguistics often held the distinctive or opposite views from the transformational grammar. Considering that scholars began to apply semiotic or linguistic methods to the studies of texts and communicative events in the 1960s, researchers regarded systematic discourse analysis as an independent orientation of research within and across several disciplines in the early 1970s. The 1970s experienced three important developments: first, sociolinguistic work paid specific attention to discourse because it focused on the importance of language variation in the sociocultural context. It diverged with the structuralist analysis whose research focus was on written language because of its emphasis on the spoken language 30. .

(40) as well as the functions a discourse serves. Second, researchers began to pay attention to speech acts since sociolinguistics valued the role of language variation and the social context. They regarded verbal utterances not only as sentences, but also as specific forms of social actions. Third, within the framework of a grammatical theory itself, it was repeatedly maintained that structural characterizations of isolated sentences should not be the only essences discussed. Therefore, the following themes about semantic macrostructures and other typical features of texts were studied by researchers: pronouns, cohesion markers, presupposition, topic and comment, and etc. Thus, the studies about actual language use appeared in the 1970s (Dressler, 1972; Petofi, 1971; Schmidt, 1973; van Dijk, 1972), and it developed to a more general, interdisciplinary, and expanding study of textlinguistics and discourse (e.g., Halliday & Hasan, 1976). Although discourse analysis is a relatively new field, different scholars still hold different definitions. Nevertheless, it is basically considered to be a general term for approaches to analyzing written, spoken, signed language use or any significant semiotic event, and its research has covered a variety of fields, such as linguistics, sociology, anthropology, social work, cognitive psychology, social psychology, international relations, human geography, communication studies and translation studies. In addition, discourse analysts not only analyze language use beyond the sentences, but also examine authentic language use.. 2.2.3 Systematic Functional Garmmar Halliday (1961) developed systematic functional grammar based on the research of J.R. Firth and the Prague School. Halliday argued that the form of language is presented based on the functional structure by grammatical patterns. He referred to his functions of language as 31. .

(41) metafunctions, and subcategorizes them into three general functions: the ideational, the interpersonal and the textual. The ideational metafunction, which we use to explain the “reality”, is the function for interpreting human experience. Halliday further divided the ideational function into two functions: the logical metafunction and the experiential metafunction. The former referred to the grammatical essences that could combine grammatical units to complex structures, and the latter referred to the grammatical essences involved in reasoning the human experience through the unit of the clause. The interpersonal metafunction describes the relationship to a text’s aspects of tenor or interactivity. It was the language use of speakers to participate in social activities, express their attitudes, make judgements, and interact with people. The textual metafunction dealt with the internal structure and the meaning delivery of a text. The interactivity in a text, spontaneity of the idea exchange and the distance of communication were included in this domain. It enabled a language to function in specific contexts, and achieve the ideational metafunction and the interpersonal metafunction through the use of a language. He proposed three textual functions: cohesion, information structure, and thematic structure. Cohesion was a textual function concerning the coherent relationship among sentences. He argued that linguistics needed to focus not only on the thematic organization of sentences, but also on the relationship between sentences and discourse. Information structure and thematic structure were textual functions discussed at the sentential level. The former was the study of how speakers structured sentences to convey new information linked to the preceding context by old information. The latter consisted of two elements, the Theme and the Rheme. The Theme, which was the point of departure of the message, was usually what the clause was concerned about. The 32. .

(42) Rheme, which was the remainder of the message, provided information about the Theme. This study analyzed students’ discourse errors based on the three discourse functions mentioned above: cohesion, information structure, and thematic structure. The following was the literature review of the three aspects.. 2.2.3.1 Cohesion Cohesion is the grammatical and lexical relationship within a text or a sentence. It is the use of explicit linguistic devices to signal the relationship between sentences and parts of texts to hold a whole text together. Halliday and Hasan (1976) maintained that coherence was what linked the discourse semantically. A text was different from a non-text due to its “texture” (p. 1), which was formed by the cohesive ties contributed to the whole unity. In other words, cohesion, which was explicitly presented, signaled underlying semantic relationships among the components of a text. In addition, Halliday and Hasan (1976) argued that a text was a passage of discourse which was coherent in the following two aspects. First, it was coherent in terms of the context of the whole article; thus, it would be consistent in register. Second, it was coherent itself; hence, the text was cohesive. Further, they identified five general categories of cohesive devices that signaled a coherent relationship in a text. They were reference, substitution, ellipsis, conjunction, and lexical cohesion. Reference, substitution, and ellipsis were grammatical; lexical cohesion was lexical; conjunction stands on the border line between the two categories. The sections that follow will discuss the five cohesive devices. 33. .

(43) Reference Reference refers to specific linguistic elements which cannot be interpreted semantically. Instead, it makes reference to something else within a discourse (Halliday & Hasan, 1994). (35) Sally preferred the company of herself.. (Wikipedia, the free encyclopedia) In (35), the subject of the second sentence “herself” referred back to the subject of the sentence “Sally”. Halliday and Hasan (1994) also distinguished several types of references as figure 1 shows:. Figure 1 Types of Reference (Halliday & Hasan 1994: 33) Exophoric reference was used to describe unconcrete existences that could not be found in a given text; thus, it was not considered cohesive. In contrast to exophoric reference, endophoric reference related to something within a given text (Halliday & Hasan, 1994).. 34. .

(44) (36) Mike: Hey John, did you just see that? John: Yes, that was amazing.. (Wikipedia, the free encyclopedia) In (36), “that” used as a reference item in the conversation was abstract; therefore, it could not be explicitly identified. To put it differently, it could not be retrieved from elsewhere in the text since it described something outside the text. In contrast, endophoric reference was employed to refer to something within the text. It could be either anaphoric or cataphoric. Anaphoric reference was aimed to avoid repetition by referring back to someone or something that had been previously identified. The example could be found in (35) in which “herself” referred back to “Sally” in the preceding sentence. Cataphoric reference occurred when something was introduced in the abstract form before it was identified. (37) There it is, my so much admired watch. (Wikipedia, the free encyclopedia). In (37), “it” refers forward to a specific element within the subsequent text.. Ellipsis Ellipsis was another cohesive device. It was characterized by “the omission of an item” to avoid repetition (Halliday & Hasan, 1994: 88). It was a relation within the text, and the presupposed item usually occurred in the preceding text (Halliday & Hasan, 1994). (38) Mary ate some chocolate chip cookies, and Robert [blank] some gummy bears. (Wikipedia, the free encyclopedia). In (38), the predicator “ate” is omitted in the second part of the sentence; however, 35. .

(45) it is presupposed because it is present in the first half of the sentence. The ellipsis is optional. Three different types of ellipsis were distinguished: nominal ellipsis, verbal ellipsis, and clausal ellipsis. Examples of these subclasses are given below. a. nominal ellipsis: (39) All photos were taken between 1980 and 1981 and nearly all (the photos) were shot in the vicinity of Petticoat Lane. (Christopher, 2003: 353). In (39), “nearly all” was used with ellipsis of “the photos”.. b. verbal ellipsis: (40) “Have you noticed?” “Yes, dear, of course I have (noticed).” (41) It might be true that Rose was the father. It might not (be true). (Christopher, 2003: 353). In (40) and (41), ellipsis of the lexical verb was used: “noticed” in the first sentence, and “be true” in the second sentence. c. clausal ellipsis: (42) “Bob is going to fly to Finland. I can tell (that he is going to fly to Finland) from his expression.” (Christopher, 2003: 354). In (42), the clausal complement of “tell” was ellipted.. Subsititution Substitution was the process in which one word within a text or discourse was substituted for another, more general word (Halliday & Hasan, 1994). (43) Jack’s car is very old and ugly. He should get a nicer one. 36. .

(46) (Wikipedia, the free encyclopedia). Example (43) showed this cohesive relationship in which the word “car” was replaced by the word “one”. The difference between reference and substitution is that the substituted items are always exchangeable by the items they refer to. In contrast, with reference the presupposed items can almost never replace the specific linguistic elements which refer to them. Examples appear below: (44) John goes fishing every other week. John is a very good fisherman. (45) *There watch is, my so much admired watch. (46) Jack’s car is very old and ugly. He should get a nicer car. (from Wikipedia, the free encyclopedia). In example (44), “he” could not be replaced by “John” because that would create ambiguity that the reader could be confused if the two “John” were the same person. As for example (45), it was incorrect to replace “it” with “watch”. However, in example (46), it was completely appropriate to replace “one” with “car” without changing the meaning of the sentence. Halliday and Hasan (1994) provided further subcategories of substitution, including nominal substitution, verbal substitution, and clausal substitution. Nominal substitution referred to the replacement of a noun by “one, ones, same”, as illustrated in (47). Verbal substitution referred to the replacement of a verb by “do”, as shown in (48). Clausal substitution was the replacement of a clause by “so” and “not”, as in (49) (p. 90). a. nominal substitution: one, ones, the same. (47) There are many backpacks here. Which one is yours?. 37. .