日籍學生中文分裂句以及準分裂句之第二語言習得

全文

(1)國立臺灣師範大學英語學系碩士論文 Master Thesis Graduate Institute of English National Taiwan Normal University. 日籍學生中文分裂句以及準分裂句之第二語言習得. Japanese Students’ L2 Acquisition of Clefts and Pseudo-clefts in Mandarin Chinese. 指導教授：陳純音博士 Advisor: Dr. Chun-yin Doris Chen 研究生：程昱瑋 Student: Yu-Wei Vivien Cheng. 中華民國一百零六年一月 January, 2017.

(2) 摘要. 本文旨在探討以日語為母語之學習者對於中文分裂句及準分裂句的學習情形，主要研究議題包含觀察學習者的分裂句及準分裂句的習得是否受到中文合法性效應、句構效應、分裂句類型效應、語境效應以及語言程度效應影響。本研究主要採用真假值測驗、接受度判斷測驗、以及問答選擇測驗。受試者分為實驗組以及控制組。前者為三十位以日語為母語且在國立台灣師範大學國語文中心學習中文的學生，並按其中文程度分為中低級及中高級兩組，每組十五名。後者則為十五位母語人士。首先，研究發現句子的合法性並不影響學習者在分裂句以及準分裂句的表現。就五種分裂片語類型來看，不同分裂片語類型也不影響學習者對分裂句及準分裂句的接受度。此外，在五種分裂句類型當中，受試者最能接受的是名詞、動詞、及介繫詞組，最後是副詞及形容詞。在語境效應方面，受試者會依語境的不同來選擇分裂句或準分裂句。此現象顯示受試者的表現的確受到語境效應的影響。最後，在語言程度效應方面，受試者之中文程度越高，其分裂句及準分裂句的表現越趨向母語者的表現，顯示受試者對中文分裂句以及準分裂句的掌握度會隨者其中文能力的提升而有所增進。. 關鍵詞 : 分裂句、準分裂句、中文合法性效應、句構效應、分裂句類型效應、語境效應、語言程度效應、第二語言習得. i.

(3) ABSTRACT The present study conducted an empirical study with an aim to investigate the Japanese students’ acquisition of Mandarin clefts and pseudo-clefts. The issues addressed in the present study included grammaticality effects, construction effects, cleft type effects, contextual effects, and proficiency effects. In addition, the study conducted three tasks: a Grammaticality Judgment Task, an Acceptability Judgment Task, and a Multiple Choice Question Task to examine the subjects’ interpretation of Mandarin clefts and pseudoclefts. Thirty Japanese students were recruited and further divided into two groups, i.e., 15 low-intermediate and 15 high-intermediate subjects, based on their Mandarin language proficiency. Moreover, fifteen Mandarin native speakers participated in the study as the control group. The overall results indicated that the subjects’ performances were affected by the grammaticality of clefts and pseudo-clefts and that no construction effects were found. As for cleft types, the subjects highly accepted the NP, VP and PP types, whereas their acceptance rates for the ADJP and ADVP clefts were the lowest of all, indicating that cleft effects were obvious, though construction effects were not. In addition, it was found that our subjects strongly associated certain cleft sentences with certain contexts. The contextual effects and construction effects were attested in this study. Finally, it was found that the Japanese learners’ cleft performances were affected by their L2 proficiency. The better their L2 proficiency was, the better they performed on clefts and pseudo-clefts.. Keywords: clefts, pseudo-clefts, grammaticality effects, construction effects, cleft type effects, contextual effects, proficiency effects, second language acquisition ii.

(4) ACKNOWLEDGMENTS I could hardly have imagined that I would finally finish this thesis, which would not have been completed without the long-term assistance of numerous people. It is my pleasure to acknowledge their guidance and assistance during the writing of my thesis and to express my gratitude for their timely support whenever I was in need. First and foremost, I would like to express my deepest appreciation to my thesis advisor, Dr. Chun-yin Doris Chen. I encountered many difficulties while working on my thesis, and Dr. Chen always provided insightful suggestions and guided me to think from various perspectives, which undoubtedly improved my thesis. When I became depressed while writing the thesis, she maintained her belief in me and gave me plenty of confidence. Her warm encouragement, constructive comments, and inspiring suggestions during this time were invaluable. I cannot find words to express my gratitude to her. Furthermore, I am immensely grateful to my considerate committee members, Dr. Rueih-Lirng Sharon Fahn and Dr. Jyun-Gwang Fred Chen. Their helpful suggestions enlightened me as to what direction I should take in the thesis’ revision and improved the quality of my thesis. I deeply appreciate all the thoughtful and insightful comments they offered me. Additionally, special thanks go to Anri Chita, Hirokazu Okuda, and Yurino Inamura, by alphabetical order. Subject recruitment would have been more difficult without the arrangement and kindly assistance of these individuals. I would like to express my sincere gratitude to the teachers of the Department of English who taught me at NTNU: Dr. Gerardo Fernández-Salgueiro, Dr. Hsiao-Hung Iris Wu, Dr. Jen-I Li, Dr. Jen Ting, Dr. Miao-Ling Hsieh, and Dr. Miao-Hsia Tammy Chang, by alphabetical order. Their skillful instruction and precise guidance introduced me to the field of linguistics and their full support encouraged me to finish my thesis. iii.

(5) Moreover, I am fortunate to have my dear friends: Bo-Fan Andrew Syue, Chia-Yin Gina Hung, Pei-Hsuan Connie Chen, Szu-Chieh Jess Huang, Yan-Cheng Shawn Lin, YuChen Eliza Cheng, and Yuko Sakashita, by alphabetical order. I could not have completed my master’s degree without their company and warm encouragement. Additional special thanks go to my good friend, Szu-Chieh Jess Huang, a kindergarten teacher in Pingtung city. During the writing of this thesis, I faced numerous difficulties and was occasionally frustrated. She was always willing to listen to my complaints and worries, and she gave me encouragement and advice whenever I was in need. She also shared in my joys and sorrows during this time and gave me much positive energy with which to overcome the difficulties and pressures I encountered. Finally, and most crucially, I would like to thank my beloved parents and brother for their constant support and encouragement. During the thesis-writing process, they tolerated my mood swings, helped me overcome some tough problems, and always had faith in me. Without their everlasting love and unconditional financial and emotional support, I would not have been able to complete this academic journey. I realize that their endless love and support were the foundation of this thesis and I will not be able to repay them for their devotion. Thus, I would like to dedicate this thesis to my dearest family.. iv.

(6) TABLE OF CONTENTS CHINESE ABSTRACT ................................................................................................................ i ENGLISH ABSTRACT ............................................................................................................... ii ACKNOWLEDGMENTS ............................................................................................................ iii TABLE OF CONTENTS ............................................................................................................. v LIST OF TABLES ................................................................................................................. viii LIST OF FIGURES ................................................................................................................... x LIST OF ABBREVIATIONS ........................................................................................................ xi CHAPTER ONE INTRODUCTION .......................................................................................... 1 1.1 Motivation ............................................................................................................... 1 1.2 Theoretical Framework ........................................................................................... 4 1.2.1 Transfer Effects ................................................................................................ 4 1.2.2 Contrastive Analysis Hypothesis ..................................................................... 5 1.2.3 Error Analysis .................................................................................................. 6 1.2.4 Markedness Differential Hypothesis................................................................ 7 1.2.5 Contextual Effects ............................................................................................ 8 1.2.6 Proficiency Effects ........................................................................................... 9 1.3 Research Questions ................................................................................................. 9 1.4 Significance of the Study ...................................................................................... 10 1.5 Organization of the Thesis .................................................................................... 11 CHAPTER TWO LITERATURE REVIEW .............................................................................. 12 2.1 Syntactic Structures of Mandarin and Japanese Clefts and Pseudo-Clefts ........... 12 2.1.1 Syntactic Structure of Mandarin and Japanese Clefts ................................... 13 2.1.2 Syntactic Structure of Mandarin and Japanese Pseudo-Clefts....................... 17 2.1.3 Comparisons between Mandarin Clefts and Pseudo-Clefts ........................... 21 2.1.4 Summary ........................................................................................................ 23 2.2 Previous Empirical Studies of Mandarin Clefts and Pseudo-Clefts ..................... 23 2.2.1 Xie (2008) .................................................................................................... 24 2.2.2 Irgin (2013) .................................................................................................. 27 2.2.3 Liao (2014) .................................................................................................. 28 2.2.4 Mai & Yuan (2016) ...................................................................................... 30 2.2.5 Summary ...................................................................................................... 36 2.3 Summary ............................................................................................................... 38 v.

(7) CHAPTER THREE. RESEARCH DESIGN ….. ........................................................................ 39. 3.1 Subjects ................................................................................................................. 39 3.2 Methods and Materials.......................................................................................... 40 3.2.1 Study I ............................................................................................................ 41 3.2.2 Study II .......................................................................................................... 43 3.2.2.1 A New Analytical Framework ................................................................. 44 3.2.2.2 The Acceptability Judgment Task ........................................................... 49 3.2.3 Study III ......................................................................................................... 50 3.3 Procedures ........................................................................................................... 53 CHAPTER FOUR. RESULTS AND DISCUSSION ….. ............................................................... 55. 4.1 Grammaticality Effects ....................................................................................... 55 4.1.1 Overall Results ............................................................................................... 56 4.1.2 Reexamination of the Grammaticality Judgment Task .................................. 61 4.1.3 General Discussion ........................................................................................ 63 4.2 Cleft Type Effects ................................................................................................. 65 4.2.1 Overall Results ............................................................................................... 66 4.2.2 General Discussion ........................................................................................ 69 4.3 Contextual Effects ................................................................................................. 71 4.3.1 Overall Results ............................................................................................... 72 4.3.2 General Discussion ........................................................................................ 75 4.4 Proficiency Effects ................................................................................................ 77 4.5 Summary ............................................................................................................... 79 CHAPTER FIVE CONCLUSIONS….. ...................................................................................... 81 5.1 Summary of Major Findings ................................................................................. 81 5.2 Pedagogical Implications ...................................................................................... 83 5.3 Limitations and Suggestions for Future Research ................................................ 84 Bibliography ................................................................................................................... 86 Appendix A The Grammaticality Judgment Task ........................................................ 90 Appendix B The Accessibility Judgment Task ............................................................ 91 Appendix C. The Multiple Choice Question Task ........................................................ 93. Appendix D The Consent Form ................................................................................... 95 vi.

(8) Appendix E The Background Information Sheet ........................................................ 96. vii.

(9) LIST OF TABLES Table 2-1. A Comparison between Mandarin Clefts and Pseudo-Clefts ...................... 21. Table 2-2 Group Choices (out of six) in Multiple Choice Questions .......................... 34 Table 2-3 Meaning Ratings in Acceptability Judgment (Telicity, 1-4 Scale) .............. 34 Table 2-4 Group Ratings in Sentence Ranking (Possible Score Range:0-6) ............... 35 Table 2-5 Major Findings and Limitations of the Empirical Studies .......................... 36 Table 3-1. Summary of the Subjects ............................................................................. 40. Table 3-2 Test Items for the Grammaticality Judgment Task ...................................... 42 Table 3-3 Test Samples of the Grammaticality Judgment Task ................................... 43 Table 3-4 Test Items for the Acceptability Judgment Task .......................................... 49 Table 3-5. A Test Sample of the Acceptability Judgment Task..................................... 50. Table 3-6 Test Items for the Multiple Choice Question Task ...................................... 52 Table 3-7. A Test Sample of the Multiple Choice Question Task ................................. 53. Table 4-1. Subjects’ Overall Performances on the Grammaticality Judgment Task ..... 56. Table 4-2. P-values for the within-Group Differences between the Two Constructions in the Two Sentence Types ................................................................................ 57. Table 4-3. P-values for the Two Constructions............................................................. 59. Table 4-4. Overall Performances for the Two Sentences .............................................. 60. Table 4-5. P-values for the within-Group Differences between the Two Constructions ................................................................................................ 60. Table 4-6. P-values for the Two Constructions............................................................. 60. Table 4-7. Subjects’ Overall Performances on the Five Cleft Types ............................ 66. Table 4-8. P-values for the within-Group Differences among the Five Cleft Types .... 67. Table 4-9. P-values for the Five Cleft Types ................................................................ 68. Table 4-10 Subjects’ Overall Performances on the Two Types of Contexts ................... 72 viii.

(10) Table 4-11 P-values for the within Group Differences in the Two Contexts .................. 73 Table 4-12 Overall Cleft and Pseudo-Cleft Performances in the Two Contexts ................................................................................................ 74. ix.

(11) LIST OF FIGURES Figure 4-1. Overall Performance of Each Group for the Grammatical Sentence Type ............................................................................................. 57. Figure 4-2. Overall Performance of Each Group for the Ungrammatical Sentence Type ............................................................................................. 58. Figure 4-3. Developmental Stages of Mandarin Clefts and Psuedo-Clefts .................. 78. x.

(12) LIST OF ABBREIVIATIONS ACC. accusative case. ADJP. adjectival phrase. ADVP. adverbial phrase. ASP. aspectual marker. CL. classifier. MOD. modifier. NP. noun phrase. PP. prepositional phrase. PTC. particle. POSS. possessive marker. PERF. perfective marker. TOP. topic. 1SG. first person singular. 3SG. third person singular. VP. verb phrase xi.

(13) CHAPTER ONE INTRODUCTION. 1.1 Motivation Cleft sentences are found in various languages including Chinese, Japanese, and English. In these languages, cleft sentences display distinct clefting strategies and are used in different manners. Clefts are noteworthy for several reasons. From a syntactic perspective, a cleft sentence is a complex sentence including a main clause and subordinate clause (Irgin 2013). In addition, clefts vary cross-linguistically in their syntactic structures (Lobo et al., 2015). Regarding the syntax-discourse interface, the syntactic structure of a cleft sentence is strongly associated with a focus reading, and distinct cleft types or structures have certain semantic-pragmatic constraints (Lobo et al., 2015). Therefore, clefts are considered challenging structures regarding the theory of grammar, syntactic acquisition, and syntax-pragmatics interface. Clefts and pseudo-clefts have been discussed together because the two constructions share some similarities. Lambecht (1995) stated that clefts and pseudo-clefts are considered pragmatically identical, and Tang (1980) claimed that the two constructions have the same meaning, proposition, and focus. Accordingly, determining the differences between the two constructions is generally considered difficult. Hence, the aim of the present study was to 1.

(14) consider both constructions and further investigate their differences from the perspective of second language (L2) acquisition. Previous studies have identified several difficulties of cleft sentences posed for L2 learners (Han 2004, Soares 2006, Irgin 2013). The structural complexity of clefts is one of the reasons why L2 learners find clefts challenging. Mandarin Chinese includes two types of cleft sentences: the shi…de cleft (1a) and the shi cleft (1b). (1) a. Hua shi wo mai de. flower SHI I buy DE ‘It is I who bought the flowers.’ b. Ta shi hen piaoliang (de), danshi wo bu xihuan she SHI very beautiful (DE), but I not like ‘She is very beautiful, but I do not like her.’ c. Shi wo mai de SHI I. ta. her. hua.. buy DE flower. These two sentences reveal that de is not always necessary in a cleft sentence. In Mandarin clefts, there are two types of de. De is required and it can function as a relative clause marker, as in sentence (1a), where a sentence that can be reconstructed into (1c). However, when de functions as a focus marker as in (1b), the sentence cannot be reconstructed into (1c), where de is optional. According to our observations, the structure of Mandarin clefts is similar to that of pseudo-clefts:. 2.

(15) (2) a. Huaping shi vase. SHI. LI XIAOJIE xihuan. de.. Li Miss. DE. like. ‘It is Miss Li that likes the vase.’. b. Ta xihuan he like. de. shi LI XIAOJIE.. DE SHI Li Miss. ‘Who he likes is Miss Li.’. (Li 1980, p.78). In (2a), a cleft sentence in Mandarin Chinese marks the focal part (i.e. Lixiaojie; “Miss. Li”) between the functional markers shi and de. In the pseudo-cleft sentence (2b), the focal part (i.e. Lixiaojie; “Miss. Li”) is placed at the end of the sentence. The examples in (2) show that the surface structures of Mandarin cleft and pseudo-cleft sentences are similar. Both sentences include shi and de. In addition, an NP focus 1 element is incorporated in the cleft and pseudo-cleft sentences. We assumed that these similarities may confuse L2 learners of Mandarin Chinese because they include the same elements in a different word order. Although structural similarities may affect L2 learners’ Mandarin cleft performance, Liao (2014) showed that their cleft performances are influenced by different cleft types that have distinct phrases in the focal part. In addition, the findings of Mai and Yuan (2016) indicated that L2 Mandarin Chinese learners had difficulties in using Mandarin declaratives and clefts in appropriate contexts. This suggests that L2 learners’ 1. In Mandarin Chinese, there are cleft types other than the NP type, such as the VP, PP, ADJP, and ADVP types. 3.

(16) performances are not merely influenced by structural similarities, but rather the meanings and functions of sentences are also factors affecting their language performance. To examine whether the possible factors ( i.e. syntax and pragmatics) actually hinder the L2 learners’ acquisition of Mandarin clefts and pseudo-clefts, we conducted empirical studies on L2 Mandarin cleft acquisition among Japanese-speaking learners of Mandarin Chinese. It is hoped that the findings of the study will be able to assist L2 learners of Mandarin Chinese with acquiring cleft and pseudo-cleft sentences.. 1.2 Theoretical Framework 1.2.1 Transfer Effects L2 learners’ native language has been shown to be a factor that affects their L2 learning. Studies from various perspectives such as discourse analysis, L2 writing and reading as well as second language acquisition studies, have shown that first language (L1) affects the L2 learners’ language production (Sasaki 1991, Kubota 1998, Chan 2004). Kubota (1998) investigated L1 and L2 transfer phenomena among the English writing essays of Japanese university students. The findings initially showed that more than half of the students used similar L1 patterns in their L2 production, revealing that the L2 learners’ L1 did influence their L2 learning. Moreover, the results identified positive rather than negative transfer in L2 learners’ production. Furthermore, Sasaki (1991) 4.

(17) studied a competition model from a psychological perspective and the results demonstrated a degree of L1 transfer. Yang (2014) also indicated that learners acquire inflection during the early stages of language learning through either L1 transfer or Universal Grammar, and it supports the Full Access/Full Transfer model proposed by Schwartz and Sprouse (1996). According to these studies, L2 learners’ native language influences their L2 learning. However, whether L1 effects are positive or negative remains unclear and requires further investigation.. 1.2.2 Contrastive Analysis Hypothesis The contrastive analysis hypothesis stated that errors made by L2 learners result from L1 interference because of the typological variation between the two language systems. The hypothesis further claimed that the greater the difference between the two language systems is, the greater the learning problem and potential area of interference are (Weinreich 1953). Word order and omission are two types of the errors commonly found in L2 learners’ language production that are caused by L1 interference. Regarding word-order effects, Xie (2008) mentioned that word order is a factor influencing Japanese learners’ cleft acquisition as in (3).. 5.

(18) (3) a. Zhe. xie. hua. shi Mali xihuan de.. this CL flower SHI Mary like ‘It is Mary that likes these flowers.’. DE. b. Hana-o suki na hito-wa Mari flowers-ACC like POSS person-TOP Mary. da. DA. The word order of the Mandarin Chinese sentence in (3a) is an object, a focus marker shi, a subject, and then a verb. However, the word order of the Japanese sentence in (3b) is a subject, an object, and then a copula da as a focus marker. Because of the typological differences between Mandarin Chinese and Japanese, Japanese learners of Chinese may make interlingual word-order errors in Mandarin clefts. We also examined the omission of shi or de. As presented in (3a), in Mandarin clefts, de often appears after shi. The Japanese sentence in (3b) includes a focus marker (i.e. da), but no Japanese elements corresponding to the Mandarin Chinese de in (3a) are present. Therefore, the appearance of de in Mandarin cleft and pseudo-cleft sentences may cause confusion to JSLs and result in their interlingual omission error.. 1.2.3 Error Analysis Error Analysis Hypothesis stated that L2 learners commit developmental errors, which are not due to L1 interference. During the process of acquiring a difficult language, L2 learners hypothesize, test, modify, and generalize the rules of the target language (TL) so that their utterances may conform to the TL (Larsen-Freeman & Long 1991). Such 6.

(19) processes suggest that errors made by L2 learners result from incorrect hypotheses regarding certain TL rules, rather than from L1 interference. If Japanese-speaking learners have not fully acquired the rules of Mandarin clefts and pseudo-clefts, their errors may result from the similarities between Mandarin clefts and pseudo-clefts, and such errors are called intralingual errors (Richards 1971). Since previous studies hardly examined these errors, one of the main purposes of the present study is to testify this hypothesis through the observation of Japanese-speaking learners’ cleft performances.. 1.2.4 Markedness Differential Hypothesis The influence of markedness has been supported by numerous researchers (Eckman 1977, Gass 1988, Gundel & Tarone 1983, Liao 2014, Lobo et al., 2015, Zobl 1983). According to Eckman (1977), the markedness differential hypothesis proposes that L2 learners face difficulty when acquiring the areas of the L2 that differ from their language. He further predicted the existence of a difficulty hierarchy in L2 learners’ language performances. The relative difficulty hierarchy reflects the marked and unmarked areas for L2 learners. Lobo et al. (2015) conducted an empirical study on the acquisition of Portuguese cleft construction. The difficulty hierarchy revealed that subject clefts appeared to be the easiest 7.

(20) cleft type for Portuguese children, whereas object clefts was found to be the most difficult. Mizumoto (2011) and Dansako and Mizumoto (2007) have also demonstrated that object clefts are more challenging than subject clefts. According to the findings of these studies, children encounter some difficulties in acquiring different cleft types. That is, some cleft types might be acquired earlier and others might be acquired later. Hence, some clefts are found to be more marked when compared with other cleft types. In the current study, the markedness. of different Mandarin cleft types was clearly displayed through the difficulty hierarchy found in the Mandarin cleft performance levels of Japanese-speaking learners.. 1.2.5 Contextual Effects Among the issues of the functional usages of cleft sentences, contextual effects have been widely discussed in previous studies (Declerck 1984, Mai & Yuan 2016, Prince 1978). Mai and Yuan (2016) revealed that English-speaking learners of Mandarin Chinese could not use Mandarin clefts in the appropriate context, thus evidencing a clear contextual effect. Moreover, Liao (2014) demonstrated that the contextual effect may be a factor influencing L2 learners’ performance. To examine this issue, the present study investigated whether Japanese-speaking learners can accurately use Mandarin clefts and pseudo-clefts in appropriate contexts.. 8.

(21) 1.2.6 Proficiency Effects L2 proficiency is a substantial issue for diverse researchers (Carrell 1991, Chan 2004, Lee & Schallert 1997, Liao 2014, Mai & Yuan 2016, Xie 2008). Carrell (1991) reported a study of L1 and L2 reading comprehension, and indicated that L2 proficiency was an important factor in L2 acquisition. Lee and Schallert (1997) conducted a study on L2 reading, and demonstrated that L2 proficiency played a substantially more critical role than L1 reading abilities. Learners with lower L2 proficiency were found to encounter more reading obstacles, whereas learners with higher language proficiency demonstrated positive language reading abilities. Chan (2004) revealed that when lower-level learners faced difficult sentence structures, they tended to rely more on their L1 than advanced learners. These studies have all indicated that L2 learners’ language proficiency affects their language learning. In the current study, this issue has been examined through the performances of two distinctive groups of Japanese learners of Chinese: the lowerintermediate (LI) and higher-intermediate (HI) groups.. 1.3 Research Questions Based on the above, four research questions were addressed to investigate the factors influencing the Japanese-speaking learners’ Mandarin cleft performance:. 9.

(22) 1) Does the grammaticality of clefts and pseudo-clefts affect L2 learners’ performances? 2) Do Japanese-speaking learners’ perform differently according to various Mandarin cleft types? If so, how do these cleft sentences subtypes affect the learners’ performance? 3) Do contextual differences influence Japanese-speaking learner’s preference for certain cleft construction? 4) Do the overall performances of low-intermediate (LI) and high-intermediate (HI) Japanese-speaking learners appear different?. 1.4 Significance of the Study The theoretical issues concerning Mandarin cleft sentences have been widely discussed by numerous researchers (Cheng 2008, Hsieh 2002, Huang 1982, Li 1980, Paris 1979, Tang 1980, Teng 1979, Tsao 1994, Simpson & Wu 2002). Although previous researchers have proposed distinct analyses for the syntactic structures of Mandarin clefts, few have discussed the acquisition of such constructions by L2 learners of Mandarin Chinese. In addition, despite the existence of studies addressing the issues of L2 learners’ Mandarin cleft acquisition (e.g. Mai & Yuan 2016, Xie 2008), only one type of Mandarin cleft construction has been investigated. Other Mandarin cleft constructions such as pseudo-clefts have not been further mentioned or discussed. Moreover, some theoretical 10.

(23) studies have touched upon the pragmatic usages of Mandarin clefts, but the interpretation and use of Mandarin clefts and pseudo-clefts among L2 learners have not been further explored. Thus, one aim of the present study was to present a more thorough picture of Japanese learners’ comprehension of Mandarin clefts and pseudo-clefts, by exploring four critical issues. This not only can address the gap found in the literature, but also present a clearer picture of L2 learners’ acquisition of Mandarin clefts and pseudo-clefts.. 1.5 Organization of the Thesis This thesis is organized as follows: In Chapter Two, theoretical issues regarding the cleft and pseudo-cleft sentences of two languages are reviewed. In addition, empirical studies of cleft sentences are shown, which demonstrate the issues of L2 learners’ cleft, and pseudo-cleft acquisition. Chapter Three presents our research design. Results, discussions, and further implications of the study are illustrated in Chapter Four. Finally, to summarize the major findings of this study and offer suggestions for future research, a brief conclusion is provided in Chapter Five.. 11.

(24) CHAPTER TWO LITERATURE REVIEW. In this chapter, some theoretical and empirical studies are reviewed. Section 2.1 presents a comparison of the structures of Mandarin and Japanese clefts and pseudo-clefts. Next, in Section 2.2, empirical studies of cleft sentences are discussed. Finally, in Section 2.3, a brief summary of this chapter is provided.. 2.1 Syntactic Structures of Mandarin and Japanese Clefts and Pseudo-Clefts In this section, the syntactic structures of Mandarin and Japanese clefts and pseudoclefts are discussed. In Section 2.1.1, the structures of clefts in Mandarin and Japanese are introduced and compared, and in Section 2.1.2, the contrasts between Mandarin and Japanese pseudo-clefts are examined. In Section 2.1.3, Mandarin clefts and pseudo-clefts are compared. Finally, Section 2.1.4 summarizes the discussed issues and provides predictions about Japanese-speaking learners’ performances on Mandarin clefts and pseudo-clefts.. 12.

(25) 2.1.1 Syntactic Structures of Mandarin and Japanese Clefts In Mandarin Chinese, a cleft sentence is a syntactic construction that marks the focal constituent between the Mandarin Chinese copula shi and the optional particle de as follows:. (1) Cleft Configuration: Subject + shi + cleaved XP + (de) (Lee 2005, pp.3-5). According to Li (1980) and Tang (1980), the structure of a Mandarin cleft is derived from two transformational rules (or movement stages): copula lowering and topicalization.. (2) a. Cleft Sentence: Ta shi he SHI. zai jiujinshan in. yujian Zhangsan de.. San Francisco meet. Zhangsan DE. ‘It was in San Francisco that he met Zhangsan.’ Deep Structure: [Ta][zai jiujinshan yujian Zhangsan][shi][de] Transformations: Stage 1: b. Copula Lowering:. Stage 2: Topicalization: [Ta] [shi][…][zai jiujinshan yujian Zhangsan][de] (Li 1980, pp.77-78 and Tang 1980, pp.184-185). Another syntactic analysis proposed by Cheng (2008, p.19) reveals that a Mandarin cleft is derived from a small clause. 13.

(26) (3) a. Shi [[SUBJ Xiaowang] [PRED zuotian wanshang lai de.]] b.[SUBJ Xiaowang] shi [[….. ][PRED zuotian wanshang Xiaowang SHI yesterday night ‘Xiaowang came last night.’. lai de]] come DE. Example (3a) displays a sentence configuration with a copula shi and a small clause; in (3b), the subject of the small clause is raised across the copula to the subject position to form a Mandarin cleft sentence (Cheng 2008). According to the analyses proposed by Li (1980), Tang (1980), and Cheng (2008), we can observe that the shi and de elements play a crucial role in a Mandarin cleft sentence. They appear to serve distinct functions and are obligatory in some Mandarin cleft sentences. According to Teng (1979), shi in the Mandarin cleft is viewed as a copula that is a focus marker. But, Tang (1980) considered shi as a main verb. Huang (1982) claimed that shi is an adverb in a Mandarin cleft sentence. From these viewpoints, we can determine that although the function of shi in Mandarin clefts is still in debate, shi is an undeniably crucial element and is required in Mandarin clefts. Tang (1980) and Lee (2005) have analyzed de as a final particle and stated that it may be optional in cleft sentences. Observing the cleft types from the pragmatic perspectives reveals that NPs, verb phrases (VPs), prepositional phrases (PPs), adjective phrases (ADJPs), and adverb phrases (ADVPs) can all be in the focal part of Mandarin clefts. In addition, Mandarin clefts have a focus reading and hold two main pragmatic functions. First, they convey 14.

(27) strong assertion from the previous topic (Liu 1996, Hsieh 2002, Lü 2002). Second, they can denote past events (Paul & Whitman 2008, Tang 1980). The preceding arguments and observations present the basic structure of Mandarin cleft sentences. Nevertheless, some problems remain to be solved. For example, the syntactic analyses of Mandarin clefts cannot clearly account for the following sentence (Cheng 2008, p.26):. (4) Zhangsan shi. zuotian. kandao. Zhangsan SHI yesterday see. Wang xiaojie (bu shi qiantian). Wang Ms.. not SHI the day before. ‘It is yesterday that Zhangsan saw Ms. Wang, not the day before yesterday.’. In (4), de in a Mandarin cleft sentence can be omitted; however, the analyses performed by Li (1980), Tang (1980), and Cheng (2008) did not provide any explanation for this type of Mandarin cleft sentence. Therefore, one more transformation such as de deletion may be required in the analyses of Li (1980) and Tang (1980). In addition, these studies did not explain why de must be deleted in some Mandarin cleft sentences. Hence, more syntactic investigation of Mandarin clefts are required. Furthermore, although previous researchers have presented some pragmatic functions of Mandarin clefts, they did not provide supportive evidence regarding the use of such constructions under specific contexts. Thus, the pragmatic functions outlined by previous researchers require further justification. 15.

(28) Japanese cleft sentences consist of an NP followed by a topic marker wa, as shown in (5). The cleaved part follows the topic marker, and the copula da is situated at the final position (Morikawa 2011).. (5) Cleft Configuration: NP + wa + cleaved XP + da (Morikawa 2011, p.361) (6) a. Akahoshi-wa Akahoshi-TOP. kyaputan da. captain. DA. ‘Akahoshi is the captain.’ b. Ano CL. toki no. yuuhi-wa. kirei. dat-ta.. time POSS. sunset-TOP. beautiful DA-PAST. ‘The sunset at that time is beautiful.’ (Fukuda 2010, p.16). The configurations in (1) and (5) reveal that the syntactic configurations of Mandarin and Japanese clefts both include a focus NP. In addition, a focus marker can be found in both Mandarin and Japanese clefts. However, the focus marker shi in Mandarin follows the NP, whereas the copula da in Japanese is placed at the end of the sentence. This demonstrates that the word orders of Mandarin and Japanese clefts are different. Furthermore, a particle such as de is optional in a Mandarin cleft,, whereas this element does not exist in a Japanese cleft sentence. By observing the cleft types, Lee (2005) have stated that NP, VP, and PP clefts can be found in Japanese clefts. These three cleft types can also be found in Mandarin clefts. 1. Some modification of the configuration was made in this study to show the consistency of the syntactic configurations. 16.

(29) However, no studies have identified ADJPs and ADVPs in Japanese clefts. This is slightly different from the cleft types of Mandarin clefts. Concerning the pragmatic usages of Japanese clefts, Kato (2000) stated that Japanese clefts could be used in a topic continuity context, implying that Japanese clefts have the function of thematic continuity. Mandarin clefts also have such function.. 2.1.2 Syntactic Structures of Mandarin and Japanese Pseudo-Clefts A Mandarin pseudo-cleft is a sentence that marks the focal constituent after de shi. De is optional under some circumstances as with Mandarin clefts:. (7) Pseudo-Cleft Configuration: Complementizer phrase (CP) + (de) + (Noun) + shi + cleaved XP. Li (1980) proposed that three transformational (or movement) stages, described in (8), are involved in the derivation of Mandarin pseudo-clefts, whereas Tang (1980) introduced a slightly different analysis of pseudo-clefts as in (9).. (8) Pseudo-Cleft Sentence: Ta xihuan de he like. DE. shi lixiaojie. SHI Li Miss. ‘The one he likes is Miss Li.’ Deep Structure: [ren][Ta xihuan ren][shi][lixiaojie] Transformations: Stage 1: Deletion (ren) and Insertion (de): [ren][Ta xihuan ren de][shi][lixiaojie] Stage 2: Relative clause proposition: [Ta xihuan de][ren][…][shi][lixiaojie] Stage 3: Head noun deletion: [Ta xihuan de][shi][lixiaojie] (Li 1980, p.80) 17.

(30) (9) Pseudo-Cleft Sentence: Wo dapo de. shi ni. de. I break DE SHI you POSS. huaping. vase. ‘What I broke is your vase.’ Original Form:[. ][Wo dapo. ][shi][ni de huaping]. Transformations: Stage 1: Relativity transformation: [Wo dapo de ][shi][ni de huaping] Stage 2: Head deletion: [Wo dapo de][shi][ni de huaping] (Tang 1980, pp.194-195). Cheng (2008) also offered an analysis of Mandarin pseudo-clefts, revealing Mandarin clefts and pseudo-clefts are derived from the same basic structure (i.e. shi with a small clause). However, the difference between the derivation of a cleft and a pseudo-cleft is that a pseudo-cleft is derived through a predicate movement, whereas a cleft is not (Cheng 2008, p.19).. (10) a. Shi [[SUBJ Xiaowang][PRED zuotian wanshang lai de.]] b.[PRED zuotian wanshang yesterday night. lai de] shi [[SUBJ come DE SHI. Xiaowang][….. .]] Xiaowang. ‘The person who came last night was Xiaowang.’. Sentence (10a) is formed using a copula shi and a small clause, and (10b) shows that the predicate in the small clause is raised to the precopula position to form the surface structure of a Mandarin pseudo-cleft sentence. According to this analysis, the de and shi elements are crucial in Mandarin pseudoclefts. According to Li (1980), Tang (1980), and Lee (2005), shi in Mandarin pseudoclefts is a copula that functions as a focus marker, and additional studies have indicated 18.

(31) that de is a relative clause marker (Lee 2005, Li 1980, Tang 1980). Without the two elements, the sentences may result in ungrammaticality. However, according to Cheng’s (2008) analyses, a pseudo-cleft sentence may be wrongly derived as in (11) because the predicate of this pseudo-cleft cannot be moved:. (11) *Zai tushuguan shi wo jie shu. in library SHI I borrow book ‘?Where I borrow books is in the library.’. In addition, the preceding analyses (Cheng 2008, Li 1980, Tang 1980) cannot account for Mandarin pseudo-cleft sentences that omit the de element, such as (12).. (12) Wo jie shu shi zai tushuguan. I borrow book SHI in library ?‘Where I borrow books is in the library.’. In this respect, more syntactic analyses regarding Mandarin pseudo-clefts are necessary. Previous studies have rarely adopted the perspectives of the cleft types and the pragmatic functions to analyze Mandarin pseudo-clefts. According to Hsieh (2002), Mandarin pseudo-clefts are sentences that separate old and new information. The former part, which is before de, presents old information, whereas the latter part, which is after shi, introduces new information. Hence, the structure of Mandarin pseudo-clefts can identify previously mentioned and new information. Moreover, although previous studies have indicated that an NP cleft type exists in Mandarin pseudo-clefts (Lee 2005, Li 1980, 19.

(32) Tang 1980,), other pseudo-cleft types were neither mentioned nor examined. This thus necessitates the further classification of Mandarin pseudo-clefts. Regarding the pragmatics, Hsieh (2002) merely indicated that Mandarin pseudo-clefts are used to emphasize certain things, but a statement like that is considered unclear. The usages and functions of Mandarin pseudo-clefts should be clearly examined under different contexts to reveal their pragmatic usages and functions. Hence, the pragmatic functions of Mandarin pseudo-clefts are still unclear and require further investigation. In Japanese pseudo-clefts, a CP clause is required at the beginning of a sentence (Ishihara 2012) along with the topic marker wa; the cleaved part is placed before the copula da. The syntactic configuration for Japanese pseudo-clefts is given in (13) and (14).. (13) Pseudo-Cleft Configuration: [CP [C no]] + wa + cleaved XP + da (Lee 2005, p.2362). (14) a. Kirei-datta. no-wa. ano. beautiful-PAST NOM-TOP CL. toki no. yuuhi. da.. time POSS sunset DA. ‘What is beautiful is the sunset at that time.’ b. Totemo very. kirei. no-wa. Ano gurasu da .. beautiful NOM-TOP CL. glass DA. ‘What is beautiful is that glass.’. 2. Some modifications of the configuration were made in this study to show the consistency of the syntactic configurations. 20.

(33) Comparing the sentences in (7) and (13) reveals that a CP clause is placed in the sentence-initial position for both Mandarin and Japanese pseudo-clefts. A copula can also be found in both languages. However, a major difference between Mandarin and Japanese pseudo-clefts is that the cleaved part of a Mandarin pseudo-cleft appears after the copula shi, whereas the cleaved part of a Japanese pseudo-cleft appears before the copula da. This reveals that the word orders of Mandarin and Japanese pseudo-clefts are different. The particle de is present in Mandarin pseudo-clefts, but there is no such particle in Japanese pseudo-clefts. By observing the cleft types in Japanese pseudo-clefts, Lee (2005) and Ishihara (2012) have also claimed that NP, VP, and PP cleft types can be found in Japanese pseudo-clefts. In Mandarin pseudo-clefts, these three cleft types are common. However, according to our review of the relevant literature, no study has claimed that ADJPs and ADVPs are acceptable in Japanese pseudo-clefts. This is rather different from the cleft types of Mandarin pseudo-clefts.. 2.1.3 Comparisons between Mandarin Clefts and Pseudo-Clefts The syntactic configurations of Mandarin clefts and pseudo-clefts are presented in Table 2-1.. Table 2-1 A Comparison between Mandarin Clefts and Pseudo-Clefts Cleft. Pseudo-Cleft. NP + SHI + cleaved XP + (DE). CP + (DE) + SHI + cleaved XP. 21.

(34) Table 2-1 shows that Mandarin clefts and pseudo-clefts are structured differently. A Mandarin cleft sentence marks the focal constituent between shi and de, whereas the focal constituent in a pseudo-cleft sentence falls after de shi. According to Cheng (2008) and Tang (1980), Mandarin clefts and pseudo-clefts pass through the same number of transformational stages. This may facilitate predicting that clefts and pseudo-clefts have the same structural complexity. Nevertheless, Li (1980) proposed that Mandarin clefts undergo two transformational stages, whereas pseudo-clefts pass through three stages. Evidently, the claims presented by Li (1980) on the syntactic relatedness of Mandarin clefts and pseudo-clefts differ from those of Tang (1980) and Cheng (2008). Therefore, to determine whether Mandarin clefts and pseudo-clefts are syntactically related, the current study examined Japanese-speaking learners’ performances on Mandarin clefts and pseudo-clefts. If these two cleft constructions involve the same syntactic complexities, JSLs should perform equally well on both constructions. However, if the subjects perform poorly on the two cleft constructions, the syntactic structures, cleft types, and the functions of Mandarin clefts and pseudo-clefts are highly likely to cause confusion for JSLs.. 22.

(35) 2.1.4 Summary This section discusses the syntactic structures, cleft types, and pragmatic usages of Mandarin and Japanese clefts and pseudo-clefts. The previous sections show that the clefts and pseudo-clefts in the two languages exhibit some similarities and differences in all three aspects. Accordingly, Japanese learners of Chinese will have difficulty acquiring Mandarin clefts because the word order of Japanese clefts differs from that of Mandarin clefts. In addition, the NP, VP, and PP cleft types can be found in Japanese and Mandarin Chinese clefts and pseudo-clefts, whereas the ADJP and ADVP cannot. Therefore, another prediction was made that Japanese-speaking learners may easily accept the NP, VP, and PP cleft types but not the ADJP and ADVP ones. The cleft type effects may be obvious. Regarding the functions of Mandarin clefts, Japanese-speaking learners may be able to acquire the thematic continuity function because Japanese clefts also have such function.. 2.2 Previous Empirical Studies of Mandarin Clefts and Pseudo-Clefts This section reviews studies on the L2 acquisition of cleft constructions in Mandarin Chinese. However, to the best of our knowledge, no empirical research has been conducted on Mandarin cleft and pseudo-cleft acquisition. In addition, few studies have focused on Japanese learners’ cleft acquisition. Thus, in this section, four studies of 23.

(36) L2 learners’ acquisition of English and Mandarin clefts are chosen for review, namely the studies by Xie (2008), Irgin (2013), Liao (2014), and Mai and Yuan (2016).. 2.2.1 Xie (2008) Xie (2008) investigated L2 learners’ acquisition of Mandarin Chinese clefts. He initially presented a classification of Mandarin clefts3 and examined the production data from an interlanguage corpus, and it included cleft sentences produced by beginners, intermediate, and advanced learners of Mandarin Chinese. The results revealed that beginners produced few cleft sentences. The cleft constructions that they produced were F1 (ADVP cleft type) and F2 (NP cleft type). Moreover, the production data of L2 learners showed that beginners often confused de with the Mandarin Chinese sentence particle le because both element can occur near the sentence-final position as in example (15) (Xie 2008, p.15).. (15) a.*Ni you. shi sheme shihou. lai le. Zhongguo?. SHI when. come-ASP China. ‘When did you come to China?’. 3. The classification of Mandarin normal clefts from Xie (2008) is as follows: F1: NP + SHR + ADV + V + DE + O F2: NP + SHR + NP + VP + DE F3: SHR + NP + VP + O + DE F4: NP + SHR + VP + (O + VP) + DE F5: NP + SHR + Phrasal V + DE F6: NP + SHR + (Modal V) + V + DE F7: NP + SHR + V + Potential Complement + DE F8: NP + SHR + ADJP + DE 24.

(37) b. Ni. shi. sheme shihou. you SHI when. lai. Zhongguo de?. come China. DE. This shows that beginners’ performances were influenced by their L2 knowledge. Moving onto the intermediate level, the learners’ usage of F5 and F6, which were both VP clefts, exhibited increases. However, they did not demonstrate a high usage rate of F7 use (clausal cleft type) because this type was more complicated for them. Xie also determined that intermediate learners often made the following type of error (Xie 2008, p.18).. (16) a. *Zhiyou yi jian qiguai de shi, shi women zai natian yudao only one CL weird POSS thing, SHI we on that day meet tongxue de. classmates DE ‘One weird thing is that we only meet our classmates on that day.’ b. Women zhiyou zai natian yudao tongxue, zheshi yi jian qiguai we only on that day meet classmates, this one CL weird de shi. POSS thing. In (16), L2 learners of Mandarin Chinese overgeneralized the rule of a Mandarin cleft and used it inaccurately. Such results reveal that although intermediate learners could produce more cleft sentences, they still had not fully acquired the syntactic rules of Mandarin cleft construction. The advanced learners could produce more types of cleft sentences than the beginners and intermediate learners. Nevertheless, they still struggled with the pragmatic usages of cleft sentences. Xie (2008) further observed learners’ correctness of each cleft type. The findings indicated that the markedness effect was a 25.

(38) crucial factor in L2 learners’ language acquisition. The highly unmarked construction was acquired first, and the marked ones were acquired later. To obtain more support for such an observation, Xie (2008) further conducted an empirical study by recruiting 71 foreigners and 30 native Mandarin speakers to complete a Mandarin test survey questionnaire, including a production task, a grammaticality judgment (GJ) task, and a sentence scrambling task. The results indicated a substantial markedness effect, which accorded with his findings from the interlanguage corpus. In summary, Xie (2008) presented a new classification of Mandarin cleft sentences and used two research methods: corpus research and an empirical investigation. L2 learners were found to face different difficulties when acquiring Mandarin clefts. The beginners’ performances were affected by L2 complexity, intermediate learners overgeneralized the rules of Mandarin clefts, and advanced learners still faced challenges regarding the pragmatic usages of Mandarin clefts. Observing the correctness among the eight cleft types revealed that the degree of markedness among different cleft types was a crucial factor affecting L2 learners’ language performance levels and development. However, the study also included some limitations. First, the participants’ backgrounds were not clearly presented. This may affect the results of the study and how the production data are examined. Second, the L1 interference effect was not clearly examined. Third, the study stated that L2 learners encountered pragmatic problems but Xie (2008) did not 26.

(39) discuss these problems. Therefore, more studies are necessary to show more cogent support regarding the issue in the L2 acquisition of Mandarin clefts.. 2.2.2 Irgin (2013) Irgin (2013) discussed the L2 acquisition of seven main types of English clefts in 61 participants. Among the participants, 44 were females and 17 were males. The results demonstrated that the participants’ pretest and posttest performances were significantly different. The minimum score of the pretest was 2.00, whereas the score of the posttest was 22.00. Moreover, the standard deviation scores of the pretest and posttest were 13.74 and 6.09, respectively. These outcomes revealed that English cleft sentences were difficult for L2 learners without explicit instructions. Irgin (2013) also indicated that syntactic differences affected L2 learners’ cleft acquisition. The findings demonstrated that cleft structures such as inferential clefts4 and it-clefts were easy for L2 learners to acquire, whereas other cleft structures such as reversed-wh clefts and pseudo-clefts were challenging for them to acquire even with explicit instruction. Accordingly, the sentence complexity effect was a factor influencing L2 learners’ performances. 4. According to Irgin (2013), the example of an inferential cleft is as follows (Irgin 2013, p.74): (i) It is not that he loves her. It's just that he has a way with her that is different. As shown in (i), an inferential cleft is formed by two cleft sentences. 27.

(40) Finally, the gender effect was examined. The p-values for different genders in the pretest and posttest did not differ significantly (p > .05), signifying that gender was not a factor affecting L2 learners’ cleft acquisition. To summarize, Irgin (2013) clearly reported that syntactic difference was a factor influencing L2 learners’ language learning, but that gender was not. Nonetheless, this study included some limitations that require further clarification. First, the study merely presented a rough finding of the cleft structure effects without further elaboration. Whether structural complexity is a factor affecting learners’ L2 acquisition remains unclear. Therefore, more research is required to gain more evidence for L2 cleft acquisition.. 2.2.3 Liao (2014) Liao (2014) discussed Taiwanese students’ acquisition of English it-clefts by observing the issues of L1 transfer, difficulty hierarchy of English it-clefts, and contrastive effects. Sixty freshmen were recruited and divided into three groups according to their English language proficiency. In addition, 20 native English speakers participated in this experiment as a control group for further observation. An acceptability judgment (AJ) task was performed. 28.

(41) The L1 transfer was found to affect learners’ L2 acquisition. The study revealed no positive transfer, but only negative transfer, in L2 learners’ performances and showed that L2 learners’ performances were affected by their L1 knowledge. Thus, the transfer effect was clear. The second crucial issue was the difficulty hierarchy of different types of English it-clefts. The results showed that L2 learners’ performances on the seven cleft types were significantly different (p = 0.000). The difficulty hierarchy of the seven cleft types is presented in (17) (Liao 2014, p.65).. (17) NP/Zero/PP > N-F CL/ADV P > FIN CL > ADJ P. Regarding the performances of each proficiency group, only the high-proficiency group exhibited near native-like performance. The low- and intermediate-proficiency groups did not show such an inclination. Liao (2014) further explained that English itcleft sentences are marked sentences for low and intermediate learners, thus making it difficult for L2 learners to demonstrate native-like performances. Liao’s findings showed that L2 proficiency was a factor influencing L2 learners’ performances on English cleft acquisition. The contrastive effects were also examined, and the results revealed that cleft sentences with contrasting information provided sentences with more cues and did not 29.

(42) affect L2 learners’ performances. As L2 learners’ language proficiency improved, they no longer needed this contrastive information to learn cleft sentences. Accordingly, the contrastive effects were not obvious in the study. In summary, the study examined four critical issues regarding English cleft acquisition: the transfer effect, cleft type effect, L2 proficiency effect, and contrastive effect. The effects of L2 proficiency and cleft type were clearly exhibited. However, the study did not consider other English clefts such as pseudo-clefts. Furthermore, Liao did not mention or examine whether the pragmatic functions of clefts were factors influencing L2 learners’ cleft acquisition.. 2.2.4 Mai and Yuan (2016) Mai and Yuan (2016) examined how semantic and pragmatic features of Mandarin cleft sentences are acquired by adult English-speaking learners. They first argued that Mandarin cleft sentences include three features ([+past], [+telic], and [+given]). Mai and Yuan claimed that Mandarin clefts have an obligatory past-tense reading, as demonstrated in (18), and that this reading is incompatible with the future-oriented temporal adverbials, as exemplified in (19). Thus, Mandarin clefts are assumed to possess a [past] feature5 (Paul & Whitman 2008):. 5. According to our observations, Mandarin clefts may denote present events, as illustrated in (i): (i)Hua shi Xiaming xihuan de. flower SHI Xiaming like DE 30.

(43) (18) Obligatory past-tense reading Mali shi. cong. Mary SHI via. Xianggang. qu Meiguo. de.. Hong Kong. go America. DE. ‘It was via Hong Kong that Mary went to the US.’ ‘*It is via Hong Kong that Mary (will) go to the US.’ (Mai & Yuan 2016, p.249) (19) Incompatible with future-oriented temporal adverbials *Mali shi. mingtian. Mary SHI tomorrow. qu Meiguo. de.. go America. DE. Intended: ‘It is tomorrow that Mary will go to the US.’ (Mai & Yuan 2016, p.249) Moreover, the VP of a Mandarin cleft must be telic (Cheng 2008, Paul & Whitman 2008). Stative VPs cause unacceptability6, as in (20).. (20) Incompatible with stative predicates *Mali shi Mary SHI. zuotian. hen shengqi de.. yesterday very angry. DE. Intended: ‘It was yesterday that Mary was very angry.’ (Mai & Yuan 2016, p.249). As Mai and Yuan (2016) indicated, the VP event in Mandarin clefts is typically discourse-given information, whereas in Mandarin declarative sentences, the VP event denotes a new change of state. Therefore, in a context where the speaker verbalizes an event and adds information about the time of the same event, a cleft sentence such as (21) is chosen. However, if the speaker wishes to describe two sequential past events, a ‘It is flowers that Xiaoming likes.’ 6. According to our committee members, the VP in a Mandarin cleft can be compatible with a stative predicate, as shown in (i), where ‘hen renzhen’ is stative: (ii) Ta shi hen renzhen. he SHI very hardworking ‘He is very hardworking.’ 31.

(44) declarative sentence such as (22) is preferred. Hence, Mai and Yuan argued that Mandarin clefts have a [given] feature, whereas Mandarin declarative sentences do not have such a feature (Mai & Yuan 2016, p.251).. (21) Same-event context Xiaozhang ba. na. tai diannao. Xiaozhang BA. that CL computer. nonghuai le. break PERF. ‘Xiao Zhang broke the computer.’ Continuation: Ta. shi zuotian. nonghuai de/ ? Ta zuotian. 3SG SHI yesterday break. nonghuai le.. DE. 3SG yesterday break. PERF. ‘It was yesterday that she broke it. She broke it yesterday.’ (22) Different-event context Wo. shangge xingqi. 1SG last. week. song le. Xiaozhang yi. tai. give PERF. Xiaozhang one CL. diannao. computer. ‘I gave Xiao Zhang a computer last week.’ Continuation: ? Ta shi zuotian. nonghuai de/ Ta. 3SG SHI yesterday break. zuotian. nonghuai le.. DE. 3SG yesterday break. PERF. ‘It was yesterday that she broke it. She broke it yesterday.’. On the basis of this observation, Mai and Yuan (2016) investigated whether Englishspeaking learners of Mandarin Chinese had acquired Mandarin cleft features. They also investigated whether these learners map their L1 cleft knowledge to their L2 Mandarin cleft acquisition. In English, [past] and [telic] features are not associated with it-clefts. Thus, if L1–L2 mapping is observed in the performance of English-speaking learners of Mandarin Chinese, the results should demonstrate that L2 learners did not acquire the 32.

(45) [past] and [telic] features. Furthermore, English it-clefts were assumed to be different from Mandarin clefts in that English it-clefts are not triggered by the [given] feature to mark the distinction between discourse-new and discourse-given information; instead, English it-clefts are triggered by the [focus] feature to mark the distinction. If Englishspeaking learners of Mandarin Chinese map the [focus] feature to Mandarin cleft acquisition, they might use Mandarin clefts as English it-clefts. The feature sets of L1 and L2 clefts are shown schematically in (23) (Mai & Yuan 2016, p.255).. (23). To test these predictions, Mai and Yuan (2016) compared data from four groups (LI, HI, advanced, and native). The subjects performed sentence completion (SC), multiple choice question (MCQ), acceptability judgment (AJ), and sentence ranking tasks. The results first indicated that LI learners interpreted the cleft sentences as past-tense sentences only 46% of the time. The percentage was similar to that of the past-tense interpretation of declarative sentences (bare shi sentences). This shows that they did not associate the Mandarin clefts with a fixed past-tense reading. However, as the L2 learners’ proficiency improved, their performances in Mandarin Chinese became more native-like.. 33.

(46) Table 2-2 Group Choices (out of 6) in Multiple Choice Questions. Second, the [telic] feature was examined. The results of the AJ task are listed in Table 2-3 (Mai & Yuan 2016, p.111).. Table 2-3 Mean Ratings in Acceptability Judgment (Telicity, Scale 1–4).. The native Mandarin speakers strongly rejected *[−telic] shi … de. Concerning the English-speaking learners of Mandarin Chinese, the HI and LI learners achieved correctness rates of only 43% and 10%, respectively. The advanced group showed a 71% target-like rate, signifying that [telic] was not typically assembled in Mandarin clefts. The findings for the [given] feature are presented in Table 2-4 (Mai & Yuan 2016, p.265). 34.

(47) Table 2-4 Group Ratings in Sentence Ranking (Scale 0–6).. Table 2-4 reveals that native speakers had a clear preference for Mandarin clefts in the same-event context and a clear preference for declaratives in the different-event context. Nevertheless, such phenomena were not clearly exhibited among the performances of the English-speaking learners of Mandarin Chinese. The two intermediate groups did not show any clear preference for clefts or declarative sentences in either context. The members of the advanced group were aware of the contextual differences and the fact that cleft sentences were more felicitous in the same-event context. On the basis of the results, Mai and Yuan (2016) concluded that English-speaking learners of Mandarin Chinese did not acquire the [past], [telic], and [given] features. These results demonstrate that Englishspeaking learners of Mandarin Chinese mapped their L1 cleft knowledge to Mandarin cleft acquisition. The [past] and [telic] features were not acquired because English itclefts do not have these features. The [given] feature was not acquired, thus suggesting 35.

(48) that the English-speaking learners of Mandarin Chinese were still using the [focus] feature to differentiate the discourse-new and discourse-given information.. In brief, the study claimed that Mandarin clefts have three features, namely [past], [telic], and [given]. The results also showed that English-speaking learners had not acquired the semantic and pragmatic features of Mandarin clefts. However, in the study, the pragmatic functions of clefts were not clearly presented and examined. Moreover, the study did not investigate Mandarin pseudo-clefts. Hence, more research would be desirable.. 2.2.5 Summary Table 2-5 presents a summary of the major findings and limitations of the empirical studies.. Table 2-5 Major Findings and Limitations of the Empirical Studies. Xie (2008). Major Findings. Limitations. 1. Transfer Effects: Yes. 1. L1 interference effect was. Beginners tended to transfer not clearly discussed. their L1 knowledge to L2. 2. The grouping was not 2. Cleft Type Effects: Yes divided according to Different cleft types yielded learners’ L1. Hence, the different difficulties for L2 analyses of L1 transfer learners. effects were problematic. 3. proficiency Effects: Yes 36.

(49) Irgin (2013). 1. Transfer Effects: Yes. 1. L1 interference and L2. L1 influence was adopted. complexity were not 2. Different Structural cleft types clearly discussed. posed difficulties for L2 2. Cleft type effects were not learners. discussed. 3. There was no control group. Liao (2014). 1. Transfer Effects: Yes 1. L2 complexity was not 2. Cleft Types Effects: Yes discussed. L2 Learners had different 2. The VP cleft type was not difficulties in acquiring added into the comparison. different cleft types. 3. proficiency Effects: Yes. Mai and Yuan 1. Transfer Effects: Yes 1. L2 complexity was not (2016) The negative transfer happened. discussed. 2. proficiency Effect: Yes 2. The pragmatic functions of clefts were not clearly presented and examined.. Table 2-5 shows that L1 interference effects were mentioned by most of the studies (Irgin 2013, Liao 2014, Mai & Yuan 2016, Xie 2008). However, these studies did not discuss this issue clearly. Because of this limitation, the current study was also conducted to determine whether L1 interference influences Japanese-speaking learners’ performances on Mandarin Chinese cleft and pseudo-cleft sentences. In addition, the structural complexity effect was not clearly examined in the studies presented in Section 2.2 (Irgin 2013, Liao 2014, Mai & Yuan 2016, Xie 2008). As mentioned in Chapter One, structural complexity might be a factor affecting L2 learners’ language acquisition (DeKeyser 2008, Housen et al., 2005, Williams and Evans 1998). Hence, we assumed that structural 37.

(50) complexity is a critical issue that requires further investigation. In addition, the cleft types and contextual effects mentioned in Chapter One were not clearly discussed in these studies. According to the mentioned limitations, the present study investigated the structural complexity, cleft types, and contextual effects by observing Japanese-speaking learners’ performance levels on Mandarin Chinese clefts and pseudo-clefts.. 2.3 Summary In this chapter, the syntactic structures of Mandarin and Japanese clefts and pseudoclefts are explored, and four empirical studies conducted on L2 learners’ acquisition of English or Mandarin clefts are reviewed. Empirical studies have shown that L1 interference is a factor affecting L2 learners’ cleft acquisition. However, structural complexity, cleft type, and contextual effects have not been clearly examined in these studies. Because few studies have investigated Japanese-speaking learners’ acquisition of Mandarin clefts and pseudo-clefts, we conducted an empirical study on these issues to determine whether the factors (structural complexity, cleft type, and contextual effects) influence Japanese-speaking learners’ performances. The research design of the current study is introduced in Chapter Three.. 38.

(51) CHAPTER THREE RESEARCH DESIGN. In this chapter, the research design of the present study is illustrated in detail. Section 3.1 presented the description of the subjects. Section 3.2 introduces the methods and materials. In Section 3.3, the experimental procedures are described.. 3.1 Subjects The subjects in the study were 30 Japanese-speaking adults: 15 low-intermediate (LI) and 15 high-intermediate (HI) and 15 native speakers of Mandarin Chinese. The Japanese-speaking adults’ Mandarin Chinese proficiency was assessed and categorized according to the Mandarin Chinese placement test from the Mandarin Training Center (MTC) at National Taiwan Normal University (NTNU), Taipei, Taiwan. They had no prior experience in learning Mandarin clefts and were studying Mandarin Chinese as their L2 at the MTC. In addition, a control group, comprising 15 native speakers of Mandarin Chinese, was recruited to facilitate the task of observing whether Japanese-speaking learners’ cleft performances were similar to those of the native speakers. Our subjects in the control group were students who chose Mandarin Chinese as their major or minor at NTNU. A summary of the subjects is presented as follows. 39.

(52) Table 3-1 Summary of the Subjects Group. Language Proficiency. Mean Age. Number. LI. Low Intermediate Level. 24.43. 15. HI. High Intermediate Level. 27.03. 15. Control. Native Speaker. 19. 15. Total. 45. 3.2 Methods and Materials Qualitative (longitudinal) and quantitative (cross-sectional) methods have often been applied in studies regarding language acquisition (Cook & Reichardt 1979, LarsenFreeman 1991, Walliman 2011). In a qualitative study, researchers observe a group of people or an individual case over a specific period. The advantages of conducting a qualitative study are that researchers can collect data over a long period and observe L2 learners’ language performances in a natural setting. However, this type of study usually takes a long time to complete and cannot ensure that the target structure can be triggered in a natural context. By contrast, a quantitative study entails measuring and expressing data numerically, and the results are thus more objective and accurate. Through a quantitative study, researchers can have a more effective control of the test items to test particular sentence structures. Because Mandarin clefts and pseudo-clefts are marked sentences that cannot be easily elicited in a natural context, the present study adopted a quantitative method to investigate JSLs’ acquisition of Mandarin clefts and pseudo-clefts. 40.

(53) As discussed in Chapters One and Two, the following three main factors may influence L2 learners’ cleft acquisition: syntax (Xie 2008), cleft types (Liao 2014), and pragmatics (Mai & Yuan 2016). To determine whether these factors affect Japanesespeaking learners’ cleft performances, three studies were conducted.. 3.2.1 Study I Previous empirical studies (Liao 2014, Mai & Yuan 2016, Xie 2008) have conducted various tasks to test L2 learners’ competence with cleft sentences. Xie (2008) used a grammaticality judgment (GJ) task to test L2 learners’ intuition of the wellformedness of cleft sentences, whereas Liao (2014) and Mai and Yuan (2016) have conducted an acceptability judgment (AJ) task to test L2 learners’ preferences for cleft sentences. In the present study, to observe whether Japanese-speaking learners’ performances can be affected by structural complexity, we investigated the learners’ intuitive judgment about two constructions (clefts and pseudo-clefts). A GJ task is effective because it can be employed to test whether L2 learners are sensitive to syntactic differences. Thus, the present study adopted this task. AJ tasks are mainly used to test L2 learners’ acceptance of sentences, but not their intuition of sentence structures. Hence, it was unsuitable for Study I.. 41.

(54) In the GJ task, the length of the sentences and the vocabulary choice were controlled to prevent the results from being affected by external factors. The task included two sentence types, namely grammatical and the ungrammatical sentences. The grammatical and ungrammatical sentence types included three and six questions, respectively. Additional four filler questions were added to the task as distractors. These questions were designed to test whether the subjects could distinguish between the wellformed and the ill-formed sentences. Table 3-2 presents the structure of the GJ task:. Table 3-2 Test Items for the Grammatical Judgment Task Sentence Types. Questions. Grammatical. 3. Qs 2, 5, 16. Ungrammatical. 6. Qs 1, 3, 6,7, 8, 10. Fillers. 7. Qs 4,9,11,12,13,14,15. Total. 16. In the GJ task, the subjects were asked to judge whether a sentence was correct on the basis of its structure. Two examples are presented as follows:. 42.

(55) Table 3-3 Test Samples of the Grammatical Judgment Task (. (. ) a. Zhangsan shi Xiaoming xihuan de.1 Zhangsan SHI Xiaoming like DE ‘It is Xiaoming who likes Zhangsan.’ ) b. Xiaoming xihuan Zhangsan shi de. Xiaoming like Zhangsan SHI DE. Example (a) is grammatical in Mandarin Chinese, whereas (b) is not. Thus, our subjects should mark (a) with an O and (b) with an X. For the complete version of the GJ task, please refer to Appendix A.. 3.2.2 Study II As mentioned in Section 3.2.1, Acceptability Judgment tasks are mainly used to test L2 learners’ preferences for different cleft types (Liao 2014, Mai & Yuan 2016), whereas GJ tasks assess L2 learners’ understanding of the syntactic structures of Mandarin clefts. By conducting an AJ task, we could investigate Japanese-speaking learners’ interpretation of different cleft sentences. Therefore, an AJ task was considered to be suitable for Study II.. 1. In the formal test, the test items were presented in Mandarin Chinese. Since our subjects have reached the intermediate level, we assumed that they were able to read the Chinese characters in this task. 43.

(56) 3.2.2.1 An Analytical Framework According to Xie (2008) and Liao (2014), L2 learners’ performance is affected by cleft type differences. Therefore, to clearly examine whether JSLs cleft interpretation is influenced by the substitution of focus phrases as previous studies have claimed, a new classification was necessary to help us conduct an in-depth investigation. In this section, analytical frameworks for the sentence configuration of Mandarin clefts and pseudo-clefts are presented in (1a) and (1b), respectively.. (1) a. Cleft Configuration: Subject + shi + cleaved XP + (de)2 b. Pseudo-Cleft Configuration: Sentence + (de) + (Noun) + shi + cleaved XP. The 2 major types and 10 subtypes of Mandarin clefts and pseudo-clefts are illustrated as follows. Type 1: S + shi + cleaved NP + de In a Mandarin cleft, the cleaved NP can be either definite or indefinite as in (2a) and (2b), respectively. (2) a. Zhe xie hua shi Mali. xihuan. this CL flower SHI Mary like. de. DE. ‘It is Mary that likes these flowers.’ b. Chuanghu shi window SHI. yige. nanhai dapo de.. one-CL boy. broke DE. ‘It is a boy that broke the window.’ (Zhang 2012, p.144) 2. In some types of cleft sentences, de is omitted. 44.