日語母語者習得華語未完成貌標誌「在」與「著」之研究

全文

(1)國立臺灣師範大學英語學系碩士論文 Master’s Thesis Department of English National Taiwan Normal University. 日語母語者習得華語未完成貌標誌「在」與「著」之研究 L2 Acquisition of the Mandarin Imperfective Markers Zai and Zhe by L1 Japanese Speakers. 指導教授：李臻儀博士 Advisor: Dr. Jen-i Li 研究生：胡淞筌 Student: Sung-chuan Hu. 中華民國一百零七年八月 August 2018.

(2) 摘要本研究旨在探討以華語為第二語言的日語母語者學習華語未完成貌標誌「在」與「著」之二語習得。根據動貌假說的相關性預測，學習者在習得第二語言未完成貌標誌的早期階段，傾向將未完成貌標誌與活動動詞相關聯，而依據動貌假說的發展性預測，學習者在後期階段，會將未完成貌標誌與動詞的相關聯延伸至有界動詞與狀態動詞。情狀體假說在許多歐洲語言中得到了廣泛的驗證，然而此類研究在中文領域則相對較少，更是缺乏以日語母語者為受測對象之研究。故本研究以檢驗該假說的兩個假設性預測為目的，探討日語母語者對中文未完成貌標誌「在」與「著」之第二語言習得是否符合動貌假說的預測，並了解日語母語者習得「在」與「著」的學習困難，探究母語轉移與語言程度效應在日語母語者習得中文未完成貌標誌中所扮演的角色。總共有六十位日語母語者參與了本研究的文法判斷測驗，受試者依據中文程度分成了兩組實驗組，另外亦含由三十位中文母語人士組成的對照組。本研究發現，日語母語者對中文未完成貌標誌「在」與「著」之習得結果不完全符合動貌假說的相關性預測，亦不完全支持該假說的發展性預測，根據研究結果表示，日語母語者對中文未完成貌標誌「在」與「著」之習得受到母語轉移、第二語言複雜度及語言程度效應等因素所影響。. 關鍵詞：二語習得、華語未完成貌標誌、動貌假說、日語. i.

(3) ABSTRACT The present study aims to investigate the L2 acquisition of Mandarin imperfective markers zai and -zhe by L1 Japanese speakers. According to the Association Prediction of the Aspect Hypothesis, L2 learners tend to associate the imperfective markers strongly with activity verbs in the early stage of acquisition; on the other hand, the Developmental Prediction of the Aspect Hypothesis predicts that as learners’ proficiency increases, they will spread the use of the imperfective markers from activity verbs to telic verbs and stative verbs. Although the Aspect hypothesis has been tested and supported in many studies on the L2 acquisition of European languages (e.g., Robison 1990, 1995, Bardovi- Harlig 1992, 1998, Bardovi-Harlig & Reynolds 1995, Giacalone-Ramat 1995, Bardovi-Harlig & Bergström 1996, Rohde 1996, Tickoo 1996, Li & Shirai 2000, Housen 2002), it has not been extensively investigated in the acquisition of Mandarin and little attention has been paid to test the hypothesis on the acquisition of Mandarin by L1 Japanese speakers. Therefore, the present study fills the gap by investigating whether the acquisition of Mandarin imperfective markers zai and -zhe by Japanese speakers adheres to the predictions of the Aspect Hypothesis. The study also attempts to find out what kind of learning difficulties Japanese learners would encounter and explores the effects of L1 transfer and L2 proficiency on the acquisition of Mandarin zai and -zhe. Sixty Japanese ii.

(4) speakers, who were divided into mid-level group and high-level group according to their Mandarin proficiency, together with thirty Mandarin controls were recruited to complete grammaticality judgment tasks. The study found that Japanese learners’ acquisition of imperfective markers zai and -zhe does not fully support the Association Prediction of the Aspect Hypothesis, nor do the findings fully adhere to the Developmental Prediction of the Aspect Hypothesis. The results indicate that Japanese learners’ acquisition of Mandarin imperfective markers zai and -zhe was affected by L1 knowledge, L2 complexity and L2 proficiency.. Keywords: L2 acquisition, Mandarin imperfective markers, Aspect Hypothesis, Japanese. iii.

(5) 謝辭我終於可以畢業了！心中充滿了無限感動與感激，回顧這條充滿挑戰的研究所求學之路，若是沒有師長、好友、同學與家人的相伴、鼓勵，或許無法堅持到最後一刻! 首先，要感謝我的指導教授李臻儀(Jen-i Li)老師。臻儀老師學識淵博、溫柔細心，從大學到研究所，我從老師身上學到了很多專業知識；擔任助理時，老師總是給予最溫暖的呵護，早上如需擔任助理工作時，老師會關心大家有沒有吃早餐，怕我們餓還會準備小餅乾，真的揪甘心！而我更要感謝臻儀老師總是細心地幫我修改論文，提供我很多建議，指點我迷津、給予我方向，像黑暗中的那盞燈塔，拯救幾度迷失在茫茫大海中的我，有緣跟著老師學習真的是穫益良多啊，如果沒有臻儀老師的耐心指導，我的論文是無法如質如期的完成，謝謝臻儀老師! 接著要感謝我的兩位口試委員：陳純音(Chun-Yin Doris Chen)老師和范瑞玲 (Rueih-Ling Sharon Fahn)老師。純音老師是一位活潑風趣又溫暖的老師，此篇論文的初步架構是在純音老師的「第二語言習得」課程產生的，老師當時給了我很多建議和鼓勵。在師大的日子裡，遇到困難時 Doris 老師的大門永遠為學生敞開，以她一派的幽默和笑容感染、鼓勵著我，讓我更勇敢、更有信心，從老師身上我學到了遭遇難題的轉念智慧與正向積極的人生觀，感念在心！瑞玲老師聲音甜美、溫和慈祥，感謝老師願意抽空大老遠從臺中北上擔任我的口試委員，給予我指導與鼓勵。真的，感謝三位老師從大綱口試到畢業口試所給予的每個寶貴建議與提點，讓我的論文更臻完整！再者，我要向研究所階段指導過我的所有師長致謝，感謝吳曉虹老師、林蕙珊老師、張妙霞老師、詹曉蕙老師、甯俐馨老師、謝妙玲老師以及蘇席瑤老師，每一位老師都超棒的，謝謝老師們用心的教學與指導，讓我能夠根基語言學的知識，進而完成這本論文。另外，我要好好的感謝研究所遇到的好夥伴、好同學們，首先是研究所這些 iv.

(6) 年來陪伴我學習與歡笑的好朋友，花尼卡成員: 陳淑芬(Amy=阿咪=阿花)與陳鈺潔 (Kabby=阿 K=卡比)。在苦悶的研究生涯當中能夠和她們互相扶持，我真的覺得很幸福! 謝謝阿咪，常常不求回報地幫助我處理很多事情，陪我聊天、聽我訴苦，一起為實習之路加油打氣。謝謝阿 K，總是在我心情不好的時候給予我最溫暖的安慰，不管是論文上、生活上、感情上都給了我很多建議，和妳一起聊天談心真的很愉快!能夠結識妳們，阿尼真的非常非常幸運。再來，我要感謝我的論文統計救命恩人: 聰明又迷人的葉惠真(Stephanie)，很感謝親愛的 Ste 在我最需要幫忙的時候幫助我，教我使用 R 統計，還有幫助我解決大大小小的論文問題，沒有 Ste 我也無法順利完成論文! Ste 我真的超感謝妳! 還有統計小天使: 超級可愛漂亮的甜姐兒 Eileen Lin 及聰明帥氣的學弟 Howard，幫助我解決統計上的問題，能有你們這麼優秀的學弟妹我真幸運!感謝幫忙我看問卷及陪我談天說地的研究所好夥伴們 Alan, Francis, Shawn, Johnny, Mark，沒有你們的友情力挺，我也是不可能完成論文的。最後，我要感謝溫暖的避風港—我親愛的家人。謝謝媽媽、爸爸總是無條件的支持與付出，在收問卷的階段，特地陪我到國語中心尋找日本人，主動出擊，一個一個詢問，就算被拒絕也不氣餒，終於快速的收完了所需份數，爸、媽不求回報的幫助與愛心，讓我真的好愛好愛你們，感謝你們提供無憂的生活環境讓我可以專心讀書、長大成人，完成我的研究所學業，因為有家這個永遠堅強的後盾，才能讓我一路走到現在，感謝我最親愛的家人。感謝師大讓我有個圓夢的機會，從大學到研究所我都在師大，我對我的母校有著很深刻的感情，在師大的這些年來，走過了種種的酸甜苦辣，有歡笑、有淚水，看到自己的蛻變、成長，收穫遠遠大於我當初所預期的，我永遠以身為師大人為榮。僅此以本論文獻給所有關心我以及愛我的人，以表達我的無限感激。. v.

(7) TABLE OF CONTENTS Chinese Abstract. i. English Abstract. ii. Acknowledgements. iv. Table of Contents. vi. List of Tables. ix. List of Figures. x. Abbreviation. xi. Chapter One Introduction........................................................................................ 1 1.1 Motivation and Background .................................................................................... 1 1.2 Purpose of this Study ............................................................................................... 7 1.3 Significance of the Study ......................................................................................... 8 1.4 Organization of the Thesis ..................................................................................... 10. Chapter Two Literature Review ............................................................................ 11 2.1 Theories of Aspectuality ........................................................................................ 11 2.2 The Aspectual System in Mandarin ....................................................................... 18 2.2.1 Situation Types in Mandarin ....................................................................... 18 2.2.2 Imperfective Viewpoint Aspect in Mandarin .............................................. 24 2.2.2.1 Imperfective Marker Zai and its Interaction with Situation Aspect ... 24 2.2.2.2 Imperfective Marker -Zhe and its Interaction with Situation Aspect . 26 2.3 Review Comments on the Classification of Situation Types in Mandarin ............ 30 2.4 Theories of Language Acquisition ......................................................................... 39 2.4.1 Aspect Hypothesis ....................................................................................... 40 2.4.2 L1 Influence ................................................................................................ 41 2.5 Previous Empirical Studies on L2 acquisition of Mandarin Aspect ...................... 42 2.5.1 Jin (2006) .................................................................................................... 43 vi.

(8) 2.5.2 Liu (2012) ................................................................................................... 44 2.5.3 Tong & Shirai (2016) .................................................................................. 45 2.6 Summary of Chapter Two ...................................................................................... 46. Chapter Three Methodology .................................................................................. 48 3.1 Participants ............................................................................................................. 48 3.2 Instrument of Data Collection ................................................................................ 50 3.2.1 Situation Types Included in the Experiment ............................................... 50 3.2.2 Cross-linguistic Comparisons between Mandarin and Japanese ................ 58 3.2.3 Questionnaire Design .................................................................................. 64 3.2.3.1 Grammaticality Judgment Tasks for Zai and for -Zhe ....................... 65 3.2.3.2 A Multiple-Choice Judgment Task for Zai and -Zhe ......................... 67 3.2.4 Scoring Policy and Statistical Analysis....................................................... 70 3.3 Procedure ............................................................................................................... 70. Chapter Four Results and Discussion ................................................................... 72 4.1 Results of the Grammaticality Judgment Tasks and the Aspect Hypothesis ......... 72 4.1.1 The Result of Zai’s Judgment Task ............................................................. 73 4.1.2 Discussion of Zai’s Result .......................................................................... 77 4.1.3 The Result of -Zhe’s Judgment Task ........................................................... 82 4.1.4 Discussion of -Zhe’s Result ........................................................................ 85 4.1.5 A Comparison between the Results of Zai and -Zhe ................................... 88 4.1.6 Discussion of the Results ............................................................................ 91 4.1.7 Summary of Section 4.1 .............................................................................. 94 4.2 The Result of the Multiple-choice Judgment Task for Zai and -Zhe ..................... 95 4.2.1 The Result of the Multiple-choice Judgment Task ..................................... 95 vii.

(9) 4.2.2 Discussion of the Result............................................................................ 101 4.3 Summary of Chapter Four ................................................................................... 105. Chapter Five Conclusion ...................................................................................... 107 5.1 Major Findings ..................................................................................................... 107 5.2 Theoretical and Pedagogical Implications ........................................................... 110 5.3 Limitations of the Current Study and Suggestions for Future Research ............. 111. References ................................................................................................................. 113 Appendix A Consent Form........................................................................................ 117 Appendix B Form of Paticipants’ Language Background ........................................ 118 Appendix C Questionnaire........................................................................................ 119. viii.

(10) LIST OF TABLES. Table 2-1 Smith’s Five Situation Types ....................................................................... 16 Table 2-2 Tai’s Three Types of Verbs in Mandarin Chinese ........................................ 19 Table 2-3 Liu’s Types of Aspectual Events in Mandarin Chinese ............................... 22 Table 3-1 Summary of Information of the Participants ............................................... 49 Table 3-2 Situation Types Included in the Questionnaire ............................................ 57 Table 3-3 Comparisons between Mandarin and Japanese on the Compatibility between Situation Types and Imperfective Viewpoints ............................... 60 Table 3-4 Distribution of the Test Items for Zai in Part I ............................................ 66 Table 3-5 Distribution of the Test Items for -Zhe in Part II ......................................... 66 Table 3-6 Distribution of the Test Items for Multiple-choice Task in Part III ............. 69 Table 4-1 Mean Scores and Standard Deviation of Zai in the Part I Task ................... 74 Table 4-2 The Result of Post-hoc Tests of the Between-group Differences on Zai ..... 76 Table 4-3 Developmental Trend of Zai ........................................................................ 77 Table 4-4 Mean Scores and Standard Deviation of -Zhe in the Part II Task ................ 82 Table 4-5 The Result of Post-hoc Tests of the Between-group Differences on -Zhe ... 84 Table 4-6 Developmental Trend of -Zhe ...................................................................... 85 Table 4-7 A Comparison of Results of Zai and -Zhe with Situation Types ................. 89 Table 4-8 The Result of the Multiple-choice Judgment Task for the Between-option Differences ................................................................................................... 97 Table 4-9 The Result of the Multiple-choice Judgment Task for the Between-group Differences ................................................................................................... 98. ix.

(11) LIST OF FIGURES. Figure 4-1 Mean Scores of Zai with the Seven Situation Types .................................. 74 Figure 4-2 Mean Scores of -Zhe with the Seven Situation Types ................................ 83. x.

(12) Abbreviation. AH. Aspect Hypothesis. AP. Association Prediction. CAH. contrastive analysis hypothesis. CL. classifier. DE. Mandarin morpheme de. DP. Developmental Prediction. DUR. durative marker. OM. object marker. PART. particle. PFV. perfective. PRO. progressive marker. RVC. Resultative verb compound. SM. subject marker. TEIRU. Japanese aspectual marker -teiru. ZAI. progressive marker zai. ZHE. durative marker -zhe. xi.

(13) Chapter One Introduction. 1.1 Motivation and Background Aspect has received a great deal of interest in the linguistic field for over half a century. It expresses characteristics of the internal structure of an event (Kearn 2000:200). The internal structure of events, according to Smith (1997), is a composite of lexical aspect and grammatical aspect. Lexical aspect of verbs/verb constellations is also called ‘aktionsart’ (Agrell 1908) or ‘situation aspect’ (Smith 1997), such as State, Activity, Accomplishment, Achievement, and Semelfactive (Smith 1997). Lexical aspect is distinguished from grammatical aspectual properties indicated by grammatical morphemes, called ‘viewpoint aspect’ by Smith (1997), such as perfective aspect and imperfective aspect. According to Smith (1997), these two types of aspect are defined independently, but interact with one another in a sentence to bring about the temporal interpretation of an event. According to Smith (1997:3), imperfective viewpoint aspect gives information about the internal stage of an event, focusing on part of a situation, an interval that excludes the initial and final endpoints. Imperfective viewpoint aspect exists in many languages of the world. According to Smith (1997:171), imperfective viewpoints occur with a variety of verb classes, available for non-stative events, and its scope excludes the initial point as well as the 1.

(14) endpoint of an event. In English, for instance, the imperfective viewpoint aspect is expressed through the progressive suffix -ing (Smith 1997:171); in Japanese, it is indicated by the suffix -teiru (Shirai 1998:237). On the other hand, the progressive zai and durative -zhe are two imperfective viewpoint aspect markers in Mandarin Chinese. The progressive zai signals the process of an event and it excludes the endpoint of an event (Li & Thompson 1981:218, Smith 1997:212), while -zhe marks a progressive event, the continuation of an activity, a durative event, or a resultant state (Li & Hsieh 2015:28). The Mandarin imperfective zai and -zhe are often translated into V-ing in English and V-teiru in Japanese. However, English imperfective -ing’s and Japanese -teiru’s compatibility with verb types is different from that of Mandarin zai and -zhe. As illustrated in (1), English -ing and Japanese -teiru can occur with Achievement verbs (or instantaneous verbs), such as ‘die’ as in (1a) and (1b), while Mandarin zai and -zhe can’t, as in (1c) and (1d). The reason why English -ing is compatible with Achievement verbs is that it accepts the preliminary stage of instantaneous verbs, which focuses on the temporal span before the initial point of instantaneous events (Smith 1997:163). On the other hand, Japanese -teiru can occur with Achievement verbs because it accepts the resultant state of Achievements, which focuses on the temporal span after the endpoint of instantaneous events (Smith 1997:27, Shirai 1998:332). However, no such temporal span of instantaneous events is acceptable to Mandarin zai and -zhe. 2.

(15) (1) a. The old man is dying.. (Riemer 2010:326). b. kare-wa sin-deiru. he-SM die-TEIRU ‘He is dead.’ c. * Tā zài sǐ. he ZAI die ‘He is dying.’ d. * Tā sǐ-zhe. he die-ZHE ‘He is dead.’. (Nedjalkov 1988:492). Such (in-)compatibility between the situation aspect and the viewpoint aspect may pose learning difficulties for L2 learners of a foreign language. For example, when learners of Mandarin learn Mandarin viewpoint markers zai and -zhe, they may have difficulty in finding out what types of verbs zai and -zhe can and cannot occur with. As a result, they may make grammatical mistakes. This issue of acquisition on aspectuality has received much attention in the literature. A well-known hypothesis addressing the issue is the Aspect Hypothesis (AH) (Andersen & Shirai 1994, 1996, Bardovi-Harlig 1999, 2000), which highlights the acquisition patterns reflecting the interactions between the situation aspect and the viewpoint aspect when learners acquire a second language. The Aspect Hypothesis makes use of the four types of aspectual verb classes proposed by Vendler (1967): states, activities, accomplishments and achievements. According to Bardovi-Harlig (2000), there are four generalizations subsumed under the Aspect Hypothesis, listed as follows: (2) Aspect Hypothesis (a) Learners first. use. (perfective) 3. past. marking. on. achievements. and.

(16) accomplishments, eventually extending use to activities and states. (b) In languages that encode the perfective/imperfective distinction, imperfective past appears later than perfective past, and imperfect past marking begins with states, extending next to activities, then to accomplishments, and finally to achievements. (c) In languages that have progressive aspect, progressive marking begins with dynamic-atelic predicates and activities, then extends to telic predicates like accomplishments and achievements, and finally to states. (d) Progressive markings are not incorrectly overextended to states. (Bardovi-Harlig 2000:227) Of the four generalizations subsumed under the Aspect Hypothesis, (2c) and (2d) are the ones concerned with the imperfective progressive aspect. In addition, according to Shirai (2004) and Robison (1995), the Aspect Hypothesis has two predictions: the Association Prediction (AP) and the Developmental Prediction (DP), covering early and later stages of acquisition of aspect respectively. In terms of imperfective aspect, the Association Prediction predicts that language learners associate the imperfective aspect marker more strongly with activity verbs than with accomplishment and achievement verbs; On the other hand, the Developmental Prediction predicts that as learners’ proficiency increases, they will spread the use of the imperfective aspect marker from activity verbs to accomplishment and achievement verbs (Shirai 2004). Putting Bardovi-Harlig’s (2000) generalizations (2c) and (2d) together with Shirai’s (2004) and Robison’s (1995) two predictions of Aspect Hypothesis (i.e., the Association Prediction and the Developmental Prediction), we can make two predictions of zai and -zhe’s acquisition pattern: Learners initially would associate the imperfective zai and -zhe with situation types that denote process (i.e., activity) only. They 4.

(17) would later expand the situation types that occur with zai and -zhe to telic verbs (i.e., accomplishment and achievement) and finally to stative verbs. Although the Aspect Hypothesis had been tested and supported in studies on the L2 acquisition of many European languages, such as English, German, Spanish, French, and Italian (e.g., Robison 1990, 1995, Bardovi- Harlig 1992, 1998, Bardovi-Harlig & Reynolds 1995, Giacalone-Ramat 1995, Bardovi-Harlig & Bergström 1996, Rohde 1996, Tickoo 1996, Li & Shirai 2000, Housen 2002), it had not been extensively investigated in the L2 acquisition of Mandarin Chinese. Among the few studies concerning the Aspect Hypothesis in the acquisition of Mandarin are Jin (2006), Liu (2012) and Tong and Shirai (2016). Jin investigates the L2 acquisition of the perfective markers -guo and -le, and the imperfective markers zai and -zhe by English speakers. Liu did a study on L2 acquisition of the progressive maker zai in Mandarin Chinese by L1 English speakers, and Tong and Shirai conducted a study on L2 acquisition of Mandarin Chinese progressive zai and perfective -le. However, in these studies, the native languages of the participants were either not controlled or only native English speakers were recruited. Besides, the studies that test the Aspect Hypothesis in the acquisition of Mandarin by L1 Japanese speakers are very scarce. According to the survey conducted by Taiwan’s Ministry of Education in 2016, among the 19,977 foreign students who came to Taiwan for learning Mandarin, Japanese students (4387) constituted the highest percentage (22%) of the total population, second being 5.

(18) students from South Korea (2568, 12.8%) and third Americans (2495, 12.5%). 1 Since Japanese speakers are the majority among the foreign students who are learning Mandarin in Taiwan, a study focusing on L1 Japanese speakers’ acquiring Mandarin can therefore contribute to the teaching of Mandarin. In addition, it is well-known that L2 acquisition may be influenced by the learners’ first language, which is so-called the ‘language transfer’ or ‘L1 effect’ (Fries 1945, Lado 1957). Mandarin progressive marker zai has a corresponding form in Japanese, which is -teiru. Interestingly, Mandarin imperfective marker -zhe also corresponds to Japanese -teiru (Huang 1990, Shirai 1998, He 2000). Therefore, Mandarin example (3a) with zai and (3b) with -zhe are both translated into (3c) in Japanese. (3) a. Ken zài. chàng gē.. Ken ZAI sing song ‘Ken is singing.’ b. Ken chàng-zhe gē. Ken sing-ZHE song ‘Ken is singing.’ c. Ken-ga uta-o utat-teiru. Ken-SM song-OM sing-TEIRU ‘Ken is singing.’. (Shirai 1998:661). Besides, Japanese viewpoint aspect -teiru has mainly two different aspect meanings: one is progressive, corresponding to Mandarin progressive zai and -zhe, and the other is durative, indicating certain continuous state corresponding to Mandarin -zhe (Huang 1990, He 2000). Such difference may pose learning difficulties to Japanese speakers learning Mandarin zai 1. The data is from Ministry of Education, Republic of China (Taiwan), Number of Foreign Students in R.O.C. (1954- 2016): http://english.moe.gov.tw/ct.asp?xItem=14530&ctNode=11432&mp=1 6.

(19) and -zhe. Given the difference between the two languages, a study of L2 acquisition of zai and -zhe by Japanese speakers can provide a more comprehensive perspective on this topic.. 1.2 Purpose of this Study The major purpose of this study is to investigate the L2 acquisition of Mandarin imperfective markers zai and -zhe by L1 Japanese speakers in order to discover what kind of learning difficulties they may encounter and to see if the findings adhere to the Aspect Hypothesis and if L1 effect plays a role when Japanese learners acquire zai and -zhe. It is known that the (in-)compatibility between the situation aspect and the viewpoint aspect varies from language to language and may hence pose learning difficulties for L2 learners. Thus, we can predict that when Japanese speakers learn Mandarin viewpoint markers zai and -zhe, they may have difficulty in finding out what types of verbs zai and -zhe can and cannot occur with because of their L1 influence, which can therefore lead to grammatical mistakes. Applying the Aspect Hypothesis to the acquisition of zai and -zhe, we can also make two predictions: Japanese learners initially would associate the imperfective zai and -zhe with situation types that denote process (i.e., activity) only. They would later expand the use of zai and -zhe to telic verbs (i.e., accomplishment and achievement) and stative verbs. To examine the L2 acquisition of Mandarin zai and -zhe by L1 Japanese speakers and to 7.

(20) test the Aspect Hypothesis, the study is guided by the following research questions: a. Does L1 Japanese speakers’ use of the two imperfective viewpoints zai and -zhe adhere to the Association Prediction of the Aspect Hypothesis in the early stage of acquiring the two markers? b. When L1 Japanese speakers’ proficiency of Mandarin increases, does their use of imperfective viewpoints zai and -zhe adhere to the Developmental Prediction of the Aspect Hypothesis? Is L2 proficiency a factor affecting Japanese learners’ acquisition of Mandarin imperfective markers zai and -zhe? c. Of the two imperfective markers zai and -zhe, which one is more challenging to Japanese speakers? What is the sequential pattern of acquisition of the two imperfective markers? d. Does L1 influence play a role in Japanese speakers’ acquisition of zai and -zhe?. 1.3 Significance of the Study Aspect has been widely discussed before, and the Aspect Hypothesis was proposed to investigate the acquisition patterns reflecting the interactions between the situation aspect and the viewpoint aspect when learners acquire a second language. Although the Aspect hypothesis has been tested and supported in many studies on the L2 acquisition of European languages, it has not been extensively investigated in the L2 acquisition of Mandarin Chinese. 8.

(21) Previous studies which investigate L2 acquisition of Mandarin aspect have not either controlled the participants’ L1 language backgrounds or only recruited L1 English speakers. Furthermore, little attention has been paid to test the Aspect Hypothesis on Japanese speakers’ acquisition of Mandarin aspect. Therefore, this study may fill the gap. This study also presents a detailed review of the situation aspect in Mandarin and provides comments on it. Although Tai (1984), Smith (1991, 1997), and Liu (2014) have classified Mandarin predicates and provided their own categories of Mandarin situation types, there are problems with their classifications. Therefore, a detailed review of the situation/ lexical aspect in Mandarin is presented. The current study’s major difference from previous studies is that rather than paying attention to investigating European languages or native English speakers’ acquisition of Mandarin aspect, the study focuses on Mandarin Chinese’s aspect system and examines Japanese speakers’ acquisition of Mandarin aspect. By conducting experiments, collecting data from Japanese speakers, and using statistic tools to analyze the data, the present study of L2 acquisition of zai and -zhe by Japanese speakers will shed light on a number of issues on the acquisition of aspectuality, including what the early stage is like, what the acquisition order is, and what the extent of L1 influence is. From this study, it is hoped that one can have better understanding on how Japanese speakers acquire Mandarin aspect and be aware of their learning difficulties. It is also hoped that the study can contribute to the Aspect 9.

(22) Hypothesis, either support or oppose it, in order to have a fuller picture of the acquisition on the aspect system.. 1.4 Organization of the Thesis This thesis consists of five chapters. Chapter One introduces the motivation, the purposes, and the significance of the present study. In Chapter Two, previous literature related the present study is reviewed, including theoretical studies and empirical studies on aspect and acquisition. Chapter Three introduces the methodology of investigating Japanese speakers’ use of Mandarin viewpoint aspect zai and -zhe. Chapter Four presents and discusses the results and findings of the current study. Finally, Chapter Five summarizes and concludes the findings of this study and makes closing remarks.. 10.

(23) Chapter Two Literature Review. In this chapter, the literature related to aspectuality and theories of language acquisition is reviewed. The theories of aspectuality are discussed in Section 2.1. In Section 2.2, the aspectual system in Mandarin is introduced. Section 2.3 provides the review comments on the classification of situation types in Mandarin. Section 2.4 presents the theories related to language acquisition, the Aspect Hypothesis and L1 influence. Section 2.5 reviews the empirical studies on L2 acquisition of Mandarin aspect. Finally, Section 2.6 provides a summary of this chapter.. 2.1 Theories of Aspectuality The aspect of a sentence contributes the temporal information and the point of view to the event that the sentence describes (Smith 1997:123). According to Kearns (2000:201), aspect expresses characteristics of the internal structure of an event, which, to some extent, are already coded in the verb or verb constellation which serves as the predicate denoting the event. For over half a century, aspect has aroused interests in the linguistic field. The most well-known aspectual classification of verbs might be Vendler’s (1957, 1967). Vendler classifies verbs into four semantic 11.

(24) categories: activity, accomplishment, achievement, and stative verbs. According to Vendler (1957:149), activity verbs describe processes going on without a terminal point. For instance, run, walk, swim, sing, and the like are activity verbs. Vendler’s accomplishment verbs refer to processes for periods of time that are unique or definite; they thus have a ‘climax’ or a necessary terminal point. For example, paint a picture, build a house, make a chair, and write a novel are accomplishments. Achievement verbs, according to Vendler, denote events which occur at a single moment and involve a necessary terminal point. For instance, recognize, realize, notice, arrive, win, and die all fall into the class of achievements. Stative verbs, Vedler’s fourth type of verbs, involve periods of time in an indefinite and non-unique sense. They are stative and can go on and on without a terminal point. For example, have, want, like, love, know, believe, be happy, and be tall are manifestly states. However, Vendler’s classification is problematic. First, Vendler does not distinguish between grammatical aspect and lexical aspect nor does he explain why verbs with endpoints, such as accomplishments, can co-occur with the progressive marker. For example, the English V-ing is a progressive form and denotes an on-going event without endpoints (Smith 1997:130); however, it can co-occur with accomplishment verbs, which denote durative events containing necessary endpoints (Smith 1997:178). In principle, the V-ing form should be incompatible with 12.

(25) accomplishment verbs based on their different natures. Such co-occurrence causes a conflict and Vendler does not explain it in his theory. The second problem is that Vendler only considers ‘verb types’ or ‘verb classes’. However, the aspectual meaning of an event is generally denoted by more than just the verb alone (Smith 1997:4). The evidence comes from sentences containing the same verb but different only in the complement or argument of the verb. For example, (1) and (2) present pairs of sentences with the same verbs which differ only in the complement and argument. (1) a. John walked in the park. (without an endpoint) b. John walked to school. (with an endpoint) (2) a. Amy ate apples. (without an endpoint) b. Amy ate an apple. (with an endpoint) Although they have the same verb walked in (1), (1a) has a locative complement in the park, which makes the event go on without an endpoint, whereas (1b) has a directional complement to school, which makes the event have a necessary endpoint. Similarly, while (2a-b) has the same verb ate, the noun phrase of (2a) refers to an uncountable quantity apples, which makes the event continue indefinitely, whereas the object of (2b) refers to a specific quantity an apple, which makes the event have a clear endpoint. Therefore, the aspectual meaning of an event is not only denoted by the main verb, but also by the complements or the arguments of the verb (Smith 1997:4). In resolving the problems that Vendler does not address, Smith (1991, 1997) 13.

(26) divides aspect into two components  ‘situation aspect’ and ‘viewpoint aspect’. The two types of aspect are defined independently, but interact with one another in a sentence to bring about the temporal interpretation of an event. According to Smith, the viewpoint aspect consists of perfective, imperfective, and neutral viewpoints. Viewpoint aspect is often conveyed by grammatical morphemes. The perfective viewpoint focuses a situation in its entirety, including both the initial and final endpoints; the imperfective viewpoint, on the other hand, focuses part of a situation, an interval that excludes the initial and final endpoints; and the neutral viewpoints includes only one endpoint and they lack a grammatical morpheme (Smith 1997:66). On the other hand, Smith’s situation aspect is conveyed by the verb constellation, including a main verb and its argument(s), such as eat an apple. According to Smith, situation aspect has five different types  State, Activity, Accomplishment, Achievement, and Semelfactive. Smith (1997:19) uses three temporal features, namely dynamicity, durativity and telicity, to classify situation types into five. Dynamicity, according to Smith, refers to events that comprise all non-stative situations which are continually ‘subject to a new input of energy’. Dynamic events take place in time and ‘they consist of successive stages which occur at different moments’ (Smith 1997:19). A durative situation, according to Smith, occupies a certain period of time; in contrast, an 14.

(27) instantaneous situation is idealized to happen at one moment in time. Telic events in Smith (1997:19) refer to situations that ‘have a change of state which constitutes the outcome, or goal, of the event’, and when the outcome or goal of the event is reached, the change of state occurs and the event is complete. Therefore, a telic event consists of a natural endpoint or intrinsic bound, whereas an atelic event does not have an inherent endpoint or the endpoint is arbitrary (Smith 1997:19). Smith’s (1991, 1997) five situation types based on the three temporal features are explained as follows. According to Smith, Activity is a dynamic and durative event without an endpoint; hence, Activities are atelic events. For instance, stroll in the park, laugh, eat cherries, walk, swim, sing are cases of Activities. Smith’s Accomplishments are dynamic activities with duration that have a necessary endpoint or change of state. For example, build a bridge, walk to school, drink a glass of wine, write a letter are all Accomplishments. Achievements, according to Smith, refer to events that are punctual, instantaneous, and telic. For instance, leave the house, reach the top, recognize Aunt Jane, win/lose a race, cross the border, be born, and die all fall into the class of Achievements. States in Smith (1991, 1997) refer to situations that are stative with duration and can go on and on without an endpoint; that is, they are atelic situations. For example, own the farm, be in Copenhagen, be tall, know the answer, believe in ghosts are States. Smith’s Semelfactives are similar to Achievements in that 15.

(28) they both are dynamic, punctual single-stage events except that Semelfactives are atelic, such as cough, knock, wink, hiccup, and pat. Table 2-1 summarizes Smith’s situation types and their temporal schemata in binary terms. Table 2-1 Smith’s Five Situation Types (Smith 1997:20) Situation types. [±dynamic]. [±durative]. [±telic]. examples. Activity. +. +. -. run, sing, swim. Accomplishment. +. +. +. build a bridge, write a letter,. Achievement. +. -. +. reach, recognize, die, win. State. -. +. -. own, know, believe, be tall. Semelfactive. +. -. -. cough, knock, wink. According to Smith (1997), the concepts of aspect play a role in all languages and the aspectual systems of different languages have a striking similarity. Therefore, Smith proposes a Universal Grammar (UG)1 account of aspect, which provides the general structure of aspectual systems across languages. The UG account of aspect suggests that every language has viewpoint aspect and similar categories of situation types and that the situation types have the same properties across languages; however, the interaction between viewpoint aspect and situation aspect varies from language to language along the parameter of aspect (Smith 1997:13). For example, the property of an Achievement event is the same across languages, i.e., instantaneous, punctual, and involving a change of state; however, the interaction between viewpoint aspect and Achievement events varies due to the parameter of aspect across languages. Take 1. The concept of Universal Grammar (UG) is proposed by Chomsky. UG is the theory of the language faculty. The basic postulate of UG is that language faculty is innate to humans, independent of other cognitive abilities and that every language has similar linguistic structures. 16.

(29) English and Mandarin for instance. Unlike English Achievement verbs, which can co-occur with the progressive viewpoint, Mandarin Achievement verbs cannot be marked with the progressive marker. As illustrated in (3), die is an Achievement verb in English, and it is compatible with the progressive viewpoint form V-ing. However, sǐ ‘die’ in Mandarin, also an Achievement verb, cannot co-occur with the progressive viewpoint marker zai, as shown in (4). (3) He is dying. (4) * Tā zài sǐ. he ZAI die ‘He is dying.’ The reason is that in some languages, like English, the progressive viewpoint allows a preliminary stage of Achievement verbs, which focuses on the temporal span before the initial point of instantaneous events, while in Mandarin it does not (Smith 1997:272). Such focus on the preliminary stage of an Achievement is marked and non-standard (Smith 1997:75). According to Smith, speakers make choices within the two components, i.e., situation types and viewpoint aspect, and standard choices are unmarked while non-standard or unexpected choices are marked (Smith 1997:10). Smith’s parametric approach of aspect can therefore account for some marked combinations of situation types and viewpoints that occur only in a particular language. It also provides a satisfactory explanation for the problems Vendler (1957, 1967) does not address, mentioned above. 17.

(30) 2.2 The Aspectual System in Mandarin Aspect in Mandarin Chinese has attracted a great deal of interest. This interest might be due to the fact that Mandarin Chinese is known as a tenseless language (Lin 2003, 2006, Wu 2009), lacking grammatical tense but rich in aspectual expressions (Chen & Shirai 2010). Mandarin viewpoint aspect is expressed by aspectual markers which are morphology-like devices and Mandarin situation aspect is manifested by different types of verb constellation. Different classifications of Mandarin situation types have been proposed by linguists such as Tai (1984), Smith (1991, 1997), and Liu (2014). In this section, Mandarin situation aspect is discussed in 2.2.1. As for Mandarin viewpoint aspect, since the present study mainly focuses on the imperfective viewpoint, only the imperfective viewpoint Aspect of Mandarin is introduced and is presented in 2.2.2.. 2.2.1 Situation Types in Mandarin In this section, different classifications of Mandarin situation/verb types, proposed by Tai (1984), Smith (1991, 1997), and Liu (2014), are presented to provide a full picture of Mandarin situation aspect. Based on Vendler’s (1957, 1967) categorization scheme, Tai (1984) classifies Mandarin verbs into three types: state, activity, and result, as shown in table 2-2. 18.

(31) Table 2-2 Tai’s (1984) Three Types of Verbs in Mandarin Chinese Type of verbs. State Activity Result. Examples gāoxìng ‘glad’, zhīdào ‘know’, xiàngxìn ‘believe’, yǒu ‘have’ xué ‘learn’, pǎo ‘run’, chàng ‘sing’, tīng ‘listen’ chīwán ‘finish eating’, xuéhuì ‘learn’, sǐ ‘die’, pò ‘break’. According to Tai, state verbs in Mandarin involves periods of time that can go on and on without a terminal point, such as gāoxìng ‘glad’ and zhīdào ‘know’. Tai’s activity verbs in Mandarin describe processes going on without a terminal point, such as xué ‘learn’ and pǎo ‘run’. Result verbs, according to Tai, have a clear result or change of state, such as xuéhuì ‘learn’ and sǐ ‘die’. In his study, Tai finds that Chinese often uses resultative verb compounds (RVCs) to describe events that English specifies with accomplishment and achievement verbs (Tai 1984:292, 293). According to Chao (1968) and Li & Thompson (198:55), an RVC is a [V + R] compound by morphologically adding a verb or an adjective (adjectival state verb) containing a result phase (R) to a verb stem (V) containing an action phase. For example, we can make the RVC chī-bǎo ‘eat-full’ (to become full after eating) by using the verb chī ‘eat’, which denotes an action, as the first component, and the ajective bǎo ‘full’, which denotes a result, as the second component. Tai argues that both Mandarin RVCs and English accomplishment verbs have the action-result semantic relation between their semantic components, but they are different in that 19.

(32) unlike English accomplishment verbs, which encode the aspectual meaning of both the action and result, the Mandarin RVCs, however, include only the result aspect in their aspectual meaning. Therefore, Tai suggests that Mandarin Chinese does not have accomplishment verbs and that Mandarin doesn’t distinguish accomplishment verbs from achievement verbs (Tai 1984:293). As a result, Tai creates a new category, called ‘result’, to include RVCs and resultative simple verbs (e.g., sǐ ‘die’ and yíng ‘win’), both of which have a conspicuous result or change of state. In addition to Tai, Smith (1991, 1997) also examines aspectuality in Mandarin and classifies Mandarin situation aspect into five types, namely Activity, Accomplishment, Achievement, State, and Semelfactive. Examples for each type are illustrated in (5). (5) a. Activity Wǒ xué fǎwén. I study French ‘I studied French.’ b. Accomplishment Tāmen chī-bǎo le. they eat-full PFV ‘They ate their fill.’ c. Achievement. (Smith 1997:285). (Smith 1997:287). Zhāngsān sǐ le. Zhangsan die PFV ‘Zhangsan died.’ d. State Mǎlì hěn gāoxìng. Mary very happy ‘Mary is very happy.’. (Smith 1997:291). (Smith 1997:292) 20.

(33) e. Semelfactive Zhāngsān késòu le. Zhangsan cough PFV ‘Zhangsan coughed.’. (Smith 1997:290). It is worth noting that Smith considers that Accomplishmants in Mandarin are often denoted by RVCs. According to Chao (1968) and Smith (1991, 1997), the second component/result phase is the complement of an RVC. Smith (1997) suggests that the complement of the RVC fall into two distributional classes: directional complements and resultative complements. Directional complements refer to the direction of an action and locate a situation from the point of view of the speaker. Resultative complements are of two types: resultative result state and resultative phase/completive. The complement of resultative result state indicates a change of state, while resultative phase/completive indicates a completion of an event. Smith considers RVCs to denote Accomplishment events because an action-result structure indicates a process with a duration and a result or change of state. Both directional and resultative complements are included in (6). (6) a. Directional complements of RVC: dao ‘arrive’, qu ‘arrive’, jin ‘in, into’, shang ‘ascend’, chu ‘out’, qi ‘up’ , etc. b1. Resultative Result State complements of RVC: bao ‘full’, qingchu ‘clarity’, kai ‘detachment’; zhu ‘fixity’, cuo ‘error’ , etc. b2. Resultative Phase/Completive complements of RVC: wan ‘finish’, hao ‘satisfaction’, cheng ‘succeed’, etc. (Smith 1997:282) Besides Smith’s (1991, 1997) and Tai’s (1984) classifications of Mandarin verbs/ 21.

(34) situation types, Liu (2012) also classifies Mandarin verb types in her study investigating L2 acquisition of the progressive maker zai in Mandarin Chinese by L1 English speakers. In her study, she identifies 10 predicate types of aspectual event in Mandarin Chinese, as shown in table 2-3 and described below. Table 2-3 Liu’s (2012:166) Types of Aspectual Events in Mandarin Chinese Situation type. Dynamic. Durative. Completive. Examples. Activity. +. +. no. kàn diànshì ‘watch TV’. +. +. no. +. +. yes. +. +. yes. +. +. yes. +. -. yes. +. +. no. Accomplishment (non-completive) Accomplishment (goal) Accomplishment (duration) Accomplishment (distance) Achievement Change with no result. kàn yī běn shū ‘read a book’ fēiqù běijīng ‘fly to Beijing’ xué liǎngnián zhōngwén ‘learn 2 years of Chinese’ pǎo yī lǐ lù ‘run a mile’ sǐ ‘die’ mànmàn shēnggāo ‘slowly go up’ xiěwán gōngkè. Result. +. +. yes. ‘write and finish homework’. State. -. +. no. State-location. +. +. no. xǐhuān zhōngwénkè ‘like Chinese class’ shuìzài dì shàng ‘sleep on the floor’. According to Liu, Activities are dynamic events without a natural endpoint (atelic). Liu classifies Accomplishments into four types: Accomplishments (non-completive) denote a dynamic process with a duration that does not necessarily. 22.

(35) have a completion, i.e., the process may not be completed; Accomplishments (goal) describe a process with duration that has a goal as the endpoint and refers to a directive process towards a certain location as the destination; Accomplishments (duration) describe a process that has a bounded time duration, i.e., the event occurs within a certain period of time; and Accomplishments (distance) describe a process that covers a certain distance in space. Achievements, according to Liu, describe events that are punctual and instantaneous with endpoints. Result in Liu (2012) refers to events which have a definite result and Liu’s Result is expressed by RVCs. Liu’s Change with no result refers to events that have change of state, and yet have no obvious or specific result. State, according to Liu, describes a situation that is stative with duration and without an endpoint. Liu’s State-location refers to those events in which someone doing a bodily posture statively in certain location. In sum, different classifications of Mandarin situation/verb types are proposed by scholars to provide a full picture of Mandarin situation aspect. Tai (1984) classifies Mandarin verbs into three types: state, activity, and result; Smith (1991, 1997) proposes five types of Mandarin situation types: Activity, Accomplishment, Achievement, State, and Semelfactive; and Liu (2014) identifies ten predicate types of aspectual event in Mandarin Chinese: Activity, Accomplishment (non-completive), Accomplishment (goal), Accomplishment (duration), Accomplishment (distance), 23.

(36) Achievement, Change with no result, Result, State, and State-location.. 2.2.2 Imperfective Viewpoint Aspect in Mandarin According to Smith (1997), viewpoint aspect presents the temporal schemata of an event. It is often expressed with a grammatical morpheme associated with the main verb of a sentence. Mandarin Chinese has several viewpoint aspectual markers, including perfective viewpoint markers -le and -guo and imperfective viewpoint markers zai and -zhe. Since this study investigates the L2 acquisition of Mandarin imperfective marker zai and -zhe by Japanese speakers, the focus will be on these two imperfective viewpoint markers in the following sections. Section 2.2.2.1 introduces the imperfective viewpoint marker zai and its interaction with the situation aspect, and Section 2.2.2.2, the imperfective viewpoint marker -zhe and its interaction with the situation aspect.. 2.2.2.1 Imperfective Marker Zai and its Interaction with Situation Aspect In Mandarin, the morpheme zai has several meanings. Zai can function as a locative marker as in zài xiūxīshì lǐ ‘in the lounge’ (Lü 1999:646), a temporal marker as in zài xiàwǔ sāndiǎn ‘at 3 p.m.’ (Lü 1999:645), and a progressive marker indicating 24.

(37) on-going events as in wǒmen zài tǎolùn ‘we are having a discussion’ (Lü 1999:672). Since the present study focuses on zai’s functioning as an imperfective viewpoint marker, in this section, zai’s aspectual meaning and its interaction with Mandarin situation aspect are reviewed. Smith (1997) and Lü (1999:672) consider the Mandarin imperfective viewpoint zai a typical progressive marker. According to Smith (1991, 1997), zai presents an internal interval of a durative but non-stative event and, consequently, Activities and Accomplishments, which are dynamic and durative, are compatible with zai (Smith 1997:271), as illustrated in (7a-b). (7) a. Zhāngsān zài dǎqiú. Zhangsan ZAI play ball ‘Zhangsan is playing ball.’. (Activity) (Smith 1997:272). b. Lǐsì zài xiě yīfēng xìn. Lisi ZAI write one-CL letter ‘Lisi is writing a letter.’. (Accomplishment) (Smith 1997:272). In addition, the dynamic feature of the progressive zai does not allow it to occur with States, which are not dynamic, as shown in (8). (8) *Tā zài gāoxìng. he ZAI glad ‘He is being glad.’. (States). On the other hand, the durative feature of the progressive zai requires an interval that has duration, and hence prevents it from occurring with Achievements, because no such interval is available internally for instantaneous events (Smith 1997:272), as shown in (9). 25.

(38) (9) * Lǎowáng. zài sǐ.. (Achievements). Laowang ZAI die ‘Laowang is dying.’. (Smith 1997:272). Semelfactives are instantaneous events and theoretically should not be compatible with the progressive zai; however, (10) is well-formed. The well-formednesss of (10) requires a shifted interpretation  it should be viewed as denoting multiple Semelfactive events, which in turn can be seen as a derived Activity with duration and without endpoints, according to Smith (1997:272). (10) Zhāngsān zài. qiāo. mén.. Zhangsan ZAI knock door ‘Zhangsan is knocking at the door.’. (Smith 1997:272). In sum, progressive zai presents an internal interval of a dynamic, durative, and on-going process of an event. Therefore, Activities, Accomplishments, and Semelfactives (derived Activities) are compatible with zai, while States and Achievements with the progressive zai are ungrammatical.. 2.2.2.2 Imperfective Marker -Zhe and its Interaction with Situation Aspect In this section, the imperfective viewpoint marker -zhe’s aspectual meaning and its interaction with Mandarin situation aspect are reviewed. In Mandarin, the imperfective viewpoint suffix -zhe presents a durative, continuous, and stable situation without regarding to endpoints (Li & Thompson 1981, Smith 1991, 1997, Liu et al. 1996, Lü 1999: 665, Li & Hsieh 2015). Therefore, it is 26.

(39) often called the durative marker (Li 1990, Lee 1996). The durative -zhe can denote a resultant state in that it can focus a resultant interval of a telic event (Li & Thompson 1981:170, Smith 1991, 1997, Liu et al. 1996, Lü 1999:665, Li &Hsieh 2015:33). For instance, (11) gives an example with a. resultant state, indicated by -zhe on the verb. (11) Zhangsan zai. chuang shang tang-zhe. Zhangsan at bed on ‘Zhangsan is lying on the bed’. lie-ZHE. In addition to denoting the resultant state, -zhe can also indicate a progressive viewpoint aspect (Lü 1999:665-6, Li & Hsieh 2015). As in (12), -zhe is a progressive marker, indicating the on-going process of the event tīng-yīnyuè (listen to music). (12) Tā zài fáng lǐ tīng-zhe yīnyuè. he at room in listen-ZHE music ‘He is listening to the music in the room.’. (Li & Hsieh 2015). On the other hand, according to Smith (1997:274), the durative -zhe can also present the internal stages of atelic durative events, as in (13). (13) Quán shìjiè de értóng dōu wán-zhe tóng yàng de yóuxì. whole world DE child all play-ZHE same kind DE game. ‘All the children in the world are playing the same games.’ (Smith 1997:274) In (13), -zhe focuses on an atelic durative event wán tóngyàng de yóuxì (play the same games) without an endpoint. In terms of the durative -zhe, we care about what types of situations it can and cannot occur with and how its distribution can be characterized. Since only durative 27.

(40) situations are relevant with durative -zhe, Activities and Accomplishments, which have duration, can co-occur with -zhe as in (14a-b). Iterative Semelfactives, according to Smith (1997:272), can be seen as a derived Activity with duration. Therefore, they are compatible with -zhe as well, as in (14c). When -zhe co-occurs with Activities, Accomplishments, and iterative Semelfactives, it can emphasize the on-going process (Lü 1999, Li & Hsieh 2015), or it can present a durative view of the event (Lü 1980, 1982). On the contrary, Achievements, which are instantaneous events with no duration, never allow the durative -zhe, as in (14d): (14) a. Tā zài he. fáng lǐ. tīng-zhe. yīnyuè.. at room in listen-ZHE. (Activity). music. ‘He is listening to the music in the room.’. (Li & Hsieh 2015). b. Tā zài fáng lǐ xiě-zhe yī-fēng xìn, zhǔnbèi míngtiān jì he at room in write-ZHE one-CL letter prepare tomorrow send chūqù. (Accomplishment) out ‘He is writing a letter in the room, preparing to send it out tomorrow.’ c. Tā zài ménkǒu qiāo-zhe mén, dànshì dōu méirén huíyīng. he at doorway knock-ZHE door, but all nobody respond (Semelfactive) ‘He is knocking the door in the doorway, but nobody responds.’. d. * Lǎowáng sǐ-zhe. Laowang die-ZHE ‘Laowang is dead.’. (Achievement). According to Yeh (1993), -zhe may also co-occur with stative situations, as illustrated in (15): (15) Wǒmen bǐcǐ shēn ài-zhe, bǐcǐ yīlài-zhe. we each.other deeply love-ZHE, each.other rely-ZHE ‘We deeply loved each other, and relied on each other.’ (Smith 1997: 273) 28.

(41) Carlson (1977) states that stative predicates which denote stable permanent properties are called individual-level predicates, while predicates which denote transitory properties are called stage-level predicates. The stative verbs in (15), i.e., ài ‘love’ and yīlài ‘rely’, according to Smith (1997), are stage-level predicates.2 They are compatible with durative -zhe. However, according to Yeh (1993:86) and Smith (1997:273), -zhe does not co-occur with individual-level predicates, which denote properties of permanence, as illustrated in (16a-b). (16) a.* Tā yīxiàng chéngshí-zhe. he always honest ‘He is always honest.’ b. * Tā zhīdào-zhe zhègè dáàn. he know-ZHE this answer ‘He knows the answer.’. (Smith 1997:273). (Smith 1997:273). To sum up, the imperfective viewpoint -zhe is considered to mark a resultant state, a progressive event, or the duration of an event. Achievements, which are instantaneous, cannot occur with the durative -zhe. Activities, Accomplishments, and iterative Semelfatives (derived Activities) are all durative and compatible with the durative -zhe. However, since States are stative by nature, sometimes there is no need for -zhe to occur on them. As stage-level States are more prone to change than individual-level States (Yeh 1993:86, Smith 1997:274), there are more chances for -zhe to occur on stage-level States.. Different from Smith’s view, we consider ài ‘love’ and yīlài ‘rely’ to be individual-level predicates. This will be discussed in more detail in Section 3.1.1. 29 2.

(42) 2.3 Review Comments on the Classification of Situation Types in Mandarin In this section, the review comments on the classification of situation types in Mandarin are provided. These comments are related to the situation types selected in the questionnaire presented in Chapter Three. As reviewed above, in terms of the classification of situation or verb types, Vendler (1957, 1967) is the first one to classify verbs into four semantic categories (i.e., activity, accomplishment, achievement, and stative verbs). Pertaining to the classification of situation types in Mandarin Chinese, Tai (1984) contends that unlike English, Mandarin doesn’t have accomplishment verbs. He claims that English accomplishment verbs and many achievement verbs are expressed in the form of RVCs in Chinese, and the difference is that an accomplishment verb in English has both the action and result aspects, which makes it able to occur with the progressive viewpoint, while an RVC in Chinese has only the result aspect, which makes RVCs clash with the progressive viewpoint. Tai considers RVCs (e.g., xué-huì ‘learn-know’) and resultative simple verbs (e.g., sǐ ‘die’) as belonging to one single category called ‘result’. Thus, Tai classifies Chinese verbs into three semantic categories: state, activity, and result. Different from Tai’s categorization, Smith (1997) considers RVCs part of Mandarin Accomplishments. She also distinguishes Accomplishments from Achievements in Chinese, and classifies Chinese situation aspect into five categories: 30.

(43) Activities, Accomplishments, Achievements, States, and Semelfactives. Tai (1984), and Smith (1997) present different classifications of Mandarin situation types and some questions thus arise. First, are RVCs a type of Accomplishments, like what Smith suggests, or do they have some specific traits that set them apart from Accomplishments? Second, is it really the case that RVCs and Achievement verbs (or Tai’s resultative simple verbs) can be grouped together as one category (i.e., result), like what Tai suggests? To answer the first question, consider the following sentences of Mandarin RVC and English Accomplishment: (17) a. Zhāngsān Zhangsan. xué-huì-le. zhōngwén.. study-understand-PFV Mandarin. b. Zhangsan learned Mandarin. (18) a. Zhāngsān Zhangsan. shā-sǐ-le. Lǐsì.. kill-die-PFV Lisi. b. Zhangsan killed Lisi. As shown in (17a) and (18a), the underlined RVCs (i.e., xué-huì ‘study-understand’ and shā-sǐ ‘kill-die’) are verbs consisting of two elements, an action phase and a result phase (Li & Thompson 1981, Tai 1984). Similarly, the English Accomplishment is also semantically composed of both a non-detachable process and a result. The Accomplishment ‘learned Mandarin’ in (17b) indicates that Zhangsan has studied Mandarin for a certain period of time (process) and he acquired Mandarin as a result. 31.

(44) Likewise, ‘killed Lisi’ in (18b) conveys that Zhangsan killed Lisi and Lisi died as a result. Therefore, Mandarin RVCs and English Accomplishments are pretty much alike in that they both semantically consist of a process/action and a result, which ends an event with a natural endpoint (Smith 1997). However, Mandarin RVCs and English Accomplishments are different in the specification of the result phase. In Mandarin, the result of the event is specified by the resultative morpheme of the RVC (e.g., huì ‘understand’ in xué-huì ‘study-understand’ and sǐ ‘die’ in shā-sǐ ‘kill-die’), whereas in English, the result specification is inherent in the meaning of the main verb, as in ‘learn Mandarin’ and ‘kill Lisi’. That is to say, Mandarin RVCs express the action-result relation explicitly and overtly (i.e., morphologically combining two elements), while English Accomplishments exhibit the action-result relation implicitly and covertly (i.e., the action-result relation is contained in the main verb) (Tai 1984). Such morphological difference might give rise to the aspectual difference between RVCs and Accomplishments. Consider the following sentences: (19) a. * Zhāngsān zài Zhangsan ZAI. xué-huì. zhōngwén.. study-understand. Mandarin.. b. Zhangsan is learning Mandarin. (20) a. * Zhāngsān zài Zhangsan ZAI. shā-sǐ. lǐsì.. kill-die. Lisi. b. Zhangsan is killing Lisi. Sentences (19a) and (20a) show that Mandarin RVCs are incompatible with the 32.

(45) progressive marker zai, while English Accomplishments are compatible with the progressive V-ing, as shown in (19b) and (20b). Tai (1984) suggests that it is because an Accomplishment verb in English is semantically durative, consisting of both an action and a result aspect, which makes it possible to occur with the progressive aspect, while an RVC in Chinese denotes only the result aspect in their aspectual meaning and behaves like verbs denoting instantaneous events, which makes RVCs clash with the progressive aspect despite the fact that their first constituent is an Activity, such as xué ‘study’ or shā ‘kill’. Based on Tai’s argument, we can answer the first question: Different from Smith’s (1997) view, RVCs are not a type of Accomplishments since their aspectual meanings are different. So far it is clear that RVCs should not be considered a type of Accomplishments. However, is it really the case that Mandarin does not have Accomplishments and that RVCs and Achievements (Tai’s resultative simple verbs) in Mandarin can be grouped together as one category ‘result’, like what Tai (1984:292, 293) suggests? In fact, it is not the case. Mandarin Chinese does distinguish Accomplishments, RVCs, and Achievements. Different from Tai’s view, Mandarin does have Accomplishments and they are often expressed in a verb + argument form denoting a process and a result, instead of in the form of RVCs. According to Smith (1997), Accomplishments are finite events with a natural endpoint and they can be denoted by a verb with a 33.

(46) countable argument (Smith 1997:29). If the argument is countable, the event is finite. Such examples in Mandarin are chī yīkē píngguǒ ‘eat an apple’ (eating as the process, and finishing eating the apple as the result), gài yīdòng fángzǐ ‘build a house’ (building as the process, and the existence of the finished building as the result), and dú yīběn shū ‘read a book’ (reading as the process, finishing reading the book as the result). The [V+NPcount] Accomplishments and RVCs both express an action-result relation, but they are different in that the [V+NPcount] Accomplishments consist of both an action and a result aspects, while an RVC denotes only the result aspect (Tai 1984). That is the reason why RVCs are incompatible with the progressive zai, while [V+NPcount] Accomplishments can co-occur with zai, as shown in (21a) and (21b). (21) a. * Wǒ. zài xiě-wán. gōngkè.. (RVC). I ZAI write-finish homework ‘I am finishing the homework (by writing it).’ b. Wǒ zài kàn yī-běn shū. (Accomplishment) I ZAI read one-CL book ‘I am reading a book.’ Since the difference between [V+NPcount] Accomplishments and RVCs in Mandarin has been clearly shown, the question left to be considered is what distinguishes RVCs from Tai’s resultative simple verbs in Mandarin. Tai’s resultative simple verbs, like Achievements, denote instantaneous and punctual events. According to Tai (1984), RVCs are thought to denote instantaneous events, which only encode the result aspect and are incompatible with the progressive zai (Tai, 34.

(47) 1984:293). That is why Tai (1984) suggests that RVCs and resultative simple verbs can be grouped together as ‘result’. Nevertheless, we consider RVCs and resultative simple verbs to be semantically different. RVCs in Mandarin obviously contain two elements morphologically and semantically, an action and a result, whereas resultative simple verbs denote Achievement events and only contain a result. Thus, different from Tai’s view, we consider that RVCs and resultative simple verbs (Achievements) are different types based on their different natures. Therefore, it is perspicuously shown that Accomplishments ([V+NPcount]), Achievements (resultative simple verbs), and RVCs are actually different from one another in Mandarin. Liu’s (2012) classification of situation type is more well-rounded and detailed among the previous research. Liu identifies ten types of predicates in Mandarin, as reviewed in Section 2.2.1. She not only distinguishes Accomplishment, Achievement, and Result, but further divides Accomplishments into four types: Accomplishment (non-completive),. Accomplishment. (goal),. Accomplishment. (duration),. and. Accomplishment (distance), States into two types: State and State-location, and adds a new type: Change with no result. Nevertheless, there are some problems in Liu’s classification. First of all, Liu’s Accomplishment (duration) type, such as xué liǎngnián zhōngwén ‘learn Chinese for 2 years’, is not a verb-complement construction. According to Smith (1997), though 35.

(48) temporally bounded events seem to be like telic events in having a finite endpoint, they are also unlike telic events because there is no change of state in them. Smith suggests that traversing time and traversing space are different in that when one traverses a certain amount of space, one arrives at a new location (thus, change of state, or result state), but no such result is assumed for time, as illustrated in (22). (22) a. Wǒ. zǒu-dào-le. xuéxiào.. I walk-arrive-PFV school ‘I have walked to school.’ b. Wǒ zǒu-le sān lǐ. lù.. I walk-PFV three mile road ‘I have walked (for) three miles.’ c. Wǒ zǒu-le sān xiǎoshí. I walk-PFV three hour ‘I have walked for three hours.’ The events expressed in (22a-b) have result states (i.e., being at school for (a) and being at a new location three miles from the old location for (b)); but the event expressed in (22c) does not. According to Smith (1997:28), people do not conventionally think of changes of time as arrivals. Thus, it is considered that temporally bounded events are different from telic events. Second, Liu’s Result type can be simply considered as equivalent to RVCs judging from the examples she provides for Result (e.g., xiě-wán ‘to write and finish’ and chī-wán ‘to eat and finish’) since no definition is given by her. It is worth noting that Liu’s Result type is different from Tai’s result type in that Liu’s Result type simply includes RVCs while Tai’s result type includes RVCs and resultative simple 36.

(49) verbs. Third, Liu’s State-location type is considered not to be a verb-complement construction. Liu’s State-location, such as zuò zài dìshàng ‘sit on the ground’, consists of a verb zuò ‘sit’ and an adjunct zài dìshàng ‘on the ground’ which is not a complement of zuò. In fact, the main verbs in Liu’s State-location type are posture verbs. According to Li and Thompson (1981), posture verbs denote body posture or position, such as zuò ‘sit’, zhàn ‘stand’, tǎng ‘lie’, ná ‘take’, bào ‘hold’, pěng ‘hold’. Thus, Liu’s State-location type is actually posture verbs with locative adjuncts. Fourth, Liu’s Accomplishment (goal) type consists of a process and the reaching of a goal as an endpoint or result. However, judging from the examples Liu gives for Accomplishment (goal), such as zǒu-dào xuéxiào ‘walk to school’ and fēi-qù běijīng ‘fly to Beijing’, it is conspicuous that Liu’s Accomplishment (goal) type is actually a type of RVCs. According to Smith (1991, 1997), RVCs fall into two distributional classes  directional RVCs and resultative RVCs, as reviewed in Section 2.2.1. Liu’s Accomplishment (goal) type is in fact Smith’s directional RVCs, which denote ‘direction of action and locate a situation from the point of view of the speaker’ (Smith 1997:282). It is worth noting that Liu’s Accomplishment (non-completive) type is a typical Accomplishment type, like Accomplishments in English, in that it has a verb 37.

(50) constellation with a countable argument [V+NPcount], consisting of a process and a result as the endpoint, such as xiě yī fēng xìn ‘write a letter’. On the other hand, Liu’s Accomplishment (distance) type, such as pǎo yī lǐ lù ‘run 1 mile’, denotes covering a distance in space, and thus can be thought of as having a process (e.g., running) and a result (finishing running one mile and being at a new location one mile away from the old location) as the endpoint. The difference between the two types is that Accomplishment (non-completive) can co-occur with the progressive zai but Accomplishment. (distance). can’t.. Liu. considers. that. Accomplishment. (non-completive) events are not necessarily completed, while Accomplishment (distance) events have necessary completion, as illustrated in (23). (23) a. Zhāngsān xiě-le. yī-fēng xìn,. kěshì méi. xiě-wán.. Zhangsan write-PFV one-CL letter but not write-finish ‘Zhangsan wrote a letter, but didn’t finish it.’ b. *Zhāngsān pǎo-le sān-lǐ lù, kěshì méi pǎo-wán. Zhangsan run-PFV three-mile road but not run-finish ‘Zhangsan ran three miles, but didn’t finish running.’ (Liu 2012) (23a) is acceptable because the Accomplishment (non-completive) event xiě yī-fēng xìn (write a letter) is not necessarily completed, whereas (23b) is unacceptable since the Accomplishment (distance) event pǎo sān-lǐ lù (run three miles) denotes necessary completion (Liu 2012). The discussion above indicates that different from Tai’s (1984) and Smith’s (1991, 1997) point of view, Accomplishments, RVCs, and Achievements should be 38.