中文否定詞「不」與「沒」之第一語言習得

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. 中文否定詞「不」與「沒」之第一語言習得 First Language Acquisition of Bu and Mei in Mandarin Chinese. 指導教授: 陳純音博士 Advisor: Dr. Chun-yin Doris Chen 研究生: 陳淑芬 Student: Shu-fen Amy Chen. 中華民國一百零七年七月 July 2018.

(2) 摘要本研究旨在探討以中文為母語之兒童對於否定詞「不」和「沒」詮釋的發展。本研究主要包含兩階段的詮釋測驗，兩個測驗皆為接受度測驗:第一階段測驗兒童對於「不」和「沒」在單一否定句及雙重否定句與不同動詞搭配時的接受程度之差異；第二階段觀察語境對於兒童詮釋「不」和「沒」在單一否定句及雙重否定句之影響。兩測驗皆探討年齡效應。研究對象包含實驗組的七十二位兒童，依其就讀年級分為四組:幼兒園大班、小學二年級、四年級、六年級，每組各十八位，以及對照組的成人十八位。研究結果顯示，對於「不」和「沒」與動詞的搭配，以及否定及雙重否定之習得隨著年齡增長，接受度逐漸與成人趨向一致。在動詞搭配上，對於兒童而言，動作動詞與「不」和「沒」搭配的接受度最高。幼兒園及二年級兒童尚未具備成人母語者的動詞與否定詞搭配之能力，但四年級兒童已可正確搭配動詞與否定詞「不」和「沒」，與成人組相同，皆對於「不」與有界動詞和「沒」與無界動詞的搭配度有較高的接受度。關於單一否定句及雙重否定句，只有六年級兒童，因已進入形式運思期，與成人組相同，皆對於「不」於單一否定句及「沒」於雙重否定具有較高的接受度，較能進行邏輯及抽象思考。最後，語境效應本研究中呈現正面效應，受試者在提供語境的測驗時，對否定句子的接受度普遍皆有提高。. 關鍵詞: 否定句、「不」、「沒」、動詞類別、第一語言習得. i.

(3) ABSTRACT The present study aims at investigating children’s acceptability of bu and mei in negative sentences. Two tasks were designed to elicit the subjects’ interpretations. In the first task, the children’ acceptability rates of bu an mei in negatives in isolation were obtained, and in the second task, context was provided to examine whether it would increase children’s acceptability rates of negatives. Age differences were also taken into consideration in the two tasks. A total of seventy-two child subjects were recruited from an elementary school in New Taipei City and they were further divided into four experimental groups (i.e., the kindergartners, Grade 2, Grade 4 and Grade 6). In addition, eighteen adults were recruited as the control group. The results showed there were age differences in the acquisition of bu and mei in negative sentences. It was found that the kindergartners and Grade 2 did not show an adult-like tendency of preference for bu and mei with verb types. However, Grade 4 started to show the adult-like tendency, and they all preferred bu to occur with atelic verbs, mei to accompy telic verbs. As for bu and mei in single and double negative sentences, the results indicated that our subjects had a stronger adult-like preference for bu and mei in negatives when they reached Grade 6, a stage where they could do logical reasoning. Finally, the subjects’ acceptability rates increased when context was provided. Keywords: negatives, bu, mei, verb types, first language acquisition. ii.

(4) ACKNOWLEDGEMENTS I cannot believe that I have finished my master’s thesis and I am glad that I have a chance to express my gratitude to those who have assisted and encouraged me during my life at graduate school. First and foremost, I would like to express my deepest appreciation to my advisor, Dr. Chun-yin Doris Chen, who has helped me in every way she could and spent lots of time, even on weekends, discussing my research with me. In addition to providing me useful comments on my thesis, she also gave me some constructive advice when I faced difficulties. Furthermore, I would also like to express my gratitude to my committee members: Dr. Jyun-Gwang Fred Chen and Dr. Rueih-Lirng Sharon Fahn. Their insightful suggestions helped me enhance the quality of my thesis. They are the very diligent committee members. I am also very grateful to many teachers, especially Ms. Yu-Mei May Zou, who helped me recruited subjects in her school. Without her assistance, I could not recruit subjects and arrange appointments with teachers successfully. Additionally, I would like to thank the teachers who taught me in the linguistic track: Dr. His-Yao Su, Dr. Hsiao-Hung Iris Wu, Dr. Hui-Shan Nissa Lin, Dr. Jen-I Li, Dr. Jen Ting, Dr. Jing-Lan Joy Wu, Dr, Miao-Hsia Tammy Chang, and Dr. Shiao-hui Chan, by alphabetical order. Their professions introduced me to the world of linguistics. In addition to teachers who helped me during the period of graduate school, I would also like to show my sincere gratitude to my dear friends and classmates at NTNU: Andrew Syue, iii.

(5) Eileen Lin, Eliza Cheng, Ka-Mei Chan, Kabby Chen, Ke-Fen Chen, Louisa Zheng, Nick Hu, Shawn Lin, and Yu-Chen Lin, listed in alphabetical order. It was fortunate that I could meet them and have their warm company. Special thanks go to Eileen Lin, Eliza Cheng, Ka-Mei Chan, Kabby Chen, Ke-Fen Chen, Louisa Zheng, Shawn Lin, and Yu-Chen Lin, who helped me to conduct the tasks and provided me suggestions about the test items. Last but not least, I would like to express my gratitude to my family. I could not finish my master’s degree without their support and encouragement. Also, I would like to thank them for loving me and supporting me no matter what decision I made.. iv.

(6) TABLE OF CONTENTS CHINESE ABSTRACT ........................................................................................................................... i ENGLISH ABSTRACT.......................................................................................................................... ii ACKNOWLEDGEMENTS ................................................................................................................. iii TABLE OF CONTENTS ........................................................................................................................ v LIST OF TABLES ............................................................................................................................. viii LIST OF FIGURES ............................................................................................................................. ix. Chapter One Introduction ....................................................................................................................... 1 1.1 Motivation ..................................................................................................................................... 1 1.2 Research Questions ....................................................................................................................... 4 1.3 Significance of the Study .............................................................................................................. 5 1.4 Organization of the Thesis ............................................................................................................ 6. Chapter Two Literature Review .............................................................................................................. 7 2.1 Theoretical Studies of Single Negation and Double Negation ..................................................... 7 2.1.1 Teng (1974)............................................................................................................................ 8 2.1.2 Li and Thompson (1981) ..................................................................................................... 10 2.1.3 Fan and Xiao (2010) ............................................................................................................ 13 2.1.4 Sun (2011)............................................................................................................................ 15 2.1.5 A Summary of Theoretical Studies of Single Negation and Double Negation .................... 17 2.2 Empirical Studies of Single Negation and Double Negation .................................................... 18 2.2.1 Jou (1988) ............................................................................................................................ 19 2.2.2 Rumain (1988) ..................................................................................................................... 21 2.2.3 Zhou et al. (2014a)............................................................................................................... 23 2.2.4 Zhou et al. (2014b) .............................................................................................................. 27 2.2.5 A Summary of Empirical Studies of Single Negation and Double Negation ...................... 31 v.

(7) 2.3 Summary of Chapter Two ......................................................................................................... 33 Chapter Three Research Design .......................................................................................................... 35 3.1 Subjects ..................................................................................................................................... 35 3.2 Methods and Materials .............................................................................................................. 37 3.3 Procedures ................................................................................................................................. 45 3.3.1 Pilot Study .......................................................................................................................... 45 3.3.2 Formal Study ...................................................................................................................... 46 3.3.3 Scoring ............................................................................................................................... 47 3.4 Summary of Chapter Three ....................................................................................................... 48 Chapter Four Results and Discussion ................................................................................................. 49 4.1 Verb Type Effects ..................................................................................................................... 49 4.1.1 Overall Findings ................................................................................................................ 49 4.1.2 General Discussion ............................................................................................................ 53 4.2 Single Negatives and Double Negatives ................................................................................... 55 4.2.1 Overall Findings ................................................................................................................ 56 4.2.2 General Discussion ............................................................................................................ 59 4.3 Contextual Effects ..................................................................................................................... 62 4.3.1 Overall Findings ................................................................................................................ 63 4.3.2 General Discussion ............................................................................................................ 66 4.4 Age Differences ........................................................................................................................ 69 4.5 Summary of Chapter Four ......................................................................................................... 71 Chapter Five Conclusion ..................................................................................................................... 72 5.1 Summary of the Major Findings ............................................................................................... 72 5.2 Limitations of the Present Study and Suggestions for Future Research ................................... 73. Bibliography ....................................................................................................................................... 75 Appendix A: Test items of the Context-free Task in the Formal Study ............................................. 77 vi.

(8) Appendix B: Test items of the Context-provided Task in the Formal Study ...................................... 86 Appendix C: Test items of the Context-free Task in the Pilot Study ................................................ 103 Appendix D: Test items of the Context-provided Task in the Pilot Study ....................................... 105 Appendix E: Consent Form .............................................................................................................. 109. vii.

(9) LIST OF TABLES Table 1-1 A Classification of Verb Types (Vendler 1957) ..................................................................... 3 Table 2-1 A Summary of the Previous Empirical Studies .................................................................. 31 Table 3-1 Background of Subjects ...................................................................................................... 36 Table 3-2 A Distribution of Test Items in the Two Tasks .................................................................. 42 Table 3-3 A Sample of Questions in a Context-free Acceptability Judgment Task ........................... 43 Table 3-4 A Sample of Questions in a Context-provided Acceptability Judgment Task ..................... 44 Table 4-1 Subjects’ Overall Acceptability Rates of bu and mei with Different Verb Types .............. 50 Table 4-2 Subjects’ Overall Acceptability Rates of bu and mei in Single and Double Negatives ....... 56 Table 4-3 The p-value for the within-group Differences between the Acceptability Rates of bu in Single and Double Negative Sentences ............................................................................... 57 Table 4-4 The p-value for the within-group Differences between the Acceptability Rates of mei in Single and Double Negative Sentences ............................................................................... 58 Table 4-5 Subjects’ Overall Acceptability Rates of bu and mei in the Context-free and Context-provided Tasks ...................................................................................................... 63 Table 4-6 The p-value for the within-group Differences between the Acceptability Rates of bu and mei in the Context-free and Context-provided Tasks ......................................................... 65 Table 4-7 Children’s Acquisition of bu and mei in Negative Sentences ............................................ 69. viii.

(10) LIST OF FIGURES Figure 4-1 Subjects’ Overall Acceptability Rates of Negators with Different Verb Types.................. 50 Figure 4-2 Each Group’s Acceptability Rating for Verb Types with bu ............................................ 51 Figure 4-3 Each Group’s Acceptability Rating for Verb Types with mei .......................................... 52 Figure 4-4 Subjects’ Overall Acceptability Rates of Single and Double Negatives ........................... 56 Figure 4-5 Each group’s Acceptability Rates of bu in Single and Double Negatives ........................ 57 Figure 4-6 Each group’s Acceptability Rates of mei in Single and Double Negatives......................... 58 Figure 4-7 Subjects’ Acceptability Rates of Negative Sentences in the Context-free and Context-provided Tasks .................................................................................................. 63 Figure 4-8 Each Group’s Acceptability Rates of bu in the Context-free and Context-provided Tasks. ........................................................................................................................................... 64 Figure 4-9 Each Group’s Acceptability Rates of mei in the Context-free and Context-provided Tasks.. ........................................................................................................................................... 64. ix.

(11) Chapter One Introduction 1.1 Motivation A negative sentence expression is very common in every language use since negation is a basic concept in logic (Nordmeyer and Frank 2013, Zhou et al. 2014b). It is used to reverse the truth value of a statement. To make it clear, consider sentence (1).. (1) Xiaohua bu. chi. haixian.. Xiaohua not eat seafood „Xiaohua doesn‟t eat seafood.‟. The verb phrase chi haixian „eat seafood‟ is in the scope of the negator bu „not‟; therefore, bu denies the positive feature of chi haixian and results in a negative meaning. A negator not only can take a scope of a verb, it also can take a scope over non-verbal elements like adverbials and manner expressions (Ernst 1995, Jackendoff 1969). Thus, the scope of a negator determines the meaning of sentences (Ernst 1995, Jackendoff 1969). As for double negative sentences, the first negator takes scope over the second negator, as in (2).. (2) Xiaohua bu hui bu Xiaohua not will not „Xiaohua will come.‟. lai. come. In (2), the second negator bu denies the verb phrase bu lai „not come‟ and generates an affirmative meaning. 1.

(12) Carpenter et al. (1999) claim that negative forms take more time to process than affirmative forms. In the acquisition of a language, it is essential to know the relationship between language forms and their corresponding meanings. This relationship is related to iconic quantity; Iconic quantity1, which states that the length and complexity of an expression influence the complexity of cognitive perspective (Humain 1983, Sun 2011). Therefore, double negative sentences are less preferred than positive sentences. From a pragmatic perspective, previous research claims that a double negative is still used to express a positive meaning but it makes the positive meaning weaker than an affirmative sentence. Children‟s acquisition of single negatives and double negatives has been discussed in previous studies (Jou 1988, Zhou et al. 2014a, 2014b). One of the popular issues in first language acquisition is age. The development of negation has been studied by previous researchers (Bloom 1970, Pea 1980, Nordmeyer & Frank 2013). Pea (1980) found that English-native children, as early as 12 month-olds, produced negative expressions for nonexistence and rejection2. However, children were found to have difficulty with negative terms such as less when they were 4 years old (Klatzky, Clark & Macken 1973). Jou (1988) and Zhou et al. (2014b) explored an age effect on Chinese children‟s acquisition of double negatives; nevertheless, they were not in agreement on when children acquire double. 1. Iconic quality is one of the concepts of iconicity, which includes iconic quality, iconic sequencing and iconic proximity. 2 Children might express nonexistence by saying “no more juice,” and convey rejection by saying “no go outside” (Bloom 1970, Pea 1980). 2.

(13) negative sentences. In addition to the age issue, Zhou et al. (2014a) investigated single negative sentences consisting of different lexical aspectual clauses of verbs, mainly following Vendler (1957), who classified verbs into four types: state, activity, achievement and accomplishment. The properties and examples of the four verb types are exhibited as in Table 1-1.. Table 1-1 A Classification of Verb Types (Vendler 1957) Verb Types. Properties. Examples. State. -endpoint, +durative, -dynamic. love, know, believe. Activity. - endpoint, +durative, +dynamic. run, swim, pull something. Achievement. + endpoint, -durative, +dynamic. win, recognize, reach, die. Accomplishment. + endpoint, +durative, +dynamic. paint a picture, build a house. Although Zhou et al. (2014a) discussed lexical aspect with single negatives, they did not investigate single negative sentences with activity verbs, such as Wo bu/mei chi yao „I didn‟t/wouldn‟t take the medicine.‟ Moreover, the influence of verb types on double negatives was not discussed. Rumain (1988) discussed how English-speaking subjects processed negative sentences and found that two cues (i.e., an article cue3 and a position cue4) helped the subjects process negatives. It is possible that some clues may be helpful for the comprehension of single. 3. An article cue means that a definite article and an indefinite article in a negative sentence help the subjects determine which element is negated (i.e., „Mary is not washing the car.‟). Since „the car‟ is a definite noun, which is not new information, it is taken outside the focus of the negation. Thus, this sentence may be interpreted as „Mary is not washing the car; she is waxing the car.‟ 4 A position cue means negators may negate different elements due to its position (i.e., „N1 Neg V N2.) Neg probably negates V, N2, or VN2 . 3.

(14) negatives and double negatives in Mandarin Chinese. Some empirical studies considered context as an important clue for acquisition (i.e., de Villiers & Flusberg 1975, Rumain 1988, Glenberg et al. 1999, Zhou et al. 2014b). For instance, deVilliers and Flusberg found that negatives were not difficult for English children to process after providing them with contexts. Therefore, the present study has taken context into account. The present study has also explored children‟s first language acquisition of bu and mei, in Mandarin Chinese and at what age they would acquire single negatives and double negatives. In order to fill the gap of previous research, the effects of context, verb types and age have been considered in this study. The aims of the present study are as follows. The first aim is to examine at what age children acquire single negative sentences and double negative sentences in Mandarin Chinese, and at what age they are aware of the relation between verb types and negators. The second is to discuss whether complexity of single negative sentences and double negative sentences will influence children‟s comprehension and whether context is helpful to determine meanings of single negatives and double negatives. 1.2 Research Questions Based on the aforementioned discussion, the research questions of the present study are as follows: 1.. Can Chinese children distinguish aspectual clauses of verbs with bu from those with mei? 4.

(15) 2.. Are Chinese children aware that the compatibility of negators and lexical aspect in double negatives is different from that in single negatives?. 3.. Can the comprehension of single negative sentences and double negative sentences be enhanced when appropriate contexts are provided?. 4.. At what age do children interpret negative sentences and double negative sentences as adults do?. 1.3 Significance of the Study Single negatives and double negatives in Mandarin Chinese have drawn numerous attentions and have been analyzed from both theoretical and empirical perspectives (Teng 1974, Jou 1988, Rumain 1988, Li & Thompson 1981, Fan & Xiao 2010, Sun 2011, Zhou et al. 2014a, 2014b). Even though theoretical issues concerning single negatives and double negatives have been widely discussed, few studies have investigated the relation between negators and verb types in double negative sentences. Children may adopt some methods to figure out the meanings of single negatives and double negatives and to judge whether the collocation of negators and verb types is possible. A contextual cue may be one of the methods they employ. However, there is a limitation of previous studies (Zhou et al. 2014a, 2014b). Context provided in the tasks was lengthy and the child subjects had processing difficulty due to limited working memory capacity. To sum up, the present study will investigate the relation between negators and verb 5.

(16) types in both single negatives and double negatives. Additionally, this study explores whether context can help children comprehend the meanings of single negatives and double negatives, and better recognize the collocation of verb types and negators.. 1.4 Organization of the Thesis This thesis is organized as follows. Chapter Two reviews some theoretical studies, including syntactic and semantic properties of single negatives and double negatives. Empirical studies on first language acquisition of single negative sentences and double negative sentences are also reviewed in this chapter. Chapter Three presents the research design of the present study, including the background of the subjects, the procedure of experiments, and the expected findings. Chapter Four reports the results and the analysis of the children‟s acquisition of bu and mei in negative sentences. Finally, Chapter Five provides the major findings of the present study and some suggestions for future research.. 6.

(17) Chapter Two Literature Review In this chapter, some theoretical studies and empirical studies of single negative sentences and double negative sentences are explored. Section 2.1 reviews characteristics of Chinese negators bu and mei, syntactic features of single negatives and semantic properties of double negatives. In Section 2.2, studies of first language acquisition of double negative sentences and single negative sentences are discussed. Finally, a summary of this chapter is provided in Section 2.3. 2.1 Theoretical Studies of Single Negation and Double Negation Negative sentences are commonly used in every language (Li and Thompson 1989). When two negators occur in a sentence, they cancel the negative meanings of each other and generate an affirmative meaning. There are many negators in Mandarin Chinese whose scope determines the interpretation of sentences. Some previous studies on semantic features and syntactic structures of bu and mei (Teng 1974, Li and Thompson 1989) are reviewed since the two negators are most frequently used in Mandarin Chinese. In addition, the characteristics of double negative sentences are also reviewed in this chapter (Fan and Xiao 2010, Sun 2011).. 7.

(18) 2.1.1 Teng (1974) Teng (1974) focuses on single negative sentences and further investigates the scope of negators in Mandarin Chinese. The difference between sentence negation and constituent negation can be distinguished by a position of a negation. For example, a negator can be placed on the left, as shown in (1a), or on the right of a modal or an adverb, as shown in (1b).. (1) a. Ta bu chang lai. s/he not often come „He doesn‟t come often.‟ b. Ta chang bu lai. s/he often not come „Often he doesn‟t come.‟ (Teng 1974:132) The negator in (1a) does not actually negate the verbal element, but it negates the entire S. The tree structures of (1a) and (1b), cited from Teng (1974: 133-134) are illustrated below. (2a). (2b). 8.

(19) In (2a), NEG immediately c-commands1 chang „often,‟ while the NEG in (2b) does not. The negator in (2b) negates the embedded S. Teng argues that two transformational rules are operated: NEG-lowering and predicate-lowering. NEG lowers to the left of the lower VP; therefore, it forms bu chang „not often‟ in (2a) and bu lai la „not come‟ in (2b). After this step, the predicate-lowering rule lowers the highest VP to combine with the embedded verb phrase. Then, bu chang lai in (2a) and chang bu lai in (2b) are generated. Sometimes, NEG-raising occurs when the NEG in the surface form is the higher predicate, as in (3).. (3) Wo jiu-shi mei zai ni jia chi-guo I certainly not in you home eat-Asp. wu-fan. lunch. a. „I have eaten lunch at other people‟s house, but not at yours.‟ b. „I have eaten other meals in your house, but not lunch.‟ (Teng 1974:139). Two interpretations can be derived from (3). First, it means “I have eaten lunch at other people‟s house, but not at yours.” Second, it can also be interpreted as “I have eaten other meals at your house, but not lunch.” Teng mentions that when NEG is raised to the higher VP and lower to the locative predicate, the second interpretation can be derived, as illustrated in (4), which is taken from Teng (1974:139).. 1. The term c-command was first seen in the dissertation of Reinhart (1976). The definition is shown as follows: A c-commands B if (1)the first branching node which dominates A also dominates B, and (2) A doesn‟t dominate B and B doesn‟t dominate A. 9.

(20) (4). Moreover, negators can be dominated by or dominate a modal at the same time to form a double negative sentence, as in (5).. (5) Ta bu hui bu shuo. s/he not have not speak „It won‟t be the case that he won‟t talk.‟ (Teng 1974: 135). To sum up, Teng claims that the scope of negation influences the interpretation of negative sentences. However, he only focuses on structures of single negative sentences. The structures and the transformational rules of double negative sentences are not discussed in his study. 2.1.2 Li and Thompson (1981) Li and Thompson (1981) investigate meanings and functions of negators in Mandarin Chinese, which consists of four frequently used negative forms: bu, mei, bie, and meiyou. The most common negative form is bu, as in (6). 10.

(21) (6) Wo bu jide ta. I not remember him „I don‟t remember him/her.‟ (Li & Thompson 1981:415). However, bu cannot be used without the modal yao „want‟ in imperative sentences, such as (7), whereas only bie can be used.. (7) *Bu/ Bie guan men! not close door „Don‟t close the door.‟ (Li & Thompson 1981: 416). The other negative form is mei, which is a abbreviation of meiyou „not have.‟ Since you „have‟ can be omitted, mei contains both negative and existential meanings. In addition to negative forms, Li and Thompson (1981) discuss the position of negators. Generally speaking, a negator often follows a subject and precedes a verb phrase. Some examples taken from Li and Thompson (1981:417-418) are shown below.. (8) Ta mei (you) kai men. 2 s/he not open door „S/He didn‟t open the door.‟ (9) Ta bu nian shu. s/he not study book „S/He wouldn‟t study.‟ Moreover, the most frequent usage of a negator is preceding the auxiliary neng „can.‟ However, the reverse position is also acceptable for a negator. To illustrate, examples taken. 2. In Li and Thompson (1981), ta is translated as 3sg. 11.

(22) from Li and Thompson (1981:421) are shown below:. (10) a. Wo bu neng qu. I not can go „I cannot go‟. b. Wo neng bu qu. I can not go „I am allowed not to go.‟ Interestingly, the scope of negation is determined by the position of negators, as shown in (10a) and (10b), which are interpreted differently. The meaning of (10a) Wo bu neng qu is „I cannot go,‟ whereas Wo negng bu qu can be interpreted as „I am allowed not to go.‟ Besides discussing the possible positions for the negator, Li and Thompson explore the functions of bu and mei(you). They (1981) argue that bu denies the existence of the state, while mei(you) denies the completion of the action, as shown in the following examples taken from Li and Thompson (1981:422). (11) a. Ta bu s/he not „S/He { b. *Ta s/he. congming. intelligent. is } not inelligent.‟ was mei(you) not. congming. intelligent. (11a) and (11b) show that mei(you) is incompatible with the stative adjective congming „intelligent,‟ whereas bu can co-occur with it. To sum up, Li and Thompson mention that negators in Mandarin have different meanings. The position of the negator depends on what it negates. In other words, the 12.

(23) element that is negated by the negator will be within the scope of the negator. Nevertheless, Li and Thompson‟s study has some limitations. First, they only describe the usage of negation, but they did not analyze the syntactic structure of negation in detail. Second, they mention that „ta mei(you) zai shui jiao 3‟ is ungrammatical in Mandarin Chinese. However, this sentence is acceptable for some native speakers. 2.1.3 Fan and Xiao (2010) Fan and Xiao (2010) investigate the derivational process to generate meanings of double negative sentences in different syntactic forms. In general terms, the concept of negation indicates that if the constituents can express negative meanings, they are all negative forms. A double negative sentence is a structure which contains two negative elements. Nevertheless, not all sentences which consist of two negative elements are double negative sentences. (12) Buyong-de shu juangei needless book donate. meiqian-de poor. haizi. children. „We can donate needless books to needy children.‟. (Fan and Xiao 2010: 78). (12) is not a double negative sentence since the affirmative meaning does not result from the two negative elements. To avoid this misinterpretation, Fan and Xiao suggest that a double negative sentence should be defined as follows: the affirmative meaning of a double negative structure is derived from one negator negates the negative meaning of the other negator.. 3. Li and Thompson (1981) mention that zai is a durative-action aspect marker which cannot occur with mei(you). Only bu can negate an ongoing activity. 13.

(24) As for the structure of double negative sentences, Fan and Xiao (2010) claim two negations cannot negate the same object in Chinese, as shown in (13):. (13) Meiyou bu he shui. not-have not drink water „Someone drinks some water.‟ (Fan and Xiao 2010: 78). In (13), meiyou „not-have‟ negates bu he shui „not drink water,‟ not meiyou bu „not –have not‟ negates heshui „drink water.‟ However, when an auxiliary occurs between two negators, as in (14), the derivational process of this sentence is more complicated, as shown in (15).. (14) Bu keneng bu lai. not may not come „It is not a case, I won‟t come.‟ (Fan and Xiao 2010: 79). (15) Bu[keneng[bu lai]] = bu [bu [yiding lai]] = yiding lai. Fan and Xiao consider that the interpretation of “keneng bu + VP” is equivalent to the negative meaning of “bu yiding + VP.” As (15) exhibits, the first bu dominates the second bu; therefore, two negators cancel each other and generate the affirmative meaning “yiding lai”. The other structure of a double negative sentence begins with bushi „not‟ as in (16a) and (16b), which are taken from Fan and Xiao (2010: 81).. 14.

(25) (16) a. Bu shi [wo bu lai]. not beV I not come „I have come.‟ b. [Bu shi wo] bu lai. not beV I not come „It was not I that did want to come.‟ Bushi in (16a) negates the whole sentence „wo bu lai‟, while bushi in (16b) functions as a focus marker. This type of double negative structures „bushi…bu…‟ conveys a negative meaning in context-free situations. However, sentences (16a) and (16b) express volition when contexts are provided (e.g. (17a) and (17b)).. (17) a. Ta bushi. bu lai,. ershi jiali. you. keren. s/he not not come, but home have guests „It is not that he won‟t come, but he has guests at home.‟ b. Wo bushi bu zancheng ni qu, shi bu tongyi ni qu I not not approve you go, beV not agree you go. „It is not that I don‟t allow, but I disapprove you to go.‟ (Fan and Xiao 2010:81). To sum up, the affirmative meaning of a double negative sentence is derived through the derivational process, as mentioned in (15) above. Fan and Xiao describe the procedure to generate meanings of double negative sentences. However, no empirical study is presented to support their argument. 2.1.4 Sun (2011) Sun (2011) analyzes double negative sentences from a cognitive perspective. The grammatical form is related to the semantic structure of language. Sun argues that the concept 15.

(26) of affirmation is not against the concept of negation. There are also some intermediate terms between them as some antonyms do. The intermediate terms between a negative sentence and a positive sentence can be expressed by using a double negative structure. Sun (2011) analyzes double negative sentences based on the theory of iconicity and compares double negative sentences of English and Chinese. To illustrate, consider sentences in (18) and (19), taken from Sun (2011:1444).. (18) a. Yiqie wei. renshizhe dou bu-neng bu zhuyi. jiaoshoufa. de. wenti.. every be teacher all not can not attention teaching of problem „Every teacher should pay attention on the problem of their teaching methods.‟ b. Yiqie wei renshizhe dou yao zhuyi jiaoshoufa de every be teacher all-have attention teaching method of wenti. problem „Every teacher should pay attention on the problem of their teaching methods.‟ (19) a. He did not deny that he had made mistakes. b. He admitted that he had made mistakes.. These examples show that double negative sentences are more positive than affirmative ones semantically, and the former is stronger in tone than the latter. On the other hand, both sentences in (20) express that the speaker has tried to help, but that there is no method available. A double negative sentence is weaker than an affirmative one semantically when it carries an indirect and euphemistic meaning, as shown in (20a-b), which are taken from Sun (2011: 1444). 16.

(27) (20) a. Bushi women bu xiang gei xianqinmen xiaolao, not we not want give fellowmen service shizai shi meiyou fazi. actually be not-have method „We really want to help fellowmen, but actually there is no method.‟ b. Women xiang we want shizai shi. gei xianqinmen xiaolao, give fellowmen service meiyou fazi.. actually be not-have method „We really want to help fellowmen, but actually there is no method.‟. The other differences between affirmative sentences and double negative sentences are the complexity and length. Based on iconic quantity, a more complicated sentence conveys a stronger meaning. That is, double negative sentences express more affirmative meanings than affirmative sentences do. To sum up, Sun (2011) claims that both affirmative sentences and double negative sentences carry positive meanings; nonetheless, double negative sentences can strengthen the degree of affirmation and it is not redundant in languages. However, no empirical evidence is provided to support that double negative sentences express stronger meanings than affirmative sentences. 2.1.5 A Summary of Theoretical Studies of Single Negation and Double Negation Previous studies have investigated syntactic features and functions of single negation and double negation in Mandarin Chinese. The scope of negation determines the meanings of negative sentences (Teng 1974, Li & Thompson 1981). For example, if a negator precedes a main verb, the negator negates the affirmative feature of the verb and forms a negative 17.

(28) sentence. In addition to the scope of negation, Li and Thompson (1981) investigate meanings and functions of negators in Mandarin Chinese. Sentences may convey different meanings with different negators. Besides, the interaction between aspects and negators in Mandarin Chinese also varies. Mei(you) cannot co-occur with –le and zai, while bu cannot co-occur with –guo. A double negative sentence is also studied by some previous researchers (Sun 2011) who argue that double negative sentences are stronger and more positive than affirmative ones. Moreover, a double negative sentence may carry an indirect and euphemistic meaning as well. Fan and Xiao (2010) investigate the process to generate meanings of double negative sentences in different syntactic forms and suggest that their meanings cannot be determined literally or from the surface form. However, no research has been focused to acquisition of single negatives and double negatives. In the next section, some empirical studies are reviewed. 2.2 Empirical Studies of Single Negation and Double Negation In this section, four empirical studies of first language acquisition on single and double negative sentences are reviewed. The development of children‟s acquisition of single negative and double negative sentences in Mandarin and English are presented (Jou, 1988, Rumain 1988, Zhou et al. 2014a, 2014b).. 18.

(29) 2.2.1 Jou (1988) Jou (1988) investigated Chinese children‟s comprehension of double negation. The aim of his study was twofold: (1) to find out when children can correctly interpret a double negative expression , and (2) to examine the development process of double negation of children. The participants were divided into eleven age groups, from age 4 to age 14. Each age group consisted of ten subjects, who were asked to act out the actions with dolls according to the sentences they heard. No context was provided to help the participants understand the sentences. The test sentences were further divided into three sets, including affirmative sentences, negative sentences, and double negative sentences. In each set, eight types of transformations were applied: (1) basic, (2) passivization, (3) subject topicalization, (4) object topicalization, (5) embedding, (6) passivization and subject topicalization, (7) embedding and object topicalization and (8) passivization and embedding. Below are examples of double negative passivization, double negative subject topicalization, and double negative object toplication, which are taken from Jou (1988:470). (21) Double Negative Passivization: Huong wawa meiyou bu bei lu wawa tuei dau. red doll doesn‟t not by green wawa struck down „It is not that the red doll isn‟t struck down the green doll.‟ (22) Double Negative Subject Topicalization: Huong wawa, ta meiyou bu ba lu wawa tuei dau. red doll, she doesn‟t not cause green doll struck down „It is not that the red doll. She doesn‟t strike down the green doll.‟ 19.

(30) (23) Double Negative Object Topicalization: Huong wawa, lu wawa meiyou bu ba ta tuei dau red doll, green doll doesn‟t not cause her struck down „It is not that the red doll. The green doll doesn‟t strike her down.‟. The results are as follows. First, it was found that sentence type was one of the effects that influenced children to interpret meanings of sentences. Children‟s performance on single negatives and double negatives showed great differences. Second, as for the age effect, only double negative structures showed significant differences. The result indicated that the younger subjects misinterpreted double negative structures to be equivalent to single negatives. By contrast, the older subjects interpreted double negative sentences correctly. Last, the subjects were divided into three groups according to the development of comprehension on double negative sentences. The subjects aged less than 7 years old all considered a double negative sentence to be the same as a single negative sentence. This group of subjects was assigned to Stage 1 in the development of comprehension of double negative sentences. The feature of the participants at Stage 2 was that these children were able to recognize the difference between single negatives and double negatives. However, they could not decode the meanings of double negative sentences. Children, aged 8, achieved Stage 3 which indicated that they could interpret the meaning of a double negative sentence as an affirmative sentence, and mastered the usage of this type of sentence. To conclude, Jou found that children were capable to differentiate single negative sentences from double negative sentences at a later age. The transition from Stage1 to Stage 2 20.

(31) revealed that the knowledge of double negative structures began to be constructed. The transition from Stage 2 to Stage 3 indicated that the concept of double negation was well-developed. However, contextual effects are not taken into consideration. Furthermore, the subject pool of each age group was small in Jou‟s study. 2.2.2 Rumain (1988) Rumain (1988) examined how children comprehended negative sentences and argued that there were two factors influencing interpretations of negative sentences. First is the article effect, as shown in (24a) and (24b), taken from Rumain (1988: 120).. (24) a. Ann is not renting the van. b. Ann is not renting a van.. The negative operator in (24b) focuses on a van, which is the new information of the sentence, and (24b) is interpreted as “Ann isn‟t renting a van; she‟s renting a car.” The negative operator is applied to a van. Nevertheless, sentence (24a) means “Ann isn‟t renting the van; she owns it”. The negative operator in (24a) negates the truth value of the verb rent. The second cue influencing the interpretation of a negative sentence is the position effect. That is, the position of nouns will affect the meaning of sentences. According to Jackendoff (1969), the scope for the negative operator is what follows it. For instance, N1 Neg N2 V. The negative operator applies to N2 , to V or to VN2 (i.e., if the negator in „Ann is not renting a van,‟ but that applies to VN2, the meaning of the sentence might be that Ann is not renting a 21.

(32) van; she is doing her income tax.) In terms of these two factors, the aim of her study was to investigate whether children and adults would use some cues to determine which constituent of a sentence that a negative operator focuses on. Also, the development of the knowledge of negation was discussed. Three experiments were conducted. In Experiment One, 20 children with a mean of 7 years old, 28 children with a mean of 10 years old, and 20 adults were recruited. Sixteen test sentences were presented in four forms, The-a (Active), The-a (Passive), A-the (Active) and A-the (Passive)4. These test sentences were read by the experimenter and the subjects only had to choose the suitable picture for the sentence. The result of Experiment One revealed that there was a significant article effect and an age effect. The interaction between age and article indicated that the dependency of article effect increased with age. A position cue was stronger than an article cue for 10-year-olds. The significant difference between 10-year-olds and adults was that 10-year-old children depended more on a position cue to understand meanings of negative sentences while the adults relied more on articles. On the other hand, the difference between 7-year-olds and 10-year-olds was that the former depended less on a position cue. Experiment Two included the same test sentences as those in Experiment One. The only. 4. Below are sentence structures used in Rumain (1988:124). (i) The-a (Active): The N1 isn‟t V-ing a N2 (e.g. The dog isn‟t carrying a frog.) (ii)The-a (Passive): The N1 isn‟t being V-ed by a N2 (e.g. The frog isn‟t being carried by a dog.) (iii)A-the (Active): A N1 isn‟t V-ing the N2 (e.g. A dog isn‟t carrying the frog.) (iv) A-the (Passive): A N1 isn‟t being V-ed by the N2 (e.g. A frog isn‟t being carried by the dog.) 22.

(33) difference was that these sentences were presented in written form. During the experiment, the subjects first read the sentence and the experimenter then repeated it again. Also, the subjects had to select the better picture from the two. The result of this experiment revealed that the performances of the subjects were not influenced by written and oral presentations. Besides these two experiments, the third experiment was conducted and the participants were asked to describe what scenario they had when they heard a negative sentence. The result showed that both article and position effects were obvious in 10-year-olds. This was same as the results indicated in Experiment One and Experiment Two. To sum up, Rumain (1988) found that there were two significant cues to help understand the meaning of negative sentences: an article cue and a position cue. As for the development of articles and position effects, the result showed that the position effect was found earlier than the article effect. Children aged 10 relied more on position cues while the adults relied more heavily on articles. Nevertheless, there are some limitations. For example, the average age of the subjects did not include 4 years old, and a contextual effect was not investigated. 2.2.3 Zhou et al. (2014a) Zhou, Crain, and Thornton (2014a) conducted research on Mandarin-speaking children‟s knowledge of interaction between lexical aspect and negation. There are two negative markers in Mandarin, including bu and mei. Zhou et al. adopted the claim of 23.

(34) previous studies (Li and Thompson 1981, Lü 1980) which mentioned that Bu is always utilized to negate habitual or volitional expressions, whereas mei always occurs in a completed situation, as shown in (25a) and (25b), taken from Zhou et al. (2014a:150).. (25) a. Wo bu chi. xilanhua.. I not eat broccoli „I don‟t eat broccoli/I will not eat broccoli/ I don‟t want to eat broccoli.‟ b. Wo mei chi xilanhua I not eat broccoli „I didn‟t eat broccoli.‟. According to Zhou et al. (2014a), these two negative markers are related to lexical aspect which is a temporal concept of verbs. And these verbs can be divided into four categories: states, activities, accomplishments and achievements (Vendler 1957). Dynamicity, durativity, and telicity are features to differentiate these four categories. Dynamicity is determined by whether or not the events change denoted by the verb. Durativity refers to the event is lasting for a period of time. Telicity means whether or not there is an endpoint of the event. Both states and activities are durative and atelic, but state verbs are non-dynamic. Sentences (26) and (27) are examples of states and activities respectively, which are cited from Zhou et al.(2014a:151). (26) Yuehan hen gaoxing. John very happy „John is very happy.‟ (27) Zhe-ben shu hen gui. This-CL book very expensive „This book is very expensive.‟ 24.

(35) The other type of verbs is called accomplishments and they are durative and telic, such as xie-le „wrote,‟ and hua-le „drew‟ (Zhou et al. 2014a:151). The difference between accomplishments and the last type, achievement verbs, is that the later type implies a state change of instantaneous events. Achievement verbs are zhao-dao „find,‟ da-po „break,‟ and kan-jian „see.‟ In Mandarin Chinese, bu accompanies states and activities, while mei occurs with accomplishment, achievement, and activity verbs. Zhou et al. focused on children‟s knowledge of interaction between negative makers and lexical aspect. Forty-six Mandarin-speaking children from the kindergarten at Beijing Language and Culture University were recruited. Before the actual test, there were two practice trails. One four years old child was excluded from the actual test because he didn‟t respond to the practice test correctly. A two-choice forced-judgment task was conducted. Every participant and two puppets had to listen to a story told by an experimenter. After the story, one puppet produced the statement with bu, and the other produced the sentence with mei. Only one of puppets produced a grammatical sentence. In the two-choice forced-judgment task, there were five pairs of sentences with state verbs, and five pairs of sentences with accomplishment or achievement verbs. Below is an example of a typical trail presented in a story. (28) A pirate was hunting for treasure in the underwater world. In the beginning, he went to the corals to find if treasure was there. However, he only found some seaweed. Then, he 25.

(36) met a mermaid. He asked the mermaid to tell him where the treasure was in a very rude way. The mermaid got very angry and left. The pirate did not give up and then he climbed up to a big shell, but he found nothing. Therefore, the pirate gave up and left. (Zhou et al. 2014a:153) Following the story told by an experimenter, puppet A produced (29a) and puppet B produced (29b). Each child was asked to judge which puppet produced a grammatical sentence.. (29) a. Haidao bu zhidao baozang zai nail pirate not know treasure at where „The pirate didn‟t know where the treasure was.‟ b.* Haidao mei zhidao baozang zai nail pirate not know treasure at where „The pirate didn‟t know where the treasure was.‟ (Zhou et al. 2014a:153). The result of this experiment showed that the proportion of correct responses increased with age. As for sentence pairs with state verbs, children aged from 3;5-3;11 responded correctly only 54.29% of the time, while children aged 4;7-5;0 did 72.50% of the time. As for sentences with accomplishments or achievements, the proportion of correct responses of the younger children was still the lowest, whereas children aged between 4;7-5;0 got the highest proportion of correct responses. This finding indicated that Mandarin-speaking children knew the interaction between negation markers and lexical aspect by age 4;7, and they could use it in an adult-like way. Nevertheless, it did not mean that the younger children did not have the knowledge of negation and lexical aspect. They just faced some problems with how negative 26.

(37) markers interact with lexical markers. Overall, Zhou et al. (2014a) investigated the development of children‟s knowledge of the interaction with lexical aspect and negators. However, negators with activity verbs were not included in the test items and only one task was conducted in their study. 2.2.4 Zhou et al. (2014b) Zhou et al. (2014b) investigated children‟s knowledge of double negation in Mandarin Chinese. The aim of their study was to find out the age at which children could master and utilize double negative structures. They conducted two experiments with plausible contexts. Experiment One was a truth value judgment task, in which thirty children were recruited from the kindergarten at Beijing Language and Culture University, including fifteen children aged from 4;2 to 5;5 and fifteen children aged between 5;6 to 5;10. In the first experiment, two experimenters were involved. One of them had to act out the story with toys, while the other played the role of a puppet as if he were an audience listening to the story. The participants had to judge whether or not the puppet described the story correctly. The materials of the first experiment were constructed by six test trails, including three „true‟ scenarios and three „false‟ scenarios5. Examples of a „true‟ scenario and a „false‟ scenario are shown in (29) and (30). (29) The „true‟ scenario. 5. The double negative structure with a true description of the story is called a „true‟ scenario, while one with a false description is named as a „false‟ scenario. 27.

(38) Today is the little hippo‟s birthday, and he invited his friends to the party. Garfield, and the little bunny bought a gift before the party. However, Winnie the Pooh was having a toothache, and he might not come to the party. But when the party began, Winnie the Pooh showed up with the present. (Zhou et al. 2014b:340-341). (30) The „false‟ scenario The little turtle, the little elephant and the little donkey went to look for their moms. Both the little elephant and the little turtle found their moms in the forest. Nevertheless, the little donkey didn‟t found his mother. (Zhou et al. 2014b:342). After the story (29), the puppet described what was going on in the story, using the sentence in (31), taken from Zhou et al. (2014b: 341).. (31) Mei-you-ren mei qu canjia shengri juhui. not-have-person not go join birthday party „Nobody did not go to the birthday party.‟ (Meaning: Everybody went to the birthday party.) (Zhou et al. 2014b:341). After the scenario (30), the puppet described what happened in the story, using the sentence in (32).. (32) Mei-you-ren mei zhaodao mama. not-have-person not find mother „Nobody did not find his mother.‟ (Zhou et al. 2014b:343). Children who rejected both „true‟ and „false‟ scenarios showed that they interpreted a double negative structure as equivalent to a single negative structure. That is, they did not have the 28.

(39) concept that two negatives can cancel each other. The result of this experiment showed that children aged between 5;6-5;10 could interpret a double negative structure as equivalent to a positive meaning. Nonetheless, children aged from 4;2 to 5;5 considered a double negative sentence as the same as a single negative structure. To conclude, the child subjects showed knowledge of double negation by age 5;6. Another question discussed in Zhou et al.‟s study was to find out when preschool children could apply double negation correctly. In Experiment Two, the participants were divided into two groups: thirteen children aged from 5;6 to 5;11, and thirteen children aged between 6;0 to 6;5. In order to see whether adults could apply double negative structures, fifteen adults were recruited as a control group. Experiment Two was an elicited production task. Like the truth value judgment task, the experimenter acted out the story, and the other played the role of a puppet who listened to the story alongside the subjects. The scenarios were the same as played in the truth judgment task. In this task, the children had to answer some questions from the puppet based on what they saw in the scenario.. (33) a. Wo kandao xiao hema yaoqing ta-de haopengyou qu I see little hippo invite his good friend go canjia shengri juhui. join birthday party „In this story, the little hippo invited his good friends to his birthday party.‟ b. Ni neng gaosu wo you-shei mei qu ma? you can tell me have-who not go „I wonder if there is anyone who did not go.‟. Q (Zhou et al. 2014b:348). 29.

(40) The result showed that the 6-year-old children performed double negative sentences as what the adults performed. These children answered the question by using double negative sentences as in (34a) and (34b).. (34) a. Mei-you not-have. yi-ge-ren one-CL-person. mei qu. not go. „Not even a single person did not go.‟ b. Mei-you-ren mei qu. not-have -person not go „Nobody did not go.‟ (Zhou et al. 2014b: 349). However, the 5-year-old children answered the question in alternative ways, as shown in (35a) and (35b).. (35) a. Dou qu-le all go-Asp „They all did.‟ b. Mei-you not-have „Not have (one).‟ (Zhou et al. 2014b: 350).. Zhou et al. (2014b) revealed that appropriate contexts helped children understand the meanings of double negative sentences in Chinese. They concluded that children aged six had the knowledge of double negation and that they could apply the concept correctly. However, subjects older than 7 years old should also be recruited because Jou (1988) found that children did not acquire the concept of double negation until 8 years old. 30.

(41) 2.2.5 A Summary of Empirical Studies of Single Negation and Double Negation This section is a summary of empirical studies reviewed. The major findings and limitations of these previous studies are displayed in Table 2-1. Table 2-1 A Summary of the Previous Empirical Studies Study Jou (1988). Rumain (1988). Major Findings. Important stages: 1. Subjects: a small pool of 1. Interpreting double negative sentences as single subjects for each group negative sentences: less than 7 years old 2. Task: only one task without 2. Interpreting double negative sentences as felicitous contexts provided affirmative sentences: 8 years old Important stages: 1. Depending less on position cues: less than 7 years old 2. Depending more on position cues: 10 year old 3. Depending more on article cues: adult controls Factors affecting interpretations of single negatives: 1. Position cues 2. Article cues. Zhou, Crain, & Thornton (2014a). Limitations. 1. Subjects: only older subjects included 2. Tasks: contextual effect not examined. Important stage: Acquiring the interaction between 1. Subjects: adult controls not negative markers and lexical aspect: Aged 4;7 recruited 2. Task: a task on negation without activity verbs included. Zhou, Crain, & Thornton (2014b). Important stages: 1. Considering a double negative sentence as a single negative one: Aged 4;2 to 5;5 2. Producing double negative sentences: 5 years old 3. Interpreting a double negative structure as equivalent to a positive meaning: Aged 5;6-5;10 4. Performing adult-like double negative structures: 6 years old Factor affecting interpretations of double negatives: contextual cues. 31. 1. Subjects: children older than 7 years old not recruited 2. Tasks: contexts are too long.

(42) Overall speaking, the age effect was mainly discussed in these studies (Jou 1988, Rumain 1988, Zhou et al. ,2014a, 2014b). Children aged from 4;2 still considered a double negative sentence as a single negative sentence (Jou 1988, Zhou et al. 2014b). They could produce double negative sentences since at the age of 5 and interpreted them as positive sentences from 5;6 (Zhou et al. 2014b). However, Jou (1988) found that children could not interpret double negative sentences as affirmative ones until 8 years old. As for the development of single negative sentences, children aged 4;7 acquired the interaction between negative markers and lexical aspects (Zhou et al. ,2014a). Interestingly, no matter what first language that children had, they all relied on some cues to help them understand single negatives and double negatives better. As for English-native speakers, children whose age were less than 7 years old relied less on a position cue than 10-year-old children. On the other hand, adults depended more on an article cue (Rumain 1988). Mandarin-speaking subjects also required a strategy to interpret double negative sentences as affirmative sentences. In terms of Zhou et al. (2014b), felicitous contexts were truly helpful for children to interpret double negative sentences. Nevertheless, these studies still have some limitations. First, the contextual effect was not investigated in these previous studies (Jou 1988, Rumain 1988). Second, Zhou et al. (2014a) did not investigate the situations when two negators occur with activity verbs. Last, the variant age group was not recruited in these studies (Rumain 1988, Zhou et al. 2014a, 32.

(43) 2014b) and the subject pool of each age group was small (Jou 1988). Therefore, the present study takes the effect of context, verb types and age into consideration. The participants will be asked to judge whether each test sentence, containing either single negators or double negators, is acceptable with or without contexts. As for the relation between negators and lexical aspect in Mandarin Chinese, the present study examines at what age children acquire the knowledge of semantic differences between “bu + activities” and “mei(you) + activities”. Also, the age of the subjects in this study will range from 4 years old to 11 years old. 2.3 Summary of Chapter Two In this chapter, the properties of single negation and double negation are discussed. As mentioned in previous studies, two negators, bu and mei, in Mandarin Chinese have their semantic features. First, bu can only occur with stative and activity verbs; whereas, mei can accompany accomplishment, achievement and activity verbs. Second, bu is considered as a pure negative marker, while mei(you) functions as the denial of completion. Moreover, the scope of negation may determine the meanings of negative sentences. The meaning of a double negative sentence cannot be derived directly from the surface form (Fan and Xiao 2010). It is more complicated than a single negative sentence. A double negative sentence is not only interpreted as an affirmative meaning, but also strengthens the degree of affirmation. Besides four theoretical studies of single negation and double negation, four empirical studies and some limitations of previous studies have been discussed. Based on characteristics of bu 33.

(44) and mei and the previous empirical studies, the research design of the present study will be introduced in Chapter Three.. 34.

(45) Chapter Three Research Design In this chapter, the subjects and the research design of the present study are introduced. Section 3.1 introduces the background of the subjects recruited in the formal study. Section 3.2 presents the methods and materials with examples. Section 3.3 describes the research procedures of a pilot study and a formal study. Finally, Section 3.4 summarizes expected findings of the present study. 3.1 Subjects Some research had investigated children‟s acquisition of both single negative sentences and double negative sentences (Jou 1988, Rumain 1988, Zhou et al. 2014a, 2014b). As for single negatives, children at ages 4-5 were found to acquire the knowledge of the compatibility of lexical aspect and negative markers (Zhou et al., 2014a). As for double negatives, researchers did not reach a consensus on the age issue (Jou 1988, Zhou et al. 2014b). Jou (1988) found that children did not understand the meanings of double negative sentences until they were 8 years old, while Zhou et al. (2014b) found that children performed in an adult-like way at the age of 6. Following the assumption of the previous research (Jou 1988, Rumain 1988, Zhou et al. 2014a, 2014b), the researcher will recruit subjects from 5 to 12 years old in order to investigate whether children at age 5 have knowledge of lexical aspect and negators and whether children aged 12 are able to interpret 35.

(46) the meaning of double negative sentences as affirmative sentences. The present study investigated the children‟s developmental process of bu and mei in negative sentences; thus, ninety subjects from different ages were recruited. They were further divided into five groups based on age. The participants of Group 1 were preschoolers, Group 2, Group 3 and Group 4 were second grade, fourth grade, and sixth grade students of elementary schools in Taipei City, respectively. Group 5 is an adult group. The backgrounds of subjects are illustrated in detail in Table 3-1.. Table 3-1 Background of Subjects Group. Age range. Number. Kindergartner. 5-6. 18. Grade 2. 7-8. 18. Grade 4. 9-10. 18. Grade 6. 11-12. 18. Control. 20-21. 18. As Table 3.1 shows, each age group consisted of eighteen subjects whose mother tongue is Mandarin Chinese. All the subjects went to school five days a week: the preschoolers had a story-telling time in the morning and before a nap time; the first graders had 23 periods of classes per week; for the third graders, 30 periods per week; for the fifth graders, 33 periods per week. Moreover, the subjects from the elementary school all had language classes: six periods of Mandarin for the first and third graders, while seven periods for the fifth graders. The subjects from the elementary school all had one period of local languages per week. Each period lasted for 40 minutes. The control group was eighteen Mandarin-speaking adults from 36.

(47) college, aged from 20 to 21. 3.2 Methods and Materials In general, there are two types of approaches to investigating children‟s acquisition: cross-sectional and longitudinal (Larsen-Freeman & Long 1991). Both approaches have their advantages and disadvantages. A longitudinal study tracks the same subject for a period of time, maybe for many years. One of the advantages of longitudinal studies is that researchers can trace the process of development and get the spontaneous speech (Larsen-Freeman & Long 1991). Researchers can also examine subjects‟ development and change directly. Nonetheless, the shortcomings of the longitudinal study include a small subject pool, inconvenience, and time-consuming (Larsen-Freeman & Long 1991). The present study aimed to compare competence of the subjects in different age groups regarding single negative sentences and double negative sentences. Children probably will not use double negative sentences to express affirmation, but this does not imply that they have not acquired the concept of double negation. It is because competence is independent of performance (Chomsky 1965). Performance refers to a speaker‟s actual use, but it does not realize the speaker‟s innate ability of language. Moreover, the finding obtained from longitudinal studies may not be generalizable to other children (Larsen-Freeman & Long 1991). Therefore, this study will adopt a cross-sectional approach, which collects data in a controlled measurement, to investigate children‟s acquisition of single negative sentences and double negative 37.

(48) sentences. There are many types of tasks conducted within cross-sectional approaches, which can be used to collect intuition data and performance data. Previous studies (Jou 1988, Rumain 1988, Zhou et al. 2014a, 2014b) conducted judgment tasks, interpretation tasks and production tasks to investigate first language acquisition of single negatives and double negatives. One aim of the present study, which emphasized more on the input processing, was to examine whether children acquire the concept of scope of negation in both single negatives and double negatives. Thus, this study will employ a three-point scale acceptability judgment task, which is regarded as an interpretation task, for collecting intuition data. The present study did not only discuss whether children can interpret bu and mei in negative sentences, but it also took some factors into account while designing test items. The first factor was verb types. Following Vendler (1957), the researcher classified verbs into four types according to their features, including state (Type 1), activity (Type 2), achievement (Type 3) and accomplishment (Type 4). Comrie (1976) terms these verb types as lexical aspect because they are inherently related to time. The collocation of negative markers is related to lexical aspect (Ernst 1995, Zhou et al. 2014a). According to Zhou et al. (2014a), bu accompanies state verbs and activity verbs, while mei occurs with accomplishment, achievement, and activity verbs. The test items of the present study were designed according to eight sentence types with different lexical aspect, and each type were classified into two 38.

(49) sub-types based on numbers of negators. Type 1 refers to negative sentences consisting of state verbs, which can be divided into two sub-types based on sentence patterns: Type 1-1 and Type 1-2. The difference between these two types is that the former is a single negative, as in (1), and the latter is a double negative, as in (2). (1) Xiaoji bu/*mei zhidao zhe shou ge. chick not know this CL song „The chick doesn‟t know this song.‟ (2) a. Xiaoji buneng bu zhidao zhe shou ge. chick not can not know this CL song „The chick has to know this song.‟ b. * Mei you ren mei zhidao zhe shou ge. not-have person not know this CL song „Everyone knows this song.‟. According to previous studies (Li & Thompson 1981, Zhou et al. 2014a), bu can accompany state while mei cannot. Zhidao „know‟ is a state verb, so it cannot co-occur with mei „not‟, as shown in (2b) above. Type 2 refers to negative sentences with activity verbs, which are also partitioned into two sub-types. Type 2-1 is activity verbs in single negatives, while Type 2-2 refers to double negatives containing activity verbs. Both bu and mei can co-occur with activity verbs in Mandarin Chinese, as shown in (3) and (4).. 39.

(50) (3) Xiaotu bu/mei shuijiao. little rabbit not sleep „The little rabbit didn‟t/wouldn‟t sleep.‟ (4) a. Xiaotu bushi bu shuijiao. little rabbit not-be not sleep „It is not that the little rabbit would not sleep.‟ b. Meiyou. ren. not-have person „Everyone slept.‟. mei. shuijiao.. not. sleep. Type 3 refers to negative sentences with achievement verbs, which are divided into single negatives (Type 3-1) and double negatives (Type 3-2). Only mei can accompany achievement verbs (Zhou et al. 2014a), as shown in (5) for a single negative, and (6) for a double negative.. (5) Xiaohua *bu/mei kanjian lushang de shitou. Xiaohua not see road poss. rock „Xiaohua didn‟t see the rock on the street.‟ (6) a. Xiaohua bu-shi bu kanjian lushang de shitou. Xiaohua not-be not see road poss. rock „It is not the case that Xiaohua didn‟t see the rock on the street.‟ b. Meiyou ren mei kanjian lushang de shitou. not-have person not see road poss. rock „No one hasn‟t seen the rock on the street.‟ Type 4 refers to a negative with accomplishments and it consists of two sub-types as well: single negative sentences (Type 4-1), as in (7), and double negative sentences (Type 4-2), as in (8).. 40.

(51) (7) Xiaozhu *bu/mei chi bao. Piggy not eat full „The piggy was not full.‟ (8) a. Xiaozhu bu-de-bu chi bao. Piggy not have to not eat full „The piggy has to eat until he is full.‟ b. Meiyou ren mei chi bao. not-have person not eat full „Nobody is not full.‟. Sentence (7) indicates that bu cannot co-occur with accomplishment verbs while mei can (Zhou et al. 2014a). Interestingly, both bu and mei can accompany accomplishment verbs in double negative sentences. In addition to lexical aspect, the other factor which may influence children‟s interpretation of a single negative and a double negative is context. Although a contextual effect was discussed in Zhou et al. (2014b), few studies have investigated both context and lexical aspect together. Some researchers (Glenberg et al. 1999, de Villiers & Flusberg 1975, Zhou et al. 2014b) claimed that contexts are helpful for interpretation. Therefore, two acceptability judgment tasks were conducted in the present study: a context-free task and a context-provided task. All in all, eight types of test items were designed according to their acceptability of co-occurring with lexical aspect, as displayed in Table 3-2.. 41.

(52) Table 3-2 A Distribution of Test Items in the Two Tasks Verb Type. Type. Context-free Task. Context-provided Task. Type 1: State. Type 1-1 single negation: bu/mei + state verb. 2. Q4, Q10. Q6, Q12. Type 1-2 double negation: bu…bu/mei…mei + state verb. 2. Q2, Q19. Q1, Q10. 2. Q9, Q17. Q5 ,Q15. Type 2-2 double negation: bu…bu/mei…mei + activity verb. 2. Q8, Q18. Q16, Q19. Type 3: achieve. Type 3-1 single negation: bu/mei + achievement verb. 2. Q5, Q13. Q9, Q20. ment. Type 3-2 double negation: bu…bu/mei…mei + achievement verb. 2. Q3, Q15. Q2, Q17. Type 4-1 single negation: bu/mei + accomplishment verb. 2. Q14, Q20. Q4, Q14. Type 4-2 double negation: bu…bu/mei…mei + accomplishment verb. 2. Q7, Q11. Q7, Q11. 4. Q1, Q6, Q12, Q8, Q3, Q13, Q18 Q16. Type2:. Type 2-1 single negation:. Activity. bu/ mei + activity verb. Type 4: accompli shment. No.. Fillers Total. 20. 20. In both tasks, a three-point scale of an acceptability judgment task was designed for each type. Table 3-3 is a sample of test items in a context-free acceptability judgment task.. 42.

(53) Table 3-3 A Sample of Questions in a Context-free Acceptability Judgment Task The subject saw:. The subject heard:. Scene 1. Xiaomao shuo, “xiaozhu bu chi bao.” „The Kitty said, “The piggy is not full.”‟. Scene 2. “Xiaopengyou, qingwen xiaomao keyi zheyang shuo ma?” „Little boys and girls, can the kitty say so?‟. In a three-scale context-free acceptability judgment task, the subjects heard a single negative or a double negative produced by the kitty that comes from America. After hearing the sentence, the subjects were told to decide which light they would like to tick. Putting a checkmark on the green light meant that the sentence was acceptable in Mandarin Chinese; a checkmark on the yellow light indicated that the sentence was a little strange, but was still acceptable; a checkmark on the red light indicated that the sentence was totally unacceptable. For a complete version of this task, please refer to Appendix A. A sample question of a context-provided acceptability task is exemplified in Table 3-4. Similarly, the subjects were asked to put a check according to their acceptability. The only different was that the subjects heard a context before hearing a target test item. For a. 43.