英語完結性相關結構之第二語言習得

全文

(1)國立臺灣師範大學英語學系碩士論. 文. Master Thesis Department of English National Taiwan Normal University. 英語完結性相關結構之第二語言習得. Second Language Acquisition of English Telicity-Related Constructions. 指導教授：陳純音博士 Advisor: Dr. Chun-yin Doris Chen 研究生：何信昌 Student: Harvey Hsin-chang Ho. 中華民國九十七年六月 June, 2008.

(2)

(3) This thesis is dedicated to my beloved parents, Shih-ming Ho and Hsiu-man Liao..

(4)

(5) 摘要. 本研究主要探討英語完結性相關結構之第二語言習得，亦即結果結構、動質結構和與格/雙賓結構。為了避免潛在的實驗效應，本研究採用兩種實驗題型，其一是文法判斷題，其二是開放式翻譯題。研究議題包括結構本身對於第二語言習得的影響、三種結構間之關係、屬性效應、組別效應以及題型效應。本研究的受試者為一百零四位以中文為母語的高中生。依據其英語程度，分為初、中和高三組。整體實驗結果顯示，此三種結構在統計上呈現極顯著之相關。此相關性並不受到屬性效應、不同英語程度和實驗效應的影響，呈現以下習得次序：與格/雙賓結構最先習得、動質結構其次，而結果結構為後。回歸分析更顯露出此三種結構間之階層性預測關係：成功習得結果結構會顯著地預測成功習得動質結構和與格/雙賓結構，且動質結構亦會顯著地預測與格/雙賓結構之習得。檢測六大屬性時，亦發現在第二語言習得結構效應優於屬性效應。而單一結構的分析亦指示屬性效應大部分對於第二語言習得並無關鍵性的影響。此外，第二語言程度的因素亦受到了證實；受試者的表現隨著其英語能力的提升而進步。在題型效應方面，受試者在文法判斷題的表現比翻譯題來得好，顯示語言理解能力較語言使用能力先習得。最後，錯誤分析也證實了此三種結構一致的關係。本研究指出，不同的第二語言複雜程度、第一語言轉移以及第二語言語料刺激可能會導致此完結性相關結構之階層性習得次序。. I.

(6) II.

(7) ABSTRACT. The present study aims to probe into the second language acquisition of three telicity-related constructions in English, viz. resultative constructions, verb-particle constructions, and dative/double object constructions. To circumvent potential methodological effects, this study drew on a comprehension task, i.e. a grammaticality judgment task, and a production task, i.e. a free translation task, to garner pertinent data. The issues addressed included construction effects, the relationships of the three constructions, property effects, group effects, and task effects. The subjects of this study consisted of 104 Chinese-speaking senior high school students, and they were further divided into three groups according to their English proficiency levels, i.e. the high-level group, the mid-level group, and the low-level group. The overall results indicated that the three constructions were significantly correlated (p<.01** throughout). The following acquisition sequence was found across different properties, tasks, and L2 proficiency levels: datives/double objects were followed by verb-particles, which in turn preceded resultatives. The regression analysis further showed a hierarchically predictive relationship of the three constructions: the success in the acquisition of resultatives well predicted the successful acquisition of the other two constructions, and verb-particles were a significant predictor of datives/double objects as well. When all the test properties were compared, it was found that construction effects consistently overrode property effects on the L2 acquisition. In the within-construction analysis, it was also shown that properties played a marginal role in influencing the L2 acquisition of the constructions in most cases. Moreover, the variable of L2 proficiency was substantiated; the subjects’ performance improved with the increase in their III.

(8) proficiency levels. With respect to methodological effects, the subjects consistently performed better on the comprehension task than the production task, implying that comprehension might precede production in acquisition. Finally, the analysis of untargeted production data also confirmed the observed relationships of the constructions. This study suggests that different degrees of L2 complexity, L1 transfer, and L2 inputs may contribute to the hierarchical acquisition order of the three telicity-related constructions.. IV.

(9) ACKNOWLEDGEMENTS. I am a perfectionist procrastinator. The completion of this thesis could not have been achieved without many generous people’s support, assistance, and guidance. It is with pleasure and gratitude that I acknowledge their contributions. I owe an incalculable debt of gratitude to my thesis advisor, Dr. Chun-yin Doris Chen, who led me into the field of second language acquisition. I benefited considerably from her extensive and profound erudition in linguistics. In the last year, she was generous with her time and advice, arranging countless meetings for me and assisting me in each step to accomplish my thesis. She was often even more concerned about the progress of my thesis than I, which clearly manifested her deep concern for her student. Her enduring support, enlightening and invaluable advice, patient supervision, and opportune mental therapy were a linchpin of the steady progress and the ultimate accomplishment of this study. I am also deeply grateful to Dr. Jen-i Li and Dr. Sharon Rueih-Ling Fahn for their benevolence as my thesis committee members. Their thorough scrutiny of my thesis proposal accounted for the esthetically consistent results of this thesis, and their meticulous reading of my manuscripts and insightful comments played a substantial role in improving the quality of my thesis as well. Besides, I wish to express my heartfelt appreciation to all the other professors who taught me during my graduate school years at the Department of English, National Taiwan Normal University (NTNU): Dr. Tammy Miao-Hsia Chang, Dr. Miao-Ling Hsieh, Dr. Hsueh-O Lin, Dr. Hsi-Yao Su, Dr. Jen Ting, and Dr. John Kwock-Ping Tse by alphabetic order. Their teaching and intellectual inspiration laid the underpinning for this thesis, and the commendable warmth of human touch prevailing in the department pumped much more color to my graduate life. V.

(10) Furthermore, I am much indebted for the invaluable help rendered by the experimental subjects recruited from Taipei Chenggong Senior High School and Taipei Jingmei Girls’ Senior High School. Special thanks go to my graduate school classmates Natasha Yu-Hsin Huang and Meng-Hsuan Claire Hsieh for their kindly assistance with data collection. But for their participation and help, this thesis would not have been possible. Moreover, my gratitude is extended to all my amicable and remarkable classmates of the MA Program in Linguistics, NTNU: Kyle Chien-Pin Chang, Mike Chan-Chia Hsu, Laney Cheng-Lan Shen, Cindy Peiru Shih, Angeline An-lin Sun, and Nefertiti Kui-Ju Tsai by alphabetic order. Graduate school life was more bustling and difficult than I had anticipated. Fortunately, my classmates’ camaraderie and rapport brought about the warm and pleasant ambiance around. I gained much benefit from their friendship, diligence, and knowledge. Their innate sense of humor gingered up my difficult graduate life, and their spiritual support smoothed my research progress as well. With a mixture of bitters and sweets, what I have experienced together with them in the phase will be a lasting memory for the rest of my life. Last and most importantly, I would like to thank my beloved parents for their unconditional love and unwavering support. Since my birth, they have devoted their whole heart and soul to me, and their love has never waned with time elapsing. They even tolerated the increasingly difficult person I had become in the last critical months of finalizing my thesis. From their constant understanding and unreserved devotion, I have experienced and realized their inestimable love and great expectations of me. Thanks to their endless love, care, encouragement, and backup, I could concentrate on my thesis and tide over the obstacles with which I was confronted. I would like to dedicate this thesis to them.. VI.

(11) TABLE OF CONTENTS Page CHINESE ABSTRACT ......................................................................................................... I ENGLISH ABSTRACT....................................................................................................... III ACKNOWLEDGEMENTS.................................................................................................... V TABLE OF CONTENTS .................................................................................................... VII LIST OF TABLES ............................................................................................................. IX LIST OF FIGURES ............................................................................................................. X CHAPTER ONE Introduction ......................................................................................... 1 1.1 Motivation ................................................................................................................ 1 1.2 Theoretical Background ........................................................................................... 4 1.3 Research Questions .................................................................................................. 7 1.4 Significance of the Study ......................................................................................... 7 1.5 Organization of the Thesis ....................................................................................... 8 CHAPTER TWO Linguistic Properties and Literature Review of Telicity-Related Constructions ...................................................................................... 9 2.1 Properties of Telicity-Related Constructions ........................................................... 9 2.1.1 Resultative Constructions in English and Chinese ......................................... 9 2.1.1.1 Stative versus Dynamic Verbs ......................................................... 11 2.1.1.2 Predication ....................................................................................... 13 2.1.2 Verb-Particle Constructions in English and Chinese .................................... 14 2.1.2.1 Stative versus Dynamic Verbs ......................................................... 14 2.1.2.2 Particle Movement ........................................................................... 16 2.1.3 Dative/Double Object Constructions in English and Chinese ...................... 16 2.1.3.1 Stative versus Dynamic Verbs ......................................................... 18 2.1.3.2 Word Order ...................................................................................... 19 2.1.4 Summary ....................................................................................................... 20 2.2 Theoretical Studies of Telicity-Related Constructions .......................................... 21 2.2.1 Hoekstra (1988) ............................................................................................ 22 2.2.2 Tenny (1994) ................................................................................................. 27 2.2.3 Snyder (1995)................................................................................................ 29 2.2.4 Baker (1997) ................................................................................................. 34 2.2.5 Summary ....................................................................................................... 38 2.3 Empirical Studies of Telicity-Related Constructions ............................................ 39 2.3.1 Snyder and Stromswold (1997) .................................................................... 39 VII.

(12) 2.3.2 Slabakova (2001) .......................................................................................... 43 2.3.3 Summary ....................................................................................................... 51 2.4 Summary of Chapter Two...................................................................................... 52 CHAPTER THREE Research Design ............................................................................ 53 3.1 Subjects .................................................................................................................. 53 3.2 Materials and Methods ........................................................................................... 54 3.3 Procedures .............................................................................................................. 60 3.3.1 Pilot Study..................................................................................................... 60 3.3.2 Formal Study ................................................................................................. 62 3.3.3 Scoring and Statistical Analysis.................................................................... 64 3.4 Summary of Chapter Three .................................................................................... 67 CHAPTER FOUR Results and Discussion .................................................................... 69 4.1 Construction Effects and Clustering ...................................................................... 69 4.2 Property Effects ..................................................................................................... 79 4.3 Task Effects ........................................................................................................... 87 4.4 Other Patterns Elicited ........................................................................................... 92 4.5 Summary of Chapter Four ..................................................................................... 99 CHAPTER FIVE. Conclusion ...................................................................................... 101 5.1 Summary of the Major Findings .......................................................................... 101 5.2 Pedagogical Implications ..................................................................................... 102 5.3 Limitations of the Present Study and Suggestions for Further Research............. 103 REFERENCES ............................................................................................................... 105 APPENDIX A The Grammaticality Judgment Task ................................................... 113 APPENDIX B The Free Translation Task .................................................................. 115. VIII.

(13) LIST OF TABLES Page Table 2-1. A summary of the properties of the three telicity-related constructions in English and Mandarin Chinese ............................................................... 21 Table 2-2. Bulgarian learners’ acquisition of aspect and the related constructions..... 49 Table 3-1. A summary of the subjects ......................................................................... 54 Table 3-2. The grammaticality judgment task ............................................................. 55 Table 3-3. The free translation task ............................................................................. 59 Table 3-4. The results of the pilot study ...................................................................... 61 Table 3-5. A general guideline for the scoring of the translation task ......................... 65 Table 4-1. Mean differences between the overall results on the constructions ........... 72 Table 4-2. Mean differences between the experimental groups’ scores on the constructions ............................................................................................... 76 Table 4-3. Correlations of the telicity-related constructions ....................................... 77 Table 4-4. Group differences on different properties of the constructions .................. 82 Table 4-5. The two ungrammatical stative verb items in datives/double objects ........ 84 Table 4-6. Frequency of other patterns of resultatives elicited .................................... 92 Table 4-7. Frequency of other patterns of verb-particles elicited ................................ 95 Table 4-8. Frequency of other patterns of datives/double objects elicited .................. 97 Table 4-9. Other patterns of the telicity-related constructions elicited ........................ 98. IX.

(14) LIST OF FIGURES Page Figure 4-1. The overall results on the telicity-related constructions ........................... 72 Figure 4-2. Each group’s scores on the telicity-related constructions ......................... 75 Figure 4-3. The acquisition of one construction vis-à-vis the other two constructions .............................................................................................. 78 Figure 4-4. Mean scores on different properties of the constructions ......................... 80 Figure 4-5. Each group’s scores on different properties of the constructions ............. 81 Figure 4-6. Accuracy on the grammaticality judgment of different verb types........... 84 Figure 4-7. Accuracy on the grammaticality judgment of different verb types (Reanalysis) ............................................................................................... 85 Figure 4-8. The overall results on the two tasks .......................................................... 89 Figure 4-9. Task effects across groups ........................................................................ 89 Figure 4-10. Task effects across constructions ............................................................ 90 Figure 4-11. Task effects across properties ................................................................. 91. X.

(15) CHAPTER ONE INTRODUCTION. 1.1 Motivation In the field of second language acquisition, research on aspect has focused on the morphological mappings of grammatical aspect and lexical semantics of verbs in the past score of years (e.g. Kumpf 1984, Kaplan 1987, Robison 1990, 1995, Andersen 1991, Bardovi-Harlig 1992, Shirai 1993, Andersen and Shirai 1996). Nonetheless, empirical studies on the relationships between aspectual meaning and certain constructions have been relatively underrepresented (cf. Slabakova 2001, Huang and Yang 2005). In English, resultative constructions are such an exemplar which pertains to aspectual interpretations. It is pointed out that the constructions denote an aspectually telic event (e.g. Hoekstra 1988, Goldberg 1991, Tenny 1994, Snyder 1995, Filip 1999, Goldberg and Jackendoff 2004). A typical example of English resultative constructions is shown in (1), in which a resultative complement red is involved.. (1) John painted the house red.. (Snyder 1995:460). Resultative constructions, nevertheless, are not a unique hallmark of the English language. Comparable to English resultative constructions, Mandarin Chinese has two ways to form resultatives, namely resultative verb compounds (RVCs) and resultative V-de constructions (e.g. Li and Thompson 1982, Huang 1988, 1992, Li 1990, 1998, 1999, Zhang 2001, Tai 2003). The following sentences illustrate these two types of resultative constructions in Mandarin Chinese.. 1.

(16) (2) Wo didi I. shuai-sui-le. yi-zhi. huaping.. (RVC). brother smash-broken-PERF one-CL vase. ‘My brother smashed a vase into pieces.’ (3) Ta he. ma-de. wo shuo. scold-DE I. bu. chu. speak not out. hua. lai.. word. come. ‘He scolded me so harshly that I could not speak a word.’. (RESULTATIVE V-DE). (Li 1999:445). The resultative element is expressed in the form of the resultative verb compound, as sui ‘broken’ in (2), and the clause wo shuo bu chu hua lai ‘I cannot speak a word’ in (3). Although both English and Chinese resultative constructions involve a resultative element, they display different morphosyntactic structures. In English resultatives, the main verb and the resultative complement are split in their syntactic positions, as shown by painted and red in (1). In Mandarin Chinese, however, the action verb and the resultative verb are adjacent and form a verb compound together, as shuai-sui ‘smash-broken’ in (2), and resultative V-de constructions are characterized by an obligatory morpheme de, as in (3). These differences may pose a challenge for Chinese learners of English as a second language in the acquisition of the telic constructions in English. In addition to resultative constructions, some researchers have associated other English constructions with telic construal, viz. verb-particle constructions and double object constructions (e.g. Snyder 1995), as illustrated in (4)-(5).. (4) The student took off his glasses.. (VERB-PARTICLE CONSTRUCTION). (5) a. The secretary gave a present to her boss.. (DATIVE CONSTRUCTION). b. The secretary gave her boss a present.. (DOUBLE OBJECT CONSTRUCTION). Given their similar aspectual property, the present study links English verb-particle 2.

(17) constructions and double object/dative constructions together, in conjunction with resultative constructions, under investigation. This study aims to examine the acquisitional difficulties of the three telicity-associated constructions in L2 English. Besides, the formation of this study is also motivated by the lively theoretical debate on the relationships of the three telicity-related constructions. Some theorists propose that the three constructions form a cluster (e.g. Kayne 1984, 1985, Hoekstra 1988, Snyder 1995), whereas some linguists suggest that they should not be subsumed under the same family of constructions (e.g. Tenny 1994, Baker 1997, Di Sciullo 1999). In spite of the theoretical disagreement, there has been a dearth of empirical evidence on the issue as to whether the three constructions belong to the same cluster. Basically, there are only two empirical studies on the issue retrieved from the literature, one L1 study (Snyder and Stromswold 1997) and one L2 study (Slabakova 2001).1 As seminal the two studies are, their methodologies are subject to some limitations. The question concerning the relationships of the three purportedly related constructions is still awaiting a satisfactory answer. Thus, further empirical research on these constructions is necessary. Moreover, according to Chomsky (1981), superficially distinct constructions will emerge around the same time in the process of acquisition if they belong to a cluster of constructions. Given the theoretical disagreement and lack of well-documented empirical evidence on the issue of clustering, the present study, based on Chomsky’s assumption, aims to explore the L2 acquisition of the three English constructions as a source of evidence. In short, the present study attempts to discover degrees of difficulty of the 1. In addition to Snyder and Stromswold’s (1997) and Slabakova’s (2001) acquisition studies on English telicity-related constructions, Huang and Yang (2005) also conducted an empirical inquiry into constructions bearing on telicity in L2 Chinese. The constructions under investigation included Chinese directionals/locatives, resultatives (resultative verb compounds and resultative V-de constructions), and double object constructions. 3.

(18) English telicity-related constructions in L2 acquisition and to shed some light on the contentious question of their relationships. With the objectives in mind, this study probes into second language acquisition of the three constructions in English by Chinese speakers.. 1.2 Research Background The term ‘telicity’, introduced by Garey (1957:106), refers to whether a situation has ‘a natural final endpoint, or intrinsic bound’ (Smith 1997:19). A telic event is true of a natural endpoint such that when an event attains a natural endpoint, a concomitant change of state takes place and the event is completed. An atelic event, in contrast, does not have such a natural outcome and thus ‘can stop at any time’ (Smith 1997:19), as shown in (6).. (6) a. John loves Mary.. (ATELIC STATIVE). b. Sam strolled in the park.. (ATELIC ACTIVITY). c. Sam coughed.. (ATELIC SEMELFACTIVE). d. Mary reached the top.. (TELIC ACHIEVEMENT). e. Sylvia drew a picture.. (TELIC ACCOMPLISHMENT) (Smith 1997:177-182). In the literature, a consensus has been reached on English telicity marking in basic SV(O) structures (e.g. Vendler 1967, Verkuyl 1972, 1999, Comrie 1976, Dowty 1979, 1991, Smith 1991, 1997, Tenny 1994, Jackendoff 1996, Olsen and Resnik 1997, Kreidler 1998, Filip 1999, Slabakova 2001). It has been found that English telicity marking is compositional and primarily hinges on the calculation of two components: the lexical semantics of the main verb and the quantization/specification status of the internal argument.. 4.

(19) The lexical meaning of a matrix verb plays a crucial role in determining the telicity interpretation of a situation. The examples in (6) illustrate telic versus atelic meanings that the main verbs contribute to. The verb in (6a) itself is a stative verb that denotes a state of a strong feeling of affection. The state has an arbitrary endpoint rather than a natural final endpoint and is regarded as an atelic situation. Besides stative verbs, dynamic verbs like stroll and cough do not encode an intrinsic bound as well. For example, one can stroll back and forth in a park for hours. Such verbs conventionally denote an atelic event per se. On the contrary, verbs such as reach and draw, together with their direct objects, signify a delimited event, as (6d) and (6e) suggest. When the natural endpoint is reached, the event is completed—a telic event. Thus, in light of the role of lexical semantics of a main verb with different aspectual interpretations, the present study takes properties of main verbs into consideration in the research design. The quantization/specification status of an internal NP argument also exerts a material influence on the telicity reading of a sentence in English. All the sentences in (7) denote past tense and perfective viewpoint aspect.. (ATELIC). (7) a. Carol ate. b. Carol ate a pear/the pear/a bag of pears.. (TELIC). c. Carol ate pears.. (ATELIC). d. Carol ate rice.. (ATELIC). In spite of the identical matrix verb ate, they depict an atelic event except for the ones in (7b). The example in (7a) contains only a verb and does not have an explicit direct object. It describes an atelic activity without any specification of what Carol ate; thus, the event of Carol’s eating is not delimited. In a similar vein, the situations in. 5.

(20) (7c)-(7d) are not delimited either. The direct object is a bare plural NP in (7c) and a mass noun in (7d). Since the internal arguments in these examples are not quantified/ specified, the events are not delimited and telic. In contrast, the sentences in (7b) contain a quantified/specified direct object. The determiners a/the and the measure words a bag of specify the quantification status of the internal argument pear(s), which yields the telic readings of the sentences. As the quantization/specification status of the internal argument is not the research focus of this study, the internal arguments in the present research design are specified direct objects so as to circumvent a potential effect induced by direct objects. In addition to the linguistic properties, L2 proficiency may contribute to variations in second language acquisition. There has been a voluminous literature on the role of L2 proficiency in acquisition (Kellerman 1983, Sasaki and Hirose 1996, Whong-Barr and Schwartz 2001, Ying 2004, among others). For example, some studies point out the U-shaped behavior exhibited by beginners, intermediate learners, and advanced ones in L2 development (e.g. Kellerman 1983), while some investigations show contrasting tendencies between high- and low-proficiency learners (e.g. Whong-Barr and Schwartz 2001). Moreover, task effects have been discussed in the acquisition research (Larson-Freeman 1976, Tarone 1985, Rothman 2007, among others). Task-induced quantitative and qualitative variations are reported in several studies (e.g. Hyltenstam 1984), whereas researchers like Chaudron (1983) claim that tasks are not a variable per se. A related issue to task effects is the relationships of comprehension and production in acquisition. It is conventionally assumed that comprehension precedes production (McCarthy 1954). However, empirical evidence points to two directions with respect to the sequence of comprehension and production (cf. Ehri 1976, Janson and Schulman 1983, Wu and Wen 1997, McKee et al. 1998). In view of the factor of L2 proficiency and task 6.

(21) effects in acquisition, this study also examines these two variables.. 1.3 Research Questions The present empirical study aims to address the following five research questions. 1) Do our subjects acquire the three telicity-related constructions in English approximately concurrently, exhibit a sequence of acquisition, or learn them in a piecemeal fashion? 2) Do different properties of the telicity-related constructions display varying degrees of difficulty and contribute to a different order of acquisition? 3) Is L2 proficiency a variable influencing the acquisition of the telicity-associated constructions? 4) Do different tasks on the telicity-associated constructions yield parallel results? If yes, does comprehension precede production consistently? 5) What other patterns are elicited with respect to the telicity-related constructions?. 1.4 Significance of the Study The. issue. concerning. the. relationships. of. English. telicity-associated. constructions has been animatedly debated (cf. Kayne 1984, 1985, Hoekstra 1988, Tenny 1994, Snyder 1995, Baker 1997), but there has been a paucity of empirical evidence on the issue. Thus, the theoretical question is worth addressing. The present study does not endeavor to provide a definitive answer to the issue. Rather, this study, adopting a quantitative and qualitative approach, aims to resort to second language acquisition of the constructions as a source of evidence and to shed some light on the theoretically controversial issue. Besides linking the three constructions together under consideration, this study looks into the acquisition of different properties of the 7.

(22) constructions in an effort to discover the interaction of construction effects and property effects in second language development. Finally, L2 proficiency and task effects (comprehension versus production) are explored in this study as well.. 1.5 Organization of the Thesis This thesis is organized as follows. Chapter Two deals with main properties of the three telicity-associated constructions in English and Mandarin Chinese, and reviews previous relevant theoretical as well as empirical studies. Chapter Three is devoted to the research design of this study, and Chapter Four reports and discusses the experimental results. Finally, Chapter Five offers concluding remarks and suggestions for further research.. 8.

(23) CHAPTER TWO LINGUISTIC PROPERTIES AND LITERATURE REVIEW OF TELICITY-RELATED CONSTRUCTIONS. This chapter consists of four sections. Section 2.1 discusses major properties of the three telicity-related constructions in English, in conjunction with comparison and contrast of pertinent characteristics in Mandarin Chinese. Sections 2.2 and 2.3 review some theoretical and empirical studies on the telicity-associated constructions in English, respectively. Section 2.4 is a summary of this chapter.. 2.1 Properties of Telicity-Related Constructions This section is concerned with key properties of the three telicity-related constructions at issue, especially those properties to which all the constructions exhibit great resemblance. Since the present study is an L2 empirical study, with English as the target language and Chinese as the mother tongue, relevant properties in the two languages are compared and contrasted accordingly. The comparison and contrast can serve as the basis for the research design of the present study. Characteristics of resultative, verb-particle, and dative/double object constructions are dealt with in the following subsections.. 2.1.1 Resultative Constructions in English and Chinese English resultative constructions have drawn a great deal of research in the fields of semantics and syntax (e.g. Hoekstra 1988, 1992, Goldberg 1991, Carrier and Randall 1992, Tenny 1994, Filip 1999, Goldberg and Jackendoff 2004). They can be categorized into three major classes on the basis of verbal properties, as in (1).. 9.

(24) (1) a. He hammered the metal red.. (TRANSITIVE). b. The butter melted to a liquid.. (ERGATIVE). c. He drank himself into the grave.. (UNERGATIVE). (Hoekstra 1988:119). In the literature, transitive resultatives are treated as part of the same telic family as verb-particle constructions and double object constructions (e.g. Hoekstra 1988, Snyder 1995). Hence, the present study focuses on transitive resultatives in English resultative constructions. In this type of resultatives, resultative complements can be expressed in three forms, namely an adjectival phrase, a prepositional phrase, and a noun phrase (Carrier and Randall 1992, Tenny 1994), as illustrated below.. (2) a. She painted the barn red.. (AP). b. She pounded the dough into a pancake.. (PP). c. She painted the barn a weird shade of red.. (NP) (Carrier and Randall 1992:183). In Mandarin Chinese, there are two ways to form resultatives, viz. resultative verb compounds (RVCs) and resultative V-de constructions (e.g. Li and Thompson 1982, Huang 1988, 1992, Jaxontov 1988, Li 1990, 1998, 1999, Zhang 2001, Tai 2003, Lin 2004, Wu 2004, Xiao and McEnery 2004), as in (3).. (3) a. Wo didi I. yi-zhi. shuai-sui-le. huaping.. (RVC). brother smash-broken-PERF one-CL vase. ‘My brother smashed a vase into pieces.’ b. Ta he. ma-de. wo shuo. scold-DE I. bu. chu hua. lai.. (RESULTATIVE V-DE). speak not out word come. ‘He scolded me so harshly that I could not say a word.’ 10. (Li 1999:445).

(25) Resultative verb compounds consist of V1 and V2 (Li and Thompson 1982:54), as shuai ‘to smash’ and sui ‘broken’ in (3a). The first verb encodes the action of the event, and the second verb signals the resultant state. Thus, resultative verb compounds are also termed ‘action-result’ verb compounds (Tai 1984:290). On the other hand, resultative V-de constructions are characterized by an obligatory morpheme de attached to the first verb.1 As (3b) shows, the action verb ma ‘to scold’ is indicative of the causing event, whereas the resultant state is encoded by the resultative clause wo shuo bu chu hua lai ‘I cannot speak a word’. Two main properties of English resultative constructions are discussed below, i.e. verb types and predication.. 2.1.1.1 Stative versus Dynamic Verbs It is pointed out that resultative constructions have the effect of telicity conversion (Hoekstra 1988, Tenny 1994, Snyder 1995, Filip 1999). The constructions could convert a sentence that denotes an atelic event into a sentence with telic construal, if the resultative predicate is felicitous with the original sentence. For example, the VP wring the towel inherently denotes an atelic event, as indicated by the durative adverbial phrase for five minutes in (4a), while the addition of the resultative complement dry coerces a delimited interpretation over the non-delimited one, as in (4b).. (4) a. He wringed the towel for five minutes. b. He wringed the towel dry in/#for five minutes.. 1. (Tenny 1994:38). There has been considerable debate regarding the matrix versus embedded relationships of the two verbs in the Chinese V-de constructions (e.g. Li and Thompson 1982, Huang 1988, Tai 2003). This issue is beyond the scope of the present study; thus, it will not be discussed here. 11.

(26) Although resultative complements have such a delimiting effect, the function does not apply to all types of verbs in English. Resultative expressions are compatible with felicitously dynamic verbs,2 as in (5). Nonetheless, they cannot apply to any stative verbs, for stative verbs cannot be converted into verbs that do denote a telic event (e.g. Slabakova 2001), as illustrated in (6).. (5) a. She wiped the table. b. She wiped the table clean. (6) a. He loved her. b. *He loved her touched.. In Chinese, resultative constructions also convey telic messages, as exemplified in (3). In contrast to English, Chinese resultatives in RVCs are compatible with some stative verbs, as in (7), and Chinese stative verbs can also appear in resultative V-de constructions when verb copying is applied, as shown in (8).. (7) a. Ta he. ai-si-le. ta.. love-die-PERF. her. ‘He loved her to death.’ b. Ta. Lisi.. hen-tou-le. he. hate-thoroughly-PERF Lisi.. ‘He hated Lisi extremely.’ (8) Ta he. ai. ta. love her. ai-de. jihu. love-DE nearly. fahuang. go.crazy. ‘He loved her so much that he nearly went crazy.’ 2. All the verbs in telic constructions are dynamic verbs, whereas not all the dynamic verbs are congruent with telic constructions. The expression ‘felicitously dynamic verbs’ is used here to avoid overgeneralizations. 12.

(27) 2.1.1.2 Predication Although English resultative predicates have three forms to modify a nominal argument, the argument of which they can be predicated is constrained. Resultative complements are predicated only of the direct object in English (Carrier and Randall 1992, Tenny 1994), as in (9). In contrast, depictives, which should be distinguished from resultatives, may describe the subject or the object, as in (10).. (9) a. Susan pounded the metali flati.. (RESULTATIVE). b. *Susani pounded the metal tiredi. (10) a. I cut the breadi hoti.. (DEPICTIVE). b. The horsei cannot run a furlong tiredi.. (Tenny 1994:58, 153-154).. Unlike English resultatives, Chinese resultative expressions are more flexible in predication (Li 1998, 1999). Resultative elements in RVCs allow ambiguous interpretations for one sentence in some cases. They can be predicated of the object, as in (11a), or the subject, as in (11b). Similarly, modification of the resultative VP in V-de constructions can apply to the object, as in (3b), or to the subject in cases like (12).. (11) Youyoui zhui-leii/j-le. Taotaoj le.. Youyou chase-tired-PERF Taotao PAR a. ‘Youyou chased Taotao and as a result Taotao became tired.’ b. ‘Youyou chased Taotao and as a result Youyou became tired.’ (Li 1998:296) (12) Woi xiang-de taj I. miss-DE her. [ chi bu. xia. fan]i.. eat not down food. ‘I missed her so much that I lost my appetite.’. 13. (Li 1999:448).

(28) 2.1.2 Verb-Particle Constructions in English and Chinese English verb-particle constructions, as the name itself suggests, consist of a main verb and a particle (e.g. Bolinger 1971), as shown in (13).. (13) a. He looked up his friends. b. They turned down my request.. In Mandarin Chinese, four directional verbs shang ‘to ascend’, xia ‘to descend’, qi ‘to rise’, and chu ‘to emerge’, and five result verbs diao ‘to fall’, zou ‘leave’, zhu ‘to fasten’, dao ‘to reach’, and kai ‘to open’ are regarded as Chinese verb particles in the literature (Teng 1977, Chen 2003), as illustrated in (14).. (14) a. Ta he. zhongyu. kao-shang-le. daxue.. eventually take.an.exam-up-PERF. university. ‘He eventually got admitted to a university.’ b. Wo xiang-chu-le I. think-out-PERF. yi-ge. banfa.. one-CL. solution. ‘I thought out a solution.’. (Chen 2003:27, 31). Verb-particle constructions in English are characterized by two major properties. One is inapplicability to stative verbs, and the other is particle movement, as explicated below.. 2.1.2.1 Stative versus Dynamic Verbs According to Bolinger (1971:85), verb particles contain two features, ‘one of motion-through-location, the other of terminus or result’. Like resultatives, the presence of verb particles can convert verbs denoting an atelic event into telic ones 14.

(29) (Bolinger 1971, Tenny 1994, Snyder 1995). For instance, the interpretation of walk the course is ambiguous between a delimited and a non-delimited reading. On the one hand, walk the course can be interpreted as a telic event, in which one walks from the onset of the course to the end. On the other hand, one can walk back and forth without covering the entire path. However, when a verb particle is added, the atelic reading is overridden by a telic interpretation, as in (15).. (15) Tom walked the course through in/#for ten minutes.. (Tenny 1994:36). Accordingly, verb particles are compatible with felicitously dynamic verbs, as the above example suggests. The delimiting function of verb particles disallows telic particles to co-occur with any stative verbs, since it is impossible to convert such verbs into verbs that do denote a telic event (Slabakova 2001), as in (16).. (16) a. She liked the man. b. *She liked the man through.. Like English verb particles, the so-called verb particles in Chinese also denote telic meanings, as in (14). Although English verb-particle constructions do not apply to any stative verbs, Chinese counterparts can change the aspectual properties of sentences with a stative verb. An example is shown in (17), where the particle shang ‘up’ contributes to an inchoative interpretation of the sentence, i.e. the starting point of falling in love (Huang and Hsieh 2008).. (17) a. Ta she. ai. na-ge. yuangong.. love. that-CL. employee. ‘She loves that employee.’. 15.

(30) b. Ta she. ai-shang-le. na-ge. yuangong.. love-up-PERF. that-CL. employee. ‘She fell in love with that employee.’. 2.1.2.2 Particle Movement Another hallmark of verb-particle constructions is particle movement. Verb particles can optionally appear on either side of an internal argument NP (Bolinger 1971, Fraser 1976), as in (18).. (18) a. He ran up the bill. b. He ran the bill up.. However, when the direct object is a pronoun, the verb particle is obligatorily placed after it, as in (19) (Quirk et al. 1985, Biber et al. 1999).. (19) a. How fast can you pick it up? b. *How fast can you pick up it?. (Biber et al. 1999:932). Chinese verb particles, in contrast, are not subject to the same particle movement as English counterparts, as in (20) (cf. (14a)).. (20) *Ta he. zhongyu. kao-le. eventually take.an.exam-PERF. daxue. shang.. university. up. ‘He eventually got admitted to a university.’. 2.1.3 Dative/Double Object Constructions in English and Chinese There has been a voluminous literature on the relationships between dative. 16.

(31) constructions and double object constructions (e.g. Chomsky 1975, Kayne 1984, Larson 1988, 1990, Aoun and Li 1989, Jackendoff 1990). Putting aside the controversial issue, this study provisionally treats dative and double object constructions as two varieties of one construction (cf. Chomsky 1975, Larson 1988, Aoun and Li 1989). English dative/double object constructions are of two types, to-datives (Goal-type) and for-datives (Beneficiary-type), as in (21)-(22).. (21) a. John gave a gift to Mary.. (DATIVE CONSTRUCTION). b. Alice baked a cake for David. (22) a. John gave Mary a gift.. (DOUBLE OBJECT CONSTRUCTION). b. Alice baked David a cake.. (Gropen et al. 1989:204). Dative constructions comprise a direct object NP, a gift/a cake, followed by an obligatory case marker to/for and an indirect object, Mary/David, as in (21). On the contrary, double object constructions are characterized by the presence of two adjacent postverbal NPs, with the indirect object preceding the direct one, as in (22). In Mandarin Chinese, two versions of dative constructions and one type of double object construction have been identified in the previous research (cf. Li and Thompson 1982, Liu 2006), as in (23).. (23) a. Wo song-le I. yi-ping jiu. give-PREF one-CL wine. gei ta.. (V NP1 GEI NP2). GEI him. ‘I gave a bottle of wine to him.’ b. Wo. song gei. ta. yi-ping. I. give GEI. him one-CL. jiu. wine. ‘I gave him a bottle of wine.’ 17. (V GEI NP2 NP1).

(32) c. Wo I. song-le. ta. yi-ping. give-PERF him one-CL. jiu.. (V NP2 NP1). wine. ‘I gave him a bottle of wine.’. (Li and Thompson 1982:376). For dative constructions, there are two variants in Mandarin Chinese, as exemplified in (23a-b.). Both types consist of a direct object and an indirect object, with an obligatory case marker gei preceding the indirect object,3 but differ in the relative word order of the two postverbal NPs. Double object constructions, on the contrary, do not have the case marker gei and comprise an indirect object followed by a direct one, as in (23c). The following subsections focuses on two characteristic properties of dative/ double object constructions in the two languages, verb types and word order.4. 2.1.3.1 Stative versus Dynamic Verbs Like resultatives and verb-particles, dative/double object constructions denote a telic message as well (e.g. Snyder 1995). The indirect object, i.e. the Goal argument, may delimit the event expressed by the constructions. All the verbs in English dative/double object constructions are ditransitive and dynamic verbs, as in (21)-(22); the constructions are incongruent with any stative verbs, as shown in (24)-(25).. (24) Carol owned some money. (25) a. *Carol owned some money to the orphanage. b. *Carol owned the orphanage some money. 3. The syntactic status of gei in dative constructions has been a focus of interest in the research on the constructions (e.g. Li and Thompson 1982, Li 1990, Chang 2004). This issue is beyond the scope of this study and will not be elaborated on here. 4 In addition to the two attributes, dative/double object constructions are famous for another eminent property, viz. dative alternation such that dative constructions alternate with double object constructions (cf. Baker 1979, Pinker 1989, Gropen et al. 1989, Sawyer 1995, Inagaki 1997, Liu 2006). Since this property is not the focus of the present study, it will not be explicated here. 18.

(33) This is the case with Chinese dative/double object constructions. They are also incompatible with any stative verbs, as illustrated in (26)-(27).. (26) Wo renshi I. know. na-wei laoshi. that-CL teacher. ‘I knew that teacher.’ (27) a. *Wo renshi-le I. na-wei laoshi. gei. Zhangsan.. know-PERF that-CL teacher GEI. Zhangsan. ‘*I knew that teacher to Zhangsan.’ b. *Wo renshi I. know. gei. Zhangsan na-wei laoshi. GEI. Zhangsan that-CL teacher. ‘*I knew Zhangsan that teacher.’ c. *Wo renshi-le I. Zhangsan yi-wei laoshi. know-PERF Zhangsan one-CL teacher. ‘*I knew Zhangsan that teacher.’. 2.1.3.2 Word Order In addition to incompatibility with all stative verbs, dative/double object constructions are typified by their characteristic word order. In English, dative constructions require an obligatory case marker to, but retention of the case marker is ungrammatical in double object constructions, as illustrated in (28).. (28) a. John gave a letter *(to) Mary. b. John gave (*to) Mary a letter.. Similarly, Chinese dative/double object constructions exhibit corresponding word order to that of English counterparts, as in (23a) and (23c), repeated here as (29a) and 19.

(34) (29c). The difference between English and Chinese lies in the absence/presence of the case marker to/gei when the indirect object precedes the direct one, as shown by the contrast in (28b) and (29b).5. (29) a. Wo song-le yi-ping jiu *(gei) ta. ‘I gave a bottle of wine to him’ b. Wo song gei ta yi-ping jiu. ‘I gave him a bottle of wine.’ c. Wo song-le ta yi-ping jiu. ‘I gave him a bottle of wine.’. 2.1.4 Summary In this section, two major properties of each of the three telicity-related constructions are discussed, as summarized in Table 2-1. English resultative constructions are compatible with felicitously dynamic verbs but cannot apply to any stative verbs. Chinese resultatives, in contrast, are congruent with some stative verbs. Besides, English resultatives are predicated only of the object, while Chinese counterparts can modify the object or the subject. In the case of verb-particle constructions, they cannot apply to any stative verbs in English but in some Chinese cases, and particle movement is available in the former but not in the latter. In the case of dative/double object constructions, English and Chinese do not allow any stative verbs to be the main verbs of the constructions, and the two languages display similar word order in dative/double object constructions—[V D.O. to/gei I.O.] and [V I.O. D.O.]. The difference resides in the absence/presence of the case marker when the indirect object precedes the direct one; the case marker is not sanctioned in English 5. In Chinese, when the indirect object is followed by the direct object, some verbs require the presence of gei, some verbs take it optionally, and others cannot appear with it (Li and Thompson 1982). 20.

(35) Table 2-1. A summary of the properties of the three telicity-related constructions in English and Mandarin Chinese6 Construction. Property. English. Chinese. Stative. ±. 9. Dynamic. 9. 9. Subject. ±. 9. Object. 9. 9. Stative. ±. 9. Dynamic. 9. 9. [V Particle NP]. 9. 9. [V NP Particle]. 9. ±. Stative. ±. ±. Dynamic. 9. 9. [V DirectObject to IndirectObject]. 9. 9. [V to IndirectObject DirectObject]. ±. 9. Verb type Resultative Predication Verb type Verb-particle. Dative/ double object. Particle movement Verb type Word order. but is permitted in Chinese counterparts. On the one hand, the comparison and contrast between the three English and Chinese constructions provide the underpinning for the research design of the present study. On the other hand, the observed L1-L2 similarities and differences are the basis for the prediction and discussion of experimental findings with respect to potential positive/negative L1 transfer.. 2.2 Theoretical Studies of Telicity-Related Constructions This section recapitulates some theoretical proposals on the relationships of the three telicity-related constructions in English, with focus on the issue as to whether the constructions belong to the same family syntactically and semantically.. 6. In Table 2-1, the tick ‘9’ indicates that the properties in question are applicable to the languages under consideration, and the cross ‘±’ represents that the languages do not exhibit the given properties. 21.

(36) 2.2.1 Hoekstra (1988) Hoekstra (1988) adopts a unified approach to dealing with double object constructions,7 verb-particle constructions, and resultative constructions. Following Kayne (1984, 1985), Hoekstra analyzes double object constructions and verb-particle constructions with recourse to small clause (SC) constructions and proceeds to treat resultative constructions as involving a small clause as well (cf. Larson 1988, Snyder 1995). Sentence (30) is a typical example of a small clause. Such an example is analyzed as containing an embedded S [Mary a genius] by Kayne (1984), in which Mary is interpreted as the subject and a genius the predicate, as bracketed in (31).. (30) John believed Mary a genius. (31) [V' [v believed] [SC [NP Mary] [NP a genius]]]. (Kayne 1984:135). Considering some resemblance between double object constructions and typical SC structures, Kayne (1984) regards the double object construction ‘V NP NP’ as involving a small clause. The double object construction, as shown in (32), is taken to be bi-sentential, with [Mary a book] constituting an embedded S, as in (33).. (32) John gave Mary a book. (33) [V' [v gave] [SC [NP Mary] [NP a book]]]. (Kayne 1984:134). Kaye adduces some parallel characteristics between double object constructions and typical SC constructions as substantiations of his SC analysis. First, both typical small 7. One point merits our note here. Although Hoekstra (1988) groups double object constructions together with resultatives as well as verb-particle constructions, he refrains from viewing dative constructions as part of the same family. Hoekstra’s analysis of double object constructions originates from Kayne (1984). Kayne (1981, 1984) advances that double object constructions and dative constructions are both base-generated, i.e. neither is derived from the other (cf. Chomsky 1975, Larson 1988, 1990, Aoun and Li 1989, Jackendoff 1990). 22.

(37) clauses and double object constructions can be paraphrased in a comparable way. The examples in (30) and (32) can be rephrased as the sentences in (34) and (35), respectively.. (34) John believed Mary to be a genius. (35) John caused Mary to have a book.. (Kayne 1984:136). Second, evidence from Russian shows that [NP NP] can function as an S, supporting the SC treatment of double object constructions. The embedded S [Mary a genius] in (30) can be found in Russian be-less copula sentences, as in (36), where only English glosses are given in the Russian word order in Kayne (1984:135). Likewise, Russian have-less constructions also exhibit a similar pattern to [Mary a book] in (32), as shown in (37).. (36) Ivan student. ‘Ivan is a student.’ (37) U Ivana krasivye glaza.8 by Ivan. pretty. eyes. ‘Ivan has pretty eyes.’. (Kayne 1984:134). Third, typical small clauses and double object constructions exhibit the same patterns concerning nominalization. They can undergo passivization, as in (38), but neither of their derived nominals is grammatical, as illustrated in (39).. 8. Although the example in (37) is cited in Kayne (1984:135), he does not elucidate whether the example is an independent sentence. Thanks to Dr. Sharon Rueih-Ling Fahn’s assistance and her Russian-speaking colleague’s native intuition, it is confirmed that (37) is an independent sentence in Russian. 23.

(38) (38) a. Mary was believed [t a genius] by John.. (PASSIVIZATION). b. Mary was given [t the letter] by her teacher. (39) a. *Mary’s belief [PP [P of [SC t a genius]] by John]. (NOMINALIZATION). b. *Mary’s gift [PP [P of [SC t the letter]] by her teacher]. (Kayne 1984:146). Building ideas from Kayne (1984, 1985), Hoekstra (1988) extends the SC analysis to verb-particle constructions (see also den Dikken 1995). The following examples suggests that verb-particle constructions are licensed in the same way as SC structures with reference to extraction and nominalization, as in (40)-(42).. (40) a. She considered [Bernard a fool]. b. She looked [the information up]. (41) a. *Who did she consider [[the brother of t] [a fool]]?. (EXTRACTION). b. *What did she look [[the information about t] [up]]? (42) a. *her consideration [of [Bernard a fool]] b. *her looking [of [the information up]]. (NOMINALIZATION) (Hoekstra 1988:107-108, 115). Along this line of research, Hoekstra (1988, 1992) further proposes that resultative constructions should be treated as SC constructions as well. In typical SC constructions ‘V [NP XP]’, although the subject NP is exceptionally case-marked by the verb, the subject NP does not bear a sensible thematic relationship to the verb. Rather, it is the subject NP and the predicate XP that form a complement taken by the verb. Hoekstra contends that resultative constructions behave on a par with typical SC constructions, as shown in (43).. (43) The joggers ran [SC the pavement thin].. 24. (Hoekstra 1988:115).

(39) In (43), ran is an intransitive verb and does not subcategorize for an internal argument. Thus, the NP the sidewalk is more thematically related to the adjectival predicate thin. In light of the similar thematic traits between SC structures and resultatives, the SC treatment of resultative constructions is borne out. Finally, Hoekstra (1988:135) claims that double object constructions are closely akin to resultative constructions. Double object constructions denote a possessive relation. As possession is argued to be the result state of an event described by double object constructions, the constructions are also regarded as ‘possessive results’ (p.135). Some Dutch examples show similar properties to double object constructions on the one hand and resultative constructions on the other hand, as in (44).9. (44) a. dat ik me that I. een ongeluk werk. myself an. (POSSESSIVE). accident work. ‘that I work very hard’ b. dat ik me that I. dood werk. (RESULTATIVE). myself dead work. ‘that I work myself to death’. (Hoekstra 1988:136). The aspectual situation in (44a) is classified as possessive as a double object construction, but the message conveyed by the sentence can be replaced by a resultative construction, as in (44b). In Kayne’s seminal work, his analysis of double object constructions is grounded on some similarities between SC structures and double object constructions with respect to semantic paraphrases, corresponding evidence from Russian, and ill-formed nominalization. The argument based on semantic paraphrases may be flimsy. The 9. Hoekstra (1988) adduces the examples in (44) to argue for the close relationship of resultative and double object constructions. However, it is not clear that the example given in (44a) is possessive. 25.

(40) sentence involving a double object construction, as in (32), is not equivalent to the paraphrase, as in (35), in their interpretation. In the case of the Russian examples, the be-less copula sentence and the have-less construction suggest that [NP NP] can be interpreted as an SC. The argument is actually predicated on the premise that the SC sentence in (30) and the double object construction in (32) semantically correspond to the be-less copula sentence in (36) and the have-less construction in (37), respectively. Nonetheless, even though the double objection construction, as in (32), denotes a possession relationship between the postverbal two NPs, the give-class constructions suggest an alienable possession relationship between the two NPs. This differs from the inalienable possession relationship between a body part and a man, as shown by the Russian example in (37), i.e. eyes and Ivan. If the assumption is not on the right track, the argument for the SC analysis needs to be reconsidered. Also, Hoekstra’s SC analysis of double object constructions might not be soundly conceived. The Dutch example in (44a) Hoekstra cites, in fact, is neither equivalent to double object constructions nor to resultatives. The analysis that double object constructions are treated in the same way as resultatives may not be tenable. As regards verb particles, Hoekstra, following Kayne, treats them as the head of an SC, as in (40b). However, a characteristic attribute of verb particles is that they can appear on either side of the postverbal NP. As the verb particle occurs on the left side of the postverbal NP, a question arises as to what licenses such a movement, no matter whether the verb particle or the postverbal NP moves. In the case of resultative constructions, Hoekstra also adopts an SC analysis. It is contended that the postverbal argument in resultative sentences shows a closer thematic relationship to the resultative predicate than the main verb, as in SC sentences. However, in sentences like (45), the postverbal argument bears one thematic role closely related to the transitive verb (cf. Carrier and Randall 1992, 26.

(41) Snyder 1995). This type of sentence seems to point to another alternative treatment of English resultatives.. (45) The worker painted the wall green.. 2.2.2 Tenny (1994) Tenny (1994) points out that English verb particles and resultative expressions exhibit close affinities with respect to semantics and syntax. She claims that these two constructions should be subsumed under the same family. As far as semantics is concerned, verb particles and resultative complements play a crucial role in determining the aspectual status of an event described by a verb. The presence of verb particles or resultative predicates can convert an atelic event into a telic counterpart. The following examples illustrate the function of telicity conversion of verb particles and resultatives.. (46) a. The woman pushed the cart for/#in ten minutes b. The woman pushed the cart over #for/in ten minutes (47) a. The worker hammered the metal for/#in an hour. b. The worker hammered the metal flat #for/in an hour.. (Tenny 1994:36-38). On top of semantic similarities, Tenny argues that verb-particle constructions are analogous to resultatives in the following two patterns. First, verb particles as well as resultative predicates exhibit parallel distributions in relation to depictives, as in (48) and (49).. (48) a. We ate the muffins up hot. b. *We ate the muffins hot up. 27.

(42) (49) a. We hammered the metal flat hot. b. *We hammered the metal hot flat.. (Tenny 1994:154-155). In (48), when a verb particle co-occurs with a depictive, the verb particle must precede the depictive. This holds true of the distribution of resultatives. A resultative complement obligatorily precedes a depictive when they appear together, as in (49). The second point of resemblance between verb particles and resultatives lies in the relative word order to a postverbal noun phrase. Verb particles can optionally appear on either side of a postverbal noun phrase, as in (50). Tenny argues that this is the case with resultative complements. Resultatives conventionally follow the postverbal noun phrase, as in (51a). When the postverbal noun phrase is a heavy NP which carries on heavy information and phonology, however, a resultative complement can appear on the left side of the noun phrase, as in (51b).. (50) a. look a name up in the directory b. look up a name in the directory (51) a. The baker beat the eggwhites into stiff peaks. b. The baker beat into stiff peaks the eggwhites that this assistant had made a special trip to a neighborhood from to get fresh eggs for.. (Tenny 1994:156). Tenny’s claim that verb particles share comparable aspectual semantics with resultatives is justifiable, but her augment from the perspective of syntax appears to be tenuous. In the distribution relative to the postverbal noun phrase, particle movement is a typical feature of verb particles. Nevertheless, such a movement is not a hallmark of resultatives. Only when the postverbal noun phrase is a heavy NP can resultative predicates appear on the left side of the NP. In effect, this phenomenon is the so-called heavy NP-shift, as illustrated in (52). 28.

(43) (52) a. Jeeves introduced ti to the guests [NP the famous detective from Belgium]i. b. My doctor told me to drink ti every night [NP two glasses of mineral water with a slice of lemon]i.. (Haegeman 1994:419). Unlike Hoekstra (1988), Tenny (1994) does not link double object/dative constructions with verb-particle and resultative together under examination. She only briefly mentions that double objects and datives are not subject to the measuring-out constraint without elaborating on the differences between double objects/datives and verb particles as well as resultatives. However, the indirect object in dative/double object constructions is a Goal argument. The Goal argument itself is an endpoint of an event and will delimit the event expressed by such constructions. Hence, dative/double object constructions do measure out.. 2.2.3 Snyder (1995) As reviewed in Section 2.2.1, Hoekstra (1988) resorts to SC constructions to group English double object, verb-particle, and resultative constructions together. Given that all these constructions not only denote a telic interpretation but also display similar structural patterns, Snyder (1995) also treats these constructions as a family. However, Snyder’s approach differs from Hoekstra’s SC analysis. Snyder, deriving much inspiration from Larson (1988, 1990),10 adopts the Larsonian VP-shell analysis to account for the three constructions. In light of telic semantics of English resultative constructions, Snyder puts forth. 10. Instead of treating double object constructions as a small clause, Larson (1988, 1990) proposes that double object constructions ‘V NP NP’ are better accounted for in terms of a complex predicate construction (i.e. Larsonian VP-shell analysis). His argument is justified based on such syntactic tests as anaphor binding, quantifier binding, weak crossover, superiority, each…the other, and negative polarity items. His complex predicate proposal is pursued by several researchers and is extended to other constructions, such as put-locatives, verb particles, and resultatives (e.g. Hale and Keyser 1992, 1993, Snyder 1995). 29.

(44) a ‘null telic morpheme’ (Øtelic) and provides a compositional account for the telic interpretation of the constructions, as formulated in (53).. (53)‖Øtelic‖(P) (e) = True, for any event e and any predicate of events P, iff for that event e' which is a subevent of e and which is the ‘natural endpoint’ of e, P(e') = True.. (Snyder 1995:459). According to Snyder, an event described by a sentence involving an English resultative construction denotes a change of state in the Theme/Patient of the event at the endpoint of this event. As characterized in (53), the Øtelic morpheme makes reference to the natural endpoint of the event such that the completion of the subevent equals to the endpoint of the event, i.e. the resultant change of state. An example of resultative constructions is shown in (54), which is diagramed in (55).. (54) John painted the house red. (55). VP John. V' V. VP DP. VP. the house. V'. V painted. XP X. AP. Øtelic. red. (Snyder 1995:460). Sentence (54) can be interpreted as ‘John painted the house until the house became 30.

(45) red’. As represented in (55), the Øtelic morpheme alludes to the natural endpoint of the event of John’s painting the house and specifies the subevent such that the house should be red at the endpoint of this event. According to Snyder’s analysis, since English licenses the Øtelic morpheme syntactically, it is likely to convert a process into an accomplishment11 by means of the Øtelic morpheme and addition of a predicative complement to this morpheme,12 as illustrated below.. (56) a. Mary laughed for/#in five minutes. b. Mary laughed John out the door.. (Snyder 1995:461). As in (56a), the verb laugh is a process verb, compatible with the durative adverbial, whereas the sentence with the predicative complement out the door added in (56b) does not denote a process but an accomplishment. In view of the delimiting effect of the path predicate,13 Snyder proposes that process verbs contain a null/overt path morpheme. As shown by the contrast in (57a), it is the overt path morpheme to that contributes to the result reading.. 11. Based on Vendler (1967) and Parson (1990), Snyder (1995) treats an accomplishment-type event as consisting of a process and a culmination subevent. The subevent refers to the culmination of an event: the subevent is the natural endpoint of the event, at which a resultant change of state is formed in the Theme/Patient of the event. 12 Snyder proposes that the aspect of a situation depicted by a sentence depends crucially on the syntax of the sentence rather than simply on the pragmatics of the situation. Only in languages that syntactically license the Øtelic morpheme can a process be converted to an accomplishment by addition of a predicative complement. Otherwise, even though the added element is pragmatically congruent with an accomplishment-type event, such an aspectual shift is not possible. To justify the aspectual role of the Øtelic morpheme in syntax, Snyder adduces supporting examples from English and Spanish. It has been found that English path predicates can convert a process into an accomplishment, as in (i). In contrast, the unavailability of the Øtelic morpheme in Spanish accounts for the impossibility of such an aspectual conversion, as in (ii). (i) John walked to the top of the hill in twenty minutes. (ii) Juan caminó hasta la cima (?/#en vente minutos). ‘John walked up-to the hilltop (in twenty minutes). (Snyder 1995:465) 13 A path predicate is noted to have the function of introducing a natural endpoint to a Theme argument, specifying the argument to reach the endpoint of the path (Snyder 1995). 31.

(46) (57) a. John pushed the box onto/#on the table. [#for result reading] b. Mary laughed John out (Øto) the door.. (Snyder 1995:462-463). Just as the case in (57a) contains an overt path morpheme to, the sentence in (56b) is also argued to involve a null path morpheme Øto, as illustrated in (57b). The null path morpheme Øto stipulates that the Theme argument John should reach the Ground14 argument the door at the endpoint of the event. In conjunction with the meaning of the preposition out, the ultimate telic meaning of the sentence obtains accordingly. Snyder applies this approach to verb-particle constructions where a Ground argument is equated with an Agent, as in (58), which is represented in (59).. (58) John took his shoes off. (59). VP John. V' V. VP DP. his. V'. shoes V took. PP P (to). XP X. PP. Øtelic. P off. (Snyder 1995:463). In this case, the Theme argument his shoes ‘reaches’ the Ground argument John, also 14. The term ‘Ground’ is used in Snyder (1995:463). However, the conventional term for the thematic role is ‘Goal’ versus ‘Source’ (Saeed 2003:150). 32.

(47) an Agent of the event, through the path denoted by the Øto morpheme, as diagramed in (59). Together with semantics of the preposition off, the result meaning of the sentence is thus derived. Along this thread of thought, Snyder proceeds to adopt this approach to accounting for English double object constructions. He capitalizes on parallels between English present-with constructions and double object constructions (cf. den Dikken 1995), 15 as in (60). Given the existence of the preposition with in the present-with construction, Snyder puts forth a corresponding null possessive morpheme Øwith for the double object construction, as illustrated in (61).. (60) a. John presented Mary with a medal. b. John gave Mary a medal. (61). VP John. V' V. VP NP Mary. V' V. XP. gave X. PP. Øtelic P. DP. Øwith. a. medal. (Snyder 1995:466). In the anslysis, Snyder uses a null telic morpheme to explain telic interpretations of English resultative, verb-particle, and double object constructions. However, his 15. den Dikken (1995) argues against viewing English double object construction as an equivalent to present-with construction. 33.

(48) analysis is subject to further discussion. To begin with, it is proposed that the Øtelic morpheme contributes to the semantics of English accomplishment-type events and that the Øtelic morpheme is characterized by an event and a culmination subevent. A question arises regarding how the null telic morpheme applies to an achievement-type event. As telic an achievement is, such an event is instantaneous and thus does not involve a culmination subevent. Moreover, Snyder draws on an implicit path morpheme and a null possessive morpheme to cope with English verb-particle constructions and double object constructions, respectively. Nonetheless, the null morpheme comes in variable forms, even in the same constructions, say Øwith or Øat in the double object constructions (Snyder 1995:466-467). The question lies in what constrains different forms of the null morphemes. Furthermore, particle movement is typical of verb-particle constructions. However, only one version of particle constructions is explained. The other version is not explicated clearly and thus calls for a sound account.. 2.2.4 Baker (1997) Dative/double object constructions suggest a transfer of possession. As Baker (1997) notes, some previous research claims that double object constructions entail a successful transfer of possession, but dative constructions do not guarantee success in such kind of transfer (e.g. Mazurkewich and White 1984, Gropen et al. 1989, Pinker 1989), as shown in (62).. (62) a. She sang a song for her dead lover. b. #She sang her dead lover a song.. (Baker 1997:89). The Goal argument her dead lover is not alive and capable of possessing anything.. 34.

(49) Thus, the sentence with the double object construction, which requires a successful transfer of possession, is semantically anomalous, as in (62b). In contrast, the sentence in (62a) involves a dative construction and is well-formed. It is suggested that double object constructions inherently impart a telic interpretation by virtue of a successful transfer of possession. Nevertheless, Baker contends that the successful transfer of possession is one kind of conversational implicature and thus is defeasible. Based on counterexamples like (63), Baker contends that double object constructions do not convey constant telic messages per se. The examples show that the implication of the transfer of knowledge and the ball can be cancelled.. (63) a. I taught the children French, but they did not learn it at all. b. I threw John the ball, but it didn’t reach him because of the strong wind. (Baker 1997:89). Besides providing contrasting evidence, Baker draws on locative alternations to justify his stance that double object constructions do not convey telic meanings themselves. In locative alternation, the thematic roles associated with the two verb complements vary with locative shifts, as shown in (64). (64) a. John sprayed this wall with paint in/#for an hour.16 b. John sprayed paint onto this wall #in/for an hour. c. John sprayed subway cars with this can of paint #in/for an hour. d. John sprayed this (whole) can of paint onto subway cars in/#for an hour. (Baker 1997:88) 16. The sentence John sprayed this wall with paint for an hour is semantically acceptable on the atelic interpretation that spaying the wall lasted for an hour but the wall was not completely sprayed with paint in the period of time. 35.

(50) The NP this wall has the Theme role that is completely affected in (64a), but has the Locative role in (64b). The thematic role associated with the NP paint shifts between an Instrument role and a Theme role. In both frames of locative alternation, the specification status of any argument in the object position slot determines the aspectual quality of the entire sentence. In (64a), the quantified theme this wall gives rise to the telic interpretation of the sentence. In contrast, when the specified argument this wall shifts with the unquantified/unspecified argument paint, as in (64b), the alternation leads to the atelic meaning of the sentence. The examples in (64c-d) also point to such an effect of locative alternation. However, the aspectual interpretation of a sentence does not vary with dative alternation, in contrast to locative shift, as in (65).. (65) a. I have read stories to the children for/#in an hour. b. I have read the children stories for/#in an hour. c. I have read the story to children ?for/in an hour. d. I have read children the story ?for/in an hour.. (Baker 1997:88). When the two object NPs alternate in word order, no change of aspectual meaning takes place, as (65a-b) and (65c-d) exemplify. Baker suggests that double object constructions themselves do not play a crucial role in influencing the aspectual reading of a sentence. Rather, it is the quantization/specification status of the Theme argument that determines the telicity meaning of the sentence. Finally, Baker proposes seven syntactic tests on dative and locative alternations to verify that the Theme argument in dative/double object constructions is the legitimate direct object, which is a determinant of the telicity interpretation. The syntactic tests include adjectival secondary predication, the domain of wh-movement,. 36.