WIWIHIIWI H WI I 1 I W 1 923$ IIIiHWHIIWIHEI” THE'SlS / ”WW/‘Wsmfi 97777773'" UanAms This is to certify that the thesis entitled COUN'I‘ERFAC’I‘UAL SENTENCES IN CHINESE AND THE INTERPRETATION OF THEIR COUNTERFACI‘UALITY presented by Bing Li has been accepted towards fulfillment of the requirements for M.A. degree in Linguistics Major professor u / Date ’70 Q‘fl’vvfl/as‘ i“ 0.7539 MS U is an Affirmative Action/Equal Opportunity Institution -_—~H‘ _4a _,_1_ ._ _ _ u LIBRARY Michigan State l University PLACE N RETURN BOX to move thin chockom from your record. TO AVOID FINES Mum on at Mon dd. duo. l -_L__|| l !!____| I __J____l [:3 C1 fil—j r__ . MSU I. An Affirmative Action/Equal Oppomnlty Initiation COUNTERFACTUAL SENTENCES IN CHINESE AND THE INTERPRETATION OF THEIR COUNTERFACTUALITY By Bing Li A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Linguistics and Germanic, Slavic, Asian and African Languages 1995 ABSTRACT COUNTERFACTUAL SENTENCES IN CHINESE AND THE INTERPRETATION OF THEIR COUNTERFACTUALITY By BingLi It has been long held in Chinese linguistics as well as in philosophy that Chinese usually does not distinguish indicative conditionals from counterfactual conditionals (Chao,1959; Li & Thompson,l981; Bloom,1981, 1984). The aim of the present thesis is to argue for and assert that there are Chinese conditional sentences that are interpreted unambiguously as counterfactuals. Two sets of data are analyzed in support of the thesis. One contains conditionals whose antecedents fit the linguistic structure of [bu-shi [p...] ], where it is claimed that shi is an anaphoric demonstrative which triggers a presupposition that the complement proposition is true. The term ’Definite Propositional Description (DPD)‘ is used to refer to this semantic phenomenon. The other case consists of conditionals containing certain indexical expressions interpreted relative to certain contextualization. The term 'Indexical Counterfactuality (IC)' is used for the counterfactual nature of these conditional sentences. Key Words: Conditional Sentences, Indicative, Counterfactual, Hypothetical, Antecedent, Consequent, Definite Description, Definite Propositional Description (DPD), Indexicals, Indexical Counterfactuality (1C) Copyright by BING L1 1995 TO MY PARENTS iv ACKNOWLEDGMENTS My deepest gratitude goes to Professor Barbara Abbott, Chair of my Thesis Committee. On her insightful comments, suggestions and criticism, I have not only better grasped the issues central to my thesis and learned the fundamental principles in conducting sound arguments, but also have understood the meaning of being a real scholar. I wish to thank Professors Grover Hudson and Yen-Hwei Lin for serving as my Committee members and for their constant help during my stay at MSU. My humble gratitude is also due to all my teachers at MSU, particularly, besides my committee members, Professors Dennis Preston and Carolyn H'arford; to the professors I have worked for as a Teaching Assistant, Yen-Hwei Lin, Roger Bresnahan of the Department of American Thought and Language, the late Joseph Lee, Dennis Preston, and Denise Troutman-Robinson; and to the stafl of the Department of Linguistics and Germanic, Slavic, Asian and African Languages, particularly Professor George F. Peters, Chair of the department, and Ms. Lauri Church; to Dr. Ken Ebert and Dr. David Homer of the Office of International Students and Scholars; and to my colleagues and classmates. Part of the present thesis is based on my presentation at the NACCL 5 at University of Delaware and I am grateful to the College of Arts and Letters for the fellowship granted for my trip. Finally, my special thanks to the Butler Library staff of Columbia University for their assistance. TABLE OF CONTENTS 1: INTRODUCTION 2: ANOTHER LOOK AT CHINESE COUNTERFACTUAL CON STRUCTIONS 2.1: The Traditional Claim and its Philosophical Consequence 2.2: A Different Claim 3: [bu-shi [p 1 1 AS A LINGUISTIC STRUCTURE FOR COUNTERFACTUALITY 3.1: Description of the Data 3.2: The Semantics of 8111': Definite Propositional Descriptions (DPD) 3.2.1: The Two Semantic Components of 3111' 3.2.2: Definite Propositional Description (DPD) 3.3: A Problem in Eifring‘s Study and Discussion 3.4: Summary 4: INDEXICAL COUNTERFACTUALITY (1C) IN CHINESE 4.1: Description of the Data 4.2: Indexicals and Context 4.3: 10 (Indexical Counterfactuality) in Chinese 4.3.1 Indexicals Related to the Speaker and Past Time Reference 4.3.2 Indexicals Related to Definite Demonstrative Reference 4.3.3 Indexicals and Counterfactual Antecedents 4.4: Summary 13 13 14 i388 s 33 $38 5: SUMMARY AND CONCLUSIONS NOTES APPENDDI I: GLOSS OF ABBREVIATIONS APPENDDI II: GLOSS OF KEY CHINESE TERMS BIBLIOGRAPHY vii 47 49 1: INTRODUCTION Conditional sentences are common syntactic configurations which express causal relationships, potentialities, possibilities, and evidence relationships in discourse concerned with explanation and argumentation. More commonly, a conditional sentence is composed of two parts, one part expressing a certain condition and the other establishing a certain logical relationship with the exprewed condition. In English (see Appiah, 1985:161. inter alia), a conditional sentence is often expressed in the form of (1) (1) If A, then C where A is called the antecedent and C the consequentl. Beside this more commonly used form for conditional sentences, Appiah (ibid.) also lists some other forms of conditional sentences in English, such as the various abbreviations like ’If John is coming, so is Mary’, or, even ‘Mary is coming, if John is', where the consequent and antecedent are not complete English sentences, nor where the antecedent has to occur before the consequent. For expository reasons, the more general structure of ’If A, then C’ will be taken as a basic form of conditional sentences throughout this thesis. Conditional sentences are generally divided into indicative conditionals and counterfactual ones (Appiah,l985; J ackson,1990; Lewis,1973; Gamut,1991; Adams,1970; Traugott, et 81.,1986). For an indicative conditional, the truth value of the sentence is dependent on the truth value of its antewdent and consequent, such that, for example in (2) below, (2) If John is coming, Mary is coming. 2 the conditional sentence is true if 'John is coming' is found to be true and 'Mary is coming' is found to be true. Or (2) is false if'John is coming’ is found to be true but ‘Mary is coming‘ is found to be false. In counter-factual conditionals, it has been commonly said that the antecedents are false. For example, given that the speaker of (3) knows that John did not come, the antecedent clause of (3) will express a counterfactual condition, (3) If John had come, Mary would have come. Nevertheless, the term ‘counterfactual’ may appear to be misleading. Karttunen 8: Peters (1979:4-8) have used the term 'subjunctive conditional sentences’, in lieu of 'counterf actual’, where the verbs use what has been commonly called ‘subjunctive' forms. According to Karttunen & Peters, while subjunctive conditionals like (3) could be one case where the assumed falsity of the antecedent follows straightforwardly from the fact that it is clear that the speaker was knowingly assuming an unreal situation, some other subjunctive conditionals are used when the consequent is known to be true but whether the antecedent is true or false is not known. Or to put it another way, we may have cases where conditional sentences are expressed in the subjunctive where the consequents are known to be true but the antecedents are not presupposed to be false. In this case, a certain hypothetical relationship between the antecedent proposition and the consequent proposition may be asserted by the subjunctive conditional, and the relationship in question cannot be confirmed without further new information coming into the play. "..., [Slubjunctive conditional sentences are well fitted by virtue of their truth conditions to proposing explanations of known facts, to explaining them on the grounds that the fact stated as the consequent clause follows from the hypothesis stated in the antecedent clause"(ibid.: 5). The example given by Karttunen & Peters (ibid:4) quoted here as sentence (4) below serves to illustrate the point well, 3 (4) If Mary were allergic to penicillin, she would have exactly the symptoms she is showing. According to Karttunen 8r Peters, the consequent clause is true by itself; however, it is impossible to judge, as what has been conventionally claimed about counterfactual conditionals, that the antecedent clause is false. Rather, a certain hypothetical relationship between the condition clause and the consequent becomes clear: it is true that Mary has the exactly the symptoms she is showing but it is not known whether she is allergic to penicillin. At the time of the utterance, the speaker may simply assume that Mary‘s showing the symptoms is because she is allergic to penicillin, but whether or not this is the case is beyond the speaker‘s knowledge. Once there is new information involved, for example, one moment later the speaker becomes aware that it is true that. she is allergic to penicillin, the intended hypothetical between the antecedent and the consequent is confirmed; or otherwise disconfirmed. Support for the above classification is seen in leech (1987 116%), where English conditionals are divided into ‘real conditions‘ and 'unreal conditions‘. The ’unreal conditions', as leech claims , are used to express ‘hypothetical meaning‘. They are normally formed by the use of ‘subjunctive forms of verbs‘, namely, the Past Tense in the condition clauses (i.e. the antecedent) and would + Infinitive in the main clauses (i.e. the consequents). Or elsewhere, ‘might + infinitive‘ , ‘should + infinitive‘, and ‘could + infinitive‘ are used when difi'ering degrees of certainty are intended. For example, (5) and (6) are contrasted by their verb forms, in accordance with the conditional meanings they purport to express. (5) a. If you‘re happy, you make others happy. b. IfJohn told you that last night, he was lying. c. If my son is a genius, I‘ve underestimated him. (1. If they left at nine, they will certainly be home by midnight. 4 (6) a. If you were happy, you would make others happy. ' b. IfJohn had told you that last night, he would have been lying. c. If my son were a genius, I‘d have underestimated him. d. If they had left at nine, they would certainly be home by midnight. (p. 1 17) And intuitively, some sentences in (6a) through (6d) may be either hypothetical or counterfactual depending on whether their antecedents are assumed to be true for the sake of argument but actually are known to be false (counterfactual) or they - only purport to establish a hypothetical situation with their consequent counterparts as illustrated above (hypothetical). For example, in (6a), given a situation where the speaker knows that the addressee is not happy at the time of the utterance, the sentence will be a counterfactual conditional in which the antecedent proposition ‘you are happy‘ is false as based on the speaker‘s knowledge. However, once the speaker knows that it is the case that the addressee is in the habit of making others happy, sentence (6a) could be felt as expressing a somewhat hypothetical message between the antecedent and the consequent, because, in this case, with the consequent being true in regards to the speaker’s knowledge, whether or not the antecedent proposition is true, namely, ‘you are happy‘, would not be able to be determined; and it is possible that the addressee makes others happy even though she herself is not happy. Understood as such, Leech‘s ‘unreal conditions‘ can be seen as covering both the hypothetical and counterfactual conditions. It may seem, therefore, that while ‘unreal conditions‘ can be both linguistically and semantically distinguished from ‘real conditions‘ or the indicative conditions, the distinction between ‘hypothetical‘ and ‘counterfactual‘ is usually not that cle ar-cut. Obviously, as the English examples show above, context plays a somewhat decisive role in this distinction, but I will not go deeply into this topic 5 here. To be consistent with the purpose of this present study, I will only concentrate on the counterfactual conditions in the strict sense, namely, those whose antecedents are unambiguously understood to be assumed to be false. Thus, wherever the term 'counterfactual‘ is used hereinafter, it refers to counterfactual conditionals in this strict sense. A claim developed in the latter part of the thesis may seem to contribute to a better understanding of the distinction between counterfactuals and hypotheticals and I will present my discussion in the conclusion part of the thesis. The main purpose of the present study is to argue that in Chinese there are cases where a conditional is understood unambiguously as being counterfactual, or the linguistic constructability of counterfactual conditionals. I will analyze some counterfactual conditional sentences in modern Chinese and propose an interpretation of their counterfactuality. The thesis is developed based on the common observation made by Gorelov (1973), King (1985), and Eifring (1988) on the Chinese sentences that contain ruguo bu-shr‘if not-be‘ in the antecedents (Chapter 3) and on data studied in Li(1993b) in which certain indexical expressions are used (Chapter 4). The abbreviations used in the Chinese examples are listed in Appendix I and the gloss of some key Chinese terms in Appendix II. 2: ANOTHER LOOK AT CHINESE COUNTERFACTUAL CON STRUCTIONS 2.1 The Traditional Claim and its Philosophical Consequence The general consensus in Chinese linguistic study and even philosophical research is that Chinese usually does not distinguish indicative conditionals from counterfactual conditionals , and many linguists believe that the language never marks counterfactuality at all (Chao, 1959; Li & Thompson, 1981; Bloom, 1981, 1984; Au, 1983, 1984; Liu, 1985; Wu, 1987; Wu, 1989). According to Li 8: Thompson (1981:646fi), conditional sentences are used to express one of the following three messages: Reality, a normal inferred causal relationship, Imaginative hypothetical, what would be true in some imaginary world, and Imaginative counterfactual, what would have been true but was not, as shown in (7), (8), and (9) respectively, (7) If you heat water to 100 degrees, it boils. (8) HI saw the queen, I’d bow (I could imagine seeing the queen). (9) If you'd taken algebra, you would know this formula (but you didn‘t take algebra). Li 8: Thompson claim that while in English these three types of conditionals are grammatically distinguished from each other "in the auxiliary verbs and tense and aspect markers"(p.647), the Chinese language has no such grammatical distinctions. Therefore, what type of message is conveyed by a Chinese conditional sentence can only be ”inferred from the proposition of the second clause [consequent] and from his/her knowledge of the world and of the context in which the sentence is being used"(p.647). Therefore, it follows that given an antecedent sentence like (10) it could well be one for all these three types of conditionals as described in (a) 6 through (c), (10) Ruguo ni gei qiche zhuang 1e, if you PAS car hit FP, a. Ruguo nr’ gei qiche zhuang Ie, ni jiu zhidao gai zenmo xiaoxin 1e. if you PAS car hit FP, you then know should how careful FP "If you are hit by the car, you will know how to be more careful." b. Ruguo ni gei qiche zhuang 1e, women zenmo ban? if you PAS car hit FP, we how do "If you were hit by the car, what would we do?" c. Ruguo ni gei qiche zhuang 1e, women zenmo ban? if you PAS car hit FP, we how do "If you had been hit by the car, what would we do?" (10a) is a conditional sentence conveying the message of ‘reality‘ once the conditional relation between the two propositions is understood to be referring to the ‘real world‘; (10b) would be ‘imaginative hypothetical‘ once the conditional antecedent is understood to be related to an ‘unreal or imagined situation‘; and finally (10c) would be a counterfactual once the antecedent could have been true but was not in reality. It would be the speaker‘s extralinguistic knowledge of the world and of the consequent condition, therefore, that will determine which reading the antecedent of (10) exactly conveys. All this led Chao to claim, quite blatantly, that "there is no way of expressing counterfactuals in Chinese“ (1959:257). As a philosophical consequence, Bloom has linked this so-called absence of Chinese counterfactual expressions to what he terms the ‘cognitive schema‘ of the Chinese speaker. Based on empirical study of the relation of this so-called inability of Chinese to express counterfactual ideas to the cognitive 8 schema, Bloom comes to the hypothesis that Chinese speakers have difficulty in understanding counterfactuals. Bloom’s study has inspired a considerable amount of philosophical exchange, particularly among philosophers like Au (1983,1984), Liu (1985), and Wu (1987) who have taken pains to refute Bloom's claim. Au, Liu, and Wu, each conducting their own study, have formed natural Chinese counterfactual sentences and given them to their informants for them to understand. Based on their empirical experiments they were able to conclude that, cognitively, the Chinese speakers express and understand counterfactuals just as well as English speakers do, given the appropriate linguistic context and the extralinguistic knowledge. Nonetheless, none of them has claimed any linguistic ground for the Chinese speakers‘ expressing and understanding counterfactual sentences. 2.2 A Different Claim Though the general consensus discussed above may have become a doctrine in the area of research and study of conditional sentences in Chinese, counterexamples have been found such that there are some conditions that are able to be understood as counterfactuals only without the confusion between the indicative, hypothetical, and counterfactual interpretations. First of all, in classical Chinese, as Lu (1945) has stated, counterfactual sentences are distinguishable from the others simply based on some lexical items, or ‘relation words‘ (Guanxici), rather than on the morphological variations of the main verbs, a feature which can be seen as pertinent to the analyticity of the Chinese language. Among the ‘relation words‘ used in the antecedents, shi, ling, danshi, xiangshi, ruling, jialr'ng, feiran, buran (generally all equivalent to if), xiangshi is exclusively used in counterfactuals, though Lu has made no explanation for this unique phenomenon. See example (1 1) below, 9 (11) Xiggghi Si-jun que ke er bu nui, shushi er if Sijun refuse guest CON not accept, unknown-people CON buying,shishi guo wufiili zi baa er tai wu not use, this CAUS nation no valuable ZI-II treasure CON nation no qiangda zi ming ye (p.149) strong ZHI name FP ”If Sijun had refused to accept guests and not used the unknown people, this would have rendered the nation without valuable treasures and the nation without strong name." Although there is no other literature immediately available for a theoretical account of this of Lu‘s finding that , intuition provides positive support for the sentence to be interpreted so in the original context. Ample evidence for the linguistic constructability of counterfactuality is also available in modern Chinese. Lu (ibid.:140-2), in classifying Chinese conditionals into 'condition sentences’ and ‘hypothetical sentences’z, as. exemplified by (12) and (13) respectively, has claimed that both sentences in (13) express false propositions in the antecedents, ‘contrary to facts‘, in his words, ( 12) Ni yao bu renshi ta, wo keyi gei ni feng jieshaoxin. you if not brow her, I may give you CL reference-letter. “If you do not brow him/her, I may give you a reference letter.” (13)a. W0 yao bu xiangxin ni, wo dao bu hui ba zhe hua gaosu ni 1e. I if not trust you, I but not willBAthiswordtell youFP “IfI had not believed you, I would not have told you about this word.“ b. Yaoshi wo renshi ta, wo hebi hai Iai qiu ni jieshao. if I brow her, I why still come ask you recommend “IfI had brown her, why would I have come to ask you to recommend me.“ 10 Similarly, Hu (197 9:367 -8) has classified Chinese conditionals into ‘hypothetical conditionals‘3 and ‘counterfactual conditionals‘, as shown in (14) and (15) respectively, (14) Ruguo mingtian bu xiayu dehua, women jiu qu canguan gongye zhanlanhui. if tomorrow not rain EMPH, we then go see industrial exhibit ”If it does not rain tomorrow, we will go to see the industrial exhibit‘." (15) Ruguo zuotian bu xiayu, women zao jiu qu canguan gongye if yesterday not rain, we early then go see industrial zhanlanhui Ie. exhibit FP "If it had not rained yesterday, we would have definitely gone to see the industrial exhibit." where Hu argued that (14) expresses a hypothetical condition which may have the truth conditions of an indicative conditional but (15) only of a counterfactual. More recent study has provided further support for the plausibility of the linguistic constructability of counterfactual conditionals. First of all, conditionals which contain the negator bu-shi ‘not-be‘ in the antecedent, as recorded in Gorelov (1973), King (1985), and Eifring (1988) are counterfactuals only. For example, sentence (16a) below, (16)a. Ruguo bu-shi ta gaosu ni, ni jiu bu hui zhi-dao zhe shi. if not-be she tell you, you then not will know this matter "If it were not the case that she told you, you would not have known this." As intuitively the case, an indicative reading for sentence (16a) would be incompatible here, particularly when contrasted with (16b) which is the normal indicative construction, ll (16)b. Ruguo ta bu gaosu ni, ni jiu bu bui zhi—dao zhe shi. if she not tell you, you then not will brow this matter "If she does not tell you, you will not brow this.” The oddness in the indicative reading of (16a) is felt even more clearly particularly when some other linguistic elements are added to accompany the indicative reading, as shown in (17) below, (17) (?) Ruguoj_ia_ng_1_ai/#m_i_nm'a_n bu-shi ta gaosu ni, ni jiu m if future/tomorrow not-be she tell you, you then definitely bu hui zhi-dao zhe shi. not will brow this matter "If she does not tell you in the future/tomorrow, you will not brow this.“ The modifications in (17) are the additions of the future time operator in the antecedent, jiarrglaningtian‘future/tomorrow‘ and the phrase expressing modality, kendingdefinitely‘. In this case, the sentence could be understood as expressing an indicative conditional relationship between the two propositions such that if in some future time she does not tell you, you will definitely not know the matter. Although the sentence appears to be doing the job in expressing the intended indicative reading, a sense of un-naturalness remains with the sentence and a more natural construction would see the elimination of the 8121' altogether, as shown in (18) below, (18) Ruguo W ta bu gaosu ni, ni jiu kenggg' bu hui if future/tomorrow she not tell you, you then definitely not will zhi-dao zhe shi. brow this matter "If she does not tell you in the future/tommorw, you will not brow this." 12 Moreover, there are sentences which have been observed (Li, 1993b) as intuitively conveying perspicuous counterfactuality, for example, sentence (19) when understood as referring to a past event, and (20) below, (19) Ruguo na-tian wo zuo zai nar, ta jiu hui bac—shou duo 1e. if that-dayl sit in there, she then will comfortable much FP ‘If I had sat there that day, she would have been more comfortable.‘ (20) Ruguo ni zuotian Iai 1e, jiu bu hui chu zhe shi 1e. if you yesterday come FP, then not will happen this matter FP "If you had come yesterday, nothing would have happened." Based on the discussion above so far, the point should have become clear: there are, in Chinese, linguistic constructions that convey counterfactual message. However, given the observations made as such, there has been no attempt in Chinese linguistics for working on this unique linguistic phenomenon. In this study, I will limit my effort to provide an account for the counterfactuality in those sentences recorded in Gorelov (1973), King (1985), and Eifring (1988), which, to the best of my browledge, has completely been left out of previous studies, and those as studied in Li (1993b), which I feel is rather prematurely concluded and needs to be further developed. 3: [bu-shi [pm] ] AS A LINGUISTIC STRUCTURE FOR COUNTERFACTUALITY 3.1 Description of the Data - While the most commonly used ‘if phrases in modern Chinese are ruguo and yao, there are some others that are often used interchangeably with ruguo or yao, such as yueshi, tangyue, tangshi, jiayue,jiaru, and yaoshi. Although none of these ‘if‘ words alone could contribute significantly to the counterfactuality of the conditional sentences they introduce, it is intuitively the case that yueshi is least natural in use with the [ bu-shi[ p ...] ] structure which I am going to study in this chapter. Again, diachronic Chinese study may provide some hint to this phenomenon, but I will not take up the discussion here. For the purpose of unity, I will use ruguo as the cover term for all these candidates for the 'if' phrase. Early evidence that shows ruguo bu-shi ‘if not-be‘ marking a conditional sentence with counterfactuality comes from Lu (1945). There, sentence (21) was said to be a counterfactual, however, it seems that Lu was not aware of the strategic function of bu-shi and consequently made no explanation. (21) Ta zao (ruguo! bu-shi tianliangjiu dengshen, na neng shangwu jiu she if not-be daybreak then set-off, how can morning then gandao.(ibid.:1 76) arrive "If she had not set off at daybreak, how could [she] have arrived in the morning?" Subsequent assertions of the counterfactuality of sentences with ruguo bu-shi have seen recorded in Gorelov (1973:218) and King (1985:33), but it is not until Eifring (1988) before ruguo bu-shi has been studied as almost a lexical marker for counterfactuality in Chinese conditionals. Eifring explored Chinese 13 l4 conditionals of the type containing negators mei (you) ‘not (have)’ or bu ‘not’. Making changes with the negator mei (you) ‘not (have)’ or bu ‘not’ for bu-shi ‘not- be’ in a conditional sentence, as he has concluded, would turn a conditional sentence into one with only a counterfactual interpretation. Thus, sentence (22) below, (22) Ruguo ta meigzou) dao meiguo qu nianshu, wojiu hui zhao ta tan-tan.(11)4 if she notghave) to USA go study, Ithen will find her chat-chat would mean (22)a. "If she has not gone to the US. to study, I will contact her to have a chat." or (22)b. "If it were not for the fact that she had gone to the US. to study, I would (have) contact(ed) her to have a chat.“ But if mei(you) is replaced by bu-shi as shown in (23), the sentence can have counterfactual interpretation only, namely, (22b), (23) Ruguo bu-shi ta dao meiguo qu nianshu (le), wojiu hui qu zhao ta tan-tan.( 12) if n_9_t_l;e she to US. go study (FP), I then will go find her chat-chat It suffices to say that this new deployment makes both a semantic and a syntactic difference from the original sentence meaning. This, unfortunately, was left unstudied by Eifring. Here, I am going to propose my explanation with the analysis of the semantics of the word shi in bu-shi and see how the difference is generated. 3.2 The Semantics of 8121' Insight into the nature of the combination of the negator bu'not' and the seeming copula shite be‘ came from Gao (1957). In his study (pp.426-40) on the negation of propositions, Gao has classified Chinese sentences into three main categories based on their predicates: nominal, adiectival, and verbal sentences. 15 Gao argues that while the negator bu‘not‘ alone can be used to negate the propositions expressed by both adjectival and verbal sentences, the unit negator of bu and the copula slri, namely, bush, is used exclusively for negating a proposition in a nominal sentence, as shown in (24) through (26) below, (24) Zai meiguo bu shou meiguo falu, zhe Ire bu shi nao zhe wanr de in US not follow US law, this EMPH not be make PM joke NOM “To be in the US and not to follow the US laws, this is not ajoking matter." (25) Qieshue na Chaogai ba ren jie le Cai dashi de shengri liwu, to-say that Chaogai eight pe0ple rob ASP Cai Master DE birthday present may xunchang xiaoabi. not to-be trivial matter ”To mention that Chaogai and seven others had stolen Master Cai‘s birthday presents, [that] is not a trivial matter.“ (26) Wu jiaoshou tingdao waimian shengyin, _l_)_r_rs_hr_' bue ren, shi wo hurriia Wu professor hear outside noise, 99133; other people, be my Hurriia he jinr (p.429) and J im- "Professor Wu heard some noise outside. [That] is not anyone else, but my Hurriia and Jim.“ The contrast between (27a) and (27b) further supports Gao‘s (ibid) argument, such that whenever a nominal predicate is not negated by the combination of the negator bu with the copula shi, ungrammaticality will occur as shown in (27b). (27)a. Ta renwei dapo 1e beizi bu §l_r_i tade cuo. she think break ASP glass not b_e_ her mistake "She thinks that it is not her fault that the glass is broken." 16 b. *Ta renwei dapo le beizi bu tade cuo. she think break ASP glass not her mistake. Given this unique requirement for the negator and the copula to be used together in negating a nominal predicate as shown above, an adequate explanation is needed and this is the main issue in the following section of the thesis. 3.2.1 The Two Components of shi Li and Thompson (1977; see also Wang, 1957) assert that shi began its career as a copula as a demonstrative pronoun. In classical Chinese, it came to be used as a pronoun resuming the Theme in a "Theme-Predication construction“, as shown in (28) below, in which the Theme-Predication construction‘ is the first part of the sentence, namely, ‘Ji ru qi sheng you ru qr” si,...‘, (28) Ji yu qi sheng you yu qi si, _s_hr_' buo ye. already wish him live also wish him die, DEM indecision FP "Wishing him to live while wishing him to die, that is indecision."(p.424) According to Li and Thompson the Theme-Predication construction‘ gradually developed into ‘a subject-predicate construction with the anaphoric demonstrative pronoun sbi being reanalyzed as a copula‘ (ibid.:424). Accordingly, it becomes a special term forming a compound predicate which requires a complement of a nominal phrase or a clause. Further, I assume that shi functions to fill in the otherwise missing copular position which bridges a sentential Theme with a nominal predicate when the Theme is overtly expressed; or it functions to serve as a sentential Theme when such a linguistic element is missing while at the same time it retains its status as a copula. Therefore, sentences (29) and (30) below are felt odd, 17 (29) *Fan le cuowu jiu gai, zhe jian hao shi. makeASP mistake andcorrect,this a goodthing “It is a good thing to correct your mistakes.“ (30) *Tingshuo youren zai zuo baogao, yuanlai Wang Laoshi. be-told someone ASP deliver lecture, after-all Wang Teacher ”We were told that someone is delivering lecture; after all, it is Teacher Wang.“ The oddness in these two sentences results from the absence of 3111' before a nominal predicate, say, jian hao shi ‘a good thin§ and Wang Laoshi ‘Teacher Wang’. In (29), slu' is needed to link the overt Theme zhe‘this‘ which is anaphoric to the proposition expressed in the preceding part of the sentence, Fan 19 cuowu jiu gai ’to correct mistakes‘, with the nominal predicate; and in (30) it is to identify the missing Theme which is an implicit reference [someone]. 3.2.2 Definite Propositional Description (DPD) Following the argument established so far, I further assume that shi in the lexicon has two semantic components: the copular and demonstrative, as shown in (31 ), (31) 8121': [+ copula (COPll [+ demonstrative (DEM)] And this is the very point to begin my exposition of the semantics of shi in [bu- shib...]]. I start my chain of reasoning with the semantic as well as philosophical notion of ’definite description‘. Russell( 1905) posits that in a proposition of the form ‘The F is G‘, ‘the F’ is a ‘definite description‘ and the sentence can be analyzed as a conjunction of following three propositions5: 18 (32) a There is at least one F; bThereisatmostone F; and c. Every F is G. And, accordingly, the existence of F is implied which follows by the use of the definite article, namely, there is exactly one F, no more and no less. By the same token, therefore, the definite article in the definite description, ‘the present king of France' in (33) would purport to imply the existence of a unique present king of France. Proposition (33a) is obviously false because there is not at least a present king of France, thus the conjunction of the three propositions will be false and thus the sentence is false. (33) ‘The present king of France is bald’ a. there is at least one present king of France; b. there is at most one present king of France; and c. every present king of France is bald. P. F. Strawson( 1990), a leading member of the descriptive analytic philosophers, however, has argued that the uniqueness in Russell‘s definite description theory is only presupposed rather than asserted. For this argument, Strawson introduced the notion of ‘implication‘ which is, used in this way, what is known in modern logic as ‘presupposition‘ (Caton, 1959), and stated When we begin a sentence with “the such-and—such’ the use of ‘the’ shows, but does not state, that we are, or intend to be, referrirrgio one_p_articular individual of the species_‘such-and-such’ which particular individual is a matter to be determined from context, time, place, and any other features of the situation of utterance. (p.325, underline added) In contrast to Russell‘s analysis of the sentence (33) repeated here as (34), Strawson’s theory can be, therefore, recapitulated as follows: (34) The present king of France is bald. First of all, (34) is a sentence, a linguistic string of words arranged in a grammatical order, hence a meaningful one here. This is Strawson’s notion of ‘sentence’, and l9 similarly, the expression of ‘the present king of France’ here is ‘an expression’. When (34) is used to make a statement it is to have its truth value and thus (34) is true if, presumably, it was uttered during the reign of Louis XIV, because at that time there was such a king of France, and he was bald; when (34) was uttered at the present time, say, February 1995, it would be neither true nor false because one presupposition of the sentence is ‘There is a king of France’ which is false, and thus the question of whether the sentence istrue or false simply does not arise. Following Russell‘s idea of definite description and Strawson’s theory of presupposition, I proceed with my reasoning. While it is obvious that the structure of such a definite description usually takes the form of a common noun phrase (CNP) plus a determiner which is a definite article, ‘the such-and-such‘ in Strawson‘s term, I assume that substituting this definite article with demonstratives like ‘this’ and ‘that’ would also yield definite description given the denoting properties of those demonstrativess, though, in some cases but not necessarily all, it would require some kind demonstration, or pointing, to accompany with the reference (Kaplan,l977 :491). Thus, I claim that in (35), (35) They went to this/that bar. “this/that bar’ will presuppose that there exists a bar. . It becomes interesting now to push this point a little further: what will happen if the CNP of a definite description becomes a PROPOSITION or SENTENCE? For ex ample, given a linguistic structure of (36), (36) [the/thislthat [PROPOSITION] what would be some speculation on the embedded proposition? I feel convinced that since a demonstrative in a definite description expressed in a sentence purports to presuppose the existence of an entity described by the definite description, such an element as used in (36) will purport to presuppose that the ' 20 PROPOSITION is true. I call this construction ‘Definite Propositional Description’ (hereinafter referred to as DPD). Coincidentally, Kiparsky 8: Kiparsky (1971) have found that there are compound sentence constructions in English where certain matrix clause predicates may presuppose the embedded propositions to be true. They have identified two kinds of predicates, 'factive‘ and ‘non-factive’, to which different semantic properties as well as syntactic properties are correlated. For example, the pairs in (37) and (38) (quoted from ibid.:l47), (37 ) It is odd that it is raining. It is likely that it is raining. ( 38) I regret that it is raining. I suppose that it is raining. According to this study, the first sentence in each pair has a factive main predicate, namely, rbe odd‘ and ‘regret‘, and the speaker of these utterances is understood to presuppose that the embedded proposition, namely, ‘it is raining’, is true. In contrast, the second sentence in the above pairs has non-factive predicate, namely, ’be likely’ and ’suppose’, and thus these two sentences do not have the semantic property as to presuppose the truth of the subsequent proposition. Factive predicates are listed as regret, be aware (of), grasp, comprehend, take into consideration, take into account, bear in mind, ignore, make clear, mind, forget (about), deplore, resent, care (about), and those that are listed as non-factive predicates include suppose, assert, allege, assume, claim, charge, maintain, believe, conclude, conjecture, intimate, deem, fancy, figure (Ibid.:l45). It seems, therefore, that both Kiparsky 8: Kiparsky‘s study and my claim of DPD have come to the same support of the idea of the presupposition of the truth of a proposition. - In the former case, this occurs in a compound sentence based on certain main 21 clause predicate, and in the DPD this occurs in a construction based on certain anaphoric demonstrative element. Now let us continue with the semantics of shi. As it has been assumed above, shi is understood as having two semantic components in the lexicon: the copular and the demonstrative. Based on the notion of the DPD, the proposition after ahi is thus presupposed to be true. Consequently, the negation of a true proposition would naturally lead to a false proposition, and to make an assumption on a false proposition would naturally lead to a counterfactual conditional sentence. Take the following for example, (39)a. Ruguo ta bu §_lu_j zhongguorerr, if she not be Chinese, "If she is not a Chinese, ..." b. Ruguo M ta fl zhongguoren, if not-be she be Chinese, "Ifit were not the case that she is a Chinese, ...“‘ or, "If she were not/had not been a Chinese, ...“ c. Ruguo ELSE ta bu fl Zbongguoren, if not-be she not be Chinese "If it were not the case that she is not a Chinese, ..." One immediate contrast among the above sentences is the fact that whenever the unit negator, say, bu-shi is seen to be followed by a clause, for example in (b) and (c), the proposition it expresses is intuitively assumed to be true. Therefore, in (b), itis assumedtobetruethat‘SheisaChinese’ andin(c)itisassumedtobetrue that ‘She is not a Chinese‘. This is not the case for (a) where shi is followed by only a nominal phrase and is only a copula without its functioning as an anaphoric demonstrative. 22 As the last step in my reasoning, since the negation of a true prOposition results in a false proposition, the negator bu would make all the above true propositions false. And a conditional sentence made on a proposition which is assumed to be false can only "be counterfactual. The general scheme, first advanced in Li(1993a), can be sketched in (40), (40) [ruguo [q hu- [3 shi [p 8 ...] l 1,... a. 8 is a proposition; b. B is a true proposition = slri presupposes 8 to be true; c. o: is a false proposition = the negation of I3; and d. (40) is a counterfactual proposition = the assumption of o: The oddness in the (41) and (42) below provides further evidence for the validity of the argumentation spelled out above: (41) *Ruguo bu-zhi bu-shi ta zuotian lai 1e meiyou, ni de cheng’i if not-know not-be she yesterday come ASP not, you DE result yexu hui geng bao maybe will more good *"If it were not the case and not brown whether John came yesterday or not, your result could have been better.“ (42) *Ruguo bu-shi ta mingtian lai M wo yexu jiu bu zheyang if not-be she tomorrow come not come, I maybe then not this zhunbei le. prepare FP *“If it were not the case that she will come tomorrow 9% I would not have prepared like this.“ The oddness in (41) and (42) lies in the fact that the phrase bu lai 'not come' (the undelined part) contradicts the semantic function of bu-shi, which, according to 23 DPD mechanism, presupposes a definite truth value, namely truth, for the subsequent proposition. Chinese lexicology provides further support for the above argument, where ruguo‘if and bu-sbi ‘notrbe’ have been taken together as an idiom, glossed as ‘if it were not the case ...’ (HanYing CiDian, 1981:804). For example, (43) YaolLugol bu-shi tamer: yr'nggandi qiangiiu guojia caicban, na sunshi kejiu if not-be they bravely rescue state property, then loss may then da le.(ibid.:p.804) enormous FP "If it were not for their courage in rescuing the state property, the loss would have been enormous“ (ibid.:804). (44) Hugo bu-shi ta zhiyin, women jiu mi lu le. (ibid.:57 9) if not-be she guide, we then miss road FP "If it were not for her, we would have gone astray." (ibid.:57 9) A final word on ruguo bu-slu’ . There are different positions where ruguo bu-shi can occur in a sentence, and the phrase can have different forms of stylistic variation. It is observed, however, that the function of its marking counterfactuality is not altered. Based on Eifring, the following generalization can be made: A. ruguo bu-shi is positioned after the subject: (45) Ta mo bu-slri dao meiguo qu nianshu (le), wojiu hui qu zhao ta tan-tan. she if not-be to US go study . (FP), I then likely go find her chat-chat. ‘If she had not gone to the US. to study, I would contact her to have a chat’. B. ruguo bu-shi is separated by the subject: (46) 12mg ta M dao meiguo qu niansbu (le), wojiu hui qu zhao ta tan-tan. fl she m to US. go sturdy (FP), I then likely go find her chat-chat 'If she had not gone to the US. to study, I would contact her to have a chat.’ 24 C. yao-bu-shi occurs together standing for bu-shi: (47) Ni Logo yao-bu-shi shuo 76.4, wo zenme hui qu dingzhuang kezhang ne you if if-rrpt-be say 76.4, 1 how will go refute department-head FP ‘If you had not said 76.4, how could I have gone to refute the department head?’ _ D. ruguo/yao can be dropped: (48) __l_3_u_-s_h_i ta dao meiguo qu nianshu (le), wojiu hui qu zhao ta tan-tan. m she to US go study (FP), I then will go find her chat-chat. ‘If she had not gone to the United States to study, I would contact her to have a chat.‘ E. yaosbi stands for yao: (49) M ta Mdao meiguo qu nianshu (1e), wojiu hui qu zhao fl she not be to US. go study (ASP), I then will go find ta unbunt her chat-chat ‘If she had not gone to the United States to study, I would contact her to have a chat.‘ 3.3 A Problem in Eifring's Study and Discussion A question that can be raised against Eifiing’s discussion on the counterfactuality marking of the phrase ruguo bu-shi is in his treatment of sentences like (50), where he claimed that (50b) and (50c) are counterfactual conditionals, but (50a) is not, (50)a. W0 bu zhidaojintian zhiban de shi shei. Hugo bu-shi women I not brow today on-duty NOM be who. if notrbe she, we iiu iin bu gu le.(15)7 then enter not go FP 25 "I do not brow who is on duty today. If it is not she, we will then be unable to get in.” b. Jintian zhiban de yiding shi ta. fiyguo bu-slri targwomen today on-duty NOM certainly be she. If not-be she, we jig jin bu gr; le. (16) then enter not go FP “The person on duty today is certainly she. If it were not she, we would then be unable to get in." c. Zhen ganxie Ding Xiansheng. _R_ugro bu-shi tg, women jiu 1i_n_ really thank Ding Mr. if not-be he, we then enter bu g le. (17 ). not go FP “We really thank Mr. Ding. If it had not been for his help, we would have been unable to get in.‘ It is felt strange as to the reason that, given the exactly same linguistic structure (the underlined parts), one sentence could be, rather arbitrarily, counterfactual and the other could not. And the conclusion Eifring has made in his study is simply that ruguo bu-shi does not always mark a conditional with counterfactuality. Actually Eifring has only ofi‘ered an incomplete picture here. It is true that ruguo bu-shi does not always do the job, for example, as what the DPD rules out, when shi is followed by a nominal phrase rather than a clause. This is shown in the above example (39) repeated here as (51), (51)a. Ruguo ta bu sin; zhongguoren, if she not be Chinese, "If she is not a Chinese, ..." 26 b. Ruguo bu-shi ta §l_r_i zhongguoren, if not-be she be Chinese, ”If it were not the case that she is a Chinese, ...”‘ c. Ruguo M ta bu gag Zhongguoren, if not-be she not be Chinese “If it were not the case that she is not a Chinese, ..." It is said above that the phrase bu-slri in (51a) is different from that in (51b) and (51c). In the former case, shi is only followed by a nominal predicate, say, zhongguoren‘a Chinese’, so that DPD would be beyond reach for any possible account as to whether the sentence is a counterfactual or not. This stands in contrast to the latter cases, (51b) and (51c), where shi is followed by clauses or propositions which DPD is supposed to account for. The problem with the sentences in (50), therefore, is not simply an arbitrary matter. According to DPD, none of the sentences can be determined only based on their linguistic structure, though the phrase ruguo bu—shi is involved, because, what immediately follows the shi is a pronominal phrase, say, ta’she’. However, the reason why (b) and (c) are counterfactual is that they involve different contextual information in their interpretation, say, ’I do not brow who is on duty today’ in (50a), ’The person on duty today is certainly she‘ in (50b), and ‘We really thank Mr. Ding’ in (50c). Obviously they provide different contextual information which, in turn, helps decide the counterfactual readings. In normal discourse circumstances, each of these preceding sentences would convey difi’erent implications which, in turn, contribute to a sentence’s being counterfactual or non- counterfactual. (Coincidentally, the problem in question here is also related to another category of counterfactuals in Chinese which I am going to study in the next chapter, namely, the indexical counterfactuality.) For example, the (b) 27 sentences in (52) through (54) are felt to be implied, among other possibilities, by the corresponding (a) sentences: (52) a. I do not brow who is on duty today. (The first part of (503)) b. It is possibly the case and it is possibly NOT the case that she is there. (53) a. The person on duty today is certainly she. (The first part of (50b)) b. She is certainly there. (54) a. We really thank Mr. Ding. (The first part of (50c) b. Mr. Ding has done something that ’we’ appreciate. This being the case, these implications will affect the interpretation of the subsequent conditionals. Thus, (52b) will presuppose a possibility by which the antecedent of (50a) is possibly true; (53b) will presuppose a necessity by which the antecedent of (50b) is almost definitely to be false; and (54b) will presuppose a counterfactuality by which the antecedent of (50c) will be deemed false. Therefore, the exactly identical Chinese conditional constructions will naturally have different interpretations, namely, (50a) an indicative, (50b) and (50c) counterfactuals. 3.4 Summary In this Chapter of the thesis, I have presented conditional sentences of the linguistic structure of [bu-shi [pm ] ] which are intuitively determined as counterfactuals. Based on this, I have analyzed the said counterfactuality by proposing a notion of Definite Proposition Description (DPD). Finally, a problem in Eifiing’s study is identified and solved, which, in turn, adds to further defense for the validity of DPD. 4: INDEXICAL COUNTERFACTUALITY (1C) IN CHINESE The counterfactual conditionals studied in the previous chapter are those that are expressed in the linguistic frame of [ bu-shi [p ...] ] where the key lexical word shi is claimed to convey a presupposition that the proposition which follows it is true. That has constituted the first set of data I have used in arguing for the linguistic constructability of counterf actuality in Chinese conditional sentences. The second set of data I am going to work on is based on Li (1993b) in which certain indexical expressions are observed to be crucial in determining the counterfactuality of certain sentences. In the first part of this chapter, I will first describe in more details the data. In the second part, I will review the relevant theories in order to establish a sound ground for my account for the observed counterfactuality, which is explained in the last part. ‘ 4. 1 Description of Data In my previous study (Li,1993b), I have found that while a conditional sentence may have either indicative or counterfactual interpretations, or even a hypothetical one, certain indexical expressions may serve to disambiguate such a sentence. For example, (55)a. Ruguo ta meiyou dao meiguo qu nianshu, wojiu hui qu zhao if she not-have to US go study, I then will go find ta tan-tan. her chat-chat "If she does not go to the US to study, I will go have a chat with her." or, "If she had not gone to the US to study, I would have gone have a chat with her.“ 29 b. Ruguo ta meiyou dao meiguo qu nianshu, yv_o 2.11M jiu hui if she not-have to US go sturdy, I_ ygs__te_rday then will an zhao ta tan-tan. go find her chat-chat "If she had not gone to the US to study, I would have gone have a chat with her yesterday." (56)a. Ruguo ta lai le, jiu bu hui chu shi 1e. if she come FP, then not will happen thing FP ”If she comes, nothing will happen.“ or, "If she had come, nothing would have happened." b. Ruguo ta lai le, jiu bu hui chu QQ shi le. if she come FP, then not will happen tl_ri§ thing FP ”If she had come, this would not have happened." (57)a. Ruguo wo zuo zai nar, ta jiu hui haoshou duo le. if I sit in there, she then will comfortable more FP "If I sit there, she will feel more comfortable." or, "If I had sat there, she would have felt more comfortable.” b. Ruguo 9352431 WO zuo zai nar, ta jiu hui haoshou duo le. if My I sit in there, she then will comfortable more FP "If I had sat there yesterday, she would have felt more comfortable." Intuitively, sentences (55a) can well be indicative, counterfactual, and hypothetical conditionals, but (56b) turns into a counterfactual only; in (56b) only the counterfactual remains; and finally, the counterfactual (57b) stands in the same contrast to (57a) as the other two cases. A closer look into the difi'erence between the (a) sentences and their (1)) counterparts in the above examples reveals that all the (b) sentences involve using indexical expressions, sometimes called ’deictic expressions‘. For ex ample, 30 the first person pronoun, namely, wo’I’; past time expression, namely, zuotian’yesterday‘; and the demonstrative pronoun, namely, zhe’this‘. It appears that these indexical expressions contribute substantially to the counterfactuality observed as such and I use the term ‘Indexical Counterfactuality' (hereinafter referred to as IC) to refer to these counterfactual conditionals. 4.2 Indexicals and Context First of all, it can be said that the meaning of sentences like (52) through (54) cannot be successfully grasped without assigning due reference to the indexical expressions. This would presuppose that the speakers properly use, and the hearers reasonably assume that the speakers do, the indexical expressions in their uttering such sentences. Furthermore, since the significance of these indexical expressions is "context dependent", it would be, therefore, necessary to understand the contextual parameters to which the indexical expressions tend to be related. In his study of context-dependent semantics, Sag (1981), following D. Kaplan, has identified five extralinguistic factors which specify a context. These factors are ’the agent of the context‘, namely, the speaker of the utterance, ’the audience’, ’a time of utterance’, ’a place of utterance‘, and ’a world of utterance‘ (p.280). In light of Kaplan‘s (1977 :499) original illustration, the truth value of such a sentence like in (58) (58) I like that. can only be decided in a given context, such that the agent of the context, namely, ‘1‘, likes whatever the speaker intends to refer to, either by act of pointing or not, at the place of the utterance and at the time of the utterance. In regards to the demonstrative pronoun ’that’, according to Sag (ibid.:283), common ‘cultural 3 1 assumptions and mutual browledge’ as shared by the language users will enable them to assign adequate reference to the indexical expression. Probably no one would doubt that the commence of indexical expressions is inherent to all human languages. Chierchia 8: McConnell-Ginet (1990), who have followed Bar-Hillel (1954), have defined indexical expressions as “... terms whose contribution to propositional content depends on the context in which they are used, and their meaning consists in specifying the aspect of context that determines what contribution to content they will make." (p. 265) Accordingly, therefore, the first person pronouns such as ’I‘ or ‘we’, for ex ample, are indexical expressions because their reference is entirely dependent on the context in which they are used. Moreover, by taking their meaning fi'om the context, which contains the speaker or speakers of a certain utterance, they contribute to the sentence being uttered with that meaning. Kaplan( 1977) has listed indexicals from personal pronouns such as ’I’, ‘my‘, ‘you’, ’he’, ’she’, ‘it'; demonstrative pronouns such as ‘that’, ’this‘; place adverbs such as ‘here’, ’there‘; time adverbs such as ‘now‘, ‘tomorrow’, ’yesterday’; and even the adjectives such as ‘actual’, ‘present’ (p.489). In the next discussion of IC (Indexical Counterfactuality) in Chinese, I will show what these indexicals do in contributing to a conditional sentence’s being counterfactual. A The alleged counterfactual conditionals that I am going to study which involve use of certain indexical expressions should be understood as occurrences of natural conversational linguistic behavior. Thus, Grice‘s (1975, 1978) Maxims of Conversation are at stake in interpreting the context and the nature of conversation in which these Chinese conditionals occur. Grice’s famous Maxims of Conversation(Grice, 1975)8 are shown as follows, 32 a. Relation: Be relevant. b. Quantity: Be only as informative as required for current conversational purpose. c. Quality: Say only what you believe true and adequately supported. d. Manner: Be perspicuous: be brief and orderly and avoid obscurity and ambiguity. To sum up what I have done in this part so far and to set out my purpose more evidently, I have reviewed the definition of indexical expressions which I am going to refer to in my discussion of IC on Chinese; furthermore, I have proposed that understanding of the Chinese conditional sentences like those in (52) through (54) is assumed to be consistent with Grice‘s Maxims of Conversation, such that other uses of utterances are limited as they are not committed to the purpose of conversational information exchange as required by the Grice‘s Maxims of Conversation. 4.3 IC (Indexical Counterfactuality) in Chinese I claim that indexicality contributes to counterf actuality in Chinese conditionals. This contribution is related to the point which is central to defining counterfactuality, namely, when the truth value of such a conditional antecedent is assumed by the speaker and properly processed by the bearer as such. A common definition of counterfactual conditionals, in contrast to hypothetical ones, is that such sentences presuppose the falsity of their antecedents, as I have shown in the first chapter of this thesis. However, there are other ways to understand a conditional sentence as a counterfactual and the strategy I use in attacking the IC in Chinese is following the common law of logic, brown as modus tollens. According to Adams (1975), the counterfactual modus tollens inference is described as follows, (59) Not B. If it were the case that A, then it would be the case that B. Therefore not A. (p.116) 33 Accordingly, once the consequent proposition is understood to be assumed to be false by the speaker, and once the event of the assumption is made clear to the hearer, the conditional sentence will be taken as a counterfactual one; consequently, the falsity of the antecedent proposition can be logically inferred and the conditional sentence will be understood counterfactually. To illustrate, look at the above example (56b) repeated here as (60), (60) Ruguo ta lai .Ie, jiu bu hui chu fig shi le. if she come FP, then not will happen gig thing FP "If she had come yesterday, this thing would not have happened." Based on the properties of indexical expressions such as ’this thing' used here, it can be inferred that both the speaker and the hearer have some shared browledge as what ’this thing’ refers to, say, something that already has happened. It follows that, upon hearing (60), the hearer will understand that the speaker was assuming the situation of the antecedent, namely, when it is the case that she came yesterday; however, the demonstrative indexical in the consequent implies that the ’thing’ under discussion has already happened. And an assumption based on this indexical implication is obtained as shown in (61) below, (61) It is not the case that this thing has not happened. It follows that the consequent in (60) must be false. By modus tollens, therefore, the hearer will naturally infer that the antecedent condition is also false, namely, (62), (62) It is not the case that she came yesterday. Therefore, sentence (57) will be processed as expressing a false proposition in the antecedent, which defines the sentence as a counterfactual conditional, to which a normal intuitive interpretation fully applies. 34 4.3.1 Indexicals Related to the Speaker and Past Time Reference I continue my illustration of the above point with example (55) above repeated here as (63). (63)a. Ruguo ta meiyou dao meiguo qu nianshu, wojiu hui qu zhao ta tan-tan. if she not-have to US go study, I then will go find her chat-chat “If she does not go to the US to study, I will go chat with her." or, "If she had not gone to the US to study, I would have gone chat with her." b.Ruguo ta meiyou dao meiguo qu nianshu,}LQ guo_tia_q jiu hui qu zhao if she not-have to US. go study, 1 yesterday then will go find ta tan-tan. her chat-chat "If she had not gone to the US to study, I would have gone chat with her. “ In (63b), there is one linguistic element added, namely, the indexical expressing a past time reference, zuotian‘yesterday’. By having this indexical element, the speaker is understood to be truthfully reporting a causal situation in which the speaker would go to chat with her yesterday but did not. The hearers of this utterance will naturally assume that the speaker brows whether or not she went to see somebody yesterday. In the case of (63b), the hearer will naturally assume that the speaker did not go to chat with her yesterday. And such an assumption will further lead to the conclusion shown in (64) below, (64) It is not the case that the speaker went to chat with her yesterday. Based on (64), the consequent of (63b) will be assumed to be false. By modus tollens, it follows that the antecedent proposition is false, namely, (65), (65)a. It is not the case that she has not gone to the US to study. i.e., b. It is the case that she has gone to the US to study. 35 Conditional sentences like (63b) above are observed to have the two indexical elements, past time reference and the first person pronoun, and in this case they are equally necessary in contributing the counterfactuality to the sentence in which they occur as shown in the following sentences, (66) Ruguo ta meiyou dao meiguo qu nianshu, Li Laoshi zuotian jiu if she not-have to US. go study, Teacher Li yesterday then hui qu zhao ta tan-tan. will go find her chat-chat ‘If she has not gone to the US to study, Teacher Li went to chat with her yesterday.‘ or ’If she had not gone to the US to study, Teacher Li would go to chat with her yesterday.‘ (67) Yaoshi Lin guniang shuo zhexie hua, wo ye he ta shengfen Ie.(Wu,l989:86) if Lin Lady say these words, I then with her angry FP ’If Lady Lin has said these words, I then will be angry with her.’ or If Lady Lin had said these words, I then would have been angry with her.‘ In (66), while the indexical past time reference exists, the absence of a first person pronoun in the consequent makes it once again ambiguous between indicative, hypothetical, and counterfactual readings. This is so because the context itself will not warrant that the hearer of this utterance will arrive at the inference that the speaker brows that Teacher Li did not go yesterday, which is only one possible meaning of the sentence. In (67), where the first person pronoun is present in the consequent but the indexical past time reference is not, the conditional resumes its ambiguity. Support for this argument can be found in English conditional sentences with both of these two indexical elements, like in (68), (68) *If ..., I went to talk with her yesterday. 36 Most commonly, (68) is felt anomalous, regardless of what condition the speaker is based on. This shows that whenever these two indexical labels are present at once in the consequent, namely, the past time reference and the first person pronoun which also plays an agentive role in the sentence, namely, the subject of the sentence, the consequent proposition is counterfactual. In the case of (68), therefore, it requires that the verb ’go’ have a subjunctive perfective form, as shown in (69), (69) If ..., I would have ggne to talk with her yesterday. In contrast, other personal pronouns or indexicals expressing temporal points other than a past one will yield different results, shown in (70) and (71), (70) If.., My went to talk with him yesterday. (71) If...,1will go to talk with him tomorrow. which may be perfectly grammatical based on certain logical conditions. For example, in the situation described by (72), (72) "If John has not gone to Beijing, Mag; went to talk with him yesterday.“ it can well be the case that the speaker does not brow whether or not John has really gone to Beijing, nor does the speaker brow whether or not Mary went to talk to John yesterday. Therefore, the speaker is understood to mean that, if John has not gone to Beijing, Mary would have gone to talk with him yesterday. Sentence (73) below describes another situation, (73) "If John has not gone to Beijing, 1 will go to talk with him tomorrow." where it is understood that if John has not gone to Beijing, the speaker will go to talk with him tomorrow. In any case, therefore, whenever the first person pronoun as the " agent of the context“ and a past time indexical are present in the consequent, the conditional sentence will be a counterfactual. Here are some more examples in Chinese, 37 (74)a. Ruguo wo yao rende waiguo zi ab, wo jiu bu lai qiou m’ le. if I if recognize foreign word FP, I then not come request you FP ‘If I recognize [the] foreign words, I will not come to request you for help.’ or ’If I had been able to recognize [the] foreign words, I would not have come to request you for help.’ b. Ruguo wo yao rende waiguo zi ah, wo n_ag'_a_n_ jiu bu laiqiou ni le. if I if recognize foreign word FP, I that-day then not come request you FP If I had been able to recogrize [the] foreign words, I would not have .come to request you for help.’ (75) Ruguo zuotian bu xiayu, m Lag jiu qu canguan gongye zhanlanhui le if yesterday not rain, is Ligigrg then go see industrial exhibit FP If it had not rained yesterday, we would have definitely gone to see the industrial exhibit.‘ (76) W0 yaoshi zhende yao taozou, [wo] sfl jiu zou diao le.(Wu,ibid.:86) I if really want escape, [I] e_2__a_r'_ly then go away FP ’If I really had wanted to escape, I would have done so long before." (74a) as it stands will not give a hearer enough clues to infer that the speaker does not brow the foreign language, though, intuitively it is more so than not. Nevertheless, while the possibility still exists for (74a) to be interpreted as hypothetical or indicative, (7 4b) turns to be counterfactual only, when the demonstrative indexical expression is understood to refer to a past time. The counterfactuality of sentence (75) has been claimed by Hu (ibid.:368) and the understanding of it lies in the indexical word zao'before‘. When this indexical is used with the first person pronoun wo‘I‘, the hearers will understand that it actually refers to a temporal point which is earlier than the moment the speaker is making the utterance. And the same explains the counterfactuality of (76). 38 f 4.3.2. Indexicals Related to Definite Demonstrative Reference The similar analytical strategy is applicable to sentence (56) above repeated here as (77), (7 7 )a. Ruguo ta lai le, jiu bu hui chu shi le. if she come FP, then not will happen thing FP b. Ruguo ta lai le, jiu bu hui chu _z_l_r_§ shi le. if she come FP, then not will happen _tl;i_s_ thing FP In my previous discussion of the DPD in Chapter 3, I have presented Russell‘s definite description theory, who argued that a definite description in a proposition, for example, ‘the F is G‘, asserts that there is one and only one (at least and at most one) F. And, Strawson, in response to Russell, argued that when a speaker utters a sentence with such an expression as ‘the such-and-such‘, the speaker presupposes that there is "one particular individual of the species ‘such-and-such‘ (Strawson, 1990:325). In (7 7), (a) and (b) are distinguishable from each other by the indexical demonstrative, zhe 'this’. In (1)), by demonstrating ’this matter’, the existence or the occurring of ’this matter‘ is implied. With the speaker‘s asserting this sentence, it will be presupposed that the matter has already happened, namely, (7 8), (78) It is the case that this matter has happened. Accordingly, it follows that the consequent proposition is false, because it is the negation of a true proposition. Applying the modus tollens law, we get that the antecedent proposition is false, namely, it is false that she came. The following is another example, (79)a. Ruguo ni guozben ai ta, jiu bu yinggai bi ta. if you really love her, then not would force her ‘If you really love her, then you would not force her.‘ or ‘If you really loved her, then you would not have forced her like this.‘ 39 b. Ruguoni guozhen ai ta, jiu bu yinggai zbeya_ng bi ta. if you really love her, then not would like-this force her 'If you really loved her, then you would not have forced her in this way.‘ Clearly, the reason why (79b) is interpreted counterfactually is because of the demonstrative phrase zheyang'like this‘ in the consequent which implies that the event of ’forcing her‘ has already happened in a way ‘like this’. And in similar argument, the antecedent proposition, say, ‘you really love her‘, is false. At this point, it is necessary to address the issue of the speaker's belief and the truth value of such a belief. The above examples, say, (77b) and (79b) are understood as counterfactuals by the he arers based on the speaker‘s belief as triggered by the indexical expressions thus used. It is likely that the speaker may be wrong or the speaker‘s belief may not be correct as a matter of fact. But in any case, when the speaker utters such a sentence as (77b) or (79b), the speaker assumes only what she believes is the case. This point can be further illustrated in the following example, (80) Ruguo Zhangsan zbende shengqi le, ta jiu bu hui zai zhemg if Zhangsan really angry FP, he then not will still like-this shuo rdaqu le talk on FP According to the IC theory discussed in this section, namely, 4.3.2., (80) should be a counterfactual conditional because the consequent contains a demonstrative indexical phrase zheyang'like this’. Nevertheless, it might not necessarily follow that it is not the case that Zhangsan is angry which is presupposed in a counterfactual conditional sentence. For example, Zhangsan may be pretending that he is not angry, but in actuality it is the case that Zhangsan is angry. To put it another way, the truth values we are talking about here are those that are assumed by the speaker, who also assumes that the hearers would come to the 40 intended interpretation of the sentence . This is in contrast to what their actual truth values are, where, for example, the antecedent proposition is false if and only if it is not the case that Zhangsan is really angry, regardless of the claim that the demonstrative indexical zheyang’like-this‘ renders the sentence counterfactual. Nevertheless, here we are looking at the issue of speakers‘ belief and the truth value of such a belief. Therefore, in making a statement like (80), the speaker is expressing what she believes to be true, namely, the speaker believes, in uttering the sentence, that Zhangsan is not really angry, for he will not continue talking like this otherwise. This is difi'erent from the truth value of the proposition, which is determined to be true if and only if it is the case that ’Zhangsarr is angry‘ and false if and only if it is not the case that ’Zhangsan is angry‘. In light of this explanation, (80) is not a counterexample to my generalization about the demonstrative indexicals in contributing counter-f actuality in Chinese, but rather since it is the speaker‘s belief that Zhangsan was not really angry so that he could talk on like this, sentence (80) is still a counterfactual sentence. This actually happens in English too, for example, when a speaker assumes that John is not angry, quite without regards to the possibility that John may be really angry, such an idea is usually expressed as in (81) . (81) “If John were really angry, he would not keep on talking li_l_r_e_tl_ri_s_." and accordingly (82) will be felt odd9, (82) ?’If John is really angry, he would not keep on talking li_lg_e__t._his;.’ which will express a different interpretation of the speaker, namely, the speaker does not brow for sure if John was really angry. More examples can be found in the literature, by the authors mentioned elsewhere in this thesis, which are claimed to be counterfactual sentences, though none of these authors have explained their counterfactuality based on any linguistic ground. (Underlines in all these examples are added for contrast.) 41 (83) Zaoxieding 1e qin, yexu bu hui chu gh_e chazi.(Hu,1979:390) earlier decide ASP marriage, perhaps not will make this; mistake "If the marriage had been decided earlier, perhaps this mistake would not have been made." (84) Yaoshi wo zuo le nimen enke ba, Lz_h_e] ye bu if I become ASP you(pl.) honored-guest FP, [_tlgs] also not guan bieren de shi. (W u,1989:86) concern other DE matter “Even if I had been honored as your guest, it was none of their business‘(ibid.:86) (85) Ruguo women konishitaiyin you ge gongchang, wojiu keyl' if we Konigstein have CL factory, I then may zai _z_b_e_ shangban. (AG:562) in hgye; work "Wenn wir eine Fabric in Konigstein batten, dann kOnne ich hier arbeiten.”(ibid.:557 ) “If we had a factory in Konistein, then I would be able to work here." (86) W0 yao bu xiangxin ni, wo dao bu hui ba gag bua gaosu ni le. (Lu, ibid.). I if not believe you, I but not will BA _tl_rl_'_§ word tell you FP "If I had not trusted you, I would not have told this to you." Applying the same argument, the above three sentences will be interpreted as counterfactual conditionals only, namely, based on the indexical expressions of zbe’this’ in (83), zhe‘this’ in (84), zhe’here’ in (85), and zhe’this’ in (86) respectively. 4.3.3 Indexicals and the Conditional Antecedents While the preceding discussion is on the analysis of counterfactuality of the consequent proposition as contributed by indexical expressions, it is also 42 possible to detect cases of such contribution by indexicals to the counterfactuality of the antecedent proposition occurs. Look at examples of ( 57 ) above repeated here as (87) and another example of (88) (87 )a. Ruguo wo zuo zai nar, ta jiu hui baoshou duo le. if . I sit in there, she then will comfortable more FP b. Ruguo gig-(Lag wo zuo zai nar, ta jiu hui haoshou duo le. if Qr_a_t-_(_l_ay I sit in there, she then will comfortable more FP (88) Ruguo wo my you Is haizi, wo jiu hui meiyou zhemo duo if I heme have ASP children, I then will not-have this much shijian lai xuexi. (Wu, 1989:96) time for study “If I had had children before, I then would not have this much time for study.“ Quite naturally, when people make statements like the above they are understood to be making assumptions of what might be, or might have been, the case if certain conditions hold. In this circumstance and in normal discourse contexts, it follows that the speakers brow whether certain conditions really hold at certain time in the past. In (87a), for example, the causal relationship holds either at a time to come or at a time in the past, say, the speaker imagines what will occur and assumes what could occur by her presence at a future or a past time respectively. In contrast, the addition of the past time reference in (87b), in lieu of the original first person speaker, or the “agent of the context“, eliminates the future interpretation possibility, thus renders the sentence a counterfactual only. And the same can be said about (88). The antecedent is interpreted as a counterfactual condition because it is quite natural to understand that the speaker does not have children by the fact of the speaker’s speculation on what could have occurred if she had had children before. It follows, therefore, that the consequent 43 proposition is assumed by the speaker to be false, namely, the speaker has assumed that she does not have this much time for study. 4.3.4 Summary In this Chapter of the thesis, I have claimed that indexicality contributes to counterfactuality in modern Chinese. First, I have presented conditional sentences which involve using certain indexical expressions and which are intuitively interpreted as counterfactual sentences. Furthermore, I have analyzed the said counterfactuality by arriving at a counterfactual consequent proposition, which, through modus tollens, eventually leads to the counterfactuality of the antecedent proposition (4.3.1 and 4.3.2). Finally, the indexicals‘ contribution to the counterfactuality of antecedent propositions was discussed (4.3.3). 5: SUMMARY AND CONCLUSIONS The enterprise that the present thesis has undertaken is to show the linguistic constructability of counterfactual conditional sentences. In doing so, I have concentrated efforts on two sets of data which have been recorded as conveying perspicuous counterfactuality. The first set of data is composed of those conditional sentences which fall into the linguistic structure of [ bu-shi [p ...] ] in which the lexical item shi is claimed to presuppose the truth of the proposition which follows it. I have used the term ‘Definite Propositional Description (DPD)‘ to refer to such a linguistic structure which functions in this way (Chapter 3). The second set of data is composed of conditional sentences which involve using certain indexical expressions which, in turn, contribute counterfactuality to the conditional sentences in which they occur. The term ‘Indexical Counterfactuality (IC)’ is used to refer to this linguistic occurrence (Chapter 4). It is shown throughout the thesis that the Chinese language does have the linguistic apparatus for expressing counterfactual constructions and for the speakers of this language it should be as natural to engage in counterfactual thinking as for those of any other language such as English. Consequently, it is necessary that this linguistic observation be properly incorporated into the general theory of Chinese linguistics. Moreover, though a side product, Bloom’s (1981; 1984) claim, that the Chinese speakers have difficulty in understanding counterfactual constructions, is seen to be too radical and arguable to be accepted for the inquiry either into the Chinese mind or the language. As I have stated in several places in the thesis, the Chinese language bears an ambiguity between hypothetical and counterfactual conditionals as 45 notoriously as that between indicative and counterfactual and hypothetical conditionals. It appears that the present study may contribute to a better understanding in the distinction between counterfactual and hypothetical conditionals in Chinese. It is commonly brown that conditional sentences are counterfactual if the propositions expressed in the antecedents are assumed to be false. In contrast, a hypothetical conditional is one in which the propositions expressed in the antecedent are not real but can be true at some future time. According to Karttunen 8: Peters (1979), a hypothetical conditional is such that its antecedent proposition cannot simply be denied; this needs to be decided as certain new information becomes available. This is where I want to state my claim: in cases where a conditional sentence can be interpreted either as a counterfactual or a hypothetical, an indexical expression of a past time reference in the antecedent may serve to disambiguate the situation. For example, (89) Ruguo wo zuo-tian kan-jian guo-wang, wo jiu hui jiu-gong. if I yesterday see king, I then will bow ’Ifl had seen the king yesterday, I would have bowed.“ (90) Ruguo wo kan-jian guo-wang, wo jiu hui jiu-gong. if I see king, 1 then will how ‘If I saw the king, I would bow.‘ The past time reference in (89), in combination with the first person sentence subject, namely, an agentive thematic role of the sentence, makes the conditional sentence counterfactual. This is explicable by the same reason based on which sentences of (87 ) and (88) above are understood as counterfactuals. Quite obviously, the speaker of (89) brows whether she saw the king yesterday or not, and, consequently, such a circumstance of the speaker‘s seeing the king yesterday is interpreted only as her assumption. Therefore, such a condition in (89) will be 46 processed by the hearer as false, which, in turn, defines (89) as counterfactual conditional sentence. In contrast, the absence of the past time operator in (90) leaves an ambiguity at stake between the hypothetical and counterfactual readings. It is hypothetical because the conditional proposition is not real at the time of utterance, namely, the speaker does not see the king at the time of uttering (90), but it is possible that she sees the bug two seconds later.10 When further information is available for the bearer to believe that the speaker will not be able tosee thekingon any account, forinstance,thekingisdead already atthetime of the utterance so that the speaker could never see the (living) king or bow to him, (90) would then be interpreted as a counterfactual. Therefore, as I said earlier, a past time reference used in combination with a first person indexical subject in the antecedent will function to disambiguate between hypothetical and counterfactual readings of a conditional sentence. In conclusion, the present study is an attempt aimed at the long- standing linguistic issue of Chinese counterfactuality, to which there has been no other study so far of similar approach. Nevertheless, it is obvious that the present study has not answered all questions related to this t0pic. For example, in regard to the claim made about IC in Chinese, since indexical expressions are so “context dependent“ and the picture of context is much more complicated than what Sag (1981) or Kaplan (1977 ) have studied, what other factor(s) could affect the validity of 1C analysis? How about the picture of DPD? What other contextual factor(s) should be considered for the claim made about the conditionals of the [ bu-shi [3...1] linguistic structure? And, if there are other types of conditionals to which the IC and DPD do not apply (which is obviously the case), then what to do with these sentences? It is doubtless that these and other questions make one hope for more from the present thesis than I have attempted here. Hopefully, however, they will inspire future research to work out answers. ’ NOTES 1hr some other literature, the terms protasis and apodosis are used to refer to antecedent and consequent respectively, see, for example, Traugott,1985, inter alia. Throughout this thesis, the first set of these terms are used for the fact that they are more commonly used in the literature. 2Clearly, this term refers both hypothetical and counterfactual conditions. 3Hu used the term "hypothetical" to diver both hypothetical and indicative conditionals. 4The number in the parenthesis is the original enumeration by Eifring of his examples. 5The three propositions are expressed in logical notations as follows, a. (3 x) Fx 8: b. (x) (y) [(Fx 8: Fy) -> x=y] 8: c. (x) (Fx --> Gx) where F abbreviates "is a present king of France“ and G abbreviates “is bald". 61 am thankful to Dr. Herbert Hendry of the Department of Philosophy for the development of this idea. 7Underlines in these three sentences are added. 8These maxims were first discussed by H. P. Grice in the William James Lectures delivered at Harvard University in 1967. 9Although ill modern English the subjunctive mood in the ancetedent verbs may have disappeared. 101 am thankful to Dr. James McCawley for pointing this out to me. 47 ASP: BA: CON: COP: DE: DEM: EMP: FP: NOM: PAS: PM: ZHI: CAUS: APPENDIX 1: GLOSS OF THE ABBREVIATIONS Auxiliary Phrase. deh us is used after a condition clause, bringing in a closer link with the following consequent clause. Aspect Chinese word BA, used in ’ba+NP+VP‘ structure, indicating the objecthood of the NP of the VP Causativizer Classifier Connective Copula Chinese word DE, used in these cases after nominal modifiers Demonstrative Emphatic Final Particle Nominalizer Passivizer Progressive Marker In classical Chinese used after nominal modifiers APPENDIX 11. GLOSS OF SOME TERMS OF CHINESE bu not‘ bu-slri ’not-be’ Guanxicr' ‘Relation words‘ mei ’not’ mei- you ’not-have’ mingtian ‘tomorrow‘ na-tian ‘that-day‘ ruguo ‘if ruguo 'bu-shi ‘if not-be‘ shi ’be‘ we ‘I‘ zhe ‘this' zbeyang 'like this’ zuotian ’yesterday‘ 49 BIBLIOGRAPHY BIBLIOGRAPHY Adams, W. 1970. ’Subjunctive and indicative conditionals.‘ Foundations of Language. vol.6, 89-94 ‘ . 197 5. The Logic of Conditionals: An Application of Probability to Deductive Log'c. Dordrecht, Holland: D Reidel Publishing Company. AG (Short for Alles Gute). 1980. Beijing: China Guoji Guangbo Chubanshe (China International Radio Press). Anderson, A. R. 1951. ’A note on subjunctive and counterfactual conditionals.’ Analysis. vol.12. Appiah. Anthony. 1985. Assertions and Conditionals. Cambridge: Cambridge University Press. Au, Terry Kit-Fong. 1983. “Chinese and English counterfactuals: the Sapir-Whorf hypothesis revised". Cognition. Vol. 15:155-87. -------. 1984. "Counterfactuals: In reply to Alfred Bloom”. Cogrition. Vol. 17:289- 302. Bar-Hillel, Y. 1954. ‘Indexical expressions.’ Mind. vol.63, 359-379. Bloom, Alfred. 1981. The Linguistic Shaping of Thought: A Study in the Impact of Language on Thinkingin China and the West. Hillsdale, NJ: Lawrence Erlbaum Associates. -------. 1984. ”Caution--the words you use may afiect what you say: a response to Au“. Cognition. vol. 17: 275-87. Caton, C. E. 1959. ”Strawson on referring”. Mind. Vol. 68, pp.539-44. Chao, Yuen Ban. 1959. ”How Chinese logic operates.“ In Answar S. Di] (ed.). Aspect of Chinese Sociolinguistics: Essays by Yuen Ren Chao. Stanford: Stanford University Press. Chierchia, Gennaro 8: McConnell-Ginet, Sally. 1990. Meaning and Grammar: An Introduction to Semantics. Cambridge, Massachusetts: The MIT Press. 51 Eifring, Halvor.1988. “The Chinese counterfactuals". Journal of Chinese Linguistics. Vol. 16. 193-217. Gamut, L.T.F.. 1991. Logic, Language, and Meaning: Volume 1, Introductidn to Logic. Chicago: The University of Chicago Press. Gao, Mingkai.1957. Hanyu Yufa Lun(On the Chinese GrammarXRevised ed). Beijing: Kexue Chubanshe. Gorelov, Vladimir I. 1973. Grammatika Kitaiskovo JazyktKGrammar of the Chinese Language). Moskva: Izdatelstvo Prosveshtshenie. Grice, H. P. 1975. ’Logic and conversation.’ In Peter Cole and Jerry L. Morgan(eds.). Syntax and Semantics, Vol. 3: Speech Acts. New York: Academic Press. -----. 1978. ‘Further notes on logic and conversation.‘ In Peter Cole(ed.). Syntax and Semantics, Vol. 9: Pragmatics. New York: Academic Press. HanYing CiDian (Chinese-English Dictionary). 1981. Beijing: Shangwu Yinshuguan (Commercial Press). I-Iu, Yushu. 1979 . XianDai HanYu(2nd ed.)(Modern Chinese). Shanghai: Shanghai J aoyu Chubanshe (Shanghai Education Press). Huang, Borong 8: Liao, Xudong. 1983. Xiandai HanyutModern Chinese). Lanzhou: Gansu Renmin Chubanshe. Jackson, Frank. 1990. “Classifying conditionals”. Analysis. Vol. 2:134-147. Kaplan, David. 1977. ‘Demonstratives‘. In Almog, J ., Perry, J. 8: Wettsteirr, H. (eds). 1989. Themes fi'om Kaplan. New York: Oxford University Press. Karttunen, Lauri 8: Stanley Peters. 1979. "Conventional Implicature.” In C.-K. Oh 8: DA. Dineen (eds). Syntactic and Semantics, Vol. 11: Presupposition. New York: Academic Press. Kiparsky, Paul 8: Kiparsky, Carol. 1971. ’Fact’. In Manfred Bierwisch 8: Karl Erich Heidolph (eds). Progress in Linguistics: A Collection of Papers. The Hague: Mouton. Leech, Geofl‘rey N. 1987 . Meaning and the English Verb(2nd ed.). London: Longman. Lewis, David.l973. Counter-{actuals Oxford: Basil Blackwell. Li, Bing. 1993a. ‘Demonstratives and propositional truth value: Data analysis of Chinese counterfactuals! Presentation at the Michigan State University Linguistics Colloquium. December 3. 52 ------. 1993b. ‘Re-Examination of counterfactuals in Chinese.‘ Paper presented at NACCL5. University of Delaware, Newark, May 14-16. Li, C.N. 1977. ’A mechanism for the development of copula morphemes.‘ In C. N. Li(ed.). Mechanism of Syntactic Change. Austin: University of Texas Press. Li, C. N. 8: Thompson, Sandra A. 1981. Mandarin Chinese: A Functional Reference Grammar. Berkeley: University of California Press. Liu, Lisa Garbern. 1985. "Reasoning counterfactually in Chinese: Are there any obstacles?” Cognition. Vol.21: 239-70. Lu, Shuxiang. 1945. ZhongGuo WenFa Yaolue (Chinese Grammar Outline). Chongqing: Shangwu Yinshuguan (Commercial Press). Marsh, Robert Charles. 1956. Logic and Knowledge: Essays 1901-1950. London: Unwin Hyman. Russell, Bertrand. 1905. “On denoting." Mind. pp.479-493. Strawson, P. F. 1990. "On referring“. First puglished in Mind, 1950, Vol. 59, pp.320-44. Reprinted in Robert R. Ammerman. (ed.). Classics of Analytic Philos0ph y. Indianapolis: Hackett. (Referrences in the text are made to this edition.) Thrane, Torben. 1980. Referential-Semantic Analysis: Aspects of a Theory of Linguistic Reference. Cambridge: Cambridge University Press. Traugott, E. C.. 1985. ’Conditional Markers.’ In John Haim@(ed.). Typologr'cal Studies in Language. Vol. 6: Iconicityin Syntax. Amsterdam: John Benrijamins Publishing House. Traugott, E. C., ter Meulen, A., Reilly, J .S., and Ferguson, CA. (eds). 1986. On Conditionals. Cambridge: Cambridge University Press. Wang, Li. 1957. Zhongguo Xiandai Yufa [Modern Chinese Grammar]. Shanghai: Zhongguo Shriju. . Wu, Kuang-ming. 1987 . ’Counterfactuals, universals, and Chinese thinking.’ Philosophy East and West. Vol. 37 :85-94. - Wu, Zhaoyi. 1989. Exploring Counter-{actuals in English and Chinese. PhD. Dissertation. University of Massachusetts. Xing, Fuyi. 1985. quu yu Guanxi Ciyu(Composite Sentences and Connectives). Harbin: Heilongjiang Renmin Chubanshe. HICHrch srnrs UNIV. LIBRQRIES i“"Will“WIHIWNWWNIWI)IWIIIHHWUI 31293014115988 WW-.. ”-mv‘... aim-...”. u __.L.. t..- w—w‘” ‘ i