Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On the structure of negation in Japanese and its related problems Goh, Ethel See Kean 1975

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1975_A8 G64.pdf [ 7.06MB ]
Metadata
JSON: 831-1.0093443.json
JSON-LD: 831-1.0093443-ld.json
RDF/XML (Pretty): 831-1.0093443-rdf.xml
RDF/JSON: 831-1.0093443-rdf.json
Turtle: 831-1.0093443-turtle.txt
N-Triples: 831-1.0093443-rdf-ntriples.txt
Original Record: 831-1.0093443-source.json
Full Text
831-1.0093443-fulltext.txt
Citation
831-1.0093443.ris

Full Text

ON THE STRUCTURE OF NEGATION IN JAPANESE AND ITS RELATED PROBLEMS by ETHEL SEE KEAN GOH B.A., Tokyo University of Foreign Studies, 1973 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n the Department of ASIAN STUDIES We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA A p r i l , 1975 In presenting th i s thesis in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary shal l make it f ree ly ava i l ab le for reference and study. I fur ther agree that permission for extensive copying of th i s thes is for scho lar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i ca t ion of th is thes is for f inanc ia l gain sha l l not be allowed without my writ ten permission. Ethel See Kean Goh Department of Asian Studies The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada Date A p r i l , 1975 i ABSTRACT Negation, a rather complex area of grammar, has so f a r been given l i t t l e attention i n the study of Japanese L i n g u i s t i c s . This thesis attempts to examine the structure of negation i n Japanese and i t s related problems, and to demonstrate what happens to sentences under negation employing the conceptual framework of a recent theory of generative grammar. I t i s hoped that the findings i n t h i s thesis w i l l provide more insights into the problems of negation, and w i l l help c l a r i f y some of the semantic and syntactic problems associated with various aspects of negation i n Japanese. Recent studies i n the area of negation i n English have provided various i n s i g h t f u l explications and analyses, and the findings from these studies w i l l serve as a basis for t h i s research. The analysis i n t h i s thesis i s based on the assumption that every well-formed grammatical sentence consists of a deep structure and a surface structure which are related by a system of transformations? and that the deep structure i s the relevant l e v e l f or determining the meaning of the sentence. This thesis i s organised i n the following way. Chapter One i s concerned with where the constituent NEG should be introduced i n the deep structure, and whether a l l types of negative expressions can be ascribed to a single underlying i i form. Related to t h i s , various negative expressions i n Japanese w i l l "be examined i n order to determine what t h e i r underlying structures are. The problems of meaning i n negation w i l l also be investigated, especially where they concern the "scope of the negative", that i s , what i s exactly being negated i n the sentence. Chapter Two deals mainly with a rule that has been proposed for English which i s c a l l e d negative transportation. This rule has been assumed to ex i s t i n many other natural languages. The chapter examines t h i s negative transportation rule i n an attempt to determine whether or not i t e x i s t s i n Japanese. The arguments presented i n t h i s chapter w i l l follow s i m i l a r l i n e s of arguments that have been presented for English. Chapter Three w i l l look into the problems and peculiar-i t i e s involved with the Japanese negative questions and the yes-no responses that they e l i c i t . The chapter w i l l attempt to give l o g i c a l explanations for the ambiguity of negative sentences i n Japanese and w i l l attempt to explain both the semantic and syntactic c h a r a c t e r i s t i c s of such questions. The r e l a t i o n s h i p between the negative questions and the responses that they e l i c i t w i l l be explained by incorporating the notion of presupposition. Chapter Four examines the semantic and syntactic structure of the p a r t i c l e s mo, wa and ga. The o r i e n t a t i o n of t h i s chapter w i l l be one of attempting to examine the facts of the Japanese language concerning the p a r t i c l e mo. B a s i c a l l y , the discussion i n t h i s chapter w i l l "be descriptive, attempting to characterize the general nature of the p a r t i c l e mo. The chapter w i l l present a general approach for the inte r p r e t a t i o n of the p a r t i c l e mo, and at the same time, w i l l also present a variety of syntactic constructions to i l l u s t r a t e the approach. The presuppositional properties associated with mo, which are relevant f o r the correct surface semantic in t e r p r e t a t i o n w i l l also be examined. The chapter w i l l also investigate the re l a t i o n s h i p between the negative and mo, especially where i t concerns the p o s i t i v e -negative r e l a t i o n s h i p between the assertion and the expect-ation underlying mo. Having provided a genexal schema for the i n t e r p r e t a t i o n of mo, the l a t t e r part of the chapter attempts to examine whether the same l i n e of approach i s also applicable to other p a r t i c l e s such as wa and ga. Chapter Five presents a b r i e f summary of that which has been discussed i n the thesis. F i n a l l y , I would l i k e to point out that l i t e r a l trans-l a t i o n s are necessary i n order to preserve the phrase by phrase meaning of the Japanese example sentences. As a r e s u l t , some translations rendered i n t h i s thesis may not always appear to be perfect idiomatic English expressions. TABLE OF CONTENTS i v Chapter One. SENTENCE NEGATION AND VERB NEGATION 1 1.1. Introduction 1 1 . 2 . Verb Phrase Negation 1 1.3« Sentence Negation 9 1.3.1. Arguments f o r Sentence Negation ... 9 1 . 3 . 2 . On Some Structures f o r Negative Constructions 19 1 .3 .2.1. Locative and Instru-mental Adv-er b i a l s ...» 19 1 . 3 . 2 . 2 . Reason Adv-er b i a l s ... JZ 1 . 3 . 2 . 3 . Time Adver-b i a l s 41 1.4. Summary 5° Footnotes 5^ Chapter Two. NEGATIVE TRANSPORTATION 55 2.1. Introduction 55 2 . 2 . Negative Transportation i n English 55 2 . 3 . Arguments f o r Negative Trans-portation i n Japanese 65 2.3.1. Simplex Sentence Con-d i t i o n f or Negative P o l a r i t y Adverbials and P a r t i c l e s 65 2 . 3 . 2 . "Confirmatory" Quest-ion Argument 72 2 . 3 . 3 . Counterexamples to above Two Arguments . 78 2.4. Conclusion 97 Footnotes 99 V Chapter Three. JAPANESE NEGATIVE QUESTIONS AND YES-NO RESPONSE 101 3.1. Introduction 101 3 . 2 . Negative Questions and Their Responses 102 3.2.1. Syntactic Ambiguity of Negative Questions 102 3 . 2 . 2 . Presuppositional Information and Conversational Implications 106 3 . 2 . 3 . Underlying Structu-res of the Negative Questions 109 3.2.4. The Logic Underlyi-i g the Relationship Between the Negati-ve Questions and t h e i r Responses ... 117 3 . 2 . 5 . Intonation 119 3 . 3 . Conclusion 124 Footnotes 126 Chapter Four. ON THE SEMANTIC AND SYNTACTIC STRUCTURE OF THE PARTICLES MO, WA AND GA 12? 4 . 1 . Introduction 127 4 . 2 . Interpretation of the P a r t i c l e mo 128 4 . 2 . 1 . Presuppositional Properties of mo .. 128 4 . 2 . 2 . The Occurrence of mo with Quantifiers 138 4.2.3- Concessive Sentenc-^ es and mo 15^ 4.3. On the Interpretation of wa and ga 156 4.4. Conclusion 166 Footnotes 169 Chapter Five CONCLUSION 170 BIBLIOGRAPHY 17 ^ 1 CHAPTER ONE SENTENCE NEGATION AND VERB NEGATION 1.1. INTRODUCTION This chapter i s concerned with where the constituent NEG (negative) should be introduced i n the deep structure, and whether a l l types of negative expressions can be ascribed to a single underlying form. With respect to t h i s , I w i l l examine various negative expressions i n Japanese, together with t h e i r related problems, and w i l l t r y and determine what the underlying structures f o r these negative constructions are. The "scope of the negative", that i s , what exactly i s being negated i n the sentence, w i l l also be examined because what i s ac t u a l l y negated i s c r u c i a l f o r determining the structures of the negative constructions. Related to t h i s , I w i l l also investigate problems of meaning i n negation, p a r t i c u l a r l y where they concern the scope of negation. 1.2. VERB PHRASE NEGATION Chomsky, i n his Syntactic Structures (1957), considers 2 the underlying structure f o r the negative sentence simply as a po s i t i v e structure, and the negative i s introduced by an optional transformation as a purely surface matter. The following i s the formula given by Chomsky f o r deriving negative sentences from affirmative sentences.* Tnot " ° P t i o n a 1 ' Structural analysis: ( i ) NP - C - V ... ( i i ) N P - C + M . . . ( i i i ) NP - C + have - ... (iv) NP - C + be - ... Str u c t u r a l changei X t - X 2 - X 3 > X x - X 2 + n't - X 3 By t h i s transformation, Tno^. operates on strings that are analyzed into three segments i n one of the above ways of ( i ) - ( i v ) . Given a s t r i n g that i s analyzed into three segments, the transformation Tno^. introduces the negative not or n't a f t e r the second segment of the s t r i n g . For example, apply t h i s formula to the following terminal strings of (a) - (d), T ^ w i l l derive the sentences on the r i g h t . (a) They - f& - eat > They - ^ + do + n't - eat ("They don't eat") 3 (b) They - ft + can - eat ^ They - ft + can + n •t - eat ( They can't'eat") (c) They - ft + have - en + eat •^ •They - + have + n't - en + eat (d) They - fi + be - ing + eat , ("They haven•t eaten") >They - # + be + n •t - ing + eat ("They aren't eating") This shows that the T n Q H. transformation simply introduces the negative not or, n't into the verb phrase. Let us examine a few examples and see how negative sentences are derived i n Japanese. The following sentences can be considered to be related by way of negative-positive p o l a r i t y . (1) a. Taroo ga k i t a . 'Taroo came.' >b. Taroo ga konakatta. •Taroo didn't come.' (2) a. Watakusi wa sake o nomu. •I drink sake (Japanese wine).' b. Watakusi wa sake o nomanai. •I don't drink sake (Japanese wine).' (3) a. Sakura no nana wa akai. •Cherry "blossoms are red. 1 b . o Sakura no hana wa akaku n a i . •Cherry blossoms are not red.' (4) a. Kono heya wa sizuka da. •This room i s quiet.* b. Kono heya wa sizuka de wa n a i . 'This room i s not quiet.* I f we subscribe to the above proposal by Chomsky, then the actual underlying structures for the negative sentences of ( l b ) , ( 2 b ) , (3b) and (4-b) would be t h e i r p ositive counter-parts, that i s , the (a) sentences, and the negative w i l l be introduced into the verb phrase by an optional transformation. Let us now look at the tree structures f o r two of the above sentences. Considering sentences (1) and ( 3 ) , t h e i r 2 (a) sentences w i l l have the following underlying structuresi 5 (6) S sakura no hana akai With the exception of l e x i c a l items, (2a) and (4a) w i l l also have s i m i l a r underlying structures as those of (5) and ( 6 ) . I f we assume the above underlying structures for the affirmative (a) sentences, then we w i l l have to assume the following underlying structures f or t h e i r negative counterparts, that i s (lb) and (3b). sakura no hana akai + NEG 6 Notice that the structures of (7) and (8) show that the negative i s introduced by transformation as a co n s t i -tuent of the VP, meaning that the scope of the negative commands only the VP. This may appear to mean that the negative i s relevant only with verb phrases. In f a c t , t h i s was b a s i c a l l y the way the rule of negation was treated by Inoue (1964). Her treatment was simply the attachment of the negative morpheme to the verb or adjective by a rule of optional transformation of the following kind:^ Negation; X - NP + wa ga ga - Y AB Vm + T - Z 1 2 3 5 wa 1 + ga - 2 - 3 + ana + 4 - 5 wa Examples 8 (a) To. ga ak ta. NP+ga-Vm+ T > To ga ak ana katta. NP-ga-Vm+ana+ T > 'The door did not open.' 'The door opened.' 7 (b) Kono mondai wa muzukasi i Kono mondai wa NP +wa- A +T NP +wa' muzukasi kuna i . A +ana +T 'This question i s d i f f i c u l t . ' >'This question i s not d i f f i c u l t . • However, t h i s transformation f a i l s to account f o r things l i k e the scope of negation, and other changes that occur when sentences are negated. S i m i l a r l y , Muraki i n h i s paper Negation i n English  and Japanese (1965)» treats the problem i n such a way that a l l negations i n Japanese are derived by the addition of the negative verbal na to an affirmative verbal. According to him, sentences l i k e i (9) Kare wa kessite hon o yomanai. *As f o r him, he never reads a book.• (10) Kimi to nanka ikanai. 'I w i l l not go with (a person l i k e ) you.* w i l l have the following underlying structures! 8 (11) hon yomu (12) Kimi to nanka iku where the negative i s generated as part of the VP, showing that negation i s relevant only with the VP. However, i f we subscribe to the hypothesis that transformational rules are meaning-preserving, then we 9 have to introduce the negative i n the deep structure, and not at the surface l e v e l . However, exactly where should the constituent NEG be introduced i n the deep structure? In the above examples, we have seen that the negative i s introduced as a constituent of the verb phrase and appears to be relevant only with the verb phrases. In the next section, I w i l l examine whether a l l types of negative sentences can be ascribed to t h i s single underlying form, or whether there i s a p o s s i b i l i t y of generating the negative i n positions other than as the constituent of the VP i n the underlying structure. 1 .3 . SENTENCE NEGATION 1 . 3 . 1 . ARGUMENTS FOR SENTENCE NEGATION To my knowledge of Japanese generative grammar, Soga (1966)^first introduced the negative i n the deep structure. However, he did not present a convincing argument f o r i t . He simply treated the negative as a sub-class of adjective which must co-occur with an embedded sentence. Apparently, Soga's treatment i s i n l i n e with sentence negation rather than verb phrase negation. In the following, I w i l l present a convincing argument f o r sentence negation. Now, l e t us consider the following sentences i n order 10 to see i f the underlying structures i n the previous pages can he maintained f o r other negative sentences. (13) Yamada-san wa kanemoti de wa nai keredomo, minna wa soo omotte i r u . 'Mr. Yamada i s not r i c h , but everyone thinks (he i s ) . ' (14) Taroo wa kessite gogaku no tensai de wa nai no : n i , zibun de wa soo omotte i r u . 'Taroo i s c e r t a i n l y not a genius for languages, but he thinks (he i s ) . * (15) Tikyuu wa kessite hirataku wa nai no n i , mukasi no h i t o - b i t o wa soo s i n z i t e i t a . 'The world i s c e r t a i n l y not round, but the people of ancient times believed (that i t was).' (16) Hanako wa ano g a i z i n to kekkon s i n a i keredomo, minna wa soo i t t e i r u , 'Hanako i s not marrying that foreigner, but everyone says (she i s ) . ' Notice that i n the above examples, there i s a negative sentence on the l e f t side and a pronominal soo on the r i g h t side, which refers back not to the negative sentence on the l e f t , but to the positive sentence corresponding to the negative sentence. This means that the soo i n (13) can be understood as Yamada-san ga kanemoti desu, which i s actually 11 the po s i t i v e counterpart of the preceding negative sentence. S i m i l a r l y , the soo i n (14) refers to Taroo ga gogaku no  tensai da, i n (15) tikyuu ga h i r a t a i desu, and i n (16) Hanako ga ano g a i z i n to kekkon suru, which are a l l positive counterparts of the preceding negative sentences. Notice that i n the above examples of (13)-(16), soo refers back to a sentential element. This would appear to mean that soo has to be a sentential-pronominal and not a NP-pronominal. In order to show that soo i s a sentential-pronominal and not a NP-pronominal, l e t us observe the following sentences $ (1?) a. Taroo wa Tolstoy no 'Sensoo to Heiwa' to i u syoosetu o yonde i t a ga, kare wa kinoo, sore o daigaku no tosyookan kara k a r i t e k i t a no datta. 'Taroo was reading Tolstoy's novel c a l l e d •War and Peace", and he borrowed i t from the University l i b r a r y yesterday.' b. * Taroo wa Tolstoy no 'Sensoo to Heiwa' to i u syoosetu o yonde i t a ga, kare wa kinoo, soo daigaku no tosyookan kara k a r i t e k i t a no datta. Notice that (17a) with sore o i s a per f e c t l y grammatical sentence, while (17b) with spo i s ungrammatical. In (17a) 12 sore o refers back to the NP element Tolstoy no 'Sensoo to  Helwa* to i u syoosetu 'Tolstoy's novel c a l l e d 'War and Peace'*. The ungrammatlcality of (l?b) points out the fact that soo cannot possibly be a NP-pronominal, but has to be a sentential-pronominal. The f a c t that soo i s a sentential-pronominal i s also evident i n the following examples« (18) Sore wa hon desu ka. •Is that a book?' Hai, soo desu. •Yes, i t i s . ' (19) Sensei wa nihonzin desu ka. •Is the teacher a Japanese?* Hai, soo desu. •Yes, he i s . ' (20) Sore wa sakura no hana desu ka. •Are those cherry blossoms? 1 Hai, soo desu. •Yes, they are.• (21) Kono z i b i k i wa takai desu ka. •Is t h i s dictionary expensive?' Hai, soo desu. •Yes, i t i s . ' (22) Anata no kasa wa akai desu ka. 'Is your umbrella red?' 13 Hai,soo desu. •Yes, i t i s . * (23) A s i t a Tanaka-san ga ikimasu ka. •Is Mr. Tanaka going tomorrow?* * Hai, soo desu. 'Yes, he i s . * (24) Sensei wa moo kaerimasita ka. *Has the teacher gone home already?* * Hai, soo desu. •Yes, he has.• (25) Kono densya wa Ginza o toorimasu ka. •Does t h i s t r a i n go through Ginza?* * Hai, soo desu. *Yes, i t does.* Soo desu i n the above examples i s used as an answer to the questions asked. Notice that soo i n the examples above ref e r back to the sentential element of the question. The soo i n (18) for example, refers back to the sentence Sore  wa hon desu 'That i s a book*. S i m i l a r l y , i n (19) and ( 2 0 ) . soo refers back to the sentences Sensei wa nihonzin desu •The teacher i s a Japanese*, and Sore wa sakura no hana desu 'Those are cherry blossoms*, and so on. Actually, a di r e c t answer to those questions above, f o r example questions (18), (19) and (20) would be 1 14 (18) a. Hai, kore wa hon desu. •Yes, t h i s i s a book.' (19) a. Hai, sensei wa nihonzin desu. •Yes, the teacher i s Japanese.* (20) a. Hai, sore wa sakura no hana desu. •Yes, those are cherry blossoms.* Thus, instead of imitating the question f o r the answer, soo desu i s used instead. From the above examples, i t i s evident that soo has to be a sentential-pronominal, and the use of soo i n the above examples i s the same as that of sentences (13)-(16). However, note the ungrammaticality of Hai, soo desu as answers to questions ( 2 3 ) , (24) and ( 2 5 ) . I t appears that when a verb i s used, i t i s normally not permissible to use soo desu as an answer. At t h i s stage, I do not know why t h i s i s so nor do I have any concrete explanation for i t . Nevertheless, i t should be noted that t h i s p e c u l i a r i t y e x i s t s . Since t h i s paper i s not a study on pronominalization, I w i l l leave t h i s problem open for future research. Several generalizations can be made from the above observation! ( i ) I t seems that whatever soo refers to has to be a senten t i a l element, for example i n sentence (13)» soo refers to the sentence Yamada-san ga kanemoti da 'Mr. Yamada i s r i c h . • Therefore, t h i s points out that soo 15 has to "be a sentential-pronominal. ( i i ) I f sop i s a sentential-pronominal, and soo i n sentences (13)-(16) ref e r s to the pos i t i v e counter-parts of the preceding negative sentences, then the negative formatives i n those preceding sentences w i l l have to negate the entire sentence, and not j u s t the VP. This would mean that the negative has to take a sentential s u b j e c t . T h i s c l e a r l y shows that we w i l l have to acknowledge the p o s s i b i l i t y of generating the negative i n positions other than as the constituent of the VP i n the underlying structure. ( i i i ) One s t r u c t u r a l p o s s i b i l i t y would be to derive the negative nai from an underlying predicate NEG, introduced optionally i n the base structure component as a verb of the higher sentence, instead of deriving the NEG i n the lower constituent, f o r example the VP, as we have done before. We w i l l then have to assume the following to be the underlying structure f o r sentence negation: (26) S NP VP NEG 16 where the negative nai w i l l negate the S which i t immediately commandsi i n t h i s case i t i s S^. In assuming the above abstract structure of (26) as the deep structure f or negative sentences, we can account f o r several thingsi (a) In the structure (26), notice that the proposition i s separated from the element NEG which negates i t , and so i t becomes possible f or the pronominal soo to r e f e r back to the proposition without the negative. (b) In analyses l i k e (7) and (8), where the negative element NEG i s i n the VP and i s part of the sentence that i t negates, i t i s not possible f o r the pronominal soo to r e f e r back to the proposition without the negative. (c) I f we adopt structure (26), sentence (13) would have the following deep structure i 17 Notice that i s i d e n t i c a l to Sy In the process of soo pronominalization transformation, w i l l consequently be replaced with soo. The pronominal soo then w i l l be c o r e f e r e n t i a l with S^. I f we consider (26) to be correct f o r representing the negative structures, we w i l l have the following underlying structures for sentences (lb) and ( 3 b ) i 18 where the negative negates the whole sentences Taroo ga  k i t a and sakura no hana wa akai, respectively. 19 1 . 3 . 2 . ON SOME STRUCTURES FOR NEGATIVE CONSTRUCTIONS 1 . 3 . 2 . 1 . LOCATIVE AND INSTRUMENTAL ADVERBIALS Let us look into more examples of negative sentences, and t r y to determine whether a l l types of negative expressions can he ascribed to a single underlying form. Now, consider the following sentences 1 (30) Hanako wa tomodati to e k i de awanakatta. •Hanako didn't meet her f r i e n d at the railway stati o n . * Sentence (30) above would be the negative counterpart to the pos i t i v e sentence of» (31) Hanako wa tomodati to e k i de a t t a . 'Hanako met her f r i e n d at the railway s t a t i o n . ' In t r a d i t i o n a l grammar, the underlying structure f o r (31) would be analyzed e s s e n t i a l l y asi 20 (32) Hanako NP Loc. V tomodati to e k i de a t t a I t i s generally assumed that eki de *at the railway s t a t i o n ' i s a locative adverb modifying tomodati to a t t a 'met her f r i e n d * , and that tomodati to e k i de a t t a 'met her f r i e n d at the railway station* forms a single deep structure constituent. I f we assume (32) to be the underlying structure f or (31)» then the underlying structure of i t s negative counterpart (30) would have to be analyzed as ( 3 3 ) . 21 (33) s NP VP S NEG tomodati t * eki de atta to However, note that sentence (30) is ambiguous in at least two ways. One reading of sentence (30) does not presuppose that Hanako met her friend, but i s simply a denial of the assertion that the meeting took place. This would be synonymous withi (3*0 Hanako ga tomodati to awanakatta no wa eki de da. •It i s at the railway station that Hanako didn't meet her friend.* Here, the negation i s associated with the main verb of the matrix sentence. 22 In another reading, i t i s presupposed that Hanako did meet her f r i e n d , but i t i s denied that the meeting took place at the railway s t a t i o n . Note that t h i s would he synonymous with: (35) Hanako ga tomodati to a t t a no wa e k i de de wa n a i . • I t i s not at the railway s t a t i o n that Hanako met her f r i e n d . * In sentence (35)» i t i s clear i n the surface structure that the l o c a t i o n of the event i s being negated, not the assertion that the event occurs. The negative appears to be semantically associated with the adverb,of l o c a t i o n , and not the main verb of the matrix sentence. Note that i t i s not possible to account f o r the way i n which we understand (35) i f we were to derive i t from the underlying structure of (33)• I t seems that one way that we could reasonably account f o r t h i s ambiguity of (30) would be to subscribe to the proposal made by G. Lakoff (1965), that adverbials such as Locative, Time and Instrumental Adverbials, are derived from verb phrases of 'higher' simplex sentences than the ones that appear as the main clauses i n the surface structures. Following L a k o f f s proposal, we could then derive (35) from the following underlying structures: 23 (36) S Hanako NP VP tomodati to a t t a i In sentences which have both negatives and adverbials, then the understood order of negatives and adverbials i n these sentences are supposed to correspond to the hierarchy of upper sentences containing negatives and adverbials. Thus the difference i n meaning between (3*0 and (35) i s r e f l e c t e d i n t h e i r deep structures of (37) and (38), respectively. 25 In (37) i t i s clear that the locative adverbial i s above the negative, and what i s actually being negated i s the embedded sentence. In (38), the negative i s above the adverbial, showing that i t i s the locative adverbial that i s being negated. Likewise, observe the following sentencesi (39) Watakusi wa kono kasa o Mitukosi Depaato.de kawanakatta. 26 •I didn't buy t h i s umbrella at Mitukosi Department Store.• (40) Taroo wa kyoositu de nenakatta. 'Taroo didn't sleep i n c l a s s . ' (41) Sakana wa kono esa de turenakatta. •The f i s h was not caught with t h i s b a i t . ' (42) Taroo wa genkotu de Hanako o naguranakatta. •Taroo didn*t h i t Hanako with his f i s t . * The above four sentences are ambiguous i n the same way as sentence (30) i s . Each of them i s ambiguous i n at least two readings, depending on what comes within the scope of the negation. Sentence (39) can be understood i n at least two ways. In one reading i t does not deny that I bought the umbrella. I t assumes that I d i d , and only denies that I bought i t at Mitukosi Department Store. This i s synonymous withr (43) Watakusi wa kono kasa o katta no wa Mitukosi Depaato de de wa n a i . • I t i s not at the Mitukosi Department Store that I bought t h i s umbrella.* Notice that the negative i s semantically associated with the locative adverb and not with the main verb. We can account for t h i s reading of (39) i f we derive i t from the abstract structure underlying (4-3). This would bet The other reading i s simply a denial that I bought the umbrella, and would be synonymous with* (45) Watakusi ga kono kasa o kawanakatta no wa Mitukosi Depaato de da. * I t i s at the Mitukosi Department Store that 28 I didn't buy the umbrella.' We can account for t h i s reading by deriving i t from the structure underlying (45), which would bei The same i s true of sentence (40), where i n one reading i t assumes that Taroo did sleep. What i s denied i s that the location of the sleeping i s i n the classroom. This would be 29 synonymous withi (47) Taroo ga neta no wa kyoositu de de wa n a i . 'It i s not i n the classroom that Taroo sl e p t . ' The other reading i s simply a denial that Taroo sl e p t , and t h i s would be synonymous withi (48) Taroo ga nenakatta no wa kyoositu de da. • I t i s i n the classroom that Taroo didn't sleep.' We can account for these two d i f f e r e n t readings of (40) i f we were to derive them from the abstract structures underlying (47) and (48) respectively. The same arguments that we have given f o r locative adverbs also apply to negative sentences containing instrumental adverbials, as i n examples (41) and (42). In one reading of (41), i t denies the f a c t that the f i s h was caught and does not assume that the catch took place. This would be synonymous withi (49) Sakana ga turenakatta no wa kono esa de da. ' I t i s with t h i s b a i t that the f i s h was not caught.' 30 The underlying structure f or t h i s would bei (50) S S kono esa de da NP VP S NEG NP VP sakana tureta In another reading i t i s assumed that the f i s h was caught, but i t i s denied that i t was done so with t h i s b a i t . This would be synonymous withj (51) Sakana ga tureta no wa kono esa de de wa n a i . ' I t i s not with t h i s b a i t that the f i s h was caught.' 31 In (51) i t i s clear that the instrumental adverbial /with t h i s b a i t * i s being negated. We can aceount f o r the way i n which we understand t h i s reading by deriving i t from the abstract structure underlying (51) . (52) S NP VP S kono esa de da NP VP sakana tureta Sentence (42) i s also ambiguous i n the same way. In one reading, i t i s not assumed that the h i t t i n g took place, and i s simply a denial of the assertion that the event took place. This would be synonymous withi 32 (53) Taroo ga Hanako o naguranakatta no wa genkotu de da. •I t i s with h i s f i s t that Taroo didn't h i t Hanako.' Note that i n t h i s reading the negative commands the main verb nagutta ' h i t ' . In another reading, i t i s assumed that Taroo did h i t Hanako, but i t i s denied that he did so with hi s f i s t . This reading i s synonymous witht (5*0 Taroo ga Hanako o nagutta no wa genkotu de de wa n a i . ' I t i s not with h i s f i s t that Taroo h i t Hanako.* Note that i n t h i s reading, i t i s the instrumental adverb genkotu de 'with h i s f i s t * , that comes within the scope of negation. We can account f o r our understanding of the two readings of (42) by deriving them from the structures underlying (53) and (5*0 respectively. 1.3.2.2. REASON ADVERBIALS Let us now consider the following sentences and see i f the same ambiguity that e x i s t s i n sentences containing locative and instrumental adverbials also e x i s t i n sentences 33 containing reason adverbials. We w i l l also examine some sentences to see whether we can bring f o r t h the same arguments f o r deriving reason adverbials i n the same manner as we derive locative and instrumental adverbials. (55) Tanaka-san wa kodomo o kawaigatte i r u kara, naguttari s i n a i . •Since Mr. Tanaka loves his c h i l d , (he) doesn't beat (him).* (56) Hahaoya ga daite ageta node, kodomo wa nakanakatta. 'Because mother hugged (her), the c h i l d didn't cry.' (57) Otoo-san ga kaette k i t a node, benkyoo sinakatta. 'Because father came home, (I) didn't study.* (58) Sono z i b i k i wa takakatta kara, kawanakatta. •Since that dictionary was expensive, (I) didn't buy . ( i t ) . ' (59) Atatakai kara seetaa o motte konakatta. •Since i t was warm, (I) didn't bring a sweater.' We have seen that sentences containing both negatives and adverbials (such as locative or instrumental adverbials) are ambiguous, and can be understood i n at le a s t two d i f f e r e n t ways, depending on what f a l l s within the scope of negation.* However, the above sentences (55)-(59)» which have both a negative and a reason adverbial, do not seem to be ambiguous 3^ at a l l . For example, sentence (55) can only mean that since Mr. Tanaka loves his c h i l d , he doesn't beat him. The negative commands only V2» that i s , the predicate phrase contained i n Sg. (From now on, and V 2 w i l l be referred to as those predicate phrases that are contained i n S^ and S 2 r e s p e c t i v e l y ) . . S i m i l a r l y , i n sentences ( 5 6 ) - ( 5 9 ) , the negative can have as i t s domain, only the verb to which i t i s attached, for example, nakanakatta 'didn't cry', benkyoo sinakatta 'didn't study', kawanakatta 'didn't buy', and motte konakatta 'didn't bring'. Sentence ( 5 5 ) , for example, w i l l have the following underlying structuret (60) S NP VP NP S NP VP Tanaka-san NP VP kodomo kawaigatte i r u kara da S NEG NP VP Tanaka-san kodomo naguttari suru 35 Except f o r l e x i c a l items, sentences (56)-(59) w i l l have s i m i l a r underlying structures to that of ( 6 0 ) . In order to confirm our i n i t i a l observation that the reason adverbials do not cause any s t r u c t u r a l ambiguity, l e t us examine a few more examples of sentences with the reason adverbials. (61) t Amari t o n a r i no heya ga urusai node, yoku nemurenai. 'Because the next room i s too noisy, (I) can It sleep w e l l . * (62) Kono m i t i ga semai node, torakku ga toorenai. 'Because t h i s road i s narrow, trucks can't pass through.* (63) Hako wa omokatta kara, motenakatta. 'Since the box was heavy, (I) couldn't carry ( i t ) . ' (64) Takusan benkyoo ga aru node, eiga o mi n i ik e n a i . 'Because (I) have a l o t of studying to do, (I) can't go and see a movie.' (65) Asi'ta o-susi o tukuru kara, kyoo wa tukuranai. 'Since (I) am making susi tomorrow, (I) won't make ;(xany) today.' (66) Kaze o h i i t e i r u kara, gakkoo e ik a n a i . 'Since (I) have a cold, (I) won't go to school.' 36 Notice that examples (61)-(66) are not ambiguous just as sentences (55)-(59) are not. Similarly, in ( 6 l ) - ( 6 6 ) , the sentence-final negative can only command Vg, and cannot extend i t s command power over (V^ kara/node v 2 ) , for example, ...urusai node,...nemurenai 'because...noisy,...can't sleep well' or ...omokatta kara. motenakatta 'since...was heavy, (I) couldn't carry (it)11 From the above observation, i t seems that the reason adverbials like kara and node prevent the command power of the negative from extending over to the l e f t of them. One can speculate the reason for this phenomenon, and I w i l l consider i t in the following pages. Now, observe the following sentences in order to see what happens i f the negative occurs outside of the main sentence. (67) Tanaka-san wa kodomo 0 kawaigatte iru kara, naguttari suru no de wa nai. 'It is not the case that since Mr. Tanaka loves his child, (he) beats (him).' (68) Hahaoya ga daite ageta node, kodomo wa naita no de wa nai. •It i s not the case that because mother hugged (her), the child cried.' (69) Otoo-san ga kaette kita node, benkyoo s i t a no de wa nai. 'It i s not the ease that because father came home, 3 7 (I) studied. 1 ( 7 0 ) Sono z i b i k i wa takakatta kara, katta no de wa nai. • I t i s not the case that since that dictionary-was expensive (I) bought ( i t ) . * (71) Atatakai kara seetaa o motte k i t a no de wa n a i . *I t i s not the case that since i t was warm, (I) brought a sweater.* Unlike ( 5 5 ) - ( 5 9 ) , sentences ( 6 7 ) - ( 7 1 ) are a l l ambiguous and can be understood i n at least two ways. For example, ( 6 7 ) i s ambiguous between ( 7 2 ) and ( 7 3 ) . ( 7 2 ) Naguttari suru no wa, Tanaka-san ga kodomo o kawaigatte i r u kara de wa nai.' •I t i s not because Mr. Tanaka loves his c h i l d that (he) beats (him).' ( 7 3 ) Naguttari s i n a i no wa Tanaka-san ga kodomo 0 kawaigatte i r u kara da. 'It i s because Mr. Tanaka loves h i s c h i l d , that (he) doesn't beat (him).' In ( 7 2 ) , i t i s presupposed that Mr. Tanaka beats h i s c h i l d , and the reason f o r beating the c h i l d i s not because he loves him. In t h i s sense of ( 7 2 ) , the command power of the negative extends to the l e f t of node and i t commands the whole 38 (V^ node V 2 ) . In (73), i t i s not presupposed that Mr. Tanaka beats his c h i l d . In t h i s sense, the negative commands only V^. Sentences (68)-(71) are a l l ambiguous i n the same way as (67) i s . The above (72) and (73) w i l l have the underlying structures of (7^) and (75) respectively. (74) s NP VP S NEG NP VP s Adv. Copula Tanaka-san kodomo naguttari suru S kara da Tanaka-san kodomo kawaigatte i r u 39 From the above observation, we can conclude that f o r some reason the sentence-final negative cannot command the reason adverbials. The reason adverbials somehow seem to block the command power of the negative from extending over them. So f a r as I can see, i t seems that t h i s p e c u l i a r i t y e x i s t s only i n the case of reason adverbials co-occurring with the negative i n a sentence. As we have observed e a r l i e r , with other adverbials such as locative and instrumen-t a l adverbials, t h i s p e c u l i a r i t y does not e x i s t . McGloin, i n her thesis Some Aspects of Negation i n Japanese (1972), suggests that perhaps the; command power of the negative might be blocked by the fa c t i v e reason adverbial clauses. 40 Compare the following sentences^ (a) Why did Max h i t anybody? (b) When did Max h i t anybody? J . Lawler (1971) argues that why i s f a c t i v e . According to Lawler's argument, he claims that the above (a) sentence presupposes that Max h i t somebody, but the (b) sentence does not. Related to t h i s , McGloin states that i f Lawler's claim i s true, then perhaps t h i s might be the explanation as to why the sentence-final negative cannot command the reason adverbials. The reason adverbials l i k e node and kara are f a c t i v e , and presupposes the f a c t i v i t y of the complement sentences. Consequently, the negative cannot extend over to the l e f t of the reason adverbials because the command power of the negative i s blocked by the fact i v e clauses. However, i t seems to me that such a claim may be subject to further investigation. On the other hand, sentences which have the nominalizer no are ambiguous, and i n one reading, the command power of > the negative can be extended to the l e f t of the reason adverbial, making i t a negation of the whole sentence. S. 7 Kuno, i n h i s paper Degrees of Subordination/ explains that i n order to enable the negative to extend i t s command or influence to the l e f t of node or kara, i t i s necessary, f i r s t to make the node or kara clause, for example, Hahaoya 41 ga daite ageta node, a noun clause by nominalizing i t with the nominalizer no, and then l e t the negative command the "(Noun-clause) no da". The above hypothesis must be considered extremely tentative, and s t i l l requires further studies and firmer v e r i f i c a t i o n . Nevertheless, a further study into t h i s problem, though i n t e r e s t i n g and challenging, i s beyond the scope of t h i s paper, and w i l l therefore, be l e f t open for future research. 1 . 3 . 2 . 3 . TIME ADVERBIALS Just as we brought f o r t h arguments to show that locative and instrumental adverbials are derived from verb phrases of simplex sentences which are higher i n the base phrase marker than the ones that appear i n the main clauses i n the surface structures, so we can bring f o r t h the same argument f o r deriving time adverbials i n t h i s manner. I t can be observed that a negative sentence l i k e t (76) Yukiko wa yoru no zyuuni-?zi made utawanakatta. •Yukiko did not sing u n t i l twelve midnight.* i s ambiguous i n at least two ways. In one reading, i t i s assumed that the singing did take place. What i s denied kz here i s the time factor, meaning that Yukiko sang, hut she stopped singing sometime before midnight. Note that t h i s would be synonymous with» (77) Yukiko ga ut a t t a no wa yoru no zyuuni-zi made de wa n a i . 'I t i s not u n t i l twelve midnight that Yukiko sang.' In another reading, i t means that Yukiko did not begin singing u n t i l midnight, that i s to say, her not singing stopped at midnight. This would be synonymous withi (78) Yukiko ga utawanakatta no wa yoru no zyuuni-zi made da• 'I t i s u n t i l twelve midnight that Yukiko did not sing.' Although on the surface structure, the negative i s attached to the main verb of the matrix sentence utawanakatta 'did not sing*, the reading of (77) shows that i t i s the time element that comes withi n the scope of negation. Therefore, the only way that we could reasonably account f o r the ambiguity of (76) would be to derive i t from the abstract structures underlying (77) and (78). The underlying struc-ture f o r (77) would be (79), and f o r (78) would be (80). Note that i n ( 7 9 ) , the NEG i s introduced outside the structure of the embedded sentence, and i s above the u n t i l phrase, i n d i c a t i n g that i t i s actu a l l y the whole sentence that i s negated. The command power of the negative extends over the whole of (V^ made V 2 ) , that i s , yoru no zyuuni-zi  made ut a t t a 'studied u n t i l twelve midnight'. In ( 8 0 ) , the NEG dominates only the embedded sentence, showing that 'Yukiko sang* i s being negated. This kind of ambiguity i s also true of sentences l i k e (81) with an adverbial of duration. (81) Taroo wa issyuukan benkyoo sinakatta. 'Taroo didn't study for one week.' Sentence (81) i s ambiguous meaning either that Taroo's not studying lasted f o r one week, in d i c a t i n g the duration of his not studying» (82) Taroo ga benkyoo sinakatta no wa issyuukan da. 'It i s f o r one week that Taroo didn't study.' or, that Taroo did study, but he did so f o r less than one week, where the time element i s negatedi (83) Taroo ga benkyoo s i t a no wa issyuukan de wa nai. ' I t i s not for one week that Taroo studied.' Although on the surface structure, the negative i s attached to the main verb of the matrix sentence benkyoo sinakatta •didn't study', the reading of (83) shows that i t i s the time element that comes within the scope of negation. The difference i n meaning between (82) and (83) can again be re f l e c t e d i n t h e i r deep structures of (84) and ( 8 5 ) . (84) S NP VP S issyuukan da NP VP S NEG NP VP Taroo benkyoo s i t a 46 The same goes for the ambiguity of the following sentences, depending on where the negative l i e s i n the deep structure, and what f a l l s within the scope of negation. (86) Taroo wa h a t i - z i made gohan o tabenakatta. 'Taroo didn't take (his) meal u n t i l eight o'clock.' (87) Yamada-san wa h i r u made hanasanakatta. •Mr. Yamada didn't t a l k u n t i l noon.' (88) Kare wa issyuukan hataranakatta. 'He didn't work f o r a week.' ^7 (89) Ano roozin wa i t i - n i t i arukanakatta. 'That old man didn't walk the whole day.' (90) Akanboo wa hitobanzyuu nakanakatta. 'The baby didn't cry throughout the whole night.' Now observe the following sentences 1 (91) S i b a i wa zyuu-zi made owaranakatta. 'The play didn't end u n t i l ten o'clock.' (92) Karera wa hutukakan hazimenakatta. 'They didn't s t a r t for two days.' (93) Kega s i t a k o t o r i wa tug i no h i made sinanakatta. •The wounded l i t t l e b i r d didn't die u n t i l the next day.' (94) Kisya wa yoru no ku-zi made tukanakatta. 'The t r a i n didn't arrive u n t i l nine i n the evening.' The above sentences are a l l unambiguous and can have only one reading. For example, (91) can be understood only a s i (95) S i b a i ga owaranakatta no wa zyuu-zi made da. •It i s u n t i l ten o'clock that the play didn't end. • Sentence (92) i s synonymous with only the reading o f t 48 (96) Karera ga hazimenakatta no wa hutukakan da. 'It i s f o r two days that they didn't s t a r t . ' F i n a l l y , the only possible reading f o r (93) and (9*0 would be (97) and (98) respectively. (97) Kega s i t a k o t o r i ga sinanakatta no wa t u g i no h i made da. •It i s u n t i l the next day that the wounded l i t t l e b i r d didn't die.' (98) Kisya ga tukanakatta no wa yoru no ku-zi made da. 'It i s u n t i l nine i n the evening that the t r a i n didn't a r r i v e . * In the above examples the negative can have i n i t s domain, only the verbs to which i t i s attached,ifor example, owaranakatta 'didn't end', hazimenakatta 'didn't s t a r t ' , sinanakatta 'didn't die', and tukanakatta 'didn't a r r i v e ' . We see that i n examples (91)-(9^)» there can be only one possible reading f or them, as opposed to examples (76)-(81) and (86)-(90) which are a l l ambiguous i n at least two readings. The l o g i c a l reading f or t h i s would be that examples (76)-(81) and (86)-(90), contain semantically durative verbs such as utau 'to sing', benkyoo suru 'to study', taberu 'to eat*, hanasu 'to t a l k * , hataraku 'to work*, aruku 'to walk', and naku 'to cry'. These semantically durative verbs 4-9 occurring with either the u n t i l phrase or adverbials of duration i n a sentence, are ambiguous i n at least two ways when negated. This shows that the command power of the sentence-final negative can extend over a wider range. In one reading, the negative commands only the V 2, such as utau 'to sing' of (?6) and benkyoo suru 'to study' of (81). In another reading, the negative commands the whole of (V^ made/adverbial of duration V 2 ) , such as zyuuni-zi made utau 'to sing u n t i l twelve midnight' of (76), and issyuukan  benkyoo suru 'to study f o r one week* of (81). This means that the command power of the negative can extend to the l e f t of either the u n t i l clause or the adverbial of duration clause. On the other hand, sentences having semantically punctual verbs (such as owaru 'to end', hazimeru 'to s t a r t ' , sinu 'to die', and tuku 'to arrive') as i n examples (91)-(9*Ot when occurring with either the u n t i l phrase or adverbials of duration, do not give r i s e to any ambiguity at a l l when negated. They can be understood as having only one reading. In t h i s case, the negative can only command V 2. However, notice also that sentences containing semantically punctual verbs cannot be used with the u n t i l phrase or adverbials of duration unless they are negated, as i s evident i n the ungrammaticality of the following sentences! (99) * S i b a i wa zyuu-zi made owatta. 50 •The play ended u n t i l ten o'clock.* (100) * Karera wa hutukakan hazimeta. "They started f o r two days.* (101) * Kega s i t a k o t o r i wa tugi no h i made sinda. •The wounded l i t t l e b i r d died u n t i l the next day.' (102) * Kisya wa yoru no ku-zi made t u i t a . •The t r a i n arrived u n t i l nine i n the evening.' This i s due to the f a c t that a semantically punctual verb becomes a semantically durative verb when negated, and hence can occur with either the u n t i l phrase or adverbials of duration. 1.4. SUMMARY At t h i s stage, I would l i k e to point out that t h i s paper i s based on the assumption that deep structures are the relevant l e v e l of grammar for semantic interpretation. The deep structures contain a l l those elements that c o n t r i -bute to meaning. Together, the deep syntactic structure of a sentence and the meanings of the words used i n that structure contribute to the t o t a l meaning of the sentence. We have seen that i t i s important to determine the posi t i o n of the NEG i n the deep structure, because the 51 structure of the negative construction and consequently, the semantic i n t e r p r e t a t i o n of the sentence i s dependent on where the negative l i e s . A negative sentence may have two or more d i f f e r e n t meanings depending on where the negative i s located i n the deep structure and on what f a l l s within •'-the command power of the negative. I t i s possible to get an ambiguity i n a sentence depending on whether the NEG lay within or outside the embedded deep structure sentences. This shows that the s y n t a c t i c a l ambiguity of the surface structures of negative sentences, resides i n what may be c a l l e d the "scope of the negative", that i s , i n what exactly i s being negated i n the sentence. To summarize: ( i ) The sentence-final negative formative na i n Japanese i s derived from a single underlying predicate NEG, which i s introduced optionally i n the underlying structure as a verb of the "higher" sentence. ( i i ) As we have pointed out before, l i k e the negative, adverbials are also generated i n the base as verb phrases of "higher" simplex sentences than the s u p e r f i c i a l main clause i n the surface structure. In sentences containing both adverbials and negatives, the understood order of the adverbials and negatives corresponds to the hierarchy of upper sentences containing these adverbials and negatives. Thus a sentence containing both a negative element and an 52 adverbial i s ambiguous, and can be understood i n at least two ways, depending on whether the negative i s introduced below or above the adverbial i n the under-l y i n g structure. This kind of ambiguity i s true of sentences containing either the l o c a t i v e , instrumen-t a l or time adverbials co-occurring with a negative. ( i i i ) I t has been noted that unlike other adverbials, the reason adverbials such as node and kara, co-occurring with a negative element i n a sentence does not give r i s e to any ambiguity at a l l . The sentence-final negative can only command V 2 and cannot command the reason adverbials. Somehow, i t seems that the reason adverbials prevent the command power of the negative from extending over to the l e f t of them, unless we f i r s t nominalize the S^ clause, that i s , the reason adverbial clause, by the nominalizer no, thus making i t a noun clause and then l e t the negative command the whole "(noun clause) no da". This p e c u l i a r i t y seems to exi s t only i n the case of reason adverbials co-occurring with the negative i n a sentence. I t has "been suggested that perhaps the command power of the negative might be blocked by the fa c t i v e reason adverbial clause. (iv) I t has also been observed that semantically durative verbs occurring with either the u n t i l phrase or adverbials of duration, are ambiguous when negated. 53 The extent of the command power of the negative can vary between either the V 2 only or the whole of (V^ made/adverbial of duration V 2 ) . On the other hand, semantically punctual verbs cannot occur with the u n t i l phrase or adverbials of duration unless they are negated. When they occur i n negated sentences, the sentences are unambiguous, and the negative can only command Y9. FOOTNOTES 1. Chomsky, Noam. 1957. Syntactic Structures. Mouton & Co. 'S-Gravenhage. (pp. 112). 2. The subject marker ga and the object marker o are not specified i n the underlying tree structures, since they are predictable and can be introduced by transformation. We are only interested i n presenting the outline of the underlying structure. 3. Inoue, Kazuko. 1964. A Study of Japanese Syntax. The University of Michigan, Ph. D. Dissertation, (pp. 116-117). 4 . Soga, Matsuo. 1966. Some Syntactic Rules of Modern  Co l l o q u i a l Japanese. The University of Indiana, Ph.1 D. Dissertation. 5. N. McGloin, i n her thesis Some Aspects of Negation i n Japanese (1972), has given arguments for the claim that i n Japanese, the negative i s an underlying i n t r a n s i t i v e verb which takes a sentential subject. 6. Lawler, John M. 1971. "Any Questions?" Papers from  the Seventh Regional Meeting of the Chicago L i n g u i s t i c  Society~, University of Chicago, I l l i n o i s . (pp. 163). 7. Kuno, Susumu. 1973. "Degrees of Subordination". The Structurefcoffthe®Japanese Language. The MIT Press. Cambridge, Massachusetts & London, England, (pp. 200-2 0 9 ) . 55 CHAPTER TWO  NEGATIVE TRANSPORTATION 2.1. 'INTRODUCTION This chapter w i l l deal mainly on a rule that has been proposed f o r English c a l l e d Negative Transportation. 1 Such a rule has been assumed to ex i s t i n many other languages as we l l , and i n t h i s chapter, I w i l l investigate whether or not the Hegative transportation rule e x i s t s and i s applicable to Japanese. The investigation into the a p p l i c a b i l i t y of t h i s rule to Japanese w i l l follow s i m i l a r l i n e s of arguments that have been presented for English. Before we examine the Japanese examples, I think i t would be worthwhile to,look into some of the arguments that have been presented for negative transportation rule i n English. 2.2. NEGATIVE TRANSPORTATION IN ENGLISH Charles Fillmore, i n hi s a r t i c l e The Position of  Embedding Transformations i n a Grammar, f i r s t proposed negative transportation as a rule for English. He 56 proposed the rule i n order to account f o r the r e l a t i o n s h i p between (la) and ( l b ) , and (2a) and (2b). (1) a. I believe John i s n ' t going to the party. b. I don't believe John i s going to the party. (2) a. I thought John didn't l i k e Mary, b. I didn't think John l i k e d Mary. Fillmore has pointed out that for at l e a s t one interpretation of the above (b) sentences, i t s meaning i s synonymous to the respective (a) sentences. Take sentences (la) and (lb) f o r example. Notice that (lb) i s ambiguous. One reading f o r i t can be understood as an ordinary negative, meaning ' I t i s not so that I believe John i s going to the party*. This i s simply a denial of the sentence 'I believe John i s going to the party', and does not commit the speaker to any b e l i e f at a l l . Another reading of (lb) i s synonymous to that of (la) where the speaker was not denying that he was b e l i e v i n g , but rather saying that he believed that i t was not true that John i s going to the party. In t h i s l a t t e r reading, the negative not actually negates the verb of the embedded sentence, although the negative appears overtly i n the matrix sentence. The same i s also true of (2b), where one,reading of which i s synonymous to that of (2a). Therefore, i n the above examples, the (a) and (b) sentences are semantically equivalent to each other. In other words, f o r a given 57 s i t u a t i o n , i f (a) i s true, then (b) i s also true and vice versa. Fillmore claimed that the (b) sentences are derived from the same structures that underlie the respective (a) sentences, and that the negative transportation rule moves the negative out of the embedded sentence to the main sentence. The following (3) and (4) are the underlying structures f o r (la) and (2a) respectively. (3) NP VP I V NP believe i t John i s going to the party 58 Notice that the negative originates i n the embedded sentence S^. To derive the (b) sentences, the NEG i s moved up the tree to the matrix sentence S Q. There have been several arguments proposed which seem to provide f a i r l y conclusive evidence that the negative transportation rule e x i s t s . One of the arguments c i t e d i n defense f o r the existence of t h i s r u l e i s the simplex sentence condition of the negative p o l a r i t y adverbial u n t i l which appears 5in the following sentences of (5) and ( 6 ) . (5) a. I thought you wouldn't leave u n t i l tomorrow, b. I didn't think you would leave u n t i l tomorrow. (6) a. I believe the plane w i l l not arrive u n t i l ten 59 o'clock. b. I don't believe the plane w i l l arrive u n t i l ten o'clock. The simplex sentence condition states that the u n t i l adverbials can only occur with a semantically durative verb, and not with a semantically punctual verb. However, the u n t i l adverbials can occur with these semantically punctual verbs such as leave and arrive only i f these verbs are negated. Hence the grammaticality of (7a) and the ungra-mmaticality of ( 7 b ) . (7) a. The plane w i l l not arrive u n t i l ten o'clock, b. * The plane w i l l arrive u n t i l ten o'clock. I t has been claimed that (5b) and (6b) are derived from the same structures that underlie f(5a) and (6a) respectively. The syntactic argument that has been considered to give c r u c i a l support to the semantic grounds f o r claiming that the (a) and (b) sentences are derived from the same under-l y i n g structure i s as follows t Notice that sentences (5b) and (6b) are not ambiguous at a l l and can have only the meaning of t h e i r (a) counterparts. This shows that the negative of (5b) and (6b) must have originated i n the embedded sentence, and then moved up to the matrix sentence by application of the negative transportation r u l e . Also, 60 the negative element that appears i n the matrix sentence on the surface of sentences (5b) and (6b) cannot be regularly assigned to the matrix sentence i n the deep structure because, i f that happens, then the deep structure complements of (5b) and (6b) would then have to be (8) and ( 9 ) , which are ungrammatical. (8) * You would leave u n t i l tomorrow. (9) * The plane w i l l arrive u n t i l ten o'clock. The above syntactic argument seems to give a f a i r l y strong support f o r the existence of the negative transportation r u l e . Let us now consider the following sentences where the verbs think and believe of (5b) and (6b) are substituted with say and claim. (10) * I didn't say that you would leave u n t i l tomorrow. (11) * I didn't claim that the plane would arrive u n t i l ten o'clock. Notice that sentences (10) and (11) are ungrammatical i f we substitute verbs such as say and claim for think and believe. Sentence (10) contains the ungrammatical sentence You would leave u n t i l tomorrow embedded as the subject of 61 say, and sentence (11) contains the ungramraatical sentence the plane w i l l arrive u n t i l ten o'clock,as the object of claim. The negative element required to make leave modified by u n t i l tomorrow and arrive modified by u n t i l ten o'clock grammatical must be i n the same embedded sentence, and not i n a higher sentence, as they are i n the (10) and (11) examples. The ungrammaticality of (10) and (11) which results from substituting verbs such as say and claim for think and believe reveals that negative transportation i s a rule which applies to a r e l a t i v e l y small class of verbs .... non-factive verbs of mental state, and one or two intransi*!--. t i v e s . Verbs l i k e think, believe. suppose, guess. expect, want, seem, and l i k e l y are some of the verbs that undergo negative transportation, while say, claim, f e e l , r e a l i z e , 2 hope and many others do not. The grammaticality and acce p t a b i l i t y of such sentences as (12)t (12) I don't believe Mary wanted John to leave u n t i l tomorrow. shows that negative transportation i s a c y c l i c r u l e . The negative i n (12) originates i n the sentence containing leave of I believe Mary wanted John not to leave u n t i l tomorrow, and i s then raised i n successive cycles f i r s t over want to give I believe Mary didn't want John to leave u n t i l tomorrow. 62 and then over believe to derive (12). Robin Lakoff (1969)i gives a strong syntactic argument i n defense of the existence of the negative transportation r u l e . Her argument revolves around the formation of the tag questions. Let us consider some of the c r u c i a l points which she examines. In general, with a positive sentence, one gets a negative tag question and vice versa. For ' examplei (13) a. Mary has arrived, hasn't she? b. Mary hasn't arrived, has she? Now, consider the following sentencest (14) I don't suppose the yankees w i l l win, w i l l they? (15) John doesn't think the yankees w i l l win, does he? which are both grammatical. Tag formation usually applies on the topmost S on the surface structure, and t h i s accounts for the sentence (15) , but not sentence (14). I t has been claimed that when the topmost S contains a performative verb^ l i k e suppose, then the tag goes with the verb i n the next S down. However, t h i s s t i l l does not account for sentence (14), since the tag i s s t i l l i n the p ositive even though the verb win has no overt negative, and hence an apparent v i o l a t i o n of the tag formation r u l e , 63 which states that a positive statement requires a negative tag and vice versa. The grammaticality of (14) suggests that the negative must have been i n the embedded S at the time that the rule of tag formation applied, and then the negative must have been moved up to suppose by a subsequent application of the negative transportation r u l e . Lakoff argues that t h i s paradox can be accounted f o r by making the following assumptions« ( i ) A performative abstract verb suppose e x i s t s , and that one underlies (15) but not (14). ( i i ) Negative transportation and tag formation rule apply c y c l i c a l l y . Sentence (16) i s considered to be the underlying structure fo r (14). (16) ((I.suppose) (NEG The yankees w i l l win)) Tag formation f i r s t applies on the embedded sentence to give the intermediate s t r i n g I suppose the yankees won't win, w i l l  they? Then negative transportation applies and the negative i s raised to suppose, deriving (14). For sentence ( 15) , "the s i t u a t i o n i s d i f f e r e n t . Lakoff assumes an abstract performa-t i v e verb suppose underlying ( 15) , and the tag i s formed without moving the negative out of the embedded sentence 64 which i s commanded by suppose. (15) i s assumed to have the underlying structure of (17). (17) (I suppose) ((John thinks) (NEG The Yankees w i l l win)) Tag question formation does not apply because there i s no performative verb suppose commanding the negative. Instead, negative transportation f i r s t applies because the verb think i s sensitive to the r u l e , to give the intermediate s t r i n g (18). (18) (I suppose) ((NEG John thinks) (The Yankees w i l l win)) At t h i s stage the negative i s commanded by the abstract performative verb suppose,,and so tag question formation must apply. Since the verb i s negative, a positive tag i s attached to the sentence to derive ( 15) . Since the verb suppose i s abstract, a further negative transportation rule does not apply. The two syntactic arguments c i t e d above, that i s , the simplex sentence condition of the u n t i l adverbial and the tag question formation arguments, seem to provide f a i r l y conclusive evidence for the existence of the negative transportation r u l e . I have also "described b r i e f l y the conditions necessary for the negative transportation rule 65 to apply. One such condition i s that the rule i s a syntactic rule which can apply c y c l i c a l l y . Another i s that the rule applies to a r e l a t i v e l y small class of verbs, that i s , non-facti v e verbs of mental state, and one or two i n t r a n s i t i v e s , which are considered to be sensitive to the r u l e . 2 . 3 . ARGUMENTS FOR NEGATIVE TRANSPORTATION IN JAPANESE In the previous section, I have b r i e f l y explained what negative transportation i s , the conditions necessary for the rule to apply, and I have also c i t e d two syntactic arguments which have been considered to give c r u c i a l support f o r the existence of the negative transportation r u l e . In t h i s section, I w i l l examine some examples i n Japanese to see i f such a rule does exi s t i n Japanese. I t has been assumed that the negative transportation rule i s applicable to many natural languages, and i f t h i s i s the case, I think i t would be worthwhile to examine some Japanese examples along s i m i l a r l i n e s of arguments that have been presented f o r English, to see i f the rule i s also applicable to Japanese. 2.3.1. SIMPLEX SENTENCE CONDITION FOR NEGATIVE POLARITY ADVERBIALS AND PARTICLES 66 One of the arguments that I w i l l examine w i l l be the simplex sentence condition argument which i s based on the evidence that the negative has to occur i n a simplex sentence, but not necessarily i n a sentence embedded i n a negative matrix sentence containing a verb or an adjective assumed to be sensitive to the negative transportation r u l e . In Japanese, there are several adverbials and p a r t i c l e s such as kes s i t e , tittomo, s i k a , t o o t e i , mettani, zenzen, made, and many more, which require the presence of the negative morpheme i n the same simplex sentence at some l e v e l of the derivation. This means that such adverbials and p a r t i c l e s can occur only with the negative, or with semanti-c a l l y negative verbs and adjectives. Consider the following sentences: (19) a. Sekiyu no nedan wa kessite agaranai to omou. '(I) think the price of o i l w i l l never go up.' b. Sekiyu no nedan wa kessite agaru to wa omowanai. '(I) don't think the price of o i l w i l l ever go up.' c. * Sekiyu no nedan wa kessite agaru. (20) a. Nyuugaku siken wa tittomo muzukasikunai to omou. "(I) think the entrance examination i s not d i f f i c u l t at a l l . ' 67 b. Nyuugaku siken wa tittomo muzukasii to wa omowanai. *(I) don't think the entrance examination i s d i f f i c u l t at a l l . ' c. * Nyuugaku siken wa tittomo muzukasii. (21) a. Sonna hanasi wa okaa-san n i s i k a iwanai de moraitai. '(I) want (you) not to t e l l that kind of story to anyone but only to my mother,' b. Sonna hanasi wa okaa-san n i s i k a i t t e moraitaku nai . *(I) don't want (you) to t e l l that kind of story to anyone but only to my mother.' c. * Sonna hanasi wa okaa-san n i s i k a i u . (22) a. Boozu wa yasai s i k a tabenai to kangaerareru. 'It i s thought that monks eat nothing but vegetables.' b. Boozu wa yasai s i k a taberu to kangaerarenai. 'I t i s not thought that monks eat anything but vegetables.' c. * Boozu wa yasai sika taberu. (23) a. Kotosi Yamada-san wa Amerika e t o o t e i ikenai to omou. *(I) think Mr. Yamada cara't possibly go to America t h i s year.' b. Kotosi Yamada-san wa Amerika e t o o t e i ikeru 68 to wa omowanai. '(I) don't think Mr. Yamada can possibly go to America t h i s year.' c. * Kotosi Yamada-san wa Amerika e t o o t e i ikeru. (24) a. Sensei n i mo kono kanzi ga zenzen yomenai to watakusi wa kangaeru. 'I think even the teacher can't read t h i s Chinese character at a l l * ' b. Sensei n i mo kono kanzi ga zenzen yomeru to watakusi wa kangaenai. •I don't think even the teacher can read t h i s Chinese character at a l l . * e. * Sensei n i mo kono kanzi ga zenzen yomeru. (25) a. Kozutumi wa a s i t a made Oosaka n i tukanai to omou. '(I) think the parcel w i l l not arrive i n Osaka u n t i l tomorrow.' b. Kozutumi wa a s i t a made Oosaka n i tuku to wa omowanai. '(I) don't think the parcel w i l l a r r i v e i n Osaka u n t i l tomorrow.' c. * Kozutumi wa a s i t a made Oosaka n i tuku. The (a) and (b) sentence pairs above are a l l grammatical and are semantically equivalent. Notice that the (b) sentences are related to the (a ) sentences i n exactly the 69 same way as (5b) i s related to ( 5 a ) . The above phenomenon can be explained i f we make the following assumptions! ( i ) Just as the English u n t i l adverbial co-occurring with a semantically punctual verb requires a negative i n the same simplex sentence, c e r t a i n p a r t i c l e s and adverbials i n Japanese, such as sjlka, kessite, tittomo. t o o t e i , zenzen. and made. also require the presence of an overt negative within the same simplex sentence. Hence the ungrammaticality of the (c) sentences which i s a d i r e c t r e s u l t of the v i o l a t i o n of t h i s simplex sentence condition. ( i i ) However, the (b) examples show that the p a r t i c l e s and adverbials can occur without the negative i n the embedded sentence, i f they are commanded by a c e r t a i n class of negated verbs or adjectives, such as omowanai 'don't think', kangaenai 'don't think* or don't consider', -te moraitaku n a i *don?t want to have someone do', which are considered to be sensitive to the negative transportation r u l e . Semantically and sy n t a c t i -c a l l y , these verbs are considered to be s i m i l a r to the class of verbs i n English within which the negative transportation rule i s applicable. ( i i i ) The fact that there i s no overt negative i n the 70 embedded sentences of the (b) examples, seems to suggest that the negative must have originated i n the embedded sentences i n the underlying structures, and i s then transported-to the higher sentences. At t h i s stage, i t seems reasonable to assume that the (b) sentences are derived from the same underlying structures of t h e i r corresponding (a) sentences, and that there i s no way to account f or the derivation of the (b) sentences without resorting to the negative transportation r u l e . The above argument seems to suggest that the negative transportation rule i s relevant and that we do need such a rule i n Japanese. I f t h i s i s so, then we have to assume that the rule i s applicable to Japanese too. At t h i s stage, l e t us examine more examples to see what class of verbs are sensitive to the negative transport-ation r u l e , and whether t h i s class of verbs are the same as those f o r English. (26) a. Hikooki wa a s i t a no asa made tuku to wa omowanakatta. '(I) didn't think the plane would arrive u n t i l tomorrow morning.' b. Hikooki wa a s i t a no asa made tuku to wa 71 kangaerarenakatta. •I t was not considered that the plane would arrive u n t i l tomorrow morning.* c. * Hikooki wa a s i t a no asa made tuku to wa iwanakatta. - *(I) didn*t say that the plane w i l l a r r i v e u n t i l tomorrow morning.* d. * Hikooki wa a s i t a no asa made tuku to wa k i i t e inakatta. *(I) didn't hear that the plane would arrive u n t i l tomorrow morning.' e. ? Hikooki wa a s i t a no asa made tuku to wa sinzinakatta. •(I) didn't "believe that the plane would arrive u n t i l tomorrow morning.* f. * Hikooki wa a s i t a no asa made tuku to wa k i t a i sinakatta. *(I) didn't expect that the plane would arrive u n t i l tomorrow morning.' g. * Hikooki wa a s i t a no asa made tuku to wa kanzinakatta. *(I) didn't f e e l that the plane would arrive u n t i l tomorrow morning.' From the above examples, i t seems that omou 'to think' and kangaeru 'to think or to consider* are sensitive to the 72 negative transportation r u l e , while i u 'to say', kiku 'to hear*, k i t a i suru 'to expect', and kanziru 'to f e e l * are not. We have seen e a r l i e r , that a verb such as expect i s considered to be a negative transport verb for English. However, t h i s f act does not seem to hold for Japanese. The corresponding Japanese verb k i t a i suru does not seem to be sensitive to the negative transportation r u l e , and does not allow the negative to be moved out of the embedded sentence to the higher sentence. I have checked with several native speakers of Japanese and they a l l seem to agree on that. Concerning (e) with the verb s i n z i r u 'to believe', native speakers of Japanese do not seem to agree. Some consider (e) as grammatical, while others do not. From t h i s observ-ation, i t appears that the semantic class of verbs within which the negative transportation rule i s applicable i s the same f o r both Japanese and English (that i s , the class of verbs belonging to the mental s t a t e ) , but the set of verbs within t h i s class that i s subject to the rule varies from language to language. While f or example, expect i n English i s subject to the r u l e , the corresponding Japanese verb i s not. 2 . 3 . 2 . "CONFIRMATORY" QUESTION ARGUMENT Another argument which can be c i t e d f o r Japanese and which i s rather s i m i l a r to Robin Lakoff*s Tag Question 73 argument, i s perhaps the Japanese "Confirmatory* Question Argument, as suggested by Soga (1972). The Japanese "confirmatory* question formation i s one i n which the speaker thinks that something i s true, and he seeks assurance or confirmation f o r what he thinks. The following are some examples of the confirmatory questions i n Japanese. (27) Sonna syuukan wa Amerika n i mo aru to, omoimasu ga, arimasen ka. L i t . *(I) think that kind of custom ex i s t s i n America too, but does ( i t ) not exis t (there)?* *(I) think that kind of custom exists i n America too, doesn't i t ? * (28) John wa o-susi o taberu to omoimasu ga, tabemasen  ka. L i t . '(I) think John eats s u s i , but does (he) not eat ( i t ) ? * '(I) think John eats s u s i , doesn't he?' (29) Raigetu Taroo wa Amerika e ikanai to omou keredo ikimasu ka. L i t . '(I) think Taroo i s n ' t going to America next month, but i s (he) going?* *(I) think Taroo i s n * t going to America next month, i s he?* (30) Taroo wa Hanako to kekkon s i n a i to omou keredo kekkon simasu ka. 74 L i t . '(I) think Taroo i s not marrying Hanako, but i s (he) marrying (her)?' '(I) think Taroo i s not marrying Hanako, i s he?' The underlined portion i n the above sentences are the "confirmatory" questions. Notice that these confirmatory questions behave s i m i l a r l y to the English tag questions. Just l i k e the English tag questions, i f a statement i s po s i t i v e , the confirmatory question i s formed by a f f i x i n g to i t the corresponding negative sentence, and vice versa. Therefore, with respect to t h e i r negative-affirmative forms, the behaviour of these confirmatory questions i s s i m i l a r to that of the English tag questions. As the tag question formation provides a strong argument i n defense of the negative transportation rule for English, i t seems approp-r i a t e that we look into some examples of the confirmatory questions i n order to evaluate the a p p l i c a b i l i t y of the negative transportation rule i n Japanese. Let us consider the following examples: (31) a. Taroo wa Amerika e ikanai to watakusi wa omou keredo ikimasu ka. L i t . *I think Taroo i s n ' t going to America, but i s J(;he) going?' * I"think Taroo i s n ' t going to America, i s he?' b. Taroo wa Amerika e iku to watakusi wa omowanai keredo ikimasu ka. 75 L i t . 'I don't think Taroo i s going to America, but i s (he) going?' 'I don't think Taroo i s going to America, i s he?' c. * Taroo wa Amerika e iku to watakusi wa omowanai keredo ikimasen ka. L i t . 'I don't think Taroo i s going to America, but i s n ' t (he) going?' 'I don't think Taroo i s going to America, i s n ' t he?' d. * Taroo wa Amerika e iku to watakusi wa omowanai keredo soo omoimasu ka. L i t . *I don't think Taroo i s going to America, but do (I) think so?' 'I don't think Taroo i s going to America, do I ? ' (32) a. Taroo wa Amerika e iku to Hanako ga omotte i n a i to omou keredo Hanako wa soo omotte imasu ka. L i t . '(I) think Hanako doesn't think that Taroo i s going to America, but does Hanako think so?' •(I) think Hanako doesn't think that Taroo i s going to America, does she?' b. ? Taroo wa Amerika e iku to Hanako ga omotte i n a i keredo Hanako wa soo omotte imasu ka. L i t . 'Hanako doesn't think that Taroo i s going to America, but does (*she) think so?' 'Hanako doesn't think that Taroo i s going to 76 America, does she?* c. Taroo wa Amerika e iku to Hanako ga omotte i r u to omowanai keredo Hanako wa soo omotte imasu ka. L i t . *(I) don't think Hanako thinks that Taroo i s going to America, but does Hanako think so?' '(I) don't think Hanako thinks that Taroo i s going to America, does she?' In sentence (31a) , the confirmatory question i s affirm-ative and the sentence i t i s formed on i s i n the negative. (31c) i s ungrammatical as w e l l as (31d) . The ungrammaticality of (31d) can e a s i l y be accounted f o r . Omou 'to think' i s a verb of mental state and therefore, i t i s impossible f o r the subject to question whether what he thinks i s true of him. Also omou i s used as a performative verb describing an action that i s carr i e d out i n the act of description, and hence i t would be i l l o g i c a l to question i t . (31b) ! i s a perfectly grammatical sentence, and yet the confirmatory question i s p o s i t i v e , as i s the sentence i t i s formed on. The grammati-c a l i t y and acceptability of a sentence l i k e (31b) rather than (31c) seems strange at f i r s t glance. However, the grammaticality of (31b) can be accounted f or i n exactly the same way as the grammaticality of sentence (14) wasiaccounted f o r . Notice also that (31b) i s s i m i l a r to (14). (31b) i s derived by f i r s t applying confirmatory question formation 77 on the s t r i n g Taroo wa Amerika e ikanai to watakusi wa omou 'I think Taroo i s not going to America*, to give the intermediate s t r i n g Taroo wa Amerika e ikanai to watakusi  wa omou, keredo ikimasu ka 'I think Taroo i s not going to America, i s he?' Then negative transportation applies moving the negative out of the embedded sentence commanded by omou, which i s s i m i l a r to the English performative verb suppose, and the s u p e r f i c i a l form of (31b) i s produced. The ungrammaticality of (31c) can now be accounted f o r . The embedded sentence to which the tag question i s a f f i x e d o r i g i n a l l y contains the overt negative nai at the time when confirmatory question formation applies. Hence a f f i x i n g a negative confirmatory question to a negative statement vi o l a t e s the confirmatory question condition, r e s u l t i n g i n an ungrammatical sentence. Sentence (32a) i s s i m i l a r i n structure to (15)• (32a) i s derived by f i r s t applying the negative transportation to the s t r i n g Taroo wa Amerika e ikanai to Hanako ga omotte i r u  to omou 'I think Hanako thinks that Taroo i s not going to America*, to give the intermediate s t r i n g Taroo wa Amerika  e iku to Hanako ga omotte i n a i to omou 'I think Hanako doesn't think that Taroo i s going to America'. Then confirm-atory question applies to produce sentence ( 3 2 a ) . Notice also that i t i s possible to apply a further negative transportation rule to (32a) with respect to the performative verb omou 'to think* of another higher sentence i n order to derive sentence 78 ( 3 2 c ) . The grammaticality of (32b) i s questionable, although as f a r as i t s structure i s concerned, i t should correspond to ( 15) . As pointed out by Soga (1972), t h i s seems to suggest that assuming the existence of an underlying abstract performative verb l i k e omou f o r Japanese i s probably u n l i k e l y . The above examples seem to show that the order of application of negative transportation rule and the confirm-atory question formation i n Japanese i s exactly the same as that f o r negative transportation rule and tag question formation i n English. I t also shows that the behaviour and structure of the Japanese confirmatory question i s s i m i l a r to the English tag question. Just as the tag question formation provides f a i r l y conclusive evidence f or the existence of the negative transportation rule i n English, the above argument and examples given so f a r seem to give support to the a p p l i c a b i l i t y of the negative transportation rul e i n Japanese. 2 . 3 . 3 - COUNTEREXAMPLES TO ABOVE TWO ARGUMENTS The two arguments c i t e d above, that i s , the simplex sentence condition f or the negative p o l a r i t y adverbials and p a r t i c l e s , and the confirmatory question; arguments, seem to provide f a i r l y relevant evidence for the existence of the negative transportation rule i n Japanese. I t i s 79 true that the examples given so f a r appear to uphold the assumption that t h i s rule i s also applicable to Japanese, just as i t i s to English. However, there are counterexamples •.—- some of which:can'be f a i r l y e a s i l y explained but some very challenging which does weaken and perhaps d i s c r e d i t our assumption that the negative transportation rule i s applicable to Japanese. F i r s t l y , the simplex sentence condition f o r the negative p o l a r i t y adverbials and p a r t i c l e s , as c i t e d i n section 2 . 3 . 1 . needs further investigation. We have shown that these negative p o l a r i t y words are constrained i n such a way that they and the negative must command each other at some l e v e l of the derivation. In f a c t , i t i s only when t h i s mutual command re l a t i o n s h i p does not hold that sentences l i k e the (c) sentences of (19)-(25) are ungrammatical. Nevertheless, there are some examples i n which these negative p o l a r i t y words can occur i n an embedded sentence without,an overt negative, and yet the embedded sentence i s commanded by a negated verb which does not belong to the omou class of verbs considered to be sensitive to the negative transpor-t a t i o n r u l e . Thus consider the following sentencest (33) a. Fuzi-san (ni) s i k a noboranakatta koto ga aru, '(I) have the experience of climbing up only Mt. F u j i . ' b, Fuzi-san (ni) si k a nobotta koto ga n a i . 80 '(I) have no other experience hut that of climbing up Mt. F u j i . ' c. * Fuzi-san (ni) sika nobotta. (3*0 a. Okane ga nakute, pan sika tabenakatta keiken ga aru. '(I) have an experience that I ate nothing but bread only, because (I) had no money.' b. Okane ga nakute, pan sika tabeta keiken ga na i . '(I) don't have (any other) experience but that of eating bread only, because (I) had no money.* c. * Okane ga nakute, pan sika tabeta. (35) a. Yamada-san wa o-sake o kessite nomanai koto ga aru. 'There are times when Mr. Yamada never drinks sake.' b. Yamada-san wa o-sake o kessite nomu koto ga na i . 'There are never times when Mr. Yamada drinks sake.' c. * Yamada-san wa o-sake o kessite nomu. d. Yamada-san wa o-sake o nomu koto ga kessite n a i . 'There are never times when Mr. Yamada drinks sake.' 81 (36) a. Taroo wa t i i s a i t o k i kara ima made zenzen byooki o sinakatta koto ga aru. 'Since (he) was small u n t i l now, there are times when Taroo was never i l l . ' h. Taroo wa t i i s a i t o k i kara ima made zenzen byooki o s i t a koto ga n a i . •Since (he) was small u n t i l now, there are never times when Taroo was i l l . • c. * Taroo wa t i i s a i t o k i kara ima made zenzen byooki o s i t a . d. Taroo wa t i i s a i t o k i kara ima made byooki o s i t a koto ga zenzen n a i . 'Since (he) was small u n t i l now, there are never times when Taroo was i l l . * (37) a. ? Tanaka-san ga tootei kekkon-dekinai to yume n i mo omou. '(I) even dream that Mr. Tanaka w i l l not possibly marry.' b. Tanaka-san ga t o o t e i kekkon-dekiru to wa yume n i mo omowanai. '(I) don't even dream that Mr. Tanaka w i l l possibly marry.' c. * Tanaka-san ga too t e i kekkon-dekiru. d. Tanaka-san ga kekkon-dekiru to wa t o o t e i yume n i mo omowanai. '(I) don't even possibly dream that Mr. Tanaka 82 w i l l marry. 1 (38) a. ? Byooki s i t a o z i i - s a n ga zenzen naoranai daroo to k i t a i - s i t e i t a . '(I) was expecting that my sick grandfather would not get w e l l at a l l . ' b. Byooki s i t a o z i i - s a n ga zenzen naoru daroo to wa k i t a i - s i t e inakatta. *(I) was not expecting that my sick grandfather would get well at a l l . ' c. * Byooki s i t a o z i i - s a n ga zenzen naoru daroo. d. Byooki s i t a o z i i - s a n ga naoru daroo to wa zenzen k i t a i - s i t e inakatta. '(I) was not expecting at a l l that my sick grandfather would get w e l l . ' Notice that i n the (b) examples above, the p a r t i c l e s i k a and the adverbials such as kessite, zenzen, and t o o t e i occur i n the embedded sentences without any overt negative, and the verbs commanding them, although having the negative, do not belong to the class of verbs considered to be sensi-t i v e to the negative transportation r u l e . The (b) sentences v i o l a t e the simplex sentence condition, and yet they are perfectly grammatical sentences. I t cannot be considered that the negative transportation applies here opti o n a l l y , for the verb aru 'to exi s t * i s not a negative transport verb. Another piece of evidence which renders the optional 83 application of the rule u n l i k e l y i s the fact that the (a) and (b) sentences have d i f f e r e n t meanings. The semantic difference between them i s quite d i f f e r e n t from that between the (a) and (b) sentences of ( 1 9 ) - ( 2 5 ) . Also, as Soga (1972) has pointed out, the ac c e p t a b i l i t y of a sentence l i k e (38a) i s questionable. This i s probably due to the nature of the verb k i t a i - s u r u 'to expect*. I t seems that t h i s verb has a feature i n d i c a t i n g that what the speaker expects has to be something good. Therefore, (38a) can be semantically acceptable only i f the speaker hates his grandfather and wishes him to die. On the other hand, (b) and (d) of (38) do not require such an interpretation. Therefore, based on t h i s observation, i t seems quite clear that we couldn't possibly derive (b) and (d) from (a) by simply applying the negative transportation r u l e . Notice too that the adverbials kessite of (35d) , zenzen of (36d) , t o o t e i of (37d) , and zenzen of (38d) , a c t u a l l y modify the verb of the matrix sentence. In ( 3 5 ) - ( 3 8 ) , the adverbials i n the (b) sentences, although they are considered to e x i s t within the embedded sentence on the surface, ac t u a l l y modify the verb of the main sentence, jus t l i k e the adverbials i n the (d) sentences do. To most native speakers of Japanese, the (d) sentences seem to be more natural than the (b) sent-ences, although the (b) sentences are also acceptable. The above counterexamples show quite c l e a r l y that we cannot resort to the negative transportation rule to explain 84 the grammaticality of the (b) sentences. I f t h i s i s the case, how then can we j u s t i f y the grammaticality of the (b) sentences? A solution to t h i s problem would be to consider these adverbials and p a r t i c l e s (such as kessite, tittomo, zenzen, and sika) and the negative as forming a unit predi-cate introduced as verb phrases of the "higher" sentences i n the underlying structure, a treatment suggested by Soga ( 1 9 7 2 ) . A s i m i l a r treatment i s also employed i n McGloin (1972). Sentence (33b) would then be considered to have an underlying structure s i m i l a r t o i Fuzi-san nobotta koto In (39), s i k a i s introduced together with the negative as a VP i n the higher sentence. By a transformational r u l e , ^ s i k a i s lowered into S^ and attached to the lower NP koto to produce the intermediate s t r i n g 1 Fuzi-san n i nobotta koto 85 s i k a nai 1 (I) have only the experience of climbing up Mt. F u j i ' . Then sika i s further lowered into S 2 and attached to the NP Fuzi-san to produce sentence (33b) . Sentence (35b) , with the adverbial kessite , can also be accounted for i n the same manner by introducing kessite together with the negative i n the "higher" sentence and then attaching kessite to appropriate verbs or adjectives i n the lower sentence. For instance, (35b) would have the l o g i c a l structure of: (40) NP VP kessite NEG N aru Yamada-san o-sake nomu koto Kessite i n (40) w i l l be attached to the verb i t immediately commands, that i s aru, to derive the intermediate s t r i n g : Yamada-san wa o-sake o nomu koto ga kessite n a i * There are never times when Mr. Yamada drinks sake*. Notice that t h i s 86 i s s i m i l a r to (35d). To derive (35b), kessite i s further lowered and attached to the verb nomu,. Sentences (3^b), (36b), (37b) and (38b) can a l l be accounted f o r i n a s i m i l a r manner. The above treatment enables us, to a cer t a i n extent, to account for the gramma-t i c a l i t y of the (b) sentences without resorting to the negative transportation r u l e . Now, l e t us look at a di f f e r e n t example where the sika i s attached to the subject of the sentence instead of the dire c t object. Consider the following sentences* (41) a. John sik a sukiyaki o tabenakatta koto ga aru. •John i s the only one who has the experience of eating sukiyaki.' b. John sik a sukiyaki o tabeta koto ga n a i . •John i s the only one who has no other experience but that of eating sukiyaki.* Just like (33)-(38), (41b) could not have been derived from (41a), by applying the negative transportation rule which moves the negative out of the embedded sentence to the matrix sentence. (4lb) is considered to have the following underly-ing structure! 87 x sukiyaki koto tabeta To derive (41b), s i k a i s f i r s t lowered into and attached to the NP John to produce the intermediate s t r i n g : Sukiyaki o tabeta koto ga aru no wa John s i k a de nai 'The one who has the experience of eating sukiyaki i s John only.* The next g step i s to apply a transformation rule which preposes John sik a to derive (41b). So f a r , Soga's treatment seems almost successful i n accounting f o r the grammaticality of sentences such as the (b) sentences of (33)-(38). and (41). However, i t should be pointed out that Soga's treatment also leaves a serious 88 problem. His treatment seems to f a i l when i t comes to explaining sentences co-occurring with phrases other than koto ga nai or keiken ga n a i . Observe the following sentences: ( 4 3 ) a. Otoo-san ga nonda o4>tya s i k a n a i . 'There i s no other (kind of) tea except that which father drinks.' b. * Otoo-san sik a nonda o-tya ga n a i . 'The tea that was drunk, i s none other than father.' c. Otoo-san s i k a nomanakatta o-tya ga aru. 'There i s the tea that nobody other than father drinks.' ( 4 3 a ) would have the l o g i c a l structure of ( 4 4 ) . 89 Following Soga's treatment, si k a can be lowered into S^ and attached to the NP o-tya to produce sentence (43a). Up to t h i s stage the treatment seems to work. Accordingly, si k a can be lowered further and attached to the NP of S 2, as i n (39) and (40). However, notice that i f sik a i n (44) i s lowered further and attached to the NP of S 2, i t w i l l produce the ungrammatical sentence (43b). Thus, Soga's treatment has t h i s defect, and ends up deriving an ungramm-a t i c a l sentence. The above observation seems to suggest that perhaps sika lowering can be applicable only to the highest NP. I f there i s a second lowering of s i k a , then the meaning of the sentence w i l l be t o t a l l y d i f f e r e n t , as can be noticed i n the semantic gap between (43a) and (43b). On the basis of the above observation, we may perhaps posit the following: ( i ) The attachment of sika to noun phrases of lower sentences i s perhaps r e s t r i c t e d . ( i i ) Soga's examples with koto ga nai and keiken ga nai should be treated as exceptional cases. The second sik a lowering should perhaps be r e s t r i c t e d to only koto ga nai and keiken ga nai . ( i i i ) Soga's generalization i s perhaps over-simplified :„ and cannot apply with respect to the above (42). 90 At t h i s stage I do believe the solution l i e s i n r e s t r i c t i n g the sika lowering to the highest NP. However, why that i s so i s unclear at present. For the other adverbs such as kessite, t o o t e i and zenzen, the same may perhaps be the case} the attachment may perhaps be applied only to the highest NP. However, at present i t i s s t i l l unclear. This problem w i l l , therefore, be l e f t open for future research. The counterexamples c i t e d above show that s i k a , k e s site, to o t e i and zenzen, which must co-occur with the negative i n a simplex sentence, can s t i l l occur i n an affirmative embedded sentence, so long as the verb of the main sentence contains a negative, even though the negated verb may not be a negative transport verb. This evidence renders the application of the negative transportation rule to Japanese u n l i k e l y and greatly weakens the evidence c i t e d i n support for the possible existence of the negative transportation rule i n Japanese. I t has been pointed out by Soga (1972) that a sentence li k e s (4 - 5 ) a. I didn't believe she was not stupid but I didn't believe she was stupid e i t h e r , was she? b. * I didn't believe she was not stupid but I believed she was not stupid either, was she? presents a strong counterexample to Lakoff's proposal f o r 91 the tag question argument. Notice that the tag question that i s formed on the second h a l f of sentence (45a) i s po s i t i v e , as i s the embedded sentence i t i s formed on. The ungrammaticality of (45b) leads us to believe that the negative of didn't believe i n the second half of the sentence could not possibly have originated i n the embedded sentence, and t h i s renders the application of the negative transporta-t i o n rule u n l i k e l y . Similar examples also e x i s t i n Japanese. Consider the following sentences* (46) a. Tanaka-san wa bimboo da to kessite omowanakatta keredo kanemoti da to mo yume n i mo omowanakatta. '(I) never thought that Mr. Tanaka was poor, c but (I) didn't even dream that (he) was r i c h e i t h e r . ' b. ? Tanaka-san wa bimboo da to kessite omowa-nakatta keredo kanemoti de wa nai to yume n i mo omotta. •(I) never thought that Mr. Tanaka was poor, but (I) even dreamt that (he) was not r i c h either.* c. Tanaka-san wa bimboo da to kessite omowanaka-t t a keredo kanemoti da to yume n i mo omowana-katta ga, Tanaka-san wa kanemoti desita ka. '(I) never thought that Mr. Tanaka was poor, 92 but (I) didn't even dream that he was rich, but was Mr. Tanaka rich?' The negative in the second half of (46a) cannot be considered to have derived from the embedded sentence by the application of the negative transportation rule. The evidence that renders the application of the negative transportation rule unlikely i s the fact that (46a) and (46b) have different meanings, where in (46b) the negative has been moved into the embedded sentence. The acceptability of (46b) is even doubtful. However, notice that in (46c), the confirmatory question i s positive in spite of the fact that the sentence i t i s formed on could not have contained the negative. How do we account for this strange phenomenon? Soga (1972) states that for such examples, the negative must be considered to be derived from the higherrsentence, and proposed that their grammaticality be explained on the Q basis of inferential co-occurrence. Consider the following! 7 (47) S > (-S) / -V where V commands S According to Soga, there exists a group of omou-like or suppose-like verbs which inferentially works with a f i r s t person subject according to rule (47) above. A sentence commanded by one of these negated verbs (the verb may be in 93 the affirmative i f i t is semantically negative, such as doubt'} receives a mild negative interpretation by inference. On the basis of (4-7), we can then account for the grammaticality of (45a) and (46c). The embedded sentences of the second half of (45a) and (46c) are interpreted as mild negative statements by inference, and hence allow a positive tag question and a positive confirmatory question to be formed on them. Sentences such ass (48) Otoo-san ga asita made kaette kuru no wa utagawasii keredo kaette kimasu ka. * It i s doubtful that father w i l l come home un t i l tomorrow, but w i l l (he)?* (49) I doubt i f they w i l l even l i f t a finger to help, w i l l they? can be easily explained in a similar way. Sentences (48) and (49) contain the verbs utagawasii and doubt, which, although are in the affirmative, are semantically negative. Therefore, the embedded sentences commanded by these verbs inferentially receiver- a negative interpretation, which in turn co-occur with the made adverbial and the l i f t a finger phrase, which normally would require that a negative be present in the same sentence in deep structure. Notice that this also accounts for the formation of the positive 94 confirmatory question and the positive tag question. The above observation seems to he c r u c i a l . On the basis of such counterexamples, i t seems doubtful that we can depend on the confirmatory question as evidence f or the existence and the a p p l i c a b i l i t y of the negative transportation rule i n Japanese. McGloin i n (1972), states that the optional ru l e of negative transportation i s required f o r nominalizers l i k e hazu 'expect' and tumori 'intend'. Observe the following sentences;* (50) a. Tanaka-san wa a s i t a made tukanai hazu da. 'It i s the expectation that Mr. Tanaka w i l l not arrive u n t i l tomorrow.' b. Tanaka-san wa a s i t a made tuku hazu de wa n a i . 'It i s not the expectation that Mr. Tanaka w i l l arrive u n t i l tomorrow.' c. Tanaka-san wa a s i t a made tuku hazu ga n a i . 'I t i s not the expectation that Mr. Tanaka w i l l arrive u n t i l tomorrow.' d. * Tanaka-san wa a s i t a made tukanai hazu ga aru. (51) a. Yamada-san wa n i t i y o o b i made konai tumori da. 'I t i s the intention that Mr. Yamada w i l l not come u n t i l Sunday.' b. Yamada-san wa n i t i y o o b i made kuru tumori de wa 95 n a i . • I t i s not the intention that Mr. Yamada w i l l come u n t i l Sunday.' c. Yamada-san wa n i t i y o o b i made kuru tumori ga nai . ' I t i s not the intention that Mr. Yamada w i l l come u n t i l Sunday.' d. * Yamada-san wa a s i t a made konai tumori ga aru. According to McGloin, the sentences l i k e (a) and (b) above should be synonymous, and that the (b) sentences are derived from the (a) sentences by the optional application of the negative transportation r u l e . However, i f we were to examine the sentences more cl o s e l y , we w i l l notice that the (a) and (b) sentences of (50) and (51) are not synonymous i n the l e a s t . In ( 5 0 a ) , "the speaker expects Mr. Tanaka not to a r r i v e , and the expectation of Mr. Tanaka's not a r r i v i n g i s being affirmed, while i n (50b) the speaker does not expect Mr. Tanaka to a r r i v e , and here the expectation of Mr. Tanaka's a r r i v a l i s being denied. The same i s true of (51a) where the intention of Mr. Yamada's not coming i s being affirmed, while i n (5H>) the intention of h i s coming i s being denied. On the basis of t h i s semantic difference, i t cannot be said that negative transportation applies here opt i o n a l l y . Another piece of evidence which renders the optional 96 applic a t i o n of the negative transportation rule to hazu and tumori u n l i k e l y i s the fact that ungrammatical sentences l i k e (d) would have to underlie the (c) sentences. I f t h i s rule does apply optionally to hazu and tumori, as McGloin has stated, then the (c) sentences should actually be derived from the underlying structures of (d) which are ungrammatical sentences themselves. Notice that the negative that i s attached to the verb of the matrix sentence i n (50c) hazu  ga nai ' I t i s not the expectation*, should ac t u a l l y be derived from the embedded sentence of (50d) Tanaka-san ga a s i t a made  tukanai 'Mr. Tanaka w i l l not arrive u n t i l tomorrow', by the application of the negative transportation r u l e . The same would be true of ( 5 1 c ) , where the negative that i s attached to the verb of the matrix sentence tumori ga nai ' i t i s not the intention', would be considered to have been derived from the embedded sentence of (51d) Yamada-san ga n i t i y o o b i  made konai 'Mr. Yamada w i l l not come u n t i l Sunday', by the applic a t i o n of the negative transportation r u l e . However, notice also that (50d) and (51d) are ungrammatical and therefore, could not possibly form the underlying structures fo r (50c) and (51c) . On the basis of the above observations, I must conclude that positing an optional application of the negative transportation rule for nominalizers l i k e hazu and tumori, as McGloin has done, i s an i n s u f f i c i e n t generalization. I t seems that the only possible way to account f o r the grammaticality of (50b) , (51b) , (50c) and (51c) , would be 97 on the basis of i n f e r e n t i a l co-occurrence. Following (4-7), i t seems possible to interpret i n f e r e n t i a l l y the embedded sentence of (50b) Tanaka-san ga tuku hazu fle wanai'It i s not expected that Mr. Tanaka w i l l a r r i v e * as something l i k e Tanaka-san ga tukanai 'Mr. Tanaka won*t a r r i v e ' which w i l l then co-exist with the a s i t a made ' u n t i l tomorrow* phrase. (51b) can be accounted for i n the same way. The embedded sentence there i n f e r e n t i a l l y receives a negative inter p r e t a t -ion, which i n turn co-occurs with the n i t i y o o b i made ' u n t i l Sunday' phrase. The same explanation can also be used to account f o r sentences (50c) and ( 5 1 c ) . 2 . 4 ..'. . CONCLUSION The simplex sentence condition f o r the negative p o l a r i t y adverbials and p a r t i c l e s such as kessite , t o o t e i , zenzen, s i k a and made, and the confirmatory question formation have been presented as arguments f o r the support of the negative transportation rule i n Japanese. However, I have also c i t e d several counterexamples which c a l l the whole argument into question. The condition which stated that kessite, t o o t e i , zenzen, s i k a and made must co-occur with the negative within a simplex sentence cannot always be maintained with regards to the embedded sentence. I t seems that they can s t i l l occur i n affirmative sentences, so long as the main verbs are 9 8 negated, even though the main verbs may not be negative transport verbs. With regards to the confirmatory question formation, the condition which states that i f the statement i s p o s i t i v e , a negative confirmatory question should be a f f i x e d to i t and vice versa, i s not always maintained„either. The counterexamples presented seem to v i o l a t e t h i s condition and yet they are grammatical sentences. On the basis of these counterexamples, i t seems that we cannot r e l y on these two arguments f o r the support of the negative transportation rule i n Japanese. The counterexamples c i t e d greatly weaken the j u s t i f i c a t i o n s f o r the existence of the negative transp-ortation rule i n Japanese. A possible solution proposed for sentences with s i k a , kessite, t o o t e i , zenzen and made, which cannot be explained by the application of the negative transportation r u l e , would be to consider them and the negative as forming a unit predicate introduced i n the "higher" sentence, and then lowered appropriately. However, i t also has been pointed out that t h i s proposal has i t s defects. For the solution, i t i s proposed that a further r e s t r i c t i o n i s necessary f o r the lowering @f s i k a . For the others such as kessite, t o o t e i , and zenzen, i t i s not clear at t h i s point. With regards to the sentences with confirmatory questions and made adverbials where we cannot resort to the negative transportation r u l e , i t i s proposed that t h e i r respective grammatical forms be explained on" the basis of i n f e r e n t i a l co-occurrence. 99 FOOTNOTES 1. This rule of negative transportation has been discussed under a variety of terms. For example, i t i s termed ; negative transportation i n Fillmore ( I 9 6 3 ) , R. Lakoff (1969) and M. Soga ( 1 9 7 2 ) , as negative absorption i n Klima ( 1 9 6 4 ) . as not-transportation i n G. Lakoff ( 1 9 7 0 a ) , and as negative  r a i s i n g i n J . Lindholm (1969) and L. Horn ( 1 9 7 1 ) . 2 . Notice that the negative transport verbs belong to several semantic classes. However, we might expect these verbs to share c e r t a i n properties so that generalizations would be possible, which would apply to a l l such verbs.. One such generalization, as observed by Paul and Carol Kiparsky (from Lakoff 1 9 7 0 a ) , i s that negative transportation never occurs with f a c t i v e verbs. Lakoff (1970a) has related t h i s to D. Bolinger's observation. According to Bolinger, negative transported sentences l i k e (lb) and ( 2 b ) , seem to convey greater uncertainty i n the speaker's mind than t h e i r non-transported counterparts of (la) and ( 2 a ) . Lakoff points out that since i t i s presupposed that the subject of a fa c t i v e verb knows that the complement of the verb i s true, he cannot be uncertain about i t . I f negative transportation conveys uncertainty, then f o r s t r i c t l y semantic reasons, the rule cannot apply with factive verbs. 3 . The performative verbs are those verbs which must be used with the f i r s t person subject and usually have second person direct or i n d i r e c t objects i n the deep structure. They must be affirmative and non-negative, they must be used i n the present tense and non-repetitively. These performative verbs belong to a large class of true verbs which includes those such as ask, beg, command, order, propose, demand, r e q u e s t s a y , require, inform, i n s t r u c t , beseech, advise, claim, o f f e r , enquire, sentence, warn, grant, enquire, and many more. r The main verb of the following examples (a).-(c) are a l l [+ performative] . (a) I order you to leave. (b) 1 promise you that I w i l l return. (°) 1 advise you to see a doctor. For a detailed discussion on performative verbs and t h e i r properties, see J.L. Austin ( 1 9 6 2 ) . See also J.R. Ross on "On Declarative Sentences" from Jacobs and Rosenbaum ( 1 9 7 0 ) . 4. As i t has been pointed out i n section 1 . 3 . 2 . 3 . , with made adverbials, the verb occurring i n the sentence must be 100 negated i f i t i s a semantically punctual verb. 5. This treatment would be somewhat s i m i l a r to that proposed e a r l i e r i n section 1.3.2.1., where adverbials such as the l o c a t i v e , time and instrumental adverbials are considered to be derived from verb phrases of "higher" simplex sentences than the ones that appear as the main clauses i n the surface structures. 6. The transformational rule which attaches sik a to a lower NP i s somewhat s i m i l a r to the wa attachment rule proposed by Shige-Yuki Kuroda. 7. I am indebted to M. Soga f o r t h i s observation. 8. This noun phrase preposing transformation i s an independently motivated rule which i s s i m i l a r to that which preposes the noun phrases of John and Ford i n (ia) and ( i i a ) to produce the surface structures of (ib) and ( i i b ) . ( i ) a. Wakaru no wa John desu. '(The one) who understands i s John.* b. John ga wakaru. •John understands.' ( i i ) a. Daitooryoo wa Ford desu. 'The President i s Ford.' b. Ford ga daitooryoo desu. 'Ford i s the President.' 9. Soga, Matsuo. 1972. "Negative Transportation and Cross-Linguistic Negative Evidence". Papers i n Japanese  L i n g u i s t i c s . University of C a l i f o r n i a , Berkeley, (pp. 116). 1 0 1 CHAPTER THREE  JAPANESE NEGATIVE QUESTIONS AND YES~NO RESPONSE 3 . 1 . INTRODUCTION This chapter w i l l look into the problems and p e c u l i a r i -t i e s involved with the Japanese negative questions and the yes-no responses that they e l i c i t . I w i l l attempt to give l o g i c a l explanations f o r the ambiguity of negative sentences i n Japanese, and w i l l attempt to explain both the semantic and syntactic c h a r a c t e r i s t i c s of such questions. Further-more, I w i l l examine various examples of Japanese negative questions and t h e i r answers represented by hai 'yes' and i i e 'no*, and w i l l attempt to f i n d out a simple and l o g i c a l way of explaining the syntactic and semantic r e l a t i o n s h i p between the questions and t h e i r answers. In recent years, several analyses have been proposed (Shibatanii 1972, Kunot 1973. Sogas 1973, Hojo» 197*0 dealing with the ex p l i c a t i o n of the Japanese negative questions and the rel a t i o n s h i p between these negative questions and t h e i r responses. Shibatani (1972) proposed that the rules for appropriate answers to Japanese negative questions involve conversational implications. Soga (1973) 102 attempts to explain the peculiar use of hai and i i e as answers to negative questions on the basis of the extra-l i n g u i s t i c phenomenon of presupposition. So f a r , the analyses proposed have a l l been purely semantically motivated. Hojo (1974), on the other hand, attempts to treat the ambig-u i t y of the Japanese negative questions as a syntactic problem, and proposed that the r e l a t i o n s h i p between the questions and the answers be explained i n terms, of the lo g i c governing, what he c a l l s , the response e l i c i t a t i o n questions. The analyses proposed have given us various i n s i g h t f u l explanations concerning the p e c u l i a r i t i e s of the Japanese negative questions and t h e i r responses. The analysis i n t h i s chapter w i l l incorporate the theories and arguments that have been proposed so f a r . 3.2. NEGATIVE QUESTIONS AND THEIR RESPONSES 3.2.1. SYNTACTIC AMBIGUITY OF NEGATIVE QUESTIONS Interrogative sentences i n Japanese can be phrased p o s i t i v e l y or negatively, such as tabemasu ka ' W i l l you eat?' and tabemasen ka 'Won't you eat?*. The negatively phrased questions are the ones that give r i s e to syntactic ambiguity and often" e l i c i t unpredictable yes-no response. Observe the following conversations by speakers A and B« 103 (1) Speaker A i Kaimono e ikimasen ka. •I t i s true that you are not going shopping?• Speaker B s Hai, ikimasen. •Yes, I'm not going.' l i e , ikimasu. •No, I'm going.' (2) Speaker A : Kaimono e ikimasen ka. •Aren't you going shopping?' Speaker B » Hai, ikimasu. •Yes, I'm going.' l i e , ikimasen. •No, I'm not going.* Notice that Japanese negative questions l i k e (1) and (2) above, although they appear to be s t r u c t u r a l l y s i m i l a r on the surface, are s y n t a c t i c a l l y ambiguous, and can e l i c i t responses l i k e those of Speaker B i n (1) as we l l as ones l i k e those of Speaker B i n (2). The hai and l i e responses by Speaker B i n (1) introduce both affirmative and negative elements i n the same single response, while those by Speaker B i n (2) are s t r i c t l y affirmative or negative. The same i s also true of the following examples! (3) Speaker A i Kono hen n i wa restoran ga arimasen 1 0 4 Speaker B » (4) Speaker A : Speaker B V (5) Speaker A i Speaker B t (6) Speaker A t ka. 'Is i t true that there i s n ' t any restaurant around here?' Hai, arimasen. 'Yes, there i s n ' t . ' l i e , arimasu. •No, there i s . ' Kono hen n i wa restoran ga arimasen ka. 'Isn't there a restaurant around here?' Hai, arimasu. 'Yes, there i s . ' l i e , arimasen. 'No, there i s n ' t . * Asa-gohan o tabemasen ka. 'Is i t true that you won't eat breakfast?' Hai, tabemasen. •Yes, I won't.' l i e , tabemasu. •No, I w i l l . ' Asa-gohan o tabemasen ka. 105 •Won't you eat breakfast?' Speaker B t Hai, tabemasu. •Yes, I w i l l . ' l i e , tabemasen. 'No, I won't.' (7) Speaker A » Kono hen wa sizuka de wa arimasen ka. 'Is i t true that I t i s not quiet around here?' Speaker B i Hai, sizuka de wa arimasen. •Yes, i t i s n ' t quiet.' l i e , sizuka desu. •No, i t i s quiet.' (8) Speaker A « Kono hen wa sizuka de wa arimasen ka. 'Isn't i t quiet around here?' Speaker B i Hai, sizuka desu. 'Yes, i t i s quiet.' l i e , sizuka de wa arimasen. •No, i t i s n ' t quiet.' The (3) and (4), (5) and (6), and (7) and (8) question pairs above are s y n t a c t i c a l l y and therefore semantically ambiguous i n exactly the same way as (1) and (2) are, and they e l i c i t two d i f f e r e n t kinds of responses, depending on how the questions are being interpreted by the hearer. I t 106 seems that i n the course of normal conversations, the hearer must i d e n t i f y c e r t a i n rules and interpret the negative questions i n such a way as to be able to disambiguate them and to give the appropriate responses. The problem at issue i s how do we relate the responses to the questions, and what are the rules governing the appropriate responses to the negative questions. I t i s clear from the above examples, that the responses cannot be predicted from the question utterances alone due to t h e i r syntactic ambiguity. I w i l l come back to t h i s question of syntactic ambiguity i n section 3.2.3. 3.2.2. PRESUPPOSITIONAL INFORMATION AND CONVERSATIONAL  IMPLICATIONS According to the analysis proposed by Soga (1973). the peculiar use of hai and i i e as responses to negative questions can be explained by the incorporation of presuppositional information. This e x t r a - l i n g u i s t i c phenomenon of presupp-o s i t i o n , that i s , the conception i n the mind of the speaker which he believes or tends to believe to be true, may be implied or present only i n the context. Hence i n the normal course of conversation, the l i s t e n e r must i d e n t i f y and interpret them i n such a way as to be able to give the appropriate response. For example, i n (1) the l i s t e n e r 107 interprets the question as containing a negative presuppo-s i t i o n , that i s , speaker A presupposes that speaker B i s not going shopping, and therefore seeks affirmation or information on the truth value of his presupposition. (1) should actually have the structure of (9)« (9) Speaker A i (Anata wa kaimono e ikanai to k i i t a ga, hontoo ni) kaimono e ikimasen ka. (Osiete kudasai). '(I heard that you aren't going shopping, but i s i t true that) you are not going shopping? (Please t e l l me). In the course of conversation, only the underlined part i s act u a l l y spoken by speaker A, while the part i n parentheses may either be implied or present only i n the context. The appropriate answers to t h i s should be Hai, ikimasen 'Yes, I'm not going' where speaker A's negative presupposition i s being affirmed, and l i e , ikimasu 'No, I'm going' where the negative presupposition i s being denied. In example (2), the question i s interpreted by the l i s t e n e r as containing a positive presupposition seeking affirmation or information. The presupposition here i s that speaker A wants to go shopping, and knows that speaker B wants to go shopping too, and so he seeks affirmation on the truth value of h i s presupposition. (2) would have the 108 structure of (10), where only the underlined part i s spoken i n the conversation while the rest i s implied i n the context. (It)) Speaker A : (Watakusi wa kaimono e i k i t a i . Anata mo i k i t a i to omoimasu ga,) kaimono  e ikimasen ka. (Osiete kudasai.) '(I want to go shopping. I think that you want to go too, but) aren't you going shopping? (Please t e l l me.) In t h i s case Hai, ikimasu 'Yes, I'm going* would be given as a positive answer i n order to confirm speaker A's positive presupposition, and l i e , ikimasen 'No, I'm not going* as a negative answer i n denial of the positive presupposition. Shibatani ( 1 9 ? 2 ) , on the other hand, proposed that the rules f o r appropriate answers to Japanese questions involve conversational implications. I t seems that i n the course of the conversation, the person to whom the question i s directed, must pay attention not only to the syntactic negative marker i n the question, but also to the conversational context. Consider sentences ( 5 ) and ( 6 ) for example. In ( 6 ) , the conversational context indicates that-the negative question should be interpreted as a suggestion, or as conveying a request for speaker B to have breakfast. In t h i s case hai and i i e would be given as a positive answer and a negative answer respectively. On the other hand, i n ( 5 ) , the negative 109 question implies that the questioner holds a negative assump-t i o n about the propositional content, that i s , speaker A assumes that speaker B won't have breakfast. Here, the negative answer with Hai, tabemasen 'Yes, I won't eat' i s used to confirm or agree with the questioner's negative assumption, while the positive answer with l i e , tabemasu 'No, I w i l l eat' i s used to negate the questioner's negative assumption. 3.2.3. UNDERLYING STRUCTURES OF THE NEGATIVE QUESTIONS It seems to me that both of the above two analyses, one incorporating presuppositional information and the other conversational implications, s a t i s f a c t o r i l y capture the semantic r e l a t i o n s h i p between the negative questions and t h e i r responses expressed by hai and i i e . However, both the analyses are primarily semantically motivated and do not provide any strong syntactic evidence towards the argument. Hojo (1974) proposed that the semantic ambiguity that i s associated with the negative questions, for example that of (1) and (2), could be accounted f o r s y n t a c t i c a l l y by positing two d i f f e r e n t deep structures f o r them. According to Hojo, the h a i - i i e question type l i k e (1), which he c a l l s a true negative question, i s considered to be derived from the alternative question type as shown below 110 i n (11). Kaimono I e ikimasen Q S i m i l a r l y , examples (3), (5) and (7), are also considered to be derived from alternative question types. I l l Example ( 1 ) , which actually contain a negative presupp-o s i t i o n , would have the underlying structure equivalent to (12) . With the exception of l e x i c a l items, ( 3 ) , (5) and (6) would also have underlying structures s i m i l a r to that of ( 12) . In (12) above, the question (Q) i s dominated by the post sentence (Post S),* and the negative i s introduced optionally i n the deep structure. The negative na or en i n the lexicon would be assigned the feature notation Q- affirmative] which w i l l take care of the negative presupposition. As i s clear from (11) , the question i s applied to a negative statement, and hence the only appropriate responses would be those of speaker B i n ( 1 ) . With respect to example ( 2 ) , we have noticed e a r l i e r 112 that (2) implies that i t has a positive presupposition underlying the question. This observation suggests that (2) could not possibly contain a negative i n the course of i t s derivation, and that the question has to be applied to a positive statement. Notice also that although there i s the presence of the syntactic negative i n the surface structure of (2), as manifested by masen, i t does not imply semantic negativity. Hojo states that because (2) contains a positive presupposition, i t could not be considered to have derived d i r e c t l y from the alternative question type as (1) did. Instead, he proposed that the negative i n (2) be treated as a sentence-final p a r t i c l e , functioning i n the same way as those sentence-final p a r t i c l e s of the assertive yo and ne, and the neutral no and ka. (2) would be considered to have the underlying structure equivalent to ( 13) . Examples (4), (6) and (8) would also have s i m i l a r underlying structures as that of ( 13) . Examples such as (2), (4), (6) and (8) above, according to Hojo, are false negative questions as opposed to those of ( 1 ) , ( 3 ) , (5) and (7) which are true negative questions. 113 In ( 13)i "both the negative and the question are dominated by the post sentence. The negative i n the lexicon would be assigned the feature notation {%- a s s e r t i v e j , which w i l l account for the positive presupposition. Notice that the question i s formed on a positive statement, and thus the only appropriate responses would have to be those of speaker B i n (2). Hojo's analysis to a c e r t a i n extent, gives us a rather convincing argument for the semantic and syntactic ambiguity that e x i s t 1 i n Japanese negative questions. However, I would l i k e to point out that Hojo's analysis also has c e r t a i n weaknesses. F i r s t l y , h i s treatment using two separate entries f o r the negative morpheme na and en i n the lexicon, 114 one with [- affirmative] f or the true negative, and another with [+ assertive] and sentence p a r t i c l e ] f o r the false negative, needs further examination. I t seems to me that assigning two di f f e r e n t semantic features to the negative does not actu a l l y solve the problem of syntactic ambiguity that exists i n the negative questions. The assignment of two d i f f e r e n t feature notations to the negative i s b a s i c a l l y with respect to semantic interpretation. a f f i r m a t i v e j feature notation i s assigned to the true negative i n order to account f o r the negative presupposition that underlies the question, and [_+ assertive] i s assigned to the fa l s e negative i n order to account f or the positive presupposition. This i s , a f t e r a l l , p rimarily a semantic solution and not a syntactic solution at a l l . Furthermore, the assignment of the feature notation [+ assertive] to the false negative i s questionable. Notice that i n Japanese, a suggestion or a request that i s phrased as a negative question as that of (2), can also be phrased p o s i t i v e l y as (14), with the assertive masyoo ' L i t . Let us'. (2) Kaimono e ikimasen ka. [+ assertive} 'Aren't you going shopping?' (14) Kaimono e ikimasyoo ka. [+ assertive] 'Shall we go shopping?' 115 In assigning \+ a s s e r t i v e j feature to the false negative i n ( 2 ) , i t leads us to believe that (2) and (14), which contain the assertive masyoo, are s i m i l a r . However, i f we were to examine (2) and (14) further, we w i l l notice that they are b a s i c a l l y d i f f e r e n t . (14) i s less p o l i t e than ( 2 ) , and i t implies the speaker's positive assumption about the suggest-ion. The effect of such an assumption obligates the l i s t e n e r to follow the suggestion. However, (2) does not necessarily imply such an obligation, and the l i s t e n e r i s l e f t with a choice of whether to follow or r e j e c t the suggestion. Secondly, po s i t i n g two d i f f e r e n t deep structures f o r the negative questions as suggested by Hojo does not actually accomplish anything that neither Shibatani (1972) nor Soga (1973) f a i l e d to accomplish. Hojo's underlying structures fo r the true negative question and the false negative question are primarily based on semantic int e r p r e t a t i o n of the question. This would b a s i c a l l y be s i m i l a r to Shibatani*s and Soga's analyses.based on conversational implications and presuppositions. Notice that Shibatani's and Soga's treatment would also require two d i f f e r e n t deep structures fo r the true and false negative questions because of the difference i n presuppositions or conversational implications which underlie the questions. Perhaps the difference would be that Hojo's underlying structures can be considered to be closer to the surface structures. F i n a l l y , Hojo's treatment of the f a l s e negative as a 116 sentence-final p a r t i c l e functioning i n the same way as the sentence-final p a r t i c l e s of the assertive y_o and ne, and 2 the neutral no and ka, i s questionable. Observe the following sentences: (2) Kaimono e ikimasen ka. 'Aren't you going shopping?' (15) Kaimono e iku no. 'Are you going shopping?* (16) Kaimono e iku yo. 'Let's go shopping.' (17) Kaimono e iku ne. 'Let's go shopping.' The underlined parts i n the above examples, are the sentence-f i n a l p a r t i c l e s . Notice that the verb iku 'to go* of (15)» (16) and (17) i s i n the root or c i t a t i o n form, that i s 'iku*. However, that of (2) i s not and i s i n the form of ' i k i - * . This observation shows that the sentence-final p a r t i c l e s of no, yo and ne a l l follow the root form of the verb, while the sentence-final p a r t i c l e masen does not. Therefore, i f we follow Hojo's claim, masen w i l l have to be considered an exception from a l l other sentence-final p a r t i c l e s . This seems to suggest that Hojo's treatment of the f a l s e negative as a sentence-final p a r t i c l e i s perhaps an over-generalization, and perhaps gives r i s e to more questions rather than of f e r i n g 117 a solution to the e x i s t i n g problem. 3 . 2 . 4 . THE LOGIC UNDERLYING THE RELATIONSHIP BETWEEN Hojo (1974) proposed that a rule of l o g i c governs the r e l a t i o n s h i p between the negative questions and t h e i r responses. According to t h i s l o g i c , the responses represen-ted by Hai and l i e a f f i r m or negate the statement i n the question. Let us examine examples (1) and ( 2 ) . (1) Speaker A s Kaimono e ikimasen ka. THE NEGATIVE QUESTIONS AND THEIR RESPONSES •Is i t true you are not going shopping?' A Speaker B t Hai, ikimasen. l i e , ikimasu. 'False, I am (2) Speaker A : Kaimono e ikimasen ka. 'Aren't you going shopping?' A Speaker B Hai, ikimasu 'True, I am going.' T A 118 l i e , ikimasen. 'False, I am not going.' F r \ J A In the above examples, the symbols A,r\JA, T, and F represents l o g i c a l terms. In asking the question, the interrogator seeks from the l i s t e n e r the assignment of a truth value to the statement A i n the question. The assignment of the truth value may be either true (T) or f a l s e (F). I f A meets the truth condition, the l i s t e n e r assigns the t r u t h value T to the statement A i n the form of Hai as the response ('A i s t rue'). On the other hand, i f A does not meet the truth condition, the l i s t e n e r assigns the truth value F to A i n the form of l i e ('A i s f a l s e ' ) . The t r u t h condition can be either A o r ^ A (the negative form), and t h i s usually follows the truth value i n the response. In t h i s manner, the syntactic and semantic relationships between the negative questions and t h e i r peculiar responses can be explained i n a l o g i c a l way. However, t h i s l o g i c underlying the r e l a t i o n s h i p between the negative questions and t h e i r responses as proposed by Hojo also has i t s shortcomings. The problem to be solved i s on what basis or with respect to what does the l i s t e n e r evaluate the truth value of the statement A? How does the l i s t e n e r decide whether the truth value i s T or F? On what i s the l i s t e n e r ' s designation of the truth value T or F based? 119 How does the l i s t e n e r decide whether the truth condition: i s A or ru A? I t seems to me that the assignment of the truth value T or F, as well as the decision on whether the tr u t h condi-t i o n i s A orro A, can only be explained with respect to the e x t r a - l i n g u i s t i c factors of presuppositions and conversational implications that underlie the questions. This suggests that the r e l a t i o n s h i p between the negative questions and t h e i r responses involve larger semantic e n t i t i e s that include presuppositions and conversational implications. A l l t h i s seems to give greater support to the semantically motivated analyses proposed by Soga (1973) and Shibatani (1972). 3 . 2 . 5 . INTONATION It has been noted that there e x i s t s a c e r t a i n overt phonological feature at the surface l e v e l that distinguishes the (1) and ( 2 ) , (3) and ( 4 ) , (5) and ( 6 ) , and (7) and (8) question pairs. This seems to be the intonations that are associated mainly with the negative morphemes na and en, as pointed out by Hojo (1974) and Kuno (1973). Examples (1) and (2) would have d i f f e r e n t intonation patterns as follows: 120 _ J — L _ J 1 s* (1) Kaimono e ikimasen ka. 'Is i t true that you aren't going shopping?' i — i S (2) Kaimono e ikimasen ka. 'Aren't you going shopping?' Notice that i n (1), there i s a c e r t a i n degree of prominence i n the intonation retained "by the negative morpheme. Hojo points out that t h i s can he either primary or secondary depending on i t s r e l a t i v e place of occurrence. On the other hand, i n (2), t h i s prominence associated with the intonation of the negative morpheme i s l o s t altogether. However, at t h i s stage, i t i s not clear as to whether t h i s phonological d i s t i n c t i o n occurs regularly or not. I t seems to me at present that the intonations used with such negative questions are d i f f i c u l t to define i n order to d i s t i n g u i s h syntactic and semantic difference. In f a c t , I have consulted a number of Japanese native speakers about t h i s , and i t seems that to some of them the intonations do not mark any syntactic and semantic d i s t i n c t i o n between (.1) and (2). Furthermore, to many of them the same intonation pattern seems'to be applica-ble to both (1) and (2) with no difference whatsoever! Next observe the following sentences! 121 _ J — L _ J — 1 _ / " (18) Speaker A : Kaimono e ikimasen desita ka. 'Is i t true that you didn't go shopping?' Speaker B i Hai, ikimasen desita. •Yes, I didn't go.• l i e , ikimasita. 'No, I went.' J—1 /• (19) Speaker A : Kaimono e ikimasen desita ka. •Didn't you go shopping?' Speaker B « * Hai, ikimasita. 'Yes, I went.' * l i e , ikimasen desita. •No, I didn't go.' T r-t_j / (20) Speaker A : Sensei n i aimasen desita ka. 'Is i t true that you didn't meet the teacher?' Speaker B : Hai, aimasen desita. •Yes, I didn't meet (him).' l i e , aimasita. •No, I met (him).• (21) Speaker A t Sensei n i aimasen desita ka. 'Didn't you meet the teacher?' Speaker B : Hai, aimasita. •Yes, I met (him).' l i e , aimasen desita. •No, I didn't meet (him).' Notice the intonation difference among (18), (19)» (20) and (21). Interrogative questions i n Japanese have the delayed r i s e type terminal intonations. (19) and (21) have the neutral interrogative intonations, that i s , of the delayed r i s e type on ka, while (18) and (20) have terminal r i s i n g intonations e a r l i e r than those i n (19) and (21). I t seems that the past tense form of (1) and (2) question pairs can only e l i c i t one type of response, those of (18)j those of (19) are ungrammatical. A l l of the native speakers with whom I have consulted agreed on t h i s point. However, many of them did not agree with the intonation pattern as those of the above examples (18) and (19)» and some of them commented that (18) can even have the intonation pattern of (19) and yet the responses e l i c i t e d would s t i l l be those of (18). With respect to (20) and (21), i t i s possible to e l i c i t two di f f e r e n t types of responses depending on intonation. However, i t seems that the intonation pattern i s i r r e g u l a r . There seems to be no agreement on the intonations among the native speakers of Japanese with whom 123 I have consulted. Some argued that the intonation pattern of (20) should ac t u a l l y he that of (21) and vice versa, while others argued that the intonations do not mark any syntactic d i s t i n c t i o n between (20) and (21) . From the above observations, I can only conclude the following s (i ) The phonological d i s t i n c t i o n based on intonation does not seem to be r e l i a b l e , because i t i s d i f f i c u l t to define s t r i c t l y . This may perhaps be a r e s u l t of d i a l e c t a l difference, but at the moment, i t i s not clear. ( i i ) The fact that we cannot r e l y on intonations to disambiguate the negative questions seems to suggest that other factors are involved. I t appears to me that the negative questions i n Japanese involve larger semantic e n t i t i e s that include not only presuppositions and conversatio-nal contexts, but perhaps also f a c i a l and body expressions, past conversation, circumstances and location of the conversation, and a l l other forms of e x t r a - l i n g u i s t i c factors. This obser-vation to a c e r t a i n extent, strengthens the analysis based on presuppositional information proposed by Soga'(1973)• 3 . 3 . CONCLUSION Negative questions i n Japanese are s y n t a c t i c a l l y ambiguous and can e l i c i t peculiar responses represented by hai 'yes' and i i e 'no'. As we have noticed, the hai and i i e i n one type of response introduce both affirmative and negative elements i n the same single response, while i n the other, the hai response i s s t r i c t l y affirmative and the i i e response i s s t r i c t l y negative. In t h i s chapter, I have attempted to explain how we can relate the questions and the answers c o r r e c t l y . We have seen that the responses cannot be predicted from the question utterances alone due to t h e i r syntactic and semantic ambiguity, and have resorted to the e x t r a - l i n g u i s t i c information such as presuppositions and conversational implications as explanations. I have shown that there e x i s t s a co r r e l a t i v e r e l a t i o n s h i p between the semantic information of the negative questions and the responses that they e l i c i t . However, t h i s i s purely semantically motivated and does not provide any syntactic evidence. This semantic ambiguity that i s associated with the negative questions, on the other hand, i s accounted for s y n t a c t i c a l l y by positing two di f f e r e n t deep structures fo r them, and assigning two di f f e r e n t types of feature notation to the negative morpheme na and en i n the lexicon. In one, the negative i s introduced optionally i n the deep 125 structure, and the feature notation [- affirmative]] i s assigned to the negative morpheme na or en. In the other, the negative i s treated as a sentence-final p a r t i c l e and i s o r i g i n a l l y not present i n the deep structure, and the negative morpheme i s assigned the feature notation assertive J which takes care of the positive presupposition underlying the question. However, I have also shown that t h i s has i t s defects. With respect to the re l a t i o n s h i p between the negative questions and t h e i r responses, i t seems that t h i s can be explained i n terms of the lo g i c that underlies the questions. According to t h i s l o g i c , hai uttered as an answer to the question affirms the statement i n the question, while i i e negates the statement i n the question. However, we have also seen that t h i s has i t s weaknesses, and that the hai and i i e as answers to the negative questions can only be explained i n terms of e x t r a - l i n g u i s t i c factors underlying the questions. We have also noticed that intonations, to a c e r t a i n extent, mark the d i s t i n c t i o n i n the syntactic ambiguity of the negative questions. However, I have pointed out that the intonation patterns are d i f f i c u l t to define s y n t a c t i c a l l y , and that they do not seem to occur regularly. Whether t h i s i s a r e s u l t of d i a l e c t a l differences or not, i s not cl e a r at t h i s stage, and requires further research. 126 FOOTNOTES 1. I think the term Post Sentence (Post S) as used i n Hojo (1974), probably refers to the sentence-final p a r t i c l e s such as those of the assertive yjo and ne and the neutral ka and no. 2. This observation was brought to my notice by M. Soga i n my discussion with him. 127 CHAPTER FOUR ON THE SEMANTIC AND SYNTACTIC STRUCTURE  OF THE PARTICLES MO, WA AND GA 4 . 1 . INTRODUCTION In recent years there have been several a r t i c l e s published, dealing with the analysis and the int e r p r e t a t i o n of the adverbial even i n English. Among some of these publications are Bruce Fraser (1969 and 1971), L. Horn (1969 and 1971), S. Anderson (1972) and R. Jackendoff (1972). These publications have given us various i n s i g h t f u l explana-tions concerning the nature and interpretation of even. The corresponding word for even i n Japanese i s generally believed to be mo, which has been variously translated into English a s even, also, too and as many (much) as. The orientation of t h i s chapter w i l l be one of attempting to examine the facts of the Japanese language concerning the p a r t i c l e mo, and to characterize the general nature of that p a r t i c l e . On the basis of the studies made by Kageyama (1973) and Soga (1975)1 I w i l l t r y to present a general approach for the inter p r e t a t i o n of the p a r t i c l e mo, and at the same time, w i l l also present a variety of Japanese constructions 128 using mo to i l l u s t r a t e the approach. Also, I w i l l examine the r e l a t i o n s h i p between the negative and mo, especially where i t concerns the positive-negative r e l a t i o n s h i p between the assertion and the expectation underlying mo. Furthermore, I w i l l examine the presuppositional properties of mo, the presence of which i s responsible for the unexpect-edness or surprise that usually accompanies the use of the p a r t i c l e , and which also i s relevant f o r the correct surface semantic interpretation. Then, i n the l a t t e r h a l f of the chapter, having provided a general schema for the interpre-t a t i o n of mo, I w i l l t r y to examine whether the same l i n e of approach can also be applied to other p a r t i c l e s such as wa and ga. 4.2. INTERPRETATION OF THE PARTICLE MO 4.2.1. PRESUPPOSITIONAL PROPERTIES OF MO This section i s concerned with the presuppositional properties that are associated with the p a r t i c l e mo, the presence of which i s relevant for the semantic int e r p r e t a t i o n of the sentence. Let us begin by observing the occurrence of mo i n the following sentencess (1) John mo nihongo no gakusei desu. 129 •John i s also a student of Japanese.' (2) Sono hon mo nihongo no hon desu. 'That "book i s also a Japanese book. • (3) Taroo wa eigo mo wakarimasu. •Taroo knows English too.' The eff e c t of the p a r t i c l e mo i n the above sentences expresses the meaning of membership within a set. In (1) for example, John, to which the p a r t i c l e mo i s attached, i s regarded as a member of a group of students of Japanese? i t i s implied that there are also,other members. Therefore, the use of mo i n (1) could imply, f o r example, that B i l l i s a student of Japanese, Mary i s a student of Japanese, Jane i s a student of Japanese, and that John i s a student of Japanese. Si m i l a r l y y i n ( 2 ) , the NP sono hon 'that book' i s considered to belong to a set of s i m i l a r tokens, that i s , Japanese books, and implies that there also e x i s t other books of the same kind. In (3), mo implies that Taroo knows other languages such as French, Spanish, German, Russian or Chinese, as well as English. Hence the use of mo i n the above examples implies or permits the l i s t e n e r to make the presupposition that the constituent in,the scope of mo ( i n the above cases, John i n ( 1 ) , sono hon i n ( 2 ) , and eigo i n (3)) must be viewed as a member of a set of s i m i l a r tokens. Although i n the above examples (l)-(3)» only one member of the set i s mentioned i n the sentence, the effect of mo implies the existence of 130 other i d e n t i c a l members within that set. Notice that the meaning of mo here would be s i m i l a r to that of top or also i n English. I f the deep structures should c o r r e c t l y predict the interpretations of mo, then they must incorporate i n them the presuppositions that are associated with the sentences. Following our observation! the i n t e r p r e t a t i o n of example (1) can be analyzed into at least two parts t (4) a. X (Y,Z) ga nihongo no gakusei da; *X (Y,Z) are students of Japanese*, b. John ga nihongo no gakusei da. •John i s a student of Japanese.' (4b) i s c l e a r l y the assertion of the sentence i n ( 1 ) , and notice that i t remains unaffected even i f mo i s not present. (4a) i s the implication of ( 1 ) , that i s to say, the presence of the p a r t i c l e mo i n ( 1 ) , implies the information shown i n (4a). The two sentences underlie ( 1 ) which can be represented by the following tree structure i f we subscribe to the abstract verb Prsp.s 131 X (Y,Z) nihongo ho. John nihongo no gakusei da gakusei da where Prsp = Presupposition This whole structure of (5) underlies ( 1 ) . The structure consists of two sentences and S 2, e x i s t i n g with the abstract verb Prsp. This Prsp relates the two sentences and S 2, meaning that i s presupposed f o r Sg. Therefore, S^ represents the presupposition and S 2 the assertion. Through the process of transformation, Prsp together with S^, are consequently deleted and, at the same time, the p a r t i c l e mo i s attached to the NP of S 2 to produce the surface structure of ( 1 ) . Sentences (2) and (3) can also be derived i n a s i m i l a r way, where the underlying structures are considered to consist of at least two sentences. ( 6 ) a. X (Y,Z) ga nihongo no hon da. •X (Y,Z) are Japanese books.' b. Sono hon ga nihongo no hon da. 'That book i s a Japanese book.' 132 (7) a. Taroo ga X ( Y , Z ) gengo ga wakaru. 'Taroo knows X ( Y , Z ) languages.' I D . Taroo ga eigo ga wakaru. 'Taroo knows English.* ,* At t h i s stage, I think i t would be worthwhile to examine further the notion of membership i n a set or the notion of s i m i l a r i t y of tokens. Consider the following sentencet (8) Ninon e i k i t a i ga, okane mo hima mo nakute wa dame desu. *I want to go to Japan, but i t i s impossible as I have neither the money nor the time f o r i t . ' Normally okane 'money* and hima 'time' would not be considered to have any i d e n t i c a l import, that i s to say, they would not be considered to belong to the same cl a s s . Yet, the noun phrases okane and hima have mo attached to them. This shows that the notion of i d e n t i t y here need not necessarily mean natural i d e n t i t y , but rather semantic i d e n t i t y . I f we look at i t from the point of view that okane and hima are necessary f o r going to Japan, then we can consider them to be semantically equivalent and belonging to the same semantic set. Hence the notion of s i m i l a r i t y of tokens would mean that the tokens with which the scope of mo i s contrasted must share at least the same co-occurrence r e s t r i c t i o n . This 133 means that i t must be semantically possible f o r the other members of the set torbe substituted f o r the contrasted constituent i n the scope of mo. In addition, there i s also a sense i n which the p a r t i c l e mo involves the notion of expectation. I f the notion of expectation accompanies the presupposition, then mo comes to assume the meaning of even i n English. I t seems that i t i s t h i s notion of expectation accompanying the presupposition that i s responsible for the surprise or unexpectedness that appears with mo. Consider the following sentences: (9) Sensei mo kono mondai wa tokenai. 'Even the teacher can't solve the problem.* (10) Kodomo mo hiragana ga yomeru. •Even a c h i l d can read hiragana.' (11) Kono omoi hako wa otoo-san mo ugokenai. 'As for t h i s heavy box, even father can't move i t . ' The e f f e c t of mo on the above sentences ( 9 ) - ( l l ) r e f l e c t s an attitude on the part of the speaker or hearer that the information contained i n the rest of the sentence would not normally be expected to be true of the constituent i n the scope of mo. The constituents that f a l l within the scope of mo i n the above sentences are the subject NP's se'hsei 'teacher' i n ( 9 ) , kodomo ' c h i l d ' i n ( 1 0 ) , and otoo-san 134 •father' i n (11) . In ( 9 ) . f o r example, the speaker or hearer expects that the teacher w i l l be able to solve the problem, although other people (perhaps those with less q u a l i f i c a t i o n s or education) are not able to. However, contrary to t h i s expectation, the fact i s that the teacher, too, i s not able to solve the problem, hence the surprise or unexpectedness. S i m i l a r l y i n ( 1 0 ) , the speaker or hearer expects other people, such as adults, to be able to read hiragana and would not normally expect a c h i l d to be able to read i t . However, contrary to t h i s expectation, a c h i l d can also read i t , thus the surprise. In (11) , the speaker expects the father, who i s perhaps considered to be the strongest member i n the family, to be able to move the heavy box, but the fact i s that the father, too, cannot move i t . This gives r i s e to the surprise. This peculiar'^property of unexpectedness or surprise that i s associated with mo can perhaps be made clearer by constructing a scale of degree, such as that suggested by Soga (1975)i and which i s shown below. Emphasis Emphasis (12) Minimum i Extreme Maximum Extreme Membership (also/too) 135 Applying (12) to sentence ( 9 ) , we have a scale of degree of int e l l i g e n c e among the members i n the set ( i n t h i s case, probably the educated group) which i s scaled i n an increas-ing order of i n t e l l i g e n c e , ending with the sensei 'teacher' who i s considered to be i n the maximum in t e l l i g e n c e p o s i t i o n of the scale. Thus the speaker expects that the teacher, who i s best q u a l i f i e d , w i l l be able to solve the problem, while other people cannot. But the fact that he, cannot gives r i s e to the surprise that i s associated with mo. In t h i s sense,, the p a r t i c l e mo i s used to emphasize the unexpected nature that i s associated with the constituent i n i t s scope. I t seems that the p a r t i c l e may be used i n extreme cases of membership i n a set, emphasizing either the maximum or the minimum extreme. In the case of ( 9 ) , i t emphasizes the maximum extreme, with the teacher considered to be the most q u a l i f i e d person r e l a t i v e to the compared group. The same treatment can also be applied to (10) by constructing a scale of degree of a b i l i t y to read, with the kodomo ' c h i l d ' f a l l i n g i n the minimum extreme p o s i t i o n r e l a t i v e to the compared group. In the case of (11) , a scale of degree of strength can be applied, with otoo-san 'father' i n the po s i t i o n of maximum extreme r e l a t i v e to the compared group;. Sentences containing mo, which spec i f i e s emphasis such as examples (9)» (10) and (11) above, can be considered to have at least three sentences underlying them. Considering 136 ( 9 ) , the interpretation can he analyzed into at leas t three parts t (13) a. X (Y,Z) n i kono mondai ga tokenai. 'X (Y,Z) can't solve the problem.' b. Sensei n i kono mondai ga tokeru. 'The teacher can solve the problem.' c. Sensei n i kono mondai ga tokenai. 'The teacher can't solve the problem.' Notice that there i s an additional piece of information expressed i n a sentence l i k e ( 9 ) , when compared to those of ( l ) - ( 3 ) discussed e a r l i e r . This additional piece of inform-ation i s represented i n (13b) which states that the speaker expects the teacher to be able to solve the problem. I t i s exactly the presence of t h i s notion of expectation that i s responsible for the surprise that may appear with the p a r t i c l e mo. Also, notice the positive-negative r e l a t i o n s h i p between the expectation (b) and the assertion ( c). Using the three-part i n t e r p r e t a t i o n of (13)» the following underlying structure f or (9) i s possibles 137 (14) X (Y,Z) kono sensei kono mondai tokenai mondai token*.;. In structure (14), S 2 presupposes which act u a l l y consists of two sentences, and S^. Just as i n (5), Prsp, and are deleted i n the process of transformation, and mo i s inserted into the NP sensei 'teacher' i n S 2 to derive the surface structure ( 9 ) . Notice that i n a l l the cases c i t e d above, the occurrence of the p a r t i c l e mo i n Japanese does not a l t e r the basic proposition of the sentence. The main assertion of the sentence with respect to who cannot do what, remains unaffected even i f mo i s not present, as shown i n (4b) and (13c). However, the p a r t i c l e mo provides additional information about the proposition, the speaker's or the hearer's viewpoint and about the state of the world. The presence of mo i n (9) f o r example, adds the information shown i n (13a) and (13b) to the information e x p l i c i t l y present i n the sentence without mo, that i s ( 13c) . 138 4.2.2. THE OCCURRENCE OF MO WITH QUANTIFIERS In t h i s section, I w i l l present and examine a variety of syntactic constructions to indicate how the in t e r p r e t a t i o n of mo follows the approach set up i n the previous section. The p a r t i c l e mo occurring with q u a n t i f i e r s provides rather, i n t e r e s t i n g examples for the inter p r e t a t i o n of mo, because they show c l e a r l y the semantic and syntactic nature of the p a r t i c l e . Let us examine some examples of mo occurring with time adverbials. (15) Taroo wa tooka-kan mo benkyoo s i t a . 'Taroo even studied f o r ten days.* (16) Okyaku-san wa hyaku-nin mo k i t a . 'Even a hundred guests came.' (17) Tanaka-san wa asa no n i - z i made mo hataraita. 'Mr. Tanaka even worked u n t i l two i n the morning.' (18) Watakusi wa Amerika e san-do mo i t t a . •I even went to America three times.' ( i . e . I went to America f o r as many as three times.) (19) Otooto wa gohan o go-hai mo tabeta. •My younger brother even ate f i v e bowls of r i c e . ' ( i . e . My younger brother ate as many as f i v e bowls of r i c e . ) 139 Notice that i n the above examples of ( 1 5 ) - ( 1 9 ) , the use of mo sp e c i f i e s both emphasis and membership. The interp r e t a t i o n f o r each of the above sentences follows from the approach given i n the previous section, 4.2.1. There are at leas t three parts to the inte r p r e t a t i o n of the above sentences, and they are as follows t (20) a. Taroo wa nan-nitikan ka benkyoo s i t a . 'Taroo studied f or a number of days.' b. Taroo wa tooka-kan wa benkyoo sinakatta. 'Taroo didn't study f o r ten days.* ( i . e . The number of days that Taroo studied didn't amount to ten days.) c. Taroo wa tooka-kan benkyoo s i t a . 'Taroo studied f or ten days.* (21) a. Okyaku-san wa nan-nin ka k i t a . 'A number of guests came.' b. Okyaku-san wa hyaku-nin wa konakatta. •Not a hundred guests came.' ( i . e . The number of guests who came did not amount to a hundred.) c. Okyaku-san wa hyaku-nin k i t a . 'A hundred guests came.' (22) a. Tanaka-san wa nan-zi made ka hataraita. 'Mr. Tanaka worked u n t i l a c e r t a i n time.' 1 4 0 b. Tanaka-san wa asa no n i - z i made wa hataranakatta. 'Mr. Tanaka didn't work u n t i l two i n the morning.' c. Tanaka-san wa asa no n i - z i made hataraita. 'Mr. Tanaka worked u n t i l two i n the morning.* (23) a. Watakusi wa Amerika e nan-do ka i t t a . 'I went to America a number of times.' b. Watakusi wa Amerika e san-do wa ikanakatta. 'I didn't go to America three times.' c. Watakusi wa Amerika e san-do i t t a . •I went to America three times.' (24) a. Otooto wa gohan o nan-bai ka tabeta. 'My younger brother ate a number of bowls of r i c e . ' b. Otooto wa gohan o go-hai wa tabenakatta. 'My younger brother didn't eat f i v e bowls of r i c e . ' c. Otooto wa gohan o go-hai tabeta. 'My younger brother ate f i v e bowls of r i c e . * Considering ( 2 0 ) , the three-part i n t e r p r e t a t i o n when taken together simply means that Taroo worked f o r a number of days, and was not expected to work f o r as many as ten days, but contrary to t h i s , he worked f o r ten days. A l l the (a), (b), and (c) sentences provide the underlying int e r p r e t a t i o n f or the surface sentences of ( 1 5 ) - ( 1 9 ) . Comparing the (b) and 141 (c) sentences, notice that the (b) sentences are a l l negative while the (c) sentences are p o s i t i v e . This positive-negative r e l a t i o n s h i p i s interpreted as r e f l e c t i n g an attitude on the part of the speaker that the information contained i n the rest of the sentence would not normally be expected to be true of the constituent i n the scope of mo. In the above examples, i t i s the (b) part of the inter p r e t a t i o n that provides the explanation f or the negative expectation associated with sentences ( 1 5 ) -(19). Therefore, we see that i n positive sentences, the expectation associated with the main clause i s negative. The interpretation of the negative sentence i s obtained i n exactly the same way as for the positive sentence. I t seems that sentence negation has a systematic effect on sentences containing the p a r t i c l e mo. Sentence negation simply negates the three parts of the in t e r p r e t a t i o n , where the negation of a negation seems to r e s u l t i n a positive statement for the (b) parts. The negation of (16) for example, would be ( 2 5 ) . (25) Okyaku-san wa hyaku-nin mo konakatta., 'Not even a hundred guests came.' The corresponding interpretation for (25) i s as follows» (26) a. Okyaku-san wa nan-nin ka konakatta. 142 *A number of guests didn't come.* b. Okyaku-san wa hyaku-nin k i t a . *A hundred guests came.' c. Okyaku-san wa hyaku-nin konakatta. 'Not a hundred guests came.* Again, observe the positive-negative r e l a t i o n s h i p between the expectation (b) and the assertion ( c ) . Thus, f o r negative sentences the expectation associated with the main clause i s po s i t i v e . According to Kageyama (1973)» the expectation associated 2 with the p a r t i c l e mo can be formalized as follows« (27) Expect [cC,{-f(x) . (/a/^ /x/)}J In ( 2 7 ) , f represents the assertion, a the stated quantity i n the assertion, while x represents the quantity expected. According to t h i s formula, the stated quantity a has to be less or at least equal to the expected quantity x. In other words, the expectation has to be greater or at le a s t equal to the stated number. Applying (27) to example (16), the expectation of (21b) w i l l be» (28) Expect [dC,{-f(x) . ( / 1 0 0 / 4 A / ) } J In (28), the speaker or hearer expects that not a hundred 143 guests or more w i l l come. Contrary to t h i s expectation, the r e a l number of guests who came turned out to be exactly a hundred, an unexpected number, and thus the surprise accompanying i t . The expectation expressed as formula (28) can be rewritten as (21b*) . (21b ' ) Okyaku-san wa hyaku-nin mata wa sore izyoo konakatta. *A hundred guests or more didn't come.' I f we were to compare (21b ') and ( 2 1 c ) , we w i l l notice that the positive-negative r e l a t i o n s h i p s t i l l e x i s t s . Formula (27) simply states that i f the assertion i s f (a positive statement), then the expectation would be - f . On the other hand,.if the assertion i s ^  (a negative statement), then the expectation would have to be - ( - f ) , which would ac t u a l l y be f. Thus, according to the formula, for a positive sentence the expectation associated with i t w i l l be a negative, while f o r a negative sentence the expectation w i l l be p o s i t i v e . However, i t must be pointed out that t h i s rule i s i n s u f f i c i e n t to account f o r the expectation underlying d i f f e r e n t kinds of sentences. I t seems that t h i s has to be supplemented by syntax. There are some cases where the formula (27) proves inapplicable. Consider the following sentencesi 144 (29) a. Taroo wa tooka-kan mo kaeranakatta. •Taroo didn't even return f o r ten days.' b. * Taroo wa tooka-kan mo kaetta. •Taroo even returned for ten days.* (30) a. Yuki wa sansyuu-kan mo yamanakatta. •I t didn't even stop snowing for three weeks.' b. * Yuki wa sansyuu-kan mo yanda. 'I t even stopped snowing for three weeks.' (31) a. Kaze wa ikkangetu-kan mo naoranakatta. •(I) didn't even recover from a cold for a month.' b. * Kaze wa ikkangetu-kan mo naotta.' •(I) even recovered from a cold f o r a month." (32) a. Si b a i wa zyuuni-zi made mo owaranakatta. •The play didn't even end u n t i l twelve o'clock.* b. * S i b a i wa zyuuni-zi made mo owatta. 'The play even ended u n t i l twelve o'clock.* (33) a. Paatee wa h a t i - z i made mo hazimaranakatta. 'The party didn't even s t a r t u n t i l eight o'clock.' b. * Paatee- wa h a t i - z i made mo hazimatta. •The party even started u n t i l eight o'clock.' Notice that a l l the (a) examples above are negative sentences. I f we were to follow formula ( 2 7 ) , then the expectation ^ would have to be the posi t i v e (b) sentences. 145 However, i t i s evident that the expectation could not possibly be the positive sentences of (b), as they are a l l ungramraatical. This ungrammaticality i s due to the nature of semantically punctual verbs l i k e kaeru 'to return*, yameru 'to stop*, and naoru *to recover from*, being such that they have to be negated when co-occurring with time adverbials.-^ Hence, for a sentence l i k e ( 2 9 ) , the corres-ponding could only be« (34) Taroo ga kaeranakatta no wa tooka-kan, mata wa sore izyoo de wa n a i . • I t was not for ten days or more that Taroo didn't return.* (34) shows that what i s predicated i s the time adverbial, and what i s ac t u a l l y being negated i s not the verb but the time adverbial. In other words, the expectation associated with (29) would be the negation of the whole statement Taroo ga  kaeranakatta no wa tooka-kan da * I t was for ten days that Taroo didn't return.' This observation shows that we have to define exactly what f, or - ( - f ) represents. This can be accounted for s y n t a c t i c a l l y i n the deep structure. (29) can be s t r u c t u r a l l y represented as» 146 Taroo kaetta S Q NEG Taroo kaetta In ( 3 5 ) . S 2 presupposes S^, and through the process of transformation, Prsp and are consequently deleted. Mo i s attached to the VP of S 2 which i s then lowered into S^Q. Here, we w i l l have to accept the view that s y n t a c t i c a l l y the morpheme l i k e NEG must he lowered. Examples (30)-(33) can a l l he accounted for i n a s i m i l a r way. (35) i s the only possible i n t e r p r e t a t i o n f o r mo i n negative sentences containing time adverbials co-occurring with semantically punctual verbs. However, for negative 147 sentences containing semantically durative verbs co-occurring with time adverbials, at least two interpretations are possible for mo. Consider the following examples: (36) Ame wa sansyuu-kan mo huranakatta. 'It didn't even r a i n for three weeks.* (37) Hanako wa tooka-kan mo benkyoo sinakatta. 'Hanako didn't even study for ten days.* (38) Kare wa zyuunizi-kan mo arukanakatta. 'He didn't even walk for twelve hours.' The above examples are a l l ambiguous i n at least two readings. For example, i n one reading of ( 3 6 ) , the presupposition i s that i t didn't r a i n u n t i l a f t e r the lapse of three weeks, while i n another reading, the presupposition i s that i t rained but that i t did so for less than three weeks. As f o r (37). one reading presupposes that Hanako didn't study at a l l dur-ing a period of ten days, while the other presupposes that she studied but for less than ten days. In the case of (38), one presupposition i s that f o r a period of twelve hours he didn't walk at a l l , while another presupposes that he walked but he didn't do so continuously for twelve hours. Applying our three-part in t e r p r e t a t i o n to ( 3 6 ) , the former reading w i l l have the i n t e r p r e t a t i o n of (39) and the l a t t e r ( 4 0 ) . 148 (39) a. Ame ga huranakatta no wa nan-nitikan ka datta. • I t was for a c e r t a i n number of days that i t didn't r a i n . 1 b. Ame ga huranakatta no wa sansyuu-kan mata sore izyoo de wa nakatta. ' I t was not f o r three weeks or more that i t didn't r a i n . ' c. Ame ga huranakatta no wa sansyuu-kan datta. • I t was for three weeks that i t didn't r a i n . * (40) a. Ame ga hutta no wa nan-nitikan ka datta. • I t was for a certain number of days that i t rained.' b. Ame ga hutta no wa sansyuu-kan mata wa sore izyoo datta. ' I t was f o r three weeks and more that i t rained.' c. Ame ga hutta no wa sansyuu-kan de wa nakatta. • I t was not for three weeks that i t rained.' Again, notice the positive-negative r e l a t i o n s h i p between the assertion and the expectation i n the in t e r p r e t a t i o n . The underlying structure f or the int e r p r e t a t i o n of (39) would be s i m i l a r to that of (35)• (40) however, would have the following underlying s t r u c t u r e i 149 (41) NP VP NP VP NP VP Ame hutta Notice that i n (41), the positive-negative r e l a t i o n s h i p between and S 2 s t i l l holds. It can be observed that the difference between the underlying structures of (35) and (41) i s the difference i n the scope of the negative. In the case of (35)» the lowest S comes within the scope of the negative, with the adverbial introduced as a VP of an S higher than the negative. On the other hand, i n (41) the adverbial i s introduced below the negative, :andVthe negative negates the whole S„. (y I t seems that mo, when used i n negative sentences containing q u a n t i f i e r s may emphasize the maximum or the minimum value depending on the presuppositions underlying 150 them. Tor example, the (39) i n t e r p r e t a t i o n of sentence (36) emphasizes the maximum value, that i s , i t emphasizes the fact that the r e a l number of days that i t rained i s greater than that expected. The (40) i n t e r p r e t a t i o n , on the other hand, emphasizes the minimum value, that i s , i t connotes that the r e a l number of days that i t rained i s smaller than expected. In the case of sentences l i k e ( 2 9 ) , the use of mo can only emphasize the maximum value and not the minimum value. This observation shows that f o r negative sentences, whether mo emphasizes the maximum or the minimum value i s dependent on the kind of verbs and adverbs involved. I f the negative sentence contains a time adverbial co-occurring with a semantically durative verb, mo may emphasize either the maximum or the minimum value. However, i f i t occurs with a semantically punctual verb, then mo can only emphasize the maximum value. With punctual verbs mo cannot emphasize the minimum value because the i n (41) cannot take an ungrammatical positive sentence. On the other hand, with positive sentences, such as examples (15)-(19.), mo can only emphasize the maximum value. Let us now examine some constructions containing q u a n t i f i e r s other than those denoting time. Consider the following sentences! (42) Okyaku-san wa zyuu-nin mo konakatta. 'Not even ten guests came.' 151 (4-3) Okyaku-san wa zyuu-nin mo k i t a . 'Even ten guests came.' (44) B i i r u wa rop-pon mo nomanakatta. '(I) didn't even drink s i x bottles of beer.' (45) B i i r u wa rop-pon mo nonda. '(I) even drank s i x bottles of beer.' In the above examples, i t can be noticed that (42) and (44), which are both negative sentences, can be interpreted i n at least two ways. One reading of (42) implies that only nine guests came and not ten, while the other reading implies that a l l of the ten guests did not come. One reading f o r (44) implies that I drank only f i v e b ottles of beer and not s i x , while the other reading implies that there are s i x bottles of beer that I did not drink ( i . e . there are s i x bottles of beer l e f t over). The former interpretation of (42) and (44) emphasizes the minimum value and w i l l have s i m i l a r underlying structures as (41). The l a t t e r i n t e r p r e t a t i o n emphasizes the maximum value and t h e i r underlying structures w i l l be s i m i l a r to that of (35). The positive counterparts of (43) and (45) are not ambiguous and the mo can only be used to emphasize the maximum value. (43) can only imply that as many as ten guests came, and (45) can only imply that I drank as many as s i x bottles of beer. The underlying structure f o r both w i l l be s i m i l a r to (35). 152 In Japanese, q u a n t i f i e r s which represent minimal amounts very often occur with mo, such as sukosi mo "not at a l l * , tittomo 'not at a l l ' , h i t o r i mo 'not a single person', or h i t o t u mo 'not a single one'. When co-occurring with mo, these q u a n t i f i e r s always require the presence of the negative. Such q u a n t i f i e r s imply that the quantity does not exceed the minimum amount, i n other words, the quantity zero. Consider the following examples: (46) Koogi n i wa h i t o r i mo konakatta. 'Not a single person came to the lecture.* (4-7) Benkyoo wa tittomo s l n a i . '(I) didn't do (my) studies at a l l . ' (48) Sensei no i u koto wa sukosi mo wakarimasen. '(I) don't understand at a l l what the teacher says.' (49) 0-susi wa h i t o t u mo tabenakatta. '(I) didn't eat a single s u s i . ' The use of mo i n the above cases, can only emphasize the minimum value. In (46) f o r example, the sentence expresses that no one came to the lecture. This i s an entailment from the presupposition that whatever the s i t u a t i o n may be, at least one person would come to the lecture. The underlying structure would be s i m i l a r to that of (41). Indefinite pronouns i n Japanese can very often occur 153 with mo, as i n the following sentences: (50) Tabako wa nan-bon mo nomanakatta. •(I) didn't smoke many cigarettes.' (51) Tabako wa nan-bon mo nonda. '(I) smoked many cigarettes.' (52) Okyaku-san wa nan-nin mo konakatta. 'Many guests didn't come.* (53) Okyaku-san wa nan-nin mo k i t a . 'Many guests came.' The negative sentences of (50) and (52) are ambiguous. For example, one reading of (50) can imply that the number of cigarettes that I smoked are not many, while another reading implies that I didn't smoke many of the cigarettes ( i . e . there are many cigarettes l e f t over). The former interpre-t a t i o n i s c l e a r l y an emphasis on the maximum value while the l a t t e r , the minimum value. However, note that the positive sentences of (51) and (53) can have only one i n t e r p r e t a t i o n , that which emphasizes the maximum value. However, i t should be pointed out that there are some exceptions to the above generalization-about i n d e f i n i t e pronouns. I t seems that i n d e f i n i t e pronouns l i k e ikura 'how many/how much', dare 'who' and doko 'where', when occur-r i n g with mo i n negative sentences are not at a l l ambiguous and have only the interpretation of the minimum value emphasis, 154 as shown i n the following examplest (54) Okane wa ikura mo nokotte imasen. *(I) have not much money l e f t . ' (55) Setumei wa dare mo kikoenakatta. •Nobody heard the explanation.' (56) N i t i y o o b i wa doko e mo ikanakatta. •(I) didn't go anywhere on Sunday.* I f we follow the approach set up i n section 4.2 .1 . , then (55)» for example, would be interpreted as Setumei wa dare  ka n i kikoeta to omou ga, dare n i mo kikoenakatta '(I) thought at least someone heard the explanation, but nobody heard i t ' , where the mo emphasizes only the minimum value. 4 . 2 . 3 . GONCESSIVE^SENTENCES AND MO In t h i s section, I w i l l deal b r i e f l y on how concessive sentences can be interpreted i n terms of presupposition, and how the three parts of the interpretation of a sentence with mo w i l l apply to concessive sentences. In addition to specifying the meaning of membership and emphasis, the use of mo can also specify concession. Observe the following sentencesi 155 (57) Benkyoo s i t e mo sotugyoo dekinai. 'Even i f (I) study, (I) can't graduate.' (58) Warukuti o i t t e mo okoranai. 'Even i f (I) c a l l (him) names, (he) doesn't get angry.* I f we apply our three-part interpretation of mo, following the approach discussed i n section 4 . 2 . 1 . , then we would analyze the interpretation of (57) and (58) as (59) and ( 6 0 ) . (59) a. Benkyoo sureba sotugyoo dekiru. ' I f (I) study, (I) can graduate.' b. Onazi yoona koto o sureba sotugyoo dekiru. 'If (I) do s i m i l a r things, (I) can graduate.' c. Benkyoo sureba sotugyoo dekiru no de wa n a i . 'I t i s not so that i f (I) study, (I) can graduate.' (60) a. Warukuti o ieba, okoru. 'I f (I) c a l l (him) names, (he) gets angry.' b. Onazi yoona koto o ieba, okoru. •I f (I) say s i m i l a r things, (he) gets angry.' c. Warukuti o ieba, okoru no de wa n a i . 'It i s not so that i f (I) c a l l (him) names, (he) gets angry.' Notice that a l l three parts of (59) and ( 6 0 ) are necessary 156 for the semantic int e r p r e t a t i o n of concessive sentences such as (57) and ( 5 8 ) . A l l three parts of ( 5 9 ) , when taken together, enables the correct interpretation that I w i l l not be able to graduate no matter how hard I study, or no matter what I do. (60) enables the correct i n t e r p r e t a t i o n that, no matter what I say, he doesn't get angry. This i s exactly the interpretation of such concessive sentences. The underlying structure f o r (57) w i l l be something l i k e (61). With the exception of l e x i c a l items, (58) w i l l also have si m i l a r underlying structure. (61) benkyoo sureba onazi yoona sotugyoo dekiru koto sureba sotugyoo dekiru 4.3. ON THE INTERPRETATION OF WA AND GA In the previous sections, I have attempted to present a general schema for the interpretation of mo. I have also 157 examined the presuppositional properties of mo, the presence of which are relevant f o r the correct semantic i n t e r p r e t a t i o n of the surface structure. In t h i s section, I w i l l attempt to determine whether the l i n e of thinking that we have adopted for the in t e r p r e t a t i o n of mo can be applied to other p a r t i c l e s such as wa and ga. Let us begin by examining the following sentence t (62) Taroo h i t Hanako. I t i s clear that a sentence l i k e (62) i s at least two ways ambiguous, corresponding to two d i f f e r e n t deep structures. (62) may be an answer to either (63) or (64). (63) a. Who h i t Hanako? b. Taroo h i t Hanako. (64) a. Whom did Taroo h i t ? b. Taroo h i t Hanako. Notice that i n ( 6 3 ) , i t i s presupposed that somebody h i t Hanako, while i n (64) i t i s presupposed that Taroo h i t somebody. Notice also that the ambiguity associated with (62) can be r e f l e c t e d i n two d i s t i n c t stress patterns, which are shown i n (63b) and (64b), and which occur i n d i s t i n c t contexts. I t i s exactly t h i s kind of presupposition that i s important for the difference between the p a r t i c l e s mo, wa and 158 ga i n Japanese. Consider the following sentences: (65) Hanako mo utatta. 'Hanako sang too/also.' (66) Hanako wa utatta. 'Hanako sang.* (67) Hanako ga utatta. 'Hanako sang ( i . e . I t was Hanako who sang).' We have observed e a r l i e r that f or a sentence l i k e (65)» we can have the following interpretation: ( 6 5 ' ) a. X (Y,Z) ga utatta. •X (Y,Z) sang.' b. Hanako ga utatta. •Hanako sang.' which can be s t r u c t u r a l l y represented as ( 6 8 ) : (68) Prsp X (Y,Z) utatta Hanako utatta x v y w 159 The information shown i n (68) can be represented by the following formulai I f , Then, According to ( 6 8 ) , the underlying structure consists of two sentences and S 2, where Sg presupposes S^. In the process of transformation, Prsp and are consequently deleted and the NP Hanako has the p a r t i c l e mo attached to i t , to derive sentence ( 6 5 ) . Notice that i n ( 6 5 ' ) , i t i s presupposed that a group of people X,Y and Z sang, and that Hanako i s viewed as a member of the group that sang. This presupposition underlies the use of mo i n sentence ( 6 5 ) , and i s responsible for the correct semantic interpretation. 1 Let us now turn to the question of whether the interpre-t a t i o n adopted f o r mo can be applied to wa i n (66) and ga i n ( 6 7 ) . F i r s t , l e t us consider sentence ( 6 6 ) . I t seems that sentence (66) answers the question ( 6 9 ) . (69) a. Hanako ga nani o s i t a ka. •What did Hanako do?' b. Hanako ga utatta. 'Hanako sang.' v = w x / y y > y + mo Prsp, S 1 >^ 160 This observation shows that f o r a sentence l i k e ( 6 6 ) , i t i s presupposed that Hanako did something and what she did was that she sang. On the other hand, (67) can be considered to be an answer to question ( 7 0 ) . (70) a. Dare ga utatta ka. 'Who sang?' b. Hanako ga utat t a . •Hanako sang.* I t i s evident from (70) that what i s presupposed i n (67) i s that someone sang, and that someone who sang was Hanako. Notice the di f f e r e n t presuppositions underlying (66) and ( 6 7 ) . In terms of the approach that we have adopted f o r the interp r e t a t i o n of mo, we would analyze the int e r p r e t a t i o n of sentences (66) and (67) to be those of (66*) and (67*) respectively. ( 6 6 ' ) a. Hanako ga nani ka 0 s i t a . •Hanako did something.' b. Hanako ga utat t a . 'Hanako sang.' (67 ' ) a. Dare ka ga utatta. 'Someone sang.' b. Hanako ga utatta. 'Hanako sang.' 161 Observe that ( 6 6 ' ) and (67 ' ) are exactly the interpretations of such sentences as (66) and ( 6 7 ) . The parts of the inter p r e t a t i o n i n ( 6 6 ' ) , when taken together, permit the correct inference that Hanako did something and what she did was that she sang. On the other hand, the parts of the inter p r e t a t i o n i n (67 ' ) together, permit the correct inference that someone sang and that someone who sang was Hanako. This c l e a r l y shows that the interpretation f o r wa and ga can be derived i n very much the same way as we derive the interpre-t a t i o n of mo. ( 6 6 ' ) can be represented by the following underlying structure 1 (71) Prsp Hanako nani ka s i t a Hanako utatta The information contained i n (71) can be represented by the following formulas 162 x = y I f , v and w = Predicate (and v has the feature of (+ I n d e f i n i t e j ) Then, y >y + wa Prsp, S-j^  Notice that i n (71) , the NP Hanako i s i n the deep structure of S^. According to (71) , S 2 presupposes S^, which, i n the course of transformation, w i l l be consequently deleted together with Prsp. The p a r t i c l e wa w i l l be inserted a f t e r the NP Hanako of S 2, which w i l l surface to derive sentence ( 6 6 ) . ( 6 7 ' ) can be s t r u c t u r a l l y represented as ( 7 2 ) . (72) Dare ka utatta Hanako utatta V — v » < 1 „ ' * y- 1 x v y w The information shown i n (72) can be represented by the following formula: 163 I f , w and x, y = NP jg (and x has the feature of [+ Indefinite] ) Then, y >y + ga Prsp, S 1 The whole structure of (72) i s considered to underlie sentence (67) . (72) consists of two sentences and S2» where S 2 presupposes S^. In the course of i t s derivation, Prsp and are consequently deleted, and the p a r t i c l e ga i s attached to the NP Hanako of S 2 to produce sentence ( 6 7 ) . The semantic int e r p r e t a t i o n of sentences containing the contrastive wa can be accounted f o r i n a s i m i l a r way by incorporating presuppositions i n the analysis. Consider the following sentences: (73) Yamada-san wa ikanakatta.^ 'Mr. Yamada didn't go.' (74) Ame wa hutte imasen. 'I t i s not r a i n i n g . ' (73) can be considered to be an answer to (75)• (75) a. Yamada-san ga i t t a ka. •Did Mr. Yamada go?* b. l i e , Yamada-san wa ikanakatta. 164 'No, Mr. Yamada didn't go.' The answer implies that the speaker i s t r y i n g to convey the notion that someone else went but, unfortunately Mr. Yamada didn't go. In other words, the int e r p r e t a t i o n would be something l i k e s (73 ' ) Yamada-san wa ikanakatta ga, Tanaka-san ga i t t a . 'Mr. Yamada didn't go but Mr. Tanaka went.* S i m i l a r l y , (74) could be an answer to question ( 7 6 ) . (76) a. Ima, ame ga hutte imasu ka. *Is i t rain i n g now?' b. l i e , ame wa hutte imasen. 'No, i t i s not ra i n i n g . ' In answering (76b) , the speaker i s t r y i n g to imply that something else i s happening, such as ' i t i s snowing'. The interpretation could be something l i k e s (74 ' ) Ame wa hutte imasen ga, yuki ga hutte imasu. 'I t i s not r a i n i n g , but i t i s snowing.' In terms of the l i n e of approach that we have adopted, sentence (73) f o r example, i s assumed to have the following 165 underlying structure: (77) Prsp Tanaka-san i t t a v X NEG Yamada-san i t t a w In (77), S 2 presupposes S^. Notice also the positive-negative r e l a t i o n s h i p between the presupposition and the assertion S 2 i n the underlying structure. In the course of i t s deriv-ation, the contrastive wa w i l l be inserted immediately a f t e r the constituent ( i n t h i s case, Yamada-san) only i f the condition |kNEGJ s and ^-OCNEG J g holds. Prsp and S 1 w i l l be consequently deleted. This can be represented by the following formula: I f , Then, v = w x / y y ^ y + wa c o n t r a s t i v e j , only i f [oCNEsJg and [-OCNEGJ-s holds. Prsp, 166 I t seems that the semantic d i s t i n c t i o n s between the sentences containing mo, wa and ga are b a s i c a l l y due to the d i f f e r e n t presuppositions underlying those sentences. The observations above seem to show that i t i s possible to apply the same l i n e of approach adopted f o r the i n t e r p r e t a t i o n of mo to other p a r t i c l e s such as wa and ga. 4.4. CONCLUSION This chapter i s b a s i c a l l y a descriptive one, attempting to characterize the general nature of the p a r t i c l e mo i n Japanese, and at the same time, attempting to present a general schema for the i n t e r p r e t a t i o n of mo. The presuppos-i t i o n a l properties that are associated with mo have been examined, and i t has been pointed out that i n c e r t a i n cases mo involves the notion of expectation. I t i s exactly t h i s notion of expectation that i s responsible for the unexpected-ness or surprise that usually accompanies mo., I t seems that when mo involves the notion of expectation, then mo has the same meaning as that of even i n English. In conclusion, I would l i k e to summarize what I believe to be the main points of my discussion, ( i ) The p a r t i c l e mo i n Japanese can be used to specify the meaning of membership within a set, or of emphasis, or of concession. 167 ( i i ) In the case of mo specifying the meaning of membership within a set, the inter p r e t a t i o n can be analyzed into at least two parts, while i n the case of mo specifying emphasis or concession, the information can be analyzed into at least three parts. This two-part or three-part inte r p r e t a t i o n underlies the d i f f e r e n t uses of mo. The parts of the interpretation, when taken together, permit the correct semantic i n t e r p r e t -ation on the surface structure. ( i i i ) I t has been observed that with regards to mo specifying emphasis or concession, there c l e a r l y e x i s t s a positive-negative r e l a t i o n s h i p between the assertion and the expectation i n the under-l y i n g structure. (iv) When occurring with q u a n t i f i e r s , the p a r t i c l e mo can emphasize either the maximum or the minimum value depending on the kind of verb, as well as on the positive or negative form of the verb i n r e l a t i o n to the q u a n t i f i e r s . However, there are cer t a i n cases where the emphasis on the maximum or the minimum value i s dependent on only the presuppositions irrespective of the verb forms. (v) With respect to negative sentences, the mo can emphasize only the maximum value when i t occurs with time adverbials and semantically punctual 168 verbs. However, when- i t occurs with time adver-b i a l s and semantically durative verbs, mo can emphasize either the maximum or the minimum value. I t has been observed that t h i s peculiar-i t y i s related to the scope of the negative i n the deep structure. (vi) As regards positive sentences, we see that mo can only specify the maximum value or quantity. ( v i i ) I t has been observed that i f we follow the same l i n e of approach that we have adopted f o r the interpretation of mo, then we w i l l f i n d that the same l i n e of approach can also be applied to other p a r t i c l e s such as wa and ga. 169 FOOTNOTES 1. Soga, Matsuo. 1975* "Kakari Zyosi 'Mo' no Koozoo n i t u i t e no Iti-Koosatu". (A Study on the Structure of the P a r t i c l e Mo). To appear i n Nihongo Kyooiku (Japanese Language Education). 2. Kageyama, Taroo. 1973. "On the Generation of Mo". Papers i n Japanese L i n g u i s t i c s . Vol. 2, No. 2. University of Southern C a l i f o r n i a . 3. For a discussion oh the nature of punctual and durative verbs, r e f e r back to Chapter One, Section 1.3.2.3. 4. I t has been pointed out i n Chapter One, that adverb-i a l s are generated i n the base as verb-phrases of 'higher' simplex sentences than the s u p e r f i c i a l main clauses i n the surface structure. In sentences containing both negatives and adverbials (such as Time, Locative or Instrumental adverbials);, the understood order of the adverbials and the negatives i n the underlying structures corresponds to the hierarchy of upper sentences containing those adverbials and negatives. For a detailed discussion, r e f e r back to Chapter One, Section 1.3.2. 5. Yamada-san wa ikanakatta. -"'Mr. Yamada didn't go.* The above sentence i s actually ambiguous between the two readings of was (a) Thematic wa s 'Talking about Mr. Yamada, he didn't go.* (b) Contrastive wa s 'Mr. Yamada didn't go, (but Mr. Tanaka went).' I t seems that i n the actual conversation, (a) i s very often distinguished from (b) by the emphatic stress on the contrastive wa. Thus, i n the discussion, the stress marker (/) w i l l be used to mark contrastive wa. 170 CHAPTER FIVE  CONCLUSION What has been accomplished i n t h i s study i s a provision of more insights into the problems of negation, a complex area of grammar which has, u n t i l recently, been r e l a t i v e l y neglected i n the study of Japanese L i n g u i s t i c s . I t i s hoped that the general observations and evidence to be drawn from t h i s thesis w i l l provide a stepping stone towards future research into the area of negation i n Japanese. A large part of the discussion i n t h i s thesis has been based on the findings i n the area of negation i n English. In so doing, i t i s hoped that the findings and evidence obtained from t h i s study may i n turn, throw l i g h t to the problems of negation i n Japanese, and perhaps serve as c r o s s - l i n g u i s t i c evidences i n support of the analyses and explications that have been presented for negation i n English. This study on negation i n Chapter One has attempted to determine the structures f o r negative constructions, and where exactly the constituent NEG should be introduced i n the deep structure. S y n t a c t i c a l l y , negation has been cons-idered as a rather simple process of attaching a negative 171 morpheme na to the verb stem. For example, Inoue i n her paper A Study of Japanese Syntax (1964), derives negative sentences by a rule of optional transformation of the f o l l -owing kinds X - NP + wa ga ga AB - Y - | Vm |+ T - Z m * 5 1 + [=]- 2 - 3 + ana + 4 - 5 However, i t has been pointed out that t h i s transformation f a i l s to account for things l i k e the scope of the negative and what other changes occur when sentences are negated. In our analysis, the sentence-final negative formative na i n Japanese i s derived from a single underlying predicate NEG, which i s introduced optionally i n the underlying structure as a verb of the "higher" sentence. The differences i n the scope of the negative i s accounted f o r i n terms of higher predicates and the r e l a t i v e heights of these predicates. For example, as discussed i n section 1 . 3 . 2 . , the s y n t a c t i c a l ambiguity of the negative sentences containing adverbials (such as Time, Locative and Instrumental Adverbials) are accounted for by the r e l a t i v e heights of two predicates, the NEG and the Adverbial. 172 Various other aspects of negation i n Japanese were examined i n Chapters Two, Three and Four. The a p p l i c a b i l i t y of the negative transportation rule to Japanese has been examined i n Chapter Two. The evidence drawn from the discussion, show c l e a r l y that we cannot r e l y on the Simplex Sentence Condition and the Confirmatory Question Formation as arguments for the support"of the negative transportation rule i n Japanese. Furthermore, counterexamples have been ci t e d which greatly weaken the j u s t i f i c a t i o n f o r the e x i s t -ence of the negative transportation rule i n Japanese. I t has been observed i n Chapter Three that negative questions i n Japanese are s y n t a c t i c a l l y ambiguous and can e l i c i t peculiar responses represented by hai *yes' and i i e •no!. The responses cannot be predicted from the question utterances alone because of t h e i r semantic and syntactic ambiguity, and therefore we have resorted to the e x t r a - l i n g -u i s t i c information of presuppositions as explanations. Like-wise, the interpretation and the generation of p a r t i c l e s such as mo, wa and ga, as discussed i n Chapter Four, were accounted for by incorporating presuppositions into the analysis. Several i n t e r e s t i n g problems have been l e f t open for future research. Some of these are, the p e c u l i a r i t y that i s associated with reason adverbials co-occurring with the negative i n a sentence, as discussed i n 1 . 3 . 2 . 2 .s the problems associated with si k a lowering and i t s subsequent attachment to noun phrases of lower sentences, as discussed i n 2.3.3.5 1 7 3 and f i n a l l y , the surface phonological feature of intonation, discussed i n 3 . 2 . 5 . i which i s considered to mark the d i s t i n c -t i o n i n the syntactic ambiguity of the negative questions. So f a r i n t h i s t h e s i s , I have not been able to account f o r the above mentioned problems s a t i s f a c t o r i l y . Perhaps future study can deal with these problems i n a f u l l e r and more s a t i s f y i n g way. Furthermore, an investigation of a f f i x a l negation i n Japanese might be a valuable future study. In Japanese, a f f i x a l negation i s formed by the use of negative prefixes such as hu-, h i - , mu-, mi-, and bu-. The following sentences are examples of a f f i x a l negation. (a) Tikara no i r u sigoto wa onna no h i t o n i wa hutekitoo da. •Jobs that require strength are unsuitable f or women.' (b) Sore to kore to wa mukankei da. 'This and that are unrelated.' A f f i x a l negation, though a very i n t e r e s t i n g topic by i t s e l f , i s beyond the scope of t h i s study, and therefore, has been l e f t open f o r future research. 174 BIBLIOGRAPHY Alfonso, Anthony. 1966. Japanese Language Patterns, Vols. 1 and 2. Sophia University LTL. Center of Applied L i n g u i s t i c s . Tokyo. Anderson, S. 1972 . "How to Get Even". Language. Vol. 48. (pp. 8 9 3 - 9 0 6 ) . Austin, John L. 1 9 6 2 . How to Do Things with Words. Harvard University Press. Cambridge, Massachusetts. Carden, Guy. 1968. "English Quantifiers". Mathematical  L i n g u i s t i c s and Automatic Translation. NSF-20. Harvard University. Cambridge, Massachusetts. Chomsky, Noam. 1957- Syntactic Structures. Mouton & Co. 'S-Gravenhage. I 9 6 5 . Aspects of the Theory of Syntax. m T*N ' ' r*' \_' • J tin ^1 l i _ J The M.I.T. Press. Cambridge, Massachusetts. Fillmore, Charles J . 1963. "The Pos i t i o n of Embedding Transformations i n a Grammar". Word. Vol. 19. Journal of the L i n g u i s t i c C i r c l e of New York. N.Y. (pp. 208-231). Fraser, Bruce. 1 9 6 9 . "An Analysis of Concessive Conditionals". Papers from the F i f t h Regional Meeting of the Chicago  L i n g u i s t i c Society. University of Chicago. I l l i n o i s , (pp. 66-75). 1971. "An Analysis of 'Even' i n English". Studies i n L i n g u i s t i c Semantics. Edited by Fillmore & Lahgendoen,. Holt, Rhinehart & Winston, (pp. 150-178). Garner, Richard. 1971. "'Presupposition' i n Philosophy and L i n g u i s t i c s " . Studies i n L i n g u i s t i c Semantics. Edited by Fillmore & Langendoen. Holt, Rhinehart & Winston. (pp. 23-42). Hojo, Michio. 1974. "Negative Questions i n Japanese". Department of East Asian Languages and Liter a t u r e . University of Pittsburgh. (M.S.). 175 Horn, Laurence R. I 9 6 9 . "A Presuppositional Analysis of Only and Even". Papers from the F i f t h Regional  Meeting of the Chicago L i n g u i s t i c Society. University of Chicago. I l l i n o i s . (pp. 98-107). 1971. "Negative Transportation: Unsafe at Any Speed?" Papers from the Seventh Regional  Meeting of the Chicago L i n g u i s t i c Society. University of Chicago. I l l i n o i s , (pp. 1 2 0 - 1 3 3 ) . Inoue, Kazuko. 1969. A Study of Japanese Syntax. Mouton & Co. The Hague, Paris. Jackendoff, Ray S. 1968. "Quantifiers i n English". Foundations of Language. Vol. 4 . (pp. 422-442). . 1969. "An Interpretive Theory of Negation". Foundations of Language. Vol. 5 « (pp. 2 1 8 - 2 4 1 7 ; . 1 9 7 1 . "On Some Questionable Arguments About Quantifiers and Negation". Language. Vol. 4 7 , No. 2 . (pp. 2 8 2 - 2 9 7 ) . . 1 9 7 2 . Semantic Interpretation i n Genera- ti v e Grammar. The M.I.T. Press. Cambridge, Massachusetts. Kageyama, Taroo. 1 9 7 3 . "On the Generation of Japanese mo". Papers i n Japanese L i n g u i s t i c s . Vol. 2 , No. 2 . University of Southern C a l i f o r n i a , (pp. 4 4 - 7 0 ) . Karttunen, Lauri. 1 9 7 3 . "Presupposition of Compound Sentences". L i n g u i s t i c Inquiry. Vol. 4 , No. 2 . The M.I.T. Press. Cambridge, Massachusetts, (pp. 1 6 9 - 1 9 3 ) . Keenan, Edward L. 1 9 7 1 . "Two Kinds of Presupposition i n Natural Languages". Studies i n L i n g u i s t i c Semantics. Edited by Fillmore & Langendoen. Holt, Rhmehart & Winston, (pp. 4 5 - 5 2 ) . Klima, Edward S. 1 9 6 4 . "Negation i n English". The Structure  of Language: Readings i n the Philosophy of Language. Edited by J.A. Fodor & J.J . Katz. M.I.T. Prentice-H a l l . Englewood C l i f f s , New Jersey. (pp. 2 4 6 - 3 2 3 ) . Kuno, Susumu. 1 9 7 3 . The Structure of the Japanese Language The M.I.T.Press. Cambridge, Massachusetts & London, England. 0 176 Lakoff, George. 1965. "On the Nature of Syntactic I r r e g u l a r i t y " . Mathematical L i n g u i s t i c s and  Automatic Translation. NSF-16. Harvard University. Cambridge, Massachusetts. . 1966a. " C r i t e r i o n f o r Verb Phrase Consti-tuency ". Mathematical L i n g u i s t i c s and Automatic  Translation. NSF-17. Harvard University. Cambridge, Massachusetts. . 1966b. "A Note on Negation". Mathematical L i n g u i s t i c s and Automatic Translation. NSF-17. Harvard University. Cambridge, Massachusetts. . 1970a. "Pronominalization, Negation and the Analysis of Adverbs". Readings i n English  Transformational Grammar. Edited by R. Jacobs & P.S. Rosenbaum. Boston. Gi n n - B l a i s d e l l . (pp. 145-165) . . 1970b. "Repartee, or a Reply to 'Negation, Conjunction, and Quantifiers". Foundations of Language. Vol. 6. (pp. 389-422J. . 1971. "On Generative Semantics". Semantics» An I n t e r d i s c i p l i n a r y Reader i n Philosophy, L i n g u i s t i c s  and Psychology. Edited by D.D. Steinberg & L.A. Jakobovits. Cambridge at the University Press, (pp. 232-296) . Lakoff, Robin. 1969. "A Syntactic Argument f o r Negative Transportation". Papers from the F i f t h Regional  Meeting of the Chicago L i n g u i s t i c Society. University of Chicago, I l l i n o i s , (pp. 140-14-7). j . 1971. " I f ' s , And's, and But's About Conjunc-tion*'. Studies i n L i n g u i s t i c Semantics. Edited by Fillmore & Langendoen. Holt, Rhinehart & Winston, (pp. 115-149). Langacker, Ronald W. 1969. "On Pronominalization and the Chain of Command". Modern Studies i n English  Readings i n Transformational Grammar. Edited by David A. Reibel & Sanford A. Schane. Prentice-Hall, Inc., Englewood C l i f f s , New Jersey, (pp. 160-186). Lawler, John M. 1971. "Any Questions?" Papers from the Seventh Regional Meeting of the Chicago L i n g u i s t i c  Society. University of Chicago. I l l i n o i s , (pp. 163-173). 1 7 7 Lindholm, James. 1 9 6 9 . "Negative-Raising and Sentence Pronominalization". Papers from the F i f t h Regional  Meeting of the Chicago L i n g u i s t i c Society. University of Chicago, I l l i n o i s , (pp. 1 4 8 - 1 5 8 ) . McGloin, Naomi H. 1972. Some Aspects of Negation i n Japanese. University of Michigan, Ph. D. Dissert-ation. M i y a j i , Yutaka. 1 9 7 1 . " H i t e i Hyoogen (Negative Expressions)". Bunron - Gendai-go no Bumpoo to Hyoogen no Kenkyuu  ( 1 ) . (A Study of Sentences - A Study on the Grammar and Expressions of Modern Japanese). M e i j i Shoo-in. (pp. 182-210). Muraki, Masatake. 1 9 6 5 . "Negation i n English and Japanese". Studies i n Descriptive and Applied L i n g u i s t i c s . B u l l e t i n of the Summer In s t i t u t e i n L i n g u i s t i c s , Vol. 3 « International C h r i s t i a n University. Tokyo, Japan. (pp. 8 0 - 9 3 ) . 1970. Presupposition, Pseudo-clefting and Thematization. The University of Texas at Austin, Ph. D. Dissertation. Partee, Barbara H a l l . 1 9 7 0 . "Negation, Conjunction, and Quantifiers : Syntax vs. Semantics". Foundations  of Language. Vol. 6 . (pp. 1 5 3 - 1 6 5 ) . Ross, John Robert. 1 9 7 0 . "On Declarative Sentences". Readings i n English Transformational Grammar. Edited by R. Jacobs & P.S. Rosenbaum. Ginn and Company. Massachusetts, Toronto, London. Shibatani, Masayoshi. 1 9 7 2 . "Negative Questions and Conveyed Meaning". Papers i n Japanese L i n g u i s t i c s . Vol. 1 , No. 2 . Japanese L i n g u i s t i c s Workshop. University of C a l i f o r n i a , Berkeley, (pp. 3 2 I - 3 2 3 ) . Smith, Steven B. 1 9 6 9 . "The Semantics of Negation (Prelimin-ary D r a f t ) " . Reproduced by the Indiana L i n g u i s t i c s C i r c l e . (M.S.). Soga, Matsuo. 1 9 6 6 . Some Syntactic Rules of Modern Collo- q u i a l Japanese"] The University of Indiana, Ph. D. Dissertation. 1 9 7 2 . "Negative Transportation and Cross-L i n g u i s t i c Negative Evidence". Papers i n Japanese  L i n g u i s t i c s . Vol. 1 , No. 1 . Japanese L i n g u i s t i c s 178 Workshop. University of C a l i f o r n i a , Berkeley, (pp. 103-119). . 1973. "Yes-No Reconsidered". Papers i n Japanese L i n g u i s t i c s . Vol. 2, No. 1. Japanese L i n g u i s t i c s Workshop. University of C a l i f o r n i a , Berkeley. (pp. 148-154). __. 1975. "Kakari Zyosi *Mo' no Koozoo n i t u i t e no Iti-Koosatu. (A Study on the Structure of the P a r t i c l e mo)". To appear i n Nihongo Kyooiku (Japanese Language Education). Stockwell, Robert P.} Schachter, Paul? Partee, Barbara H a l l . 1973. The Major Syntactic Structures of English. Holt, Rhinehart & Winston, Inc., New York. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0093443/manifest

Comment

Related Items