UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Study of Japanese relativization Ogawa, Toshimitsu Augustine 1974

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

Item Metadata

Download

Media
831-UBC_1974_A8 O39_3.pdf [ 6.4MB ]
Metadata
JSON: 831-1.0093228.json
JSON-LD: 831-1.0093228-ld.json
RDF/XML (Pretty): 831-1.0093228-rdf.xml
RDF/JSON: 831-1.0093228-rdf.json
Turtle: 831-1.0093228-turtle.txt
N-Triples: 831-1.0093228-rdf-ntriples.txt
Original Record: 831-1.0093228-source.json
Full Text
831-1.0093228-fulltext.txt
Citation
831-1.0093228.ris

Full Text

A STUDY OF JAPANESE RELATIVIZATION by TOSHIMITSU AUGUSTINE OGAWA B.A., Nanzan University, Japan, 1955 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n the Department of LINGUISTICS We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA February, 197*4-In p resen t ing t h i s thes is in p a r t i a l f u l f i l m e n t o f the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y sha l l make i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree t h a t permission for ex tens ive copying o f t h i s t hes i s for s c h o l a r l y purposes may be granted by the Head of my Department or by h is r e p r e s e n t a t i v e s . I t is understood that copying or p u b l i c a t i o n o f t h i s t hes i s for f i n a n c i a l gain sha l l not be al lowed wi thout my w r i t t e n permiss ion . Department of The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8, Canada Date i ABSTRACT A Japanese r e l a t i v e construction i s d i f f e r e n t i n some respects from that of English. This thesis i s an attempt to investigate the nature of Japanese r e l a t i v i z a -t i o n i n contrast to that of English r e l a t i v i z a t i o n . It deals with some hypotheses concerning Japanese r e l a t i v i z a -t i o n . and i t s related problems. Through the examination of those hypotheses, i t i s hoped that the nature of Japanese r e l a t i v i z a t i o n w i l l be c l a r i f i e d . The analysis i n t h i s thesis i s within the framework of a transformational generative grammar, which can be considered as a modified version of the theory proposed i n Aspects of the Theory of Syntax, but which i s not within the framework of so-called generateve semantics. It i s assumed i n t h i s thesis that every acceptable sentence has i t s deep structure and i t s surface structure, which are related by some appropriate transformations. There are seven chapters i n t h i s t h e s i s . Chapter I presents the c h a r a c t e r i s t i c s of Japanese thematization i i d iscussing the hypotheses about the underlying structure of a thematized sentence. The reason f o r t h i s chapter i s to investigate the hypothesis that r e l a t i v i z a t i o n i s c l o s e l y r e l a t e d to thematization. Chapter II presents a contrastive study of r e l a t i v -i z a t i o n between English and Japanese. It attempts to put the c h a r a c t e r i s t i c s of Japanese r e l a t i v i z a t i o n into r e l i e f . In t h i s chapter, a claim i s also examined that r e l a t i v -i z a t i o n i s a movement transformation. Further, Ross's Complex NP Constraint and Coordinate Structure Constraint are investigated i n connection with Japanese r e l a t i v i z a t i o n . Chapter III i s concerned with a copying r e l a t i v i z a -t i o n hypothesis. The underlying structure of a r e l a t i v e construction i s also examined. Chapter IV examines the hypothesis that a r e l a t i v i z e d noun phrase i n Japanese leaves i t s (r e f l e x i v e ) pronoun behind. In t h i s chapter various conditions of Japanese r e l a t i v i z a t i o n proposed so f a r are outlined i n order to give an explanation to the source of the r e f l e x i v e pronoun of the r e l a t i v e clause. Chapter V investigates the hypothesis that a noun phrase i n an adverbial clause i s r e l a t i v i z a b l e . In Chapter VI, a hypothesis i s examined that a r e l a t i v i z a b l e noun phrase i s a thematic noun phrase imme-d i a t e l y followed by the p a r t i c l e wa. In t h i s chapter, three types of examples are presented, which show that r e l a t i v i z a t i o n applies to a non-thematized noun phrase as well as a thematized noun phrase. Chapter VII i s the summary and conclusions of t h i s t h e s i s . It also presents unsolved problems. The Japanese examples are represented by using phonemic t r a n s c r i p t i o n . iv TABLE OF CONTENTS Chapter Page I. THEMATIZATION 1 Notes 31 I I . JAPANESE RELATIVE CONSTRUCTIONS AND ROSS'S CLAIMS 3^  Notes 80 I I I . JAPANESE RELATIVIZATION AND MURAKI'S HYPOTHESES 83 Notes 109 IV. KUNO'S HYPOTHESES AND JAPANESE REFLEXIVIZATION 110 Notes 139 V. RELATIVIZATION OF NOUN PHRASES IN ADVERBIAL CLAUSES H H Notes 1^9 VI. RELATIVIZABLE NOUN PHRASES AND THEMATIC NOUN PHRASES 150 Notes 170 VII. SUMMARY AND CONCLUSIONS 171 Notes 177 BIBLIOGRAPHY 178 CHAPTER I THEMATIZATION 1.1 INTRODUCTION In t h i s chapter, we w i l l deal with the theme-comment structure i n Japanese "because thematization i s claimed to be related to Japanese r e l a t i v i z a t i o n . We w i l l present some c h a r a c t e r i s t i c s of thematization and also discuss some hypotheses about the underlying structure of athematized sentence. 1.1,1 The theme-comment structure, which i s the t y p i c a l sentence structure, of Japanese, i s l i n e a r l y arranged i n the following way» (1) NP-wa + sentence '—v ' ' — - v ' NP=noun phrase (theme) (comment-part) where : w a = rthematic r- v ^T— p a r t i cle (theme-comment structure) However, there sometimes occurs an intervening p a r t i c l e 2 between the thematic NP and wa, such as kara'from', de •with' etc.. In t h i s case, the word-order of the structure i s as follows» (2) NP — - i k a r a i — wa + sentence , v ' ' s, ' (theme) (comment-part) v v — ' (theme-comment structure) 1.1.2 The theme i s usually translated into 'speaking of NP' or 'as f o r NP' . When an intervening p a r t i c l e l i k e kara ^from* or de'with' occurs, the l i t e r a l meaning of the theme i s something l i k e 'as f o r from NP' or 'as f o r with NP'. Therefore, as these l i t e r a l t ranslations show, the theme functions as a kind of"frame of reference"within which some statement i s made about that theme.''" Or the theme r e s t r i c t s the scope of the succeeding statement. In other words, the comment-part of a sentence i s some appropriate statement about the preceding theme. 1.1.3 A thematic NP i s an anaphoric NP which refers to an NP i n the preceding sentences. For example, observe the following sentences t ( 3 ) Hanako-ga eigo-o dare-ni osieta - ka? S.P. " Q.P. English who-to teach-past S.P.=subject p a r t i c l e Q.P.=question p a r t i c l e ( L i t . ) ' Whom did Hanako teach English?' (k) Hanako-wa eigo-o Taroo-ni osieta. (focus) Please note interrogative sentence ( 3 ) . The natural an-swer to ( 3 ) i s sentence ( 4 ) . In sentence ( * 0 , the theme Hanako-wa presupposes Hanako-ga of sentence ( 3 ) « Thus, the thematic NP i n NP + wa must be mentioned i n ^preceding sentences. It must be anaphoric, that i s , part of the old information presented. It must not be new information. New information introduced i n a sentence i s usually c a l l e d a focus. In sentence (^), Taroo-ni, which i s new information, functions as a focus. A focus receives strong accent. On the other hand, the rest of the sentence which i s already understood between a speaker and a hearer i s usually unstressed. Therefore, a NP, which i s part of the old information, i s unstressed i n a l l cases, except when an NP + wa has some contrastive connotation as i n (5). (5) Taroo-wa o o k i i ga Hanako-wa ookiku n a i . b i g - i s not ( L i t . ) * Taro i s big, but Hanako i s not b i g . ' Here, we follow Schachter's d e f i n i t i o n of a focus as "the focused constituent of a sentence i s the one that expressed the non-presupposed part of the propositional 2 meaning of the sentence." 1.1.4 A thematic NP may not only be a part of a previous sentence but i n some cases may be a whole sentence. Suppose we have the following assertion (7) i n connection with a presupposed sentence (6). (6) Hanako taught somebody English. (7) [JHanako-ga eigo-o o s i e t a no-wa] N P Taroo datta. one ( f o e u s^was ( L i t . ) ' The one to whom Hanako taught English was Taro.• In t h i s pseudo-cleft sentence (7)"\ the thematic NP constitutes the whole part of the presupposed sentence (6). 1.1.5 As we have already seen, a thematic NP i s anaphoric and has a feature (+ d e f i n i t e J . If i t i s not anaphoric, i t must be g e n e r i c . Please note the f o l l o w i n g example (8). (8) Zoo-wa hana-ga n a g a i . elephant t r u n k l o n g - i s ( L i t . ) 'As f o r elephants, t h e i r t r u n k s are l o n g . ' A g e n e r i c NP i s an NP which r e p r e s e n t s , without any an-tecedent* "the whole c l a s s of an e n t i t y i n qu e s t i o n , such as 'human beings i n g e n e r a l ' 'dogs i n g e n e r a l ' e t c . . 1.2 THEMATIZATION HYPOTHESES 1.2.1 Now, l e t us examine some hypotheses about themat-i z a t i o n . A c c o r d i n g t o Ross (1967),^ E n g l i s h t o p i c a l i z a t i o n i s a chopping t r a n s f o r m a t i o n which moves a t o p i c a l i z e d NP to the l e f t m o s t p o s i t i o n of a sentence as i n (9), I t can be formulated as (10). (9) I'm going t o ask B i l l t o make the o l d geezer take up these p o i n t s l a t e r . These p o i n t s I'm go i n g t o ask B i l l t o make the o l d geezer take up l a t e r . (10) X — NP — Y opt. 2 # [ l 0 3] 1 2 3 where: # i s Chomsky-ad j unct ion^ As f o r a thematic NP i n Japanese, we have two opinions. Claim 1 A theme-comment sentence i s derived from a corresponding non-thematized sentence by a movement transformation. Claim 2 A thematic NP exists o r i g i n a l l y i n the deep structure of a sentence.^ Claim 1 has the following two sub-arguments. Claim 1.a Thematization i s a kind of chopping trans-7 formation.' Claim 1.b Thematization i s a kind of copying trans-formation and i s followed by equi-NP deletion or pronominalization.^ 7 Let us examine here the nature of a chopping rule, a copying rule, and a feature-changing rule as-proposed by Ross. Ross gives us the following d e f i n i t i o n i n hi s d i s -s e r t a t i o n : ^ If the s t r u c t u r a l index of a transformation has n-terms, a,, a 2 , ... a^, i t i s a reordering trans-formation i f ms s t r u c t u r a l change has any a- as i t s k t h term, or i f a. i s adjoined to i t s fctn term, where i / k. If a transformation reorders a., and i t s s t r u c t u r a l change substitutes the i d e n t i t y element of some a k , i / k, f o r the .th term of the s t r u c t u r a l index, the transformation i s a chopping transformation. Other reordering transformations are c a l l e d copying transformations. Chopping rule S.D. a-^  — a 2 — a ^ — a ^ S.C. 1 2 3 4 (a. 1 3 2 4) (b. 1 2+3 0 4) (c. 1 0 3 4+2) (d. 4# [10 3#2 0]) etc. Copying rule S.D. the same as above =^ S.C. (a. 2+1 2 3 4) (b. 1+2 2 3 4) (c. 1 2 3 4+2) etc. Feature-changing rule By 'feature-changing rule' I mean any rule whose st r u c t u r a l index i s of the form (5.78a) and whose st r u c t u r a l change i s of the form, of ei t h e r (5. 78b) or (5.78c). 8 (5.78) a. ^'A^ •••A2 With the enumeration of Ross's three rules, l e t us examine claims l . a and l . b stated above. 1.2.2 Claim l . a According to Makino, i n order to have thematic NP at the leftmost p o s i t i o n of a sentence, we must apply two steps of transformation! ( i ) wa-insertion rule ( i i ) thematization r u l e ^ Rule ( i ) i s the process of i n s e r t i n g wa a f t e r an NP which already has another p a r t i c l e . This rule includes the process of transforming the sequence of the phrase from NP + p a r t i c l e + wa to NP 4- wa by the following r u l e i (11) C —> 0 / wa where« C = ga, _o, n i . The rule i s optional with n i but obligatory with ga/o. Here, ga i s a subject p a r t i c l e . 0 i s a d i r e c t object p a r t i c l e , and n i i s an i n d i r e c t object p a r t i c l e . Rule ( i i ) i s an optional transformation which s h i f t s the NP with wa to the i n i t i a l p o s i t i o n of a sentence. Thus, Makino gives the following rule f o r ( i i ) . (12) ( T o p ) ## X N + wa —> 0 + 2, 0 + 1 1 2 ## = sentence "boundary T s= optional trans-p formational rule I w i l l give my example of Makino*s above two steps of transformation. Sentence (14) w i l l be generated from the corresponding non-thematized sentence (13) i n the following way* (13) Hanako-ga Taroo-ni eigo-o osi e t a . (S.P.) (I.O.P.) (D.O.P.) English teach-past ( L i t . ) * Hanako taught English to Taro.' Hanako-ga Taroo-ni eigo-o-wa o s i e t a . Hanako-ga Taroo-ni eigo-wa osieta. 1 2 3 4 (14) Eigo-wa Hanako-ga Taroo-ni osi e t a . S.P. = subject p a r t i c l e I.O.P. = i n d i r e c t object p a r t i c l e D.O.P. SB d i r e c t object p a r t i c l e 10 1.2.3 Claim l . b Now, I w i l l i n t r o d u c e Muraki's a n a l y s i s of t h e m a t i z a t i o n . Suppose sentence (1*0 has Taroo as a f o c u s . Muraki's u n d e r l y i n g s t r u c t u r e of sentence (1*0 would be r o u g h l y r e p r e s e n t a b l e as (15)> 11 where i s presupposed f o r S 2 . (15) s Hanako Dum eigo o s i e t a Hanako Taroo eigo o s i e t a The necessary t r a n s f o r m a t i o n process would be as f o l l o w s J ( i ) s t r e s s - s p e c i f i c a t i o n 12 S.D. (15) r «. S.C. 1. Mark a l l the S, c o n s t i t u e n t s as £-stsj. . O b l i g a t o r y . 2. I f any Sp c o n s t i t u e n t i s i d e n t i c a l t o i t s c o r r e s p o n d i n g S, c o n s t i t u e n t , s p e c i f y i t as [.-sts} . O b l i g a t o r y . 3. S p e c i f y any unmarked c o n s t i t u e n t as C+sts] . O b l i g a t o r y . ( i i ) p a r t i c l e i n s e r t i o n ^ ^ ( i i i ) p r e s u p p o s i t i o n d e l e t i o n Ik I f S£ has no s i s t e r , d e l e t e i t , and d e l e t e the verb-Prsp. O b l i g a t o r y . 11 (iv) thematization 1. Chomsky-adjoin a non-stressed non-verbal constituent to the l e f t of the topmost sentence as the theme. If i t applies to a constituent within an em-bedded sentence, leave a pronoun behind. Obligatory i f there i s no theme i n the sentence. 2, Chomsky-adjoin the thematizer wa to the right of the theme. Obligatory. Now, i f we follow Muraki's analysis, sentence (14) would be derived from i t s underlying structure (15)» as i s roughly i l l u s t r a t e d i n the following s i m p l i f i e d tree-diagrams 8 stress s p e c i f i c a t i o n S Hanako Taroo eigo o s i e t a [+sts] p a r t i c l e i n s e r t i o n 3 j+sts] 12 presupposition deletion S ga/o deletion eigo If we formulate Muraki's thematization, i t w i l l be as follows: (16) X — NP — Y — Z =^ 2# £ 1 2 3 4] 1 2 3 4 13 According to Muraki, a f t e r copying thematization we have a kind of equi-NP deletion, as i s shown above. 1.2.4 Let us examine Makino's Claim l . a and Muraki*s Claim l.b and decide which i s correct. Observe the following sentences c i t e d from Mifcami."^ (17) (a) £[Senzitu katta] ssake-gaJ N p nokotte i r u . the other day bought l e f t - o v e r - i s ( L i t . ) • The sake which (I) bought the other day i s l e f t - o v e r . ' (b) Sake-wa. f fsenzitu kattajo Jsaket-ga 1 .jp (theme) I no J one nokotte i r u . ( L i t . ) * Speaking of sake, thejsake/which (I) t one bought the other day i s l e f t - o v e r . ' When we postulate thematization as reordering, (l?.b) i s supposed to be derived from (a). If thematization i s a chopping r u l e , the following sentence (c) and (d) w i l l be generated from (a) by i t s application, but (b) i s not derived from (a). 14 (17) (a)=> [^Senzitu katta^g sake-ga-wa) N p nokotte i r u . ^ [(Senzitu k a t t a } s sake-0-wa] N p nokotte i r u . (c) =^ [pSenzitu katta^g sake-wa] N p nokotte i r u . ( L i t . ) * As f o r the sake which I bought the other day, i t i s l e f t - o v e r . ' (a) ^ [CSenzitu katta^g sake-ga-waj N p nokotte i r u . =^ [CSenzitu k a t t a } s sake-j2f-wa] N p nokotte i r u . (d) =^ *Sake-wa, [(senzitu k a t t a ) s -J ^ O f l p nokotte i r u . Sentence (17.c) i s grammatical but i t i s not synonymous with (17.b). Sentence (17»d) i s absolutely ungrammatical. In other words, (17.b) can not be derived by chopping themat-i z a t i o n . Rather, the above examples show that thematization i s a copying r u l e . If we follow Muraki's copying themat-i z a t i o n , the d e r i v a t i o n a l process of (17.h) from (17.a) w i l l be roughly as follows: (17) (a) ([Senzitu kattajg s a k e - g a ) ^ nokotte i r u . Sake-ga-wa, [^senzitu kattajg s a k e - g a J N p nokotte i r u . 15 (b)=^Sake-#-wa, [[senzitu k a t t a ^ s sake-ga] N p nokotte i r u . Or from the above (b), by pronominalization, Sake-wa, [f senzitu k a t t a j g n o - g a l N p nokotte i r u . The following sentences are from Mikami.3"^ (18) Nittyoku^-wa, zyosi-syokuin-ga kore^-ni ataru. day-duty female-staff t h i s undertake ( L i t . ) • As f o r day-duty, the female s t a f f s (should) take charge of it> ' (19) Nihon-no gendai bungaku^-wa, sore^-ga hazimmatte Japan's modern l i t e r a t u r e i t begin kara sono^ dentoo-ga dekita to i e r u hodo-no since i t s t r a d i t i o n established that say-can degree's zikan-mo tatte i n a i . time elapse is-not ( L i t . ) ' As f o r Japanese modern l i t e r a t u r e . , since i t . started, the time has not elapsed to such an extent that we can say i t . has established i t s ^ t r a d i t i o n . ' According to Mikami, we leave pronouns l i k e sore ' i t ' , sono ' i t s j the' i n a formal s t y l e and we have also kore 16 • t h i s ' or kono ' t h i s ' at l e a s t i n an e d i t o r i a l tone of the 18 press or i n l e g a l usage. If so, Makino's chopping themati- zation can not explain sentences l i k e (18) and (19)» which leave pronouns behind a f t e r thematization. Therefore, i t i s necessary f o r us to consider themati-zation as copying rather than chopping. A f t e r copying  thematization we have some conditions of equi-NP d e l e t i o n or pronominalization. 1.2.5 Now, l e t us proceed to the examination of Kuno*s and Muraki*s hypotheses with regard to the source of a thematic NP, and decide which has a better explanation f o r Japanese thematized sentences. In 1970, there was a controversy between Kuno and 19 Muraki concerning the following sentence. 7 (20) Sakana-wa t a i - g a JLi. f i s h red-snapper ( L i t . ) • Speaking of f i s h , red-snapper i s the best.' Muraki says that the thematized sentence (20) comes from a corresponding non-thematized sentence (21) i n the following way: 17 (21) T a i - g a sakana-no-naka-de i t i b a n i i . among the b e s t ( L i t . ) ' Red-snappers are the best among the f i s h e s . ' t h e m a t i z a t i o n  = = =^ Sak ana-no- naka- de-wa t a i - g a ( i t i b a n ) i i . . , 0 ( o p t i o n a l ) d e l e t i o n of no-naka-de Sakana-0-wa t a i - g a i i . Muraki's argument f o r h i s copying t h e m a t i z a t i o n h y p o t h e s i s i s summed up i n the f o l l o w i n g way: I f thematized sentences d i d not have any corresponding non-thematized sentences, the theme NP-wa would have t o be generated by the base r u l e s independently of the f o l l o w i n g sentence. In o t h e r words, t h e r e would not be a requirement t h a t the theme be i d e n t i c a l t o some c o n s t i t u e n t of the f o l l o w i n g sentence i n the u n d e r l y i n g s t r u c t u r e . Then, we co u l d not e x p l a i n the f a c t t h a t the thematized sentence t y p i c a l l y has a s l o t i n which the theme can f i t , n or the f a c t t h a t the p a r t i c l e which can occur a t the end of the theme depends on the verb of the f o l l o w i n g s e n t e n c e . 2 0 Thus, he g i v e s us the f o l l o w i n g examples to support h i s 21 argument. (22) (a) John-kara-wa Mary-ga kane-o ka r i t a . ; , from money borrowed ( L i t . ) ' From John, Mary borrowed some money.' 18 (b) * John-kara-wa Mary-ga kane-o kasi-*-ta. l e n t ( L i t . ) * From John, Mary l e n t some money.' Judging from h i s above argument and h i s examples, the accept-a b i l i t y of (22.a) and the u n a c c e p t a b i l i t y of (22.b) cor r e -spond to the a c c e p t a b i l i t y or u n a c c e p t a b i l i t y of t h e i r a s s o c i a t e d non-thematized sentences (23.a) and (23.b). (23) (a) Mary-ga John-kara kane-o k a r i - t a . (b) Mary-ga John-kara kane-o kas?i - t a . Fu r t h e r , the theme John-kara-wa i n (22.a), e s p e c i a l l y the p a r t i c l e k a r a , does not o r i g i n a l l y e x i s t i n the u n d e r l y i n g s t r u c t u r e but i s copied from the corresponding non-themat-i z e d sentence. He seems to t h i n k that there are the same s e l e c t i o n a l r e s t r i c t i o n s between thematic sentences and the corresponding non-thematized sentences. On the other hand, Kuno says that a thematized sentence (20) does not have a corresponding non-themat-i z e d sentence, as i s shown i n h i s f o l l o w i n g sentence (24), (24) *Sakana^ ga no de } t a i - g a i i . (non-thematized sentence) Thus, he says that i n sentence (20), the theme sakana-wa e x i s t s o r i g i n a l l y as the theme i n the u n d e r l y i n g s t r u c t u r e , 19 F u r t h e r , Kuno argues a g a i n s t Muraki's d e r i v a t i o n a l process o f ( 2 0 ) , s a y i n g i n the f o l l o w i n g ways There i s no j u s t i f i c a t i o n f o r t r a n s f o r m a t i o n a l l y d e l e t i n g no u t i / naka de, which c o n t a i n s u t i and naka (both meaning ' i n s i d e ' ). Jftjdeed, no u t i / naka de i s not o r d i n a r i l y d e l e t a b l e . He g i v e s the f o l l o w i n g example t o support h i s counter-argument a g a i n s t Muraki. (26) (a) John t o B i l l t o Marv-no u t i / n a k a de Mary-ga and among i t i b a n yoku d e k i r u . most w e l l d o e s - w e l l ( L i t . ) ' Among John, B i l l , and Mary, Mary does the b e s t . ' (b) ?? John t o B i l l t o Mary-0-wa, Mary-ga i t i b a n yoku d e k i r u . 20 In h i s new book, Kuno c i t e s the f o l l o w i n g examples from Mikami t o support h i s c l a i m t h a t a thematic sentence does not always have i t s corresponding non-thematized 24 sentence. " (27) Sinbun-o yomi t a i hito-wa. koko-ni arimasu. newspaper read-want people here e x i s t ( L i t . ) ' Speaking of those who want to r e a d news-papers ,they(=newspapers)are here.' (28) Basyo-wa, okunai-setu-ga a t t o o t e k i d a t t a . p l a c e s i n d o o r - t h e o r y predominant was ( L i t . ) • Speaking of the p l a c e (of the murder), the "indoor" t h e o r y was predominant.* Kuno's argument f o r h i s s o - c a l l e d deep s t r u c t u r e  theme h y p o t h e s i s would be summed up as f o l l o w s : There are thematic sentences f o r which t h e r e are no corresponding themeless sentences. ••• Thus, we are f o r c e d t o assume t h a t themes e x i s t as themes i n the deep s t r u c t u r e o f thematic sentences. Concerning the above two hypotheses, I agree with Kuno and r e j e c t Muraki's h y p o t h e s i s f o r two reasons. F i r s t , i f we do not f o l l o w Kuno's h y p o t h e s i s , there w i l l be cases which can not be accounted f o r . F o r example, 21 Kuno's examples (2?) and (28) do not have t h e i r correspond-ing non-thematized sentences. I w i l l give two more examples to support Kuno's view. Observe the following sentences. Notice that (b) sentences are not acceptable. (29) (a) Kore-wa simatta. t h i s made a mistake ( L i t . ) * Oh, I screwed up.' (b) *Kore-ga o de n i f simatta. (non-thematized sentence) ( 3 0 ) (a) Kuwasii koto-wa hon-no go peezi-o mite detailed thing book's 5 page look-at kudasai. please ( L i t . ) ' As f o r the d e t a i l s , please look at page f i v e of the book.' (b) *Kuwasii k o t o - l n i l hon-no go peezi-o mite (gaJ kudasai. e t c # (non-thematized sentence) The above thematized sentences ( 2 9 . a ) and ( 3 0 . a ) can not be explained by Muraki*s hypothesis because they do not 22 have any corresponding non-thematized sentences, as i s shown i n (29.b) and (30.b). Second, with the ordering of a set of transformations, we could reply to Muraki's argument quoted on P. 17 i n t h i s chapter. My solution to the ordering of a set of trans-formations i s as follows. ( i ) We should have an underlying structure something l i k e (25), f o r any thematic sentence, ( i i ) part iglej-copying (a) If the thematic NP happens to be i d e n t i c a l to an NP i n the comment-part, the p a r t i c l e (attach-ed to the i d e n t i c a l NP i n the comment-part) must be copied to the thematic NP. (b) If there i s no i d e n t i c a l NP i n the comment-part, t h i s rule does not apply. ( i i i ) p a r t i c l e - d e l e t i o n If the newly-inserted p a r t i c l e i s ga. or o, i t must be deleted. The formulation of t h i s rule would be something l i k e Makino's rule (11) quoted on P. 8. ( i v ) equi-NP d e l e t i o n o r p r o n o m i n a l i z a t i o n (a) I f the thematic NP has an i d e n t i c a l NP i n the comment-part, the i d e n t i c a l NP must be d e l e t e d o r p r o n o m i n a l i z e d . (b) I f t h e r e i s no i d e n t i c a l NP i n the comment-p a r t , the r u l e does not apply. With t h i s s e t of t r a n s f o r m a t i o n s , Muraki's counter-example (22.a) and Mikami's sentence (17.b) w i l l be analyzed i n the f o l l o w i n g way. ( i ) the u n d e r l y i n g s t r u c t u r e of (22.a) N P NP-ga NP-o NP-kara V I i John Mary okane John k a r i t a ( i i ) p a r t i c l e - c o p y i n g John-kara-wa Mary-ga okane-o John-kara k a r i t a : ( i i i ) p a r t i c l e - d e l e t i o n T h i s r u l e does not apply here. ( i v ) equi-NP d e l e t i o n John-kara-wa Mary-ga okane-o $ k a r i t a . ( i ) the underlying structure of (17.a) NP NP-o V senzitu sake katta In t h i s case, we have del e t i o n r e l a t i v i z a t i o n f i r s t , by which we derive the following s t r i n g . Sake-wa [senzitu 0 katta,] sake-ga nokotte i r u . ( i i ) particle-copying Sake-ga-wa [ s e n z i t u katta] sake-ga nokotte i r u . ( i i i ) p a r t i c l e - d e l e t i o n Sake-jtf-wa [ s e n z i t u k a t t a ) sake-ga ... (iv) pronominalization (optional) Sake-wa £ senzitu k a t t a ) no-ga ... 25 Now, we can e x p l a i n the d e r i v a t i o n s of t h r e e types of thematic NP's i n the f o l l o w i n g way. ( i ) In case t h e r e i s no i d e n t i c a l NP i n the comment-part. NP + wa ( i i ) In case t h e r e i s an i d e n t i c a l NP i n the comment-part and an i n t e r v e n i n g p a r t i c l e i s ga o r o. ( In t h i s case, the i n t e r v e n i n g p a r t i -c l e i s d e l e t e d by p a r t i c l e d e l e t i o n . ) NP + p a r t i c l e + wa * 0 ( i i i ) In case t h e r e i s an i d e n t i c a l NP i n the comment-part and an i n t e r v e n i n g p a r t i c l e i s not ga o r o . ( I n t h i s case p a r t i c l e d e l e t i o n i s not a p p l i c a b l e . ) NP + p a r t i c l e + wa Thus, we are i n c l i n e d t o go along with Kuno's deep  s t r u c t u r e theme h y p o t h e s i s . We t h i n k t h a t a s e t of a p p r o p r i -ate t r a n s f o r m a t i o n s can d e a l w i t h Muraki's counter-argument, and t h a t Kuno's hy p o t h e s i s i s the best r e p r e s e n t a t i o n of 26 of a theme-comment structure because i t represents best the semantic function of the theme to r e s t r i c t the scope of the comment-part. 1.3 THE P-MARKER OF A THEMATIZED SENTENCE Let us examine the P-marker of a thematized sentence. Notice that both Kuno and Muraki claim that a thematic NP i s Chomsky-adjoined to the l e f t of a following comment-sentence. This P-marker i s quite natural as the semantic representation of a theme-comment structure and also f o r the phonological reason that we seem to have a longer pause a f t e r the thematic NP than at any other place i n a simplex sentence. This i s true also f o r syntactic reasons. A t y p i c a l example f o r t h i s P-marker w i l l be found-i n the sentence which follows, where the comment sentence i s a kind of predication of the theme. The structure of (31) i s one which equates one thing with another, such as 'A i s B', where A i s the theme and B i s the whole comment-sentence. 2? (31) Z i t uzyoo-wa, muda-ga o o i . a c t u a l s i t u a t i o n f r u i t l e s s n e s s much-is ( L i t . ) ' As t h i n g s stand, t h e r e ' s too much t h a t i s t o no a v a i l . ' We have c r o s s - l i n g u i s t i c evidence which may support our P-marker under d i s c u s s i o n here. Park claims t h a t the m u l t i p l e s u b j e c t c o n s t r u c t i o n i n Korean i s not a simple sentence, but a complex sentence c o n s t r u c t i o n , where embedded sentences f u n c t i o n as the p r e d i c a t e phrases of each h i g h e r 26 sentence. Lewkowicz a l s o has a h y p o t h e s i s t h a t i n A r a b i c a topic-comment sentence has a s t r u c t u r e — NP + S, i n which NP i s a theme and S i s a p r e d i c a t e t h a t c o n s i s t s of 27 an embedded comment c l a u s e . t o the e n t i r e sentence a l s o s u b s t a n t i a t e s our c l a i m f o r the P-marker of a thematic sentence. Observe the f o l l o w i n g 28 s e n t e n c e s , which are from Muraki. The f a c t t h a t a thematic NP r e l a t e s s e m a n t i c a l l y (32) [ k a r a i e - n i k a e t - t a . because home-to r e t u r n e d . ( L i t . ) ' Since t h a t person was s i c k , (I) went home.' 28 (b) (sono hito-wa [byooki dajg k a r a i e - n i k a e t - t a ) s ( L i t . ) * As f o r t h a t person., s i n c e he.^ was s i c k he^ went home.' (c) jsono hito^-wa [sono h i t o . p g a byooki da)g k a r a i e - n i kaet-ta} g ( L i t . ) * As f o r t h a t person., he^ went home because the person, was s i c Z . ' J Normally each sentence g i v e n above means* (32.a)*the person who was s i c k and the per s o n who went home are d i f f e r -ent.' (32,b)'The person was s i c k and went home.*, and (32.c) 'The person went home because somebody e l s e was s i c k . ' . These sentences l e a d us t o say t h a t a thematic NP r e l a t e s s e m a n t i c a l l y t o the e n t i r e sentence, whereas the NP with ga i n (32,a) and (32,c) r e l a t e s o n l y as f a r as the verb of the c l a u s e i n which i t o c c u r s . T h i s f a c t supports our c l a i m t h a t a thematic NP must be Chomsky-adjoined to the l e f t of the f o l l o w i n g comment-sentence. 1.4 SUMMARY To summarize the o b s e r v a t i o n s i n t h i s chapter, we have d i s c u s s e d the f o l l o w i n g t h r e e t o p i c s : a. Some c h a r a c t e r i s t i c s of Japanese thematized sentences b. Hypotheses about thematic noun phrases c. The P-marker of thematic sentences With r e s p e c t to ( a ) , we have seen the f o l l o w i n g c h a r a c t e r i s t i c s i n Japanese thematized sentences. ( i ) The word-order of a thematized sentence i s as f o l l o w s : (theme) (comment-part) ( i i ) The theme f u n c t i o n s t o r e s t r i c t the scope of the succeeding statement, and i s u s u a l l y t r a n s l a t e d as *as f o r NP*. • ( i i i ) A thematic NP i s an anaphoric NP o r a g e n e r i c NP. 30 ( i v ) An anaphoric thematic NP r e f e r s e i t h e r t o an NP i n a p r e c e d i n g sentence o r to the whole p r e c e d i n g sentence. Concerning t o p i c (b), we have examined three hypotheses, t h a t i s , Makino's chopping t h e m a t i z a t i o n , Muraki's copying t h e m a t i z a t i o n , and Kuno's deep s t r u c t u r e theme h y p o t h e s i s . A f t e r the v a r i o u s examples we have concluded t h a t Kuno's h y p o t h e s i s has the b e s t e x p l a n a t i o n f o r thematized sentences. A l s o , we have seen t h a t the o r d e r i n g of some t r a n s f o r m a t i o n s i s needed t o account f o r an i n t e r v e n i n g p a r t i c l e between a thematic NP and wa. F i n a l l y , we have observed t h a t the theme must be Chomsky-adjoined t o the l e f t of the f o l l o w i n g comment-sentence, because t h i s P-marker r e p r e s e n t s b e s t the under-l y i n g s t r u c t u r e of a thematic sentence, s e m a n t i c a l l y and s y n t a c t i c a l l y . 31 CHAPTER I NOTES 1. Minora. Nakau, "Some Constraints on Topicalization," Papers in Japanese Linguistics, Vol. 1 No. 1 (June, 1972), p. 79. 2. Paul Schachter, "Focus and Relativization," Language, Vol. 49. No. k (March 1973). p. kZ — 3. In English, a cl e f t sentence and a pseudo-cleft 'sentence are respectively illustrated in the following sentences i (i) Tom broke the window, ( i i ) It was the window that Tom broke, (cleft sentence) ( i i i ) What Tom broke was the window, (pseudo-cleft sentence) The above pseudo-cleft sentence ( i i i ) i s translated into Japanese as followsi (iv) (Tom-ga kowasitajg no-wal^. mado datta. where« no. means 'the onej the thing*. datta *was' is a copula verb. 4. John R. Ross, "Constraints on Variables in Syntax," unpublished Ph. D. dissertation, Massachusetts Institute of Technology, (1967) p. 115. 5. The term Chomsky-adjunction is defined in the following way t When an element B is adjoined to node A. i t i s possible to make another node A over the original node A and have the new node A dominate both node B and the original node A. The following tree diagrams are a depiction of the Chomsky-adjunction. (i) ( i i ) A T B In this case, B is said to be Chomsky-adjoined to the l e f t of the original node A. B and the original node A are said to be in sister relation in ( i i ) . 32 6 . Susumu Kuno, Notes on Japanese Grammar, (Harvard Computation Laboratory, NSF-27, 1 9 7 0 ; , Chap. 1 9 , p. 11 . 7. S e i i c h i Makino, Some Aspects of Japanese Nominaliza-tions,(Tokyo J Tokai University Press, 1 9 6 8 ) , p. 116. He did not say that thematization i s a chopping r u l e , but judging from a reading of his rule, his thematization contains an i m p l i c i t chopping r u l e . 8 . Masatake Muraki, "Presupposition, Pseudo-clefting and Thematization," (unpublished Ph. D. dis s e r t a t i o n , the University of Texas At Austin, 1 9 7 0 ) , pp. 48 & 190. 9 . Ross, o_p. c i t . , pp. 172 & 235. 1 0 . Makino, op. c i t . , p. 116, 11 . Muraki, op. c i t . , p. 116. 12. Ibid., pp. 165 f . 13. Ibid., p. 154. 14. Ibid., p. 1 6 8 . 15. Ibid., pp. 168 f . 16. Akira Mikami, Zoo wa Hana ga Nagai, 4th ed., (Tokyo t Kurosio Pub. Co., 19597, p. 72. The o r i g i n a l orthography used by each author was kept except f o r some necessary modifications, such as underlining, labeled bracketing etc., f o r the purpose of d i r e c t quotes. 17. Ibid., pp. 140-144. 18. Ibid. 19. See Muraki, op_. c i t . , pp. 231- 234 and Kuno, op. c i t . , Chap. 1 9 , pp. 9 - l > T 20. Muraki, Ibid., pp. 233 f . 21. Ibid., p. 234. 22. Kuno, jjp_. c i t . , Chap. 19, p. 11. 23. Ibid., Chap. 19, pp. 10 f . 2 4 . Susumu Kuno, The Structure of the Japanese Language. (Cambridge 1 The MIT Press, 1973), P. 253. 25. See Kuno, op. c i t . , Chap., 1 9 , pp. 9 f . 33 26. Byung-soo Park, "On the Multiple Subject Construction i n Korean," L i n g u i s t i c s i n International Review, Vol. 100 (March, 1973)t PP. 72—76. 27. Nancy K. Lewkowicz, "Topic-comment and Relative Clause i n Arabic," Language, V o l . 47, No. 4 (December, 1971), pp. 813-816. 28. Sentences (32 a.b.c) are from Muraki. He argues that (32.a) and (32.b) are ambiguous even though one of the read-ings may be more common than the other. See Muraki, op. c i t . , pp. I90-I92. — CHAPTER II JAPANESE RELATIVE CONSTRUCTIONS AND ROSS'S CLAIMS 2.1 INTRODUCTION There are some c h a r a c t e r i s t i c s which d i s t i n g u i s h a r e l a t i v e clause construction i n Japanese from that i n English. In the following, we w i l l enumerate the character-i s t i c s of Japanese r e l a t i v e constructions. ( i ) A r e l a t i v e clause i n Japanese precedes i t s head NP, as i n (1), while i n English i t follows the NP. (1) [boku-ga suwatte i t a ] i s u I sitting-was chair ( r e l a t i v e clause) (head NP) ( L i t . ) ' t h e chair which I was s i t t i n g on ' ( i i ) There i s no s p e c i a l r e l a t i v e morpheme, except f o r tokoro-no which corresponds to English 'who', 'which','that' etc., as i s shown i n example (1). Tokoro-no, whose l i t e r a l meaning i s 'of the place', 35 i s only used i n d i r e c t t r a n s l a t i o n from foreign languages, and i s not used i n ordinary conversation or writing. Therefore, tokoro-no i s out of our present study, ( i i i ) R e l a t i v i z a t i o n i n Japanese i s assumed to be a deletion process (or we could c a l l i t 0-pronominaliza-t i o n ) , which deletes the r e l a t i v i z e d NP i n an embedded clause, together with the following p a r t i c l e (or post-p o s i t i o n ) . 1 In English, however, r e l a t i v i z a t i o n may be considered to involve a movement transformation and the preposition attached to the r e l a t i v i z e d NP i s not deleted. Sentence (1) i s derived from sentence (2), where n i 'on* undergoes deletion by r e l a t i v i z a t i o n . (2) [boku-ga sono i s u - n i suwatte i t a ] i s u I the chair-on sitting-was chair [boku-ga 0 suwatte i t a j i s u ( L i t . ) ' the chair which I was s i t t i n g on ' (iv) Japanese r e l a t i v i z a t i o n i s said to be subject only to a r e s t r i c t e d version of Ross's Complex NP Constraint. It i s also said that an NP i n an adverbial clause i s r e l a t i v i z a b l e . These topics w i l l be d e t a i l e d l a t e r . 36 (v) It has also been claimed that "Japanese has no phonological, morphological and syntactic d i s t i n c -tions between r e s t r i c t i v e or non-re s t r i c t i v e 2 clauses." They are distinguished only semantically as i n the following examples. (3) [Stanley Park-ni i r u ] zoo ( r e s t r i c t i v e ) i n i s elephant ( L i t . ) >' the elephant which i s i n Stanley Park • (4) [rikuzyoo saidai-no doobutu dearu] zoo land-on biggest animals are (non-restrictive) ( L i t . ) • elephants, which are the biggest animals on land ' 2.2 THE RELATIVE CONSTRUCTION IN ENGLISH AND JAPANESE In t h i s chapter, we w i l l discuss some of the problemat-. ie points with respect to the Japanese r e l a t i v e constructions. Let us enumerate them f i r s t , ( i ) English r e s t r i c t i v e and non - r e s t r i c t i v e r e l a t i v e clauses and t h e i r counter-parts of Japanese, ( i i ) Whether or not Japanese r e l a t i v i z a t i o n i s a movement transformation, ( i i i ) Whether or not an element i n a complex NP i s r e l a t i v i z a b l e . 37 (iv) The examination of the Coordinate  Structure Constraint. Now, l e t us examine these points one by one. ( i ) English r e s t r i c t i v e and n o n - r e s t r i c t i v e r e l a t i v e clauses and t h e i r counter-parts of Japanese. In order to provide a contrast between the use of r e l a t i v e clauses i n English and Japanese, we w i l l discuss the f o l l o w i n g i (a) i d e n t i t y condition (b) the problem of a r e l a t i v e marker (c) the semantic roles of r e s t r i c t i v e and n o n - r e s t r i c t i v e r e l a t i v e clauses . In English, a r e l a t i v e construction i s divided into two types t r e s t r i c t i v e r e l a t i v e constructions — hereafter abbreviated as RR-constructions and n o n - r e s t r i c t i v e ; ones , — NRR-constructions. An RR-construction i s a kind of endo-cent r i c construction. It consists of a r e l a t i v e clause and i t s modified head NP, and the whole construction functions as an NP. In an RR-construction, a clause i s usually said to 38 be an embedded sentence i n the head NP of a matrix sentence. To generate a r e l a t i v e clause, a condition must be met that some NP within the embedded sentence be iden-t i c a l with the head NP. This condition i s c a l l e d hereafter i d e n t i t y condition. From a semantic viewpoint, we can say with Schachter that the function of an RR-construction is« to provide names f o r or ways of designating the multitude of i d e n t i t i e s that people wish to t a l k about, but f o r which there i s no established single-noun designation.-' The above two f u n c t i o n a l c h a r a c t e r i s t i c s of RR-con-structions — that i s , the modification of the head NP and naming the e n t i t i e s i n question — are common to Japanese. Nevertheless, there are some c h a r a c t e r i s t i c s which are not shared between English and Japanese. For example, i n English we have r e l a t i v e pronouns to designate the syntactic r e l a t i o n between a clause and i t s head NP. A Japanese r e l a t i v e clause, as mentioned previously, does not have these kinds of substantial markers which e x p l i c i t l y indicate the existence of r e l a t i v e elauses. For. example, observe the following sentences* 39 (5) (a) the book £1 bought the booS yesterday] =^ the book[which I bought yesterday] (b) [sono hon-wa boku-ga kinoo katta] hon the book I yesterday bought 0 boku-ga kinoo katta] hon v v 1 ( r e l a t i v e clause) (head NP) In (5.a), r e l a t i v e clause transformation adds the features [+pronoun] and [+WHJ to the embedded NP'the book' and moves i t to the front of the embedded sentence. On the other hand, the only s t r u c t u r a l change i n (5»b) i s the de l e t i o n of the embedded i d e n t i c a l NP. That i s , no s p e c i a l r e l a t i v e markers are involved i n Japanese. Also, i n English NRR-constructions are divided into two sub-types. Type 1 sentences, whose r e l a t i v e pronouns match the head NP's, are exemplified by ( 6 ) ( 7 ) and ( 8 ) , while type 2 sentences, by ( 9 ) . A type 2 sentence i s a sentence whose r e l a t i v e pronoun matches the whole preceding clause, or part of the-preceding VP as i n the following sentences 1 Tom i s t a l l , which I w i l l never be. I saw that Tom was easy to please, which I should be, too. 40 (6) My father, who i s now 80 years old, i s s t i l l working. ( 7 ) Elephants, (which l i v e ) i n India and A f r i c a , have long trunks. (8) I went to Osaka, where I happened to see Mr. Sato. (9) Tom remained s i l e n t , which made his wife s t i l l more angry. The above sentences can be translated into Japanese as follows J (10) (a) Boku-no t i t i - w a ima hatizyuu desu ga my father now 80 i s mada hataraite i r u . s t i l l working-is ( L i t . ) ' My father — he i s now 80 years o l d — i s s t i l l working.* (b) Hatizyuu-ni naru boku-no t i t i - w a mada 80 become my father s t i l l hataraite i r u . working-is ( L i t . ) • As f o r my father, who i s 80 years old, (he) i s s t i l l working.* 41 (11) (a) Zoo-wa India to A f r i c a - n i sunde i r u ga elephants and i n l i v i n g - a r e hana-ga nagai. trunks long-are ( L i t . ) ' Elephants — they l i v e i n India and A f r i c a — t h e i r trunks are long.* (b) India to A f r i c a - n i sunde i r u zoo-wa and i n l i v i n g - a r e elephants hana-ga nagai. trunks long-are ( L i t . ) • As f o r elephants, which l i v e i n India and A f r i c a , (their) trunks are long.* (12) Boku-wa Oosaka-ni i t t a , (sosite) sokode I Osaka-to went and there guuzen Satoo-san n i att a . by chance met ( L i t . ) • I went to Osaka, and there I happened to see Mr. Sato.* (13) Tom-wa damatte i t a , (suruto) sono koto-ga silent-was and then the thing okusan-o naosara okoraseta. wife s t i l l angry-cause-past ( L i t . ) ' Tom remained s i l e n t , and i t made his wife s t i l l more angry,' 42 As sentence (13) shows, Japanese does not have type 2 sentences. Instead, we express the meanings o f these sentences i n the form o f c o n j o i n e d sentences o r two independent sentences as i n (13). T h i s may be because Japanese does not have r e l a t i v e pronouns. S i m i l a r l y , because of the l a c k o f a r e l a t i v e pronoun, (12), which expresses two events a t one p l a c e i n a se-q u e n t i a l order, i s r e p r e s e n t e d o n l y by j u x t a p o s i n g the two sentences. That i s , some k i n d o f sentence i n type 1- NRR-construction l i k e (8) has no r e l a t i v e c o u n t e r - p a r t of Japanese. As f o r an i n t r a s e n t e n t i a l NRR-construction l i k e (6) and ( 7.), we have two k i n d s o f Japanese t r a n s l a t i o n s , one i s i n terms of c o n j o i n e d sentences and the o t h e r i s i n terms o f embedding. However, we must note t h a t i t i s s a i d t h a t t h e r e are no p h o n o l o g i c a l , m o r p h o l o g i c a l , o r s y n t a c -t i c d i s t i n c t i o n s between an o r d i n a r y R R - c o n s t r u c t i o n ( f o r example (5.b)) and sentence ( l O . b ) . That i s , i n E n g l i s h we can d i s t i n g u i s h between RR- c o n s t r u c t i o n s and NRR-constructions by means of d i s t i n c t i v e i n t o n a t i o n contours, the e x i s t e n c e 43 or non-existence of commas when written, and by some r e s t r i c t i o n s such as, *in case of NRR-sentences, the ob-jec t i v e r e l a t i v e pronouns must not be deleted* etc.. Since Japanese has no NRR-RR distinctions,there i s a p o s s i b i l i t y that i f , f o r example, (11.b) i s spoken to a person who does not know where elephants l i v e and what elephants look l i k e , he thinks that there are some other kinds of elephants whose trunks are not long. Thus, (11.a) i s , more often used i n ordinary speech when a speaker suspects that the information may be new to h i s hearer. Rather i t may be that we should say (10.a) and (11.a) are more natural i n Japanese. Here we must examine the semantic property of in t r a s e n t e n t i a l NRR-construction-r-for example (6) (7);— compared with that of the NRR-construction which;as attached sentence-finally as i n (8). A. Loetscher (1973)^ says that inan ( 8)-type sentence, the preceding conjunct and the sentence-final NRR-conjunct are considered to be equivalent from the point of information value. His example (14) shows that S 1 and S p are interpreted as the 44 same-level information and constitute an answer as a whole. (14) Qt Did you get anything to eat yesterday? As Oh, yes. Paul i n v i t e d us to him, * y * S l where he offered us a splendid dinner. Y ' So On the other hand, i n (6) and ( 7 ) . which involve i n t r a s e n t e n t i a l NRR-constructions, the intended messages are contained respectively i n each main clause, that i s , the part of the sentence which excludes the NRR-elause. Loetscher gives the following examples (15) Q« What are the p a r t i c u l a r c h a r a c t e r i s t i c s of a lark? Atf(a) The l a r k has a very sweet song. It builds i t s nest on the ground. (b) The l a r k , which builds i t s nest on the ground, has a very sweet song. Both of the two independent sentences i n (a) equally express the c h a r a c t e r i s t i c s of a lark, but i n (b) the NRR-clause does not give any information to the preceding question 45 and only the main clause i s considered to be relevant to the question. The reasons why we insert an NRR-clause into a main clause are, according to Loetscher (1973)» that ( i ) One can avoid t r e a t i n g a hearer as being un-informed by embedding such a sentence as a NRR-construction, thereby i n d i c a t i n g that one t e l l s t h i s sentence only as a supplementary information, i n order to r e c a l l an old f a c t , but i t i s not the a c t u a l l y intended message. ( i i ) The message one wants to convey i s contained i n the main clause, but i t needs an explanation which i s given i n the NRR-construction.7 -In other words,an NRR-construction i s used i n a s i t u a t i o n where one wants to d i s t i n g u i s h between hot jiews and l e s s hot news. As a conclusion, he says that only an i n t r a s e n t e n t i a l NRR-eonstruction i s exclusively low-ranked information, due to the p o s i t i o n i n a sentence c a l l e d the parenthesis p o s i t i o n . He also says i t i s often the case that these NRR-construc-tions can be paraphrased into simple non-relativized parenthetical sentences without changing the discourse p sense, as i n (16). 46 (16) (a) Sam, who Jack had given a blow on the head, went down and started squeaking. (b) Sam — - Jack had given him a blow on the Q head went down and started squeaking. 7 Then, how about i n Japanese? (12), which i s the tr a n s l a t i o n of (8), consists of two juxtaposed sentences, and each conjunct i s equivalent concerning information value. That i s , the conjunction s o s i t e 'and then' conjoins two conjuncts as equal information value. The only difference between (8) and (12) i s that the former uses a sentence-final NRR-construction and the l a t t e r uses a coordinate conjunction. On the other hand, i n (10.a) and ( l l . a ) , a con-junction ga seemingly conjoins the two conjuncts i n the same way as i n (12). But i t i s not true. To most native speakers, ga has mostly two meanings and functions. One mean-ing i s purely 'but'. In t h i s case, ga i s strongly stressed and functions as a coordinate conjunction which juxtaposes two conjuncts. The other ga i s unstressed and does not mean k? •but*. In t h i s l a t t e r case, the f i r s t conjunct + ga part indicates that the information there i s only supple-mentary and low-ranked. The main message which a speaker wants to convey i s contained i n the second conjunct. For example, take (11.a) again. (11) (a) Zoo-wa India to A f r i c a - n i sunde i r u (gi hana-ga nagai. ( L i t . ) * Elephants — they l i v e i n India and A f r i c a — t h e i r trunks are long.* The main information i s contained i n the underlined part and the frame-part i s subsidiary. Ga, i n t h i s case, does not function as a coordinate conjunction which conveys the two contrary messages, but functions only as a marker in d i c a t i n g that the supplementary information i s expressed within the frame. The following example also points to the same phenomenon. (17) (a) Sono gakusei-wa tetugaku senkoo da ( p the student philosophy majoring-in-is ima tosyokan-de mai n i t i sanzikan now l i b r a r y at every day 3-hours-for 48 hataraite imasu. working-is ( L i t . ) • The student •— (he) i s majoring i n philosophy — now works at the l i b r a r y f o r three hours every day. / T h e student, who i s majoring i n philosophy works at the l i b r a r y f o r three hours every day.* (b) Sono gakusei-wa [tetugaku senkoo nan desu(yoj, ; i s (I t e l l you) so s i t e and ima tosyokan-de m a i n i t i sanzikan hataraite imasu(yoi ( L i t . ) • (I t e l l you) the student i s majoring i n philosophy and (I t e l l you i n addition) he works at the l i b r a r y f o r three hours every day.' The two pieces of information i n the frames of (17.b) are equivalent, while i n (17.a) the information within the frame i s of secondary information value. Thus the above observation shows that an i n t r a s e n t e n t i a l supplementary NRR-construction i s expressed i n Japanese i n the form of the frame I — — (gai as i n (18). This may be due to the lack of r e l a t i v e pronouns. (18) NP - wa Mb i - stress - contrastivej 49 To summarize the above observationsi ( E n g l i s h ) ( J a p a n e s e ) I RR-construction (embedding) (e.g. (5.a)) I RR-construction (embedding) (e.g. (5.h)) II NRR-construction Type 1 ( i ) i n t r a s e n t e n t i a l NRR-construction ( i i ) sentence-final NRR-construction (e.g. (8)) ( i ) (a) RR-construction (embedding) (e.g. (10.b) (11.b)) (b) i n terms of I.. frame (e.g. \ix.a)) ( i i ) coordinate construction (e.g. (12)) Type 2 sentence-final NRR-construction (e.g. (9)) coordinate construction (e.g. (13)) Notice that we are not discussing here whether the deepest underlying representation f o r a r e l a t i v e con-s t r u c t i o n involves a sentence embedded into an NP, or conjoined sentences proposed by S.A. Thompson. 1 0 What I want to say i s that at l e a s t at some stage of derivation, the above mentioned embedding or conjoin-ing i s observed. It must also be said that our present study i s mainly concerned with a so-called embedded r e l a t i v e clause, 2.3 Ross's Constraints and R e l a t i v i z a t i o n 2,3,1 The CNCP and Japanese R e l a t i v i z a t i o n Let us turn to the claims of Japanese r e l a t i v i z a -t i o n proposed by Ross i n his d i s s e r t a t i o n . Ross argues i n h i s d i s s e r t a t i o n that Japanese r e l a t i v e clause formation must involve reordering, and not simple deletion because i t i s subject to the following constraints s 1 1 (a) the Complex NP Constraint (b) the Coordinate Structure Constraint (c) the Cross-over Condition proposed by P.M. Postal . 51 He gives us the following d e f i n i t i o n of the above constraintsi (19) The Complex NP Constraint (CNPC) No element contained i n a sentence dominated by a NP with a l e x i c a l head noun may be moved out of that NP by a transformation. 1 2 The above constraint i s explained i n the following diagram. (20) NP NP S (21) The Coordinate Structure Constraint (GSC) In a coordinate structure, no con-junct ^ may be moved, nor may any element i n a conjunct be moved out of the conjunct. 1-? (22) The Cross-over Condition No NP mentioned i n the structure index of a transformation may be reordered by that rule i n such a y way as to cross over a c o r e f e r e n t i a l NP. In CNPC, he thinks i t necessary f o r the theory of grammar to di s t i n g u i s h between l e x i c a l items l i k e •claim* 52 and an abstract pronoun * i t ' because t h i s d i s t i n c t i o n accounts f o r the grammaticality and ungrammaticality of the following sentences and t h i s d i s t i n c t i o n also holds even i n Japanese. (23) and (24) are from Ross. 1^ (23) (a) I believed the claim that Otto was wearing t h i s hat. (b) I believed that Otto was wearing t h i s hat. (24) (a) *The hat which I believed the claim  that Otto was wearing i s red. (b) The hat which I believed that Otto was wearing i s red. His Japanese t r a n s l a t i o n to (23) and (24) i s (25) and (26). (25) (a) Otto-ga kono boosi-o kabutte : i t a t h i s hat wearing-was to i u syutvoo-o watakusi-wa s i n z i t a . that say claim I believed 33 (b) Otto-ga kono b©osi-o kabutte i t a koto-o watakusi-wa s i n z i t a . thing ( 2 6 ) (a) *Otto-ga kabutte i t a to i u syutyoo-o watakusi-ga s i n z i t a boosi-wa akai. red (b) Otto-ga kabutte i t a koto-o watakusi-ga s i n z i t a boosi-wa a k a i . ^ As a conclusion, he argues that the Japanese nouns koto, mono, and no, which mean roughly 'thing* are non-l e x i c a l while nouns l i k e syutyoo 'claim' are l e x i c a l because the CNPC only prevents the elements with a lex-17 i c a l head NP from reordering. ' Let us observe the c r u c i a l points with regard to the CNPC; ( i ) the necessity of the existence of a fea-t u r e ^ LexJ and ( i i ) the subjection of Japanese r e l a -t i v i z a t i o n to the GNPG. F i r s t of a l l , we must note that Ross's example ( 2 6 . a ) i s perfect to native speakers. We can e a s i l y present another example against Ross. 54 (27) (a) ^[Bacon-ga 'Hamlet'-o k a i t a ] s syooko-o] N p wrote evidence 1+ Lex 3 boku-wa s i t t e i r u . I know ( L i t . ) • I know the evidence that Bacon wrote 'Hamlet*.• (b) [[Bacon-ga *Hamlet*-o k a i t a ] s k o t o - o ] N p thing boku-wa s i t t e i r u . ^" Lex} ( L i t . ) ' I know that Bacon wrote 'Hamlet*.' (28) (a) |[Bacon-ga 0 kaita^fg syooko-o] N p boku-ga L * Hamlet*] s i t t e i r u J s Hamlet* ( L i t . ) * ' 'Hamlet*, which I know the evidence that Bacon wrote ' (b) |{]]Bacon-fga 0 k a i t a ] s k o t o - o ] ^ boku-ga s i t t e i r u j s 'Hamlet*] N p ( L i t . ) ' 'Hamlet*, which I know that Bacon wrote ' It would be he l p f u l f o r us to add Kuno's following counter-examples to Ross's proposal saying that the element i n a r e l a t i v e clause can not be moved out of 5-5 the complex (29) (a) jjsono sinsi-wa k i t e iruj yoohuku-gaj the gentleman wearing-is s u i t i : yogorete i r u . £+ Lex] d i r t y - i s ( L i t . ) ' The su i t which the gentleman i s wearing i s d i r t y . * (b) |j)2f kite i r u ] s yoohuku-ga]Np wearing-is suit yogorete i r u ] s s i n s i J N p dirty-is gentleman ( L i t . ) .*• A gentleman who the suit that (he) is wearing is dirty.' (30) (a) jjsono sensei-wa osiete i t a ] s ^ e i t o - g ^ N p the teacher teaching-was student rakudaisita. [+ Lex] flunked ( L i t . ) • The student who the teacher was teaching flunked.' (b) [[p osiete i t a ] s s e i t o - g a ] N p teaching-was student 56 rakudai s i t a J s senseiJ Np flunked teacher ( L i t . ) * * the teacher who the student that (he) was teaching flunked.' If we assume Japanese r e l a t i v i z a t i o n to be a move-ment fransformation, we are tempted to doubt that the CNPC applies to Japanese. Notice that we have seen i n (26) (28) that at l e a s t i n Japanese r e l a t i v i z a t i o n , the d i s t i n c t i o n between [+ Lexi c a l ] and I- L e x i c a l J does not seem to work at a l l , though Ross claims that there i s a strong evidence f o r the existence of the d i s t i n c t i o n i n Japanese. Let us now observe Ross's claim that Japanese 19 r e l a t i v i z a t i o n i s also subject to the Cross-over Condition . fF i r s t , he argues that the Japanese version of pa s s i v i z a t i o n can not apply to a r e f l e x i v e sentence as 20 i n (31). Sentence (31) i s his example f o r his claim. (3D (a) Sono hito-^wa zibun^-o aratta. that man s e l f washed 57 ( L i t . ) • That mani washed h i m s e l f i . t (b) Zibun--wa sono h i t o ^ - n i arawareta. wash-passive-past ( L i t . ) • That man. was washed by himself..* We argue against Ross i n the following way« F i r s t , example (31.a) i t s e l f i s unacceptable to most native speakers. N.A. McCawley's Like-NP Constraint ex-pla i n s the ungrammatically of the sentence i n such a way that i Japanese has a syntactic constraint, the Like-NP Constraint, which discards the sentences as ungrammatical i f the r e f l e x i v e and i t s antecedent are i n peer r e l a t i o n s h i p . 2 1 I f we follow her, the P-marker of (31.a) would be represent-ed as (32). (32) S V sono hito sono h i t o aratta Notice NP 1 and NP2 are i n peer relat i o n s h i p , so NP, can not be r e f l e x i v i z e d nor can i t be l e f t untouched. 58 Because i f i t i s l e f t untouched, we have a reading that 'the person washed somebody else.' When we want to express Ross's English t r a n s l a t i o n i n idiomatic Japanese, we w i l l have the following sentence, L (33) Sono hito^-wa zibun^-no karada-o aratta. that man s e l f ' s body washed ( L i t . ) ' That man washed h i s body.' Second, note that the passive sentence of (32) i s neither (34) nor Ross's (31.b). (34) *Zibun^-no karada-wa sono h i t o ^ - n i s e l f ' s body that man-by arawareta* wash-passive-past ( L i t . ) ** Se l f ' s body was washed by that man.' As Muraki (1970) indicates, i t i s not yet proven that p a s s i v i z a t i o n should be a reordering transformation i n 22 Japanese. Rather many scholars claim that Japanese p a s s i v i z a t i o n i s not a reordering transformation on the ground that some passive sentences have no corresponding active sentences as i n (35). 59 (35) (a) Taroo-wa tuma-ni sinareta. (passive) (his)wife die-passive-past ( L i t . ) • Taro had (his) wife die on him.* (b) *Tuma-wa Taroo-o sinda. (active) died ( L i t . ) * ' (His) wife died Taro.' The underlying structure of (35.a) would look l i k e (36). (36) NP-ga V I i tuma sinu After the verb-raising and the subject-raising of S 2 into the matrix sentence, we could get (35*a) as a surface struc-ture. Therefore, i f the Japanese p a s s i v i z a t i o n i s not a reordering transformation, the Cross-over Condition w i l l become e n t i r e l y i r r e l e v a n t . Ross gives another example as evidence f o r the 6 0 a p p l i c a b i l i t y of the Cross-over Condition. This time, he uses Japanese r e l a t i v i z a t i o n f o r evidence. Observe hi s 2 3 following example . -^The underlined l e x i c a l items i n (37) are mine f o r the purpose of correction. (37) (a) fol/e'-ga fifa&i to it-ta h i t o ^ zibun^ ookii he t a l l that said man ( L i t . ) ' the man who said he was t a l l ' (b) * hito^-ga *i/^ eV£ to i t t a h i t o ^ ookii man t a l l that said man ( L i t . ) * the man^ who he^ s a i d was t a l l ' 2k He gives the following underlying structure of (37). (38) h i t o . o o k i i 6 1 He says, "the boxed NP can be r e l a t i v i z e d , although the c i r c l e d NP can not."" This i s not exactly c l e a r . As Muraki says, i f Japanese r e l a t i v i z a t i o n involves reorder-ing of a r e l a t i v i z a b l e NP i n the rightward d i r e c t i o n , the boxed NP can not be r e l a t i v i z a b l e . It must cross over the c i r c l e d NP i n i t s movement to the r i g h t . Rather the c i r c l e d NP should be r e l a t i v i z a b l e . Nevertheless, i n f a c t i f the c i r c l e d NP were to undergo r e l a t i v i z a t i o n , i t would generate (37.b) incor-r e c t l y , which i s ungrammatical. On the other hand, (37.a) i s acceptable when we correct ka-re ' he ' into zibun - ' s e l f ' , as i n (39). That i s , only the boxed NP can be r e l a t i v i z a b l e i n spite of the Cross-over Condition. In order to derive (39) correctly, we must postulate i t s process i n the following wayt 25 (39) NP (i?) The r e f l e x i v i z a t i o n of the c i r c l e d NP under i d e n t i t y with the boxed NP which 62 meets the Subject-antecedent Condition* The  Subject-antecedent Condition requires that J The r e f l e x i v e refers back to the subject i n the same simplex sentence or the subject i n any higher sentence, ( i i ) The r e l a t i v i z a t i o n of the boxed NP as a deletion process. If we conclude that r e l a t i v i z a t i o n i s a d e l e t i o n transformation, the Cross-over Condition has nothing to do with r e l a t i v i z a t i o n . If we assume r e l a t i v i z a t i o n i s reordering even i n Japanese, and believe i n the a p p l i c -a b i l i t y of the Cross-over Condition to Japanese (whose examination i s beyond the scope of our present study), we block a grammatical sentence (39) i n c o r r e c t l y . 27 Thus, we must r e j e c t Ross and others 'with respect to t h e i r r e l a t i v i z a t i o n as a movement transformation. If Japanese r e l a t i v i z a t i o n i s only a deletion process, the CNPC i t s e l f has nothing to do with i t . We can explain the deletion of the r e l a t i v i z e d c i r c l e d NP i n (40) by say-ing that the c i r c l e d NP i s deleted as r e l a t i v i z a t i o n 63 under i d e n t i t y with the boxed NP. (40) S NP V kabutte i t a 2.3.2 THE CSC AND JAPANESE RELATIVIZATION We w i l l now examine Ross's CSC (the Coordinate Structure Constraint). According to him, a coordinate structure i s defined "as any structure conforming to the schematic diagram i n (41)", although a conjunction i s a 64 more abstract, language-independent representation of and or or, and t h i s abstract conjunction "should be understood as either preceding a l l i t s conjuncts, as i n English, French, etc., or as following them, as i n Japanese"* 2® (41) A After the applications of conjunction copying, where a conjunction i s Chomsky-adjoined to each conjunct, and the deletion of the f i r s t conjunction to (41), we could get an ordinary coordinate construction, as i n (42). (42) I went to the store and Mike bought some whisky. Now, i f 'some whisky' i n (42) i s moved out of the second conjunct by r e l a t i v i z a t i o n , i t generates an ungrammatical sentence (43). (42) and (43) are from Ross. 2^ (43) *Here's the whisky which I went to the store and Mike bought 0 . 65 S i m i l a r l y , the underlined elements i n (44.a) can not be moved out of each conjunct by r e l a t i v i z a t i o n or question formation. (44) (a) The nurse polished her trombone and the plumber computed my tax. (b) *The plumber who the nurse polished her  trombone and 0 computed my tax. Also, an NP which i s a conjunct i n a coordinate NP structure can not be r e l a t i v i z e d nor be reordered by WH-question formation. (44) and (45) are from Ross.^ 0 (a) He w i l l put the chair between [ N p [ N p some t a b l e ] ^ a n d [ N p some s o f a ] N p ] N p (b) »What sofa w i l l he put the chair between some table and 0 ? •Thus, he proposed h i s so-called universal constraint (CSC) c i t e d on page 51 i n t h i s chapter. 66 Would t h i s CSC be applicable to Japanese? F i r s t , i t must be noted that Japanese WH-question formation i s i r r e l e v a n t concerning the CSC because a WH-question word replaces the NP i n question without involving a movement transformation, as i n (46), whioh i s the tr a n s l a t i o n of (45). (46) (a) Kare-wa teeburu to sohaa-no aida-ni he table and sofa between sono isu-o oku. the chair put dono sohaa what sofa kal Q.P. Q.P, = question p a r t i c l e We w i l l now examine the r e l a t i v i z a t i o n of an NP i n a coordinate NP construction. For example, take (46,a) and r e l a t i v i z e , respectively, the conjunct NP's. (47) (a) *[fcare-ga fteeburu to 0 - no]aida-ni sono isu-o oku} sphaa 67 ( L i t . ) *• the sofa which he w i l l put the chair between some table and jS ' (b) *[kare-ga £ 0 to sohaa—no]aida-ni isu-o oku] teeburu ( L i t . ) *' tha table which he w i l l put the chair between 0 and some sofa * Sim i l a r l y , observe the following sentences: (48) (a (b (c ( W (a (b [Taroo to Hanako-waj nihongo-o hanasu. and Japanese speak L i t . ) * Taro and Hanako speak Japanese.' * j^Taroo to 0 -ga ] nihongo-o hanasuJ Hanako L i t . ) *' Hanako, who Taro and 0 speak Japanese* to Hanako-ga ] nihongo-o hanasu J Taroo L i t . ) * ' Taro, who 0 and Hanako speak Japanese ' j"Taroo ka Hanako-ga] soko-e i t t a . there went or L i t . ) ' E i t h e r Taro or Hanako went there.' * [ f T a r o o ka 0 -ga] soko-e i t t a ] Hanako L i t . ) *' Hanako, who eithe r Taro or 0 went there ' (c) * ka Hanako-gaj soko-e i t t a ]Taroo ( L i t . ) *' Taro, who 0 or Hanako went there ' The above examples show that the CSC does apply to the r e l a t i v i z a t i o n of the conjunct NP i n a coordinate NP construction. Next, observe the following sentences t (50) (a) Taroo-wa piano-o h i k i (sosite) Hanako-wa play and uta-o utatta. song sang ( L i t . ) • Taro played the piano and Hanako sang.' (b) * fTaroo-ga piano-o h i k i (sosite) 0 uta-o u t a t t a j Hanako ( L i t . ) * * Hanako, who Taro played the piano and 0 sang.* (c) * \0 piano-o h i k i (sosite) Hanako-ga uta-o utatta JTaroo ( L i t . ) * ' Taro, who played the piano and Hanako sang.* 69 ( 5 D (a) Taroo-wa kuruma-de dekaketa ga boku-wa car-by went-out but I u t i - n i i t a . home-at was ( L i t . ) ' Taro went out by car, but I stayed at home.' (b) *£raroo-wa 0 dekaketa ga boku-wa u t i - n i i t a j kuruma ( L i t . ) * ' the car by which Taro went out and I stayed at home ' As i s shown above, the NP i n a coordinate sentence construction i s also subject to the CSC, This i s due to the f a c t that the head NP i s semantically related only to the conjunct i n which the i d e n t i c a l NP i s contained, and not to the other conjunct. For example, i n (51.b), the head NP kuruma 'car' has nothing to do with the second conjunct boku-wa u t i - n i i t a 'I stayed at home'. Si m i l a r l y , i n (48.b) the head NP Hanako has the r e l a t i o n with only part of the r e l a t i v e clause — that i s , nihongo  -o hanasu 'speak Japanese', but Taroo i n the r e l a t i v e clause does not constitute any relevant part of the 70 description of the head NP Hanako. We must conclude that the fact that the above examples show the a p p l i c a b i l i t y of the CSC to Japanese i s not evidence f o r Ross's claim that Japanese r e l a t i v i z a -t i o n i s a movement transformation. Indeed, the r e l a t i v -i z a t i o n of only one NP i n a coordinate structure does not generate a meaningful r e l a t i v e clause f o r the head NP. If the r e l a t i v i z e d NP occurs both i n the two con-juncts and i f these two NP's are i d e n t i c a l concerning gram-matical functions, then the derived sentence i s grammatical, and the r e l a t i v e clause constitutes an appropriate descrip-t i o n of the head NP as Makino s t a t e s ? 1 For example, observe the following sentences! (52) (a) Taroo-ga sono sake-o katta ka Ziroo-ga the bought or sono. sake-o Hanako-kara moratta. from received ( L i t . ) ' Taro bought the sake or Ziro received i t from Hanako.' (b) Kore-wai[Taroo-ga 0 katta ka Ziroo-ga 71 ( L i t . ) * This i s the sake which Taro bought 0 or Ziro received 0 from Hanako.' (53) (a) Taroo-wa sono hon-o katta ga Hanako-wa the book bought but sono hon-o kawanakatta. the book buy-not-past ( L i t . ) ' Taro bought the book, but Hanako did not (buy i t ) . ' (b) Kore-wa[Taroo-wa 0 katta ga Hanako-wa 0 t h i s ~~ kawanakatta] hpn desu. book i s ( L i t . ) ' This i s the book which Taro bought but Hanako did not.• This f a c t i s explained by Ross i n such a way that a r e l a t i v e clause i s not subject to the CSC when i t works 'across the 32 board*. According to Ross, an across-the-board-rule i s the one which applies to a l l the conjuncts simultan-eously when some element i s contained i n a l l the conjuncts. Therefore, as i s shown i n the following examples, r e l a t i v -i z a t i o n applies to the NP's i n both conjuncts 'across the board', regardless of t h e i r grammatical functions. (54) (a) Saisyo-ni Taroo-ga Ziroo-o nagutta ka f i r s t h i t or 72 soretomo Ziroo-ga saisyo-ni Taroo-o ketta. (or) f i r s t kicked ( L i t . ) * F i r s t Taro h i t Ziro, or Ziro kicked Taro f i r s t . • (b) ^Saisyo-ni Taroo-ga 0 nagutta ka soretomo 0 Taroo-o saisyo-ni k e t t a j Ziroo-wa kanzen-ni completely yopparatte i t a . drunk-was ( L i t . ) ' Ziro, whom Taro f i r s t h i t or who kicked Taro f i r s t , was completely drunk.' ( 5 5 ) (a) Taroo-wa t i i s a i ga boku-wa sumoo-de short but I i n Sumo kare-ni katenai. him defeat-can-not ( L i t . ) ' Taro i s short, but I can not defeat him i n Sumo wrestling.' (b) t i i s a i ga boku-ga sumoo-de 0 katenaiJ Taroo ( L i t . ) * Taro, who i s short but whom I can not defeat i n Sumo wrestling ' F i n a l l y * observe the following sentencesi 73 (56) (a) (Boku to Taroo)) and (Bokutati) -ga sakaya-ni i t t e Taroo-ga "we ^ " liquor-store to went and sake-o katta. bought ( L i t . ) • ( l a n d Taro) w e n t t Q t h e l i q u Q r g t o r e and Taro bought some sake.• Taroo-ga katta]sake desu. < L i t ' ) * This i s the sake which \ Z ^ T a r o ) went to the store and Taro bought.* One may say that (56) i s a kind of ostensibly con-joined structure from which items can be moved out. J.D. McCawley says» Ostensibly conjoined structures i n English have p a r a l l e l s i n Japanese, or rather these - o structures are more f r e e l y allowed i n Japanese. Ostensibly conjoined structures are not c l e a r l y defined i n Ross's d i s s e r t a t i o n , but judging from h i s and McCawley's examples, the structures i n question are exemplified i n (57) and (58), which are respectively from Ross and McCawley.-' 74 (57) (a) I went to the store and bought some whisky. (= I went to the store to buy some whisky.) (b) Here's the whisky which I went to the store (and bought). ( to buy ). (58) (a) Taroo-wa daigaku-e i t t e , boodoo-o okosita. university-to go r i o t caused ( L i t . ) ' Taro went to the uni v e r s i t y and started a r i o t . ' (b) Taroo-ga daigaku-e i t t e , okosita boodoo ( L i t . ) * the r i o t which Taro went to the u n i v e r s i t y and started ' In these cases, the subject NP's are i d e n t i c a l i n a l l the conjuncts. Further, according to Ross, the syntactic indications of ostensibly conjoined structures are as f o l l o w s x J J a, It i s only with a non-stative verbs as the main verb of the second conjunct that sentences (4.101.a) can be constructed. b. The second conjunct can not be negative. 75 c. There are r e s t r i c t i o n s on the tenses that may appear i n such sentences as (4.101.a). In the above quotation, the sentence (4.101.a) i s the sentence (57.b) here. These syntactic indications do not hold i n Japanese. For example, we can say i n the following wayi (59) (a) Taroo-ga sakaya-ni i t t a ga sono sake-o liquor-store-to went but the kaenakatta. buy can-not-past ( L i t . ) ' Taro went to the store, but (he) could not buy the sake.* (b) Kore-wa ^Taroo-ga sakaya-ni i t t a ga 0 t h i s kaenakatta! sake desu. i s ( L i t . ) * This i s the sake which Taro went to the store but could not buy 0 (60) (a) Taroo-wa London-ni ryuugaku s i t e i t a i n study-abroad doing was kare-wa sono kan hotondo he the time-for almost ( ga ) (s i k a s i ) but 76 noirooze datta. C+ stative J neurosis was ( L i t . ) ' Taro studied i n London, but he was ,. almost a neurotic during that period.' (b) |jzf London-ni ryuugaku s i t e i t a ga 0 sono kan hotondo noirooze datta 1 Taroo-ga tootoo i + s t a t i v e 3 f i n a l l y kaette k i t a . returned ( L i t . ) ' Taro, who studied i n London, but was nearly neurotic during that period, f i n a l l y came back (to Japan).' Furthermore, take (56) again. In (56.b) the subjects i n each conjunct are not p e r f e c t l y i d e n t i c a l . Therefore, I doubt (56) i s an ostensibly conjoined structure, and I think t h i s example constitutes a counter-example to Ross's CSC. Nevertheless, we must admit that the CSC does apply to Japanese i n most cases, and at the same time the constraint must be re-examined and revised i n order to hold to a l l the examples i n Japanese. It must be pointed out that the f a c t that Japanese 77 r e l a t i v i z a t i o n i s , i n most cases, subject to the CSC i s not the evidence f o r the claim that Japanese r e l a t i v i z a — t i o n i s not a deletion process but a movement trans-formation. Whether or not we assume i t to be a movement transformation, the ungrammaticality of the sentences which v i o l a t e the CSC comes from the semantic anomaly of the derived r e l a t i v e clauses. Further, there i s no reason nor any syntactic evidence f o r us to conclude that because the CSC i s applicable to Japanese r e l a t i v -i z a t i o n , i t must not be a deletion transformation. Rather, i t i s quite natural f o r us to believe on the basis of the investigation of the CNPC that i t i s a dele-t i o n process. 2.4 CONCLUSION What we have discussed i n t h i s chapter are as followsi a. some c h a r a c t e r i s t i c s of Japanese r e l a t i v e constructions b. the contrastive study of English and Japanese r e l a t i v i z a t i o n 78 c. Ross's hypothesis about Japanese r e l a t i v -i z a t i o n i n connection with his CNPC and CSC and Postal's Cross-over Condition, . We have shown that i n Japanese, r e l a t i v i z a t i o n must meet the i d e n t i t y condition, exactly as i n English. However, because of the lack of r e l a t i v e markers, Japanese has no sentence-final NRR-construction. Instead of that con-struction, we use a juxtaposed coordinate construction i n Japanese. Concerning an i n t r a s e n t e n t i a l NRR-construction, Japanese has no such construction, either, f o r the same reason mentioned above, namely, that there i s no r e l a t i v e marker. Instead, we use e i t h e r an RR-construction or a coordinate construction which contains the frame ga i n t r a s e n t e n t i a l l y . The semantic r o l e of t h i s coordinate construction seems to be s i m i l a r to that of the English i n t r a s e n t e n t i a l NRR-construction. In section (2.3), we have said the following. ( i ) Japanese r e l a t i v i z a t i o n i s not a movement transformation but a d e l e t i o n process. ( i i ) As Japanese r e l a t i v i z a t i o n does not involve a movement transformation, the Cross-over Con- d i t i o n i s ir r e l e v a n t to i t . ( i i i ) CNPC i s also i r r e l e v a n t to i t f o r the same reason. (iv) Japanese examples do not support Ross's d i s t i n c t i o n between l e x i c a l head nouns and non-l e x i c a l head nouns i n h i s CNPC. (v) In most (but not a l l ) cases, Ross's CSC i s applicable to Japanese r e l a t i v i z a t i o n . But that f a c t i s not evidence f o r a reordering r e l a t i v i z a t i o n hypothesis. 80 CHAPTER II NOTES 1. Every p a r t i c l e i s not deletable i n r e l a t i v i z a t i o n . Some p a r t i c l e s l i k e kara 'from' are deletable i n some cases, but not i n others. Observe the following sentences. ( i ) (a) Sono mado-kara-wa yama-ga mieru. the window from mountain v i s i b l e ( L i t . ) '(As f o r ) from that window, the mountain i s v i s i b l e . ' (b) uama-ga mieru] mado ( L i t . ) • the window from which the mountain i s v i s i b l e . ' ( i i ) (a) Vancouver-kara-wa Hanako-ga k i t a . came ( L i t . ) ' As f o r from Vancouver, Hanako came.' (b) *JHanako-ga k i t a ] Vancouver ( L i t . ) ' Vancouver from which Hanako came.' 2. Kuno, Notes on Japanese Grammar, Chap. 18, p. 2. 3. Schaehter, "Focus and R e l a t i v i z a t i o n , " p. 43. 4. Sandra A. Thompson, "The Deep Structure of Relative Clauses," Studies i n L i n g u i s t i c Semantics, ed. Charles J . Fillmore and D. Terence Langendoen (New York: Holt, Rinehart and Winston, Inc., 197D» p. 84. 5. Andreas Loetscher, "On the Role of Nonrestrictive Relative Clause i n Discourse," Papers from the Ninth  Regional Meeting. (Chicago: Chicago L i n g u i s t i c Society, A p r i l , 1973), P. 366. 6. Ibid., p. 363. 7. Ibid., pp. 365 f . 8. Ibid., p. 366. 9. Sentences (l6,a) and (l6.b) are from Loetscher. Ibid., pp. 361 & 366. 81 10. Sandra Annear, "Relative Clauses and Conjunctions," Working Papers i n L i n g u i s t i c s , (Columbus: Ohio State University, December, 1967)» pp. 8 0 ^ 9 0 . 11. Ross, "Constraints on Variables i n Syntax," p. 72. 12. Ibid., » P. 70. 13. Ibid., r p. 89. 14. Ibid. , P. 73. 15. Ibid., , P. 70. 16. Ibid., , P. 75. 17. Ibid., , p. 76. 18. Kuno, Notes on Japanese Grammar. Chap. 18. p. 8 & Chap. 19. P. ,7. 19. Ross, op. c i t . , p. 73. 20. Ross, Ibid., p. 74. 21. Noriko A. McCawley, "A Study of Japanese R e f l e x i v i z a -t i o n . " (unpublished Ph. D. d i s s e r t a t i o n , I l l i n o i s University, 1972), p.30. The notion of 'peer* i s due to Postal. He defines the NP's which are peers i n the following way: "Two NP, NP, and NP2, neither of which dominates the other nor i s co-ordinate with the other i n a phrase marker P are peers with respect to a node S., just i n case the paths between each of these NP and S. are such that they contain no NP-nodes not seperated from the s t a r t i n g point NP, NP, or NP2, by a node S." 1 See Paul M. Postal, Cross-over Phenomena. (New York: Holt, Rinehart and Winston, Inc., 1971), p. 179. 22. Muraki, "Presupposition, Pseudo-clefting-and Thematization," pp. 34 f ; 23. Ross, _op. c i t . , p. 74. 24. Ibid., 25. Ibid. 82 26. N.A. MeCawley, op. c i t . , p. 4. 27. For example, Makino assumes that Japanese r e l a t i v -i z a t i o n i s a movement transformation. His r e l a t i v i z a -t i o n rule i s as follows! where: 2 = 5 C+DCLJF s e n t e n t i a l feature {+ declarative3 T 0b= obligatory transformation See Makino, Some Aspects of Japanese Nominalizations, pp.157 f . 28. Ross, op. c i t . , p. 89. 29. Ibid., p. 93. 30. Ibid., pp. 14 & 89. 31. Makino, op. c i t . , p. 157. 32. Ross, Ibid., pp.96~98. 33. James D . MeCawley, "Japanese Relative Clauses," The  Chicago Which Hunt, (Chicago: Chicago L i n g u i s t i c Society, A p r i l , 19727, pT"207. 34. See Ross, op. c i t . , p. 93 & MeCawley, Ibid., p. 207. 35. Ross, Ibid., p. 93. 1 2 3 4 5 6 CHAPTER I I I JAPANESE RELATIVIZATION AND MURAKI'S HYPOTHESES 3.1 POINTS OF DISCUSSION In t h i s chapter we w i l l discuss Muraki*s claims (1970) which are shown below. 1 a. The underlying structure of a Japanese r e l a t i v e clause i s Chomsky-adjoined to the rig h t of the head NP, exactly as i n English. b. Japanese r e l a t i v i z a t i o n i s a copying r u l e . c. After copying, the Nearest NP Constraint applies to the r e l a t i v i z e d NP i n the clause. 3.2 UNDERLYING RELATIVE CLAUSES AND THE NEAREST NP CONSTRAINT Muraki r e j e c t s the usual P-marker of an under-84 l y i n g r e l a t i v e construction i n Japanese which i s repre-sented as (1). Instead, he claims that (2) i s the cor-rect underlying structure. ^TP± NP i One of the main reasons f o r his underlying struc-ture i s connected with Ross's claims (1967) that Japanese r e l a t i v i z a t i o n i s a movement transformation and the Cross- over Condition applies to Japanese r e l a t i v i z a t i o n . Please r e f e r to Ross's example again on p. 60 i n chapter II of t h i s t h e s i s . (3) JZ^zibun-ga takai to i t t a h i t o ^ s e l f t a l l that said ( L i t . ) ' the person^ who said he^ i s t a l l ' NP«-ga V 85 Suppose we accept Ross's hypothesis stated above. As we have already seen i n the previous chapter, i n or-der to generate sentence ( 3 ) , the NP2 i n (4) must move to the r i g h t . But t h i s i s a v i o l a t i o n of the Cross-over  Condition because the NP 2 crosses over the i d e n t i c a l NP^ i n i t s movement. In order to avoid t h i s v i o l a t i o n and generate (3)» the only a l t e r n a t i v e way i s to postulate that the NP 2 i n S^ moves not to the right but to the l e f t , away from the matrix head NP. The r e l a t i v i z a t i o n process would be: ( i ) The leftward.' movement of NP2, (which i s vacuous here)• ( i i ) The deletion of the NP 2 under i d e n t i t y with the head NP-^ . This process, according to Muraki, involves two unnatural steps; (a) the movement of the NP2 away from the i d e n t i c a l head NP^ , and (b) the unnaturalness of " the i d e n t i t y deletion which operates leftward (that i s , backward) across the constituent sentence." These two unnatural steps are his main reasons why he re j e c t s the 86 underlying structure (1). He says that i n order to avoid the unnaturalness of the above steps we must postulate (2) as the underlying structure of a Japanese r e l a t i v e construc-t i o n . The remote structure of (3) w i l l be represented as (5) i f we follow Muraki.3 (5) NP NP 1 S hito NP.-ga takai The d e r i v a t i o n a l process of (3) from (5) would bet ( i ) The r e f l e x i v i z a t i o n of NP 3 ( i i ) The r e l a t i v i z a t i o n of NP2 (a) The leftward movement of NP0 without a v i o l a t i o n of the Cross-over Condition, here, vacuous (b) The deletion of NP2 by the Nearest NP Constraint ( i i i ) An o b l i g a t o r y r e l a t i v e clause p r e p o s i n g which makes the r e l a t i v e c lause precede the head NP. As you can see from the above example, even i f we were t o assume t h a t Ross's claims are r i g h t , we co u l d generate (3) without any v i o l a t i o n s of the Cross-over  C o n d i t i o n . Here, Muraki proposes a c o n s t r a i n t known as the Nearest NP C o n s t r a i n t . T h i s c o n s t r a i n t a l l o w s the d e l e t i o n of an NP i n a r e l a t i v e c l a u s e which i s n e a r e s t to and c o r e f e r e n t i a l with the head NP. T h i s n e a r e s t NP should be " the n e a r e s t not only i n the l i n e a r arrange-ment, but a l s o i n terms of the number of i n t e r v e n i n g nodes." Now, l e t us examine h i s s o - c a l l e d u n n a t u r a l s t e p s . F i r s t , h i s argument t h a t the r e l a t i v i z e d NP must move l e f t w a r d away from the head NP i s by no means p e r s u a s i v e because he h i m s e l f admits the claim.that " s i n c e r e l a t i v e i z a t i o n i s not a movement t r a n s f o r m a t i o n , t h i s argument does not h o l d . " ^ R e l a t i v i z a t i o n as a d e l e t i o n process has n o t h i n g t o do wit h the Cross-over C o n d i t i o n because we do 88 not have to move a r e l a t i v i z e d NP. As f o r his second argument, he says that "the other argument that the backward deletion across the sentence i s unnatural s t i l l holds" even i f r e l a t i v i z a t i o n i s not a movement transformation.^ The reason why t h i s backward deletion i s unnatural i s not c l e a r to us. The only example he gives us i s the following (6) (a) *musume-ga zibun^-o korosi-ta otoko^ daughter s e l f k i l l e d man ( L i t . ) * the person^ whose^ daughter k i l l e d him^' (b) * [zibun^-no] musume-ga 0 k o r o s i - t a otoko^ s e l f ' s daughter 0 k i l l e d man ( L i t . ) the person^ whose., daughter k i l l e d him^ 1 He says the reason why sentences (6.a) and (6.b) are un-grammatical i s that these sentences v i o l a t e the .Nearest NP Constraint. Note his underlying structure (7) of (6. a,b); 8 (7) otoko^ otoko^ korosi-ta otoko. 89 Neither NP^ nor NP^ can be deleted. This i s because they do not meet the Nearest NP Constraint. S p e c i f i c a l l y NP^ i s not nearest to NP2 i n the l i n e a r arrangement, and NP^ i s not nearest, either, i n terms of the number of i n t e r -vening nodes. But i n spite of t h i s f a i l u r e to f u l f i l l the Nearest NP Constraint, NP^ i s deleted i n (6.a), and NPj^  i s deleted i n (6.b). Therefore, the derived sentences are unacceptable. According to Muraki, instead of ( 6 ) , native speakers use (8) i n t h e i r speech, where the co-r e f e r e n t i a l NP^ which meets t h i s constraint i s r e l a t i v i z e d and deleted^ Figure (9) i s his underlying structure of phrase ( 8 ) . (8) [ s [gZibun^-no] musume-ni koros-are-ta] otoko^ l s e l f ' s daughter-by killed-be-past man ( L i t . ) ' the mani who.^  was k i l l e d by h i s i own daughter ' (9) otoko^ NP, otoko koros-are NP/--no musume I 6 otoko^ 90 The above i s his explanation f o r the necessity of the  Nearest NP Constraint. As a conclusion, he says, If the r e l a t i v e clause did not follow the head noun, the constraint would have to be more complicat-ed. It would have to block the r e l a t i v i z a t i o n of any noun unless i t i s l i n e a r l y most distant from but h i e r a r c h i c a l l y (that i s , i n terms of nodes) nearest to the c o r e f e r e n t i a l head noun phrase. Since the Nearest NP Constraint i s much simpler, i t supports the argument that the r e l a t i v e clause should follow the head noun i n the underlying s t r u c t u r e . 1 0 Let us examine his argument. To begin with, i t can not be denied that (6) might be unacceptable f o r the following reasons. ( i ) When one speaks about a person, e s p e c i a l l y a sufferer, i t i s natural to talk about the person from the point of view of the victim, that i s , to speak of him as the subject of a sentence. Therefore, i n t h i s case when we want to speak of 'the man' i n the form of a r e l a t i v e con-struction, the r e l a t i v e clause w i l l be n a t u r a l l y i n the passive form with 'the man' as the sub-ject. 91 ( i i ) Please r e f e r back to sentences (6.a) and (6.b) and Muraki's tree diagram (7). Whether or not we follow his underlying structure, r e f l e x i v i z a t i o n i n (7) i s blocked anyway because i t does not s a t i s f y the Subject-antecedent Condition. 1 1 Now, we could explain the grammatical nature of (8) without the help of the Nearest NP Constraint. We consider that sentence (8) can be derived from (10) following the steps s p e c i f i e d below. (10) NP, I rareta V otoko S •2 korosu NP8-ga V musume otoko aru The n e c e s s a r y steps ( i ) The d e l e t i o n r e l a t i v i z a t i o n of NPg ( i i ) The g e n i t i v e f o r m a t i o n i n which con-v e r t s z i b u n - n i aru i n t o zibun-no ( i i i ) The r e f l e x i v i z a t i o n of NP^ under the i d e n t i t y with NP^, by the S u b j e c t - a n t e - cedent C o n d i t i o n ( i v ) The equi-NP d e l e t i o n of NP^ under i d e n -t i t y with NP^ (v) The v e r b - r a i s i n g of korosu, which combines korosu and r a r e t a ( v i ) The t r e e - p r u n i n g of Sg. which makes the complement s u b j e c t zibun-no musume a c o n s t i t u e n t of the h i g h e r sentence S^ ( v i i ) The backward d e l e t i o n r e l a t i v i z a t i o n of NP Secondly, we doubt the adequacy of the Nearest NP  C o n s t r a i n t . Does the c o n s t r a i n t c o n s t i t u t e a support f o r h i s u n d e r l y i n g s t r u c t u r e o f a r e l a t i v e c l a u s e ? Observe my f o l l o w i n g sentences: 93 (11) jjj&ibun^-ga undaj kodomo-dake-ga 0 kawaii} s e l f g a v e - b i r t h - t o c h i l d o n l y d e a r - i s Hanako^ ( L i t . ) * Hanako., t o whom only the c h i l d she. bore i s dear. / Hanako., who l i k e s o n l y the c h i l d she^ gave b i r t h t o . ' (12) [Qjcaze-de n e t u -ga 0 a t t e ] -mo 0 h u r o - n i a c o l d - w i t h f e v e r i s even bath h a i r u j koto-ga 0 s u k i n a j Taroo take t h a t l i k a b l e ( L i t . ) ' Taro., t o whom t a k i n g a bath i s l i k a b l e even I f he. has a f e v e r with a c o l d . , / Taro^, wno l i k e s t a k i n g a bath ... These sentences seem t o be p e r f e c t t o n a t i v e speakers. Take (11) as an example. The u n d e r l y i n g s t r u c t u r e would be re p r e s e n t e d as (13) i f we f o l l o w Muraki. (13) Hanako. kodomo unda 94 In figure (13), NP^ does not s a t i s f y the Nearest NP Constraint l i n e a r l y and NP^ does not s a t i s f y i t hier a r c h i c a l -l y . Therefore,; they can not be deleted by r e l a t i v i z a t i o n i f the Nearest NP Constraint i s correct. But i n f a c t , NPj^  must be deleted to generate phrase (11). Sim i l a r l y , observe (14). This represents the under-l y i n g structure of phrase, (12) when we follow Muraki's argument. (14) S-mo NPQ-ga. NPi.-ni A 3 I 4 N P c - d e N P r - g a N P„-ni V l 5 |6 i 7 i kaze netu Taroo. aru Taroo^ suki da NPR-ga NP Q-ni V i ° i 9 i Taroo. huro hairu In (14), none of the p r e f e r e n t i a l NP^, NP^, and NPg are deletable because of the f a i l u r e to s a t i s f y his con-95 s t r a i n t . In other words, the Nearest NP Constraint blocks grammatical phrases (11) and (12) i n c o r r e c t l y . I t must be noted that i n figures (13) and (14), we do not postulate that the underlying r e l a t i v e clauses con-t a i n the thematic NP's. This i s because Muraki argues against Kuno's hypothesis that a r e l a t i v i z a b l e NP must be thematic, and he himself does not use the underlying r e l a t i v e structure with a thematic NP. Now, i f we admit that the underlying structure of (11) contains a thematic NP, we can generate phrase (11) without Muraki's structure and h i s constraint. Figure (15) i s the usual underlying structure of phrase (11) and i t contains thematic NP's, NP^ and NPg. To generate (11) from figu r e (15), the necessary transformations would be as follows: ' ( i ) The deletion of NP g under i d e n t i t y with NPg. ( i i ) The deletion of NPg by r e l a t i v i z a t i o n . ( i i i ) The r e f l e x i v i z a t i o n of NP ? under i d e n t i t y with a thematic NP^ by the Theme-antecedent Condition proposed by M. Ohso, who says that the theme of a sentence can be the antecedent 12 of a r e f l e x i v e pronoun. 96 (iv) The deletion of NP^ by r e l a t i v i z a t i o n . (15) Hanako NP^-dake-ga (sono)kodomo NP 7-ga NPo-o r i Hanako kodomo V I und a Hanako kawaii Thus, we a r r i v e at phrase (11). It i s true that the thematic NP^ i s l i n e a r l y most distant from, but h i e r a r c h i c a l l y nearest to the head NP2, as Muraki indicates, and that to delete t h i s NP^, the deletion operation applies backward across S 2 and S^. But why i s t h i s kind of backward deletion unnatural? We are 97 not given s u f f i c i e n t j u s t i f i c a t i o n f o r t h i s kind of un-naturalness. Observe the following example where back-ward deletion across the sentence occurs. (16) (a) kensonsite [zibun^-ga binboo da] to humbly s e l f poor i s that Hanako-ni i t t a j koto-ga kekkateki-ni-wa said that r e s u l t i n g l y Taroo^-no s i t u r e n - n i tunagatte i r u . disappointed-love-with connected-is ( L i t . ) • That Taro humbly said to Hanako (that) he was poor i s as the r e s u l t related to his sisappointed love.* / "That Taro said humbly to Hanako that he was poor resulted i n his f a i l u r e to win her.' (b) [raroo^-ga kensonsite [zibun^-ga binboo daj to Hanako-ni i t t a ] koto-ga kekkateki-ni-wa kare^-no si t u r e n - n i tunagatte i r u . Figure (17) on the next page represents the under-l y i n g structure of sentences (16.a) and (l6.b). The necessary operations f o r deriving (l6.a) from (17) would be» ( i ) The r e f l e x i v i z a t i o n of NP 1 Q under i d e n t i t y with NP,- by the Subject-antecedent Condition. 98 ( i i ) The deletion r e l a t i v i z a t i o n of NP g. ( i i i ) The genitive formation i n (ni aru —* no), (iv) The deletion of NP^ under i d e n t i t y with NPrt Taroo-no. zibun^ Here, the backward deletion across sentence; i s supposed to occur i n ( i v ) . If we do not postulate the existence of t h i s backward deletion, we derive only (16.b) and not (16.a). Sentence (l6.b) would be d e r i v e d by the f o l l o w i n g t r a n s f o r m a t i o n s t ( i ) i—* ( i i i ) The same as (l6.a). ( i v ) The p r o n o m i n a l i z a t i o n of NP^ under i d e n t i t y with NP^. In the above step ( i v ) , r e f l e x i v i z a t i o n does not apply t o NP<j because the antecedent NP^ does not s a f i s f y the Command C o n d i t i o n of r e f l e x i v i z a t i o n which i s propos-1 3 ed by N.A. McCawley. T h i s c o n d i t i o n r e q u i r e s t h a t the antecedent of a r e f l e x i v e not o n l y be the s u b j e c t of a sentence but must a l s o command the r e f l e x i v e . The above example shows t h a t backward d e l e t i o n a c r o s s the sentence i s used i n Japanese. F i n a l l y , b e s i d e s a r e l a t i v e c o n s t r u c t i o n , we have another n o m i n a l i z i n g c o n s t r u c t i o n , where a sentence precedes a s e t o f n o n - a b s t r a c t n o m i n a l i z e r s (Makino's terminology) such as hazu ' i n e v i t a b i l i t y ; e x p e c t a t i o n ' , keikoo' tendency*, s a i t y u u 'the midst o f e t c . , o r a Ik s e t o f a b s t r a c t n o m i n a l i z e r s l i k e no, koto * t h i n g *. 100 Observe the following sentence: (18) [[Taroo-ga soko-e iku ] s hazu-ga J there go reason not-is ( L i t . ) • There i s no reason that Taro goes there. /Taro i s not expected to go there.' As Makino says , x : > t h i s kind of construction i s not supposed to be a r e l a t i v e construction because a r e l a t i v e construction must meet the i d e n t i t y condition i n the under-l y i n g structure, but i n (18) t h i s i s not the case. That i s , the embedded sentence i n (18) does not have the i d e n t i c a l NP hazu. However, i n spite of t h i s syntactic difference, t h i s construction i s s i m i l a r to a r e l a t i v e construction i n that both constructions constitute nominalizing con-structions as a whole. Therefore, i t i s natural to pos-tulate that the underlying structures of both constructions are s i m i l a r i n some respects, as i n (19). (19) (a) (b) NP NP 101 However, Muraki*s underlying structures f o r these two constructions seem to be quite d i f f e r e n t . Observe (20). (20) (a) (b) Therefore, his derivation of r e l a t i v i z a t i o n requires an extra step — — that i s , a r e l a t i v e - c l a u s e preposing. As a conclusion, we are i n c l i n e d to r e j e c t his underlying structure f o r r e l a t i v e constructions and con-sider his Nearest NP Constraint vacuous. We would rather subscribe to the structure represented i n (19.b) as the correct underlying structure f o r Japanese r e l a t i v e con-structions. 3.3 COPYING RELATIVIZATION In t h i s section, we w i l l examine Muraki's claim that r e l a t i v i z a t i o n i s a copying transformation. He says i n his d i s s e r t a t i o n , "Thematization and 102 r e l a t i v i z a t i o n are both copying r u l e s . " 1 ^ . Observe the following sentences which he uses to support his hy-pothesis ?^ (21) h i t o i £ s Ts sono hito^-no] hon-ga yoku ureru] person the person's book well s e l l hito-^ [ s Cg zibun^-nojhon-ga yoku ureru] h i t o ^ tg hon-ga yoku ureru] ^[g hon-ga yoku ureru] h i t o ^ ( L i t . ) ' t h e person^ whose^ books s e l l well * (22) h i t o - [ s [g sono hito^-no] hon-ga yoku u r e r u ] person the person's book well s e l l h i t o ^ tg Ts zibun^-no] hon-ga yoku ureru ] ^[g [g zibun^-no] hon-ga yoku ureru] h i t o ^ ( L i t . ) ' the person^ whose^ books s e l l well ' Now, do the above derivations show that r e l a t i v i z a t i o n i s a copying rule? It seems that the answer i s i n the neg-a t i v e . Notice that r e l a t i v i z a t i o n operates only when an embedded NP i s c o r e f e r e n t i a l with a matrix NP. The f i r s t h i t o 'person' i n (21) or (22) i s the NP which o r i g i n a l l y e x i s t s i n each matrix sentence, and never the NP copied from the embedded sentence. For example, suppose (21) or 103 (22) i s contained i n a matrix sentence as i n (23.c). Notice that the word-order of (23.c) i s based on his argur ment. (23) (a) Hito^-wa u r e s i i , (matrix sentence) person happy-is ( L i t . ) ' (The) person i s happy.* (b) £sono hito.-noJ hon-ga yoku ureru. (constituent sentence) (c) Hito^ [[sono hito^-no] hon-ga yoku ureru] -wa person the person's book well s e l l ures i i . happy-is ( L i t . ) 'The person whose books s e l l well i s happy.' The underlined head NP hi t o exists o r i g i n a l l y in.the matrix sentence and i s not derived by copying. If we con-sidered r e l a t i v i z a t i o n to be copying, we would have the f o l -lowing s t r u c t u r e l change (23»d).from (£3.c): (23) (d) frHito^ h i t o . [[sono hito i-no] hon-ga yoku ureru]-wa u r e s i i . 0 ( L i t . ) 'The person^ the person^ whose^ books s e l l well i s happy.' 104 However, h i s derivations do not have such a deriva-t i o n a l s t r i n g as i s shown i n (23,d). It i s u n l i k e l y that he would propose the same under-18 l y i n g structure that i s claimed by Schachter. - If we f o l -low Schachter, the underlying structure of (23.c) would be roughly represented as (24). 1 (24) sono hito hon k a i t a According to Schachter's analysis, the NP2 i n S^ sono h i t o 'the person' must "promote" to the place of the head NP by copying r e l a t i v i z a t i o n and replace a dummy 105 symbol. Then the o r i g i n a l NP 2 i n i s deleted or pro-nominal i zed. However, Muraki 1s underlying r e l a t i v e construc-tions show that he does not argue that the head NP i s a dummy symbol. He also does not say anything about the related problems which w i l l n a t u r a l l y ari s e from the 19 above Schachterian Promotion Analysis . Therefore, hi s underlying structure of (23.c) must be (25), which meets the i d e n t i t y condition. (25) sono h i t o i hon kai t a , If (25) represents his underlying structure, how does his copying r e l a t i v i z a t i o n apply to the NP^ i n S- and to which node does i t copy the embedded NP^? There are no traces of copying i n h i s derivations (21) and (22). Thus, we can say that h i s r e l a t i v i z a t i o n i s not a copying transformation i n Ross's sense of the term. Rather, judging from h i s derivations, h i s r e l a t i v i z a t i o n i s a kind of feature-changing rule because the head NP changes the i d e n t i c a l embedded NP into a r e f l e x i v e . But t h i s r e f l e x i v e i n (21) and (22) i s supposed to be derived by a d i f f e r e n t transformation and i t s antecedent i s not the head NP. This problem w i l l be discussed i n the next chapter. 3.4 CONCLUSION To summarize the observations i n t h i s chapter, we have discussed the following points. a. Muraki's underlying structure of a r e l a t i v e clause. b. His argument that Japanese r e l a t i v i z a t i o n i s a copying transformation. c. The necessity of the Nearest NP Constraint 1 0 7 With respect t o p o i n t ( a ) , we have d e t a i l e d t h a t h i s two arguments f o r h i s u n d e r l y i n g s t r u c t u r e s do not h o l d . His f i r s t argument i s not convincing because as he h i m s e l f admits, Japanese r e l a t i v i z a t i o n i s not a movement tra n s f o r m a t i o n , and the d e l e t i o n of the r e l a t i v i z e d NP has nothing t o do w i t h the Cross-over Co n d i t i o n . His second argument th a t the backward d e l e t i o n across the sentence i s unnatural and that the Nearest NP  C o n s t r a i n t i s necessary f o r the block of unacceptable sentences i s not persuasive, e i t h e r . We have shown t h a t the u n a c c e p t a b i l i t y of ( 6 .a) and (6.b) does not a r i s e from the v i o l a t i o n of the Nearest NP C o n s t r a i n t but i t i s un-acceptable f o r other reasons. We have a l s o observed t h a t the Nearest NP C o n s t r a i n t i s untenable since i t b l o c k s grammatical sentences i n -c o r r e c t l y and t h a t Japanese has the backward d e l e t i o n across the sentence, i n s p i t e of Muraki's argument. We have seen t h a t i n connection with a n o n - r e l a t i v e but n o m i n a l i z i n g c o n s t r u c t i o n , h i s u n d e r l y i n g r e l a t i v e s t r u c t u r e i s unnatural. We assert i n t h i s chapter that the r e l a t i v i z e d NP i s the nearest to the head NP only i n terras of s t r u c t u r a l hierarchy. F i n a l l y , we have discussed his copying r e l a t i v i z a -t i o n and have found that h i s r e l a t i v i z a t i o n i s not a copy-ing rule, but seems more l i k e a feature-changing r u l e . We w i l l demonstrate i n the next chapter that i t i s not a feature-changing rule, either. Also, we s h a l l discuss Kuno's hypotheses about Japanese r e l a t i v i z a t i o n . 109 CHAPTER III NOTES 1. Muraki, "Presupposition, Pseudo-clefting and Thematization," pp. 37~5k & 190. 2. Ibid., p. 42. 3. Ibid., p. 43. 4. Ibid., p. 51. 5. Ibid., p. 5^. 6. Ibid. 7. Ibid., p. 51. 8. Ibid. 9. Ibid., pp. 51 f . 10. Ibid., p. 54. 11. See p.62 i n Chapter II of the present t h e s i s . 12. Mieko Ohso, "A Remark on Japanese R e l a t i v i z a t i o n As a Process of Deletion," (unpublished paper, 1972), p. 7» 13. N.A. MeCawley, "A Study of Japanese R e f l e x i v i z a t i o n , " p. 12. 14. Makino, Some Aspects of Japanese Nominalizations, pp. 112-115. ~~ 15. Ibid., p. 106. 16. Muraki, op. c i t . , p. 190. 17. Ibid., pp. 50 f . 18. Schachter, "Focus and R e l a t i v i z a t i o n , " p . 35. 19. Schachter terms his own analysis Promotion Analysis, while he c a l l s the usual transformational analysis of r e l a t i v i z a t i o n Matching Analysis. Concerning the problems which w i l l n a t u r a l l y arise from hi s Promotion Analysis, see Schachter, "Focus and R e l a t i v i z a t i o n , " pp. 38^40. CHAPTER IV KUNO'S HYPOTHESES AND JAPANESE REFLEX IVIZ AT ION 4.1 AN EXAMINATION OF KUNO'S HYPOTHESES In t h i s chapter, we w i l l discuss whether r e l a t i v i z a -t i o n i s a feature-changing ru l e . Kuno claims that a r e l a -t i v i z a b l e NP sometimes leaves a (ref l e x i v e ) pronoun behind. We w i l l examine h i s examples and r e j e c t h is claim, based on his another hypothesis about r e l a t i v i z a t i o n and some knowl-edge of Japanese p e f l e x i v i z a t i o n . We w i l l introduce a few conditions of Japanese r e f l e x i v i z a t i o n . Kuno mentions the following i n his paper. 1 a. Both themes and r e l a t i v e clauses can, under ce r t a i n conditions, r e t a i n pronouns i n the po s i t i o n formerly occupied by the o r i g i n a l noun phrase that has been thematized or r e l a -t i v i z e d . b. It i s not the case that the o r i g i n a l noun phrase can remain as a (reflexiveT pronoun unconditionally. It i s not clea r under what conditions o r i g i n a l noun phrases can remain i n the form of (ref l e x i v e ) pronouns a f t e r thematization. I l l c. One thing that i s c l e a r i s that such noun phrases must be left-branching noun phrases. Kuno's statement above leads us to suspect that Japanese r e l a t i v i z a t i o n i s a kind of feature-changing rule, as we have discussed i n the preceding chapter. Let us examine whether or not he i s right and whether r e l a t i v i z a t i o n a f f e c t s the r e f l e x i v i z a t i o n or the pro-nominalization of the o r i g i n a l NP. Before our investigation, i t must be noted that the above mentioned left-branching NP's are defined by Kuno as: d. For example; John's i n John's brother, John's  brother's wife. John's brothers wife's s i s t e r , etc., i s a left-branching noun phrase. S i m i l a r l y , John i n John no ottoto no okusan no  imooto "John's brother's wife's s i s t e r " i s l e f t -branching. 2 Also, we must have Kuno's other hypotheses i n mind. They are as follows: e. Themes exist i n the deep structure of thematic sentences.3 f. R e l a t i v i z a t i o n i n Japanese applies, not to an ordinary noun phrase^ but to the theme NP-wa of the r e l a t i v e clause. 112 F i r s t , observe h i s following examples (1) (a) Sono kodomo-ga Mary-o butta. the c h i l d struck ( L i t . ) ' The c h i l d struck Mary.* (b) *Sono kodomo-wa ( zibun 1 -ga Mary-o butta. ) s e l f \ kare \ ^ he J (c) * ( Zibun I -ga Mary-o butta kodomo \ Kare J ~~ ( L i t . ) ' the c h i l d who himself struck Mary ' (2) (a) Sono kodomo-no sensei-ga kootuuziko-de the c h i l d ' s teacher t r a f f i c accident-in sinda. died ( L i t . ) ' The c h i l d ' s teacher was k i l l e d i n a t r a f f i c accident. ' (b) Sono kodomo-wa zibun-no sensei-ga kootuuziko-de sinda. ( L i t . ) ' Speaking of the c h i l d , his teacher was k i l l e d i n a t r a f f i c accident.' (c) zibun-no sensei-ga kootuuziko4desinda kodomo ( L i t . ) ' the c h i l d whose (his own) teacher was k i l l e d i n a t r a f f i c accident ' 113 (3) (a) Zoo-no hana-ga nagai. elephant's trunk long-is ( L i t . ) ' An elephant's trunk i s long.' (b) Zoo-wa (sono ( i t s i t h e ) ) hana-ga nagai. (•sore-no (that's) ) (»zibun-no ( s e l f ' s ) ) ( L i t . ) * Speaking of an elephant, i t s trunk i s long.' (c) (sono ) (*sore-no ) hana-ga nagai zoo (»zibun-no) ( L i t . ) * an elephant whose ( i t s ) trunk i s long ' (4) (a) Sono kodomo-ga kawaigatte i t a inu-ga the c h i l d petting-was dog sinde simatta. died ( L i t . ) ' The dog that the c h i l d was petting died.' (b) Sono kodomo-wa zibun-ga kawaigatte i t a s e l f inu-ga sinde simatta. ( L i t . ) ' As f o r the c h i l d , the dog that (he) himself was petting died.* 114 (c) zibun-ga kawaigatte i t a inu-ga sinde simatta kodomo ( L i t . ) ' the c h i l d who the dog that (he) himself was petting died ' (5) Watakusi-ga (sono hito (that person))-no T (kare (he) ) , 1 (sp_ (that) ) 8 namae-o wasurete simatta okyaku-san name have-forgotten guest ( L i t . ) * a guest whose name I have forgotten * Now, Kuno's hypotheses (e) and (f) are c r u c i a l f o r our examination of his other hypotheses (a) ^ - ( c ) , which are concerned here. If h i s hypothesis (e) i s correct, then each sentence (b) i n h i s examples (1) ~ ( 4 ) can not be derived from i t s corresponding sentence (a). I f hy-pothesis (f) i s correct, then an underlying r e l a t i v e clause must contain a thematic sentence. Take (2,b) and (2.c). The underlying structure of (2,b) would be represented as (6)t sensei, kodomo. aru As i s shown i n (6), the thematic NP.^  has an i d e n t i c a l NP,p i n the comment-part. S i m i l a r l y , NP^ has an i d e n t i c a l NPg. As the hypothesis (e) shows, these thematic NP's should exist o r i g i n a l l y i n the underlying structure of (2,b). And t h i s underlying structure (6) must be the underlying r e l a t i v e clause of (2.c) by vi r t u e of hypothesis ( f ) . Thus, (2.c) would look l i k e (7). 116 (7) NPg-ga NP 9-ni V sensei. kodomo. aru NP^ ^ i s a r e l a t i v e construction, where NP2 i s the head NP and S-^  i s the r e l a t i v e clause. In S-^ , NP^ i s the theme and S 2 i s the comment-part. In S 2, there i s another r e l a t i v e construction, where NPg i s the head NP and i s the r e l a t i v e clause which contains the thematic NPr, and the comment sentence S^. Thus, from (7), we can get (2.c) i n the following way. 117 ( i ) The deletion of NPg under i d e n t i t y with NP^ ( i i ) The deletion r e l a t i v i z a t i o n of NPr,. ( i i i ) The genitive transformation, which converts n i aru'exist' to_no' 's'. (iv) The r e f l e x i v i z a t i o n of NP^ under i d e n t i t y with NP3. We now have the following structure. (8) The f i n a l process w i l l be the following! (v) The deletion r e l a t i v i z a t i o n of NP^. The derived sentence i s (2.c). Nbtice i n (8) that i n S cycle, NPg has already changed into zibun before the r e l a t i v i z a t i o n of the thematic NP^ and that the r e l a -t i v i z a t i o n of NP^ only requires i t s deletion, and has no thing to do with the r e f l e x i v i z a t i o n of NPg. That i s , step ( i v ) generates a r e f l e x i v e zibun and the f i n a l r e l a t i v i z a t i o n i s only a deletion and i t does not leave i t s own r e f l e x i v e behind. In spite of h i s f i r s t statements, i f we follow his other hypotheses (e) and (f) consistently, we f i n d that his hypothesis (a) that r e l a t i v i z a t i o n leaves the r e l a t i v i z e d NP as a (reflexive) pronoun i s incorrect. Rather, we would say that before r e l a t i v i z a t i o n the underlying r e l a t i v e clause has already undergone the application of r e f l e x i v i z a t i o n , and that r e l a t i v i z a t i o n i s neither copying nor feature-changing.^ In the preceding chapter, we have observed that Muraki's r e l a t i v i z a t i o n i s not a copying rule but i t seems more l i k e a feature-changing r u l e . But here we must re-examine the example i n question to see whether Japanese r e l a t i v i z a t i o n i s a r e a l feature-changing r u l e . Based on Kuno's two hypothses (e) and ( f ) , we could postulate that the following example (9) has i t s under-l y i n g structure (10). (9) Zibun.-no hon-ga ureru h i t o - w a u r e s i i . s e l f ' s bqok s e l l person happy-is ( L i t . ) * The person whose book s e l l s (well) i s happy.' (10) (sono)hito (sono)hon kaita 120 The process of derivaion would be as follows. ( i ) The deletion of NPg under i d e n t i t y with NPg. ( i i ) The deletion r e l a t i v i z a t i o n of NPg. ( i i i ) The genitive formation i n S^. hon NP ? NPg V I hon hon 0 NP 5 NP ?-(ia N hon (sono; h i t o \ k a i t a no (iv) The r e f l e x i v i z a t i o n of NP^ under i d e n t i t y with NP^. ureru NP,,-no NP e •J J zibun^ hon 121 (v) The deletion r e l a t i v i z a t i o n of N P V We notice from the above that we already have a r e f l e x i v e pronoun by step (iv) and a f t e r that we have a deletion r e l a t i v i z a t i o n . Therefore, i t is not the case that the head NP 2 affects the r e f l e x i v i z a t i o n of NP.-,. Thus, we may say that Japanese r e l a t i v i z a t i o n i s never a feature-changing r u l e . I t seems that r e l a t i v i z a -t i o n operates to delete the thematic NP i n the underlying r e l a t i v e clause, which i s h i e r a r c h i c a l l y the nearest NP to the head NP. Let us return to Kuno's examples. F i r s t , observe the remote structure of (4.c). (12) We have the following transformations to derive (4.c) from (12). ( i ) The deletion 6f NPg under i d e n t i t y with NPg. ( i i ) The deletion r e l a t i v i z a t i o n of NPg. ( i i i ) The r e f l e x i v i z a t i o n of NP ? under i d e n t i t y of NP^ This r e f l e x i v i z a t i o n would be optional because we can delete NPr, instead of a r e f l e x i v i z a t i o n trans-formation. 123 (iv) The deletion r e l a t i v i z a t i o n of NP^. In t h i s case, too, r e f l e x i v i z a t i o n applies to the deeply embedded NP^ under i d e n t i t y with the thematic NP^ before the deletion r e l a t i v i z a t i o n of NPy Here, we must note that Kuno's hypothesis (c) i s not c l e a r to us that the retained pronoun a f t e r r e l a t i v i z a -t i o n i s a left-branching NP l i k e John's i n John's brother or John-no i n John-no otooto. I f he means by i t that a left-branching NP i s r e s t r i c t e d to a genitive NP, h i s hypothesis does not f u l f i l l the generalization i t i s claim-ed to have. Because as zibun-ga the subject NP i n the r e l a t i v e clause i n (4„c) shows, t h i s ostensibly retained pronoun does not always have a genitive form. Rather, we should say that at some stage of derivation, some deeply embedded NP gets r e f l e x i v i z e d or pronominalized under ce r t a i n conditions. The reason why a left-branching genitive NP undergoes r e f l e x i v i z a t i o n i s that the genitive NP i t s e l f i s o r i g i n a l l y an embedded sentence. Then, what are the conditions of r e f l e x i v i z a t i o n ? In the next section, we w i l l examine some conditions of 124 r e f l e x i v i z a t i o n . 4.2 SOME CONSTRAINTS OF JAPANESE REFLEXIVIZATION In t h i s section, we w i l l have only the introductory and b r i e f exposition of Japanese r e f l e x i v i z a t i o n which i s necessary f o r our further examination of Kuno's examples. We w i l l discuss some s a l i e n t points of Japanese r e f l e x i v i z a -t i o n . The Subject-antecedent Condition The r e f l e x i v e refers back to the subject i n the same simplex sentence or the subject i n any higher sentence.° In English, the antecedent NP can be either the subject or some other element i n the same simplex sentence, but N.A. McCawley claims that i n Japanese the antecedent NP must be the subject. (13) (a) B i l l talked to Mary^ about h e r s e l f ^ . .(b) *Bill-wa Mary-ni zibun-no koto-ni t u i t e to s e l f ' s thing about hanasita. talked 125 ( L i t . ) ' B i l l talked to Mary^ about h e r s e l f ^ ' By v i r t u e of the Subject-antecedent Condition, the only acceptable reading of (13.b) i s i B i l l ^ talked to Mary about himself^. In Japanese, the antecedent NP and i t s r e f l e x i v e do not have to be i n the same simplex sentence. (14) Taroo-wa zibun^-ga k a i t a s tegami-o s e l f wrote l e t t e r yomi kaesita; <. read again ( L i t . ) ' Taro^ re-read the l e t t e r which s e l f ^ wrote.' Q As N.A. McCawley puts i t 7 ' , Japanese r e f l e x i v i z a t i o n seems to operate i n the same way as English pronominal i z a-? t i o n i n that i t can enter into an embedded sentence, though i n r e a l i t y they do not operate exactly i n the same manner. This w i l l soon be discussed. As we can see i n (13) and (14), the Japanese r e f l e x -ive takes one form zibun ' s e l f * , regardless of person, gender,and number. The Humanness Condition 1 0 The antecedent must be human. 1 2 6 The Command Condition' The antecedent of the r e f l e x i v e not only must be the subject but must also command the r e f l e x i v e . To explain the above condition, N.A. MeCawley*s 12 following example may be considered. (15) (a) Hirosi^-ga ima gesyuku s i t e i r u i e - n i now board house-in Hirosi^-wa moo go-nen-mo sunde i r u . already 5 years l i v e ( L i t . ) ' In the house where H i r o s i . boards now Hi r o s i * has been l i v i n g as long as f i v e y e a r s . 1 (b) Zibun^-ga ima gesyuku s i t e i r u i e - n i s e l f now board house-in Hirosi^-wa moo go-nen-mo sunde i r u . already f i v e years l i v e ( L i t . ) * In the house where s e l f , boards now Hir o s i * . has been l i v i n g as long as f i v e years.* (c) *Hirosi^-ga ima gesyuku s i t e i r u i e - n i now board house-in zibunj-wa moo go-nen-mo sunde i r u . s e l f already f i v e years l i v e ( L i t . ) * In the house where H i r o s i , boards now s e l f , has been l i v i n g as long as f i v e years.' 1 2 7 The notion of command was proposed "by R.W. Langacker i n his "on Pronominalization and the Chain of 13 Command" i n such a way that» a node A "commands" another node B i f (1) neither A nor B dominates the other; and ( 2 ) the S-node that mostly immediately dominates A also dominates B. If we follow N.A. McCawley, the remote structure of the above sentence would be roughly represented as ( 1 6 ) t ( 1 6 ) gesyuku s i t e i r u Now, NP2 commands NP^, but not vice versa. There-fore, NP^ can not be the antecedent of NP2 i n r e f l e x i v i z a -t i o n . ( 1 5 . c ) i s ungrammatical because i t v i o l a t e s t h i s condition. We have seen that English pronominalization and 128 Japanese r e f l e x i v i z a t i o n operate i n the same way i n the sense that both can enter into embedded sentences* How-ever, as sentence (15.c) shows, t h e i r behavior i s not exactly the same i n that i n Japanese r e f l e x i v i z a t i o n , the notion of precede r e l a t i o n does not play so much an import-ant role as i n English pronominalization. Precede^ r e l a t i o n was also proposed by R.W. Langacker i n h i s above c i t e d paper. This means that the antecedent NP precedes the pronominal-14 ized NP i n terms of l i n e a r ordering of constituents. The like-NP c o n s t r a i n t 1 ^ From the preceding f a c t s , I would l i k e to propose that Japanese has a syntactic constraint, the Like-NP Constraint, which discards the sentences as ungrammatical i f the r e f l e x i v e and i t s antecedent are i n peer r e l a t i o n s h i p . The notion of 'peer* i s due to Postal (1970). Postal defines the NP's which are peers i n the following way.1^ Two NP, NP, and NP 2» neither of which dominates the other nor i s co-ordinate with the other i n a phrase marker P are peers with respect to a node S., just i n case the paths between each of these NP and S. are such that they contain no NP-nodes not separated from the s t a r t i n g point NP, NP-, or NP 2f by a node S. The following two examples explain the like-NP  constraint > 129 (17) (a) *Taroo»wa zibun^-o t a t a i t a . s e l f h i t ( L i t . ) ' Taro i h i t himself i.' (b) Taroo^-wa zibun^-no atama-o t a t a i t a , s e l f ' s head h i t ( L i t . ) ' Tare- h i t h i s J L head.' Sentences (17.a) and (17.h) would be represented as (18.a) and (18.b) i f we follow N.A. McCawley's analysis, (18) Taroo (b) Taroo^ t a t a i t a Taroo^ In (18.a), NP 1 and NP2 are peers, while NP X and NP^of (l&^b) are not. Therefore, NPg i n (18.a) can not be r e f l e x i v i z e d but NP^ i n (18.b) can be a r e f l e x i v e . The Highest NP Condition The antecedent of the r e f l e x i v e f o r Backward 130 R e f l e x i v i z a t i o n must be the highest NP which s a t i s f i e s the humanness condition. ••• We have to add a r e s t r i c t i o n f o r Backward R e f l e x i v i z a t i o n within sententially-complex NP subjects that i f the head NP i s human, Backward R e f l e x i v i z a t i o n i s blocked. 1' •The highest NP which s a t i s f i e s the humanness con-di t i o n * means here the highest human noun i n the s t r u c t u r a l hierarchy i n a given structure. 'A sententially-complex NP subject' i s a r e l a t i v e construction which functions 18 as a subject i n a sentence. The following sentence i s 19 T. Oyakawa's example. (19) Hosyuseitoo-no ooboosa to zibun^-no conservative party's s e l f ' s unreasonableness s i z i s i - t e k i - t a seitoo-no huhai-ga sono supported-had party's corruption the gakusei-no otooto-no sinyuu^-no sisoo-o student's brother's best friend's thought muse'ihusyugi-e katamuk-ase-ta. anarchism-to lean-cause-past ( L i t . ) 'The unreasonableness of the Conservative Party and the corruption of the Party that s e l f , had supported made the thought of the student's younger brother's best f r i e n d , lean towards anarchism.* 1 He says the relevant NP's i n (19) have the following surface 131 structure i (20) NP NP NP sono gakusei NP = antecedent NP a . In (20), the highest NP i s NP sinyuu 'best friend* which 3. * can be the antecedent of backward r e f l e x i v i z a t i o n . The Theme-antecedent Condition This condition seems to be a kind of the Subject- antecedent Condition because M. Ohso her s e l f says that the Subject-rantecedent Condition might have to be changed 20 to the Theme-antecedent Conditon. This condition requires that the antecedent of a r e f l e x i v e be the theme of a sentence. She explains the ungrammatically of Kuno's 21 example ( l . c ) c i t e d above i n the following way. 132 (1) (c) *Zibun.-ga Mary-o butta kodomo^ s e l f struck c h i l d ( L i t . ) ' the c h i l d j who himself^ struck Mary • (21) NPy-wa NPj.-ga NP.-o V kodomo^ kodomo^ Mary butta In the underlying structure of ( l . c ) , NP^ and NP^ are i n a peer relationship and so NP^ can not be r e f l e x i v -ized. However* i n ( l . c ) , NP^ i s r e f l e x i v i z e d with the r e s u l t that the derived sentence i s ungrammatical. In-stead of re"frexi"vization / r P ^ must undergo;the obligatory deletion. The above s i x conditions are used f o r our examina-tions of Kuno's remaining examples i n the following sec-t i o n . 133 4.3 THE SUBJECT-ANTECEDENT CONDITION AND THE THEME-ANTECEDENT CONDITION In section 4.1, we observed that Kuno's retained ( r e f l e x i v e ) pronoun i s not the one which i s r e l a t i v i z e d but the one which has already been r e f l e x i v i z e d or prono-minalized before the deletion r e l a t i v i z a t i o n of the nearest thematic NP. Let us return to h i s two remaining examples, where r e f l e x i v i z a t i o n does not apply. Take Kuno's example (3.c) f i r s t . (sono ) hana-ga nagai zoo j trunk long-is elephant ( L i t . ) ' an elephant whose(its) trunk i s long ' (22) NP-The s i m p l i f i e d underlying structure (22) represents (3.c). NP^ and NP^, which are i d e n t i c a l , are not peers. The antecedent NP^ i s the theme of S^, but i t does not meet the Humanness Condition. Therefore, NP^ can not be r e f l e x i v i z e d and i s pronominalized. Sentence (5) which i s Kuno's l a s t example has the following remote structure» watakusi -ga (sono hito (that person)) rno ( kare (he) ) ( so (that) ) s (*zibun ( s e l f ) ) namae-o wasurete simatta okyaku-san name forgotten have guest ( L i t . ) ' a guest whose name I have forgotten ' (23) wasurete simatta (sono) okyaku-san^ namae 135 Figure (23) shows that NP^ and NPg are i d e n t i c a l and not peers. NP^ meets the Theme-antecedent Condition and the Humanness Condition. Therefore, NPg should be r e f l e x i v i z e d , but i n fa c t , i t can not be. It can only be pronominalized or deleted. The possible answer to i t i s that i f NPg i s r e f l e x i v i z e d , the r e f l e x i v e causes a reading to the e f f e c t that 'I have forgotten my name' because NP^ occurs exactly i n the p o s i t i o n which s a t i s f i e s the Sub.ject-antecedent Condition. It i s the same case with my following example . (24) [boku-ga (sono. (the) ) nazukeoya-ni (kare--no (his) ) I (*zibOn.-no ( s e l f ' s ) ) godfather ( ~T~ — ) natt'a ~\ kodomo^ became ( L i t . ) ' the c h i l d ^ whose^ godfather I became ' (25) (sono) kodomo. nazukeoya 136 In (25). NP^ meets both the Humanness Condition and the Theme-antecedent Condition but NPg can not be r e f l e x i v i z e d . However, observe the two other examples below» (26) ya-ga (zibun.-no) ude-ni a t a t t a samurai (kare.-no ) arrow ( 0 ) arm-at shot warrior (sono. ) ( L i t . ) * the samurai^ at whose^ arm an arrow shot ' (27) Kodomotati-no booru-ga (zibun.-no) atama-ni ( T ~) children's b a l l (?kare»no ) head at a t t a otoko-wa okotte simatta. h i t man got angry ( L i t . ) ' The man whose head the children's b a l l h i t got very angry.' (28) kodomotati booru (sonoVotoko^ atama 137 Figure (28) i s the underlying structure of (27). NP^ and NPg are i d e n t i c a l . NP^ can he the antecedent of NPg i n r e f l e x i v i z a t i o n , though there i s an intervening subject NP^ between them. The above observation suggests that: ( i ) The Subject-antecedent Condition takes precedence over the Theme-antecedent Condition. ( i i ) The Theme-antecedent Condition i s suspended, by virtue of the Subject-antecedent Condition, when the subject which intervenes between a thematic NP and i t s i d e n t i c a l NP i s human, ( i i i ) If the intervening subject i s not human, r e f l e x i v i z a t i o n applies. Kuno's example (5) and mine (24) f a l l under the above case ( i i ) , while sentences (26) and (27) under the case ( i i i ) . 4.4 SUMMARY In t h i s chapter, we have introduced various conditions of Japanese r e f l e x i v i z a t i o n and discussed Kuno's hypotheses summarized below. 138 a. A r e l a t i v i z e d NP leaves a (r e f l e x i v e ) pronoun behind, c. Such an NP must be a left-branching NP. f. A r e l a t i v i z a b l e NP i s a thematic NP. When we analyze his examples based on his hypothe-s i s ( f ) , we f i n d that his hypothesis (a) i s f a l s e — — that i s , an ostensibly retained pronoun i s not derived by r e l a t i v i z a t i o n but p r i o r to deletion r e l a t i v i z a t i o n i t has been reflexivize'd, or pronominalized by i d e n t i f i c a t i o n with some c o r e f e r e n t i a l NP i n the underlying r e l a t i v e clause. Concerning his hypothesis (e), we should restate his hypothesis to the e f f e c t that such an NP i s not res-t r i c t e d to a left-branching genitive NP. It involves a deeply embedded NP and the r e f l e x i v i z a t i o n of such an NP i s determined by various condition? The conditions of r e f l e x i v i z a t i o n are necessary to account f o r whether or not r e f l e x i v i z a t i o n i s applicable to a given structure. However, we have seen that the  Theme-antecedent Condition i s suspended i n cer t a i n cases. This f a c t leads us to suspect that i t may be the case that there are some exceptions to the condition. 139 CHAPTER IV NOTES 1. Kuno, Notes on Japanese Grammar, Chap. 19, pp. 3~5. 2. Ibid., Chap. 19, p. 5. 3. Ibid., Chap. 19, p. 11. 4. Ibid., Chap. 19, p. 14. 5. Ibid., Chap. 18, p. 4 & Chap. 19, p. 4 f . 6. I am indebted to Ohso f o r t h i s analysis. See Ohso, "A Remark on Japanese R e l a t i v i z a t i o n As a Process of Deletion," pp. 1~7. 7. See example (23) on pJ.03of the present t h e s i s . 8. N.A. McCawley, "A Study of Japanese R e f l e x i v i z a t i o n , " p. 4. 9. Ibid., p. 5. 10. See Takatsugu Oyakawa, "Japanese R e f l e x i v i z a t i o n I," Papers i n Japanese L i n g u i s t i c s , V ol. 2 No. 1 (Summer, 1973), p. 95, and Susumu Kuno7 "Pronominalization, R e f l e x i v i z a t i o n , and Direct Discourse," L i n g u i s t i c Inquiry, Vol. 3, No. 2 (1972), pp. 177 f . 11. N.A. McCawley, op. c i t . , p. 12. 12. Ibid., pp. 10 f . 13. Ronald W. Langacker, "On Pronominalization and the Chain of Command," Modern Studies i n English, ed. David A. Reibel and Sanford A. Schane (New Jersey1 Prentice-H a l l , Inc., 1969), p. 167. 14. Ibid., pp. I65-I69. 15. N.A. McCawley, op. c i t . , p. 30. 16. Postal, Cross-over Phenomena, p. 179. 17. Oyakawa, op. c i t . , pp. 115 & 121. 18. Ibid., pp. 109 & 113. 19. Ibid., p. 112. 20. Ohso, "A Remark on Japanese R e l a t i v i z a t i o n As a Process of Deletion," p. 7. 21. Ibid., p 3. CHAPTER V RELATIVIZATION OF NOUN PHRASES IN ADVERBIAL CLAUSES 5.1 INTRODUCTION In t h i s chapter, we w i l l examine another hypothesis of Kuno's, e s p e c i a l l y with respect to the p o s s i b i l i t y that an NP i n an adverbial clause can be r e l a t i v i z e d under some conditions. We w i l l analyze t h i s hypothesis based on h i s claim that a r e l a t i v i z a b l e NP i s a thematic NP immediately followed by wa. We w i l l r e j e c t h is hypothesis and conclude that a r e l a t i v i z e d NP i s the thematic NP which i s s t r u c t u r a l -l y highest i n the r e l a t i v e clause. 5.2 KUNO'S HYPOTHESES 5.2.1 Let us examine Kuno's claim which f o l l o w s i a. An element of an adverbial clause can be r e l a t i v i z e d . 1 142 b. It i s not clear, however, that r e l a t i v i z a t i o n i s f r e e l y applicable to elements i n adverbial ..clauses, complex noun phrases and s e n t e n t i a l subjects. Although i t i s not clear under what conditions i t i s possible or under what condition i t i s not, i t i s c l e a r that when thematization 2 i s possible, then r e l a t i v i z a t i o n i s also possible. Since Ross himself admits that Japanese r e l a t i v i z a t i o n i s not subject to h i s Sentential Subject Constraint? we w i l l discuss here the p o s s i b i l i t y of r e l a t i v i z a t i o n of the NP i n an adverbial clause. Before our examination of Kuno's hypothesis c i t e d above, we must r e c a l l Kuno's important hypothesis which we used f o r our analysis i n Chapter IV. We w i l l c i t e i t again and c a l l i t hypothesis (c ) . c. R e l a t i v i z a t i o n i n Japanese applies, not to an ordinary noun phrase, but to the theme NP - wa of the r e l a t i v e c l a u s e . 4 Now, observe Kuno's example below which he gave us f o r his hypothesis ( a ) . ^ as evidence (1) (a) Sono hito-wa, sinda node, minna-ga kanasinda, the person died because a l l were-saddened ( L i t . ) * Speaking of that person, everyone was saddened because he died,' 143 (b) S i n d a node minna-ga kanasinda, h i t o If Kuno's hypothesis (c) i s correct, r e l a t i v e construction (l.b) contains sentence (l.a) as the r e l a t i v e clause because (l.a) i s a thematized sentence. The under-l y i n g structure of (l.b) would be roughly represented as (2). (2) sono hi t o , s inda , Notice that i n (2), S^, which i s a r e l a t i v e clause, represents the thematized sentence ( l . a ) . The head noun phrase NPg, the thematic noun phrase NP^ and the NP i n the adverbial clause, that i s , NP^ are a l l i d e n t i c a l . To generate (l.b) from figure (2), we would have the following d e r i v a t i o n a l process t 144 ( i ) The equi-NP deletion of NP^ under i d e n t i t y with NP^ ( i i ) The deletion r e l a t i v i z a t i o n of NP., J • The above example shows that at step ( i ) , the NP i n an adverbial clause i s deleted not by r e l a t i v i z a t i o n but by equi-NP deletion. At step ( i i ) , the thematic NP of a r e l a t i v e clause undergoes deletion r e l a t i v i z a t i o n . In other words, when we analyse Kuno's example, based on h i s hy-pothesis (c) that a r e l a t i v e clause contains a thematic NP, we w i l l have to say that an NP i n an adverbial clause i s not r e l a t i v i z e d . What i s r e l a t i v i z e d i s a thematic NP. We w i l l examine one more example of Kuno's. Observe the following.^ (3) hara-ippai tabetara: >^ geri-o s i t e simatta b e l l y - f u l l when-I-ate diarrhea doing ended-up okasi cookies-with • ( L i t . ) ' cookies which, when we had glutted our-selves with (them), we ended up with diarrhea * According to Kuno, i n (3) okasi i n the adverbial 145 clause hara-ippai tabe-ta-ra i s deleted by r e l a t i v i z a t i o n . His claim seems incorrect. I f we follow h i s hypothesis ( c ) , the remote structure of (3) would look l i k e (4), which contains a thematized sentence ST as the r e l a t i v e clause. watakusi hara-ippai sono okasi* tabe ta^ The relevant transformations here are * ( i ) The equi-NP deletion of NPg under i d e n t i t y with the thematic noun phrase NPy (This transformation deletes the i d e n t i c a l NP i n an adverbial clause, ( i i ) The deletion r e l a t i v i z a t i o n of NP^, which i s thematic. 146 This example also shows that a r e l a t i v i z e d NP i s never an element i n an adverbial clause, but a thematic NP which i s s t r u c t u r a l l y highest i n a r e l a t i v e clause. Notice that when the r e l a t i v i z a t i o n of the theme occurs, the i d e n t i c a l NP i n an adverbial clause has already been deleted. 5.2.2 Now, l e t us examine Kuno's hypothesis (b). F i r s t , observe his example below.^ ( 5 ) (a) Sono hito-ga dekinakereba, watakusi-ga yaru. the person if-cannot-do I do ( L i t . ) * If that man can not do i t , I w i l l do i t . ' (b) Sono hito-wa, dekinakereba, watakusi-ga yaru. ( L i t . ) * Speaking of that man, i f he can not do i t , I w i l l do i t . ' (c) dekinakereba, watakusi-ga yaru hi t o ( L i t . ) ' the person who, i f (he) con not do i t , I w i l l do i t • Why i s (5.c) ungrammatical? It i s easy to answer t h i s question, because the underlying r e l a t i v e clause, that i s , the thematized sentence (5 •*>')' i s unacceptable. In order 147 to v e r i f y t h i s , we w i l l give the underlying structure of (5«b), which i s represented as f i g u r e ( 6 ) . (6) sono h i t o . dekinai As Kuno points out? i n ( 6 ) , the comment-sentence Sg — — — dekinakereba wakakusi-ga yaru ' i f (he) can not do i t , I w i l l do i t . ' — — does not constitute an appropriate statement about the thematic noun phrase NP^ sono h i t o •the person*. Rather, i t i s a statement about something that the speaker says he w i l l do i f sono h i t o can not do i t . When the thematized sentence (5»b) i s anomalous, (5«c), which contains (5.b) as the r e l a t i v e clause, i s also unacceptable. Thus, Kuno's claim i s correct that i f thematization i s possible^theii r e l a t i v i z a t i o n i s also possible. Notice that t h i s means that r e l a t i v i z a t i o n i s applicable to the highest thematic NP i n the r e l a t i v e clause, only when the thematized r e l a t i v e clause i s acceptable. Notice again that r e l a t i v i z a t i o n never applies to an NP i n an adverbial clause. 5.3 SUMMARY In t h i s chapter, we have observed the following t a. When we consistently assume that a r e l a t i v e clause contains a thematized sentence, we must reject Kuno's hypothesis that an NP i n an ad-v e r b i a l clause i s r e l a t i v i z a b l e . The NP i n •~ ; an adverbial clause undergoes equi-NP del e t i o n before the deletion r e l a t i v i z a t i o n of the highest thematic NP i n the r e l a t i v e clause. b. The difference of r e l a t i v i z a b i l i t y between Kuno's examples, that i s , between (l.b) and (5.c) can be accounted f o r by r e f e r r i n g to whether or not t h e i r embedded thematic sentences are anomalous. 149 CHAPTER V NOTES 1. Kuno, Notes on Japanese Grammar, Chap. 18, p. 5. 2. Ibid., Chap. 19, pp. 6 f . 3. See Ross, "Constraints on Variables i n Syntax," p. 134. He defines the Sentential Subject Constraint i n the following ways "No element dominated by a S may be moved out of that S i f that node S i s dominated by an NP which i t s e l f i s immediately dominated by S." He says on p. 134, "the constraint, though operative i n the grammars of many languages other than English, can not be stated as a universal, because there are languages whose rules are not subject to i t . " He gives a Japanese sentence to support his view that the Sentential Subject Constraint i s not applicable to some languages. 4. Kuno, op. c i t . , Chap. 19, p. 14. 5. Ibid., Chap. 19, p. 6. 6. Ibid., Chap. 18, p. 5. ?• Ibid., Chap. 19, p. 8. 8. Ibid.. Chap. 19, p. 5. CHAPTER VI RELATIVIZABLE NOUN PHRASES AND THEMATIC NOUN PHRASES 6.1 INTRODUCTION In t h i s chapter* we w i l l discuss whether or not Kuno i s r i g h t i n saying that a r e l a t i v i z a b l e NP i s an NP immedi-at e l y followed by the thematic p a r t i c l e wa. In the second section, we w i l l introduce Muraki*s counter-examples to Kuno's claim. We w i l l demonstrate i n conclusion that we must revise Kuno's hypothesis to the ef f e c t that the r e l a t i v i z a b l e NP i s thematic, regardless of the existence of an intervening p a r t i c l e between the NP and the thematic p a r t i c l e wa. In the t h i r d section, we w i l l examine whether or not a non-thematic NP i s r e l a t i v i z a b l e . In connection with t h i s problem, we w i l l b r i e f l y discuss Kuno's c l a s s i f i -cation of ga and wa. We w i l l conclude that there are some 151 cases where a non-thematic NP i s r e l a t i v i z a b l e , and there-fore, we can not but discard both Kuno's hypothesis and our r e v i s i o n of h i s hypothesis. 6.2 THE REVISION OF KUNO'S HYPOTHESIS Kuno's claim about a r e l a t i v i z a b l e NP i s as follows i What i s r e l a t i v i z e d i n a r e l a t i v e clause is, not an ordicary noun phrase, but a noun phraseTimmedi-at e l y followed by the thematic p a r t i c l e wa. What he gives as evidence f o r t h i s claim i s the 2 following type of sentences: (1) syuusyoku-ga taihen na buturigaku employment d i f f i c u l t i s physics ( L i t . ) ' physics, where fi n d i n g a job i s d i f f i c u l t / (2) (a) Wio-naka-de} buturigaku buturigaku Phrase (1) i s a r e l a t i v e construction which i s acceptable. 152 In order to generate (1), we must assume that (1) i s derived from the underlying structure (2.b) and not from (2.a), because (2.a) i s ungrammatical. (2.b) contains a thematic NP immediately followed by wa, while (2.a) does not. Thus, he concludes that a r e l a t i v e clause must involve a thematized sentence, where a thematic NP i s immediately followed by wa. Nevertheless, there i s a counter-argument against t h i s claim, Muraki gives us a set of counter-examples to Kuno's above generalization. According to Muraki? we must assume that i n some cases there i s an intervening p a r t i c l e between a thematic NP and w§, f o r deriving a grammatical r e l a t i v e clause construction. There i s a r e l a t i v e clause construction f o r which a corresponding NP + wa - sentence does not exi s t , while only an NP + p a r t i c l e + wa -sentence exis t s , Muraki's counter-examples involve the p a r t i c l e de i n i t s various usages and the dative n i . The following sentences are only part of Muraki's counter-examples. ( 3 ) _de(as an instrumental p a r t i c l e ) (a) *Sono ink-wa kono tegami-o k a i - t a . that ink t h i s l e t t e r wrote ( L i t . ) ' (I) wrote t h i s l e t t e r i n that ink.' 153 (b) Sono ink-de-wa kono tegami-o k a i - t a . (c) kono tegami-o k a i - t a ink L i t . ) 'the ink i n which (I) wrote t h i s l e t t e r . ' (4) de (as a reason p a r t i c l e ) ^ (a (b (c (a (b (c *Sono riyuu-wa John-ga kesseke s i - t a . the reason absent-was L i t . ) ' For that reason John was absent.' ?Sono riyuu-de-wa John-ga kesseki s i - t a . John-ga kesseki: s i - t a riyuu L i t . ) * the reason why John was absent * (5) de (as an i n d i r e c t object p a r t i c l e / * •Sono isya-wa John-ga Mary-o syookai s i - t a . the medical doctor introduced L i t . ) ' To that doctor, John introduced Mary.' Sono isya-ni-wa John-ga Mary-o syookai s i - t a . John-ga Mary-o syookai s i - t a i s y a L i t . ) ' the doctor to whom John introduced Mary ' In the above examples ( 3 ) — ( 5 ) . each (a) sentence i s a thematic sentence with a thematic NP immediately followed by wa. Each (b) i s also a thematic sentence, 154 but i t has an intervening particle de or ni between the thematic NP and wa. (a) is unacceptable, while (b) is acceptable. Thus, in order to generate each relative construction (c), the grammatical sentence (b) must be included in the underlying relative clause. The existence of these counter-examples calls for the revision of Kuno's hypothesis that a relativizable NP is a thematic NP immediately followed by wa. In my opinion the most important point of Kuno's hypothesis i s that a relativizable NP i s thematic. Whether or not there is an intervening particle like de or n i , the NP preceding wa i s a thematic NP. Therefore, we w i l l revise Kuno's hypothesis temporarily to the effect that what i s relativ-izable in a relative clause i s an NP + wa, regardless of an intervining particle. This revision of Kuno's hypothesis does not seem to change his original claim completely because his claim that a relativizable NP is thematic i s s t i l l v alid. How-ever, i f there i s a case in which thematization i s impossible and yet relativization i s possible, i t w i l l constitute a stronger counter-example both to Kuno's original hypothesis 155 and t o our r e v i s i o n mentioned above. We w i l l p r e s e n t some counter-examples o f t h i s type i n the next s e c t i o n . 6.3 COUNTER-EXAMPLES TO KUNO'S HYPOTHESIS AND ITS REVISION In the p r e c e d i n g s e c t i o n , we r e v i s e d Kuno's hypothe-s i s . In t h i s s e c t i o n , we w i l l examine whether o r not a r e l a t i v i z a b l e NP must be the m a t i c . Before the examination, however, we must have i n mind Kuno's o t h e r hypotheses about the use o f the s u b j e c t p a r t i c l e ga and the thematic p a r t i c l e wa. The f o l l o w i n g are h i s hypotheses about g a and wai (6) Kuno's hypotheses about ga and wa a. Wa marks e i t h e r the theme o r the c o n t r a s t e d  element of the sentence. The theme must be e i t h e r anaphoric o r g e n e r i c , while t h e r e i s no such c o n s t r a i n t f o r the c o n t r a s t e d element. b. Ga as s u b j e c t .case marker, i s e i t h e r f o r ... n e u t r a l d e s c r i p t i o n (as i n Oya, ame-gai h u t t e  - i r u 'Oh, look! I t i s r a i n i n g . ' ) o r f o r 6.3.1 156 a state (but not existence) or a habitual-generic action, only the e x h a u s t i v e - l i s t i n g i n t e r p r e t a t i o n i s obtained. c. The most natural way to introduce an e n t i r e l y new event i n conversation seems to be to t a l k about the existence, or coming into existence toward the place of the speaker, of something. Since we have already observed the use of the thematic f o r neutral description. Suppose that two men are working i n a room. Then one of them happens to look outside and finds i t i s snowing. He says* ( L i t . ) • Oh, i t i s snowing.* The whole sentence above conveys e n t i r e l y new information. To explain t h i s s i t u a t i o n , Kuno saysi The entire event i s presented out of the blue, so to speak, by the speaker. These sentences are not sentences about something. They are themeless sentences," p a r t i e l e wa i n Chapter I, we w i l l introduce the use of ga (7) This kind of sentence i s named neutral description and i t s subject NP i s followed by ga. It i s used to introduce new 15? information into a discourse. Once introduced into the dis-course, i t is "recorded in the registry of the discourse," 9 and can become the theme of the following sentences in conversation. Ga used for exhaustive listing is explainable by the following interrogative sentence and its answer. ( 8 ) (a) Kimitati-no naka-de dare-ga itiban wakai you(pi.) among who the youngest desu ka? is (Lit.) 'Who is the youngest among you?' (b^Taroo -^| a|itiban wakai1 desu. (Lit.) ' Taro (and only Taro) is the youngest.' In (B.b), Taroo is new information, which can be the focus of the sentence, while itiban wakaiV desu,'is the youngest' is old information. Kuno says this NP, which can be translated to the effect that 'Taro (and only Taro)* as in (8,b), is used for exhaustive l i s t i n g . 1 0 Exhaustive l i s t - ing means listing exhaustively a l l the names of the entities 158 which are concerned to convey new information. The sub-ject NP in this use i s also followed by ga and not by wa. This subject NP + ga i s often used as the real answer-part to an interrogative sentence. 6.3.2 With the above knowledge of ga and wa, let us return to the fundamental topic of this section. I w i l l present three types of counter-examples to the hypothesis that a relativizable NP is thematic and is followed by wa. My f i r s t counter-examples are sentences (10) and (11). (9) Taihenda! John4ga( zisatusita. , good heavens k i n e ( j - s e i f ( L i t . ) 'Oh, my goodness! John has committed suicide.* (10) Taihenda' t i i s a i kodomo-J gal yane-no l i t t l e c h i l d fa^l roof-on ue-ni i r u . i s ( L i t . ) *0h, goshi A l i t t l e child i s on the roof.* (11) AJ mekura-no hito-4 ga) kuruma-ni hanergreta V oh blind persoSi—* car-by hit-was ( L i t . ) *0h, a blind man was hit by a car *J* 159 Sentence (9) i s an example of neutral description given by Kuno. 1 1 Our examples (10) and (11) also have ga f o r neutral description. Here we have the same s i t u a t i o n as we have i n sentence (7). In other words, looking out of the window, one of the two men sees something and utters something about i t which i s e n t i r e l y new information i n the form of (10) or (11). Here, wa i s not allowed to be used. The noun phrases kodomo ' c h i l d ' and h i t o 'person* are modified respectively by t i i s a i ' l i t t l e ' and mekura-no •blind', which i s the a t t r i b u t i v e adjective and the genitive form of mekura. The usual transformational account i n Japanese of these a t t r i b u t i v e s i s that they are derived from r e l a t i v e clauses by the a p p l i c a t i o n of r e l a t i v e clause reduction. This account gives us the following underlying structure f o r t i i s a i kodomo * a l i t t l e c h i l d ' . (12) kodomo. t i i s a i 160 As i s shown i n the above tree diagram, NP^ kodomo must be followed by ga and not by wa, because the entire sentence i t s e l f i s brand-new information. Therefore, NP^ kodomo i s non-generic and non-anaphoric. In other words, the r e l a t i v i z a b l e noun phrase i n ( 12) , that i s , NP^ i s never thematic. Si m i l a r l y , the a t t r i b u t i v e mekura-no i n (11) i s a reduced r e l a t i v e clause. Mekura-no i n mekura-no hit o has hito^-jga ^  mekura desu *A man i s b l i n d ' i n the underlying r e l a t i v e clause. The r e l a t i v i z a b l e NP i n t h i s underlying structure i s a non-thematic noun phrase hito-ga. The above sentences f o r neutral description constitute one kind of counter-example both to Kuno's hypothesis and sentences contain reduced r e l a t i v e clauses i n which there are no thematic NP's. This type of counter-example would be representable as t our r e v i s i o n i n the previous section, because these (13) 161 6 . 3 . 3 Another type of counter-example i s shown i n the following conversation between A and B. (14) At Dare-ga gakkoo-ni hi-o tuketa ka? who school to s e t - f i r e ( L i t . ) Who set f i r e to the school-building.' Bi Gakkoo-ni urami-o idaku darekaMga ) yatta. grudge bear somebody w aJ did ( L i t . ) Somebody did i t who bears a grudge against the school.' In (14), i t i s supposed that between the two persons (A and B) there i s a presupposition, something l i k e 'Some-body seems to have set f i r e to the school b u i l d i n g . ' . Speaker A wants to get information about the criminal but speaker B does not know d e f i n i t e l y who he i s . He only gives a conjectural answer, i n d i c a t i n g that i t was probably an incendiary bearing a grudge against the school who started the f i r e . The phrase at issue i s gakkoo-ni urami-o idaku dareka 'somebody who bears a grudge against the school'. The head NP dareka 'somebody' must be non-generic and non-anaphoric. It w i l l be c l a s s i f i e d as an NP f o r exhaustive l i s t i n g . This judgement i s supported by the f a c t that 162 dareka does not allow wa to follow i t i n (14.B). Moreover, the r e l a t i v e clause i n (14.B) contains an i d e n t i c a l i n d e f i n i t e pronoun dareka* This dareka also must be non-anaphoric and non-generic, because even i f the underlying r e l a t i v e clause i s used as an independent sentence, only ga follows t h i s dareka. For ease of explanation, we w i l l give the rough under-l y i n g structure of the phrase i n questions (15) I - generic I (- anaphoric] Notice that both NP^ ^ and NP^ are followed by ga. In other words, they are never thematic. The following conversations can be explained i n a s i m i l a r way. 163 (16) At Ano soto-no sawagi-wa nan da? that outside noise what i s ( L i t . ) ' What i s that noise outside?' Bt Yopparatta gakusei-) g a l sawaide i r u n drunk-are students J making a noise-are daroo. perhaps ( L i t . ) ' Maybe some drunken students are making a noise.' (17) At Ano ki-no ue-no mono-wa nan da? that tree's on thing what i s ( L i t . ) ' What i s the thing on the tree over there?' Bt K i - n i kakatta tako r a s h i i . t r e e - i n caught-is k i t e look ( L i t . ) ' It looks l i k e a k i t e caught i n the tree.' In (16), the noise was heard i n the room. As an answer to speaker A's question, speaker B imagines that some students are drunk and they are being boisterous. In the room neither A nor B knows anything about the r e a l state of a f f a i r s . Therefore, the students i n question are never generic nor anaphoric. It may be that speaker B himself has no idea who i s making the noise, though he gives a 164 conjectural answer. Therefore, the underlying relative clause also contains an indefinite and non-thematic NP like 'some students* followed by ga. This non-thematic NP i s relativized in (16.B). caught in the tree, i s brought into existence as new infor-mation in the discourse. The relativizable identical NP tako 'kite' in the underlying relative clause is also non-thematic. As i s shown in the above examples (14) ~- (17), we have a type of sentence containing a relative construction, whose head NP i s classified as an NP for exhaustive l i s t i n g and whose embedded coreferential NP is non-thematic. These examples constitute the second type of counter-example to the hypothesis that a relativizable NP i s thematic. It w i l l be formulated a s i The situation of (17) is similar. A kite, which i s exhaustive l i s t i n g 6.3.4 The third type of counter-example i s shown in the following conversations! 165 (19) Ai [Mado-o kowasita ] no-Jwa1 dare desu ka? window broke person who i s ( L i t . ) 'Who was i t that broke the window?' Bi fKowasital no-J wa> watakusi desu. U J frgal one I ( L i t . ) ' The one who broke ( i t ) was I.' (20) Gakkoo-de nani-ga atta ka? £Atta] koto-wa school-at what happened happened thing kakusazu-ni hanasi nasai. openly speak ( L i t . ) ' What happened at school? T e l l me exactly what happened there.* Sentences (19.A ) and (20) show that the questioner i n f e r s that 'Somebody broke the window'» or 'Something happened at school'. But he does not know who broke the window or what happened at school. More s p e c i f i c a l l y Mado-o kowasita no 'the one who broke the window' i n (19.A) i s to be regarded as related to something l i k e 'Somebody. l broke the window — that somebody^'. In t h i s case, the first*somebody' i s i n d e f i n i t e and non-thematic. This •somebody' i s r e l a t i v i z e d i n the underlying r e l a t i v e clause. 166 It i s followed by ga and not by wa. I t i s the same case with (20). On the other hand, the head noun phrases i n (19) and (20) are thematic based on the presupposed sentences. The basic meaning of (19.A) involves a presupposition and the overt meaning carried by the surface form. The pre-supposition may be considered 'Somebody broke the window'. The overt meaning of (19.A) i s 'As f o r the one who broke the window, who i s i t ? ' In t h i s case, ' i t ' presupposes 'the one who broke the window*. Therefore, no 'the one* i n the surface form i s followed by wa. The observation mentioned above i s supported by two l i n g u i s t s , though t h e i r analyses of pseudo-cleft sentences are d i f f e r e n t from each other. 12 According to Muraki, the underlying structure of (19.B) would be postulated, something l i k e (21). (21) [ g [Dum mado kowasita] g [watakusi mado window broke 1 I kowas i t a j s P r s p ] g where: Dum = dummy C s i S 0 Prsp ] means that S-, i s 2 . presupposed f o r S. 167 In (21), Dum represents 'somebody*. It i s non-anaphoric and non-thematic. Also according to Nakada^? sentence (19.B) would have the following l o g i c a l structure (22). (22) I Assume that you Want to Request us to T e l l you S-^ (who broke the window) and I T e l l you that S 2(who broke the window) was * I*. In (22),'who' i n represents no,(the one who) i n (19«B). This WH-word i s non-anaphoric and non-thematic. Our concern i s not to argue f o r or against Nakada's or Muraki's analysis, nor to observe t h e i r d e r i v a t i o n a l processes of pseudo-elefting. What I want to indicate here i s that Muraki's dummy symbol i n and Nakada's WH-word i n are both i n d e f i n i t e , non-generic, non-anaphoric and non-thematic. The above observation of sentences (18) and (19) t e l l s us that there i s a type of sentence which contains a r e l a t i v e construction whose head NP can be thematic but whose embedded c o r e f e r e n t i a l NP must be non-thematic. I w i l l formulate t h i s type of r e l a t i v e construction i n the 168 following way: (23) [[NP -ga X Yj NP -wa ) (-thematic] [+thematic J Now, based on the above three types of counter-examples (I3)t (18) and (23), we can not but discard both Kuno's hypothesis and our r e v i s i o n . Those hypotheses block our counter-examples i n c o r r e c t l y because they require that r e l a t i v e clauses contain thematic NP's followed by wa. We conclude that there are some cases where r e l a t i v i z a t i o n i s applicable to a non-thematic NP i n the r e l a t i v e clause. 6.k SUMMARY The following i s the r e c a p i t u l a t i o n of the s i g n i f i c a n t points discussed i n t h i s chapter. a. There are some counter-examples to Kuno's hypothesis that a r e l a t i v i z a b l e NP i s an NP immediately followed by wa. Therefore, we revised Kuno's hypothesis to the e f f e c t that a r e l a t i v i z a b l e NP i s thematic. b. However, i n section 3 i we presented some c r u c i a l counter-examples to our r e v i s i o n of Kuno's hypothesis. There are three types of counter-examples which do not cantain the themes i n the underlying r e l a t i v e clauses. Therefore, we can not but r e j e c t both Kuno's hypothesis and our r e v i s i o n . 170 CHAPTER VI NOTES 1. Kuno, Notes on Japanese Grammar, Chap. 19, p. 17. 2. Ibid., Chap. 19, p. 18. 3. Muraki, "Presupposition, Pseudo-clefting and Thematization," pp. 203—216. 4. Ibid., p. 204. 5. Ibid., p. 208. 6. Ibid., p. 213. 7. Susumu Kuno, "Functional Sentence Perspective : A Case Study from Japanese and English," L i n g u i s t i c  Inquiry, V o l . 3, No. 3 (summer, 1972), pp. 296 f . 8. Ibid., p. 284. 9. Kuno, op. c i t . , Chap. 1, p. 4. 10. Kuno, op. c i t . , pp. 272—276. 11. Ibid., p. 282. 12. Muraki, op. c i t . , pp. 152 f . & 166 f. 13. S e i i c h i Nakada, "Pseudo-clefts » What Are They?" Papers from the Ninth Regional Meeting. (Chicago» Chicago L i n g u i s t i c Society, A p r i l , 1973), p. 432. CHAPTER V I I SUMMARY AND CONCLUSIONS The main p o i n t s of our d i s c u s s i o n may be r e s t a t e d f o l l o w s * a. Ross's c l a i m t h a t the CNPC, the C r o s s - o v e r Con- d i t i o n , and the CSC apply t o Japanese r e l a t i v i z a -t i o n i s not always t r u e . Rather, we should say t h a t Ross's CNPC and P o s t a l ' s Cross-over C o n d i t i o n are mostly i r r e l e v a n t to Japanese r e l a t i v i z a t i o n . b. Japanese r e l a t i v i z a t i o n i s n e i t h e r a chopping r u l e n o r a copying r u l e n o r a f e a t u r e - c h a n g i n g r u l e . c. I t i s a d e l e t i o n t r a n s f o r m a t i o n by which an i d e n t i c a l embedded NP i s d e l e t e d without any movement. d. Because h i s Nearest NP C o n s t r a i n t b l o c k s gram-m a t i c a l sentences i n some cases and a l s o because Japanese has the backward d e l e t i o n a c r o s s the sentence, Muraki's argument f o r h i s u n d e r l y i n g s t r u c t u r e of a r e l a t i v e c o n s t r u c t i o n does not h o l d . e. We have d i s c u s s e d Kuno's f o l l o w i n g c l a i m s . ( i ) The r e l a t i v i z e d NP l e a v e s a ( r e f l e x i v e ) p r o -noun i n the o r i g i n a l p o s i t i o n under a c e r t a i n c o n d i t i o n . ( i i ) An NP i n an a d v e r b i a l c l a u s e i s r e l a t i v i z a b l e under some c o n d i t i o n s which are q u i t e unknown. I f we must base our r e a s o n i n g on h i s o t h e r hy-p o t h e s i s t h a t a r e l a t i v i z a b l e NP i s an NP imme-d i a t e l y , f o l l o w e d by a thematic p a r t i c l e wa, we w i l l have to r e j e c t ( e . i ) and ( e . i i ) . Here, what i s r e l a t i v i z e d s hould be the thematic NP. f . In the f i r s t h a l f of Chapter VI, we have r e v i s e d Kuno's above h y p o t h e s i s t o the e f f e c t t h a t a r e l a t i v i z a b l e NP i s thematic whether o r not there i s an i n t e r v e n i n g p a r t i c l e between the themat-i c NP and wa. g. However, i n the l a t t e r h a l f of the same chapter, we have observed t h a t there are some cases where non-thematized NP's are r e l a t i v -i z e d . Thus, we can not but d i s c a r d the above h y p o t h e s i s ( f ) . h. As a n a t u r a l consequence of the above con-c l u s i o n s , Kuno's two hypotheses ( e . i ) and ( e . i i ) would a r i s e a g a i n and c l a i m our re-examination. I f (g) i s c o r r e c t , we l o s e the ground f o r the r e j e c t i o n of these two hypotheses ( e . i ) and ( e . i i ) . i . However, i f we abandon h y p o t h e s i s ( f ) , we can not e x p l a i n why an NP i n an a d v e r b i a l c l a u s e i s r e l a t i v i z a b l e i n some cases and i t i s not i n o t h e r s . We can not account f o r the c o n d i t i o n s of r e f l e x i v i z a t i o n of a r e l a t i v i z e d NP, e i t h e r . I do not know at pre s e n t how t o r e c o n c i l e these c o n t r a d i c t o r y phenomena. However, one p o s s i b l e s o l u t i o n may be found i n Lewkowicz's o b s e r v a t i o n . Lewkowicz says that i n Arabic only s u b j e c t - i n i t i a l (including t o p i c -i n i t i a l ) sentences can be used as r e l a t i v e clauses. This means that there are two types of underlying struc-tures f o r r e l a t i v e clauses, that i s , non-thematized sub-ject i n i t i a l sentences and theme-initial sentences. 1 This may be useful f o r our analysis of Japanese r e l a t i v e clauses. In other words, though Japanese and Arabic are d i f f e r e n t , t h i s leads me to speculate on the p o s s i b i l i t y that there may also be two types of under-l y i n g structures f o r Japanese r e l a t i v e constructions. One of these involves a thematic NP and the other would not. In f a c t , each of the examples given by Kuno concern-ing (e.i) and ( e . i i ) i s supposed to contain a thematic NP i n i t s underlying structure. Each sentence requires that an embedded r e l a t i v e clause be a complex sentence which contains i n i t a deeply embedded sentence and that an NP i n t h i s deeply embedded constituent sentence be i d e n t i c a l to the head NP i n a matrix sentence. On the other hand, our counter-examples do not i n -volve the deeply embedded constituent sentences. Our counter-examples contain exhaustive l i s t i n g or neutral  description. The exhaustive l i s t i n g and neutral descrip-t i o n involve new information-part i n conversation. A new information-part must, be simple and also c l e a r . It must be expressed i n the form of a simplex sentence or i n the form of a complex sentence which has one embedded sentence. It must not be too complex and complicated. When one has much new information to convey, one can divide the information into a set of simplex sentences or a sequence of conjoined sentences. One w i l l not change i t into one complex sentence which contains a complicated s t r u c t u r a l hierarchy. Thus, I conjecture that there may be two types of r e l a t i v e clauses; one involves a thematic NP and the other does not. The above extremely tentative solution has not yet been f u l l y examined, and the question of how many deeply embedded sentences are possible i n each sentence i s quite unknown i n Japanese. That seems to be closely related-to the problem of s t y l e . It may also require some r e l a t e d research i n the f i e l d of psychology which l i e s beyond .the scope of our present study. We w i l l have to leave the solution of t h i s problem f o r the future. CHAPTER VII NOTES 1. Lewkowicz, "Topic-comment and Relative Clause i n Arabic," pp. 813—816. 178 BIBLIOGRAPHY Annear, Sandra. "Relative Clauses and Conjunctions," Working Papers i n L i n g u i s t i c s * > (December, 1967). Columbusi Ohio State University, pp. 80~ 9 9 , Kuno, Susumu. Notes on Japanese Grammar, Mathematical L i n g u i s t i c s and Automatic Translation, f  Report NO. NSF - 2 7 . Cambridge, Massachusetts i Harvard University, 1970. "Pronominalization, R e f l e x i v i z a t i o n , and Direct Discourse," L i n g u i s t i c Inquiry. Vol, 3. No.2 (1972), pp. 161 - 195. "Functional Sentence Perspective! a Case Study from Japanese and English," L i n g u i s t i c  Inquiry, Vol. 3. No. 3 (1972), pp. 269 - 320. m The Structure of the Japanese Language. Cambridge, Massachussets i The MIT Press, 1973. Langacker, Ronald W. "On Pronominalization and the Chain of Command," Modern Studies i n English, Edited by David A. Reibel ana Sanford A. Schane. Englewood C l i f f s , New Jersey! Prentice-Hall Inc., 1969. Lewkowicz, Nancy K. "Topic-comment and Relative Clause i n Arabic," Language, Vol. 47, No. 4 ( 1 9 7 D i pp. 816 - 825. Loetscher, Andreas. "On the Role of Nonrestrictive Relative Clauses i n Discourse," Papers from the  Ninth Regional Meeting, Edited by Claudia Corum.T. Cedric Smith-Stark and Ann Weiser. Chicago! Chicago L i n g u i s t i c Society, A p r i l , 1973. Makino, S e i i c h i . Some Aspects of Japanese Nominalizations. Tokyot Tokai University Press, 1908, McCawley, James D. "Japanese Relative Clauses," The Chicago Which Hunt, Papers from the Relative  Clause F e s t i v a l , Edited by"Paul M. Peranteau, Judth N. Levi and G l o r i a C. Phares, Chicago! Chicago L i n g u i s t i c Society, A p r i l , 1972. 179 MeCawley, Noriko A. "A Study of Japanese R e f l e x i v i z a t i o n , " Unpublished Ph.D. d i s s e r t a t i o n . I l l i n o i s University, 1972. Mikami, Akira. Zoo wa Hana ga Nagai. Tokyoi Kurosio Pub. "So". , 1 9 5 9 7 ~ ~ Muraki, Masatake. "Presupposition, Pseudo-clefting and Thematization." Unpublished Ph.D. di s s e r t a t i o n . The University of Texas at Austin, 1971. Nakada, S e i i c h i . Nakau, Minoru. "Pseudo-clefts1 What Are They?" Papers  from the Ninth Regional Meeting. Edited by Claudia Corum, T. Cedric Smith-Stark and Ann Weiser. Chicago L i n g u i s t i c Society, A p r i l , 1973. "Some Constraints on T o p i c a l i z a t i o n , " Papers i n Japanese L i n g u i s t i c s , V o l . 1, No. 1 (1972)i pp. 74 - 88. Ohso, Mieko. "A Remark on Japanese R e l a t i v i z a t i o n as a Process of Deletion." Unpublished paper. Columbus: Ohio University, 1972. Oyakawa, Takatsugu. "Japanese R e f l e x i v i z a t i o n I," Papers 4 ~- X -T • • i_ • 1 A * « . Park, Byung-soo. Postal, Paul M. Ross, Robert J . Schachter, Paul. i n Japanese L i n g u i s t i c s , Vol. 2,* No. 1 (1972), pp.94 - 135. "On the Multiple Subject Construction i n Korean," L i n g u i s t i c s i n International  Review, Vol. 100 (March, 1973), PP. 72 - 76. Cross-over Phenomena. New York t Holt, Rinehart and Winston, Inc., 1971. "Constraints on Variables i n Syntax." Unpublished Ph. D. d i s s e r t a t i o n . Massachusetts In s t i t u t e of Technology, 1967. "Focus and R e l a t i v i z a t i o n , " Language, Vol. 49, No. 4 (March, 1973), PP. 19~45. , Shibatani, Masayoshi. (Ed.) Papers i n Japanese L i n g u i s t i c s , Vol. 1, No. 1"T1972^. University of C a l i f o r n i a , Berkeley: The Japanese L i n g u i s t i c Workshop. 180 Shibatani, Masayoshi. (Ed.) Papers i n Japanese L i n g u i s t i c s , Vol. 2, No.l T 1 9 7 3 ) . University of C a l i f o r n i a , Berkeley: The Japanese L i n g u i s t i c Workshop. Thompson, Sandra A. "The Deep Structure of Relative Clauses," Studies i n L i n g u i s t i c Semantics, Edited by Charles J . Fillmore and D. Terence Langendoen. New York: Holt, Rinehart and Winston, Inc., 1971. 

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.831.1-0093228/manifest

Comment

Related Items