THE SYNTAX OF JAPANESE HONORIFICS JANUA LINGUARUM STUDIA MEMORIAE NICOLAI VAN WIJK DEDICATA edenda curat C. H. VAN SCHOONEVELD INDIANA UNIVERSITY SERIES PRACTICA 102 1970 MOUTON THE HAGUE · PARIS THE SYNTAX OF JAPANESE HONORIFICS by GARY DEAN PRIDEAUX THE UNIVERSITY OF ALBERTA G8 1970 MOUTON THE HAGUE · PARIS © Copyright 1970 in The Netherlands. Mouton & Co. Ν.V., Publishers, The Hague. No part of this book may be translated or reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publishers. LIBRARY OF CONGRESS CATALOG CARD NUMBER: 72-95009 Printed in The Netherlands by Mouton & Co., Printers, The Hague. PREFACE Advances in generative grammar have been rapid and far-reaching. As new data are uncovered, grammars of particular languages are being constantly revised. One of the most important discoveries resulting from such improvements is that the deep structures of sentences must often be far more abstract than had been suspected previously. Consequently, particular analyses quickly become obsolete. This ten- dency has been especially evident in English studies, but it is also true of Japanese grammars. Recent investigations indicate that several formulations presented in the present work are faulty. In particular, the rather complicated passive, causative, and benefactive rules seem to be inadequate. The benefactive construction seems to be readily derivable from existing, independently motivated rules without the need of a special benefactive transformation. The passive and causative morphemes appear to be better treated as independent verbs which require sentential complements. In this case, passive forms of both transitive and intransitive verbs could be accounted for, while in the present work only transitive passives are allowed. In spite of such deficien- cies, it is hoped that the treatment of the various topics presented here, especially the so-called honorifics of Japanese, will provide a firm foundation for further analyses. It is now clear that a proper formulation of the honorifics of Japanese is far more complex than simply the addition of a few optional rules to an existing grammar. An adequate grammar of Japanese must account for the various restrictions and quite complicated levels of politeness available to a native speaker of the language. This work is a slightly revised version of a doctoral dissertation presented to The Univeisity of Texas in 1966. I am especially grateful for the constant guidance and help of Emmon Bach, under whose direction the original work was done. I am also endebted to Robert T. Harms for his tireless assistance and suggestions. To Yasuko and Karou Takata, who acted as my principal informants, I owe special thanks for their patience and friendship. Finally, I should like to thank my wife, Ann Prideaux, for her continued encouragement and editorial assistance. The original work was made possible at The University of Texas through NDEA Fellowships, while revisions were supported by a Research Grant from the University of Alberta. This book has been published with the help of a grant from the Humanities Research Council of Canada using funds provided by the Canada Council. Tottori-shi, Japan May 1968 G.D.P, TABLE OF CONTENTS Preface 5 1. The Problem 9 1.1 Polite and Formal Language 9 1.2 Previous Studies of the Honorifics 11 1.3 Transformational Treatments of Japanese 15 1.4 Revisions in the Transformational Model 19 2. The Base Rules 23 2.1 Definitions 23 2.2 The Branching Rules 24 2.3 The Feature Rules 28 3. The Lexicon 35 3.1 The Nature of the Lexicon 35 3.2 Subcategorizational Features 37 3.3 Selectional Features 40 3.4 Additional Features 42 4. The Polite and Formal Transformations 45 4.1 The Nature of the Transformations 45 4.2 Copula and Formality Transformations 47 4.3 Politeness Transformations 54 4.4 Titles 59 4.5 Imperatives 60 5. Some Additional Problems 63 5.1 Introduction 63 5.2 Adverbs 63 5.3 Pronouns 68 5.4 The Phonological Component 71 5.5 A Second Alternative Grammar 75 8 TABLE OF CONTENTS Appendix A. Conventions 77 Appendix B. Japanese Grammar 79 1. Base Component 79 2. Sample Lexicon 81 3. Transformational Component 85 4. Phonological Component 101 Bibliography 104 Index 106 1. THE PROBLEM 1.1 POLITE AND FORMAL LANGUAGE A distinction is made in modern Japanese between the colloquial and the learned language. This distinction is based primarily "...on the inflection; all levels of the learned language use the classical verb system."1 In this study I shall consider only the colloquial language. The only inflected forms in modern colloquial Japanese are the verbs, although other classes such as the nouns may take prefixes marking them as polite. Verbs are not marked for person, number, or gender, but rather they are inflected obligatorily for tense (past or non-past) and optionally for aspect (durative, completive, prepar- atory, or transitive stative). In addition the verb may enter into a number of modal constructions and further may take one of a number of tenseless connectors which serve to join two sentences. If such a connector is used, the tense of the verb is deleted, and the tense of the conjoined string of sentences is provided by the tense of the verb of the final sentence. Forms usually considered as adjectives are better classed as verbs, since they are obligatorily inflected for tense (though not for aspect or mode). The verb charac- teristically appears in sentence final position, except for an optional final particle or connector, and the adjectives also follow this pattern (with the minor exception that the copula is added in formal constructions). Attributive constructions using adjectives and those using verbs are exactly alike. In both cases the adjective, or the verb, immediately precedes the noun being modified. For example: akai hon '(a) red book' kinoo katta hon '(the) book (I) bought yesterday' akakatta hon '(the) book which was red' In all such constructions, the verb (or adjective) which is inflected for tense imme- diately precedes the noun. 1 W. P. Lehmann and Lloyd Faust, A Grammar of Formal Written Japanese (Cambridge, Harvard University Press, 1951), p. 1. 10 THE PROBLEM The strongest reason for considering adjectives to be verbs is that such a choice simplifies the grammar in significant ways. For example, the various rules in the base which assign features to the verbs would have to be stated separately for adjectives if the latter were not considered as verbs. Also, the various transformations which adjust embedded sentences used as modifiers would be more complex if the adjectives were considered separately from the verbs. Further, the total number of lexical cate- gories is reduced, and all the rules operating on them consequently shortened if this choice is made. The savings obtained by considering adjectives as a subclass of verbs in Japanese raises the question of the independence of 'adjective' as a lexical category in other languages. (Lexical and grammatical categories make up the set of terminal cate- gories of the grammar. The term TERMINAL CATEGORY is taken here in the technical sense defined in Chapter II. A TERMINAL CATEGORY refers to an element in a branching rule of the base component of the grammar which is rewritten as a matrix of features.) Γη several Indo-European languages, the traditional adjective is inflected to agree with the noun it modifies. In Spanish, for example, the adjective agrees with the modified noun in gender and number, while in Sanskrit, the agreement is in gender, number, and case. However, these agreements can be accounted for by late trans- formations in the grammar, and there are reasons for formulating even these Indo- European adjectives as a subclass of verbs in the base component of the grammar. The agreement rules would then be a late set of transformations. Absence of the lexical category of adjective, if established on independent grounds (those of internal simplification of the grammar) within several languages, would suggest the possibility of eliminating the category from a universal inventory of lexical categories from which individual transformational grammars draw. The absence of 'adjective' and of as many other lexical catagories as possible could effect a considerable savings in the grammars of individual languages. Such a saving could be accomplished much in the same way as it is in the present grammar of Japanese. That is, if adjectives were not considered as a subclass of verbs, then those rules which apply to both adjectives and verbs would have to be stated separately, but if the adjective is considered a subclass of verbs, then the same transformations can be stated in terms of the verb and will automatically apply to the adjectives as well. Of course, the absence of the lexical category of 'adjective' does not mean that the classes traditionally called adjectives do not exist in such languages as Japanese or English or Sanskrit, but rather that if the 'adjectives' are formulated as a part of another lexical category, then the grammar can conceivably be simplified. The claim that the adjective is universally a subclass of the lexical category of verb is an empirical hypothesis and must be tested. (Empirical testing of the claim can either support or disprove it, and the testing will consist of providing alternative analyses and choosing the simpler.) For simplicity of reference, I shall continue to use the term 'adjective' to refer to the subclass of verbs containing the feature [-Vb]. In addition to the constructions mentioned above, the Japanese verb also enters