Natural Logic and the Greek Moods: The Nature of the Subjunctive and Optative in Classical Greek 9783110819465, 9789027930613

Citation preview

JANUA L I N G U A R U M STUDIA M E M O R I A E N I C O L A I VAN WIJK D E D I C A T A edenda curat C. H. VAN SCHOONEVELD Indiana University

Series Practica,

230

NATURAL LOGIC AND THE GREEK MOODS The Nature of the Subjunctive and Optative in Classical Greek

by

DAVID LIGHTFOOT McGill

University,

Montréal

1975

MOUTON THE HAGUE • PARIS

© Copyright 1975 Mouton & Co. B.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.

ISBN 90 279 3061 9

Printed in The Netherlands by Zuid-Nederlandsche Drukkerij N.V., 's-Hertogenbosch

PREFACE

Over the last few years a dispute has been waged amongst proponents of Transformational Grammar, always vigorously and often bitterly. The 'interpretivists' have maintained that syntactic rules operate autonomously of semantics and that a semantic representation for any sentence is generated by rules which 'interpret' the syntactic representations. The 'generativists', on the other hand, have claimed that syntactic rules are inextricably intertwined with semantic considerations and, indeed, that there is no useful distinction to be drawn between syntax and semantics. They visualise a grammar as consisting of one set of more or less homogeneous rules which map semantic representations into surface syntactic structures. Semantic representations will be based on a 'natural logic', a logic which will be built up to capture not only the traditional concerns of the standard logics but also linguistically significant generalisations. This book does not attempt to advocate Generative Semantics over Interpretive Semantics or vice versa. Indeed, it is by no means clear that there is any such choice to be made, or whether there is a substantive difference between these two positions. Rather, we are undertaking a detailed examination of a certain area of grammar within the framework of Generative Semantics. We claim to solve a number of problems in the description of the Classical Greek mood system, explaining the nature of the subjunctive and optative moods. In doing this, we demonstrate certain strengths of the Generative Semantics position but we also show certain important problems inherent in it. Most of these problems centre on methods of argumentation: what constitutes valid evidence for an abstract verb? how can one give a sensible definition of 'presupposition', so that the presuppositions of a given sentence are not indeterminable and infinite? Chapter III in particular is concerned with problems of substantiating abstract verbs. Chapter IV makes claims about the presuppositions of several syntactic structures, particularly reason and conditional clauses, indefinites (sentences containing any, ever, etc.), and universal quantifiers. Given the two-fold purpose of this book, to solve certain problems in the analysis of Greek and to identify problems for Generative Semantics, and for linguistic theory as a whole, it should not surprise readers to find that more than half of the example

6

PREFACE

sentences are in English. It is more convenient and in keeping with the traditions of Transformational Grammar to discuss points of theory largely in the framework of English. Sometimes I discuss Greek through English, because in many cases this makes the discussion more transparent. At the same time, there is enough Greek to prevent this from becoming irresponsible. I should mention my great debt to Robin Lakoff. I am building on the framework she adopted in her analysis of Latin and in fact it was the power and elegance of her theory of abstract verbs which led me to undertake this study in the first place. The book began life as a doctoral dissertation, although it has since been revised radically in places; Robin was my thesis advisor and I benefitted enormously from her careful comments and criticisms. Also in order is a word of public thanks to my wife Sarah. When linguists' intuitions begin to wobble, their spouse is always the first line of defence; no sooner do they wake up in the morning than they are asked if they can accept some sentence like who did the realisation that he was President disturb? Sarah's answers were always 'right'. For her an evening at Le Petit Havre. 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. I am deeply indebted to them for their financial assistance and for their referees' reports on an earlier manuscript, which helped to make this a better book. David Lightfoot McGill University, Montréal March 1973

TABLE OF CONTENTS

Preface I. Introduction: The Problem II. Greek Complementation III. Abstract Verbs in Greek A. Lakoff's Abstract Verbs B. Abstract Verbs in Greek

5 9 25 49 49 55

IV. Mood Uses in Subordinate Clauses A. The Abstract Verb [entail] B. The Interaction of Universal Quantifiers and Adverbial C l a u s e s . . . . C. McCawley's Quantification Proposals D. Will and would in English E. Subordinate Clauses in Greek

79 79 96 110 116 126

V. Conclusion

133

References

144

Index

148

I INTRODUCTION: THE PROBLEM

In Classical Greek the verb is marked morphologically for number, person, tense, aspect and mood. There are six distinct moods: the indicative, subjunctive, optative, imperative, and, according to some definitions of 'mood', the infinitive and the participle. The first three are finite and can be marked for all three persons in both the singular and the plural. This is where our primary interest lies. These moods are names for certain paradigms or sets of desinences; in fifth century Classical Greek the subjunctive endings have one of the inherently long vowels, t] or co, and the optative endings fall into two major classes but always have some kind of iota diphthong. When used in an independent clause each of these moods will carry quite distinct ideas and will be translated differently. (1) ol ap)£0VTE

Of course, such a viewpoint brings its problems, not least of which is that of handling complex sentences with adverbial clauses. This is no problem if one posits more complicated phrase structure rules which contain a node ADVERB from which one is willing to subtend all adverbial clauses by a series of ad hoc rules. All this says, however, is that languages contain adverbial clauses and it does not attempt to relate these clauses in any meaningful way which could be other than language specific. Lakoff did not try to account for the mood uses in conditional and relative clauses, but that does not necessarily discredit her theory. We shall have some suggestions to make about these clause types in Greek in Chapter IV. Lakoff pointed to the ambiguity of the Latin sentence venias, which consists of an independent subjunctive. It may be interpreted as an imperative, volitive or potential, and might mean either of: (a) (b) (c)

'come!' 'if only you would come' 'perhaps you are coming'

However, each of these three uses is subject to different syntactic constraints and may be disambiguated. The negative of (a) and (b) is ne, so non venias is unambiguous 8

For the purposes of this discussion I am comparing an overt verb of ordering in the present tense and first person singular. 9 J. R. Ross, 'Auxiliaries as Main Verbs', in Studies in Philosophical Linguistics; Series /, ed. by William Todd (Evanston, 1969).

ABSTRACT VERBS IN GREEK

55

and has only the potential reading of (c). We have seen that subjunctives of command cannot be first person singular, so veniam is ambiguous only between (b) and (c), and ne veniam is unambiguously the negative of (b). Similarly, as we have seen, stative verbs cannot occur with imperative meaning, so alius sis is ambiguous only between (b) and (c). Moreover, only (c) may occur in a question. To say that one is a true subjunctive, the second a volitive and the third a potential, as do many Latinists, says nothing about the relationship between them and why they have the same set of endings. It does not say why, if (a) and (b) have different origins, they are negated similarly and why (c) is negated differently. If mood is an element of deep structure then each of these constraints will have to be stated for the specific kind of subjunctive, so one will need special rules and constraints of something like the following form simply to account for these few facts: 1.

neg

ne / subjunctive [+imperative]] [+volitive] J

2. 3. 4. 5. 6.

neg

non / subjunctive [+potential] V -> (-stative) / subjunctive [+imperative] V (-first person singular) / subjunctive [+imperative] V -»• (+present tense) / subjunctive [+imperative] Block Q / subjunctive [ [+imperative] | [[+volitive] J

There are, of course, many possible ways of stating these facts, but the point is that the facts will have to be stated in special rules and in terms of semantic features on the particular kind of subjunctive. They are isolated facts, unrelated to whatever else we know about subjunctives. Lakoff handled all of these facts by independently motivated rules by positing three abstract verbs. The meaning class of the abstract verb gives the meaning of the subjunctive and governs the form of the negation, etc.

B. ABSTRACT VERBS IN GREEK

We have established that the Greek subjunctive and optative used in complement constructions carry no distinctive meaning of their own; that they differ from the infinitive and indicative in such constructions only syntactically, in that certain main verbs may exclude certain constructions. Thus if one wants to claim that subjunctives

56

ABSTRACT VERBS IN GREEK

and optatives are not generated randomly or as unsystematic accidents, in this formulation one will need to look at the class of verbs which govern a subjunctive mood in the subordinate clause and to determine what characterises that class of verbs. In the last chapter we suggested that all such verbs shared the property of future directedness. There are, however, certain independent uses of the subjunctive and optative in main clauses with no apparent governing verb. Here clauses with a main verb in one of these two moods will differ semantically from a clause with an indicative main verb. A sentence with its main verb in the subjunctive will denote a command or a fear or a tentative assertion; a main verb in the optative will denote a wish for the future or a potential referring to the future if it is accompanied by the modal particle We noted on page 43 that verbs of the will meaning class often undergo CP(d), and we quoted the example from Plato's Republic, SiaxeXeuovTai, arc«? Ti^copyjaeTai. So SiaxeXeiofiai will be marked in the lexicon to indicate that it may undergo CP(d). However, a number of other constraints on the lower clause need to be noted and articulated by the grammar. If the governing verb is a verb of ordering, the following constraints will apply. 1. The lower verb may not be stative. Thus it is not possible to say *8iaxeXeuovTai &TZ(I>C, 00961; east, because