THE SEMANTICS OF COORDINATION STUDIES IN LANGUAGE COMPANION SERIES (SLCS) The SLCS series has been established as a companion series to STUDIES IN LANGUAGE International Journal sponsored by the Foundation "Foundations of Language". Series Editors: John W. M. Verhaar Werner Abraham Gonzaga University & University of Groningen Spokane, WA The Netherlands Volume 9 Ewald Lang The Semantics of Coordination THE SEMANTICS OF COORDINATION EWALD LANG Translation: John Pheby AMSTERDAM/JOHN BENJAMINS B.V 1984 Authorized english translation by John Pheby (Oxford) from the German original edition "Semantik der koordinativen Verknüpfung", Berlin, 1977. Library of Congress Cataloging in Publication Data Lang, Ewald. The semantics of coordination. (Studies in language companion series, ISSN 0165-7763; v. 9). Translation of: Semantik der koordinativen Verknüpfung. Bibliography: p. 289 Includes index. 1. Grammar, Comparative and general — Coordinate constructions. 2. Semantics. I. Title. II. Series. P293.L3613 1984 415 84-14541 ISBN 90-272-3008-0 ® Copyright 1984 - John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. Contents Foreword 7 Acknowledgements 11 Introduction 13 1. The problem and its background 13 2. The specific object of investigation 17 Chapter I Towards a theory of coordination: Some basic assumptions 19 1. Raw material and definition of coordination 19 2. Characteristics of coordination 23 2.1. Preliminary considerations 23 2.2. Sentence vs. utterance, grammar vs. communication 24 2.3. Coordination as integration 27 2.4. Basic assumptions 37 2.5. Heuristic and method of enquiry 38 3. Aspects of conjoining 40 3.1. Homogeneousness of conjuncts 40 3.2. Parallelization effect of coordination 51 3.3. Further surface effects of coordination 63 3.4. The role of connectors in conjoining 66 3.5. Common Integrator (CI) of the conjunct-meanings 69 3.6. Syndesis vs. asyndesis. Symmetry vs. asymmetry 79 3.7. »Sentential conjunction' vs. ,Phrasal conjunction' 85 4. Where are the limits to the scope of a linguistic description of the semantics of coordination? 93 5. The structure of the proposed theory of the semantics of coordination 95 Chapter II Relations between conjunct-meanings 98 1. Distinctness of conjunct-meanings 98 2. Semantic inclusion of conjunct-meanings 113 3. Incompatibility between conjunct-meanings 117 Chapter III Outline of the semantics of coordinate conjoining 124 0. Introduction 124 1. Logical characterization of coordinate conjoinings 125 1.1. Inventory of examples 125 1.2. Logical characterization of semantic relations between conjunct-sentences 130 1.3. Logical characterization of the connector-meanings 132 1.4. Logical characterization of coordinate conjoinings 134 2. Linguistic interpretation of the logical characterization of coordinate conjoinings ... 137 2.1. Logical and linguistic properties of conjoinings 137 2.2. Contextual conditions for connectors 140 2.3. Components of connector-meanings 143 5 2.4. Contradiction vs. incomprehensibility 144 2.5. Tautology vs. redundancy 147 2.6. Contingent conjoinings 148 2.7. Summary 152 2.8. Aspects of connector universals 156 3. Some properties of the connectors but and for 169 3.1. but 169 3.2. for 175 4. Survey of connector-meanings 178 4.0. Preliminaries 178 4.1. Notation 179 4.2. Connector-meanings 180 4.3. Interrelations of connector-meanings: a case study on DE MORGAN'S laws 183 Chapter IV Coordinate structures with modal verbs 191 0. Introduction 191 1. The raw material 192 1.1. German modal verbs 192 1.2. The representation of the conjuncts 193 1.3. Meaning postulates 195 1.4. Interpretation of the modal predicates 199 1.5. Semantic networks 202 2. Coordinate conjoinings with modal verbs 204 2.1. Notation 204 2.2. Conjoinings with und [and] 206 2.3. Common Integrator in conjoinings with modal verbs 218 2.4. Conjoinings with aber [but] 232 2.5. Conjoinings with denn [for] 235 2.6. Conjoinings with oder [or] 236 2.7. Conjoinings with entweder — oder [either — or] 236 2.8. Conjoinings with nicht — sondern [not — but] 238 2.9. nicht — sondern vs. nicht — aber: ,correction' vs. ,contrast' 251 Chapter V Discussing and extending the notion ,Common Integrator' 263 1. »Common Integrator' in comparison to related concepts 263 1.1. »Common Denominator' in WIERZBICKA (1972) and (1980) 263 1.2. »General Formula' and »General proposition' in MCCAWLEY (1973) and (1981) 266 1.3. »Common Topic' in R. LAKOFF (1971) 268 2. Common Integrator in word formation 270 2.1. Coordinate compounds in German 270 2.2. Coordinate compounds in Chinese 271 3. Common Integrator in other connections 274 Chapter VI Some applications 276 1. Exemplification as a stylistic device 276 2. Reconstruction of ideological background 279 3. Demagogy 282 4. Jewish joke 286 Bibliography 289 Subject Index 297 Name Index 299 6 Foreword This study is an attempt to explain coordinate conjoining as a rule-governed process of establishing specific semantic relations within and between sentences. Coordination is thus conceived of both as a basic device of linguistic complex formation and as a rather fundamental principle underlying the creation of text. The many different ways in which this principle is manifested in coordinate structures are taken to be governed by the necessity of achieving semantic coherence and thus to be subject to the conditions on doing so. Coordinate structures consist of conjuncts in the presence, latent or actual, of connectors. Coordinate structures are the structural realization and coordinate conjoining is the operational realization of the principle of coordination. From the point of view of achieving coherence, coordinate conjoining is described here as an integrative process, the result of which — to adapt a well-known principle — is that the semantic interpretation of a given coordinate structure is more than the sum (or list) of its individually interpreted parts. What will be described, therefore, are the conditions governing this process, the rules according to which it takes place, in short : the complex interaction of various linguistically identifiable features displayed by coordinate structures. The new approach proposed in this study is based precisely on not regarding coordinate conjoining merely as a function of connectors. Such a view would lead to an unjustified reduction of linguistic connectors to logical functors (see Chapter III). Rather, coordinate conjoining is regarded here as the result of the interplay of three factors which belong to distinct levels of semantic description : (1) Meaning of the conjuncts (level of propositions : conceptual and referential semantics) (2) Relations between the meaning of the conjuncts (level of entailment relations between propositions or hierarchical relations between concepts) (3) Meaning of the connectors (operational meaning) The step-by-step explication' of the interaction of (1)—(3) in determining the semantic interpretation of coordinate structures forms the core of the present study. Not only what is being described but also the description itself is an integrative process: the detailed characterization of coordinate structures in terms of grammar has on the one hand to be combined with very general postulates for successful communication, because it is only against this background that acceptability evaluations of coordinate structures — serving as a heuristic key 7 to the whole — can be carried out effectively. On the other hand the grammatical characterization of coordinate structures must be related to a suitably adapted model of cognitive processing, because this is the sphere in which coordinate conjoining as a process of integration takes place. Finally, all three aspects, which for reasons of method are first of all developed separately and then connected in pairs, must be interconnected through various channels to form one unified, coherent description. In view of this process it is inevitable that this book will contain some passages which require considerable patience, namely those sections which deal primarily with problems of representation and with the interconnections between systems of interpretation. That such a long march through different systems is nevertheless worthwhile is proved in Chapter VI by the application of the analysis to some examples taken from real life. Chapter VI can in fact be read out of turn : as a way of approach to the analysis, for light relief during the reading of the more exacting parts, or as a dessert after the heavy main course. Finally, a few remarks on the relationship between this version and its German predecessor of 1977. The overall exposition has been retained. The text itself, however, has undergone substantial revisions as to form and content. The Introduc­ tion as well as Chapters I, III, and IV have been entirely redone. In the remaining parts several sections have been added and some passages have been abandoned. Drawing on some central issues in current semantic theorizing, e.g. the need to distinguish semantic from conceptual representations, or the task of specifying the role of verbal and/or situational contexts in determining the range of possible semantic interpretations of a given linguistic structure, I have attempted to sharpen the basic notions of my approach and to tighten up the network of their interrelations. The notion of Common Integrator, which I regard to be the heart of the matter, has been accorded more space than before. There are two important areas which have not been explored here in detail, though the reader will find pointers to them in many places. First, I have made no attempt to provide a formal account for (1)—(3) above in terms of, say, one of the current variants of model-theoretic semantics, because such an undertaking seemed to me to be premature. Recent proposals of Gerald GAZDAR (1980) and Barbara PARTEE/Mats ROOTH (1982), however, suggest that such an effort can, and in fact, should be made in the near future. Second, due to space limitations I did not include the presentation and discussion of some psychological data on speech production which provide independent evidence for the linguistic findings of Chapter I. 3.1.—3.7. To put it briefly: in 1982, during a stay at the Max-Planck- Institute for Psycholinguistics (Nijmegen, The Netherlands), I got acquainted with a large corpus of spontaneous speech repairs, whose structure and functioning display many striking similarities to the structure and functioning of coordinate conjoinings (cf. Willem J. M. LEVELT (1983)). Put in terms of LEVELT'S speech production model the most relevant observation reads thus: "The speaker, in making coordinations, parses his first conjunct and feeds some of its structural pro­ perties jointly with the content-information into the formulator. In this way some of 8 the formulator's procedures may become directly activated by structural properties of previous speech, over and above activation by conceptual input." (LEVELT, ibid.) Pursuing this line of thinking we may arrive at what I will call a "Template Theory of Coordination" drawing on the following assumption: the result of parsing the first conjunct is transferred to the formulator as a template consisting of categorized variables of which the terminal elements of the first conjunct are instances. The elaboration of this empirically well supported view will provide a means for explaining lots of otherwise unexplainable phenomena concerning the homogeneousness of conjuncts, their status as exemplifications of a Common Integrator etc. — see Chapter I. In other words, facts concerning the structural properties of acceptable coordinate structures may thus be related to the way in which coordinate structures are produced in speech. Thus it is not only in view of these two areas but also in view of many other problems not taken up here that the present study should be interpreted as a first volume written in such a way as to anticipate continuation. Berlin, December 1982 E. L. 9