Tài liệu Izchak m. schlesinger cognitive space and linguistic case_ semantic and syntactic categories in english (studies in english language) (1995)

This study sheds new light on the complex relationship between cognitive and linguistic categories. Challenging the view of cases as categories in cognitive space, Professor Schlesinger proposes a new understanding of the concept of case. Drawing on evidence from psycholinguistic research and English language data, he argues that case categories are in fact composed of more primitive cognitive notions: features and dimensions. These are registered in the lexical entries of individual verbs, thereby allowing certain metaphorical extensions. The features of a noun phrase may also be determined by its syntactic function. This new approach to case permits better descriptions of certain syntactic phenomena than have hitherto been possible, as Schlesinger illustrates through his analysis of the feature compositions of three cases. STUDIES IN ENGLISH LANGUAGE Executive Editor. Sidney Greenbaum Advisory Editors: John Algeo, Rodney Huddleston, Magnus Ljung Cognitive space and linguistic case Studies in English Language The aim of this series is to provide a framework for original studies of present-day English. All are based securely on empirical research, and represent theoretical and descriptive contributions to our knowledge of national varieties of English, both written and spoken. The series will cover a broad range of topics in English grammar, vocabulary, discourse, and pragmatics, and is aimed at an international readership. Already published Christian Mair Infinitival complement clauses in English: A study of syntax in discourse Charles F. Meyer Apposition in contemporary English Jan Firbas Functional sentence perspective in written and spoken communication Forthcoming John Algeo A study of British—American grammatical differences Cognitive space and linguistic case Semantic and syntactic categories in English IZCHAK M. SCHLESINGER Department of Psychology, The Hebrew University, Jerusalem CAMBRIDGE UNIVERSITY PRESS CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 2RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521434362 © Cambridge University Press 1995 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1995 This digitally printed first paperback version 2006 A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Schlesinger, I. M. Cognitive space and linguistic case: semantic and syntactic categories in English / Izchak M. Schlesinger. p. cm. - (Studies in English language) Includes bibliographical references and index. ISBN 0 521 43436 X (hardback) 1. English language — Grammatical categories. 2. English language— Semantics. 3. English language — Syntax. 4. English language — Case. I. Title. II. Series. PE1199.S35 1995 425-dc20 95-12805 CIP ISBN-13 978-0-521-43436-2 hardback ISBN-10 0-521-43436-X hardback ISBN-13 978-0-521-02736-6 paperback ISBN-10 0-521-02736-5 paperback For Avigail, who made it possible Contents Preface Introduction 1 page xiii i 1 2 3 4 5 6 Cognitive space Semantic and cognitive categories Instrument and Accompaniment - rating studies Ratings of Instrument, Accompaniment, and Manner Notions expressed by with — & qualitative study Agent and Experiencer Cognitive space and grammar 4 4 5 12 14 21 23 1 2 3 4 5 6 7 8 Agent and subject The semantics of the subject Features of the Agent Feature assignment Case assignment Linking Non-agentive subjects Converse verbs Conclusions 28 28 30 40 45 49 52 54 57 The Comitative fftfA-phrases Features of noun phrases in B^Y/r-phrases The Comitative Linking Studies of linking Conclusions 60 60 61 70 75 86 91 2 3 1 2 3 4 5 6 4 Non-comitative instruments 1 The "Instrument" in subject position 2 Constraints on subjects 92 92 94 ix x Contents 3 4 5 6 7 5 Degree of CONTROL in the inanimate subject Conjoining of subject noun phrases Passives with instruments Semantic saturation and language acquisition Conclusions Predicates 1 What is a predicate? 2 Phrasal predicates 3 Events and States 6 98 104 106 109 no in in 115 117 1 2 3 4 5 The Attributee The subjects of States Noun phrases that are both A-case and Attributee Implicit Events and States Passive sentences Conclusions 122 122 124 126 128 138 1 2 3 4 5 6 7 8 Mental verbs Experiencer and Stimulus - the problem The subjects of mental verbs The objects of mental verbs CONTROL in the Experiencer and the Stimulus Studies of rated CONTROL The Event-State dichotomy of mental verbs The linguistic realization of experiences Conclusions 139 139 140 144 146 149 151 156 162 1 2 3 4 5 6 Objects Direct objects Objects of prepositions Linking of core arguments Deletion of prepositions Features of verb phrases Conclusions 163 163 167 169 173 175 179 1 2 3 4 5 6 Verb classes and Agents Subdivisions of verbs A proposal for a subdivision Substitutability by pro-forms Pseudo-cleft sentences Accounting for the hierarchy The agency gradient 180 180 181 183 190 192 199 7 8 9 Contents xi 7 The psychological reality of verb classes 8 Conclusions 10 1 2 3 204 209 Retrospect and prospects The traditional view The present approach Prospects 210 210 211 213 Notes 215 References 227 Subject index Author index 233 238 Preface During the past twelve years or so I have been carrying out several linguistic and psycholinguistic studies on the relationship between cognitive and linguistic categories. The impetus for summarizing this work in book form came from Sidney Greenbaum. I then had to embark on the task of spelling out and developing the theoretical approach underlying my previous research work. In this venture I was supported by many people who gave me an opportunity to discuss my ideas with them. In particular I would like to mention Professor Greenbaum, who read the whole draft in instalments and saved me from at least the worst blunders, and Professor Richard Hudson and two of his doctoral students, And Rosta and Nik Gisborne, with whom I met regularly during several months on my sabbatical in 1992. Richard Hudson also read most of the chapters, and I owe much to his criticism and insightful suggestions. Much of what is good (I hope) in this book is due to them, and I cannot thank them enough for their interest and support. The studies were conducted with the help of many research assistants. Some of them were not merely helping with the technical side but were acting more in the nature of collaborators, participating in planning and taking on responsibility for data collection and analysis: Neta Bargai, Laura Canetti, Alon Halter, Dalia Kelly, Neta Ofer, Liat Ozer, Ruth Pat-Horenczyck, Anat Rappoport-Moscovich, and Smadar Sapir. If these studies had been published in journals, my assistants would have figured there as co-authors. Anat Ninio and Naomi Goldblum each read and commented on most of the chapters in draft form, and I am very much indebted to them for their labors. I also thank Moshe Anisfeld, Edit Doron, Eyal Gamliel, Ainat Guberman, Yonata Levy, Anita Mittwoch, Ruth Ostrin, Rita Watson, Vlad Zegarac, and Yael Ziv for their comments on various parts of the manuscript; and Benny Shanon and Samuel Shye for valuable suggestions. I am grateful to Annie Cerasi, who made a large number of very helpful suggestions regarding presentation of the material and checked the entire manuscript for errors in cross-references and infelicities of style. Some of the research reported in this volume was begun while I was a Fellow at the Institute for Advanced Studies, The Hebrew University. xiv Preface Support for some of the research reported here was made available by the Basic Research Foundation of The Israel Academy of Sciences and Humanities and in part by the Human Development Center, The Hebrew University. My debt to these institutions is gratefully acknowledged. My greatest debt is to my wife, who provided the ambience and physical conditions conducive to my work, often at the cost of considerable inconvenience to herself. To her this book is dedicated. Introduction This book deals with one aspect of the perennial problem of the relation between language and cognition. Language has rules stating how meanings are expressed by linguistic constructions; or, put differently, grammar describes the (often complex and indirect) mappings from cognitive space into syntactic structures. In formulating these mappings, linguists resort, inter alia, to a construct that goes by various names: case, semantic relation or role, thematic role, theta role. In this book I use the term case, which has the advantage of brevity. These days, the usefulness of such a construct is being questioned (Dowty, 1988; Ravin, 1990:112). My analysis of various phenomena of English syntax shows that cases can do a lot of explanatory work if they are conceived of in a different way, namely, not as categories in cognitive space, but rather as linguistic constructs that are defined partly in terms of cognitive concepts. There must be a level between the cognitive structure and the linguistic expression; one may call this level semantic, but I will usually avoid this term, which has come to mean several disparate things. In developing this conception it became clear that not only are the choices of the speaker limited by the resources of language, but the resources of the language spoken may determine the way the message is conceived of by the hearer. Some readers might be put off by such a Whorfian heresy, and I therefore hasten to remark that, as will be shown in due course, this is only a communicative effect of language; it does not imply anything regarding the classic Whorfian thesis of a more wide-ranging effect of language structure on cognition that goes beyond the communicative situation. The concern of this book is the analysis of noun phrases that are complements of the main verb. The proposals made are largely neutral vis-avis currently contending theories of grammar. The motivation for the linguistic work reported in this book came mainly from my interest in the theory of native language acquisition. It is difficult, if not impossible, to explain how syntactic functions and their correspondences with semantic relations might be acquired, unless one assumes that the child gains a hold on the former via the latter (Schlesinger, 1982, 1988). On this assumption, language learning becomes simpler the larger the correspon- 2 Introduction dence between cognitive categories in terms of which the child conceives of the world around her and the linguistic categories she has to master. The plausibility of such a theory would therefore be vastly increased if it could be shown that syntactic categories, like the subject, are semantically relatively homogeneous (just as Jakobson, 1936/1971, has shown that Russian cases have certain abstract core meanings). Some studies I conducted - both linguistic and psycholinguistic - suggested that this may indeed be so (see also the discussion in Pesetsky, 1990, on the issue of linguistic ontogenesis). In developing the case system presented here, the homogeneity of syntactic categories served as a working hypothesis. Semantic homogeneity also makes sense when one considers the phylogeny of language. One might assume (naively, no doubt) that to facilitate the use of language, the syntactic—semantic mapping would be maximally simple and straightforward. Now, there are many factors, historical and otherwise, that preclude such a one-to-one mapping of semantic and syntactic categories, but a theory that provides for more semantic homogeneity than others has an advantage over them. The issue of homogeneity has been broached here in order to explain why and how I went about developing the present approach; it is not adduced as support for this approach. In fact, in the course of my investigations it turned out that the homogeneity hypothesis is only partially true. The subject category, for instance, is shown in Chapter 2 to be much more homogeneous than is usually assumed (but see Chapter 6), whereas the direct object is much less so (Chapter 8). Throughout the book I have attempted to illustrate and support the linguistic argumentation with psycholinguistic studies. Many of these involve judgments of native speakers. Linguists usually rest content with referring to their own intuitions, supplemented perhaps by those of a few other people within easy reach. It has been found, however, that native speakers' judgments may differ from those of linguists more than they differ among themselves (Spencer, 1973; see also the studies by Quirk and Svartvik, 1966, and Greenbaum, 1973). It is necessary therefore to corroborate linguists' observations by data obtained from a larger sample of native informants (Schlesinger, 1977: 210—11). The first chapter explores the nature of categories in cognitive space. Its conclusions are a starting point for developing the concept of cases in Chapter 2, which is in some respects the central chapter of the book. It deals specifically with the Agent, or A-case, and its relation to the sentence subject, and Chapter 3 then applies the same approach to instrumentals and the Comitative. Chapter 4 is based on it and deals with the instrument in subject position. These two chapters are independent of the remaining chapters (and the reader who likes skipping can do so here). A new case category, which is not usually recognized as such by case grammarians, is introduced in Chapter 6, with Chapter 5 preparing the ground. This case is then resorted to in Introduction 3 Chapter 7 in dealing with the subjects of mental verbs. Chapter 8 is devoted to objects, and in particular to the direct object, which is shown not to be amenable to a treatment in terms of case categories. Chapter 9 deals with the prototypical structure of linguistic categories. In the final chapter, the results of the analyses in the various chapters are reviewed and some areas of further research discussed. Finally, let me point out in what ways this work is limited. All the analyses are of English sentences, with only desultory remarks made about other languages. Further, the units of analysis in this book are almost exclusively simple clauses. One respect in which the current approach will have to be elaborated is that of extending the analyses to complex clauses and sentences in a discourse setting. Cognitive space . . . the relations are numberless and no existing language is capable of doing justice to all their shades. William James (1892/1962: 176) 1. Semantic and cognitive categories This book deals with cases (also called semantic relations or thematic roles) and their relationship with cognition. It is often assumed that cases are conceptual or cognitive categories. Thus Wilkins (1988: 191-92) states that thematic roles " . . . are components of the mental representations of objects and concepts." This view seems to be shared by, inter alios, Nilsen (1973), DeLancey (1982), and Jackendoff(i983). It is important to be clear about what is being claimed here. On one interpretation, the equation of cases with cognitive categories means only that cases are anchored in cognition, which is tantamount to the truism that language maps meanings into sounds. This, however, is apparently not what Fillmore had in mind when he wrote: "The case notions comprise a set of universal, presumably innate, concepts which identify certain types of judgments human beings are capable of making about the events that are going on around them, judgments about such matters as who did it, who it happened to, and what got changed" (Fillmore, 1968: 24). Here the much more interesting claim is made that case categories exist in cognition independently of language, presumably also prior to language, and that the linguistic system then makes use of these independently existing categories. According to this view, there is a single cognitive-semantic level that is mapped somehow, directly or indirectly, into the level of formal syntactic constructions. One alternative would be to distinguish between a semantic level and a cognitive or conceptual one. Grammar, on this view, consists in a mapping from the cognitive level to the formal syntactic one via the semantic level. Cases belong to the semantic level, and they are of course defined in terms of cognitive categories, but they are not primitive cognitive concepts, as the previous view has it. How can a decision between these two rival conceptions be arrived at? It seems that a prerequisite for any intelligent debate about this issue is at least some general idea about the nature of categories in cognitive space. Some 2 Instrument and Accompaniment - rating studies 5 studies will be reported in this chapter which may serve to throw some light on the question of how people conceive of such case-like categories as Instrument and Agent. The term notions, rather than cases, will be used here for these case-like categories, and the term cases will be reserved for linguistic constructs that function in the grammar. The question whether the latter are primitive, universal categories in cognitive space, and their relationship with notions will be taken up again in the final section of this chapter. The studies reported in this chapter address the following questions: (i) Are the notions in cognitive space well-defined, mutually exclusive categories? (ii) Are they homogeneous categories, i.e., is membership all-or-none or is it graded? (iii) Are they mutually exclusive categories? 2. Instrument and Accompaniment - rating studies The notions dealt with in the studies reported in this section are Instrument and Accompaniment. Both can be expressed, in English, by the same prepositional phrase: (1) He opened the crate with a crowbar. (Instrument) He opened the crate with his friend. (Accompaniment) What is the nature of these notions? Suppose, for the sake of the argument, that they are identical to the cases in a grammar. Then it would be convenient for stating linguistic regularities if notions turned out to form clearly delineated categories. But a few examples suffice to show that this is not so. The noun phrases in some z^/YA-phrases are not classifiable as either Instrument or Accompaniment, but lie somehow halfway between these two notions; for instance: (2) The The The The pantomimist gave a show with the clown. general captured the hill with the soldiers. prisoner won the appeal with a skilled lawyer. hoodlum broke the window with a stone. In the first sentence, the noun in the H?^-phrase expresses Accompaniment, whereas in the last one it expresses the Instrument; the other two sentences, however, intuitively seem to lie in between these clear-cut examples. Now, a linguist's intuitions ought to be backed up by psycholinguistic studies on the judgments of native speakers under controlled conditions (Schlesinger, 1977: 210-n). There is evidence that judgments are sensitive to contextual effects (Greenbaum, 1973), and that native speakers' judg- 6 Cognitive space ments may differ from those of linguists more than they differ among themselves (Spencer, 1973). The above intuitions were therefore tested by obtaining judgments on sentences like (2) from a larger group of native speakers.1 But first a clarification is in order. Let us assume for the moment that the foregoing characterization of (2) is indeed correct; then there are two possibilities of describing the relation between Accompaniment and Instrument. One is that these are graded categories with fuzzy boundaries, that is the two notions lie on a continuum with no clear dividing line between them. In recent years such fuzzy categories have become increasingly recognized in linguistic theory (see, e.g., Ross, 1972b; Keenan, 1976). The other possibility is that these categories are partially overlapping: Rather than lying on the boundary line between Accompaniment and Instrument, the soldiers may belong to both categories, and so may a skilled lawyer. These alternatives require some elucidation, since they involve a distinction between three properties of categories: gradedness, fuzziness of boundaries, and partial overlap. While these properties often co-occur, they are logically independent of each other. Thus, a given category may have members varying in degree of membership - that is, the category is graded whereas the boundaries of the category are sharp and clearly defined (for instance, the category "low income group" is graded but may be defined as "having a monthly income below a certain sum"). Again, the boundary between two categories may be fuzzy (e.g., hirsute and bald) without the categories overlapping even partially (no one can be both hairy and bald). 2.1 First rating study — procedures In a study carried out in collaboration with Ruth Pat-Horenczyck, sentences like those in (2) were presented to a group of native speakers of English, who were asked to indicate to which extent the zz^M-phrase in each sentence was an instance of the notions Accompaniment and Instrument. The sentences were those listed in Table 1.1, below. Two rating scales were prepared, one for the notion Accompaniment - the A-scale — and one for the notion Instrument - the I-scale. Instructions for the A-scale were as follows: The following sentences each contain a with-phrast. With has several meanings. Among others it can mean ACCOMPANIMENT, as in "He went to the movies with his friend." Please read the following sentences carefully and check for each sentence to which extent with has the meaning ACCOMPANIMENT, using one of the eight spaces between "yes, definitely" and "no, definitely". Please make sure to make only one check mark for each sentence. Please do not turn over until asked to do so. For the I-scale, "INSTRUMENT or MEANS" replaced "ACCOMPANIMENT" in these instructions, and the example given was: He wrote the note with a pencil