Editorial page
Title page
Copyright page
Table of contents
Preface
PART I. TOWARD A COHERENT AND COMPREHENSIVE LINGUISTIC THEORY
Meaning and Semantic Structure
The Nature of a Grammar
Grammar as a Symbolic Phenomenon
Distribution and Predictability
Arbitrary Grammatical Markings
Grammaticality Judgments
Conclusion
A View of Linguistic Semantics
Cognitive Domains
Profile and Base
Imagery
Scale and Scope of Predication
Relative Salience of Substructures
Perspective
Conclusion
The Nature of Grammatical Valence
The Problem
Grammatical Valence: Canonical Instances
Non-Canonical Valence Relations
Further Departures from Canonical Valence Relations
Scope and Morphological Layering
A Usage-Based Model
The Network Conception
General Applicability
Distribution
Conclusion
PART II. ASPECTS OF A MULTIFACETED RESEARCH PROGRAM
1. The Nature of Grammatically Specified Notions
2. Categories of Grammatically Specified Notions
2.2 Plexity
2.3 State of Boundedness
2.4 State of Dividedness
2.5 The Disposition of a Quantity: A System of Categories
2.6 Degree of Extension
2.7 Pattern of Distribution
2.8 Axiality.
2.9 Perspectival Mode
2.10 Level of Synthesis
2.11 Level of Exemplarity
2.12 Other Categories and Processes
2.13 Nesting: An Interaction of Categories
2.14 Four Imaging Systems
3. Further Cognitive Connections
Notes
Where Does Prototypicality Come From?
2. A Case Study in Synonymics
3. Usage as Evidence for Prototypicality
4. Introspective Evidence of Prototypicality
5. The Functional Explanation of Prototypicality
6. Onomasiological and Semasiological Aspects of Cognitive Semantics
Appendix. A translation of the quotations.
The Natural Category MEDIUM: An Alternative to Selection Restrictionsand Similar Constructs
2. Co-occurrence and Categorization
2.1.2 Ascription Features
2.1.3 Transfer Features
2.2 Toward a Viable Theory of Categorization in the Lexicon
3. The Natural Category MEDIUM
3.2 The Categorial Periphery of MEDIUM
4. Conclusions and Inconclusions
Notes
Spatial Expressions and the Plasticity of Meaning
2. Insufficiencies of the \
2.2. Geometric Conceptualization
2.3. Synecdoche
2.4. Divergence from the Simple Relations
2.5. Added Constraints
2.6. Unexplained Context Dependencies
2.7. Unexplained Restrictions
3. Meaning and Use of Spatial Prepositions
3.2. Geometric Descriptions
3.3. Pragmatic Near-Principles
3.4. Use Types
4. Discussion
Notes
Contrasting Prepositional Categories: English and Italian
1. Prepositions and Polysemy
2. The Nature of Prepositional Meaning
3. The Prepositions
4. Conclusions
Notes
The Mapping of Elements of Cognitive Space onto Grammatical Relations: An Example from Russian Verbal Prefixation
Goto 337 /FitH 555.1 The Role of the Prefix and How It Is Captioned
1.0 The pere-Network
1.1 Configuration 1
1.2 Configuration 2
1.3 Configuration 3, <Interchange>
1.4 Configuration 4,  <Division>
1.5 Configuration 5,  <Thorough>
1.6 Configuration 6,  <Over>
1.7 Configuration 7,  <Bend>
1.8 Configuration 8,  <Turn>
1.9 Configuration 9,  <Mix>
2.0 The Cognitive Model in Semantic Description: A Summary
3.0 Mapping the Cognitive Configuration onto the Verbal Arguments
Notes
Conventionalization of Cora Locationals
2.0 Conventional Morphology
3.0 Choice of Verbal Prefixes
3.1 The Tree Trunk
3.2 All Burned Up
3.3 Down to the Ground
3.4 Bottles, Sticks and Pots
3.5 On the Edge of Things
3.6 High in the Sky
3.7 Comparative Anatomy
3.8 A Trip into Time
4.0 Conclusion
Notes
The Conceptualisation of Vertical Space in English: The Case of Tall
2. English Adjectives Denoting Vertical Space
2.2 High and Low
2.3 Tall
3. Collocational Preferences of Tall
4. The Status of Collocation
5. The Concept of TALLNESS
Notes
Length, Width, and Potential Passing
2. The Distribution of Length and Width
2.2 The Length and Width of Mobile Entities
2.3 The Length and Width of Immobile Multidimensional Entities
2.4 Metric Definition of Length/Width
2.5 Constraints
3. Synchronic Interpretation of the Distribution of Length and Width
3.2 Length/Width and Potential Passing
4. Interpretation of Length/Width in Logical Diachrony
5. Discrepancies between Logical and Historical Time. Evidence fromFrench
6. Conclusion
Notes
On Bounding in Ik
2. Change, Boundedness and Directionality
2.2 Directionals with Stative Verbs
2.3 Directionals with Other Verbs
3. Deictic Center with Other than Motion Verbs
4. Summary and Perspective
Notes
A Discourse Perspective on Tense and Aspect in Standard Modern Greek and English
1. Tense and Aspect in SMG: Establishing the Relevant Categories
2. The Semantics of Perfective and Imperfective
3.1 The Psychosemantic Basis of Aspect
3.3 The Iconicity Principle, Aspect and Tense
4.1 Packaging and Grinding with a Discourse Purpose in a Narrative
4.3 Tense and Discourse Representations
4.4 Aspectual Distinctions of English: Simple Tense versus Progressive Aspect
4.5 Progressives and Imperfectives
4.6 Packaging and Grinding in an English Text and Its SMG Translation
5. Conclusion
Appendix 1
Appendix 2
Notes
Semantic Extensions into the Domain of Verbal Communication
2. Underlying Concepts and Assumptions
3. Verbal Communication: Its Multi-Level Components and Its Donor Domains
4. Spatial Motion
4.2. The Components of Motion Events and Their Linguistic Expression
5. Extensions of Spatial Motion
5.2. Phrasal Verbs
5.2.2. Including a SURFACE Landmark
5.2.3. Including a POINT Landmark
5.2.4. Including a CHANNEL Landmark
5.2.5. Including an Agentive Co-Trajector or an Instrumental Trajector
5.3. Non-Phrasal Verbs
5.4. Major Patterns of Cross-Domain Correspondences
5.5. Domain Intersections
6. Conclusion
Notes
Spatial Metaphor in German Causative Constructions
2. Causative vs. Non-causative
3. Prototypical Causation
4. Events, Agents, Patients, and Instruments
5. Foregrounding and Syntactic Marking
6. Pedagogic Grammars vs. Actual Usage
7. A Survey of Prepositional Usage in German Causative Constructions
8. Prototype Combinations
9. Rare or Difficult Combinations
10. Non-prototypical Combinations
11. Causatives vs. Non-causative Transitives
12. Spatial Metaphor in Case Grammar
13. A Localist Basis for von, mit, and durch
14. A Cognitive-Spatial Account of German Causative Constructions
15. Conclusion
Náhuatl Causative/Applicatives in Cognitive Grammar
1. Preliminaries
1.2 Causation
2. Causatives
3. Applicatives
4. The Causative/Applicative Schema
5. Cases Intermediate between Causatives and Applicatives
5.2 Multiple and Mixed Construals
6. Verbalizing Usages
6.2. Verbalizations of Nouns
7. Summary and Conclusion
Notes
PART III. A HISTORICAL PERSPECTIVE
1. Language and Cognition: Aristotle's Approach
2. Mediaeval Speculative Grammar
3. Conclusion
Notes
Cognitive Grammar and the History of Lexical Semantics
2. Historical-Philological Semantics
3. Cognitive Semantics
4. Comparing the Cognitive and the Historical-Philological Tradition
5. A Survey of the History of Lexical Semantics
6. Main Lines in the History of Lexical Semantics
7. Conclusion
Notes
REFERENCES
Subject Index
The series Current Issues in Linguistic Theory

Author: Rudzka-Ostyn Brygida  

Tags: linguistics   english language  

ISBN: 0304-0763

Year: 1988

Text
                    Cur nt Issues in Linguisti Theor o
-L
EDITED BY
Br> n'dj Rud ka- si n
JOH BEN) INS PUBLISHI C CO P NY


TOPICS IN COGNITIVE LINGUISTICS
AMSTERDAM STUDIES IN THE THEORY AND HISTORY OF LINGUISTIC SCIENCE General Editor E.F. KONRAD KOERNER (University of Ottawa) Series IV - CURRENT ISSUES IN LINGUISTIC THEORY Advisory Editorial Board Henning Andersen (Buffalo, N.Y.); Raimo Anttila (Los Angeles) Thomas V. Gamkrelidze (Tbilisi); Hans-Heinrich Lieb (Berlin) J. Peter Maher (Chicago); Ernst Pulgram (Ann Arbor, Mich.) E.Wyn Roberts (Vancouver, B.C.); Danny Steinberg (Tokyo) Volume 50 Brygida Rudzka-Ostyn (ed.) TOPICS IN COGNITIVE LINGUISTICS
TOPICS IN COGNITIVE LINGUISTICS Edited by BRYGIDA RUDZKA-OSTYN University of Leuven JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA 1988
Library of Congress Cataloging-in-Publication Data Topics in cognitive linguistics / edited by Brygida Rudzka-Ostyn p. cm. — (Amsterdam studies in the theory and history of linguistic science. Series IV, Current issues in linguistic theory, ISSN 0304-0763; v. 50) Bibliography: p. Includes index. 1. Cognitive grammar. I. Rudzka-Ostyn, Brygida. II. Series. P165.T65 1988 415--de 19 87-37495 ISBN 90 272 3544 9 (alk. paper) CIP © Copyright 1988 - All rights reserved No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission of the copyright holders. Please direct all enquiries to the publishers.
To the memory of my mother
Contents Preface ix I. Toward a Coherent and Comprehensive Linguistic Theory An Overview of Cognitive Grammar Ronald W. Langacker A View of Linguistic Semantics Ronald W. Langacker The Nature of Grammatical Valence Ronald W. Langacker A Usage-Based Model Ronald W. Langacker II. Aspects of a Multifaceted Research Program The Relation of Grammar to Cognition Leonard Talmy Where Does Prototypicality Come From? Dirk Geeraerts The Natural Category MEDIUM: An Alternative to Selection Restrictions and Similar Constructs 231 Bruce W. Hawkins Spatial Expressions and the Plasticity of Meaning 271 Annette Herskovits Contrasting Prepositional Categories: English and Italian 299 John R. Taylor The Mapping of Elements of Cognitive Space onto Grammatical Relations: An Example from Russian Verbal Prefixation 327 Laura A. Janda 3 49 91 127 165 207
Vlll CONTENTS Conventionalization of Cora Locationals 345 Eugene H. Casad The Conceptualisation of Vertical Space in English: The Case of Tall 379 Rene Dirven and John R. Taylor Length, Width, and Potential Passing 403 Claude Vandeloise On Bounding in Ik 429 Fritz Serzisko A Discourse Perspective on Tense and Aspect in Standard Modern Greek and English 447 Wolf Paprotte Semantic Extensions into the Domain of Verbal Communication 507 Brygida Rudzka-Ostyn Spatial Metaphor in German Causative Constructions 555 Robert Thomas King Nahuatl Causative/Applicatives in Cognitive Grammar 587 David Tuggy III. A Historical Perspective Grammatical Categories and Human Conceptualization: Aristotle and the Modistae 621 Pierre Swiggers Cognitive Grammar and the History of Lexical Semantics 647 Dirk Geeraerts References 679 Subject Index 695
Preface Recent years have witnessed a growing interest in cognitive linguistics, a framework aiming at an adequate account of the relationship between language and cognition and as such involving human psychology, interpersonal relations, culture, and a host of other domains. Our purpose has been to present the theoretical premises of this framework and to explore its descriptive and explanatory potential with respect to a wide range of language phenomena. In pursuing this goal, we have frequently relied on corpus analyses, intensive field work, and other means of empirical verification. Crossing the boundaries of particular languages or language families has lent additional support to the findings emerging from our research. If one had to name a key notion of cognitive grammar, one would certainly point to the dependence of linguistic structure on conceptualization as well as the conceptualized perspective. Within this framework, meanings are defined relative to conceptual domains, particular linguistic choices are often found to hinge upon the vantage point from which a given situation is viewed, and category boundaries are seen as fluctuating and dependent on, among other things, the conceptualized experience or purpose. This relativism extends to the very structure of the framework. The reader will soon discover that not all contributors to the volume draw the same distinctions.Neither do they use identical descriptive tools. Their choice and nature vary with the purpose as well as perspective adopted. This inherent flexibility renders the framework exceptionally receptive to findings in other disciplines. Only some of the current interdisciplinary crossovers could be signalled here; but they suffice to show the enormous potential of cognitive grammar in capturing various facets of language. As language is such a complex phenomenon, it can be adequately described only when approached from different angles. By relativizing its own methodology, cognitive linguistics can accommodate these different angles readily and naturally.
X PREFACE To place cognitive grammar against a broader background, we have explored some earlier linguistic theories. While adding another dimension to our research, this exploration has unveiled interesting links between present and past attempts at grasping the relation of language to cognition. The project has materialized thanks to the help of several people. I am indebted to Rene Dirven who came up with the idea and suggested to me the role of editor. The conferences organized by him, first at the University of Trier and now in Duisburg, have allowed many of us to come together and discuss topics of common interest; and for this wonderful forum we remain grateful. As editor, I wish to thank the authors for their contributions, but also for their good humor and the spirit of cooperation. Much of the present volume is the fruit of an intensive exchange of ideas and materials, and also numerous revisions. In this context, a special word of thanks is due to Rene Dirven, Dirk Geeraerts, Bruce Hawkins and Pierre Swiggers, all of whom offered to act as referees on several occasions. To Peter Kelly, I owe a particular debt of gratitude for sharing with me his native-speaker intuition. Finally, I must record my great appreciation for my husband's encouragement and help at all stages of the project. Brygida Rudzka-Ostyn
PARTI TOWARD A COHERENT AND COMPREHENSIVE LINGUISTIC THEORY
An Overview of Cognitive Grammar Ronald W. Langacker University of California, San Diego Orientation Cognitive grammar (formerly "space grammar") is a theory of linguistic structure that I have been developing and articulating since 1976. Though neither finished nor formalized, it has achieved a substantial measure of internal coherence and is being applied to an expanding array of languages and grammatical phenomena (see Langacker 1981, 1982a , 1984, 1985, in press; Casad 1982a; Casad and Langacker 1985; Hawkins 1984; Lindner 1981, 1982; Smith 1985; Tuggy 1981; Vandeloise 1984, 1985a). These efforts have been prompted by the feeling tha* established theories fail to come to grips in any sensible way with the real problems of language structure, as they are based on interlocking sets of concepts, attitudes, and assumptions that misconstrue the nature of linguistic phenomena and thus actually hinder our understanding of them. It is therefore necessary to start anew and erect a theory on very different conceptual foundations. Cognitive grammar thus diverges quite radically from the mainstream of contemporary linguistic theory, particularly as represented in the generative tradition. The differences are not confined to matters of detail, but reach to the level of philosophy and organizing assumptions. I will succinctly sketch these differences as they pertain to the nature of linguistic investigation, the nature of a linguistic system, the nature of grammatical structure, and the nature of meaning. My presentation of the "orthodox" position is admittedly a caricature; I state it without the necessary qualifications for sake of brevity, and also to underscore the substantially different spirit of the two approaches. ® Ronald W. Langacker
4 RONALD W. LANGACKER With respect to the nature of linguistic investigation, orthodox theory holds that language (or at least grammar) is describable as an algorithmic system. Linguistics is thus a formal science akin to logic and certain branches of mathematics (e.g. automata theory). Of paramount importance is the construction of an all-embracing linguistic theory incorporating explanatory principles; ongoing description is considered most valuable when formulated in terms of current theory and directed towards the testing and refinement of its predictions. Discrete categories and absolute principles are sought, on the grounds that a theory should be maximally restrictive and make the strongest possible claims. Moreover, economy is a prime concern in formulating the grammar of a language: redundancy of statement implies the loss of significant generalizations. The cognitive grammar "heresy" sees biology as providing a better metaphor for linguistic research than the formal sciences. While certain aspects of language may be discrete and "algebraic", in general a language is more accurately likened to a biological organism; our expectations concerning the nature of revealing analysis and viable description must be adjusted accordingly. For instance, absolute predictability is normally an unrealistic expectation for natural language: much is a matter of degree, and the role of convention is substantial. Considerations of economy must cede priority to psychological accuracy; redundancy is plausibly expected in the cognitive representation of linguistic structure, and does not in principle conflict with the capturing of significant generalizations. Further, linguistic theory should emerge organically from a solid descriptive foundation. Preoccupation with theory may be deleterious if premature, for it stifles the investigation of non-conforming phenomena and prevents them from being understood in their own terms. In the orthodox view, the grammar of a language consists of a number of distinct "components". The grammar is conceived as a "generative" device which provides a fully explicit enumeration of "all and only the grammatical sentences" of the language. The linguistic system is self-contained, and hence describable without essential reference to broader cognitive concerns. Language may represent a separate "module" of psychological structure. Cognitive grammar views the linguistic system in a very different fashion. It assumes that language evokes other cognitive systems and must be described as an integral facet of overall psychological organization. The grammar of a language is non-generative and non-constructive, for the
AN OVERVIEW OF COGNITIVE GRAMMAR 5 expressions of a language do not constitute a well-defined, algorithmically- computable set. The grammar of a language simply provides its speakers with an inventory of symbolic resources — using these resources to construct and evaluate appropriate expressions is something that speakers do (not grammars) by virtue of their general categorizing and problem-solving abilities. Only semantic, phonological, and symbolic units are posited, and the division of symbolic units into separate components is considered arbitrary. Orthodox theory treats grammar (and syntax in particular) as an independent level or dimension of linguistic structure. Grammar (or at least syntax) is considered distinct from both lexicon and semantics, and describe able as an autonomous system. The independence of grammatical structure is argued by claiming that grammatical categories are based on formal rather than semantic properties. Speakers are capable of ignoring meaning and making discrete well-formedness judgments based on grammatical structure alone. By contrast, cognitive grammar claims that grammar is intrinsically symbolic, having no independent existence apart from semantic and phonological structure. Grammar is describable by means of symbolic units alone, with lexicon, morphology, and syntax forming a continuum of symbolic structures. Basic grammatical categories (e.g. noun and verb) are semantically definable, and the unpredictable membership of other classes (those defined by occurrence in particular morphological or syntactic constructions) does not itself establish the independence of grammatical structure. Well-formedness judgments are often matters of degree, and reflect the subtle interplay of semantic and contextual factors. Finally, it is commonplace to reject a "conceptual" or "ideational" theory of meaning as being untenable for the scientific investigation of language. It is assumed instead that the meanings of linguistic expressions are describable in terms of truth conditions, and that some type of formal logic is appropriate for natural language. It is held that a principled distinction can be made between semantics and pragmatics (or between linguistic and extralinguistic knowledge), that semantic structure is fully compositional, and that such phenomena as metaphor and semantic extension lie outside the scope of linguistic description. In the cognitive grammar heresy, meaning is equated with conceptualization (interpreted quite broadly), to be explicated in terms of cognitive processing. Formal logic is held to be inadequate for the description of
6 RONALD W. LANGACKER semantic structure, which is subjective in nature and incorporates conventional "imagery" — defined as alternate ways of construing or mentally portraying a conceived situation. Linguistic semantics is properly considered encyclopedic in scope: the distinction between semantics and pragmatics is arbitrary. Semantic structure is only partially compositional, and phenomena like metaphor and semantic extension are central to the proper analysis of lexicon and grammar. Meaning and Semantic Structure An "objectivist" view of meaning has long been predominant in semantic theory. Rigorous analysis, it is maintained, cannot be based on anything so mysterious and inaccessible as "concepts" or "ideas"; instead, the meaning of an expression is equated with the set of conditions under which it is true, and some type of formal logic is deemed appropriate for the description of natural language semantics. Without denying its accomplishments, I believe the objectivist program to be inherently limited and misguided in fundamental respects: standard objections to the ideational view are spurious, and a formal semantics based on truth conditions is attainable only by arbitrarily excluding from its domain numerous aspects of meaning that are of critical linguistic significance (cf. Chafe 1970: 73-75; Langacker in press: Part II; Hudson 1984). Cognitive grammar explicitly equates meaning with "conceptualization" (or "mental experience"), this term being interpreted quite broadly. It is meant to include not just fixed concepts, but also novel conceptions and experiences, even as they occur. It includes not just abstract, "intellectual" conceptions, but also such phenomena as sensory, emotive, and kinesthetic sensations. It further embraces a person's awareness of the physical, social, and linguistic context of speech events. There is nothing inherently mysterious about conceptualization: it is simply cognitive processing (neurological activity). Entertaining a particular conceptualization, or having a certain mental experience, resides in the occurrence of some complex "cognitive event" (reducing ultimately to the coordinated firing of neurons). An established concept is simply a "cognitive routine", i.e. a cognitive event (or event type) sufficiently well "entrenched" to be elicited as an integral whole. Cognitive grammar embraces a "subjectivist" view of meaning. The semantic value of an expression does not reside solely in the inherent prop-
AN OVERVIEW OF COGNITIVE GRAMMAR 7 erties of the entity or situation it describes, but crucially involves as well the way we choose to think about this entity or situation and mentally portray it. Expressions that are true under the same conditions, or which have the same reference or extension, often contrast in meaning nonetheless by virtue of representing alternate ways of mentally construing the same objective circumstances. I would argue, for example, that each pair of sentences in (1) embodies a semantic contrast that a viable linguistic analysis cannot ignore: (1) (a) This is a triangle. (a') This is a three-sided polygon. (b) The glass is half-empty. (b>) The glass is half-full. (c) This roof slopes upward, (c') This roof slopes downward. (d) Louise resembles Rebecca, (d') Rebecca resembles Louise. (e) Russia invaded Afghanistan. (e') Afghanistan was invaded by Russia. (f) I mailed a package to Bill. (f) I mailed Bill a package. I use the term "imagery" to indicate our ability to mentally construe a conceived situation in alternate ways (hence the term does not refer specifically or exclusively to sensory or visual imagery (cf. Kosslyn 1980; Block 1981)). A pivotal claim of cognitive grammar is that linguistic expressions and grammatical constructions embody conventional imagery, which constitutes an essential aspect of their semantic value. In choosing a particular expression or construction, a speaker construes the conceived situation in a certain way, i.e. he selects one particular image (from a range of alternatives) to structure its conceptual content for expressive purposes. Despite the objective equivalence of the sentence pairs in (1), the members of each are semantically distinct because they impose contrasting images on the conceived situation. At this point, I will confine my discussion of imagery to a single example, namely the semantic contrasts distinguishing the universal quantifiers of English. It is intuitively obvious that the sentences in (2) have subtly different meanings despite their truth-functional equivalence:
8 RONALD W. LANGACKER (2) (a) All cats are playful. (b) Any cat is playful. (c) Every cat is playful. (d) Each cat is playful. These sentences share the conceptual content of a property (playfulness) being attributed to all the members of a class (the set of cats). They nevertheless employ distinct images with respect to how one "reaches" or "mentally accesses" the class members for this purpose. An informal sketch of the contrasts is attempted in Fig. 1, where circles indicate class members, a box represents the class as a whole, and heavy lines enclose the entity designated by the quantifier. All refers to the class collectively, as an undifferentiated mass, and makes a blanket attribution of the property to its members. The other three quantifiers each refer to a single, arbitrary class member, but this member is conceived as being selected in such a fashion that the property attributed to it is similarly attributed to all the other members. The image conveyed by any is one of random selection: if one chooses a member at random, it will invariably display the property in question. Every and each are alike (and contrast with all) in attributing the property to the full set of class members on an individual (rather than a collective) basis. The difference between them is that each further suggests that the members are examined sequentially — one at a time — for this purpose. (a) ALL ANY l-ooo-l c) EVERY |-o t 1 j ??'" 1 ♦ 1 1 1 1 1 1 (d) EACH |-o < f J 1 1 1 * 1 1 1 ??•■■ 1 A i I 1 1 1 1 Figure 1 This analysis is at best preliminary and suggestive, but the putative imagic contrasts account for a variety of otherwise puzzling phenomena. First, they explicate the clear differences in meaning among the sentences in (3) (first cited by George Lakoff): (3) (a) Tonight you can see any star in the Milky Way. (b) Tonight you can see every star in the Milky Way. (c) Tonight you can see each star in the Milky Way.
AN OVERVIEW OF COGNITIVE GRAMMAR 9 Speakers readily agree that (3)(a) describes the ability to see a single star: if a particular star is in the Milky Way, tonight you can see it, whichever one you choose to look at. In (3)(b), the stars of the Milky Way are viewed simultaneously, but nevertheless stand out as individuals. This individuation is even stronger in (3)(c), which evokes the image of the viewer shifting his gaze across the sky, looking first at one star, then at another, and so on. The analysis further explains the grammatical behavior of the quantifiers, for instance the ability of any, every, and each (but not all) to occur with a singular count noun or the pro form one: (4) (a) any star (b) every star (c) each star (d) * all star (5) (a) any one of those stars (b) every one of those stars (c) each one of those stars (d) * all one of those stars Moreover, because they specify individual attribution of the property to multiple class members, every and each construe the class with a greater degree of individuation than either all (which views the class collectively) or any (which specifically invokes only a single, randomly selected member); the former two quantifiers specifically highlight the status of the class as an aggregate of distinct individuals. It is natural, then, that mass nouns (including plurals) should occur with only all and any, since the hallmark of a mass is internal homogeneity: (6) (a) all water (b) any water (c) * every water (d) *each water (7) (a) all cats (b) any cats (c) * every cats (d) * each cats The analysis also explains the judgments in (8) and (9). (8) (a) He examined each one in turn, (b) He examined every one in turn.
10 RONALD W. LANGACKER (c) ?He examined all in turn. (d) *He examined any one in turn. (9) (a) All cats have something in common. (b) ?Every cat has something in common. (c) *?Each cat has something in common. (d) * Any cat has something in common. In turn specifies tliat the examination described in (8) affects class members individually and sequentially. This is fully compatible with each and every (and reinforces something suggested by the former). Sentence (8)(c) is a bit peculiar, since the collective construal imposed by all is at cross-purposes with the individuating force of the adverbial, i.e. there is a certain amount of tension between the images evoked by different components of the expression. In (8)(d) this conflict in imagery is even more pronounced, since any specifically picks out only a single member of the class for attribution of the property, while in turn implies the participation of multiple individuals. The semantic effect of have something in common runs counter to that of in turn, for the notion of commonality requires the simultaneous conception of class members. The well-formedness judgments in (9) are thus essentially the inverse of those in (8). The collective character of all renders (9)(a) unproblematic, while the individuation suggested by every and each accounts for the relative infelicity of (9)(b) and (c), greater (as expected) in the case of each. However the sentence with any is once again the least acceptable, and for the same reason as before: only a single class member is specifically selected (at random) for attribution of the property, but have something in common (like in turn) implies access to multiple members. The notion of sequential examination may further account for the fact that each is non-generic, i.e. a sentence like (2)(d) is only used with respect to some restricted subset of class members identifiable from the context (not the set of cats as a whole). The reason, perhaps, is that sequential examination is not easily conceived as providing mental access to all members of an open-ended, indefinitely expandable class. This is one example of how the lexical and grammatical resources of a language embody conventional imagery, which is an inherent and essential aspect of their semantic value. Two expressions may be functionally equivalent and serve as approximate paraphrases or translations of one another, and yet be semantically distinct by virtue of the contrasting images they
AN OVERVIEW OF COGNITIVE GRAMMAR 11 incorporate. The imagery is conventional because the symbolic elements available to speakers are language-specific and differ unpredictably — languages simply say things in different ways, even when comparable thoughts are expressed, as exemplified by the sentences in (10). (10) (a) I'm cold. (b) Pai froid. (c) Mir ist kalt. In this respect the cognitive grammar view of linguistic semantics is Whorfian or relativistic. However it is not Whorfian in the sense that this imagery is taken as imposing powerful or even significant constraints on how or what we are able to think. I take its influence to be fairly superficial, a matter of how we package our thoughts for expressive purposes. We shift from image to image with great facility, even within the confines of a single sentence, and freely create new ones when those suggested by linguistic convention do not satisfy our needs. The Nature of a Grammar The grammar of a language is characterized as a "structured inventory of conventional linguistic units". By "unit", I mean a thoroughly mastered structure (i.e. a cognitive routine). A unit may be highly complex internally, yet it is simple in the sense of constituting a prepackaged assembly that speakers can employ in essentially automatic fashion, without attending to the details of its composition. Examples of units include a lexical item, an established concept, or the ability to articulate a particular sound or sound sequence. The units comprising a grammar represent a speaker's grasp of linguistic convention. I do not limit the term "conventional" to structures that are arbitrary, unmotivated, or unpredictable: linguistic structures form a gradation with respect to their degree of motivation, and no coherence or special status attaches to the subclass of structures that are fully motivated or to those that are fully arbitrary; speakers operate with the entire spectrum of structures as an integrated system. Finally, this inventory of conventional units is "structured" in the sense that some units function as components (subroutines) of others. Only three basic types of units are posited: semantic, phonological, and symbolic. Symbolic units are "bipolar", consisting in the symbolic relationship between a semantic unit (its "semantic pole") and a phonolog-
12 RONALD W. LANGACKER ical unit (its "phonological pole"). I claim that lexicon, morphology, and syntax form a continuum of symbolic structures divided only arbitrarily into separate components of the grammar. Symbolic units are by no means a homogeneous class — they vary greatly along certain parameters (e.g. specificity, complexity, entrenchment, productivity, regularity) — but the types grade into one another, and do not fall naturally into discrete, non- overlapping blocks. Grammar (i.e. morphology and syntax) is accommodated solely by means of symbolic units. The intrinsically symbolic character of grammatical structure represents a fundamental claim whose import and viability will be examined in the remainder of this paper. Treating grammar as symbolic in nature (and not a separate level or autonomous facet of linguistic structure) enables us to adopt the highly restrictive "content requirement", which holds that no units are allowed in the grammar of a language apart from (i) semantic, phonological, and symbolic units that occur overtly; (ii) "schemas" for the structures in (i); and (iii) "categorizing relationships" involving the structures in (i) and (ii). For example, the phonological sequence [tip] occurs overtly, so it is permitted by (i). The syllable canon [CVC] is "schematic" for [tip] and numerous other phonological units (i.e. it is fully compatible with their specifications but characterized with a lesser degree of specificity), so (ii) allows its inclusion in the grammar. Permitted by (iii) is the categorizing relationship between the phonological schema [CVC] and a specific sequence like [tip] that "instantiates" or "elaborates" this schema. I will indicate this relationship in the following way: [[CVC] —> [tip]], where a solid arrow stands for the judgment that one structure is schematic for another, and square brackets enclose a structure with the status of a unit. Hence the categorizing relationship [[CVC] —> [tip]] is a complex unit containing the simpler units [CVC] and [tip] as substructures. The schema [CVC] defines a phonological category, and [[CVC] —> [tip]] characterizes [tip] as an instance of this category. The effect of the content requirement is to rule out many sorts of arbitrary descriptive devices routinely employed in other frameworks. For example, it rules out the use of "dummies" having neither semantic nor phonological content, invoked solely to drive the machinery of formal syntax. Also proscribed is any appeal to arbitrary diacritics or contentless features (more about this later). It further precludes the derivation of an overt structure from a hypothetical underlying structure of radically different character (e.g. the derivation of a passive from an underlying active). There
AN OVERVIEW OF COGNITIVE GRAMMAR 13 is a real sense in which the content requirement is far more constraining than the conditions, principles, and restrictions which continue to proliferate (and often evaporate) in the theoretical literature. If valid, it will help immeasurably to ensure the naturalness of linguistic descriptions. In describing the grammar of a language as an "inventory'' of conventional units, I refer to my conception of a grammar as being non-generative and non-constructive. Specifically rejected is the notion that a grammar can be regarded as an algorithmic device serving to generate (or give as output) "all and only the grammatical sentences of a language", at least if the description of a sentence is taken as including a full semantic representation. The reason for this rejection is that the full set of expressions is neither well-defined nor algorithmically computable unless one imposes arbitrary restrictions on the scope of linguistic description and makes gratuitous (and seemingly erroneous) assumptions about the nature of language. Treating the grammar as a generative, constructive device requires (i) the assumption that well-formedness is absolute, not a matter of degree; (ii) the exclusion of contextual meaning from the scope of semantic description; (iii) the omission of metaphor and figurative language from the coverage of a grammar; (iv) the supposition that semantic structure is fully compositional; and (v) the claim that a motivated distinction can be made between semantics and pragmatics (or between linguistic and "extra-linguistic" knowledge). These positions are commonly taken, but not because the facts of language use cry out for their adoption. Rather, they are adopted primarily for methodological reasons, there being no other way to make language appear to be a self-contained, algorithmically-describable system. I will simply comment that the convenience to the theorist of positing a self- contained formal system does not constitute a valid argument for its factual correctness. I opt for a cognitively and linguistically realistic conception of language over one that achieves formal neatness at the expense of drastically distorting and impoverishing its subject matter. Hence the grammar of a language is not conceived as a generative or algorithmic device, nor does it construct any expressions or give them as outputs — it is simply an inventory of symbolic resources. It is up to the language user to exploit these resources, and doing so is a matter of problem- solving activity involving categorizing judgments. The results of these operations draw upon the full body of a speaker's knowledge and cognitive ability, and are thus not algorithmically computable by any limited, self-contained system.
14 RONALD W. LANGACKER The basic scheme is sketched in Fig. 2(a). The language user brings many kinds of knowledge and abilities to bear on the task of constructing and understanding a linguistic expression; these include the conventional symbolic units provided by the grammar, general knowledge, knowledge of the immediate context, communicative objectives, esthetic judgments, and so on. Both the speaker and the addressee face the "coding" problem: given the resources at their disposal, they must successfully accommodate a "usage event". The semantic pole of the usage event is identified as the detailed conceptualization that constitutes the expression's full contextual value, i.e. how it is actually understood in context. The phonological pole of the usage event is the actual vocalization employed to symbolize this conception, in all its phonetic detail. Roughly, the speaker starts from the conceptualization and must arrive at the proper vocalization, while the addressee proceeds in the opposite direction (encoding vs. decoding). However both of them, by employing the varied resources at hand, must deal with the full, bipolar usage event for its occurrence to amount to a meaningful symbolic experience. Doing so necessarily involves categorization and problem-solving more generally. (a) symbolic units (G) general knowledge knowledge of context communicative objectives esthetic judgment etc. CODING problem-solving categorization RESOURCES USAGE EVENT (o) other symbolic j_ units L^kh (inactive) SU GRAMMAR ^r^ )~ (COMPOSITIONAL VALUE) Figure 2 I USAGE \ ' **1 EVENT I ACTUAL CONTEXTUAL VALUE One facet of this coding operation, namely the relation between the grammar and the usage event, is diagrammed in Fig. 2(b). In arriving at the usage event, and evaluating it with respect to linguistic convention, the
AN OVERVIEW OF COGNITIVE GRAMMAR 15 speaker or addressee must select a particular array of symbolic units (SU) and activate them for this purpose; I take it that the cognitive reality of linguistic units (and their activation in language use) is self-evident and uncon- troversial. Moreover, since the usage event is identified as the utterance itself and how it is actually understood in context, it is obviously real — by definition the usage event does in fact occur. The only facet of Fig. 2(b) whose status is uncertain is the intervening structure, representing the "compositional value" of the expression, i.e. the value that could, in principle, be obtained by algorithmic computation based solely on the conventional values of the symbolic units employed. It is not unlikely that the language user computes this compositional value as one step in the coding process leading to the full usage event. Whether he does so consistently or only on certain occasions is something we presently have no way of knowing. What is, however, clear is that the usage event is not in general equivalent to the compositional value. It is virtually always more specific than anything strictly predictable from established symbolic units, and thus represents an elaboration or "specialization" of the compositional value; or else it conflicts in some way with the compositional value and thus constitutes an "extension" from it. This departure of the usage event from the hypothetical compositional value, whether by elaboration or by extension, reflects the contribution to the coding process of extra-grammatical resources deployed by the language user. An example of extension is a novel metaphor, e.g. the use of cabbage harvester to designate a guillotine. The compositional semantic value of this expression is roughly 'something that harvests cabbage'; its actual, contextual semantic value is 'guillotine'; the comparison of these two values, serving to register their points of similarity and conflict (and thus responsible for the expression being perceived as metaphoric), is a type of categorizing judgment; and the symbolic units activated in the coding process include the lexical items cabbage and harvester, as well as the schematic symbolic unit representing the relevant pattern of compound formation. As an example of specialization, suppose I show you a new gadget, used to sharpen chalk, and refer to it as a chalk sharpener. The compositional value of this expression is simply 'something that sharpens chalk', but in context — where you actually see the gadget — your understanding of this novel expression is far more detailed: you see that it is a mechanical device, and not a person; you note its approximate size; you may observe how it operates; and so on. All of this constitutes the expression's contex-
16 RONALD W. LANGACKER tual semantic value, the semantic pole of the usage event. As is typically the case, the compositional value substantially "underspecifies" this contextual value, i.e. the former is schematic for the latter in the categorizing relationship between the two. Suppose, now, that the gadget I have shown you becomes widely used, and that the term chalk sharpener establishes itself as the conventional term for this type of object. The conventional semantic value of this lexical item now includes many of those specifications that represented non-compositional facets of its contextual value when the expression first occurred. (This is precisely what happened with pencil sharpener, which means not just 'something that sharpens pencils', but is further understood as indicating a particular sort of mechanical device.) The implication of this type of development ought to be apparent: since the non-compositional aspects of an expression's meaning are part of its contextual value (i.e. how it is actually understood) the very first time it occurs, and further become part of its conventional value when it is established as a unit in the grammar, it is pointless (indeed misguided) to arbitrarily exclude these facets of its meaning from the domain of linguistic semantics. Moreover, conventionalization is a matter of degree; there is no particular threshold at which non-compositional specifications suddenly undergo a change in status from being extra- linguistic to being conventional and hence linguistic. The process is gradual, and is initiated when the expression is first employed. This discrepancy between compositional and contextual value is not limited to short expressions of the sort that conventionalize as lexical items; it is characteristic of novel sentences as well. Far more is contributed to the understanding of a sentence by context and general knowledge (and far less by compositional principles) than is normally acknowledged. For instance, (1 l)(a) would generally be interpreted in the fashion of Fig. 3(a), but given the appropriate context it could easily be used and understood for any of the situations in 3(b)-(g) and indefinitely many others. (11) (a) The ball is under the table. (b) He is barely keeping his head above the water. While (ll)(b) evokes the conception of somebody struggling to keep afloat while swimming, this conception cannot be derived by regular compositional principles from the conventional meanings of its component lexical items; in another context, the sentence could be understood in a radically different way. (Imagine a race over the ocean by helicopter, where the con-
AN OVERVIEW OF COGNITIVE GRAMMAR 17 testants must transport a severed head, suspended by a rope from the helicopter, from the starting line to the finish; a contestant is disqualified if the head he is carrying ever dips below the water's surface.) An expression's compositional value owes nothing to context, so in cases like these it must be sufficiently abstract (schematic) to be compatible with all the countless contextual interpretations the expression might receive — and that is very abstract indeed. Thus, if semantics is restricted to what is algorithmically computable from linguistic units, the resulting semantic representations will be so limited and impoverished relative to how expressions are actually understood that we would hardly recognize them as reasonable approximations to their meanings. The domain of semantic analysis can certainly be defined in this way, but one must then question whether its content, so delimited, is independently coherent or worthy of serious interest. Figure 3 In short, the dictum that linguistic semantics is fully compositional does not rest on empirical observation, but is rather a matter of a priori definition by theorists who wish to consider language as a self-contained formal system. Semantic structure is rendered compositional simply by defining non-compositional aspects of meaning as belonging to pragmatics or extra- linguistic knowledge. The position of cognitive grammar is that this distinction is arbitrary, and that semantics is only partially compositional. There are indeed patterns of composition, and these are described at the semantic pole of the conventional units representing grammatical constructions.
18 RONALD W. LANGACKER However, the meaning of a complex expression (whatever its degree of conventionalization) is recognized as a distinct entity not in general algorithmically derivable from the meanings of its parts. Its compositional value (assuming for sake of discussion the coherence of this notion) need not always be separately computed, and if it is, this computation is only one step in arriving at how the expression is actually understood. Grammar as a Symbolic Phenomenon The "autonomy" of grammar (and of syntax in particular) is a fundamental tenet of contemporary theoretical orthodoxy (cf. Newmeyer 1983). Grammar (or at least syntax) supposedly constitutes an independent level or dimension of linguistic structure, describable as an autonomous "component" within the linguistic system. The remainder of this paper addresses the autonomy issue and argues for the viability of a radical alternative: grammar is inherently symbolic, and therefore distinct from neither lexicon nor semantics. In view of the pivotal nature of this issue, it is surprising how little attention has been devoted to clarifying precisely what it means for grammatical structure to be autonomous, or to specifying what form a conceivable alternative might take. For our purposes here, I will understand the autonomy thesis as including the following points: (i) Grammatical description requires the postulation of tree-like structures which represent an independent facet of linguistic organization; commonly assumed are phrase trees such as Fig. 4, which specify constituency, linear ordering, and category membership (the latter through node labels), (ii) Lexical items are "inserted" into such trees, which themselves have neither semantic nor phonological content; rules of semantic and phonological interpretation apply to the content provided by the inserted lexical items, and take grammatical tree structure into account, but these trees per se are not semantic or phonological objects, (iii) Being susceptible to neither semantic nor phonological characterization, a distinct set of grammatical categories must be posited; the class memberships of a lexical item are specified by syntactic features (or some comparable diacritic device).
AN OVERVIEW OF COGNITIVE GRAMMAR 19 Figure 4 The proposed alternative does not employ phrase trees and rejects points (i)-(iii). It claims that grammar (both morphology and syntax) is describable using only symbolic elements, each of which has both a semantic and a phonological pole. The symbolic units characterizing grammatical structure form a continuum with lexicon: while they differ from typical lexical items with respect to such factors as complexity and abstractness, the differences are only a matter of degree, and lexical items themselves range widely along these parameters. I will therefore begin my presentation of the alternative view with a brief discussion of lexical units. Lexical items vary greatly in their internal complexity. We are concerned here in particular with "symbolic complexity", i.e. whether an item decomposes into smaller symbolic units. Those which do not, and are consequently minimal from the symbolic standpoint, are known as "morphemes". Examples are given in Fig. 5(a), which illustrates the notations I will use here for symbolic units: the semantic pole is shown at the top, its content indicated by capital letters (thus SHARP abbreviates the meaning of sharp); the phonological pole is shown at the bottom (orthographic representations suffice for present purposes); the line between the two poles stands for their symbolic relationship; and the box enclosing them indicates that the symbolic structure as a whole has the status of a conventional unit.
20 RONALD W. LANGACKER (a) SHARP! sharp] IPENCII pencil J en! -en] |er |-er ft) SHARP-EN sharp-en / SHARP sharp x EN -en composite structure component structures (c) "exploded" format [[SHARP Jsharp -en) "compacted" format (d) | [SHARP || sharp ~en~| -en| er] -erl (e) PENCIL^ pencil] SHARP [sharp EN -enl ER -er Figure 5 The majority of lexical units are morphemically complex, as illustrated for sharpen in Fig. 5(b). I will speak of the constituent morphemes, [SHARP/sharp] and [EN/-en], as "component structures". The structure at the top in 5(b), what I call the ''composite structure", represents the semantic and phonological value of the expression as an integral whole. I do not assume full compositionality for either the semantic or the phonological pole, so the composite structure need not be strictly computable from its components; possible discrepancies between the compositional value and the actual composite structure are not our immediate concern, however. To the extent that an expression does approach full compositionality, the composite structure reflects the "integration", at each pole, of the contents of the two component structures. Thus [SHARP-EN], the composite semantic structure, reflects the integration of [SHARP] and [EN], while the composite phonological structure [sharp-en] results from the integration of [sharp] and [-en]. Fig. 5(b) depicts the complex expression in an "exploded" format, showing the composite structure separately from its components. The "compacted" representation of 5(c) is adopted for convenience as an abbreviatory notational variant: it implies both component and composite structures, even though this latter is not shown individually. There is of course no intrinsic limit on the symbolic complexity of lexical items, since a composite structure may function as one component of a larger expression at a higher level of organization. Fig. 5(d) diagrams the integration of sharpen and -er to form the derived noun sharpener. Sharpener in turn combines with pencil to form the compound pencil sharpener, as seen in 5(e). All of these expressions are conventional units of English.
AN OVERVIEW OF COGNITIVE GRAMMAR 21 Lexical units also vary greatly along the parameter of schematicity, i.e. they are characterized at differing levels of precision and detail. This is most obvious at the semantic pole. The semantic pole of bird, for instance, is schematic for that of sparrow, the categorizing unit [[BIRD] -» [SPARROW]] indicates that the specifications of [SPARROW] achieve a finer grain (or higher resolution) than those of [BIRD], and represents the conventional judgment that sparrows instantiate the bird category. There is similarly a relationship of schematicity between two basic senses of cat, that which indicates any feline, and that which contrasts with kitten and indicates an adult feline in particular. The grammar of English thus includes the categorizing unit shown in Fig. 6(a): the conventional symbolic structure [FELINE/cat] is schematic with respect to the equally conventional symbolic unit [ADULT FELINE/cat]. Note that the locus of elaboration is confined to the semantic pole (the two symbolic units are identical at the phonological pole). (a) FELINE cat ADULT FELINE cat PL" | -lar " PL -ler J Figure 6 Variation along the parameter of schematicity is also observable for lexical units at the phonological pole. Depending on vowel harmony, for instance, the Turkish plural morpheme appears as either -lar or -ler, hence [PL/-lar] and [PL/-ler] are symbolic units that occur overtly. Presumably speakers capture the obvious generalization by extracting a schema to represent the commonality of these two expressions; this schema can be given as [PL/-lAr], where A is schematic for a and e (i.e. it is characterized simply as a non-high, unrounded vowel). The grammar of Turkish thus includes the units and categorizing relationships shown in Fig. 6(b). To accommodate the special root structure of Arabic, we can posit phonologically schematic symbolic units such as [WRITE/...k...t...b...], where '...' indi-
22 RONALD W. LANGACKER cates the possible occurrence of a vowel (but does not specify its quality). We see, then, that lexical units manifest varying degrees of schematic- ity at either the semantic or the phonological pole. Some symbolic units are schematic at both poles, e.g. Turkish [PLAlAr] ([PL] designates a set of entities belonging to the same category, but this category is identified only schematically). I suggest that schematicity, at one or both poles, is precisely what characterizes those symbolic units responsible for grammatical structure. Though a gradation can be observed, essential grammatical units are highly schematic at both poles. Moreover, the greater the schematicity of a symbolic unit, the more likely it is to fall within the traditionally-recognized realm of grammar (as opposed to lexicon). In Part II of Langacker in press, I try to demonstrate that a subjectivist view of meaning which properly accommodates conventional imagery makes possible a descriptively adequate and revealing semantic characterization of basic grammatical classes (e.g. noun, verb, adjective); reasonably precise definitions are offered that are grounded in plausible assumptions about cognitive processing. Here I will simply assume that such descriptions are possible in principle, and show where this assumption takes us. If nouns, verbs, and other basic categories are susceptible to notional definition, we may posit a schematic symbolic unit serving to characterize each class: there will be a "noun schema" expressing the commonality of all nouns, a "verb schema" defining the class of verbs, and so on. Such schemas are maximally schematic at each pole. Phonologically, we can say little more about nouns or verbs as a class than that they have "some phonological content"; I use [X] and [Y] to indicate these highly schematic phonological structures in Fig. 7(a)-(b). At the semantic pole, I claim that every noun designates a "thing", while every verb designates a "process"; note that "thing" and "process" are technical terms whose import is highly abstract (the class of "things", for instance, is not restricted to physical objects). The membership of a lexical unit in one of these categories is represented in the grammar not by any diacritic or syntactic feature, but rather by a categorizing unit. The unit depicted in Fig. 7(c) specifies that pencil instantiates the noun schema, and is therefore a member of the class that this schema defines. Similarly, the categorizing unit in 7(d) indicates that write is a verb by virtue of bearing an elaborative relationship to the verb schema.
AN OVERVIEW OF COGNITIVE GRAMMAR 23 (a) (b) [THING] I X 1 [PROCESS 1 Y noun (c) s chema ] verb 1 s chema 1 THING] 1 x 1 , * , PENCIL [pencilj 1 (d) [PROCESS 1 Y 1 f IwriteI |write)1 Figure 7 We see, then, that this framework accommodates basic grammatical categories and category membership within the restrictive confines of the content requirement: nothing has been invoked other than specific symbolic units (e.g. pencil), schematic symbolic units (e.g. the noun schema), and categorizing relationships between the two. What about grammatical rules and constructions? These are not distinguished in cognitive grammar, but are treated instead as alternate ways of regarding the same entities, namely symbolic units that are both schematic and complex (in the sense of having smaller symbolic units as components). I refer to these entities as "constructional schemas". Fig. 8(a) represents the constructional schema for deverbal nominali- zations in -er {talker, swimmer, complainer, painter, mixer, opener, and so on). Despite the abbreviatory notations, this structure is a complex symbolic unit in which component structures are integrated to form a composite structure at both the semantic and the phonological pole. One component structure is the verb schema; semantically it designates a process of unspecified character, while phonologically its content is maximally schematic. The other component structure is the morpheme -er, which is specific phonologically but semantically abstract (it designates a "thing" identified only by the role it plays in a schematic process). The composite structure (not separately shown in the compacted format of 8(a)) can be abbreviated as follows: [PROCESS-ER/Y-er]. Semantically, it designates a thing, and the process in which it plays a specified role is equated with that designated by the schematic verb stem [PROCESS/Y]. Phonologically, it specifies the suffixation of -er to the content provided by the stem.
24 RONALD W. LANGACKER (a) (b) : (process) T Y J ER |-er llTALK Htalk er~| -er] constructional schema instantiation of schema [PROCESS 1 Y 1 ER j -er) 1 i 4 & Italk] Italkl ER 1 j -er j 1 categorizing relationship (structural description) Figure 8 In short, the constructional schema is exactly parallel in formation and internal structure to any of its instantiations, such as talker, diagrammed in Fig. 8(b); the only difference is that the schematic verb stem in the former is replaced in the latter by the more elaborate content of talk. The constructional schema can therefore be regarded as a symbolically-complex expression, albeit one that is too abstract semantically to be very useful for communicative purposes, and too abstract phonologically to actually be pronounced or perceived. Instead it serves a classificatory function, defining and characterizing a morphological construction by representing the commonality of its varied instantiations. The relationship between the schema and a specific instantiating expression resides in a categorizing unit of the sort depicted in Fig. 8(c). The global categorization, indicated by arrow T, amounts to a relationship of schematicity: talker elaborates the abstract specifications of the schema, but is fully compatible with them, and thus instantiates the morphological pattern the schema describes. This global categorization can be resolved into local categorizations between particular substructures. Arrow 42' indicates that talk is categorized as a verb and elaborates the schematic stem within the constructional schema. The morpheme -er occurs in both the schema and the specific expressions that instantiate it; the double-headed solid arrow, labeled '3', marks this relationship of identity. A symbolically-complex expression can of course be incorporated as one component of an expression that is more complex still. A derived verb like sharpen can therefore function as the stem in the morphological pattern just described, resulting in sharpener, this noun in turn combines with pencil in the compound pencil sharpener, whose organization is sketched at the bottom in Fig. 9(a). In the same fashion, one constructional schema is capable of being incorporated as a component of a larger schema describing
AN OVERVIEW OF COGNITIVE GRAMMAR 25 expressions of greater complexity. At the top in Fig. 9(a), we see the higher-order constructional schema responsible for expressions like pencil sharpener, mountain climber, lawn mower, taxi driver, flamethrower, etc. One of its components is the constructional schema for nominalizations with -er, as represented in Fig. 8(a); the other component is the noun schema (cf. Fig. 7(a)). Phonologically, the higher-order schema specifies the juxtaposition of the two stems to form a compound: [X Y-er]. Semanti- cally, it specifies that the "thing" symbolized by [X] is equated with the object of the process symbolized by [Y]. The global categorizing relationship between this constructional schema and a specific compound like pencil sharpener reflects local categorizing relationships at different levels of organization. In Fig. 9(a), arrow '1' again indicates that the overall relationship is elaborative (i.e. pencil sharpener conforms to the specifications of the schema but is characterized in finer detail). At a lower hierarchical level, arrows '2' and '3' show that pencil and sharpener qualify respectively as instantiations of the noun schema and of the morphological pattern previously considered. Finally, at the lowest hierarchical level, relationship '3' is resolved into the local categorizations '4' (which classes sharpen as a verb) and '5' (an identity relation). (a) (THING! | X | 2 I \ (process 1 Y ~i~. f [PENCIL 1 [pencil ER II -er II —i i k \ I T JSHARI 1 JEN ] (sharp [ -en] r3 I—» 5 If , f "ER -er| (t) THING CHALK chalk PROCESS] ER -er| flSHARP] 1 [sharp) |EN]! |-en| Figure 9 Grammar, I claim, is nothing more than patterns for successively combining symbolic expressions to form expressions of progressively greater complexity. These patterns take the form of constructional schemas, some of which incorporate others as components. Constructional schemas have multiple functions in this model. First, they capture generalizations by representing the commonality observed in the formation of specific instantiating expressions. Second, they provide the basis for categorizing relation-
26 RONALD W. LANGACKER ships which show the status of specific expressions with respect to structures and patterns of greater generality. An expression's "structural description" is simply the set of categorizing relationships in which it participates; for instance, the structural description of pencil sharpener includes the categorizing relationships depicted in Fig. 9(a) (together with others not shown). Finally, a constructional schema serves as a template for the computation of novel instantiating expressions. Consider a previous example, namely the coinage of chalk sharpener to designate a previously unfamiliar gadget. Symbolic resources available to the speaker include the units and categorizations indicated in Fig. 9(b): the lexical units chalk and sharpen, their respective categorizations as a noun and a verb, and the constructional schema representing a compounding pattern. To compute an appropriate compound, the speaker need only co-activate these structures, i.e. carry out the elaborative operations relating chalk to the noun schema, and sharpen to the verb schema, in the setting of the schema for the construction. Of course, the composite semantic structure obtained in this way will only approximate the expression's actual, contextually-determined value. To conclude this section, let us once more consider phrase trees like Fig. 4, which specify linear ordering, constituency, and constituent types. It may now be apparent that all three sorts of information are also provided in the proposed alternative, which posits only symbolic units for the description of grammatical structure. Linear ordering (actually, temporal ordering) is simply one dimension of phonological structure. A symbolic structure — whether simple or complex, specific or schematic — specifies the temporal sequencing of its phonological components as an inherent part of the characterization of its phonological pole. Thus sharp, sharpen, sharpener, and pencil sharpener all specify, as part of their internal phonological structure, the temporal ordering of their segments, syllables, morphemes, and stems; the same is true of the schemas which these expressions instantiate, except that the phonological elements in question are partially schematic (e.g. the constructional schema of Fig. 8(a) specifies, in its composite structure, that -er follows the schematically-characterized verb stem). Syntax and morphology are not sharply distinct in this regard. The difference between them reduces to whether the construction involves multiple words at the phonological pole, or parts of a single word (compounds are thus a borderline case). Constituency is not a separate facet of linguistic structure, but merely reflects the order in which symbolic structures are successively combined in
AN OVERVIEW OF COGNITIVE GRAMMAR 27 the formation of a complex expression. We may speak of a "compositional path" leading from individual morphemes, through intermediate-level composite structures, to the highest-level composite structure representing the value of a complex expression overall. The compositional path remains implicit in "compacted" diagrams like Fig. 9(a), which fail to separately depict composite structures, but is immediately apparent when these are converted into the equivalent '"exploded" format, as illustrated in Fig. 10. On the left in Fig. 10 is the compositional path of pencil sharpener; intermediate-level composite structures show that sharpen and sharpener are constituents (in addition to the individual morphemes and the expression as a whole). On the right in Fig. 10 is the compositional path of the constructional schema that pencil sharpener instantiates. PENCIL-SHARP-EN-ER pencil sharp-en-er 1 THING-PROCESS-ER1 X Y-er PENCIL 1 pencil \ 2 THING] X SHARP-EN-ER sharp-en-er 1SHARP-EN [sharp-en ^ k PROCESS Y ] SHARP sharp enI -en Figure 10 These are not phrase trees, but simply ordered assemblies of symbolic units. They differ from phrase trees in several ways. First, every node is a symbolic structure incorporating both semantic and phonological content
28 RONALD W. LANGACKER (and nothing else). Second, these structures are not linearly ordered: each node specifies temporal ordering internally at the phonological pole, but the nodes are not temporally ordered with respect to one another. A third difference is the absence of node labels specifying the grammatical class of constituents. Class membership is specified not by labels or features, but rather by categorizing relationships, each reflecting the assessment that a structure instantiates a particular schema. In the case of grammatical structure, the categorizing schemas are symbolic, with actual semantic and phonological content (though often this content is abstract, even to the point of being essentially vacuous). Collectively, these categorizing judgments constitute the expression's structural description. Distribution and Predictability Cognitive grammar posits only symbolic units for the description of grammatical structure. Having examined the general character of such analysis, we turn now to specific phenomena that are often taken as demonstrating the autonomy of grammar. One class of arguments pertains to the impossibility of predicting the membership of "distributional classes", i.e. classes defined on the basis of morphological or syntactic behavior. For example, there is no way to predict, on either semantic or phonological grounds, precisely which English verbs form their past tense by ablauting / to a (i.e. [1] to [ae]); this pattern is possible only with a small set of verb stems (sit, swim, begin, ring, sing, etc.) that are neither coherent semantically nor unique phonologically. Nor is it possible, apparently, to predict the exact membership of the class of verbs occurring in the so-called "dative shift" construction. While the construction favors verbs of transfer that are monosyllabic (e.g. give, send, mail, ship, and buy, but not transfer, communicate, purchase, or propose), certain pairs of verbs that appear quite comparable both semantically and phonologically exhibit contrasting behavior: (12) (a) I told the same thing to Bill, (a') I told Bill the same thing, (b) I said the same thing to Bill. (b') * I said Bill the same thing. (13) (a) I peeled a banana for Bill, (a') I peeled Bill a banana.
AN OVERVIEW OF COGNITIVE GRAMMAR 29 (b) I cored an apple for Bill, (b') *I cored Bill an apple. Much more can be said about this issue, but let us assume the worst, namely that the class of dative-shift verbs is unpredictable and must somehow be listed in the grammar. In providing such information, linguists generally resort to some type of diacritic or grammatical feature. For example, sit might be marked by a diacritic indicating its membership in the class of verbs (e.g. "Class 2B") that ablaut / to a to mark the past tense. Or, in addition to semantic and phonological features, give might be attributed a syntactic rule feature such as [4- Dative Shift], which specifies its ability to undergo the dative shift rule. Whatever device is used, the marking is assumed to have neither semantic nor phonological content, since the class it identifies is not uniquely predictable on the basis of either form or meaning. The autonomy of grammar is commonly assumed to follow as an inescapable consequence: since grammatical behavior forces one to posit a distinct set of specifically "grammatical" classes and descriptive constructs, grammar must constitute an independent domain of linguistic organization. This argument is fallacious, for it confuses two issues that can in principle be distinguished: (i) what kinds of structures there are; and (ii) the predictability of their behavior. I call this the "type/predictability fallacy". It is not logically incoherent to maintain that only symbolic units are required for the description of grammatical structure, even though one cannot always predict, in absolute terms, precisely which symbolic units occur in a given construction. Some type of marking or listing is required to provide this information, but there is no a priori reason to believe that a contentless feature or diacritic is the proper device for this purpose. Indeed, little cognitive plausibility attaches to the claim that speakers possess any direct analog of empty markers like [Class 2B] or [4- Dative Shift] in their mental representation of linguistic structure. It is possible to furnish the requisite distributional information without resorting to anything other than symbolic units. To say that a particular verb, e.g. sit, marks its past tense by ablauting / to a is equivalent to saying that sat is a conventional unit of English. Similarly, to say that a particular verb, e.g. give, occurs in the dative shift construction is equivalent to saying that the schema describing this construction is instantiated by subschema having give as its verbal element. I have argued elsewhere (Langacker 1982a, in press) for a "usage-based" model of linguistic structure, wherein
30 RONALD W. LANGACKER both schemas and their instantiations are included in the grammar of a language, provided that they have the status of conventional units. Schemas capture generalizations by representing patterns observable across expressions. Unit instantiations of constructional schemas (both specific expressions and subschemas at varying levels of abstraction) describe the actual implementation of these generalizations by specifying their conventional range of application. The specification that sit marks its past tense by ablauting / to a is thus accommodated by including in the grammar of English the symbolic structures sketched in Fig. 11(a). At the top in 11(a) is the constructional schema describing the ablaut pattern in general terms. There are two component structures. The first is the verb stem, characterized semantically only as designating a process; its phonological pole is also schematic, specifying only the inclusion of /. The other component structure is the appropriate past-tense allomorph; its semantic pole is given as [PAST], and its phonological pole as [i —> a] (i.e. past tense is symbolized by an ablauting operation applied to the stem). The compacted format is used, so the resulting composite structure, namely [PROCESS-PAST/...a...], is not separately shown. Instantiating this constructional schema is the structure at the bottom in 11(a), which is nothing other than the lexical unit sat. This complex expression is naturally parallel in organization to the schema it instantiates. The two component structures are [SIT/sit], which elaborates the schematic verb stem, and the past-tense allomorph [PAST/i --> a], which also occurs in the constructional schema. Though not shown separately, the composite structure [SIT-PAST/sat] instantiates that of the schema. Other past-tense verbs that follow this pattern are also included in the grammar as conventional units, each of them combining with the constructional schema in a categorizing unit analogous to 11(a). Both the morphological pattern itself and its lexical extension are thereby specified using only the apparatus permitted by the content requirement. :| (process 1 Y 1 THING X ' ♦ ; [GIVE [give - ■■ | "THING 1 j I z l|| I' ■ j THING X L- -J J. .-. THING111 Z | Figure 11 l| PROCESS ...i... 1 PAST it i--»*a|| 1 * ; 1 x" r [SIT sit f PAST ||| 1—a||
AN OVERVIEW OF COGNITIVE GRAMMAR 31 Fig. 11(b) sketches the categorizing unit identifying give as a verb that occurs in the dative shift construction. At the top in 11(b) is a simplified representation of the constructional schema. There are three component structures, the verb schema and two instances of the noun schema (one would actually have to indicate that these are full noun phrases). At the semantic pole, the composite structure (not separately shown) specifies that the "things" designated by the noun phrases participate in a possessive relationship that results from the verbal process; its phonological pole specifies that the noun phrases are directly contiguous and follow the verb. At the bottom in 11(b) is a subschema representing the generalization that give appears in this construction. It is precisely parallel to the schema, except that the specific content of give replaces that of the verb schema at both the component- and composite-structure levels. Similar categorizing units are posited for other dative-shift verbs. There is much more that needs to be said about distribution, predictability, and the related issue of economy in linguistic descriptions, but this is not the place to pursue the matter. Our present concern is merely to observe that inability to predict the exact membership of distributional classes does not, in principle, establish the autonomy of grammar (as earlier defined). I have outlined a strategy for providing the necessary information using only symbolic units. Moreover, only constructs and descriptive devices needed for independent reasons arc employed. Arbitrary Grammatical Markings The autonomy of grammar is commonly asserted on the grounds that languages impose arbitrary requirements on the form of permissible expressions. The validity of such an argument depends on how the pivotal notions are understood, in particular the notions "form" and "autonomy". My own definition of what it means for grammar to be autonomous was presented earlier; the alternative I propose relics solely on symbolic units for the description of grammatical structure. From this perspective, the only sensible definition of "form" equates it with the phonological pole of linguistic units and expressions — specifically rejected are phrase trees like Fig. 4, which supposedly represent an independent level or dimension of "grammatical form". Linear ordering is simply one dimension of phonological space, and we have seen how the symbolic alternative accommodates both constituency and category membership.
32 RONALD W. LANGACKER Obviously, languages do impose requirements on the form of expressions. Most of these requirements have to be learned, as they are not strictly predictable from any independent factor; in this sense they are arbitrary (or at least conventional). That itself does not distinguish grammar from lexicon — after all, a lexical item such as [APPLE/apple] can be regarded as imposing the arbitrary formal requirement that the notion [APPLE] be symbolized by [apple]. Nor does it demonstrate that grammatical structure involves anything other than symbolic units. Consider the formal restriction that a simple adjective in English must precede rather than follow the noun it modifies, as in ripe apple. Although this formal relationship is essentially arbitrary, we must certainly recognize its symbolic import: the fact that [ripe] directly precedes [apple] along the temporal axis is precisely what symbolizes the fact that the property designated by [RIPE] is attributed to the "thing" designated by [APPLE]. The regularity of this relationship is captured by positing a constructional schema having a schematic adjective and a schematic noun for its component structures; the schema's phonological pole specifies the temporal contiguity and ordering of the adjective and noun, while its semantic pole specifies the fact and nature of their semantic integration. The formal restriction is therefore simply the phonological pole of a particular type of symbolic unit. Discussions of the autonomy issue generally cite the existence of "empty" markers, which appear to serve purely grammatical purposes without making any semantic contribution. Frequent examples are the auxiliary verbs be and do; certain subordinators; prepositions that are governed by a particular verb or construction (especially when no alternative choice is permitted); markings induced by agreement; and also the markers for case, gender, and similar categories. We cannot examine all these matters in detail (previous works on cognitive grammar describe the semantic value of numerous "grammatical" elements). I will nevertheless argue that they are not intrinsically problematic for the symbolic conception of grammatical structure. We may certainly concede the existence of markings that do not substantially affect the information content of the expressions in which they appear. However, there is much more to meaning than information content (or truth conditions): equally important is imagery, i.e. how that content is structured or construed for communicative purposes. Given a linguistically appropriate view of meaning, one that accommodates conventional imagery, grammatical markers like those noted above can in fact be considered
AN OVERVIEW OF COGNITIVE GRAMMAR 33 meaningful. The nature and degree of their semantic contribution varies — as a limiting case, it may be essentially vacuous; but this extreme should be regarded as the endpoint of a continuous scale, and not as criterial or prototypical for grammatical elements. Let us begin by considering certain factors that are sometimes taken as indicating that a grammatical marker is meaningless. These include: (i) being redundant; (ii) being obligatory; (iii) serving a specifiable grammatical function; and (iv) being language-specific. Granted various concepts and assumptions of the present framework, none of these factors implies the meaninglessness of a grammatical element. A redundant element may be meaningless from the standpoint of information theory, but that is only a minor concern of linguistic semantics. For natural language, it is important that we distinguish clearly between two notions that are certainly not equivalent, namely "meaningfulness" and "non-overlapping meaning". There is always some semantic overlap between two component structures that combine to form a composite expression; it is precisely because the component structures overlap in some fashion that they can be integrated to form a coherent composite conceptualization. In ripe apple, for instance, ripe specifies the maturity of fruit and its readiness for consumption, and hence introduces the notion of fruit schematically, while apple subsumes and elaborates this schematic notion by characterizing a particular type of fruit. Since all complex expressions involve semantic redundancy of this sort, the meaning of one component cannot be determined just by subtracting the meaning of its companion from that of the expression as a whole — the meanings of the component structures are not strictly complementary, so each must be described in its own terms. As a limiting case, one component structure is properly considered meaningful even if its semantic content is fully subsumed by that of the other, and therefore contributes nothing to the composite semantic structure that would not be available from the second element alone. This can be exemplified by a compound like oak tree or puppy dog. The content of tree is fully subsumed by that of oak, and the content of dog by puppy, but no one would argue that tree and dog are semantically empty morphemes. Semantic overlap is inherent to grammatical constructions, cases like oak tree and puppy dog being special only in that the region of overlap happens to exhaust the content of one component. Expressions like oak tree illustrate what I call the "schematic transparency principle": when combined, a
34 RONALD W. LANGACKER schema and its instantiation merge to form a composite conceptualization equivalent to the instantiation. Since [TREE] is schematic for [OAK], the content of the former being fully subsumed by that of the latter, the composite notion obtained by merging their specifications is identical to the instantiation [OAK]. But while tree adds nothing to the information content of oak tree, it does make a semantic contribution from the standpoint of imagery: explicit mention of the superordinate category tree makes the membership of oaks in this category somewhat more salient in the expression oak tree than it is in the simple expression oak. With this in mind, consider the auxiliary verb do and the contrast between pairs of sentences like (14): (14) (a) My cousin smokes. (b) My cousin does smoke. I analyze the auxiliary do as designating a maximally schematic process, i.e. it is a true verb that conforms to the specifications of the verb schema (Fig. 7(b)) and in fact is identical to that schema at the semantic pole. Semanti- cally, do is therefore schematic for smoke, as it is for any other verb. From the schematic transparency principle, it follows that the composite structure of do smoke is equivalent to that of smoke alone, i.e. do contributes nothing to the information content of the complex expression. One can understand why many linguists have concluded that do is semantically empty, but I have tried to show that this conclusion does not really follow. The redundancy of do in composition with a specific verb does not preclude its analysis as a symbolic unit with actual (though schematic) semantic content. Moreover, this content helps explain the semantic contrast of sentence pairs such as (14). Like that between oak and oak tree, the contrast between smoke and do smoke pertains to imagery rather than information content, and resides in the added salience accorded a superordinate category (the [PROCESS] category in this case) by virtue of explicit symbolization. In conjunction with emphatic stress, the occurrence of do + V (rather than V alone) underscores the notion of a process and thus lends a nuance of "emphasis1' or "assertion" to the verbal expression. Hence the semantic redundancy of a grammatical marker (from the standpoint of information content) is compatible in principle with its being both meaningful and describable as a symbolic unit. We can also coherently assert the meaningfulness of a grammatical marker that is obligatory, is chosen to the exclusion of other alternatives, and serves a specifiable gram-
AN OVERVIEW OF COGNITIVE GRAMMAR 35 matical function. The auxiliary verb do is once again a convenient example. If (14)(a) is negated, questioned, used as a tag, or rendered elliptic, do occurs in addition to smoke or in lieu of it: (15) (a) My cousin does not smoke. (b) Does my cousin smoke? (c) My cousin smokes, doesn't she? (d) She does. Do is generally attributed a specific grammatical function in these constructions: it carries the markings of tense and subject-agreement, which would otherwise be "stranded". However, nothing inherently prevents a meaningful element from being conventionally adopted to play a particular grammatical role, and elements with highly schematic content would seem to be ideally suited for this purpose. (By the same token, subordinators like since, because, while, although, etc. indicate the beginning of a subordinate clause and thus have a specific grammatical function, but they are clearly meaningful.) The schematicity of do with respect to the class of verbs explains its semantic "transparency" in the constructions of (15)(a) and (b), and also its ability to function as a verbal "pro form" in those of (c) and (d). What about the fact that do is obligatory in these constructions and fully predictable (i.e. no other verb can take its place)? Only granted a strict information-theoretic definition of meaning does this imply that do is meaningless; there is no such implication when meaning is related instead to conceptualization and conventional imagery, and semantic overlap is recognized as an inherent feature of grammatical constructions. A constructional schema indicates the conventionality (well-formedness) of expressions conforming to its specifications. Should such a schema mention a particular symbolic element (e.g. do) as one of its component structures, the expressions it sanctions are limited to those that incorporate the element in question. The absence of sanctioned alternatives does not prevent this element from having a semantic pole with actual content, regardless of how schematic this content might be, or how redundant it might be with respect to the other component. One facet of conventional imagery resides in the compositional path through which the composite structure of a complex expression is progressively assembled, and even redundant elements contribute in this regard, as we have seen. Finally, consider the fact that grammatical markings vary from language to language — not just their form, but also their inventory and dis-
36 RONALD W. LANGACKER tribution. This fails to establish that such markings are meaningless or non- symbolic. It merely indicates, in the realm of grammatical constructions, what is already evident from the study of lexicon: languages make different inventories of symbolic resources available to their speakers, who consequently say comparable things in different ways. To convey roughly the same content, the speakers of two languages may be forced, by the constructional schemas at their disposal, to employ expressions that differ in such factors as how precisely they specify some parameter (e.g. definite- ness, number, gender), the amount of redundancy they incorporate, or the nature of the compositional path through which they arrive at the composite semantic structure. In short, semantic structure is not universal when imagery is properly taken into account, so the non-identity of grammatical markings across languages does not itself demonstrate their semantic irrelevance. I conclude, then, that the factors considered above do not establish the meaninglessness of grammatical markers, as they are commonly assumed to do. With this background, we can now examine some specific phenomena that are almost invariably cited in support of the autonomy thesis: agreement, number, case, and gender. I must limit myself to sketching the general strategy for dealing with them in the context of cognitive grammar. Let us start with subject-verb agreement in English, as illustrated in (16): (16) (a) The kitten is playful. (b) *The kitten are playful. (c) *The kittens is playful. (d) The kittens are playful. Standard analyses attribute no semantic value to the person and number distinctions marked on the verb. Instead, some mechanism is invoked (e.g. a transformational rule) that copies the person and number features of the subject noun phrase onto the verb, where they are eventually spelled out morphologically. Being mechanically induced by morphosyntactic processes, verb agreement does nothing more than duplicate information provided elsewhere in the sentence, and is therefore semantically empty even if the person and number categories are considered semantic. We will confine our attention to number, and assume for the moment that the number category, as marked on nominal expressions, is indeed semantic. There are cogent reasons for questioning the validity of the fea-
AN OVERVIEW OF COGNITIVE GRAMMAR 37 ture-copying analysis; though it can doubtless be made to work, its implementation is far less straightforward than is generally recognized (cf. Morgan 1972), and it offers no real insight into the phenomenon. Let us consider just one type of problem, namely instances where the same subject nominal occurs with either a singular or a plural verb: (17) (a) The faculty agrees, (b) The faculty agree. (18) (a) Drinking and singing is fun. (b) Drinking and singing are fun. (19) (a) The team is playing hard, (b) The team are playing hard. In the case of (17), one might argue that faculty can be valued either ' + ' or ' — ' for the feature [PLURAL]. Some corroboration is seemingly found in the fact that only the singular demonstratives (this, that) can substitute for the definite article in (17)(a), and only the plural demonstratives (these, those) in (17)(b). For (18), one might similarly propose that nominals formed by conjunction with and are sometimes free to take either feature value (it would not, however, be easy to specify just when that is possible). This type of analysis will not work for (19) in any event, since team is clearly singular in both examples (note this team, *these team, teams). More, then, appears to be going on than is suggested by a mechanical feature-copying analysis. How can these phenomena be dealt with in cognitive grammar, which posits only symbolic units for the description of grammatical structure? The central claim is that number-marking on the verb is independently meaningful. A plural verb contrasts semantically with the corresponding singular verb by portraying its subject as exhibiting a greater measure of individuation among its subparts. Often this specification is redundant, as in (16)(d); because a plural noun like kittens designates a mass consisting of indefinitely many distinct instances of the type of entity designated by the stem, substantial individuation is implied by the morphology of the subject nominal itself. However, since "meaningfulness" is not the same as "non-overlapping meaning", such redundancy does not affect the viability of the analysis. Moreover, the deviance of sentences like (16)(b) and (c) is attributable to the semantic incompatibility of separate specifications concerning the same conceived entity: either the verb attributes substantial individuation to this entity while the subject nominal does not, or conversely.
38 RONALD W. LANGACKER Number marking on the verb is neither redundant nor contradictory with respect to the subject nominal when the entity designated by the latter lends itself to alternate construals implying different degrees of individuation. In (17), it is the number of the verb that indicates whether the faculty is construed holistically as a unified body, or whether its status as a collection of individuals is given greater weight; accordingly, (17)(a) conveys the notion that the faculty, as a unitary formal entity, agrees with some externally-generated proposal, while (17)(b) portrays the faculty agreeing as individuals, perhaps with one another. Similarly, the choice between is and are in (18) signals whether drinking and singing are being construed as two facets of a single complex activity, or as two essentially separate activities, each of which is fun individually. As for (19), it is clear that team is inherently ambivalent, implying distinct individuals who nevertheless subordinate themselves to cooperative endeavor toward a common goal. No particular importance attaches to the fact that the choice between (19)(a) and (b) is in large measure dialectally determined (American vs. British). For one thing, individual speakers may well be familiar with both patterns, and use them to make a semantic distinction (this is coming to be true of my own speech). But what if speakers have no option, so that one pattern or the other must be employed (strictly according to dialect), even though either construal is conceptually quite natural? This merely reflects the conventionality of the imagery embodied by the symbolic resources of a language: out of all the ways of construing a given type of situation, certain possibilities become conventionally established (i.e. represented in the grammar by symbolic units) to the exclusion of others. Like languages, dialects often diverge in this regard. I would argue, then, that the dialectal difference between (19)(a) and (b) is also a subtle difference in meaning — speakers of the two dialects conventionally employ slightly different images to construe the situation for expressive purposes. An argument sometimes advanced for the autonomy of grammar maintains that "semantic number" and "grammatical number" must be distinguished, with grammatical rules sensitive only to the latter (cf. Hudson 1976: 6). For instance, nouns like binoculars, scissors, pants, glasses, tongs, tweezers, pliers, shorts, and trousers are said to be semantically singular, but they show the agreement pattern of plurals: (20) (a) These binoculars are powerful. (b) *This binoculars is powerful.
AN OVERVIEW OF COGNITIVE GRAMMAR 39 The problem with this argument is that it fails to acknowledge either the phenomenon of imagery or its conventional character. Why are binoculars, scissors, pants, etc. considered semantically singular? Presumably because each designates a discrete physical object, not a mass consisting of separate discrete objects all of the same type. Yet it is surely no accident that these nouns in particular act as plurals; a prominent feature of each designated object is that it consists of two more-or-less identical parts that have no separate function or identity — they occur only as complementary halves of the object in question. Consequently, our conception of these objects embraces both unity and duality, either of which might be highlighted by the conventional units of a given language. Anything which emphasizes one facet of a complex notion at the expense of another counts as imagery. Since alternatives are conceivable, the imagery is conventional, i.e. speakers must learn the proper way of expressing these notions. Our inability to predict what pattern a language might use does not entail that the choice has no semantic basis. To summarize, the patterns illustrated in (16) (for "regular" nouns like kitten) and (20) (for nouns like binoculars) are fixed and obligatory, involve a certain amount of redundancy, and could not be predicted if one did not happen to know them. In the present framework, we can nevertheless coherently maintain that the number indications of the subject nominal and the verb represent separate semantic specifications and are independently meaningful; their contributions are sometimes consistent, sometimes incompatible, and sometimes complementary. Plural marking on the verb attributes a certain measure of individuation to the subparts of the entity designated by the subject nominal; the nature and the degree of individuation implied require further study — most likely they vary depending on the inherent properties of the subject. The plural marking that occurs on nouns is also variable in its precise semantic import, and is not necessarily identical in value to that which co-occurs on verbs. Prototypically, plural inflection indicates a mass consisting of indefinitely many instances of the type of discrete object designated by the noun stem. From this basic sense of the plural morpheme, extended senses develop to accommodate the semantic peculiarities of special types of nouns. I thus identify the final -s of nouns like binoculars as the plural morpheme, even though the remainder docs not always occur individually (this is unproblematic in cognitive grammar — see ch. 12 of Langacker in press); presumably it indicates the inherent duality of the object designated by the overall expression.
40 RONALD W. LANGACKER Like number, the categories of case and gender are often subject to "agreement", i.e. the case or gender of a particular entity is often marked in more than one place. It should now be evident that the redundancy, obligatoriness, and language-specific character of such markings do not per se demonstrate the autonomy of grammar. The pivotal question is whether a marker for case or gender can be attributed some kind of semantic value and treated as a symbolic unit. If so, the phenomena can be analyzed along the general lines suggested above for number agreement. Each occurrence of such a marker constitutes a separate semantic specification. Whether for functional reasons or out of sheer perversity, languages sometimes demand multiple specifications pertaining to the same entity, despite their full semantic overlap. Conventional agreement patterns are represented in the grammar of a language by constructional schemas which incorporate only symbolic units. These schemas embody conventional imagery. They subtly influence the meaning of the complex expressions they sanction, both by specifying the inclusion of the case or gender markings (however tenuous or abstract their semantic content may be), and also by dictating a particular compositional path for arriving at the ultimate composite structures. We must now address the issue of whether the categories of case and gender are in fact to be considered meaningful. My comments on case will be very brief and general, as this topic has only begun to be investigated in the context of cognitive grammar (cf. Smith 1985). We may distinguish between nominative and accusative case on the one hand, and "oblique" cases on the other. Basically, the question of whether markings for nominative and accusative case are meaningful reduces to the question of whether the grammatical relations "subject" and "direct object" have a semantic basis. I claim that they do — the characterization pertains to the relative prominence of relational participants (prominence being an important dimension of imagery) — but discussion and justification of this claim are beyond the scope of this paper (see Part II of Langacker in press). The assertion that oblique cases like dative, genitive, instrumental, ablative, etc. make a semantic contribution is perhaps less controversial. I suggest that they always have a semantic value of some sort, though it may be redundant or even vacuous on occasion. It is important that the basic question of their meaningfulness be distinguished from two subsidiary issues that do not bear on it directly. First is the issue of whether a given case has a single meaning in all its uses; I neither claim nor expect that it does. Lexical items in common use almost invariably have a family of distinct but
AN OVERVIEW OF COGNITIVE GRAMMAR 41 interrelated senses, and it is reasonable to assume (as I did earlier for plural markings) that the same is generally true of the more abstract symbolic units regarded as grammatical markers. Having a multiplicity of interrelated meanings is not equivalent to being meaningless. The second issue concerns the grammatical function of case markings, and the fact that a verb, preposition, or construction often governs a particular case. I have already argued that such considerations do not establish that the element in question is semantically empty. Meaningful elements can perfectly well serve specifiable grammatical functions, even when no alternatives are permitted. As with number, a distinction is commonly drawn between "grammatical gender'1 and "semantic" or "natural gender". It is recognized that gender-like categories tend to be semantically motivated for a substantial range of vocabulary (e.g. le filslla fille, le chat/la chatte, etc.), but even within that range there may be exceptions (la sentinelle), and outside that range the category membership of a noun is basically arbitrary (le fromagella viande, le crayon/la plume). While the arbitrariness of gender-like classes is sometimes overstated, the basic facts arc clear enough. It would be foolish to maintain that every French noun taking le or la does so because it is semantically 'masculine' or 'feminine' (using these terms in their normal sense). Nor can one plausibly argue that the gender categorizations of nouns are strictly predictable from their meanings in all cases. Many scholars have taken such facts as demonstrating the autonomy of grammar and grammatical classes. Nevertheless, the data is readily accommodated in a model that posits only symbolic units for the description of grammatical structure. Recall, to begin with, that a meaningful element need not have precisely the same meaning in all its uses — both lexical and grammatical morphemes typically display a network of distinct but interrelated senses (cf. Lindner 1981, 1982; Hawkins 1984; Casad and Langacker 1985). We can therefore attribute to a gender morpheme such highly specific semantic content as 'masculine' or 'feminine' in many of its occurrences, notably with the vast majority of animate nouns, without thereby committing ourselves to the claim that it retains this content in all its occurrences. Indeed, with inanimates its semantic value may well be highly schematic and contribute very little to the meaning of the composite expression. But schematic conceptions are meanings nonetheless, and markers with schematic content are properly analyzed as symbolic units.
42 RONALD W. LANG ACKER Consider the noun endings -o and -a of Spanish. With animate nouns, they generally indicate 'masculine' and 'feminine' respectively; hence perro 'male dog' vs. perra 'female dog', muchacho 4boy' vs. muchacha 'girl', etc. The complex symbolic units representing the two forms for 'dog' are sketched in Fig. 12(a)-(b); note that each component morpheme contributes something substantial to the composite semantic value. How, then, do we analyze such nouns as palo 'stick' and mesa 'table', which have no natural gender but function grammatically as masculine and feminine parallel to perro and perra? 1 1 DOG i perr masculine]] -o 1 j Fdog" 11 perr FEMINII^Ej] -a I] (stick 1 pal ihingI j -o 11 1] TABLE 1 mes THINGM -a Figure 12 A possible analysis is given in Fig. 12(c)-(d). Let us assume that speakers segment -o and -a as gender markers. What semantic value do they attribute to these seemingly empty morphemes? Possibly their content, necessarily schematic with inanimate stems, is simply to be identified as [THING], the semantic pole of the noun schema (just as the content of the auxiliary verb do is the same as [PROCESS], the semantic pole of the verb schema). The content of the ending is then fully subsumed by the stem, since [THING] is schematic for both [STICK] and [TABLE] (and any other noun). By the schematic transparency principle, the composite structures [STICK-THING] and [TABLE-THING] are equivalent to the respective component structures [STICK] and [TABLE]; this would account for why the ending appears to be meaningless. Still, only symbolic units are employed in the description. Presumably a speaker of Spanish masters all the conventional units depicted in Fig. 12. Those in 12(a) and (b) indicate that the stem for 'dog' occurs with either ending to specify its "natural" gender. Those in (c) and (d) specify that 'stick' and 'table' take -o and -a respectively, though all the essential content is provided by the stem. The categorization of these nouns as grammatically "masculine" or "feminine"
AN OVERVIEW OF COGNITIVE GRAMMAR 43 is represented by these units themselves — separate diacritics or grammatical features are not required. Grammatically Judgments Since the earliest days of generative grammar, linguists have supported the autonomy thesis by asserting that speakers are capable of judging the grammatical well-formedness of sentences out of context and without regard for their meaning (cf. Chomsky 1957). Thus (21)(a) is accepted as grammatical despite its semantic anomaly, while (21)(b) is judged ill- formed. (21) (a) Colorless green ideas sleep furiously, (b) * Green furiously ideas colorless sleep. The same point is often made with Jabberwocky, where sentences are recognized as grammatical even when nonsense forms are used for the stems, so that only the "grammatical morphemes" are familiar. Let me state at the outset that the treatment of so-called "grammatical- ity judgments" in the generative tradition has been little short of scandalous, with respect to conceptual confusion in discussing the issue, the absence of any firm empirical basis for assessing grammaticality, and the dependence of sweeping theoretical claims on judgments that are in fact quite fragile and variable. We have no reason whatever to believe that natural-language well-formedness judgments resemble the prototype suggested by automata theory, where a mechanical procedure yields a discrete decision of "grammatical" or "ungrammatical" simply by scanning a string of vocabulary symbols devoid of content. The natural-language judgments reflected in the generative literature by asterisks and comparable notations are often not discrete (hence the presence vs. the absence of an asterisk is not sufficiently delicate to represent them), and almost invariably take semantic considerations into account in one way or another (regardless of whether they are recognized as such). These assessments are also quite sensitive to contextual factors. We should seriously consider Dinsmore's suggestion (1979) that sentences are always construed relative to some context, and that a "normal" context is implicitly invoked for supposedly aeon- textual judgments. Still, there is a legitimate difference between (21)(a) and (b) that has to be accounted for. What do such examples actually demonstrate? First, they
44 RONALD W. LANGACKER show that languages impose restrictions on the form of permissible expressions, restrictions that refer in some manner to grammatical classes (e.g. N, V, ADJ, ADV, NP). Second, they show that speakers are capable of assigning words to such categories even when their meanings, integrated according to the usual compositional patterns, fail to yield a coherent composite semantic structure. These observations do not, however, establish the autonomy of grammar (as earlier defined). We saw previously that the symbolic alternative accommodates both formal restrictions and membership in grammatical classes. At least in principle, a description employing only symbolic units is sufficient to handle the phenomenon. We have been assuming that basic grammatical categories are susceptible to notional characterization, and are represented in the grammar by highly schematic symbolic units (cf. Fig. 7). Constructions, i.e. conventional patterns for assembling symbolically complex expressions, are similarly described by schemas for such expressions. A constructional schema incorporates symbolic units as component structures and specifies how they integrate to form a composite structure (Figs. 8-11): at the semantic pole, it represents a pattern of semantic composition; at the phonological pole, it determines (among other things) the temporal sequencing of the component elements. The judgments in (21) thus reflect the availability in English of constructional schemas that permit the assembly of certain [ADJ ADJ N V ADV] sequences, and the unavailability of schemas that would sanction the construction of comparable sequences with the temporal ordering [ADJ ADV N ADJ V]. The former requires at least three constructional schemas: one permitting nominals of the form [ADJ ADJ N]; a second for the modifying relationship [V ADV]; and a third allowing the integration of the resulting composite structures at a higher level of organization. (This third schema actually allows the integration of a subject and a predicate regardless of their internal composition. For sake of discussion, let us assume — quite plausibly — the existence of a subschema that takes as components a subject nominal and a predicate with [ADJ ADJ N] and [V ADV] in particular as their respective internal structures.) The requisite constructional schemas (in compacted format) are sketched at the top in Fig. 13(a). The schema describing the relevant type of nominal expressions has for component structures two instances of the adjective schema and one of the noun schema. The second constructional schema takes the verb schema and the adverb schema as component structures. The third constructional schema incorporates the first two as its com-
AN OVERVIEW OF COGNITIVE GRAMMAR 45 ponents, and specifies their integration at a higher level of constituency. This entire complex structure (which contains only symbolic units) allows the computation of an open-ended class of instantiating expressions that are coherent at both the semantic and the phonological pole (e.g. Tired young boys sleep soundly). Of course, nothing guarantees that a speaker (or linguist) will always employ the symbolic resources at his disposal with semantic coherence as his ultimate objective. It is quite possible to choose lexical items which conform to local specifications of the overall schema on an individual basis, slot by slot, but which cannot be successfully integrated in the manner that the schema requires. (a) 1 ADJECTIVAL RELATION 1 x 1 ADJECTIVAL RELATION Y ' ICOLORIESS] [colorless] 1 thing] Z 1 ' GREEK green < PROCESS 1 W f IDEAS ideas] ADVERBIAL11 RELATION u HI f 1 sleep! [sleep] f [FURIOUSLY 1 [furiously) (b) colorless green ideas sleep furiously COLORLESS colorless GREEN, green IDEAS] ideas] ? 1 sleep furiously] / SLEEP sleep / \ FURIOUSLY' furiously Figure 13 Suppose, then, that the lexical units colorless, green, ideas, sleep, and furiously are selected to instantiate the lowest-level schemas, as indicated in Fig. 13(a). What happens when these units are integrated, level by level, in accordance with the specifications of the constructional schemas? The
46 RONALD W. LANGACKER result is sketched (using the exploded format) in Fig. 13(b). At the phonological pole, this computation yields precisely the sequence Colorless green ideas sleep furiously, henee our judgment that (21)(a) is formally correct. The outcome is quite different at the semantic pole, however. When one attempts to integrate the component semantic structures in the fashion specified by the constructional schemas, a coherent composite structure fails to emerge at any level of organization (this is because the lexical units selected attribute conflicting properties to entities that the schemas stipulate as being identical). An analysis employing only symbolic units is therefore capable of accounting for the fact that (21)(a) is semantically anomalous despite its formal correctness. It also accounts for the difference between (21)(a) and (21)(b) — in addition to its semantic anomaly, the latter violates the formal specifications of the constructional schemas with respect to temporal sequencing, and could not be computed from any other available schemas. The issue, then, is not the existence of ''grammatical patterns" and ''formal restrictions11, nor the possibility of an expression being recognized as conforming to the latter despite its semantic incoherence. Rather the issue is whether the representation of those patterns and restrictions requires anything other than symbolic units. Proponents of the autonomy thesis answer in the affirmative: in addition to meaning and phonological shape, they posit a separate dimension of "grammatical form1', whose primary ingredients are linear ordering, constituency, and labeling for membership in grammatical categories (cf. Fig. 4). The symbolic alternative does not deny the validity of these factors, but interprets them in a very different way. Linear ordering is simply temporal ordering, one dimension of phonological space. Constituency is a very general notion, amounting to nothing more than the order in which simpler structures are progressively assembled into larger ones; grammatical constituency thus reduces to the path of composition followed in the assembly of complex symbolic structures. Grammatical classes are defined by schematic symbolic units, and class membership is specified by categorizing relationships holding between schemas and their instantiations. The various elements of "grammatical form11 are recognized, therefore, but they do not constitute a separate, self- contained, or individually coherent structural domain — they are viewed instead as integral facets of symbolic structures and their interconnections. Languages provide for the symbolization of ideas by observable sequences of sounds. Wc must therefore acknowledge in some fashion the
AN OVERVIEW OF COGNITIVE GRAMMAR 47 reality of meanings, sounds, and the symbolic associations between them. The status of a distinct and autonomous domain of "grammar" is on the face of it less secure, for it is hard to envisage a realm of "grammatical content" analogous to semantic and phonological content. Undeniably, there are conventional patterns for the assembly of complex symbolic expressions out of simpler ones, and "grammar" is the obvious term for referring to them. But the reality of "grammar", so defined, does not itself establish the character of the elements that specify these patterns. In particular, it leaves open the possibility that these elements are themselves symbolic in nature, schematic units which embody the commonality of well-formed symbolic expressions and serve as templates for the computation of novel instantiations. What could be more natural and straightforward? If workable, this approach affords a coherent and integrated view of the various facets of linguistic organization, and permits the adoption of the highly restrictive content requirement. I consider it a curious anomaly in the history of our discipline that this natural conception must be explained and defended, while the autonomy thesis — with all its unnaturalness and conceptual obfusca- tion — commands widespread and often unquestioning allegiance. Conclusion Certain ways of viewing linguistic structure have become so standard and deeply ingrained in recent years that scholars tend to accept their validity as self-evident and to discount the potential viability of radical alternatives. My basic objective has been to outline a very different conception of linguistic organization, and to show what things look like when viewed in this perspective. I do not pretend to have demonstrated the incorrectness of currently predominant theories, or to have proved the non-autonomy of grammatical structure. I have nevertheless considered various phenomena commonly cited as supporting the autonomy thesis, and mapped out strategies for attacking them within the type of model proposed. I have also suggested the coherence, the naturalness, and even the restrictiveness of the descriptive framework. To be sure, this overview resembles grammar itself in being neither autonomous nor self-contained. It is schematic in its description of critical notions, such as imagery, categorization, constructions, the semantic definition of basic grammatical classes, and the usage- based approach. For elaboration and justification, I must refer the reader to other works on cognitive grammar.
48 RONALD W. LANGACKER I have emphasized two fundamental issues that underlie virtually every question of linguistic theory and determine how we talk and think about language structure: the nature of meaning, and the putative autonomy of grammatical structure. I have tried to indicate that these two issues are inextricably linked. Only with an impoverished, objectivist conception of meaning and linguistic semantics do the standard arguments for the autonomy thesis have any force. Given a more inclusive and linguistically appropriate conception of semantic structure, one which treats meaning as a subjective phenomenon and accommodates conventional imagery, these arguments are seen as invalid. For example, a subjectivist semantics makes possible a notional characterization of basic grammatical categories; the supposed impossibility of notional definitions is a major pillar supporting the edifice of autonomous grammar. Another major theme has been the inappropriateness of relying on absolute predictability as criterial for determining the nature of linguistic structure and organization. I have noted the prevalence of the type/predictability fallacy, which confuses the two distinct issues of (i) what kinds of structures there are, and (ii) the predictability of their behavior. A usage- based model, which accommodates the full complexity of linguistic knowledge by incorporating in a grammar both schematic structures and unit instantiations at various levels of abstractness, accounts for distribution without resorting to structures that violate the content requirement (which allows only semantic, phonological, and symbolic units). It is commonly assumed that the autonomy of grammar is established unless one can demonstrate that every peculiarity of grammatical structure and behavior is strictly predictable from independent factors (cf. Newmeyer 1983: ch. 1); under this assumption, grammar is non-autonomous only if we can "get it for free" as automatic consequences of something else. A severely reductionist program of this sort is doomed to failure; the grammatical patterns of a language are in large measure conventional and have to be specifically learned (regardless of their degree of naturalness and functional motivation). However, the conventionality of grammatical structure does not per se prove that it is autonomous in the sense of constituting a separate level or dimension of linguistic organization demanding special, purely "grammatical" classes and elements. I have shown that conventional patterns, including arbitrary restrictions on the form of expressions, are in principle describable using only symbolic units for the representation of grammatical structure.
A View of Linguistic Semantics Ronald W. Langacker University of California, San Diego The Nature of Meaning A language enables its speakers to effect an open-ended set of symbolic correspondences between meanings and phonological sequences. The set is open-ended because speakers learn not only a vast (though limited) inventory of conventional symbolic expressions, but also an array of patterns for successively combining simpler expressions into expressions of ever greater complexity. These patterns constitute the grammatical structure of the language. The theory of "cognitive grammar" (Langacker in press) is founded on the claim that grammatical structure itself is inherently symbolic, i.e. it is fully describable by means of symbolic elements, each with both semantic and phonological content (though this content may be quite abstract). On this view, grammatical analysis and semantic analysis are indissociable: a description of grammatical structure that makes no reference to meaning is ultimately no more revealing than a dictionary providing only a list of undefined forms. As a basis for grammatical analysis, consequently, this theory must offer at least a programmatic account of meaning and semantic structure. Our initial discussion will focus on five basic claims about the nature of linguistic meaning. These are summarized in (1). (1) Thesis A: Meaning reduces to conceptualization (mental experience). Thesis B: A frequently-used expression typically displays a network of interrelated senses. ® Ronald W. Langacker
50 RONALD W. LANGACKER Thesis C: Semantic structures are characterized relative to "cognitive domains". Thesis D: A semantic structure derives its value through the imposition of a "profile" (designatum) on a "base". Thesis E: Semantic structures incorporate conventional "imagery", i.e. they construe a situation in a particular fashion. We will consider these claims in the order listed, the later ones requiring more extensive discussion. While all of them are in some measure controversial, I am more concerned with articulating a coherent view than with arguing against alternatives or reviewing the history of pertinent issues. With Thesis A, I embrace a form of the "conceptual" or "ideational" theory of meaning that semantic theorists so commonly feel obliged to challenge (e.g. Lyons 1977: 1/113; Kempson 1977: 15-20; Palmer 1981: 24-29). Conceptualization is ultimately to be explicated in terms of cognitive processing; there is nothing inherently mysterious about its basic character, nor is cognition beyond the realm of scientific inquiry. I speak of "conceptualization" in order to emphasize the "subjective" nature of linguistic meaning. I understand the term in a maximally inclusive way with respect to the domain of mental experience: it subsumes both established concepts and novel conceptions; includes sensory, emotive, and kinesthetic sensations; and extends to our awareness of the physical, social, and linguistic context. Semantic structure is therefore treated as a special case of conceptual structure (as it is in Jackendoff 1983, which otherwise shows virtually no resemblance to the present framework in its view of meaning and grammar). Linguistic expressions (including the symbolic units of lexicon and grammar) are "bipolar", each having a "semantic pole" and a "phonological pole"; a "semantic structure" is thus definable as a conceptualization conforming to linguistic convention that functions as the semantic pole of an expression. I will also refer to semantic structures as "predications", and to the semantic pole of a morpheme (regardless of type) as a "predicate". Hence all predicates qualify as predications, but the predications corresponding to symbolically-complex expressions are not predicates. An important feature of cognitive-grammar semantics is that predications of all sizes are characterized by means of a single array of descriptive constructs. Thesis B simply recognizes that polysemy is the norm for lexical units, and must therefore be accommodated by linguistic theory as a natural,
A VIEW OF LINGUISTIC SEMANTICS 51 unproblematic phenomenon. I propose a network model, where each node in the network represents one established sense of the lexical item, and each arc connecting two nodes indicates the nature of their association. Nodes are associated via "categorizing relationships", of which there are two basic types: relationships of "schematicity" (indicated by solid arrows) and of "extension" (marked by dashed arrows). The notation [[A] —> [B]] says that [A] is "schematic" relative to [B], which constitutes an "elaboration" or "instantiation" of [A]; [B] is consistent with [A], but [BJ's specifications are more detailed and achieve a finer "grain" than [A]'s (hence the relationship is one of semantic "specialization", or its inverse, "abstraction"). By contrast, semantic extension implies some conflict in value; in the relationship [[A] -~> [B]], certain specifications of the basic sense [A] must be suspended or modified to arrive at the extended sense [B]. These notions are illustrated in Fig. 1, which shows a fragment of the network describing the conventional meanings of the English noun ring. The concept 'circular entity' is schematic relative to 'circular mark' and 'circular object'; the latter is in turn schematic relative to 'circular piece of jewelry', which is further elaborated by conceptions representing jewelry worn on different parts of the body. The term ring is also applied to an arena used in boxing, wrestling, or bullfighting; the sense 'arena' constitutes an extension relative to 'circular object' or 'circular entity' because it does not incorporate the specification of circularity (boxing and wrestling rings are typically rectangular). It is certainly not assumed that all these nodes and categorizing relationships are equal in status. The nodes vary greatly in their degree of entrenchment or cognitive salience, some being far more readily activated than others. In the case of ring, for instance, the sense 'circular piece of jewelry worn around finger' is presumably the category "prototype", and is likely to be activated in preference to others in a neutral context. Moreover, categorizing relationships vary not only in their cognitive salience, but also in their "distance" (i.e. the extent to which [A] must be extended or elaborated to yield [B]).
52 RONALD W. LANGACKER [CIRCULAR ENTITY \- ~_ _ ___ ~ ^ "" ""M ARENA I CIRCULAR MARK] [CIRCULAR OBJECTK CIRCULAR PIECE OF JEWELRYl CIRCULAR PIECE OF JEWELRY WORN AROUND FINGER CIRCULAR PIECE OF JEWELRY WORN THRU NOSE Figure 1 This network, when fully articulated, represents the conventional usage of the lexical item, i.e. the range of conceptions for which it is normally employed. The precise form of the network may vary across speakers, depending on their experience and the categorizing judgments they happen to have made, but this will hardly be a problem for communication provided that enough nodes are shared. Also, we have no immediate way of knowing the degree of "delicacy" that speakers achieve in their mastery of conventional usage (i.e. how far "downward" they articulate the network into progressively more specialized applications), nor the degree of schematization (how far "upward" they build the network through successive levels of abstraction). But the specific details are less important than the appreciation that some type of network must in fact be posited: any attempt to reduce the meaning of a lexical item to a single node will in general prove unworkable. The obvious candidates for such reduction are the category prototype and the highest-level schema, corresponding respectively to the primary sense and the most abstract sense. However, we cannot presume that every lexical category has a clearly identifiable prototype, nor that speakers invariably extract a "superschema" having all other nodes in the network as elaborations (none is shown in Fig. 1); often the necessary abstractness of an all-subsuming superschema would leave it essentially void of content. Furthermore, even if a prototype or superschema were always available, it would afford no basis for predicting precisely which array of extensions and elaborations — out of all those that are conceivable and semantically plausible — happen to be conventionally established in the language. The conventional usage of a lexical item must be learned, as
A VIEW OF LINGUISTIC SEMANTICS 53 it is not strictly predictable. A speaker's knowledge of the conventional meaning of a lexical unit therefore embraces the entire network, and must be characterized as such. Each node in such a network represents a distinct predication, and together with the phonological pole it defines a distinct "semantic variant" of the lexical item. Nevertheless, since polysemy is not our immediate concern, the remainder of our discussion will focus on individual nodes. What is necessary to characterize a lexical item (or some novel expression) when it is understood in one particular way? Cognitive Domains According to Thesis C, linguistic predications are characterized relative to "cognitive domains". The basic observation is that certain conceptions presuppose and thus incorporate others, which provide the necessary basis for their emergence and characterization. For instance, the notion hypotenuse presupposes the conception of a right triangle, being incoherent without it; it thus incorporates this conception as a necessary part of its own characterization, i.e. as a cognitive domain. Similarly, the notion corner kick presupposes, as one of its domains, a considerable body of knowledge pertaining to the rules and objectives of football. We can therefore reasonably posit hierarchies of conceptual complexity, where a concept at one hierarchical level derives through various cognitive operations performed on concepts at lower levels, as sketched in Fig. 2. These operations, indicated by double arrows, include the coordination of lower-level concepts (i.e. the integration of their contents), and also adjustments pertaining to profiling and other dimensions of imagery, as discussed in later sections. For example, by coordinating multiple instances of the notions line segment and angle, we obtain the conception triangle. That, together with the concept of perpendicularity, yields the notion right triangle, which functions as a cognitive domain for hypotenuse. This latter notion is in turn a potential domain for such higher-order conceptions as midpoint of the hypotenuse, and so on indefinitely.
54 RONALD W. LANGACKER w A B y C increasing complexity- Figure 2 The central claim is that a conceptualization occupying any level in such a hierarchy can function as a cognitive domain for the characterization of a linguistic predication. In Fig. 2, for instance, [A] is depicted as a complex notion serving as the domain for predication [B], obtained by performing an operation on [A] and coordinating it with additional content. However, since a predication is itself a conceptualization, [B] can itself be invoked as the cognitive domain for predication [C], derived by further operations. The import of this claim is that the starting point for the semantic analysis of a given expression is an integrated conceptualization which may have any degree of internal complexity. Obvious though it may seem, this type of account differs significantly from certain others that are commonly assumed. First, it contrasts with the view that the meaning of an expression is directly describable in terms of a fixed vocabulary of semantic primitives; a cognitive domain need not be primitive, nor is there any fixed, limited set of them. Second, it rejects the idea that an expression's meaning is represented as a bundle of semantic markers or features; a cognitive domain is an integrated conceptualization in its own right, not a feature bundle. What occupies the lowest level in hierarchies of conceptual complexity? I do not necessarily posit conceptual primitives, being essentially neutral as to their existence. It is however necessary to assume some inborn capacity for mental experience, i.e. a set of cognitively irreducible representational spaces or fields of conceptual potential; I refer to these as "basic domains". Among these basic domains are the experience of time and the ability to conceptualize configurations in 2- and 3-dimensional space. There are basic domains associated with the various senses: color space (i.e. a range of possible color sensations), the ability to perceive a
A VIEW OF LINGUISTIC SEMANTICS 55 particular range of pitches, domains defining possible sensations of taste and smell, and so on. Moreover, special domains are doubtless required for emotive and kinesthetic sensations. I offer no specific inventory of basic domains — the important point is simply that certain types of mental experience are cognitively irreducible. Presumably we begin constructing our mental universe out of experience registered in basic domains, arriving at ever higher levels of conceptual organization by means of innately specified cognitive operations. Some linguistic predications may be characterized directly with respect to basic domains; possible examples are red (color space), hot (temperature), line (2-dimensional space), and before (time). However most predications presuppose domains representing higher levels of conceptual organization. For instance, the conception of a right triangle, the domain for hypotenuse, is non-basic (or "abstract") because it is reducible to more fundamental notions; the fact that a right triangle occupies 2-dimensional space (and thus incorporates this basic domain) does not obviate its conceptual (hence non-basic) status. Similarly, the noun tip presupposes the conception of an elongated (generally thin) object, an elbow is characterized in part by its position within the overall configuration of the human arm, April invokes for its definition the complex notion of a cycle of months exhaustively dividing a year, the verb castle presupposes substantial knowledge of the rules and strategies of chess, and so forth. There is obviously no fixed or universal set of cognitive domains. Any facet of our mental experience or conceptual world can be appropriated to function in this capacity. In principle, then, to describe a predication exhaustively one would have to furnish complete descriptions of the cognitive domains it presupposes. This would in turn require characterizations of the more fundamental conceptions that each domain itself presupposes, and so on, until finally one reached the level of basic domains. One would also need to characterize the various operations and abilities by means of which these hierarchies of conceptual complexity are constructed during the course of cognitive development. In short, a complete and definitive account of semantic structure would require an essentially full description of developmental cognition and the conceptual hierarchies it produces! This is obviously not a practical short-term goal. In practice we must do the best we can with partial and provisional accounts, describing relevant domains to the best of our ability in whatever degree of detail is minimally necessary for particular purposes. It is important, though, to realize
56 RONALD W. LANGACKER what a cognitively accurate description of semantic structure would actually require were it attainable. There is no point in deceiving ourselves by maintaining the gratuitous belief that meaning can be understood without a comparable understanding of cognitive processing and the conceptual hierarchies and knowledge systems to which it gives rise. Semantic structure is not a separate or autonomous "module" of psychological organization that can be pulled out and studied in isolation from the full, rich fabric of our mental experience. Viewing semantic structure as a self-contained, algorithmically describable component of the linguistic system is theoretically convenient, but it necessarily impoverishes and distorts the subject matter in drastic ways. Another consideration leading to a similar conclusion pertains to the number of cognitive domains presupposed by a given predication. The full set of domains required for a predications's description will be called its "matrix". The matrix is said to be "complex" if it consists of more than one domain, as it typically does. Even with basic domains, one easily finds examples where a single domain is not sufficient. For instance, a beep is a sound of short duration that to some extent approximates a pure tone; a characterization of this notion therefore presupposes the basic domains of both time and pitch. A flash is a brief, intense, and possibly diffuse light sensation; the description of this noun consequently refers to time, color space (particularly the brightness dimension), and the extensionality of the visual field. For predications invoking non-basic domains, the matrix is often quite large, and sometimes open-ended. The complex matrix of knife is sketched in Fig. 3. One specification of this predicate is a characterization of the typical shape of a knife (or a family of typical shapes), presupposing the basic domain of space. Another salient specification is the role a knife plays as an instrument of cutting; the conception of a canonical episode of cutting therefore functions as a second, non-basic domain. Other specifications, each invoking some non-basic domain, pertain to the normal size of a knife; its weight; the material from which it is made; the inclusion of a knife (together with a fork and a spoon) in a standard place-setting; the games played with knives; their cultural associations; the existence of knife-throwing acts in circuses; the role of knives in the conquest of the American frontier; and so on indefinitely. What ties these diverse conceptions together is the participation in each of them of the entity that knife "designates" (see below).
A VIEW OF LINGUISTIC SEMANTICS 57 SPACE DOMAIN 1 _j_— r^n CUTTING DOMAIN 2 "-•••... 0 i w \ H1 SILVERWARE *•• DOMAIN 3 COMPLEX MATRIX | KNIFE Figure 3 A fundamental question of semantic theory arises at this juncture. Which of the specifications in a complex matrix are part of the conventional meaning of a linguistic expression, and which ones represent extra-linguistic knowledge? In fact, can such a distinction be made at all? The standard view is that it can. Theorists generally assume that the meaning of a word like knife is not to be equated with a speaker's vast and open-ended knowledge of the objects in question, but rather with some restricted portion of this overall body of knowledge; the linguistic description of a term's conventional value more closely resembles an entry in a dictionary than an entry in an encyclopedia. A definite division is therefore presumed possible between linguistic and extra-linguistic knowledge (or between "semantics" and "pragmatics"). The problem with this view is that the facts of language offer little basis for accepting its validity. Haiman (1980) has surveyed the obvious rationales that might be invoked for drawing the line between the linguistic and extra-linguistic knowledge associated with a term, and found them all to be inadequate. In truth, the existence of a clear-cut boundary has been assumed on methodological (not factual) grounds: only under this assumption can semantic structure (and more generally, linguistic structure) be described as a self-contained system amenable to algorithmic formal treatment. Theoretical convenience does not, however, constitute empirical evidence. I see no a priori reason to accept the reality of the semantics/pragmatics dichotomy. Instead, I adopt an "encyclopedic" conception of linguistic semantics. I posit no specific boundary between our linguistic and non-linguistic knowledge of the entity designated by a term, such that all those specifications on one side of the boundary clearly fall within the purview of seman-
58 RONALD W. LANGACKER tics, while all those on the other side are safely relegated to pragmatics. Far more realistic, I believe, is to posit a gradation of "centrality" in the specifications constituting our encyclopedic knowledge of an entity: some domains and specifications are obviously more salient and linguistically important than others, and in practice we will focus our attention primarily on these, but the imposition of any precise or rigid boundary is considered arbitrary. A number of factors contribute to a specification's degree of centrality, among them whether the specified property is inherent or contingent, whether it is generic or specific, whether it is peculiar to the entity in question or shared by many others, and so on (for fuller discussion, see Langacker in press: ch. 4). At the level of cognitive processing, centrality can be explicated as the likelihood that a particular specification will be activated on a given occasion when a term is used. The prevalence of the "dictionary" view of linguistic semantics is attributable in part to the ubiquitous "container metaphor", which permeates our thought about language and is reflected in how we talk about it (cf. Reddy 1979). This metaphor portrays lexical items as "containers" for an abstract substance called "meaning"; speakers exchange meanings by sending strings of such containers back and forth. Any actual container has a definite and limited volume, so the metaphor encourages us to think of a lexical item as having some restricted set of semantic specifications — we in fact have some initial difficulty in conceptualizing the meaning of a term as extending indefinitely. We must not allow the formulation of linguistic theory to be dictated by a metaphor of dubious validity. As an alternative to the container metaphor, we can profitably regard a lexical item as providing "access" to knowledge systems of open-ended, encyclopedic proportions. Speakers do not send meanings to one another (all they transmit, in fact, is sound waves). Rather, their detection of a lexical item in the speech stream enables them to activate selected portions of the knowledge base they already possess; communication is possible to the extent that the knowledge systems of different speakers are comparable, but it is never perfect, since these systems are never fully identical. Profile and Base Thesis D holds that the imposition of a "profile" on a "base" is crucial to the value of every linguistic predication. The base for a predication is nothing more than its matrix, i.e. the set of operative domains (or more
A VIEW OF LINGUISTIC SEMANTICS 59 precisely, those portions of such domains which the predication actually invokes and requires). Some facet of the base is invariably raised to a distinctive level of prominence, and serves (intuitively speaking) as its focal point; this substructure is the predication's profile. I do not know whether profiling reduces to any independently attested psychological phenomenon. Factors that suggest themselves as possibly being relevant include figure/ ground organization, focus of attention, and level of activation, but the simple identification of profiling with any such notion is problematic. Whatever its cognitive basis turns out to be, profiling is essential and fundamental to linguistic semantics. I will further describe the profile as comprising those portions of the base which the predication "designates". (Designation, then, is not to be understood as the relation between a linguistic expression and the "world", but rather one that holds between an overall conceptualization and certain of its subparts.) For instance, the conception of a right triangle functions as the base for hypotenuse, whose profile (or "designatum") is one of the constituent line segments, as illustrated in Fig. 4(a). When predications are diagrammed, the profiled entity is depicted with heavy lines to indicate the special prominence that distinguishes it from the remainder of the base. (a) (b) (c) HYPOTENUSE Figure 4 Profiling is one of the cognitive operations responsible for the growth of hierarchies of conceptual complexity (cf. Fig. 2): a predication presupposes and incorporates its base, from which it derives a distinct, higher- order conception through the imposition of a particular profile. An expression's meaning does not, then, reside in either the base or the profile alone. Both facets of a predication are crucial to its value, which reflects the elevation of some particular part of the base to a special level of prominence and
60 RONALD W. LANGACKER functional significance. For instance, if we suppress the profiling of hypotenuse, as in Fig. 4(b), what results is no longer the conception of a hypotenuse, but simply that of a right triangle; if we suppress the unprofiled portions of the base, as in Fig. 4(c), there is no basis for identifying the remaining line segment as being a hypotenuse, which exists only in the context of a right triangle. The base of a predication can thus be thought of as the "frame" needed to establish the character and identity of the intended designatum: a person qualifies as a cousin only when linked to another individual through a series of kinship connections; a span of idle time constitutes an intermission only if it interrupts some type of performance; and only the extremity of an elongated object can be recognized as a tip. All expressions, regardless of complexity, are characterized semanti- cally by the imposition of a profile on a base. A pattern of grammatical composition must specify, at its semantic pole, precisely which entity (out of those contributed by the component expressions) is selected as the profile of the composite expressions it derives. What is traditionally known as the "head" of a given construction is simply that component from which the composite expression inherits its profile (and hence its grammatical categorization). For example, consider the nominal expressions in (2), assuming that each of them invokes for its base the conceived situation sketched in Fig. 5. (2) (a) the lamp above the table (b) the table below the lamp (c) the leg of the table below the lamp (d) the light from the lamp above the table Figs. 5(a)-(d) represent the composite values of (2)(a)-(d) respectively. These expressions contrast semantically, but not by virtue of any differences in their conceptual content — we have assumed a common base, incorporating the conception of a lamp (hence light), a table (hence legs), and the locative relationship between the two. The semantic contrasts are attributable primarily to the imposition of alternate profiles on this base; starting from the same basic conception, each expression selects a different substructure as the one it designates. It is readily observed that the entity designated by the expression as a whole is in each case the same one that is profiled by the head noun. The compositional pattern permitting a noun to be modified by a prepositional phrase describes the integration of their contents to form a coherent composite conception, and further specifies that
A VIEW OF LINGUISTIC SEMANTICS 61 the composite-structure profile is the same entity that is designated by the noun. (c) r W s Figure 5 The expressions just considered have all been nominal in character, but the notion of profiling (designation) is also considered applicable to relational predications (including verbs, adverbs, adjectives, prepositions, and others). A relational predication profiles the "interconnections" among conceived entities, where an "entity" can be either a "thing" or some other relation; interconnections can be thought of as cognitive operations assessing the relative position of entities within a domain. For exemplification, let us compare the meanings of go, away, and gone, as reflected in (3): (3) (a) You've been here long enough — please go now. (b) California is very far away. (c) By the time I arrived, she was already gone. All these words are polysemous, but we will in each case confine our attention to the single sense illustrated. As a verb, go designates a "process", i.e. a series of relational configurations followed sequentially through conceived time. Time is therefore one domain in its matrix, as indicated by the arrow in Fig. 6(a), and space is another. The diagram shows explicitly only four component "states" — including the initial and final states — out of the continuous series that the process comprises. There are two major participants, represented diagram- matically by circles. I refer to one of them as the "trajector" (tr), and to the other as the "landmark" (/m). The dotted lines signify "correspondence"; they indicate that the trajector is the same from one state to the next, as is the landmark. The heavy, dashed lines connecting the trajector and landmark within each state represent the profiled interconnections responsible for the relational character of the predication, i.e. the cognitive operations that register the position of the trajector within the domain relative to that
62 RONALD W. LANGACKER of the landmark. The initial state finds the trajector in the landmark's "neighborhood", given as an ellipse. The trajector's position changes from state to state with the passage of time, and in the final state it lies outside the neighborhood of the landmark. (a) GO Cb) AWAY (c) GONE Figure 6 Away is not a verb, and time is not an active domain for the semantic variant under examination. In (3)(b), away profiles a single spatial configuration, diagrammed in Fig. 6(b). Observe that this configuration is the same one that constitutes the final state in the profile of go: it situates the trajector outside the vicinity of the landmark. Hence the process designated by go results in a locative relationship appropriately described by away. However, we see from examples like (3)(b) that this relationship need not be construed as coming about by any motion of the trajector (presumably California is stationary for the time period under consideration). What about gone? An examination of Fig. 6(c) reveals that it differs from both go and away, but resembles each of them in a certain respect. Gone matches away in profiling a single locative relationship in which the trajector is outside the neighborhood of the landmark. These two predications differ, however, in their base: for the relevant sense of away, the base is simply some portion of the spatial domain, but the base for gone is precisely the process profiled by go. That is, something is properly described as being gone only if its position results from an instance of the process of going, hence the conception of this process constitutes its base. Go and gone are therefore identical in their basic conceptual content — they differ in the profile imposed on this content. This difference in profiling is the semantic contribution of the past participial predication (which itself has several semantic variants — see Langacker 1982a). The participial morpheme takes a schematically-characterized process as its base, and profiles
A VIEW OF LINGUISTIC SEMANTICS 63 only the final state of this process. The participial construction identifies the schematic base of the participial morpheme with the specific process profiled by the verb stem, but specifies that the profile of the former prevails at the composite-structure level (i.e. the participial morpheme is the "head" in this construction). In short, the effect of the participial inflection, at the semantic pole, is to convert a process like go into a non-processual relation that profiles only the final state of that process. Imagery The remainder of this paper deals with Thesis E. It claims that semantic structures incorporate conventional "imagery", where that term is understood in a special technical sense. By imagery, I do not refer to visual imagery or sensory imagery more generally, as studied by Kosslyn (1980), Shepard (1978), and others (though I believe this phenomenon to be both cognitively and semantically significant). I refer instead to our amazing mental ability to "structure" or "construe" a conceived situation in many alternate ways. In view of its obvious significance, it is surprising that this ability has received so little systematic attention. I would argue, in fact, that imagery is crucial to a revealing account of either semantic or grammatical structure, and provides the necessary foundation for a subjectivist theory of meaning. For the description of lexical and grammatical structure, cognitive grammar posits only symbolic units, each attributed some measure of conceptual "content". This content, moreover, is structured in a particular fashion, i.e. the unit incorporates conventional imagery, which contributes to its semantic value. Because of imagery, two expressions that have the same content, or describe the same objective situation, may nevertheless have different meanings; recall, for example, that go and gone have the same content (cf. Fig. 6), but differ semantically due to profiling. The lexical and grammatical resources of a language are therefore not semantically neutral — inherent to their nature is the structuring of conceptual content for symbolic purposes. I should emphasize that the distinction between content and imagery is drawn in part for expository convenience, and should not be thought of as a strict dichotomy. These two facets of meaning grade into one another and may well be indissociable: every conception reflects some particular construal of its content, if only by the selection of default- case options.
64 RONALD W. LANGACKER Linguistic semantics must therefore accommodate conventional imagery in a natural and integral manner. Imagery cannot be treated as an afterthought — its various dimensions are intrinsic to the conceptualization process, and a linguistic predication, by its very formation, assumes particular values along these parameters. Some important dimensions of imagery are catalogued in (4); they will be discussed in the order listed: (4) (a) level of specificity (b) background assumptions and expectations (c) secondary activation (d) scale and scope of predication (e) relative salience of substructures (f) perspective Some of these dimensions are further divisible into factors requiring separate consideration. No substantive claim is intended by either the groupings or the sequencing. The first dimension of imagery pertains to the degree of precision and detail with which an entity is characterized. Here the indissociability of imagery from content is quite apparent: in an obvious sense, a schematic notion has less content than one spelled out with finer "grain" or greater "resolution". Level of specificity has already been noted as one parameter of lexical polysemy; certain conventional senses of a lexical item are commonly schematic (or conversely, specialized) relative to others (cf. Fig. 1). The same type of relationship often holds between distinct lexical items, in which case it is common to speak of "hyponymy" or "taxonomic hierarchies". We thus find ordered series of terms like those in (5)(a)-(b), where the solid arrows indicate relationships of schematicity. (5) (a) move -> locomote -» run —» sprint (b) animal —» reptile —» snake —> rattlesnake —» sidewinder (c) long —> over 6 feet long —> about 6-1/2 feet long —> precisely 6 feet 7 inches long I also consider each expression in (5)(c) to be schematic for the one that follows: a schema introduces certain domains (a scale of length in this case), and specifies a relatively broad range of possible values within those domains; its instantiations confine the permitted values to narrower ranges (and may themselves introduce further domains). We see from this example that level of specificity, like all the other dimensions of imagery, is applicable to novel expressions of any degree of complexity, not just lexical items.
A VIEW OF LINGUISTIC SEMANTICS 65 Whenever a speaker constructs a sentence, he necessarily opts to portray the conceived situation with a certain degree of specificity. For example, he might use any of the sentences in (6) to describe the same event: (6) (a) I saw an animal and moved on. (b) I saw a long snake and ran away. (c) I saw a rattlesnake about 6-1/2 feet long and sprinted to safety. While the last sentence is certainly the most dramatic, a speaker more concerned with preserving a reputation for fearlessness might well prefer the greater schematicity of (6)(a). Relationships of schematicity also play a significant role in grammatical constructions. A construction is a pattern for integrating two or more "component" structures to form a "composite" structure. Patterns of this sort are described in cognitive grammar by means of "constructional schemas": complex symbolic structures that are fully parallel in formation to instantiating expressions, but characterized at a level of schematicity that neutralizes their differences. As the embodiment of a structural regularity, a constructional schema can be regarded as a template for the computation of novel instantiating expressions, which are obtained by elaborating its schematic components. For example, the structure on the left in Fig. 7 sketches the semantic pole of the constructional schema describing the integration of a modifying adjective with a head noun (in a full description, relations between the component and composite structures would similarly be given'for the phonological pole — in the case of English, they specify that the adjective immediately precedes the noun). A specific expression such as long snake is judged well-formed by virtue of instantiating the schema, and the categorizing relationship between the two constitutes the expression's "structural description".
66 RONALD W. LANGACKER CONSTRUCTIONAL INSTANTIATING SCHEMA EXPRESSION ADJ-N u-~© _y\ lm tr ; ADJ —^ O N LONG-SNAKE 1 I /\ LONG —^- fj]t SNAKE 1 Figure 7 In a typical construction, schematicity also figures in the syntagmatic relationship between the two component structures. Consider the constructional schema of Fig. 7. The nominal predication profiles a "thing", indicated by a circle. The adjectival predication is relational, profiling the interconnections between its trajector, also a "thing", and some kind of landmark. Every construction hinges on correspondences established between subparts of the component structures; in this ADJ + N construction, the pivotal correspondence holds between the trajector of the adjective and the thing profiled by the noun. The dotted correspondence line thus identifies the adjectival trajector with the nominal profile, but the adjectival predication generally characterizes this entity only schematically, while the noun is far more specific about its properties; this additional information is labeled 'X' in the diagram. There is consequently a relationship of schematicity between the trajector of ADJ and the profile of N, as indicated by the solid arrow. The composite structure, shown at the top, retains the profiling of the noun, but to its characterization of the profiled entity is added — as part of the base — the participation of this entity in the relation designated by the adjective. In the particular case of long snake, the adjectival predication has for its domain the conception of a length scale. The relational interconnections
A VIEW OF LINGUISTIC SEMANTICS 67 specify that the trajector, when aligned with this scale, extends into the region that lies beyond the neighborhood of the norm (n) for its category; this region constitutes the relational landmark. Long characterizes its trajector quite schematically. This schematic trajector is placed in correspondence with the specific "thing" designated by snake, which provides much fuller information about the correspondent, i.e. the noun elaborates the schematic trajector of the adjective. The resulting composite structure profiles the snake, but incorporates in its base the information furnished by the adjective concerning its length. Observe that the instantiating expression is precisely parallel to the constructional schema in the structures it incorporates and the relationships among them. In particular, the component and composite structures of long snake elaborate their respective schematic counterparts, and the integration of long and snake follows the pattern represented in the schema. The second dimension of imagery, construal relative to different background assumptions and expectations, will not be discussed in any detail. It subsumes a variety of phenomena, including the familiar contrast between (7)(a) and (b): (7) (a) The glass is half-empty, (b) The glass is half-full. These sentences are true under the same objective circumstances, but certainly they differ in meaning (cf. Tuggy 1980). Though I will not consider it here, presupposition certainly falls under the rubric of background assumptions. The same is true of discourse phenomena, such as the topic-comment asymmetry, the given-new distinction, and the occurrence in English of contrastive (or unreduced) stress on those elements considered informative or otherwise noteworthy: (8) (a) He likes LIVER. (b) He LIKES liver. (c) HE likes liver. The base of a predication can also be regarded as the background for viewing its profile. Recall that away and gone profile the same relational configuration, but construe it against different bases. One additional example is the semantic contrast between few and a few. The sentences in (9) could perfectly well be used to describe the same situation (let us suppose that precisely three friends are involved in each case), but they clearly differ semantically.
68 RONALD W. LANGACKER (9) (a) He has few friends in high places, (b) He has a few friends in high places. On intuitive grounds, one is inclined to say that few is somehow negative, while a few is basically positive. This is corroborated by the data in (10); only few provides the negative context required by any: (10) (a) Few people have any friends in high places, (b) * A few people have any friends in high places. I suggest that few specifies the negative departure of a quantity from an implicit norm or expectation; for instance, (9)(a) indicates that the number of friends is less than a well-connected person ought to have. By contrast, a few carries no special assumptions, and simply specifies that the quantity departs somewhat from the baseline of zero. The next dimension of imagery, secondary activation, pertains to the network model of complex categories, as illustrated in Fig. 1. The network model has general applicability, and is employed for such disparate classes as the alternate senses of a lexical item, the allophones of a phoneme, and the family of complex structures — including schemas, subschemas, and instantiating expressions — representing the conventional range of a grammatical construction (cf. Langacker in press: ch. 10). The network describing a complex category consists of a set of structures (nodes) connected by categorizing relationships. The structures may have any degree of internal complexity (e.g. an entire constructional schema might function as a single node). In the case of lexical polysemy, our prime concern here, the nodes comprising a network are symbolic units, which are distinct at the semantic pole but identical at the phonological pole (note that only the semantic pole is depicted in Fig. 1). Each node thus constitutes a semantic variant of the lexical item. For the most part, only one semantic variant is directly relevant when a speaker employs a particular lexical item in a specific context. Suppose a term is needed to describe the arena for a boxing match. In this context, the symbolic unit [ARENA/ring] is far more relevant than [CIRCULAR MARK/ring], [CIRCULAR OBJECT/ring], or even the prototype [CIRCULAR PIECE OF JEWELRY WORN AROUND FINGER/ring], despite its peripheral status in the ring category. The variant [ARENA/ ring] is the one that the speaker activates when he decides to use the term: he matches its semantic specifications against his conception of the entity to be described, and accepts the term as appropriate if this categorizing judg-
A VIEW OF LINGUISTIC SEMANTICS 69 ment reveals these notions to be sufficiently compatible. The node invoked in this fashion for primary categorization of the target conception will be called the network's "active node". Clearly, different nodes serve this function on different occasions. We can reasonably assume that activation tends to spread across associated cognitive structures (cf. Collins and Loftus 1975). Hence the primary activation of a network's active node tends to elicit the secondary activation of other nodes linked to it by categorizing relationships. While the active node determines the basic applicability of a term, the additional nodes activated secondarily contribute their own nuances and thereby enrich the semantic value that the term acquires on a particular occasion of its use. Let us assume, for example, that [ARENA/ring] is categorized as a semantic extension from the more basic variant [CIRCULAR OBJECT/ ring]. In this case, the primary activation of the extended variant in describing a boxing arena is likely to evoke the secondary activation of the basic variant. The subsidiary activation of [CIRCULAR OBJECT/ring] reinforces and thus renders more salient the notion of an enclosure completely surrounding the locus of activity. There is no such effect when the term arena is used instead, since [ARENA/arena] is not directly linked by any categorizing relationship to [CIRCULAR OBJECT/ring] and does not elicit its secondary activation. Metaphorical expressions are simply more extreme instances of semantic extension. For instance, the conventional usage of pig to designate a glutton implies the semantic variant [GLUTTON/pig], which is categorized as an extension from the basic variant [PIG/pig] and evokes its secondary activation. The same phenomenon is observable with morphemically complex expressions, e.g. established idioms like spill the beans or take the bull by the horns. When these expressions are used idiomatically, the extended or "figurative" sense functions as the active node — it represents the actual notion to be conveyed — while the basic or "literal" sense is activated secondarily. This two-level semantic representation accounts for the special quality of metaphorical expressions, but the enhancement of meaning by the activation of subsidiary nodes is present to some degree in virtually every expression. The metaphorical use of novel expressions has the same fundamental character, the only difference being that the extended variant and categorizing judgment have not yet achieved the status of conventional units.
70 RONALD W. LANGACKER Scale and Scope of Predication The "scale" of a predication is roughly analogous to the scale of a map. Many predications presuppose for this parameter a value that falls within a certain range. For example, a body of land completely surrounded by water is called an island, but this term would not normally be used for a handful of mud in the middle of a puddle — it is only applicable to situations on a larger scale. In similar fashion, the term continent is applied to land masses on the scale of Australia, Europe, and North America, but hardly to Ireland, Tahiti, or Luxembourg. The contrast between bay and cove is also one of scale: whereas either might refer to precisely the same land/water configuration, cove is limited to small-scale configurations, for which bay is marginal. Scale of predication is important even for relational expressions like near and close, which may assume essentially any value along this parameter. An expression of the form A is near B or A is close to B is equally appropriate for galaxies, nations, objects in a room, and atoms in a molecule, indicating in each instance that A lies in the neighborhood of B. But what constitutes a neighborhood? The absolute distances involved vary enormously from one example to the next. Obviously, the neighborhood conception must be characterized in terms of relative distances assessed with respect to relationships on a particular scale. The "scope" of a predication is its coverage in relevant domains, i.e. how much is included in the conception that functions as its base. Every predication has some implicit scope, though it need not be precisely delimited or cognitively salient. Consider the land/water configuration depicted in Fig. 8, where the outer box and the dashed lines indicate alternate scopes of predication. Scope (a) is sufficient to identify the land mass as an island, but (c) and (d) are not; to qualify as an island, a land mass must be completely surrounded by water, and neither (c) nor (d) is inclusive enough to provide this information. What about scope (b), which contains only the land together with a narrow strip of water all around its periphery? Interestingly, we could not properly identify the land as an island with only this much coverage; it is possible, for instance, that the strip of water is a moat protecting a castle, but it is not conventional to speak of the land inside a moat as an island. The water surrounding an island must have a certain expanse — it must extend outward from the island a substantial distance (at least on most sides). Though the scope of predication is not sharply delimited, it must at least be large enough to reveal this expanse. As another
A VIEW OF LINGUISTIC SEMANTICS 71 example, consider the lexeme peninsula. In this case we find that scopes (a) and (c) suffice, while (b) and (d) are inadequate for different reasons. The insufficiency of scope (b) is similar to that observed for island: a certain minimal expanse is required for the body of water into which a peninsula projects. Scope (d) fails because the finger of land constituting a peninsula must be fairly small relative to the overall land mass of which it is a part; scopes (a) and (c) are sufficient to establish the relative sizes, but (d) is not. Figure 8 The scope construct has substantial grammatical significance (cf. Casad and Langacker 1985). For one thing, the scope of a predication can be referred to explicitly in the characterization of a grammatical construction. Consider the "nested locative" construction illustrated in (11). (11) (a) The rake is in the yard by the back fence near the gate, (b) The camera is upstairs in the bedroom in the large closet on the top shelf underneath the quilt. Each sentence contains a series of locative expressions, which collectively describe the location of the entity designated by the subject. Under the relevant interpretation, each successive locative in the series narrows the "search" for this entity, confining it to a region that is properly included in the region singled out by the locative that immediately precedes. Following Hawkins (1984), I will use the term "search domain" to indicate the region
72 RONALD W. LANGACKER to which a locative predication confines its trajector (subject). Fig. 9 diagrams the syntagmatic relations connecting the three locatives of (ll)(a). Each locative has a scope of predication, and within that scope it specifies the position of the trajector relative to a landmark. It is mandatory in this construction that the trajector be the same for all the locatives in the sequence; the identity of the trajector from one predication to the next is indicated by dotted correspondence lines. Cross-hatching is used to mark the search domain: each locative tells us that the trajector is somewhere within this area. The first one, in the yard, confines the trajector to the interior of its landmark (namely, the yard). The other two landmarks, by the back fence and near the gate, are "neighborhood" predications; each situates the trajector in the vicinity of the landmark identified by its prepositional object. , s< locative 1 0997777791 :ope 1 1m locative 2 V/////4W//'*'' scope 2 locative 3 wht^im ' scope ^§W Figure 9 The "nested" character of these locative specifications derives from a second set of correspondences, marked by the two lower dotted lines. They specify that the search domain of one locative constitutes the scope of predication for the next. Hence the interior of the yard, the search domain of the first locative, functions as the scope of predication for the second, i.e. only the interior of the yard is considered for purposes of identifying the back fence in question and assessing the proximity relationship. The neighborhood of the back fence is the search domain of the second locative, and consequently the scope of predication for the third: the gate to which the trajector is proximate must be part of the back fence in particular, not some other fence outside this scope (observe that the gate may well be con- textually unique — as signaled by the definite article — only with respect to scope 3). The composite structure that results from integrating these locatives in the specified manner finds the trajector located simultaneously in all three nested search domains, each calculated with respect to a different landmark. We see from this example that highly abstract and cognitively
A VIEW OF LINGUISTIC SEMANTICS 73 non-salient constructs, notably scope of predication and search domain, are capable of participating in correspondences that are pivotal to a grammatical construction. When scopes of predication are nested in this fashion, and an entity of concern is simultaneously located in all of them, I will refer to the innermost scope (i.e. the most restricted one) as the "immediate" scope of predication. With respect to (ll)(a) and Fig. 9, the locative relationship near the gate has scope 3 as its immediate scope of predication; scope 3 figures directly in the computation of this proximity relation. By contrast, scopes 2 and 1 figure only indirectly in the characterization of near the gate: they provide the initial basis for identifying scope 3, hence scope 3 mediates their role in the predication. This nesting of scopes (with special status accorded the immediate scope) not only arises through syntagmatic combination, but is observable as well in the relationships among distinct expressions. This phenomenon is nicely illustrated by English body-part terms. It is important not to confuse the hierarchical part/whole relationships of a complex entity with the conceptual hierarchy that corresponds to it — the former pertains to the structure of the entity itself, and the latter to the structure of our conceptualizations. Thus the fact that arms, legs, head, and torso are parts of the human body does not entail that the concepts arm, leg, head, and torso are conceptual "parts" out of which the higher-order concept body is assembled. Indeed, it is more plausible to assume precisely the opposite: a crucial specification of the concept arm is the position of the designated entity within the overall configuration of the human body, hence the notion arm presupposes and incorporates the notion body, rather than the converse. In the relevant hierarchy of conceptual complexity, body is arguably more fundamental than arm, leg, etc., and functions as a conceptual component of these higher-order notions. I therefore assume that a rough gestalt representing the overall shape of the human body constitutes one domain in the matrix for such predicates as arm, leg, head, and torso, which profile particular substructures within this base. To be sure, the matrices for these terms contain many other domains, pertaining to substance, function, and so on, but these can be ignored for present purposes. The notion body is thus analyzed as the base and scope of predication for the higher-order concepts arm, leg, etc. Continuing along the conceptual hierarchy, each of these concepts functions in turn as the scope of predication for the characterization of additional terms (defined on a smaller scale) designating their subparts; the conception of an arm, for instance, is
74 RONALD W. LANGACKER the base for describing such higher-order notions as elbow, forearm, and hand. The conception of a hand then provides the scope of predication for palm, thumb, and finger, with finger giving rise to such small-scale notions as knuckle, fingertip, and fingernail. The conceptual hierarchy is therefore defined by nested scopes of predication. For knuckle, the sequence is body > arm > hand > finger, with finger the innermost layer, i.e. the immediate scope. It is intuitively obvious that knuckle is characterized directly with reference to the conception of a finger, and progressively more indirectly with respect to notions farther away along the hierarchy. It is not that these more distant notions are totally irrelevant — after all, finger presupposes hand, which presupposes arm, and so forth; all these notions are within the scope of predication, broadly construed. Clearly, though, finger is the most prominent within knuckle, and provides the immediate basis for its characterization. These constructs have linguistic consequences, one of which is illustrated in (12): (12) (a) A finger has 3 knuckles and 1 nail. (b) ?? An arm has 14 knuckles and 5 nails. (c) ??? A body has 56 knuckles and 20 nails. In sentences with have that describe part/whole relationships, there are differences in felicity that reflect the putative conceptual hierarchy (cf. Bever and Rosenbaum 1970; Cruse 1979). Specifically, such a sentence is most felicitous when the subject provides the immediate scope of predication for the object, and declines in felicity (other things being equal) as the distance between them along the hierarchy increases. Compounds designating body parts are subject to a similar restriction: (13) (a) fingertip; fingernail; toenail; eyelash; eyelid (b) *bodytip; *armnail; *footnail; *headlash; *facelid The terms in (13)(a) are the conventional English expressions for the body parts in question. Note that the first member of the compound is in each instance the immediate scope of predication for the second (in the case of eyelash and eyelid, eye must be interpreted as designating the entire eye region, not the eyeball in particular). The compounds in (13)(b) are to be construed as designating the same body parts as their respective counterparts in (13)(a). In (13)(b), however, the first member of the compound is not the immediate scope of predication for the second, but is farther removed along the appropriate conceptual hierarchy. Not only are these
A VIEW OF LINGUISTIC SEMANTICS 75 compounds non-conventional, but they seem strikingly unnatural — if new terms were needed for these body parts, the ones in (13)(b) would hardly suggest themselves as suitable candidates. While it is conceivable that languages might occasionally conventionalize such expressions, I would expect them to be relatively infrequent. Relative Salience of Substructures Unfortunately, the claim that a particular element is "salient" or "prominent" relative to others is much easier to make than to verify. We have no way at present to measure objectively the relative salience of the various facets of a conceptualization. However, detection by direct and objective means is hardly a realistic criterion for the adoption of linguistic constructs, and theorists seldom hold themselves to this standard (note the abstract, indirect basis for positing such entities as traces and zero pro forms in government-binding theory). In contrast to many theoretical constructs, the idea that varying levels of prominence distinguish the subparts of linguistic structures is both cognitively plausible and intuitively natural. Still, the notion of salience is overly vague if left unexplicated, being applicable to numerous linguistic phenomena of diverse character. To render the notion somewhat more precise and substantive, we can attempt to distinguish particular types of linguistic salience, and to examine as best we can their basis and structural significance. Just three types of prominence will be considered, the first of which — profiling — has already been discussed. As the "focal point" within a predication, the profile (i.e. the substructure that is designated) is prominent relative to the remainder of the base. A predication's profile determines its basic grammatical category: a nominal expression profiles a "thing" (defined abstractly as a "region in some domain"), while such classes as verbs, adjectives, and adverbs profile different sorts of relations. In a complex expression, that component which contributes its profile to the expression as a whole can be identified as the "head". Snake is thus the head in the nominal expression long snake, since the composite structure designates the reptile, not the relationship of length (cf. Fig. 7). A second type of salience pertains to relational predications. With few if any exceptions, relational predications manifest a subtle asymmetry in how they portray their participants: one participant is always singled out for special prominence — this is the one that I have labeled the "trajector" in
76 RONALD W. LANGACKER previous diagrams (Figs. 6 and 7). Though I will not pursue the issue here in any detail, there is some basis for supposing that the special status of the trajector is a matter of figure/ground organization (cf. Langacker in press: ch. 6); if so, a trajector can be characterized as uthe figure within a relational profile". I use the term "landmark" for salient participants other than the trajector. If there are multiple landmarks, the most prominent among them is called the "primary landmark". The terms trajector and landmark are inspired by prototypical predications of motion and physical activity (e.g. go, run, hit), where the trajector is the mover and the landmark serves as a point of reference for specifying its path. Observe, however, that the definition of these constructs refers only to relative prominence. The trajector/landmark asymmetry underlies the subject/object distinction, but they are not equivalent. The terms "subject" and "object" are generally reserved for overt nominals (i.e. noun phrases) that have a particular role in clause-level organization. The trajector/landmark asymmetry differs in several respects. First, it pertains to a relational predication's internal structure; substructures are identified as trajector and landmark regardless of whether these participants are spelled out overtly by accompanying expressions. In Fig. 7, for example, the adjectival predicate long has both a trajector and a landmark — respectively, a thing and a region along the length scale — even though only the former is capable of being elaborated by an accompanying noun. Second, the trajector/landmark asymmetry is observable at any level of organization, not just the clause level. With minor exceptions (e.g. prepositional objects), only verbs are traditionally regarded as having subjects and objects; yet verbs, adjectives, adverbs, and prepositions are all attributed trajectors and landmarks, regardless of whether they function as clausal heads. Third, only nominal expressions are normally considered to be subjects or objects, but the trajector or the landmark of a relational predication can itself be a relation. In fact, having a relational trajector is what distinguishes adverbs from adjectives. We may summarize as follows. Relational predications profile interconnections among conceived entities, some of these entities being recognized as central "participants". One aspect of conventional imagery is the level of prominence accorded the various participants in the construal of a relationship: one participant, termed the trajector, achieves the greatest prominence by virtue of being selected as the figure within the relational profile; to some extent, the remaining participants (landmarks) can them-
A VIEW OF LINGUISTIC SEMANTICS 77 selves be ranked for salience. These participants, with their differing degrees of prominence, characterize the internal structure of a relational predication, regardless of type or level of grammatical organization. Often a relational predication occurs with another expression in a grammatical construction serving to render one of its participants overt; specifically, the relational participant is placed in correspondence with the profile of the accompanying expression, and is typically elaborated by this expression (e.g. the schematic trajector of long in Fig. 7 corresponds to the profile of snake, which elaborates this trajector). If the relational predication happens to be a clausal head, and the elaborating expression a nominal, the latter is traditionally recognized as a subject or direct object. A subject is thus characterized in this framework as a nominal expression that corresponds to the trajector of a clausal head, and a direct object as one that corresponds to its (primary) landmark. What could be the conceptual basis for the trajector/landmark asymmetry? If we seek a schematic definition of trajectors, i.e. one that is applicable to all relational predications at all levels of organization, we obviously cannot focus on such notions as "agent", "controller", and "topic", which are commonly proposed for the characterization of prototypical clausal subjects. A universally valid definition must be sufficiently abstract and flexible to accommodate sentence pairs like (14)-(16): (14) (a) The painting is above the light switch, (b) The light switch is below the painting. (15) (a) A trumpet resembles a cornet, (b) A cornet resembles a trumpet. (16) (a) Iraq invaded Iran. (b) Iran was invaded by Iraq. All three pairs display a semantic contrast that I attribute to the choice of trajector within the clausal head (this trajector being elaborated by the subject nominal), yet there is no action in (14) or (15) (hence no agent or controller), while in (16) the trajector is the agent in one instance and the patient in the other. In fact, we observe the trajector/landmark asymmetry even with verbs, like resemble, that designate a symmetrical relationship. It appears, then, that the trajector/landmark asymmetry is essentially independent of a predication's content, i.e. the choice of trajector is not strictly predictable on that basis. Nor does the trajector/landmark asymmetry reduce to discourse notions such as "topic", for it is observable even in the
78 RONALD W. LANGACKER semantic contrast between pairs of words such as above and below, considered in isolation and out of context. For several reasons, defining a trajector as the figure within a relational profile is reasonable as a working hypothesis. First, the importance of figure/ground organization as a cognitive phenomenon is well established (see Talmy 1978b and Wallace 1982 for discussion of its linguistic applicability). Also, the variability and independence from content observable in the choice of relational trajector is mirrored by our ability to make alternate selections of figure within a given scene. Moreover, figure/ground organization affords a plausible analysis of problematic examples like (14)- (16). It nicely explicates the semantic contrast observable in (14), if we assimilate the notion "landmark" or "reference point" to the more inclusive notion "ground". Specifically, in (14)(a) the light switch is construed as a point of reference (landmark) for locating the painting, while the converse is true in (14)(b). Though the sentences in (15) pertain to more abstract relationships, the semantic contrast is precisely parallel: (15)(a) takes a cornet as the standard of comparison (abstract reference point), and evaluates a trumpet with respect to this standard; (15)(b) reverses these roles. For situations involving motion or other activity, the mover or the actor is most naturally selected as the figure; linguistically, the same entity represents the unmarked choice for trajector and subject, as seen in (16)(a). Marked constructions allowing some other participant as trajector and subject, such as the passive in (16)(b), can then be explicated as instances of figure/ground reversal. Finally, there is a notable cross-linguistic tendency for the participant most readily selected as figure on perceptual grounds to represent as well the unmarked choice of trajector in a relational predication. This is true not only for action sentences, but also for expressions like above and below that describe static configurations. Intuitively, it is evident that above is in some sense the unmarked (or "positive") member of the pair; in similar fashion, over and in front of are unmarked relative to under and in back of, respectively. In each instance, the unmarked member is the one whose trajector would normally be the most salient perceptually: something is more easily perceived and selected as figure within the scene when it is above, over, or in front of something else than when it is below, under, or in back of it.
A VIEW OF LINGUISTIC SEMANTICS 79 (a) (b) i V E K T 1 O" i i i o [ HORIZ ABOVE BELOW Figure 10 Such pairs also demonstrate that linguistic semantics cannot be based on content alone. Everyone agrees, for instance, that above and below are semantically distinct, yet they appear to have the same content — even the same profiling — as sketched in Fig. 10. The domain is oriented space, i.e. space organized into the horizontal and vertical dimensions. Each predication profiles the interconnections between two entities in this domain, which are differentially located along the vertical axis. Where, then, does the semantic contrast lie? It can only lie with the choice of trajector, identified as the upper entity in the case of above, and the lower one for below. Apart from the figure/ground organization imposed on the relational profile, the two predications are basically identical. The last type of salience to be considered is a concomitant of "analyza- bility". A complex expression is "analyzable" to the extent that speakers are cognizant of the contributions of component morphemes to the value of the composite whole. Analyzability need not take the form of conscious awareness (though it often does); it is sufficient that the component structures are activated at some level of processing when the composite structure is constructed or manipulated. Novel expressions are necessarily analyza- ble, but once a complex expression has the status of a familiar unit, it is conceivable that a speaker might activate its composite structure independently from its components. Intuitively, fixed expressions show a gradation of analyzability. Speakers definitely recognize the contribution of complain to complainer, but the presence of compute in computer lies at the margins of awareness, and propel may not be perceived at all in propeller. 1 i V E K T Otr 1 1 1 0lm i [ HORIZ
80 RONALD W. LANGACKER The analyzability of an expression affects its semantic value. I would argue, for example, that pork contrasts semantically with pig meat, father with male parent, gravel with pebbles, and oak with oak tree, even if we ignore any differences in connotation or information content that might distinguish the members of each pair. Specifically, I suggest that analyzability augments the salience of the semantic elements profiled by the component structures. If we assume for sake of discussion that pork and pig meat have precisely the same composite structure, they nonetheless differ in meaning because the latter expression provides individual symbolization to the conceptual components pig and meat, thereby rendering these notions more salient within the composite whole than they would otherwise be. As a derived plural, pebbles highlights the fact that the designated mass consists of multiple instances of the category pebble, whereas the monomorphemic gravel conveys comparable information about its constituency but fails to emphasize it in any way. The semantic effect of analyzability becomes more pronounced as the paraphrases become more elaborate. Thus triangle is quite different in meaning from three-sided polygon despite their identical extensions: the latter expression calls explicit attention to a factor remaining latent in the former, namely the membership of the designated entity in a larger class of geometrical figures. And few would want to claim that sink has precisely the same semantic value as passively descend through a medium under the force of gravity (even assuming the accuracy of the paraphrase). If the semantic import of analyzability is acknowledged, how is it to be accommodated? We need only recognize that the meaning of a complex expression does not reside in the composite structure alone, despite its privileged status; also included, as a subsidiary aspect of an expression's meaning, is the compositional path through which the composite structure is progressively assembled. Suppose, for example, that an expression has the composite semantic structure [XYZ], where X, Y, and Z are distinct (though possibly overlapping) facets of this conception. There are many potential compositional paths for arriving at [XYZ], some of which are diagrammed in Fig. 11; each path represents a slightly different meaning, despite the identity of content at the composite-structure level. The path in 11(a) corresponds to a monomorphemic expression, in which the entire structure [XYZ] is symbolized as an unanalyzed whole by the phonological sequence [w]. The paths in (b) and (c) correspond to bimorphemic expressions, where [X] is individually symbolized in one instance, and [Z] in the
A VIEW OF LINGUISTIC SEMANTICS 81 other. All three facets of the composite structure are separately symbolized in (d), which represents a trimorphemic expression with internal constituency. Because they have the same composite structure, the four expressions schematized in Fig. 11 are paraphrases in the sense of conveying the same content and designating the same entity. Yet they contrast semantically, their overall semantic values incorporating the following sets of structures: (a) [XYZ]; (b) [X], [YZ], [XYZ]; (c) [XY], [Z], [XYZ]; (d) [X], [Y], [Z],[YZ],[XYZ]. (a) XYZ1 00 (c) XYZ wz XY] [Z] ~w"j |zj Figure 11 (d) "XYZj xyz |xj [yzI I rYjfzi LElEj An expression's composite structure is thus construed against the background provided by its path of assembly; its meaning embraces the entire, hierarchically-organized set of component structures through which the speaker arrives at the composite conception. Although this is quite an elaborate definition of meaning, there are specific advantages to adopting it. For one thing, it resolves a classic problem of truth-conditional semantics that arises in regard to semantic anomaly. Because an anomalous sentence has no truth conditions, all such sentences are semantically equivalent. The sentences in (17) are thus treated as synonymous: (17) (a) * Existence spanked a yellow aroma, (b) * Colorless green ideas sleep furiously. Barring special interpretations of the component lexemes, such sentences are certainly incoherent, but speakers do not readily accept either their meaninglessness or their semantic identity. Furthermore, sentences like (18) are semantically well-formed (and certainly non-equivalent), despite their incorporation of an anomalous constituent:
82 RONALD W. LANGACKER (18) (a) It is nonsensical to speak of a yellow aroma. (b) It is nonsensical to speak of colorless green ideas. This is problematic for traditional compositional semantics, since the anomaly of a constituent prevents the emergence of well-formed semantic representations at higher levels of composition. These problems essentially disappear with a conceptualist view of meaning acknowledging the semantic contribution of compositional paths. A phrase like yellow aroma is indeed anomalous; when the two words are combined in accordance with the proper constructional schema, conflicting properties are attributed to corresponding entities that the schema specifies as being the same (cf. Fig. 7), hence a coherent composite conceptualization fails to emerge. However, an inconsistent conceptualization is still a conceptualization, and the expression does have a compositional path (incorporating the meanings of the individual words) even if it terminates in a composite structure that is empty or incoherent. Moreover, yellow aroma and colorless green ideas contrast semantically because their compositional paths incorporate semantically distinct components. The coherence of sentences like (18) can also be explained when meaning is identified with conceptualization. Some conceptions refer to other conceptions as substructures, and nothing prevents a defective conception from being referred to in this way. If an anomalous conception occurs in conjunction with a predication like nonsensical, which specifically comments on such anomaly, the overall conception achieves a certain coherence. Sentences that predicate the anomaly of a specific notion such as yellow aroma need not themselves be any more anomalous than words, like nonsensical, which refer schematically to an anomalous conception as part of their internal structure. Defining an expression's meaning as including its entire compositional path at the semantic pole has the further advantage of accounting for the otherwise inexplicable semantic contrast between certain pairs of expressions. Consider the words unrecover and reuncover, which feel a bit awkward but nonetheless conform to established derivational patterns. Speakers agree that they are semantically distinct: roughly, unrecover means 'undo the act of recovering', while reuncover means 'repeat the act of uncovering'. (Note that recover is to be interpreted as 'cover again', not 'get back'.) Yet the two expressions, at the composite structure level, have precisely the same base, and profile the same substructure within this base. The semantic distinction cannot be explained unless their compositional paths are included as part of their meanings.
A VIEW OF LINGUISTIC SEMANTICS 83 > (I) > 1 (II) 1 > (III) (iv) ; ?R0FII£D EVENT BASE Figure 12 The base for each word comprises a sequence of four events, labeled I- IV in Fig. 12. The first event consists in the placement of a cover on some object, and the second, in its removal; events III and IV repeat the cycle. Events I-IV must all be included in the base because they provide the minimal context required for the proper identification of the profile (event IV). To be in a position to unrecover something, i.e. to undo the act of recovering it, one must already have recovered it; this implies an original act of covering (I), a second act of covering (III), and some intervening event (II) permitting the covering to be repeated. Unrecover thus evokes the entire sequence of events in Fig. 12, though it profiles only the last. What about reuncover? To reuncover something, one must first uncover it (II) and then repeat this act (IV), and under normal circumstances each act of uncovering requires a previous act of covering to undo (I and III). Hence reuncover also presupposes the entire sequence of events, and designates only the last. Consequently, the semantic contrast between unrecover and reuncover cannot be attributed to either their base or their profile at the composite structure level. But obviously, they arrive at this complex conception through different compositional paths. Each path begins with the root cover, whose content is limited to the first event, and terminates with the composite structure, whose content includes the full sequence; let us therefore represent these structures as [I] and [I-II-III-IV], respectively. The difference between the two paths lies in the intermediate-level structure resulting from the initial derivational process. In the case of unrecover, this intermediate structure corresponds to the derived stem recover, with the content [I-II-III]. For reuncover, on the other hand, the intermediate structure corresponds to the stem uncover, whose content is simply [I-II]. Distinct sets of structures thus constitute the compositional paths of the two
84 RONALD W. LANGACKER expressions. Leaving aside the meanings of the individual derivational morphemes, the path consists of [I], [I-II-III], [I-II-III-IV] for unrecover, and [I], [III], [I-II-III-IV] for reuncover. Perspective The final dimension of imagery, perspective, subsumes a number of more specific factors: orientation, vantage point, directionality, and how subjectively or objectively an entity is construed. All these factors imply a viewer, or more abstractly, a conceptualizer whose construal of the conceived situation can be thought of as roughly analogous to a perceptual relationship (cf. Langacker 1985). As a default-case option, the viewer (or conceptualizer) is identified with the speaker (and secondarily, the addressee). The notion of orientation is largely self-evident, so I will say little about it. Certain predications, most obviously those involving left and right, crucially presuppose a particular orientation on the scene being described. Thus, if V is the viewer in Fig. 13, and the arrow indicates the direction in which he is looking, he would describe X as being to the left in the context of 13(a), but to the right in 13(b). Many predications rely on a canonical orientation that is independent of the viewer's immediate circumstances. For instance, the domain of oriented space — space organized into the horizontal and vertical dimensions — reflects the typical orientation of human observers in relation to the surface of the earth. This domain is invoked for predications like above and below (cf. Fig. 10), whereby one entity can be characterized consistently as being above or below another whether the viewer observes the scene while upright, standing on his head, or lying on his side. (c) i i 1 0 ! Figure 13 The notion of vantage point is similarly straightforward. Clearly, the use of left and right in Fig. 13 is dependent on the viewer's vantage point as
A VIEW OF LINGUISTIC SEMANTICS 85 well as his orientation. Noteworthy here is the phenomenon of "mental transfer" (Vandeloise 1984), i.e. a speaker's ability to mentally assume a vantage point other than his actual one for purposes of calculating a particular relationship. For example, from his actual vantage point the viewer in Fig. 13(c) would describe X as being to the left of Y, but he could also, by assuming Y's vantage point and orientation, say that X is to the right of Y. The same phenomenon is well known from sentence pairs like (19), either of which might be uttered by a speaker in San Diego talking on the phone to somebody in Chicago: (19) (a) I will go to Chicago tomorrow, (b) I will come to Chicago tomorrow. In (19)(a), the speaker describes the envisaged motion from his own vantage point, whereas in (19)(b) he adopts for this purpose the vantage point of the addressee. In some languages, the notion of vantage point is of greater systematic importance than it appears to be in English. A case in point is Cora, a Uto- Aztecan language of Mexico (cf. Casad 1982a; Casad and Langacker 1985). Cora has an elaborate system of locative particles and verb prefixes, some combination of which occurs in virtually every sentence, and vantage point is often essential to their characterization. A particularly striking example is provided by the sentences in (20), both of which refer to a dog: (20) (a) u-h-k*-tya-pu'u inside-vertical-short-middle-planted 'Its tail is cut short.' (b) a-h-ki-tya-pu'u outside-vertical-short-middle-planted 'Its tail is cut short.' Despite the identical English translations, these sentences are semantically distinct, and reflect the vantage point of the speaker: the first would be spoken by someone viewing the dog directly from behind, and the second when the dog is viewed from the side. The way in which the speaker's position is indicated proves both interesting and instructive. The locative elements in (20) are the contrasting prefix combinations u-h- and a-h-. They belong to a larger set of combinations describing topographic relationships, which are often extended to characterize relationships in non-topographic domains. In its basic topographic sense, -h- locates
86 RONALD W. LANGACKER some entity in the face of a slope (as opposed to the foot or the head of a slope); by extension, it locates an entity on any vertical surface. The prefixes u- and a- are ubiquitous, and form a contrastive pair with a considerable array of specific senses. The glosses 'inside' and 'outside' reflect their prototypical values. In topographic expressions, these glosses are still accurate, but must be construed relative to a very special domain. This domain involves the conception of a viewer (generally the speaker) standing at the foot of a slope and looking straight up the face of the slope toward the top. With respect to this abstract domain, we can define an entity describable as the "line-of-sight region": this region includes a relatively narrow corridor that runs along the viewer's line of sight up to the crest of the slope (and also subsumes the viewer's immediate surroundings on all sides). It is this subjectively-determined line-of-sight region that is pivotal to the topographic value of u- vs. a-, which are relational predications taking this region as their landmark. Specifically, u- indicates that the trajector falls 'inside' the line-of-sight region, and a-, that it is located 'outside' this region. The basic topographic values of u-h- and a-h- are consequently 'inside the line of sight in the face of a slope' and 'outside the line of sight (i.e. off to the side) in the face of a slope', respectively. Figure 14
A VIEW OF LINGUISTIC SEMANTICS 87 By extension, these prefix combinations apply to non-topographic domains, where they pertain to any sort of vertical surface. In (20), the vertical surface in question is some facet of a dog's torso, which is represented as a block in Fig. 14. Fig. 14(a) corresponds to (20)(a), in which the dog is viewed from behind. The trajector of the prefix combination is a relation, namely that of the tail being short; the use of u- indicates that this trajector falls within the viewer's line of sight extending upwards along a vertical surface. Because a Cora knows where a dog's tail is attached, he can therefore determine that the viewer is examining the dog from the rear. In (20)(b), on the other hand, the viewer is examining the dog from the side, as sketched in Fig. 14(b). The occurrence of a- indicates that the trajector (i.e. the tail configuration) lies outside the line-of-sight region on a vertical surface, hence the viewer cannot be examining the dog end-on. He must instead be looking at it from the side. Let us turn now to directionality, the penultimate factor listed under the heading of perspective. The main observation I wish to make is that the relevance of this notion is not limited, as one might expect, to sentences describing physical or even abstract motion. Directionality is common even in sentences describing static configurations of indefinite duration, as we see from (21): (21) (a) The roof slopes upward at a steep angle, (a') The roof slopes downward at a steep angle. (b) The road widens just outside the town, (b') The road narrows just outside the town. (c) The hill gently rises from the bank of the river, (c') The hill gently falls to the bank of the river. (d) The line from point A to point B is straight, (d') The line from point B to point A is straight. The sentences in each pair are semantically distinct despite their applicability to precisely the same objective situations. Moreover, it is intuitively clear that the meaning contrast hinges on directionality. But if nothing is conceived as moving, how does this sense of directionality arise? Evidently, it reflects the manner in which the conceptualizer (notably the speaker) construes the conceived situation: at some level of processing, the conceptualizer mentally "scans through" the static scene in a particular direction. Hence the feeling of directionality does not reflect the contents
88 RONALD W. LANGACKER of the conceptualization, but rather how it is mentally accessed; ultimately, it stems from the sequencing of cognitive events. As it stands, of course, this proposal is quite vague. A substantive analysis will have to provide a broadly-grounded body of descriptive constructs allowing the origin of the intuitively-perceived directionality to be made explicit in a motivated way (some preliminary proposals are made in chs. 3 and 4 of Langacker in press). Still, directionality must be recognized as an important facet of imagery, however dimly we understand the underlying mechanisms. The last factor subsumed under perspective pertains to how subjectively or objectively an entity is construed. In this context, the terms "subjective", "objective", and their derivatives must be understood in a particular sense (I am not presently referring to the general claim that meaning is a subjective phenomenon). The construal of an entity is said to be purely "subjective" when it functions solely as the conceptualizer, and not at all as an object of conceptualization. Conversely, the construal of an entity is purely "objective" when it functions solely as the object of conceptualization and is fully distinct from the conceptualizer. The only pure case of a subjectively-construed entity is consequently the conceptualizer himself when he is totally absorbed in the conception of some other entity, to the extent of losing all awareness of himself and even of his own role in the conceptualization process. Whether the construal of any entity is ever purely subjective or purely objective is immaterial — subjectivity is a matter of degree, and the pure cases are primarily invoked as reference points on what is probably a continuous scale. These notions are subtle and perhaps elusive, but their ultimate importance for semantic and grammatical structure is hard to overstate (see Langacker 1985 for extensive discussion). By way of brief illustration, let us consider their application to the problem of directionality. Sentence (22)(a) is a typical description of physical motion: the entity designated by the subject moves along the spatial path specified by the locative phrases. (22) (a) A boy walked across the field, through the woods, and over the hill, (b) There was a fire last night across the river, through the canyon, and over the mountain. Though a comparable series of locatives appears in (22)(b), there need be no motion involved; the speaker may simply be indicating where the fire occurred, with no thought of anybody going there. The noteworthy feature
A VIEW OF LINGUISTIC SEMANTICS 89 of (22)(b) is that the locative sequence is identical in form to one that describes an extended spatial path, yet its function is evidently to identify a single, point-like location. I analyze the locative series in both sentences as describing an extended spatial path — what differs is not the path, but the nature of the movement that follows it. The movement in (22)(a) is concrete and objectively construed: the mover is an entity (the boy) fully distinct from the con- ceptualizer (the speaker), as is the spatial path itself; the motion is concrete in that the mover physically occupies the successive points that constitute the path. By contrast, the movement in (22)(b) is abstract and subjectively construed. The path itself is only partially objective, for it extends from the speaker to the location of the fire, and is thus not fully distinct from the speaker. It is the speaker who moves along this path, but this motion is conceptual rather than physical; it consists in the speaker successively activating his conception of the ordered points or segments that the path comprises. Starting from his own position, the speaker mentally constructs a progressively more elaborate path conception whose endpoint, upon completion, is identified with the location of the fire. Why is this motion subjective? Because the speaker is not concerned with his own movement along the path — indeed, he does not think of himself as moving at all. Only from the external perspective of the analyst is the speaker construed objectively as following a path. From his own perspective, the speaker is simply conceiving and expressing the location of the previous night's fire. The motion, in other words, is not itself the object of conception, but rather an inherent facet of the conceptualization process. It lies much closer to the subjective than to the objective end of the scale, since it represents activity by the con- ceptualizer in his role as such. A similar analysis is proposed for the sentences in (21). Each of them describes a complex spatial configuration, and we can reasonably suppose that a speaker builds up to this complex conception by successively and cumulatively activating various facets of it. The order in which he does so (at some stage or level within the overall processing activity) determines the perceived directionality of the corresponding expression. In the case of (21)(d), for instance, the speaker starts with point A, and progressively builds up the conception of a line extending away from it, until the conception so arrived at includes point B as its terminus; in (21)(d'), he proceeds in the opposite direction. Once more, the speaker can be characterized as moving mentally along the directed path, but this motion is both abstract
90 RONALD W. LANGACKER and subjectively construed: what the sentences actually describe and designate is the locative configuration, not the speaker's construal of it. Conclusion The ideational view of meaning is looked on with little favor by mainstream semantic theory. I contend, however, that the heavy emphasis in current theory on truth conditions and rigorous formalization stems from a priori methodological considerations, not from any evidence that a self- contained, objectivist semantics is optimal or even appropriate for natural language. I have outlined a subjectivist approach to semantics that, while admittedly programmatic, nevertheless goes considerably beyond the simple, essentially vacuous assertion that the meaning of a word is a concept (of undefined character). For an ideational account to be substantive, one must actually describe the internal structure of the concepts at issue in an explicit, descriptively adequate manner. I have attempted to catalog some of the factors that such an account must accommodate. In particular, linguistic semantics cannot confine itself to the information content of expressions; our capacity for imagery — the ability to construe a conceived situation in alternate ways — is a fundamental aspect of meaning that is indissoc- iable from conceptual content. The view of meaning that emerges from these considerations is complex and elaborate. The phenomena we have examined will severely test the adequacy of any comprehensive model of cognitive processing.
The Nature of Grammatical Valence Ronald W. Langacker University of California, San Diego Preliminary Notions Cognitive grammar seeks to characterize a speaker's grasp of established linguistic convention. This knowledge is assumed to take the form of a "structured inventory of conventional linguistic units", where a "unit" is a structural complex that the speaker has fully mastered. Only three basic types of units are posited: "semantic", "phonological", and "symbolic". Symbolic units are "bipolar", with a semantic unit at one pole in symbolic correspondence to a phonological unit at the other. The inventory of conventional units is "structured" in that some units function as components of others. The vowel [e], for instance, is one component of the phonological unit [[b]-[£]-[djl, which is in turn a component of the symbolic unit [[BED]/ [[b]-[e]-[d]]],andsoon. Three facets of the model must be explicated in slightly greater detail: schematicity, grammatical structure, and semantic structure. "Schematic- ity" is the relation between superordinate and subordinate nodes in a taxonomic hierarchy. For example, the concept [TREE] is schematic relative to more highly elaborated concepts such as [OAK] and [PINE]; superordinate to phonological units like [a] and [i] we can posit the schematic phonological unit [VOWEL]. A schema is said to be "elaborated" by the structures subordinate to it, called its "instantiations", and this elaborative relation is indicated by a solid arrow, e.g. [TREE] —> [OAK]. Relatively speaking, a schema specifies a notion only in gross terms, while its instantiations specify it in finer detail — it is like the difference between a graph of the stock market plotted on a coarse grid, showing only general trends from Ronald W. Langacker
92 RONALD W. LANGACKER month to month, and a graph plotted on a fine grid, showing day-to-day fluctuations with greater precision. The conventional units of a grammar include both schemas and their instantiations. Schemas have a categorizing function and also a sanctioning function in the creation of novel expressions. Relations of schematicity also play a role in grammatical valence, which will be our basic concern here. Cognitive grammar does not posit special morphological or syntactic units. Grammatical structure, both morphology and syntax, is claimed to be symbolic in nature, forming a continuum with lexicon. It is therefore accommodated by symbolic units, each consisting of a semantic and a phonological pole. Grammatical patterns, or "rules", are represented in the form of schematic symbolic units, parallel in structure to the specific expressions that instantiate these patterns. These schematic units capture generalizations by representing the commonality of instantiating expressions, and also function as templates that a speaker can employ in constructing and evaluating novel expressions. For a concrete example, consider the plural noun beds. What must we say to fully describe this complex structure? First we must characterize the two component morphemes, each a symbolic unit with a semantic and a phonological pole. The second task is to specify precisely how the two morphemes combine — this is the problem of grammatical valence. For now I simply observe that the combination is bipolar and symbolic. It is bipolar in the sense that we must state not only how the phonological units [bed] and [-z] integrate to form the composite phonological structure [bed-z], but also how the semantic units [BED] and [PL] integrate to form the composite semantic unit [BED-PL]. Moreover, the phonological integration symbolizes the semantic integration: the fact that the plurality designated by [-z] pertains to [BED] in particular (rather than some other nominal concept) is conveyed specifically by the fact that [-z] suffixes to [bed] rather than some other stem. In short, the semantic relation between the meaning components [BED] and [PL] is symbolized by a specific relation (one of suffixation) between their respective phonological representations.
THE NATURE OF GRAMMATICAL VALENCE 93 SEMANTIC POLE ! PHONOLOGICAL POLE Figure 1 Fig. 1 summarizes these relations of symbolization and integration. It shows an integrative relation between the semantic units [BED] and [PL], to yield the composite semantic structure [BED-PL], and also between the phonological units [bed] and [-z] to yield the composite phonological structure [bed-z]. It shows symbolic relations between [BED] and [bed], between [PL] and [-z], and between the modes of integration of [BED] and [PL] at the semantic pole and of [bed] and [-z] at the phonological pole. The composite structures [BED-PL] and [bed-z] stand in a symbolic relationship derivative of the ones linking their components and the integration of these components. This may strike one as an overly complicated way to describe a simple case of suffixation, but in actuality it is not. I have only made explicit reference to all the facets of structure and organization that go into a stem-affix combination, since these facets must ultimately be described in full detail rather than being taken for granted, as they normally are. An account of each of the structures and relations in Fig. 1 is the minimum required for an adequate description of a simple instance of grammatical valence between two morphemes. Fig. 1 of course represents a specific instance of the general pattern of plural formation with [-z]. The pattern itself — the cognitive grammar equivalent of a morphological or derivational rule — is simply a schematic version of Fig. 1, where the specifications of bed are replaced by those of the count-noun schema. The schema for the morphological construction coexists, in a speaker's cognitive representation of the linguistic system, with those of its instantiations that have been mastered as estab-
94 RONALD W. LANGACKER lished units. Semantic structure is viewed in cognitive grammar as "conventionalized conceptual structure", i.e. the form our conceptualizations must assume for ready linguistic expression given the symbolic conventions and resources of the language. A semantic structure is a conceptualization that functions as the semantic pole of a linguistic expression, either fixed or novel. I refer to any such structure as a "predication", and to the semantic pole of a single morpheme as a "predicate". A predication is characterized relative to one or more cognitive "domains", collectively referred to as its "matrix". Some are "basic domains", i.e. primitive fields of representation not reducible to other, more fundamental conceptions; examples of basic domains include space, time, color, taste, emotive domains, etc. Most domains are reducible and hence non- basic: they are related to basic domains through conceptualization or some conceptual hierarchy. Random examples include our conception of the human face; knowing how to skin a rabbit; the calendrical cycle; the notion that objects or events can be ranked on a scale with respect to some property; and so on indefinitely. Any integrated conception or knowledge system can function as a domain for the characterization of a linguistic predication. Non-basic domains are structured hierarchically, in the sense that simpler conceptions can be coordinated or transformed to create more complex, higher-order conceptions, which can in turn combine to form still more complex conceptions, and so on. A given predication must be characterized at an appropriate level in such hierarchies of complexity. It would be pointless, for example, to try to define the geometrical notion [ARC] directly in the basic domain of two-dimensional space, since an arc exists only by virtue of its status as a segment of a circle. [ARC] is thus superordi- nate to [CIRCLE] in the hierarchy of conceptual complexity, and we must approach its characterization in two steps. First, [CIRCLE] can be defined as a configuration in the basic spatial domain. Then, since [CIRCLE] is a coherent concept, it can serve as the cognitive domain for the characterization of derivative notions like [ARC]. It is claimed that a predication always derives its value through the imposition of a "profile" on a "base". The base for a predication is its domain (or matrix), or some relevant subportion thereof. The profile is that portion of the base which the expression "designates". It is thus the focal
THE NATURE OF GRAMMATICAL VALENCE 95 point within the predication, and the special prominence it receives makes it a kind of figure with respect to the ground provided by the remainder of the base. Neither the base nor the profile is sufficient in itself to define a predication, which consists precisely in the relation between the two. Without the base, the profile cannot be identified; without the profile, the base makes no designation. These notions are illustrated for [CIRCLE] and [ARC] in Fig. 2. Heavy lines are used for profiled entities. The base for [CIRCLE], in 2(a), is the basic domain of two-dimensional space. The profile for this predicate, what the predicate designates, is a configuration in this domain. As a conceptualization, [CIRCLE] can then serve as base for the characterization of expressions like [ARC], sketched in 2(b). Observe that the conception of a circle is part of the predicate [ARC]; this configuration constitutes the base of [ARC] but is not profiled, though precisely the same configuration functions as the profile in 2(a). CIRCLE ARC Figure 2 Predicates fall into several basic types determined with respect to the nature of their profile. A "nominal" predication designates a "thing", defined as a region in some domain; this is a bounded region in the case of count nouns. Both circle and arc are count nouns, since they designate lines, a special type of bounded region in space. Yellow (as a noun, the name of a color) designates a bounded region in the color domain, defined primarily by its location along the hue dimension. More abstractly, a paragraph s a bounded region in a written passage, Tuesday profiles a bounded region in the cycle of days constituting a week, and octave designates a bounded region on a musical scale. Physical objects, as bounded entities in three-dimensional space, are prototypical members of the "thing" category,
96 RONALD W. LANGACKER but nevertheless represent a special case in the spectrum of possibilities defined by its schematic characterization. A "relational" predication profiles the "interconnections" among two or more conceived entities; we can think of interconnections as cognitive operations assessing the relative position of these entities within a domain. Typically a relational predication profiles the interconnections between two things, as illustrated by [IN] in Fig. 3(a). (Observe the notational device of using circles to indicate things.) However, one or more of the relational participants may themselves be relational in character. [FAST], for example, locates a process in a particular region along a scale of rapidity. The conception of such a scale constitutes a non-basic domain, and [FAST] situates a process in that region of the scale which lies beyond the neighborhood of the norm (n) in a positive direction. In every relational predication, one of the profiled participants is singled out as the "trajector" (tr), characterized as the figure within the relational profile; other salient participants ■(c) LEAVE Figure 3 FAST i w T ' (®) tr TIME tr P ^^in
THE NATURE OF GRAMMATICAL VALENCE 97 are referred to as "landmarks" (Im). As the terms suggest, the landmark entity can be thought of as a point of reference for specifying the location of the trajector. The relations in 3(a) and (b) are "stative", i.e. each is conceptualized as a static configuration in its domain, scanned as a simultaneously available whole. A "process" is more complex and involves the tracking of a conceived situation through time. It can be regarded as a continuous series of states occupying a continuous series of points in time and scanned sequentially; a process is thus seen as unfolding through time, and the span of time through which it is tracked is referred to as its "temporal profile". Essential facets of the processual predicate [LEAVE] are sketched in Fig. 3(c). The matrix for this predicate coordinates the basic domains of time and space; the heavy-line portion of the time arrow indicates the temporal profile of the process. The Problem We will further explore the cognitive grammar framework by examining its treatment of valence and grammatical constructions. Exemplification will be drawn from various American Indian languages of the Uto-Aztecan family (see Langacker 1977 for a survey). A common feature of these languages, and of languages generally, is the formation of complex verbs through the layering of "verb-like" derivational affixes. The Luiseno data in (1) is not untypical (the tense suffix -q indicates present or recent past). (1) (a) noo rjee-q I leave-TNS T am leaving.' (b) noo rjee-vicu-q I leave-want-TNS T want to leave.' (c) noo poy rjee-ni-q I him leave-make-TNS T made him leave/ (d) noo poy rjee-vicu-ni-q I him leave-want-make-TNS T made him want to leave.'
98 RONALD W. LANGACKER (e) noo poy rjee-ni-vicu-q I him leave-make-want-TNS 'I want to make him leave.' (f) noo poy rjee-vicu-ni-vicu-q I him leave-want-make-want-TNS 'I want to make him want to leave.' Perhaps the most intriguing aspect of such constructions is the tendency for morphological layering to correlate with "semantic scope", so that an inner layer of structure is semantically "in the scope of" the affix constituting the layer immediately external to it. One approach to this phenomenon was afforded by the theory of generative semantics, as illustrated in Fig. 4(a). Scope was explicated in terms of subordination in underlying structure; the surface complex verb was derived by repeated application of rules such as Predicate Raising and Equi-NP Deletion (cf. Langacker 1973). In retrospect, several objections can be raised to this type of analysis. One pertains to the adequacy of the posited underlying structures as representations of meaning. Another is the absence of theory-independent motivation for deriving the root and derivational affixes from clauses embedded to NP. Finally, the whole notion of deriving the surface form from a radically different underlying structure is rather suspect. (a) UNDERLYING STRUCTURE s LEAVE SURFACE STRUCTURE (RULES) M HIM V TNS /\ V MAKE V WANT I LEAVE Figure 4 TNS MAKE / \ I WANT \ LEAVE HE
THE NATURE OF GRAMMATICAL VALENCE 99 Another alternative, which avoids at least one of the objections cited above, is to posit dependency trees like that in Fig. 4(b). Such trees accommodate predicate-argument relationships without the arbitrary appeal to syntactic embedding under NP-nodes. However, they do not themselves address the matter of morphological layering; special provisions are required to provide for the grouping of predicates into words. I will argue, moreover, that predicate-argument representations of this kind are crucially inexplicit even with respect to pivotal semantic relationships. In the remainder of this paper, I approach the correlation between semantic scope and morphological layering from a broader perspective on the nature of grammatical valence. I provide a preliminary but reasonably explicit account of how morphemes combine, and of the relation their forms and meanings bear to those of the composite whole. It is suggested that predicate-argument structures like Fig. 4(b) represent only a special (though prototypical) case in the spectrum of possible grammatical and semantic relationships. The correlation between semantic scope and morphological layering receives a natural explication in a theory that treats grammatical structure as an inherently symbolic phenomenon. Grammatical Valence: Canonical Instances Grammar is the combination of symbolic structures to form progressively more complex symbolic expressions. This combination is bipolar, with the integration of semantic structures standing in a symbolic relation to the integration of their corresponding phonological structures. Our main concern at present is the nature of this integration at the semantic pole. What precisely is the character of the semantic integration that forms a composite semantic structure from two or more component predications? I emphasize at the start that I am not concerned with trying to predict the valence of a morpheme on the basis of its internal semantic structure. In fact, I do not think it is possible to predict valence in absolute terms. There is no reason to suppose, lor instance, that the semantic pole of Luiseno rjee is substantially different in any crucial respect from that of English leave, yet rjee is consistently intransitive, while leave optionally bears a valence relation to a nominal complement (this very optionality establishes the point):
100 RONALD W. LANGACKER (2) (a) The man left. (b) The man left the building. Instead of absolute predictability, we must settle for predictability of a weaker sort, one more generally appropriate for language: the semantic structure of a morpheme defines its valence potential and determines how readily it lends itself to certain kinds of valence relations exploiting that potential, but whether and how this potential will actually be realized is in some measure a function of linguistic convention. Our concern here, though, is with the nature of the valence ties (at the semantic pole) between two structures that do in fact combine in accordance with the conventions of the language. We will start with prototypical valence relations, turning subsequently to cases which deviate from the prototype in various ways. Consider the following Hopi sentence: (3) taaqa moosa-t tiwa man cat-ACC see/find The man found the cat.' (I simplify matters by ignoring the accusative inflection on 'cat' as well as definiteness and tense/aspect.) In standard predicate-argument terms, the Hopi verb tiwa is a canonical two-place predicate, for it designates a relation between two salient entities, a searcher/perceiver and the object sought/perceived. Fig. 5 is a typical predicate-argument dependency tree representation for the lexical morphemes of (3), showing the valence relations between [FIND] and its arguments [MAN] and [CAT]. FIND / \ MAN CAT Figure 5 Though Fig. 5 is perfectly acceptable as a first approximation, it is undeniably inexplicit on many crucial points. For one thing, nothing of substance is indicated about the internal structure of any of the three predicates. Second, nothing explicitly shows that [MAN] and [CAT] have different roles with respect to [FIND], or what these roles are. Third, how [MAN] and [CAT] connect to [FIND], and what permits this combination
THE NATURE OF GRAMMATICAL VALENCE 101 in the first place, are left unspecified. Finally, there is no direct characterization of the composite semantic structure that results from integrating the three components. The cognitive grammar account of grammatical valence can be regarded as an attempt to be explicit on all these points, i.e. to explicate the valid intuitions (pertaining to prototypical valence relations) that lie behind the widespread acceptance of some version of predicate-argument structure. We can begin with a characterization of the internal structure of the predicates, most crucially [FIND]. The matrix for this predicate is complex, involving not only the domain of physical space but also non-basic domains pertaining to perception and cognition, including the knowledge that perceiving individuals, at a given time, have perceptual access to a limited area and make perceptual contact only with objects located within this field. I have conflated these different domains in Fig. 6, which oversimplifies matters but is sufficient for present purposes. FIND Figure 6 [FIND] traces through time the evolution of a situation involving several entities. Two of these entities, both physical objects, are included in the relational profile as central participants: the trajector, corresponding to the searcher, functions as the figure within this profile; the object sought and found functions as the primary landmark. Less salient in this conception is a third entity, namely the perceptual field of the trajector, which I have given as an ellipse. The base of [FIND] includes a search process of indefinite duration. Only the final stages of that process are actually desig-
102 RONALD W. LANGACKER nated by the predicate and hence profiled, namely the transition into the situation where the landmark is located in the trajector's visual field. The dotted "correspondence" lines in Fig. 6 indicate that the trajector is identical for all the component states, as is the landmark. I treat [MAN] and [CAT] in much lesser detail, using only a mnemonic sketch representing their shape specifications in diagrams below. For our purposes, the relevant observation is that they designate prototypical things, bounded objects in physical space. Their full description also includes characterizations with respect to other domains (e.g. specifications of size, color, canonical activities, and so on). Let us now consider the valence relations in Fig. 5. A valence relation between two predications is possible just in case these predications overlap, in the sense that some substructure within one corresponds to a substructure within the other and is construed as identical to it. The dotted correspondence lines in Fig. 7 therefore specify that the trajector of [FIND] is construed as being identical to the object profiled by [MAN], and that the landmark of [FIND] is construed as being identical to the profile of [CAT]. Establishing these correspondences is what permits [FIND] to combine with [MAN] and [CAT] in a grammatical valence relation. The lines of correspondence can also be thought of as the integration of the component predications, i.e. as instructions for fitting these predications together to form a coherent composite structure. In Fig. 7, they specify that the specifications of [MAN] are to be superimposed on those of the trajector of [FIND], and the specifications of [CAT] on those of the landmark of MAN 4JCAI FIND (simplified) Figure 7
THE NATURE OF GRAMMATICAL VALENCE 103 [FIND]. [MAN] and [CAT] thus bear different relations to [FIND], since their profiles correspond to different substructures within it. I claim that all valence relations are based on correspondences between subparts of the component structures. This is in fact the only constant factor in valence relations. However, there is considerably more that needs to be said about canonical relations like those in Fig. 5 if we want to be fully explicit about their nature. For example, what is the nature of the asymmetry between [FIND] on the one hand and [MAN] and [CAT] on the other that leads us to put [FIND] on top in the dependency tree and the others on the bottom? What leads us to say, in predicate-argument terms, that [FIND] is the predicate, and [MAN] and [CAT] the arguments, rather than the converse? It would appear that this asymmetry is connected with the relational character of [FIND]. [FIND] introduces and organizes a scene in which salient participants interact in a specified way. [MAN] and [CAT] designate individual objects, and while it is part of our knowledge of these objects that they participate in relations with other objects, these external relationships are neither salient nor profiled within these predicates. [FIND] thus makes salient reference to two objects as part of its own internal structure — these objects function as the trajector and landmark within its relational profile — while neither [MAN] nor [CAT] profiles an external relation. I speak of [FIND] as being "conceptually dependent", while [MAN] and [CAT] are "conceptually autonomous", with respect to the correspondences in Fig. 7. [FIND] is conceptually dependent because it presupposes, as an inherent part of its own internal structure, the two things participating in the correspondences; [MAN] and [CAT] are conceptually autonomous because they do not similarly presuppose a salient external relationship. One cannot conceptualize the [FIND] relationship without conceptualizing the two things functioning as trajector and landmark of that relation (even if they are conceived only in the vaguest terms, say as blobs), but it is perfectly possible to conceptualize a man or a cat without mentally setting it in a relation with some external object. I would emphasize that conceptual autonomy and dependence are ultimately matters of degree, but in canonical instances of grammatical valence there is a fairly clear asymmetry between the dependent and autonomous predications along these lines. The dependent structure can be equated with the predicate, in predicate- argument terms, and the autonomous structures with its arguments. (Recall that in the terminology of cognitive grammar, either the dependent or the
104 RONALD W. LANGACKER autonomous structure can be a predicate — a predicate is defined as the semantic pole of a single morpheme.) Though [FIND] introduces two relational participants that correspond to the profiles of the autonomous predicates, there is obviously a difference in the degree to which the corresponding entities are specified by the relational and nominal predications. [FIND] characterizes its trajector and landmark only in schematic terms — the former only as a thing capable of searching and perceiving, the latter only as a thing capable of being found — while [MAN] and [CAT] specify the corresponding objects in far greater detail. Hence there is a relation of schematicity between each profiled participant of [FIND] and the autonomous predicate whose profile corresponds to it. We can say that the dependent predicate organizes the scene, setting up a relation between schematically specified objects, and that the autonomous predicates fit into this scene and elaborate particular substructures within it. These substructures can be called "elaboration sites" (e- sites); they are cross-hatched in Fig. 8 for ease of identification. The arrow leading from an e-site in the dependent predicate to the corresponding autonomous predicate thus stands for a relationship of schematicity, as before. MAN 4L|cai FIND (simplified) Figure 8 Two further aspects of canonical valence relations must be considered. The first is "constituency", which is treated here not as a separate dimension of linguistic organization, but merely as reflecting the order in which simpler symbolic structures successively combine to form progressively more complex ones. We may speak of a "compositional path" leading from
THE NATURE OF GRAMMATICAL VALENCE 105 individual morphemes, through intermediate composite structures at various levels of complexity, to the final composite structure representing the overall semantic and phonological value of the entire expression. Typically constituency is binary: at any particular level of organization, two component structures integrate to form a composite structure, which may in turn function as one of two component structures at the next higher level of organization, and so on. It is important to observe that constituency per se is not employed in this framework for the characterization of basic grammatical relations. For example, the subject and direct object of a clause are not defined with respect to a particular constituent-structure configuration — instead the subject is defined as a nominal whose profile corresponds to the trajector of the clausal head, and the direct object as one whose profile corresponds to its primary landmark (cf. Fig. 8), regardless of the order in which these components are assembled. Different compositional paths (orders of assembly) often lead to the same overall composite structure, with no effect on our ability to characterize grammatical relations. Inherent to this account of grammatical valence, consequently, is the potential for constituency to be fluid and variable. The alternate constituency groupings suggested by phonological considerations (e.g. intonation breaks, different phonological phrasings) generally prove workable for the semantic pole as well. Fig. 9 shows the most likely constituency grouping for the example under discussion. At the lower level of constituency, [FIND] and [CAT] are integrated to form the composite structure (FIND-CAT). The schematicity arrow indicates that [FIND] is the dependent structure, and [CAT] the autonomous one; more specifically, [CAT] elaborates the landmark of [FIND]. (Observe that the structure which I identify as conceptually dependent, in the sense described previously, is often not the same structure that is considered dependent in the sense of dependency-grammar representations.) At the higher level of constituency, [MAN] is integrated with the composite structure (FIND-CAT) to form the higher-order composite structure (FIND-CAT-MAN). At this second level, (FIND-CAT) is dependent, and the autonomous structure [MAN] elaborates its trajector. Note that rectangles and square brackets are employed to indicate that the enclosed structure presumably constitutes a unit (i.e. a fully mastered configuration) for a typical speaker; closed curves and parentheses are employed for novel structures. Thus, while all the individual predicates in Fig. 9 are obviously established units, the intermediate-level composite
106 RONALD W. LANGACKER structure (FIND-CAT) may well be novel, as is the overall composite structure (FIND-CAT-MAN). (FIND-GAT-MAN) (FIND-CAT > -[MAN] [find} *{cat] Figure 9 The underscores in Fig. 9 relate to the final aspect of canonical valence relations. Consider the lower level of constituency. [FIND], a process predicate, integrates with [CAT], which profiles a thing. What, then, will be the profile of the composite structure (FIND-CAT)? Will it designate a thing or a process? Clearly moosat tiwa 'found the cat' designates a process, not a thing; it is the core of a clause, and a clause by definition is processual in nature. However there is no inherent reason why the composite structure would have to inherit the profile of the conceptually dependent component rather than the conceptually autonomous one — the choice must be specified as part of each grammatical construction. In a canonical valence relation, therefore, one of the component structures must be singled out as the "profile determinant", which means that its profile prevails in determining the character of the composite structure. The underscores in Fig. 9 mark the profile determinants. (FIND-CAT) is thus a process, for it inherits the processual profile of [FIND]. Similarly, at the second level of constituency (FIND-CAT) is marked as the profile determinant, so the composite structure for the entire expression is also processual, and (3) taaqa moosat tiwa The man found the cat' qualifies as a clause. Fig. 9 of course abbreviates the semantic structures at each level. Fig. 10 is a somewhat more adequate representation, depicting the internal organization of these structures within the limits of the notations introduced (obviously it is still quite informal). By now Fig. 10 should be largely self- explanatory. At the lower level of constituency, [CAT] elaborates the landmark of [FIND], which serves as elaboration site and corresponds to the profile of [CAT]. [FIND] is the profile determinant — indicated by using heavy lines for the box surrounding it. The novel composite structure (FIND-CAT) is formed by superimposing the specifications of [CAT] on
THE NATURE OF GRAMMATICAL VALENCE 107 FIND-CAT-MAN lm Tit. 4—4 * TIME w a 4 TIME Figure 10
108 RONALD W. LANGACKER the schematic landmark within [FIND], while retaining the processual profile of the latter. (FIND-CAT) is thus a process with unspecified trajector. This trajector is elaborated at the second level of constituency, where it is put in correspondence with the profile of [MAN]. Since (FIND-CAT) is the profile determinant, the novel composite structure (FIND-CAT-MAN) is a process whose trajector receives the specifications of [MAN]. It should be apparent that the same composite structure would be obtained if the component structures were amalgamated in another order, i.e. if the constituency grouping were different. For instance, if [MAN] first elaborated the trajector of [FIND], the intermediate-level composite structure (FIND-MAN) would constitute a processual predication with a specific trajector and a schematic landmark. This landmark could then be elaborated by [CAT] at the second level of constituency, with the resulting overall composite structure being no different from the one shown in Fig. 10. Regardless of the order of composition, taaqa 'man' would be identified as the clausal subject, since it elaborates the trajector of tiwa 'find' (whose processual profile is inherited by the clause as a whole); and moosat would be identified as the direct object, for it elaborates the primary landmark. It is reasonable to suppose that speakers do in fact employ this alternate constituency in sentences displaying object topicalization, as illustrated in (4). (4) moosat-t taaqa tiwa cat-ACC man see/find The cat the man found.' Non-Canonical Valence Relations The essential aspects of a canonical valence relation are summarized in Fig. 11. It is a binary relation between two predications, one of which is conceptually autonomous and the other dependent. The dependent structure is relational and includes within its profile an entity, specifically a thing, which corresponds to the profile of the autonomous structure. This entity, only schematically specified within the dependent structure itself, functions as an elaboration site; this e-site bears a relation of schematicity to the autonomous structure, which serves to specify it in finer detail. Finally, the dependent structure is the profile determinant and hence imposes its relational profile on the composite structure.
THE NATURE OF GRAMMATICAL VALENCE 109 DEPENDENT AUTONOMOUS STRUCTURE STRUCTURE o ash e-site Figure 11 This is the basic type of valence relation generally assumed in predicate-argument accounts of semantics, but among the cluster of properties defining it, only the existence of a correspondence between substructures of the components appears to be a universal property of grammatical valence relations. In this section we explore some of the ways in which such relations commonly deviate from the prototype. When there is a clear asymmetry between two predications along the lines of conceptual autonomy/dependence, it is natural for the dependent structure to function as profile determinant, as it canonically does. This is quite consonant with its function of organizing a scene, of establishing relations among schematically specified entities; the autonomous structure simply fits into this scene and elaborates one of these entities. Choosing the autonomous structure as profile determinant in such a situation amounts to a kind of skewing, where one component is intrinsically suited to play a scene-structuring role, but where the perspective of the other component is adopted by the composite structure. This skewing is the basis for the cognitive-grammar characterization of the head-modifier relation. We speak of such a relation when there is a clear asymmetry between a conceptually autonomous and a conceptually dependent predication, and where the autonomous structure functions as profile determinant: the autonomous component is then the "head", while the dependent component is the "modifier". In the canonical alignment of Fig. 11, by contrast, where the dependent component functions as profile determinant, we speak instead of a "head-complement" relation. Consider the Hopi expressions in (5). o
110 RONALD W. LANGACKER (5) (a) taaqa wiipa man tall 'The man is tall.' (b) wipa-taqa tall-man 'tall man' Sentence (5)(a) is clausal, hence relational in character, and here we would normally speak of a subject-predicate relation between taaqa and wiipa, referring to taaqa as a complement of wiipa. On the other hand, (5)(b) is nominal in character, and in this case it is customary to speak of a head- modifier relation. Yet in both instances [MAN] is autonomous and elaborates the trajector of the dependent [TALL], as seen in Fig. 12(a). [TALL] profiles a relation between an object and an abstract scale of comparison. This scale pertains to the vertical extension of elongated objects when they assume their canonical vertical orientation. [TALL] specifies that the upper extreme of the trajector, with respect to this orientation, falls in the landmark region of the scale's positive end, i.e. that portion beyond the neighborhood of the norm (n). TALL MAN (IS) TALL TALL MAN Figure 12
THE NATURE OF GRAMMATICAL VALENCE 111 The relationship depicted in 12(a) is valid for both (5)(a) and (5)(b), yet these expressions are quite different in meaning. They differ precisely in that (5)(a) is relational, while (5)(b) profiles a thing and is therefore nominal. Hence the composite semantic structure in (5)(a) inherits the relational profile of [TALL], and that of (5)(b) the nominal profile of [MAN]. These respective composite structures are sketched in 12(b) and (c). Observe that in each instance a predication of tallness is made with respect to the man — the only difference lies in whether that relation is profiled (designated) by the overall expression, or whether it is merely part of the base. The difference between an adjectival and an adverbial modifier is simply that the latter has a relation rather than a thing for its trajector and e- site. Consider the expression run fast. [FAST] was diagrammed in Fig. 3(b). Its trajector is a process rather than a thing. This schematic process is elaborated by [RUN] in a way precisely parallel to the way in which [MAN] elaborates the trajector of [TALL] in tall man. [RUN] is clearly the profile determinant in run fast, since this expression designates a type of running, not a type of rapidity — the composite structure is quite analogous to Fig. 12(c). Observe that the notion "adverb" receives a very simple characterization in this framework: an adverb is a modifier whose head profiles a relation (as opposed to a thing). We have now seen two ways in which valence relations can depart from the prototype: the autonomous structure may function as profile determinant, and the e-site within the dependent structure may be a relation rather than a thing. Additional kinds of departure from the prototype can be illustrated by a series of composite locative particles in Cora, based on data and analysis provided by Eugene Casad (1982a; Casad and Langacker 1985). The particular locative particles in question are specialized for topographical relations in mountainous terrain. Representative examples are given in (6)(a)-(b). (6) (a) yuu 'right here at the foot of the slope' (b) mah 'away up there to the side in the face of the slope' (c) y 'proximal'; m 'medial'; 0 'distal' (d) u inside'; a 'outside' (e) : 'foot of slope'; h 'face of slope'; n 'top of slope' These particles prove to be essentially regular combinations of three morphemes each. The alternatives for each morpheme position are given in (6)(c)-(e).
112 RONALD W. LANGACKER The semantic structure of mah 'away up there to the side in the face of the slope' is sketched in Fig. 13. The first component, m 'medial', marks distance from the speaker (S). The proximal range, indicated by a small, light circle, essentially encompasses the area within the speaker's physical reach. The medial range, functioning here as a salient landmark and shown in heavy outline as a major relational participant, basically includes the area within the speaker's visual field. The medial predication is thus a sta- tive relation which locates the trajector within the medial range but outside the proximal area. Figure 13 The general contrast in Cora between u 'inside' and a 'outside' assumes specialized values in particular contexts. The specific versions of u and a relevant here are defined relative to a domain which finds the speaker
THE NATURE OF GRAMMATICAL VALENCE 113 standing at the foot of a slope and looking straight up the face of this slope. The slope (including the foot, face, and top) is represented by a rectangle in the middle diagram at the bottom in Fig. 13. The landmark area for u and a is defined as the area along the speaker's line of sight as he looks up the slope from this canonical position. Note that this landmark region extends along the face of the slope only to the skyline, since the speaker's line of sight from the foot of the slope cannot curve to take in the region on top. The morphemes u and a designate stative locative relations, situating the trajector either inside or outside the landmark region along the speaker's canonical line of sight; a 'outside' is of course shown in Fig. 13. Finally, h indicates that the trajector is located in the face of the slope, as opposed to the foot or the top. Its representation in Fig. 13 should be self-explanatory, as should be the lines of integration connecting the three component predicates. It is an obligatory specification of this construction that the trajectors of the three predicates correspond. The speaker is obviously the same in the first two components, and the slope is the same in the second two. The three predicates are therefore tightly integrated by correspondences connecting shared substructures. The composite structure is obtained simply by superimposing corresponding entities. The result is a complex locative relationship in which the trajector is simultaneously located with respect to three parameters and thtee landmarks, one contributed by each component predicate. Taken together these add up to a fairly precise specification that can be glossed as in (6)(b). This construction departs from prototypical valence relations in three ways. First, the construction is not binary. There is no apparent reason to break the three-morpheme sequence down into two levels of constituency; the three specifications are essentially coordinate. Second, there is no obvious sense in which any of the components is conceptually dependent relative to the others, hence there is no e-site. Finally, there is no basis for singling out any of the three component predicates as profile determinant. The composite expression mah does not designate any of the three component locative relations in particular, but rather the complex locative relation defined by coordinating the locative specifications along the three parameters. Because there is no profile determinant, nor any asymmetry in terms of conceptual autonomy/dependence, none of the component structures can be identified as a head, complement, or modifier with respect to the others.
114 RONALD W. LANGACKER Further Departures from Canonical Valence Relations For subsequent examples, I will adopt the abbreviatory notations given in Fig. 14. A line between two entities indicates that they participate in a stative relation, which may be characterized in either schematic or specific terms. Only one component state is shown explicitly in the abbreviation for a process, but it should be understood that there is one such state for every point in its temporal profile, represented by the heavy-line portion of the time arrow. The trajector of a process appears always to be a thing, as the abbreviations suggest. G □ P rj ? thing entity specified schematic specified schematic (thing or stative stative process process relation) relation relation Figure 14 Additional types of departures from canonical valence relations can be illustrated by the Cahuilla data in (7) (from Seiler 1977: 300f). (7) (a) ne-'as kiyul my-pet fish 'my fish' (a') ne-'as tamawet my-pet mockingbird 'my mockingbird' (b) ne-wes-'a navet my-plant-NR cactus 'my cactus' (b') ne-wes-'a sandiya my-plant-NR watermelon 'my watermelon' This is a type of noun-classifier construction frequently found in Uto-Azte- can possessive expressions. Instead of going directly on the possessed noun,
THE NATURE OF GRAMMATICAL VALENCE 115 the possessor prefix attaches to a more schematic noun, or classifier, to which the possessed noun stands in a sort of appositive relation. Cahuilla has a whole series of classifiers, only two of which are exemplified here. The classifier 'as is used for pets and other domesticated animals, and wes- 'a, a nominalization of the verb stem wes 'plant', is used for crop plants sown in a row. The semantic structure of (7)(a) is sketched in Fig. 15. At the lower level of constituency, [MY] functions as a modifier of [PET]. I assume that [MY] profiles a schematically-characterized stative relation, whose landmark is identified with the speaker; it simply indicates that the trajector interacts with the speaker in some unspecified way. [PET] is a relational noun: it profiles a thing, identified schematically as some kind of animal (AN), but also makes salient internal reference to another individual, who participates in an unprofiled relationship of ownership with the designated animal, and thus functions as a type of landmark (reference point) within this nominal predication. The valence relation between [MY] and [PET] involves two correspondences. First, the schematic trajector of [MY] is put in correspondence with the profile of [PET], and is elaborated by the latter predication. A second correspondence associates the landmarks of the two component predications — more precisely, the schematic relationship profiled by [MY] is equated with the ownership relation in the base of [FETj. Since [PET] is the profile determinant (head) in this construction, the composite structure [MY-PET] is nominal rather than relational: it profiles an animal whose characterization is schematic apart from the specification that it is owned by the speaker. Consider now the second level of constituency, where ne-'as 'my pet' combines with kiyul 'fish'. The profile of [FISH] corresponds to the schematically-characterized profile of [MY-PET], which is substantially elaborated by the former predication. The profile of [MY-PET] therefore functions as an e-site, and [MY-PET] can be regarded as conceptually dependent with respect to [FISH]. The composite structure (MY-PET- FISH), which designates a fish owned by the speaker, is surrounded by a closed curve rather than a box to indicate its presumed status as a novel expression (rather than a fixed unit). This latter valence relation is noteworthy in two respects. For one thing, both component predications are nominal: [MY-PET] is dependent in the sense of being elaborated by [FISH] rather than conversely (cf. Fig. 11), but neither component structure is a relational predication. The lack of
116 RONALD W. LANGACKER MY-PET-FISH Figure 15 a relational element is problematic in standard predicate-argument accounts of semantic structure (where predicates are considered n-place relations), but it poses no problems whatever in the present framework, where valence relations depend crucially only on the existence of correspondences between substructures of the component predications. The second noteworthy feature of this construction is that neither component structure can be singled out as the profile determinant. We saw this situation previously in the Cora example, but here the reason is different: since the two component structures profile corresponding entities, the same composite structure would result regardless of which one we identified as the profile determinant. Because there is no clear-cut profile determinant, we lack the basis for positing either a head-modifier or a head-complement relation; the construction is basically appositional. As a final example, Fig. 16 sketches the semantic structure of ne-wes-'a 'my plant', the classifier portion of (7)(b). At the lower level of constituency, the nominalizer -'a combines with the verb stem wes 'plant'.
THE NATURE OF GRAMMATICAL VALENCE 117 [PLANT] is a process predicate, represented with the abbreviatory notation introduced in Fig. 14. The nominalizing predication [NR] characterizes a thing by virtue of the role it plays in a process. The base of [NR] is thus analyzed as a schematic process, and its profile is identified as the landmark of this base process. In the valence relation between [PLANT] and [NR], the entire schematic process constituting the base of [NR] functions as e- site; the specific process profiled by [PLANT] is put in correspondence with the base process of [NR], so that [PLANT] spells out in specific terms the process with respect to which the profile of the nominalization is charac- MY-PLANT-NR Figure 16
118 RONALD W. LANGACKER terized. [NR] is both dependent (being elaborated by the other component) and the profile determinant, so we must speak of a complement (rather than a head-modifier) relation. The composite structure [PLANT-NR] designates a thing identified as the landmark of the process of planting. The integration of [PLANT] and [NR] illustrates two further departures from the canonical valence relation schematized in Fig. 11. For the first time in our examples the e-site is neither the profile nor a subpart of the profile — rather it is a process in the base. Second, the e-site is not a proper substructure within the dependent predicate, but is in fact exhaustive of this predicate (apart from its profile specification). It is of course not unexpected, as a limiting case, that the substructure functioning as e-site might coincide with the whole. At the second level of constituency, the trajector of [MY] corresponds to the profile of [PLANT-NR], being parallel to Fig. 15 in this regard. A second correspondence associates the landmark of [MY] (the speaker) with the trajector of the process constituting the base of [PLANT-NR]; the interaction which [MY] predicates between its trajector and landmark is thus equated with the process of planting, and the speaker is identified with the trajector of this unprofiled process. [PLANT-NR] is conceptually autonomous relative to [MY] and also the profile determinant, so the two participate in a head-modifier relationship. The composite structure [MY- PLANT-NR] is a nominal predication which designates the landmark of the process of planting, whose trajector is the speaker. The resulting expression ne-wes-'a 'my plant/what I planted' is then capable of combining with an appositional noun like navet 'cactus' at a higher level of organization, as in our previous example. We may now summarize the various ways in which a grammatical valence relation can depart from the prototype. A valence relation need not be binary, and it is not necessary that there be a clear asymmetry between an autonomous and a dependent structure. If there is such an asymmetry, the dependent structure need not be relational, and its e-site does not have to be a thing included in the profile: it can be a relation rather than a thing, and it can be an unprofiled facet of the base (or even subsume the base). Either the autonomous or the dependent structure can function as profile determinant, and in some instances the components contribute equally to the profile of the composite structure. Finally, a valence relation often involves multiple correspondences. The existence of at least one correspondence is perhaps the only invariant feature of valence relations.
THE NATURE OF GRAMMATICAL VALENCE 119 Scope and Morphological Layering We can now return to the problem posed earlier, namely the tendency for morphological layering to correlate with semantic scope, as illustrated by the Luiseno sentences in (1). The present conception of grammatical valence allows a straightforward account of the correlation and the expressions which display it. Let us focus on (l)(d), repeated here as (8). (8) noo poy rjee-vicu-ni-q I him leave-want-make-TNS T made him want to leave.' Limiting our attention for the moment to the verb, its constituency tree is given in Fig. 17. (LEAVE-WANT-MAKE-TNS) (LEAVE-WANT-MAKE )-* -[TNS] [ LEAVE-WANT}-- -{MAKE] [ LEAVE}-* -[WANT] Figure 17 Two factors combine to account for the correlation between morphological layering and semantic scope. First, at each level in the constituency hierarchy at the semantic pole, the conceptually dependent structure is also the profile determinant (hence we are dealing with head-complement rather than head-modifier relations). This is the canonical alignment, and it constitutes what was recognized as "semantic scope" in the generative semantic framework. At the lowest level, for instance, [LEAVE] elaborates the landmark of [WANT], hence [WANT] is dependent; [WANT] is also the profile determinant, since qee-vidu 'want to leave' is a kind of wanting, not a kind of leaving. Thus [WANT] imposes its profile on the composite structure, overriding the profile of [LEAVE], which serves to elaborate a substructure of [WANT]. This is the type of relationship people have in mind when they say that [LEAVE] is "in the semantic scope of" [WANT]. The second factor pertains to integration at the phonological pole,
120 RONALD W. LANGACKER which we have largely ignored until now. The notions of autonomy and dependence are equally important at the phonological pole in valence relations as at the semantic pole. In the case of word structure, they amount to the distinction between "root" or "stem" on the one hand, and "affix1' on the other. An affix is morphologically dependent in the sense that it is characterized in part by its position relative to a root or stem, and thus makes inherent reference to a schematically-specified root or stem as part of its own internal structure. This schematic stem within each affix serves as e-site in a valence relation and is elaborated by a specified stem. A root or stem is autonomous in the sense that it makes no salient internal reference to another phonological entity relative to which it is positioned. f rjeevicuniq j Figure 18 The phonological pole of tjee-vicu-ni-q 'made want to leave' is sketched in Fig. 18. At each successive level in the hierarchy, an affix (dependent) combines with a root or stem (autonomous) to form a higher-order stem or word (also autonomous). Observe, moreover, that at each level the autonomous semantic structure is symbolized by the autonomous phonological structure, and the dependent semantic structure by the dependent phonological structure. There is consequently a kind of "harmony" between the alignment of autonomous and dependent structures at the two poles, a perfectly natural phenomenon reminiscent of numerous other types of "harmonization" observable between different facets of linguistic organization (note, for instance, the tendency for linear ordering to correlate with the sequencing of events). It is hardly surprising that this parallel alignment of autonomy/dependence at the two poles represents a strong universal tendency (though exceptions can be found). When it also happens that the
THE NATURE OF GRAMMATICAL VALENCE 121 dependent semantic structure functions as the profile determinant (this too being the canonical situation), we obtain the usual correlation between morphological layering and semantic scope. As we work from the root outwards, each successive morphological increment symbolizes the introduction of a semantic predication which has the previously-constructed predication "in its scope" (i.e. they participate in a head-complement relationship). Let us conclude by examining in specific detail how the complex verb rjee-vicu-ni-q 'made want to leave' is assembled at the semantic pole. We must first consider the semantic poles of the component morphemes. [LEAVE] was diagrammed previously in Fig. 3(c). The other three predicates are sketched in Fig. 19. Each profiles a process, i.e. a series of relational configurations distributed through conceived time. Only a single configuration is explicitly represented in each diagram, which must nevertheless be interpreted as indicating the evolution of the profiled relationship throughout the temporal profile (marked by the heavy-line portion of the time arrow). Cb) MAKE (c) (a) 1m WANT 1 • I : i ^ TNS • ii • | raj speech event Figure 19 The trajector of -vicu 'want' is a schematically-characterized thing, and its landmark is a schematic process. I have used a broken-line arrow to indicate the desire held by the trajector with respect to the landmark. Luiseno -vicu differs from English want in that its trajector is obligatorily equated with the trajector of the landmark process; hence the correspondence line connecting the two trajectors in Fig. 19(a).
122 RONALD W. LANGACKER The causative predicate -ni 'make' is diagrammed in 19(b). The trajec- tor is again a thing, but there are two salient landmarks, one nominal and the other relational. The relational landmark is the process induced by the overall trajector; the causative relationship between this trajector and the landmark process is represented by the double arrow. The nominal landmark is the same entity that functions as the trajector of the landmark process. [MAKE], in other words, is one of many predicates that has the potential for taking two non-subject complements, one of them a direct object nominal and the other a relational complement having the direct object as one of its participants. (Cf. throw in He threw the clothes into the closet — the clothes is construed as both the direct object of throw and the subject (trajector) vis-a-vis the relational complement into the closet.) Lastly, the tense morpheme -q marks present or recent-past time (and also indicates that the subject is singular). In combination with -ni 'make', it receives the recent-past construal, as indicated in Fig. 19(c). For reasons beyond the scope of this paper (see Langacker 1985), it is analyzed as profiling a schematic process, specified in the base as situated prior to the time of speaking (but in temporal proximity to the speech event). The full constituency tree for the semantic pole of nee-vicu-ni-q 'made want to leave' is provided in Fig. 20, which assumes that nee-vicu represents a familiar combination and has the status of a unit, while nee-vicu-ni and nee-vicu-ni-q constitute novel expressions. Observe that the three predicates described in Fig. 19 all function as profile determinants, each at its own level of constituency. Moreover, each is conceptually dependent at its own level of organization, for it contains a salient substructure that is elaborated by the stem it combines with. Each suffix is therefore semantically a head, and the stem it attaches to is a complement to that head. At the first level of organization, the specific process profiled by [LEAVE] elaborates the schematic process serving as the landmark of [WANT], yielding the composite structure [LEAVE-WANT]. Recall that [WANT], as part of its internal structure, establishes a correspondence between its overall trajector and the trajector of the schematic process functioning as its landmark. As a result, when the specifications of [LEAVE] are superimposed on those of the processual landmark of [WANT] to derive the composite structure [LEAVE-WANT], the overall trajector of this composite predication is equated with that of the leaving, i.e. what the trajector wants is for himself to leave (rather than some other individual). The correspondence line internal to [WANT] therefore
THE NATURE OF GRAMMATICAL VALENCE 123 IZAVE-'WANT-MAKE- TNS 31 ltn lmi jL_1 @:&.'.o Figure 20
124 RONALD W. LANGACKER accomplishes what generative grammarians formerly achieved by their transformation of Equi-NP Deletion. No derivation from a hypothetical deep structure is required in this framework. At the second level of organization, [LEAVE-WANT] elaborates the relational landmark of [MAKE] to form the composite structure (LEAVE- WANT-MAKE). This structure predicates a causal relationship between the overall trajector and the process of desiring to leave. At the third level, finally, the process (LEAVE-WANT-MAKE) elaborates the schematic process profiled by the tense predication, so that the composite structure (LEAVE-WANT-MAKE-TNS) situates just prior to the time of speaking the profiled process of one individual causing another to want to leave. Observe that both individuals are characterized only schematically at this level of organization. The complex structure rjee-vicu-ni-q 'made want to leave' is categorized as a verb because its composite structure designates a process. As with any other verb, it is eligible to participate in further valence relations serving to elaborate the schematic things functioning as its trajector and landmark. The probable constituency tree for these further levels of composition is presented in Fig. 21 (it is exactly analogous to Fig. 9, specifying the integration of subject and direct object nominals with a simple verb of Hopi). At the first level of organization, [HIM] (i.e. a third-person singular pronoun) elaborates the nominal landmark of (LEAVE-WANT- MAKE-TNS); since the latter predication is the profile determinant, the composite structure (HIM-LEAVE-WANT-MAKE-TNS) designates a process, one equivalent to the composite structure of Fig. 20 except that its landmark is now more specific. At the second level of organization, the schematic trajector of (HIM-LEAVE-WANT-MAKE-TNS) is elaborated by [I]. Again we have a complement relation (i.e. the processual predication is the head), so the final composite structure (I-HIM-LEAVE-WANT- MAKE-TNS) also profiles a process, as does any finite clause. (1-HIM-IiEAVE-WANT-MAKE-TIB) [IH <HIM-UAVE-WANT-MAKE-TNS ) [HIMh* (LEAVE-WANT-MAKE-TNS) Figure 21
THE NATURE OF GRAMMATICAL VALENCE 125 Observe that the postulated constituency of this sentence, as reflected in Figs. 17 and 21, is perfectly compatible with the one suggested by phonological and morphological considerations. [I] and [HIM] are added only after the entire complex verb has been assembled, as reflected by their status as separate words: noo poy yee-vicu-ni-q. Nevertheless, the correspondences which figure in the various valence relations (and those internal to certain predications) properly establish the speaker as the trajector of the causative relationship, and the third person singular individual as the trajector with respect to both the wanting and the leaving. All of this is accomplished without deriving the sentence from a hypothetical underlying structure posited solely to accommodate its supposed predicate-argument configuration (cf. Fig. 4(b)), and without special rules like Predicate Raising and Equi-NP Deletion (Fig. 4(a)). As a general matter, this framework promises to reconcile semantic and grammatical constituency with the constituency one is led to posit on phonological grounds. Basic grammatical relations are not defined with respect to specific constituent-structure configurations, but rather through correspondences and the trajector/landmark alignment internal to every relational predication. The same correspondences, and hence the same grammatical relations, can be established through alternate compositional paths leading to the same overall composite structure. This is why the conventions of a language often permit alternate word orders or phonological phrasings for otherwise equivalent expressions. Correspondences are essential to grammatical valence relations, but constituent structure is to some degree incidental and variable.
A Usage-Based Model Ronald W. Langacker University of California, San Diego Introduction The generative tradition has strongly emphasized the importance of "generality" in linguistic description. The quest for generality is of course fundamental to the scientific enterprise; we may certainly accept its validity for linguistics. It is not necessarily obvious, however, how this abstract methodological imperative is best adapted and applied to the problems of our discipline, with its own distinctive subject matter. Generative theorists have applied it to linguistics in a specific manner that has had a powerful impact on their conception of linguistic theory and description. I believe their interpretation to be inappropriate for natural language, and its influence to have been a continuing source of difficulty in dealing with linguistic phenomena. Summarized in (1) are three basic tenets of classic generative theory. (1) (a) Economy: A grammar should account for the widest possible array of data with the fewest possible statements. (b) Generativity: A grammar is a set of statements specifying in full and explicit detail how expressions are constructed; it gives a well-defined set of expressions as "output". (c) Reductionism: If the rules of a grammar fully describe the composition of a particular structure, that structure is not itself individually listed in the grammar. © Ronald W. Langacker
128 RONALD W. LANGACKER The economy principle holds that the shortest grammar is the best grammar, other things being equal; redundancy is therefore to be avoided. The generativity principle construes the grammar of a language as a self-contained algorithmic device, consisting primarily of rules for constructing well-formed expressions. The third principle follows from the other two: if a grammar is a set of rules for constructing expressions, and contains the fewest statements possible, then any expression constructed by these rules must itself be omitted from the grammar. Separately listing an expression computable by general rules would be redundant (and redundancy is evil). Though initially this seems quite reasonable, in practice it has had some unfortunate consequences. One result is that virtually all research activity has been dedicated to searching for general rules and universal principles. Now obviously, the search for generalizations is a prime objective — the question I raise is one of balance, and whether the generalizations rest on adequate empirical foundations. It is apparent, for example, that generative grammarians have never dealt seriously with lexicon in its own terms. Nor do they often attempt, after stating a rule in general fashion, to document the actual extent of its applicability and the various factors that influence its felicity (cf. Gross 1979). The abstract systems of rules and principles constructed by theorists seldom emerge organically from the sensitive, fine-grained description of fully representative data (hence their mortality rate is high and their lifetime often tragically short). In brief, all the glory attaches to general principles and abstract theory; careful attention to the minutiae of language data is left for those without the insight and imagination to be good theorists. These practical consequences are clearly a matter of judgment, and we will pursue them no further. More significant is the issue of whether the general methodological imperative of generality is appropriately implemented in linguistics by the positions in (1). One can argue, for instance, that generativity is purchased only at the price of arbitrary assumptions and a priori boundaries that exclude from the description substantial portions of the phenomenon that is putatively being described (a case in point is the distinction between "semantics" and "pragmatics" — cf. Haiman 1980; Langacker in press: ch. 4). One can also suggest that economy and reductionism, as defined in (1), are of dubious validity in any framework that makes a serious claim of psychological reality (cf. Langacker 1982a). We will focus on this latter point. The assumptions in (1) constitute a "minimalist" conception of Unguis-
A USAGE-BASED MODEL 129 tic knowledge; the grammar of a language is reduced to the smallest possible set of statements, with all redundancy avoided. It is also a "top-down" conception emphasizing computation: anything which follows from general statements is omitted from the grammar, on the assumption that it is computed rather than being represented individually. However, we have no a priori reason to believe that the cognitive representation of a language conforms to this conception. It is plausible, psychologically, to suppose that speakers represent linguistic structures in different ways, with considerable redundancy built in. It is also reasonable to assume that many structures are learned as established units even when they also follow from general principles — the computability of a structure does not in principle preclude its learnability and inclusion as a distinct element in the cognitive representation of the linguistic system. The generative grammarian might reply that such considerations belong to the theory of performance, not a theory of competence. But at best the competence/performance distinction is unclear and problematic; as things stand, to invoke it in this manner is essentially vacuous. In actual practice, the effect of this distinction has been to insulate the framework from any possible attack based on its obvious psychological implausibility. If claims of psychological reality are taken seriously, questions of economy assume the status of empirical issues, as opposed to methodological ones. Is it in fact true that a speaker arrives at any kind of redundancy- free representation of linguistic structure? Do speakers in fact avoid learning specific structures as separate units if they happen to conform to general rules? A description of linguistic ability that answers these questions negatively cannot legitimately be attacked on the grounds that the description fails to achieve maximum simplicity: the question of simplicity only arises for two descriptions of the same range of data, but the issue at stake is precisely that of determining what the relevant data is (i.e. what are the cognitive structures that we are trying to model?). One could just as well omit phonology from the grammar on the grounds that a grammar without a phonological component is simpler than a grammar with one. For a specific example, consider such expressions as dogs, trees, toes, pins, bells, walls, and eyes, which instantiate a regular pattern of plural- noun formation. Clearly, a regular rule of plural formation can be given for English, and these expressions conform to the rule. According to the principle of reductionism, incorporating this rule in the grammar of English precludes the listing of individual plural forms like dogs, trees, toes, etc. The
130 RONALD W. LANGACKER rules allow their computation from the noun stems, hence their inclusion in the grammar would be redundant. The goal of cognitive grammar is to characterize those psychological structures that constitute a speaker's linguistic ability, i.e. his grasp of established linguistic convention. This notion inspires an alternate approach to forms like dogs, trees, toes, etc. A typical speaker uses frequently-occurring expressions like these on countless occasions; at least some of them must attain the status of "units" (i.e. familiar, thoroughly mastered structures — cognitive routines). In fact, the pattern itself can only be learned through the observation of instantiating expressions, some of which most likely become units before the pattern is extracted; it is implausible to suppose that these plural forms suddenly lose their status as familiar units when the rule is acquired, and must henceforth be computed from scratch. An alternative conception is sketched in Fig. 1. The grammar of a language is defined as a "structured inventory of conventional linguistic units". Specific expressions are included in this inventory provided that they have the status of units — a reasonable assumption for dogs, trees, etc. Also included in the grammar are "schemas" extracted to represent the commonality observed in specific expressions (both units and non-units). The schema corresponding to a grammatical pattern can be regarded as a template for the construction of instantiating expressions. Hence the schema for nouns like dogs, trees, etc. is a complex structure whose internal organization is exactly parallel to these plural forms, except that a schematic noun stem (given as [THING/X]) occurs in lieu of a specific noun stem. The schema therefore captures the pertinent generalization, and its categorization of instantiating expressions constitutes their "structural description". The coexistence in the grammar of the schema and instantiations affords the speaker alternate ways of accessing a complex but regular expression with unit status: it can simply be activated directly, or else the speaker can employ the schema to compute it. Moreover, the schema is available for the computation of novel instantiations (e.g. quagmires)', if such an expression is frequently employed, it may very well become established as a unit and thus be incorporated per se in the grammar. The specific array of instantiations having the status of units doubtless varies from speaker to speaker (and changes with experience for an individual speaker), but this is not considered problematic.
A USAGE-BASED MODEL 131 GRAMMAR SCHEMA THING [dog] |dog| [PPl l-s 1 [TREE |jtree PLjl -s J FIXED EXPRESSIONS NOVEL EXPRESSION Figure 1 Generative grammarians have normally resisted the idea that regular expressions should be listed in a grammar, on the presumption that listing entails a failure to capture significant generalizations. I refer to this attitude as the "rule/list fallacy". It is fallacious because it tacitly presupposes only two options: rules vs. lists. But nothing in principle prevents a third option, namely positing both rules (i.e. schemas) and lists, as shown in Fig. 1. By their very nature, schemas embody generalizations. The implicit assumption that rules and lists are mutually exclusive merely reflects the generative conception of economy, as described in (1), whose appropriateness for natural language is precisely what is at issue. In describing cognitive grammar as a "usage-based" model of language structure, I have in mind the "maximalist", "non-reductive", and "bottom- up" character of the general approach (as compared to the minimalist, reductive, and top-down spirit of the generative tradition). The full import of this description will gradually become apparent. For now, let us briefly examine the basic thrust of each term. The minimalist spirit of generative theory reflects an archetypal conception of the linguistic system as a self-contained and well-behaved set of general rules; though nobody believes that a language consists solely of general rules, this archetype strongly influences virtually every aspect of generative theory and descriptive practice. By contrast, the maximalist conception views the linguistic system as a massive, highly redundant inventory of conventional units. These units run the gamut from full generality to complete idiosyncrasy, and no special significance attaches to any distinc-
132 RONALD W. LANGACKER tions one might draw along this scale. Valid generalizations are sought and captured (represented in the grammar by schematic units), but fully general statements are probably a distinct minority: rather than being prototypical for language, exceptionless rules are special, atypical cases. Moreover, the maximalist outlook leads one to anticipate a gradation between linguistic and extra-linguistic knowledge (cf. Langacker in press: ch. 2); while it does not deny the possibility of innate structures specific to language, neither linguistic ability nor the grammar of a particular language is conceived as a discrete "module" with well-defined boundaries (pace Fodor 1983). Cognitive grammar is non-reductive by virtue of recognizing both rules or patterns and the individual knowledge of specific structures that conform to them. A schema and its instantiations represent different facets of linguistic knowledge, and if they have the status of units, both are included in the grammar of a language. One advantage of this approach is its ability to accommodate, with no special apparatus, instances where a fixed expression is more detailed and elaborate than the structure that a rule or schema would allow one to compute (e.g. an eraser is a particular type of object with specific properties, not just 'something that erases'). The grammar is not conceived as a constructive device giving expressions as "output'1, but simply as providing a speaker with an inventory of symbolic resources that he — the speaker — can employ for the construction of novel expressions, using all the information and abilities at his disposal. As one such resource, the schema describing a pattern of composition is not itself responsible for actually constructing an expression. Instead it serves a categorizing function: it furnishes the minimal specifications an expression must observe to be categorized as a valid instantiation of the pattern it embodies. An expression may satisfy these specifications, and thus be judged compatible with the schema, even if its characterization is more precise and fully articulated than anything predictable just from the schema and the component morphemes. Finally, the model is said to take a "bottom-up" (rather than a top- down) approach. What this amounts to is a redistribution of emphasis: instead of being almost solely concerned with general rules and principles, we must also give substantial weight to their arrays of conventional instantiations, investigating the actual extension of the patterns in question and the factors that influence it. Furthermore, since patterns are abstracted from specific instances, we need to investigate the schematization process. We know, for example, that speakers learn and manipulate specific expres-
A USAGE-BASED MODEL 133 sions; but we do not know, in any direct way, precisely what degree of schematization they achieve, i.e. how abstract and general the rules are that they manage to extract from more specific structures. I suspect that speakers differ somewhat in this regard, and do not invariably arrive at the highest-level schemas that the data would support. In any event, the omnipotence of high-level generalizations is not a matter of a priori necessity. Though regularities are obviously noted and employed in the computation of novel expressions, it is quite conceivable that low-level schemas are more important for this purpose than highly-abstract schemas representing the broadest generalizations possible. If high-level schemas are extracted, they may be of only secondary significance, serving more of an organizing function than an active computational one. The Network Conception Critical to the formulation of a usage-based theory is a coherent view of linguistic categorization. A particular model of categorization, the "crite- rial attribute" model, has generally been accepted without serious question in the Western intellectual tradition. More recently, the "prototype" model has been advanced as an alternative with greater claims to cognitive plausibility. My own proposal, the "network" model, represents a synthesis of prototype theory and categorization based on schemas. (For general discussion, see Lakoff 1982b, in press; Langacker in press: chs. 10-11.) In a strict formulation of the criterial attribute model, a category is defined by a fixed set of properties or features. These attributes are necessary and sufficient conditions for category membership, affording absolute predictability in this regard: if an entity possesses all the criterial properties, it is a member of the class; otherwise it is not. Class membership is consequently an all-or-nothing affair; there are no degrees of membership, nor does a category display any significant internal structure. The prototype model was pioneered by Eleanor Rosch, and has been presented and supported in numerous publications (e.g. Rosch 1973b, 1975, 1977, 1978). In this model, a category is defined with reference to a prototype, i.e. a schematized representation of typical instances. Entities that conform to this prototype are accepted unproblematically as "central" members of the class. Non-conforming members may nevertheless be assimilated to the category as "peripheral" members provided that they are judged as being similar to the prototype in certain respects. A class is struc-
134 RONALD W. LANGACKER tured internally by virtue of its organization into central and peripheral members; moreover, class membership is a matter of degree, reflecting the distance of a member from the prototype. Because there is no specific checklist of criterial features, membership in a category is not subject to absolute predictability (indeed, there need be no significant features that are shared by all class members). Whether an entity qualifies depends on the judgment of the categorizer, and his tolerance in accepting members that diverge from the prototype. There is no fixed limit on how far something can depart from the prototype and still be assimilated to the class, if the categorizer is perceptive or clever enough to find some point of resemblance to typical instances. The model I have adopted for cognitive grammar incorporates the prototype model as a special case. The members of a category are analyzed as nodes in a network, linked to one another by various sorts of "categorizing relationships". One kind of categorizing relationship is "extension" from a prototype, which may either be a "local" prototype or the "global" prototype for the category as a whole. The notion of extension, symbolized by a dashed arrow, implies some conflict in specifications between the basic and extended values; hence [A] > [B] indicates that [B] is incompatible with [A] in some respect, but is nevertheless categorized by [A]. A second kind of categorizing relationship holds between a "schema" and a structure that "elaborates" or "instantiates" the schema. Symbolized by a solid arrow, e.g. [A] —> [B], the relationship amounts to one of specialization: [B] conforms to the specifications of [A], but is characterized with finer precision and detail. (Note that a schema is not viewed as a set of features or criterial attributes — it is an integrated structure whose internal organization is parallel to that of its instantiations.) A third kind of categorizing relationship, [A] <— > [B], amounts to a perception of mutual similarity; it differs from extension only by lacking directionality. For any type of categorizing relationship, we can speak of the "distance" between [A] and [B], i.e. how much modification or elaboration of [A] is required to arrive at [B]. Beyond this, the nodes and categorizing relationships comprising the network vary greatly in their cognitive salience and degree of entrenchment. The network model is conveniently illustrated by the alternate senses of a polysemous lexical item. A fragment of the network required for the English verb run is presented in Fig. 2. With heavy lines (to indicate special cognitive salience) I have singled out the global category prototype, namely
A USAGE-BASED MODEL 135 the conception of rapid bipedal human locomotion; this is presumably the meaning that is acquired first, and also the one most likely to be activated in a neutral context. We cannot be certain how far "upward" a speaker extends this network through the process of abstraction (schematization), and in particular, whether he extracts a "superschema" having all other nodes as direct or indirect instantiations. Nor do we know how far "downward" a speaker articulates the network into progressively more specialized notions. Speakers may very well differ in these respects, and also in the specific set of categorizing relationships they establish between nodes. But although the precise configuration of the network is variable and even indeterminate, the need to postulate some type of network is seemingly beyond dispute: the meanings of a commonly-used lexical item define a "complex category", i.e. one that is not reducible to a single structure (node). [""rapid ""j ]_MOTIONj RAPID n-LEGGED \ LOCOMOTION! COMPETITIVE POLITICAL ACTIVITY (candidate) s RAPID [MECHANICAL MOTION s*r\ (engine) RAPID FLUID MOTION I (vater) S RAPID 2-LEGGED [LOCOMOTION (person) \ RAPID k-LEGGED [LOCOMOTION (animal) \ COMPETITIVE 2-LEGGED LOCOMOTION (race) RAPID k-LEGGED [LOCOMOTION (dog) RAPID k-LEGGED LOCOMOTION (horse) Figure 2 A strict reductionist approach would seek maximum economy by positing only a single structure to represent the meaning of a lexical category. However, if our goal is to properly characterize a speaker's knowledge of linguistic convention, any such account is unworkable. From neither the category prototype alone, nor from an all-subsuming superschema (should there be one), is it possible to predict the exact array of extended or
136 RONALD W. LANGACKER specialized values conventionally associated with a lexeme (out of all those values that are cognitively plausible). A speaker must learn specifically, for instance, that run is predicated of people, animals, engines, water, hosiery, noses, and candidates for political office; the conventions of English might well be different. Equally deficient is the atomistic approach of treating the individual senses as distinct and unrelated lexical items. The claim of massive homonymy implied by such an analysis is simply unwarranted — it is not by accident, but rather by virtue of intuitively-evident semantic relationships, that the meanings are symbolized by the same form. A network representation provides all the necessary information: an inventory of senses describing the expression's conventional range of usage; the relationships these senses bear to one another; schemas expressing the generalizations supported by a given range of values; and specifications of distance and cognitive salience. Some classic problems of lexical analysis are readily addressed in terms of the network model, namely the distinction between "polysemy" and "homonymy", and that between "ambiguity" and "vagueness". The first distinction hinges on whether the various senses associated with a given form are semantically related. In the network model, semantic relatedness is a matter of degree, so polysemy vs. homonymy does not reduce to a simple dichotomy. Two senses may be related directly, by a categorizing relationship, or else indirectly, through a chain of such relationships. Direct relationships range continuously along the distance parameter (e.g. the 'rapid 4-legged locomotion' sense of run lies closer to the prototype 'rapid 2-legged locomotion' than does 'rapid mechanical motion'); moreover, relationships at a particular distance may differ in salience (consider the gradual "fading" of conventional metaphors). Since speakers are very adept at perceiving semantic relationships, and since comparison is encouraged by common symbolization, it is seldom safe to assume that no connection whatever is established between the alternate senses associated with a form. Even a tenuous relationship, dimly perceived, is still a relationship, and though certain phenomena may presuppose a close or salient connection (so that, for instance, Tom and his dog are both running is more felicitous than Tom and the water are both running), there is no basis for positing a specific cut-off point along the scale of semantic relatedness where polysemy abruptly stops and homonymy begins. Homonymy is better analyzed as the endpoint along the cline of relatedness — it is the limiting or degenerate case of polysemy, where the only relationship between two
A USAGE-BASED MODEL 137 senses consists in their common phonological realization. Hence the actual descriptive problem is not to distinguish between homonymy and polysemy, but rather to characterize semantic networks as fully and accurately as possible. Whereas the issue of polysemy pertains to whether two or more senses are semantically related, the distinction between "ambiguity" and "vagueness" depends on whether multiple senses should be posited in the first place. A commonly-employed test concerns the number of interpretations supported by sentences like those in (2): (2) (a) Tom has an uncle, and Bill does too. (b) Tom has two ears, and Bill does too. A possible interpretation of (2)(a) is that Tom has a maternal uncle (i.e. a mother's brother) while Bill's uncle is paternal (a father's brother), or conversely; the anaphoric relationship between clauses does not require that the two uncles be of the same type. By contrast, (2)(b) cannot easily be construed as meaning that Tom has two organs of hearing, while Bill has two cobs of corn — both clauses must be interpreted as referring to the same type of ear. From such observations, it is normally concluded that uncle displays vagueness rather than ambiguity: it has the single meaning 'parent's brother' (unspecified for the gender of the linking relative) rather than the two distinct senses 'mother's brother' and 'father's brother'. On the other hand, ear is ambiguous, with anaphora sensitive to the difference between the two meanings. David Tuggy (1981) has argued that the felicity of interpretations involving a mixture of types is often a matter of degree. Thus (3)(a) seems peculiar if Tom is an artist doing a portrait while Bill is putting a new coat of paint on a fence, but the sentence is much less infelicitous if Tom is painting a massive mural instead. (3) (a) Tom is painting, and Bill is too. (b) Tom is talking, and Bill is too. (c) Tom is writing, and Bill is too. If Tom is giving a formal lecture, and Bill is outside in the corridor chatting with a friend, (3)(b) is less than perfect, but much better than (2)(b) under the mixture-of-type interpretation; the same is true of (3)(c) if Tom is a novelist at work while Bill is simply writing a letter to his mother. Moreover, while (4)(a) is marginal (unless the bacon has burst into flames), the same mixture-of-type interpretation seems relatively natural in (4)(b).
138 RONALD W. LANGACKER (4) (a) ?The fire is burning, and the bacon is too. (b) Well, the fire is still burning. Oh my god! The bacon is too! Examples like these can be multiplied indefinitely, and suggest the inadequacy of a fully discrete analysis that posits either a single vague sense or two distinct senses. Tuggy further suggests a promising way of handling these graded phenomena in a usage-based approach employing the network model. It is assumed, first, that schemas and instantiations may coexist in the grammar as different facets of a speaker's linguistic knowledge (cf. Fig. 2). Thus, if the semantic network for uncle contains the schematic node [PARENT'S BROTHER], this does not preclude its also containing the instantiations [MOTHER'S BROTHER] and [FATHER'S BROTHER]. Moreover, the units of a grammar differ in their cognitive salience, which correlates with the likelihood of their activation. For seemingly clear instances of vagueness, we can hypothesize that the schema is quite salient relative to its instantiations; the reverse is postulated for clear-cut instances of ambiguity. This type of analysis is illustrated in Fig. 3, where the supposed difference in salience is indicated by the use of heavy-line vs. broken-line boxes. Because the gender of linking relatives is not a significant factor in English kinship terminology, it is plausible to assume that [PARENT'S BROTHER] is more salient in the network for uncle than either of its instantiations. For those speakers who associate the two senses of ear, a schema may well be extracted to represent their perceived commonality. However, this schema must be quite abstract (since the similarity is so tenuous), and would seem to be of little cognitive or communicative utility, so its salience is presumably quite low. (a) -uncle ■^^■^■m fa) ear (PARENT'S | BROTHER | rMOESFs1^ .JfaSSts] I EARi L J m2 j^BPOTFffiRJ L5?2™??J |(body part)| |{ear of corn)] Figure 3 ?art)| |( Let us assume that the felicity of anaphoric expressions like those in (2) depends on the same semantic structure being activated in the construal of
A USAGE-BASED MODEL 139 each clause. In the case of uncle, the large disparity in cognitive salience ensures that [PARENT'S BROTHER] is virtually always activated in preference to its instantiations; hence the common construal of uncle in the two clauses renders (2)(a) felicitous. By contrast, the specific senses of ear are far more likely to be activated than their schema, with the consequence that the felicity of (2)(b) varies depending on whether the same sense is activated for both clauses. For relatively clear examples like these, the network analysis is roughly equivalent to one that simply posits a single sense for vague expressions and two senses for ambiguous expressions. Its advantages are more apparent for examples of intermediate status, like those in (3) and (4). For these we need only assume that the specific and schematic senses have sufficient cognitive salience to compete for activation. Talk, for instance, has a specific value pertaining to formal oral presentations, another relating to informal conversation, and a schematic sense which abstracts away from the social context and focuses on the verbal activity per se. The felicity of a sentence like (3)(b) depends on which of these structures prevails when it is constructed or interpreted on a particular occasion. If Tom is giving a lecture and Bill is chatting with a friend, (3)(b) can nevertheless be judged felicitous when one is concerned primarily with the activity itself, so that the schema is activated for the construal of both clauses. When, instead, the emphasis lies on the social aspects of the two events, the specific values are activated in lieu of the schema; this difference in the construal of talk in the two clauses renders (3)(b) problematic (its degree of deviance reflecting the cognitive distance between the two senses). In principle, then, the fluidity of well-formedness judgments for sentences like (3)-(4) can be explicated in terms of a network whose elements range along a continuous scale of salience and ease of activation. Let us conclude this initial discussion with some brief remarks on the growth and development of networks. There is an intimate connection between the "outward" growth of a network through extension, on the one hand, and its "upward" growth through schematization, on the other. The process of extension occurs because a speaker perceives some similarity between the basic value (i.e. the local or global prototype) and the extended value. This similarity-perception represents the commonality of the basic and extended values, so it constitutes a schema having the two for instantiations, as depicted in Fig. 4(a). The similarity perception per se need not be cognitively salient or achieve the status of a unit — it may be only a fleeting occurrence. Still, a category's outward extension from the
140 RONALD W. LANGACKER prototype should tend to be accompanied by a certain amount of upward growth, as schemas are extracted to generalize over a more diverse array of category members. (a) [SCHEMA} [PROTOTYPE]- -^EXTENSION) In fact, the growth of a network from the category prototype probably involves a variety of phenomena, as depicted abstractly in Fig. 4(b). Likely sorts of development include at least the following: (i) extension, with or without the implied schema achieving unit status (compare the extensions [A] > [B] and [B] > [G]); (ii) the "downward" articulation of a category as finer distinctions are made (e.g. the differentiation of [A] into [D], [E], and [F]); (iii) the extraction and interpolation of subschemas for nodes already present (note [H]); (iv) the incorporation of additional categorizing relationships, as direct comparisons happen to be made between indirectly- associated nodes (observe [C] —> [G]); and (v) adjustments in the entrenchment and salience of elements, as determined by the vicissitudes of usage and experience. General Applicability The semantic pole of a polysemous lexical item has served thus far as our prime example of a complex category, i.e. one whose characterization cannot be reduced to any single structure. I have proposed that a complex category be described as a network, where nodes with varying degrees of cognitive salience are connected by categorizing relationships. The network model subsumes the prototype model as a special case, and further accommodates taxonomic relationships based on schematization; it offers an
A USAGE-BASED MODEL 141 integrated account of these modes of categorization, and holds considerable promise of empirical adequacy. I now suggest that the utility of these notions extends beyond the realm of lexical polysemy: linguistic categories are in general complex, and networks are required for their proper description. Clearly, the network model conforms to the maximalist, non-reductive, and bottom-up spirit of the usage-based approach. In fact, it is by adopting this model of categorization that cognitive grammar achieves and implements its usage-based character. For a non-lexical example, consider the analysis of a phoneme as a complex category. Let us suppose that the phoneme /a/ (in a particular language) occurs only in the syllables /a/, /pa/, /ta/, and /ka/. Each preceding consonant induces some phonetic modification of /a/, however minor it might be. This phoneme consequently has at least four allophones, namely [a], [pa], [la], and [ka] (where [pa] is the allophone induced by /p/, and so on). The allophone [a], which stands alone as a syllable, is plausibly regarded as the "basic allophone" and equated with the category prototype; the others then function as context-induced extensions from this prototype, as diagrammed in Fig. 5. Moreover, speakers may well extract a schema to represent the commonality of the various allophones. Shown as [xa] in the diagram, the schema is neutral as to whether and how the basic vocalism of [a] is modified by a preceding consonant. Figure 5 The network model therefore reconciles two classic views on the nature of a phoneme: that which analyzes a phoneme as a set of allophones; and that which treats it as a unitary but necessarily abstract entity (i.e. a schema). The non-reductionist character of the analysis also accords with
142 RONALD W. LANGACKER traditional phonemic descriptions, which provide a list of allophones for each phoneme, and state the environments that condition each "derived" or "non-basic'1 allophone. The necessity for a non-reductionist account is readily apparent in this domain, since a speaker's phonetic ability does not reside in any single structure. A speaker who fully controls the phonetics of his language is able to pronounce not only the basic allophone, but also the full array of derived allophones, properly distributed. Each implies an articulatory (also an auditory) routine that a speaker masters as part of his internal representation of the linguistic system. These units are properly included in the grammar of a language, for they constitute one facet of a speaker's grasp of linguistic convention. The inventory of conventional units comprising the grammar of a language is "structured", in the sense that some units function as components of others. Often, in fact, a unit owes its specific character to a more inclusive structure in which it occurs, and is therefore confined to this structure (at least initially). For instance, the notation of Fig. 5 should not be interpreted as implying that [pa], [la], and [ka] are free-standing units that can occur independently; they occur only in the context of the respective syllabic units [[p][pa]], [[t][ta]], and [[k][ka]], since the preceding consonant induces their distinguishing phonetic properties. The categorizing relationship [a] ---> [pa] of Fig. 5 is thus more adequately represented in Fig. 6(a), which shows the extended variant in the environment that determines and supports it. (a) [Tf (b) [a] |m!J 0 Figure 6 DOG RUN' The process of "accommodation", whereby a structure is adjusted to make it compatible with its context, is obviously a major factor in the evolution of complex categories. In this maximalist and non-reductive framework, a variant arising through accommodation is recognized as a distinct entity, a separate node in the network describing the category, regardless of how fully or automatically the context determines its specific properties. Let us suppose, for instance, that run is first learned with reference to
A USAGE-BASED MODEL 143 the canonical, upright, 2-legged locomotion of humans. Hence the semantic pole of this verb is limited initially to a single value, [RUN] (which serves as the prototype of the fully-articulated category that eventually develops). What if the learner now observes the rapid locomotion of a dog? He may himself decide to use the verb run to describe this activity, or he may hear someone else describe it in this fashion. In either event, this usage implies and induces a sense that diverges from the prototype in readily-observable respects: the actor is canine rather than human, it uses four legs rather than two, and so on. When predicated of a dog, run thus accommodates to its subject through the semantic extension [RUN] > [RUN'], as sketched in Fig. 6(b). The accommodation is easily made, and essentially predictable from the specifications of the subject, but the extended value [RUN'] is nonetheless distinct from [RUN] and takes its place in the network constituting the semantic pole of the lexical item. In similar fashion, I would posit numerous semantic variants of eat, accommodated to the nature of the food consumed and the specific activity required for its consumption. To be sure, I have no way of knowing just how finely articulated this category is, and it is doubtful that the lowest- level variants have any substantial measure of cognitive salience. It is nevertheless a conventional fact of English usage — not something a speaker must decide anew on each occasion — that eat is employed for such diverse activities as the consumption of meat, bananas, peanuts, and soup. There are standard objections to this analysis; they pertain to the proliferation of meanings, the mixture of pragmatic and semantic considerations, and the failure to extract a unifying generalization. However, all these objections reflect the minimalist, reductive bias of contemporary semantic theory, which is precisely what is at issue, and have little force from the standpoint of cognitive grammar. In particular, the postulation of specific variants does not prevent us from expressing the unifying generalization by means of a coexisting schema, which may very well have greater salience. Like its semantic pole, the phonological pole of a lexical item is a complex category revealingly described as a network. Even an expression not generally thought of as having multiple allomorphs can nevertheless assume a variety of specific values (some of which may establish themselves as units) depending on such factors as tempo and prosody. The applicability of the network model is of course more obvious for expressions that do show allomorphic variation. Consider the noun leaf, whose phonological pole is diagrammed in Fig. 7(a). The basic allomorph, [lif], functions as the cate-
144 RONALD W. LANGACKER gory prototype (it possibly subsumes more specific variants, as just noted). Since the plural is leaves, we must also recognize the allomorph [liv] (occurring only in the context of the plural construction), which constitutes an extension from the basic allomorph; representing the commonality of [lif] and [liv] is the schema [liF] (where [F] neutralizes the voicing contrast of [f] and [v] — i.e. it is equivalent to an "archiphoneme"). Though Fig. 7(a) depicts only the phonological pole, observe that the context for [liv] requires bipolar characterization: the suffix triggering the phonological extension must be the plural morpheme [PL/z] in particular, and not the possessive. (a) uf| IM---J5tvUz1 E3~-fPl Figure 7 5karf[ iU)skarvl [z] The voicing of final [f] in the plural is of course not limited to leaf. Numerous other nouns display the same alternation, e.g. hoof/hooves, life/ lives, elf/elves, knife/knives, scarf/scarves, shelf/shelves, wife/wives, loaf/ loaves. Although the.pattern is not productive (consider///es, reefs, puffs, etc.), it does constitute a regularity that speakers may incorporate as part of their cognitive representation of the linguistic system. In the present framework, regularities are expressed by schemas. It is assumed that any configuration of structures — even a categorizing relationship — is potentially subject to schematization, should the proper circumstances arise. Here the conditions for schematization are indeed met: we find a series of nouns, all conforming to the schematic characterization [...f], which undergo parallel modifications in the context of the plural construction. The phonological network for each of these nouns includes a categorizing relationship between the basic allomorph ending in [f] and a secondary
A USAGE-BASED MODEL 145 allomorph ending in [v], this latter occurring only in combination with the plural morpheme; thus, alongside [lif] —» [liv] we find [nayf] > [nayv], [skarf] ——» [skarv], etc. A schema can be extracted to represent the commonality of these categorizations, as shown in Fig. 7(b). Though abstract, this schema is itself a categorizing relationship of phonological extension, like any of its instantiations. Also like its instantiations, it contains a specification of the conditioning environment: the extended variant [...v] occurs only with [PL/z] (again, the semantic pole is not shown). The schema depicted in Fig. 7(b) is the cognitive-grammar equivalent of a morphophonemic rule. It expresses a systematic relationship between the basic form of a noun and the special form it assumes in a particular morphological context. However, the asymmetry implied by the direction of the arrow, in either [...f] > [-v] °r its instantiations, is not interpreted derivationally, i.e. as the relationship between an "underlying" and a "surface" representation. It is interpreted instead as the asymmetry inherent to comparison and categorization: [...f] is the standard of comparison (category prototype), while [...v] is the target of comparison (a peripheral member assimilated to the category through resemblance to the prototype). Apart from the type of structures that figure in the categorization, there is no fundamental difference between [...f] > [...v] (or [lif] — —» [liv], etc.) and the semantic extension of a lexical item. The general applicability of the network model is starting to become apparent. With a limited set of constructs, this model offers a unified account of many facets of linguistic organization that are normally approached using very different techniques and descriptive devices. What distinguishes the various domains of linguistic structure is not the prevalence of complex categories requiring networks for their description, but rather the types of structures that function as nodes in these networks. In the case of a phoneme, these structures are single phonological segments (allophones and the schemas extracted from them). For a morpheme, the network at the semantic pole has individual senses as nodes, whereas allomorphs serve as nodes at the phonological pole. The network model is also adopted for the description of rules and their conventional instantiations, in which case the individual nodes of the network have a complex internal structure. In Fig. 1, which depicts a combinatory rule (one pattern of plural-noun formation), the nodes include a schema and various instantiations; each node incorporates two symbolic units as component structures, together with the composite structure (not separately shown) resulting from
146 RONALD W. LANGACKER their integration. The morphophonemic rule of Fig. 7(b) is also modeled as a network; it is a different type of rule because the nodes comprising this network have a different type of internal structure: rather than a combinatory relationship between symbolic units, each node consists in a categorizing relationship between two phonological sequences (allomorphs). The rule of Fig. 7(b) is considered morphophonemic because it depends on a morphological context. The phonological extension [...f] — -> [...v] is limited to the plural-noun construction, which therefore figures in the characterization of each node. What if a pattern of phonological extension is not restricted to any particular morphological environment? In that case, no such environment is specified in the schema or its instantiations, and we recognize the rule as being purely phonological. For example, in words like kitten, button, sentence, etc., where [t] precedes the syllabic nasal [n], the [t] is commonly replaced by a glottal stop. Hence each word has two phonological variants linked by a categorizing relationship, e.g. [kitn] — —> [ki^n]. From some array of specific categorizations of this sort, a schema is extracted to capture the regularity, as shown in Fig. 8. This particular schema (unlike the previous one) is sufficiently salient or accessible to be readily invoked for the computation of novel instantiations. When activated in the context of a specific structure containing [tn], the schematic extension [...tn...] > [...?n...] yields a variant having [,?n] in its stead; even nonsense words are likely to be pronounced in this fashion. [kxtnl Jki^ri ]...tn... *- 1 l...'n...| 1 ' r IbAtn [■- H "bA^nl s£ntns— s£n9ns Figure 8 A phonological rule is therefore analyzed in cognitive grammar as a pattern of phonological extension. Phonological and semantic extension are viewed as being directly analogous, with any differences between them attributable to inherent properties of the semantic and phonological domains. I will pursue their parallelism only to the point of noting that semantic extension is also subject to schematization and often follows con-
A USAGE-BASED MODEL 147 ventional patterns. One such pattern, illustrated in (5), involves extending the term for an animal to indicate a person who resembles that animal in certain respects. (5) (a) He's a pig. (b) You're a rat. (c) That lawyer is a real fox. Pig, rat, and fox are among the numerous animal names conventionally employed in this fashion; thus the semantic network associated withp/g, for example, incorporates the categorizing relationship [PIG] —» [PIGLIKE PERSON]. Specific relationships of this sort give rise to the schema [ANIMAL] > [ANIMAL-LIKE PERSON], which is freely used for the computation of novel instantiations. When I call somebody an ostrich, a fennec, or even a veritable brontosaurus, I am thereby conforming to the conventions of English, even if these particular terms have never before been applied to people. Distribution Questions of distribution and productivity figure prominently in contemporary linguistic theory. Many theoretical constructs have at one time or another been invoked to deal with these matters (e.g. diacritics, rule features, major vs. minor rules, disjunctive ordering). Moreover, they are often cited in support of particular analyses and claims concerning the organization of linguistic systems. For example, it is not always possible to predict, on the basis of purely semantic or phonological properties, precisely which lexical items are eligible to undergo a given rule; this absence of absolute predictability is commonly taken as establishing the autonomy of grammar and the need for special, "grammatical" features and classes. It is also sometimes assumed that full generality is criterial for syntactic rules; any rule displaying only partial productivity is assigned to some other component of the grammar (e.g. lexicon or morphology). Cognitive grammar's approach to these issues reflects its emphasis on naturalness, conceptual unification, and austerity in the adoption of theoretical constructs. The highly restrictive "content requirement" allows only three types of units in the grammar of a language: (i) semantic, phonological, and symbolic units that occur overtly; (ii) schemas for such structures; and (iii) categorizing relationships. A separate domain of speci-
148 RONALD W. LANGACKER fically "grammatical" structure is thereby precluded, as are any descriptive elements (e.g. features or diacritics) devoid of both semantic and phonological content. A unified treatment is proposed for lexicon, morphology, and syntax: all are described by means of symbolic units exclusively; these traditionally-recognized areas form a gradation of symbolic structures divided only arbitrarily into distinct "components". To assume that productivity is coextensive with a particular structural domain, or delimits a coherent, self- contained body of phenomena, is at best gratuitous. We must now consider how such a theory is capable of handling the problems of variable productivity and non-predictable distribution. Granted that certain rules are applicable only to a limited class of structures, and granted further that the membership in this class is not always predictable on semantic or phonological grounds, how does one specify the proper restrictions? As we will see, the answer lies in the usage-based character of the framework, together with the network model of complex categories. Linguistic expressions are symbolic, each residing in the relationship between a semantic and a phonological "pole". Grammar consists of patterns for combining simpler symbolic expressions to form progressively larger ones. In cognitive grammar, these patterns (i.e. grammatical rules) are analyzed as schematized expressions — they are themselves complex symbolic structures, parallel in formation to the expressions they schematize, but characterized at a level of abstraction that neutralizes the differences among them. These combinatory patterns are equivalent to grammatical constructions, so we can refer to the schemas describing them as "constructional schemas". Each specifies the way in which two or more "component" structures are integrated, at the semantic and phonological poles, to form a bipolar "composite" structure. Constructional schemas capture generalizations, and serve as templates for the assembly of novel expressions. I have already argued that constructional schemas coexist in the grammar of a language with those of their instantiations that have the status of units (cf. Fig. 1). These instantiations need not be limited to specific expressions — we can also posit subschemas at various levels of abstraction, corresponding to subpatterns discernible in the data. In fact, these structures form a network, as they are linked to one another through relationships of schematicity, and possibly through other types of categorization as well. A grammatical construction can therefore be regarded as a complex category:
A USAGE-BASED MODEL 149 it does not reside in any single structure, but rather in a family of structures connected by categorizing relationships. Internally, each node of this network is quite complex, comprising an entire constructional schema or subschema, or else a specific composite expression. These notions are illustrated in Fig. 9, which sketches the network corresponding to one pattern of past-tense marking in English. The maximal generalization is captured by the topmost schema, which says, in effect, that a verb stem containing [1] may form its past tense by ablauting this vowel to [<e]. In this abbreviatory notation, only the two component structures are explicitly indicated: [PROCESS/... 1...] is the verb stem, and [PAST/...1... — -> ...ae...] is the appropriate allomorph of the past-tense morpheme; a more complete representation would also show the nature of their integration at the two poles, and the composite structure that results (in this case [PROCESS-PAST/...ae...]). The two immediate instantiations of this schema are subschemas representing special cases of the general pattern, that in which the stem vowel [1] is followed by [t], and that in which it is followed by a nasal ([N]); note that the component structures [PROCESS/... it] and [PROCESS/...iN] elaborate the stem of the higher-level schema (the added detail is also reflected in the past-tense morpheme, which makes internal reference to the stem in specifying the ablaut pattern). Instantiating the first of these subschemas are the specific verb forms sat and spat, complex structures that I have rendered orthographically to simplify the diagram. The second subschema is instantiated by the specific verbs swam and began, and also by a lower-level subschema which identifies the nasal consonant of the stem as the velar nasal in particular. This subpattern in turn has two special cases (each with a number of instantiating expressions) that differ in whether the nasal is final or followed by [k]. The analysis is complicated, but I would argue for its cognitive and linguistic plausibility. It is certainly reasonable to suppose that forms like sat, swam, began, rang, sank, etc. are learned by speakers as familiar units (some, of course, are more frequent and deeply entrenched than others). All significant generalizations are captured, both the global generalization expressed by the topmost schema, and also the more limited generalizations reflecting the prevalence of certain types of stems as participants in this morphological construction. Whether speakers extract all available generalizations is an open question, but there is no particular reason to suppose that they proceed directly from specific forms to the highest-level schemas supported by the data, or that a category, should it develop in this
150 RONALD W. LANGACKER (PROCESS PAST [ II PROCESS 1 ...It PAST 1 .. .It--*-.. .aet] sat spat A I PROCESS [...IN... PAST I1 ...X^.-j-^-^ .aeN...[ j PROCESS l"-*9-"l PAST 1 .. .Itj. .♦--»-.. .aeij.. .1 swam "began 1 PROCESS 1 -:-*3 , PAST || ...ip--^. ..aer)| PROCESS .irfa PAST .xrjk--*-.. .aerjk Figure 9 manner, necessarily fails to undergo subsequent differentiation through the emergence of subschemas. A network like Fig. 9 brings out clearly the maximalist, non-reductive, and bottom-up nature of a usage-based approach. Revealing its maximalist character is the incorporation of structures representing all levels of generality, from specific forms to an all-subsuming schema. The analysis is non- reductive because it posits as units both schemas and specific expressions computable from those schemas. Its bottom-up orientation is reflected in the emphasis on low-level structures that provide the basis for higher-level schematization. By contrast, linguists are more accustomed to a minimalist, reductionist, top-down approach that achieves greater economy by including in the grammar only a rule equivalent to the topmost schema. Either type of approach must somehow specify precisely which verbs participate in this construction, since most stems containing [i] form their past tense in some other manner. A minimalist analysis would typically mark the proper verb stems with some kind of diacritic (e.g. a rule feature, or an indication of class membership), and condition the application of the ablaut rule on the presence of this marking. Because the diacritic has no
A USAGE-BASED MODEL 151 intrinsic semantic or phonological content, it is considered a "grammatical" construct; and because the set of verbs in question is not semantically or phonologically predictable, it is taken as constituting a grammatical (specifically, a morphological) class. This apparent need to posit purely grammatical entities is then invoked as an argument for the autonomy of grammatical structure. Diacritics are rendered unnecessary by the maximalist analysis of Fig. 9. For example, the information that sing follows the [...1... > ...ae...] ablaut pattern is provided directly, by (i) the inclusion of sang among the conventional units of the grammar, and (ii) its categorization by the schemas that define the pattern. Observe that the analysis obeys the content requirement, as it employs only symbolic units (both specific and schematic) and categorizing relationships. Though we can perfectly well speak of a grammatical (or morphological) construction, it is fully characterized in terms of symbolic relations between semantic and phonological structures — there is nothing that represents a separate domain, autonomous vis-a-vis semantics, of purely "grammatical" structure. Nor is the existence of an independent set of grammatical entities established by the impossibility of predicting the participating verb stems on semantic or phonological grounds. To assume so is to embrace the "type/predictability fallacy", i.e. the failure to distinguish between two issues that are in principle distinct: the types of structures that occur, and the predictability of their behavior. We can now observe that the apparent argument cited above for the autonomy of grammar is in reality simply an artifact of the minimalist approach to grammatical description. If specific forms that instantiate a pattern are excluded from the grammar by the thesis of reductionism, they must be "constructed" by application of the relevant rule. If the lexical items that undergo this rule are limited to an arbitrary class, they must somehow be marked to undergo it. And since the requisite marking has no intrinsic semantic or phonological content, it must represent some other realm or dimension of linguistic organization. We have seen, however, that symbolic units are capable of furnishing the requisite distributional information provided that we take a maximalist, non-reductionist approach allowing schemas and instantiations to coexist in the grammar of a language. In this usage-based framework, grammatical constructions are analyzed as complex categories. A speaker's conventional knowledge of a
152 RONALD W. LANGACKER construction is not limited to a single, abstract rule or schema standing in isolation — it further embraces his knowledge of how the rule is "implemented" with respect to more specific structures. By its very nature, a high-level schema is compatible with a broad and structurally-diverse array of potential instantiations. Typically, however, conventional usage carves out for exploitation only limited regions within this field of structural possibilities. A full linguistic description must identify these regions, i.e. it must characterize the construction's conventional domain, as articulated by a hierarchy of lower-level structures. Providing this information are subschemas and expressions with the status of units: they specify the actual array of subcases and specific instances that support and give rise to the high-level generalization. In the network for a grammatical construction, the individual nodes and categorizing relationships presumably differ in their cognitive salience and likelihood of activation, as they do in any complex category. There is no reason to assume that the highest-level schema is necessarily the most salient, or even that an all-subsuming schema is always extracted (indeed, the abstractness of a high-level schema is probably inimical to its prominence). For the network of Fig. 9, we can plausibly suppose that the highest-level schema is less readily activated (has a lesser degree of prominence) than specific forms like sat, swam, began, rang, sank, etc.; this would imply that such forms are generally accessed as units (as opposed to being computed from the stems by means of the schema). The opposite is no doubt true for other constructions: greater salience attaches to schemas (though not necessarily those at the highest levels), and relatively few instantiating expressions have the status of units; computation must therefore predominate. At least in principle, it is possible for behavioral evidence to be brought to bear on such matters — cognitive salience and accessibility are neither inherently mysterious nor beyond the reach of empirical inquiry. This notion of accessibility is crucial to a usage-based account of distribution and productivity. The general problem can be formulated as follows: granted that a construction is a complex category, and properly represented as a network, how is this network invoked for the assembly (or evaluation) of a particular instantiating expression? We cannot assume that access to the network is random, or that all nodes are simultaneously and equally activated for this purpose — the resulting chaos would afford no basis for clear judgments of well-formedness or the assignment of structural descriptions. Instead, I suggest a working hypothesis that is basically com-
A USAGE-BASED MODEL 153 patible with a "connectionist" or "interactive-activation*' model of cognitive processing (cf. Elman and McClelland 1984; Rumelhart and Zipser 1985; Waltz and Pollack 1985). This two-part proposal is formulated in (6). (6) (a) Uniqueness: When an expression is assessed relative to a grammatical construction, a single node (from the network representing the construction) is activated for its categorization; if this "active node" is schematic for the expression, the latter is judged well-formed (conventional). (b) Selection: The likelihood that a given node will be chosen as the active node for categorizing a target expression correlates positively with its degree of entrenchment and cognitive salience, and negatively with its "distance" from the target (i.e. how far the target diverges from it by elaboration or extension). The thrust of (6)(a) is that an expression's well-formedness depends on how it is structurally construed (i.e. what it is taken as being an instance of), and that a single "episode" of categorization (structural description) construes it in a particular way. Some factors that influence the choice of categorizing structure (active node) are suggested in (6)(b); they are matters of degree, and possibly antagonistic. Let us consider these matters with reference to Fig. 10, which represents the overall past-tense verb construction of English. At the semantic pole, both the topmost schema and all the subschemas specify the integration of [PROCESS] and [PAST] to form the composite structure [PROCESS-PAST] (not individually shown). However, at the phonological pole the topmost schema is so abstract that it is almost contentless; essentially, it characterizes the stem and inflection only as having "some phonological value". The reason for this extreme schematicity is that the various patterns of past-tense formation have very little in common phonologically. Schemas corresponding to four of these patterns are shown in the diagram. At the left is the subschema for the ablaut pattern [...i... > ...ae...]; this is the same structure that functions as the highest-level node in Fig. 9 (in this way the entire network of Fig. 9 fits into the more inclusive network of Fig. 10). The second subschema corresponds to the "regular" pattern of past-tense formation. At the phonological pole of the past-tense morpheme, [-D]
154 RONALD W. LANGACKER stands for a schematized suffix having [-d], [-t], and [-3d] as instantiations (i.e. it specifies an alveolar stop and is neutral as to the presence of a preceding schwa). The three subcases of the regular pattern are represented by lower-level subschemas, which incorporate specific suffixes and specify the phonological characteristics of the stems they attach to; elaborating each low-level subschema is some array of instantiating expressions having unit status. The third and fourth major subschemas describe respectively the ablaut pattern of brought, caught, fought, sought, taught, etc., and the zero pattern of verbs like cut, hit, slit, bet, spread, and bid. Also represented in Fig. 10 are rough hypotheses concerning the cognitive salience of the individual nodes. Among the schemas, the structures corresponding to the regular pattern are attributed the greatest prominence, as indicated by the heavy-line boxes; they can be regarded as the category prototype. Most of the specific expressions learned as units presumably have substantial salience as well. The schemas describing minor patterns are analyzed as having a lesser degree of prominence, while the topmost schema — considering the vacuity of its phonological pole — may well be the least prominent of all. Such differences in salience (likelihood of activation) are the device employed in this framework to implement the distinction between productive and non-productive patterns (or between major and minor rules). Since prominence is a relative matter and varies continuously, we should expect in general to find a gradation between the two types instead of a strict dichotomy. The contrast may nevertheless be quite pronounced in particular instances; a pattern that is distinctly more prominent than any potential competitor will almost invariably be selected for the construction and evaluation of novel expressions. The "regular" pattern of English past-tense formation is so identified precisely because it has this type of advantage. Consider a speaker who needs the past-tense form of flit, but happens not to have learned it as a unit. He must therefore select one of the schemas in the network for the past-tense construction to employ as the active node (basis for categorization) in assessing possible alternatives. Though all four patterns in Fig. 10 are potentially applicable, he will almost certainly choose the regular pattern by virtue of its distinctive prominence. Within the regular pattern, moreover, the subschema [[PROCESS/...T]-[PAST/-9d]] will be chosen in preference to [[PROCESS/...C]-[PAST/-t]] on the basis of distance: the former characterizes the stem with greater specificity (alveolar stop vs. voiceless consonant) and hence is ''closer" to the stem (namely flit)
jlffiOCESS i| ... PAST • ••_] 1 PROCESS .. .1. . . PAST I ... I.. . - -»-. . . ae. . . 1 1 PROCESS 1 ...T past n1 ...T--—. ..t|i ll PROCESS || ...T PASTll -?a || > G > O m 03 > m D o D Figure 10
156 RONALD W. LANGACKER of potential target expressions. With [[PROCESS/...T]-[PAST/-sd]] selected as the active node, the target flitted is judged well-formed; it is fully compatible with the active node's schematic specifications. Other target expressions, e.g. flat, flaught, and flit, conflict with these specifications and are consequently judged ill-formed, despite being computable from other schemas. Given the selection of [[PROCESS/...T]-[PAST/-ad]] as the active node, their deviance is predicted by (6)(a), the uniqueness hypothesis. The situation is quite different for specific expressions that are mastered as units, e.g. sat, taught, and hit. As units, they are themselves part of the network representing conventional knowledge of the past-tense construction, making them eligible for selection as the active node representing this construction. Suppose, for instance, that a speaker wishes to express the past tense of hit. If the past-tense expression hit has any substantial cognitive salience (which it must, as a frequently-occurring form), the distance factor virtually assures its being chosen as active node: its distance from the target (i.e. the desired past-tense form of hit) is essentially zero. Only hit itself is compatible with this categorizing unit — alternative expressions such as hitted, haught, and hat are deviant as the past-tense of hit, though computable via established schemas. In short, specific expressions with the status of conventional units sanction themselves (or occurrences of themselves) as being conventional, and thereby preempt the process of selecting an active node. Though the selection process is described only vaguely at best in (6)(b), its general import can now be appreciated. The well-formedness (conventionality) of an expression is not absolute: it depends on what unit the expression is construed as instantiating, and on its compatibility with that unit (as evaluated by a categorizing judgment). The selection of a categorizing unit hinges on the dynamic interplay of factors that are frequently antagonistic. If specificity is held constant, the determining factor is entrenchment or cognitive salience; given a range of options, one of them may establish itself as the "regular" choice (the default-case option). Holding salience constant, the determining factor is cognitive "distance", i.e. the extent to which the categorizing unit (active node) must be elaborated or extended to "reach" the target. The tendency for specific structures to prevail over patterns of greater generality has been noted in a variety of theoretical frameworks; examples include disjunctive rule ordering (Chomsky and Halle 1968), proper-inclusion precedence (Sanders 1974),
A USAGE-BASED MODEL 157 and Hudson's priority-to-the-instance principle (1984). In the present framework, this tendency is naturally accommodated as one facet of a broader theoretical perspective emphasizing a non-reductive, bottom-up, usage-based account of language structure. Adopting this general orientation enables one to countenance with equanimity a far-reaching implication of (6)(b): for the most part, specific structures and low-level schemas are more significant than high-level schemas expressing the most inclusive generalizations. Even certain linguists who would resist this implication have, in formulating the precedence principles cited above, at least partially acknowledged its empirical force. This conception gives rise to certain empirical predictions. One pertains to "irregular" forms, i.e. those that are idiosyncratic or follow a minor pattern: the more salient and deeply entrenched they are (as reflected in frequency of occurrence), the more resistant they should be to regulariza- tion. This correlation is in fact so firmly established that documentation would be otiose — let us focus instead on the basis for predicting it. Consider the past participle of drive. The regular pattern predicts drived, which a speaker immediately recognizes as being incorrect. The proper form, driven, exemplifies a minor pattern of little cognitive salience, but the form itself occurs frequently and constitutes a well-entrenched, easily accessible unit of English. If we attribute comparable salience to the unit driven and the schema describing the regular pattern of past-participle formation, the far greater specificity of the former determines its selection as active node whenever the past participle of drive is required. Driven thus sanctions itself as the correct expression, but drived is judged ill-formed (despite its regularity) when measured against this standard. By contrast, the verb thrive is relatively infrequent, and a typical speaker hardly ever has occasion to use its past-participial form. Those who use thriven must know it as a unit, but its status as such is only marginal; its rarity ensures its lack of prominence. Consequently, neither this unit nor the schema representing the regular pattern has an overwhelming advantage in the competition for selection as active node: the former is more specific, but the latter is far more salient. Hence the schema may well be selected, and if so, an occurrence of thrived will slip by unnoticed and be accepted as well-formed (which is quite unlikely in the case of drived). A second prediction is that lower-level schemas should predominate in the computation of novel expressions. This prediction is based on the distance factor: because they are "closer" than high-level generalizations to
158 RONALD W. LANGACKER the target expression, lower-level schemas should in general be selected as the basis for computation (active node). For illustration, let us return to the past-tense schemas of Fig. 9, and consider the relative likelihood of innovative past-tense forms involving the ablaut of [i] to [ae]. Though brought is well-entrenched as the past tense of bring, the sporadic occurrence of brang is at least conceivable (it is attested dialectally and in child language). We can explain this with reference to the low-level schema [[PROCESS/...ig]- [PAST/...IQ ——> ...aerj]] of Fig. 9, extracted to represent the commonality of rang, sang, and sprang; not nearly so salient as the schemas for the regular pattern, it is nevertheless quite specific concerning the phonological shape of the verb stem, and thus stands a decent chance of being activated to compute the past-tense form of another stem meeting its specifications. Similarly, shat is sometimes encountered as a jocular or euphemistic past- tense form, and is computed from the low-level schema that expresses the commonality of sat and spat. On the other hand, the innovative past-tense forms san (for sin) and hass (for hiss) are unattested, uninterpretable, and inconceivable. Observe that the lowest-level schemas available for the computation of these forms are in fact quite abstract. San (if it occurred) would invoke the schema [[PROCESS/...iN...]-[PAST/...iN... > ...aeN...]], which identifies the consonant following the ablauted vowel only as a nasal (not a specific segment), and is further neutral as to whether or not this consonant is stem-final. Hass would be computable only from the topmost schema in Fig. 9, which is maximally schematic concerning the environment of the ablauted vowel. To make the same point in another way, consider the topmost schema in Fig. 10, which represents the maximal generalization concerning past- tense verbs in English. We have already noted that this schema is essentially vacuous at the phonological pole, because the various patterns of past-tense formation have virtually nothing in common. If this high-level schema were the one to be invoked for the computation or evaluation of novel expressions, it would have the effect of allowing any kind of stem to be marked for past tense in any way whatever — a maximally schematic characterization of the construction imposes no significant constraints on its possible instantiations. The actual work is obviously done by lower-level schemas. A final example should reinforce and further clarify this notion. Postpositional endings in Classical Nahuatl vary in the type of noun stems to which they attach. Certain postpositions, among them -ko 'in' and -tew
A USAGE-BASED MODEL 159 'like', attach only to lexical nouns (e.g. siwaa-tew 'like a woman1). Others, including -waan 'with' and -pampa 'because of, occur only with pronouns (e.g. no-waan 'with me'). A third class of postpositions, exemplified by -pan 'on' and -caalan 'among', suffix to nouns of either sort (e.g. to-caalan 'among us', kwaw-caalan 'among the trees'). Each of these distributional possibilities implies a low-level constructional subschema, depicted in abbreviated form along the bottom row in Fig. 11. For instance, the box at the lower left in this diagram stands for the constructional subschema expressing the generalization that -ko occurs on lexical nouns; the box at the lower right similarly expresses the generalization that -pampa combines with pronouns. Note that -pan and -caalan figure in two such subschemas each, one for each type of noun. N -caalan N -vaan N -parcpa pro pro pro * *^ Figure 11 From these low-level schemas, certain broader generalizations can be extracted. Since -ko, -tew, -pan, -caalan, and others occur on lexical nouns, a speaker could extract the intermediate-level subschema shown on the left in Fig. 11; it specifies the existence of a compositional pattern whereby postpositions suffix to lexical nouns. Similarly, the intermediate-level subschema on the right in Fig. 11, specifying the attachment of postpositions to pronouns, is supported by the occurrence on pronouns of -pan, -caalan, -waan, -pampa, etc. Moreover, since postpositions occur on both lexical nouns and pronouns, there are grounds for extracting the higher-level schema depicted at the top, which specifies that postpositions suffix to nouns. Each schema expresses the commonality of its immediate instantiations. It is clear, however, that only the lowest-level schemas could be invoked for the computation of novel expressions — it is only at this level that the distributional restrictions are apparent. If the topmost schema were
160 RONALD W. LANGACKER activated for this purpose, it would sanction the occurrence of any postposition with either type of noun. The intermediate-level schemas would fare no better, for they specify the occurrence of any postposition with either a lexical noun or a pronoun. Conclusion Accepting the general principles of scientific inquiry does not itself resolve the more specific but nonetheless crucial issue of how these principles are appropriately applied to the problems of a particular discipline at a given stage of its development. I have argued, both on methodological and on empirical grounds, that the principle of generality has received in linguistics a commonly accepted interpretation that is in fact not appropriate to its subject matter. Current doctrine favors a minimalist account of linguistic knowledge, described in accordance with a complex array of theoretical apparatus featuring specialized devices for the various "components" of the linguistic system. By contrast, cognitive grammar pursues a maximalist account of linguistic knowledge, and tends toward austerity in the adoption of theoretical constructs; it seeks a unified treatment of the various facets of linguistic structure, attributing their differences to the content of the domains in question rather than the basic constructs invoked to handle them. Prominent among these constructs are those comprised by the network model of complex categories. The network model affords an integrated account of categorization for the varied domains of linguistic structure. It accommodates not only those phenomena generally thought of as involving categorization, but also the nearest cognitive-grammar analogs of rules, derivations, and structural descriptions. It is further responsible for the usage-based character of the framework; by tolerating the coexistence in a single network of specific expressions and schemas at varying levels of abstraction, it implements the maximalist, non-reductive, bottom-up orientation of the usage-based approach. Problems of distribution and productivity are addressed by treating grammatical constructions as complex categories. A single, high-level generalization does not exhaust a speaker's conventional knowledge of a construction. A full description must also specify how this generalization is articulated through the supporting hierarchy of subpatterns and specific expressions. In this maximalist account, the structures that occur in a given
A USAGE-BASED MODEL 161 construction are identified without the use of diacritics or other arbitrary devices. Moreover, specific expressions and low-level schemas are seen to be at least as important as higher-level schemas capturing the broadest generalizations. We have no assurance that speakers invariably arrive at high-level schemas, whose abstractness may render them essentially useless for the computation and evaluation of novel expressions; there is reason to think that lower-level schemas figure more prominently in this role. A major advantage of the usage-based conception is its ability to accommodate structures at this level of organization without the loss of valid generalizations.
PART II ASPECTS OF A MULTIFACETED RESEARCH PROGRAM
The Relation of Grammar to Cognition1 Leonard Talmy University of California, Berkeley 0. Introduction A fundamental design feature of language is that it has two subsystems which can be designated as the grammatical and the lexical (as these are characterized below). Why is there this universal bifurcation when, in principle, a language could be conceived having only a single system, the lexical? The explanation in this paper is that the two subsystems have distinct semantic functions, ones that are indispensable and complementary. To develop this account further, we must first note that we take a sentence (or other portion of discourse) to evoke in the listener a particular kind of experiential complex, here to be termed a "cognitive representation" or "CR".2 Now, the grammatical and lexical subsystems in a sentence seem generally to specify different portions of a CR. Together, the grammatical elements of a sentence determine the majority of the structure of the CR, while the lexical elements together contribute the majority of its content. Lexical elements do incorporate some of the same structural indications that grammatical elements express, but when the two are in association or in conflict within a sentence, it is generally always the grammatical elements' specifications of structure that are determinative.3 The grammatical specifications in a sentence, thus, provide a conceptual framework or, imagistically, a skeletal structure or scaffolding, for the conceptual material that is lexically specified. More generally, across the spectrum of languages, the grammatical elements that are encountered, taken together, specify a crucial set of concepts. This set is highly restricted: only certain concepts appear in it, and ® Leonard Talmy
166 LEONARD TALMY not others, as seen below. The purport of the present paper is that this set of grammatically specified notions collectively constitutes the fundamental conceptual structuring system of language. That is, this cross-linguistically select set of grammatically specified concepts provides the basic schematic framework for conceptual organization within the cognitive domain of language. Thus, grammar, broadly conceived, is the determinant of conceptual structure within one cognitive domain, language, and as such is the main object of this paper's study. But such a study directly opens out into a broader investigation across other cognitive domains, such as visual perception and reasoning, as discussed at the end of the paper. That is, the greater issue, toward which the present study ultimately aims, is the general character of conceptual structure in human cognition. The present investigation into the semantics of grammar is of a scope that follows in a progression from previous types of study. These have mostly been either an in-depth semantic analysis of a selected grammatical element (or class of elements) of particular interest within a language, e.g., the Turkish evidential suffix -mi$ (Slobin & Aksu 1982); or an exposition of the meanings and functions of all the grammatical elements of a single language, say, as in a grammar of Dyirbal (Dixon 1972); or a cross-linguistic typology of the different kinds of grammatical devices used for a single semantic function, say, to indicate the interrogative (Ultan 1978). Some previous work has also treated broader issues of grammatical meaning (Sapir 1921, Boas 1938, Whorf 1956, Jakobson 1971). But the present study is perhaps the first to address grammatical expression in language at the superordinate level, with the aim of determining the semantic and cognitive properties and functions of this structural component of language as a whole.4 The terms "grammatical" and "lexical" as employed here require some immediate elaboration. The distinction between the two is made formally — i.e., without reference to meaning — on the basis of the traditional linguistic distinction between "open-class" and "closed-class". A class of morphemes is considered open if it is quite large and readily augmentable relative to other classes. A class is considered closed if it is relatively small and fixed in membership. We can identify the particular classes belonging to these two types. The open classes of elements — i.e., the lexical classes — are the roots of nouns, verbs, and adjectives.5 Everything else is closed- class — and is here considered to be, quite generally, "grammatical". Among the overt elements of this type are such bound forms as inflections
THE RELATION OF GRAMMAR TO COGNITION 167 and derivations, such free forms as determiners, prepositions, conjunctions, and particles, and perhaps also such suprasegmental forms as intonation patterns. Included among abstract, or implicit, closed-class forms are grammatical categories and grammatical relations, word order, and perhaps also paradigms and "zero" forms. Additionally here are regular combinations of simpler closed-class forms, tending to have a unified or integrated semantic function — what are below called "grammatical complexes", including grammatical constructions and syntactic structures.6 The issues presented in this introduction are treated below in three sections. Section 1 examines the notions specified by a heuristic sampling of grammatical elements, outlines the kinds of constraints on such notions, proposes a property held in common by such notions but largely absent from excluded notions, and contrasts such grammatically specified notions with ones that are lexically specified. Section 2 presents a number of categories in which grammatically specified notions are seen to pattern, as well as broader conceptual systems in which these categories in turn participate, ending with the identification of four comprehensive "imaging systems". This section, further, examines the interactions of grammatical specifications with lexical specifications within categories and the nesting of such interactions across categories, as well as the cognitive processes that accompany these interactions. And Section 3 presents an explanation of the function of grammatical specification, as well as possibilities of its relations to other cognitive systems. 1. The Nature of Grammatically Specified Notions In this section we examine a small sampling of grammatical forms for the particular component notions that they specify. The sample will give a heuristic indication of the kinds of notions that get grammatically specified as well as of the kinds of notions that possibly never do. By contrast, it will be seen that the excluded kinds can be readily specified by lexical elements. A particular property will be seen to run through most of the grammatical notions. To indicate this property at the outset, it is preponderantly the case that grammatical specifications of structure are relativistic or topologylike, and exclude the absolute or the metrically Euclidean. Finally, a systematic difference is shown between the characteristics of grammatically specified notions and of lexically specified ones. We begin with a simple demonstration that the concepts specified by grammatical forms are constrained in two ways: as to their categories and as
168 LEONARD TALMY to the membership of these categories. Many languages have inflections on the noun that specify the "number" of the object referred to by the noun, for example its 'singularity* or 'plurality', like the English -0 and -s. By contrast, no languages appear to have inflections that specify the "color" of the object referred to by a noun, e.g., its 'redness' or 'blueness'. Here, single quotes enclose "notions", while double quotes enclose categories of notions. The "number" category can be specified grammatically and in that form is readily seen to play a structuring role in a CR.7 The "color" category is perhaps never found specified by grammatical elements, though it is readily found specified by lexical elements, e.g., English red and blue. Further, though, even within a conceptual category acceptable for grammatical expression, there are great constraints on the particular notions that can be specified. Thus, "number" notions that are expressed grammatically include little more than 'singular', 'dual', 'trial', 'plural', and 'paucal'. They apparently never include, say, 'even', 'odd', 'dozen', or 'numerable', whereas such notions, again, can be specified lexically, as shown by the words just used. Given such constraints on grammatically specifiable notions, we can seek properties that hold in common for included notions but need not apply to excluded notions. In this regard, consider a deictic like the English this or that as in This/That chair is broken. A closed-class element of this type specifies the location of an indicated object as being, in effect, on the speaker-side or the non-speaker-side of a conceptual partition drawn through space (or time or other qualitative dimension). This integral specification can be analyzed as containing the component notions enclosed by quotes in (1): (1) (a,b) a 'partition' that divides a space into 'regions'/'sides' (c-e) the 'locatedness' (a particular relation) of a 'point' (or object idealizable as a point) 'within' a region (f,g) (a side that is the) 'same as' or 'different from' (h,i) a 'currently indicated' object and a 'currently communicating' entity Notions that might at first be ascribed to such deictics, such as of distance or perhaps size, prove not to apply, on the evidence of sentence-pairs like (2): (2) a. This speck is smaller than that speck, b. This planet is smaller than that planet.
THE RELATION OF GRAMMAR TO COGNITION 169 The scenes referred to by (2a) and (b) differ greatly, involving tiny objects millimeters apart or huge objects parsecs apart. But the sentences differ only lexically, not grammatically. Hence, the scenes' differences as to the magnitude of size or distance must arise from the lexical elements, they cannot be traced to the deictics (or other grammatical elements) in the sentences. Thus, the notions specified by a this or a that are abstracted away from any particularities of magnitude and so, to this extent, are genuinely topological. Their specification of a conceptual partition remains constant, but this partition's distance can — by the characterization of topology as "rubber-sheet geometry" — be "stretched" indefinitely without challenge to any semantic constraints of the deictics. This finding about deictics alerts us to noticing whether any grammatical elements make specifications about magnitude. A spot check through English and various other languages suggests that — while there are grammatical specifications for relative magnitude8 — there are possibly never any for absolute or quantified magnitude, whether of size, distance, or other parameters. We can provisionally conclude that the referents of grammatical elements have the topological property of being "magnitude-neutral". For another case, consider the type of adposition that specifies, for a moving object, certain characteristics of path and of reference-point or -frame. An example of this type is English through as used, for instance, in / walked through the woods. In this usage, through specifies, broadly, 'motion along a line that is within a medium'. The component notions contained here include those in (3): (3) (a) 'motion' (b-e) which can be understood as 'one-to-one correspondences' between 'adjacent' points of 'space' and adjacent points of 'time' (f) motion that describes a 'line' (i.e., a 'linear extent') (g) the locatedness of a line within a 'medium' (h,i) a medium, i.e., a region of three-dimensional space set apart by the locatedness within it of 'material' that is in a 'pattern of distribution' with properties and a range of variation still to be determined It can be first observed, from a sentence-pair like (4), that the concept specified by through is indifferent to particulars of shape or contour in the linear path described by the moving object. This is evident here because, as before, the two sentences differ only lexically, not grammatically — they
170 LEONARD TALMY both use through while referring to different path contours. Another cross- linguistic spot check of closed-class elements suggests that they largely have this further topological property of being "shape-neutral". (4) a. I zig-zagged through the woods, b. I circled through the woods. With a sentence pair like (5), it can be further determined that the 'rate' of motion is not specified by through, a finding that also appears quite general among grammatical elements. And (6) shows that through, again like grammatical elements generally, excludes specification of the 'kind of material' involved — here, comprising the "medium" — and of the 'sensorimotor characteristics' attendant on executing the action involved — as, here, those attendant on wading in liquid vs. weaving amidst obstacles. Thus, it can be further held that grammatical elements are generally rate-neutral, material-neutral, and sense/motor-neutral. (5) (a/b) I crept / dashed through the woods. (6) (a/b) I walked through the water / woods. In the aim of ascertaining any properties common to grammatically specified notions, the notions examined above are gathered together in (7). For heuristic purposes, the notions are provisionally divided into two groups on the basis of their relation to topology. In group (a) are combined the notions that properly belong to the specific mathematical system of topology and, with them, the intuitively comparable notions that might belong to a language-based system of topology — one that perhaps could serve as the model for the construction of a new topology-like mathematical system.9 In group (b) are the notions that fall outside any usual conception of topological properties. The number of notions in the first group is fourteen, while the second has six — an indication of a preponderant propensity for grammatical elements to specify quasi-topological notions. The ratio in this direction is in fact improved if we consider that even several notions in group (b) — the bottom three — resemble topological ones in the sense of involving relativistic relationships between quantities rather than absolutely fixed quantities.
THE RELATION OF GRAMMAR TO COGNITION 171 (7) some notions found to be specified by grammatical elements a. topological point linear extent locatedness within region side partition or topology-like singularity plurality same different "adjacency' ' of points one-to-one correspondence pattern of distribution b. non-top ological material space time motion medium entity currently indicated/communicating In the complementary aim of ascertaining any properties excluded from grammatical specification, the categories of notions found above not to be specified by the elements investigated are listed in (8). Rather than topological, topology-like, or relativistic, these notions involve Euclidean- geometric concepts — e.g., fixed distance, size, contour, and angle — as well as quantified measure, and various particularities of a quantity: in sum, characteristics that are absolute or fixed. (8) some categories of notions seemingly rarely or never specified by grammatical elements absolute/quantified magnitude kind of material (of distance, size, etc.) sensorimotor characteristics shape/contour of line color rate The provisional conclusion to be drawn from these findings is that, if grammatical specifications generally correspond to (linguistic-) cognitive structuring, then the nature of that structure is largely relativistic or topological rather than absolute or Euclidean. This preponderant requirement for conceptual neutralities among closed-class elements is in sharp contrast with the referential freedom of lexical items, which can express not only structural abstractions but also wide-ranging specificities. For example, specificity as to magnitude is seen in nouns like inch and mile; as to shape, in nouns like circle, adjectives like square, and verbs like ricochet; as to rate, in verbs like dawdle and hurry, in material, in a noun and verb like iron and bleed; as to sensorimotor characteristics in watch and wade; and, of course, as to color by such adjectives as red and blue. To elaborate further the contrast between the grammatical and the lexical type of specification, consider the full complement of both element-
172 LEONARD TALMY types in a single whole sentence, viz., that selected in (9): (9) A rustler lassoed the steers. We first list the grammatical elements present in the sentence and the notions that they specify in (10): (10) a. -ed b. the d. -s e. a...-0 f. the grammatical category of "verb" for lasso g/h. the grammatical category of "noun" for rustler/steer i/j. the grammatical relations of "subject"/"object" for rustler/steer k. active voice 1. intonation, word order, pattern of auxiliaries 'occurring at a time before that of the present communication' 'has ready identifiability for the addressee' 'not before in discussion or otherwise readily identifiable for the addressee' 'multiple instantiation of object' 'unitary instantiation of object' 'event character' 'entity character' 'agent'/'patient' (among the possibilities) 'point-of-view at the agent' 'the speaker "knows" the situation to be true and asserts it' The lexical items in the sentence have specifications that can be characterized as in (11): (11) a complex of concepts involving: a. rustler: a person, property ownership, illegality, mode of activity b. steer: object of particular appearance, physical makeup, etc. relation to animal kingdom castration institution of breeding for human consumption c. lasso: certain objects (a body and a lasso) in particular configurations
THE RELATION OF GRAMMAR TO COGNITION 173 certain movement sequences accompanying cognitive intending, directing, monitoring, etc. In surveying the two lists, we can see these differences emerge: The grammatical elements are more numerous, and their specifications seem more spare and simpler, and more structural in function. Together, their specifications seem to establish the main delineations of the scene organization and communicative setting of the CR evoked by the sentence. The lexical elements are fewer in number, but their specifications are greater in quantity and complexity, and function more to contribute content than structure. The lexical specifications are greater in three ways: compared to a grammatical specification, each has a) more total information, b) greater intricacy of information, and c) more different types of information together. Taken together, their specifications comprise most of the conceptual content of the CR scene that is evoked by the sentence. These grammatical-lexical differences can be set into further relief by in turn varying each element-type while keeping the other constant. Thus, varying only the grammatical elements of (9), as is done in (12), seems to alter the scene organization and discourse properties of the referent event but to leave its basic contents intact: (12) Will the rustlers lasso a steer? By contrast, varying only (9)'s lexical elements, as in (13), shifts us to a new scene altogether, and yet the basic breakup of the scene and of its communicative setting seems to remain the same: (13) A machine stamped the envelopes. 2. Categories of Grammatically Specified Notions The preceding sampling of grammatical elements has yielded a set of notions helpful toward discovering common semantic properties. But the set has been small and unstructured. With a broader and more systematic investigation, patterns of organization among the notions become evident. Grammatically specified notions can be seen to pattern in categories, and the categories, in turn, in integrated systems, as presented below. And within these notional patterns can be seen certain regularities of function and process. These patterns and regularities constitute principal features of
174 LEONARD TALMY conceptual organization in language. Several such features are brought forward below. One feature is an extensive homology between the representation of space and that of time. The first category, "dimension", includes this space-time homology, and largely crosscuts the remaining categories. These categories will, in the majority, apply to both space and time, and parallel examples from each dimension will be presented side by side. Another feature is that, of the member notions of any category represented in a language, often each notion will be incorporated in at least some lexical items. Correlatively, the language will often contain grammatical forms that interact with each lexicalization type in a way that yields the expression of another notion of the category. Each such type of interaction can be regarded as a type of cognitive operation that converts the indication of one notion to that of another within the same category. A corollary feature is that a language with grammatical forms for converting from notion A to notion B frequently has forms as well for conversion in the reverse direction — that is, it can also trigger the reverse cognitive operation.10 Some of the grammatical forms in a language function specifically to perform a particular conversion operation. Others simply make structural specifications that can come into conflict with the specification of a neighboring lexical item. In such cases, the basic pattern is that the grammatical form's specification always takes precedence, and triggers a kind of operation, a "shift", in the lexical item's referent that brings it into accord.11 As a note on methodology in what follows, efforts were made to determine categories on the basis of particular grammatical meanings encountered, rather than to posit the categories as part of an a priori schema which then sought corroborative examples. In the research leading to this paper, grammatical forms were sampled from a range of languages, but an effort has been made to take most of the exemplification from English. 2.1 Dimension The category of "dimension" has two principal member notions, 'space' and 'time'. The kind of entity that exists in space is — in respectively continuous or discrete form — 'matter' or 'objects'. The kind of entity existing in time is, correspondingly, 'action' or 'events' — terms here used neutrally as to whether the entity is static or changing. These notions thus relate as in (14):12
THE RELATION OF GRAMMAR TO COGNITION 175 (14) dimension space: time: continuous matter action discrete objects events Homologies between the linguistic structuring of space and of time will be indicated in the categories that follow. But here we can indicate operations of conversion between these two main members of the dimension category. Thus, a verb root that lexicalizes expression of an event or of action as a temporal quantity can be associated with grammatical forms, including nominalizations, that signal a cognitive operation of "reification". By the semantic effect of this operation, the referent becomes conceptualized as an object or a mass, one that can participate in many of the same activities (such as being given or gotten) as a physical quantity, as well as in many of the corresponding syntactic constructions (including pluralization and modification: ...gave me two quick calls), as exemplified in (15). (A way of representing the grammatical complexes involved here and in the next operation is presented in connection with the following category.) (15) an event: reified as an object: John called me. John gave me a call. I was called by John. I got a call from John. action: reified as mass: John helped me. John gave me some help. I was helped by John. I got some help from John. The reverse conversion also occurs. A noun referring to an object or mass can be associated with grammatical forms, including verb-forming derivations, that signal a cognitive operation of "actionalizing". By this operation, the physical referent is melded together with some of the activity in which it participates, with the semantic effect that much of the referent's tangible concrete character is backgrounded, subordinated to a conceptualization in terms of a process of occurrence, as illustrated in (16): (16) object(s)/mass: actionalized as: Hail(stones) came in through It hailed in through the window the window.
176 LEONARD TALMY Ice is forming over the windshield. It is icing up over the windshield. I removed the pit from the cherry. I pitted the cherry. 2.2 Plexity The category here to be termed "plexity" is a quantity's state of articulation into equivalent elements. Where the quantity consists of only one such element, it is "uniplex", and where it consists of more than one, it is "multiplex". When the quantity involved is matter, plexity is, of course, equivalent to the traditional linguistic category of "number" with its component notions "singular" and "plural". But the present notions are intended to capture the generalization from matter over to action, which the traditional notions do not do.13 Specifications as to plexity are made by both lexical items and grammatical elements, and there is interplay between the two when they are both in association. Example English lexical items that basically specify a uniplex referent are — for matter and action, respectively — bird and (to) sigh. They can occur with grammatical elements that themselves specify a uniplexity, like those italicized in (17a) (many languages have here a more regular, overt system of markers than English). But they can also occur with grammatical elements that specify a multiplexity, as in (17b). In this association, such elements can be thought to trigger a particular cognitive operation, one of "multiplexing". By this operation, an original solo referent is, in effect, copied onto various points of space or time. (17) matter action a. uniplex A bird flew in. He sighed (once). b. multiplex Birds flew in. He kept sighing. The reverse of the preceding pattern is also found in language. First, there are lexical items that intrinsically specify a multiplexity. English examples are furniture or timber (i.e., 'standing trees') for matter and breathe for action, as used in (18a). And, too, there are grammatical forms able to appear in association with these, as in (18b), that signal an operation the reverse of multiplexing — one that can be called "unit-excerpting". By this operation, a single instance of the specified equivalent units is taken and set in the foreground of attention.
THE RELATION OF GRAMMAR TO COGNITION 177 (18) matter action a. multiplex Furniture overturned in the earthquake. She breathed without pain. b. uniplex A piece 0/furniture overturned in the earthquake. She took a breath / breathed in without pain. The English grammatical forms seen above that signaled multiplexing s and keep Ang— consisted solely of explicit morphemes. The forms that signal unit-excerpting differ in that they also include abstract elements: particular grammatical categories that require the insertion of one out of a certain set of lexical items, as represented in (19c,d). The forms can, moreover, contain two or more independent elements. These forms are here considered to be "grammatical complexes", comparable to other grammatical constructions or indeed to lexical complexes (collocations): they combine distinct elements within a structural whole serving a single overall semantic function. Actually, by one analysis, all grammatical forms are complexes, merely ranked along a cline of elaborateness. Under this analysis, a grammatical form includes not only any explicit and generic elements, but also the semantic and syntactic category memberships of its input and output forms, as represented throughout (19). Thus, the English multiplexing forms, in (19a,b), are merely at the simpler end of a continuum: (19) » r [ In + s upx N mpx e.g., bird: birds (b) keep + [ ]v + -ing upx V mpx e.g., sigh: keep sighing
178 LEONARD TALMY (c) Nunilo/ + [_]N N J UpX e.g., furniture: a piece of furniture (d) Vdummv. + [[ ]v +DERIV]N mpx upx upx e.g., breathe: take a breath (<n + PTC V upx e.g., breathe: breathe in Support is lent to the thesis that a more elaborate grammatical complex can have a semantic unity by the existence, within the same or another language, of a simpler form with the same semantic function. As an example of just this circumstance, the English unit-excerpting complex for nouns, which is rather elaborate, is paralleled in function by a simple suffix in Yiddish, either - " / or - " ele (otherwise indicating diminutives), as illustrated in (20): (20) zamd'sand': zemdl'grain of sand' groz 'grass': grezl 'blade of grass' shney 'snow': shneyele 'snowflake' 2.3 State of Boundedness When a quantity is specified as "unbounded", it is conceived as continuing on indefinitely with no necessary characteristic of finiteness intrinsic to it. When a quantity is specified as "bounded", it is conceived to be demarcated as an individuated unit entity. In application to nouns, these notions largely correspond to the traditional linguistic distinction between "mass" and "count", and in application to verbs they can correspond to
THE RELATION OF GRAMMAR TO COGNITION 179 "imperfective" and "perfective", among other terms (the closeness of these correspondences varies with different usages of the traditional terms). However, as with plexity, the concepts designated by the new terms are intended to capture the commonality between the space and time dimensions and to generalize over their usually separate analyses. Among English examples of lexical items, water and (to) sleep basically specify unbounded quantities, whereas sea and (to) dress basically specify bounded ones. These specifications are demonstrated by the fact that these words are, respectively, unacceptable and acceptable in construction with the grammatical complex "in NP _ . ", which itself specifies bounded- ° r extent-of-time ' r ness, as seen in (21): (21) matter action a. unbounded *We flew over water in 1 hr. *She slept in 8 hrs. b. bounded We flew over a sea in 1 hr. She dressed in 8 mins. As with plexity, there exist grammatical elements that can, in construction with a lexical item, shift its basic specification for state of boundedness to the opposite value. Those acting in this way on an unbounded-type lexical item, in effect, trigger a cognitive operation of "bounding", or "portion- excerpting". By this operation, a portion of the specified unbounded quantity is demarcated and placed in the foreground of attention. Examples of such grammatical elements in English are shown in (22). The reverse of the preceding pattern also exists. The English nouns shrub and panel each refer intrinsically to a bounded entity. But the grammatical elements -ery and -ing can be added to them, yielding shrubbery and paneling, forms which now refer to unbounded quantities. In effect, the grammatical elements have triggered a cognitive operation of "debound- ing" whereby the quantity formerly within bounds is now conceptualized in a form with indefinite extension. In English, however, such elements are not productive; they cannot, for example, be used with sea to yield the meaning 'pelagic water', nor with (a) tear to yield 'lachrymal fluid'.14
180 LEONARD TALMY (22) matter: N- uantity °f + I Jn bounded-quantity e. g., water: body of water action: [ 'Vunbd + f°T Nextent-of-time e.g., sleep: sleep for an hour unbd IN bd 'bd 2.4 State of Dividedness The category of "state of dividedness" refers to a quantity's internal segmentation. A quantity is "discrete" (or "particulate") if it is conceptualized as having breaks, or interruptions, through its composition. Otherwise, the quantity is conceptualized as "continuous".15 Both lexical and grammatical elements are sensitive, in their specifications, to the distinctions of this category. But there appear to be no grammatical elements that solely specify discreteness or continuity for a quantity, nor any that signal an operation for reversing a quantity's lexically specified state of dividedness. If forms of the latter type existed, we can describe how they would behave. A grammatical form for a continuous- type lexical item would signal an operation of "discretizing", whereby the originally continuous referent would become conceptualized as a particulate aggregation. Conversely, a grammatical form for a discrete-type lexical item would trigger an operation of "melding", whereby the separate elements of the original referent would be conceptualized as having fused together into a continuum. Although such grammatical forms seem lacking, there do exist certain indirect or inexplicit mechanisms for these operations. Thus, the continuity specified by the noun water can be reconceptualized as discrete with the locution particles of, as in: Water I Particles of water filled the vessel. However, the grammatical complex used here does not directly specify this shift but, like the complexes in Sections 2.5 and 2.13, comprises a several-stage sequence of other cognitive operations. In the reverse direction, there
THE RELATION OF GRAMMAR TO COGNITION 181 appears to be a general conceptual tendency for a basically discrete-type referent of a lexical root to undergo at least some degree of spontaneous melding, without the addition of any explicit grammatical forms. Thus, foliage, timber, and furniture, as contrasted with leaves, trees, and pieces of furniture, tend to evoke referents with a degree of blurring and fusion across their component elements. Because the category of dividedness has limited realization by itself, further treatment of it will be deferred until the next section, where it can be seen in interaction with the other categories. 2.5 The Disposition of a Quantity: A System of Categories The preceding four categories of attributes — dimension, plexity, boundedness, and dividedness — all pertain to a quantity simultaneously and, taken together, can be considered to constitute a system of attributes that may be termed a quantity's "disposition". The intersections of these categories form an array that can be schematized as in (23). (23) discrete continuous • • * • * ^::£ip:£::iii:JL unbounded A' B' " "l:" multiplex bounded uniplex + the distinction between matter and action, which crosscuts all of the above16
182 LEONARD TALMY Each intersection of attributes indicated here is specified by various lexical items (although one, a bounded multiplexity for action, is quite minimally represented in English). An example or two (most seen earlier) is given for each intersection in (24):17 (24) A': timber/furniture B': water (to) breathe (to) sleep A: (a) family B: (a) sea/panel (to) molt (to) empty (The bird molted.) (The tank emptied.) a: (a) bird (to) sigh Now if the particular contentful referent for which one chooses a lexical item happens to be wedded, by that lexical item, to an unwanted set of structural specifications, there generally are grammatical means available for converting this to a desired set. Such means range in directness from specifying the single relevant operation to involving a circuitous sequence of operations (cf. Section 2.13 on "nesting"). A number of starting- and ending-points for such conversions, and the means for accomplishing them, are indicated in (25): (25) A'—»A a stand of timber B' —»B a body of water breathe for an hour sleep for an hour A'—»a a piece of furniture — take a breath / breathe in A—»a a member of a family — ?molt a single feather A—»A' members of a family B—»B' paneling (A-»a-»A#) molt and molt empty and empty a—»A' trees keep sighing a—»A a stand of trees — (a-»A'-»A) sigh for a while
THE RELATION OF GRAMMAR TO COGNITION 183 2.6 Degree of Extension Implicit in the vertical dimension of the schematic arrangement in (23) is a further category that can be called "degree of extension". This category has three principal member notions, terms for which are given in (26) together with schematic representations of the notions for the linear case. Lexical items referring to either matter or action may be taken to incorporate specifications as to their referent's basic degree of extension, and three examples of these for the linear spatial case are also shown in (26):18 (26) point bounded extent unbounded extent • • • • • speck ladder river Now a lexical referent that is perhaps most basically conceived as of one particular degree of extension can, by various grammatical specifications that induce a shift, be reconceptualized as of some other degree of extension. For a first example, consider the event referent of climb a ladder, which seems basically of bounded linear extent in the temporal dimension, as is in fact manifested in (27) in conjunction with the grammatical element "m + NP , . ": extent-of-time (27) She climbed up the fire-ladder in 5 minutes. With a different accompanying grammatical form, like the "at + NP . tjmc" in (28), (as well as different contextual specifications), the event referent of the preceding can be shifted toward a conceptual schematization as a point of time — i.e., as being point-durational: (28) Moving along on the training course, she climbed the fire-ladder at exactly midday. This shift in the cognized extension of the event can be thought to involve a cognitive operation of "reduction" or, alternatively, "adoption of a long-range perspective". This shift can also go in the other direction. The event referent can be conceptually schematized as an unbounded extent by the effect of grammatical forms like "keep -ing'\ "-er and -ef\ and "as + S", as in (29): (29) She kept climbing higher and higher up the fire-ladder as we watched. Here there would seem to have taken place a cognitive operation of "mag-
184 LEONARD TALMY nification", or "adoption of a close-up perspective". By this operation, a perspective point is established from which the existence of any exterior bounds falls outside of view and attention — or, at most, is asymptotically approachable. The preceding event referent was continuous, but a discrete case can exhibit the same shifts in extension. One such case, perhaps to be considered as most basically of bounded extent, is shown with that degree of extension in (30a). But the referent can also be idealized as a point, as in (30b) (clearly, the cows would not all have died at the same moment, and yet the spread of their death times is conceptually collapsed into such a single moment). Or, the referent can be schematized as an unbounded extent, as in (30c): (30) a. The cows all died in a month. b. When the cows all died, we sold our farm. c. The cows kept dying (and dying) until they were all gone. The alternative schematizations of extension just seen as specifiable for an event referent are generally also available for an object referent. Thus, e.g., the referent of (a) box can be specified for idealization as a point or as a bounded extent (of area or volume). Some grammatical elements making such specifications are illustrated in (31). Also set forth here are the homologies between these and the event-specific elements: (31) point The box is 20 feet away from the wall. I read the book 20 years ago. bounded extent The box is 2 feet across. I read the book in 2 hours. (point within) bounded extent The ball is in the box. She left while I read the letter. 2.7 Pattern of Distribution The pattern of distribution of matter through space or of action through time is a further category of notions that can be both grammatically and lexically specified.19 For action through time — the only dimension we will be looking at here — this category together with the preceding one largely constitute the traditional category of "aspect".
THE RELATION OF GRAMMAR TO COGNITION 185 Several of the main patterns of distribution for action through time are shown schematically in (32) (the dots here, which represent situatedness in complementary states, should really be adjacent, but they are sketched apart with a connecting line to show the crossing of state-interfaces). Also shown are illustrative English verbs, both non-agentive and agentive, that incorporate these patterns. (32) one-way one-way full-cycle multiplex steady-state gradient non-resettable resettable die fall flash breathe sleep widen (intrans) kill drop hit beat carry widen (trans) One can determine that these verbs incorporate the specifications indicated by noting the grammatical forms with which they can and cannot occur (or, to put the latter case in our terms: ...grammatical forms toward whose specifications they will not [readily] shift). A full demonstration is not in order here, but a few examples show the principle: The resettable type of a one-way event is distinguished from the non-resettable type by its compatibility with iterative expressions, as in: He fell 3 times; the non-resettable type cannot occur here: *He died 3 times. This same one-way form is distinguished from a full-cycle form by its ability to appear in sentences like: He fell and then got up, which the latter cannot do: *The beacon flashed and then went off. A gradient type can appear with adverbs of augmentation, as in The river progressively widened, unlike a steady-state type: * She progressively slept. And so on. Grammatical elements can, of course, also specify differing patterns of temporal distribution, and the present form of diagramming can readily reveal some of their distinctions. Thus, the closed-class elements back and again, singly and in combination, can indicate versions of full-cycle, sesqui- cycle, and double-cycle patterns, as shown in (33):
186 LEONARD TALMY (33) one-way full-cycle sesqui-cycle double-cycle go to sleep go back to sleep go to sleep again go back to sleep again Now consider the circumstance where a verb of one distribution type appears with grammatical forms of another type. The resultant seems invariably to be that the verb shifts its specifications into conformity with those of the grammatical forms. For an example we again take die, whose basic specifications can be adjudged as point-durational one-way non-resettable — schematizable, now more precisely, as: % . This verb is used with its basic specifications in a sentence like (34a). But in a sentence like (34b), the grammatical form "be + -ing" induces a shift. In effect, the infinitesimal interval between the two states involved for die — viz., 'aliveness' and 'deadness' — is spread out, with the creation thereby of an extent-dura- tional gradient. This is the shift in the distribution pattern's structural type. But concomitantly, a shift in the basic contentful referent is engendered. Instead of 'dying', the new gradient refers to 'moribundity'. The distinction becomes clear in noting that, as the conception is structured linguistically, one can have been dying without having died, and, correlatively, one can have died without having been dying.20 (34) a. He died as she looked on. b. He was (slowly) dying as she looked on. 2.8 Axiality. The adjectives in a pair like well/sick behave contrarily when in association with grammatical forms specifying degree like slightly and almost, as seen in (35a), and they select for different readings of temporal forms like "in + NPcxtcnt_of_time"> as seen in (35b). In this, perhaps surprisingly, they parallel the behavior of certain kinds of expressions that specify spatial relations, e.g., at the border I past the border: (35) a- He\ sliphtlv / S'Ck ' PaSt thC b°rder- \ He s slightly | *weU / *at the bordei, j awake asleep
THE RELATION OF GRAMMAR TO COGNITION 187 TT , , f well / at the border. ) He s almost < 0 . . . 0 , . , > ( ysick / ?past the border. J b. He got well / to the border in 5 days. — i.e., progressively in the course of He got sick / past the border in 5 days. — i.e., first after the elapse of This behavior can be accounted for by positing that such adjectives, in referring to a more generic notional parameter, such as that of 'health', are not simply "opposites" but, rather, presuppose a schematic axis that is structured and directed in a particular way. Each adjective, then, labels a different portion of that axis. The adjectives here seem in particular to presuppose a directed line bounded at one end; well refers to the end-point while sick refers to the remainder of the line, correlating greater magnitude with greater distance along the line. These are the "axial properties", or "axiality", of the lexical items, i.e., the specific relations each has to a particular conceptual axis and to other lexical items with referents along the same axis. It is the lexicalization of such axiality that can align adjectives with expressions of spatial relation. Grammatical forms like the ones just above also have axial properties, and these can function in consonance with those of a lexical item, as in the acceptable cases of (35), now schematized as to axiality in (36): (36) past the at the border border is slightly sick / past the border is almost well / to the border got well / to the border in 5 days got sick / past the border at dawn (after 5 days) sick well
188 LEONARD TALMY In other cases, though, the axiality of a grammatical form can conflict with that of a lexical item and, accordingly, can cause the latter to shift. Thus, sick in (37) — now associated with grammatical forms that refer to an end- point — shifts from its basic "directed shaft" type of axiality, and indeed from its reference to an axis of 'health'; it now specifies the end-point of an axis pertaining to 'feeling physically bad'. (37) (After exposure to the virus, he felt worse and worse and) he was almost sick at one point. / he finally got sick in 3 days.21 2.9 Perspectival Mode As seen earlier, a particular event, whether static or changing, can have a pattern of distribution through time that is perhaps most basically associated with or intrinsic to it in its own right. But, in addition, language has the means for specifying an independent schema as to how one is to attend to the event. This schema includes the location and deployment of the perspective point one adopts from which to regard the event and the distribution of one's attention over the event. This category of specifications, here called the "perspectival mode", can either conform with or diverge from the event's own basic pattern of distribution. Two principal members of the category are characterized in (38): (38) the assuming of: a. a steady-state long-range perspective point with global scope of attention b. a moving close-up perspective point with local scope of attention For illustration, consider first an example with a basically steady-state referent, viz., objects in location. The (38a) type of perspectival mode — the one more congruent with such a referent — is invoked in (39a), multiply specified there by the set of grammatical forms shown underlined, namely, plural forms, an adverbial expression of spatial dispersion, and the locative preposition in. But these can be replaced by grammatical forms coding for the (38b) perspectival mode — as in (39b) with singular forms, an adverbial expression of temporal dispersion, and the motion preposition through. Thereby, the evoked CR is converted to one where one's perspective and attention or one's own projected location shifts in turn from object to object. In effect, a steady-state multiplexity of objects has been converted
THE RELATION OF GRAMMAR TO COGNITION 189 to a sequential multiplexity of events consisting of conceptualized encounters with the objects. (39) a. There are houses at various points in the valley. b. There is a house every now and then through the valley. For representing certain static spatial configurations, the moving-perspective mode, though non-congruent in character, is greatly favored over the steady-state mode. Thus, the ready colloquial formulation of (40b) for moving-perspective is matched in the global steady-state mode of (40a) only by a stilted scientific style: (40) a. The wells' depths form a gradient that correlates with their locations on the road. b. The wells get deeper the further down the road they are. The reverse of the preceding circumstances also exists. That is, a sequential multiplexity of events, an example of which is represented in (41a) with the more congruent moving-perspective mode, can also become the object of a fixed global viewing, as represented in (41b). Metaphorically, the effect here is as if the vertical time line is tilted up into present- moment horizontality for integrated or summational assessment. (41) a. I took an aspirin time after time during / in the course of the last hour. b. I have taken a number of aspirins in the last hour.22 2.10 Level of Synthesis The category to be considered now pertains to bounded quantities, like those schematized in the A/B row in (23). One form of locution already seen to specify such quantities is the particular type of "NP of NP" construction illustrated in (42a). Here the second NP specifies the identity of the quantity involved, itself conceptualized as without intrinsic bounds, while the first NP specifies the bounding, or "portion-excerpting"', per se of the quantity. Moreover, in addition to such a pure operation of bounding, the first NP can further specify the particular form or configuration that the excerpted portion has, as in (42b):23 (42) a. a set of trees a body of water b. a cluster of trees a puddle/drop of water The two NPs here can be seen as coding for two different "levels of synthesis". Describing this for the internally discrete case, e.g., a cluster of trees,
190 LEONARD TALMY we can say that the second NP specifies an unsynthesized multiplexity of independent elements, while the first NP specifies a particular Gestalt synthesized out of that multiplexity. Furthermore, language can mark an additional cognitive distinction here. Either level of synthesis can be placed in the foreground of attention while the other level is placed in the background. One grammatical device for marking this is the placement of the foregrounded NP at the head of the larger nominal construction (in English, placing it first), as shown in (43a). With the use of this device, moreover, predications can be made that pertain solely to one level of synthesis or the other, as seen in (43b): (43) a. the cluster of trees / the trees in the cluster b. That cluster of trees is small. The trees in that cluster are small. There are certain forms, furthermore, whose referents are keyed to applying to only one or the other level of synthesis. Thus, together (toward each other) tends to correlate with multiple objects at large, while in upon -self tends to correlate with a composite formed therefrom, as seen in (44): (44) The bricks in the pyramid came crashing together / *in upon themselves. The pyramid of bricks came crashing in upon itself / together. The preceding phenomena have involved the shift of attention from a multiplexity to a Gestalt that it can constitute, a process that can be called "Gestalt formation". Also encountered in language are means for specifying the reverse: shifting attention from a Gestalt to components seen as constituting it, in a process of "componentializing". This procedure can take place when the starting lexical item specifies an entity taken to be already at the more synthetic level, as is the case with iceberg in (45a). By grammatical devices like those in (45b), such an entity can be analytically converted from conception as a coherent whole to one of component parts and their interrelations. Again we encounter a surface form — in two — that correlates with only one level of synthesis and not the other: (45) a. The iceberg broke in two. b. The two halves of the iceberg broke apart (*in two). The two levels of synthesis with the two directions of conceptual shift applicable to them define four notional types, as indicated in (46). The term
THE RELATION OF GRAMMAR TO COGNITION 191 Figure is used here as described in Talmy (1978b, 1983). (46) example type operation cluster of trees: "composite Figure" 1 trees: "multiple Figures" I _Gestaltjf orm ation _ iceberg: "meta-Figure" I two halves of iceberg: "component Figures" T componentialization 2.11 Level of Exemplarity A further cognitive distinction can be specified for a multiplexity of objects- This distinction does not affect the basic reference to all the members of the multiplexity, but addresses how attention is directed and distributed within that multiplexity. Either the/w// complement of the multiplexity is in the foreground of attention, with perhaps individual items here and there singled out in the background of attention. Or a single exemplar out of the multiplexity is placed in the foreground of attention, with the remaining items more dimly conceived in the background of attention. This distinction as to "level of exemplarity" is specified by grammatical devices in perhaps most languages. But English stands out in the extensiveness of its specifications: there are different pairs of grammatical forms that mark the distinction for a number of different types of multiplexity. A rather full list of these pairs is indicated in (47), with examples showing first the full-complement form and then the counterpart exemplar form: (47) a. Oysters have siphons / a siphon. An oyster has siphons / a siphon.24 b. All oysters have siphons / a siphon. Every oyster has siphons / a siphon. c. All the members raised their hand(s). Each member raised his hand(s).25 d. Many members raised their hand(s). Many a member raised his hand(s). e. Some members here and there raised their hand(s). A member here and there raised his hand(s).
192 LEONARD TALMY f. Members one after another raised their hand(s). One member after another raised his hand(s). g. Hardly any members raised their hand(s). Hardly a member raised his hand(s). h. No members raised their hand(s). No member (Not a member) raised his hand(s). i. She held a gun in both hands. She held a gun in either hand.26 2.12 Other Categories and Processes A number of further notional categories and cognitive processes can be discerned in language, but there is opportunity here to present briefly only two additional examples: Scene-Division Properties. A lexical item can have particular "scene- division properties", that is, a principal breakup of its referent into parts and participants. For example, the referent of the English verb serve breaks up into an activity, an item served, and a social dyad involving the two roles of 'host' and 'guest' — this last being its particular "personation" type (Talmy 1985a) — as illustrated in (48a). But grammatical forms can also have scene-division properties. Thus, a subject-plus-reflexive-object complex has a single-role specification. When such a grammatical form occurs with a dyadic verb like serve, it triggers an operation of "monad-formation": the verb's referent shifts to one with monadic personation, as in (48b). In this shifted state, its referent is equivalent to that of an intrinsically monadic expression, like that in (48c):27 (48) a. The host served me some dessert from the kitchen. b. I served myself some dessert from the kitchen. c. I went and got some dessert from the kitchen. Associated Attributes. Lexical expressions like apartment and hotel room, in addition to their basic denotations, may be taken to have "associated attributes" — here, respectively, those of 'permanent residence' and 'temporary lodging.' Such attributes may mesh or conflict with the specifications of another element in the same sentence. The attributes of the above two nominals mesh and conflict respectively, e.g., with the closed-class directional adverb home, which specifies a permanent residence. In the case of conflict, as in (49b), a cognitive process operates on the lexical item to leave its essential characteristics intact but replace its associated
THE RELATION OF GRAMMAR TO COGNITION 193 attributes with the closed-class element's specifications: (49) a. He drove home to his apartment, b. He drove home to his hotel room. 2.13 Nesting: An Interaction of Categories A number of what can be considered "meta-properties" govern the behavior of categories of grammatical notions, in general and with respect to one another. One of these, the capacity for nesting, already illustrated in Section 2.5, can be readily presented here: The operations and shifts described throughout Section 2 need not take place singly. The output of one can serve as the input to another, thereby building up hierarchical levels of embedding. While there are a number of interesting examples of this for different types of matter and action, we will go directly to illustrating one of the longer cases in (50): (50) a. The beacon flashed (as I glanced over). b. The beacon kept flashing. c. The beacon flashed 5 times in a row. d. The beacon kept flashing 5 times at a stretch. e. The beacon flashed 5 times at a stretch for 3 hours. In (50a), the lexical verb flash appears with its basic structural specification as a point-durational full-cycle uniplex event. This undergoes the operation of multiplexing, to yield the unbounded multiplexity in (50b). This then undergoes bounding in (50c). This bounded multiplexity then first goes through the operation of reduction to become schematized as a new pointlike uniplex quantity, and this is in turn multiplexed, yielding (50d). This new unbounded multiplexity is finally then bounded in (50e). The nesting of structural specifications in this last stage can be represented schematically as in (51): (51) [( )-( ) ( )-( )] Quite analogous to this temporal nesting, except for the lack of specific numerals, is the spatial example in (52): (52) a. A duck landed on the pond. b. Ducks landed on the pond. c. A flock of ducks landed on the pond. d. Flocks of ducks landed on the pond. e. A group of flocks of ducks landed on the pond.
194 LEONARD TALMY 2.14 Four Imaging Systems Most of the preceding categories of grammatically specified notions, together with categories not discussed here, group together under four much broader conceptual systems, ones that can be understood as principal "imaging systems" of language. These are great complexes in language that organize the structuring and the "viewing" of conceptual material. The four systems outlined here (there are additional ones) are relatively independent of each other in content, with each adding a distinct conceptual parameter to those of the others, but their contributions can be coordinated and linked, at times by individual grammatical forms. The first imaging system is "structural schematization". This system comprises all the forms of conceptual delineation that can be ascribed to a quantity, or to the pattern in which two or more quantities are interrelated, whether in space or time or some other conceptual dimension. A number of the categories of notions presented above are part of this system. After "dimension", all the categories pertaining to the disposition of a quantity and its generalizations belong here, in particular, the categories of plexity, state of boundedness, state of dividedness, degree of extension, pattern of distribution, and axiality. Belonging here, too, are the category of scene-division properties and that of the "partitioning" of space or time that is specified by such deictics as this and that, described in Section 1. A further major component of this imaging system is the spatial or temporal "geometric" schematization, including the topology-like kind, that is specified especially by the adpositional systems of languages. This was only touched on here in Section l's discussion of the English preposition through, but it is an extensive domain, one treated at length in such works as Bennett (1975), Gruber (1965), Jackendoff (1977), Langacker (1986a), Talmy (1975, 1982, 1983), Herskovits (1986). The second imaging system is the "deployment of perspective". Given a structurally schematized scene, this system pertains to how one places one's "mental eyes" to look out upon that scene, including the location, the distance away, and the movement pattern of this conceptual perspective point. Belonging to this system from the discussion above is the category of perspectival mode, with its options of a steady-state or a moving perspective point. Also here is the category of degree of extension, when its alternatives are interpreted as "adopting a long-range vs. a close-up perspective". The third imaging system is "distribution of attention". Given a
THE RELATION OF GRAMMAR TO COGNITION 195 schematized scene and a vantage from which to regard it, this system pertains to the allocation of attention that one can direct differentially over the aspects of the scene. Belonging here from the discussion above are the categories of level of synthesis and level of exemplarity, as well as the component of the perspectival mode category that involves global vs. local scope of attention. In addition, a major category, not treated here, comprises the obligatory "Figure/Ground" distinctions that language imposes on a referent scene. Here, within a scene, there is ascribed to one element the status of "Figure", with its attentional primacy, and to another element the status of "Ground", with its function in the background of attention as a reference object for the localizing of the Figure (Talmy 1978b, 1978c, 1983). Additionally in this imaging system are such grammatically marked discourse concepts as focus, topic, comment, given and new. The fourth imaging system, not treated here at all, is "force dynamics", which, given a structured scene, involves the forces that the elements of the scene exert on each other. Comprehended here are the notions of force exerted by one quantity on another, as well as notions of resistance to such force, the overcoming of such resistance, blockage to the exertion of force, and the removal of such blockage. The system of force dynamics includes the traditional linguistic concepts of the "causative", but is a generalization over those concepts (Talmy 1976, 1985b). 3. Further Cognitive Connections Grammatically specified structuring appears to correspond, in certain of its functions and characteristics, to the structuring in other cognitive domains, such as that of visual perception, compared below, or those of inference and memory. In particular, perhaps the principal overarching function of the structuring common across cognitive domains is that of providing conceptual coherence, that is, acting as a means for integrating and unifying a body of otherwise disparate conceptual material. In language and, as suggested below, in vision, this fundamental function has three main global forms of realization: coherence across a conceptual inventory, coherence within a scene, and coherence through time. Across the inventory of notions available for expression within any one language, grammatical specifications bring coherence principally by constituting a classification of the vast variety of conceived and perceived material. They gather different portions of the material together into subdivi-
196 LEONARD TALMY sions distinct from each other. By this, any particular currently cognized element is associated with its implicit "subdivision-mates". An illustrative case here are the twenty-plus motion-related prepositions in English, such as through and into, which together subdivide the conceptual domain of 'paths considered with respect to reference-objects'. This domain covers a great and varied range, but any particular "path" generally falls within the purview of one or another preposition, associated there with other "paths". To a certain extent, such associations can be regarded as arbitrary or idiosyncratic. Thus, as seen earlier, classed together by through are such dissimilar cases as a straightforward liquid-parting course (walking through water) and a zig-zag obstacle-avoiding course (walking through woods). The question arises why such distinctions should be effaced by the grammatical system, while they are observed by the lexical and other cognitive systems. Why are grammatical elements — say, such prepositions — not a large and open class marking indefinitely many distinctions? One may speculate that the cognitive function of such classification lies in unifying contentful material within a single conceptual system and in rendering it manipulable — i.e., amenable to transmission, storage, and processing — and that its absence would render content an intractable agglomeration. Providing coherence within a cognized scene was the function of grammatical structuring that was originally indicated in the Introduction. There it was put forward that the grammatical elements of any particular sentence together specify the structure of the cognitive representation evoked by that sentence. Their specifications act as a scaffolding or framework across which contentful material can, in effect, be splayed or draped. It can be posited that such structuring is necessary for a disparate quantity of contentful material to be able to cohere in any sensible way and hence to be amenable to simultaneous cognizing as a Gestalt. That is, without such structuring, not only does the inventory of concepts available for expression in a language become less coherent, but also any selection of such concepts concurrently juxtaposed by a sentence would tend to be only a collection of elements, rather than elements assembled so as to convey an integrated idea or thought complex. In the course of discourse, a great welter of notions are expressed in rapid succession, posing the potential problem of an unconnected sequence of ideational elements. But grammatically specified structuring is a principal contributor to the conceptual coherence through time that is requisite here. Through such structuring, a cognitive continuity is maintained
THE RELATION OF GRAMMAR TO COGNITION 197 through this flux and a coherent Gestalt is summated over time. A language can have a great stock of closed-class elements participating in this function, for example, such English forms as "yes, but", moreover, nevertheless, besides, instead, also. Such forms direct the illocutionary flow, specify the "logical tissue" of the discourse, and limn out its rhetorical framework. That is, these grammatical forms establish a structure that extends over a span of time, and thus provides a conceptual level with temporal constancy amidst more fleeting aspects of content. The preceding three global forms of grammatically specified structuring apply over the scope of any single language but, as indicated in the Introduction, a fourth form must also be recognized that holds for language in general. While each language has to some extent a different set of grammatical specifications, there is great commonality across languages, so one can posit that each set is drawn from an innate inventory of concepts available for serving a structuring function in language. Further, though, a qualifying property of this inventory can be adduced. It can be observed that grammatically specified concepts range cross-linguistically from ones that are of extremely widespread (perhaps universal) occurrence and of broad application within a language, down to ones appearing in a scant few languages with minimal application. Thus, the innate inventory of available structuring notions that is posited here seems to be graduated as to significance for the language faculty (cf. the tabular listing of grammatical notions in Talmy (1985a: 126ff)). For example, the notions 'entity' and 'occurrence' as expressed by the grammatical categories "noun" and "verb" are of great application and probably universal distribution, the notional category "number" seems of roughly middle standing in the ranking, while notions like 'in the morning' and 'in the evening' are expressed inflectionally on the verb in just a few languages. Notably, compared to spatio-temporal structuring, the notional category of "affect" is rather low in the graduated inventory of concepts that language draws on for structuring purposes, a fact that is significant considering its importance in other cognitive domains (cf. the other cross-domain differences noted below). The affect category does have scattered representation, for example 'affection' expressed by diminutive affixes, 'scorn' by pejoratives, 'concern' by a conjunction like lest, and 'hurt' by the "adver- sive" construction (as in the English: My flowers all died on me.). But seemingly no language has a system of closed-class forms marking major affect distinctions in the way that, say, the modal system in English specifies
198 LEONARD TALMY distinctions of force opposition (Talmy 1985b). Such an affect system can easily be imagined, however. Consider a parent addressing a child in danger near an open window. Grammatical systems readily allow the parent to refer to the spatial structure in this situational complex — Get away from the window! — leaving the affectual component to be inferred. But there is no closed-class form comparable to a modal, one that we could represent as "FEAR", as in FEAR the window!, that would allow the parent to refer to the affectual component of the complex and leave the spatial component to be inferred. Comparably, to a child near a freshly painted wall and about to harm it, a parent would likely again express the spatial structure — Get away from the wall! — leaving the affect to be inferred. There is no closed-class affect form for iike, be nice to1, which we could represent as "FAVOR", that the parent could use instead — FAVOR the wall! — thereby leaving the spatial component for inference. Parallels can now be drawn between the structuring system operating in language and that in visual perception (cf. Jackendoff in press).2*. The principal function of structure to provide coherence appears common across the two cognitive domains, and the three global forms of such coherence just outlined for language correspond to comparable forms in the operation of vision. First, as proposed in cognitive psychology, the perception of any particular object is mediated by its association with related objects in a schema for that object type, and the set of such schemas constitutes a classificatory system (Neisser 1967). This posited functioning of visual perception thus parallels the classificatory function of linguistic structure across a language's conceptual inventory. Second, there is a parallel between the linguistic coherence within a referent scene and the visual coherence within a perceptual scene. The welter of optical sensations registered at any one moment from some whole visual scene is rendered coherent by the perception of structural delineations running through it. For example, one looking at, say, the interior of a restaurant from one corner of the room does not see simply a pastiche of color daubs and curves but, rather, perceives a structured whole that includes the framework of the room, the spatial pattern of tables and people, and the individuated tables and people themselves. And seeing a person in some posture involves perceiving a structural framework in the human figure, as Marr (1982) describes this in his treatment of the "3-D model" in visual perception. Children's line drawings of scenes and stick-
THE RELATION OF GRAMMAR TO COGNITION 199 figure sketches of people, animals, and objects (Kellogg 1970) demonstrate our early capacity to abstract structure from visual scenes and scene parts. Third, one can observe a parallel between the coherence through time in linguistic discourse and that in visual perception. If the viewer in the illustrative restaurant now walks through the room, the patterns in which visual stimuli and the perception of structure change give rise in turn to the perception of a coherent continuity of path and view occurring within an overall "scene-structure constancy". It is reasonable to assume that, in addition to these language-vision parallels in global structuring, a number of particular structuring devices match across the two domains. Perhaps most of the grammatically specified conceptual categories treated in this paper — including, for example, state of boundedness and level of exemplarity — correspond to structuring factors in visual perception. Further, the first three of the broader linguistic systems for conceptual organization, the "imaging systems" outlined in Section 2.14, seem also to correspond to broader systems of visual organization. One can adduce still further parallels between language and vision as to the properties of their structuring. The topology-like character of grammatical specifications may have some parallel in the character of the perceived delineations of a scene, or internal structure of a figure, or plan of a path to be followed through obstacles. Such perceptions of structure seem in certain respects to abstract away from Euclidean particularities of exact magnitude, shape, or angle, and more to involve qualitative or approximate spatial relationships. As a further parallel, the capacity of grammatical specifications to nest, one within another, and form embedded structuring seems to correspond to embedded structuring within a visual scene. The restaurant scene above was described in terms of an overall framework that embedded a spatial pattern, itself consisting of individuated objects. Marr's (1982) analysis of an object like the human figure then continues the embedding, with perceived structurings of the body ranked from its overall linearity, to its stick-figure-like limb structure, and further to its articulations of these components. Whereas the preceding has outlined a set of parallels between language and vision, significantly, each of these two cognitive domains has prominent structuring devices that play little or no role in the other domain. Thus, in visual perception, three major parameters that structure (parts of) a scene are bilateral symmetry (moving or static), rotation, and dilation (expansion or contraction) (Gibson 1966, Palmer 1983) and, if color can be treated as
200 LEONARD TALMY structural, it is a fourth. In language, by contrast, grammatical specification of symmetry is minimal, perhaps limited entirely to the notion 'reciprocal'. Closed-class indication of rotation is limited in English to the prepositions and verb particles around and over, and is barely augmented in other languages. Dilation is grammatically expressed in English by the verb particles in and out when referring to radial motion (spread out I shrink in) and, again, such notions are not greatly more elaborated in other languages. And color, of course, was this paper's original example of a conceptual category not grammatically specified. In the other direction, there are several prominent linguistic categories of seemingly little structural function in visual perception. Examples are "status of reality", as expressed, e.g., by inflections for mood, "status of knowledge", as expressed by evidentials, and "comparison of alternatives", as expressed by a category of particles that includes instead, only, and also. Further possible examples are "relative temporal location", as expressed by tense markings, "degree", as expressed by adjective inflections and modifiers (e.g., English -er, -est, almost, too), and "force dynamics", as expressed by modals (Talmy 1985b). While language may not share these conceptual structuring categories with visual perception, it may well do so with other cognitive domains. Thus, its closed-class category "status of knowledge", which distinguishes such notions as 'known as fact', 'inferred', 'deduced', and 'considered probable' is very likely related to a set of basic parameters in our reasoning faculty. And, significantly, certain conceptual categories in language have a structuring apparently similar to that of conceptual models that form part of our broader faculty for conceptualization, in particular, our naive or folk models ("mental models" — cf. Gentner & Stevens 1982, Lakoff in press) as well as models in early science. For example, Talmy (1985b) demonstrates that the way in which language structures its concepts of causation and force interactions greatly parallels the conceptual structuring of naive physics and medieval physics (cf. diSessa 1986), while all three of these forms differ conceptually in a similar way from modern physics. Generalizing from all these findings, the possibility is that there is a fundamental core to conceptual structure that is common across cognitive domains — a core that thus epitomizes the nature of conceptual structure for human cognition — but that each domain has features of structuring, and perhaps also functions for structuring, that are not shared by others. Determining the overall and particular character of conceptual structure is
THE RELATION OF GRAMMAR TO COGNITION 201 the aim of the research advanced in the present study, one requiring a cooperative venture among the cognitive disciplines. Notes 1. This paper is a moderately revised and fully rewritten version of Talmy (1978a). Since 1978, the amount of additional material on the present subject, both descriptive and theoretical, has grown to be quite extensive. The present version incorporates a certain amount of this new material, as well as bibliographic updating, but the remainder will be reserved for an entirely new paper at a later date. 2. The word "evoke" is used because the relationship is not direct. The CR is an emergent, compounded by various cognitive processes out of the referential meanings of the sentence elements, understanding of the present situation, general knowledge, etc. Our term "cognitive representation" is similar in purport to Fillmore's (1975) "scene" but is chosen over that more specifically visual term. The linguistically evoked complex can include much from other sense modalities (notably somesthetic/kinesthetic and auditory) as well as abstract conceptual aspects. More recently, Lakoff's (in press) notion of an "idealized cognitive model", or ICM, points toward a comparable mental entity. 3. For their part, grammatical elements are generally more unalloyed in their indication of structure. They can express more contentful concepts, but this is largely limited. An example of it is in English upon as used in We marched/rode/sailed/rushed upon them [e.g., the enemy]., where upon incorporates the notion of 'attack', seemingly equivalent to the paraphrase 'into attack upon'. 4. More recently, research on different aspects of this broader scope has included work by Jackendoff (1983), Bybee (1985), Slobin (1985), Morrow (1986), and Langacker (in press). 5. Not included are regular adverbs, which seem in all languages to be derived from the three open classes just mentioned, rather than to comprise in their own right an open class of specifically adverbial roots. Of possible inclusion as a type of open class are the systems of idcophones, or "expressive forms" found, for example, in a number of Asian and African languages. Also includable, at a level above that of basic elements, are "lexical complexes" (collocations) like English kick the bucket or have it in for. 6. More accurately, rather than a dichotomy between an open and a closed type of class, there appears to be a cline. A class can range from having quite few members, like that of number inflection in English, to having very many, like that of noun roots in English, and the class's properties may correspondingly range from relatively more grammatical to more lexical. There exist some mid-sized classes — e.g., the several score individual classifiers of Chinese, or the three dozen or so instrumental prefixes in the polysynthetic verb of Atsugewi (Talmy 1972. 1985a) — that appear to have properties part grammatical and part lexical in character. 7. One can note, for example, the effect on one's cognitive representation in considering first the sentence 1 looked at the dog and then / looked at the dogs. The addition of the grammatical element -5 has a major effect on the delineational breakup of — to put it vis-
LEONARD TALMY ually — the scene before the mind's eye. For example, augmentative and diminutive elements, insofar as they refer to actual size, seem to specify size relatively greater or lesser than the norm for the particular object indicated. And closed-class elements specifying distance — like English just or way, as in just/way up there — specify notions of 'near1 and 'far1 that are relative to the referent situation. The properties of the specifically linguistic form of topology require determination. In this regard, consider the English preposition in, which in one main usage specifies a plane so curved as to define a volume of space. The referent of this morpheme, as in mathematical topology, is magnitude-neutral: in the thimble I volcano; and it is shape-neutral: in the well I trench. But in addition, its referent is closure-neutral, i.e., indifferent to whether the curved plane leaves an opening or is wholly closed: in the bowl I ball. And it is discontinuity-neutral, i.e., indifferent to whether the curved plane is solid or gapped: in the bell- jar I birdcage. These last two properties would form a proper part of language's topological system, whereas they are strictly excluded from mathematical topology. In many cases, a language favors only one such direction, having much lexicalization with notion A and simple grammatical means for reaching notion B, but in the reverse direction having only little lexicalization and complex grammatical forms. Languages differ typologically in the directions they favor. This issue will not be taken up here, but is treated at length in Talmy (1985a). Shifts are actually one member of a set of "reconciliation processes" — including blends, juxtapositions, schema-juggling, and blockage — that can be triggered by the association of a grammatical and a lexical form with incompatible structural specifications. In the non-shift processes, the grammatical specification does not take precedence over the lexical one, but plays an equal role with it. Of all these processes, this paper treats mostly shifts, but an additional number are discussed in Talmy (1977). In addition to space and time, language represents other conceptual dimensions that also belong to the present category. For an example, recall from Section 1 that this and that specify a partition drawn through space — and can do so through time as well — and indicate that a referent entity is on the same or the other side of the partition as the speaker. Now consider the English pronouns you and they in their indefinite usage (akin to German man or French on). These also specify a partition, but one drawn through "identifi- cational space", understood as a new conceptual dimension. They indicate, respectively, that 'the average person1 is or is not identified with the speaker in some relevant respect, i.e., is on the same or the other side of the identificational partition as the speaker. Thus, a person who smokes that is visiting a new neighborhood can ask a passer-by about the purchase of cigarettes with you, but about the sale of cigarettes with they: (i) Where can you buy cigarettes around here? Where do they sell cigarettes around here? But a person looking for a location to open a tobacco shop would ask a business consultant in the neighborhood about purchases and sales with the reverse assignment of you and they: (ii) Where can you sell cigarettes around here? Where do they buy cigarettes around here? It is true that there are the traditional terms "semelfactive11 and "iterative" referring, respectively, to one and more than one instantiation of an event. But there is no real
THE RELATION OF GRAMMAR TO COGNITION 203 equivalent to number: "aspect" includes too much else about the temporal structure of action. And in any case, none of the traditional terms refers generically to both the dimensions. M. The mechanism actually resorted to in many such cases, including that of tear, is the use of the plural, as in: (i) Tears flowed through that channel in Hades. There seems to be a sequence of cognitive operations here in getting from a bounded to an unbounded quantity. Speculatively, the bounded quantity is first treated as a uniplex entity, it is then multiplexed, the resultant entities are conceived as spatially juxtaposed, and their boundaries are lastly effaced, thereby creating an unbounded continuum. 15. The present category may be prone to confusion with the preceding one. Contributory here is the normal meaning range of continuous, which as easily covers 'boundlessness' as it does 'internal seamlessness'. However, the two categories can vary independently. Thus, in the preceding section, the lexical examples given for unboundedness, water and sleep, happened also to be internally continuous; but the same demonstration of unboundedness could have been made with internally discrete examples like timber and breathe. In general, unbounded forms share many properties, whether continuous or discrete. Thus, mass nouns and plural count nouns, both unbounded, share many syntactic characteristics not shared by singular count nouns, e.g.: (i) a / every — book / *ink / *books; (ii) all / a lot of / more / some [unstressed] / 0 [generic] — ink / books / *book; 0 ['progressively more'] (e.g., The machine consumed ink I books I * book for an hour.) 16. For schematizing action along the one-dimensional time axis, an adaptation of the two- dimensional A', B\ A, and B diagrams would be necessary — and can be readily visualized. 17. The lexical types for several of these intersections, it should be noted, do have traditional terms. Thus, nominal forms of the a, A or A', and B' types, respectively, have been called count nouns, collective nouns, and mass nouns. And verbal forms of the a and B' types, respectively, have been called punctual and durative verbs. The matrix presented here augments, systematizes, and generalizes the traditional notions. 18. This category can be considered a generalization over the earlier category of "state of boundedness" by the inclusion of the "uniplexity" notion. It can in turn itself be generalized — becoming the category "pattern of extension" — by the further inclusion of such notions as a quantity bounded at one end but unbounded at the other (see Talmy 1983). 19. This category clearly patterns with the preceding five within a single system of notions, one that would be an expansion or generalization over "disposition of a quantity". 20. Our main purpose here is to note the shift in structural distribution type. The shift in content will doubtless prove part of a larger pattern as well, but this is not yet worked out. 21. The category of axiality can be seen as an extension of the preceding category, pattern of distribution. The two categories address temporal stasis or change, involving both spatial relations (e.g., fall there, past the border here) and qualitative states (e.g., die/flash, awake/asleep there, sick/well here). But where the preceding category focused on discrete states, the present category elaborates the notion of a scalar quantity functioning in con-
LEONARD TALMY junction with a discrete state. Due to their structural character, these two categories pattern together with all the categories after "dimension" as part of a single broad conceptual system of "structural schematization", described below in Section 2.14, and are thereby distinguished from the categories described next, which belong to different conceptual systems. The use of the perfect here points to a principal function of perfect forms in general: They can indicate the temporal counterpart of matter located within a bounded extent of space, of the type seen in (i). That is, a sentence containing the perfect, as in (ii), suggests a paraphrase like that in (hi), which is homologous with (i): (i) There were 5 aspirins on the table. (ii) I have taken 5 aspirins in the last hour. (iii) There were 5 aspirin-takings in the last hour. (In support of this interpretation, as pointed out to me by Peyton Todd, the perfect seems always to involve a temporal span bounded at both ends.) All three notions here — (a) identity of a quantity, (b) portion-excerpting from that quantity, (c) configuration of that portion — can be respectively represented by three distinct NPs together in a construction, as in: (i) a clustering (c) of a set (b) of trees (a). Many lexical items conflate the specification of two or all of these notions at once. Thus, conflating (c) and (b) is a cluster 'a clustering configuration of a set' and a drop 'a small globular form of an amount [of a liquid]'. A lexical item conflating all three notions is a tear 'drop of lachrymal fluid'. (See Talmy (1985a) for a general treatment of "conflation"). For the plural form oysters, the plural form siphons is ambiguous as to whether there are one or more siphons per oyster. All the other combinations unambiguously indicate the number of siphons per oyster. Thus, the exemplar form is always unambiguous in this regard — one of its advantages over the full-complement form. This same arrangement holds through the list. The difference between each and every arising in this analysis can now be added to those observed elsewhere (e.g., Vendler 1968). Each is the exemplar counterpart of the full- complement expression all the, but not of all without the. Thus, *Each oyster has a siphon cannot function as a generic assertion. Every is not as unilaterally aligned in this way, but does serve more naturally as the counterpart of all without the. One more pair can be added to this list by adjoining two complementary unpaired forms from two different languages. The English form some, as in some friends of mine, is a full- complement form requiring the plural and has no exemplar counterpart in the singular. The corresponding Italian form qualche, as in qualche amico mio, requires the singular and lacks a plural counterpart. Though the grammatical complex in (48b) is determinative in setting the role-number as monadic, a trace of the verb's original scene-division type does remain. In the CR, the metaphoric suggestion of a dyad is blended in, as if both 'host' and 'guest' are together present in the "I", perhaps as separate subparts of the single person. For this reason, (48b) is not the complete semantic equal of (48c). Such blending is, beside shifting, another major process of reconciliation between incompatible specifications, referred to in note 11.
THE RELATION OF GRAMMAR TO COGNITION 205 28. Clearly, the language-related faculty of the brain evolved to its present character in the presence of other already existing cognitive domains, including that of vision, and no doubt developed in interaction with their mechanisms of functioning, perhaps incorporat ing some of these.
Where Does Prototypicality Come From? Dirk Geeraerts University of Leiden/University of Leuven 1. Hypotheses about the Sources of Prototypicality Prototype theory is as it were part of the prototypical core of the cognitive paradigm in semantics, particularly in lexical semantics. I think it is safe to say that it is by now quite obvious that gradience and salience are among the linguistically relevant aspects of semantic structure. One need only recall the early experimental work by Rosch (1973b) and Labov (1973) to appreciate the importance of graduality and vagueness for the adequate description of word meaning. But what about explanatory adequacy? Can we move beyond the descriptive level and explain why prototypicality exists at all? There are at least four different hypotheses that have been proposed to explain prototypical phenomena. Each of these hypotheses has been formulated (or at least hinted at) by Eleanor Rosch herself; this is an indication that the hypotheses might well be complementary rather than mutually contradictory. I will call these four hypotheses the physiological, the referential, the statistical, and the psychological one. Let us have a look at them. The physiological hypothesis says that prototypicality is the result of the physiological structure of the perceptual apparatus (Rosch 1973b). This hypothesis has been formulated with regard to the prototypicality effects in the domain of colour terms (the first major field in which prototypicality phenomena have been observed). Particular colours are thought to be focal because the human eye is more sensitive to certain light frequencies than to others. The scope of the physiological explanation is probably fairly limited; it may only be applicable to concepts immediately referring to perceptual phenomena, or at least to bodily experiences that have a distinct physiological basis. Since this is most likely not the majority of cases, additional hypotheses will have to be invoked to explain the prototypical struc-
208 DIRK GEERAERTS ture of concepts that have no immediate physiological basis. The referential hypothesis states that prototypicality results from the fact that some instances of a category share more attributes with other instances of the category than certain peripheral members of the category (or share attributes with more other instances than these peripheral cases). The peripheral applications of a category share attributes with relatively few other cases, or share only a relatively small number of attributes with other, more central members of the category. This is the family resemblance model of prototypicality (Rosch & Mervis 1975); in psychological terms, it states that the prototypical instances of a category maximize cue validity. I have dubbed this view 'referential' because it considers prototypicality to be an automatic consequence of the structure of the range of application of a concept. Once you know what objects, events etc. a concept can refer to, you can compute differences in salience by comparing the number of shared attributes among those things. One might even say that prototypicality is a secondary phenomenon: it is a side-effect of the mutual attribute relations among the instances in the referential range of application of the concept. Statistical explanations of prototypicality state that the most frequently experienced member of a category is the prototype. At least, this is the simple form of the frequency model. It can also be combined with the family resemblance model; the weight of an attribute within a concept is then not only determined by its role within the family of applications constituting the category, but also by the relative frequency with which it is experienced (Rosch 1975). The psychological hypothesis is a functional one. It states that it is cog- nitively advantageous to maximize the conceptual richness of each category through the incorporation of closely related nuances into a single concept because this makes the conceptual system more economic. Because of the maximal conceptual density of each category, the most information can be provided with the least cognitive effort (Rosch 1977). In what follows, I would like to show that the functional explanation of prototypicality is more general than the other ones because it can explain cases of prototypicality that are counterexamples to the other models. I will elaborate the psychological hypothesis by indicating some more functional sources of prototypicality; I will try to make clear that prototypicality is the outcome of some deep-seated principles of cognitive functioning.
WHERE DOES PROTOTYPICALITY COME FROM? 209 2. A Case Study in Synonymies Dutch has a pair of synonyms vernielen and vernietigen, which both roughly mean "to destroy". Though they exhibit some degree of phonetic similarity, their origin is quite diverse. Vernielen is the older form. It is already to be found in Middle Dutch, and it is formed by means of the common verb-forming prefix ver- and the adjective niel, only a few examples of which survive, but which probably meant "down to the ground". Etymolog- ically, then, vernielen means "to throw down to the ground, to tear down". Vernietigen, on the other hand, makes its first appearance in the 16th century; it is formed by means of the same prefix ver- and the adjective nietig, which is itself a derivation from the negation particle niet (English not) and the suffix -ig (which corresponds with English -y). Vernietigen gradually replaces a third form vernieten, which is a straightforward derivation from niet with ver-, and which is extinct by the end of the 17th century. Vernietigen literally means "to annihilate, to bring to naught". The best way to study both words is to turn to the Woordenboek der Nederlandsche Taal (hence WNT), the major dictionary of Dutch that covers the period from 1500 up to 1920 and that, by the way, is still uncompleted after about a century of editorial work. This dictionary is being compiled on the basis of a huge corpus of quotations; there are as yet no equally representative corpora for contemporary Dutch, so that it is rather more difficult to get an adequate picture of 20th-century usage than it is to study the semantic history of the vocabulary of Dutch. For the purpose of this paper, this is not very important: it suffices to pick out one synchronic period and to see how both words relate to each other in that period. For a number of practical reasons (among others the amount of available material), I will concentrate on the 19th century, stretching the temporal borders of that period with approximately one decade at each end. In this way, a period from 120 to 130 years can be considered, ranging from roughly 1790 to 1910. In the light of the history of culture, this seems quite justified; we more or less envisage the cultural period from the French revolution up to the First World War: the 19th century in the broadest sense. To get a good picture of the development of vernietigen and vernielen, it would be necessary to present and discuss the entire articles that I have compiled for the WNT, and the complete set of quotations on which they are based. For obvious lack of space, I will only give illustrative quotations; translations of the quotations are given in the Appendix. The set of quota-
210 DFRK GEERAERTS tions in the table shows that vernielen and vernietigen can be used indiscriminately with the same range of application. Each numbered pair of quotations gives examples of one particular kind of usage. These examples should be studied from two points of view. On the one hand, the question has to be asked whether vernielen and vernietigen exhibit any syntagmatic differences, i.e. differences in their collocational properties. On the other hand, the question arises whether they are paradigmatically different, i.e. whether they exhibit purely conceptual differences. (More details on the analysis of both verbs can be found in Geeraerts 1985a.) Syntagmatically, we not only see that both words can be used by the same author in the same context without noticeable differences (as in (1), (2), (8) and (9)), but also that the range of application of each word can be divided into three identical major groups, which can moreover be subdivided along parallel lines. There is a set of applications in which the words are used with regard to concrete, material objects (1, 2, 3); a set in which they are used with regard to abstract objects (4, 5), and a set in which they are used with regard to persons (6, 7, 8, 9). Within the first set, frequently occurring applications relate to buildings (1), other human artifacts (2), and natural objects, in particular plants and crops (3). With regard to the abstract applications, we can distinguish between the annihilation of the existence of certain abstract objects as such (4), and applications in which the realisation or fulfillment of certain abstract notions that contain an aspect of expectation or intention with regard to the future is prevented (5). With regard to persons, (6) expresses their death as such; (7) and (8) indicate how someone's bodily or mental health, respectively, are undermined. (9) expresses how armies are beaten; this application is half-way between the abstract group (the armies cease to exist as functional entities), and the personal group (individual soldiers are killed). The existence of analogous subdivisions within each of the major groups shows that the syntagmatic equivalence of vernielen and vernietigen is not a coincidence, but that it is an essential part of their relationship. Furthermore, the examples also show that there is a paradigmatic, strictly conceptual equivalence between both: they do not only have the same collocational properties, but they also seem to express the same concepts in the same contexts. (The distinction between syntagmatic and paradigmatic meaning is used here for purposes of analysis only; it does not imply any particular view with regard to the theoretical relation between
Table 1 i With regard to 1 concrete things To demolish buildings or parts thereof To destroy other human artifacts To destroy natural objects (D (2) (3) VERNIELEN Dat huis was...cvenmin als de naburige tegen de verwoestendc veeten dier tijd bestand. Reeds onder den zoon en opvolgcr des stichtcrs werd hct ... tot den grond toe vernield (Veegens, Hist. Stud. 2, 282, 1869). Er gaat dan stroom op den daarvoor gevormden zijwcg over, waarbij genoeg warmtc ontwikkcld wordt om de draadwindingen in zecr korten tijd te vernielen (Van Cappclle, Elcctr. 214, 1908). Hoevcel het wild vernielt wordt door een Engelschman zeer gocd uitecnge- zet bij gelcgenheid van een aanval op de beschcrming die het wild aldaar ...gcnict(V61ksvlijt 1872, 175). VERMETIGEN Allcen zijn de vroegcre kruisvensters door vensterramen van nieuweren trant vervangen en hebben de vrijhcidsmanncn ' van 1795...hct wapen des stichters in den voorgevel met ruwc hand vernietigd (Veegens, Hist. Stud. 1, 125, 1864). Zonder dcze voorzorg zou het draadjc door dc enorme hittc van den glocidraad vernietigd worden (Van Cappclle, Elcctr. 295, 1908). | Bij het vernietigen van dc onkruiden door het bewerken dient op hunne voortplan- i ting en ontwikkcling tc worden gclct (Reinders, Landb. 1, 309, 1892).
Table 1 - continued to I—^ to With regard to abstract things With regard to persons To annihilate existing situations, characteristics etc. To prevent the execution of plans, hopes, intentions etc. To kill people, to take someone's life (4) (5) (6) Wei wat hamer ! Wordt door zulke sentimentele zotternyen niet al de inwendige kragt vernield ? (Wolff en Deken, Blank. 3, 220, 1789). De bergstroom in zijn grammen loop Verscheurt zijn zoom, verdrinkt de dalen: Alzoo vernielt Gij 's Menschen hoop! (Ten Kate, Job 53, 1865). Mij gendenkt ook nog dat Nicolaas Gaal...mij placht te verhalen...dat de oude man om deze ontstolen eer zich zoo ontstelde en vergramde, dat het ook scheen of hij dezen dief wel had willen vernielen (Fruin, Geschr. 1, 1974, 1888). Stel mij niet zoo hoog, zei ze onthutst, ik zou daaraan niet beantwoorden; ik zou uw ideaal vernietigen (Vosmaer, Amaz. 175, 1880). Zy is dan, van kindsbeen af, opgevoed on mynheer Daniel's echtgenote te worden, en nu is die hoop van een geheel leven vernietigd ! (Conscience, Kwael d. T. 2, 65, 1859). Dit toeval vernietigde ons geheele plan (Haafner, Ceilon 103, 1810). Intusschen heeft de Godin de Natuur besloten nu voor altijd de Drijvende Eilanden en al hun inwoners te vernietigen (Quack, Soc. 1, 246, 1875).
Table 1 - continued With regard to persons — continued To undermine someone's physical health To undermine someone's psychological well-being To defeat groups of armed men, or armies (7) (8) (9) De beroerte, die haar zwakke ' levenskrachten in een half uur tijds vernielde, had reeds in het eerste oogenblik hare spraak verlamd (Beets, CO. 206, 1840). Dc vrouwen, Lus, zijn zonen, a( de anderen blcven stom, vern'teld van ontsteltenis, op hun stoelen genagcld (Buyssc, Neef Perscyn 45, 1893). De uitslag van den stryd was ditmael hem niet gunstig: geheel zyn leger werd vernield of uiteen geslagen (Conscience, Gesch. v. Belgie 110, 1845). Hy moet rusten. Zulke driften vernietigen het sterkste gestel (Wolff en Deken, Leev. 1,290, 1784). Toen.-.antwoordde zij langzaam met een doffe stem, als vernietigd door haar eigene woorden: "Ja, indien het nog mogelijk is" (Buysse, Mea Culpa 68, 1896). Het gansch leger der Turken was i vernietigd ! (Conscience, Gesch. v. 1 Belgie 352, 1845). 3 X w w a o O H o > r n o 2 w O 2
214 DIRK GEERAERTS both aspects of lexical meaning and particularly with regard to the question whether selectional restrictions are always an automatic consequence of a concept's paradigmatic characteristics.) As a preliminary step, notice that the concept "to destroy" does not only appear as the notion "to annihilate the existence of someone or something, to cause someone or something to disappear out of existence", but that it also exhibits the weaker nuance "to undermine someone or something with regard to some aspect of his existence" (without a complete destruction or a complete removal out of existence being implied). The distinction can easily be discovered within the personal group of applications. In (6), a person is killed, taken out of existence, while in (8) (and most likely also in the second quotation from (7)) someone's existence is undermined from one point of view or another, but not entirely annihilated. Likewise, we can see that within the abstract group, (4) signifies the suppression ofihe existence of some abstract things as such, whereas in (5), plans, hopes, and expectations are undermined with regard to their realisation and fulfillment: the plan as such is not removed (at least not to begin with), but it is reduced to ineffectiveness and futility. In short, both vernielen and vernietigen express the notions of complete destruction and partial damage, that is to say, the complete removal out of existence of something or someone, and the less drastic undermining in some respect, of the existence of people or objects. According to the syn- tagmatic context, these notions receive further specifications. For instance, with regard to persons, complete destruction means killing, but with regard to concrete things, destruction signifies material demolition, and so on. (For the sake of completeness it should be added that the equivalence of vernielen and vernietigen is less straightforward in present-day Dutch than it is in 19th-century Dutch. Some of the quotations discussed here are now felt to be rather awkward; in particular, it would be difficult to use vernielen with regard to persons.) On the basis of the foregoing observations, one might be tempted to conclude that the semantic structure of vernielen and vernietigen in 19th- century Dutch is completely identical: both syntagmatically and paradig- matically, they have the same range of application. However, a number of facts testify that both words have different prototypical structures, i.e., that they have different conceptual centres. There are two sets of facts to be considered: corpus-based facts relating to the way in which both words are used in our corpus of quotations, and introspective facts relating to the way in which the words are perceived by the speakers of the language. In gen-
WHERE DOES PROTOTYPICALITY COME FROM? 215 eral, consideration of these facts will lead to the conclusion that the abstract applications are central within the structure of vernietigen, and that the material applications are central in the cases of vernielen. As such, each verb has a different semantic structure in spite of the fact that the elements of these structures appear to be the same. 3. Usage as Evidence for Prototypicality Five observations support the prototypical hypothesis. In the first place, the abstract group of applications is quantitatively more prominent within the structure of vernietigen than the material set of applications, while the reverse is true of vernielen, in which the material group is the most frequently occurring one. In both cases, the major group is represented by approximately three times the quotations of the less central group. In the second place, the differences in centrality show up in the fact that the prominent applications exhibit specifications and particular nuances that they do not have when they are peripheral within the structure of the lexical item. Thus, the material group of vernielen contains a metonymical extension of the application with regard to plants and crops, towards an application in which the fields and gardens where these plants and crops grow appear as the direct object of the verb. Likewise, the application with regard to buildings receives a figurative extension towards an application with regard to an allegorical "wall" that separates two people. These extensions are probably not impossible within the concrete set of applications of vernietigen, but the fact that they do not appear there is statistically interesting: it indicates that the concrete application is more productive in the case of vernielen than in the case of vernietigen. Conversely, the abstract group has nuances and additional specifications in the case of vernietigen that are lacking in the same group with vernielen, although it is quite easy to imagine that they would in fact occur there. For example, vernietigen has a fairly large set of applications in which social movements, institutions, activities and so on are abolished, one quotation in which it is said that railway transport destroys distances (obviously, distances do not disappear as such, they are only functionally overcome), and one quotation in which a philosopher is said to destroy the soul (again, the soul is not destroyed as such, but the idea that the soul exists is metonymi- cally abolished by the philosopher in question). None of these extensions of
216 DIRK GEERAERTS the abstract use of the concept "to destroy" can be found in the case of vernielen, which is indicative of the fact that the abstract use is less prominent in the latter verb than in the semantic structure of vernietigen. In the third place, the salience of the material kind of usage can be derived indirectly from the nominalisations of both verbs. Both vernietiging and vernieling have the verbal sense "the fact, the act or the process of destroying or being destroyed", but only vernieling exhibits the metonymi- cal extension towards the concept expressing the result of that process or that act, i.e., the concrete damage that issues from it. (In the latter case, the word is typically used in the plural: vernielingen more or less equals the notion "damage".) In the fourth place, the internal structure of the set of personal applications reflects the differences in prototypical structure between both verbs. To begin with, notice that the personal group contains concrete as well as abstract applications; to kill someone is clearly more concrete than to undermine someone's psychological well-being or his social position. If we then have a look at the mutual relationship between the abstract and the concrete subgroups of the application with regard to persons, we find that the abstract subgroup is proportionally dominant in the case of vernietigen, whereas the reverse is true in the case of vernielen. Also, we find that extensions of the concrete subgroup of the personal application with regard to other living beings than people or with regard to personifications, are not as strongly present in the case of vernietigen than in the case of vernielen. (It should be added that these observations have to be considered with more care than the previously mentioned points, since there is a general tendency throughout the centuries covered by the WTVr-material, to remove the personal application from the structure of vernielen. There are relatively less personal applications in the structure of 19th-century vernielen than in the structure of either 19th-century vernietigen or 16th-century vernielen; as has already been mentioned, it is even more difficult to use vernielen with regard to persons in present-day Dutch. In any case, the 19th-century material does seem to show that the material subgroup of the personal application of vernielen is more resistant to the tendency in question than the abstract subgroup, as can be predicted from our centrality hypothesis.) Finally, the importance of prototypicality can be derived from the fact that different nuances play a central role within the core of each concept, whereas those nuances are not particularly important within the corresponding group in the other concept. Thus, the destruction of buildings and
WHERE DOES PROTOTYPICALITY COME FROM? 217 other human constructions is prominent within the material use of vernielen, but is only rarely present within the material group of vernietigen. Within the structure vernielen as a whole, demolishing buildings is the single most frequently represented kind of usage, but within the structure of vernietigen, it is merely one among many equally important nuances of the material set of applications. In the same way, the central, abstract group within the structure of vernietigen is itself centred round applications relating to the dissolution, the cancellation, the annulment of agreements, commitments, engagements, obligations, permissions, rights, and so on, and of the laws, orders, contracts etc. in which they are contained and through which they come into existence. Whereas vernielen only rarely exhibits this kind of usage, it is the most frequently occurring sense within the abstract group of vernietigen as well as within that word as a whole. In general, these facts of linguistic usage clearly favour the hypothesis that the abstract applications of the concept "to destroy" are prototypical within the structure of vernietigen, whereas the concrete applications are prominent in the case of vernielen. Taking into account that each central group is itself concentrated round a dominant kind of usage, it seems plausible to say that the latter is the prototypical sense for each of the verbs in question. It should furthermore be noted that these prototypical phenomena seem to be connected with the etymology of the words. On the one hand, the abstract prototype of vernietigen may well be connected with the abstract character of the words niet "not", and nietig "null and void, insignificant", on which it is based. Moreover, the common phrase nietig verklaren "to declare something to be null and void, dissolve, annul something" corresponds pretty closely with the central notion within the abstract group of vernietigen. On the other hand, the centrality of the application with regard to buildings in the structure of vernielen seems to correspond with the etymological meaning "to tear down, to throw to the ground" that we reconstructed above as the original meaning of the verb. 4. Introspective Evidence of Prototypicality Before we can deal with the introspective evidence in favour of the prototypicality hypothesis, two preliminary questions have to be answered. In the first place, how trustworthy is the introspective methodology? The paradoxical fact of the matter is that it is exactly the unreliability of introspection that makes it interesting for our purposes. If introspection were
218 DIRK GEERAERTS able to yield a completely adequate picture of the facts of linguistic usage (which is doubtful), it would simply reduplicate the results reached in the previous paragraphs on the basis of a direct examination of linguistic usage. But given the presupposition that introspection yields only a partial insight into the semantic structure of the words that are investigated, we can also presuppose that it will be exactly the prototypical kinds of usage of those words, that reach the introspective consciousness of the language user. We can use the results of the introspective method as support for the prototypical hypothesis if we presuppose that prototypical kinds of usage (precisely because they are more salient than other applications) will more easily pass the threshold of conscious attention. Given this presupposition, the introspective judgements of native speakers may shed light on the question which kinds of usage are predominant within a certain concept. In the second place, how can the introspective method be used with regard to historical material? There are no 19th-century speakers of Dutch around to be asked what they think is the meaning of particular words, so how are we going to get introspective judgements at all? The fact is that we do have information on how the 19th-century speakers of Dutch perceived the near-synonyms that we are investigating, viz. in the form of synonym dictionaries. Synonym dictionaries (at least the older ones) are notoriously unreliable as descriptions of actual patterns of usage; most of the time, the compilers of synonym dictionaries rationalise away the actual identity of words by imposing distinctions that cannot be discovered in the actual facts of usage. However, these rationalisations need not always have proceeded out of the blue: it seems quite plausible that they were guided by the introspective judgements of the compilers. So, if we like to know something of the introspective insights of the 19th-century speakers of Dutch, we can have a look at the synonym dictionaries of that time to see whether the distinctions they make between vernielen and vernietigen (however inadequate as a picture of the complete set of possible kinds of usage) do indeed reflect the differences in prototypical structure of both words. And indeed, the 19th-century synonym dictionaries of Dutch do distinguish between vernielen and vernietigen along lines that fit into our hypothesis. On the one hand, there are those that draw the line syntagmat- ically, such as Weiland & Landre (1825), who state that vernielen can only be used with regard to "lighamelijke dingen" (material things), whereas vernietigen is more widely used, in particular also with regard to "menschelijke instellingen" (human institutions). De Beer (1897) expresses
WHERE DOES PROTOTYPICALITY COME FROM? 219 an analogous point of view. On the other hand, there are those that describe the distinction along paradigmatic lines, so that there would be an actual notional difference between the verbs in question, rather than merely a distinction in selectional restrictions. Whereas vernietigen is defined as "to bring to naught, to annihilate", vernielen is defined as "to damage, to smash to pieces, to tear down". In this sense, vernietigen implies a complete annihilation whereas there may be some pieces left of the original object in the case of vernielen. It is easy to see that this paradigmatic point of view, which can be found among others in Pluim (1894), is connected with the previous, syntagmatic one: it is precisely because vernielen relates to material things that the notion of remaining debris comes to the fore. Likewise, a complete annihilation (in which the original objects disappear completely) is less likely in the material world of concrete objects, so that the restriction of vernietigen to abstract objects will tend to be related to the notion of complete annihilation. This is in fact done by Weiland & Landre (1825), though not all proponents of the paradigmatic distinction adhere to the syntagmatic distinction. For instance, De Flines (1810) mentions that vernietigen can in fact be used with regard to material objects, but that there is a difference with vernielen in the degree of damage achieved. By and large, these views faithfully reflect the insight into the prototypes of vernielen and vernietigen that we have gained by considering the actual facts of linguistic usage. Syntagmatically, it is recognised that the material context is more important for vernielen, whereas abstract objects are predominant in the case of vernietigen. Paradigmatically, this is reflected by the fact that vernielen carries overtones of material destruction and damage (think of the relationship between the prototypical usage of vernielen with regard to buildings, and the definitions of that word that bring to the fore the act of smashing and demolishing things), whereas vernietigen calls forth the idea of complete annihilation (as it were, wiping something off the face of the earth). As such, the stubborn efforts of the compilers of synonym dictionaries to find semantic differences among near- synonyms seem to be not entirely gratuitous. To the extent that they try to capture the characteristics of the most salient kinds of usage of both lexical concepts, they strenghten our own hypothesis about the differences in prototypical structure among the verbs.
220 DIRK GEERAERTS 5. The Functional Explanation of Prototypicality There are a number of interesting conclusions to be derived from the above analysis of the near-synonyms vernielen and vernietigen. First, prototypicality is an interesting new point of view in the study of synonyms. It is traditionally well-known in lexical semantics that there are relatively few true synonyms in natural languages, and the ways in which near-synonyms differ can be very diverse. Our discussion of vernielen and vernietigen shows that there is one more factor to be added to the list of differentiating factors: near-synonyms may be distinct with regard to the prototypical structure imposed on an otherwise identical range of application. Once again, the importance of prototype theory for the traditional concerns of lexical semantics becomes apparent (cf. Geeraerts 1983a, 1983b, 1984, 1985b); prototype theory opens up new perspectives in the study of synonyms. Secondly, there are some indications that introspective judgements in lexical semantics relate to the prototypically salient instances of concepts rather than to the full range of actual usage possibilities. If this can be confirmed by additional comparisons between introspective perceptions of lexical meanings and actual usage patterns, more will be known about the value of both methodologies (introspective and corpus-based) in lexical semantics. Also, if we maintain the classical view of modern linguistics that it is one of the goals of linguistic theory to account for the introspective judgements of native speakers, and if these judgements appear to be influenced by prototypical phenomena, yet one more reason presents itself for incorporating prototype theory into lexical semantics. Thirdly, the fact that vernielen and vernietigen have the same conceptual and collocational range of application, and yet differ with regard to the core and the periphery of their categorial structure, indicates that there are at least some cases of prototypicality that cannot be explained by means of the referential model. Vernielen and vernietigen refer to the same set of acts and processes; as such, the differences in their prototypical structure cannot be the automatic consequence of their referential range, as is implied by the family resemblance hypothesis. In addition, the physiological and the statistical explanation will not be of much avail either. There is no particular organ or mechanism for the perception of processes of destruction, and even if there were, we would still need two different physiological structures to explain the distinction between both verbs, which is beyond all intuitive plausibility. The statistical explanation is inapplicable for the same reason
WHERE DOES PROTOTYPICALITY COME FROM? 221 as the referential hypothesis: since the range of application of both verbs is the same, the frequency of occurrence of the processes referred to is the same for both verbs. That is to say, the frequency with which the demolishing of buildings occurs in reality, relative to the frequency with which, say, agreements are cancelled in reality, has exactly the same effect on both verbs, since these refer to the same objective reality. Because they denote the same things in reality, the structure of reality (either with regard to the frequency of occurrence of its elements, or with regard to the mutual resemblances among those elements) cannot be invoked to explain the distinction in semantic structure between vernielen and vernietigen. In short, we can reject all materialistic explanations of the prototypicality effects observed in the verbs under consideration. Indeed, the physiological, the referential, and the statistical hypotheses have this in common: that they try to explain prototypicality on the basis of materialistic data, either the material structure of the human perceptual apparatus, or the material characteristics (statistical or otherwise) of the referential range of the concepts involved. Given that we have to reject these materialistic hypotheses, we can provisionally choose, by elimination, for the psychological, functional explanation of prototypicality. To support this choice, I would like to make clear that the functional hypothesis has some additional advantages, besides the fact that it avoids the problem of the materialistic hypotheses. First, however, three remarks have to be made. To begin with, it might be claimed that a statistical explanation of the prototypicality effects in vernielen and vernietigen can indeed be given, if we take into account, e.g., that the material sense occurs much more frequently with vernielen than with vernietigen, or that the abstract specification of the notion "to destroy" is statistically much more prominent in the latter verb than in the former. However, the frequencies that are mentioned here are linguistic frequencies, not referential frequencies, i.e., they are frequencies of occurrence of words, not of the things those words refer to. Because the frequency at stake here is linguistic rather than referential, it can hardly be invoked to explain prototypicality; as an aspect of linguistic usage, it is one of the things we have to explain, not one of the things that are themselves part of the explanation. We can use linguistic frequencies to determine what instances of a concept are prototypical (that is what we did in section 3), but explaining prototypicality on the basis of linguistic frequency is putting the cart before the horse. Some kinds of usage are not prototypical because they are more frequent; they are more frequent
222 DIRK GEERAERTS because they are prototypical. The apple is not a prototypical fruit because we talk more about apples than about mangoes, but because we experience apples more often than we encounter mangoes (and this fact, in turn, may be the reason why we talk more about apples). Frequency of linguistic occurrence may be a heuristic tool in the pinpointing of prototypes, but it is not the source of prototypicality as meant in the statistical hypothesis. The second remark has to do with the fact that criticism with regard to the referential, family resemblance model of prototypicality has already been formulated elsewhere. This has been the case in the work of Pulman (1983) and — in more stringent fashion — in the well-known article by Armstrong et al. (1983). They argue that gradience can be observed in concepts with rigid boundaries (their examples relate to natural numbers), so that family resemblances cannot be invoked to explain the differences in salience among numbers. There are two reasons, however, why their argumentation is less relevant than they assume. First of all, they more or less equate prototype theory and the family resemblance model of the sources of prototypicality, whereas it is quite clear that the family resemblance model is merely one of a number of hypotheses concerning the sources of prototypicality: ruling out one hypothesis does not mean that one can ignore the others. And also, I do not think that Armstrong et al. are successful in presenting a counter-example to the family resemblance model. Even if a concept has rigidly defining characteristics, family resemblances may exist among the non-defining characteristics of the instances of that category. Since cognitive semantics is basically encyclopaedist in its approach, these non-defining, "encyclopaedic" attributes should be incorporated into the computation of degrees of shared attributes. As Lakoff (1982a) has shown, such encyclopaedic, experiental factors do indeed occur with regard to numbers, and they can be used to explain the prototypicality ratings found by Armstrong et al. My third remark is this: my criticism of the materialistic hypotheses should not be overgeneralised. The fact that they do not work in the case of vernielen and vernietigen clearly does not imply that they do not work in any case, but merely makes clear that next to the physiological, the referential, and the statistical model, there will have to be at least one other source of prototypicality. Let us now come back to the functional model of prototypicality and try to elaborate it. Remember that the psychological hypothesis involves requirements that the cognitive system is to comply with if it is to function
WHERE DOES PROTOTYPICALITY COME FROM? 223 efficiently: prototypicality exists because it is cognitively advantageous. As we have seen, Rosch has specified this functional advantage in terms of the economical effect of informational density; prototypical categories enable one to reach the most information with the least cognitive effort. This functional line of reasoning can be supplemented with some additional (and perhaps even more fundamental) functional reasons for having prototypical categories. We can base the discussion on one of the fundamental insights of cognitive psychology, viz. that cognition should combine structural stability with flexible adaptability. On the one hand, cognition should have a tendency towards structural stability: the categorial system can only work efficiently if it can maintain its overall organisation for some time, if it does not change fundamentally any time new information has to be incorporated. At the same time, however, it should be flexible enough to be easily adaptable to changing circumstances. To prevent it from becoming chaotic, it should have a built-in tendency towards structural stability, but this stability should not become rigidity, lest the system stops being able to adapt itself to new and unforeseen circumstances. This necessity of flexibility is one of the aspects of lexical semantics that was recognized by the prestructuralist tradition of historical semantics, but that has been more or less lost in the meantime, as a result of the structuralist attention for fixed synchronic structures. Be that as it may, it will be clear that prototypically organised categories are particularly well suited to fulfil the double demand for flexible adaptability and structural stability. On the one hand, the fact that slightly deviant nuances can be developed within a particular category indicates that categories have the dynamic ability to cope with changing conditions and changing expressive needs. On the other hand, the same fact (that marginally deviant concepts can be incorporated into existing categories as peripheral instances of the latter) proves that these categories have a tendency to maintain themselves as holistic entities, thus maintaining the overall structure of the categorial system. Prototypical categories maintain themselves by adapting themselves to changing circumstances and new expressive needs; at the same time, they function as expectational patterns with regard to reality: new facts are interpreted in terms of information that is already at the disposal of the individual. The flexibility of the cognitive system does not only show up in the fact that it can adapt itself to new experiences, but this flexibility is supplemented with the fact that existing categories have a formative influence with regard to experience; new experiences are fitted into the expectational patterns pro-
224 DIRK GEERAERTS vided by the existing categorial system. Along these lines, prototypicality appears to be the outcome of some fundamental, deep-seated principles of cognitive functioning. The form of the conceptual system appears to be determined by a set of basic functional requirements, and prototypically structured concepts admirably meet these requirements. If this is correct, the same basic principles should also have a role to play in other cognitive disciplines. That is to say, if prototypicality is an emanation of some basic characteristics of all cognition, we should be able to find analogies of the prototypical idea in other fields of cognitive science, next to lexical semantics. I have tried to prove at length elsewhere (1985b) that this is in fact the case: the importance of interpretative schemata mediating between experience and existing knowledge is an idea that can be traced in a number of cognitive disciplines. It is very much apparent in Artificial Intelligence (Minsky's frame notion); it can be found in cognitive psychology, particularly in the work of Bruner, and to some extent in that of Piaget; it can be related to some of the views of the early, Husserlian phenomenological movement in philosophical epistemology; and it has some important similarities with the paradigmatic conception of scientific enquiry inaugurated by Thomas Kuhn. These are exciting parallels because they suggest that the functional, psychological hypothesis concerning the sources of prototypicality can at the same time be the basis for a truly integrated cognitive science in which the insights of linguistics, Artificial Intelligence, cognitive psychology, philosophical epistemology, and the philosophy of science can be brought together under a common denominator. In this respect, the functional model of prototypicality, even if it does not rule out the possible importance of the physiological, the referential, or the statistical explanation, does seem to be more general than the latter, not just because it is based on fundamental principles of cognition, but also because similar views have been put forward in other branches of cognitive science. 6. Onomasiological and Semasiological Aspects of Cognitive Semantics Unfortunately, the optimistic perspective of the previous paragraph does not solve everything. To round off the discussion, I would like to show that a complete explanation of all questions to be raised with regard to ver- nielen and vernietigen is far from available. The picture we have reconstructed so far looks like this: apparently, the linguistic community at some
WHERE DOES PROTOTYPICALITY COME FROM? 225 point in its development finds it convenient to have two distinct categories for the concepts of material and abstract destruction. Thus, a pair of etymologically distinct words becomes available, originally vernielen and vernieten, later on vernielen and vernietigen', their compound character ensures that one of them signifies material destruction, the other abstract annihilation. Gradually, the flexibility that is inherent in all human categorisation extends these concepts beyond their etymological usage; as a result, they have the same range of application in the 19th century. But now consider the original situation in which these flexible extensions have hardly begun taking place. Is it then not irrational to use vernielen to express abstract annihilation, when you already have vernietigen or vernielen to do so? The question can be put in terms of global and local efficiency. As we have argued, the global efficiency of the conceptual system commands its flexible, prototypical organisation. But there is, in the case of vernielen and vernietigen, also a local efficiency principle that says that it is uneconomic to have two terms expressing the same things. We are then forced to ask: why does not the local efficiency principle stop the application of the global principle? Why is not the prototypical extension of vernielen towards abstract forms of destruction checked or prevented by the consideration that you already have a lexical category expressing abstract destruction? There is yet another way of formulating the problem: prototype-based flexibility is necessary because of the expressive needs of the speaker: he may want to express concepts for which no specific term is available. But why then would he use these flexible mechanisms of semantic extension if such a specific term is indeed available? One kind of answer might simply be that the global principle is stronger than the local principle; the global principle simply supersedes the local principle to the extent that local inefficiencies are created. We are then saying that the global principle is so general that its strength overrules the local principle, and that it applies even where it is not strictly necessary. Still, this does not tell us why the local principle is weaker than the global principle. Also, it is rather awkward to explain a mechanism that is unfunc- tionally overproductive on the basis of functional considerations. Couldn't we therefore find a more rational explanation of the flexible extensions? The way out, as far as I can see, is to take into account other kinds of expressivity than the purely conceptual one. Using vernielen to express a concept that is commonly expressed by vernietigen may be conceptually superfluous, but that does not mean that doing so may not serve particular
226 DIRK GEERAERTS expressive purposes. On the level of the linguistic form, for example, it may be quite functional to use another word than the usual one. The varieties of such a formally expressive synonymy are well-known in traditional lexical semantics; near-synonyms may exhibit connotational and emotional differences (as in euphemisms), stylistic differences (as in popular words versus poetic terms), or sociolinguistic differences (as in learned words versus common words). Perhaps we can even say that speakers have an urge for stylistic variation as such, even if the formal variants do not carry specific overtones; variation may well be governed by a straightforward desire to avoid monotony, to create new ways of expressing oneself, to experiment with unexpected innovations as a way of stressing one's own individuality. Moreover, it may well be that the importance of metaphor in natural language is determined precisely by its stylistic expressivity; metaphorical expressions would then be created primarily to add expressive weight to the message one wants to convey. (See the contribution of B. Rudzka-Ostyn to this volume.) It is quite plausible, then, that factors such as these have governed the extension of vernielen and vernietigen beyond their original meanings and into each other's etymological range of application. For instance, using vernielen to express a process of abstract cancellation may have been stylistically particularly expressive, because the process of material destruction normally denoted by vernielen carried overtones of physical violence that were less marked in the case of vernietigen. The extended use of vernielen would then have been a case of metaphorical hyperbole. It is, however, very difficult to pinpoint exactly which form of expressivity is the relevant one with regard to the two verbs that we are concerned with here; our historical material for the earliest (Middle Dutch) history of vernielen and vernielen, for instance, is very hard to interpret with regard to such questions. Still, some clear cases may in fact be found. For instance, the first quotation of (5) in Table 1 clearly carries more overtones of violence, force, and intensity than the second quotation in that pair of examples. (This is mainly made apparent by the presence of a simile, marked by alzoo.) This suggests that the verbs highlight slightly different aspects of the situation described, or rather, represent the situation from different points of view (determined by the prototypical core of each verb). (In Langacker's terminology, the distinction between the two verbs, when used with regard to the same process, might then be characterized as a figure/ground-distinction: vernielen takes the violent process as figure, and vernietigen the destructive result.)
WHERE DOES PROTOTYPICALITY COME FROM? 227 This is not an altogether implausible hypothesis, but it is unfortunately hard to confirm for the simple reason that the historical texts used here do not give us enough clues to discern such subtle differences in stylistic or emotional overtones. On the whole, then, what can we conclude from our discussion of this additional problem? On the one hand, it inspires caution with regard to our attempts to explain prototypical phenomena: the linguistic materials at our disposal do not always allow completely satisfactory answers with regard to the questions at stake to be formulated. On the other hand (and this is, I think, the more important conclusion), the discussion suggests that prototype formation may be influenced by other factors than purely conceptual ones. Stretching the meaning of a lexical item may be motivated by the desire to use another form than the one that is usual to express the idea in question; stylistic, sociolinguistic, connotational expressivity rather than purely conceptual needs may determine the flexible use of a category. In such a case, the conceptual coherence of the prototypically structured category (i.e., the fact that the new, peripheral kinds of usage have to be accessible from the prototypical core) constitutes a limit to the desire for formal variation: you can use a particular lexical item to express an idea that is usually signified by another word, but only on the condition that the idea in question is part of the prototypical potentialities of that lexical item. Basically, you stretch an item's meaning to express something conceptually new, but you can also stretch it to express something conceptually old in a formally new way. This is a very important suggestion, because it implies a warning against a tendency that is a natural characteristic of cognitive semantics: the tendency, in fact, to look for purely cognitive or conceptual explanations of the facts one encounters. Taking the cognitive, experiential, encyclopaedic nature of linguistic signs seriously should not imply looking only for strictly conceptual explanations. Language is not just content: it is also form, and its formal side has an expressivity of its own, which does seem to create lexical configurations that can hardly be explained if we only take into account the conceptual expressivity of language. In the traditional terms of lexical semantics, this means that the explanation of prototypicality should not restrict itself to the semasiological perspective (in which each category is considered on its own), but that the onomasiological point of view (in which it is studied how several items may express similar or identical concepts) should be taken into account as well. Conceptual expressivity is basically a factor connected with the semasiolog-
228 DIRK GEERAERTS ical explanation of prototypicality, whereas the onomasiological influences on prototype formation seem to refer to other kinds of expressivity, as was suggested by our study of vernielen and vernietigen. The incorporation of the onomasiological approach does not mean that cognitive semantics moves away from the functional perspective advocated in the previous section; non-conceptual expressivity is just as much a functional principle as purely conceptual expressivity and cognitive efficiency. Rather, the incorporation of onomasiology implies that cognitive semantics moves much closer to the rich tradition of lexical semantics, in which onomasiological mechanisms and configurations have been thoroughly studied (see Geeraerts 1986: ch. 1). Such a link with traditional approaches can only strengthen the linguistic attractiveness of cognitive semantics. To summarise: I have tried to argue, on the basis of a case study involving the Dutch near-synonyms vernielen and vernietigen, that the functional point of view is the most encompassing, most promising one for studying prototype formation, though it should not be restricted to purely conceptual expressivity and efficiency, but should also take into account the kinds of functional mechanisms that have traditionally been studied by the onomasiological approach to lexical semantics. The fact, however, that prototypicality may come from a number of diverse sources, also implies that an adequate explanation of conceptual structures will not be easy. Appendix. A translation of the quotations. The references given in the table follow the standard abbreviations used in the WNT. Full references can be found in: C. Kruijskamp & A. Persijn, Bronnenlijst WNT, met aanvullingen (1943- 1966; Den Haag: Nijhoff, Leiden: Sijthoff). (1) Like the neighbouring one, this house was not able to stand up against the destructive quarrels of the age. Already under the son of the founder, it was demolished down to the ground. — Only, the earlier cross-windows have beeen replaced by windows in a newer style, and in 1795, the freedom fighters demolished the founder's arms in the facade with their rough hands. (2) Electric current is then transferred to the diversion that has been construed to that end, in which case enough warmth is produced to destroy the coils of the wire in a very short time. — Without this precaution, the wire would be destroyed by the enormous heat of the filament. (3) How much is destroyed by game is aptly expressed by a certain Englishman on the occasion of an attack on the protection these animals enjoy in his country. — During the destruction of weeds by cultivating the land, one should bear in mind their reproduction and development.
WHERE DOES PROTOTYPICALITY COME FROM? 229 (4) By golly! Does not such sentimental foolishness destroy all our inner strength? — Do not put me on a pedestal, she said disconcertedly, I would not live up to that; I would destroy your ideal. (5) The mountain stream in its angry course rends its banks, drowns the valleys: thus, Thou destroyest the hope of Man! — She has been raised from childhood to become Master Daniel's wife, and now this hope of a lifetime has been annihilated! — This coincidence annihilated our entire plan. (6) I still remember that Nicolas Gaal was fond of telling me that the old man used to get so angry and upset about this stolen honour, that it seemed that he would have liked to kill that thief. — Meanwhile, the goddess Nature has decided to destroy the inhabitants of the Floating Islands once and for all. (7) The stroke that destroyed her weak life force in half an hour, had from the first moment paralysed her speech. — He has to take a rest. Such passions undermine the strongest constitution. (8) The women, Lus, his sons, all the others remained silent, destroyed by disconcertedness, nailed to their chairs. — Then she answered slowly, in a dull voice, as if struck down by her own words: "Yes, if it is still possible". (9) This time, the result of the battle was not favourable to him: his entire army was destroyed or dispersed. — The entire army of the Turks was destroyed!
The Natural Category MEDIUM: An Alternative to Selection Restrictions and Similar Constructs Bruce W. Hawkins Illinois State University 1. The Problem* The prepositions italicized in (1) form a natural class in English on the basis of certain shared semantic content.1 I call this shared content [medium]. (1) a. John is in the room. b. John is out of the room. c. John ran into the room. d. John ran out o/the room. e. John ran through the room. f. John scattered papers throughout/all through the room. This paper investigates the semantic phenomenon I am calling [medium] by examining structures involving the prepositions in (1) primarily but not exclusively in their respective spatial senses.2 The purpose of this investigation is to provide an explicit and accurate characterization of [medium]. The approach that I will take in confronting this descriptive problem is a rather unconventional one. The conventional approach to investigating the content common to members of a semantic natural class such as (1) is to examine contrasts across natural classes formed on a similar basis. Analysis of [medium], for instance, would involve investigating how the prepositions in (1) systematically contrast with those in (2) and (3). (2) a. John is on the white carpet3. b. John is off the white carpet3. c. John ran onto the white carpet3.
232 BRUCE W. HAWKINS d. John ran off the white carpet3. e. John ran across the white carpet3. f. John scattered papers all across the white carpeta. (3) a. John is at the front doora. b. John is away from the front doora. c. John ran to the front doora. d. John ran from the front doora. e. John ran (to his car) via the front doora. I have rejected this conventional approach precisely because most, if not all, previous attempts to characterize the semantic phenomenon I am calling [medium] have adopted it and all have failed to meet the test of descriptive adequacy. The accounts given by Bennett (1975) and myself (Hawkins 1984), for example, fail the test of explicitness. Borrowing a definition from WIIP, I characterized [medium] as a condition, atmosphere, or environment in which something may function or flourish. This definition is much too vague. It suggests that [medium] has some significant featural and functional properties but completely fails to specify what those properties might be. Bennett, however, is even less explicit. Appealing to the term [interior]4 rather than [medium], Bennett (op.cit.: 4) assumes that the term is self- explanatory and, therefore, requires no explicit definition. Consequently, we are free to interpret Bennett's descriptive construct in any way we choose. The result of inexplicitness in each case is that the descriptive construct cannot be subjected to a meaningful test of accuracy. This is simply unacceptable. Leech (1969), on the other hand, offers a definition which does pass the test of explicitness. To account for the semantic content common to the prepositions in (1), Leech proposes the descriptive construct [3dime] which he characterizes as 'two/three dimensional'. While the disjunctive nature of this characterization appears at first to be a significant problem, it actually is not. It will be shown in section 2 that Leech's disjunctive characterization of [3dime] is an informal description of its effect rather than a formal definition of its content. Within the descriptive system Leech adopts, the formal definition can easily be stated non-disjunctively. The problem with Leech's account is that, despite its considerable intuitive appeal5, it ultimately fails the test of accuracy. Let us examine this accuracy problem briefly. In doing so, we can gain useful insight into the complexity of the descriptive task at hand.
THE NATURAL CATEGORY MEDIUM 233 Leech cites evidence like that in (4) and (5) in support of his characterization of [3dime]. He observes that the structures italicized in (4) refer to three-dimensional locations (or volumes) while those italicized in (5) refer to two-dimensional locations (or areas). (4) a. Mother never let Father smoke in the house. b. He figured he could scratch as efficiently in an airplane... c. He remembered trudging up the ramp from the smelter of Track 24 into the oven of Grand Central Station. d. An orange-black butterfly danced in the air. (5) a. They were out in the field ... b. He was also a Soviet KGB agent and had been specially selected and trained in Russia. c. I decided to stroll in the park. d. He walked in the garden. Structures like those in (6), however, pose a significant problem for Leech's characterization. These structures clearly demonstrate that the one-dimensionality characteristic of line and circle must also be accommodated by any definition of the phenomenon which I call [medium] and Leech calls [3dime]. (6) a. The road in front of the house ran due north through that line where the last glacier had melted. b. Graham watched three pelicans fly in a line. c. He started to crawl in a circle. The structures in (7) are also problematic. (7) a. Jethro placed a cutting, eye upward, in the spot hollowed out for it. b. Gurney stood in the same spot. c. I did not go through an immigration checkpoint... If one attempts to assign a dimensionality value in these structures, it very likely would be the admittedly abstract property of zero-dimensionality (of which the geometric abstraction POINT is probably the most 'concrete' example). In such a case, the problem with the structures in (7) would be similar to that in (6). On the other hand, if one opts not to assign a dimensionality value to the structures in (7), then they pose basically the same problem as the structures in (8).
234 BRUCE W. HAWKINS (8) a. Do you keep this pearl in a safe place? b. You have got to be in a position of strength. These structures seem to ascribe no dimensionality to the locations identified. How can such non-dimensional items be reconciled in a set of items characterized as 'two/three dimensional1? We will see in section 2 that Leech has an ingenious solution to this problem. However, Leech's solution ultimately proves to be unworkable because it still cannot avoid the problem posed by the well-formedness of structures like those in (6). Finally, the structures in (9) and (10) introduce an entirely different set of problems. (9) a. He cried out in the dark for water. b. Kino stared into the dimness of the little office. c. They cross the coast of Honshu in full daylight. (10) a. She was a dark, elegant, alert girl in black. b. They dress in dark blue or gray. c. a woman in yellow seated at the table. In these structures, language of a more figurative nature is used to describe relations which are essentially spatial. Can such figurative language, whether used to describe relations as concrete as those in (9) and (10) or relations of a more abstract nature, be reconciled with a description like 'two/three dimensional' which was originally intended to account only for the literal use of language? And if so, how? We have seen that the conventional approach to investigating [medium] (or [interior] or [3dime]) has failed in three separate attempts to yield a satisfactory characterization of the phenomenon. This failure is due precisely to the fact that [medium] cannot be understood completely or characterized adequately simply by examining semantic relations within and across natural classes of prepositions; [medium] is more than just a semantic feature common to the prepositions in (1). As we have just seen, the test of accuracy for any account of [medium] inevitably involves checking the account against the content of nominals which occur naturally as objects/landmarks of the prepositions in (1). What makes this test a valid one is that there is a significant lexical category in the grammar of English consisting of nominals which acceptably co-occur with the prepositions in (1) as object/landmark of those prepositions. It is this set of nominals, referred to herein as the category MEDIUM, which will be the central object of investigation in our quest herein for an adequate characterization
THE NATURAL CATEGORY MEDIUM 235 of [medium]. To understand the utility of this alternate approach to the task of defining [medium], it is necessary to recognize that there is a significant grammatical relation between [medium] and MEDIUM. In effect, [medium] is a set of conditions for acceptable co-occurrence with the prepositions in (1) as object/landmark and the category MEDIUM is the set of items which meet those conditions. In other words, [medium] is some sort of abbreviated representation of the category MEDIUM. Thus, an explicit definition of [medium] requires an explicit account of the principles by which nominals become associated with the category MEDIUM. My alternate approach, therefore, involves close examination of the structure of the category MEDIUM. The goal is to give the most explicit and accurate account possible of the structure of this category. Only by reaching this goal will it be possible to provide an explicit and accurate definition of [medium]. The basic argument that I will be making is that attempts to provide an explicit finite definition of [medium] have failed not because investigators have overlooked or misinterpreted crucial data but because the phenomenon itself is not susceptible of such a definition. Any quest for a finite definition of [medium] involves implicit assumption of the classical model of category structure. That is, it is assumed that the structure of the category MEDIUM can be characterized on the basis of a finite set of defining features, i.e. conditions which are individually necessary and jointly sufficient for membership in the category. I contend that any attempt to provide a finite definition of [medium] is doomed to failure precisely because this assumption is a mistaken one. Consequently, I will propose an alternate model of categorial association to account for the structure of MEDIUM. 2. Co-occurrence and Categorization This redefinition of the descriptive task at hand forces us to confront some significant theoretical issues. In recognizing that [medium] is essentially a set of conditions for acceptable co-occurrence with the prepositions in (1), we come face-to-face with the question of what form such co-occurrence conditions must take in a linguistic theory. In seeking to characterize the descriptive construct [medium] by investigating the lexical category MEDIUM, we find ourselves confronting the fundamental issue of categorization, a central theoretical issue throughout the cognitive sciences.
236 BRUCE W. HAWKINS 2.1 Classical Category Structure and Co-occurrence Conditions The specific purpose of this section is to demonstrate that in suggesting that the classical model of category structure is inapplicable in the case of MEDIUM, we are in effect rejecting three distinct proposals for dealing with co-occurrence conditions in a linguistic theory. Following a brief summary of the significant characteristics of a classical category, we will analyze the co-occurrence mechanisms embodied in the selection restrictions of the Katz-Fodor-Postal semantic theory, the ascription features proposed in Leech (1969), and the transfer features proposed in Weinreich (1966). It will be shown that, despite their differences, all three proposals embrace the classical model of category structure. As mentioned earlier, the structure of a classical category is characterized by a set of conditions for membership in the category. Each individual condition is necessary and the set as a collective whole is sufficient for category membership. In the standard application of this model, these defining features actually perform two complementary functions — an associative function and an exclusionary function. Any item which exhibits the defining features of some category can, indeed must be associated with that category. This is the associative function. The exclusionary function is exactly the opposite; any item which does not exhibit the defining features of a category must be constrained from association with that category. As a direct result of these complementary functions, any set of defining features determines not one, but two complementary categories — one composed of items that qualify for association with the defined category, the other of items constrained from such association. This standard application of the model predicts that the boundary between categories is always a very clear one. Any given item either exhibits all of the defining features of some category or it does not. Consequently, that item will qualify for membership either in the associated category or its complement, the excluded category, but not both. Furthermore, it is predicted that all members of the associated category have significant features in common, those being the conditions for association. No such prediction is made in the case of the excluded category, however, since it is defined by an elsewhere condition. That is, the only factor necessarily shared by all members of the excluded category is the failure to meet the criteria for categorial association. As a result, diversity is characteristic of the excluded category. In contrast, the associated category is characterized by equivalence. Although diversity exists within this group, categorial
THE NATURAL CATEGORY MEDIUM 237 association has the effect of highlighting equivalence and essentially filtering out diversity. Consequently, any given member of the associated category is as representative of the category as any other member. There are no degrees of membership within the associated category.6 2.1.1 Selection Restrictions Among the proposed types of co-occurrence conditions, selection restrictions represent the standard application of the classical model of category structure. This is evident in the definition given in (11). (11) Each reading in the dictionary entry for a lexical item must contain a selection restriction, i.e., a formally expressed necessary and sufficient condition for that reading to combine with others. Thus, the selection restriction attached to a reading determines the combinations with readings of other lexical items into which that reading can enter when a projection rule is applied (Katz and Postal 1964:15).7 Selection restrictions effectively divide the set of possible collocations with a particular lexical item into two complementary categories, those that are semantically well-formed, and those that are not. To be associated with that category of items which enter into a semantically well-formed collocation with some item A which bears a selection restriction, item B must exhibit the content of the selection restriction borne by item A. Consequently, all members of the associated category share this significant semantic content. For the purpose of comparison with the other types of cooccurrence conditions, it is useful to recognize that selection restrictions have the characteristics summarized in (12). (12) a. Selection restrictions have both associative and exclusionary functions. b. The associative function is explicit. That is, membership in the associated category is characterized on the basis of explicit necessary and sufficient conditions. c. The exclusionary function is not explicit. That is, the excluded category is defined simply by an elsewhere condition.
238 BRUCE W. HAWKINS 2.1.2 Ascription Features The fundamental problem with selection restrictions that led both Leech and Weinreich to propose alternatives is that they are too restrictive. That is, their associative function is so narrowly defined that selection restrictions often predict that certain naturally occurring collocations should be semantically ill-formed when, in fact, they are perfectly acceptable. Leech outlines his solution to this problem in (13). (13) A solution to the problem of "weakened" selection restrictions is to state them by means of ASCRIPTION FEATURES which do not actually form part of the componential content of the adjacent terminal cluster, but rather are matched against its content, such that co-occurrence with a systematically contrasting component is marked as a violation (Leech 1969:66). Like selection restrictions, Leech's ascription features have both an associative and an exclusionary function. In the case of ascription features, however, it is the exclusionary function that is explicit, not the associative function. This is reflected in Leech's statement that it is not necessary for items associated with the well-formed-collocation category to exhibit the content of the ascription feature. Instead, items excluded from the well- formed-collocation category must exhibit content which systematically contrasts with that of the ascription feature. The effect is to turn the associated category into one defined by the elsewhere condition. Consequently, we could expect that certain members of this category would have nothing in common. In this way, Leech succeeds in proposing co-occurrence conditions with an associative function less narrowly defined than that of selection restrictions. Notice, however, that lie accomplishes this task while still adhering strictly to the classical model of category structure. He has simply applied that model in a way different from its standard application. This analysis of ascription features reveals why Leech's disjunctive characterization of the descriptive construct [3dime] is not really problematic. This construct is one of only three members of a taxonomic system of dimensionality posited by Leech. Given in (14) is Leech's characterization of this system. (14) 'dimensionality' [ldime] 'at - no dimension relevant' [2dime] 'on - one/two dimensional' [3dime] 'in - two/three dimensional' Class of system:multiple taxonomic (op.cit.:\6l)
THE NATURAL CATEGORY MEDIUM 239 The constructs introduced in (14) are proposed by Leech as ascription features characteristic of certain sets of English prepositions. The disjunctive characterization of [3dime] (as well as that of [2dime]) is not really a definition of its content, but rather an informal statement of its associative function. From the definition in (13), it follows that a precise definition of [3dime] must stipulate the semantic content that systematically contrasts with [3dime]. Given Leech's three-part system of dimensionality, it would conceivably be possible to define [3dime] non-disjunctively as 'contrasts with [ldime]\ Leech's solution is valid only as long as he can maintain a three-part dimensionality system. Within a four-part system of dimensionality, [3dime] cannot be defined non-disjunctively and still have the associative function Leech intended. In section 3 I will argue that, in fact, a four-part system is required to account adequately for the semantics of English spatial prepositions. That being the case, not only does [3dime] have the descriptive problems discussed in section 1, but it also becomes inappropriate for the purpose of defining a classical category. Multiple conditions for category membership must be conjunctive rather than disjunctive. 2.1.3 Transfer Features Like Leech, Weinreich proposes a co-occurrence mechanism with an associative function more broadly defined than that of selection restrictions. Weinreich's approach to this problem, however, is quite different from that of Leech. (15) The [proposal] is to adopt a more powerful conception of the semantic interpretation process, in which features of selection restriction of a word Z would be transferred into the path of another word, W, when it is constructed with Z (Weinreich 1966:114). In clarification, Weinreich draws an explicit contrast between the respective functions of selection restrictions and transfer features. (16) The transfer features of the present theory correspond, it seems, to Chomsky's selectional features; the difference lies in the fact that whereas Chomsky's grammar merely ascertains whether the selectional features of the verb correspond to the inherent features of the nouns in its environment (and, in the negative case, discards the incongruous expressions), our theory functions more actively, by transferring the feature from the verb to the nouns (ibid.\\4\).
240 BRUCE W. HAWKINS Weinreich's proposal deviates from the classical model of category structure in two ways. First, transfer features, in and of themselves, have no exclusionary function. Second, while the transferring process described in (15) and (16) constitutes the associative function, it is an associative function different from that of either selection restrictions or ascription features. Note that the content of the item receiving the transfer feature is irrelevant. By virtue of the collocational juxtaposition of Z and W, the transfer of features from Z to W would take place under any of the circumstances described in (17). (17) a. W exhibits the content of the transfer feature even before the transfer takes place. b. W exhibits neither the content of the transfer feature nor contrasting content before the transfer takes place. c. W exhibits content contrasting with that of the transfer feature before (and after) the transfer takes place. In the case of selection restrictions, acceptable collocation would be possible only under the circumstances in (17a). Ascription features, on the other hand, would allow collocation also under the conditions in (17b). Only in the case of transfer features could collocation also take place under the circumstances in (17c). While Weinreich ascribes no exclusionary function to transfer features themselves, his grammar does include a mechanism which performs this function. He refers to it as the evaluator. Once a transfer of features occurs, that collocation becomes subject to review by the evaluator. The evaluator determines the degree of deviance between the feature transferred to W and the inherent semantic features of W. If the resulting deviance quotient is below some criterion, the evaluator marks the collocation acceptable. If the deviance surpasses the criterion, however, the collocation is disallowed. Unfortunately, Weinreich provides no further details on this mechanism. Consequently, we are left to wonder what the acceptability criterion is in any given instance and how it is established. Although transfer features represent a significant departure from the standard application of the classical model of category structure, Weinreich's proposal respects the central tenet of the model — the claim that all items associated to a given category share the defining features of that category. Notice that while items to be associated with a given category via transfer features need not have any significant features in common, all
THE NATURAL CATEGORY MEDIUM 241 items that have been associated to the category share the content of the transfer feature. Thus, transfer features are like selection restrictions in predicting that all members of a category like MEDIUM have common content. Ascription features, on the other hand, make no such prediction. In predicting significant diversity within such a category, however, transfer features are more like ascription features than selection restrictions. In this regard, note that in the case of a collocation which took place under the circumstances described in (17a), Weinreich's evaluator would assign a deviance quotient of zero. The deviance quotient of a collocation resulting from a situation like (17b) would be greater than zero but still lower than that assigned in the case of (17c). 2.2 Toward a Viable Theory of Categorization in the Lexicon Although selection restrictions and ascription features differ in how they apply the classical model of category structure, both adhere quite strictly to that model. Despite significant deviations from the classical model, Weinreich's transfer features represent a modification of the model rather than an outright rejection of it. If, as I claim, attempts to provide an adequate definition of [medium] (or notational variants thereof) have failed precisely because the structure of the lexical category MEDIUM does not conform to the classical model, then we have no choice but to reject these theoretical constructs in dealing with the facts of co-occurrence with the prepositions in (1) and to propose an alternate model of category structure by which these facts can be accounted for. I must point out, however, that, in confronting this task, I do not assume that the inability to characterize MEDIUM as a classical category entails outright rejection of the classical model itself. Indeed, I accept without the slightest reservation that, alongside categories like MEDIUM which do not exhibit classical category structure, there are some lexical categories which do exhibit such structure. The fact that the classical model is inapplicable in the former case does not mean that it should not be applied in the latter case. The challenge, then, is to propose a theory of categorization which predicts that the classical model will be applied whenever possible but which also provides a viable alternative model of category structure which can be appealed to whenever the classical model appears to be inapplicable. The classical model assumes that a category is, in some sense, bounded and, therefore, characterizable by a finite set of necessary and sufficient conditions.8 Upon rejecting this assumption in any particular case, the task
242 BRUCE W. HAWKINS becomes to propose a model of category structure which can account for the facts of categorial association but which also allows for indeterminate categorial boundaries. At present, the most widely recognized alternative to the classical model is that proposed by Rosch (cf. Rosch 1973b, 1978), commonly referred to as the prototype model. However, Rosch's model is not the only alternative that has been proposed. Indeed, Smith and Medin (1981) discuss no less than four distinct alternatives. Other alternatives have been outlined by Langackcr (cf. Langackcr 1982a, in press) and Jackendoff (1983) and still another will be proposed here. All of these alternative models (including that to be proposed herein) attempt to account for the facts of categorial association on the basis of central tendencies within a category rather than defining features.tJ Consequently, I will refer to this as the central tendencies approach to categorization. Any particular model of category structure arising from the central tendencies approach will be referred to as a natural category.10 In attempting to develop a model of natural category structure which can be reconciled with the classical model in a coherent theory of human categorization, I find it useful to appeal to the principles of categorization proposed by Rosch (1978). However, I have taken the liberty of extrapolating from and supplementing Rosch's principles in certain ways (that Rosch might not agree with). As a result, the natural category model that I will propose differs somewhat from Rosch's prototype model. The principle that provides the foundation for both models (and, for that matter, probably every model of category structure) is the PRINCIPLE OF COGNITIVE ECONOMY, which is stated in (18). (18) The task of category systems is to provide maximum information with the least cognitive effort (op.cit.: 28). There is a significant corollary to this principle. Rosch's characterization of the corollary is given in (19). (19) Cognitive economy dictates that categories tend to be viewed as being as separate from each other and as clear-cut as possible (ibid.:35). This corollary, which I will refer to as the PRINCIPLE OF MAXIMAL DISTINCTIVENESS, is of fundamental importance in demonstrating that a category which conforms to the classical model meets the demands of cognitive economy more effectively than one that does not. If we make the reasonable assumption that the principle of cognitive economy governs hypothesis-testing, which lies at the heart of the human learning
THE NATURAL CATEGORY MEDIUM 243 process, then we must conclude that the goal of hypothesis-testing in the case of human categorization is to discover a single, GLOBAL ASSOCIATIVE PRINCIPLE for any category. In other words, a category which can be characterized adequately by a global associative principle serves cognitive economy more effectively than a category which can be characterized only by some disjunctive set of non-global associative principles. The necessary and sufficient conditions which characterize a classical category constitute a global associative principle; these conditions are a conjunctive set which predict all and only the members of the given category. The distinctiveness of classical categories is not just maximal, it is optimal because it is absolute. As it stands, then, the present theory of categorization predicts that the classical model of category structure constitutes the cognitive ideal and, therefore, will be applied whenever it can be applied. The classical model would be sufficient only if the goal of hypothesis-testing is always achieved. However, it seems considerably more realistic to assume that the goal of hypothesis-testing is not always achieved. Failures to achieve the goal could occur either because of human fallibility or because there really is no global associative principle for some given category. In either case, we must conclude that human beings can and do operate with categories defined in some way other than by a global associative principle. Consequently, an adequate theory of human categorization must provide some sort of default mechanism for dealing with these problematic cases. To determine what this default mechanism must be like, we turn once again to the fundamental principles of categorization. The principle of cognitive economy dictates that the cognitive ideal must be applied whenever it can be applied. The same principle also dictates that when the cognitive ideal cannot be applied, we must approximate the ideal as closely as possible. The principle of maximal distinctiveness dictates that classical category structure constitutes the cognitive ideal (in categorization). What is necessary at this point, then, is to recognize what the principle of maximal distinctiveness dictates when the cognitive ideal is not achievable. To that end, it is useful to recognize that, as it applies to a classical category, the principle of maximal distinctiveness can be restated as in (20). (20) a. The global associative principle for a given classical category must predict all members of that category, b. The global associative principle for a given classical category must predict only the members of that category. That
244 BRUCE W. HAWKINS is, the global associative principle of one classical category must not apply to members of any other category. I will refer to (20a) as the ABSOLUTE INCLUSIVENESS CONDITION and to (20b) as the ABSOLUTE EXCLUSIVENESS CONDITION. Obviously, the associative and exclusionary functions of the classical category model are direct reflexes of these two conditions. I propose that when the absolute distinctiveness of classical categories is unachievable, the principle of maximal distinctiveness still involves inclu- siveness and exclusiveness conditions similar to those in (20), but they are not absolute. I suggest that these conditions apply only to a special central portion of the category and, therefore, any category which does not conform to the classical model must have such a categorial center. The necessary restatement of the principle of maximal distinctiveness is given in (21). (21) a. CENTRALITY CONDITION: Any category which does not conform to the classical model must have center-periphery structure. The categorial center will be defined by a single, CENTRAL ASSOCIATIVE PRINCIPLE. b. CENTER INCLUSIVENESS CONDITION: Category membership must be concentrated as much as possible in the categorial center. That is, the central associative principle for a given category must predict as many members of that category as possible. c. CENTER EXCLUSIVENESS CONDITION: The center of one category must be as distinct as possible from the center of any neighboring category. That is, the central associative principle of one category must apply to as few central members of other categories as possible. The centrality condition dictates the center-periphery structure common to all natural category models. The other two conditions draw an important parallel between the center of a natural category and the entirety of a classical category. Taken together, the center inclusive condition and center exclusiveness condition dictate that the closer the center of a natural category conies to exhausting the membership of that category and to meeting the absolute conditions in (20), the more effectively that natural category serves cognitive economy. In this regard, a classical category can be
THE NATURAL CATEGORY MEDIUM 245 seen as the limiting case in which the center of the category actually exhausts the category and meets those absolute conditions. In effect, the three conditions in (21) constitute conditions on the well- formedness of any category which is characterized by a set of non-global associative principles rather than by a global associative principle. However, of the three, only the centrality condition is an absolute condition on well-formedness. That is, it dictates that the only acceptable alternative to a classical category is a category which exhibits center-periphery structure; any other structure would be ill-formed. The other two conditions are relative. That is, what they establish is not a criterion for distinguishing categories that are well-formed from categories that are ill-formed. Instead, they establish a metric for evaluating how closely a given natural category approximates classical category structure.11 The fundamental claim of any natural category model is that the items which constitute the categorial center are generally more representative of the category than are items in the periphery of the category. In the present natural category model, this results from a fundamental difference between the one (and only) non-global associative principle which defines the categorial center and the other non-global associative principles which collectively define the periphery. The special status of the categorial center is directly attributable to the fact that a central associative principle is INDEPENDENT. That is, a central associative principle is always of the general form in (22). (22) If item 1 exhibits property(ies) [x], item 1 is associated to category X. The global associative principle of a classical category is also independent, which may explain the parallel drawn earlier between the center of a natural category and the entirety of a classical category. Adequate characterization of any particular independent associative principle, be it central or global, requires nothing more than specifying the category and the particular property(ies) which are criterial for association to that category. In contrast with the independent status of central and global associative principles, the non-global associative principles which define the periphery of a natural category are DEPENDENT. A dependent associative principle must be of the general form in (23). (23) Given some item 1 whose association to category X is independently motivated and some other item 2 whose association to cat-
246 BRUCE W. HAWKINS egory X is not already established, item 2 can be associated to category X if item 1 exhibits property(ies) [y] and item 2 also exhibits property(ies) [y].12The former item will be referred to as the SPONSOR and the latter item as the CANDIDATE. Properties such as [x] in (22) and [y] in (23) will be referred to as criterial properties. Adequate characterization of any particular dependent associative principle must demonstrate (i) that the association of the sponsor to the particular category is accounted for on the basis of some principle for which there is independent motivation and (ii) that there exists between the sponsor and the candidate some perceivable relation of identity. Notice that the present theory places a severe constraint on what can sanction categorial association. It stipulates that all categorial associations result from perception of a significant relation of identity. This is as true of a natural category as it is of a classical category. All items associated to a category by an independent associative principle share the criterial property specified by the particular principle. All instances of association to a natural category on the basis of a dependent associative principle result from perception of a relation of identity between some item (or set of items) already associated with the category and some other item (or set of items) not already associated. 3. The Natural Category MEDIUM I propose that the concrete property of three-dimensional extension (henceforth [3-d]) is criterial for association to the categorial center of MEDIUM. For the moment, the central associative principle for MEDIUM will be stated informally as follows: any nominal with a semantic structure exhibiting prominently the property [3-d] is associated to MEDIUM. Listed in (24) is a representative sampling of nominals which co-occur naturally with the prepositions in (1) and which qualify for association with MEDIUM on the basis of this explicit criterion of dimensionality. (24) ambulance, anteroom, apartment, arm, armchair, ballroom, barroom, basement, bathroom, bomb bay, beams, bedroom, chamber, classroom, clump, compartment, courtroom, dam, darkroom, diaphragm, dormitory, embers, environment, fragments,
THE NATURAL CATEGORY MEDIUM 247 glass, gymnasium, infirmary, lamb chops, limousine, machine, mailbox, market, masses, maze, meat, mire, missile, mist, mole, monastery, mouth, mouthful, mugs, musicbox, omelette, palm, pram, pyramid, rectum, room, sanatorium, smokehouse, smuggler, stateroom, stomach, storeroom, submarine, supermarket, swamp, temple, tumblers, washroom, woman, wood, workroom In the theory of categorization proposed in the preceding section, assessing the well-formedness of a natural category involves determining the inclusiveness of its center relative to that of its periphery and the exclu- siveness of its center relative to the centers of neighboring categories. With respect to the center inclusiveness condition, I can offer no compelling evidence for (or against) the well-formedness of MEDIUM with its center defined by the criterial property [3-d]. I will report that this criterial property accounts for roughly 60% of the data in my corpus, but this does not constitute much of an argument. The evidence that is needed to make a strong case requires carefully controlled statistical measures which I am in no position to perform at the present time. These measures would include, at least, determining the relative numerosity within MEDIUM and frequency within MEDIUM of items associated to the category by [3-d] and items associated to it by some other property. Statistical measures are not necessary to assess the well-formedness of MEDIUM with respect to the center exclusiveness condition. What is necessary is to determine how distinct [3-d] is from the criterial property of the central (or global) associative principle of any category neighboring MEDIUM. Consequently, it is first necessary to recognize which categories are in the neighborhood of MEDIUM. For this purpose, let us return to the contrasting sets of prepositions in (1), (2), and (3). Recall that the conventional approach to investigating [medium] involves examining how the prepositions in (1) contrast with those in (2) and (3). This approach implicitly recognizes that within each of the latter two natural classes of prepositions there is some shared semantic content comparable to [medium]. Let us call the content common to the prepositions in (2) [surface] and that common to the prepositions in (3) [node]. Just as [medium] has been reanalyzed as the conditions for association with the lexical category MEDIUM, i.e. the set of nominals which acceptably co-occur with the prepositions in (1) as the object/landmark, [surface] and [node] can be similarly reanalyzed. Thus, corresponding to [surface], there is a lexical category SURFACE composed of those English nominals which acceptably co-occur with the prepositions
248 BRUCE W. HAWKINS in (2) as the object/landmark. Similarly, corresponding to [node], there is a lexical category NODE composed of nominals which acceptably co-occur with the prepositions in (3) as the object/landmark. SURFACE and NODE are two of the categories neighboring MEDIUM. At least at the level of semantic description, [medium], [surface], and [node] are notational variants of Leech's ascription features [3dime], [2dime], and [ldime] respectively. From this, it follows that Leech would claim that SURFACE and NODE are the only categories neighboring MEDIUM. Otherwise, Leech's taxonomic system of dimensionality would have more than three members. However, only by ignoring data like those in (25) can Leech maintain this three-part dimensionality system. (25) a. You'll find John along the road somewhere3. b. John ran along the road*. c. John scattered papers all along the road*. These data are often overlooked in studies focusing on the contrast sets in (1), (2), and (3), probably because there are significant gaps in this set. Notice that there are no structures in (25) comparable to (lb), (lc), and (Id). Nonetheless, (25a) is comparable to (la) in basically the same way that (2a) and (3a) are, while (25b) is comparable to (le), (2e) and (3e) and (25c) is comparable to (If) and (2f). On the basis of these contrasts, I propose that the set of nominals which acceptably occur as the object/landmark of the preposition along must also be considered among the categories neighboring MEDIUM. Let us call this lexical category CHANNEL. I have not gathered extensive data on the categories SURFACE, CHANNEL, and NODE. Consequently, I cannot be as confident in proposing central associative principles for these categories as I am in proposing [3-d] as the central associative principle of MEDIUM. Nonetheless, my earlier research on prepositional semantics (which led to the proposal of the descriptive constructs [surface], [channel], and [node]) suggests that it is quite reasonable to hypothesize that the central associative principle in each of those categories is also a strict criterion of dimensionality. Thus, I propose that the central associative principle of SURFACE is two-dimensional extension and that of CHANNEL is one-dimensional extension. As Leech suggests, members of the category NODE are not internally analyzed with respect to dimensionality. This has the effect of filtering out any dimensional characteristics they might otherwise exhibit. Thus, I suggest that the central associative principle of NODE can be characterized accurately as zero-dimensional extension.
THE NATURAL CATEGORY MEDIUM 249 Note that this analysis allows us to account quite nicely for the failure of the natural class in (3) to exhibit a structure comparable to those in (If), (2f), and (25c). The relationship in each of these examples can be characterized as one of co-extensiveness. Obviously, such a relationship can exist only when there is some significant dimensional extension. Consequently, co-extensiveness can involve a three-dimensional MEDIUM as in (If), a two-dimensional SURFACE as in (2f), or a one-dimensional CHANNEL as in (25c). Co-extensiveness cannot occur in the absence of dimensional extension, which is precisely the environment provided by a zero-dimensional NODE. I conclude, therefore, that the explicit dimensionality criteria — [3-d], [2-d], [1-d], and [0-d] — are plausible hypotheses for the central associative principles of the four neighboring categories — MEDIUM, SURFACE, CHANNEL, and NODE, respectively. These four central associative principles are optimal with respect to the center exclusiveness condition. That is, if a certain item qualifies for association to MEDIUM by [3-d], it cannot also qualify for association to SURFACE by [2-d], to CHANNEL by [1-d] or to NODE by [0-d]. It is important to note, however, that the item in question is not simply an unanalyzed nominal or the unanalyzed semantic structure of a nominal. If either of these were the relevant item, the examples in (26) would constitute a clear violation of center exclusiveness, hence a counterexample to my claim. (26) a. The dog is sitting in the boxa. b. The dog is sitting on the boxa. A nominal becomes associated to a category like MEDIUM or SURFACE precisely because some specific portion of that nominal's semantic structure is identical to the criterial property of one of the associative principles that define the category. One portion of a nominal's semantic structure may qualify that nominal for association to one category while a different portion of that semantic structure qualifies the same nominal for association to a different category. This is the case in (26); the three-dimensional interior of the box qualifies box for membership in the categorial center of MEDIUM while any one of the two-dimensional sides of the box qualifies box for membership in the categorial center of SURFACE. At this point, it becomes apparent that adequate discussion of the category MEDIUM requires descriptions of individual semantic structures at least precise and explicit enough to reveal what portion of a given semantic structure is relevant in a given case of association to MEDIUM. To meet this requirement, I will appeal to three fundamental concepts from cogni-
250 BRUCE W. HAWKINS tive grammar — profile, base, and active zone. The profile/base distinction lies at the heart of the cognitive grammar system of semantic description. Proponents of cognitive grammar admit the obvious but still controversial fact that semantic structures are inherently and inextricably tied to human knowledge, which is encyclopedic and includes both accurate and erroneous information about the real world as well as conjectures, beliefs and simple flights of fancy. Any given semantic structure is essentially a particular portion of this encyclopedic human knowledge highlighted in a certain way against the background of knowledge directly or indirectly related to it. That portion of human knowledge highlighted in a particular semantic structure is the profile. The knowledge which constitutes the background against which the profile is highlighted is the base. The most salient aspect of any semantic structure is its profile, but the informative effect of the profile lies in its relationship to the particular base against which it is profiled.13 For purposes of illustration, consider the respective semantic structures of classroom and bedroom. There are obvious differences between the two structures, but their respective profiles are virtually identical. Each profile involves a three-dimensional space enclosed by six quadrilateral sides, each of which meets four other sides at right angles. The important differences between the two structures lie in their respective bases. Prominent in the base of classroom is its association to a school, students and teachers, and such things as desks, chalkboards, and books. The base of bedroom, on the other hand, includes such information as its association to a residence, a rather small number of occupants with some familial or quasi-familial relationship to one another, and such things as beds, closets and/or armaries, and dressers. In structures like those in (27), classroom and bedroom qualify for association to MEDIUM by virture of the fact that the enclosed spaces in their respective profiles exhibit the criterial property [3-d]. (27) a. He had walked out of the classroom. b. He went into the bedroom. The profile of box is quite similar to those of classroom and bedroom. As in (27a) and (27b), it is the three-dimensionality of the enclosed space in that profile that qualifies box for association to MEDIUM in (26a). In (26b), on the other hand, it is a particular portion of the material enclosing that space that exhibits two-dimensionality, thereby qualifying box for
THE NATURAL CATEGORY MEDIUM 251 association to SURFACE. The significance of this contrast is that it demonstrates that a nominal's association to categories like MEDIUM and SURFACE is determined not by its profile per se, but by an even more specific portion of its semantic structure. That particular portion of a nominal's semantic structure which qualifies the nominal for association to a col- locationally-defined lexical category like MEDIUM is identified in cognitive grammar as the active zone.14 Active zones and profiles are very similar semantic phenomena in that both involve highlighting a particular portion of encyclopedic human knowledge for specifically linguistic purposes. The two phenomena must not be confused, however. The profile/base semantic structure of a nominal exists whether that nominal is encountered in a particular linguistic context or in isolation. However, there can be no active zone within the semantic structure of a nominal unless that nominal enters into a collocational relation with some other lexical item. In certain collocational relations, the nominal's profile will indeed constitute the active zone. In other instances, like (26a) and (26b), the active zone is only a portion of the nominal's profile. In still other instances, the active zone is not part of the nominal's profile at all, but rather part of its base. We will see a number of such examples later. Returning to the issue of the well-formedness of MEDIUM, a violation of the center exclusiveness condition occurs only when the exact same active zone within a particular nominal's semantic structure qualifies for association to MEDIUM by [3-d] and to one of the categories neighboring MEDIUM by the criterial dimensionality property defining the categorial center of that neighboring category. The examples in (26) do not constitute a violation because the collocational relations in (26a) and (26b) involve different active zones within the semantic structure of box. 3.1 The Categorial Center of MEDIUM For reasons that will become apparent later, the semantic structure which qualifies a nominal for association to the categorial center of MEDIUM (i.e. the relevant active zone) must be in that nominal's semantic profile. The central associative principle of MEDIUM can now be stated formally. (28) MEDIUM: central associative principle If some active zone within the semantic profile of an English
252 BRUCE W. HAWKINS nominal exhibits the property [3-d], that nominal is associated to the category MEDIUM. This associative principle independently defines the categorial center of MEDIUM. This categorial center is not an amorphous mass of nominals all exhibiting the criterial property [3-d] somewhere in their respective semantic profiles. It is, instead, structured into subcategories each of which is defined by some property (or combination of properties) other than [3- d]. Any such property, though not criterial within the categorial center of MEDIUM, is still significant within the category itself because it may become criterial in the categorial periphery. The concrete property of boundedness is a perfect example. As we will see later, boundedness is criterial in many cases of peripheral association to MEDIUM. At this point, however, we are interested only in the role boundedness plays in the categorial center. To that end, consider the examples in (29) and (30). (29) a. when you're in the cabinet... b. The attendant vaulted into the car. c. I live in a house over there. d. I went off to meet a man in a hotel room. (30) a. Two planes crashed in the air. b. The very small children rolled in the dust. c. He walked uphill in the mire. d. He took a small oyster and threw it in the water. Each of the nominals italicized in (29) and (30) is associated to MEDIUM on the basis of the central associative principle. All of these nominals are significantly similar in that their respective profiles have a particular active zone which exhibits the criterial property [3-d]. There is, however, an obvious contrast across the two sets. This contrast is attributable precisely to the property of boundedness. Each of the nominals italicized in (29) corresponds to a discrete object, semantically identifiable as such by the existence of a complete external boundary in the semantic profile of the nominal. This boundary is not the active zone for association to the categorial center of MEDIUM, but it has the effect of making that active zone a bounded three-dimensional space. Each of the nominals in (30), on the other hand, corresponds to continuous masses with no external boundary in the semantic profile. These are instances in which the nominal's profile itself (not a portion thereof) constitutes the active zone that associates the
THE NATURAL CATEGORY MEDIUM 253 particular nominal to MEDIUM. Among the central members of MEDIUM that exhibit a complete external boundary in their profile, there is another significant distinction to be made. It is apparent in the contrast between the nominals italicized in (29) and those italicized in (31). (31) a. They were cutting steel rods into bullets. b. too much crushed pineapple in the carrot cake c. He chomped into an apple. The nominals in both groups profile three-dimensional objects with complete external boundaries. The contrast is a matter of internal consistency. The discrete objects profiled by the nominals in (29) can be characterized as shells or frames with a solid external boundary circumscribing a three-dimensional interior with a consistency quite different from that of the boundary. The nominals in (31), on the other hand, profile solid chunks of material with little or no contrast in consistency between the external boundary and the material it circumscribes. The presence or absence of a significant contrast in consistency between external boundary and interior has an interesting effect on the relationship between a nominal's profile and the active zone that qualifies it for association to the categorial center of MEDIUM. When there is no contrast, as in (31), the profile and the active zone coincide. However, when there is a significant contrast, only a portion of the profile constitutes the active zone. So far, we have seen only cases in which the active zone includes the three-dimensional bounded interior but excludes the solid boundary. There are, however, instances in which the situation is reversed. That is, the active zone can include the solid boundary and exclude the three-dimensional space it circumscribes. This raises the possibility that two distinct active zones within the semantic profile of a given nominal might both qualify that nominal for association to the categorial center of MEDIUM. The examples in (32) clearly demonstrate that such instances actually do occur naturally. (32) a. She lets the fire die in the stove. b. the cracks in the stove. In both examples, stove profiles the same discrete, three-dimensional object. In (32a), stove is interpreted in the same way as the nominals in (29), with its active zone including the three-dimensional bounded space
254 BRUCE W. HAWKINS and excluding the solid material which constitutes the external boundary of that space. Notice that the fire in this situation is located in the empty space not in the solid material circumscribing it. In (32b), on the other hand, we are concerned only with that solid chunk of material constituting the external boundary. The fact that it circumscribes some empty space is irrelevant to the relation between the cracks and the stove. In this instance, then, the active zone of stove includes the solid external boundary and excludes the space it circumscribes. To this point, all of the examples of boundedness we have been considering involve a boundary that is complete. There are, however, many nom- inals which gain association to the categorial center of MEDIUM by virtue of a three-dimensional-space active zone which is only partially bounded. Listed in (33) are just a few of the nominals in my corpus which exhibit this property of partial boundedness. (33) arms, ashtray, basin, bathtub, bed, bowl, cave, chair, corner, groove, gutter, hand, inkwell, mug, nostril, pocket, seat, sink, teeth, tray, wastebasket, well The range of data in (34) illustrates a significant point concerning the virtual criteriality of boundedness in certain cases of association to MEDIUM by the central associative principle. (34) a. a flower in a vase b. I picked the bottle full of gin off the floor and poured some into a glass. c. She put four lumps in the cup and stirred vigorously. d. She put four drops of the fluid in a petri disha. e. She placed Mary Jane's drink securely in its coaster. Each of these nominals profiles a vessel open at the top. The active zone which qualifies each nominal for association to MEDIUM is a three-dimensional space bounded by the closed bottom and vertical wall of the vessel but unbounded at the top.15 The gradation from (34a) to (34e) is a matter of decreasing length and salience of the vertical wall separating the open upper extreme from the closed bottom. For our present purpose, the significant example is that in (34e). When confronted with this particular example, one native speaker of English (not myself!) noted that the preposition in is appropriate only if the coaster has a perceivable 'lip'. This lip is obviously the vertical wall mentioned earlier. Without such a lip, the appropriate preposition would be on. We can conclude, therefore, that, in this
THE NATURAL CATEGORY MEDIUM 255 instance, the vertical boundary (lip) is actually criterial for perception of a partially-bounded, three-dimensional-space active zone in the semantic profile of coaster. 3.2 The Categorial Periphery of MEDIUM We have now laid sufficient foundation to account for a large number of peripheral associations to MEDIUM. Let us consider first a significant example discussed by Bennett. (35) The men were standing in a circle. Bennett observes that this structure is ambiguous. In one reading, the men are so aligned as to form a circular figure. We will return to this particular reading a little later. Right now, we are interested in that reading of (35) in which the men are standing together in some area circumscribed by a circular line. In this reading, the active zone in the semantic structure of circle involves the concrete property [2-d] (as reflected in the locution 'area circumscribed by a circular line'.) Thus, the association of circle to MEDIUM in this particular instance must be accounted for on some basis other than the central associative principle. The solution to this problem is self-evident. Although this instance of circle contrasts in dimensionality with central members of MEDIUM, it shares with many of those central members the property of boundedness. Thus, I propose that such instances of peripheral association to MEDIUM can be accounted for naturally and realistically on the basis of a dependent associative principle in which these central members of MEDIUM are the sponsors and boundedness is the criterial property. (36) MEDIUM: dependent associative principle 1 (DAP 1) Given the existence of nominals which exhibit the property [boundedness] in an active zone qualifying such nominals for association to MEDIUM by the central associative principle, any other nominal which exhibits [boundedness] in its semantic profile can be associated to (the categorial periphery of) MEDIUM. This principle plays a crucial role in accounting for nearly half of the data in my corpus not obviously associated to MEDIUM on the basis of the central associative principle. In few of these cases, however, is the violation of the criterion of [3-d] as clear as in the case of circle. Consider, for example, the data in (37).
256 BRUCE W. HAWKINS (37) a. Kaz could see the first snowflakes of winter through the cracks between boards on the roof. b. He stood in the doorway. c. paintings in great gilded and ochre frames d. the cartridges in the loops of his tunic front e. He stared straight out through the porthole. In each case, one could conceivably argue that the relevant active zone does involve the property [3-d], thereby qualifying the nominal for association to MEDIUM by the central associative principle. It is significant, however, that no such argument could be made in the case of circle. Thus, these data do not provide the basis for an argument against the dependent associative principle in (36). It is also significant that no such argument is necessary for those data already accounted for by the central associative principle. The point is that while [3-d] is perceived clearly in some semantic structures and [2-d] is perceived clearly in others, there are structures like the active zones in (37) which may be perceived as involving either [3-d] or [2- d]. What this suggests is that violation of the central associative principle is not a simple question of yes or no, but a matter of degree. Bounded active zones which clearly exhibit [2-d] are closer to the categorial center of MEDIUM than are those which clearly exhibit [1-d] or [0-d] but not as close to that center as those in (37) which can be perceived as either [2-d] or [3-d]. We saw a similar gradience phenomenon in the discussion of the data in (34). In both cases, as the criteriality of [3-d] begins to fade, that of [boundedness] becomes more important. From this, it seems reasonable to hypothesize a causal relation between the existence of a dependent associative principle such as (36) and gradience within a category. That is, a continuum of category membership will result from the interaction of a dependent associative principle and the principle (central or dependent) which accounts for the categorial association of the sponsor in that dependent principle. The prediction is that wherever there is such an intersection of associative principles, gradience will be found. The sponsor constitutes the extreme of the gradience continuum closest to the categorial center, hereinafter referred to as the central extreme of the continuum. The principle associating that sponsor to the category determines the parameter of variance within that continuum while the associative principle dependent upon that sponsor determines the parameter of constancy within the continuum. The gradience continuum that includes circle and the data in (37) involves boundedness as the con-
THE NATURAL CATEGORY MEDIUM 257 stant and dimensionality as the variable. The gradience in (34) is from a continuum which involves the same constant but a more complex variable resulting from the interrelation of dimensionality and completeness of boundary. The DAP 1 continuum (in which dimensionality alone is the variable) figures quite prominently in the structure of MEDIUM. At one extreme of this particular continuum we find nominals like those in (29) and (31), in which the active zone is three dimensional and completely bounded in all dimensions. A discrete step away from this central extreme is the instance of circle we have been discussing. The active zone in that case is two dimensional and completely bounded in both dimensions. The active zones associated to MEDIUM in (37) are situated somewhere between circle and the central extreme. Those active zones can be interpreted either two- or three-dimensionally, but they are completely bounded only in two dimensions. On the other side of circle we find examples like those in (38), in which the active zones are probably interpreted as being two-dimensional but are completely bounded only in one dimension. (38) a. Madeleine genuflected in the aisle. b. In the driveway Ron was racing the motor of the Volkswagen. c. The soldier had shouted at two tiny figures far ahead in the road. d. a small dry mustached man in the front row e. He has his driver tagging along out in the street. Finally we reach the non-central extreme of this continuum at which we find examples like spot in (39). The semantic structure associated to MEDIUM in this case is of questionable dimensionality, but completely bounded nonetheless. (39) a. Jethro placed a cutting, eye upward, in the spot hollowed out for it. b. Gurney stood in the same spot. Among the peripheral associations to MEDIUM in this particular continuum, those between circle and the central extreme are easily the most prevalent. Foremost among these are geographical and geopolitical areas with well-defined complete boundaries in two dimensions. Given in (40) is a representative sampling of these very common but nonetheless peripheral instances of MEDIUM.
258 BRUCE W. HAWKINS (40) America, camp, city, clearing, country, field, garden, Harlem, Los Alamos, Hiroshima, Germany, Montreal, St. Mary Mead, VietNam, Birmingham, West Palm Beach, Baltimore, Bermuda, Cambridge, Camden Town, Columbus, East Germany, Joplin, Larchmont, Ludeyville, Marin County, Massachusetts, Mexico, Michigan, Minneapolis, Missouri, Mobile, Moscow, Mount Vernon, New Mexico, Northumberland, Oklahoma, park, state, town, Vermont, village, West Germany Each of the nominals in (40) profiles a configuration with well-defined complete boundaries in two dimensions. What makes it possible to identify these configurations as geographical or geopolitical areas is that they are profiled against a base in which knowledge of the heavenly body Earth figures prominently. With this description in mind, consider the examples in (41). (41) a. He doubted that there were such wooded hills or winding creeks in the cornbelt of the north. b. the $2 billion doomsday machine sitting in the desert c. It had repercussions, naturally, in every Negro ghetto. d. The car curved around into the small hills. e. Negro soldiers received their military training in the South. Like the nominals in (40), those italicized in (41) arc recognizable as geographical or geopolitical areas. However, the respective active zones of the nominals in (41) do not exhibit well-defined external boundaries. Consequently, the association of these particular nominals to MEDIUM cannot be accounted for on the basis of DAP 1. Instead, I suggest that they can be accounted for most naturally with a dependent associative principle in which the nominals in (40) constitute the sponsor and the criterial property is profiling relative to a base in which knowledge of Earth figures prominently. (42) MEDIUM: dependent associative principle 2 (DAP 2) Given the existence of nominals which exhibit [profiling relative to Earth as a base] and which qualify for association to MEDIUM by DAP 1, any other nominal which exhibits [profiling relative to Earth as a base] can be associated to (the catego- rial periphery of) MEDIUM. In essentially the same part of the DAP 1 continuum as the nominals in (40), we find nominals like those italicized in (43), which involve discrete,
THE NATURAL CATEGORY MEDIUM 259 observable parts of the human body. (43) a. He pushed the palms of his hands into his eyes. b. Sugar Ray landed a clean left jab in the mouth*. Notice that there is a distinct difference in terms of association to MEDIUM between the instances of eye and mouth in (43) and (44). (44) a. There was trust as well as inquiry in his clear, hazel eyes. b. The classic ways to smuggle things are in the mouth, ... The association of the relevant active zones in (44) to MEDIUM can be accounted for on the basis of the central associative principle. Those in (43), on the other hand, are probably best accounted for on the basis of DAP 1. It is significant that each of the nominals italicized in (43) profiles a configuration with well-defined boundaries in two dimensions against a base in which knowledge of the human body figures prominently. With this description in mind, consider the examples in (45). (45) a. He kicked me viciously in the chest. b. That didn't keep him from trying to kick me in the groin. c. Get a little broadside and shoot him in the neck. d. Then, he reared back and let me have it in the stomach*. The contrast between the nominals italicized in (43) and (45) is directly analogous to that between the relevant nominals in (40) and (41). In this instance, the association to MEDIUM of the nominals in (43) can be accounted for by DAP 1, but the nominals in (45) must be associated to MEDIUM by a dependent associative principle similar to DAP 2. (46) MEDIUM: dependent associative principle 3 (DAP 3) Given the existence of nominals which exhibit [profiling relative to the human body as a base] and which qualify for association to MEDIUM by DAP 1, any other nominal which exhibits [profiling relative to the human body as a base] can be associated to (the categorial periphery of) MEDIUM. DAP 2 and DAP 3 raise two important points, one concerning the nature of the criterial properties in those principles and the other concerning the categorial association of the sponsors in each case. With regard to the former, recall that the present model of category structure assumes that all instances of categorial association are based on perception of some relation of identity. The criterial property of any given associative principle represents the specific identity relation upon which some set of categorial
260 BRUCE W. HAWKINS associations is based. In this respect, it is significant that the respective criterial properties of DAP 2 and DAP 3 are quite similar to one another in nature and considerably different in nature from the criterial properties of the central associative principle and DAP 1. What this indicates is that the identity relations which sanction categorial association fall into certain general classes. The identity relations represented by the criterial properties of the central associative principle and DAP 1 are both featural in nature. Those represented by the criterial properties of DAP 2 and DAP 3, on the other hand, are quite different. Indeed, they are rather complex, involving both identity of a non-featural, holistic property (unanalyzed profile) and a significant relation of contextual identity (profiled relative to some specific domain of human knowledge in the base). With respect to the sponsors in DAP 2 and DAP 3, notice that their association to MEDIUM is accounted for not by the central associative principle but by a dependent principle, DAP 1. This forces us to recognize the possibility of peripheral members of the category gaining association by virtue of identity to members associated to the category by some principle other than the central associative principle. In this way, the present model of natural category structure provides another means of accounting for gradience in category structure. Nominals like those in (40) and (43), which are associated to MEDIUM on the basis of DAP 1, are less representative of the category than nominals associated to it by the central associative principle but more representative than nominals associated to it by DAP 2 or DAP 3. None of the associative principles proposed earlier can account for the association to MEDIUM of the nominals in (47), (48) and (49). (47) a. She was a dark, elegant, alert girl in black. b. They dress in dark blue or gray. c. a woman in yellow seated at the table (48) a. a girl in sleazy saiin b. I floated away in a white organdy. (49) a trim little girl in pink plaid It is necessary, then, to propose yet another associative principle. The nature of the criterial property of that principle is probably difficult to recognize but the identity of the sponsor should be obvious. If it is not, the paraphrases in (47'), (48') and (49') make it obvious.
THE NATURAL CATEGORY MEDIUM 261 (47') a. She was a dark, elegant, alert girl in a black dress. b. They dress in dark blue or gray suits. c. a woman in a yellow dress seated at the table (48') a. a girl in a s/eaz)> satin dress b. I floated away in a white organdy dress. (49') a trim little girl in a pink plaid suit The sponsors are nominals which profile articles of clothing. Any article of clothing constitutes a boundary which circumscribes a space canoni- cally occupied by the wearer. This space, which must be three-dimensional if it is to be occupied by a human being, is the active zone which qualifies nominals like those in (50) for association to MEDIUM by the central associative principle. (50) uniform, suit, shirt, coat, pants, jacket, boots, dress, hat, sandals, slacks, sweater, blouse, jeans, socks, bathrobe, Bermudas, bloomers, bonnet, britches, cap, diapers, dungarees, glasses, glove, gown, kilt, kimono, knickers, overcoat, shawl, shoes, shorts, skirt, slippers, tuxedo, undershirt, wig, apron, cloak As in the case of the nominals in (34), the boundary circumscribing the active zone in each of these nominals is incomplete. Very few articles of clothing constitute a complete boundary; most have significant openings from which protrude portions of the body which either remain uncovered (e.g. head and hands) or are covered by some other article of clothing (e.g. feet). To understand the criterial property in the dependent principle associating the nominals in (47), (48) and (49) to MEDIUM, it is necessary to understand first the nature of any semantic structure identifiable as clothing. Identifying anything as an article of clothing necessitates profiling some structure against a base crucially involving knowledge of the human body as well as of the social and/or physical requirement that the body be covered in certain environments. In canonical instances, the structure profiled is the item of clothing itself while global properties of clothing, such as color, material or pattern remain in the base. However, the examples in (47), (48) and (49) illustrate that one or more of those global properties may actually be brought into the profile in place of the item of clothing itself, and the resulting profile/base structure is still identifiable as some (perhaps generic) article of clothing.16 As in DAP 2 and DAP 3, then, the significant identity relation that unites (47), (48) and (49) with (50), thereby sanctioning association to MEDIUM, is contextual in nature.17
262 BRUCE W. HAWKINS (51) MEDIUM: dependent associative principle 4 (DAP 4) Given that nominals referring specifically to articles of clothing qualify for association to MEDIUM by the central associative principle and that those nominals involve [profiling against a base involving the human body and the social and/or physical requirement that it be covered in certain environments], any other nominal which refers to clothing by virtue of [profiling against a base involving the human body and the social and/or physical requirement that it be covered in certain environments] can be associated to (the categorial periphery of) MEDIUM. Now consider the examples in (52) through (59). (52) a. He cried out in the dark for water. b. Kino stared into the dimness of the little office. c. They cross the coast of Honshu in full daylight. (53) I sat down in the sun to dry. (54) a. its head under her waist out of the cold b. those who wished to sit in the coolness of a spring night c. The two were gossiping happily in the warmth of the morning. (55) a. He did his best thinking in the scent of oil vapors. b. I was enveloped in a peculiar mixture of a pungent and fragrant smell. (56) They had accustomed themselves to living in silence. (57) a. He stood for an hour in the icy afternoon. b. The bomb was gone. Ferebee looked up from his sight and watched it fall through the morning. c. Misao stumbled out into the day. (58) a. I rolled out of range. b. I poured another slug of the gin and held it out of his reach. c. There wasn't a fighter or a puff of flak in sight. (59) a. His friend Bonnard generally played the fool in public. b. I signed it thus in the presence of his lieutenants. c. You can read anything you care to, on your own time and in privacy.
THE NATURAL CATEGORY MEDIUM 263 The dependent principle by which the nominals italicized in these examples are associated to MEDIUM involves a contextual identity relation similar to that in DAP 4. The sponsors in this case are central members of MEDIUM which profile either bounded or unbounded three-dimensional spaces. Thus, among the sponsors are nominals like room, atmosphere and environment. The profiles of these nominals include no information concerning the conditions that exist in the profiled space. However, among the information included in the base are the facts that the space will be illuminated in some way, it will have a certain temperature, it may be filled with a particular odor or sound, it will be experienced at a certain time of day, and it may or may not be occupied by certain individuals or groups. The nominals italicized in (52) through (59) have the semantic effect of raising one or more of these environmental conditions into the profile and relegating to the base the particular three-dimensional space in which the condition exists. This feature of the base is the criterial property in DAP 5. (60) MEDIUM: dependent associative principle 5 (DAP 5) Given that nominals profiling a [three-dimensional space] qualify for association to MEDIUM by the central associative principle, any nominal which profiles an environmental condition against a base involving a [three-dimensional space] can be associated to (the categorial periphery of) MEDIUM. Notice that if the central associative principle had not been restricted to information in the profile, DAP 5 would be unnecessary because the nominals in (52) through (59) exhibit the criterial property [3-d] in their base. However, my native intuitions tell me that these nominals are not as representative of MEDIUM as nominals with a [3-d] active zone in profile. For this reason, the restriction on the central associative principle seems necessary. Let us conclude this discussion of the categorial periphery of MEDIUM by returning to the reading of (35) in which the men are so aligned as to form a circular figure. In this reading, the active zone associated to MEDIUM is the solid, circular boundary formed by the men, not the empty space circumscribed by that boundary. In accounting for categorial association in this instance, it might be possible to argue that the configuration of bodies is three-dimensional, thereby qualifying for association by the central associative principle. On the other hand, the boundary of a circle is generally considered to be linear, i.e. one-dimensional. In that
264 BRUCE W. HAWKINS case, this example would require a dependent associative principle. When faced with such a situation, it is wise not to rush out and search for arguments on one side or the other. In effect, such a pursuit amounts to an attempt to argue away questionable judgments. A central lesson of this discussion of natural category structure is that such judgments are valuable pieces of evidence in and of themselves. When confronting such evidence, it is wisest to consider the possibility that the questionable judgment arises because the data in question are located near but not at the central extreme of some gradience continuum not already recognized. What we should be looking for, then, are sets of data on either side of those evoking the questionable judgment. If there is a continuum to be found, we will find data on one side which qualify unquestionably for association by the central associative principle and data on the other side which must be associated to the category in some other way. The principle which accounts for these associations will be revealed by the factor common to all the data in this particular continuum. With this discussion in mind, we return to the data and find examples like those in (61). (61) a. He was curiously clever at all kinds of things that couldn't be the slightest use ... cutting a banana or some such thing into a dancing doll. b. sandstone which had been weathered into a sort of bal- coney c. the hair in a bun d. the straight hair screwed up into a knob e. bodies heaped in great piles f. The men were gathering in groups to talk. Each of the nominals italicized in (61) profiles a three-dimensional object referred to earlier as a solid chunk because its interior is of the same solid consistency as its external boundary. This profile constitutes the active zone which qualifies the nominal for association to MEDIUM by the central associative principle. Nominals like these constitute the sponsor in the dependent principle by which the relevant reading of circle in (35) is associated to MEDIUM. To recognize what the criterial property is, consider the example in (62). (62) He could see the little image of the consecrated candle reflect in the soft surface of the pearl.
THE NATURAL CATEGORY MEDIUM 265 In this instance, surface profiles a solid chunk of material that is two- dimensional rather than three-dimensional. It is this two-dimensional chunk that is the active zone associated to MEDIUM by some dependent principle. The external boundary in the profile of circle can be interpreted as a one-dimensional solid chunk and qualify for association to MEDIUM by the same dependent principle. This is true also of the linear configurations profiled by each of the nominals italicized in (63). (63) a. He started to crawl in a circle. b. She began to move around the floor in graceful circles. c. She whirled in a half-circle. d. Graham watched three pelicans fly in line. e. Lesser men in perfect lines strode along with guns. It becomes apparent at this point that these data do not require a new dependent principle; they can be accounted for naturally and realistically by DAP 1, the criterial property of which is boundedness. I have reserved consideration of these data for the end of this discussion of MEDIUM for an important reason: they draw attention to abstract dimensions of the category. In this respect, notice that the relations depicted in (61), (63) and relevant reading of (35) essentially elaborate upon the relation X IN THE FORM OF Y. This leads to the realization that there is a significant functional identity uniting certain nominals associated to MEDIUM. Many such nominals, especially those which profile a chunk (be it one-, two- or three-dimensional), share the property of being able to function as Y in some elaboration of X IN THE FORM OF Y. Consider now the data in (64), (65) and (66). (64) a. her brown hair rolled in flat invisible-pinned curls b. Their hair stood in spikes. c. two long poles that ended in wisps (65) a. Part of the far shore disappeared into a shimmer. b. The deer watch, the mule ears alert; in silhouette they look like giant hares. c. The name in sparkling lights was Bella's. d. The name Enola Gay survives in faded black paint. (66) a. She spoke in a forlorn, pleading note. b. Your voice will remain in a monotone. c. "DANGER" signs in English and Spanish
266 BRUCE W. HAWKINS d. a 16x16 shack known in Army argot as a "hutment" e. That the employer called the tune may be seen in the stark terms of my salary. f. Chani saw the sadness in Jessica s expression. g. Her lip curled in a sneer. All of these examples involve a relation quite similar to X IN THE FORM OF Y. The instantiations of this relation in (64) and (65) are almost as concrete as in the examples in (61) and (63). The relations in (66), however, are considerably more abstract. With respect to the association to MEDIUM of the nominals italicized in (66), I suggest that the necessary dependent principle involves a criterial property which captures the functional identity that has been pointed out. I am not presently in a position to formulate this principle. For our present purposes, however, the important point is not how the principle would be formulated but the fact that it introduces into MEDIUM a gradi- ence continuum which reaches into the domain of metaphor. I could present other examples of MEDIUM reaching into this abstract domain in other ways but I will refrain from doing so. It was not my original intention to confront the issue of metaphor in this paper and, consequently, I haven't enough relevant data to do so productively. Nonetheless, the examples that have been presented are sufficient to demonstrate that no account of MEDIUM can be complete without confronting metaphorical associations to the category. However, given that one of the leading approaches to dealing with metaphorical relations depends crucially on the concept of similarity (cf. Paivio 1979 and Ortony 1979a), I suggest that metaphorical associations to MEDIUM can be accounted for in exactly the same way as any other peripheral association to the category, i.e. by means of a dependent associative principle. Empirical verification of this hypothesis is an interesting problem which can be confronted in some other forum. 4. Conclusions and Inconclusions This examination of the category MEDIUM has led to the proposal of six reasonably explicit associative principles, one central and five dependent. Given that we have not confronted the issue of metaphor, thereby leaving a significant portion of the category unexamined, it is natural to wonder what has been accomplished. I suggest that we have accomplished a great deal indeed. With respect to the original descriptive problem of
THE NATURAL CATEGORY MEDIUM 267 defining the semantic phenomenon I call [medium], this examination of the category MEDIUM clearly reveals the source of the difficulties encountered by those linguists who have tried to give an accurate, explicit, and concise definition of [medium] (or notational variants thereof). In this regard, we appear to have two options. Either we conclude that [medium] is to be equated with the sum total of principles that account for associations to MEDIUM or that [medium] can be equated with the central associative principle with the understanding that other principles are involved. If we choose the first option, our characterization of [medium] could be accurate and explicit, but not concise. On the other hand, if we choose the second option, our characterization of [medium] is explicit and concise, but only because it is incomplete. I will leave the judgment of which of these options is preferable to those better informed than I concerning the facts of mental representations.18 With respect to the theoretical issues of categorization and co-occurrence, we have developed an explicit, principled model of natural category structure which can account not only for the facts of categorial association, but also for the related phenomena of centrality, gradience and indeterminate boundaries. Furthermore, we have laid the foundations of a theory of categorization in which this model can be reconciled with the classical model of category structure. Such a theory is necessary if all lexical categories are to be dealt with in a manner that is true both to the observed facts and to the fundamental principle of cognitive economy. Before closing, it is necessary to point out a pair of issues that need to be confronted in any follow-up examination of MEDIUM. The first concerns the composition of MEDIUM. The present investigation has been strictly data-driven. Throughout the discussion, I have assumed that cooccurrence with any of the prepositions in (1) is evidence that a nominal is associated to the category MEDIUM but there is a significant problem with this assumption. If a given nominal is truly a member of MEDIUM, shouldn't it co-occur with all of the prepositions in (1)? This points out the need to take a conceptually-driven, data-based approach in any further examination of MEDIUM. Such an examination might well reveal that MEDIUM is actually the intersection of six slightly different collocational- defined categories which we can call MEDIUM. , MEDIUM ,, ° m9 out of7 MEDIUM , MEDIUM p> MEDIUM, „ and MEDIUM, u . If into7 out of7 through7 throughout this turns out to be the case, then it would seem necessary to examine closely both how and why these categories differ.
268 BRUCE W. HAWKINS The second issue concerns the matter of constraints on a model of natural category structure. The model that has been proposed here places inclusiveness and exclusiveness conditions on the categorial center, but no conditions whatsoever on the categorial periphery. With respect to the inclusiveness of the periphery, might there be some limitation on the number of dependent associative principles in a given category or on the nature of the items that can serve as a sponsor of a dependent principle or on how far from the categorial center a given gradience continuum may reach? With respect to exclusiveness, we should expect the peripheries of two neighboring categories to intersect, but might there be a constraint on how much they can intersect? Furthermore, might there be some constraint on how much the periphery of one category can encroach on the center of a neighboring category. The category MEDIUM provides a perfect environment for confronting this particular question since many nominals exhibit an active zone which qualifies for association to MEDIUM by DAP 1 and to SURFACE by its central associative principle. As in the case of metaphor, these theoretical questions and the methodological question in the preceding paragraph are matters to be taken up in another forum. Notes * I am grateful to the following people for their insightful comments on an earlier version of this paper: Dirk Geeraerts, Noriko Kawasaki, Brygida Rudzka-Ostyn, Pierre Swig- gers, John Taylor, and Claude Vandeloisc. I owe a special debt of gratitude to my colleague Kevin R. Gregg. This paper benefitted greatly from lengthy discussions we have had at various points in the development and subsequent presentation of the analysis. Of course, any remaining shortcomings are strictly the responsibility of the author. 1. This observation is based on native-speaker intuitions, primarily but not exclusively my own. I assume that the observation is not a controversial one since various investigators of English prepositional semantics have reported similar observations. Quirk et al. (1972) recognize exactly the same natural class as that presented in (1). Others have recognized basically the same class but have omitted (not excluded) one or two of its members. Leech (1969) and Dirven (1981), for instance, both omit throughout while Catford (1959) omits both through and throughout. 2. The present analysis is drawn primarily from a corpus of roughly 3300 instances of the prepositions in (1). These data were collected from recent novels, short stories, and magazine news features. Most, but not all of the authors whose works were scoured for data are (or were) American. (I wish to express my gratitude to Toshiya Matsuo and Kyoko Fukuda for their assistance in gathering these data.) As much as possible, the data presented herein will be examples that actually appeared in print. In a number of instances, this investigator has found it useful to supplement this corpus with examples which were not found in the works surveyed. This, in fact, is true of the examples in (1).
THE NATURAL CATEGORY MEDIUM 269 For the remainder of the paper, all such artificial evidence will be identified explicitly as such. The superscript a (for artificial) at the end of any example indicates that that particular example was created by this investigator (a native speaker of English) specifically for the purpose of this discussion. To insure that the investigator had not conveniently altered his own native intuitions for the purposes of a nice analysis, at least one other native speaker of English was consulted for judgments on each piece of artificial evidence. 3. Webster's Third New International Dictionary, p. 1403. 4. Technically, Bennett does not recognize a semantic natural class composed of the prepositions in (1). The semantic content I am calling [medium] is captured by Bennett in two supposedly distinct semantic components — [interior] and [exterior]. The former is attributed to in, into, and through, the latter to out of. (Like Leech (1969) and Dirven (1981), Bennett does not consider the spatial sense of throughout.) It is intuitively obvious, however, that there is a close semantic relationship between these two components. Indeed, Bennett implicitly recognizes it in defining [exterior] as the converse of [interior] (Bennett 1975:78). Cooper (1968) demonstrates that [exterior] is, in fact, superfluous. It can be characterized quite nicely as separation from [interior] in: location(x, interior(y)) out of: separation(x, interior(y)) Elsewhere (Hawkins 1984 and 1985) I have given detailed discussions of the reasons for favoring Cooper's analysis over Bennett's. 5. Dirven (1981) appears to have arrived independently at precisely the same characterization as Leech while Quirk et al. (1972) and Lyons (1977) seem simply to have adopted Leech's proposal. 6. However, it has been demonstrated that the classical model of category structure can easily be modified to allow for degrees of membership within the associated category. Cf. Zadeh (1965) and Smith and Medin (1981). 7. Cf. also Katz and Fodor (1963:191-2). 8. Although still widely accepted, this assumption has been subject to increasing criticism in recent years, most notably in the work of Rosch. Given the context of the present discussion, it is worth noting that studies in lexical semantics have played a major role in maintaining the challenge. Significant examples include Labov (1973), Kay and McDaniel (1978), Coleman and Kay (1981), Lindner (1981), Brugman (1981, 1984) and Coates (1983). 9. This paper is not the first to suggest that the facts of linguistic co-occurrence are best accounted for on the basis of strong tendencies rather than inviolable rules. As revealed in the passage that follows, this very point is the central thesis of Haas (1973). This is not to state a rule; it is to state what happens 'as a rule':which implies that there is no rule about it. We are here concerned with something akin to probability-statements; though, since we have to rely on judgments of acceptability and comparative normality rather than on statistical measures, we might refer to habits and tendencies, rather than probabilities. That norm, then, which has been mis-stated as a rule, is in fact a tendency certain verbs have towards co-occurring with 'animate' subject-nouns. It is this tendency that accounts for the metaphorical flavour of contravening sentences {pp. cit.: 146).
270 BRUCE W. HAWKINS 10. For an extensive overview of the significance of natural category structure in linguistic theory, see Lakoff (1982b). 11. These relative well-formedness conditions appear to have an effect quite similar to Jackendoffs preference rules. Cf. Jackendoff (1983:128-158). 12. It should be obvious that this characterization was inspired by Wittgenstein's notion of family resemblances among certain sets of members of a given category. JLangacker's notions of partial schematicity and partial sanction also provide significant insights into virtually the same phenomenon. 13. For a more in-depth discussion of profile and base, cf. Langacker in press. 14. Langackcr (1984) presents a detailed discussion of the active-zone phenomenon and its grammatical significance in English. 15. Vandeloisc (I985f) suggests that in cases such as those in (33) and (34), the perceiving being actually closes the incomplete boundary. Furthermore, he suggests that this perceptual closure is necessary for associating open containers in the same category as closed containers. I suggest that this is a case of putting the cart before the horse. That is, if Van- dcloise's initial suggestion is correct and the perceptual closure actually occurs, it is a result of categorial association, not a prerequisite for it. Categorial associations are made strictly on the basis of identity. It is as a result of categorization that features of dissimilarity can be ignored such that unlike items (open vs. closed boundary) can be treated as being alike (container). 16. Given this characterization, it becomes apparent that these examples illustrate the phenomenon identified in cognitive grammar as conventional imagery, which refers to the human cognitive ability to structure a given portion of knowledge in a number of different ways. 17. The semantic relationship uniting the nominals in (47), (48), (49) and (50) is commonly identified as metonymy. This appears to create a conflict with the present analysis, since metonymy is defined as involving the relation of contiguity (cf. Ullmann 1962), which has traditionally been in opposition to the relation of identity or similarity (cf. Hume 1748). Contiguity is seen as involving some external relationship while similarity/identity involves an internal relationship. The profile/base distinction allows us to reconcile the problem by characterizing similarity/identity as internal to the semantic profile and contiguity as alternate profiling relative to the same base. Given this characterization, contiguity and contextual identity become virtually synonymous. 18. Rosch (1978:40) suggests a preference for the second option.
Spatial Expressions and the Plasticity of Meaning Annette Herskovits Wellesley College 1. Introduction* The problem of lexical meaning is the skeleton in the closet of both artificial intelligence and logical semantics. The lexical representations used in artificial intelligence rarely even approximate speakers' uses, and in a majority of cases it is impossible to provide — as logical semantics requires — necessary and sufficient conditions to define word meanings. Often, pragmatics is called upon to explain the discrepancies between meaning as produced by the theory and actual use, but this pragmatics is mostly uncharted. And those semanticists who pay attention to the complexities of word meaning, and have made many subtle observations about word use, have failed to tie them all together into a theory that explains composition or entailment. In this article, I will propose an analysis of spatial prepositions that respects the intricacy and seeming capriciousness of their behavior, and yet accounts for the meaning of complex locative expressions, and for their use in context. This analysis will be based on a prototype — a better term would be "ideal" — meaning for the prepositions. This ideal is only part of the information needed in the lexical entry and deviations from the ideal will be precisely described. Prototypicality makes possible an account of polysemy; it also underlies some pragmatic processes of "allowance," or "tolerance." I will call "locative expression" any spatial expression involving a preposition, its object, and whatever the prepositional phrase modifies (noun, clause, etc.), such as: the spider on the wall Jenny is at the playground.
272 ANNETTE HERSKOVITS He is washing the dishes in the sink. Whether one approaches these expressions from the perspective of linguistics or artificial intelligence, two fundamental questions present themselves: (a) Given a locative expression used in a particular situation, can one predict what it conveys, how it will be interpreted — that is, provided it has been used appropriately? If not, can one explain the inappropriate- ness? (b) Given a situation with two spatial objects, can one predict the locative expression(s) that can be used truly and appropriately to describe their spatial relation? At first sight, the problem appears relatively straightforward. We picture the world as made up of objects, each with a well-defined shape and position in space at a given time. Nouns denote the objects. Prepositions are assumed to denote simple spatial relations between two objects ("containment," "support," "contiguity," etc.), or sometimes among three objects (e.g., between). The meaning of a locative expression (say a copulative expression, such as The mouse is in the hole) is a proposition predicating the relation denoted by the preposition of the objects referred to by the noun-phrases. So, for instance, the meaning of in might be:1 In(X,Y) iff Included (Part(X),Y) (X is in Y if and only if a part of X is spatially included in Y). For on, at least two meanings are needed, one for three-dimensional examples, like the book on the desk: On^Y) iff Contiguous(X,Y) and Support(Y,X) and one for two-dimensional examples like the house on the lake: On2(X,Y) iff Contiguous(X,Boundary(Y)) Then, given The mouse is in the hole, one can apply a simple compositional rule: insert symbols denoting the referents of the noun-phrases as arguments of the relation defining the meaning of in above. This yields: ln(Mouse,Hole) iff Included (Part(Mouse),Hole) which is the relation that holds true in the situation described by The mouse is in the hole. That formula is "the meaning" of the sentence. But this "simple-relations model" is inadequate.2 Such simple relations together with the compositional rule yield wrong or insufficient predictions about the set of situations (intensionally the situation type) described by the
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 273 locative expression. Current pragmatics gives us no clues on how to bridge the gap; and to dispose of the problem by labeling the uses not predicted by the model "idiomatic'1 often sounds like obfuscation — it fails to account for speakers' frequent intuitions that such uses are motivated. This article will have two main purposes: first, to present evidence of the rather unexpected unruliness of locatives; second, to propose a model of their meaning and use that accounts for the empirical data, and allows for the surprising plasticity of meaning. Here is a first sketch of this model: 1. The prepositional category as a whole is organized around a focal relation, the ideal meaning. The whole set of uses of the preposition can then be subcategorized into use types (corresponding roughly to different senses), each manifesting the ideal meaning, but usually after some transformation. 2. A level of "geometric conceptualization" mediates between our naive representation of the physical world and the application of locative terms; prepositional meaning applies in effect to schematic images (geometric descriptions) matched onto the objects. 3. A number of contextual factors bear on the choice and interpretation of a locative expression: relevance, salience, tolerance, and typicality. There are pragmatic principles relating to these that explain many characteristics of the situation of use. One question will recur as this model takes shape: which aspects of the meaning of locative expressions are arbitrary, and which are motivated? A precise account of the respective scopes of arbitrariness and motivation will eventually emerge, and each will find its proper place. The article will conclude with a precise description of the difficulties involved in building computer models for spatial expressions and suggest consequences of the analysis proposed for semantic theory. 2. Insufficiencies of the "Simple-Relations Model" Before introducing my own proposal, I will present some of the shortcomings of the simple-relations model. 2.1. Figure and Ground Consider a simple locative expression: James is near the lemon tree.
274 ANNETTE HERSKOVITS I will call James the subject, and the lemon tree the object of the locative expression. Although many prepositions have converses (above/below, behind/in front of), and some are symmetric (near), the two possible assignments of the roles of subject and object are not equivalent (Talmy 1978b), a fact that the simple-relations model does not predict. Compare: The house is behind the church. The church is in front of the house. Prototypically, the speaker intends to inform the addressee of the location of the located object (or "Figure;" referred to in the subject position of the locative expression); and the addressee either knows the location of the reference object (or "Ground;" referred to in the object position of the locative expression) or could easily discover it. Because of this, typically the reference object will be large and fixed with respect to the located object. Pairs of objects which greatly violate these conditions are unacceptable, even if the choice of reference object is perfectly rational in the current context: * Mary's house is near the bicycle.3 * The cognac bottle is the one under the cap. are not acceptable, even if the addressee knows where I parked the bicycle, or if there were several similar bottles, and only the cognac bottle had a cap. One must say instead: Mary's house is near where the bicycle is. The cognac bottle is the one with the cap on it. In the first example, a fixed reference entity, i.e., a "place," is substituted for the typically very mobile bicycle. In the second, a periphrase is constructed using with (which is not basically locative, and does not demand a fixed and/or salient reference object). 2.2. Geometric Conceptualization The simple relations do not in any case apply to the objects themselves, but rather to parts of space, to geometric images matched onto the objects, to what I call geometric descriptions of the objects. And even with the same preposition, different geometric descriptions may be applicable. For instance, in the water in the vase, the water is actually in the "volume of containment" of the vase, corresponding to the cross-hatched region in Figure 1(a). But in the crack in the vase (Figure 1(b)), the crack is in the "nor-
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 275 mal volume'' of the vase, that is, the volume the vase would occupy if it were uncracked (to talk about a hole in some object is implicitly to refer to a "normal" volume, the volume before the hole was there). Finally, in the bird in the tree (Figure 1(c)), the bird is in the interior of the outline of a part of the tree. (a) (b) (c) Figure 1 In the simplest case, the applicable geometric description is the space occupied by the object, which I call the place, or canonical description, of the object. This is the case with the located object in all the examples above. Another important type of geometric description involves viewing or conceptualizing an object as a point, a line, a surface, or a strip — in other words as some simple geometric figure. So in: the village on the road to London the village is viewed as a point, and the road as a line. Or in: The top of the cloud cover is at 3000 feet. the top of the cloud cover is viewed as a horizontal plane, by some process of approximation. Another non-trivial geometric description is illustrated by the crack in the surface where the surface is conceptualized as some very thin three-dimensional lamina, not as an actual two-dimensional surface without thickness. 2.3, Synecdoche The geometric descriptions considered so far correspond to a level of geometric conceptualization that mediates between language and perceptual representations or mental images of scenes. But other kinds of trans-
276 ANNETTE HERSKOVITS fers to geometric descriptions arise from more classical types of metonymies, in particular synecdoche, where the whole stands for a part. Thus, in Figure 2, only the base of the house is above the base of the building; house and building are used as metonymic substitutes for their bases. the house above the building Figure 2 Or consider the use of under. In: the key under the rug the key is under the bottom surface of the rug. But in: One could see shiny white carp under the water. Ants dig burrows under the sand. carp and burrows are under the top surface of water and sand, respectively. In other words, the water and the sand stand for their top surface. 2.4. Divergence from the Simple Relations Another shortcoming of the simple-relations model is that the simple relations are often not true, not even of the geometric descriptions. So, for example, the pear in Figure 3(a) is in the bowl, although it is not in the interior of the bowl. In Figure 3(b), "Ulysses" is on the desk, although it is not contiguous with the desk, and is only indirectly supported by it. And in Figure 3(c), A is to the right of B although it is not in the cross-hatched the pear in the bowl "Ulysses" on the desk A is to the right of B (a) (b) (c) Figure 3
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 277 area, where it should be according to the most plausible simple-relation meaning for to the right (if told Put the chair to the right of the desk, I would be expected to place the chair so). Or take at to mean: At(X/Y) iff Coincides(X,Y) (where X and Y are points) as in: The center of the circle is at the intersection of the axes. But in: Mary is at the gate. Mary and the gate are not coincident; they are only very close together. So there is an admissible tolerance for deviation from the simple relations. In all these examples, the shift from the simple relation can be conceived of as gradual, measurable by some distance or angle, with perhaps a threshold beyond which it is inappropriate to use the preposition. But the shift cannot always be conceived of in this way. So in: the wrinkles on his forehead a wrinkle is not an object contiguous with and supported by the forehead, but it appears to be, and this resemblance does motivate the use of on. Thus one would not say: *the ruts on the road because the depth of the surface accident here prevents one from seeing a resemblance with a situation of contiguity and support. Where the surface accident is deep, in is required, expressing the relation of a gap to the "interrupted" object: the deep wrinkles in his forehead the ruts in the road Or consider: the apple on the branch the medal on a chain The relation centrally expressed here is one of attachment. But, most often, attachment co-occurs with contiguity and support; this would ordinarily be the case in the scenes described by these examples. That attachment is implied, however, is shown by the fact that the first example would be appropriate if the branch were on the ground and the apple supported by
278 ANNETTE HERSKOVITS the ground instead of the branch. We see that, although the meaning expressed is not the simple relation defining on, it is closely connected to that simple relation. So the simple-relation meaning is sometimes, one could say, "almost true," and sometimes shifted to another, though closely related meaning. This suggests a prototypical meaning structure for the prepositions. 2.5. Added Constraints There are sometimes added constraints; that is, the relation implied by a certain use is more specific than the simple relation. So, in: the man at the desk man and desk are very close to each other, but also we assume the man is working at his desk, or at least is in a position to do so. If he were doing somersaults on top of his desk, he would not be at his desk. No doubt, in this example as in some that follow, the deviation from the simple-relations model could be described as exceptional and the expressions themselves as idiomatic. Yet, we do perceive that the choice of preposition is motivated, that is, related to other uses of the same preposition. 2.6. Unexplained Context Dependencies In: The gas station is at the freeway. the gas station is at the intersection of the freeway with a crosspath to be determined by context. Ordinarily, speaker and addressee would be on that path. In Figure 4(a), one could say that A is to the right of X. But if an object B were introduced as in Figure 4(b), then, in many contexts, only B is to the B [3—s A is to the right of X B is to the right of X (a) (b) Figure 4
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 279 right of X (in some contexts, one might say A and B are to the right of X, but the building to the right of X would be identified as B, and The school is to the right of X would not leave the addressee wondering whether the school is A or B). So the use of the expression A is to the right of X depends upon the presence of objects in the environment other than A and X. Consider also: Jim is at the supermarket. The appropriateness of this sentence does not depend solely on Jim's position with respect to the supermarket, because if both speaker and addressee are in the supermarket the sentence is inappropriate — one would then be likely to say instead: Jim is in the supermarket. 2.7. Unexplained Restrictions Finally, there are restrictions that cannot be explained in terms of the simple-relations model. For example, one cannot say: * Draw a line in the blackboard^. * The table is in the lawn.4 although lawn and blackboard are areas in which something could in principle be included. But one can say: Draw a line in the marginl The table is in the garden. Or take Figure 5. One would not ordinarily say ""thepotato in the bowl * the bulb under the socket to describe these scenes, although, according to the simple-relations meaning of the prepositions, both phrases should be true. the potato under the bowl the bulb in the socket (a) (b) Figure 5
280 ANNETTE HERSKOVITS * * * In conclusion, the simple relations together with the compositional rule do not constitute an adequate model, as they fail to predict the uses of a great many expressions. 3. Meaning and Use of Spatial Prepositions Locative expressions show systematicity in two respects: 1. Each preposition has an ideal meaning, a kind of prototype, manifested in every use of the preposition; such an ideal meaning, or some transformation of it, applies to schematizations (geometric descriptions) of the objects. 2. There are a number of pragmatic "near-principles" involving salience, relevance, tolerance, and typicality that explain certain prepositional choices and metonymies. Outside these patterns of motivation, one finds idiosyncratic aspects of meaning which must somehow be specified in the lexicon as properties of the preposition. The set of use types will collect the necessary information. Detailed descriptions of these elements follow. 3.1. Ideal Meanings It is no coincidence that most linguists have proposed the same simple spatial relations as meanings of the prepositions. These simple relations indeed play an important role, but as something akin to prototypes, not as truth-conditional meanings. I prefer the term "ideal meaning" to "prototype," first because there is little agreement about the nature of prototypes; second, because at least for natural kinds, prototypes appear to be rather concrete: they correspond to some "best" instance and function as a central model to shape the category. But in the case of prepositions, what seems to shape the category is a very abstract relation. For instance, the topological relation of inclusion, whatever the dimensionality, size, or kind of the reference object, shapes the whole category of uses of in — not the best example of use of in (generally described by informants as "containment in a medium-size box"). The idea of an ideal meaning is that it is manifested in all uses of the preposition, although shifted or distorted in various ways. The whole prepositional category is structured by resemblance to an ideal relation; the
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 281 spatial relation fulfilled in any given use of the preposition is distinctly derived from that ideal relation (it may be identical to it). So, given a locative expression: "NPl Preposition NP2" if NPl refers to Ol and NP2 refers to 02, then a relation like the following will hold: [T(IM)](G1(01), G2(02)) Gl and G2 are the geometric descriptions applying respectively to Ol and 02, IM is the ideal meaning of the preposition, and T is a transformation applying to the ideal meaning (possibly the identity), yielding a new relation [T(IM)]. (When the ideal meaning is a three-place relation, we get an equivalent three-argument relation.) I call such a formula the geometric scene representation. I will give below several specific examples of such representations; they appropriately capture the meaning of an expression, evaluated in terms of truth in the world, but one should be aware that the 'T(IM)" and "Gl" and "G2" are frequently fuzzy. Transformations applying to the ideal meaning are of two kinds: (a) shifts to another relation — let us call these "sense shifts" (these will give rise to different "use types"); (b) gradual shifts such that the ideal meaning, or sometimes the ideal meaning tranformed by a sense shift, is "almost true;" these will be called "tolerance shifts." No general principle regulates sense shifts, but the derivation from the ideal meaning is usually quite transparent. Consider for instance: The chest of drawers is on the left wall. the pear on the branch the wrinkles on his forehead In the first example, the support condition is dropped from the ideal meaning of on, which involves contiguity and support. In the second example, on means attachment which most often co-occurs with contiguity and support. And in the third, the transformed ideal meaning is related to the ideal meaning by a complex process involving resemblance in appearance. Thus, in the case of the prototypical structure of prepositional categories, resemblance with the prototype involves more complex processes than the sharing of features.5 As an example of geometric scene representation, take The chest of
282 ANNETTE HERSKOVITS drawers is on the left wall. It can be represented as: Contiguous^ (Place(Chest),OuterSurface(Place(LeftWall))) where Contiguous is a predicate derived from the ideal meaning of on in two steps. First we get the predicate Contiguous by dropping the support condition from the conjunction of contiguity and support that defines on. Second, we allow some tolerance; the subscript M" in Contiguous indicates that contiguity need be only approximately true — if a tapestry were hanging between the wall and the chest, the chest would still be on the left wall. The composite predicates applying to the two objects Chest and LeftWall yield the geometric descriptions relevant to this example (in this and other examples, I use, for perspicuity, variable and predicate names that are full words or even phrases). In summary, the category denoted by a preposition is shaped by an ideal meaning, which is subjected to transformations: some represent conventional extensions of the range of use of the ideal meaning; others represent allowances for graded deviations from either the ideal meaning or from those conventional extensions. But the arguments of such ideal meanings, or shifted ideal meanings, are not simply the referents of the subject and object noun-phrases; they are the geometric descriptions. 3.2. Geometric Descriptions Geometric descriptions are parts of space associated with the objects by ordinary synecdoche, or by processes of geometric conceptualization. The functions that relate an object to the relevant geometric descriptions are geometric description functions. In some cases, the applicable geometric description is the space occupied by the object, its "place." I mean by "object" anything that could be referred to in a locative expression; this will include a whole range of spatial entities besides ordinary solid objects (e.g., parts, holes, geographical entities, etc.). The definition of a place will thus often require a method for representing vagueness, as with: He stopped at the curve in the road. The child hid in the folds of her skirt. Beyond the canonical geometric description function, I have come upon 21 elementary geometric description functions occurring in the uses of the various prepositions (Table 1).
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 283 Table 1 Elementary geometric description functions (1) (2) parts: - three-dimensional part - edge - base - oriented total outer surface - oriented free top surface - top surface - lowest surface - underside idealizations: - approximation to a point - approximation to a line - approximation to a surface - approximation to a strip (3) (4) (5) (6) good forms: - shape - outline adjacent volumes: - interior - volume/area associated with a vertex - lamina associated with a surface axes: - main axis - associated frame of reference projections: - projection on plane at infinity - projection on ground The list should cover all the more common occurrences. These functions map a certain part of space (a geometric construct) onto another; the canonical function, say Place, applies first to the spatial entity, mapping it onto the space it occupies at the time at which the locative expression applies. Several "elementary" functions can then apply in succession, their composition yielding the relevant geometric description. Several examples follow. Consider the path along the ocean Assume the ideal meaning of along to be: for two lines to be parallel and "very" close to each other or coincident. Note that the path could vary in height following the edge of a cliff. So the phrase indicates that the projection onto an horizontal plane at sea level of a line approximating the path is parallel and close to a line approximating the edge of the ocean:6 AlongA (HProj(LineApprox(Place(Path))), LineApprox(Edge(Place(Ocean)))) The subscript 4\4" on the ideal meaning allows for parallelism to be only approximate. HProj is the function projecting onto sea level. LineApprox is one of the functions that map a geometric construct onto some idealization
284 ANNETTE HERSKOVITS of it. Functions that map a geometric construct onto some associated "good form," obtained by filling out irregularities ("normal shape" and "outline") implement the Gestalt principle of "closure" or "good form." Thus: The bird is in the bush. can be represented as follows: Included(Part(Place(Bird)), Interior(Outline(VisiblePart(Place(Bush))))) Finally, consider: The North Star is to the left of the mountain peak. In this use of to the left, there is an implicit observer. A frame of reference is mapped onto that observer. Peak and star are viewed as points. To the left is true of the projections on the horizon (a line in the plane of view) of the apparent positions of star and peak in the plane of view (that is, the positions of their projections on the plane at infinity). ToTheLeft {HorizonProj{InfinProj{Place{NorthStar))), PtApprox(InfinProj(Place(Peak)), RefFrame{Place{Observer))) 3.3. Pragmatic Near-Principles Are there principles that can be used to predict the applicable geometric descriptions and the acceptable shifts from the ideal meaning? Or are these a matter of convention? One can ask a similar question about all the apparent irregularities described in Section 2. Can all or any of these phenomena be explained by inferences from world knowledge, conversational principles, or other pragmatic principles yet to be formulated? I believe that the place to look for principles, for regular behavior, is at some fundamental properties, principally salience, relevance, typicality, and tolerance. But although one can discern some systematicity, it is actually very hard to formulate rigorous principles and find rigorous inferential paths from them; I will thus speak of "near-principles." These near-principles are not predictive; they embody necessary but not sufficient conditions for the appropriateness of a certain use, and are formulated in terms of factors for which we lack a formal account. Yet these near-principles, somewhat in the manner of Grice's conversational maxims (1974), reveal some of the forces that shape language use.
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 285 (a) Salience Salience explains the direction of metonymic shifts. Here is a near- principle that motivates certain synecdoche: "One can use a noun which basically denotes a whole object to refer to a part of it that is typically salient." To illustrate, in: a waiting line at the counter a functionally salient part of the line, its head, should be substituted for the line itself; the phrase is appropriate only if the head of the line, and not any other part, is very close to the counter. In: the cat under the table table stands for the table top, a functionally salient part of the table. This near-principle has a corollary: "One can use a noun-phrase denoting an object to refer to a part of it that is typically visible." So, in: There is a rabbit under the bush. the rabbit is not under the whole bush, which would be under the roots. A second corollary is: "The geometric description applicable may be the base of the object (i.e., its area of contact with the ground plane)." This would account for: the house above the building in Figure 2. Also, whenever we describe an object as being "in an area," only the base of the object is in the area. For instance, one would describe Figure 6(a) by: The block is in the circle. although, in fact, only the lowest surface of the block is within the area of the circle. This has come to be so natural to us that it is hard to think that it involves any special process such as metonymy. Yet, if we consider the analogous situation in two dimensions, a similar metonymy does not pertain. One could not say, in Figure 6(b):
286 ANNETTE HERSKOVITS * The rectangle is in the line. (a) (b) Figure 6 One cannot say: * Mary was standing in the edge of the cliff. * The town is in the border. although one can express inclusion in a line, as in: a notch in the edge Typically, knowing where an object is located involves knowing on which part of the earth's surface that object sits; the location of the base is what matters. Two other near-principles authorize metonymies on the basis of salience: "The geometric description applicable may be the projection of the object on the plane at infinity." In other words, one can locate according to appearance rather than fact, as in: The North Star is to the left of the mountain peak. This near-principle relates to salience, since appearance is a salient aspect of one's experience of objects. Another near-principle is: "The geometric description applicable may be the projection of the object on the ground plane." The caption attached to Figure 7 is appropriate, although the painting is actually higher than the chair. Horizontal coordinates stand out in our experience of objects, because we are normally bound to the ground, and we move on it to approach objects. So one often ignores altitude when using prepositions like to the right, in front of, etc.
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 287 E5I —n— The painting is to the right of the chair Figure 7 (b) Relevance Relevance has to do with communicative goals, with what the speaker wishes to express or imply in the present context, while salience refers to the foregrounding of objects or object parts that arises in our interactions with and perception of our environment. Take for instance something in a container, like milk in a bowl. The body of milk is contiguous with, and supported by the bowl; yet one cannot use on, one must use in to describe its relation to the bowl: *the milk on the bowl To account for this restriction, one could posit a hierarchy among prepositions, such that when both in and on apply, the expression of containment — therefore in — has priority. This would reflect the perceptual salience of containment. But in Figure 8, on is appropriate, although there is containment. the oil in/on the pan the dust in/on the bowl (a) (b) Figure 8 Relevance helps explain these examples. One uses in or on according to whether containment or contact is most relevant. If one is most concerned with contact and its consequences, such as whether the surface of the pan is oiled, or the surface of the bowl is soiled, then one will use on. If one is more concerned with containment and its consequences — such as, the milk can be carried around, etc. — then one will use in. Usually, with milk in a bowl, there is no reason to focus on contact, which is why the use of on sounds very odd.
288 ANNETTE HERSKOVITS Another example of the role of relevance is provided by the bulb and socket in Figure 5(b). Both in and under are strictly speaking true in this case. But generally concern is with the objects' function — one wants to know if the bulb will work, and this can be inferred from in, but not from under, as in is associated with the bulb functioning. Therefore, under sounds odd, though not impossible — precisely in contexts where bulb function is not the topic, for instance: Place bulb and socket assembly in such a way that the bulb is under the socket. Similarly, the potato in Figure 5(a) is under, not in the bowl, because in is typically associated with the bowl functioning as a container. If it is important for the addressee to know that the bowl is not in a position to function as a container (if function is most relevant), then one must not use a preposition that will mislead the addressee into assuming that it does. (c) Tolerance There is a certain tolerance for deviation from the truth of the ideal meaning or the sense-shifted ideal meaning. I am concerned here with gradual deviations measurable in terms of an angle or a distance. For example, I want to ask: how far apart can two objects be so that one can still say that one is at the other? How close to the gate must Mary be, for the sentence Mary is at the gate to be true? How close to the right axis of object X (given some observer) must object A be, for A is directly to the right of X to be true? How does the required accuracy vary with different kinds of objects, say houses on a block, chess pieces on a board, or silverware on the table? First, there is a basic indeterminacy stemming from the nature of the objects. Consider: Put the napkin directly to the right of the plate! How precisely can one define the right axis of a plate? That axis would normally be parallel to the edge of the table, and that edge cannot be defined with more than a certain accuracy, given the surface texture of any physical object. Another source of indeterminacy is perception, and the degree to which two positions can be distinguished. For instance, the truth of: The white horse is at the finish line. depends on the means available to tell how close a horse is to a finish line. But primarily, tolerance depends on what the speaker deems to mat-
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 289 ter, on relevance. So, for instance, with chess-pieces on a board, if the game is what matters, then the grid of the board defines the relevant precision. If I say: The Knight is directly to the right of the Queen. the allowed distance from the right axis that preserves truth is defined by the size of the squares. Possibly, given (a) some physical description of the objects, (b) some account of perceivable differences in location, (c) a theory of relevance and (d) a representation of the speaker's purposes in discourse, one could, at least in some cases, predict what distance or angular deviation is tolerable in various contexts. But often, the tolerance reflects an accumulation of practices, of interactions with the objects, making prediction impossible. In our constant intercourse with the objects in our world we have integrated into our knowledge strategies that allow us to count or discount some fact according to context; what those strategies are is still very much a mystery. Yet tolerance is one direction in which the search for systematicity could proceed. One phenomenon related to tolerance does show some clear systematicity. If two objects, A and B, are placed in relation to a reference object in such a way that the ideal meaning of a preposition is truer of A than of B, then one can use that preposition to discriminate A from B — so that the locative phrase will be assumed true of A but not of B. I will give two examples of this "shifting contrast" near-principle. Consider again the use of to the right in Figures 4(a) and 4(b). One would say that A is to the right of X in 4(a), but if B is placed as in 4(b), then the ideal meaning applies better to B than to A, and, in many contexts, only B is to the right of X; to the right can be used to discriminate A from B. Consider also the pear and bowl of Figure 3(a). One would ordinarily say The pear is in the bowl, but a photographer might say: Move the pear so that it is in the bowl! The ideal meaning of in applies better to the apples than to the pear; one can use in to contrast their location, although most often that contrast will be ignored. (d) Typicality It is quite difficult to enunciate even near-principles in this area, but it
290 ANNETTE HERSKOVITS is easy to see how pervasive the influence of typicality is. Its role in motivating pragmatic inferences has been widely studied in artificial intelligence, but it has rarely been observed that typicality also motivates metonymies and certain linguistic choices. As an example of a pragmatic inference, consider The fountain is behind the city hall. One will assume that the fountain is close to the city hall, in fact that it is next to the city hall, meaning that no other salient architectural object is between the city hall and the fountain. This "nextness condition" is not part of the meaning of behind, because in: The treasure is buried 600 feet in a straight line behind you. no such proximity should be assumed. Still, we find it most useful to relate objects by means of behind when they are close together for several reasons: the direction defined by behind is otherwise difficult to evaluate, and it is generally more rational to choose a reference object close to the object one wishes to locate. So, most typically, objects will be close together. As a consequence, "nextness'' is the default case, the inference one should draw, unless one has evidence to the contrary. As for metonymies, the first salience near-principle made an explicit reference to typicality: metonymies are allowed on the basis of typical properties of the objects — which part is typically salient, typically visible, etc. The corollary stating that one can use a term referring to a whole object while strictly speaking locating its base also depends on typicality. Typically, objects sit somewhere on the ground, and we focus on the location of their base. We have also seen how typicality restricts the assignments of the roles of Figure and Ground. Note that the characteristics whose typical value we need to consider are not only physical characteristics. Thus, the typical use of, and intercourse with, an object play a role in allowing for metonymies. In fact, one must even consider typical characteristics of the act of locating as well: the fact that reference objects are usually close to the located object, the fact that the location of the base of an object is generally what matters, and the fact that certain kinds of objects are not normally used as references for other kinds.
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 291 3.4. Use Types The pragmatic near-principles could be part of a pragmatics that complements the simple-relations model. But this would still leave many phenomena unaccounted for. For instance, there is no way to predict that one cannot say: * the cat in the lawn or that it is possible to use in with the following kinds of geometric descriptions: the volume of containment, the normal volume, and the outline. There is no explanation for the fact that at can be used indexically, as in The gas station is at the freeway. Consider also: Jim is at his desk. The implication that Jim is using his desk might seem obtainable as a pragmatic inference: if Jim is at his desk, then he is very close to his desk, and he is very probably using it. But since Jim is byI {just next to) his desk. do not give rise to the same implication, there is an element of conventionality involved in this use of at. Similarly, the kinds of divergence from the ideal meaning that I have called sense-shifts are not deducible from the ideal meaning and pragmatic information about the objects. So all such conventional facts of use will have to be somehow specified in the lexicon, as characteristics of additional senses of the preposition or of idiomatic forms. The kind of lexical representation I suggest, the use type, will preserve the relation of the various uses of the preposition to its ideal meaning. It will also be a uniform representation, one that does not distinguish between senses and idioms. A use type will be a phrase pattern centered around a preposition, together with the interpretation, or meaning, associated with the pattern. Most often, the phrase pattern is simply a preposition with selection restrictions for the subject and object of the preposition, but sometimes it involves a specific word as object of the preposition. So there will be a use type: "Person 'at' Artifact" with the interpretation that the person is using the artifact, for examples such as Jim is at his desk and the man at the stove. But there will also be a use type:
292 ANNETTE HERSKOVITS "Spatial Entity 'at sea"' where the very word sea must occur as object of the preposition, for examples like The Titanic will never be at sea again. As a last example of use type, consider "Object 'on' a large vehicle" for phrases like the man on the bus. The vehicle must be large since *the man on the taxi is not acceptable. The interpretation of the use type specifies that the vehicle must be in the course of travel, since ?* The family lives on a converted bus is odd. I see use types as complex entities, more elaborate than senses as linguists usually conceive of them. One might specify in a use type a whole range of elements of meaning, anything that holds true for any phrase generated by the use type pattern, including conditions on the context and conditions that might ordinarily be called pragmatic. For example, consider Jean is at the park. Jean is in the park. Although they may denote identical positions with respect to the park, they would not be used in the same context. The preposition at is: (a) preferred if the speaker's knowledge is indirect; (b) unacceptable if speaker and hearer are in the park; (c) unacceptable if the speaker wishes to contrast inside and outside of the park. These conditions follow (loosely speaking) from taking a remote viewpoint on the situation. This viewpoint condition should be included in the relevant use type. The idea that one's linguistic competence consists of a repertory of syntactic patterns, each associated with its own partially idiosyncratic semantic interpretation and pragmatic information, with "regular" cases just patterns among others, runs counter to traditional linguistic theory. But recent work by Fillmore, Kay, and O'Connor (1983) on "minor" grammatical constructions (e.g., expressions involving let alone, as in He never made first lieutenant, let alone general) supports this view.7 4. Discussion Implicit in most computational approaches to the problem of locatives is the following belief: given a description of a scene, or an environment, in terms of the shape and location of the objects it contains (the canonical description), one could generate appropriate locative constructions, using
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 293 more or less complex spatial relations as meanings of the prepositions. But in fact, the spatial objects we relate are mental constructions, geometric conceptualizations. These geometric images are neither contained in, nor directly inferable from, the canonical description. Certainly, the selections, approximations and idealizations that give rise to the geometric descriptions are not strictly linguistic: they are inherent in our spatial knowledge, implicit in the way we move in the world among and toward objects, in the way we interact with the objects; they also certainly play a role in perception (Herskovits 1986). Geometric conceptualization introduces ambiguity (with the same preposition, several geometric descriptions are frequently possible), vagueness (e.g., in Mary is at the gate, the points in the conceptualized space corresponding to Mary and the gate are very imprecisely defined), and occasional context dependency (e.g., for what size of box and what locative purpose is it appropriate to "see" the box as a point, as required by The box is six feet from the wall). Then, various contextual factors play a role in determining an appropriate locative expression, requiring the representation and use of knowledge beyond that contained in the canonical description. For instance, salience interacts with typicality to determine the direction of metonymic shifts; to predict such shifts, one needs to know which part of an object is foregrounded for someone who uses the object, and more generally interacts with it, in a typical manner. But the corresponding metonymies do not apply in all contexts; for instance, it is not always the case that the top surface of a large body of water is salient, so that one can say It is under the water of an object that is in the water. Most people do not accept *Fish swim under the water, because in the associated mental image, the surface of the water is completely out of focus. In One could see shiny white carp under the water however, the phrase One could see brings that surface into focus. A formal characterization of such distinctions is of course very difficult. Typicality also allows for some pragmatic inferences. Drawing inferences based on typical rather than necessary conditions is one central problem of artificial intelligence. Research on this problem proceeds in two main directions: the development of default logics (e.g., Reiter 1980); and the search for models of memory such that appropriate inferences fall out of general processes of manipulation of memory structures (e.g., Quillian 1968; Minsky 1980; Schank 1980). Neither approach has so far yielded results approximating human performance. Note that it is not enough to represent the typical properties of object
294 ANNETTE HERSKOVITS categories, even including information about how we normally interact with the objects. We found several cases where typical characteristics of the act of locating played a role in determining certain metonymies, inferences, and linguistic choices. Relevance explains certain prepositional choices, and in conjunction with tolerance allows certain relational approximations. Relevance requires us to take into account what matters, to whom, and in what circumstances. Underlying any communication, there is a network of concerns shared by speaker and addressee, which they need not make explicit. This becomes apparent when one imagines contexts in which concerns are changed. This dependence on underlying concerns is generally overlooked in artificial intelligence work; only situations based on the most typical concerns are considered. It is, in fact, not clear how it is possible to make an explicit representation of all relevant concerns. The relational approximations allowed by the interplay of tolerance and relevance require a spatial knowledge beyond the canonical description of a scene (and one which is not explicit in the meaning of the prepositions). For example, with Mary is at the gate, defining the space within which Mary must be for the sentence to be true requires specific knowledge about the positions and distances involved in arriving at and waiting at a gate. As with the geometric conceptualizations of objects, such relational approximations are implicit, grounded in our physical involvement in the world. Although I have offered a general description of the information use types should contain, their actual representation presents many difficult problems. For instance, a use type should specify the way context-dependent parameters are computed, but such specification is in fact no simple matter. Take The Knight is to the right of the Queen: the relevant right axis would not normally be that of the Queen herself (except in Through the Looking Glass), but an axis defined by the chess board and one of the two players' positions; this follows from the fact that, in that context, the game is what matters. Here again relevance plays an important role, and one must appeal to the shared concerns of speaker and addressee. No simple predetermined function of the context can be used to assign a value to the variable right axis. Vagueness also makes the representation of use types problematic. What counts as a large vehicle in expressions of the type John is on the bus/ boat/subway/....? More than size is involved. For instance most speakers
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 295 accept I came on the jitney, though jitneys may be no bigger than a taxi; but the fact that the jitney is public transport like a bus reduces the size threshold. Vagueness often involves not simply a zone of uncertainty on a single scale, but depends on many interacting parameters. Finally, comprehension and production would involve respectively matching a situation with a given use type, and the converse, matching a use type with a given situation. The necessary matching processes would actually be very complex, beyond our present expertise.8 In summary, language is thoroughly context-dependent and pervaded with vagueness. We rely on implicit shared beliefs and concerns for communication to a degree which has not been sufficiently recognized in artificial intelligence. Pronouns and other indexicals represent only the most conspicuous cases; a closer look reveals a wealth of hidden context dependencies. To characterize precisely the functions of context involved is most difficult. In fact, it may be impossible to define a fixed set of parameters upon which the literal meaning of a given utterance must depend. Possibly, for any such set of parameters, one can come up with a context such that a parameter outside the set matters. This is the kind of argument that Searle makes in "Literal meaning" (1979). Another way to express the same thing is to say that every utterance takes its meaning from a background of assumptions and beliefs that cannot all be made explicit. Winograd and Flores (1985) express a similar idea. As for vagueness, it remains without satisfactory formal treatment. Zadeh's "fuzzy logic" (1974) is not the answer; the problem is not simply that of allowing degrees of truth, but of understanding how we integrate the interplay of many vaguely-defined parameters into some definite utterance. It is not clear that this can be solved by making explicit the set of relevant parameters, and attaching an appropriate weight to each, because, in effect, it is not clear that a list of such parameters and weights can be specified. Naturally, such problems spell trouble for logical semantics as well. From a linguistic point of view, the central question is that of deriving the truth-conditions of a locative sentence, given lexical representations of the component morphemes, a syntactic description of the sentence, and the context. I have shown that a specification of those truth-conditions would consist of a relational formula, with a shifted ideal meaning applying to geometric descriptions — not to the referents of the subject and object of the preposition themselves. Conventional facts of use (captured by the use
296 ANNETTE HERSKOVITS types) as well as pragmatic principles determine these shifts and reference transfers. Clearly, if the principle of compositionality (i.e., the meaning of an expression is a function of the meaning of its parts and of their mode of combination) applies here, it will not do so in a direct and simple way, as it does with the simple-relations model. Two elements, the ideal meaning and a use type, are candidates for entering into the combination process; but clearly only the use type fully defines the contribution of the preposition to the truth-conditions. Plausibly then, a theory could be devised that maintained compositionality and accounted formally for the way a use type combines with the meanings of the other lexical items in the sentence so as to produce those truth-conditions. Such a theory would have to cope with: (a) pragmatically controlled shifts of relational meaning and reference transfers — concealed indexical- ity is surprisingly widespread, and, moreover, it is difficult, maybe even fundamentally impossible, to fully specify the parameters controlling the values that the indexical variables take in particular contexts; (b) the complex interplay of motivation and convention — in other words, the lack of a clear division of spatial expressions into idiomatic and regular ones; (c) vagueness. No present theory deals with these basic facts of language. A somewhat impressionistic, but revealing, view of language would be as follows. Word meanings are defined in an ideal world — a world of lines, points, surfaces, definite relations of inclusion, contact, intersection, etc. But our need to describe, to communicate facts efficiently about a complex and imperfect world, forces us to bend and stretch the concepts. This bending and stretching, however, is not to be done at the speaker's whim; while it may not be all the result of "linguistic rules" proper, it is certainly constrained to preserve mutual comprehension, so our references to objects and facts are, usually, successful. Only with such a view of language can we make sense of all the apparent irregularities a careful look at any semantic domain reveals. We can then proceed to rigorously describe the ways in which ideal conceptual objects are distorted by their immersion in a linguistic and situational context. Notes This paper is a condensed version of a paper published in Cognitive Science, vol. 9. The original version was written while the author was a post-doctoral fellow in the Sloan Foundation Cognitive Science Program at the University of California at Berkeley.
SPATIAL EXPRESSIONS AND THE PLASTICITY OF MEANING 297 Thanks go to Barbara Grosz, Paul Kay, Ageliki Nicolopoulou, Martha Palmer, Eleanor Rosch, Ivan Sag, Elizabeth Traugott, Robert Wilensky, and Terry Winograd who helped at various stages of the research and writing. Thanks also go to Kris Holmes and Rene Yung for the drawings. 1. The representations that follow are adapted from Cooper (1968) and Miller and Johnson- Laird (1976). 2. The simple-relations model underlies most work on locatives within the computational paradigm (Cooper 1968, Winograd 1972, Miller & Johnson-Laird 1976, Boggess 1979, Waltz 1981). Boggess takes into account an additional level of complexity, namely the role that a few perceptual features of the objects play in determining the interpretation of locative expressions, but, as we will see, this is insufficient. In linguistics, early work on locatives came mostly from structuralists (Leech 1969, Bennett 1975). Although they do not require meaning to express the sentence truth-conditions, but only to account for intralinguistic relationships, their representations resemble the propositions just described. Implicit in Fillmore (1971a) and Talmy (1975) is a view of spatial prepositions as expressing simple relations between physical objects. The work of Brugman (1981), Lindner (1981), Talmy (1983), and Vandeloise (1984), is more concordant with the views expressed here, but it gives neither a precise account of the relation of locative expressions to the real world of physical objects, nor of the respective roles of motivation and convention. 3. The symbol "*"* signifies syntactic, semantic, or pragmatic ill-formedness. This example is adapted from Talmy (1978b). 4. This example is adapted from Fillmore (1971 a). 5. In effect, it is unlikely that features are generally adequate for expressing resemblance. Goldmcier (1972) shows that it is not the case in perception; it would be unlikely to be the case with conceptual structure. 6. In all such representations, I will use the name of the preposition itself as the ideal meaning of the preposition, except for at, on, and in, for which the following ideal meanings are proposed: in: spatial inclusion on: contiguity and support (for three dimensional objects) contiguity (for one- or two-dimensional reference objects) at: coincidence of two points 7. Fillmore, Kay, and O'Connor (1983) find that a large part of a speaker's competence consists of such "peripheral" constructions; given that the machinery required to handle these would be powerful enough to handle the regular structures, they suggest a uniform treatment for both. 8. For a detailed description of what building computer models of comprehension and production would imply, see Herskovits (1986).
Contrasting Prepositional Categories: English and Italian John R. Taylor University of the Witwatersrand 0. Introduction This paper summarises the results of a contrastive study of a small set of prepositions in English and Italian.1 The prepositions examined are those which — alongside other uses — designate the higher location of one entity with respect to another, i.e. English on, over and above, and their approximate Italian equivalents su, sopra and al di sopra (di). The study was undertaken against the background of recent treatments of prepositions within the Cognitive Grammar paradigm (e.g. Hawkins 1984, Vandeloise 1984), and benefited greatly from Brugman's pioneering monograph on over (1981). It was hypothesised at the outset that the only partial overlap between the uses of the prepositions in English and Italian, far from being unsystematic, is in fact the consequence of different categorisations of spatial relationships in the two languages. This working hypothesis contrasts strikingly with the traditional view of prepositions. It is well known that prepositional usage varies enormously from language to language, even between historically closely related languages, to the never-ending frustration of foreign language learners and their teachers. Prepositions thus present contrastive analysis — especially a pedagogically oriented contrastive analysis — with a particularly urgent challenge. This challenge does not appear to have been taken up. If we turn to an authoritative survey of contrastive analysis, such as Carl James (1980), we find only the most sporadic reference to prepositions. The reason for this lack of interest may be inferred from the treatment given to prepositions in most pedagogical and reference grammars. In these works, prepositional usage is generally presented as something essentially arbitrary and idiomatic, something, in other words, that "just has to be learnt".2 To
300 JOHN R. TAYLOR quote from one reference grammar of English: "There are not many rules to help you choose correctly (...) so you have to learn each expression separately" (Swan 1980: 483). With few exceptions, mainstream theoretical linguistics, with its overriding concern with syntax, has not paid much attention to prepositions, either, and no doubt for the same reason: prepositional usage, along with other "idiosyncratic''' phenomena, have been relegated to the lexicon, where they cease to be of interest to the syntactician. One consequence of Brugman's analysis of over, and of similar works, such as Lindner's (1981) study of out and up, is that it is now more difficult to dismiss prepositional usage as arbitrary and unsystematic. The fact remains, however, that prepositional usage does vary, in an at first sight unpredictable manner, from language to language. If we wish to argue that prepositional usage is systematic, we will find it salutary to approach the issue cross-linguistically. 1. Prepositions and Polysemy The problem with prepositions, as already hinted, is that a particular preposition is likely to be used in a very wide range of different contexts, and in a variety of different senses; furthermore, its distribution is unlikely to match the distribution of any one preposition (or other language form) in another language. Prepositions, in other words, are highly polysemous, and their polysemy is language specific. Polysemy, as Langacker (this volume) notes, must be regarded as the normal state of affairs in lexical semantics. Langacker also observes that the more frequently used a lexical item, the more senses it is likely to possess. Prepositions, being items of extremely high frequency, are thus likely to display an unusually high degree of polysemy. Any analysis of prepositional usage therefore hinges on how one is to deal with polysemy. Polysemy is the association of distinct, though in some way related meanings to a single phonological form. The point at issue is: how are these various meanings related? Two main approaches can be distinguished. The first is the "core sense" approach (see Hawkins 1984: 180ff. for a detailed discussion of this approach, and reasons for rejecting it). Faced with an array of the different senses of a particular lexical item, the analyst attempts to abstract from them some semantic content which is common to each. Very often, this semantic core has been represented in terms of a matrix of putatively universal semantic primitives (i.e. semantic features which are
CONTRASTING PREPOSITIONAL CATEGORIES 301 not further analysable into other features). These semantic features specify the necessary and sufficient conditions for the appropriate use of the lexical item. Faced with such highly polysemous items as prepositions, the core sense approach breaks down, for failure to meet one (or both) of two conditions. Firstly, the approach demands that the core meaning of an item is sufficiently general such that all the uses of the item are accounted for; at the same time, the core meaning must be sufficiently specific so as to uniquely distinguish the item from all others. To give an example. After examining several occurrences of over one might propose the core meaning "at a higher location than, but not in contact with", i.e. [ +vertical], [—contact].3 But neither verticality nor contact are at issue in a sentence like He lives over the river. This, and many other uses, must now be listed as "idiomatic", "arbitrary", "unsystematic", etc. (cf. the pedagogical treatment of prepositions mentioned in the Introduction). The other problem is that the same core sense, i.e. [+vertical] , [—contact] would presumably be required for above. How then can one explain the only partial overlap in the distribution of the two words? Difficulties are magnified as one increases the range of languages under discussion. Sopra is distributed differently from either over or above, but, once again, the same core sense seems to be indicated. The alternative to the core sense approach is the "meaning chain" analysis developed by Brugman (1981) and Lakoff (1982a: 72ff). (See also Hawkins, this volume.) For each preposition, we recognise a central, or prototypical sense. The prototypical sense, rather than being highly general, may well profile a very specific configuration.4 Polysemy comes about when the preposition is used in a sense which is closely related to, but distinct from, the prototypical instance. For example: a condition which is essential to the prototype might not be met; a feature which is optional to the prototype now assumes central importance, or vice versa; or some additional feature might be required. By the same process, this derived meaning may in turn give rise to a further extension, and so on. The various senses of the word thus radiate out from the central prototype, like the spokes of a wheel. Senses at the periphery might well have little in common, either with each other, or with the central sense; they are merely related by virtue of the intervening members of the meaning chain. Figure 1 illustrates a meaning chain analysis, and contrasts it with a core sense approach:
302 JOHN R. TAYLOR (a) core sense (b) ® © <§> © © © Q) Q> ® ® ® <z)—©—® ® © Figure 1: The "core sense" and the "meaning chain" approaches to polysemy. In (a) the core sense fully sanctions each attested use of an item, (b) reproduces Lakoff s (1982a: 82) representation of the structure of over. The prototypical senses are enclosed in the box. In this instance, one can expect meaning 12 to have virtually nothing in common with meaning 16. The related senses of a lexical item together constitute what Lakoff calls a "natural category" (Lakoff 1982a). Contrasting the distributions of roughly equivalent lexical items in different languages (or, for that matter, in the same language) thus involves contrasting the internal structures and extensions of categories. Categories might fail to overlap because of
CONTRASTING PREPOSITIONAL CATEGORIES 303 (i) differences in the central, prototypical sense; and (ii) differences in nature and extent of the meaning chains which radiate out from the central sense. We shall see in section 3 that these factors can account in a highly systematic manner for the non-equivalence of the prepositions under discussion. This fact, in turn, provides additional empirical support for the validity of a prototype approach to polysemy. 2. The Nature of Prepositional Meaning It is a truism that one can only compare like with like; any attempt to compare two entities presupposes a tertium comparationis, a constant against which differences and similarities can be identified. A contrastive language study thus inevitably raises the question of universals. Generativists propose that the lexical items of each language select from a universal, presumably innate set of semantic primitives (cf. Dirven & Taylor, this volume); these form a natural point of departure for comparing semantic structures in different languages. Cognitive Grammar takes no such strong position on universality. Linguistic expressions associate a phonological form with a semantic structure, i.e. with a conventionalised conceptualisation. There is every reason to expect that different languages will embody different, perhaps even highly idiosyncratic construals of certain conceived states of affairs (Langacker 1983: 1/56). But if semantic structures themselves are likely to be language specific, the conceptualisation processes which give rise to these structures might well be universal. Of particular relevance in this connection is the distinction between figure and ground, and the notion of a relational profile. The confines of this paper require us to restrict our attention largely to the spatial uses of the prepositions. In their spatial use, prepositions serve to locate one entity (the Trajector, or TR) with reference to another entity (the Landmark, or LM) within three-dimensional space. Prepositions, in other words, profile a relation between two entities against a base of physical space. Of course, prepositions are not the only linguistic units which can perform this function. Spatial relations can also be profiled by means of adverbs, case markers, verbs, etc. Prepositions in English and Italian specify the TR-LM relation in various ways. In section 3 we shall refer to various aspects of the TR-LM relation that are relevant for an adequate description of the prepositions under
304 JOHN R. TAYLOR discussion. Let us here briefly survey some of the kinds of things that may need to be taken into account. The following points are to be taken as a sort of heuristic check list; in describing any particular prepositional sense, not all of these parameters need be relevant: (i) the nature of the LM, i.e. its shape, size, dimensionality, texture, etc. Very important for some English and Italian prepositions is whether the LM is construed as a bounded area or enclosure, or as an unbounded line of surface (ii) the nature of the TR, also in comparison with the LM. For example, it might be a requirement that the TR is smaller, or larger than the LM (iii) contact. Is the TR in contact with the LM? If not, the distance between them might be important (iv) the orientation of the TR with respect ot the LM. Is the TR in the horizontal plane of the LM? Orientation can also refer to other TR-LM configurations, such as total or partial enclosure, interlocation, etc. (v) static vs. dynamic relation. If the TR-LM relation is a static one (i.e. it is instantiated at some point in time), the preposition designates the Place of the TR. The relation may also be dynamic. A dynamic relation is instantiated over some stretch of time, i.e. the TR moves with respect to the LM. Three kinds of dynamic relation can be distinguished: Goal, Source and Path. These are, in fact, elaborations of the more basic Place relation. Goal focuses on a Place which is the end-point of the TR's movement; Source focuses on the initial point; while Path specifies a Place which defines the trajectory traced by the TR. Path prepositions are likely to be particularly numerous, since the prepositions might specify either all or some part of the trajectory, or might specify a trajectory of a particular shape, e.g. circular, rectilinear, arc-like etc. (vi) the role of an observer. Some prepositions might be strongly deictic. Others might make reference to perceptually prominent aspects of an entity, or require that both TR and LM be in the perceptual field of an imaginary observer. Also, it should be emphasised that the various aspects listed under (i) to (v) do not necessarily have to do with objectively verifiable features of a situation, but rather with how that situation is construed. For
CONTRASTING PREPOSITIONAL CATEGORIES 305 example, one and the same entity might be construed either as a surface or as an enclosure. An example is the contrast in English between on the field and in the field. Finally, it must be pointed out that no check list for the components of prepositional meaning is likely to be exhaustive. Given the conventional, language-specific nature of conceptualisations, one must always keep open the possibility that a particular usage may profile some highly idiosyncratic aspect of a TR-LM relation. 3. The Prepositions In the limited space available, it would be impossible to undertake a comprehensive account of the three English and the three Italian prepositions. I will therefore restrict my attention to some of their major senses, concentrating on the similarities and differences between the languages, and attempting to account for the latter in terms of different structurings of prepositional categories. Let us begin with English on. (1) gives a sample of the range of senses of this preposition: (1) (a) The book is on the table (b) The picture hangs on the wall (c) There's a fly on the ceiling (d) Don't write on the wall (e) Sign your name on the dotted line (f) There's a snail on the bottom of the bucket (g) He put shoes on his feet (h) She has a ring on her finger On may designate either Place, e.g. (a), or Goal, e.g. (g). The LM may be a surface (a), a line (e), or even a point; container LM's, however, are ruled out. Furthermore, the TR makes contact with the LM. Having said this, we have exhausted what is common to the eight uses illustrated in (1). Have we thus identified the "core sense" of on? The answer must be no, since we have failed to distinguish on from other prepositions which designate a Place/Goal relation of contact with a line or surface, such as underneath, against, etc. (1), in fact, provides compelling evidence for a prototype approach. In addition to contact, on can profile a number of other aspects of the TR-LM relation. These include:
306 JOHN R. TAYLOR (i) The TR is spatially superior to the LM; i.e. the LM is the upper surface/edge of the profiled entity (ii) The LM is the perceptually most salient surface (which may or may not be the upper surface) of the profiled entity (iii) The LM supports the TR (iv) The TR is smaller than the LM, such that the LM can be perceived as the background to the TR. Each of these features is present in some, but not all uses of on (Table 1). No particular combination of features is necessary for the appropriate use of on, although certain combinations sanction specific instances. In other words, we need to regard the sentences in (1) as instantiations of eight different, though closely related senses. Note that all the features listed above are present in (a). There is thus good reason to regard this sentence as an example of prototypical on. (Note also the ambiguity of (f). The bottom of a bucket has two surfaces, an "inside" surface and an "outside" one. Which is the more prominent? It depends on whether one views the bucket from the inside or the outside. The snail could be in contact with either.) Italian su may be used in a roughly equivalent range of senses. (Note that (T) (f) is ambiguous in the same way as the English sentence.) There is, though, one instance of non-equivalence: (h). (1') (a) // libro sta sul tavolo (b) // quadro sta sulla parete (c) C'e una mosca sul soffitto (d) Non scrivere sul muro (e) Metta la firma sulla linea punteggiata (f) Ce una lumaca sul fondo del secchio (g) Si e messo le scarpe sui piedi (h) Porta un anello *sul dito lal dito How can we account for the acceptability of su in (g), but not in (h)? The relevant difference seems to be that whereas a shoe can cover a foot (g), a ring does not cover a finger. Notice that the LM's in (g) and (h) are a little different from the LM's in the first six sentences. In the latter, the LM is one specific surface or edge of the profiled entity, while in (g) and (h) the LM is the total surface. In other words, su displays disjunctive sets of senses. In one set, the TR is in contact with the perceptually most prominent (in the prototypical case, the upper) surface/edge of the profiled entity; in this case, the TR is usually smaller than the LM. In the other
CONTRASTING PREPOSITIONAL CATEGORIES 307 book on the table picture on the wall fly on the ceiling write on the wall sign on the line snail on the bottom of the bucket shoes on his feet ring on her finger TR smaller than LM, LM is background to TR + + + + - + - + LM supports TR + + 7 - - + + + LM is perceptually most prominent surface/edge of entity + + + 7 + + n.a. n.a. TR superior to LM + - - - + +/- - - Contact between TR and LM + + + + + + + + Table 1: Some uses of on and their meaning components, (n.a. means that the component is not applicable to the sense.)
308 JOHN R. TAYLOR sense, the LM is the total surface, and the TR must be at least as large as the LM. We shall see later that the sense "TR covers LM" represents one important meaning extension of su. On, as we have seen, requires that there is contact between the TR and LM; also, in the prototypical instance, the TR is superior to the LM. If the TR is superior to the LM, but there is no contact, English uses over or above. Italian sopra and al di sopra are similar: (?) The lamp hangs over the table/above the table La lampada pende sopra il tavololal di sopra del tavolo We have seen that Italian su is roughly equivalent to English on. Are we now entitled to claim that sopra/al di sopra are equivalent to over/above? Not at all. Su (unlike on) may be used even in the absence of contact between TR and LM; the TR, though, must be superior to the LM: (3) La lampada pende sul tavolo The lamp hangs *on the table Conversely, sopra (but not al di sopra) may be used even if there is contact; in this case, again, the TR must be superior to the LM: (4) (a) Appoggio la mano sopra il mio braccio He put his hand on Cover) my arm (b) Si costrui la casa mattone sopra mattone He built the house brick on Cover) brick An immediate problem now is how to distinguish between su and sopra. In point of fact, the two words are in very many cases interchangeable, as noted in Zingarelli's Vocabolario delta lingua italiana (1970) under the entry for su. Both may frequently be used for a TR which is superior to a LM-, whether or not there is contact between them. (If the TR is not superior to the LM, as in many of the examples in (1), then sopra may not be used.) The two sentences just cited in (4), for example, are acceptable with su: (4') (a) Appoggio la mano sul mio braccio (b) Si costrui la casa mattone su mattone Some more examples of the interchangeability of the two words: (5) (a) Mise il libro sulla scrivanialsopra la scrivania He put the book on Cover) the writing desk (b) Siediti su quella sedialsopra quella sedia Sit on Cover) that chair
CONTRASTING PREPOSITIONAL CATEGORIES 309 (c) Giurd sul Vangelolsopra il Vangelo He swore on Cover) the Bible The interchangeability of linguistic units is a very common phenomenon; we shall have occasion to mention other instances in the course of this discussion. Interchangeability, however, does not necessarily mean that the units are perfect synonyms. I would like to suggest that su and sopra do differ with regard to the relations they profile. The profiled relations, however, are such that they may well be applicable to the same real-world configurations of entities. A similar case is discussed by Geeraerts (this volume) in connection with the apparent interchangeability of vernielen and vernieti- gen in nineteenth-century Dutch. Geeraerts argues that the prototypical senses of the two verbs, while different, likewise permit a rather similar distribution of the two items. One hint of a difference in meaning between su and sopra is given by the different frequencies of occurrence of the two words. Su is one of the highest frequency words in Italian. In a corpus of 500 000 words (Bortolini et al. 1972) its rank position was 28, and the number of occurrences 2159. Sopra had a rank position of 428, and the number of occurrences was only 111.5 This suggests that sopra is profiling some very specific aspect of a scene, while with su the focus is more blurred. With sopra, it is the superior location of the TR which dominates; this may (but need not) involve lack of contact with the LM. What about sul We have already drawn attention to disjunction in the senses of su. One set of senses, shown in (l')(a) - (0> requires contact with the LM; the TR, however, need not be superior to the LM, although in the prototypical instance it is. The second set profiles a covering relation; here, verticality is not at issue. Now, we have a third disjunctive set. The superior location of the TR now becomes the vital feature, and the requirement of contact with the LM is dropped. In this third set of senses, another meaning component seems to be coming to the fore, one already foreshadowed in a sentence like (l')(b). There, the LM could be perceived as the background to the TR. In the absence of contact between the TR and the LM, an important condition for the use of su is the "perceptual unity" of the TR and the LM. Both TR and LM must be in the field of vision of a potential observer (this point has been made by Parisi & Castel- franchi 1970). As an illustration, consider a phrase discussed by Cristari (1971): una casa sul mare "a house on the sea". The referent of una casa sul mare need not be in contact with the sea, or even be at sea level. (The expression thus has a wider reference than the English gloss.6) What is
310 JOHN R. TAYLOR important, according to Cristari, is that a potential observer (e.g. someone approaching the land from the sea) can perceive both the house and the sea, and the relation between them. Una casa sul mare thus contrasts with una casa al mare "a house by the sea1', where the perceptual unity of house and sea is not a requirement. The house, for example, might not be visible from the sea because of an intervening line of trees or sand dunes. Nonetheless, una casa sul mare could actually be further from the sea, both on the horizontal and vertical axes, than una casa al mare. Una casa sul mare also contrasts with una casa sopra il mare. This latter expression profiles the superior location of the house to sea; an appropriate English gloss might be "a house overlooking the sea". It is the perceptual unity of TR and LM that accounts for the acceptability of (3): (3) La lampada pende sul tavolo "The lamp hangs over the table" Here, lamp and table are in the field of vision of an observer. In the unlikely event of one wishing to describe a situation in which the table is on the first floor of a building, and the lamp was hanging vertically upwards from it, but on the second floor of the building, then one could not say that the lamp was hanging sul tavolo. (Neither could one say, in English, that the lamp was hanging over the table.)This relationship could, however, be expressed by sopra, or, better, by al di sopra. This does not mean that with reference to the normal alignment of lamps and tables, su in (3) is not also replaceable by sopra and al di sopra. The fact that sopra and al di sopra highlight the higher vertical location of the TR does, however, exclude their use in (l')(b): // quadro sta sulla parete 'The picture hangs on the wall" — unless, of course, one wished to describe the odd situation of a picture hanging above, but not supported by, a wall. In spite, then, of their interchangeability in some contexts, the different (though sometimes compatible) relations profiled by su and sopra can give rise to very clear semantic contrasts. Let us consider a more complex example. (6)(a) and (b) could (but need not) encode the same intention, i.e. the desire to reach the top of the tower: (6) (a) Voglio salire sulla torre "I want to climb on (i.e. up) the tower" (b) Voglio salire sopra la torre "I want to climb onto (the top of) the tower"
CONTRASTING PREPOSITIONAL CATEGORIES 311 Su, in (a), is ambiguous. It could be a preposition of Place, i.e. it designates the location of the climbing. Salire profiles an upward movement, and nothing excludes that this movement will continue till the top of the tower is reached. Alternatively, su designates the Goal of the climbing, i.e. the top-most point of the tower. In (b), sopra is likewise a Goal preposition, and, perhaps even more strongly than su brings the upper surface of the tower into profile. There is a similar difference in emphasis if a Source relation is expressed: (7) (a) Si gettd dalla torre "He threw himself from the tower" (b) Si gettd da sopra la torre "He threw himself from the top of the tower" The expression salire sulla torre reminds us that su can also be used in Italian as an adverb, with a meaning roughly equivalent to the English up. (Diachronically, the adverb is older than the preposition.) While this paper is concerned principally with prepositions, it is of interest to note that sopra also occurs as an adverb, and again the two words are sometimes interchangeable: (8) Vieni su! Vieni sopra! "Come up!" (e.g. upstairs) Su, though, is not always replaceable by sopra: (9) (a) / bambini vengono su/ *sopra bene "The children are growing up well" (b) I prezzi vanno su/ * sopra "Prices are going up" The reason seems to be that in (8) there is a LM located at a superior point on an implied three-dimensional entity (e.g. a staircase) of finite vertical extent, while in (9) the LM is simply the vertical dimension as such, either spatial as in (a), or metaphorical as in (b). As an adverb, sopra seems to require that there are concrete entities in its profile, entities which have upper surfaces, parts which are higher than others, etc. How does al di sopra fit into the Italian prepositional system? As hinted in the discussion of (3), al di sopra profiles a relation of superiority, with no contact between the TR and LM. One could go further and state that al di sopra emphasises the separateness of the TR from the LM. This aspect emerges very clearly in metaphorical uses. English, in comparable instances, has above (not over) or beyond:
312 JOHN R. TAYLOR (10) (a) La discussione era al di sopra della mia comprensione "The discussion was above my comprehension", i.e. "The discussion was above me/beyond me" (b) Vive al di sopra dei suoi mezzi "He lives above/beyond his means" (c) Un cittadino al di sopra di ogni sospetto "A citizen above/beyond all suspicion" Let us summarise so far. We have been able to characterise some important senses (or clusters of senses) for on, su, sopra and al di sopra. Leaving out much vitally important detail, we can say that (a) on requires contact between TR and LM. In the prototypical instance, the LM is the upper surface of a profiled entity (b) su has three major groups of senses: (i) there is contact between the TR and the LM, similar to on\ (ii) the TR is superior to the LM, with or without contact; and (iii) the TR covers the LM (c) sopra very strongly highlights the superior location of the TR, with or without contact to the LM (d) al di sopra emphasises the separation of the TR from the LM; the TR must be superior to the LM. Note that the relations profiled by su, sopra and al di sopra could well be compatible with certain real-world configurations of entities, resulting in the apparent interchangeability of the prepositions in some contexts. Wc now turn to the two English prepositions that we have not yet adequately discussed: above and over. In some cases, these prepositions, too, appear interchangeable, as in the English sentences in (2). Another example: (11) (a) He lives over me (b) He lives above me Over and above, at least in their prototypical senses, profile the higher location of the TR with respect to the LM, as well as the absence of contact between the TR and the LM. Some native speaker informants detect no difference in meaning between the sentences in (11) and the English sentences in (2). In this connection, it is of interest to note that Hawkins (1984: 367, 380) gives identical schemata for above and for one of the senses of over. I would like to suggest, however, that there is a difference — or at least that there is a potential difference. Whereas above merely designates that the TR is at a higher level than the LM, over can carry a number of more
CONTRASTING PREPOSITIONAL CATEGORIES 313 specific meaning nuances. Some senses of over have exploited and magnified these nuances, resulting in a weakening, even in the obliteration of the components of the central meaning "superior to, not in contact with". Above, on the other hand, comes to profile not just the absence of contact between a higher TR and the LM, but the separateness of the TR form the LM, as in the English glosses in (10). In other words, as we move away from the central meanings of the prepositions, over and above diverge more and more. Over is perhaps the most complex preposition in English (Lakoff 1982a: 72). The various meaning chains which radiate out from its central meaning have been intensively studied by Brugman (1981). Here, we need give a selective account only, concentrating on how our analysis differs from Brugman's, and making the necessary comparisons with Italian. What are the meaning nuances present in (ll)(a) but not in (ll)(b)? He lives over me seems to imply that the other person is not just living on a higher floor than me, but that he lives directly vertical to me (cf. Dirven 1981: 116). Also, the distance between us is not likely to be very great. Note the unacceptability of over in (12): (12) He lives ten floors above (*over) me (There is though an alternative explanation for the occurrence of above in (12). See below.) I am more likely to be disturbed by noise from people living over me than from people living above me. In other words, with over the TR is capable of "affecting" the LM. A similar kind of difference is present in the English sentences in (2). The lamp hangs above the table merely locates the lamp with respect to the table. Over, in the same context, implies in addition a functional relation. By virtue of its position, the lamp is also able to "affect" the table, i.e. it can illuminate it. Imagine a writing table fitted with an adjustable reading lamp. I want the position of the lamp to be adjusted so that the working surface of the table is fully illuminated, i.e. is fully under the "influence" of the lamp. In such circumstances, I am likely to say sentence (13), using over, but not above: (13) Move the lamp down over (*above) the table Italian, in equivalent circumstances, permits both su and sopra, but not al di sopra (13'). The exclusion of above and al di sopra is in keeping with the earlier observation — see the discussion of (10) — that these prepositions profile the separateness of the TR and the LM, i.e. the absence of any kind of "influence" between the two entities.
314 JOHN R. TAYLOR (13') Abbassa la lampada sul tavololsopra il tavolol *al di sopra del tavolo (Needless to say, the acceptability judgements given for (13) only apply to the sentences in the described context. If above the tablelal di sopra del tavolo are understood as reduced relative clauses, e.g. Lower the lamp (which is) above the table, then above and al di sopra are acceptable.) It is the meaning nuance "TR affects LM" that permits over to be used in the extended sense "TR covers LM"; covering is but a special instance of "affecting". This is a meaning extension which has important implications for a theory of polysemy, since a covering relation involves the suspension of a meaning component essential to the central sense, i.e. lack of contact between TR and LM. A covering relation may also involve the suspension of the other component of the prototype, i.e. the superior location of the LM. In other words, it is no longer possible to specify the necessary and sufficient conditions for the use, both of the central sense, and of the extended sense. In regarding "TR covers LM" as an extension of "TR affects LM", our analysis differs from Brugman's. Lakoffs summary of Brugman — see Figure 1 in Lakoff (1982a: 82) — presents (15) as an extension of (14), i.e. the two-dimensional TR in (14) becomes three-dimensional in (15). But the required extension of (14) is surely not (15), but something like (16). We still have to explain how an essential component of the prototype — i.e. [-contact] — ceases to be of importance. (14) The powerline stretches over my yard (15) She spread the tablecloth over the table (16) The rain clouds hung over the city Consistent with our discussion of (13) and (13'), it comes as no surprise that both su and sopra can be used, like over, in the sense of covering. (We have already met this meaning extension of su. See the discussion of (l')(g) and (l')(h).) Predictably, neither above nor al di sopra, which profile the separateness of TR and LM, permit this extension. In the case of over and sopra it is no longer necessary that the TR is spatially higher than the LM, nor, in the case of over, is contact between the TR and the LM excluded: (17) (a) He put his hands over Con) his face i.e. "He covered his face with his hands" Si mise le mani sopra il viso/sul viso (b) He wore a pullover over (*on) his shirt Portava una maglietta sopra la camicialsulla camicia
CONTRASTING PREPOSITIONAL CATEGORIES 315 For a covering relation to be effected, it is of course necessary that the covering entity has a surface equal in area to, or exceeding the area of the covered surface. The sentences in (18) merely describe a relation of contact, similar to that in (4)(a). A hand — note the singular noun in (18) as opposed to the plural in (17)(a) — is not large enough to cover a head. Consequently, over is impossible. (18) He put his hand on Cover) his head i.e. "He made contact between the surface of his hand and the surface of his head" Si mise la mano sulla testa/sopra la testa While both over and sopra can encode a covering relation, it seems that the English preposition has developed this sense to a greater degree than the Italian preposition. This is suggested by certain instances of non- correspondence between over and sopra. In distinguishing su and sopra we mentioned that sopra highlights very strongly the spatial superiority of the TR. Sopra seems to retain sufficient of this component of its central sense to prevent its use in cases where a downward movement or downward extension of the TR is profiled. There is no problem with (19). The TR not only covers the LM, the TR also moves in an upward direction: (19) Si tiro i calzini fin sopra le ginocchia* He pulled his socks up over his knees But in (20) (a) there is a downward movement of the TR, and in (20)(b) a downward extension of the TR. Here, sopra is much more problematic. It is as if there were a clash between the described state of affairs and a still active component of the central sense. (20) (a) Si abbassd il cappello sugli occhil* sopra gli occi He pulled his cap down over his eyes (b) // vestito le arriva sotto il ginocchio/* sopra il ginocchio Her dress comes down over her knee Su, which has no such strong component of verticality, is used in (20)(a), while (20)(b) prefers the converse of sopra, i.e. sotto "under", "below". Of course, both sopra and over occur in (21), where the component of spatial superiority again comes to the fore: (21) Si tiro il vestito fin sopra il ginocchio She lifted her dress up over her knee Before proceeding to discuss another meaning chain radiating out from prototypical over, it might be mentioned that the correspondences noted
316 JOHN R. TAYLOR between over, sopra and su also exist in metaphorical expressions. Let us briefly examine one metaphorisation of the meaning component "TR affects LM". Both English and Italian conceptualise status in society in terms of verticality. A person of superior status is above those of lower status; in Italian he is al di sopra degli altri. Here, above and al di sopra merely designate position in hierarchy. But a person of superior status normally also has the power to influence and control those beneath him. Since spatial over can imply that the TR affects the LM, it comes as no surprise that relations of power and control can be expressed with over. A person with power exerts this power over {"above) others. Consistent with what has been said of Italian prepositions — see (13') — we expect both su and sopra to express a power relation. This is in fact the case: regnare sulsopra "reign over", trionfare sul sopra "triumph over". Some more examples (note that English permits only over in these cases): (22) (a) Non ha nessun potere su di melsopra di me9 He has no power over me (b) Ha un grande ascendente sul ministro/sopra il ministro He has considerable influence over the minister The power of the TR over the LM may express itself in a relation of threat or aggression. Italian again permits both su and sopra; English, in comparable cases, has only over: (23) (a) La minaccia pende sul suo capolsopra il su capo The threat hangs over his head (b) Ha un vantaggio sul nemicolsopra il nemico He has an advantage over the enemy We started this discussion of over by noting some contexts in which over and above were interchangeable. We suggested that over possessed certain meaning components not shared by above. These made possible the meaning extensions "TR covers LM" and "TR controls LM". Su and sopra displayed similar extensions. Let us now examine another potential difference between over and above. Over possesses a potential meaning component which permits the development of a very extensive meaning chain. This time, none of the Italian prepositions under discussion share a similar development. On one reading, the following sentences are virtually synonymous: (24) (a) The plane flew over the Atlantic (b) The plane flew above the Atlantic
CONTRASTING PREPOSITIONAL CATEGORIES 317 The expressions over the Atlantic and above the Atlantic designate a bounded region perpendicular to the Atlantic, within which the plane's flight is located. The synonymy imposes itself more strongly if the verb is in the progressive aspect. The progressive (the plane was flying) profiles the continuation over a period of time of the Place relation encoded by the preposition. All three Italian prepositions may be used in an equivalent sense: (25) (a) Vaereo void sulV Atlantico (b) Vaereo void sopra VAtlantico (c) Vaereo void al di sopra dell'Atlantico The verb in (25) is in the punctual past tense. Consistent with the temporal extension of a static Place relation, the imperfect (volava) might be preferable. On another reading, the sentences in (24) are not synonymous, (a), unlike (b), could mean that the plane crossed the Atlantic. The expression over the Atlantic designates a bounded region which the plane entered, flew through, and then left. What about the Italian equivalents of (24)? Informants agree that al di sopra, like above, can only designate the Place of the plane's flight. There is some uncertainty as to whether su and sopra can convey the notion of crossing. One informant preferred to express the plane's crossing of the Atlantic by means of a verb prefix rather than a preposition, i.e. the crossing of the Atlantic is construed as a temporally dynamic process, rather than as a spatially dynamic relation: (26) Vaereo sor void V Atlantico (The prefix sor- is used in a few other instances to designate the crossing of a LM, e.g. II fiume ha sormontato gli argini "The river has risen over, i.e. burst its banks".) (26), however, also permits the insertion of su: Vaereo sorvold sulV Atlantico. While one dictionary (Devoto & Oli 1971) defines sorvolare as "attraversare in volo", i.e. "cross in flignt", another (Zingarelli 1970) gives the gloss "volare sopra", i.e. "fly over". Furthermore, sopra and su can both be used with passare: (27) Le truppe passarono (il fiume) sopra il ponte/sul ponte "The troops passed, i.e. crossed (the river) over the bridge" This use has a metaphorical extension similar to English: passare sopra una cosa "pass over, i.e. ignore something". Do these examples demonstrate that sopra and su possess the meaning component "TR crosses LM"? I think not. The notion of crossing seems to be contained in the verb [sorvolare, passare), the preposition merely desig-
318 JOHN R. TAYLOR nating the Place of the crossing. Over is quite different. Over has acquired the sense "TR crosses LM", and has become a full-fledged Path preposition. The semantic extension has gone even further. Over not only designates a Path (28), but also the end-point of the Path (29): (28) He walked over the road (29) He lives over the hill The Italian equivalents of these sentences make use of the Path preposition attraverso "across" (or the verb attraversare "to cross"), or, in the case of (29), the Place/Goal preposition oltre "beyond, at/to the other side of": (28') Attraverso la strada (29') Abita oltre la collina In this connection, it is of interest to observe that whereas the sentences in (30) are roughly equivalent, those in (31) definitely are not: (30) (a) He jumped on the wall (b) Saltd sul muro (31) (a) He jumped over the wall (b) Saltd sopra il muro The Italian sentences are rather similar to those in (7). In (30), sul muro (and on the wall) designate the Goal of the TR's movement, i.e. a part of the wall which could (but which need not) be its upper surface. Alternatively, su and on could be Place prepositions, i.e. sul murolon the wall designate the part of the wall on which the jumping takes place. This interpretation emerges more clearly in the English sentence He jumped up and down the wall. In Italian, the use of the imperfect would similarly favour this reading: Saltava sul muro. (31)(b) is very similar to (30)(b), except that the Place/Goal designated by sopra il muro is obligatorily the upper surface of the wall. (English can capture a similar restriction with regard to the Goal relation by means of the preposition onto: He jumped onto the wall.) (31)(a) is quite different. Here, over is a Path preposition. The LM is an obstacle which the TR passes by means of an ascending- descending movement. Italian construes this situation differently, by focussing rather on the end-point of the TR's movement: Saltd oltre il muro "He jumped beyond, to the other side of the wall". Alternatively, the scene could be construed as a process: Saltd il muro, a possibility which also exists in English (He jumped the wall). Over as a Path preposition is the starting point for a further meaning chain which, predictably enough, finds no correspondence in the distribu-
CONTRASTING PREPOSITIONAL CATEGORIES 319 tion of any of the Italian prepositions under discussion. A TR which goes over a vertical obstacle, as in (31)(a), traces a curved, arc-like trajectory. Over thus comes to designate movement along a circular, semi-circular or curved path, as in the following sentences: (32) The vase fell over (33) He turned over in bed (34) Turn over the page In these examples, over is of course an adverb, not a preposition. Its sense, though, is clearly an extension of the preposition as used in (31)(a). Italian encodes these processes through verbs, rather than verbs plus adverb: (32') // vaso si rovescid (33') Si rigird nel letto (34') Volta la pagina English and Italian, then, as a consequence of the different semantic structures that are available in the two languages, may well construe the same objective state of affairs in rather different ways. Consider the following two sentences: (35) (a) He pushed her over the balcony (b) La spinse giu dal balcone (35)(a), designates a curved trajectory of a rather special shape. There is an optional initial portion, during which the TR traces a short, upward movement, followed by a more extended movement downwards and away from the Agent. Italian does not focus on the configuration of the movement, but on its Source {dal balcone "from the balcony") and its direction (giu "down"). It may be noted that in some of its uses that we have considered, e.g. (29), over is strongly deictic, i.e. it implies the other side of the LM from the vantage point of an observer. This deictic component comes to the fore in expressions like the following: (36) The book's over here (37) Look over there where over denotes the end-point of an imaginary, horizontal path originating from the addressee. Needless to say, su, sopra and al di sopra cannot be used in a comparable deictic sense. It is worth noting, however, that the German cognate of over — uber — has developed a range of Path, Place and deictic uses very similar to English over, uber die Strafie gehen "walk
320 JOHN R. TAYLOR over the street", fiber der Strafie wohnen "live over the street", hieriiber "over here", etc. In fact, German has gone even further than English in the semantic extension of tiber, and uses tiber as a Path preposition corresponding to English via, past: (38) Ich fahre nach Hamburg uber Bremen "I'm going to Hamburg via Bremen" Let us conclude this comparison of English and Italian prepositions by considering their use with measurements of vertical distance and with scales of verticality. Possibly as a consequence of the various meaning nuances attaching to over, English usee above if merely the verticality of the TR to the LM is being profiled, cf. (11) (b). If the extent of vertical distance between the TR and the LM is at issue, the use of over is excluded: (39) The village is 1000 metres above (*over) sea level (This is the alternative explanation, alluded to earlier, for the unacceptabil- ity of over in (12).) There is the same restriction if vertical distance is measured from some arbitrary reference level in a more abstract domain: (40) The temperature is 50 degrees above (""over) zero Above is also used to locate a higher (or, more abstractly, an earlier) portion of a written (or spoken) text: (41) The above-mentioned (*over-mentioned) examples In comparable instances, Italian has no such restrictions. (39) can be expressed with su, sopra or al di sopra (although su is perhaps more usual). (40), however, prefers sopra; a possible reason for the exclusion of su will be mentioned later: (39') // paese si trova a 1000 metri sul livello del mare/sopra il livello del mare/al di sopra del livello del mare (40') La temperatura e 50 gradi sopra zero/*su zerolal di sopra dello zero (41') Gli esempi succitati/sopraccitati If the LM is an arbitrary reference level, English, as we have seen, uses above. On the other hand, if the LM is a certain critical value, which is exceeded, over must be used: (42) (a) It cost over (*above) a million pounds (b) A man over (*above) sixty years old
CONTRASTING PREPOSITIONAL CATEGORIES 321 The profiled value, while exceeding the critical value, is still fairly close to it. If something costs two million pounds, one would not say it costs over one million. This calls to mind our earlier claim — see the discussion of (11) — that with over the TR, though not in contact with the LM, is nevertheless likely to be close to it. Above has no implications of proximity; in (39) and (40) the distance from the reference level might be very considerable. For the sense "exceeding a critical value" Italian uses sopra: (42') (a) Costava sopra un millione (b) Un uomo sopra i sessant'anni Here, sopra is not replaceable by su. Su denotes "approximate value": Un uomo sui sessant'anni "a man of around sixty". This is perhaps why su is impossible in (40'); su zero (or sullo zero) would be equivalent to "around zero". The adverb su, though, can be used in a sense rather close to (42), as in uomini di sessant'anni in su "men of sixty upwards", i.e. "men of sixty and over". Note that on is of very limited use in connection with values on a scale: *a man on sixty is impossible. We can, though, say a man just on sixty, where we refer to the precise (not the approximate, as in Italian) value. It is tempting to relate these contrasting senses with the central meanings of the prepositions. On requires contact between the TR and the LM, and so naturally encodes precise position on a scale. Su is relatively indifferent to contact, so comes to mean "approximate position". With sopra, as always, it is the higher location of the TR that dominates. 4. Conclusions This necessarily brief and selective account of English and Italian prepositions has hopefully gone some way to convincing the sceptical reader of the systematicity of prepositional usage.10 The frustrations of the foreign learner are due, not so much to the "idiomaticity" of prepositions as such, but rather to the language-specific manner in which prepositional categories are structured. This structuring of categories makes no appeal to "core senses". It would be fruitless to try to account for prepositional usage in terms of a core sense which (a) is present in each occurrence, i.e. it fully sanctions each occurrence, and which (b) uniquely distinguishes that preposition from every other lexical item in the language. Over, of course, provides the "classical" refutation of the core sense approach. While it might be feasible to identify a core sense for above and al di sopra — prepositions which display a relatively restricted range of uses — the core sense
322 JOHN R. TAYLOR approach manifestly fails with other prepositions, e.g. su and sopra, if only because these prepositions have uses in which some presumably "essential" meaning component is absent. Thus, one would presumably posit "TR is higher than LM" as a component of the core sense of sopra; how, then, does one account for (17)? Prototype theory offers itself as a viable alternative to the core sense approach to polysemy. Indeed, the necessity of some such theory to a study of prepositional semantics has been amply demonstrated by Brugman (1981) and others. My aim in this paper was not so much to argue for a prototype approach — I take the need for a prototype approach for granted — but rather to demonstrate its relevance to a cross-language comparison. In what ways, then, has a prototype approach thrown light on the sometimes very different distributions of similar lexical items in different languages? In section 1 it was suggested that characteristics of the prototype itself may be in part responsible for the particular distribution of a preposition. A comparison of over and sopra is instructive in this respect. In many ways, over and sopra are very similar. Both can designate the same kind of spatial relation (2), both can have the meaning component "TR affects LM" (13), and both undergo the meaning extension "TR covers LM" (17). There are, though, cases in which a covering relation is encodable by over, but not by sopra (20). This lack of correspondence can be explained if we posit "TR is vertical to LM" as a very strong component of prototypical sopra. The requirement that the TR is vertical to the LM can indeed be relaxed, as in (17), but only if there is no clash between the notion of higher location and the profiling of a downward movement or downward extension, as in (20). Elsewhere in our discussion, e.g. when distinguishing sopra from su, we have found it necessary to refer to the very high salience of this component of sopra.(See also note 7.) The very specific "image" associated with sopra (and indeed with other prepositions) means that instances of perfect synonymy (and of perfect translation equivalence) are probably exceedingly rare. Throughout this paper, we have drawn attention to the fact that even though two prepositions might be interchangeable in a certain context, it does not necessarily follow that they encode identical construals of a particular state of affairs; rather, the prepositions profile relations which, though different, are nevertheless compatible with the real-world situation. A nice example is provided by the sentences in (2): The lamp hangs above the table and the lamp hangs over the table. The different "images" encoded by the two pre-
CONTRASTING PREPOSITIONAL CATEGORIES 323 positions are no longer compatible with a slightly different real-world situation, as shown by the unacceptability of *Move the lamp down above the table (13). As argued in section 3, it is the different images associated with over and above that permit the different extensions of the two words. The meanings "TR covers LM" and "TR controls LM" capitalise on the meaning component "TR affects LM" of over, while above comes to highlight the absence of interaction between the TR and the LM. Meaning extensions of the type alluded to can be represented schematically as follows. A polysemous lexical item has a central meaning A. Meaning B develops from A, and meaning C develops from B: A > B > C The development of meaning C is obviously dependent on the prior existence of meaning B. Suppose a lexical item in another language also has a central meaning A (or one very close to it). If this item also has a meaning equivalent to C, we predict that it will have meaning B. Conversely, if meaning B is not attested, then meaning C will be absent.11 This very simple model accounts for the parallel development of the meanings "TR covers LM" and "TR controls LM" in over, sopra and su. This semantic extension is dependent on the prior existence of the meaning "TR affects LM". But we also encountered a meaning chain for over which had no counterpart in Italian. The explanation is that none of the Italian prepositions has undergone a generalised extension to "TR crosses LM" — a precondition for extension along the lines of over. In other words, the lack of equivalence between English and Italian, as shown in (28), (29) and (31) to (34), far from being random, reflects a systematic difference in the structuring of prepositional categories. If the meaning chain approach can account very elegantly for the different distributions of semantically rather similar lexical items, both within the same language and across different languages, it does nevertheless raise a number of unresolved issues. The most pressing is: Why is it that certain items develop sometimes quite extensive meaning chains, while other, apparently similar items, undergo extension in a different direction, or perhaps do not undergo extension at all? A related point concerns constraints. If meanings can be chained on the basis of similarity between contiguous members, then in the absence of constraints anything can ultimately be associated with anything else. Lakoff (1982a: 17) has suggested that the extension of a category might be influenced by the existence of neighbouring or contrasting categories.
324 JOHN R. TAYLOR Certainly, it is tempting to regard the existence of above and on in English — the first profiling verticality without contact, the second verticality with contact — as a factor which facilitates the extension of over away from the domain of spatial verticality. Although we have no diachronic data to support this view, one might speculate along similar lines that the semantic extension of sopra created a "gap" that was filled by the formation of the prepositional phrase al di sopra. However, the Italian data that we have examined suggests that lexical categories do not necessarily interact in this way. Both su and sopra are interchangeable in a wide range of contexts, and the fact that sopra can mean "TR covers LM" does not prevent su from developing in the same way. Neither — to return to English — does the existence of the prepositions across and beyond prevent over from encroaching on their semantic space. But perhaps the desire to constrain meaning chains is misguided. Lan- gacker writes that "all manner of entities will be assimilated to a category if a person can find any plausible rationale for construing them as extensions from prototypical members" (1983: 1/13). While the development of a meaning chain is certainly motivated, it does not follow that it is predictable. We should therefore expect different languages to categorise human experience in sometimes highly idiosyncratic ways. We have already seen in (35) how the same objective situation may be conceptualised very differently in two languages. Notes 1. I am indebted to Rene Dirven, Bruce Hawkins, Mario Leonardi-Lollis and Brygida Rudzka-Ostyn for comments on an earlier draft of this paper. Inaccuracies and infelicities, of course, are my own. 2. A noteworthy exception is Dirven (1981). See also the forthcoming Learner's Grammar of English, edited by R. Dirven. 3. This is essentially the procedure adopted by Bennett (1975), who extracts the core sense [locative superior) for over. 4. Langacker is very explicit on this point: There is nothing in (prototype theory) which intrinsically encourages one to seek the shortest possible list of criterial properties that will correctly pick out all and only the members of a class. Instead it encourages the fullest possible characterization of prototypical instances. (Langacker 1983: 1/ 13) 5. The data for sopra necessarily also include those for al di sopra. The real frequency of the
CONTRASTING PREPOSITIONAL CATEGORIES 325 simple preposition sopra is thus likely to be lower. If one arrived at a hotel which had been described in a publicity brochure as being "on the sea", and found that one had to make a steep 200 metre descent to get to the beach, one would certainly have cause for complaint. Italian sul mare does not necessarily indicate such close proximity. In metaphorical expressions, su and sopra are again often interchangeable, as in (22) and (23). Sometimes, though, semantic contrasts are possible. Thus, piangere su and piangere sopra both correspond to English "mourn, weep over1'. Sopra, however, strongly suggests the spatial verticality of the mourner, e.g. with respect to a loved one (i)(a). If the LM is abstract, this spatial orientation is not possible, and su is preferred (b): (i) (a) Piangeva sopra il pardre morto "He wept over (the body of) his dead father1' (b) Piangeva sulle sue sventure "He bemoaned his misfortunes" Fin "till" in this sentence, and in (21), is optional. It highlights the Goal relation expressed by sopra. Note the obligatory occurrence of di in su di me, sopra di me. Most Italian prepositions (exceptions include a "to", da "from", di "of" and per "for") require the insertion of a second preposition, usually di, before a pronoun object. Thus verso la casa "towards the house" but verso di noi "towards us". This obligatory marking of a pronoun object by means of a preposition recalls a similar phenomenon in Spanish, where a [+human], [+definite] Direct Object must be preceded by the preposition a: El director busca el carro "The manager is looking for the car" but El director busca al empleado "The director is looking for the employee". Comrie (1981a: 120ff.) explains this phenomenon in terms of prototypical clause structure. In a prototypical transitive sentence, the Subject is a [+human] Agent, the Direct Object a [—human] Patient. The Object is usually also New Information, and Indefinite. A Direct Object which is [+human] and [+definite] (and therefore probably also Given) diverges considerably from the prototype, and thus favours explicit case marking. A similar kind of explanation may be offered in connection with Italian prepositions. The LM in a locative expression is prototypically immobile, [-human], and specified by a full, definite NP. A pronoun, especially a personal pronoun, is an untypical LM, and so requires special marking. A more comprehensive discussion would have to deal with the wide range of abstract and metaphorical uses of the prepositions, as well as with over- and sopra- as noun and verb prefixes. At this point, an observation on the distinction between "spatial" and "metaphorical" uses is perhaps in order. In this paper, I proposed to limit my attention to spatial uses. On some occasions, however, it was found advisable to make reference to metaphorical extensions. (22) and (23) seem particularly apt for elucidating the meaning component "TR affects LM" claimed to characterise over, sopra and su in (13) and (13'). But the meaning component "TR affects LM" is itself somewhat "abstract", not to say metaphorical. In other words, a clean distinction between literal and metaphorical uses might not always be desirable, or even possible. As Langacker (1983: 1/1-2) remarks, language is essentially figurative in nature, and attempts to separate the figurative from the literal, and to treat the figurative as subsidiary to the literal, are bound to lead to an impoverished description.
326 JOHN R. TAYLOR 11. This is, of course, an idealised representation. Meaning B could become obsolete, leaving meaning C stranded.
The Mapping of Elements of Cognitive Space onto Grammatical Relations: An Example from Russian Verbal Prefixation Laura A. Janda University of Rochester 0.0 Introduction This paper demonstrates the use of a cognitive model to describe the semantics of Russian verbal prefixes, as applied to the prefix pere-. The cognitive model provides a valid description not only of the semantic contributions of the prefix, but of the syntactic relationships between the resulting verb and its arguments as well. Thus, as will be shown below, a single set of basic concepts can be used in descriptions on both the semantic and syntactic levels. 0.1 The Role of the Prefix and How It Is Captioned The role of the Russian verbal prefix is comparable to that of a director; it establishes a setting, gives a general plot to the action and casts arguments of the verb in specific roles. The prefix is an organizer which processes the "raw materials" of the semantics of the verb and its relationship to its arguments. This semantic contribution of the prefix to a sentence is captioned in terms of configurations, which are roughly equivalent to the profiles used by Langacker (1983; see also Lindner 1981 and Brugman 1981 in which English up and out, and over, respectively, are described in terms of cognitive models and Rudzka-Ostyn 1983 which uses cognitive models to compare the semantics of Dutch uit and Polish wy). One of the configurations associated with a prefix usually serves as an Idealized Cognitive Model (ICM, see Lakoff 1982b) or prototype of the prefix and therefore as the central member of the network of related configurations which describe
328 LAURA A. JANDA the prefix as a whole. The use of a network of configurations captures both the diversity of instantiations of a prefix and their semantic unity. Previous descriptions of Russian verbal prefixes (specific examples of which are given below) have been based on a set-theory model of organization in which diversity and unity are diametrically opposed. Traditionally, lists of seemingly unrelated meanings have been ascribed to prefixes (promoting diversity at the expense of unity; see Slovar' 1950-65 and Boguslawski 1963), whereas in the structuralist school each prefix was given a single vague designation (stressing unity at the expense of diversity; see van Schooneveld 1978, Gallant 1979 and Flier 1975). The cognitive model does not restrict the salience of either of these aspects of prefixal semantics. In the present cognitive model configurations are drawn in what might be termed "cognitive space". This is not necessarily three-dimensional space as it is understood by physicists, but rather our mental perception of it, and may have one, two, or three dimensions. Prototypically a configuration consists of a landmark (labeled LM, sometimes called a domain), and a trajector (TR) which moves in relation to it. The trajectory (TRy) is a profile of this movement, usually with respect to time (cf. the use of trajectories in the profiles in Langacker 1983). First the semantic role of these elements will be examined briefly and then their relevance to the syntax of the prefixed verb and its arguments will be discussed. The landmark and trajector may take a limited range of different shapes and relative sizes and, like the cognitive space in which they are drawn, can refer to many things other than physical objects. The landmark can refer to space, time, an act, or a state, and the trajector likewise may be an actual or abstract object. Like pere-, any other Russian prefix has several configurations, each of which may have several applications, or submeanings. Usually at least one such submeaning will be spatial and the rest will be metaphorical extensions of that submeaning, created by varying the referents of the landmark and the trajector. Of the configurations associated with a given prefix, one typically emerges as the central or prototypical configuration, to which all others are related by means of a series of links. These links represent the minor transformations by which the configurations differ from one another and comprise a small, closed set. All of the links associated with pere- can be found in the structure of other prefixes as well (for more on inter-config- urational linkage and the limitations on its typology, see Janda 1986), which
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 329 Configu rati on 7 <bend> LM = TRy Configuration 6 cover > Configuration 9 <mix> LM mass -> count Configuration 8 <turn> TR = part of LM LM2-D->3-D [ r^ '/ / L" [ V_y \r 1 Configu rat ion <d1vision> ILM 1 A ►TR TR0-D->2-D Configuration 5 <thorough> Configuration 3 <interchange) TR1 _ LM rTP2 Reduplication of TR LM 1-D -» 2-D Configuration 1 <trensfer> <duration> Configuration 2 superior ity><excess> 4- LM -^•TR iQuantificationof LM i TR LM Figure 1
330 LAURA A. JANDA indicates that these links are neither random nor arbitrarily concocted. Links exhibit transformations of quantification (i.e., distinctions in dimensionality, conception of configurational elements as count versus mass objects) and identification (of parts or wholes of elements with each other). 1.0 The pere-Network Figure 1 contains the network of configurations associated with the prefix pere-, which is by far the most versatile of the Russian verbal prefixes and which performs about the same function as the English verb particle over in approximately two-thirds of its submeanings. The submeanings captioned by each configuration are given in brackets. These bracketed terms are intended only as convenient labels for the submeanings and should not be construed as semantic features. A brief characterization of the linking transformation (described in more detail in the text below) appears between neighboring configurations. Configuration one serves as the ICM of pere-'s semantic network, from which all other configurations in this network trace their origin, as indicated by the arrows in Figure 1. 1.1 Configuration 1 Configuration-1 <transfer> <duration> H—s—**TR Figure 2 The prototype of this network is numbered one. In this one-dimensional configuration, which is associated with the submeanings <transfer> and <duration>, the trajector traverses the length of the landmark, a stretch of unidimensional cognitive space which may refer to distance, an object, or time. 1.1.1 <Transfer> Cognitive space indexes a distance between two points of reference in
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 331 either a spatial or a non-spatial domain. The trajector is transported across the intervening distance and arrives at the terminal boundary of the landmark. In concrete instantiations the landmark refers to a physical distance. An example of a metaphorical extension of this submeaning involves the social domain of loyalty, in which individuals can be "transferred" to a new leader or firm without being physically transported. In the examples the letters (A) and (B) refer to relationships between the verbal arguments and configurational elements. These relationships will be discussed in section three. (1) Ona prikazala pereletet'(A) na druguju ploscadku. She ordered pere-fly-infin to other square "She ordered me to fly to another airfield." Also: perenesti (A) cemodan v druguju komnatu per e-c&rry-mim suitcase in other room "transfer (by carrying) the suitcase to another room" peredat'(A) sol/operu pere-give-infin salt/opera "pass the salt/broadcast an opera" perejti (B) ulicu pere-walk-infin street "cross the street" 1.1.2 <Duration> Here, cognitive space refers to time and therefore the landmark is a period of time during which the trajector pursues a given activity and the trajectory represents its progress through time. The activity usually involves waiting or killing time in some way. (2) Oni otpolzli v storonu, ctoby perezdaf(B) obstrel. They crawled-off in side to pere-wait-infin firing. "They crawled off to one side to wait for the firing to finish." Also: perezimovaf (A) pere-winter-infin "spend the winter"
332 LAURA A. JANDA perekurit' (B) /?ere-smoke-infin "smoke during a break" pereboiet' lixoradku (B) pere-\\\\r\. fever "get over a fever" 1.2 Configuration 2 Configuration 2 <superiority><excess> 1 TR r LM Figure 3 In this configuration the landmark is quantified and the movement of the trajector is measured against it. To signal the scalar function of the landmark in this configuration it appears with a vertical orientation. There are three submeanings associated with this configuration: <superiority>, <excess>, and <redo>. Of these submeanings, only <superiority> makes occasional reference to the spatial domain, and in most instances the landmark is a performance. 1.2.1 < Superio rity > In this meaning the landmark is a given performance of an activity which serves as a quantitative scale against which the activity of the trajector is measured. The trajector crosses the endpoint of the landmark and therefore exceeds the performance referenced by the landmark.
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 333 (3) Kto kogo perepet (B) ? Who who pere-drink-3rd sg "Who will outdrink whom?" Also: perekricaf (B) pere-shout-infin "outshout" perexitrit' (B) pere-cunning-infin "outsmart" 1.2.2 <Excess> As in <superiority>, the landmark refers to a prescribed performance, but in this case it is an ideal or canonical performance. By overstepping the endpoint of the landmark the trajector produces a performance which is evaluated as excessive. (4) Dobrynin toze peresidel (A) v Vasingtone. Dobrynin also pere-sit-past-masc sg in Washington. "Dobrynin has also been sitting around too long in Washington." Also: peresolit' (A) pere-sa\t-\nf'm "oversalt" perestarat'sja (A) pere-try-infin-reflex "try too hard" 1.2.3 <Redo> The landmark here again refers to an act, which is this time specified as previously completed. The trajector retraces the path of the action, in many instances making repairs or corrections along the way. (5) My ne rasslysali ego imja, We not hear-distinctly his name "We did not catch his name..."
334 LAURA A. JANDA Also: no peresprasivat' (B) bylo neudobno. but /?ere-ask-imperf-infin was awkward .. but it would have been awkward to ask again" peredumat' (B) /?ere-think-infin "rethink" perestroit' (B) /?ere-build-infin "restructure" 1.3 Configuration 3, <Interchange> Configuration 3 <1nterchange> Figure 4 Configurations one and two are the semantic center of gravity for the entire network and are therefore laden with a number of submeanings. The remaining configurations are less central and have only one submeaning each. Configuration three, which captions the submeaning <interchange> can be produced by reduplicating the trajectory in configuration one. Thus the verbal action is volleyed back and forth across the landmark. (6) V vetrennom sume perelaivalis' (A) sobaki. In windy noise /?ere-bark-past-pl-reflex dogs "In the roar of the wind the dogs barked at each other." Also: perepisat'sja (A) /?ere-write-infin-reflex "correspond"
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 335 1.4 Configuration 4, <Division> Configuration 4 <division> Figure 5 Here the landmark of configuration one is extended to two dimensions, and it is therefore bisected by the one-dimensional trajector. This action divides the landmark into two pieces. The landmark may be either a physical object or may refer to an "object" in another, non-spatial domain, as in perebif "interrupt (a conversation)." (7) Peresekaem (B) liniju fronta. pere-cut-lst pi line front "We cut across the front line." Also: peregruzit' (B) pere-gnaw-infin "gnaw in two" pererubif (B) pere-chop-infin "chop in two" 1.5 Configuration 5, <Thorough> Configuration 5 <thorough> Figure 6
336 LAURA A. JANDA This confguration is derived from configuration four by extending the dimensionality of the trajector. The trajector becomes a two-dimensional object1 which spreads over the landmark, covering it. The landmark refers to some object, all of which is subjected to the action of the trajector. (8) Katja vdrug upala i peremazalas' (A). Katja suddenly fell and pere-smear-past-fem/sg-reflex "Katja suddenly fell and got herself completely dirty." Also: perezjabnut' (A) pere-chill-infin "get chilled through" peretrusif (A) pere-coward-infin "become altogether cowardly" 1.6 Configuration 6, <Over> Configuration 6 <over> Figure 7 A further extension in the dimensionality of the landmark in configuration four produces the solid, three-dimensional landmark of configuration six. The trajector is no longer locked in a plane of two dimensions and can now reach the other side of the landmark by hopping over it rather than cutting through it. In a degenerate version the landmark may be no more than a line, as in perestupit' granicu "cross the border," which, however, still retains the vertical dimension in the trajectory. (9) Samoletu ne udajetsja pereskocif (B) gornuju grjadu. Airplane not manage pere-jump-infin mountain ridge "The plane will not be able to leap over the mountain ridge." Also:
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 337 perestupit' (B) /?ere-step-infin "step over" perelezt' (B) pere-climb-infin "climb over" 1.7 Configuration 7, <Bend> Configuration 7 <bend> Figure 8 Configuration seven is arrived at by integrating features of both landmark and trajector in a single entity. The product of this identification of the landmark with the trajector produces a three-dimensional solid which approximates the curve of the trajectory in configuration six, i.e. a bent object. This configuration and the two which follow it are rather limited in terms of the number of verbs which are associated with each of them, and indeed they are peripheral in pere-'s cognitive network. (10) Koska peregibaet (A) spinu / zmuritsja. cat pere-bend-3rd sg back and squints "The cat hunches up its back and squints." 1.8 Configuration 8, <Turn> Configuration 8 <turn> Figure 9
338 LAURA A. JANDA Configuration six serves as the source for this configuration as well. One end of the landmark traces the path of the trajector, causing the landmark to flip over. Although the landmark is usually a physical object, it may reference "objects" in non-spatial domains, as in u nego zizn per ever- nulas' "his life has been turned upside-down." (11) U-2 zacepilsja kolesami za provoda U-2 caught wheels behind wires "The wheels of the U-2 got caught on the wires..." .../ perevernulsja (A). and pere-turn-past-masc sg-reflex "...and flipped over." 1.9 Configuration 9, <Mix> Figure 10 <Mix> can be conceived of as a degenerate version of the preceding configuration. In the transformation between configurations six and seven the solid landmark is decomposed into a group of objects. Thus when the edge of the landmark traces the trajectory the result is a shuffling of the landmark's composite parts. (12) Koren'ja i luk peremesivajut (A) i podzarivajut. spices and onion pere-m\x-?>xd pi and saute "Spices and onion are mixed together and sauteed." 2.0 The Cognitive Model in Semantic Description: A Summary The preceding section has presented an example of the application of a cognitive model to semantic description. Despite the semantic diversity of the prefix pere-, all of its meanings were integrated into a single coherent network, which illustrates the relationships that hold the category together. A traditional treatment (using set theory and/or semantic features) of such Configuration 9 <m1x>
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 339 a rich semantic category would be unable to capture its underlying unity since the main focus of such a description is the identification of divisions rather than relationships. A structuralist description would conversely attempt to assign a single vague characterization to the entire category, suppressing the very real distinctions which exist within the category. Without the cognitive model, semantic diversity cannot be reconciled with underlying unity and the result is inevitably a list of unrelated meanings, giving the impression that pere- (or any other semantically complex lexical item) is merely a haphazard group of homonyms. The position of a configuration in the network is also relevant. Configurations which are central to the network index submeanings which are semantically central to the prefix and which are more frequently used than those which appear on the periphery of the network and which are indeed peripheral, such as pere- <bend>, <turn>, and <mix>. The network thus offers a scale of centrality and typicality of submeanings. In addition to providing an elegant description of the semantics of the prefix pere-, the cognitive model is endowed with a certain amount of predictive power. Each submeaning is associated with certain semantic groups of verbs. Thus, given a prefix and a verb, one can predict what submeaning will occur. For example, verbs denoting longitudinal transportation are associated with pere- <transfer>, verbs denoting transportation having a vertical component are associated with <over>, verbs used in cooking are associated with <excess>, and verbs of cutting and breaking are associated with <division>. In theory (once all of the prefixes have been subjected to this description) one could likewise select the proper prefix to produce the desired submeaning with a given verb. Such prediction can also be made available with a traditional semantic description, although to a more limited extent, since the relationships between the submeanings and between the verb and prefix are not as clear as they are in the cognitive model. 3.0 Mapping the Cognitive Configuration onto the Verbal Arguments An important by-product of the use of the cognitive model in the semantic description of prefixes is that it also reveals a system of relationships between the prefixed verb and its arguments. The verbal arguments are assigned the roles of trajector and landmark2 according to two specific patterns, outlined below. (Throughout the text above the letters (A) and (B) have been used to identify which pattern is represented by each of the
340 LAURA A. JANDA examples.) Pattern A: TR = intrans subj or DO LM = distance to terminus specified in PP or indirectly Pattern B: TR = trans subj LM - DO To elaborate, according to pattern A, which is the dominant pattern, the role of trajector will be played by the subject of the sentence when the verb is intransitive, or by the direct object when the verb is transitive. The opposite endpoint of the landmark in this case will be identified in a prepositional phrase or specified by the context. In some cases the identification of the landmark is included in the verb itself, as in a number of the <dura- tion> verbs e.g., perezimovat' (pere-winter — "spend the winter"), perenocevat' {pere-n\g\\\ — "spend the night"). Pattern B assigns the role of trajector to the subject of a transitive verb, and that of landmark to its direct object. To illustrate how the configurational elements are mapped onto verbal arguments, I will review some of the examples given in the text above (which here appear in abbreviated form). (la) Ja perelecu (A) na druguju ploscadku. "I will fly to another airfield." The subject of the intransitive verb acts as the trajector and the terminus of the landmark is specified in the prepositional phrase. (12a) My peremesivaem (A) koren'ja. "We mix spices." This example also illustrates pattern A, but in this case the verb is transitive and it is therefore the direct object which fills the role of trajector. The trajector in this configuration is defined as part of the landmark, thus the landmark is indirectly identified as the remaining spices. (7a) My peresekaem (B) liniju fronta. "We cut across the front line." Again the verb is transitive, but this is an instance of pattern B, in which the subject, we, is the trajector and the direct object, the front line, is the landmark.
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 341 In a survey of all the submeanings of four prefixes (see Janda 1986) it was found that all examples of verbs conform to one of these patterns, of which pattern A is the most common, and that the patterns appear in complementary distribution. Most submeanings are specific to a single pattern, and in those instances where both patterns appear in a submeaning, transitive verbs were found to use pattern B and intransitive verbs were found to use pattern A. Of pere-'s submeanings, five conform to pattern A (<excess>, <bend>, <interchange>, <turn> and <mix>), four conform to pattern B (<superiority>, <redo>, <division>, <over>) and three of the sub- meanings use both patterns, the distribution of which is determined by whether the verb is transitive or intransitive (<transfer>, <duration> and <thorough>). Intransitive verbs are restricted to pattern A, in which their subjects act as trajectors. This is reasonable, since in a statement made with an intransitive verb only the subject is capable of any movement, be it literal or metaphorical. Transitive verbs, however, conform to either pattern A or B, and the role of trajector is played by the direct object or the subject, respectively. The trajector is the active, moving part of the configuration and likewise indexes the most active and salient argument of the verb (there are examples of transitive verbs with which both the subject and the object follow the motion described by the trajectory, cf. perenesti cemodan cited in 1.1.1 above; in such cases the object is the trajector, since it is the patient). Thus the selection of pattern A suggests that the subject is inactive and that the direct object is active, whereas the converse is true for pattern B. This distinction between active and inactive subjects implies that there are subtle distinctions to be made in the seemingly unified categories of subject and direct object. Furthermore, since entire submeanings select a single pattern for their transitive verbs, it seems that the distinction between pattern A and pattern B for transitive verbs is made on semantic grounds. Thus, distinctions made between salient, active subjects and inactive ones arise from differences in the semantic freight of verbs. This confirms Langacker's (1983) observation that there are varying degrees of subjectivity, and that distinctions are made between what he has termed the "optimal viewing arrangement," which parallels pattern A, in which the object is more involved in the action, and the "egocentric viewing arrangement," parallel to pattern B, in which the subject is more prominent. Since in pattern A direct objects are equated with subjects of intransitive verbs (i.e., both are identified as the trajector in the configuration),
342 LAURA A. JANDA transitive pattern A verbs will henceforth be identified as "para-absolu- tive," as opposed to pattern B verbs which will be labeled "non-absolutive." This terminology has been adopted because pattern A assigns equivalent roles to subjects of intransitive verbs and direct objects, reminiscent of the case distinction made in ergative-absolutive languages (although Russian is a nominative-accusative language, this does not preclude the existence of ergative-absolutive distinctions in the language, cf. Comrie (1981a: 104-108) who cites languages which use both distinctions and states that "it is misleading to classify a language as being either ergative or not"). Significantly, verbs which conform to pattern B have often undergone trans- itivization as a by-product of prefixation, probably due to the close relationship between the subject of the newly-transitivized verb and the verbal action. Below is a list of para-absolutive and non-absolutive verbs prefixed by pere-, in which transitivized verbs are marked with an asterisk. Para-absolutives: Non-absolutives: peredavaf *perepit' pere-give-infin pere-drink-infin "convey" "outdrink" peregruzit' peresit' pere-load-infin pere-sew-infin "overburden" "resew" perekinuf *pereskoc'it' pere-throw-infin pere-jump-infin "toss over" "jump over" Although pit' "drink" can, of course, be used transitively, it is the intransitive use which is selected by the prefix and transitivized, i.e. while it is possible to say On perepil gostej "He outdrank the guests", it is certainly not possible to say *On pil gostej "*He drank the guests". The transitivization of verbs through prefixation in Russian has a parallel in the use of verb particles in English: the simplex verb laugh, for example, is intransitive, but can be transitivized by the addition of the particle out, as in When John presented his new theory, his colleagues laughed him out of the room. As mentioned above, the terms "non-absolutive" and "para-absolutive" were chosen because Russian makes distinctions similar to those made in ergative languages. In ergative languages the focus is on identifying the patient of the verb, which results in a distinction between patient (and therefore salient) and non-patient subjects. Russian verbal prefixes also
COGNITIVE SPACE AND GRAMMATICAL RELATIONS 343 select patients (as trajectors) and distinguish them from non-patients. Thus although no ergative distinction is morphologically encoded in Russian, it is nevertheless present and in turn reveals that the grammatical relations of subject and object may be complex rather than primitive concepts, and that ergative-absolutive distinctions may be more universal than case typology indicates. The fact that such conclusions can be drawn from a study of pre - fixal meaning points to a dynamic interaction between semantics and syntax, two levels of linguistic study which were integrated in the present cognitive model. Notes 1. The referent of the trajector may be three-dimensional, but since its thickness is irrelevant, it is construed as two-dimensional. 2. The terms "trajector" and 'landmark1' signify the concepts used by Talmy, rather than the more abstract trajectors and landmarks characteristic of Langacker's work.
Conventionalization of Cora Locationals Eugene H. Casad Summer Institute of Linguistics 1.0 Introduction Locational and directional notions are frequently elaborated to a surprising degree by many of the world's languages.1 A close examination of the ways a given language employs spatial terms often uncovers a plethora of peculiarities and puzzles. The usages of spatials often seems inconsistent and arbitrary. In this paper I show that many of the usages of the locational-direc- tional system of Cora, a language of Mexico, are based on broad perceptual and cognitive principles.2 In part, these usages are determined by linguistic conventionalization. By conventionalization, I mean the way in which speakers of a language have collectively arrived at a concensus on how to express thoughts and symbolize them in the grammatical structure of their language. Conventionalization operates at all levels of awareness and language use. Thus, for Langacker, semantic structure is the conventional expression of conceptual structure. Syntax and morphology are the conventional symbolization of semantic structure (1983: 1/6, llOff.). The data discussed in this paper show such conventionalizations at work at several different levels of analysis in Cora grammar. My analysis is based on the theory of Cognitive Grammar exemplified by Langacker (1979, 1982a, and 1983). This theory sees no intrinsic difference between general human cognitive abilities and more specific linguistic ones (1982a:7). A primary purpose of this paper is to thus examine the various cognitive and perceptual abilities that appear to partially determine the conventionality of the Cora spatials. Visual imagery is seen to be an important, but by no means exclusive, determinant of the usages of the topog-
346 EUGENE H. CASAD raphic adverbs that I discuss in section 2.0, as well as many of the usages of verbal prefixes and stems that I discuss in later sections of this paper. This correlates well with the recent work of Kosslyn (1980) and Shepard (1978) on the nature of the mental image and the capacities that people have for manipulating them. Kosslyn (1980) has demonstrated the viability of visual imagery as a testable cognitive construct as has Shepard (1978). Miller and Johnson-Laird have suggested quite plausibly that visual imagery can provide valuable support to logical reasoning (1976:241). In no case do these researchers imply or state that visual imagery plays the predominant role in language use, but rather they grant it a viable status in this aspect of human endeavor (Kosslyn 1980:456). The data in this paper suggest an analogous role for it. Various usages of Cora spatials are explainable in terms of transformations that people perform on mental images. A set of distinct points, for example, can become assimilated to an entire configuration. On the other hand, a speaker can fade out a portion of a mental image and focus on the portion that remains in his field of awareness. Thus, Cora speakers at some point began to ignore the intrinsic vertical orientation in the meaning of the prefix sequence antyi- "up at the highest point of" and extended that meaning to "at the tip of any elongated entity". Cognitive Grammar views semantic structure as conceptual structures that have been shaped in particular ways common to a culture so that speakers can express their thoughts linguistically with a reasonable expectation of success (Langacker 1983:11/1). Meaning is therefore central to linguistic analysis and description. In this paper I show clearly two facets of meaning pinpointed by Cognitive Grammar that are crucial for providing an adequate characterization of the usages of Cora locationals. For one, there is no direct link between a given usage and the objective scene that it describes. Rather, the same objective scene can be viewed in several different ways, with various aspects of that scene overlooked or included for particular communicative purposes (ibid.:ll/15). Shepard (1978) shows that this is a general cognitive ability. In addition, the characterization of a concept usually requires reference to more than one conceptual domain. Thus, to characterize the individual meanings of the Cora topographic adverbs, I have had to speak of conceptual complexes that relate the speaker to (1) a distance domain, (2) a boundary domain, (3) the land gradient, (4) the domain of the hill and (5)
CONVENTIONALIZATION OF CORA LOCATIONALS 347 the domain of the river. In short, Cognitive Grammar takes an encyclopedic view of meaning, rather than a narrowly defined dictionary meaning (Langacker 1983:11/63).3 The meaning of linguistic expressions is context defined and includes all aspects of the information that are relevant to describing particular conceived situations (ibid.: 11/65). As the analyses discussed in this paper show, it is often necessary to specify what the speaker's vantage point is from which he views a scene (cf. Casad and Langacker 1985:281). In other instances, it is necessary to specify the particular aspects of the scene that the speaker selects for discussion or to indicate the relative degree of detail the speaker brings to the description of that scene (Langacker 1983:11/23). Often, particular usages can most readily be explained in terms of their representing extensions from prototypical usages.4 Finally, perceived similarities in shape figure prominently in certain extensions of Cora locationals. With the preceeding considerations in mind, I now turn to an examination of the data, selected from an almost bewildering array of spatials.5 The various morphological sets of Cora locationals appear to be fully productive on several accounts. Simple morphological units combine with others to form more complex ones. For example, a set of slope-oriented locative particles, themselves complex morphologically, combine in various ways with a set of topographical suffixes to derive a set of sixty-three topographic adverbs. This derived set constitutes a complex, but highly flexible system for mapping out relative positions within the domain of the lay of the land (Casad 1982a: 135-215). Cora also has a set of seventeen prefixes of location and direction. These attach to verbs either as a single prefix or in strings of two or three prefixes. The prefixes and combinations specify schematically the spatial framework within which a process unfolds or a state exists (ibid.:221). The prefixes of location combine with one another in numerous overlapping sets. There are at least sixty-seven prefix combinations, each of which can occur with an indeterminate number of verb, adjective or noun stems. Most verb stems can take one or more prefix combinations; some stems occur with up to thirty such combinations. Each distinct prefix combination leads to a semantically distinct global meaning for the prefix combination + verb stem sequence (Casad 1982b). There are limits, however, on all this seeming productivity. Sometimes the morphological sets are defective; i.e. not all the possible combinations occur. In such cases, a form that is conventionally restricted in one way car-
348 EUGENE H. CASAD ries over into another context in which the use of a different morpheme would seem appropriate. In other cases, distinct verbs can be used to describe analogous situations. However, the prefix combination used with one stem is not used with the other. Sometimes the nature of the stem itself determines which prefix sequence will occur. Linguistic conventionalization thus results in the nonpredictability of both the prefix sequence and the stem. Finally, the non-prototypical extension of the meaning of a prefix sequence may carry over into novel situations. In the rest of this paper, I present examples of both the locative particles and the prefix sequences to illustrate each of these limitations on the productivity of Cora locationals. 2.0 Conventional Morphology The morphological sets of slope-oriented locative particles and topographic suffixes provide a clear example of how the conventionalization of linguistic forms limits the extent to which members of one morphological set can combine with those of another. There are five morphological sets of slope-oriented particles. As (1) shows, each set consists of three members. They are all morphemically complex, having a three-fold left-to-right organization. This organization reflects distance, boundary, and slope predications, respectively. The canonical interpretation of the distance predication is as follows: Proximal distance, marked by y, corresponds to the speaker's position. Medial distance, marked by m, corresponds to the addressee's position, whereas distal, unmarked, signals a location distinct from either the speaker's or addressee's. The boundary orientation places an entity either inside of a bounded area or outside of it. The inside location is marked by w, and the outside location is marked by a. The slope orientation is symbolized in one of three ways: "foot of the slope" is marked by a long u or a vowel; "in the face of the slope" is marked by syllable-final h; and "head of the slope" is marked by syllable-final n. (1) Foot of Slope In Slope Head of Slope Inside Outside Inside Outside Outside here y-u-u y-a-a y-u-h y-a-h y-a-n there m-u-u m-a-a m-u-h m-a-h m-a-n off there u-u a-a u-h a-h a-n Table 1
CONVENTIONALIZATION OF CORA LOCATIONALS 349 As can be seen in (1), the slope-oriented locative particles are further organized into paired "inside-outside" morphological sets. The "head of the slope" is the exception; it contains only a single "outside" series. There are no "inside" forms *yun, *mun or *un. Instead, "head of the slope" locations are conventionally construed as involving an external perspective. In the canonical situation, shown in Figure 1, the speaker as viewer(S) is at the foot of the slope; his line of sight extends from his own position and follows the vertical axis of the slope, ending at the horizon. As a natural restriction it does not curve around to cover the head of the slope (Casad 1982a: 101). This restricted view point area is shown imposed on a generalized slope base in Figure 1. This base is the necessary context that defines the distinct gradient areas as well as the line of sight bounded area. Figure 1. Slope as Base for Line of Sight. The slope-oriented particles of (1) combine with the set of six topographic suffixes of (2). This yields a set of sixty-three topographic adverbs that specify the location of an entity somewhere within the broad lay of the land. (2) (a) (b) -tye "downstream" -ki "downhill" -tyapwa "upstream" -tyi(vi) "uphill" -ta(va)n "cross stream" -ka'i(wa) "over hill" Table 2 The topographic suffixes take as their points of reference the flow of a river and the gradient of the land surface. Both of these concepts presuppose a context in which the speaker as viewer also figures as a reference point. There are salient restrictions on which slope-oriented particles can
350 EUGENE H. CASAD combine with particular topographic suffixes. These restrictions are summarized in (3). The slope-oriented categories occur down the left side of Table 3, whereas the topographic suffixes appear along the top. The combination of a particle and a suffix is indicated by an A" in a cell of Table 3. An asterisk, then, indicates that a given particle does not combine with the suffix shown at the top of a column. Two nearly full rows of asterisks highlight the absence of a "head"-"inside" parameter as well as the absence of the basic yu and ya series forms of the locative particle in most categories.6 Every instance of an X in Table 3 represents a set of three topographic adverbs built on a combination of a particle and a suffix. Each asterisk represents a possible, but non-occurring combination. All non-occurrences, except for those with simple u and a, are based on perceptual factors relevant to the lay of the land. (3) -t*e -ki -tyi -t*apwa -tan -ka'i ya yu yaa yuu yah yuh yan yun (-slope) (foot) (face) (head) * * X X X * X * * * X X X * X * * * * * X X X * * * * * X X X * X X * * X * X * * * * * X X X * Table 3 The morpheme uh, whether it is used as a free particle or as a locative prefix sequence, typically designates movement along a vertical path that takes its departure from a horizontal base. When applied to the lay of the land, it designates positions uphill from the speaker. Thus uh is semanti- cally incompatible with -tye "downstream" and -ki "downhill". Both of the areas these two suffixes designate tend to be below the speaker's location. On the other hand, aa "off out in the flat" and uu "there in an enclosed area of the flat" refer to locations below the speaker's level while the contour of the land implied by -tyi "uphill", -tyapwa "upstream" and -ka'i "over the hill" is generally upward going and is above the speaker's canonical viewing position. Thus the meanings of aa and uu are semantically incompatible with these three suffixes. The notion of slope gradient is only loosely connected with the mean-
CONVENTIONALIZATION OF CORA LOCATIONALS 351 ing of the -ta(va)n adverbs. This is partly because -ta(va)n simply relates two polar points. This allows -ta(va)n to occur with the short-voweled basic u and a particles. This also at least partly accounts for why there are no -ta(va)n adverbs built on uu, aa or uh. The upper profile of the banks of a river is often prominent enough that the land on the other side is perceived as an "on top" location. This perception of the scene is apparently contradictory to the meaning that speakers would attach to a potential form *uhtavan. The latter form, were it to be interpretable at all, would have to designate a path that leads upward from some unseen place off on the far side of the river. In short, uh seems to contribute overly specific and irrelevant information to what is needed to characterize -ta(va)n locations. The specialized meanings of u and a, when they combine with -ta(va)n reinforce the idea that an "upwards going" path is irrelevant to the meaning of -ta(va)n. Thus, in many cases, utavan locations are outside the speaker's field of vision, whereas atavan locations are visually accessible to the speaker (Casad 1982a: 163). This is depicted diagrammatically in Figure 2. The canonical viewing position(CV) is at one side of a solid line which represents a barrier that delimits the viewer's field of vision. The dotted arrows are a shorthand representation of the extent covered by the observer's visual field. The area on the viewer's side of the line is therefore an a area, whereas that on the far side of the line away from the viewer is an u area. /u/ Figure 2. "Inside" and "Outside" with -tavan. The non-occurrence of uu and aa with -ta(va)n can be accounted for in similar terms. Essentially, the non-occurrence of uu-, aa-, and uh- is motivated by both perceptual and morphological considerations and is thus multiply motivated. The relative positions of the speaker and his topographical reference points are summarized in Figure 3, 'The Topographic Base'. This pictorial representation attempts to highlight the perceptual factors mentioned above that relate to the lay of the land.
352 EUGENE H. CASAD Figure 3. The Topographic Base. Key to Figure 3 -tye: downriver from the Speaker. -tyi(vi): uphill from the Speaker. -tyapwa\ upriver from the Speaker. ki: downhill from the Speaker. -tavan: across the river from the Speaker. -ka'i(wa): over the hill from the Speaker. The conventionalization of "head of the slope" locations as necessarily external areas carries over into various situations where one might otherwise expect an internal *un form. For example, uh, ah and an all combine with the suffix -ka'i "over the hill" to indicate distinct locations at or behind the horizon or hill. Essentially, locations at the side of a hill, where one's range of perception ends, are ahka'i locations (external), whereas those directly behind the hill are uhka'i (internal) ones. Obviously, there can be additional knolls and ridges with "on top" locations of their own that are hidden by the mass of the hill the speaker sees. This potentially allows for an "up on top over behind the hill" location to be construed as an inside one, e.g., *unka'i. However, the potential for forming a topographic adverb is limited to the morphological sets of locative particles that actually do combine with the topographic suffixes. Since the "head of the slope" locative particles conventionally exclude "interior" forms, the potential for "interior" forms of "head of the slope-over the hill" topographic adverbs is also automatically precluded; i.e., the conventionalization of the an series of locative particles carries over into all of its combinations with other mor-
CONVENTIONALIZATION OF CORA LOCATIONALS 353 phemes, as is indicated in Table 3. 3.0 Choice of Verbal Prefixes Conventionalization also results in cases in which one might expect a particular prefix sequence to occur with a given stem for describing an event that takes place in a particular domain. This expectation would be based on the observed use of that prefix sequence with another stem that describes an analogous event that takes place in that domain. I illustrate this from the domains of the tree, a burning candle, things falling to the ground, the human body, astronomy, and time. 3.1 The Tree Trunk The prefix sequence aiitya- "in the middle of a surface" conventionally designates the trunk of a tree. (Other parts of the tree are designated by other prefix sequences.) Examples (3) (a and b) illustrate this typical use of aiitya-P (3) (a) a- ii-ty a-uh-tutah i kiye outside-path-middle-REFL-curve ART tree "The tree trunk curves back upon itself". (b) kiye nu a-ii-ty aa-tu} a pole I outside-path-middle-hit "I struck a pole (with my machete)". In (3)(b) the speaker asserts that he struck the side of an old rotten log with a machete. One might expect that were he to go on to say that he cut it in two, using the stem veihce, he would use (4)(a). It turns out, however, that he actually would use the form given in (4)(b), which employs the prefix sequence antyi "up at the highest point". (4) (a) nya-r-a-'i-tya-veihca-kaya I-it-outside-path-middle-cut-PERF *"I cut the log in two", (b) nya-ra-'an-tyi-veihcha-ka'a I-it-top-up-cut-PERF "I cut the log in two". The use of antyi- "up at the highest point" may be rather surprising since the speaker seems to be in the same position with respect to the log regardless of whether he is simply chopping at it or is chopping it in two.
354 EUGENE H. CASAD However, notice that the meaning of (4)(a) is starred, but the form itself is not. (4)(a) is a perfectly well-formed Cora word; it simply means either "I split the log right down the middle" or "I struck the log along its side". The conventional meaning of nyara'ityaveihdaka'a, then, in part, forces the selection of an unexpected prefix in referring to the chopping of the log across its axis. Other factors also influence the choice of antyi- in this situation. For example, one other prefix sequence, anta- "on top-across", seems appropriate to use for the situation we are describing. Sentence (5) shows that anta- can designate the cross-section of the stem by which an orange attaches to a tree. (5) ra-'an-ta-huura-ka'a i maraanka it-top-across-pick-PERF ART orange "He picked the orange". However, there is a rather distinct perceptual difference between the result of plucking an orange and the chopping of a log in two: the plucking frequently leaves a clean break straight across the stem, whereas the chopping usually leaves a fairly pointed configuration at the end of the log. This is especially true if the log is somewhat large and must be turned over so that the woodsman can chop at it from all sides. The result of this cutting is thus a prominent fresh-cut area approximately conical in shape. This makes it natural to use a non-vertically-oriented version of antyi- in the situation being described. 3.2 All Burned Up A burning candle is typically placed vertically with the zone of burning located at its highest point. It is natural to view the tip of the candle as being analogous to the tip of a standing tree, the ridge of a gabled roof, or the top of a mountain peak. The prefix sequence antyi- "up at the highest point" is used to designate all these locations, as examples (6)(a - c) show. (6) (a) an-tyi-nyeeri-'i i cVi hapwa top-up-visible-STAT ART house on "It is all lit up there at the top of the house". (b) an-tyi-tyee an-na in hapwa top-up-long top-SG hill on "It's a long way up to the top of the hill". (c) na-'an-tyi-raa i kiye hece I-top-up-go:PERF ART tree in "I climbed to the top of the tree".
CONVENTIONALIZATION OF CORA LOCATIONALS 355 However, as (7)(a and b) together show, antyitaasin is not the appropriate form for speaking of the prototypical position of a burning candle; instead, the appropriate form includes the prefix sequence uka- "downwards through the interior". (7) (a) an-tyi'-taa-sin i kantiira top-up-burn-DUR ART candle "*The candle is burning" (oriented vertically), (b) u-ka-taa-sin i kantiira inside-down-burn-DUR ART candle "The candle is burning". Although a candle is typically positioned vertically, and the flame is situated at its uppermost point, Cora speakers conventionally select two other aspects of that scene as the basis for their assertion that the candle is burning. In the first place, the burning originates at the candle wick, which is in the middle of the body of the candle and parallels its vertical axis. Secondly, it proceeds downward along its vertical extension. Thus the appropriate prefix sequence is u-ka~ "inside-downwards", as in (7)(b). The choice of appropriate locative prefixes may also depend on non- prototypical extensions of the basic meanings of these prefix sequences. In turn, such extensions carry over into novel situations. As noted above, antyi- typically refers to the upper extremities of vertically-oriented entities such as houses, steeples, hills, trees, and flagpoles. The notion of verticality may fade out of the picture completely and allow antyi- to apply to prominent projections of any sort, such as the tip of a pencil, the fingertips, the tines of a fork, or the front end of a vehicle, as in (8)(a - d).8 This extension of antyi- to horizontally-oriented projections leads to the acceptability of (7)(a), but gives it a distinct meaning. Whereas antyitaasin does not refer to a candle in its normal standing position, as the gloss of (7)(a) suggests above, it does refer to a burning candle lying on its side. (8) (a) an-tyi'-pityi i laapi top-up-pointed ART pencil "The pencil is pointed". (b) na-'an-Pi-suhca'arii y-en nya-sitye hece I-top-up-split open here-top my-finger at "The end of my finger is split open". (c) an-tyi-pityi-hmee i tenidoori top-up-pointed-COLL ART fork "The fork has several points on it".
356 EUGENE H. CASAD (d) mwa''an-ty-uh-tu'a i kaarru they-top-up-REFL-hit ART auto "The cars collided head on". To complicate the matter somewhat, prefix sequences other than antyi- would also seem to be plausible candidates for applying to this situation. For example, both a-uu- "distal-that way" and wa-ta- "extensive-straight ahead" designate essentially horizontally oriented paths, among other things. Either auutaasin or watataasin seem more appropriate than antyi- for talking about the burning candle lying on its side. However, auutaasin is conventionally taken to designate burning that occurs off on the horizon. On the other hand, watataasin appears to be a nonce form.9 The choice of antyitaasin, then, is likely to be perceptually based: when a candle tips over, the wax at its tip melts and forms a pool into which sooner or later the wick falls and the flame goes out. Usually this is confined wholly to the area relatively close to the candle tip. Figure 4 illustrates diagrammatically the path that the semantic extension of antyi- takes in acquiring the meaning "at the tip of an elongated object". The prototypical meaning of antyi- "at the highest point of" is seen in 4(a). The base includes a schematic verticality scale composed of a horizontal baseline and an arrow directed vertically upwards. The triangle represents the mass of a vertically oriented body. The horizontal broken line marks the indefinitely defined boundary between the upper and lower halves of the entity under consideration. Finally, the precise antyi- location is indicated by the boldface circle at the apex of the triangle. The fading out of aspects of the image associated with prototypical instances of antyi- is shown in Figure 4(b). Broken lines are used for the verticality scale and the bottom half of the vertically oriented entity. The final stage of the extension is represented by an elongated horizontal mass which has a pronounced narrowing to a point at one end. No notion of verticality whatsoever is present in this version of antyi~. Instead, the entire narrowed area is construed as the appropriate antyi- location. Both antyitaasin, and nyara'ityaveihcaka'a (Sec. 3.1) thus show how the conventionalization of linguistic forms leads to an unexpected, but not arbitrary, choice of locative prefixes in situations that are seemingly analogous to some other one.
CONVENTIONALIZATION OF CORA LOCATIONALS 357 (a) (b) A- (c) UL .-.,-> Figure 4. Extensions of antyi~. 3.3 Down to the Ground One might expect that things falling to the ground would tend to be viewed the same way. Thus, (9)(a and b) show that snow and discrete objects follow downward paths which are designated by the prefix sequence aha- "downward to an external surface". (9) (a) m-a-ka-h-vice mi nieve they-outside-down-slope-fall ART snow "The snow is falling", (b) a-kaa-vaci i tyetye hece outside-down-fall ART rock from "It fell to the ground from off the rock wall". From this, one might predict that the downward path of a discrete object falling into a fire on the ground would also be conventionally symbolized by aka-, possibly even by the same form that occurs in (9)(b). Yet (10)(a - c) show that such is not the case. Rather, the prefix sequences ana "off at the periphery", and wa- "Extensive(EXT)" occur with vaci "to fall". The difference is that the path relevant to anaavaci is a horizontal one, i.e., the wind pushes a piece of paper along the ground until it crosses the edge of the burning area. On the other hand, a vertically oriented downward path is part of the situation described by wavaci (10c). The selection of wa- over aka- may be linked to the use of the medial form of the locative particle. In particular, the speaker's view comprehends simultaneously both the starting point and the end of the path of the falling entity. The duration of the fall is quite short, making it natural for the speaker to characterize the entire situation wholistically.
358 EUGENE H. CASAD (10) (a) *a-kda-vaci i tdih hece outside-down-fall ART fire to "*It fell into the fire". (b) a-nda-vaci i tdih hece outside-perimeter-fall ART fire to "It fell into the fire". (c) m-an pu wa-vdci i tdih hece MED-top SUBJ EXT-fall ART fire in "It fell into the fire from right up there on top". To summarize, in one pair of situations that seem analogous to one another, the same verb stem and postpositional phrase is used for the sentences that refer to the scene being described. However, the appropriate form appears with unexpected prefix sequences in two of the sentences. Linguistic conventionalization is behind all this: notice that the conventionalization that rules out the use of a particular morpheme or morpheme sequence does not in itself determine what the appropriate choice will be. A slightly different scene will bring an additional factor to an already complex picture. Here I consider the scene in which someone forcibly propels an object to the ground. As (ll)(a and b) show, aka- does not even figure in either sentence. (11) (a) ti'iki dya-'a-rd-hraii aihna i takwaci CNJ there it-outside-face-throw DEM ART ball "And then he threw his ball to the ground". (b) ti'i niiu i kin tyi-ra-'a-vd'a-tu'a CNJ QUOT SEQ with DISTR-it-outside-covering-hit aihna i pa'ari'i i ru-takwdci kime'e DEM ART child ART REFL-ball with "And then that boy struck the ground good and hard with his ball". (ll)(a and b) both relate to the same scene. They are taken from the same text and refer to the same person and the same ball. In this case, neither choice of prefix sequence used with one verb stem forces the particular selection made for the other one. As Langacker points out, people can construe particular scenes in various ways (1983:11/15). The freedom to choose which details of a scene to discuss, the ways in which particular components of the scene can be characterized and the kinds of relationships that hold
CONVENTIONALIZATION OF CORA LOCATIONALS 359 between the components of a given scene all interact to determine the appropriate prefix choice (cf. Casad and Langacker 1985:280). Thus, in (ll)(a), the path that the ball follows is conceived to originate at the boy's location and to follow the natural direction that he faces as he propells the ball away from his own position. This interpretation is substantiated by the meaning of raii "to cast away". The object nominal aihna i takwaci "that ball", in (ll)(a), highlights the ball, whereas ara- designates both the origin of the ball's path and its directionality based on the boy's natural motion. The use of ara- in (ll)(a) is similar to the use of ra- to designate the notion "downstream"; i.e., ra- "downstream" indicates the natural direction of the flow of water (Casad 1982a:307). The verb stem tu'a "to hit" refers saliently to contact between a pair of entities. The contact point is construed as occurring within an upper, visually accessible surface. Typical surfaces that are treated as both upper and visually accessible include the ground, a table top, a ceiling, the top of one's head, and even the location of a layer of clouds in the sky. To express the notion of squashing something with one's foot, for example, one would say "nyara'ave'ecF. Thus, avaa- is selected as the appropriate sequence (Casad 1982a:296). The instrumental phrase of (ll)(b) highlights the point of contact between the ball and the ground at the end of the ball's path. To close, these prefix sequences highlight the two most salient points of a directed path (cf. Miller and Johnson-Laird 1976:405). The contrast in the mental imagery lying behind these two usages is striking when viewed pictorially. In both cases they reflect extensions from prototypical variants. For both, the base for defining the relationships that they designate is a three dimensional solid. For ra-, this solid has one of its vertical surfaces profiled in boldface. This surface is construed as its "face". The bottom surface of the solid is extended outwards from the foot of the highlighted surface to give the base a front and back orientation. This is shown in Figure 5(a). The "face" can be construed as the point of origin for either a line of sight or a path along which something physically moves. The former is depicted in Figure 5(b), whereas the latter is depicted in 5(c). For ease in presentation, I represent the "face" of the base solid as a single vertical line in both 5(b) and 5(c). The only difference between the two is that in 5(b) the line of sight is indicated by a broken arrow, but in 5(c) the path for motion is indicated by a solid one. Both versions of ra- are compatible with the meaning oiya'arahrai. The semantic contribution of the verb stem, then, is partially to superimpose the notion of a real path on a scene that
360 EUGENE H. CASAD also includes a virtual path, (a) (b) (c) ^ hh Figure 5. Varieties of ra-. For va'a-, the horizontal top surface of its base is profiled, and the vantage point(VP) from which the whole is perceived is overtly indicated (Figure 6(a)). The use of va'a- in tyira'ava'atu'a represents a fading out of the vertical extensiveness of the base and a fading out of the boundedness of its horizontal upper surface (Figure 6(b)). Thus, va'a- comes to designate any flat surface whatsoever. The semantic role of the verb stem -tu'a9 then, is to bring a downward path into contact with that flat surface, as seen in Figure 6(c). (a) (b) A^7 Figure 6. Extensions of va'a-.
CONVENTIONALIZATION OF CORA LOCATIONALS 361 3.4 Bottles, Sticks and Pots Another glimpse of the freedom to structure a scene is afforded by observing a bottle cap that is cross-threaded onto its container. Grounds for using a phrase employing antyi- appears to be garnered by noticing the use of anfi- with the stem -kuuna "to open" in (12)(a). Nevertheless, as (12)(b) suggests, antyi- does not figure in describing the cross-threaded dilemma. (12) (a) pa-ra-'an-tyi'-kuuna i limeeta you-it-top-up-open ART bottle "Open the bottle". (b) *cere'e pit an-tyi-yeihsi limeeta hapwa oblique SUBJ top-up-sit bottle on *"The cap is cross-threaded onto the bottle". (c) cere'e pu u-ty a-muuku'u oblique SUBJ inside-middle-hat "The cap is cross-threaded onto the bottle". The appropriate phrase, given in (12)(c), shows again that Cora speakers can ignore the objective position of an entity at the uppermost part of the entity that it is attached to. Furthermore, they view the bottle cap not as having the shape of a stopper (itu'upuari), but rather as having that of a "hat" (muuku'uci). Because of this, the prefix sequence is determined from the way a hat is placed on one's head. Essentially, the human head fits within the crown of the hat (an "inside" location). The crown itself is situated within the horizontal expanse of the brim of the hat (an "in-the- middle" location). In short, once the bottle cap is viewed as being a "hat", then the fit of the bottle into the cap is viewed as being analogous to the way the human head fits into a broad-brimmed hat.10 The choice of stem, therefore, determines what the prefix sequence shall be. Shape specifications partly determine prefix selection in describing a stick that is placed horizontally crosswise at the top of a pot or bucket. As (13)(a and b) show, the prefix sequence wata- is commonly employed to indicate the horizontal crossing of a bounded area. In (13)(a), the bounded area is delimited by the opposite walls of a room, whereas in (13)(b) the space between two poles laid parallel to one another on the ground determines this. These two situations are represented diagrammatically in Figures 7(a) and 7(b).
362 EUGENE H. CASAD (13) (a) alaampi hece ti m-an wa-ta-vi cable on SUBR MED-top EXT-straight-hang "on that wire, which is stretched from side to side right up overhead" (b) tya-taa-sin i kiye ti m-an middle-burn-DUR ART pole SUBR MED-top wa-ta-kaa EXT-straight-lay:SG "The pole that is laid flat across on top of the other two is burning". The area marked off by the rim of a pot is clearly a bounded area. In addition, the pot in use is normally set upright so that the area circumscribed by its rim is roughly horizontal. Thus it would seem reasonable to use the word wataka'a "it is laid flat across it" to talk about the stick placed across the top of the pot,as in Figure 7(c). (b) Figure 7. wata- vs. anka- The similarity between the situations depicted by 7(b) and 7(c) is clear. Yet (14) shows that the proper sequence in 7(c) is anka- "going downwards from the top". The stem -ka'a "to lay stretched out" is still appropriate. (14) siicira'a i kiye ti m-an an-kaa-kaa straight ART pole SUBR MED-top top-down-lay:SG
CONVENTIONALIZATION OF CORA LOCATIONALS 363 i saari hapwa ART pot on "The pole which is laid flat across the top of the pot is long and straight". The use of an- "on top" in (14) is straightforward; the overall position of the stick is at the top of the pot. Since the stick is laid out horizontally across the rim of the pot, the use of ka- "down" seems anomalous. It turns out that the sequence anka- has a conventionalized usage with respect to the human body that helps unravel this puzzle. In the domain of the human body, anka- designates the mouth, as in (15). (15) na-'an-kaa-tu'a y-en nye-ty mye-e me-top-down-hit here-top my-mouth-in "He hit me right here in the mouth". The use of anka- to designate the mouth seems to be based on the perceptual salience of the mouth, the normal standing position of the human creature, and the downward path followed by objects being placed in the mouth. With these factors in mind, the transfer of the notion "mouth" to entities like cooking pots is fairly transparent (cf. Friedrich 1971:33-35). Note also that the salience of the "mouth" of the pot overrides the objective horizontal extension of the stick. In short, (15) can just as well be glossed "The stick laying across the mouth of the pot is long and slender." 3.5 On the Edge of Things Alternate perspectives are also evident from the way two different prefix sequences apply to the edge of cutting instruments. As (16)(a and b) show, both anta- "straight along the top" and aiira'a- "back and forth along a path" designate location along the cutting edge of a knife or machete blade. (16) (a) an-ta-nyetyi i nawaa top-straight-dull ART knife "The knife is dull", (b) a-ii-rci a-tunye i cun outside-path-corner-doubled ART machete "The machete blade is all knicked up". In (16)(a), dullness is viewed as extending uniformly along the cutting edge of the knife. The "on topness" of this edge possibly comes from the
364 EUGENE H. CASAD typical way one turns a knife cutting edge upwards in order to gauge its sharpness manually. The distinct straightness of the knife edge all along the length of the blade is sufficient grounds for selecting ta-. On the other hand, the knicks in the machete blade are probably not equally distributed all along the machete's cutting edge, but are rather probably unequal in size and irregularly distributed along that edge. They may well be more densely congregated along one particular region of it, however. Both anta- and ail- appear to designate identical paths along an instrument's cutting edge; however, anta- apparently summarizes the entire unbroken path (Figure 8(a)), whereas ail- is more schematic in this regard; i.e., aii- can designate either an entire path or a single point along it. The prefix raa-, for its role, designates subsidiary "back and forth" or "side to side" paths that run at right angles to the cutting edge, as in Figure 8(b). (a) (b) .& •$& Figure 8. Two Views of an Edge. The semantic role of the individual verb stems is to make explicit what the contrast is all about. Thus, "dullness" extends more or less uniformly throughout a normally incise configuration, whereas "doubling-up" refers to a more well-defined and restricted configuration. In short, another aspect of the conventionalization of locative prefix uses is the freedom to see the essentially same configuration from two different perspectives (cf. Shepard 1978:130). Cora terminology for sharpening work tools reinforces this latter point. As one might expect, the prefix sequence anta- "straight along the top" is appropriate for talking about sharpening a knife. In this situation, the stem mumite'e is employed (17)(a). It seems plausible to think that the Cora would talk about sharpening machetes in the same terms as they use to discuss the sharpening of knives. Although both anta- and mumitee are appropriate to the task, that is not the whole story. The prefix ka- can turn up on a different stem -ce, which means "to move" (17)(b).
CONVENTIONALIZATION OF CORA LOCATIONALS 365 (17) (a) r^a-ra-'an-ta-miimi-te-'e i nawaa I-it-top-straight-RDP-edge-CAUS-APPLIC ART knife "I am sharpening the knife", (b) ny-au-ce'e nye-£im kaa-ce I-LOC BASE-CONT my-machete down-move "I'm still sharpening my machete". The explanation for this is not a syntactic one at all. Instead, it is bound up in the features that people focus on in the conceptual scene. In (17)(a), the speaker's attention is placed on the shape of the knife blade itself, as in Figure 8(a). In (17)(b), on the other hand, the speaker's focus is placed on the interaction of the machete blade and the moistened grinding stone surface across whose flat surface he moves the machete. The directional ka-y therefore, refers to the area of contact between the machete blade and the grinding stone.11 3.6 High in the Sky Predicting the Cora phrase that describes a ring around the sun or moon has its pitfalls also. There are at least four prefix sequences that can be roughly glossed "going all the way around". The first is kura'a- "around- corner". In (18) it places a set of origin points along a virtual path that circles the speaker's location, which is the simultaneous goal point of all the paths implied by the verb stem -kih "to leave, to go". (18) ma-ku-rtia-kih-si they-around-corner-leave-PAST:DISTR "They came from all directions".12 In (19), the prefix sequence an-ku-ra'a- "top-around- corner" designates a path cut all the way around a tree trunk. (19) ra-yan-ku-reye-ve\ca-hraa it-top-around-corner-cut-PAST "He chopped a ring around the girth of the tree". In (20), the prefix sequence ah-ku-raya- "slope-around-corner" traces a path that starts at the speaker's side of a hill and goes past the edge of the hill to a distal point out of sight. (20) a-h-ku-rcCa-raa ah-kcCi in hece outside-slope-around-corner-go slope-overhill hill at "He went off over the edge of the hill".
366 EUGENE H. CASAD As (21) shows, both a-na- "outside-at periphery" and a-ta- "outside- straight" bear some relation to a path that follows the edge of a bounded area. (21) a-na= 'a-ta-ye'i vi'ira'a-ta'a outside-perimeter=outside-straight-walk field-in "He was walking around the edge of the corn patch". By itself, ata- can designate a path that goes clear around the edge of a bounded area. (22) illustrates the use of ata- to designate the location of a row of nails that hold the rubber soles and leather insole together on a pair of sandals. (22) a-td-ta-taii-wa mi klaavu outside-straight-RDP-nail-PASS ART nail "The nails are nailed down in a row running all along the edge (of a sandal)". Finally, a nominal form tutah can be prefixed by ata-. The combination is conventionally taken to mean that someone has a ring around his head, for example, a paper crown from a fast food establishment, as in (23). (23) a-ta-tutah miiu-raan hece outside-straight-ring head-his at "There is a ring around his head". The examples above, then, suggest that any of the forms such as kura'atutah, ankura'atutah, ahkura'atutah, {l)ana atatutah or anaatutah might be appropriate for describing a ring around the sun. As (24) shows, however, none of the above prefix sequences nor the nominal tutah are used in this way. Instead, the proper verb stem is -vi "to hang, to be strung out", the locative prefix sequence is ata-, and the proper nominal is the subject noun kuusai "halo, rainbow". (24) ii pwa,a kuusai a-ta-vih-ka SUBR COND halo outside-across-hang-HAB * sika hece ART sun at "If there is a halo (ring) around the sun..." The choice of locative prefixes in this case is only partly determined by overt perceptual factors. The nominal kuusai is the personification of a mythological entity to the Coras and the conceived spatial relationships are at least partly based on the Coras' view of ritual space in both the earthly
CONVENTIONALIZATION OF CORA LOCATIONALS 367 and heavenly spheres.13 The ring around the sun or moon, for instance, is mirrored in the ritual by the counterclockwise path the dancers follow around a fire in the middle of the dance area. 3.7 Comparative Anatomy The locative prefix sequences also have conventionally established meanings when they designate areas on the human body. Taken together, they form a mapping of the entire body area that is both exhaustive and discrete. Crucially, these locative prefix sequences are not homophonous with or even remotely related phonologically to the nominal forms that serve as names of the body parts. In fact, the relative mappings provided by both locative prefixes and body part nominals are not congruent.14 By saying that the prefix sequences have conventionally established meanings in the domain of the human body, I intend to convey two notions. First, adjective and verb stems that refer to states and events localized at a particular body part area are typically marked by the same prefix sequence. Second, these typical usages are shared by at least most of the members of Cora society. It turns out, however, that the typical prefix marking is not invariant. In the following paragraphs, I examine two cases of non-prototypical marking. These relate to the prefixes that designate the midriff and the feet. 3.7.1 On Midriffs The prefix sequence a-uu-(ta-) "outside-that way-(straight)" can be glossed "going off horizontally from edge to edge" (Casad 1982a:315). In the domain of the human body, auu(ta)- takes on the meaning "at the midriff", as well as "going clear around a joint" (ibw/.:388).15 (25)(a and b) illustrate the use of auu(ta)- as it applies to the midsection.16 (25) (a) a-uu-pi-pua outside-that way-RDP-skinny "He is skinny at the waistline", (b) n-a-'u-ta-n-tu'uny i-ty e-'e-sin I-out-away-straight-REFL-pants-CAUS-APP-DUR "I'm going to put on my pants". The expressions that refer to a male's ample abdomen do not employ auu(ta)-. One involves the incorporated noun huka "stomach", the other
368 EUGENE H. CASAD makes use of saye "entrails". In both cases, (26)(a and b), the speaker selects the prefix sequence ana- "at the periphery" for conveying a meaning functionally equivalent to English "he has a pot-belly". (26) (a) a-nd-huka outside-perimeter-stomach "He is pot-bellied", (b) a-nd-saye outside-perimeter-entrails "He really has a gut on him". The choice of ana- in (26)(a and b) is partly forced by the fact that the form auutahuka already exists. It is conventionally construed as meaning "she is pregnant". A rather more salacious version auutapu'uta "she is all puffed up in the middle" reinforces the association of auuta- with pregnancy. Although the conventional meaning of auutahuka forces some other prefix choice for conveying the notion "pot-bellied", that does not of itself determine what the particular choice will be. In this case, anahuka is probably partly motivated by the potential the speaker has for shifting his perspective or vantage point for viewing a scene (cf. Miller and Johnson- Laird on the salience of one's vantage point, 1976:228). ana- typically marks position at the perimeter of a bounded area or at the front of it. Since a person's stomach is at his frontside, a particularly salient stomach configuration is easily construed as being out in front of him. The selection of ana- is thus also seen to be multiply motivated. 3.7.2 On Feet Two prefix sequences typically designate areas on the foot. The more general one is ana- "outside-at the periphery". It designates the top of the foot (and hand) and the lower leg below the knee, as in (27a). The other one is a-h-ta- "outside-slope-straight", which designates the sole of the foot, as in (27b). (27) (a) ru-mwayati mu a-na-tapV-iwah-me-n REFL-pure they outside-perimeter-tie-PASS-PERIP-PRTC wa-ka'akai EXT-sandal "To a man, they all go around wearing sandals with ties anchored at three places".
CONVENTIONALIZATION OF CORA LOCATIONALS 369 (b) heiwa pit a'ih pwa'a a-h-ta-'in lots SUBJ DEM bad outside-slope-straight-smear i ru-iika hece ainyi dpi pii ART REFL-foot on since papaya SUBJ a-ve'e-ci ti petyi outside-covering-squash SUBR rotten "The bottom of his foot is all mucky because he stepped on a rotten papaya". Examples to corroborate this distinction between the top and bottom of the foot are fairly common. Apparent counterexamples are not difficult to find either, as (28)(a and b) demonstrate. (28) (a) hayi-tye yauh-mwa'a pit na-'a-na-waa ant-PL son-PL SUBJ me-outside-perimeter-rub i nye-capwaace-ye i nyi-taata ART my-sole of foot ART my-father "My father rubbed ant larvae on my feet", (b) u-'u-tya-haa-ni nii'u ru-iika hece NARR-inside-middle-swell-FUT QUOT REFL-foot at "He says that his foot is going to swell up". (28b) presents a third prefix sequence, utya- "inside-middle", that also designates a location in the foot area. This includes at least part of the lower leg. The role of u is to pinpoint the interior of the leg as the source of swelling. On the other hand, tya- pins it to a perceptible surface. Finally, the phrase ruiika designates all of the foot and lower leg.17 (28a) actually does follow the pattern. The rubbing is specifically located on the top of the feet. Thus, capwari encompasses both the top and sole of the foot, and ana- and ahta- divide up the area capwari designates. 3.8 A Trip into Time Cora, like many other languages, takes elements with spatial meanings and extends their usages into the temporal domain.18 These elements include both prefix sequences and adjective stems. The examples in (29) give the prefix sequence auu-, the stem -tyee "spatially extended" and -titfi "physically elongated" in their strictly spatial usages. (29) (a) a a-uu-tyee a-h-ka'i there distal-that way-long outside-siope-overhill "It's a long way off to the horizon".
370 EUGENE H. CASAD (b) kih pu tyee i kantiira little SUBJ long ART candle "The candle is short and stubby", [lit. "it is a little bit short".] (c) eekan pu titi'i i kauhna-ri lots SUBJ elongated ART rope "The rope is really long". The prefix sequence auu- is also used to express the notion of time going along, which is clear from (30). This use of auu- represents a path of extension from a strictly spatial meaning of "all over a flat surface" to "all along a horizontal path in space" to "throughout a period of time". In its meaning "all over a flat surface", auu- overlaps with wa- "Extensive" (Figure 9(a)). It has come to designate many surfaces that are elongated along one dimension (Figure 9(b)). Thus, an implicit path orientation has become associated with many of its usages. This notion of path has become salient enough to the meaning of auu- that it is now used to measure distance, as sentence 29(a) shows. This path version of auu- is given in Figure 9(c). The final extension of auu-, then, is its extension into the domain of time, illustrated by (30). (a) A m (b) y ■ Figure 9. Extensions of auu-. (30) da pu'u i-(y)-a-uu-ma-'a-ka there SUBJiPAUS NARR-DIST-that way-go-IMPERF-PRTC ami sika five day "And so, just as the five day limit was approaching..."
CONVENTIONALIZATION OF CORA LOCATIONALS 371 The temporal usages of -tyee and titi'i do not employ auu-. Instead they are derived by combining with a'- "area".19 (31) (a) *aacii a-uu-ty ee-vi-i... somewhat distal-that way-long-PRTC * "a little while later..." (b) ctatii aya-tyee-vVi... somewhat area-long-PRTC "a little while later ..." (c) aacii-nyi a'a-tih-tii i avion ti QNT-Q area-RDP-elongated ART plane SUBR wa-me-'en EXT-go-PRTC "How long does the plane take to get there"? Probably the temporal use of a'a- with these stems indicates a conceptual differentiation from the view that motivates the temporal use of auu- in (30). One plausible account correlates with the observation that the notion of a path is particularly salient to the meaning of auu~, whereas the notion of a localized area is more salient to the meaning of a'-. The localized meaning of a'a- is illustrated in (32)(a and b). (32) (a) a'aZii-nyi u aya-tyee-r-e QNT-Q there area-long-CAUS-APPLlC "How long is he going to stay there"? (b) ka-nit iy=ee-tyee-r-e NEG-I here=area-long-CAUS-APPLIC "I'm not going to stay here a long time". The semantic representation of a'- takes as its base a region of space I call the distance domain. It is calibrated three ways with respect to the speaker. At the speaker's position, entities are marked as being proximal to the speaker, whereas entities that may be in eyeshot of the speaker or are otherwise particularly within the scope of his attention are marked as being at a medial distance from him. Entities completely removed from his purveyance are treated as distal to his position. The distal category is also the unmarked one. For this reason it is particularly appropriate to the representation of a'-, given below in Figure 10.
372 EUGENE H. CASAD Figure 10. a'-. I suggest,then, that auu- represents an imperfective view of the passing of time, whereas a'- represents a perfective view. In Comrie's terms, these two prefixes signal a semantic aspectual contrast, a temporal distinction marked lexically, rather than morphologically (Comrie 1976:7). The connection between time and aspect is, of course, easy to see. Situations, events, and processes are all embedded in the matrix of time, even minimal, momentary ones. Aspect, then, turns out to present a view of "situation-internal" time (ibid.:5). It can be characterized as different ways of looking at the internal structure of a situation. This internal structure has distinct phases. In particular, it has beginning, middle, and end points, any of which can be singled out for linguistic purposes. In addition, it is possible to refer to a situation in its totality without making reference to any of the phases of the evolution of that situation (ibid.: 18-20). All of this is highly reminiscent of the notion of the directed path schema, discussed in both Miller and Johnson-Laird (1976) and Casad (1982a). The temporal path followed by a situation throughout its development is precisely such a path, albeit an abstract one extended from the domain of physical space into the aspectual domain. Considerations such as these led Langacker to include a temporal profile as part of his general characterization of processes (Langacker 1978:861-2; 1982a:48). The use of this basic conceptual schema in the area of aspect is a prime example of the human creature's ability to make abstractions, symbolize them for purposes of linguistic expression and weave them intricately into grammatical structure. To close, Comrie also notes that the notions of perfectivity and imper- fectivity do not represent the objective reality about a scene since that scene can be viewed by the same speaker, both perfectively and imperfec- tively, without contradicting himself (1976:4). Examples (29)-(32) show, therefore, both the conceptual ability to transfer a concept from one domain into another, and the ability to view a highly abstract concept in more than one way.
CONVENTIONALIZATION OF CORA LOCATIONALS 373 4.0 Conclusion In this paper I have examined various cases in which Cora locationals show unexpected uses in particular contexts. I have tried to show that there are broadly based principles which limit the productivity of these locationals. Linguistic conventionalization is at work at various levels of analysis. That it is not restricted to just a single area of Cora grammar can be taken as an argument for Langacker's claim that all of grammar is symbolic in nature and that there is no real autonomy between the lexicon, morphology and syntax (1983:1/6,14). A complex of perceptual and cognitively based abilities partly underlie the usages I have cited here. For one, speakers have the freedom of choice to structure a scene in various ways. They can choose to include certain aspects of a scene to the exclusion of other aspects of it. They can fade out aspects of the prototypical orientation on a scene so that noncanonical interpretations arise or at least more schematic versions can be utilized. They can perceive the shape of an entity and use that as the basis for comparing and categorizing the shape of something else. Data cited in this paper also show that speakers of human language can perform highly abstract conceptual transformations such as that of inferring from a set of points to a coherent configuration or that of taking an abstract configuration from one domain and extending its use to the semantics of linguistic elements used to refer to something else in another domain. Finally, we have seen that previously established meanings may, at least partially, force the choice of prefix sequence that occurs with a given stem. The type of analysis that I have presented in this study diverges sharply from most present day approaches to semantic description. A possible alternative analysis would begin by entering each locative prefix plus stem combination into the lexicon as unanalyzable wholes. This would be done simply because so much unpredictability resides in the usages and meanings of these words (cf. Wasow 1977:330). Each entry would then be given a semantic characterization, stated in terms of finite sets of componential features that are assumed to have universal validity. Many of the factors that I have discussed in this paper, such as the speaker's vantage point, would be relegated to the area of pragmatics and would not even be a part of the lexical representation. Relations between individual lexical entries would be indicated through lexical redundancy rules.
374 EUGENE H. CASAD It is easy to see that such an alternative analysis would not allow a unified treatment of the data this paper contains. To begin, the individual components in a set of features used in a lexical entry are typically treated as discrete members of an unordered set. In our view, however, each feature is taken as a shorthand label for a conceptual complex of some kind; all the conceptual complexes that the feature bundle represents are then integrated in specific ways to yield a composite, coherent semantic whole (Langacker 1983:1/16). The alternative assumption that the locative prefix plus stem combinations are unanalyzable wholes runs into numerous problems of its own. Such an analysis would be plausible only for the most highly lexicalized locative prefix plus stem combinations of the language. The unitary analysis would overlook the fact that Cora speakers can focus on the analyzability of these prefix combinations and can even form puns based on them. It would also fail to point out systematic similarities in the meanings of words involving a given locative prefix sequence and systematic contrasts between the meanings of pairs of words built up from distinct prefix combinations. Although certain idiosyncrasies can be accounted for by the use of lexical redundancy rules, many of them cannot, simply because of the way the rules are formalized and because of the interpretation that is given to them. Thus, Wasow puts a "local" constraint on them, restricting them to only those noun phrases that bear particular grammatical relations to a given verb (1977:330). Notice that redundancy rules as Wasow formulates them are, apparently, not allowed to refer to the individual locative prefixes that I discuss in this paper. Anderson's reformulation of the structural conditions under which lexical redundancy rules can operate seems more promising. However, to even reformulate them so that they could refer to the locative prefixes would require us to drop the unitary analysis. The use of redundancy rules, however reformulated, would still not take into account one of the most central facets of this study, i.e., the particular factors that motivate the semantic extensions that I have documented here. At best, redundancy rules note that lexical entry X. shares a stretch of morphemic material with entry X.. This is basically construed as an accidental occurrence, a construal we find less than satisfying (cf. Aronoff 1978:14). I conclude that a model of grammar that characterizes the semantic structure of lexical items as unordered bundles of features, with redundancy rules relating the lexical items in question, is inadequate for characterizing the central and pervasive phenomenon of
CONVENTIONALIZATION OF CORA LOCATIONALS 375 semantic extension. Although I have employed a minimum of formal apparatus in presenting my analyses of Cora data, it is worth mentioning that there is no intrinsic incompatibility between holding to an encyclopedic view of meaning and making a formal description of language. A case in point concerns the work of Roger Schank and his associates involved in research into artificial intelligence. In a stimulating paper titled "Identification of Conceptualizations Underlying Natural Language", Schank attempts to construct a model of a human sentence parser (1973:189). For Schank, the human parser examines the content of both his syntactic structures and his conceptualizations and goes on to make expectations at both the levels of syntax and meaning. The parser's set of expectations is partly based on the use of context to determine events, specify who is involved in those events, guess what the result of the event will be and select a proper response to it all. Other expectations that Schank cites are the assumed purpose of a conversation and the parser's world view of the unfolding situation. Finally, the parser carries with himself an entire set of cultural norms (1973:189-90). In short, several of Schank's "expectations" fit in well with an encyclopedic view of meaning; such information can also be characterized in a rigorous manner. To close, the Cognitive analysis of Cora data shows that the non-predictability of locative markings and the conventionality of their usages entails no claim that they are meaningless and must be entered into the lexicon as single units. Instead, the choice to view meaning as encyclopedic in its scope has allowed me to treat in a unified manner what would otherwise seem to be a contradictory puzzle. There are broad perceptual and cognitive principles behind these complicated usages. The conventionality of locative usages in Cora simply means that speakers of Cora must learn how things are said and learn to exploit the richness of their semantic reservoir for saying comparable things. Notes 1. For some of the studies made on locational systems, see Friedrich (1970), Denny (1978) and Casad (1982a). Clark (1978), Traugott (1978) and Brown (1982) all present typological studies based on data from numerous languages. 2. Cora pertains to the southern group of Uto-Aztecan languages. It is spoken in the state of Nayarit, Mexico by approximately 15,000 people. The data in this study were collected
376 EUGENE H. CASAD from various speakers who live in the village of Jesus Maria, municipio of El Nayar while I have been engaged in a field program under the auspices of the Summer Institute of Linguistics. The period of research covers the years from 1971 to the present. I would like to express my gratitude to the Cora speakers who have been my teachers in much more than just language learning. This paper was first given at the Friends of Uto-Aztecan Working Conference held at the University of California at San Diego, June 28-29, 1984. Discussions with Doris Bartholomew, Bruce Hollenbach, and David Tuggy, as well as comments by Ron Langacker and Brygida Rudzka-Ostyn have been very helpful in bringing this paper to its present form. 3. For a theoretical treatment of this distinction, see Haiman (1980). 4. Rosch and others have shown clearly the role that prototypical instances serve in the ways that people categorize objects (Rosch 1975). 5. The multiplicity of spatial forms in Cora reflects a number of influences, both linguistic and nonlinguistic. For one, the Coras have, for hundreds of years, lived in rugged, mountainous terrain. Thus, characteristics of the lay of the land, e.g. hills and rivers and plains, have taken on special salience in their awareness of landmarks by which to calculate their position, while travelling, for instance. It is interesting to note that the Tarahumara and Guarijio, who live in the Sierra Madre Occidental of Chihuahua state in Mexico, have also developed a set of topographic adverbs, whereas the Yaqui, who live on the coastal plain of Sonora, apparently never did develop such a morphological system. In addition, the Cora share an inheritance from Uto-Aztecan. Thus, many Uto-Aztecan languages are characterized by rather elaborate deictic systems that relate the speaker in various ways to some other entity. Hopi, Papago, Luiseno and Northern Paiute all have spatially oriented deictic systems of substantial complexity. Huichol shares a locative prefix system with Cora, albeit the Cora system appears to be the more elaborate of the two. Western Mexico itself seems to have been a cultural area in which the groups living there placed special emphasis on spatial concepts. Thus Cora and Huichol reflect much of the same kind of complexity as has been so delightfully described by Friedrich for Taras- can, another language of western Mexico. Finally, the multiplicity of locationals in Cora suggests strongly that the notion "Lo- cation(LOC)" is not a semantic primitive, but rather, is a label for a broad range of conceptual complexes. 6. The morphological conventionalization of the locative particles accounts for the absence of CV particles with the topographic suffixes. Topographic adverbs require that the particle be overtly marked for three categories. 7. The following abbreviations are used in this paper. APPLIC: ART: CAUS: CNJ: COLL: COND: DEM: DISTR: DUR: EXT: Applicative Article Causative Conjunction Collective Conditional Demonstrative Distributive Durative Extensive NARR: NEG: PASS: PAST: PAUS: PERF: PERIP: PL: PRTC: QNT: Narrative Negative Passive Past Pausal Perfective Peripatetic Plural Participle Quantifier
CONVENTIONALIZATION OF CORA LOCATIONALS 377 FUT: HAB: IMPERF: LOC: BASE: MED: Future Habitual Imperfective Locative Base Medial QUOT: RDP: REFL: STAT: SUBJ: SUBR: Quotative Reduplication Reflexive Stative Subjunctive Subordinator 8. This metaphor of the fading out of some aspect of a mental image is adapted from Kosslyn (1980:141). 9. Suppose the burning were to affect a string of banners hanging on a wire stretched from one side of a room to the other. Rather than use the expected *watataasiny we find that watyataasin is chosen. Apparently, the semantics of ~taa "to burn" are incompatible with those of wata-. Coras also can speak of a string of lightbulbs hanging from a wire stretched out horizontally. Once again, however, *watataasin is inappropriate. Instead, they use the form a'ataatimee "the lights strung out along the wire are all burning brightly". I found it necessary to select another verb stem in order to associate wata- with the process of illumination. For describing the effects of a neon light that runs horizontally across a wall, a Cora speaker can use (i), in which wata- is used with -taca "to illuminate brightly" and a?a- occurs in a relative clause with -taa "to burn". (i) m-an pit wa-ta-taca mi taih ti MED-top SUBJ EXT-straight-shine ART light SUBR d'a-taa mi fooko hece ti situra area-burn ART light bulb at SUBR straight "The light which is burning up there all along the neon bulb is brightly lighting the room". 10. This perspective also ignores completely the lateral turning associated with securing the lid onto the bottle. 11. In contexts such as this, "down" comes to mean "to impinge upon from the outside" (Casad 1982a:281). 12. This example illustrates another conceptual ability of the speaker. The suffix -Si applies to each individual subject entity who followed a distinct path from a distal point to a common goal. The role of kura'a- is to assimilate a set of points to a full configuration (cf. Miller and Johnson-Laird 1976:53-54). 13. Eliade observes that earthly rituals are often patterned on archetypes of phenomena believed to occur in the heavenly sphere (1974:21). 14. Thus the Cora system is highly reminiscent of the Tarascan system described by Friedrich (1969a, 1969b, 1970, 1971). 15. This second meaning of auuta- is based on a perceived geometrical shape associated with each of the body appendages. 16. Both the uu of (25a) and the 'u of (25b) represent the same morpheme. A glottal stop is inserted between the two locatives when a subject prefix is added to the verb word. The insertion of the glottal stop is accompanied by a shortening of the prefix vowel. 17. Since all of the examples in 3.7.2 originate with the same Cora speaker, I think that dialect differences do not in any way affect these data. 18. For a particularly spectacular example of this in Hopi, see Malotki (1983).
378 EUGENE H. CASAD 19. The morpheme that I gloss here "area" has the same shape as the locative prefix a'-, which I gloss variously as "away, distal". It also has the same shape as the plural posture verb form a a- "to lay flat, be located in a general area". I think it probably arises from the posture verb stem, but is falling together in the verb structure of Cora with the prefix a'-, which may come from a different historical source.
The Conceptualisation of Vertical Space in English: The Case of Tall Rene Dirven University of Duisburg John R. Taylor University of the Witwatersrand Introduction One of the basic assumptions of Chomskyan linguistics has been that language constitutes an autonomous cognitive domain. Although language is seen as an essential component of the mind, even as the key to its understanding, no direct link is postulated between linguistic categorisation processes and other cognitive processes of categorisation. In this respect, Chomsky is maintaining that fundamental split between linguistic knowledge and encyclopaedic knowledge first postulated by Bloomfield (1933: 139ff.). One of Chomsky's main achievements was to have located this autonomous linguistic functioning at the cognitive level, without, however, permitting any direct channel to exist between linguistic and other faculties of the human mind. The issue can be represented schematically as follows. Given that language and conceptualisation are domains of human cognition, what relationship exists between mental concepts and the meanings of linguistic forms? (i) Cognition Language Conceptual system Form Meaning Concepts ??' What link exists between concepts and meaning?
380 RENE DIRVEN AND JOHN R. TAYLOR The question posed in (1) has been taken up by Manfred Bierwisch in two recent papers (Bierwisch 1984, 1985). Bierwisch rejects a view of language as a means of directly linking phonetic substance to conceptual structure; on this view, the meaning of a linguistic form would itself have the status of a conceptual structure. Instead, semantic structure is viewed as an 'interface' between the linguistic system and conceptual knowledge. Bierwisch represents the two viewpoints as follows. In (2)(a) — the viewpoint adopted by Bierwisch — semantic structure is independent of conceptualisation, while in (2)(b) semantic structures constitute a component of the conceptual system. (2) (a) FS (b) L FS C L = a natural language FS = semantic form C = conceptual knowledge In a much earlier paper (Bierwisch 1967), Bierwisch had already espoused the Chomskyan divorce of linguistic from other conceptual processes. In this paper, Bierwisch undertakes an analysis of German adjectives, including those denoting spatial relations. He states that a semantic analysis of a lexical item is finished only if it leads to a combination of basic elements, that are true candidates for the universal set of semantic markers (op.cit.: 35). Bierwisch is at pains to stress that semantic markers are formal linguistic entities, which are not necessarily amenable to "a simple interpretation in terms of physics, geometry, biology and so on" (ibid.: 34). While the structuring of these semantic markers in a lexical item "may be dependent on the apperceptive structure", they must be viewed as "part of the innate capacity for language learning". Otherwise, Bierwisch asks, how can we explain that a child can acquire the semantic structure of his language "before learning even the smallest bit of physics, geometry, and so on" (ibid.: 35)?
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 381 The tensions inherent in this approach come to the fore in the discussion of the collocational possibilities of hoch "high, tall" and lang "long" (Bierwisch 1967). Both adjectives may be predicated of nouns like Wagen "car, cart" and Stange "pole"; with Wagen the adjectives pick out different properties of the referent, while with Stange they refer to one and the same property. Only hoch, however, may be predicated of Turm "tower", and only lang of Zigarette "cigarette". According to Bierwisch, these examples show that in the case of Wagen the contrast between 'height' and 'length' must be inherent in the noun, while with Stange the contrast is induced by the adjectives. Turm and Zigarette, on the other hand, show that the "semantic structure of the adjectives [...] is tied up in a rather complicated way with the structure of the modified nouns" (op.cit.: 14). That is to say, in order to account for the collocational restrictions of Turm and Zigarette Bierwisch must attribute semantic markers to the nouns, such that there is congruity between noun and modifying adjective. While Bierwisch concedes that the universal set of semantic markers may be dependent on the human apperceptive apparatus, at no point in his analysis does he allow apperception to interfere in the process whereby adjectives are combined with nouns. Linguistic categorisation is kept distinct from conceptual categorisation. Consequently, the possibility that linguistic and conceptual categorisation are dependent on the same cognitive processes, or that they might ultimately be one and the same process, is totally excluded. In his 1967 paper, then, Bierwisch is arguing that the collocational restrictions of hoch and lang are a consequence of the semantic structure of the adjectives and of the nouns which they may (or may not) modify. There is, however, an alternative explanation. Surely, it is our everyday experience of cars, poles, towers and cigarettes, and the way we conceptualise these objects, rather than some abstract properties of the nouns which denote these objects, which guides us in our use of language. A car has at least two environmentally important dimensions, length and height. A pole has two possible positions: it may be lying on the ground, in which case we would describe it as long; in its canonical erect position, it becomes tall or high. Towers, by their very nature, are upright structures, and are described in terms of height, while cigarettes are thought of as lying down, and are described in terms of length. If we cannot speak of a *hohe Zigarette "*high cigarette", this is simply because we see no purpose in categorising cigarettes as vertically standing objects. Even if we were to
382 RENE DIRVEN AND JOHN R. TAYLOR take a long cigarette and stand it upright on its end, we would still describe it as long. Conversely, as Herbert Clark (1973: 44) has pointed out, a tall person lying on the beach is still tall, not long, since we do think of human beings as upright entities. Even a person's shadow, though it typically falls on a horizontal surface, can be described as tall, since it reflects the tallness of a human body: (3) "Who is he, the voice chanted, who stands here in the morning sun? Who is he with so tall a shadow beside him?"1 In the above paragraph, we have referred to Clark's (1973) study of the semantics of space. In many respects, Clark's approach is diametrically opposed to that of Bierwisch. Clark sees no need for innate semantic markers. Rather, according to Clark, innate perceptual mechanisms, operating within an environment common to all human beings, confer a certain universality and naturalness to the processes of human categorisation. Bierwisch invokes innate semantic markers in order to account for the (to him) paradoxical fact that a language user can acquire a category like "vertical extension upwards from the earth's surface" without having the least knowledge of physics and geometry. Clark points out that we all exist on a (for most practical purposes) flat world; we and the objects around us are subject to the force of gravity, which operates perpendicular to the earth's surface; and we have a natural upright posture. These facts, accessible to all human beings from a very early age, lead naturally to the emergence of the cognitive-linguistic category "vertical extension". In the present paper, we reject the Chomskyan notion of an autonomous linguistic structure. Linguistic structures, we believe, are cognitively, and ultimately, perceptually based. Linguistic categorisation, in other words, is merely an aspect of human categorisation in general. What are the implications of this approach for the question of universalism in language? The hypothesised innateness and universality of perceptual processes does not lead necessarily to universalism of semantic (i.e. linguistic) categories. The point has been stressed by cognitive linguistics, who are reviving and developing the notion of linguistic relativity of the tradition of Humboldt, Sapir, Whorf, Weisgerber and others. Hawkins (1984: 42) sums up the claims of cognitive linguistics — especially as developed by Langacker (1983:I/3ff.) —as follows: (i) Semantic structures are characterised as conceptual structures that have become conventionalised for purposes of linguistic symbolisation
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 383 (ii) Given the conventional nature of semantic structure, it follows that while conceptual structures may (or may not) be universal, semantic structures definitely are not. Universalism, then, if it exists, is to be found at the perceptual, cognitive level, not at the linguistic, semantic level. A fundamental contribution in this respect is Berlin & Kay's (1969) study of colour terminology in 98 languages. Berlin & Kay found that if a language has only two colour terms (no language has only one), these will be 'black' and 'white'. If it has a third term, this is always 'red'. The fourth term is either 'green' or 'yellow', the fifth term being the other member of the pair 'green' and 'yellow'. The sixth term is 'blue'; the seventh is 'brown'; while the next four select randomly from 'pink', 'purple', 'grey' and 'orange'. These facts suggest that the human mind is not so much selecting from a universal set of innate colour categories, but is following a 'universal path' in dividing up, categorising and labelling the colour spectrum. Berlin & Kay's study has been criticised on methodological grounds, e.g. for its undue reliance on second-hand reports of colour terminology, and for the use of bilingual informants (see the reviews by McNeill 1972 and Conklin 1973). Nevertheless, we believe that their basic insight could provide a valuable paradigm for future linguistic research. It might turn out that different languages also follow a 'universal path' in making ever finer and more subtle distinctions in the domain of vertical space. In this paper, our aims are more modest. We restrict our attention largely to the four basic level2 adjectives for the vertical dimension in English: high, tall, low, and short, focusing especially on the case of tall. What is interesting about the English system is the existence of not one, but two positive polarity terms: high and tall. This sets English apart from all other languages with which we are familiar. Indeed, we would be extremly surprised to find a language which made exactly the same distinction as English in this respect. We hope that our analysis will show that it would be barely feasible to claim that in differentiating tall and high English is selecting from a putatively universal set of semantic primitives. But if the products of categorisation processes (i.e. the categories themselves), as well as the components of these categories (i.e. semantic features), are hardly likely to be universal, human strategies for categorisation might well be. One likely candidate here is Langacker's (1983) distinction between figure and ground. When perceiving and categorising the world around him, a human observer highlights certain aspects of the scene, while other aspects become the context
384 REN£ DIRVEN AND JOHN R. TAYLOR for the prominent aspects. Certainly, the notions of figure and ground provide a valuable heuristic device for discovering the meanings of such closely related items as high and tall. 2. English Adjectives Denoting Vertical Space 2.1. General Characteristics As a first approach, let us consider the distribution of high, tall, low and short in English, and contrast these adjectives with similar predicates in other languages. In English, buildings can be tall, high and low, but not short; people can be tall or short, never high or low (at least, not with reference to their stature); the ceiling of a room can be high or low, but hardly tall, and certainly not short. We can summarize this as follows: (4) [+ polarity] [- polarity] high tall short low buildings + + + people + + ceiling + + Like English, a number of other languages are reluctant to describe the stature of a person with the same words as are used to describe the height of an inanimate object, i.e. these languages reject equivalents of *high man, *low man (cf. Dirven 1976). In German, people are never hoch "high" and niedrig "low", but grofi "big" and klein "small". Dutch groot and klein are used in a similar way, although Dutch can also nickname someone een lange "a long one, i.e. a tall person". (In English, too, a long, thin person is marginally acceptable3.) French is like Dutch and German, in that people are grand and petit, not haut and bas. Note that while these languages reject equivalents of *high man, they do not have a second adjective exclusively for the vertical dimension. Other languages make no distinction between the height of a person and the height of an object. Italian alto and basso, Polish wysoki and niski, and Hebrew gavoha and namux can be used for both. Zulu, on the other hand, has no adjectives reserved exclusively for the vertical dimension. The adjectival stems -de and -fishane merely pick out the most prominent dimension of an entity, whether in the vertical or horizontal plane. Thus -de, which can be applied equally well to people and things, translates English tall, high, long, and even deep, while -fishane
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 385 translates low and short. English is unusual, not only in having a second positive polarity term for the vertical dimension, namely tall, but also in its use of short. Short is given in (4) as the converse of tall. The word has two meanings: "of little height" and "of little length". The former gloss applies to short people, the latter to short grass. Strictly speaking, short grass contrasts with long grass rather than tall grass (cf. short hair : long/*tall hair). In fact, it is doubtful whether tall as applied to non-human entities has a true antonym, and small (a term which, like short, is borrowed from another spatial domain) serves as a general substitute. The converse of tall tree is small tree, rather than *short tree. These various converse relationships are shown in (5). It is to be noted that whereas in German, for example, there is a simple one-to-one relationship betwen antonyms, the English adjectives form a much more complex network. (5) German: groG hoch klein niedrig English: long taj[ high (large) short low (small) 2.2 High and Low Teller (1969: 203) has pointed out that in their spatial uses, both high and low are polysemous. Both items can refer, either to vertical extent, or to vertical position. A high window could be a window of great vertical extent (high^), or a window (of whatever size) situated a great distance from the ground or floor (high2). Similarly, a low window is ambiguous between a window (irrespective of its location) of small vertical extent (low ), or a window located close to ground or floor level (low2). All uses of high and low imply some reference to a norm. A high window is one whose vertical extent exceeds the norm, or whose vertical location is above the norm. Low implies that the upward extension of an entity fails to reach the norm, or that an entity is located lower than the norm:
386 RENE DIRVEN AND JOHN R. TAYLOR (6) (a) NORM -h- highj ZERO | LINE (extensional) (positional) A lowx dress is a dress with a low2 neckline; the upward extension of the dress fails to meet some expected norm. A low dress thus contrasts with a long dress. A long dress is one with a low2 hemline, i.e. the lower limit of its vertical extent falls below some expected norm. With both highx and lowv vertical extent is measured from the base of the entity itself. Very often, the base of the entity corresponds with ground (or floor) level, as in high/low mountain, high/low building, high/low doorway, but this need not be the case (e.g. high window, high church-spire). Also, a constituent part of an entity can be characterised as high{ or low{. We can speak of a man with a high forehead, or a chair with a high back. Here, we say nothing about the overall size of the man or the chair, we only refer to the greater than expected vertical extent of a constituent part, namely the forehead and the back. Note that a chair with a high back is not necessarily a high chair. A high chair is one with a high2 seat; we say nothing about the dimensions of the back. Positional high2 and low2 likewise often imply reference to ground or floor level, as in high ceiling, low clouds. But the 'zero point', or 'origin', with reference to which an entity is located in vertical space, may be given by some more specific domain. Thus, high cheekbones are located high within the domain of the human face; high shoulders are high within the domain of the human torso; a high waistline and a low neckline are high and low with reference to an article of clothing; a high shelf is high with respect to, e.g., a bookcase, etc. Sometimes, it is the base of the entity itself which is raised with respect to the norm. Thus, a high table could be a table with long legs (cf. high chair); more normally, a high table is one of normal (b) high2 NORM low. ,t lOWj __\ ZERO ^LINE
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 387 dimensions which has been elevated by being placed on a raised platform. The same interpretation would be given to high altar. 2.3 Tall In their non-metaphorical uses, high and low always designate either vertical extent, or location in vertical space. Whatever subtleties attach to the use of these two words, their meanings are still relatively clear-cut. The only restriction is that in their extensional sense, the words may not be applied to certain animate creatures, especially human beings. In rejecting *high/*low man, English is like some other Western European languages. Unlike these languages, English has a special lexical item for designating the greater than average height of a human being, namely tall. At first blush, we would wish to represent the meaning of tall in the same way as highv cf. (6)(a), i.e. a tall entity is one whose vertical extent, as measured from the base of the entity itself, exceeds some norm. Yet, as we have seen (cf. (4)), tall and highl are distributed differently in English. It is not just that tall and high are in complementary distribution, i.e. that tall is used just in those cases where high may not be used; many entities may be described equally felicitously (or, as the case may be, perhaps not so felicitously) as both tall and high, usually with subtle semantic distinctions. Coming to grips with the meaning of tall has been the most challenging aspect of our study of the basic level adjectives of vertical space in English. For this reason, the remainder of this paper will be devoted mainly to a detailed study of this one lexical item. On the basis of de Saussure's dictum that language is "un systeme ou tout se tient", it is to be anticipated that a proper understanding of tall will enable us to better define the meaning of highv and to account for the otherwise puzzling restriction on the use of highY and low1 with human beings. 3. Collocational Preferences of Tall Let us first examine the collocational preferences of tall. What kinds of entities are most typically described as tall? With what kinds of entities is the use of tall possible, though infrequent? And when is the use of tall excluded? Data from the Birmingham University corpus of 7.3 million English words, which Professor Sinclair has kindly made available to us, confirm that tall is used overwhelmingly for the description of people, especially adult or adolescent males. The kinds of entities to which the 373
388 RENfi DIRVEN AND JOHN R. TAYLOR Table 1: The kinds of entities to which the 373 tokens of tall in the Birmingham University Corpus of English refer. Kind of entity people trees plants buildings poles, columns, etc. animals pile, mound, etc. birds miscellaneous metaphorical uses unclear cases1 Number of instances 219 33 26 24 18 6 6 2 27 6 6 Percentage 58.7% 8.8 7.0 6.4 4.8 1.6 1.6 0.5 7.2 1.6 1.6 1) In these cases, the computer print-out contains insufficient text to determine the entity to which tall refers. tokens of tall in the corpus are applied, are listed in Table 1. Apart from people, tall is applied most readily to trees, and to certain kinds of plant (especially grass). Buildings, and artefacts of a certain shape (i.e. poles, posts, pylons, columns etc.), constitute further sub-groups. It is of interest to note that the group labelled 'miscellaneous' in Table 1 consists entirely of man-made artefacts, including such disparate items as hat, fence, gate, axe, canvas, typewriter, trumpet, candle, figurine, etc. It comes as something of a surprise to find that all non-metaphorical4 uses of tall in the corpus refer either to living things (people, trees, plants, etc.), or to man-made artefacts; there is not a single instance of tall applying to natural phenomena.5 We return later in this section to this rather unexpected collocational constraint. To obtain further information on the range of entities that may be described as tall, acceptability judgements were elicited from 40 English- mother-tongue students at the University of the Witwatersrand. Subjects were given a questionnaire which listed 67 nouns, each qualified by attributive tall. The questionnaire asked: "How acceptable do you find the following expressions?" Responses were given on a seven-point scale: 1 = very good, 2 = good, 3 = less good, 4 = doubtful, 5 = bad, 6 = very bad, 7 = extremely bad. Mean acceptability ratings are listed in Table 2.
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 389 Table 2: Acceptability ratings of collocations with tall. B A D V E R Y B A D E X T R E M 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. L 62. Y 63. 64. B 65. A 66. D 67. tall dome tall shadow tall wall tall wardrobe tall door tall bridge tall chair tall bird tall seedling tall mountain tall box tall house tall precipice tall infant tall bungalow tall baby tall dog tall window tall cavernous room tall pint tall stalactite tall ceiling tall table tall cave tall room tall hill tall cloud tall hair 4.22 4.25 4.39 4.39 4.52 4.55 4.56 4.83 4.83 4.86 4.86 4.87 4.88 4.99 5.05 5.07 5.12 5.25 5.52 5.53 5.78 5.93 6.04 6.09 6.25 6.37 6.40 6.94 G O O D L E S s G O O D D O U B T F U L 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. tall person tall boy tall girl tall fir-tree tall cypress tall child tall tree tall pine-tree tall poplar tall building tall oak-tree tall beanstalk tall reeds tall elm-tree tall pillar tall lamp-post tall jacaranda tall apple-tree tall spire tall glass tall willow tall grass tall plant tall block of flats tall tower tall telegraph pole tall grain silo tall factory chimney tall electricity pylon tall stalagmite tall mealies tall church tall book-case tall skyscraper tall ostrich tall cocktail-glass tall TV aerial tall pinnacle tall pyramid 1.05 1.17 1.22 1.40 1.65 1.65 1.66 1.68 1.71 1.83 1.90 1.92 1.97 1.98 2.00 2.02 2.12 2.18 2.28 2.29 2.39 2.41 2.41 2.44 2.47 2.53 2.58 2.58 2.76 2.80 3.05 3.10 3.17 3.22 3.23 3.26 3.32 3.39 3.97
390 RENE DIRVEN AND JOHN R. TAYLOR The two sets of data — those from the corpus and those from the questionnaire — both present a fairly consistent picture. Tall, like highv refers to vertical extent; examples of vertical position (e.g. *tall ceiling) are either not attested, or are judged very bad. There are, though, some constraints on tall which do not apply to high. The entities whose vertical extent is qualified by tall are seen as solid, three-dimensional structures. Whereas we can speak of a high cave and a high room, *tall cave and *tall room6 are judged very bad. The reason is that in these two cases, the vertical dimension is occupied by empty space, not by some solid, vertically-rising object. The overall shape of the entity is also important. Tall seems to apply preferentially to slender objects, whose verticality is the most striking feature of their appearance. Adult humans, certain kinds of tree, grass, poles, pylons, and certain buildings optimally fullfil this requirement. Thus, babies and infants do not qualify for tallness. In the case of trees we find an interesting grading of preferences (see Table 2). First, poplars and cypresses are tall trees par excellence; trees of a 'fatter' shape, such as oaks, elms, jacarandas, apple-trees and willows (in that order), are judged to be slightly less typically tall. Certain kinds of plant, especially reeds, grass and mealies (i.e. maize, defined by Collins Dictionary as a "tall, annual grass") are also natural candidates for tallness. Animals, on the other hand, are not usually described as tall. A doberman is a big dog, not a tall dog, since the front-to- back dimension is at least as prominent as the bottom-to-top dimension. The giraffe is, perhaps, an exception. Collins describes it as "the tallest mammal", thus focussing, not on the animal's overall size, but on the long extended neck. Similarly, one might conceivably describe the ostrich as a tall bird.7 With regard to buildings, the salience of the vertical dimension again seems to be of prime importance. Tall block of flats, with a rating of 2.44, is judged better than tall pyramid (3.97) and tall dome (4.22), since in these latter cases the prominence of the horizontal dimension counteracts the overall impression of verticality. Tall, then, profiles the vertical dimension as the most prominent aspect of the total appearance of a three-dimensional entity. It is perhaps not too fanciful to trace the notion of prominence to the etymology of the word. Over the past 500 years, tall has undergone a remarkable sense shift. According to the OED, older and now obsolete senses included "quick, prompt, active" and "good at arms, valiant", while the meaning "grandiloquent, highflown" was current from the 17th to the 19th century.8 Although the word is now used almost exclusively to refer to vertical space,
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 391 we can unify past and present meanings in terms of a shift of prominence from the psychological to the spatial domain. Very probably, it is the component of vertical prominence which prevents the use of tall with very small entities. To some extent, tallness is a relative attribute, and implies a comparison with other, similar-sized entities. A tall man is taller than average, yet in absolute terms he is shorter than a tall tree, which in turn is smaller than a tall skyscraper. Even so, we can scarcely talk of a *tall seedling,9 while *tall hair10 is quite impossible. Tall entities have a number of other things in common, in addition to those already mentioned. We have pointed out that the Birmingham corpus contains no instances of tall applying to natural phenomena; tall applies exclusively to living things, and to human artefacts. These are entities whose vertically is in a sense 'extendible'; the entities have reached their stature through a process of natural growth, or through a process of construction. Their vertical stature is not 'given' by nature. We can characterise this component of tall as 'dynamic'. High, in contrast, seems to describe in a much more neutral, 'static' fashion the extent of vertically. There is thus a subtle, though real, difference between a tall building and a high building. A tall building is a prominent structure which is seen to tower up over other, similar structures. When looking down over a city from a plane which is about to land, one would be more likely to comment on the tall buildings than on the high buildings. It is the absence of dynamic, prominent vertically which seems to be at issue in some of the more doubtful collocations on the questionnaire. Consider articles of furniture. While it is possible to speak of a tall bookcase, or a tall wardrobe, these are not typical instances; acceptability ratings were 3.17 and 4.39 respectively. A tall chair (4.56) is even more marginal, while *tall table (6.04) is completely ruled out. Likewise, receptacles like cups and mugs, bottles and buckets do not normally qualify for tallness, although a glass (e.g. a cocktail glass) might do so.11 Although the corpus contains no instances of tall applying to natural phenomena, the questionnaire shows that such collocations are possible in English. Interestingly, the collocations are acceptable just in those cases where it is possible to regard a natural phenomenon as a dynamic, upwardly-rising structure. An obvious instance is tall stalagmite, with an impressively high acceptability rating of 2.80. Stalagmites have the appropriate slender shape. Furthermore, they do, in a sense, 'grow', albeit very slowly. Tall stalagmites, which rise upward from ground level, contrast
392 REN£ DIRVEN AND JOHN R. TAYLOR strikingly with the barely acceptable * tall stalactite (5.78), which protrudes downwards. Although tall mountain is not a typical collocation (acceptability rating: 4.86), one could imagine a tall mountain (though not a *tall hill) as a prominent physical mass towering upward from a flat plain. R.L. Stevenson even speaks of tall precipices. The following quotation is from a passage in which Edinburgh is personified as a figure sitting on the top of a rock: (7) "From her tall precipices and terraced gardens, she looks far and wide on the sea".12 There is a further, very important characteristic of tall entities. Tall entities are seen as whole structures, physically discontinuous with the background against which they are viewed. Tall buildings, for instance, are typically silhouetted against the sky. As already pointed out, it is possible to speak of a tall bookcase. Again, a bookcase can be viewed against the background of the wall in front of which it is standing. But it is difficult to imagine a *tall door, or a *tall doorway, even though the frontal dimensions of the door might be identical to those of the bookcase. The reason is that a door is an integral part of a larger structure (i.e. a wall); it is physically continuous with the wall, and cannot be viewed against a distinct background. Tall, then, unlike high, cannot describe a component part of a larger structure. We cannot speak of a man with a * tall forehead, or a chair with *tall legs. Sometimes, however, a component part can be seen as a complete entity in its own right, especially if located at the upper extremity of a larger structure. Examples are tall hat, tall head-dress, tall heads(of flowers), all attested in the Birmingham corpus. Another example comes from Iris Murdoch. Even though our informants gave a very low acceptability rating to tall window (5.25), a window can be seen as a complete structure in itself: (8) "The room was full of yellow light, and the three tall windows were wide open on to the garden".13 In brief, tall profiles an entity in its entirety, as a Gestalt. In spite of the 'dynamism' implied by the word, it is significant that tall cannot combine with rise, a predicate which profiles upward movement on the vertical axis. *A tall-rising building and *The building rises tall into the sky are impossible. While it is possible to say The boy grew tall (where tall denotes the end- point of the growth process), we cannot speak of a * tall-growing boy (where growing focuses on the growth process itself).
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 393 It should now be clear why people are obligatorily described as tall, never as high. By the very nature of their animateness and their potential mobility, human beings constitute a supreme example of entities which are discontinuous with their environment. We see ourselves as freely-moving Gestalts, dominating the landscape in virtue of our upright posture. The anthropocentrism of our perceptions makes us reluctant to see ourselves merely as objects in the landscape. 4. The Status of Collocation Section 3 has been a detailed study of the collocational possibilities of the adjective tall. We have not only examined the frequency with which tall combines with other lexical items, we have also attempted to isolate the features of those entities which are preferentially described as tall. What, though, is the theoretical status of collocation? In stating the collocational range of tall, have we exhaustively defined the meaning of the item? Or should we rather regard collocational range as symptomatic of the concept TALLNESS, a concept which we still have to elucidate? The term 'collocation' is associated mainly with the British linguists Firth and Halliday. It is defined by Lyons (1977: 612) as the "tendency [of lexical items] to co-occur in texts". For Firth (1949), the meaning of a lexical item is indeed definable in part in terms of its collocations. The fact that tall regularly co-occurs with man, less frequently with bookcase, and not at all with hair, in itself constitutes part of the meaning of tall. Firth's approach neatly captures the fact that word combinations are not simply 'acceptable' or 'unacceptable'; rather, the fit of a word with another is a matter of 'more typical' and 'less typical', a limiting case being zero cooccurrence. Firth's approach, however, fails to explain why certain combinations are less typical. If asked why one cannot say *tall hair, a Firthian would probably reply "Because tall does not collocate with hair" — which, of course, explains nothing. An explanation for collocational possibilities is proposed within the Chomskyan paradigm, namely in terms of selectional restrictions. A lexical item is decomposed into semantic features (cf. Bierwisch's semantic markers). The lexical entry may also specify selectional restrictions, i.e. the semantic features which must obligatorily be present in lexical items which the item modifies. Thus, one of the semantic features of both tall and high{ would be [+vertical extent], and a selectional feature would be [+vertical
394 RENE DIRVEN AND JOHN R. TAYLOR dimension]. The discussion in section 3 has shown that, in order to adequately account for the distributional difference between tall and high the semantic and selectional features, especiallly of tall, would have to be considerably more complex. The selectional features of tall, for instance, would have to include [vertical dimension is many times greater than the other dimensions], [the entity is viewed as a Gestalt], [the entity can be silhouetted against a background], to name but a few. Some of the selectional features would be optional, such as [vertical dimension is a result of natural growth]. These features look very different from the 'semantic primitives' that have usually been postulated in componential analysis, like [human], [animate], etc. It seems that, once we move away from simplistic data of the type John admires sincerity and *Sincerity admires John, and try to account for the subtleties of real language use, it becomes increasingly unlikely that word meaning can be adequately defined in terms of semantic primitives which are not subject to further decomposition. This, in turn, strikes a blow at the very heart of Chomskyan linguistics, namely the innate- ness and the universal hypotheses. If semantic features turn out to be cogni- tively quite complex, and possibly have to be defined afresh for each language, how can they be learnt by the child, given the supposedly autonomous nature of the human linguistic faculty? These problems disappear within the cognitive grammar framework. The selectional features that we have suggested are aspects of the figure- ground relationship profiled by tall. They have to do with the manner in which the world is perceived, and their apparent complexity — almost, one might say, their provincialism — can be explained quite naturally in terms of modes of visual perception. Far from being unlearnable, the features are readily accessible to any normal-sighted human being. But there is another, more fundamental way in which the cognitive approach differs from the Chomskyan (and also from the Firthian). According to Langacker (1982a: 75), selectional restrictions are semantic, and amount to nothing more than compatibility in the specifications of corresponding entities that are superimposed in syntagmatic combinations to form a composite whole. Within the generative paradigm, selectional restrictions have to do with properties of lexical items; Langacker claims that combinatorial possibilities are a matter of semantic (i.e. conventionalised cognitive) structures. Recall the earlier discussion of *hohe Zigarette. We claimed that the expression is unacceptable, not because of any properties of the lexical
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 395 items hoch and Zigarette, but because of our everyday experience with (and hence our conceptualisation of) cigarettes. Similarly, we do not talk of *tall hair, not because of features of the lexical items tall and hair, but because of the incompatibility of the concepts TALL and HAIR. Collocation is a matter of fitting the figure-ground relationship symbolised by tall to the concept HAIR. To the extent that the fit is congruous, the combination will be judged meaningful, and therefore acceptable; otherwise it will be rejected. This approach to collocation throws light on an issue which up to now we have by-passed. Both the corpus and the questionnaire showed that tall combines preferentially with a certain range of nouns, e.g. person, tree, grass, pole, etc. Other combinations, like tall window, tall wardrobe etc., are attested less frequently, and were judged by informants as less acceptable. But what do we mean by saying that a collocation is 'less acceptable'? (Note that the classical Chomskyan paradigm does not allow degrees of acceptability — either the selectional features match or they don't. The Firthian approach, however, does better on this score.) Is tall window really less acceptable, i.e. less good English, than tall man? On the basis of her use of tall window in (8) would we wish to impute to Iris Murdoch an imperfect command of the English language? Surely not. With the exception perhaps of the items at the bottom of the scale, Table 2 does not really rank expressions in terms of their grammaticality, but in terms of the goodness of fit between the concepts symbolised by the nouns and the concept of TALLNESS. Because TALLNESS and WINDOW are not optimally compatible, it does not follow that tall window is 'bad English'. Eleanor Rosch's work on concept categorisation is relevant here (see Rosch 1977). Subjects are unanimous that a robin is a better example of a bird than is an ostrich. The concept ROBIN optimally coincides with the concept BIRD, the concept OSTRICH fits less well. Yet an ostrich is just as much a bird as is a robin; only a robin is a prototypical bird. Similarly, the concepts MAN, TREE, etc. fit the concept TALLNESS better than do WINDOW arid BOOKCASE — but windows and bookcases can still be tall. To put it another way: to understand (i.e. to conceptualise) what is meant by tall window and tall bookcase, one has to perform more cognitive work than in understanding tall man and tall tree. Hence the lower frequency of these expressions, and the lower acceptability ratings given to them. The 'cognitive' approach to collocation is of particular relevance in an evaluation of informant judgements. We have found that if informants are
396 RENE DIRVEN AND JOHN R. TAYLOR asked in an interview situation "Is it possible to say a tall X in English?", their responses are often quite unlike the clear-cut intuitions which the Chomskyan paradigm predicts. Instead, their replies are frequently vague, hesitant, and highly variable. The informant perhaps realises that the collocation is an unusual one; not being able to recall having heard it before, he replies "No, that's not English". On the other hand, he might realise that with a little cognitive work on his part the expression does make sense, so he replies "I guess it's OK". Conflicting responses like this present the linguist with a serious methodological problem. They show the need to give serious thought to the theoretical status of acceptability judgements. They also underline the need for data obtained by alternative procedures. Corpus-based data are, of course, a valuable antidote to intuitional judgements. In the following section, we describe a questionnaire which, we believe, represents a considerable innovation in this regard, in that it purposely refrains from asking for acceptability judgements. 5. The Concept of TALLNESS If frequency of occurrence and acceptability ratings of collocations are reflexes of the speaker's representation of the concept TALLNESS, what then is this representation? We have investigated this matter using a technique similar to that initiated by Rosch, i.e. we searched for 'good' and 'bad' examples of a category. Our investigation represents a major extension of Rosch's procedure, in two respects: (i) Rosch investigated categories designated by nouns, whereas we operated with a more abstract category, one designated by an adjective, i.e. tall (ii) Rosch required subjects to make a judgement as to whether a given entity, e.g. a robin, was a good or bad instance of a category, e.g. BIRD. We required subjects to judge how well the category TALLNESS applied to a given noun. 50 students from the University of Birmingham participated in the investigation. They were given the following instructions: "We can all think of objects that we would readily call 'tall'. But some would be better examples of 'tallness' and others less good or even bad examples. How do you think 'tall' applies to the following examples?"
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 397 Answer sheets listed 65 nominals in alphabetical order, and judgements were made using the same 7-point scale of the previous investigation. Mean ratings, in descending order of goodness, are given in Table 3. Table 3: TALLNESS scalarity. Mean judgements of how well the concept TALLNESS applies to the listed entities. G 0 0 D L E S s G 0 0 D 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. a sky-scraper the Eiffel Tower the Empire State Building a block of flats a spire a telegraph pole a tower a pylon a fir tree an elm a tree a beanstalk a building an oak tree a poplar a pyramid a lamp-post a cypress people St. Peter's in Rome a mountain a boy St. Paul's Cathedral reeds a church pinnacles a TV aerial an ostrich a willow a cocktail glass the Golden Gate Bridge a pillar stalagmites 1.42 1.56 1.60 1.64 1.83 1.87 1.94 1.98 2.07 2.09 2.13 2.17 2.28 2.29 2.30 2.32 2.33 2.49 2.61 2.62 2.66 2.70 2.82 2.87 2.96 3.14 3.18 3.24 3.28 3.28 3.36 3.37 3.40 D O U B T F U L B A D V E R Y E X T R. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. B 59. A 60. D 61. 62. 63. B 64. A 65. D a silo a book-case a house plants a child a bridge a dog a person's shadow a dome a wardrobe a glass a door a ceiling a wall grass a hill maize a cavernous room a room a box a chair a cloud stalactites a pint a table a seedling an infant a bungalow a cave a bird hair a baby 3.42 3.42 3.51 3.53 3.54 3.63 3.78 3.79 3.81 3.89 3.98 4.03 4.07 4.08 4.20 4.21 4.23 4.51 4.64 4.66 4.67 4.96 4.98 4.99 5.03 5.17 5.18 5.26 5.32 5.81 6.13 6.24
398 RENE DIRVEN AND JOHN R. TAYLOR Superficially, this and the previous questionnaire were very similar. There is, though, a vital difference. The first questionnaire had explicitly asked for acceptability (i.e. linguistic) judgements; here, we requested judgements concerning the goodness of fit between the nominal categories and the category TALLNESS. No doubt, some of the Birmingham students may have approached the task by assessing the linguistic acceptability of an adjective-plus-noun combination. While some nominals were indeed given very similar rank positions in the two investigations, there are some very striking differences. Note, for example, the very different rankings given to person!people, pyramid, mountain and skyscraper. This suggests that the two groups of subjects were in fact performing different tasks. The first group of entities (1-8) in Table 3, those which were judged to be very good to good instances of TALLNESS, are all artefacts. They all exemplify extreme prominence of the vertical dimension. However, the first four items {a skycraper, the Eiffel Tower, the Empire State Building, a block of flats) combine extreme vertical prominence with an overall impression of massiveness and bulk, whereas in the next four items {a spire, a telegraph pole, a tower, a pylon) only the component of vertical prominence seems to be of importance. These two (partially overlapping) senses can be summarised as follows: (9) . tal\ tall2 T extreme prominence of I vertical dimension; I massive, bulky; strikingly high in L the environment Does this finding indicate that we need perhaps to recognise, not one, but two prototypical representations of TALLNESS? The data discussed in section 3 would appear to belie this suggestion; in section 3 we stressed the association of vertical prominence with slenderness (i.e. tall2) rather than with bulkiness (tallx). Nevertheless, a closer examination of some of the responses to the first questionnaire does suggest that some of our respondents, at least, may have been operating with tally Consider the distribution of the 40 ratings given to the collocation tall mountain: (10) rating 12 3 4 5 6 7 No. of 0 2 10 8 2 8 10 subjects extreme prominence of "I the most salient, i.e. I vertical dimension; I thin, slender J 14
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 399 The responses are distributed bi-modally. Those subjects who responded with 6 and 7 (i.e. 'very bad' to 'extremely bad') presumably had the representation tall2 (i.e. vertical prominence and slenderness). Those whose responses clustered around 3 (i.e. 'less good') may have had a representation of TALLNESS which permitted the notion of bulkiness (tall^. To return to Table 3. The two senses of tall which we have posited are present, albeit in slightly attenuated form, in the next group of items (from 9, a fir tree to 25, a church), i.e. those entities that fall within the judgement scale 'good' to 'less good'. It is here that we find most tree names. In contrast with Table 2, the more 'bulky' trees (i.e. elm and oak: tallj are rated slightly higher than the more 'slender' ones (poplar and cypress: tall2). Amongst this second group of items, however, we seem to be witnessing a blurring of the two senses given in (9). Whereas tall^ (prototypical application: skyscraper) profiles the extreme prominence of the vertical dimension in combination with an overall impression of massiveness, and tall2 (prototypical application: spire) profiles vertical prominence and slenderness, with many of the items in this second group both vertical and horizontal prominence seem to be more in balance. This is the case with pyramid, St. Peter's, St. Paul's and church. With one or two exceptions (people, boy, reeds, beanstalk), all the items that were rated between 'very good' and 'less good' instances of TALLNESS are large-sized entities. Again with one or two exceptions (e.g. the Golden Gate Bridge), this feature is absent from the entities rated 'doubtful' and worse. This suggests that in its most typical uses, tall is associated with some absolute notion of impressive largeness. In fact, a cursory glance at Table 3 might lead to the conclusion that subjects had merely ranked the 65 items in terms of their decreasing size. This conclusion would be false, in that strict ranking according to size is interrupted by several misplaced items. Interesting examples are ostrich (28) and cocktail glass (30). These relatively small-sized entities cluster together with much larger objects, like church, pinnacle, willow-tree and the Golden Gate Bridge. An ostrich is a tall instance of the category bird (63), which itself is a very bad example of TALLNESS. Likewise, a cocktail-glass is a tall instance of the category glass (44), which comes at the limit of 'bad' examples. In other words, alongside the feature of absolute largeness, the notion of relative largeness seems to be of relevance in some instances. There are several other examples in the data. Beanstalk (12) and reeds (24) perhaps owe their high ratings to the fact that they constitute tall instances of the category plant (37).15 Again, St. Peters (20) is rated better than church (25), and the Gol-
400 RENE DIRVEN AND JOHN R. TAYLOR den Gate Bridge (31) better than bridge (39). We see that an adjective can sometimes be predicated more readily of a specific exemplar, or of a more specific sub-category, than of the general category. The ratings given to human beings —people (19) and boy (22) — also break up ranking according to actual size. Here subjects were probably operating, not with absolute or relative size, but with a conventionalised standard. While people and boy are good examples of TALLNESS, child (38) is less good to doubtful, infant (66) is bad, and baby (65) very bad. In other words, the standard of TALLNESS for human beings is given by adults and adolescents. Young children fall below this standard, and TALLNESS cannot be predicated of them. There is insufficient space in this paper to discuss in detail the less good and bad examples of TALLNESS. One thing does, however, call for comment, namely the discrepancies between the ranking of some of the entities in Table 3 and the acceptability judgements in Table 2. Is it not contradictory to say that grass is a 'doubtful' example of TALLNESS, while the collocation tall grass is given a 'good' acceptability rating? Skyscrapers are very good instances of TALLNESS, yet tall skyscraper has an acceptability rating between 'doubtful' and 'bad'. Subjects were practically unanimous in giving tall person an acceptability rating 'very good', yet in Table 3 people occupies position 19, with a rating between 'good' and 'less good'. The discrepancies are a result of the different tasks the two groups of of subjects were required to perform. In the first investigation subjects judged the acceptability of a noun modified by attributive tall, e.g. tall grass. The other group were judging on the basis of a predication. They were asked to what extent tall applied to a particular entity, or category of entities, e.g. they had to judge to what extent grass is in fact tall. In terms of the prototypical representation of TALLNESS which this investigation has brought to light — i.e. extreme prominence of the vertical dimension optionally combined with an overall impression of massiveness — then grass is indeed not a particularly good instance of TALLNESS. This does not exclude the possibility that particular instances of grass, e.g. grass which exceeds the height of an adult, cannot be appropriately described as tall. Similarly, people, as a category, are not optimal instances of TALLNESS. Yet tall can, and in the corpus the majority of attested tokens of the word do, refer to human beings of more than average height — no doubt because, with our typical anthropocentrism, we regard tall human beings as particular good instances of entities which tower above their immediate
CONCEPTUALISATION OF VERTICAL SPACE IN ENGLISH 401 environment. Skyscraper is another interesting example. As a category, skyscrapers are optimal instances of TALLNESS. For this very reason, we do not normally talk of tall skyscrapers. Since skyscrapers are, in their very essence, tall, it is redundant to attribute TALLNESS to any particular example of the category.16 Notes 1. Laurence van der Post, The Heart of the Hunter, Harmondsworth: Penguin Books, 1965: 89. 2. We therefore exclude from consideration adjectives like lofty, lanky, elevated etc. The term 'basic level' stems from Berlin & Kay (1969), and was refined by Verschueren (1981). 3. The Birmingham University Corpus attests the collocation of long and tall: long tall Brenda. There is also an instance of long men. 4. Non-spatial uses of tall in the corpus {tall order, tall tale, tall story) are very few, whereas metaphorical uses of high are numerous. 5. In one or two cases (e.g. tall earth mound) it is perhaps arguable whether the entity is man-made or not. 6. The corpus contains the expression two tall classrooms. Here, the entities can be seen from the outside, as solid, three-dimensional structures. 7. The kinds of birds and animals described in the corpus as tall include bear, elephant, heron and water-sprite. 8. The expressions tall order, tall tale, tall story and tall talk should perhaps be regarded as relics of this earlier sense, rather than as metaphorisations of spatial tall. 9. Tall is, of course, appropriate in sentences like Transplant the seedlings when they are two inches tall. This is due to the more general phenomenon whereby an unmarked, positive polarity term occurs in quantifying expressions. 10. Even though *tall hair is quite impossible, the expression a tall growth of hair exists. 11. The surprisingly high rating given to tall glass (2.29) may be due to association with the expression a tall drink. Both collocations occur in the corpus: a tall cool drink, two tall glasses. 12. Edmund Gosse, ed., The Works of Robert Louis Stevenson. Vol. I: Edinburgh: Picturesque Notes, London, 1906: 289f. 13. Iris Murdoch, The Sandcastle, Harmondsworth: Penguin Books, 1957: 25. 14. These specifications of the meanings of tall are not to be seen as the 'feature bundles' of the Chomskyan paradigm, but merely as convenient paraphrases of the complex senses of the lexical item. 15. The high rating given to beanstalk may be due to association with the expression as tall as a beanstalk, colloquially used to describe a tall, thin person.
RENE DIRVEN AND JOHN R. TAYLOR We wish to thank a number of people who helped us during the course of this investigation: Brygida Rudzka-Ostyn (Leuven) for her inspiring comments on earlier drafts of this paper; Dirk Geeraerts (Leuven) for very insightful suggestions; John Sinclair (Birmingham) for providing the data from the Birmingham Computer Corpus; Malcolm Coul- thard (Birmingham) and his co-workers for their friendly assistance with the investigations; the students and staff from Birmingham and Johannesburg who participated in the investigation and Andreas Kocks (L.A.U.D., Duisburg University) who processed the text on computer.
Length, Width, and Potential Passing Claude Vandeloise University of Antwerp 1. Introduction Mathematical concepts, because they are independent of the context of enunciation, have often been considered the more desirable metalanguage to describe spatial terms.1 In his description of German adjectives of dimension, Bierwisch (1967) distinguishes length and width by their sizes and perpendicularity only: while the former is a primary, maximum non-vertical dimension, the latter is judged to be secondary. Length and width, then, are inherent to the shape of the entity described and characterized by the innate semantic markers [±Maximum] and [±Second]. It is important for the verisimilitude of the innateness hypothesis that the semantic markers should not be dependent on contingent factors and should be independent as far as possible of our experience of the world. The above markers fulfil this requirement. In contrast, I contend that space is categorized in human cognition and in language on the basis of a varying number of factors, of which the actual shape of entities is only one. A careful scrutiny of the distribution of the length/width pair shows that the usage of these words depends also on factors such as the motion of the object, its function, the orientation of the speaker and, ultimately, the passing of the object by or through another element defined by the context. Bierwisch overlooks these factors, and in consequence describes only a limited set of usages of length and width.2 This paper, part of a larger study on the description of space and movement (Vandeloise 1985b), begins with an examination of the distribution of length and width. Four aspects of the dimensions referred to by these items are described by separate usage rules, specifying the directions along which they are evaluated. The first rule concerns linear objects and paths whose
404 CLAUDE VANDELOISE length is evaluated along the actual shape (§ 2.1). The second (§ 2.2) applies to mobile entities. Length and width can also depend on the speaker's position (§ 2.3). According to the fourth usage rule (§ 2.4), the length/width are the greatest/smallest non-vertical dimensions of an entity. Bierwisch's description applies solely to this metric aspect.3 Speakers are usually not aware of the complexity of the distribution of length and width. This feeling of uniformity may be explained synchronically by the relationships (pragmatic bridges) connecting the four aspects of these dimensions (§ 3.1). Length and width, then, are complex categories (Langacker in press: ch. 10) described by schematic networks. The four usage rules of these dimensions categorize relationships at a rather specific level of schematiza- tion. At a more abstract level, a highest schema involving the potential passing of the described object through or by a second entity delineated by the context of speech (§ 3.2) subsumes the three first usage rules. The metric usage rule, however, resists this generalization. Thus, the global highest level schema cannot be found to characterize all the usages of length/width. Neither is there a set of usages of these words functioning as a category-prototype, i.e. "that is substantially more salient than any other and functions as the apparent basis of more extensions" (Langacker ibid.). Diachronically, one may wonder how a symbolic connection has been established between the phonological poles of the length/width pair and their complex semantic poles. In this paper (§ 4) I will develop a system of logical diachrony in which, as a postulate, a complex category must be the result of a development from a simple category which I call impulsion.4 Even if, hypothetically, there are impulsions for each lexical category, this device is only of interest for general linguistics when its development toward the present complex category is systematic and can be established on a synchronic basis. From my experience (Vandeloise 1984, 1985b) this usually seems to be the case for spatial terms. Concerning the length/width dimensions, I endeavour to demonstrate in (§ 4) that a logical diachronic development from the highest schema subsuming the three first usage rules to the fourth metric rule is more plausible than the reverse. 2. The Distribution of Length and Width I will deal with four aspects of the dimensions referred to by these words. First, I will describe their more characteristic usages and propose usage rules governing them. Next, I will present the constraints operating
LENGTH, WIDTH, AND POTENTIAL PASSING 405 on these usage rules. 2.1 The Length and Width of Linear Objects and Paths Truly linear objects have no width. The evaluation of their length departs dramatically from the length and the other dimensions of multidimensional objects. Indeed, whereas these dimensions are always evaluated along straight directions fixed by their usage rules, the length of linear objects is evaluated on the basis of their actual shape. Thus, the length of the thread in Figure 1 and of the worm in Figure 2 is a distance going from A to B, following the curves of the former object and the waves of the latter: Figure 1 Figure 2 Roads and paths share an important property of linear objects: their length is evaluated along their actual shape. Some rigid linear objects, however, constitute an exception to this rule. Indeed, the length of the poker in Figure 3 is evaluated along its projection on a straight line. The shape of the object cannot explain this exception. Even though the thread beside it has exactly the same shape as the rigid poker, its length is evaluated along its actual shape. I will propose an explanation for this exception in section Figure 3 In turn pipes are rigid objects whose length is calculated along their actual shape. Thus, the length of the pipe in Figure 4 takes into account its different sections. The same holds true for the pipes of the old stoves in former school buildings. This discrepancy can be explained because pipes are paths in some way. The only difference between them and roads is that, instead of carrying living beings and their vehicles, they transport gases or liquids.
406 CLAUDE VANDELOISE A LB Figure 4 In contrast to linear objects, paths have a width, evaluated along a direction perpendicular to their length. Rule L(ength) governs the use of length for both linear objects and paths. Rule W(idth)1 only applies to paths. LjAVj.- The length/width of a linear entity or of a path is its extent evaluated along its actual shape/a direction perpendicular to its actual shape.5 By extent, I mean the greatest distance between two points of the object. 2.2 The Length and Width of Mobile Entities The length and width of mobile entities are evaluated respectively along directions parallel and perpendicular to their direction of movement. This is the case, for instance, with the car in Figure 5: Figure 5 Movement introduces a specific way of evaluating the length of multidimensional entities. Thus, the concept of length cannot be applied to either the immobile square or the immobile cube in Figures 6 and 7. The mobile square kite and the rolling cube in the same figures, in contrast, even though they have the same shapes as these objects, have a length and a width. Like the length of cars, these are evaluated according to rule L2/ W2:
LENGTH, WIDTH, AND POTENTIAL PASSING 407 fer w ^- Figure 6 Figure 7 L2/W2: The length/width of a mobile entity is its extent evaluated along a direction parallel/perpendicular to its direction of movement. Note that the length and the width of the kite and the rolling cube can only be justified by movement and do not coincide with the edges of these objects. It is worth noting that, for aerodynamic reasons, mobile objects are usually longer along the direction parallel to their movement than along their perpendicular direction. Therefore, their length is usually greater than their width, in comformity with Bierwisch's analysis. For functional reasons, however, some objects need a certain resistance to their movement. This is the case with razor blades, ploughs and rollers. The width of these objects is their greatest horizontal dimension. According to Bier- wisch, they can only be treated as unaccountable exceptions. Git^ W -■ ^ Figure 9 Movement has an influence on linear entities as well. Indeed, if a flexible linear object moves without modifying its shape like the sea-worm in Figure 10, its length is no longer evaluated along its actual shape but along a parallel to its movement, in keeping with rule L2. Figure 10
408 CLAUDE VANDELOISE 2.3 The Length and Width of Immobile Multidimensional Entities For many speakers, the length and the width of a relatively symmetrical entity may depend on their position relative to the object. Thus, for the priest in Figure 11, the length of the house may correspond to its width for the president and vice versa.6 A ~B J n L Figure 11 The length/width of the house, then, is the dimension evaluated along a parallel/perpendicular direction to the general orientation7 of the priest or the president. These usages are governed by rule L3/W3: L3/W3: The length/width of an immobile multidimensional entity is its extent along a direction parallel/perpendicular to the speaker's general orientation. Zubin and Choi (1984) note the same phenomenon for a relatively symmetrical table. Definition W holds true for the greatest horizontal dimension of a couch, usually perpendicular to the general orientation of the speaker. This point, noted by Clark (1973), constitutes another unaccountable exception to Bierwisch's analysis. 2.4 Metric Definition of Length/Width There is a last type of usage of length/width independent of the movement of the object and of the speaker's position. These usages are governed by rule L4/W4: L/W4: The length/width of an entity is its greatest/smallest non- vertical extent. According to rule L4/W4, side AB of the razor blade in Figure 8 becomes its length whereas side BC is called its width. For the house in Figure 11, the dimensions attributed by rule L4/W4 coincide with the dimensions used by the priest. Rule L4/W4 is preferred to the other usage rules for length/width in geometry because it is independent of the movement of the object and of the speaker's position; it therefore has the objective character required by
LENGTH, WIDTH, AND POTENTIAL PASSING 409 this science. The power of mathematics over its former pupils might explain why the metric definition of the length/width dimensions often prevails when speakers are asked for an explicit judgement on the meaning of length/width. Note, however, that the objective character of usage rule LJ W4 only holds true for geometric forms like rectangles, parallelograms, etc. Here, the directions along which length/width are evaluated are clear enough. With actual objects, however, many directions can compete for the evaluation of length/width and rule L4/W4 must be elaborated (cf. § 4). It is worth noting that this choice is always made between the directions defined by rule L2AW2 (the parallel/perpendicular dimension to movement) and rule L3/W3 (the general/lateral orientation of the speaker). It should not be surprising, then, if these latter rules prevail in our knowledge of length/width. 2.5 Constraints I will now turn to the main constraints bearing upon the length/width of objects. Three types of objects cannot have length: a) Some objects such as windows and doors whose height is the greatest dimension: Objects with vertical surfaces do not have length if they are usually in front of the speaker and if their height is their greatest dimension. Characteristic objects in this category are normal windows and doors. *The length of a window and *the length of a door, indeed, are normally unacceptable. Picture frames belong to this category as well, whenever their height is their greatest dimension. ^L^T L'W Figure 12 Figure 13 Thus, the picture in Figure 12 has no length. In contrast, since the greatest dimension of the picture frame in Figure 13 is horizontal, it can be called either its length (rule L4) or its width (rule W3). b) Objects with a length that would be smaller than their width or equal to it: Rules L2 and L3 sometimes assign to an object a length that is smaller than
410 CLAUDE VANDELOISE the width attributed by rules W2 and W . This is the case with the razor blade in Figure 8 and with the house in Figure 11, described from the priest's point of view. In both cases, application of rules W2 and W3 to the smallest side is blocked. If the speaker wants to qualify two horizontal sides of this type of object simultaneously, he will use the metric set of rules LJ W4 instead. The greatest side of the razor blade or the house, then, will be called its length. Usage of length is blocked as well when this dimension is equal to width. This is the case with squares and cubes, hence *the length of a square or *the length of a cube. Only width can be used in these cases. For mobile objects, however, equality of length and width is possible, as is evidenced by the square kite in Figure 6 and by the mobile cube in Figure 7. It remains impossible, though, to speak of *the length of a ball, even if it is moving. c) Objects without a significant extension along the direction of length: If context does not provide an appropriate scale of measurement it would be odd to speak of Ithe length of a flea. This third constraint is not specific to length only. It holds true for the other dimensions as well. For example, it is usually impossible to speak of *the width of a hair and, generally, of the width of a linear object. A scientific inquiry in a laboratory, however, might make speaking of the width of a hair more appropriate. So far we have looked at objects without length. The first two constraints bearing upon length do not concern width. In particular, even though a square or cube has no length, each of its edges can be called its width. It is nevertheless difficult to speak of Ithe width of a ball, except if it has to enter a narrow pipe. Apart from this restriction, it always seems to be possible to speak of an object's width, provided that constraint c is met, i.e. that the object has a significant extension along the direction corresponding to width. 3. Synchronic Interpretation of the Distribution of Length and Width In section 2, four usage rules were needed to describe words usually believed to be monosemic by speakers. How can we reconcile this feeling of uniformity with the diversity of the facts? In this section, I will present pragmatic bridges linking these different aspects of length/width. Next, I will propose a highest schema subsuming the three first usage rules presented in section 2. Synchronically, these connections suffice to explain the sense of uniformity given by the length/width pair.
LENGTH, WIDTH, AND POTENTIAL PASSING 411 3.1. Pragmatic Bridges between the Usages of Length/Width Pragmatic bridges are particular categorizing judgements involving our knowledge of the world beyond simple formal similarity or direct proximity. Neither of these connections stands, for instance, between the usages of behind in sentences 1 and 2: <C^S I -T 1 1 UrW Figure 14 (1) The rabbit is behind (the president) (2) The cat is behind (the wall) However, the application of behind to both the rabbit and the cat is not arbitrary since (Vandeloise 1984), for different reasons pertaining to our view of the world, both animals are inaccessible to perception. Here, inaccessibility to perception, is the pragmatic bridge connecting these usages of behind. Different pragmatic bridges between different aspects of length and width will be examined in this subsection. (a) Pragmatic bridge between rules Lj/Wj and L2/W2: The movement implied by roads allows an easy transition between these rules. The length of a straight path (L{) and the length of a mobile object running along it (L2) are evaluated along the same direction. A similar connection holds true for winding paths as well, at least for a sober driver obeying traffic rules. The length of the vehicle, here, is parallel to a tangent to the path in each point of its trajectory, as illustrated in Figure 15. L Figure 15 (b) Pragmatic bridge between rules L /W2 and L4AV4: This bridge has already been touched upon. It is based on the aerodynamic character of moving objects. Indeed, an object offers less resistance to air if its longer
412 CLAUDE VANDELOISE dimension (its length according to L4) is parallel to movement (the direction along which the length is evaluated according to L2). Exceptions like razor blades and ploughs are, indeed, necessary. The purpose, here, is to get the maximum resistance to the beard or to the earth. Width, in such cases, is greater than length. (c) Pragmatic bridge between rules L2AV2 and L3/W3: This pragmatic bridge involves a potential passing between the described object through or by a second element delineated by the context of speech. Because this bridge leads directly to a higher schema subsuming rules Lj/Wj, L2/W2 and L3/W3,1 will elaborate on it in the next subsection. 3.2 Length/Width and Potential Passing The importance of both relative movement and potential meeting to lexical categories has already been established for the French prepositions avant/apres (Vandeloise 1984, 1985c, 1986). Depending on which element of the relative movement is conceptualized as mobile (and on which element is taken as the immobile reference point), two different types of usages of these prepositions are obtained. The length of mobile entities is directly connected to these spatial relations since their length is, roughly, the distance between their avant (front part) and their arriere (rear part). The role of relative movement in the description of the length of mobile entities, then, is obvious. The potential passing rather than the potential meeting is the salient feature involved in the justification of the usages of length, as will be seen below. This concept also plays a unifying role between: (1) Linear objects and paths (2) Mobile and immobile multidimensional entities. 3.2.1. A Connection between Linear Objects and Paths According to rule L /Wj, the length of both linear objects and paths, in contrast with all the other objects, is evaluated along their actual shape. In this section, I will look for an explanation of this similarity. In the same way, I will try to account for the exceptions to rule L1/W]. But first, let me recall the categories of entities useful for the discussion. In the first place, the movement of a linear object can be either independent, as for instance in the case of a worm, or dependent, as in the case of a thread. Among non-
LENGTH, WIDTH, AND POTENTIAL PASSING 413 self-moving linear objects, the length of flexible ones is calculated along their actual shape whereas the length of rigid ones, being an exception to rule L /W , is calculated along their projection on a straight line. Finally, self-moving objects can move relative to themselves, like common worms, or be immobile relative to themselves, like sea-worms. Sea-worms constitute an exception to rule L{ since their length is evaluated along a straight line. All these points will prove useful in the following discussion. I summarize them in the tree below: entities paths mobile relative to itself common worm immobile relative to itself sea-worm L. non-self-moving flexible rigid thread poker L. "1 "i "i ^i In this section, I will explore what all these categories have in common. A. Perceptual hypothesis: In contrast to the evaluation of volumes, the length of paths and most linear entities is evaluated along their actual shape, even if this is not a straight line. Indeed, whereas volumes can be scanned along several directions, linear objects can only be scanned along their actual shape. Therefore, no conventional direction has to be imposed on them: their length is equated to the only way in which they can be scanned. Similarly, the length of a road is evaluated along the easiest way of going along it. Consequently, it is very tempting to establish a pragmatic bridge between roads and linear objects along the following lines: while roads are paths imposed on mules and donkeys, linear objects impose paths
414 CLAUDE VANDELOISE on the line of vision. The lengths of both categories are evaluated along the easiest way of going entirely through them, either on foot or with the eyes. This pragmatic bridge therefore is, at least for linear entities, perceptually based. Still, this solution cannot provide a complete explanation of the distribution of linear entities. It cannot for one thing explain the discrepancy between the length of the thread and the poker. While the easiest way of scanning the flexible thread and the rigid poker of Figure 3 is the same, their lengths differ. Moreover, although no scanning of the thread is possible in the case of the reels of Figure 16, the speaker is able to decide which thread is longer. Scanning, then, is not a necessary condition for speaking of length, and other connections have to be found between paths and linear objects. S 7 \ / Figure 16 B. Functional hypothesis: What enables the speaker to decide which thread is longer in Figure 16 is the knowledge he has of the function of reels. Depending on the purpose of the speaker, the threads in Figure 17 and 18 can be assigned different lengths: B B Figure 17 Figure 18 If the tailor plans to sew with a simple thread as in Figure 17, the length of the thread is AC. But if he needs a double thread, its length will rather be AB. In the same way, if the rope in Figure 18 is intended to tie a prisoner to a tree, its length is AC. But suppose that it is a rope in a hangman's hand, then, its length is AB. From this example, one might conclude that there is not a univocal relation between a multipurpose object and its length but rather that there is a univocal relation between its length and its differ-
LENGTH, WIDTH, AND POTENTIAL PASSING 415 ent functions. Much in the same way, the length of a pocket-knife varies according to whether it is closed in one's pocket or open in a tin of sardines. Coming back to the poker and the thread in Figure 3, then, one notes that their length is determined by an evaluation of their most likely function. Whereas the thread will probably be sewn into a piece of cloth, the poker is used for the purpose of stoking a fire. The first operation implies an extension of the thread, and consequently, its length is evaluated along its entire shape. With the poker, the housekeeper is interested rather in getting at the coal without burning her fingers. The length of the poker, its projection on a straight line, tells her exactly the poker's reach and at what point she has to be cautious. The functional nature of length allows us to give an explanation of the discrepancies between flexible and rigid objects. After acquiring the habit of calculating the length of flexible objects along their actual shape and the length of rigid ones along their projection on a straight line, the length of these objects can be evaluated, even if their function is unknown. Through their function, then, movement is involved in the contrast between the lengths of flexible and rigid linear objects. As will be seen below, relative movement and potential passing can also explain the similarity between the lengths of linear objects and paths and the discrepancy between the lengths of self-moving entities mobile or immobile with respect to themselves. C. A connection based on relative movement and potential passing: In the perceptual hypothesis, the line of sight is moving through the linear object in the same way as a moving element going along a path. In this section, I will consider linear objects and paths as, respectively, the mobile and the immobile element of a potential passing. The extent of passive/active passing is the distance between the first and the last contact (or maximal proximity) of one point of the mobile/immobile entity with an immobile/ mobile element.8 Even if they are provisionally immobile, most linear objects can be viewed as potentially moving along a fixed trajectory because of the anticipation of their function. Thus, whereas the length of the thread or the poker is the extent of their complete passing through a piece of cloth or through a stove, the length of a path is the extent of the complete passing by a mobile element. Width, the dimension perpendicular to the direction of passing, is proportional to the easiness of the passing.
416 CLAUDE VANDELOISE The discrepancies between the modes of evaluating the lengths of flexible and rigid linear objects on the one hand, and those of self-moving entities mobile or immobile relative to themselves on the other hand, can be explained by different ways of passing. When going along a path, a mobile element passes successively each point of the path. Therefore, the length of paths is always evaluated along their actual shape. A point of an immobile landmark, in contrast, can be passed by the points of a mobile trajector successively, simultaneously or in a mixed way. Flexible linear objects as well as creeping animals pass the point in the successive manner. Therefore, as in the case of paths, their length is evaluated along their actual shape. Simultaneous passing is illustrated by a vertical object moving on a horizontal plane, much in the same way as human beings walking. These entities do not have a length. Rigid linear objects like pokers and self-moving entities immobile relative to themselves illustrate the intermediate case. As illustrated in Figures 19 and 20, when several of their points simultaneously pass the immobile landmark, they only count for one point. In other words, all these points are projected on the direction of movement and their length is evaluated along a straight line. ft (T? Figure 19 Figure 20 Similarities between paths and linear entities as well as dissimilarities between different types of linear objects have now been explained by different ways of potential passing. In brief, the length of paths is related to a passive passing (in which the path is passed) whereas the length of linear objects is related to an active passing (in which the linear object passes). Amongst linear entities, rigid objects and self-moving elements immobile relative to themselves are related to a mixed passing, successive in some parts, simultaneous in other parts. The potential passing subsumes the perceptual hypothesis (passing of
LENGTH, WIDTH, AND POTENTIAL PASSING 417 the eye-sight along the object) and the functional hypothesis. The function of the object prevails over scanning in the choice of its length. The potential passing, as will be seen, also allows us to make a connection between the length of mobile and immobile multidimensional objects. 3.2.2 Pragmatic Bridge between Mobile and Immobile Multidimensional Entities In the case of mobile multidimensional objects (Figure 21), the immobile landmark relative to which the passing is evaluated is usually the speaker. The length of this type of object behaves in the same way as that of self-moving linear objects immobile relative to themselves. In each case, all the points in a plane perpendicular to the direction of movement pass the speaker simultaneously. The length of immobile multidimensional objects, on the other hand, is evaluated relative to a potential movement of the speaker (conceptualized as a point) passing them (Figure 22). He does not pass through them, as is the case for paths, but rather beside them, at a close distance along a straight line. ~4 . ^_ movement of the object potential movement of the speaker Figure 21 Figure 22 The concept involved in rule L3/W3, i.e. general orientation, is a family resemblance concept whose features are the direction of movement, the line of sight, the frontal direction, etc. Amongst these features, the first, namely the direction of movement, proves to be essential. A pragmatic bridge between rules L2AV2 and L3/W3 hinges on this feature: whereas rule L2/W2 implies a potential passing of the speaker by the object, rule L3AV3 involves a potential passing of the object by the speaker. While for both rules length is estimated along the direction of potential passing, width is estimated along a perpendicular to this direction. Rule LAV, then, subsumes rules Lj/W,, L/W2 and L3/W3: LAV: The length/width of an entity is its extent along a direction parallel/perpendicular to its direction of potential passing (through, beside or by an object suggested by the context of speech).
418 CLAUDE VANDELOISE In sum, pragmatic bridges have been established between usage rules Lj/W, and L/W2, L2/W2 and L/W4 and L/W3. Synchronically, this complex network of relationships is fairly representative of what speakers feel length/width have in common. 4. Interpretation of Length/Width in Logical Diachrony The command of the network representing a word can be explained as resulting from memorizing lists of its usages and schemas expressing what they have in common at different levels of schematicity (Langacker in press). However, a symbolic association cannot occur simultaneously between the word's phonological pole and all the nodes characterizing its semantic pole. The first access to the semantic pole of a word, either in historical time or in personal acquisition, must occur with a simple portion of this network. Let us dub this portion historical impulsion of the lexical category in the first case and personal impulsion in the latter. In the development of many words, either in historical time or in personal acquisition, impulsions lose their prominence. They are, in these cases, of little interest for a general theory of the lexicon. Starting from the supposition that a systematic evolution from the impulsion of a word toward its present distribution is not a rule in the lexicon, my inquiry into one of its most basic domains, i.e. spatial terms, led me to the conclusion (Vandeloise 1984) that a predictable and systematic development of the distribution of these terms is not only likely but that, moreover, some vestiges in their usage leave traces of their evolution. For instance, I took advantage of some constraints on the French preposition sous (under), usually judged to be arbitrary, to explain its development (Vandeloise 1985c) on the basis of its impulsion, i.e. the relationship bearer/burden. Such an impulsion, labelled logical impulsion, is a node in a semantic network, such that synchronic evidence shows that a development from this node to the other nodes of the network is more likely than other developments. This developmental approach must explain, as far as possible, some seemingly arbitrary aspects of the distribution of spatial terms and most notably their selection restrictions. In a systematically evolving language, logical impulsions should parallel historical impulsions. However, since many accidental changes occur in historical time, discrepancies between both types of impulsion are likely to emerge. Historical linguistics and language acquisition, though, can be
LENGTH, WIDTH, AND POTENTIAL PASSING 419 very helpful when no synchronic argument allows a choice between two putative developments. Geeraerts (1983a) pushes as far as possible, I believe, the contribution of historical linguistics in his exemplary study of the Dutch verb vergrijpen. When there is a common group of personal impulsions for the same word, language acquisition can also provide important clues. Such "common group" is not the rule but occurs often, I feel, with spatial terms. The evolution of the distribution of a word is not completely determined by its impulsion. There are a number of options open and different languages can make different choices (Vandeloise 1985d). Thus, the logical evolution of a word is not completely predictable from its impulsion. Even with the present distribution of a given word, when the language under study has made all its choices, several alternatives often offer themselves when one tries to explain the diachronic evolution of a word with synchronic arguments. Several models for the organization of its usage rules can then be proposed. They often rely on cognitive postulates that cannot be completely evaluated in a partial study. Information about a larger number of words in a larger number of languages is required for this task. In this section, I will present two guesses for the logical development of the length/width dimension depending on whether the following cognitive postulate is accepted: Postulate: The direction along which a dimension is evaluated is only established at the time of (or a short time before) the linguistic acquisition of this dimension. With this postulate at hand, I can propose a complete logical development of the length/width dimension. This task cannot be completely achieved without this postulate. At any event, however, there is synchronic evidence indicating that the functional and subjective rules L/Wj, L2AV2 and L3AV3 precede the objective rule L4/W4. (A) In the first model for the development of the distribution of length and width, usage rule Lj/Wj will be chosen for impulsion. It develops successively toward rules L2/W2 and L3/W3 and next to rule L4/W4. This choice is based on the following line of argument: (1) Rules L2/W2, L3/W3 and L/W4 elaborate rule L/Wr (2) Rule L/W, elaborates rules LjW, and L/W,. v/ 4 4 2233 If one admits that an elaboration is posterior in logical diachrony to the rule
420 CLAUDE VANDELOISE it elaborates, the following model can be proposed for the distribution of length/width. L/W, L2/W2~ ~L3/W3 4 4 As there is no elaboration between rules L2/W2 and L3/W3, they are placed at the same level. Note, however, that whereas an actual movement is involved in the former rule, this movement is usually potential in the latter. In many cases, actual movement takes precedence over potential movement and it might be the case for rule L2/W2 over rule L3/W3. I will attempt now to justify the elaborations upon which this development is based. First elaboration: The length of paths/linear entities is the extent between their first and their last point in a complete passing of an element designated by the context of speech. Less abstractly, recall that the length of these entities is evaluated along their actual shape when the passing of their points is successive. Their constitution does not allow different trajectories, either for the vehicles passing through the path in a normal way, for the passing of a linear object exerting its function in a landmark or for the line of sight scanning a linear object. Therefore, it is not necessary to fix a direction along which the extension of these dimensions will be calculated. Volumes, in contrast, can be scanned along many directions. Therefore, the three later rules of length and width must state explicitly along which direction the extension of these dimensions will be evaluated. In sum: - rule Lj/Wj has to specify an extent - rules L2AV2, L3/W3 and L4/W4 not only specify an extent but a direction as well. In this sense, the latter three rules are more complex than the first and elaborate it. Second elaboration: Rules L2/W2 and L3AV3 explicitly choose the directions parallel or perpendicular to the movement of the object and the directions
LENGTH, WIDTH, AND POTENTIAL PASSING 421 parallel or perpendicular to the speaker's general orientation respectively. Even though the choice of a direction does not appear in rule L /W explicitly, it plays an implicit role in this usage rule. Indeed, if it were applied to the square in Figure 23 carelessly, the length of the square would be AB (the greatest straight distance between two points of the squares edges). This holds true under the condition that length has to be calculated along a straight line. Things would be even worse without this condition. Rule L4/W4, in this case, would choose the greatest curve joining two points of the edges of the square. Length and width, of course, are not evaluated in such a fancy way. Figure 23 For regular geometric figures, rule L4/W4 normally makes a comparison between the edges of the figure. But in the case of actual objects, this rule chooses between the directions imposed by L2/W2 and L /W . Thus, rule L /W is established upon information provided by these two latter rules and elaborates them. (B) The argumentation on which the above elaborations are based would however lose its force if the child at a prelinguistic stage were aware of the directions along which non-vertical dimensions are evaluated. Suppose, for example, that the direction of potential passing with an object and its perpendicular constitute a very pervasive way of orienting objects at a very early stage. Long before knowing the meaning of length and width, the child would then be aware of these directions. Since their acquisition does not need a further effort when learning the names of dimensions, the elaboration proposed in (A) no longer holds true. In this case, the first attempt of a child who already knows the directions under discussion when he speaks about dimensions might be aimed at the dimension that has the maximum extent along any of these directions. Even if the above arguments are open to doubt, there remain two pieces of evidence showing that the development of length/width goes from rule LAV to rule L4/W4 rather than the reverse. This process, then, witnesses an evolution from an egocentric and functional conception
422 CLAUDE VANDELOISE of language toward a more objective one separated from the speaker's location and independent of the object's usage. (i) The pragmatic bridge between rules L2 and L4 works better, I feel, from L2 to L4 (model A) than from L4 to L2. This pragmatic bridge states that, for aerodynamic reasons, the extent of mobile entities along a direction parallel to their usual movement is almost always larger than their extent along a direction perpendicular to this movement. In Figure 24, the extension of mobile objects is represented by set A and the extension of the objects of which length is the larger non-vertical dimension by set B. The mobile objects of C (the objects of which width is the larger non-vertical dimension) like razor blades are very rare. Set C, therefore, is much smaller than sets A and B. Figure 24 The pragmatic bridge between rules L2 and L4 is established by the objects of A n B. If it goes from A (rule L2) to B (rule L ), the objects of C are irrelevant exceptions in the original set A. They do not contribute to the pragmatic bridge and can be neglected. What happens now if the pragmatic bridge goes from B to A? The application from L4/W4 to A n B is obvious. But how can the exceptions in C be explained? Has rule L2/W2 only been established for their sake? How and why did these exceptions to rule L4/W4 develop this sudden application of the width to the largest non- vertical dimension? By the arbitrariness of language? Such a conclusion can only be reached after the failure of all attempts to justify set C. As far as the width of a razor blade is concerned, I am given the choice between an arbitrary usage of this dimension (L4/W4 —> L2/W2) and a justified one (L2/W2 —> L4/W4). All other things being equal, I will choose the second solution. (ii) There is a more imperative reason to prefer a development from LAV (and the rules it subsumes) to L4/W4 to the opposite. Even if the above postulate proves to be wrong, in an actual situation of discourse the extent of all the dimensions of an object is not always known to the speaker. What happens then if he has to name a non-vertical dimension of an object with-
LENGTH, WIDTH, AND POTENTIAL PASSING 423 out knowing the other dimension? Will he abstain from naming it as long as he is not provided with the indispensable piece of information? Such a situation often occurs when facing the buildings in a street. Do you feel unable to speak of the length or the width of these houses without entering them to evaluate their second non-vertical dimension?9 I do not believe you do. If nobody has ever been confronted with such a dilemma, this means, I feel, that rule L4/W4 does not predominate in the use of length and width. Naming a non-vertical dimension of an entity of which one sees only one side is a very common strategy. A last example will show that, even if the extent of both non-vertical dimensions of an object is available, naming an object along one dimension prevails over naming it along two dimensions. According to rule Wv the speaker will certainly name the side of the table in front of him width. If asked afterwards to name the smaller side of the table he will be puzzled. Indeed, rule L4 exerts a selection restriction over rule L It prevents the speaker from naming the smaller non-vertical side of the table length. A common strategy with containers (or container-like) objects is to name this dimension depth (Vandeloise 1979). For the table in Figure 25, the only way for the speaker to name its two non-vertical edges will be to take a step backward. He has to rename the larger side of the table length according to rule L4, in this way keeping width available for the smaller side of the table. Rule L4/W4, thus, eventually comes to the speaker's mind. Note, however, that it is only triggered by more complex situations, when two non-vertical dimensions have to be simultaneously named. AB = Width 3 Length 4 BC = Width 4 J§L Figure 25 Thus, even if the postulate proposed at the beginning of this section is not accepted, a development from L/W to L4AV4 seems more likely than the reverse. If this development is accepted, one aspect of the distribution of length and width calls for explanation. Although L2 and L3 precede L4 in logical
424 CLAUDE VANDELOISE diachrony, this latter rule imposes constraints on the other ones. The purpose of these constraints, however, might be to preserve regularities previously observed in rules Lj and Lr Remember that length in these rules is almost always the largest non-vertical dimension. Furthermore, the constraints operate only in the complex case where the speaker describes the object along two non-vertical dimensions. Thus, this might be a retrospective effect of rule L4 on rules L2 and L3 at a more elaborate stage of description of the object involving two dimensions. A putative model of diachronic logical evolution is represented in the following diagram: Li/Wi 5. Discrepancies between Logical and Historical Time. Evidence from French Besides its spatial usages, the French adjective large (wide) can designate a generous man, as in sentence 3: (3) Le president est un homme large "The president is a generous man" Historically, the spatial and the "generous" meanings of large have different sources, the Latin adjectives latus and largus, respectively. Suppose that an inquiry in logical diachrony overlooks this information (as well as some synchronic clues such as the contrast between les largesses d'un homme large and *les largeurs d'un homme large). It might try to relate all the usages of large to the same impulsion and ... succeed! Indeed, connections between the spatial meaning of large and its "generous" meaning are easy enough to find. For example, a wide table bears more food than a narrow one; generous men in paintings always have un large sourire and open their
LENGTH, WIDTH, AND POTENTIAL PASSING 425 arms largement on a huge belly. A pragmatic bridge between both types of usage, then, might be readily postulated. Should this discrepancy between the logical and the historical evolution of large be used as an argument against logical diachrony? I do not think so. The aim of logical diachrony is to describe the systematic evolution of the distributions of words, its independence of historical time is thus not surprising. In fact, historical evolution does not only depend on systematic factors but on accidental ones as well. Therefore, historical and logical diachrony can differ. In the case of the two versions of large, what logical diachrony predicts is that because they are connected, an extension of the spatial usages of large to its generous meaning is likely. Therefore, one or several languages in the world might rely on these connections to extend the distribution of this word. Whether or not this development will actually occur is beyond the grasp of logical diachrony. The case of large in French is even more vexing. There, the predicted development does occur but not for the reasons alleged. As a matter of fact, the use of the same adjective for spatial and "generous" meanings is due to a phonological fusion. Note, however, that even if a semantic derivation of the "generous" usage of large from its spatial usages were historically wrong, it does not mean that semantic resemblance has not played a potential role in the evolution of this word. If semantic connection between the usages of latus and largus had been impossible, resistance to their phonological fusion might have been strong enough to prevent their fusion. Thus, the logical derivation predicted by logical diachrony is not completely without foundation. 6. Conclusion The metric definition of length/width does not stand up to an exhaustive analysis of the distribution of these dimensions. Their complete description establishes the decisive interference of a functional factor, the potential passing, with their distribution. This concept, whose different aspects are slowly and progressively acquired, cannot be said to be innate, as Bierwisch claimed for semantic markers like [±Maximum] and [± Second]. For heuristic purposes, I pushed the justification of the usages of length and width as far as possible, notably by establishing their logical diachrony
426 CLAUDE VANDELOISE on synchronic and cognitive considerations. This attempt certainly involves the risk of looking for justification where there is only accident and arbitrariness. A developmental approach to lexical categories, on the other hand, might allow an explanation of their formation when there is neither a prototypical node in the network nor a global highest schema subsuming them. Impulsions, then, might be a third device illuminating the structure of complex categories. Notes 1. I dealt with longueur and largeur (length/width) for the first time in 1979, at the University of Paris VIII. I was advised at that time by D. Carter and G. Fauconnier. This paper was next presented at the Katholieke Universiteit Leuven in April 1985. I would like to thank all the participants at this lecture and most particularly H. Cuykens, M. Dominici, O. Leys, C. Plantin, and B. Rudzka-Ostyn. I am also indebted to R. Dirven and B. Rudzka-Ostyn for reading through a first version of this paper. 2. For other criticism of Bierwisch (1967), see Dirven and Taylor (this volume), Vandeloise (1979, 1984, 1986) and Zubin and Choi (1984). 3. The third and fourth usage rules correspond roughly to the orientation and gestural imaging systems of the length/width in the terminology of Zubin and Choi. The relationship of length/width with movement is not noted by these authors. 4. Impulsion should be understood here as a technical term, the exact meaning of which will hopefully become clearer as my research proceeds. 5. This formulation is short-hand for: L : The length of a linear entity or of a path is its extent evaluated along its actual shape. W : The width cf a linear entity or of a path is its extent evaluated along its actual shape. As it has been explained above, rule W only applies to paths. 6. A common strategy in describing a house is to call its front width and its internal dimension depth, as in Figure 26. Many people, especially when walking along a house, also speak of its length. Zubin and Choi note that out of 100 respondents requested to compare houses A and B below, 38 answered that A was longer, 12 were unsure, and 50 preferred wider. cL T3 width street / A / B Figure 26
LENGTH, WIDTH, AND POTENTIAL PASSING 427 7. General orientation (Vandeloise 1984) is a family resemblance concept (Wittgenstein 1953) whose principal features are the line of sight, the direction of movement, and the frontal direction, characterized by the anatomical features of the human body. The features of lateral orientation include the lateral direction parallel to the line of the shoulders and a direction perpendicular to general orientation. 8. Even though both elements of a passing can be mobile, its linguistic representation construes this event in such a way that one element of the passing is taken as an immobile point of reference. 9. Cf. note 6.
On Bounding in Ik Fritz Serzisko University of Cologne 1. Introduction In his discussion of the semantics of the English verbs come and go Fillmore states that: Sometimes we find, in descriptions of exotic languages, grammatical categories associated with, say, directional affixes of some sort, that are named 'Toward Speaker' or 'Away from Speaker'.(1971a:54) It is these 'directional affixes of some sort' that will be discussed in this paper. We will use as an example the 'exotic' language Ik.1 Directionals or 'directives' are quite common in African languages and throughout the languages of the world (see e.g. DeLancey 1982 and 1985 for a description of directives in Tibeto-Burman). As indicated by the quotation from Fillmore, the semantics of these directional affixes is commonly described as identical to the semantics of the motion verbs come and go and the transfer verbs bring and take, for which Fillmore gives the following characterization: Go and take indicate motion toward a location which is distinct from the speaker's location at coding time. (1971a:55) Come and bring indicate motion toward the location of either the speaker or the addressee at either coding time or reference time, or toward the location of the home base of either the speaker or the hearer at reference time. (ibid.:6l) The following examples from Ik show that the category 'directional' is to a large extent a functional equivalent of come and go: (I)2 Nat-eta naa ima. "The boy came running." run-YEN PAST boy (towards speaker)
430 FRITZ SERZISKO Nat-uKota naa ima. "The boy ran away." run-AND PAST boy (away from speaker) Furthermore, both directionals and come and go can be used in non-spatial expressions, in which case come and go refer to a change of state rather than to physical motion. (For a detailed description of this and other types of idiomatic usages of come/go see Clark 1974.) (2) He went out of his mind. He came round very slowly. The motor went dead. The motor came alive again. (op. cit.:320) That directional affixes can be used exactly in the same way is demonstrated in the following examples taken from some Eastern Nilotic languages.3 (3) MOTION VERBS STATIVE VERBS KAR abON-un "to come back" a-jok-un "to become very good" abON-Or "to go back" ajok-ar "to become good" MAA a-suj-u "to follow hither" a-rok-u "to become black" a-suj-aa "to follow thither" TUR a-dOk-uni "to climb up towards S" a-bob-o-uni "to become sweet" a-dOk-ari "to climb up away from S" a-poti-ori "to become heavy" ATE a-pet-un "to kick towards S" a-la-Un "to become clean" a-pet-Or "to kick away from S" a-kwangi-ar "to become white" The occurrence of directionals in non-spatial expressions is usually explained in terms of metaphorical linkage. Thus e.g. Otaala (1981:54) states that: "The reading 'to become' is, at a more abstract level, connected with 'come' in the active verbs. The notion 'becoming' involves an involuntary motion from one state to another." But this statement provides no explanation of the exact link between 'becoming' and the spatial 'coming'. The setting up of a spatial metaphor requires a characterization of both the source and target domain as well as a description of the mapping between them (see Langacker 1986b:8). Moreover, in languages where directionality is expressed by means of verbal affixes the situation is different from that of English where we still have two lexical elements with independent lexical meaning which can be used metaphorically. In the languages under investigation here, the difference in directionality is expressed by grammatical elements which cannot occur independently. This is not to say that they may not have developed originally from verbs for 'come' and 'go', as DeLancey (1985) has shown for Tibeto-Burman Ian-
ON BOUNDING IN IK 431 guages. Also in Heine and Reh (1984) numerous examples are given for the development of these verbs into grammatical elements. In this paper, however, we will not investigate the diachronic origin of the directional affixes. We will concentrate on the present-day situation and try to give a unified functional description of directionals which accounts for both their spatial and non-spatial usage. As already mentioned, the language described in this paper is Ik, a member of the Kuliak language family, spoken by about 3000 speakers in north-eastern Uganda. Ik is a predominantly VSO language with a dominant Determinans - Determiner word order. It has suffixal case marking and a highly developed verbal derivational system. In addition to causative, habitual, passive and intentional derivational suffixes, there are two affixes - -uKot and -et - which serve the function of directionals. 2. Change, Boundedness and Directionality Langacker (1986b) makes a distinction between stative relations, which consist of "only one state"(p.4) and complex relations, which are to be represented as "a sequence of states"(ifc/d.). Complex relations can further be divided according to whether they are scanned summarily, i.e. whether they are construed "atemporally and accessed as a simultaneously available whole"(1985:lll), or sequentially, whereby "the process is seen as unfolding through time stage by stage" (ibid.). Given these distinctions, verbs can be analyzed as profiling a process, where a process implies the sequential scanning of a complex relation whose component states are taken as being distributed through time. (1986b:6) An ordered sequence of states can be schematically characterized as in formula (Fl), where r stands for a stative relation, t for conceived time, T for processing time and C for the conceptualizer. L c > C J {ibid.-.l)
432 FRITZ SERZISKO In this paper, I will use a simplified formalization in which a simple line represents the development in time and the states are abbreviated as S. (F2) t, S>S>S>S t 1 n Within the class of verbs two basic subclasses have to be distinguished "depending on whether they profile a perfective or an imperfective process'' (Langacker 1986b: 7), the difference being that In a perfective, the component states constitute a bounded series, and generally they involve some change through time.... By contrast, imperfectives are not specifically bounded, and all the component states are construed as being identical, (ibid.; my emphasis) The important notion here is 'boundedness', already introduced by Talmy (1978a). Imperfectives are unbounded, and they do not indicate a change of state. Using our simplified formula we may represent this by (F3) Perfectives are bounded, and there are at least two states A and B, where A =/= B. We represent this by: (F4) A B s, > s, > s 1 2 n 2.1. Directionals with Motion Verbs Motion verbs can be of two types: (a) those where the trajector, i.e. the moving entity, is the actor; e.g. 'come', 'run', 'swim', and (b) transfer verbs, where the trajector is the undergoer; e.g. 'buy', 'give', 'throw' etc. We exclude here transfer verbs of the type 'to carry' which entail a movement of both actor and undergoer. Langacker regards motion verbs as denoting a "special sort of perfective process" (1986b:8), which implies that he restricts himself to 'bounded motion' that can be characterized as having a starting point and an ending point, an origin and a destination — what I have been calling Source and Goal.(Fillmore 1971a:52) The process is considered as 'special' because "each component state
ON BOUNDING IN IK 433 specifies the relation between the mover and his immediate location" (Langacker 1986b:8). Moreover, as DeLancey observes, any linguistic report of a motional event will be made at a specific location, and the motion will be either in the direction of or not in the direction of that location. Thus deixis for motion verbs ... is an integral part, not of the event itself, but of any linguistic coding of the event. (1985: 380; my emphasis) The 'specific location' is not necessarily identical with the location of the speaker but can also be "somehow associated with either the central character of the narrative or a chosen reference point" (Fillmore 1971a:67). Following Fillmore we will use here the term DEICTIC CENTER (DC). Deixis with motion verbs can then be described as the relation between the location of the trajector and the DC, where SOURCE and GOAL as the location of the trajector at the initial or final stage of the process are of special importance. We will assume that in any description of a process of 'bounded motion' the DC is related to either source or goal, from which it follows that an indication of deixis also implies an indication of 'bounded- ness'. In Ik, deixis is expressed by the verbal affixes -uKot for 'motion away', which corresponds to 'go', and -et for 'motion towards', which corresponds to 'come'. In the following, the terms Andative (AND) for -uKot and Ven- itive (VEN) for -et are used as labels for the morphological category. Before we come to describe the use of these affixes, however, it should be pointed out that in Ik, as in most of the languages which have the verbal category 'directional', the verbs for 'come' and 'go' are not derived forms but are distinct lexical items. (4)4 na kono odo-we ats-uo loNotoma aw-e when one day-GEN come-3s enemy home-GOA nts-ie 3s-GEN "And one day there came an enemy to his house." (5) Kay-oo koto loNotoma go-3s then enemy "Then the enemy went." These verbs never occur with either an AND or a VEN affix, which indicates that a specification for deixis is part of the conceptual structure of these verbs, i.e., they are inherently bounded.
434 FRITZ SERZISKO The use of the directional affixes with other verbs of movement is demonstrated in the following examples: (6) bot-et-ia nyomotoka gwar-iao nda Nat-et-oni jump-VEN-ls car top-SOU and run-VEN-INF "I jumped off the car and came running." (7) na louk-wa en-ie roBa ats-at-ie when hyena-ACC see-SBJ people come-3p-SBJ Nat-uKot-uo run-AND-3s "When the hyena saw that people were coming it ran away." The verb Nates "to run" is neutral with respect to deixis, i.e., it is unbounded. It denotes, in its unmarked form, an imperfective process. If it is used, as in Ex.(6) and (7), to denote a perfective process, deixis has to be specified by means of the directional affixes. The verb forms Natetoni and NatuKotuo can therefore be regarded as being externally bounded.5 It should be noted that in the verb form botetia in Ex.(6) the VEN affix -et, which indicates 'motion towards', is used with a 1st person subject. This is additional evidence for our claim that speaker is not necessarily identical with the DC. Transfer verbs, unlike motion verbs, are always internally bounded, but they can also be either inherently specified for deixis or neutral, the term 'deixis* being used in a very broad sense. An example of a deictically neutral verb is dzwigwes "to trade" from which we derive (a) the source- oriented form dzigwesuKot "to sell", or (b) the goal-oriented form dzig- wetes "to buy". Examples of transfer verbs with inherently specified deixis are ma-es "to give" (source-oriented) and kanes "to take" (goal-oriented). They differ from motion verbs, however, in so far as they can occur with directional affixes, as will be seen in the following. (8) ma-ida-ka give-2s-PRF "You have given it(to him)." me-et-ida-ka give-VEN-2s-PRF "You have given it(to me)." (9) maa-ka I maa-Kota-ka give-PRF/ give-AND-PRF "He has given it(to someone)."
ON BOUNDING IN IK 435 me-eta-ka give-VEN-PRF uHe has given it(to me)." A source-oriented verb like maes "to give" entails the notion 'away', and therefore the use of the AND-affix does not add anything to the meaning of the verb. It may optionally be used to express a certain emphasis. The use of the other directional affix, however, introduces a notion not already expressed by the verb. In (8) and (9) the VEN is the only indication that the speaker is the goal of the action. This seems to argue for a direct relation DEICTIC CENTER = SPEAKER, if not explicitly stated otherwise. The following examples, however, show that such a relation can only be contex- tually determined and does not follow automatically. (10) me-et-ia lotoba bi-ke give-VEN-ls tobacco 2s-GOA "I gave you tobacco." maa-Kot-ia rag na bien-e give-AND-ls bull DET 2s-GOA "I gave this bull to you." In Ex.(10), where speaker and actor are identical, both directional affixes can be used. This implies that DEICTIC CENTER =£ SPEAKER. The meaning difference between these two expressions can only be determined if we take the whole context into consideration. This, however, would be beyond the scope of this paper. The following examples show that the goal-oriented verb kanes "to take" can also be used with opposing directionality. (11) kan-et-ia lotob. take-VEN-ls tobacco "I took a cigarette for me." kan-uKot-ia lotob take-AND-ls cigarette "I took a cigarette (and went)." Again, the use of the VEN corresponds to the goal-orientation of the verb 'take' and does not add anything to the meaning of the expression. The use of the AND affix, however, adds some additional information, in this case, the movement of the speaker AWAY from the place where the action took place. This example may also show that the actual interpretation of the directional forms is highly dependent on the context.
436 FRITZ SERZISKO To summarize: We have investigated motion verbs which denote a 'perfective process of bounded motion'. Deixis is an integral part of the linguistic report of such a process, where the DEICTIC CENTER may be related either to the source or to the goal of the motion described. If the DC is related to source then we receive the notion of 'motion away'; if it is related to the goal, the resultant notion is 'motion towards'. Verbs can be inherently specified for deixis in which case they have been characterized as either source- or goal-oriented. If they are deictically neutral they can be externally specified by means of the directional particles AND and YEN. Schematically we can represent this by (F5) A AND DC = motion away VEN source-oriented Ka-es "to go" ma-es "to give" NatuKotes "to run away" dzigwesuKot "to sell" B DC = motion towards goal-oriented ats-es "to come" kan-es "to take" Natetes "to run towards" dzigwetes "to buy" The function of the directional affixes AND and VEN can then be described as 'relating the DC to a boundary' which in the case of an imper- fective verb like Nates is functionally equivalent to 'setting a boundary' or simply bounding. This, however, is primarily a static notion. The traditionally ascribed dynamic interpretations 'motion away/towards' result from the combination of AND and VEN with motion verbs but are not an intrinsic part of their meaning. 2.2 Directionals with Stative Verbs There is no category of adjectives in Ik. Qualities are expressed by what we call here stative verbs. This is basically a semantic classification. So far we have no grammatical criterion to determine whether a verb is stative or not, as in English where stative verbs do not normally occur in the progressive form. The predicative and adnominal use of stative verbs is shown in the following examples: (12) TSagwa ema raw meat "The meat is raw."
ON BOUNDING IN IK 437 nta beD-i-i ema na TSagwa NEG want-ls-NEG meat REL raw "I don't like raw meat." (13) weena nyomotokaa fast car "The car runs fast." hona nyomotokaa na weena drive car REL fast "He's driving a fast car." The static situations represented by stative verbs can be conceived of "as existing, rather than happening, and as being homogeneous, continuous and unchanging throughout [their] duration"(Lyons 1977:11/483). They can therefore be related to unbounded, i.e., imperfective processes, where Sx = S r Schematically we represent this by (F6). (F6) Stative verbs can be bounded to indicate that a static situation is terminated. The formal means by which this bounding in Ik is expressed are the directional affixes, as the following examples show. (14) yoda nyokopo-ika full cup-PL "The cups are full." yod-aak-eta nyokopo-ika full-PL-VEN cup-PL "The cups are becoming full." (15) bara ida sour milk "The milk is sour." bar-uKota ida sour-AND milk "The milk became sour." (14) and (15) show that both AND and VEN can be used to derive inchoative6 verbs from stative verbs, i.e. we derive a perfective from an imperfective verb by indicating what has been called deixis with relation to motion
438 FRITZ SERZISKO verbs. With perfective verbs the function of directionals was described as 'relating the DC to a boundary', with imperfectives as 'bounding'. This interpretation is supported now by the usage of directionals with imperfec- tive stative verbs. It seems that directionals in Ik are what Talmy (1978a: 17) described as: grammatical elements suitable for co-occurrence with unbounded-type lexical items which therewith, in effect, trigger a cognitive operation of "bounding". By this operation, a portion of the specified unbounded quantity is demarcated and placed in the foreground of attention. The question of 'foregrounding' will be discussed in section 2.3., here we will restrict ourselves to the process of 'bounding'. Schematically we can represent this process by simply marking a point on the timeline whereby we derive (F6)' from (F6): (F6)' A SI>S = S1 = S0 = S -1 12 n This schema should be interpreted as a continuous series of states S _ > S j which leads to a point A representing the state S from when on any subsequent state Sx+1 is identical to Sx. In a less formal manner we can say that the bounding of a stative verb implies that the speaker conceptualizes the state as having a starting point with the preceding period becoming part of the base, but not of the profile. The description given here for inchoative verbs can be compared with the one given by Langacker for the participle broken as in a broken cup or this cup is broken: The base for these stative relations is a process predication such as break, which designates a continuous series of states distributed through time .... The stative participle derived from it, however, designates only the final state in the overall process. The evolution of a situation through time is thus a prominent facet of the meaning of the participle, but it is confined to the base and left unprofiled. (Langacker 1983:11/128) The main difference between participles and inchoatives lies in the fact that the former are derived from process verbs and profile the notion of final state while the latter are derived from stative verbs and profile the notion of initial state.
ON BOUNDING IN IK 439 (F7) process state S X break > broken S X . become full < full In the following section we will show that in Ik both stativization and dynaminization are expressed by the same means. The question arises how the notion of DEICTIC CENTER, which is bound to the distinction between AND and VEN, can be applied to the descriptions of stative verbs. This will be discussed in section 3. 2.3 Directionals with Other Verbs Directionals not only occur with motion and stative verbs but they occur with practically all verb-classes. The only exceptions known so far are the already mentioned verbs for 'come' and 'go'. Directionals even appear to be the most productive verbal affixes in the language, since there are hardly any sentences in a text which do not contain at least one such affix. But while the usage with both motion and stative verbs seems to be rather systematic, with other verbs the function of AND and VEN is more difficult to determine. This is partly due to their discourse related function, as was already seen in the discussion of examples (9) and (10). In some cases it may also be due to the fact that directional affixes have become lexicalized, e.g.: (16) tam-es "to think" tam-et-es "to remember" Apart from these rather rare lexicalized cases, however, we have found the following regularities: 1. The already described deictic usage with motion verbs. 2. The inchoative function, whereby inchoatives are derived from stative verbs. 3. A passive-like function, comparable to the 'foregrounding' of the Patient. This is mainly found with effective verbs. In some cases a distinction between the inchoative meaning and the passive is difficult to make due to the lack of clear defining criteria for stative verbs. Two alternative interpretations are possible: The AND-form either means: "X has become V-ed" (change of state) or "X has been V-ed" (pas-
440 FRITZ SERZISKO sive). Further research into the nature of verb classes in Ik is needed. The following complex sentence may serve as an example of this passive-like function: (17) a nts-uo iDits-an-e say-aa Jik PART 3s-COP beat-an-SBJ some-ACC even iDits-uKota be saa rob-e tumedo beat-AND PAST some people-GEN there "And some of them were beaten there also/some people were beaten there." In this example neither deixis nor change-of-state is involved. The profiled scene is identical in both cases, but there is a grammatical difference. In the first case, which could be explicitly paraphrased as: "It is that (they/he/someone) beat some there", the Patient saa-yaa is overtly marked as such by the ACC form. This implies that there is an Agent, expressed in this predication by the impersonal pronoun an. In the second case, no subject is marked in the verb and the object remains unmarked. Since case marking in Ik would be a topic of its own, it may suffice to say that absence of ACC marking does not necessarily imply that promotion, i.e. a change of grammatical relation, took place. All we can say is that the only participant in the sentence is the Patient and that the passive translation, given by my informant, seems to be justified. The overall impression is that this construction is used to put some emphasis on the fact that the Patient was affected by the action. We can regard this as a foregrounding of the Patient. 4. A resultative function, where directionals are used to emphasize the result or the completion of an action. (18) nKaKot-uo feta tiriebpiri bul-uKot-uo kiJ eat-AND-3s sun completely empty-AND-3s land "It was eaten by the sun ; the land was finished." The translation in this example is also passive, but there is a syntactic difference. In (17) above, the only participant expressed overtly in the sentence was the Patient, in (18) it is the Agent. The meaning of nKaK-uKot may best be paraphrased as "eaten up", as in the following example: (19) wet-uKota cue drink-AND water "The water is completely drunk/is finished."
ON BOUNDING IN IK 441 Observe that 'resultativity' and 'emphasis on Patient' are related notions. They both have to do with the final state of the action or process, i.e. with its boundary. All the apparently quite divergent uses of directional particles can now be explained along the same lines. They all are instances of bounding; in those cases where the verbs denote a perfective process, i.e. where they are inherently bounded, bounding means 'reference to a boundary' while with imperfective verbs, bounding means 'introducing, setting up a boundary'. 3. Deictic Center with Other than Motion Verbs So far we have shown that directionals are used to indicate bounding. What remains is the question: How does the difference in directionality come in? To answer that question we must take another look at the deictic function of directionals. With regard to motion verbs deixis actually involves two points: (1) the Deictic Center, i.e. the point where the observer/speaker actually is or which has been chosen as the vantage point from which the motion is decribed and (2) the Referential Center, i.e. the place where the moving entity (the trajector) is at time t . In an abstract construal of a motional event we can identify the RC with the starting point (Source) since the minimal requirement for movement is to leave the initial state. The difference in deixis results from the relation between RC and DC. If the scene is observed from the Source, then RC = DC, and in this case the notion 'away from' is evoked. If the scene is described from the end- point, the Goal, i.e. if RC =£ DC then the motion is seen/described as 'towards'. Schematically we represent this as follows: (F8) A B motion away RC=DC motion toward RC =£ DC If we apply this to stative verbs we encounter a problem insofar as with inchoative verbs only one boundary is involved. This boundary, in turn, can be regarded as either Source or Goal depending on the viewpoint of the conceptualizer. This may be demonstrated in the following examples, which are not made-up sentences but taken from texts.
442 FRITZ SERZISKO (20) buDam-et-es-uo kUa dark-VEN-INT-3s land "The world was getting dark." (21) budam-uKota kUa nci-e dark-AND world Is-GOA "The world is completely dark to me." The semantic difference between these two sentences becomes clear if one considers the larger context in which these expressions occur. In (20) the speaker is talking about the process of 'becoming dark': light is fading away but it is still not dark. In (21) the state of darkness has already been reached. Given that all points in an imperfective process are identical, we can reduce the state expressed by budam 'to be dark' to the point S , from which point on the expression It is dark is true. This point is taken as RC. The placement of the Deictic Center depends on whether the speaker conceives this state as (a) not yet achieved, in this case RC =£ DC, or (b) as already achieved, i.e. RC = DC. This difference in viewpoint is expressed by the use of the appropriate directional affix. In the first case the VEN affix indicates that the state has not yet come about; in the second case, the AND indicates that the state has been achieved and that it now holds. Schematically we can represent the two situations as follows: (F9) S = budam -(20) DC =£ RC = (21) RC = DC The difference in viewpoint expressed by the AND and VEN affixes in Ik is thus reduced to the question of identity or non-identity of RC and DC, which corresponds to what Clark found out about the idiomatic use of come and go in English evaluative statements. She says that "Movement toward is related to presence at the deictic center, while movement away is related to absence from the deictic center" (1974:331). The Deictic Center in this case is the 'normal state', and in an expression like He went mad the normal state is considered to be the starting point of the given process (i.e. RC = DC). By contrast, in the sentence He came down (from a 'high') the starting point of the process, i.e. the Referential Center, is different from the normal state (i.e. RC =£ DC). The following schema demonstrates that identity or non-identity of RC and DC can account for such cases as well.
ON BOUNDING IN IK 443 (F10) go mad RC = DC ^ come down from a high RC ^DC 4. Summary and Perspective Directionals are traditionally defined as denoting 'Motion Towards/ Away from Speaker'. We have argued that (i) the reference to Speaker should be substituted by reference to a Deictic Center which, if not expli- citely defined otherwise, may be identical with Speaker; (ii) Motion is not an integral part of the conceptual structure of directionals, it only occurs in combination with motion verbs. The deictic meaning expressed by directionals is primarily a static one. They denote the correlation between the Deictic Center, ideally the Speaker, and the Referential Center, which is identified with the starting point of a state or process, and the difference in 'deixis' results from either identity or non-identity of DC and RC. The function of directionals has been described as indicating bounding, i.e. an operation by which an unbounded process is demarcated. Lexical items, in our discussion of predicates, may be inherently bounded. In this case bounding implies 'reference to a boundary'; if the predicate is inherently unbounded it implies 'setting a boundary'. The discussion of the use of directionals with other than motion or sta- tive verbs has shown that their usage may be described within at least three different domains of traditional grammar. 1. Deixis: This is the domain to which directionals traditionally belong. 2. Aspect: 'Resultative', 'Completion' belong to the domain of Aspect or Aktionsart rather than to Deixis, but as Fuchs (1986) has quite convincingly shown, aspect may be related in a direct manner to deixis. Rudzka-Ostyn, in her description of Slavic aspects, comes to the conclusion that "perfectivity is basically a matter of bounding" (1986:48). Compare also the following statement from Hopper & Thompson: An action viewed from its endpoinl, i.e. a telic action, is more effectively transferred to a patient than one not provided with such an endpoint. (1980:252)
444 FRITZ SERZISKO A development of directionals into aspect markers during a process of grammaticalization would therefore not be unexpected. 3. Voice: The correlation between directionality and aspect has already been shown by DeLancey, but he has also pointed out that "there is a semantic parallel between the structure of aspectual and voice distinctions" (DeLancey 1982:168), involving a category which he calls viewpoint. Both aspect, especially telicity, and voice are directly related to the degree of "effectiveness with which an action takes place" (Hopper & Thompson 1980:251), i.e. transitivity, and in some languages directionals tend to become a marker of transitivity. They may be used to 'attract' the patient or an otherwise affected participant to the predicate. This seems to have happened in Murle, although our data is rather sparse, where directionals have developed into transitivizers. "When the directional suffix is used on the verb, then there is no location suffix on the location word. If the location marker is used on the noun, then there is no directional marker on the verb" (Arensen 1982:79). Further research is needed to determine more fully the function of the directionals in Ik. The different verb classes have to be defined and the correlation between directionals and case marking more satisfactorily investigated. Notes 1. The data presented here were collected during a field trip in 1985. The research was sponsored by the Deutsche Forschungsgemeinschaft to which I want to express my gratitude. I would also wish to thank the members of the Cologne Institute for African Studies for valuable comments. My special thanks go to Brygida Rudzka-Ostyn for her comments and constructive criticism. 2. The transcription used here is a preliminary one. The phonology and especially the question of tonology need further investigation. The following conventions and abbreviations have been used: AND = Andative(Itive) GOA = Goal REL = Relative ACC = Accusative INT = Intentional s/p = sing./plural COLL = Collective LOC = Locative SBJ = Subjunctive COP = Copulative PL = Plural SOU = Source GEN = Genitive PRF = Perfect VEN = Venitive(Ventive) CAPITAL CONSONANTS CAPITAL VOWELS Raised Vowels = Ejectives = Open vowels = silent vowels
ON BOUNDING IN IK 445 3. KAR = Karamojong, see Mantovani (1985) MAA = Maasai, see Tucker & Mpaayei (1955) TUR = Turkana, see Dimmendaal (1983) ATE = Ateso, see Otaala (1981) 4. Most of the examples given here are taken from texts and the translation is the one given by the informant. I did not change these translations, which may account for some akwardness. 5. For the distinction between internal and external bounding, see also Rudzka-Ostyn 1986. 6. We use 'inchoative' here in a restricted sense referring only to constructions of the type "to become X".
A Discourse Perspective on Tense and Aspect in Standard Modern Greek and English Wolf Paprotte University of Miinster 0. Introduction* This paper offers only a partial explanation of the Tense - Aspect - Mood - Voice system of English and Standard Modern Greek (SMG). Mood and voice are not analysed although language-specific, hierarchical orderings of the grammatical categories of the verb point to the interdependence of these categories. The spatial metaphor of the time line and its mathematical explication1 will not be accepted as the basic semantic structure underlying tense and aspect. As Schnelle (1981a) notes, a tense semantics ought to be superseded by psychologically relevant semantic notions. This is due to the fact that the form and the structure of a language and its grammar are constrained by basic operational and organisational principles of the human perceptual apparatus and the cognitive system. Thus, while avoiding the mathematical explanation of time, a link to underlying perceptual factors and their conceptual reflexes is sought: 'Temporal concepts relevant to the semantics of tense are only acceptable insofar as their representation is specified in terms of the organisation of mind" (Schnelle 1980:332).2 In the majority of recent studies, tense and aspect are considered as sentential operators. Not many researchers have looked into the discourse function of aspect and tense, although traditional stylistic rules such as con- secutio temporum, and requirements on temporal anaphoric relations should have pointed to the necessity of analysing the discourse function of these categories. An attempt is made to draw a line joining the © Wolf Paprotte
448 WOLF PAPROTTE psychosemantic bases of tense and aspect to their functions in discourse as factors promoting coherence. The evolving picture is necessarily incomplete and in need of a more finely grained analysis; however, it is such that it will easily accommodate the complexities of Mood and Voice as categories interacting with Tense and Aspect. 1. Tense and Aspect in SMG: Establishing the Relevant Categories SMG verbs are morphologically marked for the aspectual distinction between perfective and imperfective.3 SMG verbs have two stems, the so called present stem (imperfective), used in the derivation of imperfective past and non-past verb forms of the active and passive voice, and the aorist stem (perfective) from which derive past and non-past perfective forms of the active voice and the singular perfective imperative of the passive voice. The different moods of SMG are also aspectually distinguished. Morphologically, aspect is the overriding verbal category of SMG, a category which is generally not subject to category shifting.4 Tense, however, may be shifted. The range of uses of the SMG imperfective present is as varied as the English simple present (cf. Mackridge 1985: 103). Mirambel (1942) notes that SMG aorist (perfective past) forms can be used to refer to non- past events while the corresponding paratatikos forms (imperfective past) cannot be used in such a manner; cf. xatdtaxfkx (perf.past) "I understand"; te^eicooa (perf. past) rr| Soi^eid "I now stop my work"; versus te^ei- cova (impf. past) tr] Soutaid "1 am about to stop my work"; Jieivaoa (perf. past), "I am hungry". The aspectual qualification clearly overrides the shiftable temporal marker. For this and further reasons, Mirambel (1942: 38) concludes "on ne peut concevoir en grec un temps qui serait exprime en dehors de l'aspect" and suggests an ordering of the categories of the SMG verb in terms of an Aspect > Voice > Mood > Tense scale. For English, the scale Tense > Mood > Voice > Aspect is implicit in many descriptions (e.g. Givon 1984 : 295; cf. the traditional "Aux" rewrite rule in generative grammar). Tense and aspect in SMG and English thus have reciprocal values on the scale of verbal categories. Traditional grammars of SMG treat forms such as va x™Jir|Ott), va Jiidaco "I should hit"; "that I catch" (perfective stem plus non-past, indicate active ending) as aorist subjunctive. As a consequence, the SMG verb paradigm is defective in these descriptions and lacks a perfective present tense form (cf. Tzermias 1969). The rationale behind the aorist subjunctive
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 449 draws on a curious constraint. Such forms are restricted to occur in subordinate clauses and are always preceded by one of a small list of preparticles, "converters"5 like av, otav, ojtoiog, aqpou etc. However, SMG perfective and imperfective future forms are also preceded by a tense and mood formant (fta). A more consistent interpretation (Barri 1981) would therefore treat va as a modal particle signaling "subjunctivity" and consider the remaining verb form as a perfective, indicative, present tense. There is thus an exact interpretive parallel between future fta + present indicative form (either perfective or imperfective) and volitive, deliberative, potential va + present indicative form (either perfective or imperfective). The limited distribution of the perfective present, indicative forms (xTimr|aa), judaca) in subordinate clauses depends on co-occurring converters like va, \ir\, av etc. With this neat reanalysis of the SMG verb paradigm, all tenses, moods and voices are either perfectively or imperfectively marked. There is a further complication that is difficult to solve: do we deal in SMG with an aspectual triad of imperfective, perfective and perfect? The SMG perfect has been variously considered as a tense, an aspect or a mixed category: an aspectually marked complex tense or an aspect with temporal implications (Armstrong 1981, Comrie 1981b). Comrie (1976) and Mack- ridge (1985: 116), who define aspects as ways of representing the internal temporal constitution of an event-situation, say of a perfect that it is "an aspect in a rather different sense from the other aspects" (Comrie 1976:52). It is said to express a relation between two time-points; its aspectual meaning is given as "continuing relevance", "prior being the case of a situation", "persisting existence of a situation", and "temporal closeness, recentness". This analysis has the serious drawback of inconsistent criteria in the definition of aspect: perfect is defined with respect to external time-points, perfective and imperfective are defined relative to the internal temporal contour of an event. However, Comrie might argue for a mixed category and his argument may be strengthened by considerations of form: the standard form of the SMG perfective consists of a tense marked form of exo) "1 have" and the perfective participle as perfect formant. 'Exoa has no perfective stem and could therefore be considered as a pure imperfective and the perfective participle as the perfect formant. The question now is: Will the conjoined aspects neutralize each other and in fact force a reinterpretation of the designated event in terms of relating it to external time points, or will the meaning of the perfect simply be a merger of two prototypical aspectual meanings?
450 WOLF PAPROTTfi Mackridge (1985) assumes a meaning merger (p. 118) but his discussion of the grammatical meanings involved again makes use of the notion of external time points (p. 116). His decision for assuming perfect to be an aspect and for attempting the merged meanings approach may be supported by the fact that perfect forms can be replaced by perfective forms only; the converse is not true (Tzermias 1969: 139; Mackridge 1985: 118). Taken together, these points suggest that the assumption of a third aspect, "perfect", possibly as a subcategory of the perfective, is more consistent with the facts of SMG than treating it as a tense or a mixed category. Furthermore, the use of perfect forms seems restricted to "perfective" (Aktionsart) verbs in more modern terminology, to activity, accomplishment and achievement predication types, as Comrie noted (1976: 62). Friedrich (1974), in his masterful and perspicacious study of Homeric aspect, describes perfect as an aspectual category signifying "a state of the subject resulting from a realization of the process referred to by the verb" (p.16), and suggests renaming it "consequent" (p.17). The feature REALIZATION, which in his taxonomy distinguishes aorist and perfect, turns out to be a complex feature that combines tense, (perfect forms inherently imply that some event was completed in the past or at least begun in the past and leads up to the present through an extended series of acts), viewpoint aspect (completion), Aktionsart or situation type (state), and voice (intransitivity). Because of these differences between perfect and the basic aspects, the perfect is called a mixed aspectual category (p. 19). However, on semantic and typological considerations, Friedrich groups it with the aorist rather than with the imperfective. His analysis also holds for the SMG perfect which is a semantically multidimensional category intersecting tense, voice, situation type and aspect. Furthermore, although the perfect seems to neutralize the perfective-imperfective distinction, it is relatively close to the perfective in that it never denotes progressive or iterative event-situations. Its strong resultative component presupposes temporal characteristics of an anterior onset of action but signals consequent aspect. I feel therefore justified in considering the perfect to be one of three SMG aspects. SMG tense categories: From a morphological perspective, past and present tense forms share similar inflectional processes but are distinguished by contrasting sets of endings and the presence/absence of an augment. Forms marked for modality and futurity contrast with past and present tense forms in their periphrastic construction; they depend on the pre-
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 451 sence of semimorphological formatives fta, va. This morphological fact supports Lyons' discussion of futurity as a modal notion (subjunctivity of likelihood, of possibility) rather than a treatment of future as a deictic tense (1977 : 688). The argument is based on an empiristic epistemology and as such is impeccable; however, epistemological implications are also carried by past and present tense, so that a legitimate and sound case can also be made for future as a tense. To simplify and shorten matters here, I opt for the traditional classification of future as a tense.6 2. The Semantics of Perfective and Imperfective Verbs belonging to one of the well known predication types: activity, state, achievement, and accomplishment predications, have or can have perfective and imperfective forms in SMG: (1) paint the nativity (activity): ^(OYQdcpi^e (impf. past) ttvv yevvTjoT] xov Xqioxov J;a)YQd<piae (perf. past) ttvv yevvx\or\ xov Xqioxov (2) love her husband (state): ayajtouae (impf. past) xov avbQa xr\c, aydmjae (perf. past) xov dv6Qa xtjc; (3) reach the summit (achievement): ecpftave (impf. past) arnv xoQu<pr| ecpftaae (perf.past) ottiv xoQu<pr| (4) play a Mozart sonata (accomplishment): ejtai^e (impf. past) ^iia aovdxa xov Mozart ejtai^e (perf. past) ^iia aovdxa xou Mozart Neither the Aktionsart potential of SMG lexical verbs (e.g. swim vs. die) nor the predication type of the verb, its arguments and other syntactic constituents will determine the morphologically marked choice of aspect. Notice for example the activity-accomplishment conversion brought about by the prepositional phrase in (5) (5) (a) John walked, (activity) (b) John walked to the station, (accomplishment) (Dowty's example, 1986:39). Yet both sentences could occur with either of the SMG aspects. The speaker's choice of perfective, imperfective and perfect seems to be the main factor determining the aspectual class of a sentence. Thus, in (6)
452 WOLF PAPROTrf: (6) Exeivr) \iaq ^ayeiQeue, cpoovxi^e xa £coa, cpouovi^e r\ enXeve xa Qov%a \ia<; axo taxyxd&i. (all verbs impf.past). "She cooked for us, looked after the animals, baked or washed our clothes far down in the valley." all four predications are marked for intrinsic duration and homogeneity; even washed our clothes, which would appear a clear case of an accomplishment like build a house, acquires an iterative or an habitual status of a repeated activity. Only with a perfectivity marking would predicates such as (pQOVxiae xa £coa or en\vve xa Qov%a become inhomogeneous accomplishments. In Standard Modern Greek, the choice of aspect also forces state- achievement and state-accomplishment transitions. Dominate, love, want, hate generally count as state predications. Perfectively used, they become achievements, respectively accomplishments: (7) (a) Ajio jiai&i xov Fiawr) xov ayajiouaa. (impf.past) "Ever since childhood, I (have) loved John." (b) Tov ridvvri xov aydjiriaa (perf.past) \i6Xiq xov ei&a. "I loved John the moment I saw him." (c) Ajio xo 1680 ni/Qi xo 1820 o xouoxixog axo^og xuoiao- Xouae (impf.past) xo Aiyio. 'The Turkish fleet dominated the Aegeian Sea from 1680 to 1820." (d) Tr] oxiy\ir\ jiou xi)QidQXT)ae (perf.past) xo \iioo, otax 6Xa XaOrjxav (perf.past). "The moment hate (started to) dominate, everything was spoilt." Figuratively speaking, perfectivity transforms an event-situation into a contour-enveloped, bounded figure unless the Aktionsart potential of the predication already displays this contour; so that as a rule, state and activity predicates will become accomplishments or achievements if used in the perfective aspect. Perfective use of accomplishments and achievements does not change the predication type but functions as an explicit signal for it. In general, the imperfective aspect transforms accomplishments into activities or states while states and activities remain unaffected, cf. (8): (8) (a) ;Oxav r\\iovv hixqoc; (impf. past) ave(3atva xa axa^ojidxia xadrmeoivd. "As a child, I ascended these stairs daily." (accomplishment)
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 453 (b) O r\kioq fiaoikevEi (impf. non-past) Suxixct. "The sun sets in the West." (state) But note (9): (9) (a) *H JteiQCt yrvnovoe (impf.past) xa xujiaxa. "The stone hit the waves." (achievement) (b) Mexti xy]v exQY]^Y], Jtexoec; yrvnovoav xa xtijiaxa (imp. past) yia oaoec;. "After the explosion, stones hit the waves for hours." Achievement-state transitions brought about by the imperfective aspect present some difficulties. If the designated event-situation is per- formable only once as in 9(a), the perfective is obligatory unless a (contex- tually induced) reinterpretation of the verb is possible. The punctuality of a single act as in hit the wave, combined with a definite, singular subject also seems to demand a perfective verb form. On the other hand, pluralized arguments of the verb, an adverbial of indefinite frequency count or duration (e.g. for hours) as in (9b), will oblige the speaker to use the imperfective; the achievement predicate will be transposed into an activity (or process). The examples of (9) show that there are limiting factors for the freedom of the SMG speaker to choose a viewpoint aspect: in the case of achievement predications, the possibility of choosing the imperfective is partly dependent on the availability of a contextual reinterpretation of the actual characteristics of the situation denoted. In a default interpretation achievements will demand the perfective aspect. However, contextual information which cancels default assumptions will render examples such as 9(b) natural. It has been observed that perfective forms co-occur with adverbials denoting a point in time and adverbials of cardinal count. Sentences containing perfective verbs will readily admit insertion of such adverbials. Imperfectives on the other hand co-occur with adverbials of indefinite frequency count or duration7 or such adverbs can be inserted into sentences with imperfective verbs (ll).8 (Mourelatos 1981; Armstrong 1981; Mack- ridge 1985). Many scholars have noted the close parallel between the mass-count distinction and the distinctions between perfective and imperfective aspect, and between state and activity predications on the one hand, and event (achievements and accomplishments) predications on the other (Allen 1966; Armstrong 1981; Mourelatos 1981; Bach 1986). These parallels are assumed on two grounds:
454 WOLF PAPROTTE 1. characteristics of mass and count nouns are shared by state and event predications; 2. identical underlying cognitive mechanisms (packaging and grinding) hold for nominal as well as for verbal expressions. Ad 1: Let me assume that aspectual distinctions, predication types and the distinction between count and mass nouns are based on a general, human cognitive capacity of conceptualizing entities and event-situations as individuals or as stuffs/masses. Count terms, specify individuation and divide their references (Quine 1960:91); they also include an inherent determination of what "counts" as a single instance, and what counts as another instance. Due to this inherent individuation of count terms, the sum of a number of individuals is a plural set. Cumulative reference does not hold: the single individuals do not add up to another individual. Similarly, events (achievements and accomplishments) and perfectively marked verbs divide their reference. Mourelatos (1981), therefore, suggests a nominalization test for event predications in English. (10) (a) Vesuvius erupted three times. < > There were three eruptions of Vesuvius. (b) Mary capsized the boat. < > There was a capsizing of the boat by Mary. In the transcription equivalents, the occurrences are either counted or governed by expressions or terms that presuppose that the occurrences are identifiable and countable (Mourelatos 1981: 204). Count - quantifiability is thus used as a simple and abstract criterion for distinguishing predication categories: all and only event predications allow for count quantified transcriptions. Unfortunately, this nominalization test does not work for SMG. Nominalizations are not possible. However the co-occurrence restriction of perfective verb forms with cardinal count adverbials may be a substitute for the Mourelatos test. Mass terms refer cumulatively. It is true of a mass term X that any sum of parts which are X is also X (Quine 1960: 91). This rests on the hypothesis of homogeneous reference: from the point of view of linguistic use, mass terms are not supposed to have minimal parts. Water simply divides into water; (up to a certain point of division; cf. Gathercole's discussion of the Minimum Parts Hypothesis, 1986). Mass terms thus have no inherent determination as to what counts as a single instance. Mass terms provide only a means of measuring for amounts of that stuff.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 455 Transposed to the realm of event-situations, Mourelatos' results fit well into this picture. In the case of state and activity predications, the transcriptions lack the indefinite article in the nominalization equivalents. They are mass quantified in that they resemble existence statements of the type There is snow on the roof, or There is bauxite in these mountains, and can be modified by adverbials such as little, much, enough etc. In English, state/activity predications are typically transcribed with deverbative nouns: John hates liars, (state predication) < > There is hate by John of liars. (Mourelatos 1981: 206 f.) The strong co-occurrence restrictions of SMG perfective and imper- fective forms and adverbials can therefore be explained by analysing SMG perfective forms as count-quantified event predications and imperfectives as mass-quantified state or activity predications. Consequently, perfective SMG aspect may be said to mark situations as occurrences of individuated events and to denote countable action. Imperfectives, on the other hand predicate of event-situations that they behave analogously to nouns denoting stuffs. Whereas adverbs of cardinal count may be assumed restricted in occurrence to perfectives because aorist forms are by themselves countable events, indefinite frequency count or duration adverbials qualify imperfec- tive predications in an analogous manner as amount expressions qualify non-count nouns. The widely accepted general definition of aspect "as a way of conceiving the passage of action" (Friedrich 1974:2) as well as more specific definitions such as aspect "deals with the temporal values inherent in the activity or state itself" (Jakobson, quoted in Friedrich 1974:2) will also allow the mass-count distinction to be translated into defining criteria which make use of the notion of the truth of a sentence with respect to temporal intervals (cf. Dowty 1986:42). For an explication of the perfective, I agree with Dowty in not distinguishing accomplishments and achievements because at least for SMG this distinction does not play an important role. The temporal extent of accomplishments vis a vis "punctual" achievements is not, as Dowty points out, a clear-cut criterion and furthermore relates to characteristics of the real world situation rather than to the aspectual marking. On similar grounds, states and activities ought to be grouped together in SMG. Such a regrouping will reflect the clear parallel between indivisibility and divided reference in the realm of count-entities and additivity of mass entities: No proper part of a man is a man; and no proper part of building-a-house is a
456 WOLF PAPROTT£ building a house. The addition of some water to water is water (but the addition of a horse to a horse is two horses); and "the fusion of two runnings is a running but no two dyings are a dying" (Bach 1986:10). Ad 2: It has frequently been observed that any mass and count noun can be used as a count, respectively mass noun, (11) (11) (a) Eivai koXv outoxivtito y^a ta hecpxd tou. "That's a lot of car for the money." (b) H Bauaoia eivai cprmia^ievr) Yia TlS TcoXkeg ^ijujoec; tr)c;. "Bavaria is famous for its many beers." The processes enabling speakers to switch back and forth between the categories have been dubbed grinding and packaging (Bach 1986:10). The SMG verbal system can be said to employ aspects as morphological markers which signal packaging or grinding of event-situations. On the basis of the evidence for predication type transpositions and nominal category switches, we may assume that packaging and grinding are universal processes spanning the word-class differences between verbs and nouns. A minimal assumption about the conceptual consequences of these processes is that the concepts are prototypically structured and allow for degrees of countability or degrees of "massiness". A bag of apples may thus seem to be more count and less mass than a bag of flour; knocking at the door more count than running a mile. While packaging and grinding account for the fact that verbs and nom- inals can be switched from their most natural category to another including non-standard uses of aspectual grinding (cf. Smith 1986), the assumed pro- totypicality characteristics may explain the difference between rather stable, central, category membership of predications and a continuum of cases mainly induced by the other constituents of the sentence: thus notice as an achievement will not normally indicate duration. / noticed it for a while may at best be interpreted as iterative (Carlson 1981:35f.). But Some time passed before all were noticing the recurrent blips, sounds much more acceptable and has all the characteristics of an activity predication. Packaging and grinding can now be used to explain how an SMG speaker can talk about an event situation from more than one viewpoint and choose (almost freely) between different perspectives.9 The conceptual consequences of these processes, I suggest, are ICMs (idealized cognitive models, G. Lakoff s term) that represent highly abstract characteristics of mass or count quantification which are superimposed on the lexical "mod-
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 457 els" of the predicate, its arguments, adverbials etc. thus producing pro- totypicality effects; that is, some event-situations will be easier to grind than others which normally resist the process. Achievements are a case in point: it is difficult to grind them by using the imperfective aspect. Conceptual representations of event-situations are neither limited to proposi- tional structures nor to primitive semantic atoms; they include perceptual information (e.g. imagery), procedural notions, episodic facts etc. Recent attempts at providing an imagery based analysis of aspect (Langacker 1982b; Jessen 1974) therefore demand special attention. I also assume that concepts may be made to vary as to depth and level of specificity, i.e. as to complexity or informational load on a given occasion of use. For further assumptions about characteristics of concepts cf. Paprotte (1985). 3.1 The Psychosemantic Basis of Aspect Sentential aspect presents an event-situation as being either mass- quantified or as countable. Situation type (Aktionsart) contributes little in SMG towards specifying the possible aspectual perspective a speaker chooses. In that sense, aspect is not an extralinguistic category which is reflected in language, similar to natural gender, number, size or shape. Rather, it results from cognitive acts of a speaker and is what has been called the viewpoint aspect. That is, the speaker chooses an aspectual perspective and determines an event-situation as either mass or count quantified. When looking for human information processing analogues of aspect, the mechanisms postulated should be general enough to explain both sentential aspect and the discourse functions of aspect. Schnelle (1980; 198la,b) was one of the first scholars to explain tense and aspect phenomenologically and relative to the human information processing system. Although I agree with his approach, I wish to depart from his results in two points: I retain tense as a deictic category and consider aspect a predicate of event-situations; and I base aspect on a different mechanism from that assumed to be the basis of tense. The qualitative distinctions which aspects may predicate of event situations are countability (divided reference, partitivity) and "massiness" (cumulative reference, additivity), reducible to underlying perceptual and cognitive factors. Traditional terminology may lead the way: note that the perfective aspect has also been called bounded and the imperfective unbounded. Greek grammarians talk about the synoptic value of the aorist; it serves to "package" an event situation and to present it as a form that is viewed as a whole, as a totality. The
458 WOLF PAPROTTE imperfective aspect on the other hand presents an event-situation as homogeneous stuff that extends but begins and ends only by implication. To explain this difference I draw on a parallel between aspect and the well-known figure-ground distinction. When a figure (in the sense of an intuitive notion of "recognizable form") is drawn inside a frame, it specifies a bounded bit of space inside that frame as positive, and sets it against the unbounded, complementary shape that is its negative space, the "ground" or "background" to that figure. From an information processing point of view, the figure-ground distinction plays a major role in explaining how bits of information can be selected and focused upon and processed. As there is always more information available, more to know, to see and to hear than is necessary, the significance of informational stimuli is (in part at least) determined by the figure-ground distinction. Some parts of the perceptual field stand out as the figure; others recede to form the ground. Informational figures also form units, have borders, and cohere. They are differentiated from and located against a ground. Portions of the figure are treated perceptually as belonging to the same object; portions of the ground appear to be independent; but, as Ehrenstein (1954) points out, figure and ground unite functionally into a coherent whole. Three factors interact in determining what becomes a figure, what is thus found more salient and important to a human user: (1) higher level control and intentional monitoring; (2) design principles of the primary perceptual system; and (3) characteristics of the stimuli, such as changes in the visual field (movement) or changes in the stimulating energy potentials and higher levels of energy (bright light, loud sounds, strong odours). Ehrenstein cites such factors as form, colour, extension/size and Gestalt regularity affecting the relative strength of figures (1954: 304 ff.); similar acoustic factors affecting auditive figures have been found. What will become a focussed figure also depends on higher level control and monitor factors. Neisser (1976) points out that a perceiver allocates attention (or selects) on the basis of anticipatory schemata and active exploration of the informational input. Schemata thus act as format specifying kinds of information that are to be interpreted coherently. They also act as plans for finding out about perceptual situations in order to fill in the schema; information that does not fit a schema goes unused due to the inherently selective nature of perception. Neisser, therefore, takes attention and perception as identical, directed, goal oriented activities: We choose
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 459 what we see, hear and understand by anticipating the structured information the stimuli will provide. One important kind of structuring is achieved by the figure-ground specification in two ways. First, stimuli that are marked with respect to their salience, relevance and coherence are figures which fill in or alter the schema. Figures have informational significance. Secondly, figures are structured as informational Gestalt, that is, a figure is individuated and discrete and remains categorially stable over time. In that sense, foregrounding and packaging are the same. Countability (possibly also definiteness) results, therefore, from figure characteristics, but will not sufficiently explain the psychosemantic basis of aspect unless the salience function of figures within anticipatory schemata10 is recognised. Aspect, I assume, not only increases the probability of noticing relevant event-situations but directs the integration of anticipated event-figures into the discourse schema (discourse representation) of the hearer. Aspect may, therefore, be seen as a strategic discourse control and planning mechanism — rather than as a sentential countability indicator — that shapes event-situations as conceptual figures. To sum up: Aktionsart or predication type is a concept that is relatively insignificant in the explanation of SMG aspectual distinctions. Co-occurrence restrictions between types of adverbials and aspectually marked predications point to the count-mass distinction as the conceptual basis for aspect: perfective forms can be explained as count-quantified event-situations which a speaker portrays as bounded, discrete, event-situation individuals which will not divide into individuals of the same kind; imperfective forms, on the other hand, are mass-quantified event-situations portrayed by a speaker as unbounded, homogeneous event-"stuff" which divides into parts that are of the same "stuff. Two cognitive processes, packaging and grinding, are assumed to account for these aspectual differences as well as for the mass-count distinction of nominals. Packaging and grinding rest on processes that effect the perceptual and conceptual figure-ground distinction and effect salience, relevance and coherence markings. 3.2 Tense and Its Psychosemantic Basis Let me assume that tense is a deictic category which grammaticalizes relationships that primarily hold between an extralinguistic anchor, the moment or period of speech, which is identical to the deictic now (Reichen- bach's speech time), and the verbally represented event-situation (Reichen- bach's event time). In the canonical, default situation, the moment of
460 WOLF PAPROTTE speech coincides with the experiental now of the speaker and hearer. But note, this now may range from a point in time to a time period of almost arbitrary size: now may denote this minute of writing, or the time taken to finish this paper, or the century we live in etc. It is also possible to impose an explicit metric onto this now by using appropriate temporal adverbials. It is 12:29 p.m., September 19, 1986 and I am writing down this sentence. Generally, however, the notion of time with respect to which now can be explained is not that of a metrically ordered time line. Imagine rather the following comparison to hold between an everyday notion of time and now, the moment of speech or writing. A broad stream of time with varying speeds of flux carries downstream a wooden board which can vary in size and which has fuzzy edges. This board can be likened to speech time and it carries downstream with it the speaker (and the hearer). Let the size of board be determined by the extent necessary for the performance of an act or a series of acts that cohere, for example producing a discourse. Due to the linear characteristics of the articulatory channel — (a speaker cannot produce two simultaneous utterances) — and due to the movement of the board, utterances will by default be successively ordered even if no overt markers of the sequence of utterances are given. They will also be understood as denoting successively ordered acts if, in my simplified example, they denote only what can be seen along the banks of the river at the moment of utterance. A tense marker will suffice to indicate whether or not the speaker talks about the board he is presently travelling on, or about other boards further upstream or downstream. 'The problem posed by time in narration is that of explaining why e.g. the events described ... are understood as occurring in sequence" (Ner- bonne 1986:83). In fact, this problem is easily explained, at least in my simplified scenario, by an Iconicity Principle that describes the default case: Iconicity Principle (IP): Utterances that follow each other in discourse denote successive event-situations, unless otherwise marked. (Cf. Dowty's Temporal Discourse Interpretation Principle, 1986:45.) The functioning of the IP is obvious in (12) (12) (a) Peter married Ann. Ann had a baby, (b) Ann had a baby. Peter married Ann. Let me assume that tense markers simply serve to locate the discourse events relative to a "board" further upstream or downstream to the one
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 461 presently occupied by the participants in the discourse. Ideally, there would be three such tense markers which consistently signal one type of board. 'Present' would mark the discourse events denoting event-situations occurring simultaneously with the board occupied by speaker and hearer (NOW); 'past' will mark discourse events that denote event-situations occurring simultaneously with a board that had previously been occupied (NOT NOW AND GIVEN); and 'future' will mark discourse events that denote event-situations occurring relative to a board that is not yet occupied (NOT NOW AND NOT GIVEN) but which may be assumed to be accessible. Unfortunately the functions of natural language tense markers, especially of present tense are much more complex. Tense markers thus seem to specify what Reichenbach called event time, in other words the time of those event-situations an utterance purports to denote. Reducing them to sentence operators will lead into Ner- bonne's problem of how to establish the correct succession of denoted event-situations. However, this problem is easily solved by the IP. Only the question of how to determine the temporal location of the discourse topic frame (Paprotte & Sinha 1985) is still open. The notion is related to Kamp's and Hinrichs' discourse representation structure (cf. Hinrichs 1981; Partee 1984; henceforth I use the term discourse representation (DR)). I shall assume, that besides their function as sentence operators, tense markers redundantly mark the temporal location of the discourse as a whole while the IP by default takes care of the correct succession of discourse events inscribed in the DR. A change of tense markers will then either indicate a relocation of the DR relative to speech time, or will locally invalidate the IP. That is to say, a change of tense markers will set a temporal location for a discourse event which either serves locally as a reference point from which another discourse event is viewed, or it introduces a reference point from which to view the following — and sometimes even the preceding — discourse as a whole. I believe that Reichenbach's notion of reference time will allow this kind of reinterpretation. In my simplified account, tense markers will situate the discourse representation (DR) relative to the moment of speech. A resetting of the time location of the discourse can be achieved by a change of tense markers. Regardless of the temporal location of the discourse representation, the IP holds. That is, each inscription of an utterance in the discourse topic frame by default marks the denoted event-situation as successive to the one previously inscribed — unless otherwise indicated. (The use of a complex tense marker (present perfect, past-perfect)
462 WOLF PAPROTT£ may be a language specific means of indicating that the Iconicity Principle does not apply locally.) I shall look at a psycho-semantic argument for the above analysis of tense before arguing that by far the most important morphological means of validating or invalidating the IP is aspect. In the search for an appropriate "psychosemantic" basis of tense, linguists have turned to Husserl's phenomenological analysis (Schnelle 1980, 1981a) or have at least looked at its feasibility (Schopf 1984). In a cursory reinterpretation of Schnelle's work, we may consider TIME a high order concept resulting from cognitive manipulations over basic pre-concepts such as 'nowness' and 'becoming' which refer to the perceptual intake and retention of real world stimuli. The use of tense markers can then be based on the commonplace experience that an impression of a present event is superseded by another event which occurs now, comes into being and thus becomes perceptually accessible. 'Nowness' and 'becoming' are the elements basic to temporal deixis insofar as they themselves are organisational features of the experiential system. They are not linked to the spatial metaphor of time, the time line, but concern the intake of an original, momentary NOW, a field of event-stimuli into Very Short Term or Short Term Memory, its retention for a limited time, (possibly in the form of a series of increasingly weaker retentions as suggested by Husserl), its upgrading, and its eventual supersedure. Impressions and the sequence of decaying retentions which have been superseded may be stored away for some longer time, a fact basic to the notion of no-longer-now and the higher order concept of PAST. It is certain that the information that is stored away for some longer duration undergoes a transformation which will be the more substantial the longer the period of storage. A reactivation of stored stimuli ensures the difference between a presently given stimulus or field of stimuli and one that is "no-longer-now" but again representationally given. In a way, it is a condition on no-longer-now that VSTM be completely cleared and its content entered into STM or LTM before reactivation. Different durations of storage could be used as a gradient for the amount of pastness. The relation holding between present and future (not now and not given) is analogous to the relation holding between past and present. Many acts of everyday reasoning, believing and assuming, rest on default assumptions that kinds of event-situations will be in the future what they are now, will be predictably changed or will predictably occur as a consequence of a present state of affairs. In that way, what is to be behaves as what is now, when this now developed from a previous state of affairs.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 463 In distinguishing these three notions of 'now and given' (PRESENT), 'not now and given' (PAST), and 'not now and not given' (FUTURE) the question of which data or stimuli are taken in, retained, stored or expected has little import. Higher level judgements of relevance of data, of perceptual selection, and shaping of stimuli for further processing are related to the psychosemantic basis of aspect. Tense, at all events, is taken to have as its psychosemantic basis the analogue processes of information intake, retention and supersedure in a cognitive and perceptual apparatus rather than the time line and its mathematical explication. 3.3 The Iconicity Principle, Aspect and Tense I have argued that packaging and grinding, or viewed from a slightly different perspective, foregrounding and backgrounding, are at the basis of aspect. I have also proposed that a speaker of SMG can choose a viewpoint aspect (perfective, imperfective, and perfect) almost regardless of the predication type (Aktionsart) of the verb and its satellites. But what makes a speaker choose perfective rather than imperfective aspect, and which consequences will result for the IP from the aspectual packaging or grinding of a discourse event? Secondly, if foregrounding of an event in discourse also affects the information structure of a discourse representation, what consequences of aspectual marking will be felt both in the informational and the temporal structure of a discourse? The IP entails that inscriptions of utterances into a discourse representation preserve the order of the production sequence in the succession of designated event-situations. No problem ought to arise for a sequence of perfectively marked predications in that these, in any case, appear as discrete, bounded, individualized, "packaged" events. Temporal overlap between successively inscribed discourse events will be excluded, - unless a hyponymy relation holds between successive predications that marks a succeeding event as a subevent of its predecessor; and - unless pragmatic considerations and common sense reasoning indicate simultaneity of events; or - unless adverbials explicitly call for a simultaneity interpretation or any other reordering. In a sequence of a perfective and an imperfective event-situation, in the simplified default case, successivity will not hold. When an imperfective predication is asserted in a discourse, it is asserted as an event-mass, as tern-
464 WOLF PAPROTT£ porally grated, ground and unbounded; potentially co-occurring adverbials will neither set a countability restraint nor exclude that the imperfective event situation has lasted prior to its position in the sequence of events and will continue to last past its present position. Imperfectives will therefore both appear as simultaneous to, or overlapping with count-quantified per- fectives as well as in general overlapping with each other. Temporal orientation adverbials, however, can be used to cancel default assumptions and effect a successivity ordering. In pure sequences of imperfective event-situations the IP seems to interact in a more complex way with the aspectual marking unless the sequence of denoted states can be interpreted as a causal chain of states where the underlying principle effecting the changes of states is left implicit. An imperfective state predication followed by an imperfectively marked activity, accomplishment or achievement will also mark simultaneity; an imperfective sequence of activities or achievements / accomplishments may mark locally successive, non-overlapping event-situations, for example in the context of habitual actions performed by one and the same agent (B. Rudzka-Ostyn, personal communication). A sequence of imperfectively marked state predications will indicate simultaneity. But note that even such a succession of imperfective event-situations is marked as overriding the IP vis-a-vis preceding or following perfective event-situations. It seems then that the IP will in general induce a successivity interpretation for sequences of perfectively marked event-situations, and for a subset of imperfective event sequences, but will be overridden by the occurrence of a switch of aspectual markers. The first discourse function of aspect, therefore, can be said to be refining the mechanism of the IP and spelling out successivity or simultaneity of events in the succession of utterances. Although temporal reference and aspectual predication have the sentence/clause as their basic unit, multi-propositional discourse imposes conditions that affect their role in the thematic structure and the textual coherence of a discourse representation. I assume that the concepts of tense and aspect embrace functional notions which relate to the topic comment articulation, i.e. to the provision of familiar anchors for the chain of sequenced events (the function of tense) as well as the foregrounding or backgrounding of events in terms of salience and relevance (the second discourse function of aspect). In the subjective construal and plan of a discourse representation, a perfective aspect marker obtains its value from predicating figure characteristics of an event situation, and from according the event a high
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 465 level of prominence against a background of imperfectives. Perfectives thus signpost "landmark" events for their inscription as salient and relevant, foregrounded discourse events and their arguments (discourse entities) into a DR. As many researchers have noted, perfectively marked events in a narrative will sketch the story or plot line; imperfectives mark supportive, descriptive material that temporally overlaps with the foregrounded discourse events occurring immediately prior to or following the imperfectives, or material that provides general background information, introductory stage setting, locale, etc. To sum up: The effects of packaging or focussing (high identifiability, Gestalt characteristics) will jointly spell out countability and discreteness. Together with the IP, this ensures that perfectively marked discourse events are understood to occur as bounded event individuals, in succession, and to be those parts of the discourse's informational structure which are salient and relevant and ensure textual coherence. Imperfectives as backgrounded discourse events present stage setting or descriptive filler material. These hypotheses will be tested and discussed with respect to a small textual basis. 4.1 Packaging and Grinding with a Discourse Purpose in a Narrative The text (Appendix 1) "To ndaxa tt]c; Yicxyiac;" is a simple narrative about a Greek family awaiting the return of their father, a captain, for the Easter celebration. It is told by the captain's granddaughter who recounts a story she was told by her grandmother and deals with an episode of the grandmother' s youth. The sentences of the text are numbered (1-133); simple and complex sentences with perfective main verbs occur in the left-hand column; simple and complex sentences with imperfective main verbs in the right-hand column; complex sentences with perfective main verb and imperfective subordinate clause, or complex sentences with imperfective main verb and perfective dependent verb are marked with a double + + . Sentences with tense shifts are marked by a double asterisk. Finally, sentences with direct speech will be treated as complex sentences but will not be analysed. A first look shows that the aspectual distinctions are evenly distributed across paragraphs. However, in the main, aspectual distinctions are patterned: there are several blocks of text with only or mainly imperfective verbs (cf. 4-11; 23-33; 42-54; 67-74; 84-89; 93-99); several mainly or purely perfective text passages such as 12-22; 34-41; 55-66; 87-92; 100-133; and mixed passages such as 75-89. Compare now the resumes of some aspectu-
466 WOLF PAPROTTE ally pure passages: (A) Imperfective verbs (4-11): My father was a captain. We had our own boats. Often he was absent from the house for months. At that time, he had been absent for a long time. (B) Perfective verbs (12,13,22,24): Father wrote to mother that he would come for Easter. The days of the Easter week passed. He had not come. Marigo took me with her to light some candles in Agio Nikola. That hour, a terrible storm began. (C) Imperfective verbs (20, 23-33): For days the weather had been bad. (The storm) pushed us and pulled at our skirts. Marigo held on to my hand. From the small path, we saw the sea beat against the shore. Marigo murmured prayers all the way for those at sea. The weather worsened and we could not see ahead. She grasped me more firmly and pulled me along. I cried, afraid of the weather; she said blessings for those who were away. (D) Perfective verbs (39-41; 69; 81; 90-92; 100-133): A boat (had) capsized close to the harbour. The good weather restored mother' s hopes. We went to the cave to watch father arrive. The happy noises of the Easter celebration filled the air. The noise died down. We relapsed into silence. We went back home. Madame, why is the light switched on? We opened the front door. He took us into his strong arms. (A) and (C) present descriptive, stage setting material, narrative sidelines and commentary. The imperfective verbs thus appear to mark subordinate events which are backgrounded to the main story line. Within the narrative, the imperfective aspect has an "epic" function of providing background information "atmospheric" material, descriptive details of scenery and natural phenomena that function as narrative ornament. The imperfective passages of this narrative would, if given by themselves, require extensive inferencing and bridging assumptions but would still not enable the reader to know what happened. All information relevant and necessary to the plot is contained in the perfectively marked sentences of (A) and (D). I therefore accord a "dramatic" function to the perfective verbs in the narrative; perfective aspect marks focused plot line events which move the action forward in the narrated sequence of events (cf. Talmy 1978b; Hopper 1979b). It was proposed that these two functions which support or override the Iconicity Principle are consequences of their mass or count quantification. It is a task of the reader or listener to set up a discourse representation on
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 467 the evidence of what is said or written. By the Iconicity Principle, each incoming sentence will be inscribed into the DR later than the previous one and the denoted event-situations will be taken as occurring successively. However, a perfectively marked discourse event (my "landmarks" on the cognitive map of a discourse representation) with its count-quantification characteristics denotes an event-individual that is not divisible into parts that are of the same kind. Adding another event-individual of the same kind will give a plural set of this kind of events. In other words, if such an individual is asserted to obtain in a particular position of the sequence of events inscribed into the DR, it is asserted only for this particular position. Perfectives will thus enforce the Iconicity Principle by implicating that the preceding and following position are not taken up by the same event individual. Thus compare the sequentiality holding between (12), (13); (16), (18); (19), (21); (81), (82), (83), (87); (90), (92);. However, there is not always a relation of temporal succession between successive perfectives: (21) ... a strong wind began to blow ... (22) He found us up on the mountain; or (100) The bell could be heard; (101) The sound filled the air; (102) It frightened and overwhelmed us; (103) We were made to get up, hearing the commotion in the village. (104) Joy filled the air. I suggest that these examples are cases of temporal anaphora: semantically and pragmatically, relations between (100), (101); (102), (103); and also (21), (22); spell out the same event-situation; it is elucidated by an assertion of an obvious consequence (101); it is expanded by a definite specification (22); or simply repeated by a new lexical label (103). These cases thus seem to work like nominal anaphora: The man ... he .... that drunkard in the corner ... the fooL As to perfectives, cases of this kind seem to be the only exception to the working of the Iconicity Principle. Figuratively speaking, the IP and perfective verbs mark salient events that can only hold one position in the sequence of events told; this position is marked as "one landmark further down the road" with respect to the previous one, the currently-last event on the 'plot line'. "One further down the road" corresponds closely to a temporal relation of strict precedence holding between each successive discourse event as hypothesized in Partee (1984). The order of discourse events beginning with the landmark (12) He wrote to my mother finally culminates in (129) He took us into his strong arms. Perfective verbs here locate each main event relative to the preceding and succeeding one. Imperfectively marked verbs do not necessarily locate
468 WOLF PAPROTrf- discourse events but mostly specify temporal simultaneity of states and processes (or activities) that hold while main events happen, or that overlap with the whole narrative (cf. (4), (7)). Thus, sentence (20) specifies a weather condition which is only redefined in (52) and which overlaps with or remains simultaneous to a series of events of the plot line. Figuratively, the imperfective aspect marks "districts" or "regions", even plots detours around landmarks. Again, the reasons for the "epic" discourse function of imperfectives follow from their mass-quantification characteristics: they mark referentially distributed event-material, any proper part of which is of the same material and any fusion of its parts will yield the same kind of event-material. Any imperfective that occurs in a sequence of discourse inscriptions can be said to occupy a temporal position that extends beyond the preceding and the following event. This follows from the fact that every imperfective can be thought of as being a proper part of a mass-quantified event which, without changing its material set-up extends by implication further than the currently asserted event-part. Imperfectives thus endure beyond the contextually fixed position in the sequence of events; they overlap and mark temporal simultaneity (cf. Dowty 1986:49). If we look at some aspectual switches across sentences, the above ideas are in the main corroborated: (111) Bv$ioTf\xa\x£ (perf.) jidXi otr] oloojit), oto oxoxabi. "We relapsed again into silence in the darkness." (112) HOLtdCafie (impf.) rn ftakaooa. "We looked at the sea." (116) Yla\ie (impf.), euie (perf.) r] uriteoa. "Let's go, Mother said." (117) Me Jirjoe (perf.) djio to %eqi. "She took me by the hand." (118) H Magi/yob ecpeyye (impf.) \ie to cpavdoi. "Marigo shone the flashlight." (128) Meoa arnv toajie^aoia fuxgjieoL^eve (impf.) o Jiateoac;. "In our dining room, father was waiting." (129) Mag dQJia!;e (perf.) ota bvvaxa xov \xjzQ&xoa xca \xaq cpiXouae (impf.). "He held us in his strong arms and kissed us." Also here, the perfectives mark focused discourse events as single, individual, bounded landmarks which occur in succession (cf. 116, 117). The
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 469 imperfectives denote for the most part activities of temporal extent which are simultaneous to the preceding (or the following) event, even to series of events: Marigo's shining the flashlight temporally extends over the narrated time that starts with (116) and lasts for all of the walk from the cave to the house; note that three events are mentioned which occur within this span of narrated time: the child says something; the child starts to cry; the mother keeps her silence. These are perfectively marked and occur against the background of Marigo shining the flash light while the three persons walk along the mountain path. The imperfective here guides the reader's imaginative reconstruction of a walk at night, at one point of which the mother does not answer her daughter. (128) and (129) do not show temporal overlap between successive sentences. (128) indicates a state that had been enduring for some time; (128) is a case of backwardly established simultaneity. Again, the full import of aspectual usage becomes accessible only in a consideration of a whole series of narrative events. 4.2 SMG Aspect in Complex Sentences There are two competing explanations of intrasentential aspect switching (1) temporal coincidence (Seiler 1952; Bakker 1970); (2) scope and scenario (Newton 1979). Seiler (1952) and Bakker (1970) assume that the choice of imperfective aspect in subordinate clauses indicates temporal coincidence of the verbal actions denoted by matrix and subordinate verb. Perfective aspect in the subordinate verb serves to indicate lack of temporal coincidence. Sentences in which va-subordinates have no temporal relation to the matrix verb present some problems to this otherwise interesting explanation. Clear cases in point are matrix modals with either perfective or imperfective subordinates, epistemic modality matrix verbs like believe, want, or main verbs like stop, begin which may have imperfective subordinates but exclude temporal coincidence on semantic grounds. The concept of temporal coincidence may, therefore, better be seen as limited to the discourse functions of aspect in SMG. Newton (1979) assumes that the perfective-imperfective distinction reflects differences of the logical forms of sentences. His basic notion is that of time as 'occasion' (t), as a discrete temporal 'entity' devoid of the sortal characteristics which Mourelatos attributes to his events. 'Occasion' is sharply distinguished from the tense logic moments of time along a linearly ordered infinite axis governed by the axioms of denseness and continuity. It
470 WOLF PAPROTTE amounts to natural language equivalents such as SMG cpood; French fois\ German Mai; as in ein Mai, zwei Mai; etc. It is a variable ranging over the 'times' situations occur. A universally quantified list of occasions which specify antecedent conditions for the occurrence or realization of event situations denoted by a verbal predication (consequent) is called a scenario expression with the following logical form: (13) V t (p at t d q at t) ('It is the case that, at all times, if p is true then q is also true'). The basic rule given by Newton for scenario structures that are overtly11 mapped onto surface structures in SMG, is that both antecedent and consequent are assigned imperfective aspect (14): (14) Kofre cpooa kov \xz pXejtei/epXejie (impf.) \xz Pql^el/^i'ePql^e (impf.). "Everytime he sees/saw me, he insults/insulted me." Scenario structures successfully explain a large class of expressions whose matrix (often modal) verb can be aspectually marked independently of the complement.12 That is, modalized sentences containing rate expressions are seen aspectually marked according to scope relations within their equivalent logical forms: universally quantified propositional and modal terms will surface as imperfective verbal forms; whenever a modal operator is outside the scenario structure, a perfective will surface. Modal operator expressions, 'Modal p at t' (M*p) yield three kinds of scenario expressions (15): predicted aspects of p, q: (15) (a) MVt(ptDqt) impf. p; impf. q (b) V t (M p d q) Perf- Pi impf. q (c) VT(ptDMtq) impf. p; perf. q As most elegant explanations, this one is beset with some problems, too. The main one, as Newton also points out, is that many such sentences are ambiguous between a wide and a narrow scope modal interpretation. The latter requires a perfective subordinate verb. The factors which impose a wide or narrow scope reading depend on the semantic characteristics of the modal: e.g. njtood): deontic reading: "have standing permission to"; and physical ability reading "have permanent ability to" denote durable, temporally undistributed states of wide scope modality; in its weak epis- temic reading "it may be the case that" it refers to a transient, temporally distributed state, accordingly takes a narrow scope modality interpretation,
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 471 and demands perfective subordinates. The semantic reading of the modal is, however, affected by the speaker's representation of real world event characteristics of the complement and by contextually imposed pragmatic factors; the pragmatics of event situations will here determine aspectual restrictions. Another set of problems arises in cases of "antecedent incorporation" (cf. Newton 1979: 155 ff.). Typical cases involve surface NP objects which are seen to originate in LF antecedents but are mapped onto surface consequents. The suggested logical forms and their antecedents are given predicates that either originate in frame- or script-like knowledge structures denoted by the nouns or have to be recovered from representations of the linguistic and non-linguistic context. In either case, encyclopaedic and pragmatic knowledge is used to resolve aspect assignment. But this clearly runs counter to Newton's explanatory scheme which intends to have aspect determined at the level of logical form. Furthermore, circularity arises whenever recourse to a surface reading of a sentence is necessary to determine which LF is to be chosen as a correct description. In these cases, LF will not determine aspectual choice but will itself be determined by considerations of the import of the aspects chosen. We may, therefore, take Newton's results as post hoc explications which account for no more than can be accounted for by the heuristics of adverbial co-occurrence restrictions. However, as we are equally justified in claiming that the semantics of aspectually marked predications imposes selectional restrictions on the kind of temporal adverbials that can be used, the problem of aspectual choice in complex sentences is still unsolved. To sum up: I have suggested that aspects in SMG have the discourse function of packaging or grinding. These functions can be viewed in terms of an analogy between a discourse representation and a cognitive map. Aspect marking serves the function of guiding textual exploration by setting up foregrounded discourse events against a background of discourse states and processes. Vis-a-vis preceding or following perfectives, the background has been conceptually ground; it is mass-quantified and non-countable. Foregrounded events are packaged and countable. They serve as textual landmarks. The sequence of foregrounded discourse events can normally not be changed without changing the represented sequence of the denoted course of events. As a result of the linearity characteristics of speech, of the Iconicity Principle and of prototypical characteristics of perfectives, foregrounded events are represented one after the other; as a consequence, a
472 WOLF PAPROTTE text will spell out a path from each "landmark" to the next. Imperfective aspect locally invalidates the Iconicity Principle due to its prototypical par- titivity characteristics; it marks regions surrounding landmarks or charts routes into the district surrounding a landmark. Backgrounded clauses are thus like detours taken for exploratory reasons. They allow for temporal displacement, are non-sequential and simultaneous to main foregrounded discourse events and thus have an "epic" function. Sequentiality constraints follow from the characteristics of text production and reception but not from perfective aspect. For that reason it is not necessary to define aspects relative to the temporal characteristics of predication types. At best, aspect supports or overrides the Iconicity Principle which by default parallels the sequence of denoted events with the sequence of their production and inscription into a discourse representation. The salience of perfectively marked discourse events as individuated event figures contributes to their relevance to the plot and to coherence relations between the foregrounded events. Relevance to the plot is likened to "importance for plot development" (Fleischman 1985), whereas coherence relations may be linked to the traditional idea of "unity of plot". If these two notions are said to make up the salience of perfectively marked discourse events, an element of newness and unpredictability may be said to adhere to the characteristics of foregrounded events (Fleischman 1985). 4.3 Tense and Discourse Representations "Discourse representation" must be a dynamic notion. Due to the constant upgrading of information in a DR, a cohesive thematic structure of anaphorically chained discourse referents and the coherence of informational structure given with the sequence of discourse events must be achieved and maintained with each new utterance. Theoretically tense may have two functions in the maintenance of informational coherence: (1) the deictic function of locating a whole DR relative to the moment of speech; (2) the function of temporally ordering and relating discourse referents and events within the temporally situated DR. This latter function has been discussed as temporal anaphora by Partee (1984); Dowty (1986) tries to account for it in terms of an interaction of a discourse principle and the predication type of a sentence. Ad (1): With respect to the SMG text (Appendix 1), temporal deixis is
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 473 the result of a complex process of relating the DR "Grandmother' s Easter" to the moment of reading, i.e. of embedding it into increasingly deeper layers of pastness, cf. (16); ( < 'precedes' ; ^ 'included in') (16) (a) I am now sitting here reading a story (rt); (b) The writer wrote the story prior to my reading it = rt ; where rt2 < rtQ (c) The story was told to the writer when she was a child = rt2; where rt2 < rtj < rtQ; (d) The events happened to the (writer's) grandmother when she was a child = rt ; where rt3 < rt2 < rt] < rtQ; (e) The events happened one Easter week in the grandmother' s childhood = rt- 4' where rt, ^ rt, < rt, < rt, < rtn 4 3 2 10 Common sense reasoning establishes the temporal location of each current DR with respect to another DR with a temporal index of being prior to or of temporally including the current one. A chain of such embed- dings eventually situates the whole discourse with respect to rt . Lexical and grammatical markers, tensed verbs, adverbials, and nouns like childhood, Easterweek in combination with encyclopedic knowledge provide notions of relative extension and overlap of rt and rt and provide a temporal frame for the series of narrated discourse events that begins with (12) "He wrote to my mother." The deictic function of tense thus serves to locate a DR with respect to the moment of speech or understanding of utterances. It appears that the notion of time underlying this reasoning is not one of time points that are linearly ordered. Rather, time is taken to be some kind of mass out of which portions can be measured such as a lifetime, a childhood, an Easterweek. These simply spell out a mass inclusion relation (cf. Bach 1986:7). Tense markers locate these portions with respect to the reader's perceptual actuality of 'now'. I shall assume that a succession of tokens of one sentential tense operator redundantly marks the temporal location of the discourse representation relative to rtQ. Ad (2): If a hearer were only interested in establishing the temporal relation obtaining between utterance tokens, the order of mention and the order of their inscription would automatically spell out precedence/succes- sivity relations. However, his focus certainly is on establishing the temporal
474 WOLF PAPROTTE ordering among the denoted events. The Iconicity Principle provides only one mechanism for establishing temporal order by default: whereas a speaker's utterances have to occur in succession, denoted event situations can also occur simultaneously, or in partial overlap; furthermore, an actual course of events may verbally be represented in a different order (cf. a hysteron proteron: "Ihr Mann ist tot und laBt Sie griiBen..). If aspect, as I assume, asserts countability and thereby implies temporal succession, partial overlap or simultaneity, what is the function of tense markers with respect to the discourse internal order of events? A look at tense switches will provide a first answer. 1. A switching of past into present tense as in (3) Nat, fiQfre (perf.past) akXa axou va Seic; (perf. pres.) Jtcbc; fiQfre (perf. past) "Yes, he came but listen to see how he came (back)" forces a change in the temporal index of the current discourse represention from rt4 to rt The reader is moved into a period of time later than the course of events during the Easter week, i.e. to the time when the grandmother narrates to her grandchild. In (8) and (9), the imperfective present fru^ia^ai and the perfective future fta ^exocaca take the rt4 index forward to rt3; from there the future tense points to an unspecified period between rt and rt Thus, tense markers not only set a DR's temporal index, they also effect a resetting of this index. 2. A DR internal change in the sequence of discourse events or a temporal relocation of one event (or series of events) relative to another is effected by past-pastperfect switches, cf. sentences (35), (39); (44), (55); (69). However, a switch from past to pastperfect does not always serve this function: cf. (34) and (35): 'Oxav cpxaaa^ie (perf. past) arr|V exxXriaia, Sev P^ejta^e (impf.past) Jtia xrj ftaXaaaa. (35) Mia |3aoia xaxaxvia xrjv et/B axeJt&aei (perfect past) "When we stopped at the church we did no longer see the sea." (35) "A heavy fog had covered it." Here, aspect competes with tense; the pastperfect indicates a state that lasted and still lasts as a cause for the imperfective. Thus, causal process, and simultaneity of states are signalled by this particular tense switch. 3. Frequent present to past switches in direct speech do not indicate temporal relocation and updating. They simply mark the difference between quoted parts of direct speech and the rest of the narrative. 4. Frequent tense switches occur in complex clauses, mainly va-com- plements, which contain a present tense while the matrix verb is past. Thus (1) xofrtaa^e (perf.past) va xov jteoi^Evou^ie (impf.pres.); "we sat down to wait for him;" (7) Exava^e (impf.past) ^if|vec; va xov Soii^e (perf.pres.) "we
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 475 spent months to see him again;" cf. also (11); (36) xrjv axovya\ie (impf. past) [xovo va uxyuyKQi^ei (impf. pres.) ajto xaxoa "We only heard it roar from below." (39) Bi%av Set (perfect past) eva xaixi va Jieovaei. (impf. pres.). "They had seen a boat perish."; cf. also (37); (38); (72), (86), (91), (94), (105), (119) ..dQXiaa (perf. past) va xtaxta) (impf. pres.) "I began to cry." All these cases involve "scenario expressions" in which the relation of the two verbs in terms of aspectual marking is of foremost interest. Aspect not only neutralizes the semantic influence of the tense marker; in the above constructions va + imperfective past would signal a modal past of an action that has not taken and is not taking place (Mackeridge 1985:277). The present tense forms thus do not refer temporally but serve to exclude a modal interpretation and call for pragmatic considerations to disambiguate temporal location. The English infinitival complements as translation equivalents behave similarly. 5. An aspectual contrast is implied by the tense switch between ei%e / r|iav and eivai in (48) and (49); note that exo) and ei|Kxi have no morphological aspect marking so that context and tense switching will here take care of necessary aspectual implications. 6. Tense switches from past to future also relocate the position of a discourse event in the series of events cf. (62). Whereas past to pastperfect switches put the pastperfect event into the position prior to the position of the past tense one, events marked for future generally allow for other events to have occurred in positions following the past discourse event. In both cases, only events that are mentioned will occur in between pastperfect and past or past and future. If nothing is mentioned, the events may be viewed as having occurred in succession. One consequence of these sparse remarks is the finding that present tense in SMG only rarely deictically refers to the moment of speech (rt ). While pointing to a time ("context of evaluation") at which discourse events happen, the index of a DR may be set by present tense markers as anterior to rtQ, as rt() and as posterior to rt . Similarly, discourse internal tense shifting by present tense markers may effect complex relocations of discourse events in the course of the denoted events. Pragmatic considerations and common sense reasoning will in all cases be necessary to help dissolve temporal ambiguities and decide whether the DR has been given a new index or whether local reorderings of discourse events have been achieved. To sum up: It has been argued that the grammatical category of tense involves the indexical, shifting relation of the representation of narrative
476 WOLF PAPROTTE events to the temporal and spatial coordinates of the speech event (rt0). Tense will indicate anteriority, simultaneity or posteriority of the narrated events relative to rt . I assume that each discourse representation is temporally indexed; consistent tense marking yields the default assumption that the temporal index of DR remains constant. For many reasons, single discourse events or parts of a DR need to be temporally relocated; this is achieved by tense shifting. While the succession of the same tense markers redundantly marks the temporal location of DR as unchanged, tense shifting may indicate a reindexing of a DR (e.g. in the form of embedding it in another DR) or provide new "reference points" (in the sense of Reichen- bach) by signalling that a denoted event must be reordered against the sequence of mention. In particular, the SMG present tense is a multifunctional marker which is often neutralized by an overriding aspect. 4.4 Aspectual Distinctions of English: Simple Tense versus Progressive Aspect In a serious cross linguistic sense, aspect belongs to the set of universal grammatical categories. This does not mean to say that an aspectual Gesamtbedeutung or core meaning could be posited to hold for English as well as for SMG. Not even the discourse function of English progressives need be the same as that of SMG imperfectives. Also, the relation of English aspectual distinctions to the indexical function of tense might vary from that of SMG. These are the points which will be discussed in the rest of the paper. The morphological difference between English and SMG is quickly summarized. (1) English progressive aspect is morphologically signalled by a combination of the -ing participle, and tense and voice marked forms of the auxiliary be; SMG marks the perfective verb stem. (2) SMG verbal themes which are marked for aspect correspond to English verbs being marked for tense distinctions. This indicates different positions of the grammatical categories on the scale. Note, for example, that whereas SMG mood (subjunctive and imperative) is marked for aspect, English imperatives are not so marked. Secondly, whereas SMG mood is not marked for tense distinctions, tense distinguishes voice, mood, and aspect in English. SMG aspect, therefore, has a higher rank on the scale than SMG tense; in English tense is the category with widest distribution (cf. Mirambel 1942), that is finite verb forms in English are generally marked for tense. SMG finite verb forms are marked for aspect. It seems natural then that predica-
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 477 tion types play a prominent fole in the simple tenses of English. If only progressive forms are overtly marked for aspect, the so-called simple tenses must be considered as either not marked for aspect at all, as implying an aspectual marking by zero morphemes, or else, as carrying the aspectual distinctions implied by the predication types of the verbs. It is this latter position that seems to fit the facts better. English provides elaborate examples of how predication type characteristics of a verb are determined by spatial, temporal or other quantitative features of adverbial phrases, by transitivity, by definiteness, specifications of subjects and direct objects, and other factors (cf. Friedrich 1974; Schopf 1984; Dowty 1986). Predication type transpositions are achieved by adjuncts in the VP and by the arguments of the verb. For example, activity verbs of movement like run, walk will be converted into accomplishments by a PP or NP specifying quantity or extent: (17) John ran. (activity) John ran a mile, (accomplishment) John walked to the station, (accomplishment) John ran for some time, (activity) He played a Bach concerto, (accomplishment) He played Bach concertos, (activity) He played a Bach concerto for hours, (activity) The "accusativus effectivus" in He wrote a poem will also change an activity write into an accomplishment (Friedrich 1974: 5). Similarly, the collaboration of Aktionsart characteristics with an indirect object and with number and definiteness marking will result in a change of a verb's predication type.13 In the absence of a morphologically marked aspect, the predication type of a sentence will be the only aspectual meaning component. There are roughly two consequences for the discourse internal sequencing of events brought about by the simple past tense in English when the tense marker indexes the DR as past: In the case of state and activity predications, the Iconicity Principle will be overriden by the predication-type characteristics of states and activities. That is, mass-quantified state and activity predications of simple tense sentences will denote discourse events which may overlap with the preceding, the following or with both discourse events. This point is born out by Dowty's findings. He notes that denoted states are normally assumed to extend before and after their asserted position in the sequence of discourse events except when
478 WOLF PAPROTTE pragmatic reasons indicate the contrary (1986:51). Just as with states, the overlap of activities with surrounding discourse events is determined by pragmatic considerations of the length of a specific kind of activity. Only accomplishment and achievement predications in the simple tenses of English share characteristics of the perfective aspect. They have "undistributed reference", they are indivisible in that no proper subpart of either an achievement or an accomplishment is of the same kind. It is therefore only achievements and accomplishments which will naturally support the Iconicity Principle and will denote a sequence of events that corresponds to the order of mention. There will of course be exceptions as in the case of SMG perfectives: repeated mention of the same event in successive utterances as well as anaphoric backchaining via verbal synonym and mention of "subevents" of a superordinate verb will not push forward the sequence of discourse events. In view of these findings, Smith's hypothesis that the simple tenses of English have a prototypical central meaning close to the traditional notion of perfectivity ("simple aspect presents a situation in its entirety" 1986:103) and that this meaning is interpreted differently for each predication type, does not seem convincing. It rather appears to be the case that the simple tenses of English are unmarked for view-point aspect and only show the Aktionsart distinctions inherent in the predications. I claim, therefore, that there is no equivalent of SMG perfective aspect in English. At the best, a family resemblance between the perfective aspect and English achievements and accomplishments as predication types can be found. Although it does not make for a nice binary distinction, English appears to have only one viewpoint aspect, the progressive. A quick look at the simple present may corroborate this claim. The function of the simple present as a tense has been questioned to the point of considering it as the mood of "diminished reality" (Bailey 1984). Generic uses of the simple present, its use as historic present and in place of future forms, point to the fact that the simple present is not a straightforward tense in the same way as the past is. However, as long as there is an appreciably strong connection between the use of a present tense form and the "now" of the moment of speech, I suggest following its traditional categorization as a tense; the strong and complex interrelations of present tense, mood, and Aktionsart are not to be denied. Let me note that in cases of a "true present" Cjust now at the moment of speech'), state predications tend to occur in the simple present whereas activities, accomplishments and achievements will preferably take the progressive aspect to denote absolute
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 479 nowness of the event. In fact, putting state verbs into the progressive present will in most cases effect a state-activity transposition plus an altered reading of the verb, thus cf. (18),(19): (18) (a) John loves Mary. (b) I like this job. (c) We know the answer. (d) We care for your patient. (e) Amy resembles her uncle. (f) That cake looks done. (19) (a) John is loving Mary. (b) I'm liking this job. (c) We are knowing the answer. (d) We are caring for your patient. (e) Amy is resembling her uncle. (f) That cake is looking done. Smith (1986), some of whose examples I used, points out that even in the non-standard uses that are fairly frequent in spoken language, the progressive suggests an activity, a temporariness, and, I would add, an actuality component that is characteristic of events rather than states (cf. Bach (1981) who remarks that (unconventional) progressives of state verbs denote "temporally limited manifestations (or realizations)" p. 78). The present tense stative predications seem to work well because of their distributed temporal reference: Their mass-quantification characteristics coincide with the loose connection of the present tense with the moment of speech. A present tense form which may denote past, present and future events thus parallels the "distributed reference" characteristics of a state predication which will overlap with prior and succeeding discourse events. If this hypothesis can be supported by further facts, it, too, would call into question the position that simple tenses of English are marked for perfectiv- ity. It would rather support the view that only the predication type characteristics play a role in the simple tenses. As the present tense is only loosely connected to rt , both the implied interpretation of 'routine' of such utterances as (20) (20) (a) The bus leaves at twenty past the hour. (b) He goes to work at 8:30. (c) I buy my shirts at Harrod's.
480 WOLF PAPROTTfi and the gradual shading into the generic present in cases of stative and dispositional predications as in (21) (21) (a) This plastic preserves its flexibility. (b) This factory makes watches. (c) Sea water tastes salty. (d) Water boils at 100° at sea level. will speak against the view that the simple tenses are inherently marked for perfectivity and show a strong interaction between this prototypical idealized aspectual model and the verbal Aktionsarten, as is assumed by Smith (1986). Note finally that the so-called historic present shows a preference for achievement and accomplishment predications while activities are preferably presented in the past progressive (for an intended actuality component). Thus (22) (22) A strong wind was blowing. The cable snaps. John doesn't even blink and he says to me... In this example, the temporal index of the discourse representation is kept unchanged due to pragmatic and contextual considerations. In fact the stativity marking of was blowing implies that the following discourse events overlap with it. However, the past transposition frees an accomplishment or achievement of the temporal/modal implications of present: "of unlimited or timeless validity and relevance to rt " and allows the predication types to figure prominently as aspectual markers. Smith (1986) points out that achievements and accomplishments will not allow conjoining their sentences with phrases such as and still does so which will work for states and cancel the conversational implication of completion. Thus (23a) is definitely better than (23b,c) if a habitual reading is excluded: (23) (a) John loved Mary and for all I know still does so. (b) *John walked to the station and still does so. (c) *John walked a mile and may still do so. The conjunctions of (23b,c) seem strange because the added phrase cancels a packaging characteristic of accomplishments and achievements: non- additivity. The fusion of two packaged, individuated accomplishments will yield two exemplars not one; the added phrase suggests an activity or a portion of a process. The unacceptability is thus a result of fusing an individuated event with an event - stuff. But why should the conjunction not work in the case of activities, as Smith suggests (p. 103)?
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 481 (24) (a) *He played a Bach concerto for hours and still does so. (b) *He played Bach concertos for hours and still does so. (c) * John ran for some time and still does so. (d) When I saw her, Lily strolled along the beach and probably, she still does so. In the above examples, the unacceptability effect is due to the adverbials which denote limited duration whereas the conjoined phrase implies continuation of the activity. In the absence of such adverbials, activities can be conjoined with and still does so. Smith's conjunction test thus provides no criterion distinguishing states and activities. It will not explain why the progressive does not comfortably occur with state predications but allows activities. We may also note the puzzling treatment of the progressive as a stative operator (Vlach 1981; Dowty 1986). Both states and activities are mass - quantified. Why should the progressive attribute to an activity what is already inherent in its predication type? The difference between states and activities must therefore be sought somewhere else. Activities are generally processes predicated of animate agents who have to spend energy to keep the process going. A concept of 'animate agent' will contain the notion that only a limited amount of energy is available to each agent and that in many cases he can intentionally allocate the available energy. It may then well be that the difference between states and activities rests on the fact that activities imply processes that start and can continue homogeneously only up to the point when the agents have spent their energy and the activity has to end; it can obviously be intentionally terminated. The notion of state does not carry a similar notion of termination. On this assumption, activities can be regarded as virtual event figures that can be ground. States, on the other hand, display various degrees of inertia, that is they continue to endure unless something or someone brings them to an end in order to effect a change of state. Thus if The pyramid is on the small yellow block, occurs in a discourse, the denoted state must be assumed to obtain unless it is mentioned that the pyramid was removed from the block. An activity can be presupposed to come to an end even without an expressly mentioned interference by an agent. The use of the progressive with activities may thus be said to override the presupposition of termination by asserting local "stativity", that is by asserting continuation of the activity relative to adjacent discourse events and availability of energy (for a similar position cf. Bach 1981).
482 WOLF PAPROTTE 4.5 Progressives and Imperfectives It seems clear that the simple tenses of English cannot be equated with SMG perfective aspect. Meaning components that correspond to the perfective aspect result from an interaction between the morphologically unmarked simple tense and predication type characteristics of accomplishments and achievements. As I showed with respect to the present tense, there are frequent and normal uses of these predication types which cannot be subsumed under perfectivity. In this they resemble the so-called gnomic aorist (cf. Friedrich 1974). In English, state predications do not comfortably occur with the progressive. The relative frequency of "non-standard aspectual choices" (Smith 1986:98) such as (25) (a) I am understanding more about aspect each day. (b) Amy is really resembling her uncle. points to the fact that there is a difference of meaning between the aspectual meaning of a state predication-type and the progressive aspect which, however small it may be, a speaker may choose to signal even at the expense of violating a grammatical restriction. Being non-standard choices, they also signal a basic equivalence of statives and progressives (Dowty 1977; 1986; Smith 1986; Mourelatos 1981; Vlach 1981). This equivalence rests on the fact that both state predications and progressives pass the tests for mass-quantification: If someone is V-ing, then it is true that any part of this V-ing is also V-ing; and adding a portion of V-ing to another portion of V-ing is also V-ing. Similarly, if a state S is said to endure, then any moment of S's enduring is also S. The part-whole relation of mass-quantified states and progressives is thus different from the part-whole relation that is typical of count-quantified accomplishments and achievements.14 In this respect the progressive aspect will turn sentences into stative sentences. It is on a par with the SMG imperfective but also on a par with simple tense state predications. Note the parallels in the interaction with point adver- bials denoting the moment of speech: states and progressives seem to denote an extended "temporal frame encompassing something else" (Jes- persen 1961:180). (26) Max was here when I arrived. (26) will be true if Max had been there for some time before the arrival of the speaker and stayed there at least up to the arrival of the speaker. Similarly (27):
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 483 (27) Max was writing when I entered. will be true if Max was writing prior to the moment of utterance and continues to write even after the moment of utterance (cf. Vlach 1981:273; Jes- persen 1961:180). I will, therefore, assume this to be the one meaning component of the progressive aspect. It denotes a discourse event which temporally extends around the moment of speech and therefore overlaps with both preceding and following events in the sequence of denoted events. However, the non-standard uses also point to a difference between a state predication and a progressive form. It seems that whereas a state predication will only allow the pragmatic inference of temporal extent and overlap, the progressive is a means of explicitly asserting it. Historical precursors of the progressive be on/at V-ing (where V-ing was a gerundive NP) already had a metaphorical reading for the preposition that can be rendered as 'be in the process of V-ing' (cf. Jespersen 1961:178). This would mean, that anyone engaged in such a process of V-ing is neither at an initial point of beginning nor at an endpoint of terminating the process but somewhere in the middle of this process. I believe that the "actuality" component attributed to the progressive is due to this interpretation: the progressive denotes a process that is witnessed as actually obtaining and continuing. As Vlach (1981) points out, a progressive achievement denotes the final portion of the activity that completes the achievement (1981:289). A progressive activity or accomplishment on the other hand, denotes an interval before which the activity or accomplishment already endured and which has not yet finished to last. No progressive denotes the initial point or interval of an event-situation whereas states do allow an inceptive interpretation. (Dowty 1986:56) makes this difference the defining one, and it fits well the above picture. In view of these short remarks on the progressive it may be assumed that the SMG imperfective corresponds to the progressive and to the subgroup of simple tense state predications. Both will allow the denoted events to overlap with surrounding discourse events so that the Iconicity Principle will be invalidated by progressives (and state predications) in roughly the same way as by the SMG imperfective. The difference between SMG imperfective and the progressives lies in the progressives actuality interpretation which also distinguishes progressives from simple tensed state predications. A look at an English text and its Greek translation (cf. Appendix 2) will serve to refine the present picture.
484 WOLF PAPROTTE 4.6 Packaging and Grinding in an English Text and Its SMG Translation In this short survey, it has been shown that the simple tenses of English while serving the deictic function of tense are aspectually marked by their predication types. Only the progressive functions as a viewpoint aspect in the sense of SMG imperfectives: it can be said to grind count-quantified accomplishments and achievements. In its application to (already stative) activities we find its special semantic import in denoting some intermediate intervals of an ongoing mass-quantified process and in asserting actual continuation. This discussion thus assumes that the semantic properties of a) verbs, adverbials, and definite or indefinite nouns (the predication type) and b) the viewpoint aspect progressive ultimately determine the aspectual properties of English sentences. Sentences are made to function in a discourse and speakers have the possibility of choosing between alternative verbal presentations of event-situations. Non-standard uses are just one example, progressive "grinding" of events a second, English verbo-nominal phrases a third, e.g.(28).15 (28) He laughed. He was laughing. He gave a laugh. He drinks. He is drinking. He took a drink. She flies. She is flying. She is in flight. These and further linguistic means enable the speaker to encode idealized, prototypical models of aspect which function in structuring a sequence of discourse events in terms of their temporal succession, overlap, simultaneity and relocation in the series. In this respect SMG perfective and imperfective aspects serve the same function as the variety of linguistic means which signal in English aspectual distinctions. As an English text that was also available in a SMG translation, I chose Milne's Winnie-the-Pooh (Appendix 2). The quoted passages are given in the original version and their translation equivalents. Two chapter introducing paragraphs provide good examples of the general discourse function of the progressive (cf. Appendix 2, (1), (2)). It sets the stage,' and describes the background for the highlighted discourse events. The plot is never forwarded by a progressive sentence. The foregrounded, plot advancing discourse events are simple tensed accomplishment predications. The Greek equivalents of these progressives are imperfectives. Where subordinate perfectives seem to disturb this equivalence, they are clearly bound to an iterative reading by the imperfective main verb; the subordinate perfective seems to be due to the real-world, punc-
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 485 tual achievement-characteristics of the event-situations. Progressives and imperfectives present in the linear order of mention a number of discourse states which hold simultaneously, or else overlap for some part of the time (cf.(l) "the streams were tinkling, ...the cuckoo was trying over his voice .."). The Iconicity Principle clearly does not hold in these cases. In paragraph-internal use, the progressives function as local backgrounding devices. That is, the states or processes they denote are made to appear as overlapping with surrounding discourse events. Temporal clauses (When ..; as ...: while ...) also demonstrate these local cases of temporal overlap; (cf. Appendix 2, (3) - (10)). The SMG equivalents of these examples show a more complex patterning. He was getting rather tired by this time (4) is translated by pastperfect + va-complement with an imperfective verb; the temporal reference point is thus relocated prior to the current discourse event while the imperfective signals continuity of persisting state. In (5), the deception which we are practising is rendered by an imperfective main verb, (Jorifrouoec; plus va - complement with a perfective deceive. So the single act, synoptic sense of the perfective acquires an iterative reading. In (6), (7), (9), (10), however, there is the straightforward predicted equivalence between imperfective and progressive, and perfective and simple past. (8) again indicates the enormous range of alternatives a speaker has for marking aspectual distinctions: / have just been thinking is rendered as to^a |a,6Xtg oxeqprr|xa. The adverbials xcoga "just now"; \i6Xic, "hardly" together, signal something like "this present single act is hardly complete". Thus even a perfective in combination with an adverbial can be used to approach the actuality, "in-the-process-of" quality of the progressive. But things are by no means as clear-cut as the above examples make them seem. For one, English simple tenses serve both the "dramatic" foregrounding and the "epic" stage-setting function. In (11), the plot advancing series of discourse events is given by simple tense forms, He put in, he corked, he leant out and he threw. In (12) on the other hand, simple tense forms are also used to convey background information; that is only the simple tensed state predications are so used. The SMG translation (11) reflects this difference but also marks other points. Three imperfective tokens of [jjiOQOiJoe denote temporally unbounded (physical) ability; the English modal can ambiguously also allows for this interpretation. The imperfective aspect in jictQaxoXoufroiJoe ajio^axQTJVOxav, v6|j,i£e indicates the temporal distribution of the corresponding activities, whereas xaxd^aPe (perf.past)
486 WOLF PAPROTT£ precisely conveys the single act implications of a sudden insight. Similarly in (13): the activity walk up and down is translated by an imperfective which is also syntactically marked as the temporal background to the English accomplishment sing the song. The discourse function of foregrounded event and predication characteristics of count-quantification here converge in a perfective translation form. As a rule then, simple tense forms that are count-quantified in English due to their predication types, and that denote successive events in the sequence of discourse events, have perfective SMG equivalents. The per- fectives as well as the corresponding English forms thus seem to signal a temporally interpretable conjunction 'and then'. (14) and (15) support this finding. (15) also gives a nice example of co-occurrence restrictions of an indefinite frequency adverbial (xtide Tqitti) with the imperfective. 5. Conclusion In this paper I have reduced the complex interrelations of tense, aspect, voice and mood in the notional space of the SMG and English verbal system to an outline of the interdependence of tense and aspect in their discourse functions. The explanations given will obviously need to be revised. In particular, the strong links between tense and mood and aspect and transitivity call for further investigation. Tense and aspect are traditionally treated as representing two different points of departure in our experience of time. In my argument for a psychosemantic basis of tense and aspect, two points have emerged: 1) In grammatical explanations of tenses the notion of the time-line and of time as an ordered series of time points or intervals ought to be supplanted by the phenomenological notion of time as flux of perceptual intake in the here-and-now (present), and as stored and reactivated intake that is now-and-given-but-not-here (past). The main discourse function of tense is that of locating a discourse representation with respect to the here- and-now of the moment of speech. An Iconicity Principle accounts for the sequentiality of sentences in a discourse (of inscriptions into the DR). Tense switching may effect relocations of the DR, e.g. embedding it into another DR with a different temporal index. Only present perfect and past- perfect (providing they are categorized as tense rather than aspect) seem to call for a local relocation of individual discourse events in the series of inscribed events.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 487 2) Aspect is traditionally seen as a way of representing the internal temporal contour of an event-situation, e.g. the notion of perfectivity is linked to the notion of boundedness of time intervals. In my paper, this appears as a consequence of the psychosemantic basis of aspect. The figure- ground distinction and the related mechanisms of "packaging" and "grinding" adequately explain the difference between mass- and count-quantification of predication types and the perfective and imperfective aspect. While predication types denote characteristics of event-situations, aspect denotes a speaker's perspective, an idealized cognitive model (with prototypicality effects), which at least in SMG can be superimposed upon any predication type. Both predication types and aspects can be distinguished according to whether they meet the criteria of partitivity and additivity. I suggested that not only the semantics of count- or mass-quantified predication will determine the choice of aspect but that this choice is strongly influenced by discourse considerations. Every discourse is also structured in terms of foreground and background. Foregrounded, focussed discourse events which rest on the idealized cognitive model of a count- quantified, individuated event-figure, have meaning characteristics which do not disturb the Iconicity Principle. Only insofar as the Iconicity Principle is linked to the problem of temporal ordering and sequentiality of discourse events (things happening inside a DR), will aspectual marking be implicated in the representation of the internal temporal contour of events. Presenting discourse events as mass-quantified will by implication denote temporal characteristics of discourse events that override the Iconicity Principle and force a reordering of discourse events in the series of inscriptions into DR. The interesting consequence of linking predication types and aspect only by implicature to characteristics of the internal temporal composition of event situations is the following: with the same degree of justification, spatial characteristics of predication types and aspect can be seen as implied by their idealized cognitive models. Although it is generally regarded as a universal that utterances are marked for their temporal characteristics in terms of tense and aspect (the latter minimally as predication type/Aktions- art and maximally as morphologically signalled aspect), their spatial characteristics are completely neglected. The proposed treatment of aspect is open to linking spatial characteristics by implication to the idealized cognitive models; that is treating temporal and spatial characteristics in the same manner. I thus account for the fact that in our presuppositional knowledge whatever happens takes time, but whatever happens also hap-
488 WOLF PAPROTT£ pens somewhere. Basing, as I did, aspect on a mechanism of human information processing (countability marking due to focussing) will correctly explain the problem of sequential inscription of sentences into a discourse representation if the Iconicitiy Principle is also assumed as valid. 'The internal-temporal-contour' reading of aspect will result from an interpretive exploitation of presuppositions and implicatures; spatial characteristics of aspect will equally follow from an exploitation of the geometric analogues of focussing and backgrounding. Drawing on a psychosemantic basis for tense and aspect simply and naturally also explains informational structures in discourse which are larger than the single proposition. The "epic" or "dramatic" effects of aspect contribute to informational coherence and relevance patterning via informational chunking. Tense anchors or locates a discourse representation. It therefore makes sense to regard tense and aspect as abstract categories or categorization schemata that figure prominently in the communicative contract between the participants of an interaction: "Mark stuff versus sorts (on the propositional level) and inclusion or exclusion in the relevant salient, foregrounded information (on the discourse level) by choices of aspectual categories. Use tense markers to provide for a deictic location of the discourse representation." Let's say that redundant tense markers will serve to indicate unity of time while aspectual marking will point to unity of action. Finally, an important contrastive finding of this article is that SMG perfective aspect is roughly equivalent to simple tensed accomplishment and achievement predications in English while SMG imperfective roughly equals progressives and simple tensed state and activity predications. Furthermore the synoptic, single act view expressed by SMG perfectives does not inhere in English accomplishments and achievements. The parallel between these categories rests on both participating in the prototypical idealized model of count-quantified event figures. Similarly, the English progressives with their meaning component of 'being in the process of are close to SMG imperfectives only in participating in the idealized prototype of a mass- quantified, backgrounded state. The actuality component of the English progressive is not given for the SMG imperfective. Where English lacks in an overt marking for perfectivity, the predication types of simple tense sentences will imply aspectual categorization. It is not surprising then that tense and aspect in both languages figure differently on the scale of verbal categories.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 489 Appendix 1: AijxiTaa neiQaxr]. "To ndaxa trie; YiaYidg." 1. Perfective main verbs 2. Imperfective main verbs (1) ++*-*--Kot—KadioauE (perf. past) va tov TtepLu£vouue (impf. pres.) Uta vuxia oav xnv aTtoiHvn, uovd- X.eq, TtapaKaXcivTaQ xn ddXaaaa va uaQ tov £ava<p£pEL (perf. pres.). (2) Kau fads (perf. past) y lo-y ud; (3)++** Nat, fipdE (perf. past) aAAd dxou (impf. imp.) va 6elq (perf. pres.) ttgjq fipdE (perf. past): (4) 0 Ttax£paQ uou nxav (impf. past) KarcExdviOQ. (5) Elxclue (impf. pastlkaikia 6i*d uas hl 6\o xagi6EUE (impf. past). (6) Epx6xav (impf. past) tt6xe tt6xe, Rat y^Ul^e (impf. past) xo otclxl uaQ UE xo uEYdXo kopul xou, ua.g y^Ul^e (impf. past)uE xa xd6ua xou, Kai £a~ vdcpEUYE (impf. past). (7)++** EvtdvauE (impf. past) u^veq va tov 6o6ue (perf. pres.). (8)** Eyoi auxd u6A.us xa duuduai (impf. pres.). (9)** AAAd xnv anoiKvri LOiopia 6e da xnv gExdau) (perf. fut.), hl Evtax6 xpovia va TtEpdaouv (perf. pres.) a>t6ua. (10) t6xe eXeltie (impf. past) au6 >taup6. (11)++** Mn'vEQ elxAue (impf. past)va xov 6ol>ue (perf. pres.). (12)++ 'EypcuJje (perf. past) axnv unx£pa uou 6tl da fpx.6xav (impf. cond. past) yia xo ndaxa nau 6pL- oe (perf. past) vtai xnv nu£pa. (13)** n^paaE (perf. past)6ua)Q tcov Ba'Cwv, n MEYdXn AEux£pa, n Tptxn, n TExdpxn, vti o Ttax£pag 6ev elxe <paveu (perfect past) . (14)++** MnxEpouXa, da fpdEi (perf. fut.) o Ttaxipas; xnv pcoxouaa (impf. past). (15)++ 9a 'pdEL, da 'pdEL (perf. fut.), uou 'A.EYE (impf. past) ue 3E3aL6xnxa. (16) Kai TtoxE da fpdEu (perf. fut.); (17)++**0kou Hai va 'vat (impf. pres.), £cpxaaE (perf. past). (18) 'HpdE (perf. past) nau n MEYdXn riEun:xn. (19)++** 'YaxEpa an6 to ueonue- PL, n MaptYco ue TtripE (perf.past) va TtduE (perf. pres.)v' avd^ouuE (perf. pres.) xa KavinAia axov 'Aylo NL>t6A.a - £va EVtviXncrdviL uou 1 xav (imof. past) \\ir\\a oio 13ouv6, oe Eva yhpeup Ttdvoj an6 xn ddXaaaa.
490 WOLF PAPROTT£ 1. Perfective main verbs 2. Imperfective main verbs (21)++ AAAd exeivh thv wpa dpxt- ae (perf. past) svac 6uvaT0c a£pas nou ar'ixuvE (impf. past) xai tlq TtETpec otTtO xduu. (2 2) Mas Pp^xe (perf. past) rcdvu qto 3ouv6. (20) And nipEQ fjiav (impf. past) xaLpdc xaAaauevoe. (34)++ 'Oxav (pTdaaue (perf. past) axnv ExxXnaud, 6e 3^-ETtaue (impf. past) nia xn OdXaaaa. (35)*Mia ftapid xaxaxvid thv elxe axETtdaEu ( perfect past). (23) Mac Tpa$ouoe(impf.past), uac EQEpvE (impf. past) ia cpouaTavia uaC/ uac QTpd3covE (impf. past) . (24) H MapuYci ue wpaiouae (impf. past) acpix^d and to x.£Pi- xau npoxwpouaaue (impf. past). (25)** And to \\>T)\6 uovondTi 3A.ETtauE ~ (impf. past) xdxo) ir\ OdXaaaa va x"tu- nuETai (impf. pres.) (26) H MapLYcS OXo 4jlOupl£e (impf. past) : (27)** 'Ayue NixOXa uou, wpa xaA.rj yia xelvouc nou fvai (impf. pres.) oth OdXaaaa. (28) Ma o xaipOc xeipotepeue (impf. past (29)++** Ae pXeuaue (impf. past) til a UTipoaTa uac xau 6ev gepaue (impf. past) nou va Ttpoxwp^aouue (perf. pres.). (30)++**KoupdYio,eXeye (impf. past) n MapLYci, va rcdue (perf. pres.) v'avd- 4jouue (perf. pres.) ra xavTrjA.ia. (31)** 0 TtaTEpas aou Ta£u6EUEi, (impf. pres.). (3 2) Kai uou 'acpLYYE (impf. past) tilo noXu to x^Pi/ hcxi ue Tpa3ouaE (impf. past) qto 6pduo, ttou 6ev tov 3^-£n;auE (impf. past) nua. (33) Eycj ExXauYa. (impf. past) tpouoly- uevn and to xaxd tou aspa, xat xeuvh 6A.0 eXeye (impf. past) npoaEux^S Yici xelvous nou EA.Eunav (impf. past). (37) 'YoTEpa and u)pE£ Yupuaaue (perf. past) oto otcutu. (36)** Tnv axouYQ-ue (impf. past) udvo va uouYHpuCei' (impf .pres.) and xdxco. (38) 'OXos o xdauos rjTav (impf. oto Ttd6i, xaTa) oto yia\6. past) (39)++** ELxav 6el (perfect past) Eva xal'xt va TtEpvdEU (impf. pres.) y.e'aa axnv Tpixuiaia, \±e ia navid tou tiou xovTEuavE (impf. past) va axia- touv (perf. pres.) an' to cpoua- xwua.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 491 1. Perfective main verbs 2. Imperfective main verbs (40)++** To eC6dv (perf. past) va TTEpvdei (impf. pres.) aav aaxparcri, viuvnYHUEvo art6 xov vtaip6, wai va xdvexaL (impf. pres.) uee axnv ouiX.- (41)** nepaae (perf. past)noA.u vtovxd an' xo XuudvL uac. X«*>PLC va unopEaEL (perf. pr»s.)va utieu (pert. pres.). (42) 'AA.A.OU A.EyavE (impf. past) ttcoq past) xo 6lk6 uac., hl 6l\- (44) Eviel paOid tiou elxe XO-Oel (perfect past) xo kgllki... (45) Me TifipE (perf. past) moll yu- pilaauE (perf. past) oxo otiltl. T^xav (impf. A.OL dAAo. (43) H untEpa uou, ue xa ua-A.A.ud tiou xns xa 'oepve (impf. past) o aEpag, UE xa udxia xpouaYU^va, vtouxaCE (impf. past) avt6ua. au£A.nxn xn OdXaaaa. (47)++ Aev nxav (impf. past) xo 6uvt6 uac., uou eltie (perf. past). (55)+ + ** To ' xclv YvcopuOEu (perfect past) vtai 6ev gEpavE (impf. past) av da fx.E ocjOel (perfect cond. past). (57) 'Oxav yupLOE (perf. past) au.6 xnv EvivtA-ncTLd, n Mapuyci) xn pci- xnoE (perf. past): (58) KupCa, da 3d^ouue (perf. fut.-) x' auYd; (59)** Kau Y^axL va unv xa pdi^Eic. (perf. pres.); ^EcjxiviaE (perf. past) n unTEpa uou aav TpeA.fi. (60) H MapLY^ xddnvtE (perf. past) xpouo-YUEvri. (46)++** Exa udxLa xnc TioA.Euouaa (impf. past) va 6La3aau (perf. pres.) xnv an6cpaan • (48)** Aux6 e£x.e (impf. past) u^YdXa Tiavid" xo 6i\i6 \iclq 6ev Eivai (impf. pres.) exol. (49)** Aux6 fixav (impf. past) £evo, oute xou x^plou uac. 6ev eLvai(impf. pres) (50)++** 0 6e6c. va xo (puA.d~ei (perf. pres.) ,'eA.eye (impf. past) n MapLYu vtA-aiYOvxac. (51) To MEYdA.o Edppaxo xo tipul nxav (impf-. past) XApd Geou. (52) H OdXaaaa vtaxaYdXavn t YO-A-fivia, garcA.cov6xav (impf. past) u^poaxd uclq vtdxco arc6 xov fiA.uo tiou eAxxutie (impf. past). (53) Ma 6e cpauv6xav (impf. past) XLTIOXE, OUXE XCuTlll, OUXE Tiav I . (54)++ 'OXoc o vt6auog eA.eye (impf. past) Hue. xo ua'iui tiou TiEpaaE (perf. past) fixav (impf. past) xo 6lm6 \lolq. (56) H unTEpa upu 6ev eA.eye (impf. past) xurcoxa vti 6Xo e\til£e (impf. past).
492 WOLF PAPROTT£ 1. Perfective main verbs 2. Imperfective main verbs (61)++** nou EioaL, rcou ELaai(impf. pres.) (pd)vaSe (perf. past) n unxe- pa uou Aypua. (62)** To 3pd6u Oa kdvouuE (perf. fut. on contextual grounds) Avdaxa- on ue xov acpevxn, x' dkouaEC, (perf. past). (63)** F.xnv Avdoxaon Oa 'vat (impf. fut.) e6u. (64) 8a fpOei (perf. fut.) an6vliE a>c xn vuxxa. (65) 6a 'pOel, Oa 'pOel (perf. fut). (66) Nai, Oa "'pdei (perf / fut.) . (67)++** H 6uoxuX.Lauevn i*|OeA.e (impf. past) vol &E3oLL(ioEL (perf .pres) rcpcoxa xov Eaux6 xnc kl uaxEpa Eudc,, ticoc aux6 rcou <po3oxav (impf. past) 6ev n"xav (impf. past) aXfiOELa. (68) 'Ox.l, 6ev urcopo6aE (impf. past) vol ■ xav (impf. past) aXnOuvo. (69)** 0 KaXoe Kaip6e elx.e £a)vxa- ve'JyiEL (perfect past) xn un.xepa uou. (70) '6\t\ xnv nuepa n.xav (impf. past) XOLpouuEvn. (71)++ Mae E3E3aucovE (impf. past) 6xl kdrtou nxav (impf. past) o rcaxe- pas, kdnou apYorc6pnaE (perf. past) UE xov Kaip6. (72)++** Kai Oa 'pxoxav (impf. cond. pres.) coc, xo 3pd6u, Yua vol \lt\ u^G acpriaEL (perf. ores.) uovax.EC, xo ndoxa. (73)++**n,Oa 'pOel (perf. fut.), eXe- YE (impf. past) eyco xov ^epco (impf. pres.)KaXd. (74)++** Kai yeuloe (perf. past)rtdA.L xo otilxl uols an6 EA.ru6a, Kai koixd- £au£ (impf. past) oXoEva arc6 xo 6coua, arc6 xa rcapdOupa, xn OdXaooa, 3e3aLEC. 6xl Oa 6ouue (perf. fut.) xa rcavLd uag, 6xl Of akouaouuE (perf. fut.)XLg ciXiivEQ xou Tiaxepa. (75)++ To Kak6 nou 6\oq o k6ouoc, eXeye (impf. past) 6xl ndOauE (perf. DastJELxe axauax^OEL (perfect past) £Ew and xnv n6pxa u^Cr 6ev uae ei"x.e au6ua cxyyC'Eel (perfect past) . (76) AAAd n uepa tiepvouge (impf. past) KaL n unxepa YLvdxav (impf. past) 6\o KaL tiuo aiconnXn.. (77)++**'Oxav dpxiOE (perf. past) va vuxxcovEi (impf. pres.), ue nn.pE (perf. past) arco xo xepu KaL xpa3n- EauE (perf. past) eEco an6 xo xupi6. (78) To anuXL uas n.xav (impf. past) axnv dkpn dkpn, Kovxd axn OdXaaoa.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 493 1. Perfective main verbs 2. Irnperfective main verbs (79) rUo nspa &px.lCcxv (impf. past) to. 0pdx.La, xau oto teA.oq fiTav (impf. past) £vcxq TxeXcopLOS xdftoe. (80) Ewel ipXertee (impf. past) 6ko to tc£A.o.yo£' axn pCCa tou fiTav (impf. past) u^a ueYciA.ri cmriA.ud YEudTn an6 cpuxia £spd. (81) MTifiKaue (perf. past) \i£oa axn cmnA-Ld xau xadtaauE(perf.past) (82)** Atx6 6co da 6ouue (perf. fut.)' tov TtaTepouA.n nou da ' pdEL (perf. fut.), uou eitie (perf. past) n unTspa. (83) KadiaauE (perf. past) xovxd xovTd axa £spd (puvtia. (84) BA-snauE (impf. past) 6kr] tti dd- Xaaaa uTipoaTd uct£» u£ rcspa nou ylv6- xav (impf. past) £va ue to oupav6. (85) Ae cpaiv6Tav (impf. past) TLTtoTa, OUTE TiavdXL OUTE TIOUA.1. (86)++** KoixdCauE (impf. past) ewel, Xcopie va ULXdue (impf. pres.) cocmou vuxTuae (perf. past) xaA.d. (86) ++** 'HOeXe (impf . past) va ue rcd- Pel (perf. pres.). (89)++** Eycj xXauYovxae TtapaxaAouaa (impf. past) tti unTspa va u1 acp^aei (perf. pres .) va ueivu (perf. pres.) exel uo-Cl xns. (90)** H UHTEpa, x^PLS va Ouuuaei (perf. pres), eotelXe (perf. past) xn MapLYco va (pspEi (perf. pres.) ULa xou(3spTa au.6 to otiltl. (91)** Ki auxi8!, x^P^S va. ulXt^oel (perf. pres.) £(J)uye (perf. past). (92)'Oxav YupLoe (perf. past) ue xnv xouftspTa, p.E TuA.L^avE (perf. past) \i£oa xau xadiaauE (perf. past) XL OL TpELQ EXEL. (37) 'Hpds (perf. past)xai n Ma- piY^ UE to (pavdpL. (93) AvduEaa axr| unTEpa xaL axn MapiYto, 6e ULA-ouaa (impf. past) xat 6e adA.Eua (impf. past). (94)**Aev ^Epco (impf. pres.) n6aE£ ojpeq UEivauE (perf. past)exel. (95) Aev xoiuououva (inpf. past) oute fiuouva (impf. past) EurtVLa. (96)'AxouYa (impf. past)xnv xap6id uou nou xtutiouoe (impf. past)xaL xnv avau- vot8) Tnc un^Epae.
494 WOLF PAPROTTE 1. Perfective main verbs 2. Imperfective main verbs (97) 'AHOuya (impf. past) rn OdAaaaa tiou un:auv63Ya.uve (impf. past) n"auxa axis KOUcpdA.es Tcov ftpdxcov, xdxco and xa n66ua uasx aav va avdnvee (impf. past)xu avxf\ Tiepuuevovxas ua£u ue uag. (93) Kouxa£a (impf. past) xo auoxd6u xns OdXaaaag, tiou 6ev <pauv6xav (impf. past) xaveva (pug, xau gacpvLxd vou^Ca. ( impf. past) tlcjoq TioXXd (pcoxdxoa TiXeave (impf. past) \iioa, ncog epxovxav (impf. past) npoc; eudg, Ticog T*ixav (impf. past) uau xo HaiML ua>g* dxouya. (impf. past) xn cpcovT1! xou uaxepa nou (pcivaCe (impf. past) axoug vauxeg. (99) Ma uaxepa 6\a xdvovxav (impf. past) hl 6\a xa cpcixa r^xav (impf. past) xf ao- xpa axov oupav6. (100) Sacpvixd, n xauTidva axoua- xnne (perf. past). (101) 0 n"xoe xnc, xapouuevoc, ye- uuae (perf. past) xov aepa. (102) Mae xpoua.Ee (perf. past) xau uag auvenT^pe (perf. past). (103) SeTLexaxx^xaue (perf. past) axouyovxae xn <paaapia an6 xo x^puo. (104) Mia xa-pd Y^ULCJe (perf. past) xov aepa. (105)++** 0 Xpuax6g avaaxaCvexau (impf. pres.)eurte (perf. past) n Map UYci hl exave (impf. past) xo axaupo xne« (106) ++**EuA.OYnu£vog va 'vau (impf. pres.) eune (perf. past)n unxepa" uou. (107) Anouaaue (perf. past)Yia tloAAt1! cipa xug xaundves, xn xa-pd xne Avda- xaane. (108) nepau nuaaxe (impf. past) ornv exxXncad ue xov rcaxepa, ue xig Xaurcd- 6eg u^g avauueveg. (109) Ku uaxepa eueva ue Tinyefperf. past) axnv aYxaA.Ld xou axo anuxu. (110) Eiyd auyd enaiH (perf. past) n cpaaapCa and xo xupud. (111) BuOuax^xaue (perf. past) TtaXu axn oiunr\, axo axoxd6u. (112) KoixdCaue (impf. past) xn OdXaaaa. (113) Kupia, eiTte (perf. past) aiyd n MapLyti, Tidue (perf. pres.) ma axo ortuxL. (114) Hepdaave (perf. past) xa ueod- vuxxa.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 495 1. Perfective main verbs 2. Imperfective main verbs (115)** 0 acpevTnc. Sa 'p&el (perf. fut.) oio onLiL uoXlc, ep&Eu(perf. fut.)*6e &a 'pdei (perf. fut". )e6u. (116)** ndue (perf. pres.) eltie (perf. past) n un"cepo.. (117) Me unpE (perf. past) and to X.EPL. (118) H Mapuyu ecpEYYE (impf. past) ue to cpavdpu . (119)++**Aev i*ipdE (perf .past) o na- Tepac,, etna (perf. past) Eyco hu dp- X.uoa (perf. past) va xXaLu (impf. pres.) . (120) H untEpa oute uCXn°"e (perf. past), oav va un u' dvtouaE (perf. past). (121) ©TdoauE (perf. past) xovTd OTO OTILTL. (122) ++** 'Kupia, to (po)C elvoll (impf. pres.) avauuEvo, eltie (perf. past) n Mapuyco ^acpvuaauevn. (123)** Eyco to 'x.a a3naEL (perfect past). (124) H uriTEpa uou 6e uUnoe (perf. past) rcdXi viaddXou. (125) Kupia, ttgoq dva^E (perf. past) to (pcos; (126) 'Avol£e (perf. past) tpeuov- Tac. Tnv e^corcopTa. (127) H uriTEpa rcpoxcopouoE (impf. past) oav va unv >taTa\d0auvE (impf. past). (128) MEaa oinv TpartECapCa \iolq tiepC- UEve (impf. past) o naTEpac.. (129)++ Mac dpna^E (perf. past) OTa fiuvaid rou urcpdToa xau ua-C <PL~ Xouoe (impf. past). (130)++ H uriTEPOL uiaoTiEdauEvri ExXaLye (impf. past) a.Tnv aynaXid tou, EHXauye (impf. past)£Tau orccoc, tiote 6ev Tnv gavaEioa (perf. past). (131) To vtal'vtu uas elxe pouXidEeu (perfect past), \ia o naTEpac elx.e ocoOeu (perfect past) kou eixe yu- pCoel(perfect past) and Tn x^oa ue Ta n66ua. (132)++ Elxe yuploel (perfect past) 6ucos, hu auT6 EcpTave (impf. past) OE UO-G- (133) Kai ndvaue (perf. past on contextual grounds) ndoxa u^Cl, oticjq to ' x.e tteu (perfext past) n un"c£pa.
496 WOLF PAPROTTfi Appendix 2 Excerpts from: A.A. Milne, Winnie-the-Pooh. New York: Dell (1979) A.A. MiAv, H-romvv'u-o-no'iHp. Adryva 1982: Obvooeac.. Translated by Paulina Pampouthi (1) One day when the sun had come back over the Forest, bringing with it the scent of May, and all the streams of the Forest were tinkling happily to find themselves their own pretty shape again, and the little pools lay dreaming of the life they had seen and the big things they had done, and in the warmth and quiet of the Forest the cuckoo was trying over his voice carefully and listening to see if he liked it, and wood-pigeons were complaining gently to themselves in their lazy comfortable way that it was the other fellow's fault, but it didn't matter very much; on such a day as this Christopher Robin whistled in a special way he had, and Owl came flying out of the Hundred Acre Wood to see what was wanted. p. 147 MIA MEPA,6tclv ma 6 f\XiOQ zlyjz gavayuPLoei am' Oean tou ndvco an'to Aaoog^pvovTae ua£i tou tt)v euo6id tou Mdn,xi 6Aa tol pudxia tou Adaouc yXouyAduxi£cxv euTUxiou£va nou gava3pfixav ndXi to oyopcpo oxn- ua TOU£/>a ol XiuvouXee 6veipeuovTav tt) &ori nou eixav x&vei^xai uee otti Seoraoid xcu tV)v fiouxia toG Adooue,6 xoOxoc 6ovtLpa£e npoaexTiKd rn cpwvri tou kl ohouye vd 6ei dv toG apeae i,xai Td dYPionepiarepa na- panoviovTouoav euyevivta orov eauT6 tou£,u£ t6v tzurzXimo xai dveTo tootio tioO ouvndi£ouv,>a eAeYav tuoq xdnoLoe a*Ao£ <piAapdHO£ cpTaiei, pa 6ev TieipdCei noa noAu,vud TeTOia u^pa Aoinov adv xi axsrt\,b Kpt-_ orocpep Poutilv ocpupige u'evav £161x6 tootio nou f\^zpzrMi r\ Kouxouftd- Yia node TteTcovTac dno t6 'Ehcltooto STpeupa toG Adaoue vd 6ei tl tt\v ndeAav.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 497 (2) Edward Bear, known to his friends as Winnie-the-Pooh, or Pooh for short, was walking through the forest one day, humming proudly to himself. He had made up a little hum that very morning, as he was doing his Stoutness Exercises in front of the glass: Tra-la-la, tra-la-la as he stretched up as high as he could go, and the Tra-la-la, tra-la-oh, helpMa, as he tried to reach his toes. After breakfast he had said it over and over to himself until he had learnt if off by heart, and now he was humming it right through, properly. Well, he was humming this hum to himself, and walking along gaily, wondering what everybody else was doing, and what it felt like, being somebody else, when suddenly he came to a sandy bank, and in the bank was a large hole. p. 22 and 24 O APKOYAOE "Evtouopvt,yvcjJOt6c otouq cpiAouc tou odv TVrouLWU-6-TIoup, n TToup ylcx ouvTOULa,TLepnaToC3oe uid uepa oro AdaoQ TpoYou&tfVTac. uoup- uoupLcrrd ot6v eauTO tou.Me wavidpi.Etxe cpTidgei eva tpolyoooioto uoup- uoupioua npcoL npcoi moO&jq £xave unpoord ot6v hoOp&ptti ttiv 'AvTpeiajuevri ruuvcxoTLKn tou: Tpd Ad Ad,Tpd Ad Ad,uoupuoupi£e moOcdg TevT0Jv6Tav 6ao til6 tioAaj UTtopoOoe,KaL ueTd: Tpd, Ad Ad, Tpd Ad-6x 6xox-Ad,Mo8o[>s npo- anaSoGoe vd cpTooei okuPovtoc. tlq uutcq tgjv tio6igjv TOu.MeTd t6 npoYeu- pa t6 *xe tlel ndAi xai ndAi otov eauTO tou uexpi nou t6 eixe dnoom- 3iaei,xai Tapa to vioupuoupi£e 6Aoaarjra,6ncjL>s enpene. To uoupuoupiCe Aoinov otov eauTO tou xai nepTrarouoe xapouuevoe *l avapcoTLOTav tl vd 'xavav apaYe oi aAAoi,xai odv tl Od 'uoia£e auxo nou Qd 'voicoQe dv titov xdnoioc oAAoc.,6tov gacpvixd ecpTaae ae uid au- uouoepn 6xSri xcxl ornv 6xSti olutti urapxe U^d ueYdAn Tpuna. (3) One day when he was out walking, he came to an open place in the middle of the forest, and in the middle of this place was a large oak-tree, and, from the top of the tree, there came a loud buzzing-noise. Winnie-the- Pooh sat down at the foot of the tree, put his head between his paws and began to think. p. 5
498 WOLF PAPROTTE Mid u£pa,vta8c&£ Sxave tov Tiepinaro tou,£(ptooe: a1 £va ££<jzjOto orri u£- oti toO Adaoug xai ani picrrj tou i^cpcoTOu fiTocv uid ueydVn (3eAavi6id, wa£ An1 t/)v HOpucpT*) xne 6xouY6Tav £va 6uvar6 (3ouiopa. 'H-rouLWu-6-noup KdOioe cml pi£a tou 6£vtoou ,dxouuTtrpe: to xecpd- Ai dvdueaa otlq unpooriv^c; tou narouoec; kl apxioe vd aK&pTeTai. (4) He was getting rather tired by this time, so that is why he sang a Complaining Song. He was nearly there now, and if he just stood on that branch ... Crack! p. 8 ECxe dpxioei vd xoupd£eTai,Yi' olut6 *ucli TpayouSouoe clut6 t6 xd- mug napccnovidpiHO TpaYOu6i.BpiaKOTav oua)C xovTd tgpcc,hi dv orexo- tclv Alyo a' £xeivo to xAafii... KpaTQ! (5) "I wish you would bring it out here, and walk up and down with it, and look up at me every now and then, and say Tut-tut, it looks like rain.' I think, if you did that, it would help the deception which we are practising on these bees." p. 15 f. Med 'deAa,dv unopouoee vd vf) cp£pei£ £63, vd xfiv xpaTd£,vd TiepnaTde ndvco xdTa),xaL xdfte tooo vd koltoq \\x\Xa xai vd Xee:"Te T£,Oappa) mug da 3pe£ein.Noui&o nLog dv to graves auT6,0d u£ 3orido0oee vd geyeAd- oco auTee tlq \ie\iodsq" . (6) One fine winter's day when Piglet was brushing away the snow in front of his house, he happened to look up, and there was Winnie-the-Pooh. Pooh was walking round and round in a circle, thinking of something else, and when Piglet called to him, he just went on walking. p. 35 f. Mid ouopcpn xeLucovidTixri u£pcx, xadd)£ 6 UiyyiXzT xa3dpi£e to xi6vi tlou eixe xAeioei Tnv tioptcx tou oultlou tou, ef'Tuxe vd onxcioei to xecpdAi xai vd 6ei Tn-rouivvu-T6v- noucp. 'O Iloucp, uepTiaTOuoe yupco yupco, xdvovTCXc; kukXouq hcli xdTL aAAo oxecpTOTav, ylolti l5tcxv t6v cpcovaF;e o UiyyiXex, auToe ££axoAou3rioe vd TiepnaTdei.
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 499 (7) The bees were still buzzing as suspiciously as ever. Some of them, indeed, left their nest and flew all round the cloud as it began the second verse of this song ... p. 17 Ot u£Aiooe£ 3oul£ov duoua,t6 l6lo xaxurioTiTa.Mepixec ud\Lora,&pr|- oav ttj cpcjiUd touq xl dipxiaav vd Tiexouv yixxo drc' to ouwecpdKL,xaQcjL)c; exeUvo dpxiCe vd Tpayouoa if\ Seuxepri orpocpn tou tpoyou6lo0 tou. (8) "I have just been thinking, and I have come to a very important decision. These are the wrong sort of bees," p.17 "Itipa uoAlc, ox&pmxa xl §3YaAa £va noAu £voLacp£pov ouun^paoua. Autec, £63 eivai AdSoe etooc \iiXiaosQ". (9) One day, when Christopher Robin and Winnie-the-Pooh and Piglet were all talking together, Christopher Robin finished the mouthful he was eating and said carelessly: "I saw a Heffalump to-day, Piglet." "What was it doing?" asked Piglet. "Just lumping along," said Christopher Robin. "I don't think it saw me." p. 56 MIA MEPA,xo&0£ 6 Kpiorocpep Poutilv xaL* f)-rouLVVU-6-IIoup xcll 6 IIlyhAet xou(3evtlcCclv 6Aol paCt,6 KpLorocpep Poutilv t^Aelcooe \ii ttjv unouxLd nou expcaye xcu sine dveueAa: "Etoa anuepa evav ecpdXevTa". "Tl exave;" pdrrnae 6 TIiyhAzt. "EdAeue frapLd xca xeuTi^ALxa" eine 6 KpLorocpep Poutilv. "Ae voul&o na)Q atJTOg u£ el&e". (10) "That's what we're looking for," said Christopher Robin. They went back to the others. Piglet was lying on his back, sleeping peacefully. Roo was washing his face and paws in the stream, while Kanga explained to everybody proudly that this was the first time he had ever washed his face himself, and Owl was telling Kanga <ut Interesting Anecdote full of long words like Encyclopaedia and Rhododendron to which Kanga wasn't listening. p.!22f.
500 WOLF PAPROTTE "Auto etvoa nou ilxxxvouue" elm 6 Kptlorocpep Poutilv. rupLOCiv tiloca),otouq 5AAouq/0 UiyyiXeTffjca-O^XLevoQ dvdoxeAa,KOLy6TCCV eCpnvLxa.To P6 eTiAeve if] uouoouoa xax tlq naTOuaeQ tou oto pudxL, Lvti r\ Kdyxou £&iYOuae a' oAoug uepTVixxva, ttio£ rJTav n upcoTn cpopd rtou TiAevoTav u6vo tou,xl evco r\ Kouxou3dyLa 6LriYOTav ottiv Kdyxou £va *Ev- 6Lacp£pov 'Av£x6oto 6X0 uaxpLec, Ae£e:LC,,6Tia)s 'EYxuxAoTxa.L6e:La xax TTl- xpooatpvri/Tiou r\ Kdyxou 6ev to axouye. (11) Then he put the paper in the bottle, and he corked the bottle up as tightly as he could, and he leant out of his window as far as he could lean without falling in, and he threw the bottle as far as he could throw-splash!- and in a little while it bobbed up again on the water; and he watched it floating slowly away in the distance, until his eyes ached with looking, and sometimes he thought it was the bottle, and sometimes he thought it was just a ripple on the water which he was following, and then suddenly he knew that he would never see it again and that he had done all that he could do to save himself. MeTd z&xke to xoptl u£oa ottiv unouxdAa,TT| (touPuoae 600 tilo ocpLXTa uno- pouae ue t6 cpeAAOjEoxujje e&o arto t6 -napdOupo 000 tilo tioAu unopouae xu~ pLQ vd TieaeLfXaL tt)v ti£tcl£e: ooo tilo paxpLd,u'nopouae - TtAdTC,! -E£ Xiyo f] uTiouxdAa cpdvnxe nd?iL navcd or6 vepo.Kl £ue:lve: tn£i xaC ttiv napaxoAou- Oouae ttdu dTtoviaKpuvoTav TiAeovTac, 6pYd,u£xPL tiovj tov Tioveonv Ta paTLa an'to xoLTayua.KaC ttote: voul£e: txcjc (3A£ti£l tt)v utiouxoAo. xatl tiote: ttloq f^Tav uovo £va xupcrudxL oto vzpd clut6 rtou napaxoAoudouae.KaL ueTd,xa- TaAa&e: iacpvLxd ttios 6£ Od ttiv £avd3Aeue ttotg xcll ttio£ elxe xdveL 6,tl UTiopouae yid vd CTjoScl. (12) The Piglet lived in a very grand house in the middle of a beech-tree, and the beech-tree was in the middle of the forest, and the Piglet lived in the middle of the house. Next to his house was a piece of broken board which had: "TRESPASSERS W" on it. When Christopher Robin asked the Piglet what it meant, he said it was his grandfather's name, and had been in the family for a long time. Christopher Robin said you couldn't be called Trespassers W, and Piglet said yes, you could, because his grandfather was, and it was short for Trespassers Will, which was short of Trespassers Wil-
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 501 Ham. And his grandfather had had two names in case he lost one — Trespassers after an uncle, and William after Trespassers. "I've got two names," said Christopher Robin carelessly. p. .34 0 IHFKAET £oOae o1 £va noXu yEYaXdnperto on it i,an*i u£on Uioq 6£i&q xai t\ 6£i& 0plok6tgcv ott*i u£oti toO Adoouc hi 6 THiyhXzt Bploh6tov ani yean toO onltloO tou.AitiAo. ax6 on it i Tou,urcopxe uui anaou^vri Tayrt£Aa nou gYPOOJG nduo TTiQ 0 ITAPABATHS TI. wOtov 6 Kpiorocpep P6univ pcoxnoe t6v niYxXeT tl orfljaive auT6,ToO dnavxrioe ncoc flxav tou nannou tou T'ovoya xai t6 elxav £66 nau noXu xaip6 an\v oixoY^veia. *0 Kpiorocpep P6umv eJjte,TT£OQ 6£ Od unopoOoe xa- veie vd Xiyexai 0 ITAPABATHS TI,hi 6 niYxXeT 3e&aioxre TT£0Q,vai,3d unopoOoe, dipou 6 nannouQ tou Xcyotov gtol,6 napa^dxTiQ Ti,nou ^tov t6 x&i- 6euTix6 tou 6v6yaro£ *0 riapcx^dTTiQ Tiy65eoe.Elxe 6u6 6voyaTa 6 ttocttrouq tou/Y id TTiv nepCnTCjan tiou 6d 'xave t6 Eva -I!apa3dTTiQ Xgy6tov yuxci gtol XeY6TOv Evae Ogloq TOu,xai Tiy69eoQ XeY^xav Y^aTi Xgy6tov napa- 0dTTlQ. "Ki ty& £x& 6u6 6v6yaTa" elne 6 Kpiorocpep P6yjtiv,dv£y£;\a. (13) So, while you walked up and down and wondered if it would rain, Winnie-the-Pooh sang this song: p. 17 Ki eTOi,xa8cjj£ eou nepncnroOoeQ ndvto xdTa) hi d\x3pUTid>oouv y£Ya\&pcdva dv 3d Bp^geL^-rouLWU-^TIoucp TpaYouSrpe auT6 t6 TpaYou6i: (14) Eeyore, the old grey Donkey, stood by the side of the stream, and looked at himself in the water. "Pathetic," he said. "That's what it is. Pathetic." He turned and walked slowly down the stream for twenty yards, splashed across it, and walked slowly back on the other side. Then he looked at himself in the water again. p. 72
502 WOLF PAPROTTE 0 AXBAX,6 yxpiZpQ Yepo-Ydi&3pO£,cn:EH6Tav ornv 6x$n tou puaxiou xai KOLxa^e tti cpaxod tou uee end vepo; "Euyklvtitlko Trp&Yua" erne."Nd tl Ervcu:EuYHivnTiH6n. rupioe mC Tiepndrnoe 6pYd 6ttoAou8cjJVTa£ t6 pudxi.Kauid eiHoaapid ViETpa til6 >c&tcd, Tt£paoe &Tt£vavTi TLAaTooupL£ovTaQ/xaL REpnonxioE naAi apYd/MdvovTae t6v C6lo &p6\sd,otAv 6XXr\ &x&r\. "Erie i to. oxdb9r)Ke Hai gavaxoiTaEe: Tt\ cpdToa tou pie aro vepo. (15) So Kanga and Roo stayed in the Forest. And every Tuesday Roo spent the day with his great friend Rabbit, and every Tuesday Kanga spent the day with her great friend Pooh, teaching him to jump, and every Tuesday Piglet spent the day with his great friend Christopher Robin. So they were all happy again. p. 109 *H K&yxou Mat t6 P6 Aoinov^ueivav end Adooe.Kai xd&e TpiTr^Td P6, Tiepvouae tt*i y^pa. tou kovtol aro vzydXo <piAo tou t6v KouveAo/xaC xd&e TpLTTl/T^ KdY^OU UEPVOUOE TTl U^PO. TT|£ KOVTd OTO U^YO^O (pCAo TT|£ TOV no&p,ixb9aivovTdc tov vd my&a mou ndtSe: TpiTTi 6 UiynXsT nepvouae tti Vi£pa tou KOVTd ar6v Kpiarocpep Popjtiv.Ki etoi,6Aoi fiTav ndAi eutuxl- OU^VOL- Notes * I would like to thank Vicky Alexandridis for comments and suggestions on a previous draft as well as for meticulous proofreading of the Greek examples. I am also indebted to Wendy Bonk for suggested improvements of my English and to Brygida Rudzka-Ostyn who pointed to necessary clarifications. 1. The metaphor of the time line is explicated in terms of the set of positive and negative real numbers and axioms of denseness, continuity, discreteness etc. Even in discourse oriented studies such as Partee (1984), Hinrichs (1981; 1986) tense and aspect are ultimately reduced to designating sequences of temporal points and intervals. 2. I share Schnelle's conviction "that ... the meanings of linguistic utterances are not determined unless with respect to the parameters of mental organisation" (Schnelle 1980:322). For a similar point of view see also Chafe, W.L. (1973) "Language and Memory." Language 49.261-28. 3. There are only a few exceptions to the obligatory marking of aspectual distinctions on the SMG verb. Unfortunately, these present some difficulties for a neat analysis: The auxiliary have and some of the most frequent verbs are aspectually defective. Without perfective stem:
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 503 eium "I am"; exco "I have"; £eqcd "I know"; Jieoiuivco "I wait, await, expect"; xdvco "I make"; jiqejiei "it is necessary"; jiqoxeitcii "it is a question of"; avrpuo "I belong"; XQCDotcicd "I owe". Without imperfective or imperfective generally not used: xQr)\Lat\£,oi "I serve as"; aJioxoiuieum "I fall asleep"; biayiyvwoKW "diagnose"; (cf. Mackridge 1985:103 f.). These forms could be treated as aspectually neutral, as always marking the morphologically signalled aspect thus simply being defective forms, or as aspectually ambiguous such that context (adverbs, complements etc.) and the verb's Aktionsart potential collaborate to disambiguate the intended aspect on each occasion of use. Substitution tests strongly support the latter view, thus in xca xdvau.e (impf.past) fldaxa u.a£i, ojicdc; to 'xe jiei (perf. past) r\ \it\teqcl. the morphologically imperfective form xdvau£ can be substituted only by perfectives, e.g. by yioQiaoa\i£. 4. The only exception to the rule that aspect can not be shifted are cases of the gnomic aorist: the use of a perfective form to express general truths, beliefs, sayings, and proverbs (cf. A.A. Ttaoitavo-u. 1946. NeoeUrjvtxi] £vvra£t£. Adrjva^. Vol. A. 272 f.). We can also consider the large class of SMG deponentia a case of voice shift. Thus, strictly speaking, neither SMG tense, nor aspect, nor voice has a stable and invariant meaning . 5. "Converters": a class of verbal, semimorphological preparticles causing transposition of the whole clause into a subordinate clause (Barri 1981). 6. The categories and their interdependence as discussed in the text are summed up in the following table: MOOD ASPECT \ 6 c W5 •»-> e g imperfective active/passive present past future present past present perfective VOICE active/passive present past future present past present perfect active/passive present past future present past future W/////A 7. Combinations of these classes of adverbials are possible: coordinate cardinal count and point of time adverbials co-occur with the perfective. In combination with an indefinite frequency count adverbial, a point of time or a cardinal count adverbial is included in the scope of the indefinite frequency count or duration adverbial as in daily at six o'clock, every day at noon; the imperfective is therefore obligatory as in Kdfre pica jidco va ijjamacD oxic, e£i. "Every day at six, I go out to shop." Furthermore cardinal count or
WOLF PAPROTT£ point of time adverbials may be understood as being elliptically included in the scope of an indefinite frequency count adverbial, or contextual considerations may induce such an interpretation: (:iav6a) Epxoiav otic, e§i. "He (always) came at six." = "He used to come at six." The perfect allows either class of adverbs but may not co-occur with adverbials denoting intervals or periods of time prior to the moment of speech such as xiic, "yesterday"; jipoxieg "the day before yesterday"; xr\v jiepaou-EVT] Kupiaxr) "last Sunday"; etc. This observation points towards a treatment of perfect as a tense rather than as an aspect of SMG. However, the fact that in most cases perfect forms may be substituted by perfective ones suggests a treatment of perfect as an aspect. There is an interesting analogue of "packaging" in traditional grammars of SMG. The Aorist is considered a "synoptic" tense: "The aorist differs from paratatikos in the following way, although the paratatikos signifies an action which belongs to the past, it presents it in its development, that is, as not yet complete; the aorist presents the same action synoptically in its totality, that is to say as already complete." (A.A. T^api^avou. 1946. NeoeIIy\vixy\ 2ijvxa|ig. Vol. A . 271; 177; transl. W.P.) "Vorwegnahme eines potentiellen Figurinhalts" (Ehrenstein 1954 : 322). Newton (1979) uses "scenario" to deal with expressions denoting multiple events of the form 'Every time p, then q' where the antecedent is either overtly marked in the surface structure as in Whenever he sees me, he insults me. or implicit in always, never, and non- conventional quantifiers such as rarely, often, usually, e.g. John always drinks coffee, can be translated as 'Whenever John drinks anything at all, he drinks coffee.' As imperfectives suggest an unlimited series (of repetitions) of acts or events, while the perfectives signal single acts, the aspectual distinction is treated as on a par with the distinction between every (perf.) and any (imperf.). In modal contexts any may be taken to represent the universal quantifier with wide scope, while every and all indicate narrow scope in relation to the modal. Take now You can come to me on every Saturday and we shall speak Greek. POSS V t (it is Saturday at t d you come at t) You can come to me on any Saturday and we shall speak Greek. V t (it is Saturday at t d POSS at t (you come)). In the first case, the universal quantifier has narrow scope and yields a typical scenario expression in which antecedent and consequent verb forms are associated with a universally bound t and therefore imperfective. In the second case, the universal quantifier has the wide scope associated with any and the modal buffers the consequent from the quantifier. Hence, the perfective aspect in SMG obtains whenever the complement does not contain a universally bound t. The predication type of a sentence is therefore conjunctively signalled by 1) adverbial adjuncts; 2) transitivity; 3) definiteness of the verb's arguments, all or some of which serve to override 4) the underlying Aktionsart characteristics of the verb. Dowty (1986 : 44) gives a set of examples which nicely demonstrate a verbs's aspectual flexibility and the interrelation between these parameters: (a) John noticed the rare seashell on the beach, (achievement) (b) John noticed rare seashells on the beach, (activity) (c) Tourists noticed the rare seashell/rare seashells on the beach, (activity)
A DISCOURSE PERSPECTIVE ON TENSE AND ASPECT 505 14. In terms of an interval semantics the test for mass-quantification can be captured by Dowty's rule (1986 : 44) "(PROG y) (i.e. the progressive form of y) is true at I if there is an interval I' properly containing I such that y is true at I'." Dowty mentions also that the English progressive should be given a modal interpretation that involves only some of the possible intervals containing I (lack of the strict subinterval property). A reason for this resides in progressive sentences which are not true at the moment of utterance: Vlach 1981 gives the following example: someone walks into a theatre, points to an empty seat and asks Is anyone sitting here? For the intended interval I'(length of the performance) the utterance is true while false for I (moment of utterance). Note that the lack of the strict subinterval property also holds for simple presents in a generic, routine or habitual reading (Doctor, this tooth hurts, (but no actual pain); My car makes a strange noise, (but no noise can be heard when the mechanic tries out the car). Pragmatic bridging assumptions will take care of possible misinterpretations. 15. This pattern works for a large class of verbs, e.g. give a: chuckle, smile, cackle, giggle, squeak, howl, bark, cry, snort, sigh, gulp, cough, look, glare, kick, shudder, nod etc. 2. take a: drink, look, breath, stride, pull, gulp, draught, dip, peep, prowl, glance, puff, cut, guess, turn, notice, leave, departure, offense etc. Thus, general purpose verbs such as give, take, make, have, be, do, get, put will combine with nouns of actions to form truly event - quantified predications (Rensky 1966).
Semantic Extensions into the Domain of Verbal Communication Brygida Rudzka-Ostyn University of Leuven 1. Introduction In his (1979) study of the conduit metaphor, Reddy finds English, as a metalanguage, "its own worst enemy" (p.286) in that it biases thought processes towards conceptual frames that are both misleading and blinding. Under the influence of one such frame, linguistic units are viewed as containers for ideas. Under the influence of another, language is conceptualized as pipelines conducting thoughts and feelings from speaker to hearer (p.290-291). Both frames manifest themselves in a broad range of expressions commonly used to refer to language and language use. From Reddy's vantage point, the biasing effect of these expressions on their users appears as an obstacle to successful communication and cultural growth. One wonders, however, what in the first place gave rise to the frames which the expressions instantiate. Why should speakers of English conceptualize communicative acts in terms of physical motion and view language units as concrete objects? Can this phenomenon be related to other, more general, cognitive processes? And if so, how? Further questions concern the expressions themselves, in particular their internal structure. What motivates the choice of particular items as components of given composite expressions? The underlying conceptual frames cannot be said to play a determining role here since the acceptance of these very frames has led to different choices in other languages. The purpose of this paper is to suggest answers to these and related questions. The discussion will be based on several hundred sentences compiled from novels, magazines, and dictionaries, or elicited from native speakers. Excluded from the corpus are expressions related to written com-
508 BRYGIDA RUDZKA-OSTYN munication or to the products of speech and writing; rather, the focus here is on the act of verbal exchange as such, and then only on its oral manifestation. Regarding the linguistic categories treated, the center of attention remains throughout the verbal predicate and its accompanying particles and prepositional phrases, but obviously all of the verb's arguments are regularly taken into account.1 2. Underlying Concepts and Assumptions The approach to sense extensions advanced here assumes primarily the framework developed by Langacker. It relies on the claim that semantic structure equals conventionalized conceptual structure and the meaning of an expression can be defined only with respect to some domain. Concerning the latter, it is assumed that "any knowledge system or conceptual complex — regardless of its size and scope — is capable of serving as a domain" (Langacker 1984: 173). This approach to meaning is determined by essential features of human conceptualization, one of which is figure/ground alignment (cf. Langacker 1983; Talmy 1978b). We perceive and recognize entities by contrasting them with others. These other entities function as ground against which the former stand out. In a similar manner the figure/ground contrast manifests itself on the level of designation. According to Langacker, The semantic value of a linguistic expression ... derive(s) from the imposition of a profile on a base. The base (or scope of predication) consists of those facets of pertinent domains that are directly relevant to the expression's characterization, hence necessarily accessed when the expression is used. The profile for an expression is a substructure which the expression designates, making it maximally prominent within the base (1985: 110). In view of our subsequent discussion, it should be stressed that the figure/ground contrast may operate also within a profile. For example, the verb enter includes in its profile an entity, the landmark, which serves as a ground element for locating another entity, the trajector. Such complex profiles are characteristic of relational categories, whether verbs, adjectives, adverbs, prepositions, or verbal particles.2 Members of all these categories designate relations holding between two or more entities in some domain. This is what puts them in marked contrast with nouns. Nouns are claimed to designate "things," where a thing is "a region in some domain;" in the case of count nouns, the region is said to be
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 509 "bounded" (Langacker, this volume).3 Whether the domain relates to perceptually accessible objects or to abstractions, man appears to have the ability to discern in it different parts and recognize the boundaries that mark or preserve their identity. In other words, he is able to recognize not only the concept of a house as a bounded area, but also notions such as [ATOM], [CENTURY], [DUTY] or [THOUGHT]. Even in the case of a homogeneous unbounded mass, his cognitive apparatus allows him to conceptualize it in terms of bounded portions. Many language devices bear, in fact, witness to such compartmentalizations. Very revealing in this respect is Talmy's discussion of various quantifying expressions (this volume). In addition to the notion of bounded region, another concept introduced by Langacker will prove useful, namely that of primary domain (1983: II/B). A primary domain is one that is necessarily activated by the meaning of a given lexical item. For instance, in the case of the verb/7y, the domain of motion through physical space must be activated whereas in the case of think the primary domain is that of mental processes. A primary domain is often a cluster of different domains, some of which form an integral part of it, while others are of peripheral importance. We will identify the former as central component domains. To give an example, the verb peddle evokes the domain of trade as its primary domain, with physical movement being one of its central component domains. Needless to add, characterizing the meaning of an item in relation to its primary domain may pose formidable problems. For one thing, it presupposes a clear picture of the structure of the given domain, and such a picture is often beyond one's reach. Matters are already complicated when the primary domain is space, but even more problems emerge when this domain results from an interplay of several domains, some quite inaccessible to precise description. The field of verbal communication is such a domain. 3. Verbal Communication: Its Multi-Level Components and Its Donor Domains In visual terms, verbal communication can be located at the intersection of many quite disparate domains. As Figure 1 illustrates, one of the relevant domains involves concept formation, another one pertains to perceptual processes, and still another concerns cultural and social conventions.
510 BRYGIDA RUDZKA-OSTYN ^V $r ^ .^ >. w lo-Ps 7 7 ^ / -e&* > '^N w \ \ Figure 1
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 511 Clearly, Figure 1 not only oversimplifies matters, but, to a certain extent, also distorts them. Given the fact that sound waves travel through physical space, the circle representing the spatial domain could cross that of acoustics. Moreover, arguments might be adduced in favor of treating time as an all-embracing domain and thus in favor of including the remaining domain- circles in the time circle. It should also be pointed out that each of the domains has its own complex structure comprising a variety of elements. The domain of acoustics, for example, includes the dimensions of volume, pitch, and duration. In turn, the domain of social conventions embraces a whole gamut of scales and norms interacting with various behavioral patterns. Obviously, not all of the domains mentioned are activated by every verb of speaking. While beseech automatically invokes the speaker's emotional state, aswer or say has no direct link with the emotive domain. Depending also on the context, the same verb may activate a different set of domains. Compare in this respect the use of say in He said the play was boring and Say "aa," please. The subtle interplay of domains becomes accessible only through concrete acts of communication, and any attempt at describing the semantics of a verb of speaking (VS) must take into account the relevant components of such acts. In recent years, they have received a great deal of attention as a result of which increasingly refined models of speech events have been proposed.4 For the purpose of our discussion, we will adopt a very simple model, partially based on Verschueren (1984). It can be presented schematically as in Figure 2. Important here is the distinction between a primary speech situation and that of a reporter5 for a speaker can use language not only to express his own ideas but also to report someone else's speech act. In so doing he can include information about who said what to whom, when it was said, and how. In other words, any element of the situation in which the original speech act took place may be the input to his report. The report itself is also part of some situation6 which must be distinguished from the original one, otherwise there is the risk of attributing to the primary utterance features which properly belong to the reporter's speech act.7 Our knowledge of different situations in which a man can find himself serves here as a domain against which a speech event can be characterized as unique and distinct from e.g. a game of chess, a concert, a swim, or a fox hunt. In turn, our knowledge of different speech events allows us to sub- classify them into inquiries, orders, requests, threats, or promises. This
512 BRYGIDA RUDZKA-OSTYN PRIMARY SPEECH SITUATION interpersonal - relations speaker beliefs intentions utterance channel \ I propositional I I fcontent -hearer beliefs intentions -V \ I linguistic form \ » j manner of presentation / '■■} \ \ / / \ \ \ / \ 4 speaker ^ (reporter) channel beliefs intentions utterance (report) -^hearer beliefs intentions interpersonal, relations SECONDARY SPEECH SITUATION Figure 2 knowledge functions as a domain wherein the different subclasses can be identified or bounded. Similarly, our knowledge of inquiries can become a domain for a further characterization of their subtypes. In each case a bigger knowledge structure is invoked to characterize, i.e. set off or bound, one of its segments. Subject to such cognitive bounding are whole speech situations as well as any of their elements. An utterance can be set off against a whole class of utterances as a bounded region and so can any of its components, whether on the propositional or acoustic level. A word, an
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 513 idea, a sound can all be isolated as self-contained entities within appropriate domains. As is obvious, this is just another manifestation of the already mentioned figure/ground alignment. And the nature of the ground does not matter. The ground may be more concrete and measurable or quite abstract. What really matters is our ability to set off conceptually entities against other entities. It is this ability that will be constantly invoked in our discussion of metaphoric processes in the domain of speech acts. To come back to the distinction between primary and secondary speech situations, the secondary speaker can focus on any component of the primary speech act in his report. As the interrupted lines in Figure 2 are meant to suggest, he may focus on the primary speaker's attitude vis-a-vis the addressee or the message, on his manner of speaking, on the proposi- tional content of what was said, or on the impact of the utterance on the primary addressee. The reporter may also decide to reveal his own point of view as well as his emotional attitude towards any component of the reported situation. Depending on what he intends to highlight or hide, he can choose between direct or indirect enunciation, between neutral category labels and emotionally charged expressions, between literal and figurative language. In addition, there is a whole gamut of specific lexical items to choose from, each of them activating different domains or different portions of some domain. Since our concern here is with figurative expressions, we will now take a brief look at the domains that these expressions activate. The domain most frequently evoked will be examined more carefully in the next section. A great number of figurative expressions framing speech acts contain verbs or combinations of verbs and particles/prepositional phrases whose primary or central component domain involves space or, more precisely, spatial motion. The present corpus of 700 sentences8 furnishes 374 such examples, only 7 of which denote stative locative relations. Spatial elements may also enter the domain of speech acts by means of prepositional phrases or particles alone. Examples (l)-(3) illustrate the different possibilities. (1) He broke the news that she had died. (2) "Well, I don't care," the chaplain flung back with undiminished zeal. (C-22: 458) (3) He told the office boy off for making so many careless mistakes. (ADE: 906) In all three cases, the spatial domain can be considered the source of
514 BRYGIDA RUDZKA-OSTYN transfers into speech acts. Following Kittay and Lehrer (1981), we will call this and any other source of semantic extensions donor domain and the target of such extensions recipient domain. In addition to space, there are other domains that extend into verbal communication. One of them involves various subdomains which can be subsumed under the general term of acoustics. (4) She leant towards me and hissed: "What a dreadful hat Mr. Tom- kins is wearing!" (5) "Next?" barked Milo's deputy mayor officiously. (C-22: 242) (6) ... Mondale began harping on Reagan's competence to lead ... (TM, N 5, 1984: 10) (7) "Ha! Assassino!" she hooted, and raced joyously across the room ... (C-22: 405) (8) The officer rasped out an order. (DPV: 485) The domains of acoustics and space frequently interact when they extend into speech acts. The interaction is brought on either by a spatial particle or preposition (cf.(6)&(8)), or a verb denoting discharge of some substance: (9) Come on, cough it up, we know you're guilty. (DPV: 110) (10) "You're not 11 points down," he erupted, "you're over 20 points down." (NW, N/D, 1984: 86) Another domain which emerges from the database as a source of transfer is that of cognition. Verbs such as muse, reflect, speculate or reason are used in place of true VSs, as in: (11) "How old is Sapphire, I wonder?" mused the senator, creating a welcome diversion. (C&M: 61) (12) "I'm probably just as good an atheist as you are," she speculated boastfully. (C-22: 183) (13) "Your country doesn't need your help any more," Yossarian reasoned without antagonism. (C-22: 454) Finally, the domains of perception, emotion, temperature, cooking, torture, and a few others (sometimes in conjunction with the spatial domain) can give rise to semantic extensions: (14) "At least we found out what he dreams about," Dunbar observed wryly. (C-22: 133)
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 515 (15) "Then I really do have to fly the fifty missions, don't I?" he grieved. (C-22:60) (16) ... they accompanied him ... grilling him with possible debate questions ... (TM, 0 29, 1984: 9) The following table summarizes the survey of the corpus: o ining 1 > perta verb SPACE + MOTION exclusively or partially ACOUSTICS COGNITION OTHER DOMAINS VERBAL COMMUNICATION total without particle or prepositional phrase ISIlllllSli 32 13 2 96 with particle or prepositional phrase ^?:-::'::'::::^:^:^:-:^:^:^:'::?:^:-,.:^ Hill ^Ki^P liii!ilfi?9iil!lll 604 together 374 84 14 12 216 700 %of total number 53 12 2 1.7 30.8 Altogether as many as 653 constructions (see the shaded boxes), i.e. 93% of all the data, refer to the spatial domain, either via the verb or the particle/prepositional phrase, or both. Moreover, 98% of the 374 verbs grounded in space denote motion, not static relationships. In this context, one may indeed wonder why reports of speech acts draw so heavily on spatial motion and not on other domains. What is it that makes the area such an interesting donor domain? The rest of the paper will center around this question. We will focus on the 367 examples that are related to physical movement through the verb or
516 BRYGIDA RUDZKA-OSTYN both the verb and the accompanying particle/prepositional phrase. Prior to discussing them, we will take a closer look at the nature of motion events. As a first step, let us consider the relationship between our perception of physical motion and acts of comparison. 4. Spatial Motion 4.1. Spatial Motion and Acts of Comparison It has long been observed that a metaphoric sense extension rests on a similarity comparison. According to Gentner (1982), such a comparison may involve either object attributes or inter-object relationships. In the latter case, the objects themselves may be quite dissimilar. This is also true of comparisons linking speech acts with motion events. The participants in these events as evoked by the present data are invariably three-dimensional, concrete objects. Their counterparts in the domain of communication are opinions expressed, themes and topics, attitudes, and various states in which the primary speaker or hearer may find himself. In other words, the donor domain provides us with concepts related to concrete objects whereas the recipient domain, notwithstanding the fact that speech-act participants are as concrete as can be, offers mainly abstractions for the comparisons drawn. But the relationships holding between the concrete objects may be perceived as similar to those obtaining between abstract entities, and here another problem arises: what exactly warrants the perception of similarity? This brings us to Langacker's study of acts of comparison and our conceptualization of motion (1983 II). An act of comparison involves, as Langacker observes, a standard, a target, and an evaluation of the target with respect to the standard. The standard and the target are asymmetrically related, that is a familiar or simply prior event serves as a standard for evaluating an unfamiliar or later event, not the other way around. The evaluation does not necessarily have to concern the events in their entirety but may be limited to one particular aspect. Moreover, acts of comparison may themselves be compared, thus participating in higher-order comparisons (op.cit.: 5). Acts of comparison are critically involved in how we perceive and conceptualize spatial movement. Assuming that our conceptualization of some entity e] at location 1. and time tk constitutes cognitive event (1) and the conceptualization of entity e] at 1. + 1 and tk + { constitutes event (2), registering the change in time and location will result in a higher-order event. This
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 517 event will be an act of comparison, with event (1) serving as standard and event (2) — as target. In case more events of the same nature follow, event (2) will be the standard of comparison for event (3), and so on, giving rise to a whole "chain of transitions" (op.cit.: 76) as in: (1)>(2)>(3) ... where > symbolizes the operation of evaluation. In its essence, spatial motion is then nothing else than a series of consecutive changes in the relationship of location holding between a given object and its domain. Such changes are captured via acts of comparison of the kind discussed above, and commonly expressed by verbs of movement. An examination of the spatial meanings of these verbs and their extended senses, as e.g. in This milk is about to go sour (op.cit.: 77), has led Lang- acker to an important claim, namely that spatial motion is only a special case of a very general and much more abstract phenomenon, which he calls "abstract motion" (idem). Both spatial and abstract motion involve changes in the chain of interactions between the given entity and its domain. The following formula characterizes this common base: nm/r0]tl > [NrJt Il>[Hr2]t 1 > L c Jt() L c Jtj L c Jt2 (op.cit.: 79) (t ,t t ... = consecutive points in conceived time T T,T...= consecutive points in "processing time," i.e. the time needed for the conceptualization of a given event C = conceptualizer) If m is the moving object, the relationship fm/rj specifies ra's location within the domain of space; if it is the milk, fm/rj indicates the milk's "locus" in the domain of taste, i.e. "a particular taste sensation potentially induced in someone sampling it" (idem.). This general concept of motion provides already a partial answer to the question of why so many verbs of movement and spatial particles are used in the domain of communication. Both movement through physical space and speaking concern events ordered in time. The conceptualization of both relies on a chain of comparison events registering change. Since a moving object manifests changes between consecutive events in a more tangible manner than other entities interacting with their proper domains, spa-
518 BRYGIDA RUDZKA-OSTYN tial motion should be a particularly apt representation for whatever undergoes or accompanies transition.9 Before, however, the details of the linkage between spatial movement and speaking are discussed, it may be useful to examine briefly the components of motion events and their linguistic expression. 4.2. The Components of Motion Events and Their Linguistic Expression At a very schematic level, the structure of motion events is quite simple. Among the components that come to the foreground are the moving object, its trajectory, and the entity with respect to which the trajectory is assessed. The moving object emerges as the figure against the rest of the scene, i.e. the ground. As physical motion often constitutes part of a causative event, the entity which sets another in motion becomes also a relevant component. This component assumes the status of figure vis-a-vis the motion event, the latter being now relegated to the position of ground. Motion and causation may, of course, merge into one event, and then the figure is the entity that both causes and executes motion. In the situation illustrated by, e.g., John went home, John is such a figure. Finally, an agent that brings about a motion event may do so by means of another entity. Although of an instrumental nature, this entity also seems to be conceptualized as figure vis-a-vis the event.10 When discussing motion predicates, we will use the term trajector (tr) to refer to the entity that merely executes movement and agentive trajector (ag tr) to identify the one that also, or only, causes it. The entity that is used by another to cause motion will be called instrumental trajector (instr tr). Regarding the ground, we will use the term primary landmark or just landmark (lm) to refer to its most salient constituent. In the case of a motion event per se, the entity against which the moving object's successive locations are evaluated is such a constituent. In the case of a motion event becoming part of a causative event, the moving object emerges as the most important element of the ground, against which the causative relationship is assessed. The next problem that interests us here is how the trajector/landmark distinction relates to clause structure. In keeping with Langacker's claim (1979, 1983, and this volume) that within every profiled relationship one entity is selected as figure against some relevant ground, this distinction can be said to pertain as much to the clause level as to the level of prepositional
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 519 phrases or that of adjectival or adverbial modifiers. At the clause level, Langacker equates what he identifies as trajector and primary landmark with what have traditionally been specified as clausal subject and direct object. While following this lead, we will add information on the causative, non-causative, or instrumental role of the entity functioning as trajector. Sentences (a)-(g) illustrate several manifestations of such roles, all relevant to the discussion of metaphoric transfers. Note that interrupted arrows are here used to indicate covert, non-lexicalized elements. ag tr <—j I > lm (a) John flung a stone at the boys. tr< 1' >lm ag tr< 1 I >lm (b) John threw it out. 0 tr<—' L_-^lm ag tr <-i . ->lm (c) He spat 0 at them. tr*—'I >lm ag tr« 1 | >lm (d) Finally the baby spat the button out. tr< " I lm< I ag tv<—j |—> lm (e) John hit Andrew with a stick. ' -I I Mnstrtr ag tr« 1 r > lm (f) Why are they running? 0 tr< 1 |—»lm (g) A brick hit me. In (a)-(g), the agentive trajector or just trajector, for lack of the former, coincides with the clausal subject whereas the direct object assumes the role of landmark at the clause level. Whenever another relational predication combines with the finite verb, this landmark may emerge as trajector within the profile of that predication. And if this verb designates movement, the
520 BRYGIDA RUDZKA-OSTYN dynamic element permeates the whole composite structure. Compare at this point examples (a)-(d). In isolation, the preposition at, for instance, profiles a static relationship highlighting close proximity or contact between two entities, one of which assumes the status of figure. However, when combined with a motion verb, at loses its static character and becomes a means of marking, among other things, the figure's path vis-a-vis some landmark. Similar dynamization affects out and any other particle or preposition. While in (a) the landmark of at is the entity designated by the boys, in (b) the landmark of out remains covert (hence 0), and as such is assumed to be any location within the reach of the agentive trajector.11 This lack of overt expression does not thus mean that the landmark is absent from our conceptualization of the scene; it simply means that it is either too obvious or rrot worth bothering about. Covert entities are nevertheless interesting because of their role in metaphoric transfers (see next section). To pass now to example (c), the situation here is somewhat different. The prototypical sense of spit includes the concept of sending saliva out of one's mouth, so no overt reference to this substance is necessary and, as a result, no clausal landmark (direct object) appears in (c). But when, as in (d), spit is used in the more general sense of "send out of one's mouth," the clausal landmark needs to be marked overtly. As in the previous cases, this landmark assumes the role of trajector within the relation profiled by out. At the level of this relation, the landmark constitutes part of the agentive trajector. Synecdoche is at work here since it was not the whole baby but only its mouth where the button was located. The use of with in (e) introduces an instrumental trajector. Unlike out or at, with seems to profile a relationship whose landmark always coincides with the clausal landmark, and the nature of the linkage between the instrumental trajector and its landmark is specified by the finite verb. Still different construals are expressed in (f) and (g), a phenomenon which is again tied up with what the speaker considers relevant and worth mentioning in the given context. As the focus in (f) is on the purpose of an agent's activity, and not on the ground element — which in all probability is known — only the agentive trajector is lexicalized and the landmark remains covert. In the situation conveyed by (g), the motion event has been dissociated from any causal factors, and as a consequence the trajector can assume the role of clausal subject. Brief as they are, these observations reveal that in its very essence clause structure reduces to one or more figure/ground relationships. It is worth noting that the approach adopted allows us to describe not only each
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 521 of the relationships in isolation but also the way they interact with one another. As examples (a)-(d) make clear, the profile of the verbal predicate "supplies" the trajector for the profile of the particle/preposition whereas the latter makes explicit, among other things, this trajector's path. In a word, the two types of profile complement each other in their linkage to the structure of motion events. A closer look at the meanings of the lexical items brought together in (a)-(g) indicates that this structure lends itself to a number of different instantiations. In fact, any component of a motion event can be elaborated against a whole range of domains and dimensions, often intersecting. Such elaborations hold an enormous potential for metaphoric transfers. The next section will be devoted to probing this potential. 5. Extensions of Spatial Motion 5.1. Metaphor vs. Metonymy Underlying the present analysis is the notion of metaphor as a way of understanding one concept in terms of another. Following Lakoff and Johnson (1980), we assume that metaphor is de facto a specific conceptual principle. By metonymy we will understand any association based on contiguity,12 with contiguity covering the relationships holding between objects and their locations, producers and products (e.g. speakers and their utterances), causes and effects, etc., but also between parts and wholes as well as wholes and parts. Metonymy will thus include what has traditionally been identified as "synecdoche."13 It is important to distinguish between metaphor and metonymy as conceptual principles and the linguistic expressions that instantiate them. In this paper, the terms metaphor and metonymy will be used to refer to both the principles and their instantiations; it is hoped, however, that context will make the distinction clear. As pointed out by several students of metonymy and metaphor,14 the former involves conceptualizations confined to one and the same domain or, more precisely, to entities that become part of a higher-order domain by virtue of appearing together. Metaphor, on the contrary, involves crossing domain boundaries. An essential element of its mechanism is the establishment of correspondences between two or more domains. To see how this can be done, we will first examine metaphoric expressions containing verbs of movement accompanied by particles or prepositional phrases, hence-
522 BRYGIDA RUDZKA-OSTYN forth called "phrasal verbs." 5.2. Phrasal Verbs The expressions assembled can be subclassified along different lines depending on whether the meaning of the verb or that of the particle/preposition becomes the main criterion. In each case several intersecting groupings emerge; it appears, however, that a more economical classification is the one that starts with particles and prepositions. Given the fact that their profiles overlap in many respects, particles and prepositions will be discussed jointly.15 We will confine ourselves only to those properties of the profiled elements that are directly involved in the metaphoric mappings demonstrated by the database.16 It has been observed that most English prepositions can be categorized according to the type of landmark they profile. Hawkins (1984) identifies the types as MEDIUM, SURFACE, CHANNEL, and INDETERMINATE. An insightful discussion of MEDIUM is presented in Hawkins (this volume). We will extend this categorization to include particles but use the term point to refer to INDETERMINATE landmarks as it seems to capture better the basis for some metaphoric transfers. In the case of MEDIUM, Hawkins (this volume) finds "three-dimensional extension" essential for association to the center of this category. With regard to the more peripheral associations to MEDIUM, the state of boundedness is found to be a criterial concept. In the case of the categories SURFACE, CHANNEL, and POINT (= NODE in Hawkins, this volume), Hawkins proposes "two-dimensional," "one-dimensional" and "zero-dimensional extension" as their respective central associative principles (idem). It will soon become apparent that often aspects other than the landmark's dimensionality are at the base of metaphoric transfers into speech acts. In the discussion that follows, the properties of the given landmark that do not participate in cross-domain correspondences will be disregarded. 5.2.1. Including a MEDIUM Landmark Over one-third of all the particles and prepositions accompanying verbs of movement furnished by the database code a landmark that can be associated to MEDIUM. These are primarily out, into and m, and to a much lesser degree through. While in the case of the former the landmark
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 523 manifests itself usually as some sort of enclosure or container, through seems to highlight in the role of landmark an entity that surrounds the trajector from different sides and as such can hamper its movement. Combinations with out commonly give rise to the following extensions into the domain of speech acts: (1) She was so shocked that she could hardly bring out a word. (DPV: 53) (2) At last the prisoner faltered out his story. (DPV: 179) (3) In spite of his fear, he jerked out his request. (DPV: 308) (4) Mary flung out a curt remark and left. Since the mouth is a container almost par excellence, the landmark associated with out provides here a bridge between the spatial domain and verbal communication. But, as is often the case, this bridge becomes accessible only indirectly, via the agentive trajectors. At the clause level the agentive trajectors (subjects) of (l)-(4) correspond to the primary speakers; however, at the level of the relationships profiled by out, they assume — by metonymy — the role of landmarks. In addition to selecting a particular type of landmark, the profile of out also specifies the changing orientation of the trajector with respect to the landmark. Prototypically, the trajector's path begins inside the landmark and terminates outside its boundaries. And again an analogy with speech events shows up. Observe that whatever is to be communicated changes its relation vis-a-vis the speaker by the mere fact of being uttered; physical motion and speaking can thus be correlated in a straightforward manner. Likewise, correspondences can be drawn between a particular speech unit, i.e. the utterance, and the spatial trajector of out, i.e. the moving object. Such correspondences are established in almost 80% of all the phrasal verbs with out provided by the corpus. While the particle out introduces into the communicative domain a particular trajector/landmark orientation, the verbs themselves bring in several parameters which are directly relevant to descriptions of speech acts in that they allow the reporter to focus on the manner in which the primary utterance is produced. It has been signalled in section 4.2 that any component of a motion event can be characterized against a whole array of domains or dimensions and thus become an object of different acts of comparison. For example, physical motion per se is commonly classified along scales pertaining to the effort required to set an object in motion, the speed of movement, its continuity, or the force exerted on the moving object. As
524 BRYGIDA RUDZKA-OSTYN is obvious, mode of execution or, simply, manner is an important dimension in the perception of not only physical movement but also speaking. This being the case, attributes of motion can be ascribed to the way in which a speaker delivers his utterance. By analogy with the agent that controls the movement of an object by throwing, pushing or pulling it in a certain way, the speaker is conceptualized as determining the 'flow' of his words. To return to examples (l)-(4), if he hesitates, falter will provide the right descriptive features, if he speaks with difficulty, alternating bursts and pauses, jerk will be appropriate, and if he is quick and impulsive, fling can be used. As both physical motion and speaking require time, the transfer of manner attributes from the former to the latter is greatly facilitated by temporal correspondences. Another common ground that provides a bridge across these domains is the purpose that they may serve. Speech acts, like motion events, are often means of indicating emotional states of the participants. For instance, aggression and anger commonly find an outlet in sudden, forceful movements. In a verbal exchange, the same attitudes can be expressed through acoustic elements, e.g. by raising one's voice, or the choice of particular lexical items. Correspondences may therefore be established between the latter and a violent, forceful movement. The use of fling in (4) above provides an illustration. For other examples, see the subsection on at. Sometimes it is not the nature of the movement as such that determines transfers into the domain of speech, but the value judgment associated with the given motion event. It will become apparent on several occasions that value judgments are important factors in motivating as well as constraining cross-domain mappings. Consider, for instance, the case of spitting and vomiting. In our part of the world, both activities are negatively charged, and, when verbs denoting them come to replace verbs of speaking, it is required that in the recipient domain these verbs combine with objects that also carry negative value judgments. While the collocations exemplified in (5)-(6) are perfectly legitimate, those given in (7)-(8) are not, unless the context is very much out of the ordinary. In any event, it is significant that the corpus does not provide examples like (7)-(8). (5) "Pig!" she spat out at him viciously... (C-22: 162) (6) There he stands, vomiting out lie after lie. (DPV: 704) ?(7) spit out ?(8) vomit out \ compliments, endearments, words of admiration
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 525 Regarding the question of why a reporter should use all these metaphors, observe that they offer a descriptive potential that an ordinary VS does not have. Assuming that in the case of (6) the reporter chose the verb tell, he could practically focus only on the content of the primary utterance; by choosing vomit he can not only label what was said but also convey the primary speaker's attitude as well as his own disgust at it. In short, by using a metaphoric expression he can both describe and judge, and since the description draws on another, more concrete, domain, the image created also gains in concreteness. Although less apparent so far, equally important is the semantic contribution of out. We have already pointed out that this particle profiles a very specific trajector-landmark orientation: at the initial stage the trajector is enclosed in the landmark, and leaving this landmark necessarily involves crossing its boundaries, actual or potential. The boundaries themselves can be viewed as an obstacle or something very useful, something that prevents an undesirable spread or spill of the enclosed trajector. The various assessments of a physical boundary may carry over to the domain of communication, allowing us to look at speaking as an act requiring a special effort or as an undesirable or painful transgression of the bounds of fear, politeness, discretion, or diplomacy. Compare in this respect examples (l)-(3) with (9)- (10) below: (9) Jane let out where she had hidden her father's birthday present. (DPV: 354) (10) I didn't mean to tell you his name, it just slipped out. (DPV: 582) As already mentioned, the notion of a boundary is here reinforced via references, albeit indirect, to the most conspicuous speech organ. If, however, man has the ability to conceive of all countable things, whether concrete or abstract, as bounded areas in their proper domain, then we should expect any entity that makes up the domain of verbal communication to be conceptualized as some sort of container. The database suggests that this indeed is the case. The boundaries evoked by the particle out may refer not only to the speaker's mouth but also to periods of silence, speech as such, or any state in which the speaker may find himself. (Very similar observations are made in Lindner (1981: ch. 2); cf. also Lakoff & Johnson (1980: ch. 16).) The remaining 20% of the phrasal verbs with out express such abstract boundaries. For instance in (11) below, it is the speaker that is said to have come out, not his words, and as a result the notion of container
526 BRYGIDA RUDZKA-OSTYN evoked by out can be associated with the state of silence which he broke when voicing his opinion.17 (11) The speaker came out in support of the government. Other associations appear to be at work in the following example taken from Newsweek's coverage of the 1984 presidential election: (12) He was best off, as one argued, ladling out the oatmeal — his mythic America of fluttering flags, mom's apple pie and the squirrel tail hanging from the rear-view mirror. (NW, N/D, 1984: 88) The spatial landmark here is some big food container and the trajector is the oatmeal. As the trajectors profiled by out almost invariably correspond to the speaker's utterance, the oatmeal can be easily related to what the President said, and the kitchen utensil (the instrumental trajector) evoked by the verb ladle to the linguistic expressions he used. The fact that this utensil is not a coffee-spoon but a ladle undoubtedly helps capture the gusto with which the addresses were given. By presenting them in terms of food distribution, the reporter creates a humorous effect bordering on the grotesque. In turn, the symbolic specifications of the "oatmeal's" components allow him to render in a very succinct way the mixture of grandiose patriotic aspirations, homeliness, and sentimentality expressed in some of the President's speeches. It is thanks to these different semantic transfers that the reporter can combine a vivid description with an element of humor and criticism — and all this in one sentence. This example also shows how important the nature of the trajector is. When the trajector is some amorphous stuff able to form a homogeneous flow, it can be correlated with utterances where continuity or monotony comes tc the foreground. A case in point is (13): (13) He poured out his tale of misfortunes. (AED: 663) Pour typically profiles a trajector that is inherently unbounded and can thus form a continuous stream, easily to be correlated with a tale, a type of narrative stressing duration and thematic uniformity. Observe that the analogy would immediately break down if tale in (13) were replaced with word or another item denoting a short unit of speech. While phrasal verbs with out are used to refer primarily to the speaker and conditions related to him, verbs which combine with the preposition into focus on the hearer, his or a third party's position, or the verbal
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 527 exchange as such. Why this should be so is determined by the meaning of into and the very structure of speech acts. The spatial configuration profiled by into is the reverse of the one coded by out: the trajector does not leave but enters the landmark. The landmark itself, however, is once more a container, at least in the most typical cases. Since the hearer is the recipient and not the originator of the utterance, and the most salient hearing organ is a container, the use of into is fully justified. But also our ability to conceptualize the hearer and any of his parts as a bounded region in some domain comes into play. If his cognitive apparatus can be viewed as a bounded area in the domain of cognition, then whatever is to reach this apparatus becomes analogous to a concrete object entering a real container. Such an analogy is drawn in examples (14)-(15): (14) That teacher is wasting his efforts trying to hammer useless facts into such children. (DPV: 272) (15) Parents often try to drill into their children how much they owe them. (DPV: 151) The choice of verbs denoting forceful, repetitive motion is not accidental either. We have already observed that translating physical force into communicative terms involves an intensity scale which may operate on the phonic level or the expressive power of the words chosen. It may also operate on various rhetorical devices, repetition being one of them. In the cases under consideration, all three factors might be at work. Again, ordinary VSs like teach, tell or instruct do not allow the reporter to convey also the manner in which the teaching was done. Moreover, the fact that the landmark profiled by hammer and drill is an object with a hard surface, difficult to penetrate, enables the reporter to focus on the hearer's reduced recep- tiveness to what is being said. These and similar metaphors pertain primarily to the context of teaching, a domain in which the difficulty of imparting knowledge is felt most acutely. Whereas in examples (14)-(15), the trajector profiled by into can be correlated with the primary speaker's utterance, the situation changes when the verb of movement is intransitive, as in (16): (16) The director laid into Bob for not working hard enough. Here the director corresponds to the trajector of into and Bob to the landmark, at least overtly. Underlying this metaphor is a metonymy by which Bob's name comes to denote his autonomous state, his sense of integrity and self-respect. Analogously to other entities, this sense of autonomy is
528 BRYGIDA RUDZKA-OSTYN conceptualized as a bounded area. And where there are boundaries, there is also the possibility of transgression. In this particular instance, a parallel can be drawn between transgressing the boundaries of one's autonomy or integrity and an act of scolding. But why scolding and not, for example, praising? When compared with other speech acts, scolding normally implies that the speaker is more powerful than the hearer and thus enjoys the right of imposing his interpretation of facts on the hearer. Scolding, like invading someone's territory, is an act of imposition, and this is what appears to motivate the metaphoric extension in (16). Another motivating factor is, of course, the affinity between speech and physical motion as pointed out on earlier occasions. The same mechanism can be at work when the landmark whose boundaries are violated is not the hearer but any third party. In (17) below, the landmark of into corresponds to the government whose members need not be actually present at the given session. (17) The opposition speaker made a lively speech, lashing into the government for its recent inaction. (DPV: 334) The focus of the reporter's attention may, moreover, be on the effect of the primary speaker's utterance. Any emotional or intellectual state created by this effect can also be conceptualized as a bounded area — e.g. in the domain of emotional/intellectual experiences — and thus be dissociated from the experiencing subject. Example (18) is a good illustration; the crowd corresponds here to the trajector profiled by into whereas the state of excitement brought about by the words uttered functions as the landmark. (18) The speaker soon whipped the crowd into a fever of excitement. (DPV: 717) About one-third of the combinations with into perspectivize on the hearer or a third party and on conditions related to them. The remaining two-thirds center on the verbal exchange between the speech participants. This exchange is then viewed as an integral whole bounded in time, i.e. set off from periods of silence, which a fortiori also emerge as bounded entities. Hence, beginning to speak or interrupting an ongoing flow of conversation presents itself quite naturally as a transgression of boundaries. Here are just two of the many examples found: (19) She sat silently, tossing the odd word into the conversation from time to time. (DPV: 683)
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 529 (20) Mary again broke/barged into the conversation. Similar mechanisms seem to govern the use of in when accompanied by a motion verb, as in: (21) "What about the parades?" Colonel Scheisskopf broke in. (C-22: 330) (22) "But I was there!" Jane burst in. (DPV: 60) (23) Can I bung in a word yet? (DPV: 59) Characteristically, the landmark of in, unlike that of into, is not lexicalized, which seems to relate to what Langacker (1985) identifies as a subjective construal. The role of particles in promoting this type of construal certainly deserves further attention. Another area worth examining is the kind of enunciation allowed by the given construal. At this point we can only note that phrasal verbs with in, but not into, combine freely with direct speech (see (21) and (22)). To return to the different uses of into, the database further demonstrates that not only the verbal exchange as a whole but also any of its elements or aspects can be perceived as a bounded region, bounded by its acoustic or graphic form, by its internal logic or theme, or by its conformity to various linguistic and other conventions. Good examples are provided in (24)-(29). (24) There's no need to go into details yet; just give me the general idea. (DPV: 250) (25) As usual, he retreated into circular arguments. (26) ... he wandered instead into a random line-item defense of his record. (NW, N/D, 1984: 90) (27) The speaker really waded into the government's lack of action on housing, didn't he! (DPV: 705) (28) He dropped!fell into a country form of speech when talking to his mother on the telephone. (DPV: 155/170) (29) Once the English course was over, they all lapsed into their native languages. Again, the choice of a particular motion verb permits stressing the mode of speaking or the attitude of the hearer/reporter. A number of correspondences are established between the parameters operating in the donor and recipient domains. These correlations are in their turn evaluated along a scale of desirability/undesirability, politeness/impoliteness and the like.
530 BRYGIDA RUDZKA-OSTYN To go back to examples (17)-(22), lashing, whipping, barging, flinging, breaking, or bursting all involve the use of force, which can be correlated with a high degree of eagernes or determination on the part of the speaker as well as abruptness of speech. Moreover, breaking and bursting imply damage, not respecting the form which bounds a given object. Now, interrupting the flow of conversation also involves a violation of its boundaries, so an analogy with physical damage is easily established. Whether material or not, an act of violation is often perceived as something undesirable. This element of undesirability is another common ground that facilitates some of the extensions discussed above. It is interesting to note the different means that English exploits to indicate negative value judgments. One of them is undoubtedly a downward trajectory, as associated with the verbs fall, drop, and lapse. Why this should be so becomes clear when we examine the prepositional objects of examples (28)-(29). In many cultures non-standard variants of a language are stigmatized forms of expression. The same low degree of esteem is frequently accorded to the native languages of minority groups or participants in a foreign-language course. The reporter who uttered (28)-(29) can express this low esteem by presenting the primary speakers' shift from one language (variant) to another as downward movement. Although for some English speakers it is the ease of execution that is primarily associated with the downward trajectory in (28) and (29) (Peter Kelly, p.c), many agree that in this context downward motion may express disapproval, however attenuated and camouflaged. The importance of a downward trajectory in making negative value judgments will become even more evident when verbs combining with the particle down are discussed. Passing now to the preposition through, we notice that in the few examples provided by the database, its landmark corresponds invariably to the primary speaker's utterance, as in: (30) ... the Reverend Frank Barber raced through the grace ... (C & M: 3) (31) The politician floundered through his speech, making mistakes ... (DPV: 198) (32) Reagan stumbled through a first question on ... (NW, N/D, 1984: 93) Within the domain of space, the landmark profiled by through is not so much a container-object as some sort of bounded substance or area which
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 531 the trajector is supposed to traverse, usually from end to end or side to side. The verbs combining with through make of course their own semantic contribution and may highlight different aspects of the concepts associated with this preposition. The use of race brings out the agentive trajector's desire to reach the opposite end of the landmark and stresses the speed of motion, which the reporter of (30) related graciously to the Reverend's manner of praying. The choice of flounder or stumble focuses more on the landmark as an obstacle hampering motion. The spatial sense of flounder, for instance, evokes a scene in which the agentive trajector's progress is hindered by the resistance of the landmark, usually deep snow or mud. Difficulty and impediment emerge as the main characteristics of this motion event. But these are also attributes applicable to speech acts, e.g. in situations when the speaker cannot find the right word or simply does not know what to say. The linguistic units or information that he is unable to retrieve assume then the function of a landmark impeding his communicative fluency. As in the previous cases, the choice of a motion verb plus an appropriate preposition allows the reporter to render such complex events in a very economical way. 5.2.2. Including a SURFACE Landmark The next most common grouping of phrasal verbs, which accounts for 22% of all the relevant data, involves particles and prepositions whose spatial landmark is a surface. Of these the most frequently used is on. In the overriding majority of cases, the landmark coded by on corresponds to the communicative flow or to any specific element of the subject matter raised as well as any point of view expressed. In the remaining instances, the landmark corresponds to the hearer (see (37)). (33) Mondale got stuck on the budget issue. (NW, N/D, 1984: 90) (34) ... the other side was banging hard on the age issue ... (NW, N/ D, 1984: 92) (35) "I give a great deal of it to Dunbar," Yossarian went on. (C-22: 63) (36) She rambled on for hours. (37) Mother landed/jumped on Bob for being rude to Mrs. Tomkins. In typical situations, the trajector profiled by on comes into direct contact with the landmark, either by moving along its surface or by being
532 BRYGIDA RUDZKA-OSTYN placed on top of the landmark. If the surface poses too much resistance to the moving object, all activity ceases. So an analogy can be drawn between such a landmark and an issue which is too difficult for the speaker to handle and which thus prevents further communication. A good example can be found in (33). As in so many other cases, attributes of physical motion are transferred to speech. While the use of bang in (34) brings out its repetitive and forceful character, the choice of ramble in (36) allows the reporter to characterize the speaker's performance as incoherent and without a clear purpose. The landmarks profiled by on in (35) and (36) remain covert, and as such can be related to the subject matter, i.e. the ground "covered" in a speaking activity. Like a moving object which covers more and more surface, continued speech "covers" successively more time and subject matter. This analogy explains why on can be used to mark mere continuation of speech. By combining on with land/jump, as in (37), one can create yet another image and highlight the top location of the trajector vis-a-vis its landmark. Since the landmark corresponds here to the hearer, it is easy to conceptualize him as an object of some act of imposition or suppression. English links such conceptualization with acts of scolding. Another SURFACE particle that participates in interesting metaphoric extensions is off. Unlike on, it stresses trajector-landmark detachment and often profiles trajectors that constitute part of the given landmark (cf. cut off a branch, chop off the head). When extended into the domain of speaking, these trajectors/landmarks correlate primarily with the utterance or its parts. The part that precedes another is normally viewed as the landmark from which the latter, i.e. the trajector, gets detached. The landmark may also correspond to the entire, potential utterance as intended by the speaker, and then the trajector becomes that part which the speaker managed to produce before being interrupted. The act of interrupting an utterance is thus conceptualized as analogous to physical detachment of a part from a whole. Consider in this respect the following examples: (38) And he tossed off one stinging statistic: the 21.5% prime interest rate in the U.S. under the Carter Administration. (TM, 0 29, 1984: 7) (39) But the quarrel ran inconclusively on until Mondale cut it off. (NW, N/D, 1984: 86)
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 533 (40) You're wandering off the point again, so please keep to the subject. (DPV: 710) (41) Reagan smiled ... when Moderator Edwin Newman chopped him off... (TM, 0 29, 1984:7) In (38), the trajector of off corresponds to the stinging statistic whereas the landmark relates to whatever the speaker (here President Reagan) said earlier. And as the spatial trajector of toss is usually a light object, the choice of this particular verb may also imply some casualness, if not downright 'Schadenfreude', on the part of the President when he was bringing this statistic to his opponent's attention. In (39), the extended landmark is the quarrel as it could have developed and the trajector is what was actually said before Mondale's intervention. Different correspondences obtain in (40) where the point expressed serves as landmark and the speaker's actual statements correlate with the trajector. By metonymy, the role of this trajector is overtly assigned to the speaker himself. A similar metonymic conflation takes place in (41) between Reagan and his utterance. It might be interesting at this point to compare briefly off with in/into (hence in(to)). Both can be used to mark speech interruption, but each brings a different element into focus. While off emphasizes trajector-land- mark detachment, in(to) typically stresses a relationship of inclusion. Moreover, as off can code trajectors which constitute part of their landmarks, it allows us to relate particular utterances to the ongoing flow of conversation as a whole and to highlight the disruption of that flow. In the situations reported in (39) and (41), one speaker says something in order to prevent another one from continuing. This is not the case when in(to) is used. Unlike off] in(to) normally extends to contexts where a speaker intends to add something to the ongoing verbal exchange. As examples (19)-(23) indicate, the focus is on his desire to voice his opinion and not to stop the exchange. Apart from on and off, over is another SURFACE predicate that occurs not infrequently in reports of speech acts. One version of this predicate codes vertically oriented landmarks and trajectories leading from one side of the landmarks to the other (as in e.g. climb over a wall). Because such landmarks can be seen as obstacles to the moving object, they are commonly extended to ideas or forms of expression which pose difficulty for the speaker. Appropriate illustrations are provided in (42) and (43). (42) John's stumbling over every other word got on my nerves.
534 BRYGIDA RUDZKA-OSTYN (43) The speaker answered most questions well, except one that he tripped over because he didn't have the necessary facts. (DPV: 691) Some uses of over disregard the vertical position of the landmark and highlight the fact that the trajectory spans locations on opposite sides of the landmark, whatever its orientation vis-a-vis the moving object. Given our ability to conceptualize any piece of discourse as a bounded area, the opposite sides of the landmark can easily be associated with the beginning and the end of that piece. Such an association is at work in (44). (44) The police asked the woman to go over her story once more. 5.2.3. Including a POINT Landmark Nearly 21% of the phrasal verbs under discussion include a particle/ preposition whose landmark is some point rather than a dimensional entity. One of the most common of the prepositions used is at. Its landmark can correspond to the hearer, a third party, or anything that causes discontent or controversy. The following illustrations are quite representative: (45) ... Glenn and Mondale were throwing words like "baloney" and "gobbledygook" at one another ... (NW, N/D, 1984: 36) (46) Mondale flung the magic words "social security" at him ... (NW, N/D, 1984: 90) (47) The committee kept firing questions at her. (48) They hurled insults at him. (49) Mondale hammered at Reagan as the first President since Hoover not to have met with a Soviet leader. (TM, N 12, 1984: 26) (50) Reagan ... lashed out at the "evil empire" ... (TM, N 12, 1984: 26) By focusing on a point rather than a surface or boundaries, the preposition at is particularly well suited to refer to situations in which the landmark becomes a target, i.e. the end-point of the trajectory followed by a moving object, usually an opponent or his weapons. Once an analogy between objects in motion and utterances has been established, it is easy to correlate physical targets with entities subjected to criticism, reprimand or questions. Such correlations have in fact given rise to (45)-(50). As elsewhere, the reference to the utterance itself may be explicit (as in (45)-
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 535 (48)) or implicit ((49)-(50)), and the attributes of motion like suddenness and violence may be correlated with forceful speech. Another preposition commonly used to refer to speech acts is to. Unlike at, however, in some contexts it seems to lose its metaphoric ring. A possible explanation of this impression may be the fact that at often profiles a landmark which is just an object to be used, hit, or destroyed, and not so much a recipient or partner. Quite revealing in this respect is the following example: (51) I like her because she talks to you and not at you. By contrast with at, to either neutrally points to some goal or connotes a recipient-landmark. Since speech acts include directionality and the notion of goal as their inherent components, to is expected to appear in this domain, particularly when the goal coincides with the hearer. As a result, the metaphoric load in (52) is felt to depend on the verb drift but hardly on to. (52) ... word drifted back to Caddell from the Mondale tour that Reilly was calling him crazy ... (NW, N/D, 1984: 84) The impression changes, however, when the landmark profiled by to corresponds to the utterance or some of its parts, as in: (53) ... he returned to the subject a half-dozen times ... (NW, N/D, 1984: 90) (54) Please come to the point. In both cases, spatial proximity between the trajector and landmark is evoked to express thematic linkage between one item of information and another. The trajector actually corresponds to a portion of the speaker's utterance but by metonymy becomes associated with the speaker himself. While at and to profile trajector-landmark proximity or contact, away expresses their separation. It is like out in that its landmark corresponds in most cases to the speaker, and the correspondence is established by metonymy. Yet unlike out, away does not emphasize the container-like nature of the landmark but rather its role as the initial, and often proper, location for the trajector. The latter usually corresponds to an element of the utterance. Focus on the initial, or proper, location of the trajector allows the reporter to use away in contexts where the beginning of speech is to be marked (see (55)), or where the utterance constitutes an undesirable disclosure of information (56) or an act of articulatory carelessness (57).
536 BRYGIDA RUDZKA-OSTYN (55) Fire away, we're all listening. (DPV: 189) (56) The thieves promised each other not to give away (to the police) where the jewels were hidden. (DPV: 233) (57) You chucked away that last line. (DPV: 78) Among the factors relevant to cross-domain extensions is also the shape of the trajectory. When the trajectory is winding or circular and thus requires more time and effort to cover than a straight one, reference to it may imply the trajector's difficulty in reaching the landmark-destination. This difficulty can be correlated with the difficulty a speaker experiences in expressing himself. Combinations with around provide interesting illustrations. Here is one of them: (58) Your question seems to have unnerved the speaker; he's been floundering around ever since you spoke. (DPV: 198) Focus on the speaker's poor performance is here additionally reinforced by combining around with flounder. Combinations with other verbs may bring out very different aspects of the given speech act. In the context of kick, as in: (59) The committee kicked your suggestion around for over an hour before deciding that they could not accept it. (DPV: 321) the trajector's path is conceptualized as curving with respect to several agentive trajectors functioning also as presupposed landmarks. By being kicked the trajector constantly changes its orientation vis-a-vis these landmarks and exposes different parts of its surface to view. In accordance with the structure of speech acts, the spatial landmarks can be related to the members of the committee and the trajector to the suggestion. This being the case, the different parts of the trajector may be correlated with the different aspects of the suggestion considered by the committee. Such many- sided considerations take time, which in turn explains the co-occurrence of around with adverbials stressing duration. 5.2.4. Including a CHANNEL Landmark Over 10% of the verbs of motion provided by the database combine with up and down. For these two particles the basic domain is vertically oriented space. In his discussion of their prepositional counterparts, Hawkins (1984: 165ff.) identifies the landmarks profiled by them as CHANNEL. The trajector's path coincides in each case with the CHANNEL and
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 537 the vertical axis with which the CHANNEL is in correspondence. While the path profiled by up points to the positive pole of this axis, the one profiled by down points to the negative pole. The latter is equated with the source of gravity (op.cit.: 149). As Hawkins further observes, "oriented physical space is necessarily oriented relative to something in physical space" (op.cit.: 142), and polarity, whether positive or negative, can be defined also relative to this thing. In other words, the vertical axis can be divided into an upward and downward part only relative to some entity or location which serves as a point of reference. As the phenomenon of polarization manifests itself not only in physical space but also in many other domains, including that of verbal communication, it certainly merits closer scrutiny.. In fact, we can see spatial polarization as a special case of much more general and abstract antinomic relationships. It is special in that it is the most concrete, most perceptible, expression of polarization. Thanks to this perceptibility, it can be used to represent all sorts of abstract antinomies. Once spatial polarization along the vertical axis is correlated with polarization along an abstract scale, correspondences can be established between different portions of the axis and that scale. Furthermore, the location or entity relative to which upward or downward movement is measured can be correlated with any norm with respect to which one calculates increase or decrease in value. Before we see how such correlations affect speech acts, another look at the spatial meanings of the particles up and down is in order (see Lindner (1981) for insightful discussion).18 It appears that the CHANNEL landmark is also part of their profiles, just as in the case of the corresponding prepositions. For lack of lexicalization, this landmark can in fact be identified with the vertical axis. But there is also a secondary landmark which may correspond either to one of the end-points of the axis or to any location along the axis that is higher/lower than a given point of reference. When these particles co-occur with destination-oriented verbs like come or bring, it is the end-points that seem to be invoked and put in correspondence with the location of the speaker or hearer. In appropriate contexts, come down/up and bring it down/up express trajectories that terminate not just anywhere but at the place or level where the speaker/hearer finds himself at a given moment (Fillmore 1971a). Likewise, when the particles combine with verbs like go, take, or get, the trajectory evoked terminates at the location of some goal, whatever it may be. These facts, as we will see, have some interesting consequences for meaning extension. Beginning with down, we notice that its trajectors correlate with the
538 BRYGIDA RUDZKA-OSTYN utterance and, by metonymy, with the hearer or the speaker. The points of reference relative to which transitions are calculated correlate either with the speaker's initial position in argument or the hearer's location' along the respectability scale. In both cases metonymy is at work. Consider as examples the following: (60) He put Mary down in front of her students. (61) John quickly climbed down when he heard her arguments. The analogy operating here involves two scales, one of them being the vertical axis along which the location of the given trajectors is assessed. The other pertains, at least in (60), to interpersonal relations, more specifically the importance the speaker is willing to attribute to the hearer and the appreciation or respect he is ready to show. As a low degree of respect shown is considered something negative, correspondences with a low location in oriented space are easily established. To take a closer look at (60), the trajector profiled by down overtly corresponds to Mary but by metonymy to Mary's worth and respectability. By reproaching her or making curt remarks, the speaker lowers the degree of respect she would like to enjoy. Expressions like brush/dust sb down (DPV: 56/157), cut sb down (DPV: 119) and others provided by the database carry similar derogatory overtones. In each case the person corresponding to the trajector undergoes some form of degradation, whether on the emotional or intellectual level. The latter may take place when he realizes the superiority of someone else's argumentative power and as a result gives in and abandons his position. Example (61) is relevant here. In both (60) and (61), an act of depreciation is put in correspondence with downward movement. A downward trajectory may lose somewhat its negative connotations when its end-point is part of the conceptualized scene. Such conceptualization is often activated by destination-oriented verbs and further reinforced by overt mention of the trajector's terminal landmark, e.g. via a nominal combined with the preposition to. In (62) and (63) below, the expressions specifics and main topic of interest code such a landmark. (62) His talk on Government-Binding Theory became clearer when he came down to specifics. (63) He finally got down to the main topic of interest. Like down, up brings into the conceptualization of speech events polarity scales. However, contrary to what might be expected, the connota-
SEMANTIC EXTENSIONS JNTO VERBAL COMMUNICATION 539 tions are not always positive. Over half of the relevant examples with up correlate the trajector with the utterance. The next most common correlations involve the hearer, and, by metonymy, his state. The landmarks, primary or secondary, remain covert. When up combines with destination-oriented verbs, usually no value judgments are expressed. Up simply indicates that the trajector (corresponding here to the utterance) has reached or is meant to reach the secondary landmark, corresponding in turn to the hearer and, by metonymy, to his awareness and receptivity. Notice that the latter need not be overtly mentioned. Since under normal circumstances the hearer is the receiver of the utterance, he is the only candidate with whom the trajectory's terminus can be correlated. By remaining covert, the terminus points not to any potential hearer but to the one that is directly involved in the speech act. Once again, lack of overt mention serves the purpose of subjective alignment. In view of this, the reporter can use up to focus on contexts where the given utterance is presented to the hearer for consideration. Sentences (64)-(66) exemplify such contexts. (64) Ailes ... had brought the whole thing up. (NW, N/D, 1984: 93) (65) Your question came up at the meeting. (DPV: 98) (66) I'll ask the next speaker to take up your suggestion. (DPV: 657) The information brought to the level of the hearer's awareness may be damaging, unpleasant or embarrassing, and in this case the terminus of the trajectory profiled by up acquires negative overtones. It becomes associated with a level at which not just any information but undesirable information is accessed. These associations are often reinforced by the meaning of the accompanying verb. For instance, in (67) below, the direct object of drag usually refers to a heavy thing, a thing that resists being moved easily. By correlating it with the content of the utterance, the latter can be presented as something that is not easy to disclose and as such would normally not be revealed. (67) They always drag up his war past. In the cases discussed above, the trajector profiled by up corresponds invariably to the primary utterance. It can also correspond to the hearer, and then up helps convey the effect the utterance had on him. For instance, when combined with the verb whip, it expresses a high level of excitement and enthusiasm (see (68)) and as such can be positively charged; in conjunction with chew, however, it reinforces the destructive aspect of the
540 BRYGIDA RUDZKA-OSTYN words uttered and qualifies the speech event as an act of scolding (69). (68) The speaker soon whipped the crowd up until they were ready to march. (DPV: 718) (69) The director chewed Jim up for being late again. (DPV: 76) This divergence stems from the different relations that up can bear to vertically oriented space. As pointed out at the beginning of this section, up can refer either to any location that is higher than a given point of reference or to the end-point of a vertical trajectory which can originate far below the point of reference. This referential ambivalence can be extended to other domains and correlate with either a greater or a limiting value along a given scale. In the domain of emotional experiences like those evoked in (68), up is available as a marker of hightened intensity. In the domain of destructive processes, it offers a means for expressing arrival at the point where destruction is effected. This means is used in (69), albeit to a peculiar end. Since the speech act is here reduced to a low-level physical activity and the destructive effect is clearly exaggerated, the metaphor allows the reporter to present the whole event with a degree of flippancy — and all this in a very condensed way. A similar condensation of attitudes and images can be found in most of the metaphors discussed. In (68), the meaning of whip brings a sense of force and energy into the description of the speaker's performance plus an image of a particular instrument with which his words can be associated. Such instrumental trajectors are quite common in extensions into speech acts. However, while in (68) and, earlier, in (12, 14, 15, 18 and 49) the instrumental trajectors are part of the profile of the verb, in the majority of cases it is the preposition with that codes them. The next section will present a few examples. 5.2.5. Including an Agentive Co-Trajector or an Instrumental Trajector One version of the preposition with resembles to in highlighting the presence of the hearer. However, in the case of to, the hearer corresponds merely to a landmark whereas with reinforces his role as agentive co-trajec- tor and companion in a speech event. For instance, in the situation described in (70), the delegate corresponds to the landmark as recipient of Jane's words, but as contributor to the conversation he rises to the status of agentive trajector.
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 541 (70) Although the delegate was in a hurry, Jane managed to exchange a few words with him. It should be noted here that the version of with which stresses companionship rarely combines with verbs of motion. In 95% of the examples found, it is the instrumental sense of with that comes to the foreground. Ideas, facts, lies, words of praise, or questions are conceptualized as instruments which the speaker uses with respect to the landmark, often correlated with the hearer. The choice of a particular verb enables the reporter to emphasize different aspects of this use. For instance, in (71) the image of repeated blows evoked by pummel allows him to bring out the intensity and harshness of Stockman's preparatory sessions. (71) ... Stockman ... pummeled him with facts, figures and accusations to a point where the president lost his patience ... (NW, N/ D, 1984: 88) Sometimes it is not the hearer that emerges as landmark but the utterance. In this case, particular expressions or ideas are conceptualized as noteworthy details added to the given piece of discourse to enhance its value, aesthetic or cognitive. Consider in this respect: (72) The politician's speech was interlarded with informal expressions that he hoped would attract the ordinary voter. (DPV: 304) (73) The witness embellished her story with dramatic episodes. Because the instrumental trajectors elaborated by interlard and embellish are distinct from the physical landmarks they are meant to enrich or adorn, they are readily extended into speech situations in which a contrast between the speaker's utterance as a whole and its parts is to be stressed. Focus on the contrast allows the reporter to question the desirability or truth-value of the added parts. Like an ornament that makes a physical landmark different from what it was originally, the dramatic episodes referred to in (73) introduce additional elements into what presumably happened and as such can be regarded as untrue. Again, borrowing a verb related to the physical domain allows the reporter not only to report but also to pass value judgments, however mild and disguised. As elsewhere, such borrowings depend critically on our recognition of the similitude between physical and more abstract manifestations of change as well as our ability to conceptualize entities as parts of larger wholes and to dissociate mentally the parts from the wholes.
542 BRYGIDA RUDZKA-OSTYN 5.3. Non-Phrasal Verbs When not accompanied by particles or prepositional phrases, verbs of movement focus mainly on the speaker-utterance relationship. One of the very few examples that make overt reference to the hearer is (74). (74) McPherson also lambasted Ethiopian President Mengistu Haile Mariam. (TM, N 12, 1984: 21) As in many previous cases, an analogy is here drawn between administering blows and expressing criticism, part of the common ground being the agen- tive trajector's intention to hurt or destroy the landmark. Again, the verb lambast has the advantage over many literal expressions in that it highlights the manner in which the speaker presented his criticism. Manner of speech is invariably brought to the foreground when the verb denotes discharge of a solid, liquid, or compressed air, as in (75)-(76): (75) ... John Glenn accused Mondale of spouting "the same vague gobbledygook of nothing" ... (TM, N 12, 1984: 26) (76) "That dirty goddam midget-assed, apple-cheeked, goggle-eyed, undersized, buck-toothed, grinning, crazy sonofabitchinbas- tard!" Yossarian sputtered. (C-22: 156) The trajector, i.e. what is discharged, corresponds in each case to the utterance. The speakers (Mondale and Yossarian) can in turn be correlated with the landmarks, i.e. the objects that contain the trajectors. In the context of speech events, these landmarks acquire also the role of agents causing the events and as such are elevated to the status of agentive trajectors.19 Notwithstanding the same pattern of correspondences, in each of the two cases a very different image emerges. Regarding the scene described in (75), as spouting typically involves force, shooting out jets of liquid, often upwards, it lends itself to being associated with a manner of speech that highlights noticeability. English exploits this association in contexts where pomposity and excess of rhetoric characterize the speaker's utterance. In the case of sputtering, as in (76), it is not a forceful stream that catches the eye but little drops and particles spat or thrown out in an explosive manner. Such an image fits a situation when the speaker is excited or angered, and as a consequence his utterance is more of a staccato than a continuous flow of words. This staccato quality comes out admirably in (76). Indeed, each of Yossarian's invectives can be correlated with a little particle released in an act of sputtering. Still different aspects of manner of speech are focused on in (77)-(78),
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 543 and again some particular feature of the given motion event motivates the transfer into communicative acts. (77) Snapped Mondale: "Well that's what a President is supposed to know." (TM, 0 29, 1984: 7) (78) He didn't mince his words. In (77), this crucial feature is the suddenness and sharpness of the breaking or snatching involved, whereas in (78) it is the rendering of the minced objects as more palatable, more refined. Lack of such refinement correlates straightforwardly with bluntness of speech. To pass to other types of mappings, observe that motion events may combine to form higher-level events with several trajectors interacting and assuming new roles. For instance, in an act of evading an agentive trajector tries to get out of the way of another moving entity which stands out as trajector within its own motion event but assumes the role of landmark visa-vis the agent that tries to avoid it. Like other languages, English exploits this interaction to highlight the way one speaker responds to the utterane of another. A good example is (79): (79) John evaded most of the questions raised by the chairman. Once we correlate John with the agentive trajector profiled by evade, the questions asked must be correlated with the trajector-landmark that the agent tries not to confront. And since evading typically implies reliance on some stratagem, further correspondences can be drawn with the speaker's attempt not to answer the questions fully or honestly. As elsewhere, the choice of a motion verb makes it possible to capture much more than any single VS can express. We have already observed that cross-domain transfers allow for stylistic variations, either in the direction of down-to-earth ordinariness or in the opposite direction, towards the more formal and elevated. Example (79) may again serve as an illustration. In English, questions can be "asked," but they can also be "put," "shot," or "raised." The verb raise profiles an upward trajectory which, as pointed out in section 5.2.4, can be associated with the speaker's conscious attempt to reach the level of the hearer's awareness. This association seems to add to raise a formal ring that the other verbs do not have. The style becomes even more formal when the borrowings from the domain of spatial motion involve verbs of Latin origin. Drawing on another domain and another language allows the reporter to depict more ritualized relationships. It is interesting to note that in the sav-
544 BRYGIDA RUDZKA-OSTYN age and unconventional world of Catch-22, utterances are "spat out," but in the context of, for example, formal debates or lectures, they are "delivered." Likewise, proposals are "dismissed" and arguments are "advanced." For lack of space, many details of the stylistic load of metaphoric expressions have had to be disregarded. It is clear, however, that cross-domain borrowings enable the reporter to adapt his language to the vagaries of the speaker-hearer relationship or, more precisely, to his perception of this relationship. 5.4. Major Patterns of Cross-Domain Correspondences We started to investigate the database by staking out areas of overlap in the structure of motion events and speech acts. It soon became evident that the distinct roles played by the speaker and the hearer in verbal communication together with the inherent linearity of speech parallel the relationships holding between a moving object, its landmark and the agent causing the motion. On the basis of these parallelisms, components of speech acts can be put in correspondence with those of motion events. We have examined such linkages as they are coded by particular groupings of phrasal and non-phrasal verbs. Upon closer scrutiny, it turns out that the groupings intersect in giving rise to several patterns of correspondences, each focusing on a different aspect of the speech act. When the focus is on the speaker as originator of the utterance, only predicates that profile trajector-landmark separation can be used. The trajector corresponds then to the utterance whereas the landmark can be related, by metonymy, to the speaker's mouth or his state of being the source, the 'locus' of the information imparted. Among the items that admit such correlations are verbs combined with the particles out or away, and non-phrasal verbs like belch, spout, erupt, or sputter. The ensuing pattern of correspondences can be represented as in Figure 3. MV stands here for a motion verb, and the interrupted lines indicate covert components of the given motion/speech events. To keep the representations as concise as possible, elements of conceptual structure are lumped together with linguistic units. As the identity of the spatial instantiations of agtr, tr and Im does not seem to pose major problems, references to them will be omitted from the remaining figures. Predicates stressing trajector-landmark separation can also be used when the reporter intends to emphasize the isolation of one part of the utterance from the rest, whether actually produced or potential. As Figure 4
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 545 agent causing motion location within or near agent or in a container related to agent T lm< moving object agtr< 1 (a) speaker MV speaker's mouth and/or his state of silence, knowledge, etc. t 1 out/away ->tr H>lm i utterance -*tr agent causing/ affecting motion starting point of motion t (semi-)agtr i (b) speaker speaker's mouth and/or state t llTK belch/erupt spout/sputter, etc. Figure 3 moving object r ►tr1 >lm i utterance t ► tr shows, the isolated part corresponds in such cases to the trajector and whatever precedes or is supposed to follow this part, to be identified here as "communicative flow," assumes the function of landmark (cf. he tossed off
546 BRYGIDA RUDZKA-OSTYN one stinging statistic). Characteristically, the landmark remains covert, but see example (40) above. agtr<- speaker MV -»lm utterance T tr « off communicative flow T ->lm L_ Figure 4 When predicates expressing trajector-landmark contact or proximity are used, a shift in perspectivization takes place. First of all, the hearer is brought into the picture. The utterance's impact on him as well as his position vis-a-vis the speaker may come to the foreground. The same possibility holds for any entity that qualifies as the goal of the given communicative act and as such can constitute the topic, the subject matter of the utterance. And when the speaker's role is highlighted, he is no longer conceptualized as the orginator or possessor of what he says. Rather, emphasis falls on what he does with a communicative unit or on how it affects him. This shift in perspectivization manifests itself in the case of such disparate items as lambast, pummel + with, bombard + with, and a whole range of verbs taking at, into, or on, all of which allow the reporter to focus on the hearer. Depending on the meaning of the particular item, different aspects of the hearer's role will be stressed. The constant element in his role appears to be his association with the landmark of a motion event. As the preceding sections abundantly illustrate, this landmark can also be correlated with any third party or any thing — including the utterance itself — that may become an object of the speaker's attention or a target of his verbal attacks. Whether the speaker "bombards the hearer with questions" or "interlards his utterance with informal expressions" the pattern underlying the correspondences drawn is the same. Consider in this respect Figure 5.20 Regarding the hearer, he may be conceptualized not only as a passive recipient of the utterance but also as an entity undergoing changes. These changes usually affect his opinion or mood and are brought about by the speaker's persuasive power, manipulative skill, enthusiasm, and the like. In this case, the hearer is invariably accorded the status of trajector as well, his
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 547 agtr<— i (a) speaker MV -^Im 4 utterance at/into tr<- hearer/ another entity t -Mm agtr< 1 i I (b) speaker MV tr^- intolonl out + at hearer/ another entity t »lm agtr< 1 1 > lm 4 II 4 (c) speaker pummel/ hearer/ with bombard/ another entity embellish, etc. utterance -»instr.tr agtr^- i (d) speaker lambast, etc. Figure 5 —»lm 4 hearer/another entity new state of mind or heart assuming the role of landmark. Figure 6 summarizes just two of the emerging correspondence patterns. In the case of 6(b), observe that the extent to which combinations with up or down emphasize trajector-landmark separation or contact depends very much on context and one's vantage point. For instance, in the situation described by example (68) above, the hearers' (tr) reaching the desired level of enthusiasm (lm) appears to receive much more attention than their 'separation' from their initial emotional state. Conversely, in (60) it is Mary's forced 'detachment' from her state of respectability that seems to be
548 BRYGIDA RUDZKA-OSTYN agtr< I (a) speaker MV -H>lm i hearer tr«- into a given state (e.g. enthusiasm) t >\m agtr< I (b) speaker MV ->lm i hearer T tr< up a given state T Figure 6 more important than her arrival at another state. But in the context evoked by John, come down, John's reaching the goal, i.e. the place where the speaker is at the moment, is what matters. Predicates highlighting the relationship of contact or spatial proximity hold also a great potential for reports intended to capture many subtleties of the speaker's handling of language and thought as well as his interaction with other speech participants. Such reports usually exploit the correspondences that may be established between the spatial landmark related to the given predicate and an idea, linguistic expression, style, question posed by another participant, type of discourse, theme, or topic. When the status of landmark is accorded to the verbal exchange between several participants, the relationship of contact will involve different utterances. Relevant here are the patterns presented in 7(b), (c), and (d). agtr< 1 1 I (a) speaker MV through/over/ into/to tr<- utterance/ its component/ associated element t »lm
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 549 (b) (c) agtr* 1 1 1 speaker MV agtr< i speaker MV >im i utterance tr^ II > Irr > in ^ I utterance tr< ' into in 1 1 L another's utterance T ^ i > im another's utterance t >lm agtix 1 I I (d) speaker MV tr<- in I another's utterance T >lm Figure 7 The above figures present only the more common patters of correspondences that involve the domains of space, spatial motion and speaking. Notwithstanding their limited scope, they demonstrate beyond dispute the important role played by the relationships of spatial detachment or attachment and proximity in establishing these correspondences. 5.5. Domain Intersections The groupings of metaphoric expressions examined also cross one another with respect to the nature of the verbs' primary domains. For some of these verbs, the primary domain is just movement through physical space. For others, however, the primary domain lies somewhere at the intersection of spatial motion and another domain, e.g. acoustics. In many instances, what emerges as primary domain goes beyond mere physical motion, though motion constitutes its integral part. As the database demonstrates, among such more complex domains that commonly extend
550 BRYGIDA RUDZKA-OSTYN into speech acts are the domains of (a) food preparation (evoked e.g. by verbs like mince, hash, interlard) (b) feeding and food distribution (ladle, spoon-feed) (c) cleaning (brush, dust) (d) weaving (interlace) (e) decoration (embellish, garnish) (f) physiological processes (vomit, belch, chew) (g) labor (hammer, drill) (h) trade (peddle) (i) war and fighting (retreat, shoot, fire). Several of them intersect further with the domains of human purposes, moral judgments, aesthetics, or various social and cultural conventions. Since the latter also interact with the domain of communication (see Figure 1), additional bridges between motion events and speech acts are readily established. One can say "additional" because there is always one element that links spatial movement, no matter what its manifestation, with speaking. This element is the change in the relationship that one entity bears towards another. 6. Conclusion It has become evident in the course of the discussion that the numerous extensions from the domain of physical movement into speech acts are intrinsically linked with the way man conceptualizes the universe. They follow naturally from his ability to isolate parts from a whole and to recognize what they have or do not have in common and how they interact. This very ability enables him to juxtapose entities within, as well as beyond, a particular whole and thus make comparisons of an increasingly higher order. In effect, anything can be related to anything else. But, in the case of metaphor, such comparisons are not unconstrained. On the contrary, as we have demonstrated, cross-domain comparisons are constrained by the structure of both the donor and the recipient. And they are highly motivated. Extensions into speech acts draw so frequently on spatial movement because the latter empitomizes change in the most concrete and also most diversified way. Any component of a motion event offers a whole range of standards for acts of comparison. Since these standards often invoke scales
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 551 and dimensions that are also applicable to speech acts, there is a common ground that further motivates the comparisons drawn. Regarding the diversity of motion events, the movement itself can be fast or slow, smooth or jerky, light or violent, single or repeated, and each of these parameters correlates with a different aspect of speaking. Also the shape and direction of the moving object's path hold an interesting potential for metaphoric transfers. No less important are the properties of both trajector and landmark. In the case of the former, it makes all the difference whether it is solid or liquid, light or heavy, a tool or an ornament while in the case of the latter, the fact that it is a container, a surface, or just a point may be relevant. As the database demonstrates, the two most commonly invoked landmark categories are MEDIUM (33%) and SURFACE (22%), which again seems to point to the importance of the notion of a bounded region in conceptualization. Readers familiar with the literature on verbal particles and prepositions will have noticed that many properties characterizing the given landmarks do not affect transfers into the domain of speaking. As students of metaphor have repeatedly observed (see e.g. Kittay and Lehrer (1981)), domain boundaries function like grids or filters that let through only elements relevant to the internal organization of the recipient domain. By examining the conceptual structure of particular spatial predicates we tried to identify the factors that facilitate and thus motivate their reference to speech events. We have also observed that this very structure determines the collocational properties of the predicates examined. There is an obvious dependency relationship between the meaning of the given verbal/ phrasal expression and the type of enunciation allowed (cf. ex. 13, 21, 75, 76). Some of the predicates allow both direct and indirect speech, others combine mainly with category labels (lexicalized by e.g. word, issue, lie), still others prefer parentheticals. This problem certainly deserves further, systematic study. Another problem that has been mentioned at different stages but has not been explored is the role of covert elements in forging metonymic links. Many other areas remain unexplored. What has become clear, however, is that borrowings from domains grounded in physical motion offer the reporter an extremely powerful means of expression. They allow him to focus on different speech-act components simultaneously. They allow him to evoke vivid images, to pass value judgments, to create humorous and other effects, and to vary his style.
552 BRYG1DA RUDZKA-OSTYN From Reddy's perspective, English, as a metalanguage, may indeed be "its own worst enemy." However, from the point of view adopted in this paper, it is its own best friend. Notes 1. I wish to thank Louis Goossens, Peter Kelly, Paul Pauwels, Yvan Putseys, Anne-Marie Vandenbergen, and Johan Vanparys for their valuable comments on an earlier version of this paper. 2. See Langacker (1983 II) for an extensive discussion and illustrations. 3. See also Langacker (1983: II/C). 4. Cf. Dirven et al. (1982), Leech (1980), and Verschueren (1984), (1985), (in press). 5. The distinction comes from Leech's study of metalanguage included in his (1980) collection. Note the modifications in terminology. 6. Throughout this paper, reference to the secondary speech situation will be indicated by the use of the terms reporter and secondary. Without the modifier secondary, the terms speaker, utterance, and hearer/addressee will always refer to the primary speech situation. The expressions speech act, speech event, and communicative act will be used interchangeably. Likewise, domain of (verbal) communication, domain of speech acts/events, and (domain of) speaking will be used to express the same concept. 7. The different relationships holding between the two situations and their impact on some linguistic choices are discussed in my (1982) paper "The Verb Ask and the Scene of Linguistic Communication." University of Leuven. MS. Figure 2 illustrates only one possible relationship, presumably the most typical one. 8. The following abbreviations will be used to refer to the sources: ADE = Hornby, A.S. 1974. Oxford Advanced Learner's Dictionary of Current English. London: Oxford University Press. C-22 = Heller, Joseph. 1968. Catch-22. New York: Dell. C & M = McCarthy, Mary. 1979. Cannibals and Missionaries. New York: Harcourt Brace Jovanovich. DPV = Courtney, Rosemary. 1983. Longman Dictionary of Phrasal Verbs. Harlow: Longman. NW = Newsweek ( O = October, N = November, etc. TM = Time Magazine I No date is provided in the case of the November/ I December(N/D) special issue of Newsweek covering the ^ 1984 presidential election. WEB = Gove, Philip B., ed. 1971. Webster's Third New International Dictionary of the English Language. Unabridged. Springfield, Mass.: Merriam. Lack of specifications means that the examples were provided by native informants. 9. That this indeed is the case is demonstrated not only by the present database but also by
SEMANTIC EXTENSIONS INTO VERBAL COMMUNICATION 553 the numerous uses of verbs of movement to indicate the passage of time, emotional experiences, weather conditions, oscillations in temperature, cognitive and biological processes, and many other phenemena. Langacker (1986b) also points to the linkage in the conceptualization of change and motion. I discovered this after I had completed my own investigation into the topic. 10. See Talmy (1975) and (1985a) for a much fuller and somewhat different treatment of motion events and their linguistic manifestations. One important difference involves Figure and Ground. In this paper, they are accorded a higher degree of relativity, also with respect to the mobile/static distinction. 11. This case appears to be another manifestation of what Langacker (1985) subsumes under 'subjective construal.' 12. See Ullmann (1962) for a discussion and numerous illustrations. 13. Arguments for such an inclusion are presented in Eco (1985: 114-117). 14. Cf. Lakoff & Johnson (1980) or Ullmann (1962). Ullmann also recognizes the intimate relationship between metaphor and metonymy and the difficulty of separating the two tropes. In this context, see also Culler (1981) and Eco (1985). 15. The differences and similarities between these two categories are discussed at length in Lindner (1981). It appears, however, that the main distinction between particles and prepositions lies in the kind of construal they promote. As I see it, a particle is used when a subjective viewing arrangement in the sense of Langacker (1985) is desired, whereas a preposition occurs when an objective construal is preferred, and it may well be that some of the differences noted by Lindner and others are a consequence of this basic distinction. 16. For an in-depth description of the semantic structure of particular particles and prepositions see Brugman (1981), Hawkins (1984), Herskovits (1986), and Lindner (1981). 17. This interpretation is also compatible with the scene described in example (50) in the subsection on at. Given the fact that lash can function as a transitive or intransitie verb, the landmark of out can correspond equally well to the speaker's mouth as to the barriers of silence. 18. Despite some similarities of thought, the observations presented in this section were made independently of Lindner's study. 19. A very similar conflation of roles has been observed in (5)-(6), but while in (5)-(6) it is due to the presence of a particle, here it stems from the semantic structure of the verbs. In certain contexts, the landmarks of these verbs can be construed not only as the trajec- tors' initial locations but also as semi-initiators of the trajectors' movement. For discussion of this class of verbs, and a slightly different interpretation, see Talmy (1985a: 96ff). 20. Observe that in 5(b) the speaker overtly assumes the role of agentive trajector and trajec- tor. However, in accordance with the structure of speech acts, the role of trajector is covertly assigned to the speaker's words. Similar metonymy is at work in 5(d) and many other cases for that matter. A full description of the correspondences drawn in 5(d) will certainly include a covert instrumental trajector to be correlated with the speaker's words.
Spatial Metaphor in German Causative Constructions Robert Thomas King University of Frankfurt 1. Introduction Localist approaches to semantics are not new (cf. Anderson 1971, 1977), but they have recently received considerable support from research in cognitive linguistics (cf. Lakoff 1982a, Langacker 1982a, 1983, etc.). In particular, it has been argued that even the non-locative usages of certain prepositions and particles are grounded in spatial metaphor (Lakoff and Johnson 1980, Lindner 1981, Brugman 1981). Among other things, I will attempt in this paper to show that the German prepositions von, mit, and durch have spatial interpretations even when they are used to mark agents, instruments, or causes, respectively. I will indeed argue in favor of a cognitive interpretation of agentivity, instrumentality, and causativity in terms of spatial relations. To do so, it will first be necessary to establish a clear definition of the term "causative construction" which will provide an unequivocal basis for distinguishing causatives from non-causatives. 2. Causative vs. Non-causative The semantic distinction between causative and non-causative constructions can best be expressed in terms of entailment relations. A construction is defined in this paper as "causative" if and only if the proposition containing it entails the following propositions: (a) There is an event er (b) There is an event er (c) Event e1 precedes event e2. (d) Event e2 presupposes event er
556 ROBERT THOMAS KING If any one of these propositions is false, the expression containing the construction is not considered causative. Propositions (a) and (b) are both necessary if a reasonable distinction is to be made between causative expressions and descriptions of physical events in general. An expression is regarded as causative in this paper only if two distinct events (a cause and its effect) are semantically entailed. If only one event is described in a linguistic expression, the expression should not be called causative, even if the event involves physical causation. Propositions (c) and (d) reflect the metaphysical aspects of causation. Although not all cases of physical causation are described in language by causative constructions, a causative construction cannot be truthfully asserted if there is no causal relation between the two events. Proposition (d) perhaps requires some explanation: it does not imply that event e2 would never have occurred in the absence of event e ; rather, it means that event e2 would not have occurred when it did if event cl had not occurred. If I know that an electrically operated door opens automatically every hour on the hour and I push on it precisely when it begins to open automatically, it cannot be truthfully asserted that I have opened it. This definition of "causative construction" in terms of entailment relations captures the essence of the hypothesis of event causation (McCaw- ley 1971a, 1972, Fillmore 1971b, Dowty 1972, King 1974a,b, Shibatani 1976a). The sentence (1) John killed Harry is a now-classical example of a causative expression; it entails two events, namely that John did something to Harry, and that as a result Harry changed from the state of being alive to the state of being dead. However, (2) John hit Harry is not a causative. It certainly entails that John did something to Harry, since it entails itself, but it does not entail that Harry underwent a change of state, although Harry is certainly affected by John's action. It is important here to define what is meant by "state" and "change of state". A state is a condition which is regarded by the speaker as not changing over a certain period of time; a state can be expressed linguistically in English as a combination of the non-progressive form of the copula to be and a predicate adjective. (I do not consider the term "state" here to refer to continuous actions; see McCawley 1976.) For instance,
SPATIAL METAPHOR IN GERMAN CAUSATIVE 557 (3) Harry is dead is a state-description. Every lexical causative in English (or German) is semantically related to a predicate adjective construction describing the resultant state. The adjective may be morphologically distinct from the causative verb, as with dead and kill; it may be morphologically similar to the causative, or even identical, as with open (the adjective) and open (the transitive verb); or it may be the past participle of the causative verb, as with broken and break. Notice that (4) The window was broken has two possible interpretations: one is a state-description and the other is a passive construction. (Cf. Langacker 1982a: 59ff.) However, the sentence (5) Harry was hit can only be a passive construction. There is no adjective describing a state which a person is in as a result of being hit. One might object to this analysis by saying that (2) does indeed entail that Harry underwent a change of state: namely, from the "state" of "not having been hit by John" to the "state" of "having been hit by John"; this objection, however, confuses state-descriptions with histories, and is untenable. 3. Prototypical Causation Lakoff (1977: 244, 1982a: 47) argues that prototype theory provides a better characterization of the causative vs. non-causative distinction than is offered by the classical approach discussed in the last section. Prototype theory is based on natural categorization (cf. Rosch 1973a) and interactional properties. According to Lakoff, prototypical causation is direct manipulation, and it characteristically involves "an agent that does something", "a patient that undergoes a change to a new state", "a transfer of energy from the agent to the patient", and "some instrument" used by the agent (Lakoff 1982a: 47). Most of the examples of causative expressions discussed in this paper fit Lakoff s characterization of prototypical causation. In general, we can say that lexical causatives are used if the agent and the patient interact immediately and directly, whereas periphrastic causatives are preferred when no direct interaction is described. Lakoff and Johnson (1980: 131) argue that this is a consequence of the cognitive metaphor "closeness is strength of effect": a lexical causative subsumes
558 ROBERT THOMAS KING both the agent's action and the patient's change of state in one word; this is possible only if the two events are immediately connected. If the cause and the effect are only indirectly connected, they must be described by different lexical items. Also, expressions denoting the agent and the patient must be more syntactically "distant" in the surface structure than is the case with direct causation. 4. Events, Agents, Patients, and Instruments Prototype theory complements the semantic entailment analysis of causative constructions by providing it with a cognitive basis. The prototype for event e is an action by an agent. Although non-agentive processes can constitute causes, as in (6) John was killed by a falling rock most typical examples of causative constructions involve agents. An agent is an entity which is considered capable of initiating and carrying out an action without being under the direct control of some other entity. Although agents are "typically animate", natural forces, as in (7) The wind blew the door open and even machines which follow an operational program may be linguistically called agents. In the most common examples of causatives event e2 is a change of state of a patient, which may be animate or inanimate. A patient may be affected by the action of an agent without however undergoing a definite change of state if the change is not entailed by the description of the event. In the sentence (8) John kicked the door the door is certainly affected in some way (physically contacted by John's foot) but we cannot conclude that it has undergone a definite change of state. It may be open, dented, or knocked off its hinges, but (8) entails none of these. If the resultant state is to be indicated, it must be overtly expressed, making the construction causative, as in (9) John kicked the door open If event e is an action by an agent, the use of an instrument is usually implied even if it is not explicitly mentioned. An instrument is usually a
SPATIAL METAPHOR IN GERMAN CAUSATIVE 559 physical object, and is always under the control of an agent, even if the agent is not overtly mentioned. In the sentence (10) The key opened the door the key is assumed to be used by an agent. If the key were not controlled by an agent, the sentence would seem strange indeed. Since an instrument is controlled by an agent, it must also accompany the agent. Hence it is no accident that the preposition marking an instrument is the same in many languages as the preposition used to mark accompaniment. For Lakoff and Johnson (1980: 134) this is an example of "metaphorical coherence in grammar" and reflects the metaphor "an instrument is a companion." 5. Foregrounding and Syntactic Marking Most examples of causative constructions in this paper involve five "participants": 1) The causing event e1? which is usually an action; 2) The caused event e , which is usually a change of state; 3) The agent of event e:; 4) The patient of event e2; 5) The instrument controlled by the agent. These may be foregrounded in various ways; in (1), the agent, the patient, and the change of state of the latter are overtly expressed. An action event is entailed, but its exact nature is left open; that is, we do not know what John did to Harry in order to kill him. No instrument is mentioned, but the use of some instrument (if only John's bare hands) is necessary if a physical change of state like forcible death is to be effected. Most cases of simple lexical causatives are like (1). The exact nature of the causing event can be specified by adding a gerundive phrase: (11) John killed Harry by stabbing him to death The instrument can be indicated by adding a phrase like "with a knife" to either (1) or (11). The content of (11) can also be expressed by asserting the action and the resultant state: (12) John stabbed Harry to death Without the resultant state, (12) would describe only an action and would therefore not be a causative. In a passive construction, only the patient and
560 ROBERT THOMAS KING the change of state need overtly be expressed, although an unspecified action is entailed: (13) Harry was killed If the causation is indirect, the change of state of the patient may be indicated in the main clause and the causing event in a gerundive clause: (14) Harry died as a result of being poisoned (by John) Occasionally, the causing event is expressed in a subordinate clause, although this form is a bit awkward in English: (15) Harry died as a result of the fact that John had poisoned him The causing event may also be expressed by a nominalization: (16) Harry died in the explosion Here, the word explosion can best be regarded as a nominalization of the event description "something exploded." Thus, if the causing event is overtly described, it may be expressed in the main clause of the sentence, in a subordinate clause, as a verbal noun (a gerund), or as a nominalization. The syntactic marking of the five participants in a causative expression is achieved in English through the use of word order and prepositional phrases. In an active sentence, the agent is the grammatical subject, the patient is the direct object, and the instrument, if overtly expressed, appears as the object of the preposition with. In a passive sentence, the patient is the subject, and the agent, if overtly expressed, occurs as the object of by. If an instrument but no agent is overtly expressed, the instrument is the subject in an active sentence, but still the object of with in a passive sentence. (The hierarchy of case markings is discussed in detail in Fillmore 1968). If the causing event is expressed by a verbal noun or a nominalization, the syntactic marking in English varies. If the causation is intentional, the causing event is expressed as the object of by, as in (11). But if the causation is unintentional, another preposition must be used, as Talmy (1976: 86) points out. In general, the use of by + personal agent implies an intentional result, which, in the case of causation, implies an intentional action. An unintentional action, like sneezing, cannot have an intentional result. Hence, the sentence (17) John blew the papers off his desk by sneezing at them can have only the somewhat implausible interpretation that John can sneeze at will.
SPATIAL METAPHOR IN GERMAN CAUSATIVE 561 The preposition through can be used to indicate mediating events or causes which lead to a final result, although this usage is relatively uncommon in modern English: (18) John became ill through negligence The case marking of agents, patients, and instruments in English does not provide us with syntactic evidence for distinguishing causative constructions like (1) from non-causative transitives like (2), even in their passive forms. Furthermore, intentional causing events are marked in the same way as personal or natural agents (with by in the passive). The primary syntactic evidence in English for the semantic distinction discussed in Section 2 of this paper is the predicate adjective argument. Since it would be desirable to have a set of syntactic criteria for the causative/non-causative distinction which would agree with the semantic criteria discussed above, I turned to German, which seems to present clearer syntactic evidence than English. 6. Pedagogic Grammars vs. Actual Usage Pedagogic grammars of German are often incomplete and even misleading in their descriptions of the syntactic marking of agents, instruments, and causes in passive sentences. For instance, Lohnes and Strothmann (1968: 603-604) state that: If a personal agent is mentioned, von is used: (19) Jerusalem wurde von den Romern zerstort "Jerusalem was destroyed by the Romans" Mit is used for the instrument "handled" by a personal agent: (20) Abel wurde von Kain mit einem Stein erschlagen "Abel was slain by Cain with a rock" Abstract causes, impersonal causes, and impersonal means of destruction are introduced by durch: (21) Sie wurden durch ein neues Gesetz gezwungen, das Land zu ver- lassen "They were forced by a new law to leave the country" (22) Lissabon wurde durch ein Erdbeben zerstort "Lisbon was destroyed by an earthquake"
562 ROBERT THOMAS KING (23) Dresden wurde durch Bomben zerstort "Dresden was destroyed by bombs" (My numbering) It is also argued that two binary distinctions are involved, between von and durch on the one hand, and between durch and mit on the other. Schmitz (1962: 10) asserts that "beim Passiv nennt von den Urheber und durch die Ursache" ("in the passive, von names the initiator and durch the cause") and that the "medial" use of durch must be distinguished from the "instrumental" use of mit. That is, personal agents are marked by von, causing events by durch, intermediaries by durch, and instruments by mit. For Schmitz, an intermediary can be a person, as in (24) Ich schicke Ihnen das Buch durch meinen Boten "I'll send you the book by my messenger" (25) Ich habe sie durch meinen Freund kennengelernt "I met her through my friend" or an abstract concept, as in (26) Durch Zufall habe ich erfahren, dafl ... "Coincidentally, I have learned that ..." (27) Durch Beharrlichkeit kommt man ans Ziel "One reaches one's goals through perseverance" Other authorities on German grammar reach the same conclusions (cf. Curme 1922, Griesbach and Schulz 1965, etc.) In general, the combinations durch + personal agent and durch + instrument are either not at all mentioned, or are explicitly proscribed (Drosdowski, Grebe, and Muller 1965: 473). That the above description has little in common with normal usage is readily evident upon perusal of German newspapers, magazines, books, etc., or after conversations with native speakers. The combination durch + personal agent occurs fairly often, as in (28) Die Verhandlung wurde durch den Richter abgebrochen "The hearing was broken off by the judge" (29) Er wurde durch seinen Lehrer aufgehalten "He was delayed by his teacher" (30) Seine Mobel wurden durch den Gerichtsvollzieher gepfandet "His furniture was impounded by the sheriff"
SPATIAL METAPHOR IN GERMAN CAUSATIVE 563 The combination durch + instrument is also not infrequent: (31) Er wurde durch dieses Medikament geheilt "He was healed with this medicine" (32) Sein Gesicht war durch die Maske verdeckt "His face was hidden by the mask" (33) Er wurde durch den Strang hingerichtet "He was executed by hanging" (lit. "by the rope") When asked about the inconsistencies between the prescriptive grammars and actual usage, several native speakers expressed the opinion that both von and durch could be used with personal agents with no difference of meaning, and that the situation was the same with durch and mit applied to instruments. This is also the opinion of some modern grammarians (Helbig and Buscha 1974). To help clarify the matter, I decided to conduct a survey of actual usage. 7. A Survey of Prepositional Usage in German Causative Constructions Over the past several years, I have conducted an extensive survey of causative constructions in German. Part of the survey is concerned with the prepositional marking of the participants in a causative construction. The original goals of the study were to determine the facts of actual usage (as opposed to the sometimes biased opinions of grammarians) and to find syntactic correlates for the semantic definition of causative in terms of entailment relations. As a side effect, I have discovered that the data are quite amenable to an analysis in terms of prototype theory and spatial metaphor. The data presented in this paper are a representative selection from the analysis of over 50 questionnaires and approximately 30 interviews with native speakers of German. About three-fourths of the informants come from an academic background, both students and colleagues of mine at the University of Frankfurt. The remainder is made up of people from outside the university environment. All of the informants speak standard German (Hochdeutsch). In the survey on prepositional usage, I was more interested in contex- tualization processes than in mere acceptability judgments. In general, I hold that a binary distinction between acceptable and unacceptable (or even "grammatical" and "ungrammatical") strings is untenable in describing linguistic behavior. Rather, acceptability should be considered a graded
564 ROBERT THOMAS KING concept, based primarily on difficulty of contextualization. Of course, gross violations of syntax may make a string of words virtually incomprehensible and therefore unacceptable, but many examples of unacceptable utterances in the linguistic literature are not necessarily ungrammatical; it may require considerable effort, however, to contextualize them. This applies primarily to cases of semantic anomaly. The gradedness of acceptability was borne out in my data: whereas some sample sentences were readily accepted by almost every informant and some sentences were uniformly rejected, many sentences fell somewhere in between. Often, a sentence would be initially rejected by an informant only to be later accepted after the informant had exercised some effort in contextualizing it. Based on this observation, I constructed several questionnaires which permitted graded judgments and comments. I asked the informants to classify the sample sentences as (A) immediately acceptable with little effort of contextualization; (B) acceptable after some effort of contextualization; (C) unacceptable even after considerable effort of contextualization. For sentences which fell under (B), I asked my informants to describe the contexts which would be required to make the sentence acceptable. Also, if two sentences which differed in only one preposition were both judged as immediately acceptable, I asked the informants to explain the difference in meaning, if any, between the two sentences. The data discussed in this paper come primarily from a questionnaire and several interviews devoted to a contrastive analysis of von, mit, and durch. The informants were presented with twenty-five sentence triples. The sentences in each triple were identical except for the preposition {von, mit, or durch) marking the agent, instrument, or cause. As indicated in the last paragraph, I requested graded acceptability judgments and comments on each sentence. 8. Prototype Combinations Of the nine possible combinations (von + agent, von + instrument, von + cause, mit + agent, mit + instrument, mit + cause, durch + agent, durch + instrument, and durch + cause), three were almost universally accepted (von + agent, mit + instrument, and durch + cause):
SPATIAL METAPHOR IN GERMAN CAUSATIVE 565 (34) Casar wurde von Brutus getotet "Caesar was killed by Brutus" (35) Casar wurde mit einem Dolch getotet "Caesar was killed with a dagger" (36) Casar wurde durch einen Dolchstofi getotet "Caesar was killed by means of a dagger-thrust" (Dolchstofi can be regarded as a nominalization of "jemand stofit jemand mit einem Dolch"; it is thus an event description.) Sentences (34) and (35) were accepted by all my informants without hesitation. Sentence (36) was classified as completely acceptable by 89% of my informants; the remainder said it required some contextualization, but was certainly possible. No one rejected it outright. The other sample sentences with these combinations were graded as totally acceptable by 90- 100% of the informants. Interestingly, no distinction whatever was discernible between personal agents and natural agents; both of the following sentences were immediately accepted by 100% of the informants: (37) Das Fenster wurde vom Wind zugeschlagen "The window was slammed shut by the wind" (38) Lissabon wurde 1755 von einem Erdbeben zerstort "Lisbon was destroyed by an earthquake in 1755" The variants with mit were rejected totally by over 95% of the informants. Perhaps this fact will finally lay to rest the recurrent suggestion (since Fillmore 1968) to treat natural forces as instruments. There is no evidence for it in the syntax of English or German. While the combination durch den Wind met with mixed results (one-third of the informants found it totally acceptable, about half accepted it after some contextualization, and the remainder rejected it outright), the combination durch ein Erdbeben was immediately accepted by everyone. This can be explained by pointing out that Wind is not a nominalization in German, but Erdbeben "die Erde bebt" is. Thus, the word Erdbeben can be understood in two ways: as a natural agent (a force) and as an event (a process). Another example of such a "dual role" is provided by the word Essen: (39) Von ubermafiigem Essen wird man dick "You get fat from too much food"
566 ROBERT THOMAS KING (40) Dutch ubermafliges Essen wird man dick "You get fat from eating too much" Sentence (39) was totally accepted by over 90% and sentence (40) by 100% of the informants. In sentence (39) the word Essen refers to food, too much of which can be the source of excess body fat; in sentence (40) Essen refers to the action of eating, and is thus a causing event in becoming overweight. The universal acceptability of the combinations von + agent, mit + instrument, and durch + cause should come as no surprise, since these are precisely the combinations sanctioned by the pedagogic grammars. They can even be considered prototypes in terms of which the acceptability ratings of the other combinations can be evaluated. That is, the informants attempt to interpret the object of von as an agent, the object of mit as an instrument, and the object of durch as a cause. 9. Rare or Difficult Combinations Of the nine possible combinations discussed in the last section, three were widely rejected: von + instrument, von + cause, and mit + agent. To the extent that they were accepted, massive efforts at contextualization were necessary, or the meaning of the original was changed. For instance, the combination mit + person was often reinterpreted in the sense of accompaniment; the sentence (41) Cdsar wurde mit Brutus getotet "Caesar was killed with Brutus" was said by some informants to mean that Caesar and Brutus were killed together, at the same time (a historical falsity). The interpretation of Brutus as a physical instrument used in Caesar's murder was offered only by students who had studied case grammar with me. The sentence (42) Der Verbrecher wurde mit der Polizei beobachtet "The criminal was observed with the police" was interpreted by over two-thirds of the informants to mean that the criminal and the police were acting together and were being observed by someone else. The combination mit + natural force was totally rejected by over 90% of the informants. Those that accepted sentences like (43) Das Fenster wurde mit dem Wind zugeschlagen "The window was slammed shut with the wind"
SPATIAL METAPHOR IN GERMAN CAUSATIVE 567 or (44) Lissabon wurde 1755 mit einem Erdbeben zerstort "Lisbon was destroyed with an earthquake in 1755" said that it committed them to a belief in either weather control or a (malevolent) Supreme Being. Fire, however, can evidently be considered both as a natural force and as a controlled object: (45) Karthago wurde von Feuer zerstort "Carthage was destroyed by fire" was accepted, with contextualization, by slightly over half the informants; and (46) Karthago wurde mit Feuer zerstort "Carthage was destroyed with fire" was also accepted by over half the informants under the interpretation that the fire was intentionally laid (presumably by the Romans). The variant with durch, treating fire as a process or an impersonal cause, was immediately accepted by every informant. The combination von + physical object was totally rejected by over 80% of the informants if the physical object was something (like a dagger) normally under the control of an agent: (47) Casar wurde von einem Dolch getotet "Caesar was killed by a dagger" The contextualizations offered for this sentence (by those who were marginally willing to accept it) all involved a dagger under the control not of an agent but of a natural force; they tried to imagine a situation in which the dagger had fallen off a high shelf and happened to land point-down in Caesar's back. This latter interpretation is conditioned by the fact that (48) Hans wurde von einem Stein erschlagen "Hans was killed by a (falling) rock" is perfectly acceptable under the interpretation that the stone had fallen from natural causes, or had accidentally been kicked loose (in mountain climbing). Finally, the combination von + causing event was rejected totally by over 95% of the informants. Evidently, it is virtually impossible to contex- tualize an event as an agent. All but one of my informants said that the following sentence was, quite simply, grammatically incorrect:
568 ROBERT THOMAS KING (49) * Die Demonstration wurde vom Eingreifen der Polizei aufgelost "The demonstration was broken up by the intervention of the police" 10. Non-prototypical Combinations The remaining combinations (mit + cause, durch + agent, and durch + instrument) were widely accepted, in spite of the fact that they are non- prototypical and would be rejected by prescriptive grammarians. The first combination can be explained as a reinterpretation. Instead of being seen as a cause, the event phrase which is the object of mit in the sentence (50) Die Demonstration wurde mit dem Eingreifen der Polizei aufgelost "The demonstration was broken up with the intervention of the police" was interpreted by most of the informants not as a cause, but as a contemporaneous event; that is, the demonstration dissolved just as the police began to intervene, so that the intervention by the police could not be seen as the cause. The most intriguing combinations are durch + agent and durch + instrument. They were both immediately accepted by over half of the informants, and more than a third accepted them after some effort at contex- tualization. Only about 10% rejected them outright. Furthermore, unlike the combinations discussed in the last section, no reinterpretation of case roles proved necessary. The object of durch in (51) Casar wurde durch Brutus getotet "Caesar was killed (through) Brutus" was interpreted as a personal agent, and the object of durch in (52) Casar wurde durch einen Dolch getotet "Caesar was killed (through) a dagger" was interpreted as an instrument. Although many informants professed to see no difference in meaning between (34) and (51) or between (35) and (52), more than half of them, when pressed for a contextualization, offered something like the following explanation. Brutus in sentence (51) is indeed an agent who acts to kill Caesar, but he is not seen as the primary agent. Rather, Brutus functions as an intermediary agent who carries out someone
SPATIAL METAPHOR IN GERMAN CAUSATIVE 569 else's instructions. He is thus under the control of another and is really the indirect means by which the murder is performed. The direct responsibility of Brutus in sentence (51) is reduced, as is shown by the fact that the sentence (53) Cdsar wurde durch Brutus ermordet "Caesar was murdered (through) Brutus" has a much lower acceptability rating, since it implies strong intentionality on the part of Brutus. Other examples are perhaps clearer. Since the police can be regarded either as personal agents acting under their own control or as the extended arm of state authority, both of the following sentences should have roughly the same acceptability rating, and indeed they do: (54) Die Demonstration wurde von der Polizei aufgelost "The demonstration was dissolved by the police" (55) Die Demonstration wurde durch die Polizei aufgelost "The demonstration was dissolved by means of the police" The sentences differ in meaning, however; in (54) the initiative is seen as coming from the police themselves, whereas in (55) they receive their orders from higher authority. The difference between von + agent and durch + agent is not just one of authority and responsibility, but also of control and volition. In the sentence (56) Der Abschluji des Verfahrens wurde vom Richter hinausgezogert "The conclusion of the trial was delayed by the judge" the authority of the judge to influence the proceedings (both effectively and intentionally) is emphasized; in (57) Der Abschlufi des Verfahrens wurde durch den Richter hinausgezogert "The conclusion of the trial was delayed because of the judge" the emphasis is on the cause of the delay; the judge is no longer necessarily an agent in an action, but a factor, perhaps among others, delaying the trial. The distinction becomes clearer if we replace "the judge" by "new witnesses":
570 ROBERT THOMAS KING (58) Der Abschlufi des Verfahrens wurde von neuen Zeugen hinausgezogert "The conclusion of the trial was delayed by new witnesses" (59) Der Abschlufi des Verfahrens wurde durch neue Zeugen hinausgezogert "The conclusion of the trial was delayed because of new witnesses" Sentence (58) implies that the new witnesses were actively involved in slowing down the trial, normally a prerogative of the prosecution or the defense. In sentence (59) the new witnesses are the reason why the trial was delayed, although they may not have intended it at all or been in any way actively involved. For example, the defense attorney may have been actively trying to delay the conclusion of the trial by calling new witnesses. The sense of durch 4- agent is thus actually durch 4- cause, and the meaning of (56) is more precisely described as (60) Der Abschlufi des Verfahrens wurde durch das Eingreifen des Richters hinausgezogert "The conclusion of the trial was delayed because of the judge's intervention" and (59) is better expressed as (61) Der Abschlufi des Verfahrens wurde durch die Vernehmung neuer Zeugen hinausgezogert "The conclusion of the trial was delayed because of the interrogation of new witnesses" That is, the object of durch may not be the agent of the action, but the patient instead. Whenever the phrase durch 4- person is used, it should be understood as a reduced form for durch 4- cause, where the personal object of durch functions as a participant in the causing event. In (53) the emphasis is on the action of killing Caesar and not on Brutus' agentivity, and in (55) the emphasis in on the action of the police and not on their ultimate responsibility. The combination durch 4- instrument lends itself to a similar analysis. The sentence (62) Casar wurde durch einen Dolch getotet "Caesar was killed on a dagger"
SPATIAL METAPHOR IN GERMAN CAUSATIVE 571 describes a causing event in which the dagger is involved, but the emphasis is not on its instrumentality. In (35) the dagger must be controlled by an agent, but in (62) this is not the case. To describe a situation in which Caesar stumbles and accidentally falls on his dagger, thereby killing himself, my informants stated that only (62) is possible. The combinations durch + agent and durch + instrument are the rule in a special class of "weak" causative constructions in German in which the change of state of the patient is strongly foregrounded: (63) Hans starb durch Peter "Hans died because of Peter" (64) Hans starb durch seine eigene Hand "Hans died by his own hand" (65) Hans starb durch eine Kugel "Hans died from (being struck by) a bullet" When asked to explain the meaning of these sentences, my informants said that (63) implies that Hans died as a result of some perhaps inadvertent action of Peter's; they rejected (66) Hans wurde von Peter getotet "Hans was killed by Peter" as a possible paraphrase. In (64) the responsibility of Hans for his own death is de-emphasized through the reference to his hand (an instrument), and in (65) the bullet almost seems to embody an impersonal force. The prepositions von and mit were rejected by all of my informants for these sentences. Although the combinations durch f agent and durch + instrument are relatively common in German (in spite of their dislike by traditional grammarians), their interpretation presents difficulties since they do not seem to constitute clear cases of agentivity or instrumentality. The problem lies perhaps in the attempt to define unambiguous categories of agent and instrument. Prototype theory provides the means of analyzing and grading categories by comparing their members to some ideal set of features, and it is applicable here. If we assume that the primary (non-concrete) function of durch in passive sentences is to indicate causes, then this is most clearly expressed when the object of durch is a clause, a verbal noun, or a nominalization, and these constructions are universally accepted by native speakers of German. The combination durch + cause can therefore be con-
572 ROBERT THOMAS KING sidered a prototype. When native speakers are asked to produce an example of the use of durch in this sense, they are most likely to combine it with a verbal noun or a nominalization. It is no accident that precisely these constructions occur as examples in the grammars. When the object of durch is a personal agent or an instrument, however, the prototype no longer directly applies, but it can serve as a guide for interpretation. I propose to treat such cases as reduced representations of causing events. 11. Causatives vs. Non-causative Transitives If the primary non-concrete function of durch is to mark causes, then it should not be expected to appear in the passive forms of non-causative constructions. To test this assumption, I included in the survey a number of causative and non-causative transitive verb phrases. The semantic criterion for causativity was the definition in Section 2 of this paper: if the proposition expressed in a sentence failed the entailment test, I considered the construction non-causative. For purposes of conciseness, I restricted the sample sentences to descriptions of simple actions involving agents, patients, and instruments. In the case of causatives the caused event was a change of state of the patient. German lends itself particularly well to such an analysis, since it has several pairs of transitive verbs which clearly show the above semantic distinction: one verb of each pair describes an act of physical violence directed by an agent against a patient where no change of state of the patient can be inferred; the second verb of each pair, which is derived from the first by prefixation, adds the information that the patient died as a result of the violent interaction. Some of these pairs are listed below: Non-causatives schlagen "hit", "strike", "beat" drucken "press" wiirgen "choke" schieften "shoot" stechen "stab" Causatives erschlagen "beat to death" erdriicken "crush to death' erwurgen "strangle" erschieften "shoot to death erstechen "stab to death"
SPATIAL METAPHOR IN GERMAN CAUSATIVE 573 (The verbs schiefien and stechen are not used if the direct object refers to a human patient.) I tested the acceptability of the combinations von 4- agent, durch + agent, mit 4- instrument, and durch 4- instrument with the pair schlagenl erschlagen. (In a preliminary survey, I determined that no consistent divergences in data could be observed for the other verb pairs.) There was no difference between causatives and non-causatives in the relative acceptability of the von and mit combinations: (67) Hans wurde von Peter geschlagen "Hans was struck by Peter" (68) Hans wurde mit einem Stock geschlagen "Hans was beaten with a stick" (69) Hans wurde von Peter erschlagen "Hans was beaten to death by Peter" (70) Hans wurde mit einem Stock erschlagen "Hans was beaten to death with a stick" These sentences were considered immediately acceptable by all informants. The evidence is similar to English: there is no difference between causatives and non-causatives in the marking of agents and instruments. Whereas the acceptability rating of the durch phrases in the causative constructions was comparable to that discussed in the last section, their acceptability rating in the non-causatives was essentially nil. The sentences (71) Hans wurde durch Peter erschlagen "Hans was beaten to death (by, because of?) Peter" and (72) Hans wurde durch einen Stock erschlagen "Hans was struck to death by a stick" were rated as immediately acceptable by over half the informants and acceptable after contextualization by a third; 10% rejected (71) and 15% rejected (72). But the sentences (73) ?Hans wurde durch Peter geschlagen and (74) ?Hans wurde durch einen Stock geschlagen were regarded as totally unacceptable by over 95% of the informants. Only two informants out of 54 considered these sentences normally acceptable,
574 ROBERT THOMAS KING and only two or three were willing to try to contextualize them. The results for the other verbs were the same; there was almost 100% correlation between the acceptability of durch phrases and causativity defined on semantic grounds. As a control, I "causativized" the non-causatives by adding an adjective describing the resultant state of the patient (see Section 2): (75) Hans wurde durch Peter bewufitlos geschlagen "Hans was beaten senseless by Peter" (76) Hans wurde durch einen Stock bewufitlos geschlagen "Hans was beaten senseless with a stick" The acceptability ratings of (75) and (76) were similar to those of (71) and (72). I also tested non-causative transitive verbs which do not have a causative counterpart: (77) Der Jager wurde von seinem Hund gebissen "The hunter was bitten by his dog" (78) ?Der Jager wurde durch seinen Hund gebissen Sentence (77) was marked immediately acceptable by all my informants, and over 95% totally rejected (78). The results were similar with the experiential verb sehen "see": (79) Hans wurde gestern von Maria im Park gesehen "Hans was seen by Maria in the park yesterday" (80) ?Hans wurde gestern durch Maria im Park gesehen "Hans was seen through Maria in the park yesterday" Sentence (79) was rated immediately acceptable by all but one of my informants, indicating that there is no distinction in German between agent and experiencer marking (as in English). Sentence (80) was rejected by 95% of my informants; purely experiential verbs cannot be causatives. Three of my informants said that (80) was totally acceptable: they said it meant that Hans was seen through Maria's arms or legs by someone else. This was clearly due to the interpretation of durch in a more concrete locative sense. The semantic distinction between causatives and non-causatives on the basis of entailment relations thus has a direct syntactic correspondence in German: cooccurrence with the non-concrete sense of durch. If a construction describes both an action by an agent and a resultant change of state of a patient, the combinations durch + agent and durch + instrument are
SPATIAL METAPHOR IN GERMAN CAUSATIVE 575 widely accepted by native speakers. If, on the other hand, the construction describes only an action directed against a patient, the combinations with durch are almost universally rejected. 12. Spatial Metaphor in Case Grammar The prepositions used in German to mark agents, instruments, and causes also have locative senses, and until fairly recently no one considered that the "concrete" and "non-concrete" uses of these prepositions might have a common cognitive basis (but see Mackenzie 1977). In my dissertation (King 1974a) I specifically differentiated between the locative and non- locative uses. I treated the locative prepositions as logical predicates in a locative case relation with their objects and I considered the non-locative uses to be surface markings of underlying cases. But the participant roles in a causative construction have a natural locative interpretation, and this has been brought out in several works on case grammar (see esp. Anderson 1971, 1977, Nilsen 1972, 1973, King 1974b): The agent is the source of the action which is directed against the patient as its goal. Although Source and Goal as case roles were originally intended (Fillmore 1969: 116) to mark physical locations from or to which an object moved, they were soon extended to cover metaphorical "locations": Fillmore (1971b) describes Goal as "the end result of some action or change"; it thus absorbs the earlier Factitive. Most importantly, the analysis of causatives was revised by Fillmore to accommodate the events involved. The causing event was assigned the case role Instrument and the caused event was regarded as Goal. Feature analyses of cases (Anderson 1971, 1977, Nilsen 1972) have argued for treating source and goal as features which can be assigned to particular case roles. (Anderson uses slightly different terminology.) For example, Nilsen (1972: 34-37) suggests that Agent be assigned the features [+Controller, + Cause, + Source] and Patient the features [+Effect, + Goal]. 13. A Localist Basis for von, mit, and durch The case relation Source is marked in German in two different ways. The essential locative use of the preposition von "from" is to indicate
576 ROBERT THOMAS KING motion away from a point or away from a surface; it contrasts with aus "from, out of", which is used to describe a change of location from the inside of an object or container to the outside: (81) Er stand vom Tisch auf "He stood up from the table" (82) Er kam aus dem Haus "He came out of the house" Von can be used only if the previous location could be indicated by an "at" (touching), auf "on", or bei "at" (proximity). If the previous location was described as in something, only aus may be used to describe the change of location. This distinction is much stricter than that between from and out of in English: (83) Er kam aus dem Wohnzimmer "He came from the living room" is a normally acceptable sentence in German, but (84) ?Er kam vom Wohnzimmer is extremely difficult to contextualize, since it seems impossible to be at a room without being in it. (Vom is a contraction of von dem, "from the".) Von is, of course, also used to mark agents. Unless we consider this to be a remarkable coincidence, the most reasonable assumption would be that these two uses of von have a common cognitive basis, namely in the localist account of agentivity. The other "non-locative" uses of von can similarly be shown to have a localist basis. In (39), the food is seen as the source of the excess weight, and in (85) Hans is mude von der Arbeit "Hans is tired from work" the work is the source of Hans' exhaustion. (This and the following three examples are taken from Schmitz 1962: 38-42.) The qualitative sense, as in (86) ein Buck von 500 Seiten "a book of 500 pages" has the interpretation that the material (the pages) provides the source of the object as a whole (the book). One of the most common uses of von in modern German is as an alternative for the genitive case, which has fallen into disuse in the spoken language:
SPATIAL METAPHOR IN GERMAN CAUSATIVE 577 (87) das Haus von Herrn Mutter "Mr. Muller's house" The localist interpretation is apparent if we regard possession as a figurative force which emanates from the possessor and envelops his possessions; the cognitive prototype for possession is holding something. (Compare the etymology of English o/in Onions et ah 1966: 624.) The locative use of mit "with" is to express accompaniment. This is a closer relation than mere physical nearness; it implies that two or more individuals or entities undergo the same change of state or carry out the same action together: (88) Ich freue mich mit Ihnen "I am happy for (lit. with) you" (89) Kommen Sie mit mir? "Are you coming with me?" Fillmore (1968: 81-83) relates the "comitative" construction with the preposition with to the coordinate conjunction of noun phrases, proposing the same underlying structure for both constructions, but the three sentences (90) Hans ging mit Maria spazieren "Hans went for a walk with Maria" (91) Maria ging mit Hans spazieren "Maria went for a walk with Hans" (92) Hans und Maria gingen gemeinsam spazieren "Hans and Maria went for a walk together" differ in meaning in a manner that goes beyond differences of topicaliza- tion. Although both Hans and Maria "walk" in the sentences above, only in (92) are they equal "walkers"; in (90) Maria is the principal walker, and is accompanied by Hans; in (91) the relationship is reversed. (Cf. the discussion of "landmark" and "trajector" in Langacker 1982a and 1983.) Similarly, in (88) the person addressed is the primary rejoicer, and the speaker joins in. Semantically, we may say that the person or entity which is the referent of the noun phrase object of mit or with is "closer" to the state or event described than the person or entity which is referred to by the subject of the sentence. Thus mit and with do not describe symmetrical relations in grammatical or semantic structure. The commonality of the notions of accompaniment and instrumentality
578 ROBERT THOMAS KING has long been recognized. In discussing proto-Germanic, Priebsch and Col- linson (1968: 91) state that The so-called instrumental case appears to have been primarily a comita- tive or social case, indicating someone or something in connection or association with another person or object. From this arises the notion of co-operation and hence instrumentality. (See also Nilsen 1973.) Lakoff and Johnson (1980: 134-135) argue that it is no accident that the same preposition is used in English to mark accompaniment and instrumentality, since this is merely a reflection of the metaphor (discussed earlier) that "an instrument is a companion." English and German are "coherent" with this metaphor, which reflects the "logic" of the languages. Lakoff and Johnson, however, base their argument on personification of instruments (giving them names), and on coordinate structures with agents and instruments: (93) Sleezo the Magician and his Magic Harmonica will be performing tonight at the Rialto But coordinate structures and with- (or mit-) constructions are semantically distinct, so Lakoff s and Johnson's examples cannot help us much. Instead, we should look at the semantic properties of with- (or mit-) constructions which express accompaniment and compare them with those which express instrumentality. In the former, the referent of the noun phrase object of with or mit has the following properties: (a) it is located in the same physical or cognitive space as another entity which functions as an agent, patient, experiencer, etc.; (b) it is involved in the same state, process, or action as the other entity; (c) it is semantically "closer" to the state, process, or event than the other entity is (in the terminology of Langacker 1983, etc., it is the "landmark", and the other entity is the "trajector"); (d) it may not be exchanged with the other entity without affecting the truth value of the expression in cognitive terms. An instrument has the following semantic properties: (a') it is located in the same physical or cognitive space as an agent; (b') it is involved in the same activity as an agent; (c') it is semantically "closer" to the activity than the agent is, because it is the means by which the agent affects the patient or causes a change of state of the patient;
SPATIAL METAPHOR IN GERMAN CAUSATIVE 579 (d') it may not be exchanged with the agent without affecting the truth value of the expression; (e') it is controlled by the agent. Since the relation of instrumentality includes all of the properties of the relation of with- (or mit-) accompaniment, we can conclude that the class of instruments is a subclass of the class of companions. The only difference between an instrument and other companions is that an instrument is under the control of an agent. Notice that we have reached the same conclusions as Lakoff and Johnson, but on the basis of a comparison of semantic properties and not on the basis of arguments from personification or coordinate structures, which can be applied only to a few examples of instruments. The locative sense of durch "bezeichnet eine Bewegung durch einen Raum oder Korper von einem Ende zum andern" ("describes a motion through a space or body from one end to the other" — Schmitz 1962: 9). It may be thought of as describing a "three-state journey" (Mackenzie 1977) from one location (the source) via an intermediate location (referred to by the object of durch) to a third location (the goal). The object of durch must refer to an extended space or substance. It cannot be a dimensionless point. The motion must begin and end outside the space or substance. The space or substance must be entered, traversed, and exited. If any of these conditions fail, the preposition durch cannot be used in its locative meaning. The locative use of durch is thus identical with the locative use of its English cognate through. In English and German, the prepositions used to indicate moments or periods of time almost all have locative as well as temporal senses (possible exceptions are seit, since, and for). This usage reflects the metaphorical conception of time as space; we even speak of a "Zeitpunkt" or a "point in time." A change of state is a transition from an initial state at time tj to a final state at time t2, where time tj is earlier than time t2. If time is viewed as a kind of space, then a change of state may be conceptualized as a journey from the initial state (the source) to the final state (the goal). If the change of state is mediated or caused by another state or by an event, the cause is expressed in German as the object of durch (see Section 6): (94) Hans hat sein Vermogen durch Betrug erworben "Hans gained his fortune through fraud" In his initial state at time t , Hans has no fortune; in his final state at time t2, he has a fortune. In the time period intermediate between t2 and t2, he
580 ROBERT THOMAS KING practices fraud. This period may be regarded as an intermediate "location" in time on the journey from the initial to the final state. In this use of durch, the object of the preposition is most commonly a verbal noun or a nominali- zation, one of the prototype combinations of Section 8. If the object of durch denotes a person or a physical object, it represents an event involving the person or object (see Section 10): (95) Hans hat Peter durch einen gemeinsamen Freund kennengelernt "Hans met Peter through a common friend" Hans "goes" from the initial state at time t2 of "not knowing Peter" to the final state at time t2 of "knowing Peter"; the intermediary is the common friend, or more precisely, the act of the friend introducing Peter to Hans, which takes place at a time between \x and tr The medial and causal uses of durch can be given a localist interpretation if we postulate that events be regarded as periods of time (metaphorically, spaces), within which agents may act, instruments may be used, and patients may be affected. A causative expression describes two events, e: and e . Event e2 is typically a transition from an initial state at time "location" tj to a final state at time "location" t2. Event ea takes place in the time interval ("space") between \x and tr If the change of state involved in event e2 is caused by event e1? then e2 is a necessary intermediary. It is not possible to reach the final state "location" from the initial state "location" without "going through" the intermediate event "space", at least at this juncture of time. For example, in the sentence (96) Der Stein wurde durch die Einwirkung des Eises gesprengt "The stone was split by the force of the ice" the initial state is that the stone is whole; in "moving" through time, it is affected by the expansion of the ice (the causing event); through this event space, it reaches the final state of being split. In the localist account of causation, the patient "journeys" from an initial state "location" to a final state "location" through an event space in which it is acted upon by an agent or force. A similar analysis can be applied to all example sentences in this paper which contain causative constructions with durch. 14. A Cognitive-Spatial Account of German Causative Constructions The localist interpretation of the "non-concrete" uses of German prepositions makes it possible to propose a theoretical account of causative
SPATIAL METAPHOR IN GERMAN CAUSATIVE 581 constructions in terms of cognitive-spatial metaphor. This approach offers many advantages over previous treatments in that it provides a common basis for the explanation of what have previously been considered disparate phenomena. It further shows how semantic analysis in terms of entailment relations can be consistent with results from prototype theory. Most of the examples in this paper have dealt with fairly simple constructions involving agents, patients, and instruments. An agent performs an action directed against a patient by means of an instrument; in the case of a causative construction, the patient undergoes a resultant change of state. The action is the causing event (e^; the change of state is the caused event (e2). The evidence from prepositional usage in German supports the following interpretation in terms of spatial metaphor: 1) An event is an extended space; 2) An agent is a point source of energy within an event-space; 3) An action is an event-space within which an agent produces energy; 4) A patient is the goal of the energy emanating from an agent; 5) An instrument is a conduit for the energy transfer; 6) A state is a location in space at which a patient is situated; 7) A change of state of a patient is an event-space within which the patient moves from an initial state-location st to a final state- location s2; 8) In a causative construction, event-space e2 is topologically disconnected; the only path from state-location s1 to state-location s2 is via event space er In cognitive-spatial terms, an agent can be regarded as a point source of energy within the extended space of an event. The energy emanates from the agent, which, however, has no internal structure. Thus the preposition used in German to mark agentivity in passive constructions is the same as the preposition used to mark point or surface sources of movement. The action of the agent is a production of energy; if a patient is mentioned, the energy is transferred from the agent to the patient. A transfer of energy, however, requires a medium; something must carry it. This function is borne by the instrument, which is under the control of the agent: it receives energy from the agent and transmits it to the patient. Therefore the instrument is "closer" semantically to the event than the agent, because the instrument "contacts" both the agent and the patient. The patient is the
582 ROBERT THOMAS KING energy goal; it absorbs the energy emanating from the agent and being carried by the instrument. In a causative expression the patient undergoes a change of state as a result of the agent's action. The change of state of the patient can cognitively be regarded as "movement" from one state location to another state location. The change of state is an event space which contains both the initial and the final state-locations of the patient. If the change of state is caused by another event (an action), then the only way to get from the initial state-location Sj to the final state-location s2 is through the event space of the action. In the terminology of topology, we can say that event space e2 is disconnected (Henle 1979: 30); no point of the subset which contains the initial state-location s: is near any point of the subset which contains the final state-location sr The two subsets are separated by event space el9 and the only connected path from s} to s2 must traverse er (See Figure 1.) Event space e Path of patient I S2 undergoing a y\ change of state ^/ x^ Event space e2 ^^_ _^^ (disconnected) Figure 1: The topological space of a causative construction The cognitive model described in the last paragraph provides a framework for understanding the data which I have collected on German causative constructions (see Sections 8-11). The prototype combinations are those which correspond most closely to the cognitive model. Agent marking with von is consistent with regarding agents as sources. Marking instruments with mit is consistent with the treatment of instruments as controlled companions. Marking causing events with durch is consistent with regarding events as extended spaces through which final states may be reached. The rare or difficult combinations are those which do not at all fit the cognitive model. Since instruments only carry energy, they cannot be sources; thus they do not occur as objects of von, except when they are under the control of impersonal forces such as gravity. The combination
SPATIAL METAPHOR IN GERMAN CAUSATIVE 583 which was most thoroughly rejected in my surveys was von + event. If events are regarded as extended spaces, it is clear that von would be ungrammatical; for a space to be regarded as a source, the correct preposition would have to be aus. Such combinations do occur: (97) Aus Verzweiflung sprang er von der Briicke "Out of desperation he jumped from the bridge" Here the state of desperation constitutes a space from which the motive for the suicide comes. The combination mit + person was almost always interpreted in the sense of accompaniment. This is consistent with the interpretation of instruments as controlled companions; a person is usually not regarded as a possible conduit of energy but as either a source or a goal of an energy transfer. Thus the feature of control by an agent (e') does not apply. But the other four features of instrumentality are identical with the specification of accompaniment. In other words, an object of mit which is not regarded as controlled by an agent is interpreted as a companion. The non-prototypical combinations which were widely accepted provide some of the strongest evidence for the cognitive-spatial account of causativity. The combination mit + event was partly accepted in some cases and almost totally rejected in others. When accepted, it was interpreted not as causation but as contemporality. But contemporality of events is equivalent to companionship: two events which occur in the same time period without influencing or controlling one another can metaphorically be called temporal companions. In cases in which contemporality was difficult to imagine, the informants rejected the combination. The most interesting non-prototypical combinations were durch + agent and durch + instrument. These were invariably interpreted as reduced representations of events, almost as a kind of abbreviated language. Normally, it would be considered impossible to go through an agent, since agents are point sources of energy. But in the combinations durch + agent, the agents were not regarded as the original energy source, but rather as intermediaries. In this respect they serve as secondary sources of energy, and thus do not have the same degree of responsibility as a primary agent. A secondary agent has some of the energy-carrying attributes of an instrument, but it is not a companion of the primary agent, who is not present in the description of the causing event. In the combination durch + instrument, the instrument is not primarily under the control of an agent, but is instead the carrier of some impersonal force. Since there is no agent expressed, the instrument
584 ROBERT THOMAS KING cannot be a companion, and so the preposition mit is impossible. In Section 10 I established that the acceptability of durch combinations in passive constructions correlated almost exactly with the semantic definition of causative: the combinations with durch were widely accepted in causative constructions but almost totally rejected in non-causatives. This correlation also has an explanation in terms of cognitive-spatial metaphor. If a construction describes only an autonomous event with no resultant change of state of a patient, then only one event space is set up; no other state locations are mentioned. Since no other states are described, they cannot be reached through the event space. Since the event space is not traversed from one state location on one side of it to another state location on the other side of it, the preposition durch cannot be used. Furthermore, the participants remain in one event space. They do not enter or leave it. Notice that in the case of a change of state the patient moves from an initial state location, enters the event space of the action, traverses it, and exits it in order to reach the final state location. (See Figure 1.) But consider sentences (67) and (73). The event which is described is the action of Peter striking Hans; there is no other event involving Peter through which he strikes Hans; only one event is described. Since the participants are all enclosed within the one event space, the preposition durch cannot be used. The space is not exited by the patient. (See Figure 2.) Figure 2: The topological space of an action with no resultant change of state 15. Conclusion The study of prepositional usage in German causative constructions has led to several insights into the metaphorical interpretation of semantic relations. The fact that locative prepositions are used to mark relations such
SPATIAL METAPHOR IN GERMAN CAUSATIVE 585 as agentivity, instrumentality, and causativity in German is seen not as a coincidence, but as a reflection of an underlying semantic and cognitive unity: agents are sources, patients are goals, instruments are conduits, and events are spaces. Semantics is topology.
Nahuatl Causative/Applicatives in Cognitive Grammar David Tuggy Summer Institute of Linguistics 0. Introduction A striking feature of Aztecan morpho-syntax is the use of verbal suffixes which function both as causatives and as applicatives.1 In the causative constructions a subject causes a direct object to do the process designated by the verb stem; in the applicatives a subject does the process designated by the verb stem to (or for, or from, or against, etc.) a direct object. These suffixes also form transitive verbs from nouns, adjectives, and postpositions. Tetelcingo Nahuatl (henceforth TN)2 has seven such suffixes; in each case the suffix has a constant phonological shape and constant morphological properties such as position-class in the verb, conditioning of stem-formation rules, and pattern of tense formation, which make it desirable to treat it as one suffix in spite of its different functions and meanings. An important theoretical problem such data raise is this: can causatives and applicatives and the verbalizations associated with these suffixes be analyzed in a way that shows their relatedness, accounting for the great, repeated overlap, or not? Many current theories of syntax do not allow for this: they would lead us to posit a cluster of accidentally homophonous suffixes which are quite separate from each other in terms of their meanings (if in fact they have any) and of their syntactic behavior. Langacker's Cognitive Grammar model (CG — see papers and references in this volume), however, makes possible an analysis which shows close relationships among all these usages, presenting the suffixes in question as integrated (though complex) units rather than arbitrary collections of accidentally homophonous forms. In this paper I present a CG analysis of representative usages of these suffixes, describing the usages in some detail, and setting forth schematic
588 DAVID TUGGY structures which embody the generalizations obtaining among them and show their relatedness. 1. Preliminaries 1.1 Schematic Hierarchies (motor) RUN (animate thing) RUN (cyclic limb motion causing rapid linear motion) Participate in flowing Z^x Be linearly Flow Be channel for flow Z^TS^^X (river) RUN (road) RUN (water) RUN (faucet) RUN (nose) RUN Figure 1: Schematic Hierarchy for run. Lexical items typically have a semantic structure like that in Figure 1, with a core or prototypical meaning or set of meanings, and with related meanings corresponding to different usages. In CG all these meanings are linked to each other by relationships of schematicity, where a schematic concept covers the same semantic territory as its elaboration or instantiation, but in less detail. Thus in Figure 1 the prototypical meaning of run is the notion of a human running; also very salient is the notion of an animal running; and the two are subsumed under a schema which neutralizes the distinctions between them. (Schematicity is symbolized by an arrow from the schematic concept to its elaboration; the degree of prototypicality by the width of the line forming the box around the concept's representation.) The schema neutralizing animal and human running is a sub-case of schemas involving cyclic motion and (rapid) linear motion; under the one it is sister to such notions as that of an engine running, and under the other to such as those of a river or road running. And so forth: even one's nose run-
nAhuatl causative/applicatives 589 ning is not unrelated or unrelatable. Such structures, are, of course, language specific, and essentially unpredictable. This does not mean, however, that they are arbitrary. There is not a strict dichotomy between the arbitrary and the predictable: rather there is a continuum. The absolutely predictable and the totally arbitrary are its theoretical endpoints, but most of language lies somewhere in between, exhibiting varying degrees of unsurprisingness or reasonableness without attaining the inevitability of the truly predictable. Under CG, this same type of structure is expected of morphological and syntactic units. The causative/applicative suffixes will bear a range of meanings corresponding to different usages and related in a schematic hierarchy similar to that in Figure 1, meanings not predictably, but reasonably related. With such an analysis we can certainly distinguish among those meanings, but we can also see them all as related, and the suffix as a single, though complex, unit. 1.2 Causation Causation is a notion crucial to our discussion, so we must deal briefly with the vexed question of its nature. I assume that, like many other complex concepts, it consists in the coincidence in prototypical cases of a number of characteristics, many or perhaps any of which may be violated or instantiated less than fully in non-prototypical cases.3 In the cases we will be discussing, the following two characteristics always hold true; they may be considered definitional for causation: (i) Causation is a relation uniting two relations rather than just two Things.4 Often some Thing in one or other of the component relations is so prominent that, if we factor out all elements below a certain level of salience, it will seem to enter directly into the causal relation. But if we lower the threshold of salience, it will be seen that some relation in which that Thing is involved is what actually does the causing or is caused. These relations we will for convenience call Relation A and Relation B. (ii) The probability of Relation A contributes to that of Relation B; to the extent that A happens or is likely to happen, so does B happen or so is B likely to happen. Two other characteristics that virtually always hold are the following: (iii) The trajector (i.e. the most prominent Thing, the Thing which will correspond to an external subject if there is one) of Relation A is also
590 DAVID TUGGY trajector of the causal relation as a whole. This trajector we call the causer, (iv) The trajector of Relation B is the primary landmark of the causal relation (i.e. it occupies the position of secondary prominence, and will correspond to an external direct object.) There are also a number of other characteristics that are true of the prototypical cases (and the majority even of the non-prototypical cases). Among them are (v) Relations A and B are both processes; (vi) A precedes B temporally; (vii) as a further development on the theme of (ii), A is both necessary and sufficient for B; (viii) the landmark of the causal relation (i.e. the trajector of relation B) is passive or resistant to the actualization of B; (ix) the trajector of the causal relation (and of A) is a sentient being who wills the actualization of B; (x) A involves physical action; and so forth. One final characteristic is (xi): the causal relation itself is construed as a process. This of course is not universal: English because and so construe a causal relation non-processually; however it is true for all the causative/ applicative cases we will be discussing. A 1 time ^caus B 1 1 Figure 2: Causation The causal relation will be represented diagrammatically as in Figure 2, by a two-headed arrow marked "caus";5 it will always exhibit characteristics (i), (ii), and (xi), and (with the exception of Figures 14 and 22) (iii) and (iv); the other characteristics will usually (but may not always) be manifested. Causatives (i) ni-miki I-die "I die" ni-k-mik-tiya I-him-die. perf-caus "I kill him"
NAHUATL CAUSATIVE/APPLICATIVES 591 A typical TN causative usage is given in (l):6 The stem mik-tiya means "kill someone". The following facts should be reflected in our analysis of it: (a) mik is a (perfective) verb stem designating a process of dying. The rest of the meaning of mik-tiya can be presumed to be -tiya's contribution, (b) mik-tiya designates a process in which a causer, prototypically a person, causes another entity, again prototypically a person, to undergo the process of dying (designated by mik). (c) The causer is the trajector of mik-tiya: he will be identified with the subject that is expected to combine with the stem, and (d) the person who undergoes the process of dying is the landmark, and will be identified with the expected direct object. We deduce, then, the following semantic structure for tiya: it designates a process (i.e., it is a verbal element) in which one relation causes another. The trajector of the causing relation (A) is trajector of -tiya as well, and the trajector of the caused relation (B) is its primary landmark. (These are characteristics (i) to (iv) of section 1.2.) When -tiya and mik are combined, the process designated by mik is identified with Relation B. Thus the trajector of mik automatically corresponds to the primary land- Figure 3: mik-tiya
592 DAVID TUGGY mark of -tiya. In this way the nature of the caused relation is explicated: it is one of dying. The nature of Relation A is left unspecified; we do not know exactly what the causer does to cause the landmark to die. Finally, the specifications of -tiya take precedence over those of mik in the composite structure: mik-tiya designates the same causal relation as tiya, with the nature of Relation B elaborated; it does not designate the process of dying profiled in mik. In CG terms, then, -tiya is the "profile determinant" in the structure. All these claims are represented diagrammatically in Figure 3.7 One aspect of mik-tiya that we have not discussed is the fact that it itself is a transitive verb stem. That is: (a) there exists a schematic direct- ob]ect-mik-tiya construction, in which some Thing elaborates the primary landmark of mik-tiya, and (b) mik-tiya is so closely associated with that construction that activating mik-tiya automatically activates the construction. Figure 4: Subject-Object-m//c-frya Construction
nAhuatl causative/applicatives 593 This creates the strong expectation that the construction will in fact be used, i.e. that mik-tiya will occur in construction with a direct object elaborating its primary landmark. This direct-object-mik-tiya construction is diagrammed in the circled portion of Figure 4. This construction in turn (like all other TN verb stems) is closely tied to a subject-verb stem construction (Figure 4), in which some Thing elaborates its trajector, so that a subject is also strongly expected. These expectations I have represented in Figure 3 and subsequent diagrams by cross-hatching the representation of the entities (the trajector and landmark) which are expected to be elaborated; this is to be understood as an abbreviation for claiming that the stems do not exist independently of constructions analagous to those in Figure 4. Specific structures such as ni-k-mik-tiya "I kill him" ((1)), in which the elaborations have in fact taken place, are sanctioned by those schemas. mik-tiya as we have represented it in Figure 3 is a rather schematic version of that stem; there exist various instantiating structures which specify more details of the killing process. In what is probably the prototypical version an episode of shooting causes the death. This presumably comes from an established instantiation of mik, in which shooting is specified as the cause of the designated episode of dying. (Positing such an instantiation is @™ O- O ••••• Semantic / Space / / Phonological Figure 5: mik-tiya-with-a-gun (Prototypical Construal)
594 DAVID TUGGY equivalent to claiming that it is common knowledge that one of the things that may cause death is getting shot.) This non-profiled shooting episode is then put in correspondence with Relation A in -tiya, explicating the nature of that relation. This structure, which directly instantiates that of Figure 3, is diagrammed in Figure 5. There are many other verb stem-tiya constructions in TN which parallel mik-tiya, differing in that some other process is designated by the verb stem and thus identified as the caused Relation B in the composite structure. Some examples are given in (2) below.8 (2) ni-pdki ni-k-pdk-tiya I-rej oice I-him-rej oice. perf-caus "I am happy" "I make him happy" ni-nesi ni-k-nes-tiya I-appear I-it-appear.perf-caus "I appear" "I find/discover it" ni-md-kisa ni-k-ma-kis-tiya I-hand-emerge I-him-hand-emerge.perf-caus "I escape" "I rescue him" ni-koci ni-k-koci-tiya I-sleep I-him-sleep-caus "I sleep" "I put him to sleep" wiksi ni-k-wiksi-tiya cook I-it-cook-caus "it cooks" "I cook it" For each of these stems diagrams parallel to Figures 3-4 could be drawn, embodying analyses directly analogous to that given for mik-tiya. The composite stems often bear semantic specifications which are not (or not as strongly) associated with either the component verb stem or with -tiya. For instance, the prototypical version of kocl-tiya specifies the means of causing sleep (rocking in a hammock-cradle, perhaps singing a lullaby), much as the prototypical version of mik-tiya (Figure 5) specified shooting as the means of causing death. Similarly, nes-tiya (appear-caus) specifies that the causer must be among those to whom the landmark appears; the stem means "discover" or "find", but not "show" or "demonstrate". This is, of course, what CG leads us to expect: complex concepts are not just the product of their components, but typically contain meaning of their own.
NAHUATL CAUSATIVE/APPLICATIVES 595 The generalization subsuming all these structures is a schematic structure with the following characteristics: (a) A verb stem (i.e. a symbolic structure designating a process) precedes -tiya phonologically. (b) -tiya profiles a process in which one relation (Relation A) causes another relation (B) to obtain. The causer (i.e. the trajector of Relation A) is trajector of -tiya as a whole, and the trajector of Relation B is primary landmark, (c) The verb stem is integrated with -tiya by being identified with Relation B, the caused relation. Thus the trajector of Relation B (the primary landmark of -tiya) is identified with the trajector of the verb stem, (d) -tiya is profile determinant. An automatic consequence is that the stem-tiya construction is processual (i.e. verbal), designating the same process of causation that -tiya designates, (e) The stenw/ya construction is transitive. These specifications are diagrammed in Figure 6, the Causative Verb-tiya construction. Semantic Space / Phonological / Space Figure 6: Causative Vevb-tiya Construction (3) ni-k-ihta I-it-see "I see it" ni-k-ihti-tiya I-him-see-caus "I show him (s.t.)"
596 DAVID TUGGY ni-k-mati ni-k-mac-tiya I-it-know I-him-know.perf-caus "I know it" "I teach him (s.t.)" ni-k-mati-s-tiya I-him-know-fut-caus "I announce (s.t.) to him" ni-k-il-namiki-tiya I-him-memory-meet-caus "I remind him (of s.o.)" In (1) and (2) -tiya attaches to intransitive verb stems: it also may attach to transitive stems, as the examples in (3) illustrate. When it does so, the stem-tiya construction is not transitive with respect to the same landmark the stem expected an object for (the Thing seen in ihta, the Thing known in mati, and the Thing remembered in il-namiki.)9 This is a natural (though not inevitable) result of the stems' use in a stem-riya construction: -tiya is profile determinant, and its specifications (which do not include transitivity with respect to the landmark of its Relation B) take precedence over those of the stem. ni-k-il-namiki I-him-memory-meet "I remember him" (4) ni-mewa I-arise "I arise" ni-k-mewi-liya I-him-arise-caus "I raise him" Xa$ilini ring "it (bell) rings" ni-Xaka-ti I-man-inchoative "I am born" ni-k-Xa^ilini-liya I-it-ring-caus "I ring it (bell)" ni-k-Xaka-ti-liya I-him-man-inchoative-caus "I give birth to him" -tiya is not the only causative suffix in TN: in (4) are given forms with -liya, and similar lists could be given for -Itiya, -wiya, -(i)ya, -owa, and -a: For each stem structures similar to those diagrammed in Figures 3 and 4 would be posited, and for each suffix a structure like that of Figure 6, differing only in the phonological shape of the suffix. We can express the commonality of all these causative structures by the highly schematic structure of Figure 7, the CG analogue of a rule of causative formation. This structure has as direct instantiations Figure 6 and the parallel structures for the other suffixes; among these -tiya and -Itiya are prototypical, as they are far and away the most common (and apparently
NAHUATL CAUSATIVE/APPLICATIVES 597 the only productive) structures. These relationships are represented in Figure 8: note the resemblance of this structure to that posited for English run in Figure 1. / / Figure 7: Causative Verb-suffix Construction Figure 8: Schematic Hierarchy of Causative Verb-Suffix Constructions
598 DAVID TUGGY 3. Applicatives (5) ni-tesi ni-k-tesi-liya I-grind.corn I-her-grind.corn-applic "I grind corn" "I grind corn for her" (5) is a typical applicative usage of -liya. The stem tesi alone means "grind corn" (it is intransitive), and tesi-liya means "grind corn for s.o.". tesi designates a process in which the trajector (usually a woman) grinds prepared corn (ne$-tamal-i), which results in someone (usually herself) possessing a quantity of dough (tes-ki) for tortillas or tamales or whatever. This resultant possession is not part of the profile: only the grinding episode is actually designated by the stem. This possessive relation I have termed "resultant"; I could as well have said it is caused by the episode of grinding. It fulfills all the characteristics we have given for causation except (viii)--the possessor of the tes-ki is not usually resistant to coming to possess it. In tesi-liya the same episode of grinding is included in the profile, and the person doing the grinding is still retained as trajector of the composite stem, but the profile is expanded to include the causing of possession. The person who comes to possess the tes-ki functions as primary landmark, in accordance with characteristic (iv), and the stem is transitive with respect to it. If, as we did with -tiya in mik-tiya, we assign to the suffix all the semantic material that is not contained in the stem, we deduce the following structure for -liya: it designates a process in which a Relation A causes a Relation B. The trajector of A is overall trajector, and the trajector of B is overall primary landmark. (These are, of course, exactly the same specifications we posited for -tiya.) When -liya is combined with tesi, the process designated by tesi is identified not with the caused Relation B, but with the causing Relation A. Thus the trajectors of tesi and tesi-liya automatically correspond to each other, and the nature of the causing relation is explicated: it is one of grinding corn. The non-profiled caused relation of possession in tesi is put in correspondence with Relation B, explicating the nature of that relation, just as the non-profiled causing relation of shooting was put in correspondence with Relation A in Figure 5. Finally, -liya (again paralleling -tiya) is profile determinant for the composite structure. These specifications are diagrammed in Figure 9.
NAHUATL CAUSATIVE/APPLICATIVES 599 1 lime r 1 litnc CilUS*' ". —> ► Semantic Space / Phonological / Space Figure 9: tesi-liya There are many other examples of stems with -liya which can be analyzed in a manner directly parallel to that used for tesi-liya. Some of them are given in (6). (6) ni-k-sipewa I-it-peel "I peel it" ni-#ah#i I-shout "I shout" ni-k-kdwa I-it-leave "I leave it" ni-k-kiS-tiya I-it-emerge-caus "I take it out" ni-ka-pafka I-unspec-squeeze "I milk" ni-k-sipewi-liya I-him-peel-applic "I peel him (a fruit)" ni-k-^ah^i-liya I-him-shout-applic "I shout to him" ni-k-kdwi-liya I-him-leave-applic "I leave him (s.t.)" ni-k-kiS'ti-liya I-it-emerge-caus-applic "I take (s.t.) out of it" ni-k-Aa-pdtfki-liya I-it-unspec-squeeze-applic "I milk it (cow/goat)"
600 DAVID TUGGY ni-Aa-temowa ni-k-ka-temowi-liya I-unspec-search I-him-unspec-search-applic "I search (for s.t.)" "I frisk him" In some of the examples in (6) the stem to which -liya is added is transitive; just as with the examples in (3), and for the same reasons, the object with respect to which the stem is transitive is no longer the primary landmark in the composite stem; rather it is displaced by the trajector of Relation B. Note also that the nature of the caused Relation B varies from case to case: sometimes it is possession, sometimes it is loss, sometimes it is benefit, sometimes detriment, the choice depending, to some extent, on what resultant relation is prominent in the semantics of the stem with which -liya combines. About all that is common to all the cases is that the primary landmark is somehow affected (i.e. some relation involving that landmark is caused) by the process designated by the stem.10 To capture the generalization uniting all these forms, we posit a structure like that of Figure 9, but with the semantic and phonological details of the stem despecified. This structure, diagrammed in Figure 10, characterizes the prototypical usage (and thus the prototypical meaning) of -liya. / / Figure 10: Applicative Verb-liya Construction
nAhuatl causative/applicatives 601 (7) ni-kTlka I-sing "I sing" ni-k-toka I-it-bury "I plant it*' ni-k-Xapowa I-it-open "I open it" ni-Xa-poh-polowa I-unspec-rdp-lose "I erase (things)" ni-k-kwika-tiya I-him-sing-applic "I sing to him" ni-k-toki-ltiya I-it-bury-applic "I plant it for him" ni-k-Xapo-lwiya I-him-open-applic "I open it for him" ni-k-Xa-poh-pol-wiya I-him-unspec-rdp-lose-applic "I forgive him" -liya is not the only applicative suffix in TN, however, -tiya is occasionally (though rarely) used in the same way, and certain other suffixes are quite common with this usage. Some examples are given in (7). To the extent that such applicative usage is established with any of these suffixes, it will have a structure like that in Figure 10 associated with it. And a i. A timer time _ caus B mtXM ^^^ ^^P ► Figure 11: Applicative Verb-Suffix Construction
602 DAVID TUGGY schema at the same level of abstraction as the Causative Verb-Suffix structure of Figure 7 will capture the generalization uniting all these Applicative Verb-Suffix structures. This schema is diagrammed in Figure 11. 4. The Causative/Applicative Schema The Causative and Applicative Verb-Suffix constructions (Figures 7 and 11) have a great deal in common. In both cases the suffix profiles a causation process, with the trajector of Relation A as overall trajector, and the trajector of B as primary landmark. The suffix rather than the stem is profile determinant. The composite structure is transitive, i.e. it carries with it the expectation that the landmark (as well as the trajector) will be elaborated syntagmatically. The difference between the two constructions is whether the stem is identified with the causing Relation A (with the natural consequence that its trajector corresponds to that of the suffix and thus to the trajector of the composite structure), or with the caused Relation B (with the result that its trajector corresponds to the landmark of the suffix and therefore of the composite structure). Figure 12: Causative/Applicative Verb-Suffix Construction
NAHUATL CAUSATIVE/APPLICATIVES 603 Figure 12 represents the schema which embodies the commonality of these two structures. The ease with which this schema can be extracted is a very important result; we have here a direct and powerful expression of the close relationship of causatives to applicatives. Figure 12 represents the generalization uniting all the causative/ applicative suffix usages. Instantiating it are schemas for each of the suffixes that function both as causatives and applicatives, like Figure 12 but with the phonological shape of the suffix specified. Thus there is a Causa- tive/Applicative-//ya construction, which has as its prototypical instantiation the Applicative-//^^ construction (Figure 10) and as a less salient instantiation the Causative-//ya construction. Similarly the Causative/Applicative- tiya schema will generalize over the prototypical Causative-f/ya construction (Figure 6) and the less strongly entrenched Applicative-f/ya construction. A schematic hierarchy exhibiting some of these relationships is given in Figure 13. Figure 13: Schematic Hierarchy of Causative/Applicative Verb-Suffix Constructions
604 DAVID TUGGY 5. Cases Intermediate between Causatives and Applicatives The relationship between causatives and applicatives is seen to be even closer when certain intermediate cases are considered. We here will examine two types of such cases:11 the "because-of" cases, and cases where a structure can reasonably be analyzed as either a causative or an applicative or as both at once. 5.1 "Because-of' Constructions The "because-of" cases include such forms as ni-k-we$ki-liya (I-him- laugh-applic?) "I laugh at him" or ni-k-coki-tiya (I-him-cry-applic?) "I mourn him". To the extent that in laughing at someone or mourning him one intends that person to hear, these stems could be viewed as applicatives of the communication variety (parallel to e.g. ni-k-$ah$i-liya "I shout to him" in (6) or ni-k-kwika-tiya "I sing to him" in (7)), with perhaps a twinge of the malefactive idea in the first case. However, I would claim that at least equally important, if not primary, is the idea of laughing or crying because of the person laughed at or mourned.12 "Because of" is essentially the same notion as "cause", but with the trajector and landmark flipped: i.e. "X because Y" is almost identical in meaning to "Y cause X". We can, then, represent the structure of a "because of" suffix by the familiar causation schema, but instead of the trajector of Relation A being overall trajector it is the landmark, and the trajector of B is overall trajector.13 The process designated by the stem (crying or laughing) is identified with the caused Relation B (i.e. it is construed as caused by something else). As in the case of m/A>frya-with-a-gun (Figure 5), the particular stem involved will include non-profiled (background) information about what is likely to cause its occurrence: people in the TN culture cry when someone dies, and they laugh when someone looks, does something, or says something, ridiculous. Such information is identified with the causing Relation A (again after the manner of Figure 5), and the resultant composite structure has some relation pertaining to a landmark (his death, or his ridiculousness) causing a process of crying or laughing to occur, with the trajector of that process being overall trajector. This analysis for coki-tiya is diagrammed in Figure 14.
NAHUATL CAUSATIVE/APPLICATIVES 605 / Figure 14: cokl-tiya The structure of Figure 14 shares with causatives (Figure 7 and its instantiations) the identification of the stem with the caused Relation B rather than the causing Relation A. With applicatives (Figure 11 and its sub-cases) it shares the identification of the stem with the relation whose trajector is overall trajector. Thus, like causatives, the process designated by the verb stem is construed as caused; like applicatives, the trajector of the verb stem is trajector of the overall structure as well. These insights can be captured by including in our schematic hierarchy the network diagrammed in Figure 15.
606 DAVID TUGGY Causative Verb-tiya | (Figure 6) Figure 15: Schematic Hierarchy for Causative, "Because-of", and Applicative 5.2 Multiple and Mixed Construals In certain cases, a causing action is essentially of the same type as the action it causes. Whenever this is the case, a causative of the action can also be construed as an applicative. To take an English example, when you befriend someone, are you making him your friend, or are you being his friend? In German be-atmen (be-breathe) "give artificial respiration mouth- to-mouth", do you breathe for (or to) the other person, or do you cause him to breathe? Some TN stems can be viewed in this ambivalent way as well. We analyzed mac-tiya "teach" as a causative (know-caus). Yet when you consider what the nature of the causing Relation A is, a salient component of it is the teacher's knowledge. To that extent an analysis is reasonable where mat "know" is identified with Relation A rather than, or as well as, with B. A very interesting form in this light is ni-k-asi-tiya (I-it-arrive-caus?/ applic?) "I arrive with it". Here we have again the familiar A-causing-B structure, but with both the causing action and the caused action being arrivals. Thus the stem elaborates Relation A (I arrive), and also Relation B (it arrives), with Relation B still separate from Relation A (it is a different instance of arrival, with a different trajector) and caused by it. Interestingly, Characteristic (vi) of Section 1.2 is violated: the causing process coincides with rather than precedes the caused process. This construal is dia-
nAhuatl causative/applicatives 607 grammed in Figure 16; it should be clear that it is an instantiation of both Causative and Applicative Verb-Suffix constructions (Figures 6, 7, and 11), further underscoring their close relatedness. / / Figure 16: asi-tiya 6. Verbalizing Usages -tiya and -liya and the other causative/applicative suffixes are often suffixed to non-verbal stems, converting them into transitive verbs. Space prohibits more than a quick overview of a few typical constructions. Yet it will be clear that these are closely related to the causatives and applicatives we have been examining: it is anything but arbitrary that the same suffixes serve both usages.
608 DAVID TUGGY 6.1. Verbalizations of Stative Relations (8) weyi ni-k-weyi-liya big I-it-big-caus "big" "I enlarge it" pi#o-ti-k ni-k-pi#o-ti-liya pig-adj-adj I-it-pig-adj-caus "ugly, dangerous" "I uglify it" se-sel-ik ni-k-tla-se-sel-tiya rdp-soft-adj I-him-unspec-rdp-soft-caus "(very) calm, mild soft" "I pacify him" In (8) are several verbalizations of adjectives. CG considers adjectives, adverbs, and adpositions to be stative, or non-processual, relations. Thus, for instance, the adjective weyi "big" profiles a static "in" relation between the furthest extension of a trajector along some relevant dimension(s), and the region beyond the norm for that/those dimension(s). The composite stem weyi-liya, however, is processual rather than stative: it profiles our familiar process of causation, with the caused relation specified to be one of bigness.14 As the diagram in Figure 17 makes clear, this could be a subcase of the Causative Verb-Suffix construction (Figure 7), except that the stem is a stative rather than a processual relation. We would posit, then, a Causative Stative-Relation-Suffix structure, like Figure 7 but without the profiled time arrow in the semantic specification of the stem. This would of course be joined with Figure 7 under a still higher schema stating only that the stem profiles a relation but not specifying whether it is processual or stative. Applicative-type verbalizations are rare, for complex reasons (probably having to do with the difficulty of perceiving non-processual relations as causing things to happen), but they do exist.15 ni-k-tikawi-liya (I-it-strong- applic) means "I light into it (a job), work hard(er) at it". The job (the overall landmark) is not caused to be strong; rather the trajector is (or waxes) strong, which affects how the job gets done. ni-k-dikdwi-liya should be compared with ni-k-cikdw-a, in which the same stem is verbalized by the suffix -a. ni-k-cikdw-a has two meanings: one is a causative "I strengthen/toughen it"; the other is an applicative "I get tough with him, I resist him". This latter construal is very similar to that of ni-k-cikdwi-liya, but it differs in the nature of the landmark (a person rather than a job) and what the effect is on the landmark (frustration of intentions
NAHUATL CAUSATIVE/APPLICATIVES 609 Scale of 5£k| Semantic / Space / / Phonological f Space weyi — liya Figure 17: weyi-liya rather than being accomplished). Both are reasonable usages for an applicative verbalization of the concept "strong"; neither is strictly predictable: they are meanings associated by convention with the composite structures rather than with the components. (9) n-is-pa(n) ni-k-is-pan-tiya my-eye-on I-it-eye-on-caus "in front of me" "I exhibit, display it" no-ma-k ni-k-ma-k-tiya my-hand-loc I-it-hand-loc-caus "in my power" "I surrender it" In (9) are two causative verbalizations of postpositions. The complex postpositions is-pan and md-k are transitive stative relations, expecting as object a possessive prefix, and as subject a noun or verb phrase. When verbalized they are construed as caused to occur, and the trajector (the Thing
610 DAVID TUGGY that is in front of the landmark, or in the landmark's power) is construed as the primary landmark of the verb. As usual, the trajector of the causing relation is trajector of the verb as well. The landmark of the postposition is not a central participant of the verb; it is not transitive with respect to it. This exactly parallels the causatives of transitive verbs in (3); the only difference is that those are processual relations and these are stative. We simply include the examples in (9) as further sub-cases of the Causative Stative- Relation-Suffix schema. 6.2. Verbalizations of Nouns (10) pantalo(n) trousers "trousers" Xal-i earth-abs "earth, land, dirt" tokd-yo-X name-possd-abs "name" ndmik-Xi husband-abs "husband" yeydn-Xi safe-abs "treasury,safe" atemi-k louse-abs "louse" ni-k-pantalon-tiya I-him-trousers-caus "I betrouser him" ni-k-Xdl-tiya I-him-earth-caus "I endow him with land" ni-k-Xdl-iya I-it-earth-caus "I put it down (on the ground)" ni-k-tokd-yo-tiya I-him-name-possd-caus "I name him" no-ndmik-tiya I. refl-husband-caus "I (woman) get married" ni-k-yeydn-tiya I-it-safe-caus "I put it away safely" ni-k-atemi-liya I-him-louse-caus?/applic? "I de-louse him" In (10) are given a number of noun verbalizations, all of which may be plausibly classified as causative. In the semantics of each noun are backgrounded relations in which it typically figures as landmark: some such relation is identified with Relation B of -tiya, and the Thing profiled by the noun stem is identified with the landmark of B (which is a sub-landmark
NAHUATL CAUSATIVE/APPLICATIVES 611 with respect to -tiya as a whole). For instance, an important (though unpro- filed) specification of pantalon is that the designatum is worn over the legs and lower trunk of a human. That relation of wearing is, in pantalon-tiya, construed as caused, with the person causing the wearing taken as trajector, and the person wearing the trousers as landmark. This construal is diagrammed in Figure 18. Figure 18: pantalon-tiya The other forms can be analyzed similarly. In kal-tiya the relation of owning land is caused; in kal-iya a different relation of earth to a person is construed as caused. The difference between the two meanings is a matter of convention: both are reasonable; neither is strictly predictable. In toka- yo-tiya the relation is the conventional association of a name with a person, in ndmik-tiya it is the marriage relation, in yeydn-tiya it is a locative (and protective) relation; in atemi-liya it is the cessation of a locative (and detrimental) relation that is caused. (The contrast between e.g. lal-tiya "give land", and atemi-liya "take away lice", is paralleled by the contrast in English between be-ribbon and be-head, or between planting and weeding a gar-
612 DAVID TUGGY den.) All of these construals may be subsumed under a schema which we can call the Causative Noun-Suffix construction: it is diagrammed in Figure 19. In calling this a causative construction, we are not claiming that the Thing designated by the noun stem is construed as caused (i.e. caused to exist), but rather that it functions within the relation (Relation B) which is caused. Semantic / Space . / Phonological ' Space Figure 19: Causative Noun-Suffix (11) no-mah-pil my-hand-child "my finger" mawi-lis-Xi respect-noun-abs "fear, respect" ni-k-mah-pil-wiya I-him-hand-child-applic "I beckon to him (with my finger)" ni-k-mawi-liya I-him-respect-applic "I fear/respect him"
NAHUATL CAUSATIVE/APPLICATIVES 613 Applicative verbalizations of nouns, like those of adjectives, are not easy to come by; the examples in (11) are fairly clear, however. In each case the Thing designated by the noun stem appears to figure more saliently in the causing relation than in the caused relation. Thus in mah-pil-wiya the trajector does something with his finger, which causes the landmark to understand a message that the trajector desires to convey. Similarly in mawi-liya the trajector feels respect or fear: this results in his treating the landmark appropriately. Figure 20 gives a schema for such applicative-like usages. / TT77UJ / I I I I / / Figure 20: Applicative Noun-Suffix Actually many causative noun-suffix structures may be construed under this applicative construal also. For instance, in atemi-liya the louse is not only the landmark of the caused relation of freedom, it is also the landmark of the causing relation: the trajector directly manipulates the louse in order to free the primary landmark of it. Similarly the trousers in pantalon- tiya figure to some extent in the causing relation as well as the caused: they
614 DAVID TUGGY typically come to be on the landmark by the trajector directly handling them in order to put them there. Similarly in the examples in (11), the finger and the fear may be viewed as fairly important elements in the relation which is caused: the landmark sees the finger, or senses the respect. It is thus a matter of degree which of the two analyses is prominent for these forms, and the close historical and synchronic association of causatives with applicatives becomes even more understandable. 7. Summary and Conclusion Figure 21: Schematic Hierarchy of Causative/Applicative Stem-Suffix Constructions All of the forms we have examined can be united into one great schematic network, a portion of which is represented in Figure 21. The causative and applicative structures are all very closely related semantically in a structure of the same sort as that uniting the different senses of run in Figure 1. The topmost schema, which is represented in Figure 22, embodies the commonality of all the usages, which is as follows: The suffix profiles a causation process. The stem is somehow integrated with that process: just how is of course the main parameter along which the usages differ. The suffix is profile determinant: the stem is fitted into its specifications rather than
nAhuatl causative/applicatives 615 vice versa. The composite structure is transitive, expecting to be used in an object-stem (and a subject-object-stem) construction. In the vast majority of cases the trajectors of the two Relations A and B are trajector and landmark of the suffix, respectively (with "because-of" forms like coki-tiya, Figure 14, being exceptions.) Beyond that the main split is between cases in which the stem is associated (primarily) with the caused Relation B (causa- tives) or the causing Relation A (applicatives). While such differences clearly exist and are clearly expressed, the whole structure is also clearly unified. Figure 22: Causative/Applicative Stem-Suffix The CG framework has been important to achieving this result. No other framework I know of offers the flexibility provided by the interactions of such notions as conventionality, prototypicality, schematicity, encyclopedic meaning, profile vs. base, internal specification of participants (e.g. trajector and landmark) as well as their external counterparts (subject
616 DAVID TUGGY and object), and multiple sanction, which are all crucial to various aspects of the analysis. Most other theories would either not capture the similarities or not express the differences nearly as well.16 Under the CG analysis these suffixes are very natural pieces of human language, and their historical development and persistence need not be viewed as accidental. Notes 1. Similar phenomena occur elsewhere, notably in other Uto-Aztecan languages. Langacker (1977:144 ff.) suggests five proto-forms, three of them ancestral to suffixes discussed here. Germanic be- (as in English bespatter, be-wail, be-ribbon, German be-enden "cause to end" and be-kochen "cook for s.o."), Indonesian -kan, and Lushai high to low tone shift are further examples; Comrie (1981a:177) speaks of the phenomenon as widespread, and gives Wolof -al as an example. 2. TN is spoken in the town of Tetelcingo and two adjoining colonias, about five miles north of Cuautla, Morelos, and about sixty miles southwest of Mexico City. Works describing it include Brewer and Brewer (1962), Pittman (1948, 1954), and Tuggy (1979, 1981). 3. Thus big in the prototypical case involves exceeding the norm for size along all three spatial dimensions; yet a physically big man could be short, as long as he was much wider and heavier than normal, or thin, if he was very tall, etc. 4. This concurs with e.g. Dowty (1972:62) "CAUSE invariably takes a sentential subject rather than an individual", or Shibatani (1976a: 1) and Talmy (1976:53), for whom causality holds between two Events (i.e. processual relations: cf. my Characteristic (v) below). 5. The following traditional CG diagramming conventions should be mentioned: circles represent the schema Thing, circles (or boxes) connected by a line represent the schema Relation. A box enclosing other symbols represents unit status of the enclosed material. Bold-facing indicates profiling (designation). The label TR identifies the trajector of a relation; LM identifies the primary landmark; tr and lm indicate sub-trajectors or sub- landmarks. A bold-faced arrow labelled "time" marks processes. Further conventions will be noted as they are introduced. 6. The following abbreviations are used in glossing: abs = absolutive; adj = adjectival ending; applic = applicative; caus = causative; perf = perfective; possd = possessed; rdp = reduplication; refl = reflexive; s.o. = someone; s.t. = something; unspec = unspecified object marker. The third person singular non-honorific object marker klki- is glossed "him", "her", or "it", according as a human, female human, or non-human object is judged more likely. All TN words are pronounced with penultimate stress. The "long/ short" marking on vowels is phonetically misleading but cross-dialectically helpful. A is a voiceless lateral tl affricate; the other orthographic symbols should be unproblematic. 7. The following conventions are illustrated in Figures 3 and ff.: semantic space is divided off from phonological space by a dashed line. Symbolic units are represented by a solid line connecting a unit in semantic space with one in phonological space. Co-occurring units are ranged side by side, and the manner of their integration is indicated by dotted "lines of correspondence", indicating identification of the entities joined. Arrows (as in
NAHUATL CAUSATIVE/APPLICATIVES 617 Figure 1) indicate a relation of schematicity, with the instantiation at the head and the schema at the tail of the arrow. As explained below, cross-hatching is used to indicate the expectation that an entity will be elaborated by a syntactically linked symbol: e.g. the cross-hatched trajector of mik in Figure 3 indicates that mik expects to be constructed with a subject, and the cross-hatched Relation B of -tiya indicates that -tiya expects to be put in construction with a stem which elaborates that relation. Composite units are represented above their components, with lines of correspondence and schematicity arrows indicating the manner of their integration. Profile determinance is indicated by an arrow of schematicity from a component to the composite structure, and also (as is traditional) by boldfacing the box enclosing the component. Relatively elaborate concepts like mik "die" are represented by crude diagrams hopefully iconic enough to certain visual aspects of the extremely complex specifications to jog the readers' minds enough for identification. They are at least as good as writing something like DIE in capital letters; better in that they permit identification of trajectors and landmarks, which will be important to us. Stick-men are used to represent Things that are prototypically human. A succession of dots indicates schematic (unspecified) phonological content. The integration of the phonological components and their relationship to the composite phonological structure is indicated only by straight lines; it is in many ways parallel to semantic integration, but it is not particularly relevant to our concerns and is thus ignored. 8. Several morphophonological alternations are exemplified in (2), including palatalization of apical obstruents, "lengthening" of the vowel /, and formation or failure to form a (truncated) perfective stem for use with -tiya; future examples (e.g. in (3)) will show a stem-final a changing to /. Most of these changes do not seem to correspond systematically to any discernible semantic differences, and they may all be ignored for our purposes. 9. These landmarks may be elaborated by clausal constructions (as can other secondary landmarks), but there is no strong expectation that they will be: the verbs as they are given form perfectly acceptable clauses without any such secondary objects. 10. A number of complexities in the applicatives are here glossed over. In particular some applicatives need not be analyzed as involving extension of the profile to include a previously non-profiled caused relation, but may be construed as simply switching the primary landmark status to an otherwise secondary landmark, ni-k-kis-ti-liya "I take (s.t.) out of it" in (6) above is an example. This and other matters are discussed in Tuggy (1981:419ff.). 11. There are more: see Tuggy 1981 (e.g. 444 ff.) for discussion. 12. Sullivan (1976:216-217) makes this ("action motivada por alguien") the third meaning of the Classical Nahuatl applicative; the other two are "action which benefits someone" and "action which deprives someone". 13. This is of course a non-prototypical construal of causation, violating Characteristics iii and iv of section 1.2. It is also somewhat natural for these particular cases: causers (and trajectors generally) are typically volitional, wishing the accomplishment of the process they cause (Characteristic ix); in these cases, the ridiculed person usually does not, and the dead person presumably never does, want to be laughed at or cried for. Thus they are less than prototypical causers, and weaker candidates for trajectorhood.
DAVID TUGGY The norm, as is nearly universal in such causatives, is equated with the LM's previous size, weyi-liya does not mean "make it big (absolutely)", or even "make it bigger than the norm for its type", but "make it bigger than it was." The pattern appears to hold cross-linguistically. Germanic be- is usually applicative when used on verbs; with adjectives and nouns it is strongly causative: e.g. be-little, be-foul, befool, be-doctor. In Tuggy (1981:472-484) a Relational grammar analysis of some of these data was presented. Causatives and applicatives had to be produced by very different syntactic mechanisms (Clause Union and Advancement to Direct Object, respectively); in the one case the suffix was a predicate meaning "cause"; in the other it was a marker introduced by the syntax, showing that Advancement had occurred. Many forms (e.g. the "because- of" forms and most of the verbalized forms) were not produced by the syntax at all, but were consigned to the lexicon. Instead of there being one suffix -tiya or -liya, embracing a number of closely related meanings, there were accidentally homophonous suffixes for the two types, with other accidentally homophonous suffixes for cases relegated to the lexicon. The theory is undergoing some promising changes since then, but the problem still remains: causatives and applicatives are, and must be treated as, separate kinds of syntactic beasts, and there is no explanation for why a language would use the same forms to code them. The Government-Binding theory, insofar as I understand it, would have the same problem accounting for such data. Comrie (1981a: 176) sees them as unified in that they both increase the valence of the verb to which they attach: that is fine as far as it goes (though some cases change the valence rather than increase it), but it does not express all the similarities (e.g. the importance of causation to both notions is not expressed, nor the essential sameness of the composite structure whether the suffix is placed on a verb, adjective, postposition or noun) nor does it account sufficiently for the differences.
PART III A HISTORICAL PERSPECTIVE
Grammatical Categories and Human Conceptualization: Aristotle and the Modistae Pierre Swiggers Belgian National Science Foundation/University of Leuven 0. Introduction Cognitive linguistics — defined as the study of human language in its relation to perception and knowledge — has a long-standing tradition, with three salient pre-XXth century landmarks: Aristotelian hermeneia (the expression or, better, interpretation of a thought), mediaeval speculative grammar, and XVIIth and XVIIIth century philosophical grammar (gram- maire generale or raisonnee). It is interesting to note that each of these major achievements is coupled with a semiotic theory, the core-logic of which lies at the basis of a broad scientific program. This is at the same time an empirical proof of the intrinsic validity of a cognition-based linguistic theory: its axioms and theorems have a scope extending beyond the description of language, and the thus realized understanding of language at the same time helps us to unravel basic processes of human thinking. The purpose of this paper is to offer an analysis of the two first mentioned landmarks in the history of cognitive linguistics. Elsewhere I have dealt with the principles, goals and achievements of XVIIth and XVIIIth century general grammar.1 The present account of the linguistic theory of Aristotle and the modistae is intended to show how many affinities there are between their views and those of modern cognitive linguistics. Apart from the basic similarities in viewing meaning as conceptualization restructured in terms of a conventional linguistic system, there are other striking parallels: with respect to the semantic load of content units in their relation to the accidentia of things; with respect to the semantic definition of parts of speech (in this, modistic grammar is much closer to cognitive linguistics than to generative grammar), and with respect to the view of semantic
622 PIERRE SWIGGERS information as contained within grammatical patterns. The universalistic claims of the modistae's grammatical theory should not so much be compared to those of Chomskyan grammar,2 despite similarities on the surface; Chomskyan grammar has hardly investigated the content of word classes, and this was the central concern of the modistae. Moreover, modistic grammar is not a generative algorithm for the concatenation of formatives, and for operations on linear sequences, but constitutes an in-depth account of the hierarchical principles and schemata through which meaning is embedded in lexical units and their propositional combination. Modistic grammar also provides an interesting example of how an experientalist cognitive theory can be erected on an objectivist ontology.3 On the other hand, there still remain divergences between modistic grammar and modern cognitive linguistics. Prototypical semantics, for instance, is only implicit in modistic grammar, but as shown by the recurrent examples given as "models" for word classes or constructions, it is not entirely absent from modistic theory. The latter lacks a theory of mental spaces and of imaginative schemata or processes: for the modistae schematization is nothing but categorization. Modern cognitive linguistics goes well beyond that, stressing the importance of various image-schemata (containers, paths, links, part-whole, etc.) and imaginative processes. This divergence should be seen in the light of an important difference in heuristic strategy and explanatory teleology between modistic grammar and modern cognitive linguistics: whereas the latter purports to explain grammar as a symbolic structure, through the analysis of discursive patterns and units, modistic grammar deals with grammatical structures as formal frames conveying conceptual meaning, without integrating their discursive use. But — as will be shown in the following — beneath these divergences there remain basic similarities. 1. Language and Cognition: Aristotle's Approach Aristotle, contrary to Plato's treatment of linguistic utterances in terms of truth values, approached language from the semiotic point of view: language is the symbolic expression of psychological states (ta pathemata), and these concern our relation with the outer world. This relation, which finds its expression in conventional signs, cannot be understood in isolation from the set of underlying cognitive principles which articulate human perception and knowledge.4 Human language, therefore, is situated at the cross-roads of psychology (the psyche being the receptacle of sensory images), epis-
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 623 temology (all perceptions are categorized according to invariant cognitive principles), and logic (the expression of our mental states is/should be articulated according to the general formal laws of thought).5 It is important to stress here that Aristotle provides room for language phenomena which cannot be studied from the point of view of propositional logic: not every discursive unit is an assertion, and it makes little sense to analyse prayers in terms of truth and falsity (Peri Hermeneias 17a3-7).6 Now, language as a symbolic expression of experience hinges on the possibility of categorization: this categorization is a flexible one in that it allows the speaker to adapt — e.g. through the use of a past or future tense — his statement to the flux of reality. It is interesting to note that this flexibility is marked by the accidentia (tense, mood, case), which characterize content-terms such as verbs and nouns.7 Both the content-terms and their accidentia are the symbolic correlates of basic ontic categories. Aristotle (Topics 1,9) recognizes ten of these: substance, quality, quantity, relation, place, time, position, state, action and passion (for a detailed treatment, see Aristotle's Categories). These constitute the fundamental parameters with respect to which the terms of a proposition can be described. The proposition itself, as a meaningful combination of terms, must satisfy the conditions of higher-order predication, and should correspond to one of the four modes of predicative attribution (i.e. as a definition, or as the signalling of the accident(s), the species and/or the proprium). Such a theory presupposes a theory of knowledge, accounting for the possibility of (universal) categorization. Here, Aristotle claims that there is an in-built mechanism of universal principles which is, logically and causally, anterior to the individual acquisition of knowledge.8 In other words, individual ontogenesis of cognitive contents is based on pre-existing phylogenetic principles of knowledge (causality, relative order, consequence, etc.). The function of these principles is to allow human beings to abstract away from particulars, and to recognize essence as distinct from (forms of) existence. The latter distinction cannot be arrived at through the body of knowledge accumulated by the various sciences: it is only through a holistic vision of reality, in which essence and existence are intrinsically tied up to each other, that the essence of things can be defined (Metaphysics 1025b 1-18).
624 PIERRE SWIGGERS 2. Mediaeval Speculative Grammar Aristotle did not elaborate a linguistic theory, and his grammatical apparatus was not very sophisticated. But it was his merit to have correlated language with cognition, and to have laid the foundations of the speculative grammar of the modistae.9 This group of scholars, who produced an important number of treatises between 1250 and 1350, were highly indebted to Aristotle's work, made accessible in the Latin translations of Boethius and William of Moerbeke. Aristotle's work was a crucial influence together with that exerted by philosophers such as Abelard and William of Conches. It is thanks to the speculative grammars produced by the modistae that linguistic categories were seen as corresponding to those structuring reality and our apprehension of it, both materially and formally.10 The result of this systematic reflection on grammar was the elaboration of an impressive theory and metatheory. The central concept within modistic theory is that of modus significandi: every linguistic entity has a modus significandi, conferred on it by the intellect, which in turn is correlated with a modus essendi of the thing referred to, and with a modus intelligendi of the apprehending subject. These relationships must be viewed in a dynamic way: first, because each of the above-mentioned modi can be viewed from the reference point of the intellective and signifying act,11 and also from that of the act itself (and its linguistic expression).12 Secondly, there is no strict determination imposed by the ontic mode: substantial properties can be attributed to actions (cf. nominalized infinitives), to syncategorematic terms (in metalinguistic utterances), and they can even be attributed to non-objects (e.g. nothing). The crucial point then is to understand the role of the intellect in this semiotic process, going from the material object-in-the-world (or extra animam) to linguistic forms. With respect to this key problem, the doctrine of the modistae underwent an important change. According to the first generation of speculative grammarians, there is a causal, but non-deterministic chain starting from the object endowed with its properties and going to the signifying units of language. The object presents itself to the apprehending intellect, which elaborates a conceptual representation. This process is made possible through an inherent intellective capacity, the ratio intelligendi.13 As to the linguistic expression, this also results from the activity of the intellect, and it consists of meaningful sound stretches (voces sig- nificativae). It should be noted that within this view the role of the intellect
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 625 is more passive than active;14 as a matter of fact, the intellect acts post rem, and its major role lies in the association of a mode of being and a mode of signifying, and in abstracting away from particulars. Nulla res extra realiter loquendo applicabilis est pluribus, sed intellectus concipit earn ut applicabilem pluribus (...) et sic res intellecta habet modum intelligendi per modum communis, et ab hoc modo intelligendi accipitur modus significandi per modum communis (Martin of Denmark: 20-21). Later generations of modistae — chiefly Siger of Courtrai — insisted on the active part taken by the intellect. It is the intellect that seizes upon the properties of the things-in-the-world, and this grasping is achieved by the modus intelligendi activus.15 As rightly noted by Bursill-Hall (1971: 93), "the mind apprehends the property, i.e. modus essendi of the thing, and it is by means of this that the intellect can signify and comprehend the properties of the thing". The activity of the modus intelligendi activus is pre-gram- matical in that it extracts significance out of experience without relating it to content segments on the linguistic level. Next to the modus intelligendi activus, there is a passive intellective mode, the content matter of which is the property of the thing: Modus intelligendi passivus est ipsa proprietas rei, prout ab intellectu apprehensa (Thomas of Erfurt: 142). In other words, the passive intellective mode is the property of the thing insofar as it is intelligible. The active and the passive modes of intellection are aspects of the same process, and they are formally identical as manifestations of the ratio intelligendi. Materially, however, they are very different: the passive mode of intellection is constituted by the property of the thing (as it is understood by the mind), whereas the active intellective mode is in substance a property of the intellect.16 We have seen that the ratio intelligendi stands midway between the modes of being and symbolic expression. The latter level, which brings in grammaticalization, divides, once again, into an active and a passive mode of signifying. Whereas the latter refers to the thing (or modus essendi) as it is signified (hence the material identity of the modus essendi, the modus intelligendi passivus and the modus significandi passivus), the active mode of signifying is constitutive of the precise linguistic expression of concepts. The dimensions of the active mode of signifying are (1) essential vs. accidental (modes) (2) absolute vs. respective (modes).
626 PIERRE SWIGGERS The essential mode can be further subdivided into submodes, such as general vs. specific, or in a triadic scheme: most general — subaltern — (most) specific. The essential mode is designed to express the essence of the pars [= part of speech, P.S.] in question, while the accidental mode is designed to treat those attributes of the pars which do not belong to its essence, though they occur by reason of its essence. The absolute mode refers to features of the pars which render it discrete from the other partes but without resorting to comparison with features of another pars. The respective mode deals with those features of the pars which permit it to operate in significant concord, both governmental concord and concord of gender, number, case, etc., with other partes orationis at a level higher than that of the word-class (Bursill-Hall 1971: 99-100). It is important to bear in mind that the modus signijicandi not only ensures symbolic expression, but that it achieves grammatical functionalization in three basic respects: (1) by linking a notional meaning with a linguistic segment (2) by adding to this segment a formal (grammatical) meaning (3) by establishing principles of combination between various segments or, in other words, by rendering the segments morphosyn- tactically functional. The views of the modistae on the semiotic process leading to meaningful speech are summarized in Figure 1. RATIO 1NTELL1GEND1 modus intelligent activus - conceptualization modus significandi activus grammatical functionalization THING (properties) INTELLECT WORD linguistic properties real level: modus essendi (inherent properties) primary grammatical semantico-syn- meaning meaning tactic valence relational properties: - intelligibility (modus intelligendi passivus) - expressibility/effability (modus significandi passivus) Figure 1
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 627 The grammarian's preoccupation is with the linguistic level, as it is shaped by the modus significandi activus. This level is constituted by functionally marked words, viz. the parts of speech. These result from the combination of a ratio significandi and a ratio consignificandi, a distinction which parallels that between ratio intelligendi and ratio cointelligendi. The latter refers to the apprehension of things and their properties (these are "co- apprehended" with the things: intellectus intelligit rem cointelligendo eius proprietates, the intellect understands things while co-intellecting their properties). The distinction between ratio significandi and ratio consig- nificandi is of utmost importance because it allows to separate the level of content units (corresponding to what today we call lexical morphemes), and that of grammatically functioning words or parts of speech (combinations of lexical and grammatical morphemes). As a symbolic expression of an entity, the part of speech signifies a thing (with its properties), but essentially it is a grammatical unit. As such the pars orationis consists of three hierarchic components: (1) linguistic shape (written /vocal materiality); corresponding term: vox. (2) significance (primary meaning);17 corresponding term: dictio. (3) class-meaning and grammatical properties: at this level we have a fully constituted pars orationis. This distinction is an important one because the same primary meaning can be distributed formally over a variety of parts of speech: dolor, doleo, dolens, dolenter, heu indeed signify the same thing — they have the same significatum (John of Denmark: 217) — but they are functionally divergent in that the idea of pain is expressed here, respectively, as a fixed quality {dolor), as a person-oriented process {doleo "I have pain/I suffer"), as a process {dolens "suffering"), as a processual circumstance referring to the cause {dolenter "painfully"), and as a processual circumstance referring to the affected patient {heu "ow"). And, of course, each of these "gram- maticalized meanings" has specific possibilities and limitations of syntactic distribution: Ce n'est qu'en tant que la chose est associee a telle ou telle propriete, done qu'elle est signifiee sur tel ou tel mode, que l'entite linguistique, munie d'un mode de signifier actif, aura une existence pour la grammaire: dans une construction, on trouve en effet soit souffrance, soit souffrant. Ce fait est parfois rendu obscur par une imprecision terminologique. Dictio a sou- vent le sens non technique de "mot", unite linguistique minimale, puisque
628 PIERRE SWIGGERS Ton dit parfois que la dictio a un mode de signifier et un significat, ce qui est contradictoire avec la definition stricte que tous les Modistes donnent de la dictio. D'autre part, pars orationis designe parfois une unite linguis- tique minimale, done achevee, parce que munie d'un significat et d'un mode de signifier, parfois une partie du discours, au sens ou nous l'enten- dons, e'est-a-dire comme categorie grammaticale (ambiguite: blanc est une partie du discours/le nom est une partie du discours): il y a confusion entre le niveau linguistique et le niveau metalinguistique. Ceci s'explique par le fait que les Modistes ne considerent, dans leur etude du fonctionnement du langage, que les categories grammaticales, en negligeant ies unites lin- guistiques concretes (Rosier 1983: 57). So far, we have examined the general implications of the cognitive viewpoint adopted by the modistae: as we have seen, the pivotal role given to the ratio intelligendi accounts for the possibility of human language, as a cognition-based activity relating sound-shape (or its graphic transposition) to reality, by using symbolic units endowed with referential meaning and grammatical properties (the latter being intrasystemic, although this is not pointed out by the modistae, who focus their explanatory account on Latin). We have to discuss now the more specific embedding of the cognitive viewpoint, as it pervades the description of morphosyntactic categories. We have seen that the parts of speech correspond to the combination of shape, significance, and grammatical meaning. It is precisely the concrete implementation of the latter that constitutes the various parts of speech. On this point, the modistae developed a hierarchized theory of grammatical meanings, as incorporated within words. The fundamental distinction is that between grammatical meanings focused on the object, and those focused on the relational nature of the object (its dispositio or cir- cumstantia). This bipartition accounts for the separation of Noun, Pronoun, Verb and Participle on the one hand (these parts of speech signify per modum rei, and can therefore be considered absolute terms), and Adverb, Interjection, Preposition and Conjunction on the other hand (these relative terms signify per modum dispositionis rei sive circumstantiae rei). The parts of speech that signify per modum rei can either express a state {habitus) or a process (per modum fieri). This distinction corresponds to the opposition between autonomous nominal terms (nouns and pronouns), and verbal terms (verbs and participles). Each of these has a specific mode of signifying. The noun is the modus ends, expressing a (permanent) state [material aspect], through a determinate apprehension [formal aspect]. The latter
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 629 aspect — which distinguishes the noun from the pronoun — refers to the property of context-free referentiality of nouns. The ancient grammarians had opposed nouns and pronouns on the basis of a binary feature: + or — qualitas. In this view, the noun signifies a particular substance with its qual- itas (= bundle of descriptive characteristics), whereas the pronoun expresses only substance. The modistae oppose noun and pronoun with respect to extensional scope: nouns have fixed (classes of) referents, whereas pronouns have indeterminate referents. As shifters, pronouns have an open domain of reference (or first-level suppositio):™ "Pronomen ad omne sup- positum pertinet, id est, ad omne significatum se extendit".19 Both verb and participle signify per modum rei, and they do so in the processual mode (in contradistinction with nouns and pronouns). The difference between verb and participle lies in a binary characteristic defined as +/- distance. This spatial metaphor used by the modistae refers to the linear order of the sentence, and to the composition(ality) of its terms. The verb has a predicative function which can only be explained by an essential property of this part of speech: its direct "incomposability" with a noun (or nominal) functioning as the subject term. We have then here two extremes20 of categorization, the noun and the verb. As these can be combined at the sentence level, the enunciative uniting presupposes an essential distance.21 In discourse, the essential incomposability is overriden: at the level of symbolic expression — which must be considered "indirect" in contrast to the "direct" level of reality — all the components of a sentence are relationally construed, whatever the basic composable/non-composable relationship of their significata. With respect to the verb, this implies the presence of an accidental mode of signifying, i.e. compositio. Composition is a relative accidental mode which allows the verb to be predicated of a subject, and to enter in composition with a substance. It is especially in Thomas of Erfurt's speculative grammar that the notion of compositio is elaborated: Thomas took the unusual step of declaring this accidental feature of compositio to be a modus communissimus, and there is considerable justification for placing it at the head of all the accidental modes. Aristotle had argued that a verb does not itself indicate anything but implies a copulation, and Thomas argues from this that the copula element is to be found in every verb; as a result compositio is inherent in every verb, and serves to link the verb to its suppositum. This is a very important concept, since the essential mode of the verb declares discreteness of the verb from the suppositum — Michel de Marbais had indeed argued against discreteness
630 PIERRE SWIGGERS being considered an essential mode of the verb since it would produce precisely that situation which compositio had to remedy — to be one of its most important characteristics. For this reason, Thomas cannot make of compositio an essential mode since it is in diametric contrast to the essential mode of separation, and it is equally not a feature of the modus esse of the verb nor does it constitute a specific type of verb; it is therefore an accidental mode of the verb, which states the fundamental syntactic relationship, i.e. of S and P, between the nomen-suppositum and the verbum- appositum. (...) We find then governmental concord between the compositio, which is the property of linking the verb to the suppositum, and the substantival element in the suppositum, and once this relationship has been stated, the other accidental modes will follow (Bursill-Hall 1971: 218- 219). In Thomas of Erfurt's definition, compositio is the accidental mode of signifying of the verb by means of which the verb consignifies the property of inhering in terms of its being and by means of which the verb, separate from the subject , is first and principally linked to the subject (Thomas of Erfurt: 221). The participle signifies "per modum fieri indistantis". The non-distant aspect of the participle accounts for its direct composition with substances. According to Martin of Denmark, the participle has been invented to signify in the processual mode, in direct combination (i.e. in an endocentric construction) with the suppositum (or with the nominal nucleus of any Noun + Participle construction). Thomas of Erfurt, in his attempt to group the declinable parts of speech with respect to their formal signification, points out that the modes of distance and non-distance entail (1) adjacency of the nominal elements (and in this respect, the participle is a nominal element, although materially its essential mode of signifying is like that of the verb), and (2) composition in the verb. Modus autem indistantis a substantia, seu modus uniti substantiae sumitur ab eadem rei proprietate in participio, a quo sumitur modus adiacentis in nomine, et compositio in verbo; et haec est proprietas inhaerentis alteri secundum esse.22 Each of the four parts of speech already dealt with (noun, pronoun, verb, participle) has a variety of essential and accidental modes of signifying. As declinable parts of speech, they have two very general essential modes of signifying: a material one and a formal one. These are represented in Figure 2.
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 631 ESSENTIAL MODES OF SIGNIFYING 1 1 I. MOST GENERAL 1 MATERIAL FORMAL noun modus entis determinate apprehension pronoun modus entis indeterminate apprehension verb modus esse/ modus fieri distance participle modus esse/ modus fieri non-distance Figure 2 Next to the most general essential modes of signifying, there are the "subaltern", the "less general" (this category applies only to nouns), and the "specific" essential modes of signifying.23 These modes serve to account for (1) subtypes of a particular part of speech, and (2) for a number of semantico-syntactic features.24 This results in the subdivisions given in Figures 3 and 4. In addition to the essential modes of signifying, there are the accidental modes of signifying. While the former are defined as those "by means of which the part of speech simply possesses its essence either in accordance with its class or species", the accidental modes are "what happens to the part of speech after its essence has been completed and does not naturally confer essence to the part of speech either in terms of its class or species".25 The accidental modes of signifying account for the morphological characteristics (the traditional accidentia) of the parts of speech, and they are further subdivided into absolute and relative (respective) modes, the former referring to the functioning of terms according to the properties of the thing, the latter referring to the functioning of terms with respect to each other. It is interesting to note that in the case of the nominal and pronominal forms, the absolute accidental modes concern derivational data (this also holds for the figura of the verb and the participle). Figure 5 offers a synoptic survey of the accidental modes of signifying of the four declinable parts of speech.
632 PIERRE SWIGGERS ESSENTIAL MODES OF SIGNIFYING NOUN PRONOUN II.SUBALTERN 1. mode of the common common noun 2. mode of the proper proper name //. SUBALTERN 1. A. Demonstration of the referent (1) presence ad (2) presence ad B. Relation: obliqi the referent ///. LESS GENERAL 1. modus per se stantis substantive 2. modus adjacentis adjective in direct presence sensum intellect um je presence of anaphoric pronoun 2. A. Primitive B. Derived 3. A. Adjacent B. Relative IV. SPECIFIC 1. 5 subdivisions: general specific patronymic diminutive collective 2. 24 subdivisions 4 subdivisions . SPECIFIC possessive, demonstrative Figure 3
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 633 ESSENTIAL MODES OF SIGNIFYING VERB //. SUBALTERN 1. mode of general being substantive verb to be 2. denominative mode of being vocative verb 3. mode of specific being [ adjective verb PARTICIPLE //. SUBALTERN 1. mode of general being participle of substantive verb (ens) 2. denominative mode of being participle of vocative verb (vocans/nominans) 3. mode of specific being participle of adjective verb ///. SPECIFIC adjective verb 1. with active meaning i amo | 2. with passive meaning amor 3. with neuter meaning sto 4. with both active and passive meaning criminor (te) i criminor (a te) III. SPECIFIC 1. active (amans) 2. passive (amatus) 3. neuter (stans) 4. common (criminans + criminatus) Figure 4
634 PIERRE SWIGGERS NOUN PRONOUN VERB PARTICIPLE ACCIDENTAL MODES OF SIGNIFYING /. ABSOLUTE 1. primitive derived nouns 2. simple compound surcompound 1. simple compound surcompound 1. primitive derived (inchoative, frequentative, desiderative, ...) 2. tense 1. simple compound surcompound 2. tense //. RESPECTIVE 1. gender (M/F/N) 2. number (sg/pl) 3. case 4. person26 1. gender 2. number 3. case 4. person 1. composition 2. mood (indicative/imperative/ optative/dubitative/ indefinite) 3. accidental signification27 4. voice (diathesis) 5. person 1. accidental signification 2. voice (diathesis) 3. gender 4. number 5. case 6. person Figure 5 So far, we have discussed the declinable parts of speech. All of these signify with respect to the mode of the thing (per modum rei). The other parts of speech — adverb, preposition, conjunction, and interjection — signify with respect to the dispositional mode or the circumstance of the thing. They are further subdivided in a binary way. The first dichotomy is that between the expression of a circumstance/disposition of a thing which exists
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 635 in a processual way, and the expression of a circumstance/disposition of a thing with indifferent existential status. This corresponds to the distinction between adverb and interjection on the one hand, and preposition and conjunction on the other. Each of these is further subdivided according to the reference-focus - of its determinate content: thing (adverb) vs. affected subject (interjection); - of its linking capacity: action (preposition) vs. extreme terms (conjunction). The modistae's description of the indeclinable parts of speech is summarized in Figure 6. PER MODUM DISPOSITIONIS/CIRCUMSTANTIAE RE1 thing existing per modum fieri modus determinantis actum adverb modus afficientis interjection thing with indefinite existential status modus retorquentis casuale ad actum preposition modus coniugentis 2 extrema conjunction Figure 6 The general essential mode of signifying of the adverb is the mode of signifying by means of the mode of adjacency, by means of the mode of being, signifying it simply and determining it in absolute terms; The general essential mode of signifying of the interjection is the mode of signifying by means of the mode of determining something, which is the verb or participle representing states of mind. For since the mind is influenced by the emotions of pain, joy, fear, and the like, this is expressed by means of interjections. Hence, the interjection determines the verb or participle, not in a simple fashion, but in relation to the mind, expressing the state of the mind; The general essential mode of signifying of the preposition is the mode of signifying by means of the mode of adjacency to some case form, linking it and turning it back to the act; The general essential mode of signifying of the conjunction is the mode of signifying by means of the mode of joining two polarities. This mode of signifying is derived from the prop-
636 PIERRE SWIGGERS erty of joining and uniting to be found in things in the outside world. Therefore the conjunction is a part of speech which signifies by means of the mode of conjoining two polarities.28 It is clear that all the declinable parts of speech — with the exception perhaps of the interjection — have an important role in sentence construction. Interestingly, the participation in sentence construction is defined as an essential mode of signifying of these parts of speech which depend on the presence of other constructible terms. As such, these parts of speech are essentially characterized as signifying in an adjacent way. As in the case of the declinable parts of speech, the subtypes and morphological characteristics of the indeclinable words are correlated with subaltern and specific essential modes of signifying, and with accidental modes of signifying. These give way to the various classes of adverbs (of place, quantity, quality, time, modality — all informative or interrogative), of conjunctions (copulative, disjunctive, causal, consecutive), of prepositions, and of interjections (of pain, joy, admiration, fear). The accidental modes of signifying account for the morphological features of adverbs (comparison; simple — compound; primitive — derived), and conjunctions (simple — compound; primitive — derived; postposed — preposed). As will be clear from this description, the subclasses of these parts of speech are generally defined in contentive terms, and the semantic partitioning is based on the traditional distinctions (place, quantity, etc.) made by grammarians. A more interesting categorization is that found within the discussion of conjunctions. The semantic subtypes of this part of speech are reduced to two sets, the conjunctions per vim and the conjunctions per ordinem. The conjunctions per vim are either copulative or disjunctive: The mode of signifying by means of the mode of relating two polarities by means of meaning is the mode of signifying, uniting two polarities which are not interdependent, such as two substantives, or two adjectives, or two sentences, which have no shared order. This can be divided into two special essential modes, the first of which is the mode of signifying by means of the mode of relating two polarities to one another and to any third and this mode constitutes the copulative conjunctions. The second mode is the mode of signifying by means of the mode of relating two polarities together, distinguishing them from a third, and this mode constitutes the disjunctive conjunctions (Thomas of Erfurt: 259). The conjunctions per ordinem comprise the causal and consecutive conjunctions:
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 637 The first is the mode of signifying by means of the mode of relating two polarities derived in order from a preceding part of speech. This mode constitutes the causal conjunctions which derive from the antecedent, which is the cause of the consequent, as in: Socrates currit, ergo movetur [Socrates runs, so he moves]. The second is the mode of signifying by means of the mode of conjoining two polarities in terms of the relationship which they derive from the following member, and this mode constitutes the rational conjunctions; they are properly called rational because they refer rather to the consequent whose function depends on the antecedent, as when we say: terra interponitur inter solem et lunam; ergo luna eclipsatur (Thomas of Erfurt: 261). The dichotomy between these two sets of conjunctions not only corresponds to divergent logical operations — combinatory (conjunctive/ disjunctive) and inferential reasoning — but it also corresponds to a divergent distribution of truth values. As a matter of fact, the conjunctions per vim do not construe a new, single truth value, but combine two independent truth values, whereas the conjunctions per ordinem bring about a new truth value holding for the complex sentence. Another interesting point in the modistae's description of the parts of speech concerns the nature of the preposition. The preposition, signifying in the adjacent mode, is intrinsically connected with a property of things: that of determination and limitation (coarctatio).29 This view illustrates the dynamic character of the syntactic relationships construed by indeclinable parts of speech: there is a clustering of perspectival meanings associated with particular prepositions, and each of these constitutes a specific determination of states of affairs, which can be delimitated in various respects. In the modistic treatises the section on morphology is often followed by a shorter treatment of syntax, the focus of which is the sentence. A sentence is a meaningful construction of terms (belonging to specific word classes); the joining of the "constructibles" is precisely the subject matter of syntax (or diasynthetica). In general, syntax is divided into four parts: prin- cipia construendi, constructio, congruitas, and perfection This division reflects a hierarchized view of the sentence as (1) a combination of terms (2) in an appropriate way (3) to convey a meaningful and complete concept. In this view, contentive (material) and relational (formal) aspects of sentence construction come together,31 and each of the four stages corresponds to an Aristotelian causal principle:
638 PIERRE SWIGGERS - material (members of the potential construction) - formal (the construction itself) - efficient (congruity: concord and government, but also contextual appropriateness) - final (the expression of a complete thought). The basic idea behind this descriptive and terminological apparatus is that the construction of syntactic units must be regarded as a movement (motus), by which two constructibles are linked. These two fundamental constructibles are called dependens and terminans. Their union in the discursive utterance corresponds to a movement of compounding, which actualizes the association of a material "dependent" constructive {con- structibile dependens est in potentia, quia habet se sicut materia) with a term "achieving" the construction:32 Nam sicut ex materia et forma, quorum unum est in actu, alterum vero in potentia, fit per se compositum in natura; sic ex ratione dependendi et ter- minandi fit per se constructio in sermone (Thomas of Erfurt: 280). This implies that Noun + Verb constructions {Socrates currit "Socrates runs") will be analysed as terminans + dependens, the verb expressing an action which can be attributed to any (animate) substance. This distinction is coupled with another one, viz. that between prin- cipium and terminus. The principium refers to inherent or immanent actions (which do not go beyond the causal instance or agent), whereas the terminus refers to the object in which an action finds its achievement (in the same way as the house is the terminus of the action of building). This distinction accounts for the opposition between absolute (intransitive) and transitive constructions: Alius est actus qui non transit in materiam aliquam extrinsecam, sed est in suo agente et non requirit aliquam materiam extrinseca, in quam transeat, et iste est actus absolutus, ut patet in hoc actu curro\ cursus enim est in cur- rente; Alius est actus, cuius natura talis est, quod sibi non sufficit agens nee est in agente suo sicut in subiecto, sed requirit materiam extrinsecam aliam a suo agente, et talis actus transit in materiam extrinsecam, et talis est actus transiens ut percutio Socratem, proiicio lapidem (Boethius of Denmark: 188-189). In this view, active and passive are complementary perspectival expressions of the same motus: II faut noter que, comme le dit le philosophe, Physique V, Taction et la passion sont un seul mouvement, et ce mouvement, considere sous deux
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 639 formes differentes, peut etre dit action et passion. Si on le considere en relation avec 1'agent, ce mouvement est dit action; si on le considere en relation avec le patient, ce mouvement est dit passion. De la meme ma- niere je dis que tout verbe signifie sur le mode du devenir et du mouvement. II s'en suit que ces verbes peuvent etre dits actifs et passifs selon deux proprietes differentes. Si Ton considere ces verbes en relation a la substance agent, ces verbes sont dits actifs, par exemple par moi il est lu. Mais si Ton considere ces verbes en relation avec la substance patient, alors ces verbes sont dits passifs, par exemple le livre est lu par moi (Simon de Dacie: 158). En fait, ce n'est pas reellement la relation a 1'agent ou au patient qui determine si le verbe est actif ou passif, mais la relation au prin- cipe de l'acte: si celui-ci est agent, le verbe est dit actif, si celui-ci est patient, le verbe est dit passif. Ainsi, dans le dernier exemple du texte ci- dessus, le verbe lire est considere comme en relation avec le livre, le prin- cipe, qui est patient, et non en relation avec par moi, le terme, qui est agent. Le principe, nous Tavons dit, est le point de depart de Taction, le terme son point d'aboutissement. Ces notions ne sont pas confondues avec celles d'agent et de patient. En effet 1'agent et le patient sont definis une fois pour toutes, pour une action determined. Ainsi, pour Taction de lire, Tagent est moi, le patient livre. A partir de la, Tagent comme le patient peut etre principe ou terme de Taction, rendant les constructions passives ou actives (Rosier 1983: 185-186). Now the concepts of transitivity/intransitivity are combined with those of identity (concord) and government (rection) relations, and this results in a descriptive framework which comes close to that of Immediate Constituents analysis. As a matter of fact, the sentence is decomposed into a number of segments which manifest relationships of concord or government, and this in a hierarchical way. The basic relation is that of concord between subject and verbal predicate, the latter being transitive or intransitive: (a) Socrates legit librum "Socrates reads a/the book" (b) Socrates currit "Socrates runs" In the former case, the verb builds a transitive construction with the terminus (also patient) librum. The minimal sentence consists of an intransitive construction, in which the subject term (suppositum) and the verbal predicate (appositum) agree: S - concord[Supp. - App.J For transitive constructions the formula must be expanded: S - COnCOrd[SupP. - App. (governmentVerb _ Terminus)] Each of the main segments of a sentence can then build further transitive
640 PIERRE SWIGGERS and intransitive constructions, as can be seen in the following sentence: Summa doctrina Tullii rhetoris instruit subtiliter artem rhetoricam "The great doctrine of Tullius (Cicero) the orator has subtly instructed the rhetorical art" 1. Intransitive construction: Suppositum - Appositum (summa) doctrina (Tullii rhetoris) - instruit (subtiliter artem rhetoricam) 2. Transitive construction within the Appositum: VP - Terminus instruit - artem (rhetoricam) 3. Intransitive (qualifying) constructions within Suppositum and Terminus: summa - doctrina (Tullii rhetoris) artem - rhetoricam 4. Transitive construction within suppositum (head - determinating clause) doctrina - Tullii (rhetoris) 5. Intransitive construction within determinans: Tullii - rhetoris 6. Intransitive construction within verbal phrase: verb - adverb instruit - subtiliter (modus instructionis = subtilis) The importance of this descriptive schema lies in its (1) formal universality,33 (2) its recursiveness (applying to the various levels of linguistic description), and (3) its contextual application (relationships are defined with respect to constructions, and not with respect to word classes). Whatever construction we find, it can be analysed as (1) transitive or intransitive, and (2) involving a dependent and a terminating constructible. The constructions can be defined with regard to actions (percutio Socratem "I hit Socrates"; Socrates currit "Socrates runs"), or with regard to persons (filius Socrati "the son of Socrates"; equus albus "white horse"). The grammatical relationships built by terms with respect to each other correspond to compatibility relations between their (essential and accidental) modes of signifying. Sentences fulfilling the conditions of congruity (defining e.g. the "composition" of verbs with their subjects/specific types of agreement) are well-formed grammatically, but this does not suffice to make them "complete" sentences. We can exclude, on the Jbasis of the principles of congruity, strings such as *cappa albi ("the hat of white"), or *misereor albi ("I feel pity with white"), since in these strings the terminating constructibles
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 641 should signify in the independent mode, and not in the adjacent mode. Congruity thus defines grammaticality, and not semantic acceptability: a construction such as cappa categorica ("categorial hat") conforms only to the principles of congruity. In fact, the problem of semantic acceptability (or "convenience/non-convenience of significations") is left to the logician: Sed convenientia vel repugnantia significatorum specialium a grammatico per se non consideratur sed magis a logico (Thomas of Erfurt: 308). However, a sentence must be complete, and this involves, in addition to the presence of a Suppositum - Appositum chain, and the application of the principles of congruity, a third condition: the "achievement" of all relationships opened by terms (such asfilius "son of ...") or syntagms (e.g. clauses (e.g. Si Socrates currit "If Socrates runs, ..."). From this, it is clear that there are three criteria required for the completion of the sentence. First is the suppositum and appositum, because, since the complete construction finally ordered by means of the verb may exist for the expression of the compound concept of the mind, it behooves that, just as there is a separation between compound concepts of the mind, so there will be separation in the combination of constructibles. But this separation can be only between suppositum and appositum, and further, only the verb can be the appositum since it exists by means of the mode of separation. Secondly, conformity of all the modes of signifying is required, just as it is required for any congruous statement. Thirdly, it i: required on the part of the construction that no dependence should be left incomplete which might hold it back from its final purpose, which is to express a compound concept of the mind and to generate perfect sense in the mind of the hearer (Thomas of Erfurt: 315).34 As noted by the modistae, a complete sentence should generate a corresponding complete meaning in the mind of the listener, but the generating of the latter is "not punctual, but has a degree of perfection in terms of plus and minus" (sed haec perfectio sensus in animo auditoris non est punctualis, sed habet gradum perfectionis secundum magis et minus, Thomas of Erfurt: 316). 3. Conclusion In this paper I have attempted to show how grammatical categories were defined in mediaeval speculative grammar, with respect to both their structure and their syntactic functioning. The orientation is basically a cognitive one, and it is cast in a particular framework: that of modes of signifi-
642 PIERRE SWIGGERS cation, based themselves on modes of intellection. The result is a unified description of word classes and sentence patterns, specifically in their formal aspects. The latter focus is typical of modistic grammar (and also, to a lesser extent, of XVIIth and XVIIIth century rational grammar): the semantic (or semantic-pragmatic) load of individual sentences is hardly taken into account as such. This has not so much to do with the medium of analysis (Latin as the universal language for scientific purposes), or with the unavailability of analytic tools for the study of language use, let alone ways of speaking. This focal interest ultimately goes back to the form-oriented program defined by Aristotle in the Categories, the Peri Hermeneias and the Topics', the search for grammatical principles is based on how words function as signs, and not on what they stand for in a particular discursive context. It is the formal apparatus of language or, better, the organization of the conventional linguistic system that constitutes the core object of grammar: "it should be known that expression, insofar as it is expression, is not considered by the grammarian, but insofar as it is a sign, it is, since grammar deals with the sign of things, and because the expression is the most suitable sign among other signs, therefore expression, insofar as it is a sign, is considered by the grammarian before other signs of things" (Thomas of Erfurt: 149).35 The great merit of the modistae lies in the cognitive exploitation of the grammatical "equipment" of language — in fact of one language, viz. Latin, but the insights are transposable to the other Indo-European languages — by showing what types of categorization are implied by word classes, what kind of perspectives (directness - obliqueness; adjacency, etc.) are signified by subaltern modes of signifying, and what kind of formal types of meaning are expressed by grammatical accidents.36 Notes 1. See Swiggers (1981, 1982, 1984a, 1984c, 1986). 2. For a comparison between modistic grammar and Chomskyan grammar see especially Godfrey (1960, 1966). 3. As a consequence, Lakoffs (1986) strict dichotomy between an objectivist and an experientalist view of cognition should be relativized: the methodological apparatus of a particular theory is not strictly determined by the latter's philosophical foundations. Modistic grammars share with "experientalist cognitive linguistics" the view of syntax as "1. providing grammatical categories and constructions that are semantically-motivated, 2.
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 643 giving the semantic and functional motivations for those categories and constructions, and 3. indicating the relationships among the constructions — relationships based both on form and on meaning" (Lakoff 1986: 5). 4. On these aspects of Aristotle's theory, see Larkin (1971: 41-43) and Owens (1960); an extremely useful work is Ackrill's (1963) translation with commentary of the Categories and the Peri Hermeneias. Aristotle can be read in the English edition of his complete works (Ross ed. 1910-1928). 5. On this "language of thought" (or "mentalese"), see Nuchelmans (1972). 6. A similar attitude underlies Aristotle's views on metaphor (for an analysis, see Swiggers 1984b). Of course, such a view has wide-ranging implications for epistemologies based on truth and falsity; for powerful criticisms of these, see Lakoff and Johnson (1980) and Putnam (1981). 7. In Aristotle's view and that of the modistae, the accidentia (number, gender, person, ...) are meaningful units, and their inherent meaning has various syntactic consequences. This is very similar to Langacker's views on the grammatical markers of nominal and verbal categories (number, gender, case, person, tense, etc.); see esp. Langacker {in press, and "Overview", this volume). 8. On Aristotle's theory of knowledge (in relation to language), see Ross (1937), Hamelin (1931), De Corte (1934), and Larkin (1971). 9. On the speculative grammars of the modistae one can read the following syntheses: Bur- sill-Hall (1963, 1966, 1971), Covington (1984), Pinborg (1967, 1972: 102-126, 1975, 1982), Rosier (1983). The delimitation 1250-1350 is a rough one: the first full-fledged modistic product is Martin of Denmark's De modis significandi (ca. 1270), see also Bursill-Hall (1972: 26-27). In the following the modistic grammars will be referred to on the basis of published critical editions (see in the bibliography under Martin/Boethius/Simon/John of Denmark; Siger of Courtrai; Thomas of Erfurt). The texts are mostly cited in the original Latin (I have often paraphrased the quotations in the portion of the text that precedes or follows; if not the quotations are translated into English). 10. This coincided with the promotion of grammar within the Artes curriculum: "What is beyond dispute is that grammar came to hold a position of considerable privilege in the programme of university studies in the Middle Ages; it achieved this not only because of its intrinsic importance, but because it became the key to what the mediaeval schoolmen considered higher forms of learning, i.e. dialectic and theology. The curriculum of the mediaeval university was built around the Liberal Arts which were divided into the Trivium and Quadrivium; grammar was the first subject of the Trivium and naturally assumed the position of pre-requisite to all the Liberal Arts, being the necessary study for the reading and writing of the "universal" language of learning, i.e. Latin. Thus grammar came to be studied for its own sake as part of human culture and for the study of classical authors, particularly of the Latin Bible (Bursill-Hall 1971: 22). On this evolution, see also Pare-Brunet-Tremblay (1933), Hunt (1941-43, 1950) and Chenu (1936, 1957). 11. See e.g. the definition of Martin of Denmark (p. 8): "modus significandi est proprietas rei consignificata per vocem" [the mode of signifying is the property of the thing consignified by the word]. 12. See the definition by Siger of Courtrai (p. 90): "modus significandi existens in intellectu est illud mediante quo vox significat modum essendi circa rem" [the mode of signifying
644 PIERRE SWIGGERS existing in the intellect is that by which the word signifies the way of being with respect to the thing]. 13. "II nous faut a ce point preciser une distinction importante: c'est l'intellect en tant que matiere qui a la ratio intelligendi. Cet intellect muni de la potentialite d'intelliger, est bien distinct de l'intellect qui est 1'agent, tant de l'operation d'intellection que de l'operation de signification. Cette distinction entre matiere, forme et agent renvoie a la theorie aris- totelicienne des quatre causes: l'intellect est la cause mate>ielle du concept, de meme que l'airain est la cause materielle de la statue; la ratio intelligendi est la cause formelle du concept de meme que le modele de la statue est la cause formelle; l'intellect est, comme la main qui fait la statue, la cause efficiente du concept; quant a la cause finale qui est, pour la statue, la production d'une certaine forme esthetique, elle serait, pour le concept, de produire une intellection" (Rosier 1983: 48). 14. "Potentia passiva per se et sine motore non vadit ad actum. Sed intellectus noster poten- tia passiva est. Ergo per se non vadit ad actum" (John of Denmark: 238). 15. "Modus intelligendi activus est modus quo intellectus comprehendit modum essendi seu proprietatem ipsius rei" (Siger of Courtrai: 94). 16. In Thomas of Erfurt's definition of the modus intelligendi activus, the distinction between this mode and its formal principle, the ratio intelligendi is abandoned: "modus intelligendi activus est ratio concipiendi, qua mediante, intellectus rei proprietates significat, concipit vel intelligit" (Thomas of Erfurt: 140). As noted by Rosier (1983: 211 n. 86), "ici ratio et modus sont a peu pres equivalents: il revient au meme de dire que l'intellect est muni de la potentialite d'intelliger ou de la propriete d'intelliger. Ratio et modus sont distingues lorsqu'ils s'appliquent a la meme entite envisagee a deux moments differents. Par exem- ple, la propriete de la chose est munie de la potentialite (ratio) d'etre signifiee. L'operation de signification realisee, la propriete de la chose devient partie du discours munie de la propriete de signifier" (modus). 17. This primary meaning is the conceptual structure associated with (the form of) a given word. 18. On the theory of suppositio (the relation of a sign to what it stands for), see Arnold (1952), Boehner (1958), Covington (1984: 113-119), de Rijk (1962-1967), and Pinborg (1972: 61-65, 129-148, with further references). 19. "A pronoun belongs to every suppositum, i.e. it extends to every significatum" (Thomas of Erfurt: 198); see also Boethius of Denmark (p. 239) and Martin of Denmark ("per ipsa [pronomina] nihil determinate apprehenditur", p. 13). 20. "Cum verbum sit alterum extremum in oratione, distans a supposito, et inclinetur ad suppositum, huius inclinationis principale et communissimum principium, inter ceteros modos accidentales verbi, est compositio" (Thomas of Erfurt: 220-222). 21. "Omne enim, quod de alio enuntiatur, enuntiatur de eo ut distans" (Boethius of Denmark: 186, "everything which is enuntiated of another thing, is predicated of it as being distant from it"). 22. "But the mode of non-separability from the substance or the mode of separability from substance is derived from the same property of the thing in the participle from which the mode of adjacency in the noun and composition in the verb are derived; this is the property of inhering to another in terms of its being" (Thomas of Erfurt: 240).
GRAMMATICAL CATEGORIES AND CONCEPTUALIZATION 645 23. These distinctions do not hold for all the modistae; cf. Bursill-Hall (1971: 98-113) and Rosier (1983: 97). 24. The definition of the parts of speech incorporates the semantic content of the word-classes, and no strict separation can be made, within the modistae's descriptions, between formal and semantic aspects of word classes. For a similar view, and for arguments in favour of a semantically based definition of the parts of speech, see Langacker (in press: part II, "Overview", this volume). 25. "Modus significandi essentialis est, per quern pars orationis habet simpliciter esse, vel secundum genus, vel secundum speciem. Modus significandi accidentalis est, qui advenit parti post eius esse completum, non dans esse simpliciter parti, nee secundum genus, nee secundum speciem" (Thomas of Erfurt: 148). 26. See especially the treatment of person in the noun by Martin of Denmark, and the remarks formulated by Rosier (1983: 110-111). 27. On the accidental signification (corresponding to the valency) of the verb, see Thomas of Erfurt (paragraph 58). 28. Thomas of Erfurt (p. 247, 257, 263, 269). 29. As rightly noted by Thomas of Erfurt, prepositions are not necessarily signs of transitivity: "Certain verbs require prepositions although they have no transitiveness, either quick or slow, as when we say: sum in domo. Furthermore, by saying: annulus ex auro, the noun annulus requires the preposition but it does not, however, have any transitiveness. Therefore, it can be said that the preposition was in fact invented on behalf of case forms, not just any, but the accusative and ablative. For, just as it was stated for the noun, that the accusative is the mode of signifying "whom" (quern), and that is delimitable by means of a mode "to whom" (ad quern), "in whom" (in quern), "near whom" (prope quern), and "close to whom" (iuxta quern), and so on; similarly, the ablative is the mode of signifying by means of the mode "with, from, or by whom" (quo), delimitable by means of a mode "with, in, without whom" (a quo, in quo, sine quo), and the like" (Thomas of Erfurt: 263). 30. As noted by Bursill-Hall (1971: 288), this type of syntactic theory constitutes an important change with respect to the work of the predecessors of the modistae (e.g. Peter Helias, Robert Kilwardby), who conceived their grammars as commentaries on Priscian's Institutiones grammaticae: "The Modistae, whatever their failings and inability to escape their environment, did attempt to create something quite new in the form of a syntactic theory of their own in which we can see the burgeonings of a syntactical theory which contains features to be found in the analytical procedures of certain modern linguists". For a careful study of the four parts of modistic syntax, see Covington (1984, with a synoptical "preview" on pp. 30-40). 31. Both the principia construendi and the constructio have well defined material and formal aspects: principium materiale (the word) vs. principium formale (the pars orationis) in the case of the former, and the constructibilia vs. unio constructibilium in the case of the latter. 32. "Semper duorum constructibilium, quae referuntur ad se invicem, sub forma constructi- bili sunt duo principia, scilicet principium dependentiae in uno constructibili quo unum dependet ad alterum, et in alio principium seu ratio termini, ita quod aliquo modo per rationem exigentiae intelligatur ratio dependentis, quia illud quod dependet exigit aliud
646 PIERRE SWIGGERS sicut illud quod suam dependentiam terminat; vel per rationem exigentiae intelligitur ratio terminantis dependentiam" (Martin of Denmark: 88-89). 33. This formal universality corresponds to the bundle of basic principles within cognitive linguistics: hierarchy (see especially Langacker "Grammatical valence", this volume), constructional schemata (Langacker in press, "Overview" and "Usage-based model", this volume), and categorizing relationships (Langacker in press, "Overview", this volume). The latter notion is more akin to the basic intuitions of the modistae than is the notion of prototypes (Rosch 1975, 1977, 1978). 34. "Ex his patet, quod tria requiruntur ad perfectionem sermonis. Primum est suppositum et appositum; quia cum constructio perfecta sit ad exprimendum mentis conceptum com- positum secundum distantium finaliter ordinata, oportet, quod sicut est distantia inter conceptus mentis compositos, sic etiam sit distantia in constructibilium unione. Sed haec distantia solum est inter suppositum et appositum, ex hoc quod solum verbum est appositum, quod per modum distantis se habet. Secundo, requiritur omnium modorum significandi conformitas, prout ad congruitatem requiribatur. Tertio, requiritur ex parte constructionis, quod nulla dependentia sit non terminata, quae retrahat ipsam ab eius fine, qui est mentis conceptum compositum exprimere, et perfectum sensum in animo auditoris generare" (Thomas of Erfurt: 314). 35. "Item sciendum est, quod vox, inquantum vox, non consideratur a grammatico, sed inquantum signum, quia grammatica est de signis rerum; et quia vox est habilissimum sig- num inter alia signa, ideo vox inquantum signum prius consideratur a grammatico, quam alia signa rerum" (Thomas of Erfurt: 148). 36. The author is indebted to Brygida Rudzka-Ostyn and Dirk Geeraerts for comments and suggestions on a first draft of this paper.
Cognitive Grammar and the History of Lexical Semantics Dirk Geeraerts University of Leiden I University of Leuven 1. Introduction De semasiologie of semantiek is dus dat deel der taalwetenschap, dat zich ten doel stelt de levensgeschiedenissen van alle woorden stelselmatig te verzamelen, te schiften en te klassificeeren, om daarna te trachten al de diepere psychologische en sociale oorzaken van die levensgeschiedenissen te achterhalen niet alleen, maar weer onderling te vergelijken, om zoo tot de kennis van eenige weinige algemeen geldige semasiologische wetten te komen. Daar nu de woorden alleen leven van ons menschelijk psychisch en sociaal leven, zullen dit dus psychische en sociaal-psychische wetten moeten zijn1 (Van Ginneken 1912: 6). (Translations of quotations are given in the notes.) The view of semantics expressed in this quotation has an immediate appeal to proponents of the new school of cognitive grammar that is exemplified by the studies in this volume. Indeed, studying the semantic history of words to learn something about the "laws" of the human mind can easily be appreciated as a particular realisation of the more general attempt to study linguistic phenomena at large as an emanation and con- cretisation of human cognition and its working principles. Is Van Ginneken merely an incidental precursor of cognitive grammar, or are his views typical of a broader tradition of semantic research that can be shown to have some remarkable affinities with cognitive grammar? In this paper, I will try to demonstrate the latter. I will try to point out some revealing similarities between cognitive grammar and the historical- philological tradition of semantic research to which Van Ginneken belongs (albeit with the restriction that I will be concerned exclusively with lexical semantics). Also, I will try to make clear that cognitive grammar may profit from a greater awareness of the historical-philological tradition of seman-
648 DIRK GEERAERTS tics, and I will place the observed similarities against the background of a brief survey of the history of lexical semantics as a branch of linguistics. 2. Historical-Philological Semantics The historical-philological approach can be exemplified by major works such as Paul (1880), Darmesteter (1887), Breal (1897), Wundt (1900), Nyrop (1913), Wellander (1917-1928), Sperber (1923), Carnoy (1927), and Stern (1931). It can be seen as the first truly linguistic form of semantics, i.e. as a form of semantic research that is embedded in linguistics as an autonomous empirical discipline, rather than being a part of speculative philosophy or of rhetoric (as is the case with most of the semantic studies carried out before the 19th century). The historical-philological approach is characterized by three prominent features.2 In the following, these will be illustrated with quotations from Breal (1897), not because Breal is the first or the most important exponent of historical semantics, but because his work clearly expresses the major methodological features of the historical-philological approach.3 It should be mentioned that the three characteristics listed below need not be simultaneously present in all of the works belonging to the historical-philological era. Individual differences of emphasis notwithstanding, they do, however, adequately characterize the basic methodological outlook that is shared by most of the semantic studies in this period. 1. Already on the first page of Breal's Essai de semantique, the historical orientation of semantics is indicated as an intuitively obvious matter of fact. Talking about linguistics, he notes: Si Ton se borne aux changements des voyelles et des consonnes, on reduit cette etude aux proportions d'une branche secondaire de 1'acoustique et de la physiologie; si Ton se contente d'enumerer les pertes subies par le meca- nisme grammatical, on donne 1'illusion d'un edifice qui tombe en ruines; si Ton se retranche dans de vagues theories sur 1'origine du langage, on ajou- te, sans grand profit, un chapitre a l'histoire des systemes. II y a la, il me semble, autre chose a faire ... La linguistique parle a 1'homme de lui- meme: elle lui montre comment il a construit, comment il a perfectionne, a travers des obstacles de toute nature et malgre d'inevitables lenteurs, malgre meme des reculs momentanes, le plus necessaire instrument de civilisation4 (1897: 1-3). It is even the case that an adequate understanding of words in their contem-
HISTORY OF LEXICAL SEMANTICS 649 porary meaning requires a thorough knowledge of their semantic history: "L'histoire peut seule nous donner aux mots le degre de precision dont nous avons besoin pour les bien comprendre"5 (1897: 124). This historical orientation is, of course, a general aspect of 19th century linguistics. As an autonomous empirical discipline, linguistics comes into being as a form of historical research, and the birth of historical linguistic semantics in the 19th century is merely one more aspect of the overall diachronic outlook of the first phase in the development of modern linguistics. It should be noted, however, that the birth of semantics within that young linguistic science was not just a question of completeness, but rather of necessity. The study of meaning was not simply taken up out of a desire to study linguistic change in all of its aspects, but a thorough knowledge of the mechanisms of semantic change appeared to be a prerequisite for adequate historical investigations into the formal aspects of languages. Indeed, one of the most important branches of historical linguistics was the etymological reconstruction of the history of words and phrases. The methodology of comparative reconstruction requires that the word forms from different languages that are to be compared with each other, be semantically related. But such a relationship is not always obvious. For instance, throughout the older Germanic languages, there is a fairly systematic formal resemblance between words for the concept 'beech', and words for notions such as 'book' and 'letter'. (Compare, e.g., Old High German buohha 'beech' and buoh 'book', or Old Saxon boka 'beech' and bok 'book, writing tablet'.) Now, in order to justify a reconstruction of these forms as being related to the same Proto-Germanic root, their semantic relationship has to be clarified. (In this particular case, an awareness of the frequent metonymical relationship between names for substances and the name of objects made of those substances will have to be combined with archaeological evidence showing that wooden tablets were used for writing purposes.) Considering a number of lexical forms as cognate requires that their semantic relationship can be plausibly established, and this in turn requires an overview of the regular mechanisms of semantic change. As such, diachronic semantics was not merely taken up as an end in itself, but also as an auxiliary discipline with regard to etymological studies.6 It should also be noted that the enormous success of historical semantics in the 19th century and the first decades of the 20th, does not only show up in the large number of articles and monographs devoted to it, but also in the major lexicographical projects undertaken during that period.
650 DIRK GEERAERTS Voluminous historical and etymological dictionaries such as the Deutsches Worterbuch, the Oxford English Dictonary, and the Woordenboek der Nederlandsche Taal testify that the early interest in the semantic histories of words has led to an hitherto unsurpassed amount of empirical semantic research. 2. The psychological orientation of historical-philological semantics has two aspects to it. In the first place, lexical meanings are considered to be psychological entities, i.e., (a kind of) thoughts or ideas: "(Le langage) objective la pensee"7 (Breal 1897: 273). The mental status of lexical meanings links up directly with the overall function of thinking, i.e., with the function of cognition as a reflection and reconstruction of experience. As such, language expresses thought as a storeroom of encyclopaedic knowledge: "Le langage est une traduction de la realite, une transposition ou les objets figurent deja generalises et classifies par le travail de la pensee"8 (1897: 275). Language, then, is not autonomous; it is embedded in and linked with the total set of cognitive capacities that enable men to understand the world with ever more refined conceptual tools. In the second place, historical-philological semantics tries to explain meaning changes as the result of psychological processes. The general mechanisms of semantic change that can be derived from the classificatory study of the history of words constitute patterns of thought of the human mind. Breal calls these mechanisms "les lois intellectuelles du langage", but he hastens to add that 'law' means something different here than in the natural sciences: a law of semantic change is not a strict rule without exceptions, but it represents a tendency of the human cognitive apparatus to function in a particular way. In a passage that opposes the restriction of linguistics to the study of the formal aspects of language, he remarks: Nous ne doutons pas que la linguistique, revenant de ses paradoxes et de ses partis pris, deviendra plus juste pour le premier moteur des langues, c'est-a-dire pour nous-memes, pour l'intelligence humaine. Cette myste- rieuse transformation qui fait sortir le frangais du latin, comme le persan du zend et l'anglais de l'anglo-saxon, et qui presente partout sur les faits essentiels un ensemble frappant de rencontres et d'identites, n'est pas le simple produit de la decadence des sons et de l'usure des flexions; sous ces phenomenes ou tout nous parle de ruine, nous sentons Taction d'une pen- see qui se degage de la forme a laquelle elle est enchainee, qui travaille a la modifier, et qui tire souvent avantage de ce qui semble d'abord perte et destruction. Mens agitat molem9 (1897: 338-339).
HISTORY OF LEXICAL SEMANTICS 651 The moving force of the human mind also shows up in the fact that the fundamental factor that brings the psychological mechanisms of semantic change into action, consists of the communicative needs of the language user. Languages change because people try to express their thoughts as accurately and satisfactorily as possible. Le but, en matiere de langage, c'est d'etre compris. L'enfant, pendant des mois, exerce sa langue a proferer des voyelles, a articuler des consonnes: combien d'avortements, avant de parvenir a prononcer clairement une syl- labe! Les innovations grammaticales sont de la meme sorte, avec cette difference que tout un peuple y collabore. Que de constructions maladroites, incorrectes, obscures, avant de trouver celle qui sera non pas Texpression adequate (il n'en est point), mais du moins suffisante de la pensee10 (1897: 8). Note, however, that the expressive, communicative intent that lies at the basis of linguistic change need not always have the conscious, voluntary characteristics attributed to it by Breal. Sperber (1923), e.g., stresses the importance of subconscious emotional factors. Also, the overall psychological orientation of historical-philological semantics leaves room for authors such as Meillet (1906), who stresses the social rather than the psychological background of semantic change. Likewise, the earliest classifications of semantic changes, relying heavily on traditional rhetorical classifications (as is still the case in, e.g., Darmesteter 1887), were much less psychologically orientated than later ones.11 3. Methodologically, historical-philological semantics is a hermeneutical discipline. In the following quotation, Breal does not simply repeat the point that semantics is a historical science, but he also has something to say about the way in which that scientific project is put into practice: Si Ton admet une difference entre les sciences historiques et les sciences naturelles, si Ton considere l'homme comme fournissant la matiere d'un chapitre a part dans notre etude de 1'univers, le langage, qui est l'oeuvre de l'homme, ne pourra pas rester sur Tautre bord, et la linguistique, par une consequence necessaire, fera partie des sciences historiques12 (1897: 278). Clearly, the natural sciences also study historical processes (as in geology or the study of biological evolution), and that is why the difference between the natural and the human sciences that is mentioned in the quotation has to be sought on the methodological level rather than on the level of the object matter of both approaches. The distinction made by Breal probably refers to the theories of the German philosopher Wilhelm Dilthey, whose
652 DIRK GEERAERTS views on the relationship between the natural and the human sciences (Naturwissenschaft versus Geisteswissenschaft) were widely popular near the end of the 19th century.13 The methodological autonomy of the human sciences with regard to the natural sciences resides in the fact that they try to understand, by means of an empathic process of interpretation (Ein- fuhlung), the cultural forms of expressions in which men have, throughout history, laid down their experience of the world. The natural sciences, on the other hand, try to explain the characteristics of the material world by means of rigid laws. Next to being historically and culturally orientated, the human sciences in the Diltheyan sense are hermeneutical par excellence: they try to reconstruct the original experience that has lain at the basis of particular forms of human expression that have been transmitted from earlier times to the present day; they look for the expressive intention behind historical forms of expression. The connection between the Diltheyan conception of the human sciences and the kind of linguistic semantics sketched above will be clear: through its historical approach, through its cultural, experiential orientation (language is an expression of a particular experience of the world — see the quotation given above from Breal 1897: 275), and through the importance it attaches to the expressive intentions of language users as the source of linguistic change, historical-philological semantics fits nicely into the Diltheyan view of the human sciences. This is reflected on the methodological level. Because linguistic semantics is a historical discipline, its primary material consists of texts from dead languages or from previous stages in the development of a living language. Its basic methodological act is therefore the interpretation of those texts; only afterwards can changes between periods (and the mechanisms guiding them) be recognized, classified, and explained. The primary methodological step of the historical semantician is that of the historical lexicographer and the philological scholar: to interpret historical texts against the background of their original context by trying to recover the original communicative intention of the author. 3. Cognitive Semantics With Rosch's prototypical conception of categorial structure as one of its leading notions (cf. Rosch 1977, Mervis & Rosch 1981), cognitive semantics is exemplified by studies such as Fillmore (1977a, 1977b), Lakoff
HISTORY OF LEXICAL SEMANTICS 653 (1982a, 1982c), Lakoff & Johnson (1980), Langacker (1983), Talmy (1983, 1985b), Coleman & Kay (1981), Lindner (1981), Brugman (1981, 1984), Hawkins (1984), Sweetser (1984), Vandeloise (1985a), Dirven (1985), Rudzka-Ostyn (1985), Geeraerts (1983a, 1985a). The characteristics of the cognitive trend in lexical semantics can be summarized by contrasting it with a view of categorization that has, implicitly or explicitly, been fashionable for some decades. This view is sometimes referred to as the classical or 'Aristotelian* view (cf. Rosch 1975, 1983). As far as the history of lexical semantics is concerned, this is a somewhat misleading terminology. The purpose of this paper is precisely to show that the 'classical' approach to lexical structure has not been dominant throughout the whole history of lexical semantics as a branch of modern linguistics, but is rather typical of the structuralist period between the older historical-philological era and the birth of cognitive semantics. As far as Aristotle is concerned, the predicate 'Aristotelian' is justified by the Aristotelian (and scholastic) metaphysical distinction between accidental and essential attributes, with the latter constituting the true definition of a concept. One should not forget, however, that in his poetical and rhetorical treatises, Aristotle also has a lot of attention for topics such as metaphor, analogy, and the use of concrete examples or 'paradigms' in rhetoric, that come close to the interests of cognitive semanticists. The following list of characteristics gives only a rough overview of cognitive semantics. I trust that the other papers in this volume will give the reader an exemplary view of the cognitive approach. I will not try to prove or defend the theses listed below; I merely present them as a summary of what seem to be the major tendencies in cognitive semantic research on the lexical-semantic level, a summary that may serve as an aide-memoire to the comparison of cognitive semantics with the historical-philological tradition that will be the subject matter of section 4. It should be mentioned that there is a lot of variation within cognitive semantics; the views listed below need not all be held by the specific studies that can be situated within the cognitive framework. In this respect, cognitive semantics as a theoretical paradigm corresponds with its own views on categorization. One of its basic tenets is that lexical concepts are not rigidly defined by a set of essential attributes shared by all the instances of those concepts. Likewise, a rigid definition of cognitive semantics, consisting of an enumeration of attributes shared by all cognitive semantic studies is hard to give. Notice that this is in keeping with the paradigmatic conception of sci-
654 DIRK GEERAERTS ence defended by Kuhn, Lakatos et al., which considers scientific research programmes to be clusters of interconnected theories. (The methodological consequences of this analogy between the theoretical views of cognitive semantics and the metatheoretical views of the Kuhnian, paradigmatic theory of science are explored in Geeraerts 1985b.) What, then, are the central theses of cognitive semantics? 1. Lexical concepts have vague boundaries, in the sense that they contain peripheral zones round clear conceptual centres (versus: lexical concepts are discrete, well-defined entities). — The prototypical view of categorization implies that categories may have marginal instantiations that do not conform rigidly to the central cases. This extensional fact is parallelled by the intensional phenomenon that it may not be possible to define a concept by means of a set of attributes applicable to all the instances in the extension of that concept. Intensionally as well as extensionally, then, concepts may be characterized by the existence of a chiaroscuro-zone, either with regard to the category membership of instantiations of a category, or with regard to the definitional status of attributes. 2. Lexical concepts are polysemous clusters of overlapping semantic nuances (versus: the various senses of a lexical item can always be strictly separated from each other). —The vagueness that characterizes conceptual categories as a whole is mirrored by their internal structure. The blurred edges of lexical concepts as a whole are parallelled by the possible absence of clear boundaries among the senses that constitute the category as a whole. The most common expression of this view is the family resemblance model of categorial structure (Rosch & Mervis 1975). 3. The distinction between analyticity and syntheticity, and that between essential and accidental attributes cannot be rigidly maintained (versus: there is a strict distinction between analytical and synthetic statements, and between essential or accidental attributes). — This is a definitional consequence of the previous points. There need not be a set of attributes that is applicable to all the members of a category (and only those), and that defines the 'essence' of that category (all other attributes being accidental). Since the Aristotelian distinction between essential and accidental attributes is parallelled by the Kantian distinction between analytic and synthetic statements (at least on the lexical level), the dichotomous distinction
HISTORY OF LEXICAL SEMANTICS 655 between the latter cannot be maintained either, although there may be a gradient pattern of more or less inherent attributes. 4. Lexical concepts may be disjunctively defined (versus: lexical concepts are necessarily defined conjunctively). — This is again a consequence on the level of definitions of points 1 and 2. In the classical view, a concept is defined by a set of essential attributes all of which can be predicated of the members of the category in question. Within a prototypically organized category, however, categorial membership may be based on sufficient similarity rather than identity; some of the allegedly necessary ('essential') attributes of the central exemplars of the category may appear to be optional at the periphery. As Labov (1973) has remarked, traditional lexicography very often uses disjunctive rather than conjunctive definitions, a fact that clearly indicates the practical importance of disjunctive definitions for semantic description. 5. Attributes within (or exemplars of) a category may have different degrees of salience (versus: all attributes within the definition, or exemplars in the extension of a concept have an equal degree of salience). — Attributes not shared by all the members of a category (or a majority thereof) are less important definitionally than attributes that do appear in all or most of the exemplars of a category. These frequent attributes carry more weight within the semantic structure of the category. Analogously, some members of a category carry more weight as better exemplars of the category, in contrast with the peripheral, less salient exemplars. 6. Lexical concepts function in a flexible and analogical manner (versus: lexical concepts function in a rigid, algorithmical fashion). —The fact that category membership can be defined by similarity rather than identity entails that conceptual categories can be used in an extremely flexible way; making the criteria for the use of a category less strict enhances the usage potential of that category. One aspect of this extremely important characteristic of lexical concepts is discussed in my other contribution to this volume; there, I try to make clear that the flexibility inherent in prototypical categories as interpretative schemata with regard to experience, may well be one of the major functional, psychological reasons for having such a type of categorial structure. Another point to be mentioned is this: the flexibility of prototypical concepts bridges the gap between synchrony and diachrony
656 DIRK GEERAERTS (a moot point in structuralist conceptions of language) by incorporating the dynamism that is usually considered to be characteristic for a diachronic conception of language among the synchronic features of lexical concepts. This is, of course, a point that is of particular interest with regard to the comparison of historical-philological semantics and cognitive semantics. 7. Lexical concepts have to be studied as a proper part of human cognition at large (versus: lexical concepts have to be studied as a part of an autonomous linguistic structure). — What is at stake here is the fundamentally structuralist claim that the lexicon of a natural language contains a proper semantic structure that is independent of the broader cognitive organization of the human mind, and which it is the task of linguistic semantics to reveal; this implies that semantics is an autonomous discipline.14 By contrast, cognitive semantics claims that lexical concepts can only be studied adequately against the background of the human cognitive capacities at large, and in particular, that there is no specifically linguistic- semantic organization of knowledge, separate from conceptual memory in the broadest sense. If there is no autonomous linguistic-semantic memory, lexical semantic research cannot be done autonomously, but it has to be conducted in close cooperation with other sciences studying the human mind. Of these, psychology is obviously the most important one, but there are others such as Artificial Intelligence, neurophysiology, and cultural anthropology. 8. There is no distinction between semantic and encyclopaedic knowledge (versus: the semantic definition of a lexical concept is to be distinguished from the encyclopaedic data that can be connected with the latter). —This point is a specification of the former. The distinction between semantic and encyclopaedic knowledge (which parallels that between essential/analytic and accidental/synthetic information) is based on the distinction between an autonomous linguistic conceptual structure on the level where lexical items are defined, and conceptual memory in the broadest sense. If the latter distinction is discarded, the former also falls. 9. Semantic studies cannot ignore the experiential and cultural background of the language user (versus: semantic phenomena should be studied apart from user- or culture-specific background data). —This is a further consequence of the non-autonomous approach mentioned in point
HISTORY OF LEXICAL SEMANTICS 657 7. If language is one of the basic cognitive tools of man, it should not be studied autonomously, but it should be considered in the light of this cognitive function, i.e. of interpreting, ordering, retaining, and expressing human experience. This experience may, of course, be individually or culturally specific. This implies, e.g., that attention will be given to cultural differences in the metaphorical patterning of experience (see Lakoff & Johnson 1980), and that semantic universals will be seen as general strategies for coping with experience rather than as substantive universals. 4. Comparing the Cognitive and the Historical-Philological Tradition In the light of the foregoing sections, it should not be too difficult to recognize the similarities between the cognitive and the historical-philological approaches, particularly when they are set off against the 'classical' approach. 1. A first point of primary importance is the psychological orientation of both approaches. Both consider meanings to be primarily psychological entities, that is to say, they identify 'meanings' and 'mental concepts', as against an approach that strictly separates linguistic meanings and (encyclopaedic) psychological concepts. At the beginning of his Griechische Bedeutungslehre, Max Hecht sums up the disciplinary position of historical-philological semantics: Insofern sie zugunsten der Lexikographie die Bedeutungen in zeitlicher Folge ordnet und im Interesse der Etymologie die Gesetze der Bedeutungsanderung aufstellt, hat sie sprachwissenschaftlichen Wert. Soweit sie aber diese Gesetze aus der Natur des Geistes herleitet und eine Geschichte der Vorstellungen gibt — Bedeutungen sind Vorstellungen -, fallt sie auf das Gebiet der empirischen Psychologie15 (1888: 5). In a vigorously polemical article that is properly to be regarded as the first theoretical statement of structural semantics (with Trier's 1931 monograph on semantic fields being the first descriptive achievement), Leo Weisgerber (1927) blames Hecht for relegating semantics to the field of psychology. Treating meanings as psychological 'representations' {Vorstellungen), says Weisgerber, entails that meanings cannot be studied autonomously, as a proper part of linguistic structure.16 This structuralist desire for an autonomous approach to linguistic semantics is then again discarded by cognitive semantics in favour of a whole-heartedly psychological approach in which
658 DIRK GEERAERTS the study of linguistic structure is required to be in accordance with what is known about cognitive processing in general (see Langacker 1983: 1/6-7). An immediate consequence of the rejection of an autonomous approach is the fact that neither the historical-philological nor the cognitive tradition imposes a strict distinction between encyclopaedic and purely semantic data in lexical analyses. One may say that, at the time of historical-philological semantics, the distinction had not yet gained wide acceptance, whereas it is no longer generally accepted in contemporary lexical semantics (see Haiman 1980, Langacker 1983: 11/63). Another consequence of the psychological approach is the fact that experimental psychological research will have an influence on linguistic conceptions of categorial structure. Taking the cognitive nature of language seriously entails taking psychological studies of meaning seriously. It is, in fact, possible to identify major psychological influences on the philological and the cognitive research programmes. The latter's conception of categorial structure has to a large extent been shaped by Eleanor Rosch's experimental work on prototype semantics (cf. supra), whereas the former has been profoundly influenced by Wilhelm Wundt's apperceptionalist psychology (Wundt 1900). It is true, though, that Wundt's views have not been generally accepted by all semanticists of the historical-philological persuasion, but even when they are systematically criticised (e.g., Marty 1908), the tendency is rather to substitute another psychological conception for Wundt's, than to abandon the psychological outlook as a whole. A final consequence of the psychological orientation of the research programmes that we are trying to compare is their common attention for the experiential background of semantic phenomena. In historical-philological semantics, this is embodied in what I have called its hermeneutic character: it considers lexical meanings as an expression of a particular (individual or cultural) view of the world. As for cognitive semantics, its experiential nature has probably been brought to the fore most vigorously by George Lakoff in his attack on the classical, 'objectivist' conception of categorization (but see also, e.g., Langacker 1983: 1/6-7). In the following quotation, the link between the experiential nature of semantics and its psychological, non-autonomous orientation is clearly made apparent: After a generation of research in which it was implicitly assumed that language could be described on its own terms, it has become more interesting to ask how much of the structure of language is determined by the fact that people have bodies with perceptual mechanisms and memory and proces-
HISTORY OF LEXICAL SEMANTICS 659 sing capabilities and limitations, by the fact that people have to try to make sense of the world using limited resources, and by the fact that people live in social groups and have to try to communicate with each other (Lakoff 1982c: 155). Comparing the importance of experiential factors in historical-philological and in cognitive semantics, there may be a shift in emphasis from meanings as the expression of subjective experience to conceptual categories as interpretative schemata shaping experience, but the basic point is the same: language is an integral part of the intellectual, emotional, and social life of human beings, and this is reflected by the nature of linguistic categories. As such, these cannot be considered apart from the human endeavours to make sense (literally) of the world. 2. Next to the fundamental correspondence in the methodological orientation of historical-philological and cognitive semantics, it should also be mentioned that there is a lot of similarity between both approaches with regard to their conception of the nature of semantic categories, although this theoretical similarity is not as complete as the basic methodological one. One point that is quite clear, is the importance of polysemy for catego- rial structure. Both approaches tend to conceive of lexical concepts as polysemous clusters of senses, and to study them accordingly. In particular, if the main descriptive efforts of historical-philological semanticists are directed at the classification of semantic changes, this is because the mechanisms of semantic change are responsible for the polysemy of lexical items. (One might say: mechanisms of change such as metaphor and metonymy are the structural links between the various senses of an item.) It is no surprise, then, that the cognitivists' interest in polysemy leads to a revival of the research into some of the earlier semanticists7 favourite topics. This is, of course, most clearly true for the study of metaphor, but the phenomenon is not restricted to it. (There is, e.g., the rediscovery of sound symbolism by Lakoff 1983; see Kronasser 1952: 160-168 for an introduction to the older tradition's treatment of it.) Furthermore, the notions of vagueness and centrality that figure so prominently in prototypical semantics are far from absent in historical- philological theories. They may not receive the same emphasis as in present-day theories, but they are clearly present in a number of works in the historical-philological vein. They can be found, e.g., in Paul's influential classification of semantic changes (1880), and in Erdmann (1910), whose
660 DIRK GEERAERTS particularly lucid formulation of the characteristics of lexical concepts justifies an extensive quotation. Derm welche Theorien iiber Wesen, Bedeutung und Entstehung der Be- griffe man auch vertreten mag: vom Standpunkt der Logik wird man immer fordern miissen, dass sie eine unzweideutige, klare Grenze aufweisen, dass sie einen bestimmten Inhalt und Umfang haben. Und Begriffe dieser Art werden durch Worte nicht ohne weiteres bezeichnet. Worte sind vielmehr im allgemeinen Zeichen fur ziemlich unbestimmte Komplexe von Vorstel- lungen, die in mehr oder minder loser Weise zusammenhangen. ... Die Grenzen der Wortbedeutung sind vervaschen, verschwommen, zerflies- send. Treffender aber noch wird meines Erachtens der Sachverhalt geken- zeichnet, wenn man uberhaupt nicht von Grenzlinien des Umfangs redet, sondern ... von einem Grenzgebiet, das einen Kern einschliesst. Ver- anschaulicht man sich gewohnlich den Umfang eines logisch vollkomme- nen Begriffs durch eine scharfe Kreislinie, wie sie annahrend ein gutge- spitzter Bleistift erzeugt, so kann man sich die Abgrenzung eines Wortum- fanges durch einen mehr oder minder breiten, in sich zuriicklaufenden Streifen versinnlichen, wie ihn ein in Farbe getauchter Pinsel auf einer Flache hinterlasst^ (1910: 4-5). The correspondence with the contemporary prototypical view is too obvious to need much comment. The basic correspondence established in the previous points does not imply that there are no differences between the two research traditions compared here. The basic difference is the shift from a diachronic to a synchronic approach to semantics, mirrored by the fact that the emphasis shifts from the mechanisms of semantic change to the internal prototypical structure of polysemous lexical concepts. This shift is, of course, part of the legacy of structuralism, but one that is not accepted without fundamental alterations. In fact, the extent to which cognitive semantics differs from structuralism with regard to the diachrony/synchrony-distinction is more important than its primary adherence to that distinction. A strictly synchronic approach as favoured by structuralism tends to overlook the fact that languages are in constant flux.18 By incorporating the dynamic principles that structuralism can only conceive of diachronically into the syn- chronically flexible nature of lexical concepts, cognitive semantics can avoid the structuralist paradox that the true nature of linguistic structure is considered to be a steady state, whereas such a steady state can hardly be found in linguistic reality. The prototypical view of natural categories is so to say the synchronic side of the diachronic coin; what appears as diachronic change from one point of view, is merely the realization of syn-
HISTORY OF LEXICAL SEMANTICS 661 chronically available possibilities from another. In short, flexibility is the intersection of diachrony and synchrony at the level of lexical structure, and that is a point of view that may not have been completely absent in historical-philological semantics, but that still deserved to be emphasized as it is now done in cognitive semantics. In this respect, then, cognitive semantics is a fundamental improvement over the older tradition. But does this relationship also work the other way around, i.e., can cognitive semantics profit from the historical- philological tradition? The latter constitutes such a huge body of observations, classifications, and explanations (indeed, its gigantic proportions may well be the very cause of its present-day neglect), that it is hard to imagine that it contains nothing of interest for cognitive semantics. That a confrontation with historical-philological semantics may be very fruitful is a point I would like to illustrate on two different levels, first with regard to a very specific topic of research, and second with regard to a rather more general, and more fundamental aspect of semantics. a. Both Van Ginneken (1912) and Sweetser (1984) study the figurative transfers of (among many others) expressions connected with auditory experience. Both do so from an essentially cross-linguistic, etymological point of view, but their perspective is slightly different. Sweetser concentrates on words for the act of hearing itself, and notices among other things that verbs for the concept 'to hear' shift towards the notion 'to listen, to heed', and further to 'to obey'. She also mentions that derived words coming from 'hear'-roots regularly mean 'tale, report, fame, glory, news'. Van Ginneken considers attributes for the things we hear, and points out that auditory characteristics may be transferred, among others, to visual sensations (loud colours, a loud dress), to tactile sensations (Dutch dovenetel 'dead nettle' from doof, 'deaf'), or to emotions (to quieten someone's fear). A confrontation of Van Ginneken's and Sweetser's results yields a particular question for further research: do the patterns of figurative transfers in both sets of expressions correspond with each other, and if they do not, what cognitive reason can there be for conceptualizing one set differently from the other? For instance, if 'loud' is an attribute for things that can be heard well, and if 'to be heard' shifts towards 'to be famous', is it reasonable to expect words for the notion ioud' to be applicable to fame or famous things and persons? Conversely, if attributes for noises are applicable to non-auditory perceptual sensations, is it also possible to express the notion
662 DIRK GEERAERTS 'to experience a non-auditory perceptual stimulus' with words that etymologically mean 'to hear'? These questions need not, of course, be answered here; the important thing is to see how they arise as specific research hypotheses from a confrontation of parallel studies in cognitive semantics and in the prestructuralistic tradition of historical semantics. b. One of the major questions in prototypical semantics concerns the internal coherence of natural categories. What keeps a prototypically organized category together, and more particularly, how far can the senses in a polysemous cluster be apart from each other or from the prototypical centre before they 'drop out' of the category? The question can be reformulated in terms of the relations that hold among those senses: what kind of relations are acceptable? Are they gradable, and if so, to what extent can they hold? So far, cognitive semantics has answered these questions almost exclusively in terms of similarity (obviously a gradable notion); think of a notion such as 'family resemblance', of the statement that prototypically organized concepts are clusters of overlapping senses (the overlapping senses are similar to the extent that they have a number of features in common), and of the cognitivists' interest in metaphor. Now, it would not be implausible to equate the relations that can hold among the diverse senses within a natural category, with the mechanisms of semantic change identified by diachronic semantics. In fact, if the flexibility of prototypical categories is indeed a phenomenon of the same type as the semantic changes of old (as I have argued above), the same type of mechanisms should govern both. The mechanisms that have been studied by historical-philological semantics as types of diachronic change are then incorporated into synchronical categories as the mechanisms that underly their flexibility and that specify the relations between the original senses and the new ones that arise from the flexible use of the category. These mechanisms determine and restrict the range of possible extensions of a concept, given a prototypical centre or an existing cluster of applications. As such, they are interpretative principles, strategies that help one get from existing knowledge to new, contextually functional applications of a concept. But historical-philological semantics has revealed the existence of much more mechanisms of change than mere similarity (though the latter is undoubtedly one of the major types). The associationist principles leading from existing senses to new ones also comprise relations of contiguity,
HISTORY OF LEXICAL SEMANTICS 663 which form the basis of what is traditionally known as metonymy (including synecdoche), and relations that might be called mimetic because they presuppose the existence of another expression that already has the new sense of the word we are investigating, and with which the investigated word is associated for one reason or another. A typical example of this mimetic type of change is semantic borrowing: the Gothic word daupjan is associated with the Greek word baptizein through their common basic meaning 'to cause to go down, in particular, to submerge', and next copies the secondary sense 'to baptize' of the Greek word. Other examples of mimetic change are ellipsis (in which case the association is syntactic), the semantic merging of homonyms (where phonetic associations are involved), and parallel developments among antonyms (again a case of an initial semantic association). (See Geeraerts 1986: 38-41 for a more thorough treatment.) I believe it will be one of the major tasks for lexical semantics in the next few years to study the relevance of these traditional associationist mechanisms of diachronic change for the cognitive theory of categorization. Do they have the same cognitive status as associations through (literal or metaphorical) similarity, in providing principles that guide the interpretative efforts of the hearer, enable the speaker to use lexical concepts flexibly, and define the structural coherence of natural categories? Or do prototypical categories only allow certain types of relations, so that semantic extensions based on principles of contiguity or lexical mimesis immediately lead to the creation oj^another category, distinct from the original one? Again, the question need not (in fact, could not) be answered here. It suffices to see that the confrontation with the historical-philological tradition opens up a major area of research for cognitive semantics. In short, then, the affinity between cognitive semantics and historical- philological semantics also seems to entail that the former may profit from a closer acquaintance with the latter. To deepen our insight into the affinities between both approaches, we shall now have a closer look at the history of lexical semantics as a whole. 5. A Survey of the History of Lexical Semantics Near the end of his book A history of semantics, W.T. Gordon remarks that semanticists have tended to take a tabula rasa approach to their discipline, either out of a conscious rejection of previous theories, or because
664 DIRK GEERAERTS they were simply unaware of the work of their predecessors (1982: 244). However, Gordon's remark is more indicative of his own failure to identify major schools and traditions in the history of semantics, than of the actual absence of such major trends. At least in lexical semantics (and let us not forget that up to the incorporation of semantic problems in transformational grammar, semantics had been primarily lexical), it is possible to point out the existence of a number of global methodological perspectives that shape the semantic endeavours of a majority of researchers within a particular period. Figure 1 gives an approximation of what such a methodological history of lexical semantics might look like. The justification for the figure can only be dealt with summarily here; a more detailed treatment of the major historical trends in lexical semantics can be found in Geeraerts (1986). Next to the historical-philological and the cognitive approach, the figure identifies two other basic methodological approaches to lexical semantics. The logical approach to lexical semantics is clearly a minor tradition. When it is not primarily a formalisation of descriptive ideas developed within structuralist semantics (cf. Dowty 1979), its descriptive contribution to lexical semantics resides mainly in the analysis of lexical items corresponding with logical operators (such as quantifiers and connectives), and in the analysis of the syntactic meaning of word classes (the 'types' of Montague semantics). The proponents of Montagovian semantics (which is the most sophisticated form of logical semantics) have themselves expressed serious doubts with regard to the applicability of injensional logic to word meaning beyond the level of the syntactic meaning of word classes (cf. Thomason 1974: 48). Regardless of the results achieved by logical semantics in other domains, its contributions to lexical semantics are as yet too restricted to consider it a major trend in the history of lexical semantics. Structuralist semantics is undoubtedly the most diverse of the four approaches mentioned in the figure. Disregarding the distributionalism of, among others, Dubois (1964) and Apresjan (1966), which was never very successful, there are three main approaches to be mentioned within structuralist semantics. Two of the alternatives are clearly present in the debate between the proponents of a decompositional and an axiomatic approach to the formalisation of lexical meaning, such as was held within the framework of transformational grammar in the course of the 1970s (see, e.g., Katz & Nagel 1974 versus Fodor, Fodor & Garrett 1975). For reasons to be explained in the following section, I consider transformational seman-
PERIOD Fromc. 1870 toe. 1930 Fromc. 1930 toe. 1975 Fromc. 1970 onwards Fromc. 1975 onwards THEORY HISTORICAL- PHILOLOGICAL SEMANTICS STRUCTURAL SEMANTICS LOGICAL SEMANTICS COGNITIVE SEMANTICS OBSERVATIONAL PRINCIPLE Language as the self-expression of an individual or a community Language as an autonomous structure Language as reference to possible worlds Language as a cognitive tool EMPIRICAL DOMAIN Statements about meaning change Statements about paradigmaticand syntagmatic semantic relations Statements about the truth conditions of propositions Statements about the flexibility, internal structure, experiential nature, encyclopaedic character of lexical concepts EXPLANATORY PRINCIPLE Principles of expressivity and efficiency Meaning decomposition, or meaning postulates Algorithmical compositionality Natural categorisation (prototypicality) and cognitive strategies j Figure 1
666 DIRK GEERAERTS tics to be the methodological culmination of structuralist semantics. The axiomatic and decompositional alternatives within transformational semantics, then, are related to earlier forms of structural semantics. Componen- tial analysis is an outcome of the semantic field approach made popular by Trier (1931), in the sense that the conceptual dimensions from which components are derived, are primarily the distinctive oppositions within a semantic field. The field origin of componential analysis is less clearly visible in Katz & Fodor's introduction of the method into transformational grammar (1963), but it is quite obvious in Lounsbury's (1956) and Goodenough's (1956) anthropological studies, in which the componential method for meaning analysis was applied for the first time, and from which Katz & Fodor seem to have borrowed the idea. The field-theoretical origin of componential analysis is equally apparent in the introduction of the method by European linguists such as Pottier (1963) and Coseriu (1964). It is not quite clear, incidentally, whether this European branch of componential analysis came into being independently of Lounsbury, Goodenough, and/or Katz & Fodor. On the other hand, the axiomatic (or 'meaning postulate') approach is methodologically related to earlier purely structuralist work by Lyons (1963, 1968), in which relations such as hyponymy, synonymy, and antonymy among unanalyzed lexical items are taken to be the core structure of the lexicon (rather than analytic relations among the items in semantic fields). It is extremely important to realize that the decompositional and the axiomatic approach are compared here, not as formal methods of description, but as methodological perspectives on word meaning. In particular, they methodologically define different observational bases for lexical semantics: the componential approach focusses on the structural relations among the items in semantic fields, whereas the axiomatic approach focusses on structural relations of the type represented by hyponymy and synonymy. As such, the point to be made here is simply that structural semantics may focus on different types of semantic structure. Notice in this respect that componential analysis and meaning postulates as formal methods of description are not exclusively tied up with the different observational approaches with which they are primarily connected. In fact, componential analysis can be used to explain the synonymy, antonymy etc. of lexical items (this is the Katzian position), whereas meaning postulates may serve as a descriptive method for indicating distinctive oppositions within
HISTORY OF LEXICAL SEMANTICS 667 semantic fields. Also, as methods of formalisation, the origins of the decompositional and the axiomatic approach do not lie with structural semantics, but in phonology and logic, respectively. In short, the methodologically important point is not the formal method of description, but the difference in observational basis with which it is usually connected: structuralists may focus on semantic field structures, or on structural relations such as hyponymy, antonymy, and the like. Next to these two types of paradigmatic structures, they may direct their attention at a third type of structure, viz. the syntagmatic semantic relations that have become known in transformational grammar as selec- tional restrictions, but that had actually been discussed already by Porzig (1934)." There will be much more to say about structural semantics in the following section, but first, our brief discussion of Figure 1 has to be concluded with a note on the chronological periods distinguished in it. These indicate the heyday of each of the traditions mentioned, i.e., the period in which each of the methodological perspectives was most intensively applied and in which it constituted (one of) the major approach(es) to lexical semantics of that era. This implies that studies within a particular methodological paradigm may well be found outside the periods mentioned in the figure, at a time when that paradigm is no longer (one of) the major one(s) in lexical semantics. Thus, componential approaches to meaning description can still be found today (see, e.g., Lipka 1985), but it would be misleading to say that structuralist componential analysis is the leading, theoretically most advanced conception of lexical semantics. Likewise, the fact that the classification takes its starting-point around 1870 does not mean that semantic studies are completely absent in the previous period, but merely that it is around that date that semantics begins to receive frequent and continuous attention, and that lively discussions on semantic matters begin to take place. Still, I would like to stress that the classification is intended to cover only the history of linguistic lexical semantics, i.e., semantic studies that can be situated within linguistics as a discipline that combines empirical description with theoretical analysis. In this sense, the older (roughly, pre-19th century) traditions of semantics (speculative philosophy, practical lexicography without theoretical interests, and the rhetorical classification of stylistic tropes) are not envisaged here.
668 DIRK GEERAERTS 6. Main Lines in the History of Lexical Semantics The survey of Figure 1 is based on a methodological analysis of the various approaches. The three headings used to distinguish among them (empirical domain / observational principle / explanatory principle) are well- known notions in the philosophy of science. They correspond roughly with the questions 'What kind of phenomena does a theory deal with?', 'From what point of view are those phenomena described?', and 'What are the basic principles used in explaining the empirical observations?*. As with any classificatory perspective, this methodological approach entails certain choices that might be reversed from another point of view. For instance, Generative Semantics and Interpretive Semantics are not mentioned separately because they are distinguished mainly by the position they want a semantic component to take within a transformational grammar, but otherwise share the basic methodological assumptions of transformationalist semantics. Even the latter is not singled out for separate treatment, in spite of its mentalist views, which seem to contradict the initial anti-psychological position of structuralism. Initially, however, the mentalist self-definition of transformational semantics hardly played a major methodological role, among other things because introspection had always been tacitly accepted as one of the methodological techniques of semantics, even by structuralists. (Distributionalists are the possible exception, but then, their failure merely shows the difficulty of avoiding the introspective method.) Methodologically speaking, transformationalist semantics is a culmination of, rather than a departure from, the structuralist approach. The Katz- ian definition of the methodology of lexical semantics (1972) is typically structuralist in that it assumes the existence of an autonomous level of semantic structure that is distinct from cognition at large and that underlies the linguistic judgements of native speakers on such matters as implication, homonymy, synonymy and the like. The basic methodological tenet of structuralist semantics, the fundamental point by which it defines itself in contrast with the older historical-philological tradition is the following (see Weisgerber 1927): meaning is not primarily a psychological, but a linguistic phenomenon; hence, it is by studying linguistic phenomena that semantic structures can be discovered. The whole point is easily illustrated by the notion 'distinctive opposition'. Given a number of semantically related words within a lexical field, a structural analysis will have to reveal the dis-
HISTORY OF LEXICAL SEMANTICS 669 tinctive oppositions among the items in question, and stop there, discarding other attributes as linguistically irrelevant. In a restricted set of items such as ram : ewe :: stallion : mare, the opposition ±MALE is distinctive. The fact that ram and ewe are covered with relatively long hair from which clothes can be made, whereas stallion and mare are not, is then considered to be linguistically irrelevant: it is part of our encyclopaedic knowledge of the world, but it is not a part of the structure of language. Structural semantics originates by making precisely that distinction, viz. between semantic structure (a linguistic phenomenon involving necessary and sufficient conditions for explaining lexical and semantic relations) and cognition (a psychological phenomenon involving encyclopaedic knowledge). By adhering to the idea that there is such an autonomously linguistic level of semantic structure, transformational semantics is basically structuralist. Next, what makes it a culmination of the structuralist approach is the range of linguistic facts it indicates as the observational basis of lexical semantics. In fact, it brings together the three major kinds of facts to which structuralist semanticists had addressed themselves: the semantic field relations with which structuralist semantics originated and which dominated structuralist semantics up to the beginning of the 1960s, the lexical relations highlighted by Lyons (1963), and the syntagmatic aspects of semantic structure which had once been indicated by Porzig (1934), but which had more or less20 been forgotten until they were taken up simultaneously in the com- ponential theories of Katz & Fodor (1963) and Pottier (1963). It is of course to these syntagmatic aspects of semantic structure, embodied in the notion of selection restrictions, that most of the transformationalist attention was directed, at least in the sense that it was precisely the role of semantics with regard to syntax that engendered the debate between Generative and Interpretive Semantics. However, the methodologically basic point is that the Katzian definition of the methodology of lexical semantics (1972) incorporates into the observational basis of semantics all three kinds of structural relations that had been studied separately by pretransformationalist structuralists. It is the linguistic judgements of native speakers with regard to these relations that reveal the autonomous semantic structure of language; conversely, a formal description of that autonomous level of semantic structure explains the behaviour of the native speakers. It is true, though, that Generative Semantics gradually moved towards a position in which the autonomy of linguistics appeared to be threatened. The growing importance given to semantic factors within the transforma-
670 DIRK GEERAERTS tional framework runs parallel with a move away from the original ideal of linguistic autonomy (an ideal that is no less paramount in the Chomskyan conception of grammar than in the structuralist conception of lexical semantics). One need only recall, e.g., Lakoffs incorporation of pragmatic considerations in the form of transderivational constraints (1970) to appreciate the extent to which the Generative Semantics movement slid away from the autonomistic ideal. To the extent that Generative Semantics threatened to violate that basic ideal, it is right to say, following Newmeyer (1980), that Generative Semantics destroyed itself, but only with the proviso that it destroyed itself as a branch of transformational grammar. It is remarkable, in fact, that some of the most prominent figures of Cognitive Grammar (in particular, Lakoff and Langacker) were formerly much closer to the Generativist than to the Interpretivist approach. As an approach within the transformational paradigm, Generative Semantics disappeared because it became incompatible with some of the basic assumptions of transformational grammar, but the primacy it gave to meaning now reappears within a new school of thought that has abandoned the autonomistic creed. We can see, then, that the transition from transformational to cognitive semantics does not occur at random, but is determined by substantial points of debate. Let us therefore have a closer look at the developmental lines that interconnect the traditions identified in Figure 1. The transition from the first to the second stage has already been described: in trying to establish an autonomously linguistic approach to the study of meaning, structuralist semantics rejected the historical-philological 'confusion' of encyclopaedic cognition and semantic structure. And of course, the structuralist primacy of synchronic analysis also favoured abandoning the older, primarily diachronic approach. To see how structural semantics gave way to the logical paradigm, it is once more the incorporation of lexical semantics into transformational grammar that has to be taken into account. In line with the general tendency of generative grammar, this incorporation led to an increased interest in the formalisation of semantic analyses. Justified criticism with regard to Katz's formalisations (see, e.g., Bierwisch 1969) induced people like McCawley (1971b) and Lakoff (1972) to take over the more sophisticated formalisms of predicate logic. When it was recognized that logical and semi- logical formalisations alike have to be referentially interpreted (see e.g. Lewis' 1972 well-known sarcasm with regard to the uninterpreted 'Mar-
HISTORY OF LEXICAL SEMANTICS 671 kerese' of earlier transformationalist semanticists), the gradually spreading interest in logic eventually led to the adoption of the intensional-logical format of Montague grammar. The transition from the structuralist to the cognitive paradigm can be attributed to two factors, one of which has already been mentioned: the Generative Semanticists' tendency to broaden the empirical scope of generative grammar contained a movement away from the autonomy aspired at by Chomskyan linguistics. Next to this, cognitive grammar can be considered as taking seriously the methodological consequences of the mentalism proclaimed so vigorously by transformational semantics. Although (as we have seen) the basic methodological position of transformational semantics is in line with the autonomistic views of structuralism, its mentalist pretentions did become methodologically relevant at a particular point in its development. In fact, the mentalist strand in transformational semantics became important in the debate between the decomposi- tional and the axiomatic approach to meaning representation when the choice between both approaches came to be seen as relying on the psychological reality of a set of primitive, non-decomposable semantic features (the basic vocabulary of componential representations). As might be expected, the discussion about the psychological reality of the set in question moved away from the transformational framework to a more purely psychological approach. That is to say, through the appeal to experimental psychological evidence, the discussion on meaning representation was taken over by psychologists pure and simple. (See, e.g., the debate between Glass & Holyoak 1975 and Smith, Shoben & Rips 1974.) But psychologists obviously do not bother about the ideal of linguistic autonomy: in their survey of psycholexicology, Smith & Medin (1981) explicitly suggest that a psychologically adequate representation of linguistically encoded concepts may have to rely on a complete representation of all the cognitive capacities of man. In this sense, taking th? psychological implications of the transformationalist mentalism methodologically seriously carries one away from the autonomist approach of transformational grammar. This is, of course, precisely the position taken by cognitive semantics. If semantics (as grammar in general, according to Chomsky) is indeed a psychological discipline, linguistic meaning can no longer be studied in isolation from the other cognitive capacities of man, but it will have to be considered in the light of its basic psychological function, e.g. as one of the basic cognitive tools of the human mind.
672 DIRK GEERAERTS In this sense, then, transformational semantics represents the grandeur et decadence of structural semantics. On the one hand, it is the methodological culmination of the latter, but on the other, its predilection for formali- sation and its mentalist pretentions embody the springs of its own dissolution. In particular, by giving up the earlier anti-psychological views of structuralism in favour of a mentalist self-conception, transformational semantics undermines its own autonomistically structuralist methodology. The developmental lines set out here are summarized in Figure 2; in bringing out the pivotal position of transformational semantics, the figure clearly refines the classification of Figure 1, in which transformational semantics is incorporated into the structural approach without further nuances. On a more general level, we can now see that the history of lexical semantics exhibits an oscillation between an autonomous and an experiential, psychological approach. Structuralist semantics reacted against the psychological perspective of historical-philological semantics, but nowadays, the pendulum is swinging back (partly out of an attempt to take the mentalist claims of neostructuralist transformational semantics methodologically seriously). It should furthermore be noted that, with regard to this basic dichotomy, logical semantics ranges with structural semantics for two interconnected reasons. First, logical semantics is definitely not a psychological approach. Its subject matter (the truthfunctional relationship between linguistic expressions and the world) is defined regardless of the cognitive subject: the truth conditions of a proposition exist regardless of the person(s) entertaining that proposition, and can be described accordingly. Second, by abstracting away from subjective, psychological factors, structuralism and logical semantics alike tend to be objectivist approaches. In both cases, meaning is primarily defined as an objective entity that is independent of the individual mind: the supra-individual structure of language, or possible-world truth conditions. As opposed to this, historical-philological semantics and cognitive semantics take into account the subjective, individual Sitz im Leben of language. The opposition between 'objectivist' and 'subjeetivist' approaches leads us back to the Diltheyan distinction between, respectively, 'natural' and 'human' sciences (see section 2). Against the background of the history of lexical semantics, the fundamental methodological affinity between cognitive and historical- philological semantics seems to reside primarily in the fact that, through their experiential, psychological orientation, both are subjective, 'human' sciences in the Diltheyan sense, in contrast with the objectivist tendencies
HISTORY OF LEXICAL SEMANTICS 673 HISTORICAL- PHILOLOGICAL SEMANTICS meaning should be studied from a synchronic, non- psychological, autonomous point of view STRUCTURAL SEMANTICS paradigmatic and syntag- matic relations reveal the autonomous semantic structure of language TRANSFORMATIONAL SEMANTICS mentalism formalisation a psychologically adequate study of meaning requires a non-autonomistic approach a formally adequate representation of meaning requires a model- theoretic interpretation \ COGNITIVE SEMANTICS LOGICAL SEMANTICS Figure 2
674 DIRK GEERAERTS of structural and logical semantics, which seem to take their methodological models primarily from the 'natural' sciences. This paper is obviously not the place to decide the issue between both basic approaches; the important thing is merely to see how such a major methodological dichotomy underlies the similarities between prestructuralist and cognitive semantics that we have described in section 4. 7. Conclusion In the foregoing pages, the following theses have been developed. Primo, Lakoffs statement that "The classical theory of concepts and categories has been accepted in the West for two thousand years" (1982a: 99) overstates the novelty of the cognitive approach. Within the history of lexical semantics as a branch of linguistics, it is rather a partial return to the methodological position of the prestructuralist, historical-philological tradition of semantic research. Secundo, the major methodological alternative in the history of lexical semantics is a Diltheyan dichotomy between semantics as an objectivist, autonomistic, and as an experiential, psychological approach. Tertio, the historical-philological tradition contains a huge amount of empirical observations, classificatory mechanisms, and explanatory hypotheses that may be extremely valuable for the further development of cognitive semantics. It is precisely because of the large quantity of studies in the historical- philological paradigm, that this article has had to be restricted to some main lines. Nevertheless, I hope that it will have become clear that a closer look at prestructuralist semantics will be worthwhile. This conclusion should not, however, obscure the fact that cognitive semantics also takes some fundamental steps forward in comparison with the older tradition. For instance, Gestalt psychologists have long been aware of the figure/ground alignment in perception, but its importance for lexical-semantic analysis has come to be appreciated only recently through the work of Langacker and Talmy. Similarly, the prototypical conception of lexical structure opens up hitherto scarcely explored possibilities for bridging the gap between the synchronic and the diachronic approach to linguistic analysis. Although cognitive semantics shares its fundamental methodological position with historical- philological semantics, it is much more than a mere return to prestructuralist semantics.
HISTORY OF LEXICAL SEMANTICS 675 Notes 1. "Semasiology, or semantics, is that part of linguistics that tries to systematically collect the histories of all words, to select and classify them, in order not only to try to discover the deeper psychological and social causes of those histories, but also to compare these among each other, so that an insight into some few generally valid semasiological laws can be attained. But because words can only exist and develop through our human psychological and social life, these will be psychological and socio-psychological laws." 2. It may be useful to point out that I use the term "philological" mainly in a methodological sense; I intend it to be the indication of a methodological approach to the study of semantic phenomena, and not just of a topic or field of research. In this sense, the studies grouped under the label "philological" need not be concerned with the study of a particular text or a particular literary tradition; rather, they are by and large supposed to share some important methodological characteristics with philological studies in the narrow (literary) sense. In particular, the most typical representatives of this line of semantic research are concerned with the historical interpretation of linguistic material as the expression of thought, philology in the narrow sense being concerned mainly with the historical interpretation of literary works as the expression of an individual author's views. (Needless to say, this formulation refers to the characteristics of literary philology in its pre- structuralist stage; in our own century, literary hermeneutics has changed just as much as linguistic semantics. The changes are at least partly similar: by moving away from a romantic insistence on the individuality of the author to a modernistic interest in the literary work as an autonomous "textual object", the development of literary scholarschip parallels the linguistic transition from the historical towards the structural study of meaning-) 3. Neither does my treatment of Breal imply that his ideas are completely novel within the history of linguistics at large. As was pointed out to me by Pierre Swiggers (to whom I am grateful for a number of interesting remarks with regard to an earlier draft of this paper), it may be revealing to link Br6ars views with 18th century thought. In view of the chronological limits of this paper, this is a line which will not be pursued here. 4. "If one restricts oneself to the study of the changes of vowels and consonants, this discipline is reduced to a secondary branch of acoustics and physiology, if one merely enumerates the losses suffered by the grammatical mechanism, one creates the illusion of a building tumbling into ruins; if one hides behind vague theories about the origin of languages, one adds, without much profit, a chapter to the history of systems. There is, it seems to me, something else to be done... Linguistics talks to man about himself: it shows how he has constructed, how he has perfected, through difficulties of all sorts and in spite of an inevitable inertia, in spite even of temporary retreats, the most indispensable tool of civilisation." 5. "Only history can give to the words the degree of precision that we require to understand them adequately." 6. Different approaches within the etymological study of words might be used to identify particular schools, trends, and developments within the historical tradition of lexical research that have not been mentioned separately in this article. Thus, the Worter & Sachen-movement, and the dialect-geographical method introduced by Gilli6ron deserve explicit treatment in a complete history of lexical semantics.
676 DIRK GEERAERTS 7. "Language makes thought objective." 8. "Language is a translation of reality, a transposition in which particular objects only appear through the intermediary of the generalising and classificatory efforts of thought." 9. "We do not doubt that linguistics, giving up its paradoxical prejudices, will treat more fairly the primary forces in languages, i.e. ourselves, human intelligence. The mysterious transformation that makes French grow out of Latin (just as Persian out of Zend, and English out of Anglo-Saxon), and that everywhere shows a remarkable set of similarities and parallelisms with regard to its essentials, is not simply the product of the decay of sounds and the wearing off of endings. Behind these phenomena in which everything seems to speak of destruction, we feel the active efforts of human thought liberating itself from the form in which it is constrained, trying to modify it, and very often turning to its advantage what at first sight appears to be mere loss and decay. Mind moves matter." 10. "The goal, as far as language is concerned, is to be understood. During months, the child execrcises his tongue to produce vowels, to articulate consonants: how many failures, before he can clearly pronounce a syllable! On the grammatical level, innovations are of the same sort, with this difference that an entire people is involved. How many clumsy, incorrect, obscure constructions, before the one is found that will be, not the perfect expression of thought (there is none), but at least a sufficient expression of it!" 11. Still, readers who may have doubts as to the pervasiveness of the psychological outlook in diachronic semantics will find Weisgerber's attack on historical semantics in general and psychologism in particular (1927) extremely instructive. 12. "If one admits that there is a distinction between the historical and the natural sciences, that is, if one considers man as being the subject matter of a separate chapter of our study of the universe, language (which is the product of man), cannot stay on the other side, and linguistics will inevitably be a branch of the historical sciences." 13. At the same time, Breal's insistence on the fact that linguistics is a human instead of a natural science, is a reaction against the biological naturalism of Schleicher, according to which language is an organism evolving independently of man. 14. This formulation explicitly uses the phrase fundamentally structuralist claim instead of fundamental structuralist claim to account for the fact that the Bloomfieldian, behaviourist branch of structuralism does not make the claim, but instead discards the semantic description of linguistic forms from the domain of linguistics. The point is this: to the extent that structuralists accept the feasibility of a semantic description of language, they typically posit the existence of an autonomously linguistic, non-encyclopaedic lexical-semantic structure, but this feasibility is not accepted by all forms of structuralism. References to structuralist semantics in this paper obviously exclude those structuralists that do not accept the possibility of a linguistic semantics. (Moreover, it should be kept in mind that not all semantic studies by linguists that take an overall structuralist approach need be typically structuralist on the methodological level: see note 20.) 15. "(Semantics) is linguistically valuable to the extent that it chronologically classifies meanings in the interest of lexicography, and writes down the laws of semantic change in the interest of etymology. To the extent, however, that it derives these laws from the nature of the mind and that it writes a history of ideas — meanings are ideas —, it falls within the realm of empirical psychology."
HISTORY OF LEXICAL SEMANTICS 677 16. There are two additional remarks to be made with regard to Weisgerber's comments on Hecht. In the first place, 19th century linguists sometimes make a distinction between Begriff (concept in the most neutral sense) and Vorstellung (concept as a psychological entity, mental representation). As such, Hecht's use of the term Vorstellung is itself an indication of his psychological outlook. In the second place, Weisgerber's major objection against the older tradition of semantics is the following: identifying meaning and cognition entails that meanings exist and develop independently of language; meanings (thoughts) have an independent psychological life, and words are simply attached to them as labels. It may be good to mention that this objection is ill-founded. Identifying semantic and encyclopaedic knowledge need not imply that language may not help to shape and support thought, nor does it entail that cognition develops autonomously, without the help of the symbolical tools provided by language. On the contrary, it is precisely the acceptance of an intimate connection between language and cognition that makes it possible to envisage a relationship of reciprocal influence between them. (For an elaboration of this point, see Geeraerts, 1985b, sections 1.4. and 1.6.) 17. "But whatever theories about the essential nature, the meaning and the origin of concepts one may adhere to, from the point of view of logic one will always have to require that they exhibit unambiguous, clear boundaries, that their extent and content are clearly specified. But words do not simply indicate concepts of that kind. Words in general are rather signs for fairly unspecific complexes of psychological representations that belong together more or less loosely. ...The boundaries of word meanings are vague, unclear, indeterminate. The situation is, I think, even more adequately described if one simply does not talk about the borderline of the range (of a word), but ... if one talks about a border area that includes a central area. If one usually represents the range of a logically perfect concept by means of a sharp borderline, such as is approximately drawn by a well-sharpened pencil, one can represent the boundary of the content of a word by means of a more or less wide, gradually narrowing band, such as is produced on a flat surface by a brush that has been dipped into paint." 18. To avoid misunderstandings: the "synchronic" character of the structural method does not exclude diachronic studies, but conceives of change as a succession of distinct synchronic stages. As such, the investigation into synchronic structures is taken to be methodologically basic with regard to the study of the transition between synchronic stages. However, such stages are practically very hard to identify, since linguistic systems change neither en bloc nor abruptly. 19. The structuralist views of Reichling (1935), which come close to the prototypical view (see Geeraerts 1983c), will not be discussed as they do not belong to 'mainstream' structuralism. 20. "More or less": the Firthians might be mentioned (with some provisos) as an exception.
REFERENCES Ackrill, John L. 1963. Aristotle: Categories and De Interpretatione, translated and annotated by J.L. Ackrill. Oxford: Clarendon Press. Allen, Robert L. 1966. The Verb System of Present-Day American English. The Hague: Mouton. Anderson, John M. 1971. The Grammar of Case. Towards a Localistic Theory. Cambridge: Cambridge University Press. 1977. On Case Grammar. Prolegomena to a Theory of Grammatical Relations. London: Croom Helm. Anderson, Stephen R. 1977. "Comments on the Paper by Wasow". Formal Syntax ed. by Peter W. Culicover, Thomas Wasow & Adrian Akmajian, 361-377. New York: Academic Press. Apresjan, Juri. 1966. "Analyse distributionnelle des significations et champs semantiques structures". Langages 1. 44-74. Arensen, Jon. 1982. Murle Grammar (= Occasional Papers in the Study of Sudanese Languages No. 2). Juba: University of Juba Press. Armstrong, David. 1981. "The Ancient Greek Aorist as the Aspect of Countable Action". Tedeschi & Zaenen 1981. 1-12. Armstrong, Sharon L., Lila R. Gleitman & Henry Gleitman. 1983. "What Some Concepts Might not Be". Cognition 13. 263-308. Arnold, E. 1952. "Zur Geschichte der Suppositionstheorie". Symposion 3. 1-134. Aronoff, Mark. 1978. "Lexical Representations". Papers from the Parasession on the Lexicon 12-25. Chicago: CLS. Bach, Emmon. 1981. "On Time, Tense, and Aspect. An Essay in English Metaphysics". Radical Pragmatics ed. by Peter Cole, 63-81. New York: Academic Press. . 1986. "The Algebra of Events". Linguistics and Philosophy 9. 1. 5-16. Bailey, Charles-James N. 1984. "How Can Grammar Be Learned if the 'Present Tense' Is neither Present nor Tense"? Working Papers in Linguistics. TU Berlin. 228-245. Bakker, W.F. 1970. "The Aspectual Differences between the Present and Aorist Subjunctives in Modern Greek". Ellinika 23. 78-108. Barri, Nimrod. 1981. "Aorist Present 'Subjunctive' and Converters in Modern Greek". Studies Presented to Hans Jakob Polotsky ed. by Dwight W. Young, 1-20. Beacon Hill, Mass.: Pirtle & Poison. Bennett, David C. 1975. Spatial and Temporal Uses of English Prepositions - An Essay in Stratificational Semantics. London: Longman. Berlin, Brent & Paul Kay. 1969. Basic Color Terms: Their Universality and Evolution. Berkeley: University of California Press. Bever, Thomas G. & Peter S. Rosenbaum. 1970. "Some Lexical Structures and Their Empirical Validity". Readings in English Transformational Grammar ed. by Roderick A. Jacobs & Peter S. Rosenbaum, 3-19. Waltham, Mass.: Ginn.
680 REFERENCES Bierwisch, Manfred. 1967. "Some Semantic Universals of German Adjectivals". Foundations of Language 3. 1-36. . 1969. "On Certain Problems of Semantic Representations". Foundations of Language 5. 153-184. . 1984. "Dimensionsadjektive: Semantische Struktur und begriffliche Interpretation". Zeitschrift fiXr Phonetik, Sprachwissenschaft und Kommunikationsforschung 37. 490-512. 1985. "La nature de la forme semantique d'une langue". DRLAV. Revue de Linguistique 33. 5-24. Block, Ned, ed. 1981. Imagery. Cambridge, Mass.: MIT Press / Bradford. Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart, Winston. Boas, Franz. 1938. "Language". General Anthropology ed. by F.Boas et al., 124-145. Boston: Heath & Co. Boehner, Philoteus. 1958. "A Medieval Theory of Supposition". Franciscan Studies 18. 240-289. Boethius of Denmark (ed. 1969). Modi significandi, sive Quaestiones super Priscianum maiorem ed. by Jan Pinborg & Harry Roos. (Corpus philosophorum danicorum Medii Aevi IV) Copenhagen: Gad. Boggess, Lois C. 1979. Computational Interpretation of English Spatial Prepositions. Doctoral dissertation, University of Illinois. Boguslawski, Andrzej. 1963. Prefiksacja czasownika we spolczesnym jezyku rosyjskim. Wroclaw: Zaklad Narodowy im. Ossolinskich. Bortolini, U., C.Tagliavini & A. Zampolli. 1972. Lessico difrequenza della lingua italiana con- temporanea. Milan: Garzanti. Breal, Michel. 1897. Essai de semantique. Paris: Hachette. Brewer, Forrest & Jean G. Brewer. 1962. Vocabulario mexicano de Tetelcingo, Morelos (= Serie de vocabularios indigenas Mariano Silva y Aceves 8.) Mexico City: Instituto Linguistico de Verano. Brown, Cecil H. 1982. "Where Do Cardinal Direction Terms Come From"? MS. Brugman, Claudia M. 1981. Story of OVER. Master's thesis, University of California, Berkeley. Trier: LAUT, 1983. 1984. "The Very Idea: A Case Study in Polysemy and Cross-Lexical Generalizations". Papers from the Parasession on Lexical Semantics 21-38. Chicago: CLS. Bursill-Hall, Geoffrey Lesley. 1963. "Mediaeval Grammatical Theories". Canadian Journal of Linguistics 9. 39-54. 1966. "Aspects of Modistic Grammar". Georgetown University Monograph Series on Languages and Linguistics 19. 133-148. 1971. Speculative Grammars of the Middle Ages: The Doctrine of Partes Orationis of the Modistae. The Hague: Mouton. 1912. Thomas of Erfurt: Grammatica speculativa. An edition with translation and commentary by G.L. Bursill-Hall. London: Longman. Bybee, Joan L. 1985. Morphology: A Study of the Relation between Meaning and Form. Amsterdam: Benjamins. Carlson, Lauri. 1981. "Aspect and Quantification". Tedeschi & Zaenen 1981. 31-63. Carnoy, Albert. 1927. La science du mot: traite de semantique. Leuven: Universitas. Casad, Eugene H. 1982a. Cora Locationals and Structured Imagery. Doctoral dissertation, University of California, San Diego. 1982b. "Cora Images". MS. & Ronald W. Langacker. 1985. "'Inside' and 'Outside' in Cora Grammar". International Journal of American Linguistics 51. 247-281.
REFERENCES 681 Catford, J.C. 1959. "The Teaching of English as a Foreign Language". The Teaching of English ed. by Randolph Quirk & A.H. Smith, 137-159. London: Oxford University Press. Chafe, Wallace L. 1970. Meaning and the Structure of Language. Chicago: University of Chicago Press. Chenu, Marie-Dominique. 1936. "Grammaire et theologie aux Xlle et Xllle siecles". Archives d'histoire doctrinale et litter aire du moy en-age 10. 5-28. . 1957. La theologie au Xlle siecle. Paris: Vrin. Chomsky, Noam. 1957. Syntactic Structures. The Hague: Mouton. & Morris Halle. 1968. The Sound Pattern of English. New York: Harper & Row. Clark, Eve V. 1974. "Normal States and Evaluative Viewpoints". Language 50. 316-332. 1978. "Locationals: Existential, Locative, and Possessive Constructions". Greenberg 1978, vol. 4. 85-126. Clark, Herbert H. 1973. "Space, Time, Semantics, and the Child". Cognitive Development and the Acquisition of Language ed. by Timothy E. Moore, 27-63. New York: Academic Press. Coates, Jennifer. 1983. The Semantics of the Modal Auxiliaries. London: Croom Helm. Coleman, Linda & Paul Kay. 1981. "Prototype Semantics: The English Verb Lie". Language 57. 26-44. Collins, Allan M. & Elizabeth F. Loftus. 1975. "A Spreading-Activation Theory of Semantic Processing". Psychology Review 82. 407-428. Comrie, Bernard. 1976. Aspect. Cambridge: Cambridge University Press. 1981a. Language Universals and Linguistic Typology. Oxford: Black well. 1981b. "Aspect and Voice: Some Reflections on Perfect and Passive". Tedeschi & Zaenen 1981. 65-78. Conklin, Harold C. 1973. "Color Categorization". American Anthropologist 75. 931-942. Cooper, Gloria S. 1968. "A Semantic Analysis of English Locative Prepositions". Report No. 1587, Bolt Beranek & Newman Inc. Cooper, Robin. 1986. "Tense and Discourse Location in Situation Semantics". Linguistics and Philosophy 9. 17-36. Coseriu, Eugenic 1964. "Pour une s6mantique diachronique structurale". Travaux de linguis- tique et de literature 2. 139-186. Covington, Michael A. 1984. Syntactic Theory in the High Middle Ages. Modistic Models of Sentence Structure. Cambridge: Cambridge University Press. Cristari, Maurizio. 1971. "Le preposizioni semplici italiane: un approccio semantico". Gram- matica trasformazionale italiana. Pubblicazioni della Societa Linguistica Italiana III, 97-116. Rome: Bulzoni. Cruse, D. A. 1979. "On the Transitivity of the Part-Whole Relation". Journal of Linguistics 15. 29-38. Culler, Jonathan. 1981. "The Problem of Metaphor". Language, Meaning and Style. Essays in Memory of Stephen Vllmann ed. by T.E. Hope et al., 5-20. Leeds: Leeds University Press. Curme, George O. 1922. A Grammar of the German Language. New York: Macmillan. Darmesteter, Arsene. 1887. La vie des mots etudiee dans leurs significations. Paris: Delagrave. De Beer, Taco. 1897. "Tijdschriften". Noord en Zuid 20. 460. De Corte, Marcel. 1934. La doctrine de Vintelligence chez Aristote. Paris: Vrin. De Flines, Quirijn. 1810. Proeve van Hollandsche of Nederduitsche Synonyma. Amsterdam: Bolt. DeLancey, Scott. 1982. "Aspect, Transitivity, and Viewpoint". Hopper 1982. 167-183. 1985. "The Analysis-Synthesis-Lexis Cycle in Tibeto-Burman: A Case Study in Motivated Change". Haiman 1985. 367-389.
682 REFERENCES Denny, J. Peter. 1978. "Locating the Universals in Lexical Systems for Spatial Deixis". Papers from the Parasession on the Lexicon 71-84. Chicago: CLS. De Rijk, Lambertus Maria. 1962-1967. Logica Modernorum. A Contribution to the History of Early Terminist Logic. Assen: Van Gorcum. Devoto, Giacomo & Gian Carlo OH. 1971. Vocabolario illustrato della lingua italiana. Milan: Felice le Monnier. Dimmendaal, Gerrit. 1983. The Turkana Language. Dordrecht: Foris. Dinsmore, John. 1979. Pragmatics, Formal Theory, and the Analysis of Presupposition. Doctoral dissertation, University of California, San Diego. Dirven, Rene. 1976. "A Redefinition of Contrastive Linguistics". IRAL 14. 1-14. 1981. "Spatial Relations in English". Kasusgrammatik und Fremdsprachendidaktik (— Anglistik &. Englischunterricht 14) ed. by Giinter Radden and Rene Dirven, 103-132. Trier: Wissenschaftlicher Verlag. 1985. "Metaphor as a Basic Means for Extending the Lexicon". The Ubiquity of Metaphor ed. by Wolf Paprotte & Rene Dirven, 85-119. Amsterdam: Benjamins. , Louis Goossens, Yvan Putseys & Emma Vorlat. 1982. The Scene of Linguistic Action and Its Perspectivization by SPEAK, TALK, SAY and TELL. Amsterdam / Philadelphia: Benjamins. diSessa, Andrea. 1986. "Knowledge in Pieces". Constructivism in the Computer Age ed. by George Forman & Peter Pufal. Hillsdale, N.J.: Erlbaum. Dixon, Robert M.W. 1972. The Dyirbal Language of North Queensland. London: Cambridge University Press. Dowty, David R. 1972. "On the Syntax and Semantics of the Atomic Predicate CAUSE". Papers from the Eighth Regional Meeting of the Chicago Linguistic Society 62-14. 1977. "Toward a Semantic Analysis of Verb Aspect and the English Tmperfective' Progressive". Linguistics and Philosophy 1. 45-77. . 1979. Word Meaning and Montague Grammar. Dordrecht: Reidel. 1986. "The Effects of Aspectual Class on the Temporal Structure of Discourse: Semantics or Pragmatics"? Linguistics and Philosophy 9. 37-61. Drosdowski, Giinther, Paul Grebe & Wolfgang Miiller. 1965. Hauptschwierigkeiten der deutschen Sprache. Mannheim: Dudenverlag. Dubois, Jean. 1964. "Distribution, ensemble et marque dans le lexique". Cahiers de Lexicologie 4. 5-16. Eco, Umberto. 1984. Semiotics and the Philosophy of Language. London: Macmillan. Ehrenstein, Walter. 1954. Probleme der ganzheitspsychologischen Wahrnehmungslehre. Leipzig: Barth. Eliade, Mircea. 1974. The Myth of the Eternal Return, or Cosmos and History. Princeton, N.J.: Princeton University Press. Elman, Jeffrey L. & James L. McClelland. 1984. "Speech Perception as a Cognitive Process: The Interactive Activation Model". Speech and Language, vol. 10, ed. by Norman Lass, 337-374. New York: Academic Press. Erdmann, Karl Otto. 1910. Die Bedeutung des Wortes. Leipzig: Haessel. Fillmore, Charles J. 1968. "The Case for Case". Universals in Linguistic Theory ed. by Emmon Bach & Robert T. Harms, 1-88. New York: Holt, Rinehart, Winston. 1969. "Types of Lexical Information". Studies in Syntax and Semantics ed. by Ferenc Kiefer, 109-137. Dordrecht: Reidel. 1971a. Santa Cruz Lectures on Deixis. Bloomington: IULC, 1975. 1971b. "Some Problems for Case Grammar". O'Brien 1971. 35-56.
REFERENCES 683 1975. "An Alternative to Checklist Theories of Meaning". Proceedings of the First Annual Meeting of the Berkeley Linguistics Society 123-131. 1977a. "Scenes-and-Frames Semantics". Linguistic Structures Processing ed. by Antonio Zampolli, 55-81. Amsterdam: North Holland. . 1977b. "Topics in Lexical Semantics". Issues in Linguistics ed. by Richard Cole, 76-138. Bloomington: Indiana University Press. . 1978. "The Organization of Semantic Information in the Lexicon". Papers from the Parasession on the Lexicon 148-173. Chicago: CLS. , Paul Kay & Catherine O'Connor. 1983. "Regularity and Idiomaticity in Grammatical Constructions: The Case of 'let alone'". MS. Firth, John R. 1949. "Modes of Meaning". Papers in Linguistics 1934-1951 by J.R. Firth, 190- 215. London: Oxford University Press. Fleischman, Suzanne. 1985. "Discourse Functions of Tense-Aspect Oppositions in Narrative: Toward a Theory of Grounding". Linguistics 23. 851-883. Flier, Michael. 1975. "Remarks on Russian Verbal Prefixation". Slavic and East European Journal 19. 218-229. 1985. "Syntagmatic Constraints on the Russian Prefix pere-\ Issues in Russian Morphosyn- tax (= UCLA Slavic Studies 10) ed. by Michael Flier & Richard Brecht, 138-154. Columbus, Ohio: Slavica. Fodor, Janet, Jerry Fodor & Merrill Garrett. 1975. "The Psychological Unreality of Semantic Representations". Linguistic Inquiry 6. 515-531. Fodor, Jerry A. 1983. The Modularity of Mind. Cambridge, Mass.: MIT Press / Bradford. Friedrich, Paul. 1969a. "On the Meaning of the Tarascan Suffixes of Space". International Journal of American Linguistics, Memoir 23. 1969b. "Metaphor-like Relations between Referential Subsets". LingualA. 1-10. . 1970. "Shape in Grammar". Language 46. 379-407. 1971. The Tarascan Suffixes of Locative Space: Meaning and Morphotactics (= Indiana University Language Science Monographs, No. 9). Bloomington: Indiana University Press. 1974. "On Aspect Theory and Homeric Aspect". International Journal of American Linguistics, Memoir 28. Fuchs, Anna. 1986. "Aspekt". Paper read for the Cologne Universal and Typology Project, 18 April 1986. Gallant, James. 1979. Russian Verbal Prefixation and Semantic Features: An Analysis of the Prefix VZ- (= Slavistische Beitrage 135). Munich: Sagner. Gathercole, Virginia C. 1986. "Evaluating Competing Linguistic Theories with Child Language Data: The Case of the Mass-Count Distinction". Linguistics and Philosophy 8. 151-190. Geeraerts, Dirk. 1983a. "Prototype Theory and Diachronic Semantics: A Case Study". Indoger- manische Forschungen 88. 1-32. 1983b. "Reclassifying Semantic Change". Quaderni di Semantica 4. 217-240. 1983c. "Lcxicografie en Iingui'stiek: Reichling gerehabiliteerd". Tijdschrift voor Neder- landse Taal- en Letterkunde 99. 186-206. 1984. "Type en prototype". Tijdschrift voor Tekst- en Taalweienschap 4. 69-8 1985a. "Preponderantieverschillen bij bijna-synoniemen". De Nieuwe Taalgids 78. 18-27. . 1985b. Paradigm and Paradox. Explorations into a Paradigmatic Theory of Meaning and Its Epistemological Background. Leuven: Leuven University Press. . 1985c. "Cognitive Restrictions on the Structure of Semantic Change". Historical Semantics. Historical Word-Formation ed. by Jacek Fisiak, 127-154. Berlin: Mouton. . 1986. Woordbetekenis. Een overzicht van de lexicale semantiek. Leuven: Acco.
684 REFERENCES Gentner, Dedre. 1982. "Are Scientific Analogies Metaphors"? Metaphor: Problems and Perspectives ed. by David S. Miall, 106-132. Brighton, Sussex: Harvester Press. & Albert L. Stevens, eds. 1982. Mental Models. Hillsdale, N.J.: Erlbaum. Gibson, James J. 1966. The Senses Considered as Perceptual Systems. Boston: Houghton Mifflin. Givon, T. 1973. "The Time-Axis Phenomenon". Language 49. 890-925. . 1979. On Understanding Grammar. New York: Academic Press. . 1984. "Prolegomena to Discourse Pragmatics". Journal of Pragmatics 8. 489-516. Glass, Arnold & Keith Holyoak. 1975. "Alternative Conceptions of Semantic Theory". Cognition 3. 313-339. Godfrey, Robert G. 1960. "The Language Theory of Thomas of Erfurt". Studies in Philology 57. 22-29. 1966. "Late Mediaeval Linguistic Meta-Theory and Chomsky's Syntactic Structures'*. Word 21. 251-266. Goldmeier, Erich. 1972. Similarity in Visually Perceived Forms. New York: International Universities Press. Goodenough, Ward. 1956. "Componential Analysis and the Study of Meaning". Language 32. 195-216. Gordon, W. Terence. 1982. A History of Semantics. Amsterdam: Benjamins. Gove, Philip Babcock, ed. in chief. 1961. Webster's Third New International Dictionary of the English Language. Springfield, MA: Merriam. Greenberg, Joseph H., ed. 1978. Universals of Human Language, vol. 3: Word Structure, vol 4: Syntax. Stanford, Calif.: Stanford University Press. Grice, Paul H. 1974. "Logic and Conversation". Syntax and Semantics, vol. 3, ed. by Peter Cole & Jerry L. Morgan, 41-58. New York: Academic Press. Griesbach, E. & D. Schulz. 1965. Grammatik der deutschen Sprache. Munich: Hueber Verlag. Gross, Maurice. 1979. "On the Failure of Generative Grammar". Language 55. 859-885. Gruber, Jeffrey S. 1965. Studies in Lexical Relations. Doctoral dissertation, MIT. Reprinted as part of Lexical Structures in Syntax and Semantics. Amsterdam: North-Holland, 1976. Haas, William. 1973. "Meanings and Rules". Proceedings of the Aristotelian Society 1973. 135- 155. Haiman, John. 1980. "Dictionaries and Encyclopedias". Lingua 50. 329-357. , ed. 1985. Iconicity in Syntax. Amsterdam /Philadelphia: Benjamins. Hamelin, Octave. 1931. Le systeme d'Aristote. Paris: Alcan. Hawkins, Bruce W. 1984. The Semantics of English Spatial Prepositions. Doctoral dissertation, University of California, San Diego. Trier: LAUT, 1985. . 1985. "An Introduction to Prepositional Semantics: Perspectives on the Data". MS. Hecht, Max. 1888. Die griechische Bedeutungslehre, eine Aufgabe der klassischen Philologie. Leipzig: Teubner. Heine, Bernd & Mechthild Reh. 1984. Grammaticalization and Reanalysis in African Languages. Hamburg: Buske. Helbig, A. & J. Buscha. 1974. Deutsche Grammatik. Leipzig: VEB Verlag Enzyklopadie. Henle, Michael. 1979. A Combinatorial Introduction to Topology. San Francisco: Freeman & Co. Herskovits, Annette. 1986. Language and Spatial Cognition: An Interdisciplinary Study of the Prepositions in English. Cambridge: Cambridge University Press. Hinrichs, Erhard. 1981. Temporale Anaphora im Englischen. Thesis, Universitat Tubingen. 1986. "Temporal Anaphora in Discourse of English". Linguistics and Philosophy 9. 63-82.
REFERENCES 685 Hopper, Paul J. 1979a. "Aspect and Foregrounding in Discourse". Discourse and Syntax. Syntax and Semantics, vol. 12, ed. by T. Givon, 213-241. New York: Academic Press. 1979b. "Some Observations on the Typology of Focus and Aspect in Narrative Language". Studies in Language 3. 37-64. , ed. 1982. Tense - Aspect: Between Semantics and Pragmatics. Amsterdam / Philadelphia: Benjamins. & Sandra A. Thompson. 1980. "Transitivity in Grammar and Discourse". Language 56. 251-299. Hudson, Richard A. 1976. Arguments for a hi on-Transformational Grammar. Chicago: University of Chicago Press. 1984. Word Grammar. Oxford: Blackwell. Hume, David. 1748. "An Enquiry Concerning Human Understanding". The English Philosophers from Bacon to Mill ed. by Edwin A. Burtt. 1939, 585-689. New York: Modern Library. Hunt, Richard W. 1941-1943. "Studies on Priscian in the Eleventh and Twelfth Centuries (I)". Mediaeval and Renaissance Studies 1-2. 194-231. (Repr. Hunt 1980: 1-38.) 1950. "Studies on Priscian in the Eleventh and Twelfth Centuries (II)". Mediaeval and Renaissance Studies 2. 1-56. (Repr. Hunt 1980: 39-94.) . 1980. The History of Grammar in the Middle Ages: Collected Papers ed. by G.L. Bursill- Hall. Amsterdam: Benjamins. Isacenko, Alexander. 1960. Grammatiteski] stroj russkogo jazyka v sopostavlenii s slovackim, pt. 2: Morfologija. Bratislava: Slovenska Academia Vied. Jackendoff, Ray. 1977. "Toward a Cognitively Viable Semantics". Georgetown University Round Table on Languages and Linguistics ed. by Clea Rameh, 59-80. Washington, D.C.: Georgetown Unversity Press. . 1983. Semantics and Cognition. Cambridge, Mass.: MIT Press. In press. Consciousness and the Computational Mind. Cambridge, Mass.: Bradford / MIT Press. Jakobson, Roman. 1971. "Boas' View of Grammatical Meaning". Selected Writings of Roman Jakobson, vol. 2, 489-496. The Hague: Mouton. James, Carl. 1980. Contrastive Analysis. London: Longman. Janda, Laura A. 1985. "The Meaning of Russian Verbal Prefixes: Semantics and Grammar". The Scope of Slavic Aspect (= UCLA Slavic Studies 12) 26-40. Columbus, Ohio: Slavica. -—. 1986. A Semantic Analysis of the Russian Verbal Prefixes ZA-, PERE-, DO-, and OT- (= Slavistische Beitrdge, Band 192). Munich: Sagner. Jespersen, Otto. 1961. A Modern English Grammar on Historical Principles. London / Copenhagen: Allen & Unwin / Munksgaard. Jessen, Marilyn E. 1974. A Semantic Study of Spatial and Temporal Expressions in English. Doctoral dissertation, University of Edinburgh. John of Denmark (ed. 1955). Johannis Daci Opera ed. by Alfred Otto. (Corpus Philosophorum Danicorum Medii Aevi I /1-21) Copenhagen: Gad. Katz, Jerrold. 1972. Semantic Theory. New York: Harper & Row. & Jerry A. Fodor. 1963. "The Structure of a Semantic Theory". Language 39. 170-210. & Richard Nagel. 1974. "Meaning Postulates and Semantic Theory". Foundations of Language 11. 311-340. & Paul M. Postal. 1964. An Integrated Theory of Linguistic Descriptions. Cambridge: MIT Press. Kay, Paul & Chad K. McDaniel. 1978. "The Linguistic Significance of the Meanings of Basic Color Terms". Language 54. 610-646.
686 REFERENCES Kellogg, Rhoda. 1970. Analyzing Children's Art. Palo Alto: Mayfield. Kempson, Ruth M. 1977. Semantic Theory. Cambridge: Cambridge University Press. King, Robert Thomas. 1974a. Modern Semantic Theory: A Critical Analysis and an Application to the Study of German Prepositions. Doctoral dissertation, Rice University, Houston, Texas. . 1974b. "The Hypothesis of Event Causation". Unpublished working paper, University of Frankfurt. Kittay , Eva & Adrienne Lehrer. 1981. "Semantic Fields and the Structure of Metaphor". Studies in Language 5. 31-63. Kosslyn, Stephen Michael. 1980. Image and Mind. Cambridge, Mass: Harvard University Press. Kronasser, Heinz. 1952. Handbuch der Semasiologie. Heidelberg: Winter. Labov, William. 1973. "The Boundaries of Words and Their Meanings". New Ways of Analyzing Variation in English ed. by Charles-James N. Bailey & Roger W. Shuy, 340-373. Washington, D.C.: Georgetown University Press. Lakoff, George. 1970. "Global Rules". Language 46. 627-639. 1972. "Linguistics and Natural Logic". Semantics of Natural Language ed. by Donald Davidson & Gilbert Harman, 545-665. Dordrecht: Reidel. 1977. "Linguistic Gestalts". Papers from (he Thirteenth Regional Meeting of the Chicago Linguistic Society 236-287. . 1982a. "Categories and Cognitive Models". Trier: LAUT. 1982b. "Categories: An Essay in Cognitive Linguistics". Linguistics in the Morning Calm ed. by the Linguistic Society of Korea, 139-193. Seoul: Hanshin. 1982c. "Experimental Factors in Linguistics". Language, Mind, and Brain ed. by Thomas W. Simon & Robert J. Scholes, 145-156. Hillsdale, N.J.: Erlbaum. 1983. "Elements of the Sound Symbolism System". Lecture at the CODOC symposium on Cognitive Linguistics. Brussels, 26 March 1983. 1986. "Cognitive Semantics". Berkeley Cognitive Science Report No.36, University of California, Berkeley. . In press. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago/London: University of Chicago Press. & Mark Johnson. 1980. Metaphors We Live By. Chicago: University of Chicago Press. Langacker, Ronald W. 1973. "Predicate Raising: Some Uto Aztecan Evidence". Issues in Linguistics: Papers in Honor of Henry and Renee Kahane ed. by Braj B. Kachru et al., 468-491. Urbana: University of Illinois Press. . 1977. An Overview of Uto-Aztecan Grammar (= Studies in Uto-Aztecan Grammar, vol. 1). Dallas: Summer Institute of Linguistics and University of Texas at Arlington. 1978. "The Form and Meaning of the English Auxiliary". Language 54. 853-882. . 1979. "Grammar as Image". Linguistic Notes from La Jolla 6. 87-126. 1981. "The Integration of Grammar and Grammatical Change". Indian Linguistics 42. 82- 135. 1982a. "Space Grammar, Analysability, and the English Passive". Language 58. 22-80. -----. 1982b. "Remarks on English Aspect". Hopper 1982. 265-304. 1983. Foundations of Cognitive Grammar. I: Orientation, II: Semantic Structure. Trier: LAUT. 1984. "Active Zones". Proceedings of the Tenth Annual Meeting of the Berkeley Linguistics Society 172-188. 1985. "Observations and Speculations on Subjectivity". Haiman 1985. 109-150. . 1986a. "An Introduction to Cognitive Grammar". Cognitive Science 10. 1. 1-40.
REFERENCES 687 1986b. "Abstract Motion". To appear in Proceedings of the Twelfth Annual Meeting of the Berkeley Linguistics Society. In press. Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequisites. Stanford, Calif.: Stanford University Press. Larkin, Miriam Thercse. 1971. Language in the Philosophy of Aristotle. The Hague / Paris: Mouton. Leech, Geoffrey N. 1969. Towards a Semantic Description of English. Bloomington: Indiana University Press. 1980. Explorations in Semantics and Pragmatics. Amsterdam: Benjamins. Lewis, David. 1972. "General Semantics". Semantics of Natural Language ed. by Donald Davidson & Gilbert Harman, 169-218. Dordrecht: Reidel. Lindner, Susan J. 1981. A Lexico-Semantic Analysis of English Verb-Particle Constructions with UP and OUT. Doctoral dissertation, University of California, San Diego. Trier: LAUT, 1983. 1982. "What Goes Up Doesn't Necessarily Come Down: The Ins and Outs of Opposites". Papers from the Eighteenth Regional Meeting of the Chicago Linguistic Society 305-323. Lipka, Leonhard. 1985. "Inferential Features in Historical Semantics". Historical Semantics. Historical Word-Formation ed. by Jacek Fisiak, 339-354. Berlin: Mouton. Lohnes, Walter F. W. & Friedrich W. Strothmann. 1968. German: A Structural Approach. New York: Norton & Co. Lounsbury, Floyd. 1956. "A Semantic Analysis of Pawnee Kinship Usage". Language 32. 158- 194. Lyons, John. 1963. Structural Semantics. Oxford: Blackwcll. 1968. Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press. 1977. Semantics, vol. 1 & 2. Cambridge: Cambridge University Press. Mackenzie, J. Lachlan. 1977. "The Marking of Agents and Instruments in Contemporary Standard German". Work in Progress No. 10. Edinburgh: Department of Linguistics, University of Edinburgh. Mackridge, Peter. 1985. The Modern Greek Language. A Descriptive Analysis of Standard Modern Greek. Oxford: Oxford University Press. Malotki, Ekkehart. 1983. Hopi Time: A Linguistic Analysis of the Temporal Concepts in the Hopi Language. Berlin: Mouton. Mantovani, Fr. Mario V.F. 1985. A New Karamojong Grammar. Moroto: Nadiket Seminary. Marr, David. 1982. Vision. San Francisco: Freeman & Co. Martin of Denmark (ed. 1961). Martini Daci Opera ed. by Harry Roos. (Corpus philosophorum danicorum Medii Aevi II) Copenhagen: Gad. Marty, Anton. 1908. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie. Halle: Niemeyer. McCawley, James D. 1971a. "Prelexical Syntax". O'Brien 1971. 19-33. 1971b. "Where Do NP's Come From"? Semantics ed. by Danny D. Steinberg & Leon A. Jakobovits, 217-231. Cambridge: Cambridge University Press. 1972. "Syntactic and Logical Arguments for Semantic Structures". Bloomington, Indiana: IULC. 1976. "Remarks on What Can Cause What". Shibatani 1976b. 117-129. McNeil, N.B. 1972. "Colour and Colour Terminology". Journal of Linguistics 8. 21-33. Meillet, Antoine. 1906. "Comment les mots changent de sens". Annee sociologique 9. 1-38. Mervis, Carolyn B. & Eleanor Rosch. 1981. "Categorization of Natural Objects". Annual Review of Psychology 32. 89-115.
688 REFERENCES Miller, George A. & Philip N. Johnson-Laird. 1976. Language and Perception. Cambridge, Mass.: Harvard University Press. Minsky, Marvin. 1980. "K-Lines: A Theory of Memory". Cognitive Science 4. 117-133. Mirambel, Andre. 1942. "De l'emploi de l'aoriste en grec moderne". Transactions of the Philological Society 43. 15-39. . 1956. "Subordination et temps en grec moderne". Bulletin de la Societe de Linguistique de Paris 52. 219-253. Morgan, Jerry L. 1972. "Verb Agreement as a Rule of English". Papers from the Eighth Regional Meeting of the Chicago Linguistic Society 278-286. Morrow, Daniel. 1986. "Grammatical Morphemes and Conceptual Structure in Discourse Processing". Cognitive Science 10.4. 423-455. Mourelatos, Alexander P.D. 1981. "Events, Processes, and States". Tedeschi & Zaenen 1981. 191-212. Neisser, Ulric. 1967. Cognitive Psychology. New York: Meredith. 1976. Cognition and Reality. Principles and Implications of Cognitive Psychology. San Francisco: Freeman. Nerbonne, John. 1986. "Reference Time and Time in Narration". Linguistics and Philosophy 9. 83-95. Newmeyer, Frederick J. 1975. English Aspectual Verbs. The Hague/Paris: Mouton. 1980. Linguistic Theory in America. New York: Academic Press. 1983. Grammatical Theory : Its Limits and Its Possibilities. Chicago: University of Chicago Press. Newton, Brian. 1979. "Scenarios, Modality, and Verbal Aspect in Modern Greek". Language 55. 139-167. & Ioannis Veloudis. 1980a. "Necessity, Obligation and Modern Greek Verbal Aspect". Lingua 50. 25-43. & Ioannis Veloudis. 1980b. "Intention, Destination and Greek Verbal Aspect". Lingua 52. 269-284. Nilsen, Don L.F. 1972. Toward a Semantic Specification of Deep Case. The Hague: Mouton. 1973. The Instrumental Case in English. Syntactic and Semantic Considerations. The Hague: Mouton. Nuchelmans, Gabriel. 1972. "Het mentaals. De opvattingen van het denken als vorm van spre- ken in de antieke en middeleeuwse taaltheorie". Leuvense Bijdragen 61. 295-309. Nyrop, Kristoffer. 1913. Grammaire historique de la langue francaise IV. Semantique. Kopenha- gen: Gyldendalske Boghandel. O'Brien, Richard J., ed. 1971. Georgetown University Series on Languages and Linguistics, Monograph 24. Washington, D.C.: Georgetown University Press. Onions, C.T., et al. 1966. The Oxford Dictionary of English Etymology. Oxford: Clarendon Press. Ortony, Andrew. 1979a. "The Role of Similarity in Similes and Metaphors". Ortony 1979b. 187- 201. , ed. 1979b. Metaphor and Thought. Cambridge: Cambridge University Press. Otaala, Laura A. 1981. Phonological and Semantic Aspects of Ateso Derivational Verbal Morphology. Doctoral dissertation, University of Nairobi. Owens, Joseph. 1960. "Aristotle on Categories". Review of Metaphysics 14. 73-90. Paivio, Allan. 1979. "Psychological Processes in the Comprehension of Metaphor". Ortony 1979b. 151-171. Palmer, F.R. 1981. Semantics. 2nd. ed. Cambridge: Cambridge University Press.
REFERENCES 689 Palmer, Stephen. 1983. "The Psychology of Perceptual Organization: A Transformational Approach". Human and Machine Vision ed. by Jacob Beck, Barbara Hope & Azriel Rosen- feld, 269-339. New York: Academic Press. Paprotte, Wolf. 1985. "Linguistische Aspekte der Begriffsentwicklung". Begriffs- und Wortbedeutungsentwicklung ed. by Thomas Bernhard Seiler & Wolfgang Wannenmacher, 75-101. Berlin: Springer. & Chris Sinha. 1985. "A Functional Perspective on Language Development". Social and Functional Approaches to Language and Thought ed. by Maya Hickmann. New York / London: Academic Press. (In press). Pare, Gerard, Adrien Brunet & A. Tremblay. 1933. La renaissance du Xlle siecle, les ecoles et I'enseignement. Paris: Vrin; Ottawa: Institut d'Etudes medievales. Parisi, Domenico & Christiano Castelfranchi. 1970. "Analisi semantica dei locativi spaziali". La sintassi. Pubblicazioni della Societa Linguistica Italiana II, 327-366. Rome: Bulzoni. Partee, Barbara H. 1984. "Nominal and Temporal Anaphora". Linguistics and Philosophy 7. 243-286. Paul, Hermann. 1880. Prinzipien der Sprachgeschichte. Halle: Niemeyer. Pinborg, Jan. 1967. Die Entwicklung der Sprachtheorie im Mittelalter. Munster: Aschendorff; Copenhagen: Frost/Hansen. . 1972. Logik und Semantik im Mittelalter. Stuttgart-Bad Cannstatt: Frommann - Holzboog. 1975. "Die Logik der Modistae". Studia mediewistyczne 16. 39-97. 1982. "Speculative Grammar". The Cambridge History of Later Medieval Philosophy, 254- 269. Cambridge: Cambridge University Press. Pittman, Richard S. 1948. "NahuatI Honorifics". International Journal of American Linguistics. 14. 236-239. 1954. A Grammar of Tetelcingo (Morelos) NahuatI. Language 30. vol. I Supplement (= Language Dissertation No. 50). Pluim, Teunis. 1894. Keur van Nederlandsche synoniemen. Purmerend: Muusses. Porzig, Walter. 1934. "Wesenhafte Bedeutungsbeziehungen". Beitrdge zur Geschichte der deutschen Sprache und Literatur 58. 70-97. Pottier, Bernard. 1963. Recherches sur Vanalyse semantique en linguistique et en traduction mecanique. Nancy: Faculte de Lettres et Sciences Humaines. Priebsch, R. & W.E. Collinson. 1968. The German Language. London: Faber & Faber. Pulman, Steve. 1983. Word Meaning and Belief London: Croom Helm. Putnam, Hilary. 1981. Reason, Truth and History. Cambridge: Cambridge University Press. Quillian, M.Ross. 1968. "Semantic Memory". Semantic Information Processing ed. by Marvin Minsky, 227-270. Cambridge, Mass.: MIT Press. Quine, Willard von Orman. 1960. Word and Object. Cambridge. Mass.: MIT Press. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik. 1972. A Grammar of Contemporary English. London: Longman. Reddy, Michael J. 1979. "The Conduit Metaphor - A Case of Frame Conflict in Our Language about Language". Ortony 1979b. 284-324. Reichling, Anton. 1935. Het woord. Zwolle: Tjeenk Willink. Reiter, Raymond A. 1980. "Logic for Default Reasoning". Artificial Intelligence 13. 81-132. Rensky, Miroslav. 1966. "English Verbo-Nominal Phrases". Travaux Linguistiques de Prague 1. 289-299. Rosch, Eleanor. 1973a. "Natural Categories". Cognitive Psychology 4. 328-350. 1973b. "On the Internal Structure of Perceptual and Semantic Categories". Cognitive Development and the Acquisition of Language ed. by Timothy E. Moore, 111-144. New York: Academic Press.
690 REFERENCES . 1975. "Cognitive Representations of Semantic Categories". Journal of Experimental Psychology: General 104. 192-233. . 1977. "Human Categorization". Studies in Cross-Cultural Psychology, vol. 1, ed. by Neil Warren, 1-49. New York: Academic Press. . 1978. "Principles of Categorization". Cognition and Categorization ed. by Eleanor Rosch & Barbara B. Lloyd, 27-48. Hillsdale, N.J.: Erlbaum. & Carolyn B. Mervis. 1975. "Family Resemblances". Cognitive Psychology 7. 573-605. Rosier, Irene. 1983. La grammaire speculative des modistes. Lille: Presses universitaires de Lille. Ross, William David, ed. 1910-1928. The Works of Aristotle. Oxford: Clarendon Press. . 1937. Aristotle. 3rd ed. London: Methuen & Co. Rudzka-Ostyn, Brygida. 1983. "Cognitive Grammar and the Structure of Dutch UIT and Polish WY". Trier: LAUT. . 1985. "Metaphoric Processes in Word Formation. The Case of Prefixed Verbs". The Ubiquity of Metaphor ed. by Wolf Paprotte & Rene Dirven, 209-241. Amsterdam: Benjamins. . 1986. "The Semantics of Perfective and Imperfective Expressions". Proceedings of the Third April Conference of University Teachers of English. Cracow 1984 ed. by Teresa Baluk- Ulewiczowa & Maria Korosadowicz, 31-50. Cracow: Jagiellonian University Press. Rumelhart, David E. & David Zipser. 1985. "Feature Discovery by Competitive Learning". Cognitive Science 9. 75-112. Sanders, Gerald A. 1974. "Precedence Relations in Language". Foundations of Language 11. 361-400. Sapir, Edward. 1921. Language. New York: Harcourt, Brace & Company. Schank, Roger C. 1973. "Identification of Conceptualizations Underlying Natural Language". Computer Models of Thought and Language ed. by Roger C. Schank & Kenneth Mark Colby, 187-247. San Francisco: Freeman & Co. . 1980. "Language and Memory". Cognitive Science 4. 243-284. Schmitz, Werner. 1962. Der Gebrauch der deutschen Prapositionen. Munich: Hueber Verlag. Schnelle, Helmut. 1980. "Pre-Tense". Time, Tense and Quantifiers ed. by Christian Rohrer, 329-354. Tubingen: Niemeyer. . 1981a. "Phenomenological Analysis of Language and Its Application to Time and Tense". Possibilities and Limitations of Pragmatics ed. by Herman Parret, Marina Sbisa & Jef Ver- schueren, 631-655. Amsterdam: Benjamins. . 1981b. "Introspection and the Description of Language Use". A Festschrift for the Native Speaker ed. by Florian Coulmas, 105-126. The Hague: Mouton. Schopf, Alfred, ed. 1974. Der englische Aspekt. Darmstadt: Wiss. Buchgesellschaft. . 1984. Das Verzeitungssystem des Englischen und seine Textfunktion. Tubingen: Niemeyer. Searle, John R. 1979. "Literal Meaning". Expression and Meaning. Cambridge: Cambridge University Press. Seiler, Hansjakob. 1952. L'aspect et le temps dans le verbe neo-grec. Paris: Societe d'Editions 'les Belles Lettres'. . 1977. Cahuilla Grammar. Banning, Calif.: Malki Museum Press. Shepard, Roger N. 1978. "The Mental Image". American Psychologist 33. 125-137. Shibatani, Masayoshi. 1976a. "The Grammar of Causative Constructions: A Conspectus". Shibatani 1976b. 1-40. , ed. 1976b. The Grammar of Causative Constructions ( = Syntax and Semantics, 6). New York: Academic Press. Siger of Courtrai (ed. 1977). Sigerus de Cortraco: Summa modorum significandi. Sophismata. New edition based on G. Wallerand's editio prima, with critical notes, an index of terms and
REFERENCES 691 an introduction by J. Pinborg. Amsterdam: Benjamins. Simon of Denmark (ed. 1963). Simonis Daci Opera ed. by Alfred Otto. (Corpus philosophorum danicorum Medii Aevi III) Copenhagen: Gad. Slobin, Dan I. 1985. "Crosslinguistic Evidence for the Language-Making Capacity". The Crosslinguistic Study of Language Acquisition, vol. 2, ed. by D.I. Slobin. Hillsdale, N.J.: Erlbaum. & Ayhan A. Aksu. 1982. "Tense, Aspect and Modality in the Use of the Turkish Evidential". Hopper 1982. 185-200. Slovar' sovremennogo russkogo literaturnogo jazyka, 17 vols. 1950-1965. Moscow / Leningrad: Adakemija Nauk. Smith, Carlota S. 1980. "Temporal Structures in Discourse". Time, Tense, and Quantifiers ed. by Christian Rohrer, 355-374. Tubingen: Niemeyer. . 1986. "A Speaker-Based Approach to Aspect". Linguistics and Philosophy 9. 97-115. Smith, Edward E. & Douglas L. Medin. 1981. Categories and Concepts. Cambridge, Mass.: Harvard University Press. , Edward J. Shoben & Lance J. Rips. 1974. "Structure and Process in Semantic Memory: a Featural Model for Semantic Decisions". Psychological Review 81. 214-241. Smith, Michael B. 1985. "Event Chains, Grammatical Relations, and the Semantics of Case in German". Papers from the General Session at the Twenty-First Regional Meeting of the Chicago Linguistic Society 388-407. Sperber, Hans. 1923. Einfuhrung in die Bedeutungslehre. Bonn: Schroeder. Stern, Gustav. 1931. Meaning and Change of Meaning. Goteborg: Elanders. Stiker, H.-J. 1972. "Une theorie linguistique au moyen age: 1'ecole modiste". Revue des sciences philosophiques et theologiques 56. 585-616. Sullivan, Thelma D. 1976. Compendio de la gramdtica ndhuatl (= Serie de cultura ndhuatl, Monografias 18, del Instituto de Investigaciones Historicas). Mexico City: UNAM. Swan, Michael. 1980. Practical English Usage. Oxford: Oxford University Press. Sweetser, Eve. 1984. "Semantic Structure and Semantic Change: English Perception Verbs in an Indo-European Context". Trier: LAUT. Swiggers, Pierre. 1981. "La theorie du signe a Port-Royal". Semiotica 35. 267-285. . 1982. "Theorie de la grammaire et theorie des signes chez les encyclopedistes". Semiotica 40. 89-105. . 1984a. Les conceptions linguistiques des encyclopedistes. Etude sur la constitution d'une theorie de la grammaire au siecle des Lumieres. Heidelberg: Groos; Leuven: University Press. 1984b. "Cognitive Aspects of Aristotle's Theory of Metaphor". Glotta 62. 40-45. , ed. 1984c. Grammaire et methode au XVlle siecle. Leuven: Peeters. . 1986. Grammaire et theorie du langage au XVIIIe siecle. Lille: Presses universitaires de Lille. Talmy, Leonard. 1972. Semantic Structures in English and Atsugewi. Doctoral dissertation, University of California, Berkeley. 1975. "Semantics and Syntax of Motion". Syntax and Semantics, vol. 4, ed. by John P. Kimball, 181-238. New York: Academic Press. 1976. "Semantic Causative Types". Shibatani 1976b. 43-116. 1977. "Rubber-Sheet Cognition in Language". Papers from the Thirteenth Regional Meeting of the Chicago Linguistic Society 612-628. 1978a. "The Relation of Grammar to Cognition - a Synopsis". Proceedings of TIN LAP - 2 (Theoretical Issues in Natural Language Processing) ed. by David Waltz, 14-24. New York: Association for Computing Machinery.
692 REFERENCES 1978b. "Figure and Ground in Complex Sentences". Greenberg 1978, vol. 4. 625-649. 1978c. "Relations between Subordination and Coordination". Greenberg 1978, vol. 4. 487- 513. 1982. "Borrowing Semantic Space: Yiddish Verb Prefixes between Germanic and Slavic". Proceedings of the Eighth Annual Meeting of the Berkeley Linguistics Society 231-250. 1983. "How Language Structures Space". Spatial Orientation: Theory, Research, and Application ed. by Herbert Pick & Linda Acredolo, 225-282. New York: Plenum Press. . 1985a. "Lexicalization Patterns: Semantic Structure in Lexical Forms". Language Typology and Syntactic Description, vol. 3: Grammatical Categories and the Lexicon ed. by Timothy Shopen, 57-149. Cambridge: Cambridge University Press. 1985b. "Force Dynamics in Language and Thought". Papers from the Parasession on Causatives and Agentivity at the Twenty-First Regional Meeting of the Chicago Linguistic Society 293-337. Tedeschi, Philip J. & Annie Zaenen, eds. 1981. Tense and Aspect (= Syntax and Semantics 14). New York: Academic Press. Teller, Paul. 1969. "Some Discussion and Extension of Manfred Bierwisch's Work on German Adjectivals". Foundations of Language 5. 185-217. Thomas of Erfurt (ed. 1972). Grammatica speculativa of Thomas of Erfurt. An edition with translation and commentary by G.L. Bursill-Hall. London: Longman. Thomason, Richmond. 1974. Formal Philosophy. Selected Papers of Richard Montague. New Haven: Yale University Press. Traugott, Elizabeth Closs. 1978. "On the Expression of Spatio-Temporal Relations in Language". Greenberg 1978, vol. 3. 369-400. Trier, Jost. 1931. Der deutsche Wortschatz im Sinnbezirk des Verstandes. Heidelberg: Winter. Tucker, Archibald N. & J. Tompo Ole Mpaayei. 1955. A Maasai Grammar. London: Longmans, Green & Co. Tuggy, David. 1979. "Tetelcingo Nahuatl". Studies in Uto-Aztecan Grammar, vol. 2: Modern Aztec Grammatical Sketches, ed. by Ronald W. Langacker, 1-140. Arlington, TX: University of Texas at Arlington and Summer Institute of Linguistics. . 1980. "jEthical Dative and Possessor Omission Si, Possessor Ascension No!" Work Papers of the Summer Institute of Linguistics, University of North Dakota 24. 97-141. . 1981. The Transitivity-Related Morphology of Tetelcingo Nahuatl: An Exploration in Space Grammar. Doctoral dissertation, University of California, San Diego. Tzermias, Pavlos. 1969. Neugriechische Grammatik. Bern/Miinchen: Francke. Ullmann, Stephen. 1962. Semantics. An Introduction to the Science of Meaning. Oxford: Blackwell. Ultan, Russell. 1978. "Some General Characteristics of Interrogative Systems". Greenberg 1978, vol. 4. 211-248. Vandeloise, Claude. 1979. Les termes de dimension enfrancais. Memoire de Maitrise, Univer- site de Paris VIII. . 1984. Description of Space in French. Doctoral dissertation, University of California, San Diego. Duisburg: LAUDT, 1985. 1985a. "La preposition dans et la relation contenant / contenu". Leuvense Bijdragen 74. 193-208. 1985b. La representation linguistique de Vespace et du mouvement. These de doctorat, Ecole des Hautes Etudes en Sciences Sociales de Paris. . 1985c. "Les prepositions avant I apres et la rencontre potentielle". Recherches Linguis- tiques\3. 110-129.
REFERENCES 693 1985d. "Au-dela des descriptions geom6triques et Iogique de I'espace: une description fonctionnelle". Linguisticae Investigations XI. 1. 109-129. 1985e. "La preposition suret la relation porteur/porte". Leuvense Bijdragen 74. 451-484. 1985f. "The Preposition in and the Relationship Container / Contained". MS. . 1986. Vespace en francais. Paris: Seuil. Van Ginneken, Jac. 1910-1912. "Het gevoel in taal- en woordkunst". Leuvense Bijdragen 9. 265- 356; 10. 1-156, 173-273. van Schooneveld, Cornelis. 1978. The Cardinal Semantic Structure of Prepositions, Cases, and Paratactic Conjunctions in Contemporary Standard Russian (= Semantic Transmutations, 3). Bloomington: Physsardt. Vendler, Zeno. 1968. "Each and Every, Any and All". Linguistics and Philosophy. Ithaca: Cornell University Press. Verschueren, Jef. 1981. "Basic Linguistic Action Verbs". Cahiers de linguistique francaise (Geneve) 2. 71-88. 1984. "Basic Linguistic Action Verbs: A Questionnaire". Antwerp Papers in Linguistics No. 37. . 1985. What People Say They Do with Words : Prolegomena to an Empirical-Conceptual Approach to Linguistic Action. Norwood, N.J.: Ablex. In press. "Metapragmatics and Universals of Linguistic Action". Linguistic Action: Some Empirical-Conceptual Studies ed. by Jef Verschueren. Norwood, N.J.: Ablex. Vlach, Frank. 1981. "The Semantics of the Progressive". Tedeschi & Zaenen 1981. 271-292. Wallace, Stephen. 1982. "Figure and Ground: The Interrelationships of Linguistic Categories". Hopper 1982. 201-223. Wallerand, Gaston. 1913. Les oeuvres de Siger de Courtrai. Etude critique et textes inedits. Lou- vain: Institut superieurde philosophic Waltz, David L. 1981. "Generating and Understanding Scene Descriptions". Elements of Discourse Understanding ed. by Aravind K. Joshi , Bonnie L. Webber & Ivan A. Sag, 266-282. Cambridge: Cambridge University Press. & Jordan B. Pollack. 1985. "Massively Parallel Parsing: A Strongly Interactive Model of Natural Language Interpretation". Cognitive Science 9. 51-74. Wasow, Thomas. 1977. "Transformations and the Lexicon". Formal Syntax ed. by Peter W. Culicover, Thomas Wasow & Adrian Akmajian, 327-360. New York: Academic Press. Weiland, Petrus & George Landre. 1825. Woordenboek der Nederduitsche synoniemen. 's-Gravenhage: Allart. Weinreich, Uriel. 1966. Explorations in Semantic Theory. The Hague: Mouton. Reprinted in Uriel Weinreich on Semantics ed. by William Labov and Beatrice Weinreich. 1980, 99-201. Philadelphia: University of Pennsylvania Press. Weisgerber, Leo. 1927. "Die Bedeutungslehre — ein Irrweg der Sprachwissenschaft"? Ger- manisch-Romanische Monatschrift 15. 161-183. Wellander, Erich. 1917-1928. Studien zum Bedeutungswandel im Deutschen. Uppsala: Berlings Boktryckeri. Whorf, Benjamin Lee. 1956. Language, Thought, and Reality ed. by John B. Carroll. Cambridge, Mass.: Technology Press of MIT. Wierzbicka, Anna. 1980. The Case for Surface Case. Ann Arbor: Karoma. Winograd, Terry. 1972. Understanding Natural Language. Edinburgh: Edinburgh University Press. & Fernando Flores. 1985. Understanding Computers and Cognition. Norwood, N.J.: Ablex.
694 REFERENCES Wittgenstein, Ludwig. 1953. Philosophical Investigations, translated by G.E.M. Anscombe. Oxford: Black well. Wundt, Wilhelm. 1900. Volkerpsychologie I: Die Sprache. Leipzig: Engelmann. Zadeh, Lofti A. 1965. "Fuzzy Sets". Information and Control 8. 338-353. 1974. "A Fuzzy-Algorithmic Approach to the Definition of Complex or Imprecise Concepts". Memo No. ERL-M474, Electronic Research Laboratory, University of California, Berkeley. Zingarelli, Nicola. 1970. Vocabolario della lingua italiana. Bologna: Zanichelli. Zubin, A. David & Soonja Choi. 1984. "Orientation and Gestalt: Conceptual Organizing Principles in the Lexicalization of Space". Papers from the Parasession on Lexical Semantics 333- 345. Chicago: CLS.
Subject Index Ablaut, 28-30, 149-51, 153-56, 158 Abstraction, 51 Accessibility, 152f Accommodation, 142-47 Activation, 68f Active node, 69, 153-58 passim selection of, 153 Active zone, 250-59, 261, 263-65 and metonymy, 270 Adjective, 32f, 44-46, 65-67, 76, HOf, 424f axial properties of, 186-88 for vertical space, 381-401, 403 Adverb, 44f, 111, 201, 634-36 of degree, 186-88 Affix, 120 derivational, 97f diminutive, 197 directional, 429f, 433-44 Agent, 558, 575, 581-84 Ambiguity, 136-39 Analyzabihty, 79-84 Anaphora, 137-39 Anomaly, semantic, 43, 46, 81f Applicative construction, 587, 598-607, 612- 18 Artificial intelligence, 293-95 passim, 375 and lexical meaning, 271f, 294 and prototype phenomena, 224 Aspect, 184-86, 372, 443f, 447-51 perfective /imperfective, 451-57 psychosemantic basis, 457-59 and simple/progressive form, 476-81 Autonomous vs. dependent structure, 103- 6, 108-13, 115-24 Autonomy thesis, 5, 18, 28f, 47f, 147, 151, 379-82 arbitrary/empty grammatical markings, 29, 31-43, 48 grammaticality judgments, 43-48 semantic structure, 56f, 380-82, 650, 656-58, 668-73 Base, 50, 58-63, 66f, 70, 75, 82f, 94f, 250f, 508 Bipolarity, llf, 50, 91-93, 148. See also Phonological pole; Semantic pole Bounded region, 95, 509, 551 Boundedness, 178-82, 203, 252-59, 261, 263-65 and count/mass distinction, 178-80 and perfective/imperfective distinction, 178-80, 432, 436 Bounding, cognitive operation of, 179, 189- 91,438, 512f, 528f Cahuilla, 114-18 Case marking, 36, 40f Categorial association, principles of, 242-46 Categorization, 379-82 cognitive ideal in, 243f classical model, 133, 236f, 267, 623, 653-58 passim natural-category model, 242-46, 267 network model, 133-40, 152f, 160 prototype model, 133, 207f, 220-28, 242, 652-60 passim, 662f Categorizing relation, 12, 19-28, 46, 68f, 130ff types of, 51, 134, 139f Category: classical, 236-46, 653-55 passim complex, 51-53, 134-61, 328-39, 404, 588,596f, 603,605f, 614 development/extension of, 139f, 142ff, 323f grammatical, 5, 18-28, 32-46, 75, 148- 61,591-616,623-46
696 SUBJECT INDEX of grammatically specified notions, 165-67, 173-93 lexical, 50-53, 68f, 134-40, 142f, 234- 41,246-68,588 natural, 242-46, 302f prototype-based, 133f, 223f, 322f, 327f, 654f, 658, 660, 662f Category, complex, 134-61, 596f, 603, 605f, 614 lexeme, 51-53, 134f, 302, 328-39, 588 morpheme, 143-45 phoneme, 141f Causation, 555-61 prototypical, 557, 589f. See also Force dynamics Causative construction, 119-25, 555-59, 566-85,587,589-97,602-18 syntatic marking, 559-63 Class, 22, 28, 46 closed/open, 166-98, 201-3 distributional, 28-31 Classical Nahuatl, 158-60, 617 Classifier, 114-18 Clause structure, 98-108, 119-25, 518-21, 637-41 patterns, 339-43, 639f prototypical, 325. See also Causative construction Coding, 14f Cognitive distance, 134, 153, 156-58 and computation of novel expressions, 157f Cognitive domain, see Domain Cognitive event, 6 Cognitive grammar: and Aristotle's views, 621-24, 643, 653f basic claims and concepts, 3-6, 11-15, 49f, 91-97, 130-35, 165f, 194f and generative grammar, 3-6, 12f, 17- 19, 48, 98f, 127-33 (see also Autonomy thesis; Lexical semantics) and modistic grammar, 621f, 641-43, 646 and other linguistic frameworks, 653- 74 and other disciplines, 195-201 (see also Artificial intelligence; Cognitive psychology) and Whorf's views, 11 Cognitive model, 327-38 predictive power, 339. See also Idealized cognitive model Cognitive operation, 53f, 174 actionalizing, 175f bounding, 179, 189-91, 438, 512f, 528f componentialization, 190f debounding, 179 discretizing, 180f Gestalt formation, 190f magnification, 183f melding, 180f multiplexing, 176-78 profiling, 59f reduction, 183 reification, 175 unit-excerpting, 176-78 Cognitive processing, 5f, 8, 152-58 connectionist model, 153 Cognitive psychology, 132f, 198, 207f, 223f, 654f, 658, 660, 662f Cognitive representation, 165, 188f content specification, 165-73, 195f structure specification, 165-73, 195-201 Cognitive routine, 6, 11, 130 Cognitive semantics, 49-90, 652-57, 665 onomasiological and semasiological aspects, 224-28 Cognitive space, 328, 330f Collocation, 237f, 381 selectional restrictions, 393-95 status of, 393-96. See also Co-occurrence Color terms, 383 Compacted representation, 20, 24f, 27, 30, 45 Comparison, 78 standard/target asymmetry, 145, 516f Competence/performance, 129 Component structure, 20-28, 30f, 148f autonomous/dependent, 103-25 Componential analysis, 373f, 393f, 666f, 671 Composite structure, 20, 65-67, 72-75 analyzability, 79-84 formation, 20-34, 44-46, 92f, 102-25, 148 profile of, 60-63, 66f, 82f, 106ff, 590ff Composition, 60, 629f
SUBJECT INDEX 697 Compositional path, 27, 35f, 46, 80-84, 104f, 125 Compositionality, 13-18, 295f and analyzability, 79-84 full, 13, 20 partial, 15-18 Compound, 20, 25f, 33f, 74f, 80 Concept, 6 lexical, 654-56 Conceptual hierarchy, 53f, 73f, 91, 94 Conceptual organization: in language, 166, 194-201 in other domains, 195-201 Conceptualization, 5f, 14, 50, 54-56, 88f, 94, 624-28 (see also Cognitive operation; Cognitive processing) geometric, 274f, 282, 293 of length/width, 403-18, 423, 426 of vertical space, 381-401 Conduit metaphor, 58, 507 Configuration, 327-39 mapped onto verbal argument, 339-43. See also Profile Conflation, conceptual, 204 Congruity, 640f Conjunction, 634-37 Constituency, 18-28, 44-46, 104ff, 115ff variability, 105, 125 Constituency tree, 105f, 119, 124f Construal: alternate, 38, 63, 90, 304f, 322f, 358f, 373 objective, 88f subjective, 88-90 Container metaphor, 58 Content requirement, 12f, 23, 30, 47, 147, 151 Conventional linguistic unit, 11, 19-21, 29f, 91f, 130, 156 Conventionality, 14-16, 35, 38-40, 44-48 passim, 152, 345, 348-78 passim, 609, 611 Co-occurrence, 234f, 267, 269 selection restrictions, 236-41, 393-95 ascription features, 236, 238f, 241, 248 transfer features, 236, 239-41. See also Collocation Cora, 85-87, 111-13,345-78 Correspondence, 49, 6lf, 72f, 77, 102ff cross-domain, 544-49 multiple, 118 symbolic, 49, 623 Count/mass distinction, 178f, 453-57 Dative Shift, 28-31 Decoding, 14 Deictic center, 433, 441 vs. referential center, 441-43 vs. source/goal, 435f vs. speaker, 434f, 443 Deixis, 433f, 441 Demonstrative, 168f Dependent structure, 103-6, 108-13, 115-24 Derivation, 82-84 Designation, 59f, 94f, 508. See also Profile Diachrony: in lexical semantics, 647-51, 660-63 logical, 404, 418-26 Direct object, 40, 76f, 105, 108, 122, 340- 43, 519f, 590 Directionality, 84 conception of, 87-90 Distance, see Cognitive distance Distribution: and predictability, 28-31, 147- 58 passim and productivity, 147-61 Domain, 53-63, 94-97, 346f, 508-16 passim, 544-50 passim abstract, 55, 95 basic, 54, 94 configurational, 94f defined, 53f, 94, 508 donor, 514 intersections, 509-11, 549f and meaning, 50, 53-63, 94, 508 non-basic, 55, 94 physical motion, 516-18 primary, 509 recipient, 514 sensory, 54f, 94 space, 54-56 passim, 84, 94f time, 54-56 passim, 61f verbal communication, 509-11 visual perception, 195, 198-200. See also Matrix; Search domain Dummy, 12 Dutch, 209-28, 309, 384, 419 Dyirbal, 166 Eastern Nilotic, 430
698 SUBJECT INDEX Economy, 4, 127-31, 135, 150 Elaboration, 12, 15, 21, 30, 51, 91, 120, 122, 124, 134 Elaboration site, 104, 106-9, 111, 113, 115, 117f, 120 Emphasis, 34 Encoding, 14 Encyclopedic semantics, 57f, 222, 250, 347, 375, 656, 658, 669 Entity, 61, 114, 174, 197 bounded, 178f, 189-91 Entrenchment, 6, 12, 134, 140, 157 Ergative/absolutive distinction, 342f Equi-NP Deletion, 98, 124f Exploded representation, 20, 27, 45f Extension, 15f, 51, 134-36, 139f, 142-47 metaphorical, 15, 69, 316f, 328, 331-38 phonological, 142, 144-46 semantic, 5f, 15, 51, 69, 135f, 143, 146f, 215-17, 225-28, 314-24, 356f, 360, 363, 369-73, 661-63 and similarity perception, 134, 136, 139, 516f. See also Metaphor; Metonymy; Synecdoche Family-resemblance model, 208, 220, 222 Feature, 148, 281, 297, 373f ascription, 236, 238f, 241, 248 grammatical, 29, 43, 147, 150 phonological, 29 semantic, 29, 54, 300f, 393-95 syntactic/morpho-syntactic, 22, 36f transfer, 236, 239-41 Feature copying, 36f Figure, 96,191, 195 Figure/ground, 76, 78f, 195, 200, 226, 273f, 383f, 508, 513, 518, 674 and aspect, 458f, 487 Force dynamics, 195, 200 Form, 31 grammatical, 46 French, 41, 384, 418, 424f Fuzzy logic, 295 Gender, 36, 40-42 Generality, 127f, 131, 147, 160 Generalization, 128-33, 160f Generative grammar, 127-29 and cognitive grammar, 3-6, 12-19 passim, 98f, 129-33. See also Autonomy thesis Geometric description, 273-76, 280, 282-84, 292f, 295 German, 380f, 384, 403, 561-85 Gestalt, 196f, 459 formation, 190f Goal, 432f, 575 Gradience, 256f, 260, 266f Grammar: as determinant of conceptual organization within language, 165-67, 194-200 as discrete module, 132 as generative device, 4, 13, 127, 622 as modes of signifying, 624-45 as symbolic phenomenon, 5, 11-29, 46- 48, 91f, 99 Grammatical class, 22, 29, 46 basic,5,22f, 44. See also Parts of speech Grammatical complex, 167, 177f Grammatical construction, 23, 26-28, 65, 71-73, 77 as complex category, 148-61 Grammatical structure, 5, 11-29, 46-48, 91f, 99, 165-204 function of, 165-67, 195-97 and lexicon, 5, 12, 18f, 32, 92, 165-73, 174-96 passim and semantic structure, 5, 18, 49f Grammatical system, see Grammatical structure Grammatically judgment, 43-47 and congruity, 640f and contextualization, 43, 563f Grammatically specified notions, see Notions, grammatically specified Greek, Standard Modern, 448-57, 465-72, 474-78, 482-95, 502-5 Ground, 76, 78f, 96f. See also Figure/ground Head, 60, 75 clausal, 76f Head-complement construction, 109f, 118f, 122, 124 Head-modifier construction, 65f, 75, 111 vs. head-complement construction, 109,116
SUBJECT INDEX 699 Hebrew, 384 Hierarchy: conceptual, 53f, 73f, 91, 94 of linguistic units, 25f, 64, 73, 134-60 passim, 588, 597, 603, 605f, 614 Homology, space/time, 174f Homonymy vs. polysemy, 136f Hopi, 100-11, 124 Hyponymy, 64, 666 ICM, see Idealized cognitive model Iconicity and tense/aspect, 460-67, 471f, 477f, 486-88 Idealized cognitive model, 201, 327f, 456f, 487 Ik, 429-44 Image, 7-11 alternate, 38f schematic, 273 Imagery, 7, 32-36 passim, 322f, 345f, 622 conventional, 6-11, 22, 38-40, 48, 63- 90, 270 dimensions of, 64-90 sensory, 7, 63 Imaging system, 167, 194f Impulsion, 404, 418f, 426 Innateness hypothesis, 394, 403 Instantiation, 12, 24, 51, 66, 91, 130-35, 138f Instrument(al), 558f, 575, 577-79, 581-84 Interactive-activation model, 153 Interconnection, 61, 66f, 76, 79, 96 Interjection, 634-36 Introspection, reliability of, 217-20 Irregularity, 153-57 and entrenchment and salience, 157 Italian, 204, 299,306-25 Landmark, 61f, 66f, 72, 110-25 passim, 303- 5, 328, 508, 518-20 primary, 76, 101, 105, 108 and verbal arguments, 339-41. See also Trajector/landmark Language and other cognitive domains, 166, 195-201 Latin, 627-42, 645 Lexical semantics, 207-28 and Aristotle's views, 621-23 passim, 653-55 passim cognitive, 652-57, 665 cognitive vs. historical-philological, 657-63, 672-74 historical-philological, 647-52, 670, 672-74 logical, 664f, 670-73 and prototype theory, 207f, 217-28, 301,322-24 structuralist, 657, 660, 664-74, 676 transformationalist, 666, 668-74 Lexical system, see Lexicon Lexical unit, 18-26 as complex category, 51-53, 68f, 134f, 302, 328-39, 588 Lexicon, 5, 128, 165f function of, 165-67, 201-4 passim and grammar, 5, 12, 18f, 32, 92, 165- 73, 174-96 passim Linear ordering, 26, 28, 31, 46, 120 Locative construction, 78f, 85-90, 271-96, 345-78 nested, 71-73 simple-relations approach, 272-80, 292f, 296f truth-conditional approach, 295f Luiseno, 97-100, 119-25 Mass/count distinction, 178f, 453-57 Matrix, 56, 73, 94, 97 complex, 56f, 101 Meaning, 5-11, 32f,379f as conceptualization, 5f, 32-35, 49-90, 345-47,508, 621, 650f,657f core, 300-2, 305, 321f ideal, 271, 273, 280-97 passim lexical {see Lexical semantics) linguistic vs. extralinguistic, 57f (non)overlap, 33-37, 40f primary vs. grammatical, 626-28, 644 prototypical, 214-19, 271, 280f, 301f, 305f, 322, 328-30, 659f Medium: as co-occurrence conditions, 231- 35,241,247f, 267 as lexical category, 234-41, 246-68 vs. surface/node/channel, 247-51, 268 Mental transfer, 85 Metaphor, 5f, 69, 266 defined, 521
700 SUBJECT INDEX novel, 15 spatial, 517f, 523-51, 555, 575, 580-85 Metonymy, 260-63, 270, 276, 285f, 521, 527,533,538,663 and typicality, 290, 293f Mimesis, 663 Modifier, 65f, 75, 109-11, 115, 118f adjectival vs. adverbial, 111 Modistic grammar, 621f, 624-46 language and cognition, 624-28, 641-44 passim parts of speech, 627-37, 644f syntax, 637-42, 645f Morpheme, 19-27, 39, 41-43, 92f, 111-13 as complex category, 143-46 grammatical vs. lexical, 41f, 166f nominalizing, 23f past-tense, 28-30, 149-51, 153-58 root/stem vs. affix, 120-22. See also Affix; Prefix; Stem; Suffix Morphological layering, 97-99, 119-25 Morphology: and lexicon/syntax, 5, 12 productivity of, 347-72 Morphophonemic rule, 145f Motion: abstract, 89f, 517 physical, 88f, 516-18 Nahuatl, see Classical Nahuatl; Tetelcingo Nahuatl Nesting, 181f, 193, 199 in perception, 199 Network model, 68f, 133-61 and lexemes, 50-53, 134-40, 143, 328- 30, 588 and other complex categories, 141-46, 149-59, 596f, 602f, 605f, 612-15 and rules, 145-60, 596f Nominal expression, 44-46, 60f, 75-77 Nominal predication, 95 vs. relational predication, 60f, 75-77, 110f Nominalization, 23-27 Notions, grammatically specified, 167-74, 192-95 axiality, 186-88, 203f boundedness, 178-80, 203 degree of extension, 183f, 203 dimension, 174-76 dividedness, 180-82, 203 level of exemplarity, 191f level of synthesis, 189-91 pattern of distribution, 184-86 perspectival mode, 188f, 194 plexity, 176-78, 181f, 188-93 passim quantity's disposition, 181f, 203 Noun, 5, 22f, 44f, 65f, 73-76, 92f, 114-18, 159, 628-34 collective, 37f count, 95, 129-31 count/mass, 178-82, 203 derived, 23f, 79f, 82-84 designation of, 22f, 508f pluralia tantum, 38f for spatial dimensions, 403-26 Noun-suffix construction: applicative, 612- 15 causative, 610-12, 614f Novel expression, 26, 79, 139f computation of, 26, 130-33, 146f, 157- 61 passim Number, 36-38, 197 grammatical vs. semantic, 38f Participle, 629-34 past, 62f, 157f Particle, 111-13, 347-52, 519f. See also Phrasal verb Part-whole relationship, 73f Parts of speech, 627-37, 644f Past participle, 62f, 157f Past-tense formation, 28-30, 149-51, 153-58 Patient, 77, 558, 575, 581-84 foregrounding of, 44 Perspective, 84-90, 188f directionality, 87-90 long-range/close-up, 183f, 194 orientation, 84 steady-state/moving, 188f, 194 subjective/objective construal, 88-90 vantage point, 84-87, 347, 368, 373 Phoneme, 141f, 144-46 as complex category, 141f Phonological pole, 12, 14, 19-28, 30f, 143- 46, 153-55 integration at, 20, 92f, 119f valence relations, 92f, 120f
SUBJECT INDEX 701 Phonological rule, 146 Phonological schema, 12, 21f, 141f Phonological space, 46 Phonological structure, 26 temporal ordering, 26, 46 Phrasal verb, 522-41 Phrase tree, 18f and constructional schema, 26-28 Plexity, 176-78, 181f, 188-93 passim and singular/plural distinction, 179 Plural, 21f, 36-39, 129-31, 168, 176-78 phonological pole, 92f, 131, 144f semantic pole, 92f, 131 Pole: phonological, see Phonological pole; semantic, see Semantic pole Polish, 327, 384 Polysemy, 659 core-sense approach, 300-2, 32If vs. homonymy, 136f network model, 50-52, 68f, 134-36 prototype approach, 209-19, 301-3, 305, 322f, 588 Possessive, 114-18 Postposition, 158-60 Pragmatic near-principles, 280, 284 relevance, 280, 287f salience, 280, 285f, 290, 293 tolerance, 280, 288f typicality, 280, 289f, 293 Pragmatics vs. semantics, 5f, 17, 57f, 128 Predicate, 50,94 types of, 95f Predicate Raising, 98, 125 Predication, 50, 53-56, 94 adjectival, 66f, 75-77 nominal, 59f, 65f, 75-77, 95, 115-18 relational, 61f, 66f, 75-80, 96ff scale of, 70, 73f scope of, 70-75 verbal, 61, lOlf Predictability, 4, 28-32, 35 absolute, 4, 48, 100, 133f Prefix, 327-43 sequences, 85-87, 353-74, 377f Preposition, 76, 78f, 169f, 196, 200, 271-97, 299-325, 555, 634f, 637 in causative constructions, 561-85 ideal meaning, 271, 273, 280-82, 283-97 passim medium-oriented, 227-32, 242-63, 265 use types, 273, 281, 290-92, 294-96. See also Phrasal verb Presupposition, 67 Primary domain, see Domain, primary Primitive, semantic, 54, JuOf, 303, 383, 394, 671 Process, 22f, 30, 114,431, 590 conceptualization of, 97 perfective/imperfective, 432, 434, 436 trajectorflandmark, 61f, 96f, lOlff, 121-23, 590 Profile, 58-63, 75-79, 82f, 94-97, 327f, 508 and active zone, 250f nominal, 60f, 95, HOf, 115-18 nominal vs. adjectival, 66f, 75-77 relational, 61-63, 66, 75-79, 96f, 101 24 passim, 303f temporal, 97, 121, 372 Profile determinant, 106, 109-13, 124, 592 Progressive, 476-88 Pronoun, 629-34 demonstrative, 37, 168f indefinite, 202 Property: accidental/essential, 625f, 653 criterial, 133 Prototype effects: in applicatives, 600 causatives, 557, 564-66, 582, 593f, 596f clause structure, 325 lexical/lexemic structure, 5If, 134-36, 214-28, 251-66, 280-82, 301f, 322f, 328-30, 388-93, 396-400, 652-54, 659f morphemic structure, 143f past-tense formation, 154f phonemic structure, 14lf valence relations, 99-109 Prototype theory, 133f, 207f, 217-28, 301, 322-24, 654f, 658, 660, 662f and family-resemblance model, 218 and natural-category model, 242-46, 267 Prototypicality, 133f, 207-28, 652-55, 659f, 662f and category growth, 139f, 143, 225- 28, 323f global/local prototype, 134f, 139 introspective evidence, 217-19
702 SUBJECT INDEX physiological hypothesis, 207, 218-22, 224 psychological (= functional) hypothesis, 207f, 220-25, 228 referential hypothesis, 207, 220-22, 224 statistical hypothesis, 207, 220-22, 224 usage as evidence, 215-17 Psychological reality, 128f Quantifier, 7-10, 67f Reductionism, 127-29, 132, 135, 141-43, 150f, 160 Redundancy, 4, 33-40, 128-30 Redundancy rule, 374 Referential center, 441-43 Regularity, 153-57 Relation: categorizing, 12, 22-28, 30f, 51, 65, 68f, 134ff composition, 20-27, 92f integration, 20, 92f schematic, 12, 21-26, 51, 64-66, 91f, 108f symbolic, 19f, 92f, 151 Relational predication, 61 f, 66f, 75-79 internal structure, 76f, 96f vs. nominal predication, 75-77 processual, 96f, lOlf, 114, 431 f, 590 stative, 96f, 114,431,436-39 Relevance, 273, 280, 287f, 294 Root, 120f Rule, 23f, 145-60 Rule/list fallacy, 131-33 Russian, 327-43 Salience, 152-54, 207, 280, 285f, 290, 293 types of, 75-84 Scale of predication, 70, 73f Scanning, 87f sequential/summary, 97, 431 Schema, 12, 21-28, 46, 51f, 64, 91-93, 130f, 140 constructional, 22-26, 30f, 65f, 130f elaboration/instantiation of, 12, 21-26, 51f, 66,91, 130f noun/verb, 22-25, 30f sanctioning function of, 92 Schematic hierarchy, 51f, 130f, 134f, 139ff, 588f, 597f, 603, 606, 614f Schematic transparency principle, 33-35, 42 Schematicity, degree of, 19-23, 64-67, 91f and hyponymy, 21, 64f and polysemy, 51f, 134-36, 588 Scope of predication, 70-75 immediate, 73f Scope, semantic, see Semantic scope and morphological layering Search domain, 71-73 Selection(al) restriction, 236-41, 393-95, 667 Semantic change, 224-28, 650f, 659, 662f Semantic extension, see Extension, semantic Semantic marker, 54, 380-82, 403, 425 Semantic overlap, 33-35, 37, 40 Semantic pole, 11, 14, 19-28, 30f, 44-46, 50, 65f, 91-93 integration at, 20, 92f of noun, 20-27, 92f of verb, 22-27. See also Predicate Semantic primitive, 54, 300, 303, 383, 394, 671 Semantic scope and morphological layering, 98f, 119-25 Semantic structure, 5f, 49-63, 94, 249-51, 379f, 508 and grammatical structure, 4f, 49f, 63 and imagery, 6, 32-35, 50, 63-90, 345f Semantic universal, 657 Semantics: autonomous, 56f, 380-82, 650, 656-58, 668-73 cognitive, 49-90, 224-28, 652-57, 665 encyclopedic, 57f, 250, 347, 375, 656, 658, 669 generative/interpretive, 668-71 logical, 5f, 81, 271, 295f, 664f, 670-73 objectivist, 6, 90 vs. pragmatics, 5f, 17f, 57f, 128 subjectivist, 6f, 22, 48, 50-90 Sentence, 637-41 Shift, 174, 183-88 passim, 190, 192, 202f Sound symbolism, 659 Source, 432f, 575 Space, 54f, 94f, 328 oriented, 79, 84, 384-87, 396-400 partitioning of, 168, 194 Spanish, 42f, 325 Spatial dimensions, 403-10, 412-26
SUBJECT INDEX 703 Spatial metaphor: in case grammar, 575 in causative constructions, 580-85 Spatial motion, 516-18 metaphoric extensions, 517, 521-51 Specialization, 15f, 51, 134f Speech act/event, 511 vs. motion event, 523-51 primary/secondary, 511-13 Standard/target asymmetry, 145, 516f State, 114, 431f,556f, 581-84 Stative relation, 96f, 114, 431 Stem, 120 combined with affix, 20, 24, 92f, 120, 130f, 590-618 Structural description, 26, 28, 65f, 130 Subject, 40, 44f, 105, 340-43 vs. object, 76f Subjective/objective construal, 88-90 Suffix, 92f, 154-60, 587, 590-618 topographic, 348-52 Superschema, 52, 135 Symbolic complexity, 19f Symbolic unit, 5, 14f, 19-32, 49-51, 68 phonological/semantic pole, llf, 19f, 92f, 148f Synecdoche, 275f, 285f, 521. See also Metonymy Synonymy, 209-28, 309, 322f Syntax, 18f, 637-41, 645f and lexicon, 5, 12, 18f and morphology, 5, 12, 19, 26 and semantics, 5f, 12, 18 Tense, 447-51 and discourse representation, 461f, 472-76, 488 psychosemantic basis, 459-63 Tetelcingo Nahuatl, 616 causatives/ applicatives, 587, 589-618 Thing, 22, 30f, 114 defined, 75, 95 Time: conceived/processing, 517 partitioning of, 168, 194 Tolerance, 273, 280, 288f, 294 Topology: grammatical specification of, 167, 169-73, 199, 202 and causatives, 575-85 Trajector, 61f, 66f, 75-78, 303-5, 328, 508 agentive, 518-20 instrumental, 518-20 and verbal arguments, 339-41 Trajector/landmark, 61f, 96f 303-5, 508, 518, 577f, 589f and figure/ground, 76, 78f and place/goal/source/path, 304 and subject/object, 76-78, 519f Trajectory, 328 Transitive/intransitive construction, 638-40 Transitivity, 639f, 645 Truth-conditional semantics, 5f, 81, 90, 295f, 672 Turkish, 21f, 166 Type/predictability fallacy, 29, 48, 151 Typicality, 273, 280, 289f, 293 Unit, 11, 130 phonological/semantic/symbolic, 5, 11- 15, 19-32,49-51,91-93, 147f Unit status, 11, 130 Universalism, 382f, 394 Usage-based model, 29f, 48, 131-61 Usage event, 14-16 Vagueness, 293-95 vs. ambiguity, 136-39 Valence relation, 92-125 autonomous/dependent structure, 103- 24 constituency, 104ff elaboration site, 104, 106-11, 120 profile determinant, 106, 109-13, 124 Vantage point, 84-87, 347, 368, 373 Variant: phonological, 142 semantic, 53, 68f, 143 Verb, 5, 22f, 28-31, 61f, 96f, 149-51, 169- 203 passim, 629-34 auxiliary, 32, 34f causative, 556-60, 572 inchoative, 438f of motion, 429f, 432-36, 517, 522-49 para-absolutive, 342 perfective/imperfective, 178f, 432, 437f transitive/intransitive, 340-42 Verb-particle/-preposition combination, see Phrasal verb Verb-suffix construction: applicative, 598- 602, 614f causative, 590-97, 614f
704 SUBJECT INDEX causative/applicative, 602f, 614f Verbalization, 607-14 Viewpoint, 292. See also Vantage point Visual perception, 195 and language structure, 198-200 Voice, 444, 638f Well-formedness, 13, 35 judgment of, 5, 43, 139, 152-56, ■ 96, 399f of natural category, 245-47, 251 Word class, 627-37, 644f content of, 622 Yiddish, 174 Zulu, 384
CURRENT ISSUES IN LINGUISTIC THEORY E. F. K. Koerner, Editor Zentrum fur Allgemeine Sprachwissenschaft, Typologie und Universalienforschung, Berlin efk.koerner@rz.hu-berlin.de Current Issues in Linguistic Theory (CILT) is a theory-oriented series which welcomes contributions from scholars who have significant proposals to make towards the advancement of our understanding of language, its structure, functioning and development. CILT has been established in order to provide a forum for the presentation and discussion of linguistic opinions of scholars who do not necessarily accept the prevailing mode of thought in linguistic science. It offers an outlet for meaningful contributions to the current linguistic debate, and furnishes the diversity of opinion which a healthy discipline must have. A complete list of titles in this series can be found on the publishers' website, www.benjamins.com 281 VERMEERBERGEN, Myriam, Lorraine LEESON and Onno CRASBORN (eds.): Simultaneity in Signed Languages. Form and function, viii, 344 pp. + index (incl. CD-Rom). Expected February 2007 280 HEWSON, John and Vit BUBENIK: From Case to Adposition. The development of configurational syntax in Indo-European languages. 2006. xxx, 420 pp. 279 NEDERGAARD THOMSEN, Ole (ed.): Competing Models of Linguistic Change. Evolution and beyond. 2006. vi, 344 pp. 278 DOETJES, Jenny and Paz GONZALEZ (eds.): Romance Languages and Linguistic Theory 2004. Selected papers from 'Going Romance', Leiden, 9-11 December 2004. 2006. viii, 320 pp. 277 HELASVUO,Marja-Liisa and Lyle CAMPBELL (eds.): Grammar from the Human Perspective. Case, space and person in Finnish. 2006. x, 280 pp. 276 MONTREUIL, Jean-Pierre Y. (ed.): New Perspectives on Romance Linguistics. Vol. II: Phonetics, Phonology and Dialectology Selected papers from the 35th Linguistic Symposium on Romance Languages (LSRL), Austin, Texas, February 2005. 2006. x, 213 pp. 275 NISHIDA, Chiyo and Jean-Pierre Y. MONTREUIL (eds.): New Perspectives on Romance Linguistics. Vol. I: Morphology, Syntax, Semantics, and Pragmatics. Selected papers from the 35th Linguistic Symposium on Romance Languages (LSRL), Austin, Texas, February 2005. 2006. xiv, 288 pp. 274 GESS, Randall S. and Deborah ARTEAGA (eds.): Historical Romance Linguistics. Retrospective and perspectives. 2006. viii, 393 pp. 273 FILPPULA, Markku, Juhani KLEMOLA, Marjatta PALANDER and Esa PENTTILA (eds.): Dialects Across Borders. Selected papers from the 11th International Conference on Methods in Dialectology (Methods XI), Joensuu, August 2002. 2005. xii, 291 pp. 272 GESS, Randall S. and Edward J. RUBIN (eds.): Theoretical and Experimental Approaches to Romance Linguistics. Selected papers from the 34th Linguistic Symposium on Romance Languages (LSRL), Salt Lake City, March 2004. 2005. viii, 367 pp. 271 BRANNER, David Prager (ed.): The Chinese Rime Tables. Linguistic philosophy and historical- comparative phonology. 2006. viii, 358 pp. 270 GEERTS,Twan,Ivo van GINNEKEN and Haike JACOBS (eds.): Romance Languages and Linguistic Theory 2003. Selected papers from'Going Romance' 2003, Nijmegen, 20-22 November. 2005. viii, 369 pp. 269 HARGUS, Sharon and Keren RICE (eds.): Athabaskan Prosody. 2005. xii, 432 pp. 268 CRAVENS, Thomas D. (ed.): Variation and Reconstruction. 2006. viii, 223 pp. 267 ALHAWARY, Mohammad T. and Elabbas BENMAMOUN (eds.): Perspectives on Arabic Linguistics XVII-XVIII. Papers from the seventeenth and eighteenth annual symposia on Arabic linguistics. Volume XVII-XVIII: Alexandria, 2003 and Norman, Oklahoma 2004. 2005. xvi, 315 pp. 266 BOUDELAA, Sami (ed.): Perspectives on Arabic Linguistics XVI. Papers from the sixteenth annual symposium on Arabic linguistics, Cambridge, March 2002. 2006. xii, 181 pp. 265 CORNIPS, Leonie and Karen P. CORRIGAN (eds.): Syntax and Variation. Reconciling the Biological and the Social. 2005. vi, 312 pp. 264 DRESSLER, Wolfgang U, Dieter KASTOVSKY, Oskar E. PFEIFFER and Franz RAINER (eds.): Morphology and its demarcations. Selected papers from the 11th Morphology meeting, Vienna, February 2004. With the assistance of Francesco Gardani and Markus A. Pochtrager. 2005. xiv, 320 pp. 263 BRANCO, Ant6nio, Tony McENERY and Ruslan MITKOV (eds.): Anaphora Processing. Linguistic, cognitive and computational modelling. 2005. x, 449 pp. 262 VAJDA, Edward J. (ed.): Languages and Prehistory of Central Siberia. 2004. x, 275 pp. 261 KAY, Christian J. and Jeremy J. SMITH (eds.): Categorization in the History of English. 2004. viii, 268 pp.
260 NICOLOV, Nicolas, Kalina BONTCHEVA, Galia ANGELOVA and Ruslan MITKOV (eds.): Recent Advances in Natural Language Processing III. Selected papers from RANLP 2003. 2004. xii, 402 pp. 259 CARR, Philip, Jacques DURAND and Colin J. EWEN (eds.): Headhood, Elements, Specification and Contrastivity. Phonological papers in honour of John Anderson. 2005. xxviii, 405 pp. 258 AUGER, Julie, J. Clancy CLEMENTS and Barbara VANCE (eds.): Contemporary Approaches to Romance Linguistics. Selected Papers from the 33rd Linguistic Symposium on Romance Languages (LSRL), Bloomington, Indiana, April 2003. With the assistance of Rachel T. Anderson. 2004. viii, 404 pp. 257 FORTESCUE,Michael, Eva Skafte JENSEN, Jens Erik MOGENSEN and Lene SCH0SLER (eds.): Historical Linguistics 2003. Selected papers from the 16th International Conference on Historical Linguistics, Copenhagen, 11-15 August 2003. 2005. x, 312 pp. 256 BOK-BENNEMA, Reineke, Bart HOLLEBRANDSE, Brigitte KAMPERS-MANHE and Petra SLEEMAN (eds.): Romance Languages and Linguistic Theory 2002. Selected papers from'Going Romance', Groningen, 28-30 November 2002. 2004. viii, 273 pp. 255 MEULEN, Alice ter and Werner ABRAHAM (eds.): The Composition of Meaning. From lexeme to discourse. 2004. vi, 232 pp. 254 BALDI, Philip and Pietro U. DINI (eds.): Studies in Baltic and Indo-European Linguistics. In honor of William R. Schmalstieg. 2004. xlvi, 302 pp. 253 CAFFAREL, Alice, J.R. MARTIN and Christian M.I.M. MATTHIESSEN (eds.): Language Typology. A functional perspective. 2004. xiv, 702 pp. 252 KAY, Christian J., Carole HOUGH and Irene WOTHERSPOON (eds.): New Perspectives on English Historical Linguistics. Selected papers from 12 ICEHL, Glasgow, 21-26 August 2002. Volume II: Lexis and Transmission. 2004. xii, 273 pp. 251 KAY, Christian J., Simon HOROBIN and Jeremy J. SMITH (eds.): New Perspectives on English Historical Linguistics. Selected papers from 12 ICEHL, Glasgow, 21-26 August 2002. Volume I: Syntax and Morphology. 2004. x, 264 pp. 250 JENSEN, John T: Principles of Generative Phonology. An introduction. 2004. xii, 324 pp. 249 BOWERN, Claire and Harold KOCH (eds.): Australian Languages. Classification and the comparative method. 2004. xii, 377 pp. (incl. CD-Rom). 248 WEIGAND, Edda (ed.): Emotion in Dialogic Interaction. Advances in the complex. 2004. xii, 284 pp. 247 PARKINSON, Dilworth B. and Samira FARWANEH (eds.): Perspectives on Arabic Linguistics XV Papers from the Fifteenth Annual Symposium on Arabic Linguistics, Salt Lake City 2001. 2003. x, 214 pp. 246 HOLISKY, Dee Ann and Kevin TUITE (eds.): Current Trends in Caucasian, East European and Inner Asian Linguistics. Papers in honor of Howard I. Aronson. 2003. xxviii, 426 pp. 245 QUER, Josep, Jan SCHROTEN, Mauro SCORRETTI, Petra SLEEMAN and Els VERHEUGD (eds.): Romance Languages and Linguistic Theory 2001. Selected papers from 'Going Romance', Amsterdam, 6-8 December 2001. 2003. viii, 355 pp. 244 PfeREZ-LEROUX, Ana Teresa and Yves ROBERGE (eds.): Romance Linguistics. Theory and Acquisition. Selected papers from the 32nd Linguistic Symposium on Romance Languages (LSRL), Toronto, April 2002. 2003. viii, 388 pp. 243 CUYCKENS, Hubert, Thomas BERG, Rene DIRVEN and Klaus-Uwe PANTHER (eds.): Motivation in Language. Studies in honor of Giinter Radden. 2003. xxvi, 403 pp. 242 SEUREN, Pieter A.M. and Gerard KEMPEN (eds.): Verb Constructions in German and Dutch. 2003. vi, 316 pp. 241 LECARME, Jacqueline (ed.): Research in Afroasiatic Grammar II. Selected papers from the Fifth Conference on Afroasiatic Languages, Paris, 2000. 2003. viii, 5 50 pp. 240 JANSE, Mark and Sijmen TOL (eds.): Language Death and Language Maintenance. Theoretical, practical and descriptive approaches. With the assistance of Vincent Hendriks. 2003. xviii, 244 pp. 239 ANDERSEN, Henning (ed.): Language Contacts in Prehistory. Studies in Stratigraphy. Papers from the Workshop on Linguistic Stratigraphy and Prehistory at the Fifteenth International Conference on Historical Linguistics, Melbourne, 17 August 2001. 2003. viii, 292 pp. 238 NUNEZ-CEDENO, Rafael, Luis LOPEZ and Richard CAMERON (eds.): A Romance Perspective on Language Knowledge and Use. Selected papers from the 31st Linguistic Symposium on Romance Languages (LSRL), Chicago, 19-22 April 2001. 2003. xvi, 386 pp. 237 BLAKE, Barry J. and Kate BURRIDGE (eds.): Historical Linguistics 2001. Selected papers from the 15th International Conference on Historical Linguistics, Melbourne, 13-17 August 2001. Editorial Assistant: Jo Taylor. 2003. x, 444 pp. 236 SIMON-VANDENBERGEN, Anne-Marie, Miriam TAVERNIERS and Louise J. RAVELLI (eds.): Grammatical Metaphor. Views from systemic functional linguistics. 2003. vi, 453 pp.
235 LINN, Andrew R. and Nicola McLELLAND (eds.): Standardization. Studies from the Germanic languages. 2002. xii, 258 pp. 234 WEIJER, Jeroen van de, Vincent J. van HEUVEN and Harry van der HULST (eds.): The Phonological Spectrum. Volume II: Suprasegmental structure. 2003. x, 264 pp. 233 WEIJER, Jeroen van de, Vincent J. van HEUVEN and Harry van der HULST (eds.): The Phonological Spectrum. Volume I: Segmental structure. 2003. x, 308 pp. 232 BEYSSADE, Claire, Reineke BOK-BENNEMA, Frank DRIJKONINGEN and Paola MONACHESI (eds.): Romance Languages and Linguistic Theory 2000. Selected papers from 'Going Romance 2000, Utrecht, 30 November-2 December. 2002. viii, 354 pp. 231 CRAVENS, Thomas D.: Comparative Historical Dialectology. Italo-Romance clues to Ibero-Romance sound change. 2002. xii, 163 pp. 230 PARKINSON, Dilworth B. and Elabbas BENMAMOUN (eds.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume XIII-XIV: Stanford, 1999 and Berkeley, California 2000. 2002. xiv, 250 pp. 229 NEVIN, Bruce E. and Stephen B. JOHNSON (eds.): The Legacy of Zellig Harris. Language and information into the 21st century. Volume 2: Mathematics and computability of language. 2002. xx, 312 pp. 228 NEVIN, Bruce E. (ed.): The Legacy of Zellig Harris. Language and information into the 21st century. Volume 1: Philosophy of science, syntax and semantics. 2002. xxxvi, 323 pp. 227 FAVA, Elisabetta (ed.): Clinical Linguistics. Theory and applications in speech pathology and therapy. 2002. xxiv, 353 pp. 226 LEVIN, Saul: Semitic and Indo-European. Volume II: Comparative morphology, syntax and phonetics. 2002. xviii, 592 pp. 225 SHAHIN, Kimary N.: Postvelar Harmony. 2003. viii, 344 pp. 224 FANEGO, Teresa, Belen MfiNDEZ-NAYA and Elena SEOANE (eds.): Sounds, Words, Texts and Change. Selected papers from 11 ICEHL, Santiago de Compostela, 7-11 September 2000. Volume 2. 2002. x, 310 pp. 223 FANEGO, Teresa, Javier PfeREZ-GUERRA and Maria Jose LOPEZ-COUSO (eds.): English Historical Syntax and Morphology. Selected papers from 11 ICEHL, Santiago de Compostela, 7-11 September 2000. Volume 1. 2002. x, 306 pp. 222 HERSCHENSOHN, Julia, Enrique MALLliN and Karen ZAGONA (eds.): Features and Interfaces in Romance. Essays in honor of Heles Contreras. 2001. xiv, 302 pp. 221 D'HULST, Yves, Johan ROORYCK and Jan SCHROTEN (eds.): Romance Languages and Linguistic Theory 1999. Selected papers from 'Going Romance' 1999, Leiden, 9-11 December 1999. 2001. viii, 406 pp. 220 SATTERFIELD, Teresa, Christina M. TORTORA and Diana CRESTI (eds.): Current Issues in Romance Languages. Selected papers from the 29th Linguistic Symposium on Romance Languages (LSRL), Ann Arbor, 8-11 April 1999. 2002. viii, 412 pp. 219 ANDERSEN, Henning (ed.): Actualization. Linguistic Change in Progress. Papers from a workshop held at the 14th International Conference on Historical Linguistics, Vancouver, B.C., 14 August 1999. 2001. vii, 250 pp. 218 BEND JABALLAH, Sabrina, Wolfgang U. DRESSLER, Oskar E. PFEIFFER and Maria D. VOEIKOVA (eds.): Morphology 2000. Selected papers from the 9th Morphology Meeting, Vienna, 24-28 February 2000. 2002. viii, 317 pp. 217 WILTSHIRE, Caroline R. and Joaquim CAMPS (eds.): Romance Phonology and Variation. Selected papers from the 30th Linguistic Symposium on Romance Languages, Gainesville, Florida, February 2000. 2002. xii, 238 pp. 216 CAMPS, Joaquim and Caroline R. WILTSHIRE (eds.): Romance Syntax, Semantics and L2 Acquisition. Selected papers from the 30th Linguistic Symposium on Romance Languages, Gainesville, Florida, February 2000. 2001. xii, 246 pp. 215 BRINTON, Laurel J. (ed.): Historical Linguistics 1999. Selected papers from the 14th International Conference on Historical Linguistics, Vancouver, 9-13 August 1999. 2001. xii, 398 pp. 214 WEIGAND, Edda and Marcelo DASCAL (eds.): Negotiation and Power in Dialogic Interaction. 2001. viii, 303 pp. 213 SORNICOLA, Rosanna, Erich POPPE and Ariel SHISHA-HALEVY (eds.): Stability, Variation and Change of Word-Order Patterns over Time. With the assistance of Paola Como. 2000. xxxii, 323 pp. 212 REPETTI, Lori (ed.): Phonological Theory and the Dialects of Italy. 2000. x, 301 pp. 211 ELSIK, Viktor and Yaron MATRAS (eds.): Grammatical Relations in Romani. The Noun Phrase, with a Foreword by Frans Plank (Universitat Konstanz). 2000. x, 244 pp. 210 DWORKIN, Steven N. and Dieter WANNER (eds.): New Approaches to Old Problems. Issues in Romance historical linguistics. 2000. xiv, 235 pp.
209 KING, Ruth: The Lexical Basis of Grammatical Borrowing. A Prince Edward Island French case study. 2000. xvi, 241 pp. 208 ROBINSON, Orrin W.: Whose German? The ach/ich alternation and related phenomena in standard' and colloquial'. 2001. xii, 178 pp. 207 SANZ, Montserrat: Events and Predication. A new approach to syntactic processing in English and Spanish. 2000. xiv, 219 pp. 206 FAWCETT, Robin P.: A Theory of Syntax for Systemic Functional Linguistics. 2000. xxiv, 360 pp. 205 DIRVEN, Rene, Roslyn M. FRANK and Cornelia ILIE (eds.): Language and Ideology. Volume 2: descriptive cognitive approaches. 2001. vi, 264 pp. 204 DIRVEN, Rene, Bruce HAWKINS and Esra SANDIKCIOGLU (eds.): Language and Ideology. Volume 1: theoretical cognitive approaches. 2001. vi, 301 pp. 203 NORRICK, Neal R.: Conversational Narrative. Storytelling in everyday talk. 2000. xiv, 233 pp. 202 LECARME, Jacqueline, Jean LOWENSTAMM and Ur SHLONSKY (eds.): Research in Afroasiatic Grammar. Papers from the Third conference on Afroasiatic Languages, Sophia Antipolis, 1996. 2000. vi, 386 pp. 201 DRESSLER, Wolfgang U, Oskar E. PFEIFFER, Markus A. POCHTRAGER and John R. RENNISON (eds.): Morphological Analysis in Comparison. 2000. x, 261 pp. 200 ANTTILA, Raimo: Greek and Indo-European Etymology in Action. Proto-Indo-European *agE-. 2000. xii, 314 pp. 199 PUTZ, Martin and Marjolijn H. VERSPOOR (eds.): Explorations in Linguistic Relativity. 2000. xvi, 369 pp. 198 NIEMEIER, Susanne and Rene DIRVEN (eds.): Evidence for Linguistic Relativity. 2000. xxii, 240 pp. 197 COOPMANS, Peter, Martin EVERAERT and Jane GRIMSHAW (eds.): Lexical Specification and Insertion. 2000. xviii, 476 pp. 196 HANNAHS, S.J. and Mike DAVENPORT (eds.): Issues in Phonological Structure. Papers from an International Workshop. 1999. xii, 268 pp. 195 HERRING, Susan C, Pieter van REENEN and Lene SCH0SLER (eds.): Textual Parameters in Older Languages. 2001. x, 448 pp. 194 COLEMAN, Julie and Christian J. KAY (eds.): Lexicology, Semantics and Lexicography. Selected papers from the Fourth G. L. Brook Symposium, Manchester, August 1998. 2000. xiv, 257 pp. 193 KLAUSENBURGER, Jurgen: Grammaticalization. Studies in Latin and Romance morphosyntax. 2000. xiv, 184 pp. 192 ALEXANDROVA, Galina M. and Olga ARNAUDOVA (eds.): The Minimalist Parameter. Selected papers from the Open Linguistics Forum, Ottawa, 21-23 March 1997. 2001. x, 360 pp. 191 SIHLER, Andrew L.: Language History. An introduction. 2000. xvi, 298 pp. 190 BENMAMOUN, Elabbas (ed.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume XII: Urbana-Champaign, Illinois, 1998.1999. viii, 204 pp. 189 NICOLOV, Nicolas and Ruslan MITKOV (eds.): Recent Advances in Natural Language Processing II. Selected papers from RANLP '97. 2000. xi, 422 pp. 188 SIMMONS, Richard VanNess: Chinese Dialect Classification. A comparative approach to Harngjou, Old Jintarn, and Common Northern Wu. 1999. xviii, 317 pp. 187 FRANCO, Jon A., Alazne LANDA and Juan MARTfN (eds.): Grammatical Analyses in Basque and Romance Linguistics. Papers in honor of Mario Saltarelli. 1999. viii, 306 pp. 186 MISESKA TOMIC, Olga and Milorad RADOVANOVIC (eds.): History and Perspectives of Language Study. Papers in honor of Ranko Bugarski. 2000. xxii, 314 pp. 185 AUTHIER, Jean-Marc, Barbara E. BULLOCK and Lisa A. REED (eds.): Formal Perspectives on Romance Linguistics. Selected papers from the 28th Linguistic Symposium on Romance Languages (LSRL XXVIII), University Park, 16-19 April 1998.1999. xii, 334 pp. 184 SAGART, Laurent: The Roots of Old Chinese. 1999. xii, 272 pp. 183 CONTINI-MORAVA, Ellen and Yishai TOBIN (eds.): Between Grammar and Lexicon. 2000. xxxii, 365 pp. 182 KENESEI, Isrvan (ed.): Crossing Boundaries. Advances in the theory of Central and Eastern European languages. 1999. viii, 302 pp. 181 MOHAMMAD, Mohammad A.: Word Order, Agreement and Pronominalization in Standard and Palestinian Arabic. 2000. xvi, 197 pp. 180 MEREU, Lunella (ed.): Boundaries of Morphology and Syntax. 1999. viii, 314 pp. 179 RINI, Joel: Exploring the Role of Morphology in the Evolution of Spanish. 1999. xvi, 187 pp. 178 FOOLEN, Ad and Frederike van der LEEK (eds.): Constructions in Cognitive Linguistics. Selected papers from the Fifth International Cognitive Linguistics Conference, Amsterdam, 1997. 2000. xvi, 338 pp.
177 CUYCKENS, Hubert and Britta E. ZAWADA (eds.): Polysemy in Cognitive Linguistics. Selected papers from the International Cognitive Linguistics Conference, Amsterdam, 1997. 2001. xxviii, 296 pp. 176 VAN HOEK, Karen, Andrej A. KIBRIK and Leo NOORDMAN (eds.): Discourse Studies in Cognitive Linguistics. Selected papers from the 5th International Cognitive Linguistics Conference, Amsterdam, July 1997.1999-vi, 187 pp. 175 GIBBS, JR., Raymond W. and Gerard J. STEEN (eds.): Metaphor in Cognitive Linguistics. Selected papers from the 5th International Cognitive Linguistics Conference, Amsterdam, 1997.1999. viii, 226 pp. 174 HALL,T. Alan and Ursula KLEINHENZ (eds.): Studies on the Phonological Word. 1999. viii, 298 pp. 173 TREVINO, Esthela and Jose LEMA (eds.): Semantic Issues in Romance Syntax. 1999. viii, 309 pp. 172 DIMITROVA-VULCHANOVA, Mila and Lars HELLAN (eds.): Topics in South Slavic Syntax and Semantics. 1999. xxviii, 263 pp. 171 WEIGAND, Edda (ed.): Contrastive Lexical Semantics. 1998. x, 270 pp. 170 LAMB, Sydney M.: Pathways of the Brain. The neurocognitive basis of language. 1999. xii, 418 pp. 169 GHADESSY, Mohsen (ed.): Text and Context in Functional Linguistics. 1999. xviii, 340 pp. 168 RATCLIFFE, Robert R.: The "Broken" Plural Problem in Arabic and Comparative Semitic. Allomorphy and analogy in non-concatenative morphology. 1998. xii, 261 pp. 167 BENMAMOUN, Elabbas, Mushira EID and Niloofar HAERI (eds.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume XI: Atlanta, Georgia, 1997.1998. viii, 231 pp. 166 LEMMENS, Maarten: Lexical Perspectives on Transitivity and Ergativity. Causative constructions in English. 1998. xii, 268 pp. 165 BUBENIK, Vit: A Historical Syntax of Late Middle Indo-Aryan (ApabhramEsa). 1998. xxiv, 265 pp. 164 SCHMID, Monika S., Jennifer R. AUSTIN and Dieter STEIN (eds.): Historical Linguistics 1997. Selected papers from the 13th International Conference on Historical Linguistics, Dusseldorf, 10-17 August 1997. 1998. x, 409 pp. 163 LOCKWOOD, David G., Peter H. FRIES and James E. COPELAND (eds.): Functional Approaches to Language, Culture and Cognition. Papers in honor of Sydney M. Lamb. 2000. xxxiv, 656 pp. 162 HOGG, Richard M. and Linda van BERGEN (eds.): Historical Linguistics 1995. Volume 2: Germanic linguistics.. Selected papers from the 12th International Conference on Historical Linguistics, Manchester, August 1995.1998. x, 365 pp. 161 SMITH, John Charles and Delia BENTLEY (eds.): Historical Linguistics 1995. Volume 1: General issues and non-Germanic Languages.. Selected papers from the 12th International Conference on Historical Linguistics, Manchester, August 1995. 2000. xii, 438 pp. 160 SCHWEGLER, Armin, Bernard TRANEL and Myriam URIBE-ETXEBARRIA (eds.): Romance Linguistics: Theoretical Perspectives. Selected papers from the 27th Linguistic Symposium on Romance Languages (LSRL XXVII), Irvine, 20-22 February, 1997. 1998. vi, 349 pp. + index. 159 JOSEPH, Brian D., Geoffrey C. HORROCKS and Irene PHILIPPAKI-WARBURTON (eds.): Themes in Greek Linguistics II. 1998. x, 335 pp. 158 SANCHEZ-MACARRO, Antonia and Ronald CARTER (eds.): Linguistic Choice across Genres. Variation in spoken and written English. 1998. viii, 338 pp. 157 LEMA, Jose and Esthela TREVINO (eds.): Theoretical Analyses on Romance Languages. Selected papers from the 26th Linguistic Symposium on Romance Languages (LSRL XXVI), Mexico City, 28-30 March, 1996.1998. viii, 380 pp. 156 MATRAS, Yaron, Peter BARKER and Hristo KYUCHUKOV (eds.): The Typology and Dialectology of Romani. 1997. xxxii, 223 pp. 155 FORGET, Danielle, Paul HIRSCHBUHLER, France MARTINEAU and Maria Luisa RIVERO (eds.): Negation and Polarity. Syntax and semantics. Selected papers from the colloquium Negation: Syntax and Semantics. Ottawa, 11-13 May 1995.1997. viii, 367 pp. 154 SIMON-VANDENBERGEN, Anne-Marie, Kristin DAVIDSE and Dirk NOEL (eds.): Reconnecting Language. Morphology and Syntax in Functional Perspectives. 1997. xiii, 339 pp. 153 EID, Mushira and Robert R. RATCLIFFE (eds.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume X: Salt Lake City, 1996.1997. vii, 296 pp. 152 HIRAGA, Masako K.,Christopher SINHA and Sherman WILCOX (eds.): Cultural, Psychological and Typological Issues in Cognitive Linguistics. Selected papers of the bi-annual ICL A meeting in Albuquerque, July 1995.1999. viii, 338 pp. 151 LIEBERT, Wolf-Andreas, Gisela REDEKER and Linda WAUGH (eds.): Discourse and Perspective in Cognitive Linguistics. 1997. xiv, 270 pp.
150 VERSPOOR, Marjolijn H., Kee Dong LEE and Eve SWEETSER (eds.): Lexical and Syntactical Constructions and the Construction of Meaning. Proceedings of the Bi-annual ICLA meeting in Albuquerque, July 1995. 1997. xii, 454 pp. 149 HALL, T. Alan: The Phonology of Coronals. 1997. x, 176 pp. 148 WOLF, George and Nigel LOVE (eds.): Linguistics Inside Out. Roy Harris and his critics. 1997. xxviii, 344 pp. 147 HEWSON, John: The Cognitive System of the French Verb. 1997. xii, 187 pp. 146 HINSKENS, Frans, Roeland van HOUT and W. Leo WETZELS (eds.): Variation, Change, and Phonological Theory. 1997. x, 314 pp. 145 HEWSON, John and Vit BUBENIK: Tense and Aspect in Indo-European Languages. Theory, typology, diachrony 1997. xii, 403 pp. 144 SINGH, R.K. (ed.): Trubetzkoy's Orphan. Proceedings of the Montreal Roundtable on "Morphonology: contemporary responses" (Montreal, October 1994). In collaboration with Richard Desrochers. 1996. xiv, 363 pp. 143 ATHANASIADOU, Angeliki and Rene DIRVEN (eds.): On Conditionals Again. 1997. viii, 418 pp. 142 SALMONS, Joseph C. and Brian D. JOSEPH (eds.): Nostratic. Sifting the Evidence. 1998. vi, 293 pp. 141 EID, Mushira and Dilworth B. PARKINSON (eds.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume IX: Washington D.C., 1995.1996. xiii, 249 pp. 140 BLACK, James R. and Virginia MOTAPANYANE (eds.): Clitics, Pronouns and Movement. 1997. 375 pp. 139 BLACK, James R. and Virginia MOTAPANYANE (eds.): Microparametric Syntax and Dialect Variation. i996.xviii, 269 pp. 138 SACKMANN, Robin and Monika BUDDE (eds.): Theoretical Linguistics and Grammatical Description. Papers in honour of Hans-Heinrich Lieb. 1996. x, 375 pp. 137 LIPPI-GREEN, Rosina L. and Joseph C. SALMONS (eds.): Germanic Linguistics. Syntactic and diachronic. 1996. viii, 192 pp. 136 MITKOV, Ruslan and Nicolas NICOLOV (eds.): Recent Advances in Natural Language Processing. Selected Papers from RANLP '95.1997. xii, 472 pp. 135 BRITTON, Derek (ed.): English Historical Linguistics 1994. Papers from the 8th International Conference on English Historical Linguistics (8 ICEHL, Edinburgh, 19-23 September 1994). 1996. viii, 403 pp. 134 EID, Mushira (ed.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume VIII: Amherst, Massachusetts 1994.1996. vii, 261 pp. 133 ZAGONA, Karen (ed.): Grammatical Theory and Romance Languages. Selected papers from the 25th Linguistic Symposium on Romance Languages (LSRL XXV) Seattle, 2-4 March 1995.1996. vi, 330 pp. 132 HERSCHENSOHN, Julia: Case Suspension and Binary Complement Structure in French. 1996. xi, 200 pp. 131 HUALDE, Jose Ignacio, Joseba A. LAKARRA and R.L. TRASK (eds.): Towards a History of the Basque Language. 1996. 365 pp. 130 EID, Mushira (ed.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume VII: Austin, Texas 1993.1995. vii, 192 pp. 129 LEVIN, Saul: Semitic and Indo-European. Volume I: The Principal Etymologies. With observations on Afro-Asiatic. 1995. xxii, 514 pp. 128 GUY, Gregory R., Crawford FEAGIN, Deborah SCHIFFRIN and John BAUGH (eds.): Towards a Social Science of Language. Papers in honor of William Labov. Volume 2: Social interaction and discourse structures. 1997. xviii, 358 pp. 127 GUY, Gregory R., Crawford FEAGIN, Deborah SCHIFFRIN and John BAUGH (eds.): Towards a Social Science of Language. Papers in honor of William Labov. Volume 1: Variation and change in language and society. 1996. xviii, 436 pp. 126 MATRAS, Yaron (ed.): Romani in Contact. The history, structure and sociology of a language. 1995. xvii, 208 pp. 125 SINGH, R.K. (ed.): Towards a Critical Sociolinguistics. 1996. xiii, 342 pp. 124 ANDERSEN, Henning (ed.): Historical Linguistics 1993. Selected papers from the 11th International Conference on Historical Linguistics, Los Angeles, 16-20 August 1993.1995. x, 460 pp. 123 AMASTAE, Jon, Grant GOODALL, M. MONTALBETTI and M. PHINNEY (eds.): Contemporary Research in Romance Linguistics. Papers from the XXII Linguistic Symposium on Romance Languages, El Paso/Juarez, February 22-24,1992.1995. viii, 381 pp. 122 SMITH, John Charles and Martin MAIDEN (eds.): Linguistic Theory and the Romance Languages. 1995. xiii, 240 pp. 121 HASAN, Ruqaiya, Carmel CLORAN and David G. BUTT (eds.): Functional Descriptions. Theory in practice. 1996. xxxvi, 381 pp. 120 STONHAM, John T.: Combinatorial Morphology. 1994. xii, 206 pp.