CONTRASTIVE LEXICAL SEMANTICS
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 Editoroal Board Henning Andersen (Los Angeles); Raimo Anttila (Los Angeles) Thomas V. Gamkrelidze (Tbilisi); John E. Joseph (Edinburgh) Hans-Heinrich Lieb (Berlin); Ernst Pulgram (Ann Arbor, Mich.) E. Wyn Roberts (Vancouver, B.C.); Danny Steinberg (Tokyo)
Volume 171
Edda Weigand (ed.) Contrastive Lexical Semantics
CONTRASTIVE LEXICAL SEMANTICS
Edited by
EDDA WEIGAND University of Münster
JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.
Library of Congress Cataloging-in-Publication Data Contrastive lexical semantics / edited by Edda Weigand. p. cm. -- (Amsterdam studies in the theory and history of linguistic science. Series IV, Current issues in linguistic theory, ISSN 0304-0763 ; v. 171) Includes bibliographical references and index. I. Semantics, Comparative. I. Title. II. Series. P325.5.C6W45 1998 401'.43--dc21 98-48761 ISBN 90 272 3676 3 (Eur.) / 1 55619 887 6 (US) (Hb; alk. paper) CIP © Copyright 1998 - John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O.Box 75577 · 1070 AN Amsterdam · The Netherlands John Benjamins North America · P.O.Box 27519 · Philadelphia PA 19118-0519 · USA
CONTENTS
Foreword The Lexical Item John Sinclair
vii 1
Contrastive Lexical Semantics Edda Weigand
25
The Vocabulary of Emotion A contrastive analysis of ANGER in German, English, and Italian Edda Weigand
45
Sprachvergleich als Textvergleich Christian Schmitt
67
Zur kontrastiv-semantischen Analyse von Emotionen Semantische 'Ärgerdörfer' im Russischen und im Deutschen Valerij Dem 'jankov
95
Equivalence in Contrastive Semantics The effect of cultural differences Henning Westheide
119
Sprachsystem und Sprachgebrauch in der kontrastiven lexikalischen Semantik Eckhard Hauenherm
139
Composition Principles within the Word and within Ways of Use of Words Claude Gruaz
163
VI
CONTENTS
Lexical Items and Medium-Transferibility in English and German Jürgen Esser
173
Types of Lexical Variation Christoph Schwarze
187
Non-overt Categories in Russian Partitive and Pseudo-partitive Constructions Anita Steube & Andreas Späth
209
Russian and German Idioms from a Contrastive Perspective Dmitrij Dobrovol 'skij
227
Word Sense Disambiguation An experimental study for German Wolf Paprotté
243
General Index
263
List of Contributors
269
FOREWORD This volume contains most of the papers given at an International Workshop on "(Contrastive) Lexical Semantics" at the University of Münster in May, 1997. A few papers on the same topic are added. Questions of lexical semantics in general and of contrastive lexical semantics in particular were addressed from different perspectives, from the pragmatic perspective of a corpus-oriented approach as well as from the model-oriented perspective of sign theoretic linguistics. The pragmatic per spective is crucial to a project on language comparison which aims to an alyse and describe the whole vocabulary-in-use in the area of emotion (see the papers by Weigand, Schmitt, Dem'jankov, Westheide, in part also Hauenherm and Gruaz). After the pragmatic turn, lexical semantics can no longer be seen as a discipline on its own but has to be developed as an integral part of a theory of language use. Essential features of individual languages can be discovered only by looking beyond the limits of our mother languages and including a contrastive perspective. Thus also lexical seman tics of individual languages is considered to be in part contrastive semantics. The project is characterized not only by the features 'pragmatic' and 'con trastive' but also by the feature 'corpus-based' which has been gaining ground in recent years. Semantic conventions can no longer be justified by native competence alone; instead, they have to be verified by "hard, measur able evidence" (Sinclair, Introduction to the Cobuild Dictionary) on the basis of representative text corpora of languages-in-use. Within a pragmatic, corpus-oriented approach essential new ideas are discussed, mainly the insight that single words can no longer be considered to be the lexical unit. It is the complex multi-word lexical unit a pragmatic approach has to deal with. The papers by Sinclair and Weigand address this multi-word lexical unit from different starting points: from the point of a formally and automatically retrievable unit and from the point of a functionally and syntactically defined unit of use.
VIII
FOREWORD
Arguing for a complex lexical unit has its roots in the history of lexical research. Bierwisch had already noted in the sixties that there were some difficulties resulting from specific relations between the word and the context. Wunderlich, like Bierwisch participating in the workshop, also stressed the importance of the context for the analysis of lexical meaning very early on. This line of taking as lexical unit the single word and of considering it within the context reached its climax, for instance, in the work by Pustejovsky, to whom the generative approach by Wunderlich is indebted. Necessarily however it reaches its limits in so far as the rule-governed model-oriented approach, in principle, cannot tackle all the varieties and idiosyncrasies of language use and therefore remains restricted to a subset of examples. The pragmatic, corpus-oriented approach, on the contrary, makes this claim and tries to develop a new methodology for the new object of vocabulary-in-use. The categories 'model- versus corpus-oriented' express only priorities. Naturally, the model-oriented perspective also deals with empirical material and the corpus-oriented perspective tries to structure the material according to a model. Besides the pragmatic model, other models are addressed in this volume such as the two-level model by Bierwisch, Lang, and Wunderlich in the papers by Schwarze and Steube & Späth, the structural model in the papers by Gruaz and Esser, a model of contrastive idiom analysis by Dobrovol'skij and a computerlinguistic model by Paprotté. It is however the general ground of fundamental questions regarding lexical meaning which gives rise to a fruitful and inspiring discussion. We all have to deal with the same underlying theoretical questions which refer to the rules, conventions, and principles that guide us in language use. At the workshop, such a discussion of fundamentals led to the proposal by Wunderlich to found a "Contrastive Circle". In focusing on the vocabulary of natural languages in all its complexity, the corpus-based view seems to be quite different from the model-oriented view. While the corpus-linguists are struggling with large corpora and with difficult material which occasionally seems to defy any analysis, the theorists are confronted with models of clear logical lines which are allowed to keep to their inherent logic. At this cross-road, the scientific interest of the individual researcher determines the route to be taken. Perhaps, one day, the "Con trastive Circle" will meet again and consider old and new relationships. Finally, there remains the pleasant duty to thank those who helped us to make the workshop and the publication of the papers possible, especially the
FOREWORD
IX
Ministry of Science and Research of North Rhine-Westphalia for financial support in the initial phase of the project and Prof. E.F.K. Koerner for accepting the volume in his series and for helping straightforwardly and quickly whenever questions of any type arose. I would also like to thank Eckhard Hauenherm, Larissa Wunderlich, and Giuseppina Giordano for formatting the papers and producing a unified volume and for compiling the index.
Münster, August 1998
Edda Weigand
THE LEXICAL ITEM JOHN SINCLAIR University of Birmingham & The Tuscan Word Centre 1.
Introduction In the early days of computational lexicology, some forty years ago, it was felt important to distinguish between a 'word' and a 'lexical item'. This was a period for emphasising the complexity of language, and proposing abstract categories of language form to escape from the confines of surface phenomena; procedural models, like Immediate Constituent Grammar (e.g., Stageberg 1966:262ff.) were held in suspicion because position and sequence were non-negotiable. On the other hand, Firth (1957) made a distinction between 'sequence', the physical positioning of linguistic events relative to each other, and 'order', realised mainly by sequence, which was an abstract representation of language form.1 Hence an orthographic word was recognised as a string of characters lying between spaces, but the equivalent lexical unit had greater freedom, sometimes being more than a word, and possibly even less than a word in extent, with some variation and discontinuity. At the time it was felt that in most cases the physical and the formal categories would coincide exactly, but there was a small amount of superficial evidence to justify the distinction (Sinclair 1996b). Words such as another, maybe, wherever were so obviously concatenations of otherwise separate words that little phrases such as in order to, as if, of course could easily be interpreted as essentially the same 1 The problem is still with us in the attempts to write grammars that are explicit enough to drive machine applications. Chomsky (e.g., 1965) managed to combine freedom from 'surface structure' with a procedural model of sentence generation. This was appealing on presentation, but when attempts were made a few years later to build in some of the complexities it exposed, the problem of the ordering of rules became one of the main preoccupations of the period; present-day grammars in this tradition are notoriously incapable of analysing open text.
2
JOHN SINCLAIR
structures with the word space(s) retained. Compounds and (in English) phrasal verbs were so important that large sections of the morphology - for compounds - and substantial digressions in the syntax - for phrasal verbs were regularly devoted to these in descriptive publications. The phrasal verb resisted all efforts to accommodate it in a description where word and lexical item were fused. Idioms and other phraseological conventions were regarded as im portant by the powerful language teaching lobby, but as little more than a nuisance by grammarians and lexicographers. Is there any point in analysing Don't count your chickens before they're hatched? On the one hand, it looks like a well-formed sentence of two clauses, but on the other hand there seem to be hardly any alternatives to the succession of word choices, and a gram matical analysis in such circumstances has little value if one believes that meaning arises from choice.2 At this time, computers were just beginning to cope with text rather than numbers, and they were ill-equipped for the task, both in hardware provision and suitable programming languages. One of the few tasks they could perform reliably was the identification of orthographic words, and to leave aside the distinction between word and lexical item seemed at the time an innocent simplification in practice, since the distinction had been clearly recorded for future attention. The simplification was not challenged; in fact it was supported by two other contemporary models of language description. One was the dictionary model, which for practical reasons - ease of consultation - if nothing else, has always been based on the rough equation of a word and a unit of meaning. Phrases and idioms warrant a note at the end of some entries, but there is no attempt at all to articulate a theory that allows for them, and for the relationship between the 'independent' and the 'dependent' uses of a word (Sinclair 1987a). The other model of the period was the lexical component of a trans formational grammar (e.g., Katz & Fodor 1963), which was clearly wordbased. The representation of text, for example as the output of the phrase 2
Now that the behaviour of such phrases can be studied in large corpora, it is clear that there is a substantial amount of variation, but of a lexicosemantic or phraseological nature rather than a grammatical one. Of the 16 instances of chicken and count in a corpus of 200 million words, only one instance of the presumed canonical form is to be found. Two others have the before clause, and three have another timing expression; eleven have a possessive adjective in front of chickens, and there is one instance each of no and any; the remaining three are in the structure of counting chickens, which refers to the idiom rather than quoting it.
THE LEXICAL ITEM
3
structure component, was as a string of words, each of which became an entry in the lexicon. Although the argumentation in support of this decision was not provided, the unrecognised assumption that the word is the unit of lexical meaning remains largely unchallenged. As a result of this insensitivity to lexis, the last generation of linguistics has seen grammar going through many stages of sophistication, whatever the theoretical model of preference, while the unit of lexis has remained fixed and concrete, pinned to the surface of language. The apparatus of gram matical description has acquired many components, levels, categories and scales, while lexical description has nothing but feeble surface categories like word and collocation.3 The need for a more detailed and abstract model for lexical description became clear when lexical information began to be extracted from multimillion word corpora in the early eighties.4 Several long-accepted conven tions in lexicography were called into question - for example the idea that a word could inherently have one or more meanings. The working assumption was that when these meanings were explicated (or translated, in a bilingual dictionary) and, in the better dictionaries, exemplified, the lexicographer's job was done. This practice proved to be incapable of organising the strong, recurrent patterns that were shown by corpus analysis to be present in the way words were used in texts; the importance of the surrounding language far outweighed the question of how many meanings and how they were related to each other. A dictionary is a practical tool, and no place to introduce a new theory of meaning, even if one had been available. So the first dictionary derived from a corpus (Sinclair & Hanks 1987) was conservative in its design, in relation to the disturbing nature of the evidence encountered by the lexico graphers. Despite the massive organising effect of cotext, pride of place in an entry was usually awarded to one of the meanings least dependent on the cotext. At one time in the design phase of Cobuild it was suggested that the normal pattern of an entry should be turned upside down, since the phrasal uses, which are habitually summarised crudely at the end of entries, if at all, were so important; but this was judged to be unhelpful for normal users, who 3
I do not refer here to the theories and descriptions of semantics, which have often been richly complex; however, since they are not systematically related to language in use, they are therefore not relevant to the present stage of this argument. 4 See Sinclair (1987c) for a detailed account of the way in which this problem was en countered and dealt with in lexicography. Chapter 4 is particularly relevant.
4
JOHN SINCLAIR
might become very confused, no matter how theoretically sound the structure might be.5 As research progressed in the following decade, it became clear that the original distinction between a word and a lexical item was the key to a more helpful description of the vocabulary, and a more accurate account of mean ing. The 'innocent simplification'6 obscured two basic facts: (a) many, if not most, meanings require the presence of more than one word for their normal realisation; (b) patterns of co-selection among words, which are much stronger than any description has yet allowed for, have a direct connection with meaning. 2.
Meaning The establishment of a lexical item as an abstract category distinct from the word requires us to take meaning into account, and it is important not to overburden the argument with all the problems that the concept of meaning brings along with it. Only one problem will be tackled, a central structural problem that can be discussed without raising the others.7 The problem is as follows: in dictionaries, lexicons, thesauruses etc., meanings are linked primarily with words; how can the appropriate meaning be identified in a text, at the point where a word occurs? If a word has fifty different meanings in a dictionary, how is an occurrence of the word related to just one of those meanings? And what happens if none of the fifty mean ings precisely covers the case? It is contended here that the theoretical frameworks that linguists use are inadequate to solve this problem - or even to state the problem in such a way that a solution could be attempted. So there must be adjustments made to our theoretical perspectives, after which the problem can be restated. 5
That early response to corpus data has since been repeated many times as more lexico graphers encounter corpus evidence. 6 There was more than one parameter of simplification; for example the organisation of meanings around 'headwords' - lemmas in computational linguistics - carries an assumption that, by and large, the inflected forms of a word do not have distinctive meanings. This view is now regarded as rather suspect (Tognini-Bonelli 1995), and it is to be expected that a new generation of dictionary will arise where the indexing is through the form and not the lemma. 7 In particular, this paper does not address the general matter of what is meant by stretches of language; where it is necessary for presentation of the argument to indicate meanings or distinctions of meaning, the statements are quite informal. Only when the meaningful units of a language have been reliably identified will it be useful to examine this matter thoroughly, and then, since the links between form and meaning will not have been broken, the task of description will be different from current work in semantics.
THE LEXICAL ITEM
5
The initial statement of this problem of meaning has some similarity with Chomsky's well-known view of a grammar as a finite set of rules that specifies the non-finite set of sentences of a language. Chomsky was con cerned primarily with syntactic well-formedness, but the same problem arises with the lexis. The problem in both cases is how to relate a finite resource to an unlimited set of applications; in the case of syntax the set of rules is finite and the set of sentences is not; in the case of lexis the set of meaningful items is finite and the set of meanings in use does not appear to be limited. Chomsky's solution to the problem of linking the rules and the sen tences was to introduce recursive rules, but recursion is not a useful concept in the lexicon. Recursion is one of the simpler types of combination; a single rule-form in the grammar8 provides the link between finite and unlimited sets. For the combinatorial relations of the lexicon, a more complex relationship needs to be defined. That is the purpose of this paper. A lexicon, or a dictionary, consists of a list of words, to each of which is attached a number of statements, features etc., which together express what can be said about the meaning of the word. The words are arranged according to the rules of grammar to make text.9 The problem is that a text is a unique deployment of meaningful units, and its particular meaning is not adequately accounted for by any organised concatenation of the fixed meanings of each unit. This is because some aspects of textual meaning arise from the particular combinations of choices at one place in the text, and there is no place in the lexicon-grammar model where such meaning can be assigned. Since there is no limit to the possible combinations of words in texts, no amount of documentation or ingenuity will enable our present lexicons to rise to the job. They are doomed. However, the meaningful combinations can now be described in new ways which make them much more tractable. At the time lexicons of the familiar kind were being designed, much of the regularity of the combi nations was obscured by the inadequate means of observation of the data. It was impossible to gather together a sufficient quantity and selection of data in 8
One where the same symbol appears on both sides of the operator, like 'X→XY', meaning 'rewrite X as XY'. Obviously the second X can also be expanded, and introduces another X, and so on indefinitely. See Bach's (1964) treatment of recursion. 9 It is conceded throughout this argument that occasionally a multi-word phrase may be used as an item in the wordlist. However this raises a further problem - when two words appear together in a text, how do we know whether they realise one meaning or two? The process known as tokenisation in computational linguistics is a relatively straightforward matter when the orthographic word can be trusted, but expands infinitely as soon as multi-word units are recognised.
6
JOHN SINCLAIR
which the underlying patterns could be identified. The application of powerful computers to large text corpora has begun to improve our methods of observation. 3.
Reversal To give just one example of the inadequacies of a lexicon built by established methods, they do not take into account the common phenomenon of semantic reversal.10 Situations frequently arise in texts where the precise meaning of a word or phrase is determined more by the verbal environment rather than the parameters of a lexical entry. Instead of expecting to understand a segment of text by accumulating the meanings of each successive meaningful unit, here is the reverse; where a number of units taken together create a meaning, and this meaning takes precedence over the 'dictionary meanings' of whatever words are chosen. If the two meanings (the one created by the environment and the item, and the one created by each item individually) are close and connected, the text is felt to be coherent; if they do not, some interpretation has to be made - perhaps the meanings of the items are neutral with respect to the semantic demands of the environment, perhaps there is a relevant metaphorical interpretation, or an irony, or a very rare meaning of an item, or a special interpretation because of what the text is about at this point. Whenever the meaning arises predominantly from the textual envi ronment rather than the item choice, it is considered to be an instance of semantic reversal. The flow of meaning is not from the item to the text but from the text to the item. In practice, the flow is rarely in one direction only. The textual environment will nearly always have some effect on the meaning of a unit, and the accepted features of the meaning will not often be totally ignored.11 So the problem of the description of meaning can be traced back to the rigidity and frequent irrelevancy of the lexicons that supply the meanings. Not only do they (a) only supply some of the meaning, but also (b) they often supply meaning components that do not fit the particular environment; in addition (c) there is no mechanism that I am aware of for adapting a lexical
10
Reversal at a propositional level is introduced in Sinclair (1987b) but not given a name. 11 Presumably the often invoked and rather vague criterion for a compound or idiom (e.g., Cruse [1986:37]: "an expression whose meaning cannot be inferred from the meanings of its parts") applies to the more extreme cases of the effect of the textual environment.
THE LEXICAL ITEM
7
entry to suit a particular recurrent set of circumstances, far less an individual instance. The effects of reversals can be seen in dictionaries and lexicons when a word is frequently found in collocation with another, and this has an effect on the meaning. For example, white wine is not white, but ranges from almost colourless to yellow, light orange or light green in colour. That is to say, the meaning of white when followed by wine is a different colour range than when it is not. Traditional dictionaries tend to obscure this point by using encyclo paedic information to explain the meaning, for example: (of wine) made from pale grapes or from black grapes separated from their skins (Collins English Dictionary, 1991).
This assumes that the user already knows roughly what colour white is when collocated with wine. Such examples are familiar enough. But when the word holy is interpreted as an abnormal mental state, as in the example The ambience borders on the holy ... we must assume that this semantic feature is assigned not by any lexicon. There is no lexicon that sanctions such a meaning, and indeed, if there were, it would be a distraction with reference to most of the occurrences of the word holy. The meaning is created in the collocation with borders on. Whatever follows this phrase indicates the limits of normality by specifying a mental state that lies just outside normality. When the adjective is obsessional, the feature of abnormality is already present in the meaning of the word, and the co-ordinated choice will be felt to be coherent; in the case of holy, the required feature has to be added by reversal. And if there were an instance of borders on the normal, this would be interpreted as fully ironical, suggesting that the normal is unexpected. The way in which such a semantic problem is tackled is parallel to (or even identical with) a wide range of potential difficulties of interpretation; for example if in a conversation someone says Wasn 't that awful what happened to Harry?, and if you, as the receiver of this query do not know a Harry to whom something awful has happened (or you know more than one unfortunate Harry), you have two basic courses of action. You can either challenge the presupposition that there is no identification problem, by point ing out your ignorance and/or your interlocutor's lack of clarity, or you can use a reversal technique - try to pick up from the following conversation what you need to know. Or if you are reading a book that is outside your
8
JOHN SINCLAIR
normal area of expertise, and come across items that the author assumes you know the meanings of - like abbreviations - you can either break off in your reading and consult a specialised glossary, or plough on with whatever understanding you can glean from reversal techniques. After a while either your interpretation will break down altogether or you will survive the particular passage because none of the unresolved meanings are critical to your overall understanding. 4.
Theory Adjustment 'Reversal' is one of the new descriptive categories that will be required to account adequately for the data. The problem goes deeper, however, and requires some reorientation of theory. As a start, here are three hypotheses that, in various ways, cut across established theoretical norms and assumptions. Without the acceptance of these hypotheses, or their refutation, progress towards a better account of meaning will be slow. I
Language text is not adequately modelled as a sequence of items, each in an environment of other items.
We normally accept an underlying model of language as 'itemenvironment' . At any point in a text we can interpret the occurrence of an item in terms of what other choices are possible, given the environment. Hence each item is both an item in its own right and a component of the environment of other items. This model must be examined carefully, because it seems inherently implausible. Each item would have to be interpretable, simultaneously, as having many different meaning-relations with other items, equally multifaceted. As each item came to be processed for its contribution to the meaning of the text, every other item nearby would change not only its meaning, but the basis of its meaning - whether central or peripheral to the node in focus. This would lead to a huge multiplicity of meanings, and the need for a complex processor to relate them to each other, discard irrelevant ones, etc. Further, since we have no reason to believe that the interpretation will proceed on a strict linear basis, the model would become extremely complex. Such complexity reflects the interdependence between words and their environments, and makes it clear that all the patterning cannot be described at once; some elaboration of the model is needed in order to disperse the complexity. This could take the form of a diversification of units into
THE LEXICAL ITEM
9
different types, as grammar has nouns, verbs etc. - we already recognise 'grammatical' and 'vocabulary' words, without being able to distinguish them formally, and we use equally informal terms such as 'idiom', 'figurative' and 'metaphorical' in lexicography. It could also take the form of erecting a hierarchy of units, a rank scale in Halliday's (1961) terms, where structural patterns of different types and dimensions are arranged in a taxonomy. Exactly what model will emerge is not possible to predict at present; my aim in this paper is to establish the need for it, and to put aside some well-respected assumptions about language that may be hampering our thinking - such as the imbalance at present in favour of the independence of the word, rather than its interdependence on its 'cotext', or verbal environment. To put these notions aside is not to discard them for ever, or to attack their intellectual integrity, or the competence of those who uphold them; it is merely to suspend their operation temporarily to see if the picture that emerges from a rearrangement is more satisfying than the present one. This point is particularly important with respect to the second hypothesis that I would like to put forward: II
Ambiguity in a text is created by the method of observation, and not the structure of the text.
If a word is likely to be intricately associated with the words that occur round about it, then the consequences of studying its meaning in isolation are unpredictable. Dictionaries, which have little choice than to organise their statements of meaning around the word, present a picture of chaotic ambiguity. Words have many meanings, and there is no way of working out in advance which one is appropriate in a text. However, if we extend the viewpoint to two or three words (which is normal when lexicographers recognise a relatively fixed phrase) much of the ambiguity drops away.12 People use this extended viewpoint so naturally in reading and listening, and language teachers labour the importance of con centrating on the broad aspects of meaning and not the particulars of a single word. Despite almost universal accord with the position that the environment of occurrence is important in text structure, every machine lexicon I know persists in starting with the inappropriate unit, the word. When such a 12 This can be made clear in a bilingual context; see Sinclair et al. (1996:173-174) for a practical demonstration.
10
JOHN SINCLAIR
lexicon is applied to a text, all the possible meanings of a word are listed, including whatever phraseological meanings have been noticed, as if all were potentially relevant on each occasion of the occurrence of the word. Having created quite fictitious ambiguities, the researcher then multiplies them with similarly complex possibilities for the next word, and the next ... leading to the most innocuous sentence having many thousands of possible 'meanings'. Here is a superficial example.13 In Collins English Dictionary, words are assigned a number of meanings as follows, approximately: cat mat
24 17
sit the
18 15
on
25
The number of possible combinations of meaning, on these figures and multiplying the in twice, for The cat sat on the mat is thus 41,310,000. In this desperate situation, there arises a need for sensitive algorithms to filter out all the meanings that do not apply in the particular instance, all except one. The mess is so serious by now that this cannot be achieved by automatic process alone; humans must be trained and employed to clear it up. This process must be compared with the normal linguistic activity of an ordinary person. All day long, effortlessly, this person interprets passing sentences, usually correctly, and often against a background of high levels of distraction. He or she is not even aware that any of them are potentially ambiguous. The discrepancy that we perceive between human and machine be haviour is so gross that the model behind the machine's performance must be questioned. The human, perhaps, works with a better notion of meaningful units, and does not encounter ambiguity.14 This position casts doubt on the relevance to language study of all the work over many years in Artificial Intelligence and Natural Language Processing which concentrates on the resolution of ambiguity. The ambiguity that is studied is evidenced in carefully contrived short utterances, often of a kind which would be very unlikely to occur in texts. This possible objection 13 I hope nobody actually does this, because I have deliberately exaggerated the problem to show how close it is to absurdity. 14 Here as elsewhere one cannot make absolute statements. Accidental ambiguities, that may be irresolvable even when the viewpoint is extended to the limits of practicality, are bound to arise, but very occasionally indeed; linguistic communication would be severely strained if it was more common than the sort of coincidence that happens once or twice in a lifetime. Ambiguity above the phrase - at propositional or pragmatic levels - is no more common than ambiguity at word level, but needs separate treatment which would not be appropriate here.
THE LEXICAL ITEM
11
to them is dismissed on the grounds that grammar deals with potential utterances, and whether or not an utterance actually occurred is uninter esting.15 Now that large corpora can provide, with growing reliability, state ments of regularity and norms in language usage, the marginal status of the work on artificial ambiguity can be clearly seen.16 This is not the place for a thorough examination of ambiguity and associated phenomena; the aim here is merely to open a case for reorientation of our attitudes to a very firmly established viewpoint. A representative account of the current NLP positions on the topic can conveniently be found in Monaghan (1996b). Most of the types presented there can be classified as one of: (a) Created by the observer's perspective - this is the commonest; (b) suggested by generality or vagueness of reference; one of the strengths of human language is its ability to avoid having to discriminate (Channell 1994); (c) created by reference to a formally marked distinction in another language - a genuine problem for translators but not an ambiguity in any one communication system; (d) created by a grammar which lays claim to too much potential meaning (i) combinations which might be contrastive if they occurred, but are inhibited by other factors; (ii) contrasts which are not formally distinguished in the language system and therefore can only be a matter for interpretation outside the system. In other words, none of these count as ambiguities that have to be resolved in language description. Indeed, it can justifiably be claimed that a model of language is inappropriate if it obliges the description to make distinctions in a particular text segment which are not necessary for the interpretation of the text as a whole. III
The form of a linguistic unit and its meaning are two perspectives on the same event.
15 Monaghan (1996a: 12) goes as far as to say that the cases where corpus evidence does not work "will probably be the most interesting and most crucial ones, since these will be the rarest in most corpora". 16 Marginal for the theory of how language works, that is. There may be relevance, for example, to research in cognitive psychology.
12
JOHN SINCLAIR
At first sight, most scholars in a structuralist tradition would not see anything exceptional in this statement. It is accepted that form and meaning are very closely related, and that variation in one normally leads to variation in the other. Even for those whose viewpoint leads them to perceive ambi guity, the close alignment of form and meaning would not be undermined; one would have to accept a lot of overlap, which is not the same as either a confusion between categories or a fusion of them. indeed in retrospect it is clear that the association between form and meaning has had to be somewhat loose, to allow for such notions as transformation (Harris 1957). As has been argued in the case of words, syntactic patterns may seem to vary independently of the meaning, as long as the cotext is kept to a minimum. So if active and passive constructions are presented bereft of cotext, their similarity of meaning ("who did what to whom") is highlighted, and their differences, which show in the higher organisation of the discourse, are not obvious at all. The position adopted in this paper is intended to be in sharp contrast to the approximateness of the traditional view. It is asserted that form and meaning cannot be separated because they are the same thing. Considered in relation to other forms, a lexical item is a form; considered in relation to other meanings, it is a meaning. It follows from this tighter statement that ambiguity must in practice be very close to zero, or the statement would have to be seriously weakened. Also, the form of a lexical item must include all the components that are realised in the example. Meaning cannot inhere satisfactorily in just a selection of the components of an item when there are other components left in the cotext, but requires them all to be assembled together, and a way of stating the structure of an item has to be devised. It follows from the requirement that all the components of a lexical item must be included in its specification, that these genuine meaning-bearing items will have very little connection with their cotexts; all the choices that depend substantially on other choices will be grouped together in the item, and the text will be represented essentially as a succession of relatively largescale and independent choices. No doubt the reality will be a good deal more complicated than this sketch supposes. Discontinuity of lexical items is a strong possibility, and various kinds of embedding cannot be ruled out. The method of work, which is based on studying and processing concordances of words and phrases, draws out and highlights those choices that contain an element of coselection
THE LEXICAL ITEM
13
or conditioned selection. Other choices, which may also be structural, are less obvious in the data studied at present because they are sporadic with reference to the pattern of choices. 5.
The Axes of Patterning In order to restate the problem of meaning, we must draw some general implications from the three hypotheses that have been discussed above. The main one, which pervades the whole argument, is that the tradition of linguistic theory has been massively biased in favour of the paradigmatic rather than the syntagmatic dimension. Text is essentially perceived as a series of relatively independent choices of one item after another, and the patterns of combination have been seriously undervalued. It is easy to understand how this has happened; once again it depends on the nature of the observations and the stance of the observer. The difficulty has been how to cope with the large range of variation that is apparent in most uses of language. In presenting structure, traditional lin guistics puts most of the variation to one side through the device of sepa rating grammar and semantics at the outset. This then obscures most of the structural relevance of collocation, and removes any chance of the precise alignment of form and meaning. It also presents the semantic level with the kind of problems that this paper is discussing. The opportunity to observe recurrent patterns of language in corpora has shown how choices at word rank co-ordinate with other choices round about in an intricate fashion, suggesting a hierarchy of units of different sizes sharing the realisation of meaning. The largest unit will have a similar status to the sentence in grammar (and may coincide with sentence boundaries in many instances) in that it will be relatively independent of its surroundings with respect to its internal organisation (see the distinction between rank and level in Halliday 1961). Meaning appears to be created by paradigmatic choice; this is within the orthodoxy of most theories, whether or not it is explicit. This perception also relates meaning to the information of Information Theory. However, the mechanism of paradigmatic choice is so powerful that constant vigilance should be exercised to make sure that it is not misapplied. Sometimes in the actual use of language there is less choice than the paradigm is capable of creating, as in the example of counting chickens given earlier; in these cases to present the paradigm unqualified is to distort the description by claiming more meaning in an expression than is actually usable.
14
JOHN SINCLAIR
It seems that this happens as a matter of routine in most published descriptions, and that a language is characterised as having hundreds of thousands of meanings that are not in fact available, because they are constrained by the need for other choices in the environment. By giving greater weight to the syntagmatic constraints, units of meaning can be identified that reduce the amount of meaning available to the user to something more like his or her normal experience; the balance between the two dimensions will more accurately represent the relation between form and meaning. Such a conclusion calls for nothing less than a comprehensive redescription of each language, using largely automatic techniques. Problems remain, particularly one concerning the inability of the paradigmatic and the syntagmatic dimensions to relate to each other. They have no contact with each other, they are invisible from each other, and to observe one, the other has to be ignored. The phenomenon is similar to the observational problems that led to Heisenberg's famous Principle of Uncertainty in atomic physics. An atom can have both position and momentum, but these cannot be observed simultaneously, because the techniques for observing one cut out the possibility of observing the other. Similarly, a word gives information through its being chosen (paradigmatic) and at the same time it is part of the realisation of a larger item (syntagmatic); in order to observe either of these, however, we lose sight of the other. Unless the requirements of the cotext are precisely stated, the word as a paradigmatic choice will be invested with far too much independent meaning; on the other hand when observed purely as a component of a larger syntagmatic pattern, it can have very little freedom, and therefore can give very little information; it might be no more meaningful than a letter in a word, serving only the purpose of recognition. A means must be found of relating the two dimensions in order to give a balanced picture. We are now in a position to restate the problem of meaning in tractable terms by means of the following hypothesis: IV
The meaning of a text can be described by a model which reconciles the paradigmatic and syntagmatic dimensions of choice at each choice point.
The model is set out in a preliminary fashion in Sinclair (1996a). Five categories of co-selection are put forward as components of a lexical item; two of them are obligatory and three are optional. The obligatory categories
THE LEXICAL ITEM
15
are the core, which is invariable, and constitutes the evidence of the occur rence of the item as a whole, and the semantic prosody, which is the deter miner of the meaning of the whole, as we shall see in the example below. The optional categories realise co-ordinated secondary choices within the item, fine-tuning the meaning and giving semantic cohesion to the text as a whole. The optional categories serve also as a means of classifying the members of a paradigm, and thus the two axes of patterning, the paradigmatic and the syntagmatic, are related; the relationship is in principle capable of automation, and is quantifiable. The three categories that relate words together on either dimension are collocation, colligation, and semantic preference. The first two are Firth's (1951, 1957) terms. Collocation (at present) is the co-occurrence of words with no more than four intervening words17 (given the arguments of this paper, the word is no more reliable as a measure of the environment than it is as a unit of meaning, so this measure will have to be revised, but it is at present the only measure in general use). On the syntagmatic dimension, collocation is the simplest and most obvious relationship, and it is fairly well described. On the paradigmatic dimension it is defined rather differently, because items can only collocate with each other when present in a text, and two items in a paradigm are by that arrangement classed as mutually exclusive. The relationship is that of mutual collocation, i.e., that they both collocate (on separate occasions, usually) with the same item or items. So whereas manual and restoration are both significant collocates of work, they themselves do not co-occur significantly. Colligation is the co-occurrence of grammatical phenomena, and on the syntagmatic axis our descriptive techniques at present confine us to the co occurrence of a member of a grammatical class - say a word class - with a word or phrase. Colligation as a paradigmatic concept is displaced, like collocation, to that of a mutual relationship; so a possessive may colligate with a particular noun, and the so-called 'periphrastic' construction, the ... of ... may occasionally occur as an alternative. So your (etc.) true feelings is the norm, but the true feelings of people is an example of the less common structure in this phrase. 17 The calculation of the optimal size of the collocation span for English was first published in Sinclair, Jones & Daley (1970). At that time corpora were very small, so the figure was recently recalculated with reference to a much larger corpus, and with surprisingly little change. Oliver Mason programmed the recalculation, and calls it lexical gravity; it is a function in the CUE system of corpus query language.
16
JOHN SINCLAIR
Semantic preference is the restriction of regular co-occurrence to items which share a semantic feature, for example that they are all about, say, sport or suffering. This feature is relevant in the same way to both syntagmatic and paradigmatic phenomena. The three categories are related to each other in increasing abstraction; collocation is precisely located in the physical text, in that even the inflection of a word may have its own distinctive collocational relationships. To ob serve colligation one has to assign a word class to each word under examination; where there is a preponderance of one particular word class, this is colligation. Within the abstraction of the word class, of course, there may be one or more collocations. Semantic preference requires us to notice similarity of meaning re gardless of word class; however there may well be found within a semantic class one or more colligations of words which share both the semantic feature and a word class. There may also be collocates, specific recurrent choices of word forms carrying the semantic preference. 6.
Example The word budge in English poses a problem for dictionaries; for example: to (cause to) move a little (Longman Dictionary of Contemporary English)
Leaving aside the tortuous syntax of this sort of definition, it is easy to demonstrate that there is no feature of the meaning of budge that restricts movement. Budging is the overcoming of a resistance to movement, and so it concentrates on the beginning of movement. Even a little movement constitutes a budging, which might explain the definition above, without justifying it, because something once budged (i.e., set in motion) might move a great deal. The point is that English does not talk much about budging at all, but about not budging, where the quantity of movement is irrelevant. The two examples that follow the definition above are indeed both negative, but the entry reads as if the lexicographers had not noticed this primary fact of usage, and the user, of course, has no idea how to evaluate the presence of the negatives since they are not referred to. It would be difficult to find an instance of this word which is semantically positive. The figure (see next page) gives 31 instances from a corpus, all that there are in almost 20 million words. Of these I propose to
1.ight be out of his mind and refuse to 2.ergencies . But Mr Volcker has yet to 3.off scrubbers ' hands before it would 4. to do so , but she knew. she could not 5. he recognizes it , he '11 refuse to 6. side , but still the snake will not 7. away louder than ever . I would n't 8.now . We wo n't none of us be able to 9. blow . The virus fanciers refused to 10.sat in a corner ; I determined not to 11.hen neither death nor ? disease could 12.it with my shoulder , but it will not 13.ng the following months and would not 14.ooden door of the museum . It did n't 15. another snail near him he refused to 16.me into the dining room , refusing to 17.It was a dismissal . Bonasera did not 18..9o caliber pezzonovante . You ca n't 19.fternoons when the thermometer wo n't 2 0.be so heavy that two horses could not 21. between the duellists and refuse to 22. the coroner himself are gawn ' ' t ' ' 23.omise up to a point but he refuses to 24. . The humanity here just refuses to 25. out of the packet . When it did not 26.ed at the doorknobs the doors did n't 27. and hesitated . He knew he could n't 28.at they might , the British would not 29.pressure any delegation . They wo n't 3 0.the wings of the eagle and refused to 31. tried the idea on him . He would n't
budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge budge
. In that case , the Vice-President on changing his controls over dornest . It was rumoured to be make-work to me from my view . We spent several ν off that stool where he 's sitting η . He keeps banging it on the head wi either , or come back , till a boy w tomorrow . '' They sat at their tea . Whatever the diagnosis , my recove from it until closing-time . I also her . She wrote a cheque for more th . I go to the backdoor . I find that - " What 's done can not be undone . Hastily , I looked round for a bel , even in the mating season . I ofte , so that no one else budged , and s . Finally , sighing , a good-hearted him , not even with money . He has b above minus twenty . '' '' And those it even in moist earth . Although Wa . Often to everyone ' s great relief on that . In the firrst place , d '' on design principles he knows to be . '' '' That 's ridiculous , '' says he shook it more fiercely like ' a t or even rattle . '' Oh , my God ! ' Ben Canaan . He walked to the alcove from their immigration policy . In m from that position . '' '' What a ti . after three thousand years of wait . He seemed to have already faded aw
THE LEXICAL IT
18
JOHN SINCLAIR
ignore one, tenth from the bottom, which is clearly written to represent the dialect of a region. Most of the indications of colligation with a negative are to be found to the left of the central, or 'node' word; immediately to the left we find eight instances of words ending in η 't and eight of not - together making slightly over half the total. Most of the others show the word to in this position, and by examining the word previous to that, there is a strong collocation with forms of the lemma refuse - nine in all. Although not a grammatical negative, refuse can reasonably be considered as a lexicalisation of the kind of non-positive meaning that characterises budge. There are, then, just five remaining instances that do not follow one of the three prominent ways of expressing negativity. The eighth line has a double negative in an extended verbal group, the tenth has determined not to, the eleventh has a neither/nor construction. The two remaining instances show neither grammatical nor lexical negation; the second line expresses the refusal aspect with has yet to - implying that Mr Volcker refuses to budge, and the third line draws attention to a presumably long and unpleasant period preceding eventual budging (the extended cotext of this line is so deep with caustic dirt that skin would come off scrubbers' hands ...). The negative quality of the phrase centred around budge is thus expressed in different ways, but with a predominance of collocations refuse to (and inflections), wouldn't, didn't, couldn't. Colligation is with verbs, with modals (including able to) accounting for half the 30 instances. From this point I will not attempt to describe comprehensively the two instances above that imply rather than express negativity (lines 2 and 3). I will include them when they conform to a choice pattern, and ignore them si lently when they diverge. They are sufficiently conformant with the general usage of budge to be given low priority, so little will be gained by an exhaus tive account of their minor deviations; but also I am concerned to establish a methodology that concentrates in the first instance on recurrent events rather than on unrepeated patterns. When the habitual usages of the majority of users are thoroughly described, we will have a sound base from which to approach the singularities, which may of course include much fine writing. Budge is an ergative verb, in that whatever is to be moved may figure either as subject or object of the verb; subject in an intransitive clause, and object in a clause where the subject is the person or thing making the attempt to move. A guide to these alternatives can be found by looking at whatever immediately follows budge. Intransitive clauses may well end with the verb, so where there is a punctuation mark following budge we may expect that the
THE LEXICAL ITEM
19
clause is intransitive. Twelve times there is a full stop, twice a comma and once a dash - fifteen instances in all, or half of the total. The distinction between won't and can't draws attention to two dif ferent reasons why people or things do not budge, refusal or inability. Refusal is ascribed to whoever or whatever is not budging (won't, wouldn't ... etc., and refuse ...), while inability is usually ascribed to someone who is trying to get something or someone to budge (can be expressed by can't, couldn't). Of the first type, there are twenty instances expressing refusal or interpreted as implying it, all intransitive; where the subject is non-human and the verb modal (a snake, a quotation and a thermometer) we anthropomorphise (which is a kind of reversal). Of the second type, there are four examples; the non-budging is ascribed to inability, the 'agent' is in subject position, the clause is transitive and the person or thing that is not budging is named in the object. One instance has indications of both possible reasons; the modal cluster is won't be able to, and the clause is intransitive. This is a prediction of a future inability to budge, and won't does not indicate refusal. There remain five instances, of which four are didn't or did not. This usage is neutral with respect to refusal and inability; the structure is intransitive and so suggests that an agent is not important, but a person energetically trying to move a physical object is apparent in adjoining clauses in three of the cases. In the fourth the subject of budge is a person; the cotext makes it clear that he is under pressure to move, so it is closer to refusal than inability. To summarise this matter, we note that twenty-five of the thirty instances are intransitive, with the item that does not budge as the subject; where this item is animate, it strongly collocates with refuse, and the semantic pref erence of refusal is found in most of the other instances, mainly through col ligation with certain modals. In the transitive instances the non-budging item is object, the agent of movement is in the subject, the semantic preference is inability and there is strong colligation with the modals of ability. A minor optional element of the cotext of budge is the expression of the position from which there is to be no budging. There are eight instances, all beginning with a preposition; from four times, on twice, and above and off once each. Most are of the 'refusal' type. At this point we may argue that most of the patterning in the cotext has been accounted for, with the possible exception of the word even, which oc curs four times to the right of budge, and links semantically with yet. Some-
20
JOHN SINCLAIR
thing fairly extreme is being referred to. We consider why people use this word, why they do not just use the common verb move, with which any use of budge can be replaced. Something does not budge when it does not move despite attempts to move it. From the perspective of the person who wants something moved, this is frustrating and irritating, and these emotions may find expression, because this is the 'semantic prosody' of the use of budge. The semantic prosody of an item is the reason why it is chosen, over and above the semantic preferences that also characterise it. It is not subject to any conventions of linguistic realisation, and so is subject to enormous variation, making it difficult for a human or a computer to find it reliably. It is a subtle element of attitudinal, often pragmatic meaning and there is often no word in the language that can be used as a descriptive label for it. What is more, its role is often so clear in determining the occurrence of the item that the prosody is, paradoxically, not necessarily realised at all. But if we make a strong hypothesis we may establish a search for it that will have a greater chance of success than if we were less than certain of its crucial role. For example we can claim that in the case of the use of budge the user wishes to express or report frustration (or a similar emotion) at the refusal or inability of some obstacle to move, despite pressure being applied. Then there is an explanation for even and yet, and other scattered phrases from the immediate and slightly wider cotext of the instances. A selection of the evidence for pressure and frustration is given below, with reference to the figure. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
the President may be out of his mind should no longer ... intervene ... but... stained so deep she knew she couldn't < prediction based on experience > blows his pipe furiously especially if ordered to do so ... indignant and thunderous < prediction based on present circumstances > the polio faction ... the virus fanciers < decision based on prior events > neither death nor disease nudge it with my shoulder, but ... stuck in his mind and ... leant against the heavy wooden door even in the mating season
THE LEXICAL ITEM
(16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31)
21
naughty ... ignoring requests ... forced to ... capricious It was a dismissal. not even with money < typicalisation of experience > two horses could not ... thrust himself between the duellists and ... [not considered] he may lose a client That's ridiculous he shook it more fiercely no matter how hard he tugged he knew he couldn't do what they might, < diplomatic pressure > no amount of arguing I tried the idea on him
The above quotations and remarks are taken from a wider cotext than that printed in the figure, but still a small one. The evidence is probably enough to convince many human readers that the prosody exists and is expressed, implied or alluded to in most of the instances. But the range of expression is apparently without any limit, and the amount of inference necessary to identify it as evidence is often great; moreover in several in stances there is no actual piece of language that can be quoted, but a more general appeal is made to experience. This amorphous collection is an unlikely starter for being related to a structural category, and yet the claim is made that it is the most important category in the description. Without a very strong reason for looking, a computer would find virtually no reason for gathering this collection, but if we can predict a structural place for it, then at the very least the computer could pick out the stretch of language within which a prosody should lie, and whose absence was as significant as its presence. There is good reason to believe, also, that as the number of instances available rises, so do regu larities appear that were not reliably shown in smaller sets. The occurrences of even are already obvious, and the phrase he/she knew recurs even in this small sample; there is similarity in the phrases blow ... fiercely, shook fiercely, hard ... tugged, and money is mentioned in the wider cotext of (11) (Money would send her home when neither ...) as well as (18).
22
JOHN SINCLAIR
It is impossible to predict how much or what will be left over at the end of an extensive study of this usage, but it certainly looks unlikely to be neat and tidy; hence we need a clear and strong hypothesis about the nature and structure of the lexical item of which budge is the core, in order at least to search for indications of the semantic prosody. The core gives us the starting point, in the case of budge one that anticipates the prosody fairly clearly; the optional patterns of collocation, colligation and semantic preference bring out relevant aspects of the meaning, and the prosody can then be searched for in the close environment. It is not surprising that this is a very common structure in language, because it allows the flexibility that was identified earlier in this paper as essential for an adequate lexical item. The prosody is normally the part of an item that fits in with the previous item, and so needs to have virtually no restriction on its formal realisation, whereas the core, often in the middle or at the end of an item, is buffered against the demands of the surrounding text so that it can remain invariable. An item of this shape and structure makes it possible for the lexicon to have finite entries which are adequate to describe the way the meaning is created by the use of the item. In this lengthy description of the lexical item whose core is NEG budge I have not had reason to make a distinction that most lexicographers would regard as primary - the literal and figurative uses of the word. For example: (cause to) move very little, make the slightest movement; (fig) (cause to) change a position or attitude (Oxford Advanced Learners Dictionary)
It is easy enough to go through the examples and pick the eleven that show the figurative use, where views, opinions, policies, principles etc. Oc cur instead of doors, stools and thermometers. Where the option to express position is taken, the preposition on seems to be restricted to the figurative use, while from occurs with both. In a few instances the distinction hardly seems necessary, because both the literal and figurative aspects of the mean ing of other words are also relevant; for example the wider cotext of the first instance, about the President being out of his mind, includes the use of office and sitting; moving the President from his office simultaneously requires his physical relocation and the cancellation of his authority.18 18 W.E. Louw (personal communication) argues that 'literal' and 'figurative' are points close to the extremities of a continuum of 'delexicalisation'. Words can gradually lose their full lexical meaning, and become available for use in contexts where some of that full meaning would be inappropriate; this is the so-called figurative extension. Louw points out that a
THE LEXICAL ITEM
23
7.
Conclusion I have tried to show that there is a seriously weak point in the auto mation of language description, in the design of a lexicon. Current models do not overcome the problem of how a finite and rigidly formalised lexicon can account satisfactorily for the apparently endlessly variable meanings that arise from the combination of particular word choices in texts. I have suggested that the word is not the best starting-point for a description of meaning, because meaning arises from words in particular combinations. The term 'lexical item', used to mean a unit of description made up of words and phrases, has been dormant for some years, but is available for units with an internal structure as outlined above. Elements in the surround ing cotext of a word or phrase are incorporated in a larger structure when the pattern is strong enough. The lexical item balances syntagmatic and paradigmatic patterns, using the same descriptive categories to describe both dimensions. The identification of lexical items has to be made by linguists supported by computational resources, and in particular large general corpora. The impact of corpus evidence on linguistic description is now moving beyond the simple supply of a quantity of attested instances of language in use. It is showing that there is a large area of language patterning - more or less half of the total - that has not been properly incorporated into descriptions; this is the syntagmatic dimension, of co-ordinated lexicogrammatical choices. Because of the great range of variation in realisation, the regularities of this dimension have been overlooked, whereas from the perspective of a com puter they become both more obvious and easier to describe. Acknowledgements The instances of language in use that I quote in my work come from a number of sources, in particular The Bank of English in Birmingham and The British National Corpus. I am grateful for permission to make use of these corpora. Some of the arguments in this paper were first put forward at the Second European Seminar of TELRI in Kaunas, Lithuania in April 1997, and a very short account was included in the Proceedings (Markinkeviciene & Volz 1997) under the title "The Problem of Meaning". I profited a lot from the discussion there and in Münster, and I would also like to thank Anna Mauranen and Mike Stubbs for their helpful comments on the written version. writer, especially a literary writer, must exercise vigilance so that the meaning of each word is interpreted at the intended point on the continuum. Such features as collocation are part of the control mechanism available to the writer. So in this example the use of several other words that could be interpreted literally keeps available the physical meaning of budge, while the overall interpretation of the passage will be institutional.
24
JOHN SINCLAIR
References Bach, Emmon 1964. An Introduction to Transformational Grammars. New York: Holt, Rinehart and Winston. Channell, Joanna 1994. Vague Language. Oxford: Oxford University Press. Chomsky, Noam 1965. Aspects of the Theory of Syntax. Cambridge, Mass.: MIT Press. Cruse, D.A. 1986. Lexical Semantics. Cambridge: Cambridge University Press. Firth, J.R. 1957[1951]. "Modes of Meaning". Papers in Linguistics 1934-1951, 190-215. London: Oxford University Press. 1968[1957]. "A Synopsis of Linguistic Theory 1930-55". Selected Papers of J. R. Firth 1952-59 ed. by F. R. Palmer, 168-205. London: Longman. Halliday, M.A.K. 1961. "Categories of the Theory of Grammar". Word 17:3.241-292. Harris, Zellig S. 1957. "Co-occurrence and Transformation in Linguistic Structure". Lan guage 33:3.283-340. Katz, Jerrold J. & Jerry A. Fodor 1963. "The Structure of a Semantic Theory". Language 39:2.170-210. Markinkeviciene, Ruth & Norbert Volz, eds. 1997. Language Applications for a Multilingual Europe. Proceedings of the Second European Seminar, TELRI. Kaunas, Lithuania: Institut für deutsche Sprache, Mannheim. Monaghan, A.I.C. 1996a. "Do we Need to Resolve Ambiguities?". Monaghan 1996b. 1-16. ed. 1996b. CSNLP 1966 Dublin, Ireland. Natural Language Group, Dublin City University. Sinclair, John M. 1987a. "Collocation: A progress report". Language Topics: Essays in honour of Michael Halliday ed. by Ross Steele & Terry Threadgold, 319-331. Amsterdam & Philadelphia: John Benjamins. 1987b. "Fictional Worlds". Talking about Text ed. by R. Malcolm Coulthard, 43-60. Birmingham: ELR Monograph, University of Birmingham. (Rev. version, "Fictional Worlds Revisited". Le Trasformazioni del Narrare ed. by Erina Siciliani et al., 459482. Brindisi: Schena Editore, 1996.) ed. 1987c. Looking Up: An account of the Cobuild project in lexical computing. London: HarperCollins. 1996a. "The Search for Units of Meaning". Textus 9:1.75-106. 1996b. "Beginning the Study of Lexis". In Memoriam J. R. Firth ed. by C. E. Bazell, J. Catford, M. A. K. Halliday & R. H. Robins, 410-430. London: Longman. , S. Jones & R. Daley 1970. English Lexical Studies. Birmingham: Report to the Office of Scientific and Technical Information. & Patrick W. Hanks, eds. 1987. Collins Cobuild English Language Dictionary. London: HarperCollins. , Jonathan Payne & Chantal Pérez Hernández, eds. 1996. "Corpus to Corpus: A study of translation equivalence". International Journal of Lexicography 9:3.171-216. Stageberg, Norman . 1966. An Introductory English Grammar. New York: Holt, Rinehart & Winston. Tognini-Bonelli, Elena 1995. "Italian Corpus Linguistics: Practice and Theory". Textus 8:2.391-412.
CONTRASTIVE LEXICAL SEMANTICS
EDDA WEIGAND University of Münster
1.
The problem of verification in semantics There have been some important developments in linguistics in the last few decades which we cannot ignore when we nowadays address lexical se mantics. I will briefly mention two major changes of view. First, we have left structural linguistics with its autonomous areas. Accepting the pragmatic turn we have to face the complex fact that our subject matter, language, turns out to be language use in dialogic interaction. Lexical semantics represents only one component in the new communicative unit of the action game (cf. Weigand 1997b). Second, our reference point can no longer be the compe tence of an ideal speaker. There are different human beings acting and react ing in the action game whose communicative competence cannot be assumed to be the same (cf. Weigand 1998b). Under such changed conditions we have to pose once again the question which is essential for every discipline, the question of how our description can be verified. The different reference points used for verification in lexical research enable us to distinguish differ ent types of approaches. First, there is the basic opposition between the view of language use as completely based on conventions and the non-conventionalist view of mean ing, which maintains that lexical meanings are individually established, not guided by conventions (cf. Davidson 1985). I will not go into details but only mention an essential point that leads me to the conventionalist position in general. If we consider language as dialogic interaction in which not only the speaker is engaged, there must be, in my opinion, conventions of cooperation which are the basis for the dialogic process of coming to an understanding. Everything depends, however, on how we understand the term 'convention'.
26
EDDA WEIGAND
I would like to make a distinction between rules and conventions in so far as rules can be seen as independent of the user, for instance, mathematical rules or rules of generative grammar. By contrast, 'conventions', following etymology, are in principle dependent on the user or on groups of users. In the final analysis, conventions are not definite, they may always be overridden by specific circumstances. Rules of morphology, for instance, are fixed, defined, to a large extent user-independent. In this fixed sense, too, the term convention was used in structural linguistics for the relation between the signified and the signifier of the sign. Such a concept of sign, however, does not lead us very far in pragmatics. Conventions as an orientation for a group of users are mainly based on choice as Lewis (1969) has told us and as we have seen in various ways by studying dialogic action. Conventions are either valid for a set of cases or for specific single routines. For instance, you may use the adjective thick in a variety of cases - thick wall, thick slice, thick carpet, etc. - but you have to know the single case of heavy drops/dicke Tropfen (for a complete analysis cf. Weigand 1996a: 161-162). In this sense of restricting free choice, the term convention comes near to the term rule. It is, however, different in so far as a rule always covers a set of cases. The next question we have to tackle is the question of whether language is totally governed by rules and conventions as is assumed in structural and generative linguistics, or whether we have to go beyond rules and even beyond conventions. As long as we were dealing with language as an abstract system, our subject matter was by definition governed by rules. Also some pragmaticians, who deal with language in use, keep strictly to rules and conventions. I remember Searle's (1969) speech act theory and the model of dialogue grammar to which I referred in my book "Sprache als Dialog" (Weigand 1989). Recent research, however, has taught me to include yet another aspect. On the one hand, there are approaches in the literature, for instance, by Dascal (1994), that stress the point that we have to take account of presump tions and suggestions, as well as conventions, in order to come to terms with dialogic interaction. On the other hand, research on misunderstanding and emotion confirms that language use cannot be described along the lines of mathematical-logical models (Weigand 1996b). We have to accept two things if we do not want to be restricted by artificial models: first, there are con cepts guiding dialogic interaction that are not strictly conventional concepts but are valid ad hoc or based on probability, such as preferences, habits, emotions, and individual presumptions about our memory. All these points
CONTRASTIVE LEXICAL SEMANTICS
27
may lead to misunderstanding (Weigand 1998b). Second, as a consequence, our model can no longer be a doubled speaker model. We have to account for the fact that it is different human beings, with different cognitive back grounds, who act and react in the action game. Understanding therefore can not be presupposed, it is negotiated in dialogic interaction. Consequently, language use can no longer be described by a closed system of rules and conventions. How can such a complex subject matter be approached? We are in need of a methodological technique that allows us to come to terms with cases of probability. Such a technique, in my opinion, can be seen in the concept of principles used by the interlocutors. As interlocutors we guide ourselves by principles in social interaction. Principles give us a line of orientation and help us to come to terms with the multiple and sometimes indefinite aspects of the action game. On the basis of the Action Principle and the Dialogic Principle there are various other principles such as the Sequencing Principles, the Principle of Suggestion, the Rationality Principle, etc. which can all be seen as structuring the action game (cf. Weigand 1998a). The model of dialogic action games therefore will be an open model that contains not only the definite and quasi-definite of rules and conventions but also the indefinite and cases of probability on which in the end dialogic action is based. Now we have to ask for the methodological technique that is relevant for lexical semantics as a part of language use. Speaking of language use does not necessarily mean that we refer to performance or spoken language. We refer to the level of communicative competence that guides language use, be it in the written or spoken medium. We therefore do not attempt to discover cases of individual, occasional meanings of words in spoken lan guage that are not repeatable, but primarily try to find the conventions that determine how words are used. There remains the problem of how we can discover these conventions. More concretely, this is the question of how we can verify our semantic analyses. In the past, we mainly relied on dictionaries and on questioning native speakers. They remain valuable sources, but alone they are not satisfying. We can no longer be content with the position that leaves semantics to introspection and intuition. Results in semantics should be controllable, reliable, repeatable. I share Sinclair's conviction that we have to find "hard, measurable evidence" according to the criterion of frequency (Sinclair 1987:XV). Hard, measurable evidence can be gained from large text corpora by the use of advanced computer technology. The corpus itself,
28
EDDA WEIGAND
however, gives us lists of uses, nothing else. These lists still have to be analysed within the framework of a model of language use. In a strict sense, there is no empirical evidence as such. We have to pose the questions that are indicated by our model. That means that we have to analyse the use of words according to our model and to check our results by the corpus. Therefore, besides dictionaries, native speakers and corpora, the ana lysing linguist comes in. As linguists we consult dictionaries and ask native speakers when analysing foreign languages according to our model. In the end, we have to check our assumptions about semantic conventions against representative text corpora. Analysing vocabulary in lexical semantics on the basis of text corpora is different from describing units of the lexicon in a generative model. It is mainly the criteria of the model that verify the de scription of the lexicon. 2.
Semantics in a pragmatic model Concentrating on a model for interlingual lexical analyses, we have first to clarify how to deal with lexical semantics in general, second we have to establish a level of comparison for contrastive-interlingual analyses. After the pragmatic turn, lexical semantics in general has lost its autonomous status. It has become part of language use and is integrated into the functional structure of the utterance which can be represented by Searle's formula F(p) and be seen as the essential basis for a theory of language use. Within a pragmatic model meaning in general will be a pragmatic concept, namely the purpose for which expressions are used by the interlocutors. We know the purpose for which utterances are used and we have called it action function. The action function represents the superordinate function F which dominates the propositional function p. We may distinguish semantics proper or propositional semantics as part of the utterance being dependent on the action function. As we already know from Aristotle, propositional semantics can be divided into reference and predication. We therefore have at the very beginning three fundamental types of meaning: the action function, the referential function and the predicative function. The next problem we have to tackle is a difficult one. We have to find out how the linguistic means are correlated to these fundamental types of meaning. Up to now we have mostly defined lexical semantics by referring to specific types of expression, namely expressions of open sets, leaving aside the question whether there is a specific type of meaning for these expres sions. As I have explained elsewhere (cf. Weigand 1993), it is the predicative
CONTRASTIVE LEXICAL SEMANTICS
29
function that is in general expressed by lexical items. Even in the case of speech act verbs, for instance, in utterances like I predict that ... it is the grammatical structure of the explicit performative utterance which constitutes action function and it is the lexical item of the speech act verb predict that predicates what type of action we are dealing with. Thus we have arrived at a level where we can analyse vocabulary as an area which is to some degree independent of the other parameters of the action game. It remains to be clarified what 'predicate' means. In a model of dialogic action games human beings or their language cannot be differentiated from the world; on the contrary, human beings are seen as part of the world. The world can only be experienced by means of human abilities. By means of lexical expressions, the words-in-use, human beings predicate how they see the world to which they belong. The predicative function is, in accordance with the formula F(p), subordinate to the action function. By integrating lexical semantics functionally into a comprehensive pragmatic theory in this way, we get round the problem caused by the structuralist assumption of autonomous levels. Structuralists and pragmaticians agree in the assumption that it is lan guage use which is the genuine subject matter for linguists. There is consen sus also in the next step in so far as both agree that we have to abstract from 'parole', from the feature 'actual, not repeatable' in order to arrive at a mod el for language use. Abstraction however cannot be arbitrary as is the case when structuralists jump from 'parole' to sign system and thus totally lose their initial subject matter. Abstraction has to follow criteria. Even intending to abstract from the feature 'not repeatable', we nevertheless intend to keep to language use. Thus we arrive at the underlying level of our communicative competence. Language use does not mean that there are autonomous texts which are sent like packages or complex signs from speaker to hearer as is represented in simple communication models. Language use means use of language by human beings who try to achieve their communicative purposes with commu nicative means. Communicative means are not only linguistic means but are based on all human abilities together. How could we assume that we use our ability to speak separately from our ability to think and add a third ability to see what goes on in the speech situation? We are dialogically acting and reacting in the action game (Weigand 1989, 1991, 1995). The parameters that all play a role in the action game are correlated in such a complex way
30
EDDA WEIGAND
that it is absurd to assume that we could come to terms with the notion of sign. Having assumed three fundamental types of meaning, we are now faced with the problem of structuring in more detail the components of each type. Concerning action function and reference function, I will restrict myself to indicating in a figure (from Weigand 1996a: 160) how they can be differen tiated and will concentrate on the problem of structuring predicative function. communicative means (linguistic, visual, cognitive)
MEANING EQUIVALENCE OF EXPRESSIONS
Fig. 1: Model of the correlation between meaning and expression side of the utterance
For our pragmatic notion of meaning as the purpose for which expressions are used, it is interesting to see that the correlation between meaning and expression side is not guided by a fixed relation of correlating expression and meaning within the sign but by a principle of meaning equivalence used by the interlocutors. It remains to be clarified what the units on the expression side are (cf. Weigand 1993). Though the two-level-model of Bierwisch, Lang, Wunderlich and others (cf. for instance Bierwisch 1983, Lang 1989, and Wunderlich 1996) seems quite different from the model above, I am inclined to see parallels in fundamental points: In our model we also have the semantic level, repre sented by meaning positions, and we have the conceptual level, represented in the cognitive base.
CONTRASTIVE LEXICAL SEMANTICS
3.
31
The universal level of contrastive studies Having established our model of lexical semantics as a pragmatic model, we must now come to terms with its second feature: that it is a contrastive-interlingual model. Starting from an individual language and analysing it semantically results in a semantics of that language. Thus we have meaning dictionaries of the English language, of the French language and so on. If we correlate two languages we will get two different diction aries depending on the language which is taken as starting point, one diction ary English-French and another French-English. If we intend to correlate expressions of different languages we have to assume some sort of equivalence or interlingual synonymy between these expressions. The concept of equivalence cannot be simply stated by intuition or native competence, it must be defined by semantic structure, which is nothing other than making competence transparent and explicit. If we deal with expressions from different languages, semantic structure must be a structure which is valid for more than a single language. We take this quasiuniversal structure as the common basis for our contrastive analyses with reference to which we can decide what counts as equivalent expressions. In accordance with our position that there is no independent reality, only reality in the eye of the observer, the universal structure contains the ways by which human beings perceive the world, not the structure of the world itself. Structuring the world itself would be an impossible undertaking because there is no definite criterion, independent of human beings, according to which a specific structure could be laid down. The set of hypo thetical criteria and structures would be infinite. Relating structure to human beings and their abilities gives us criteria for distinguishing different areas. In so far as the ways by which human beings perceive and act in the world depend not only on general human conditions but on special social conditions as well, it might indeed be the case that, starting from a universal level, meaning positions, for instance, of social behaviour have to be introduced that are valid for specific cultures only. In this sense, our way of life deter mines our way of using language and this is equivalent with and not at all opposite to the reversed relation of having a language from which one might derive insights into a way of life. In so far as universal structure is interesting for us not on its own but as a level on which different languages can be compared, its units have to be considered as heuristic units without any ontological status. Just as there is no independent reality, neither there is ontology. The fact that we consider
32
EDDA WEIGAND
universal structure to be a heuristic structure and not an independent notional or ontological structure constitutes a fundamental difference to onomasiological models such as that by Wierzbicka (1972). She takes semantic primitives as given and is of the opinion that they could be defined as inde pendent concepts. Such concepts are artificial ones and the thesis of their existence remains vacuous. In contrast to Wierzbicka, we introduce universal semantic units by referring to human abilities of perceiving and acting in the world and by looking at the ways in which we express our doing in different languages. Universal structure can thus be considered as the cognitive heuris tic basis of our model of language use. As a consequence, the problems we have to solve refer on the one hand to the structure and units of universal semantics, and on the other to the ways-of-use in different languages: ways-of-use in language 1 universal s t r u c t u r e w a y s - o f - u s e in language 2 etc. Fig. 2: Contrastive grammar of predicating
4.
Structuring the lexical-predicative part In our model of dialogic action games, everything has to be related to human beings and their cognitive and physical abilities. The structural attempt to establish and structure lexical fields independently was not able to achieve its purpose because a reference point for comprehending and dividing the whole was missing. Tracing the complexity of the world back to its centre, human beings, makes it possible for us to structure the whole picture. We assume as a working hypothesis and a starting point the following predicatingfields: universal structure based on human abilities
AWARENESS
MOTION
ACTION
RATIONALITY
EMOTION
SOCIAL BEHAVIOUR
Fig. 3: Universal predicating fields
These fundamental fields are intended to comprehend all the types of human abilities which are the basis for predicating. We therefore consider them to be the cognitive basis of our theory of lexical semantics. They might be
CONTRASTIVE LEXICAL SEMANTICS
33
compared with the conceptual frames used by Schwarze & Schepping (1995) in a two-level semantics. Fundamental fields are divided into partial fields and these are structured according to predicating positions which are considered to be the minimal units of meaning or the meaning positions in the lexical area. Thus, for instance, the fundamental field of AWARENESS is divided into the partial fields of the five senses: awareness by vision, by hearing, smelling, tasting and touching, to which one might add the field of consciousness. Similarly, ACTION comprises the whole field of intentional action which might be practical, linguistic, physical action as well as mental, visual and other intentional activity. EMOTION represents another fundamental field which includes as partial fields feelings, moods and urges. Having arrived at partial fields of fundamental abilities, we have to come to terms with the problem of giving structure to these partial fields. At this level, we have to introduce predicating positions as minimal heuristic units of lexical meaning. In characterizing them as heuristic units, we dissociate ourselves - as already mentioned - from approaches which are based on independent semantic primitives. In contrast, we try to establish universal structures by starting from human abilities and relating to them what seems to us essential semantic concepts and configurations. In the end these semantic essentials are built up from heuristic predicating positions, for instance, the positions of BE, BECOME versus LOSE, and CAUSE, which we need for structuring vocabulary. These predicating positions might in part be established by theoretically reflecting on possible semantic configurations, in part however they are empirically established not only by differentiating semantically between the expressions of a single language but also by comparing the vocabulary of different languages. Here we have to address the problem of how far mean ing positions should be differentiated. In this respect it is once again impor tant to observe the general principle that the universal semantic level is in tended as heuristic level, not as representation of the whole meaning, intend ed only to give structure to the vocabulary as far as distinctions are needed. The remainder of meaning is left to the interlingual relation of equivalence or quasi-synonymy between expressions of different languages. Predicating positions correspond roughly to 'semes' or 'atomic predi cates' in structural and generative linguistics. In contrast to 'semes' and 'atomic predicates', which can be considered as units of semantic decompo-
34
EDDA WEIGAND
sition starting from signs, predicating positions represent units of semantic composition starting from human abilities. Besides analysing the meaning of signs into minor units by semantic decomposition, there is another way of indicating meaning, namely by means of paraphrases or synonyms which we find in monolingual dictionaries such as the Cobuild dictionary, even if it stresses the notion of use. Both ways are intended to explain meaning. Both start from expressions of individual languages, which is not the best starting point for contrastive semantics. Having established meaning positions on a universal semantic level, we no longer need paraphrases or synonyms as meaning classifiers. We include them on the expression side as ways-of-use. 5.
Units on the expression side Having introduced meaning positions as units on the semantic level, we now have to tackle the problem of what the units on the expression side are. We admit that notional systems can be constructed from single words, but this is not our object. Our object is to find the minimal lexical units that function on the level of utterance, i.e. that are used for predicating. In language use, we do not take single signs and insert them in syntactic positions of the sentence according to rules of compositionality. On the level of utterance, we find syntactically defined phrases, i.e. words-in-use or waysof-use, nothing else. These syntactically defined phrases are the lexical units of our communicative competence. They are stored in our memory as com plex units, as we can check when examining possible combinations of words. We are ill at ease when confronted with ways-of-use that are not wellestablished and we feel relieved when we arrive at a conventional phrase stored as such in our memory. Therefore, in our pragmatic model, the units on the expression side are not single words but words-in-use or ways-of-use. These ways-of-use have a different syntactic status. They may be NPs, PPs, VPs, Adverbs or even whole utterances. Thus, for instance, in order to describe the semantics of the adjective thick/dick, we have to consider the NP as a unit of expression (Weigand 1996a, 1997a). One might think of considering only the reduced NP, the combination of Adj + Ν, as lexical unit. The Determiner in some cases does not seem to play a role on the lexical level, in general however it cannot be neglected as we will see in our analyses in the article "The Vocabulary of Emotion" (below). There are NPs whose components seem freely combinable. The NPs therefore have the status of examples: thick wall, thick slice, thick carpet/dicke Wand, dicke
CONTRASTIVE LEXICAL SEMANTICS
35
Scheibe, dicker Teppich, etc. Then we are confronted with NPs that are more or less fixed, such as (1)
thick forest thick crowd heavy drops heavy traffic swollen ankle swollen cheek
dichter Wald dichte Menge dicke Tropfen dicker/dichter Verkehr dicker Knöchel dicke Backe
The situation becomes even more complex if we include the corresponding NPs in French or Italian: (2)
forêt épaisse foule serrée gouttes lourdes circulation dense genou enflé bouche enflée
bosco fitto grande/molta folla goccie grosse traffico pesante caviglia gonfia guancia gonfia
In this way the whole vocabulary of natural languages has to be considered as a network of ways-of-use which to some degree are freely combinable and to some degree are more or less fixed. The "principle of free choice", as Sinclair (1991:109-110) has called it, comes to an end and we cannot indicate the reason. We simply have to list the idiomatic ways-of-use individually and completely. These idiomatic ways-of-use in the individual languages, for instance, thick forest, dichter Wald, forêt épaisse, bosco fitto can be found more easily by including a contrastive perspective. In this sense, essential features of our mother language become evident by comparing it with a foreign language. Like Sinclair (in this volume) we take as lexical units 'multi-word' expressions. Whereas Sinclair tries to gain these complex lexical units automatically and formally from the corpus considering them as words in the context of other words which occur before and after them in the corpus, we define them functionally and syntactically as ways-of-use or phrases. These ways-of-use are the lexical units that function on the level of the utterance in so far as they are the units that have predicating function. On the level of syntax they are the units from which the utterance is built up. The category of parts of speech represents a distributionally defined sub-part of ways-ofuse, no longer the main syntactic reference point.
36
EDDA WEIGAND
Distinguishing between phrases as the units that have predicating function and parts of speech as distributional sub-parts of phrases, we can now address the problem of the role of parts of speech which includes the problem of whether nouns are a specific part of speech that should be represented as a specific concept in the universal structure. We started in universal structure from human abilities and did not represent objects. Human abilities refer to objects which might be concrete visual objects in the case of visual perception or abstract objects in the case of rationality and reason. In this way we can include objects, too, by attaching them to the respective abilities. The fundamental concepts 'space' and 'time' could also be assigned to human abilities such as 'awareness', 'motion', and 'rationality', as is hinted at by Klein (1993), when he distinguishes between 'normal perceptual space' on the one hand and 'various types of space', which are to be thought of as more abstract, on the other (cf. also Wunderlich 1985). Reflecting on the different expression types of vocabulary, we have, however, to take account of the difference between adjectives and nouns (cf. also Weigand 1978). Adjectives characterize, predicate, and do nothing else, but nouns predicate and additionally are able to refer, i.e. more precisely, they are able to be used as the core of NPs or they are able to become objects of referring. Reference, however, is not the function of nouns, it is the function of NPs within the whole utterance, i.e. of referential ways-of-use. In so far as objects of referring, for instance, child, danger, eyewitness, awareness, etc. can be traced back to VPs, to be a child/to be childlike, to be a danger/to be dangerous, etc., we are once again justified in not representing them in universal structure, at least not in the initial stages (cf. Sasse 1992). Now one might object that one could nevertheless start with single words and indicate precisely their collocations. This technique results in a list of collocations, ways-of-use, which are subsumed under a single word. Even if we can thus achieve a valuable alphabetically ordered index of the words analysed, in theoretical respects such a technique has to be judged as problematic. First, this way of representation is theoretically deceptive in so far as it obscures the fact that it is not the single word that contains all the meanings listed below. These meanings are a result of the way-of-use in which they occur. They are no longer readings of a polysemous sign but the meaning of a complex lexical unit which normally is no longer polysemous. As native speakers we are not aware of the multiple readings that might be
CONTRASTIVE LEXICAL SEMANTICS
37
assigned to single words. We know their meaning when referring to ways-ofuse. Second, it is impossible to indicate all paraphrases and synonyms, either mono- or multilingually. There is no way from thick to heavy if you refer only to single words. Focusing on ways-of-use you will discover that thick forest and heavy traffic are synonyms with regard to the meaning position SHAPE/DENSE. The same is valid for different languages. If you contrast thick, dick, épais as single words, you will not get to know that thick crowd is equivalent to dichte Menge and equivalent to foule serrée. Again, the reference point has to be the way-of-use. Universal structure is correlated to ways-of-use in different languages. Because of the syntactic definition of ways-of-use, syntax finds its place in the description of language use. It becomes possible to start from universal semantics and to construct utterances of individual languages out of ways-ofuse. The model can be called a cognitively based theory of use which relates cognition to use. 6.
The principle of meaning equivalence As we know, formal and semantic structure are not directly correlated; each of them has its own principles that do not fit 1:1. Starting from ways-ofuse, we have to look for a principle of correlation between universal struc ture and ways-of-use that allows us to give semantics and the expression side their own respective structures. As is the case with the action function and the referential function, in the predicative field, too, it is the principle of meaning equivalence of expressions that establishes correspondences between configurations of meaning positions and units of expression considered as ways-of-use. For instance, related to the semantic structure of SHAPE/DENSE we have, among others, the following ways-of-use as equivalent or synony mous ways-of-use in the respective languages: MEANING POSITIONS
↔
{ways-of-use}
Synonymy SHAPE/DENSE
thick forest, heavy traffic dichter Wald, dicker Verkehr forêt épaisse, circulation dense bosco fitto, traffico pesante
Fig. 4: Principle of meaning equivalence
Normally not only one way-of-use will be correlated to a specific meaning structure but a few or several ways-of-use are considered equivalent, such as
38
EDDA WEIGAND
thick and heavy in thick forest, heavy traffic or dick and dicht in dicker Verkehr, dichter Wald. We have already stressed that universal structure in our model represents only a heuristic structure. Thus we are not obliged to differentiate meaning to the finest degree. Consequently, we assume a relatively rough notion of equivalence. Then, however, the set of expressions that is cor related to these meaning positions contains expressions which are not strictly, not absolutely equivalent, as can be seen more clearly with ways-of-use that are more complex than our NPs with thick/dick. In the predicative field of emotions, for instance, we use some configurations of meaning positions which remain rather global, such as EMOTION/ANGER + BE. Naturally, there will be many different ways-of-use in the individual languages that are not equivalent in a strict sense but are grouped together within the predicative field of ANGER + BE, for instance, to be annoyed/ angry/cross, to feel annoyed, to have trouble or ärgerlich/verärgert sein, sich ärgern, Ärger empfinden. The predicating fields and positions on the universal level represent a heuristic structure for correlating expressions of different lan guages in the sense of quasi-paraphrases or quasi-synonyms. The problem of further differentiation can also be considered a problem of the level we are addressing. In so far as completeness in lexical research remains an illusion, it seems reasonable for contrastive studies to leave aside the level of the ideal native speaker and to address the level of the advanced learner. The concept of the advanced learner has to be considered not so much as a didactic concept than as a safeguard against unrealistic theoretical demands. 7.
Signs and rules at the crossroad In our pragmatic contrastive model we have abandoned the single sign and taken as our lexical unit a complex multi-word-expression which is according to Sinclair (1991:109-110) - built up in part by free choice and in part by idiomatic choice. Thus we have also left the rule-governed method ology of a model of language as sign system. The point at which the pragmatic approach differs from the model-oriented, rule-governed approach is in the view that the scope of the rules cannot be determined. There are rules, or possibilities for freely combining individual words, for example, ein Spiel vorbereiten/prepare a game, einen Vortrag vorbereiten/prepare a lecture, ein Treffen vorbereiten/prepare a meeting, but the scope of the rules is arbitrary from language to language. Thus one can, in Italian, preparare la
CONTRASTIVE LEXICAL SEMANTICS
39
valigia or preparare il tavolo, while we, according to German and English conventions den Koffer packen/pack the suitcase und den Tisch decken/lay the table. Changing the term rule to convention indicates that we are dealing with 'rules' which are to some degree dependent on the user and may refer only to single cases, i.e., we are dealing with routines or conventions. In recent years the problem of lexical rules has been repeatedly addressed under the term of 'collocation'. Mel'cuk & Wanner (1994:91), for instance, concede: "idiosyncrasies in collocations abound and have to be simply listed". These idiosyncrasies run through the whole use of language like a subtle but firm network. Competent speakers have to know these subtle differences in the use of different languages which are - being single cases based on conventions not on rules, such as: (3)
ein erstes Argument dagegen einen Traum haben einen Lauf machen wie überall
one main argument against faire une rêve fare la corsa come da altre parti
Bierwisch (1967) and others have already pointed out the difficulties with which the sign theory approach is confronted. However, we should not allow ourselves to be held back by the resigned remark made by Mel'cuk & Wanner (1994:120) that:. "After all, language is notoriously capricious and inconsistent." Rather, we should accept the complexity of natural language as a challenge and try to find a new adequate methodology for it. For a long time - with a grain of salt in lexical field theory and in structural semantics, too - the individual word has been accepted unquestioningly as the unit of lexical description. In connection with the question: what is the meaning of a word? it was, however, quickly recognised that it was not enough to analyse the individual word. Nowadays we generally start from the assumption that the word and its meaning must be investigated in its context. This formulation, word and context, however, still emphasizes the individual word as a unit which is to be investigated in changing contexts. It also emphasizes traditional model-oriented methodology, according to which the relation between word and context has to be described in a rule-governed way, even though well-known linguists who dealt primarily with the matter under investigation rather than with models, have again and again pointed out how 'capricious' natural languages are (e.g., Wandruszka 1969). A pragmatic methodology also follows the principle that the methodology
40
EDDA WEÏGAND
should be derived from the object under investigation, and not the other way round. Such an object-oriented approach should once again tackle the problem of the lexical unit and question the dogma of absolute ruleconformity (cf. Sinclair 1996). If we nowadays look at the 'classics' of the literature on semantics written in the 1960s and 1970s, we realise that the problem was recognised from the beginning, but was, at first, pushed aside by the scientific trends of the time. Thus structuralism attempted, from the beginning, to demonstrate that the meaning of a word has to be analysed not only in its paradigmatic, but also in its syntagmatic relations. In their early works Bierwisch and Wunderlich pointed out the problem and hoped that it might have been solved by a deeper and more differentiated analysis in the future. Thus as early as 1967 in his seminal essay "Some Semantic Universals of German Adjectivals" Bierwisch points to a number of examples which resist analysis and expresses the hope that he will be able to include them in future work. Wunderlich's work emphasizes again and again the role of context, and especially so in his workbook on semantics (1980), and this should be taken into account when dealing with the description of the meaning of a word. The problem of context is also in the forefront of interest in later works on the 'generative lexicon', which build upon Bierwisch's two-level semantics (e.g., 1983) and upon Pustejovsky's qualia model (e.g., 1991). With Pustejovsky, Bergler & Anick (1993) it becomes the central question. They want to analyse the relations between words in a text beyond simple co-occurrence relations in a more differentiated way within the framework of a semantic theory, and in doing so go beyond purely statistical models. The qualia structure, in which the different aspects of word meaning are specified, gives an insight into the combinability of lexical units in texts. Over the long term methods are to be developed which will enable us to determine the relations between words automatically from on-line corpora. This is also the aim of Sinclair's (1996) corpus-based analyses. What is, however, decisive for Pustejovsky, Bergler and Anick is the attempt, within the framework of a formalised, completely rule-governed model, to deter mine the relations of word meaning which go beyond the individual word. Mel'cuk & Wanner (1994) also go along the route of investigating the 'word in context' by attempting to describe the rules governing the relations between 40 selected nouns and 20 selected verbs in a relatively comprehen sive study of the nouns of emotion in German. The theory that forms the basis of this study is the 'meaning-text theory', which has as its starting point
CONTRASTIVE LEXICAL SEMANTICS
41
the same premise as Pustejovsky's semantic theory: that the meaning of words can be described in a completely rule-governed way and that units of more than one word can be derived from the meaning of individual words and the lexical functions between these individual words. But in doing so Mel'cuk & Wanner are clearly aware that they can only describe a small section of the lexis using this rule-governed procedure. While Pustejovsky confidently places his trust in even more intensive research within the framework of his model, Mel'cuk & Wanner finally recognise with a certain resignation the limits of their approach: that the 'capricious' nature of natural language cannot be grasped using a rule-governed model. We have thus reached the point at which we have to analyse our own research aims and interests: do we want to continue using an already existing rule-governed model of individual words and their connections and thereby neglect essential characteristics of natural languages, or do we want to gain insights into the object of our research, language in use, and then as a result, be prepared to look for a new adequate methodology. While Pustejovsky, Mel'cuk & Warmer and others attempt to derive larger multi-sign units in a rule-governed way from the meaning of individual signs, the pragmatic approach is serious in its view that larger multi-word units cannot generally be derived from individual words in a rule-governed way, if the semantic theory is to do justice to its claim to describe the whole lexis. A rulegoverned model of the sign system cannot, on principle, cope with com plexities of the whole lexis. Mel'cuk & Wanner analyse a relatively large part but it necessarily remains only a part. They are fully aware of this (Mel'cuk & Wanner 1994:90): "Obviously, this is merely a partial study; an exhaustive analysis would require the coverage of all emotion noun lexemes and all their collocate verbs, but such an effort goes far beyond the scope of our work. " In contrast to this, a pragmatic approach is quite clearly directed to lexis-in-use as the empirical basis and as the object of its studies. We are thus, for the first time, confronted with the problem of coming to grips with the wealth of the lexis both theoretically and lexicographically, and of deriving from it conclusive insights for a new conception of dictionaries and for foreign language learning. In doing so we will have to dispense with some highly-regarded methodological principles such as the ideal of com pleteness and explicitness of semantic description. Instead, as for example Schmitt (1991, 1995, 1997 and in this volume) emphasizes, we will even
42
EDDA WEIGAND
have to go beyond the complex multi-word lexical unit in some cases and investigate lexical meaning in larger textual contexts. 8.
Concluding remarks Having thus put our contrastive model in the context of the history of lexical research, we will finally look once again at the whole field of the lexicon and lexical semantics. In recent years, much progress has been made in lexical semantics of individual languages by addressing the whole vocabulary on a large corpus base. We have thus achieved new and better dictionaries, especially the Cobuild Dictionary of the English Language. Much progress has also been achieved in the field of studying the lexicon in specific models such as that of generative grammar. In contrast to these studies, however, the field of contrastive pragmatic analyses of vocabulary has not yet been settled. This can be seen quite clearly in bilingual diction aries that are by no means adapted to language use. We are very much at the beginning and there is still a great deal to be done. In the long run the aim could be to contrast the whole vocabulary of the main languages on a pragmatic basis and thus to achieve new pragmatic contrastive dictionaries that will help the next generation to communicate in different languages and to acquire multi-lingual action competence. Pure research should be combined with application in so far as, primarily, fre quent and essential uses have to be compiled. Leaving it open whether it will be possible to achieve such a far reaching and difficult goal, it will certainly be possible to begin in the short run with one predicating field, for instance, the field of emotion in the main European languages. If we concentrate on one field, experience in dealing with problems can be gained. For the beginning, it is most important to be prepared for the fact that complications and problems will never end when we are dealing with the whole vocabulary of a language. We will be prepared if we are no longer frightened by complexity but consider it to be the most fascinating aspect of our subject matter. References Bierwisch, Manfred 1967. "Some Semantic Universals of German Adjectivals". Foundations of Language 3.1-36. 1983. "Semantische und konzeptionelle Repräsentation lexikalischer Einheiten". Untersuchungen zur Semantik ed. by Rudolph R???žička & Wolfgang Motsch, 69-99. ( = Studia grammatica, 22.) Berlin: Akademie-Verlag.
CONTRASTIVE LEXICAL SEMANTICS
43
Dascal, Marcelo 1994. "Speech Act Theory and Gricean Pragmatics: Some differences of detail that make a difference". Foundations of Speech Act Theory: Philosophical and linguistic perspectives ed. by Savas L. Tsohatzidis, 323-334. London & New York: Routledge. Davidson, Donald 1985. "Communication and Convention". Dialogue: An interdisciplinary approach ed. by Marcelo Dascal, 11-25. (= Pragmatics & Beyond. Companion Series, 1.) Amsterdam & Philadelphia: John Benjamins. Hoinkes, Ulrich & Wolf Dietrich, eds. 1997. Kaleidoskop der Lexikalischen Semantik. { — Tübinger Beiträge zur Linguistik, 428.) Tübingen: Narr. Klein, Wolfgang 1993. "Some Notorious Pitfalls in the Analysis of Spatial Expressions". Theorie und Praxis des Lexikons, ed. by Frank Beckmann & Gerhard Heyer, 191-204. Berlin, New York: de Gruyter. Lang, Ewald 1989. "Semantik der Dimensionsauszeichnung räumlicher Objekte". Dimen sional Adjectives. Grammatical structure and conceptual interpretation ed. by Manfred Bierwisch & Ewald Lang, 263-417. Berlin, Heidelberg, New York: Springer. Lewis, David K. 1969. Convention: A philosophical study. Cambridge/Mass.: Harvard University Press. Mel'čuk, Igor A. & Leo Wanner 1994. "Lexical Co-occurrence and Lexical Inheritance. Emotion lexemes in German: a lexicographic case study". Lexikos 4.88-160 (AfrilexReeks/Series, 4.) Stellenbosch/Republik Südafrika. Pustejovsky, James 1991. "The Generative Lexicon". Computational Linguistics 17.409-441. , Sabine Bergler & Peter Anick 1993. "Lexical Semantic Techniques for Corpus Analysis". Computational Linguistics 19.331-358. Sasse, Hans-Jürgen 1992. "Das Nomen - eine universale Kategorie?" Arbeiten des Sonder forschungsbereichs Theorie des Lexikons, 282. Schmitt, Christian 1991. "Kontrastive Linguistik als Grundlage der Übersetzungswissen schaft. Prolegomena zu einer Üersetzungsgrammatik für das Sprachenpaar Deutsch/ Französisch". Zeitschrift für französische Sprache und Literatur 101.227-241. 1995. "Distanz und Nähe romanischer Sprachen: Zum Beitrag des Übersetzungsver gleichs, dargestellt an den Sprachenpaaren Deutsch-Französisch/Spanisch". Konvergenz und Divergenz in den romanischen Sprachen. Romanistisches Kolloquium VIII ed. by Wolfgang Dahmen et al., 349-380. (= Tübinger Beiträge zur Linguistik, 396.) Tübin gen: Narr. 1997. "Bedeutung, Pragmatik und Lexikographie". Hoinkes & Dietrich 1997. 115-128. Schwarze, Christoph & Marie-Theres Schepping 1995. "Polysemy in a Two-Level-Seman tics". Lexical Knowledge in the Organization of Language ed. by Urs Egli, Peter E. Pause, Christoph Schwarze, Arnim von Stechow & Götz Wienold, 283-300. (= Current Issues in Linguistic Theory, 114.) Amsterdam & Philadelphia: Benjamins. Searle, John 1969. Speech Acts. An essay in the philosophy of language. Cambridge: At the University Press. Sinclair, John 1987. "Introduction" to: Collins COBUILD English Language Dictionary ed. by John Sinclair, xv-xxI. London, Glasgow & Stuttgart: Collins & Klett. 1991. Corpus, Concordance, Collocation. ( = Describing English Language). Oxford: Oxford University Press. 1996. "The Search for Units of Meaning". Textus IX.75-106. Wandruszka, Mario 1969. Sprachen, vergleichbar und unvergleichlich. München: Piper.
44
EDDA WEIGAND
Weigand, Edda 1978. Die Zuordnung von Ausdruck und Inhalt bei den grammatischen Kategorien des Deutschen. (= Linguistische Arbeiten, 58.) Tübingen: Niemeyer. 1989. Sprache als Dialog. Sprechakttaxonomie und kommunikative Grammatik. (= Linguistische Arbeiten, 204.) Tübingen: Niemeyer. 1991. "The Dialogic Principle Revisited. Speech acts and mental states". Dialoganalyse III. Referate der 3. Arbeitstagung, Bologna 1990 ed. by Sorin Stati, Edda Weigand & Franz Hundsnurscher, vol. ï, 75-104. (= Beiträge zur Dialogforschung, 1.) Tübingen: Niemeyer. 1993. "Word Meaning and Utterance Meaning", Journal of Pragmatics 20.253-268. 1995. "Looking for the Point of the Dialogic Turn". Future Perspectives of Dialogue Analysis ed. by Franz Hundsnurscher & Edda Weigand, 95-120. (= Beiträge zur Dialogforschung, 8.) Tübingen: Niemeyer. — - 1996a. "Words and their Role in Language Use". Weigand & Hundsnurscher 1996. I. 151-167. — - 1996b. "The State of the Art in Speech Act Theory. Review Article on: Savas L. Tsohatzidis, ed. Foundations of Speech Act Theory: Philosophical and linguistic perspectives. London & New York: Routledge, 1994". Pragmatics and Cognition 4.367-406. 1997a. "Semantik und Pragmatik in der lexikalischen Beschreibung". Hoinkes & Dietrich 1997. 129-145. 1997b. "The Unit beyond the Sentence". Dialogue Analysis: Units, Relations and Strategies beyond the Sentence. Contributions in honour of Sorin Stati's 65th birthday, ed. by Edda Weigand, 3-12. (= Beiträge zur Dialogforschung, 13.) Tübingen: Niemeyer. 1998a. "Emotions in Dialogue". Dialoganalyse VI. Dialogue in the Heart of Europe. Proceedings of the 6th International Congress on Dialogue Analysis, Prague 1996, ed. by Svétla Čmejrková, Jana Hoffmannová, Olga Müllerová & Jindra Světlá, vol. I, 3548. (= Beiträge zur Dialogforschung, 16.) Tübingen: Niemeyer. 1998b. "Misunderstanding: the Standard Case". Journal of Pragmatics. & Franz Hundsnurscher, eds. 1996. Lexical Structures and Language Use. Proceedings of the International Conference on Lexicology and Lexical Semantics, Münster, September 13-15, 1994, 2 vols. (= Beiträge zur Dialogforschung, 9/10.) Tübingen: Niemeyer. Wierzbicka, Anna 1972. Semantic Primitives. (= Linguistische Forschungen, 22.) Frank furt/M.: Athenäum. Wunderlich, Dieter 1980. Arbeitsbuch Semantik. Königstein/Ts.: Athenäum. 1985. "Raum, Zeit und das Lexikon". Sprache und Raum ed. by Harro Schweizer, 6689. Stuttgart: Metzler. 1996. "Models of Lexical Decomposition". Weigand & Hundsnurscher 1996. I. 169183.
THE VOCABULARY OF EMOTION A CONTRASTIVE ANALYSIS OF ANGER IN GERMAN, ENGLISH, AND ITALIAN
EDDA WEIGAND University of Münster On the basis of the contrastive pragmatic model of lexical semantics de scribed in the article on "Contrastive Lexical Semantics" above, I will analyse as a first example the predicative field of EMOTION and demonstrate in the case of ANGER in German, Italian, and English how the methodology works in detail. The analysis is divided into three parts: first a short overview of the literature, second the description of the universal structure and, third, the ways-of-use in the individual languages. In the subsequent articles by Schmitt, Dem'jankov, and Westheide the analysis is applied to other languages. 1.
Overview of the literature After having for a long time been excluded from linguistics in a narrow sense, emotions have, in the last few years, been the topic of increased interest. We have here to differentiate between two questions; first, the question in the area of dialogue analysis: how do emotions influence the course of the dialogue? Second, in the area of lexis, the question: how do we express emotions? or: how do we predicate in the area of feelings, on which our contrastive analysis is based? The article by Schwarz & Ziegler (1996) in the journal "Lexicology" gives a survey of the research on emotions in linguistics and psychology. From this it becomes clear that almost all the work here deals with the central question regarding a structuring of the whole area of emotions, based on the assumption that there are basic emotions. It also becomes clear that there is no agreement in the literature on this matter, that the individual
46
EDDA WEIGAND
authors assume intuitively and mostly without any explicit justification different basic emotions, and also that there is no clarity about what the nature of these basic emotions is. In some cases, above all in psychology, basic emotions are investigated as biological or mental phenomena; in other cases - in linguistics - this ontological approach becomes a conceptual one. In contrast to this, basic emotions represent for us heuristic concepts which structure the predicative field of emotions. There are, on the one hand, individual studies on certain emotions, e.g., the study by Bergenholtz on the semantic field 'fear' (1980), as well as more general studies which survey different emotions, such as, for example, Kövecses (1989) or Wierzbicka (1992). Wierzbicka has developed a 'reduc tive' analysis based on her 'semantic primitives' (1972), while Kövecses gains access to emotional meaning above all via metaphorical expressions. The study by Bresson & Dobrovol'skij (1998) on "Syntax und Semantik der ANGST-Ausdrücke" contains, for German, French and Russian, some interest ing and differentiated observations, in some points paralleling our approach. Thus Bresson & Dobrovol'skij attempt to structure the expressions of the basic emotion ANGST/FEAR according to features and to embed them in semantic constellations which they call 'Sachverhaltsschemata' "schematas of facts" and are represented as deep case structures. However, they pay too little attention to the vital phenomenon of collocation and address the problem of lexis only on the basis of examples, and not systematically. Instead, similar to the work of Stéphanie Bensa (1998), attempts are made to derive a semantics and syntax of expressions of emotion by means of rules from a deep case model. The work of Mel'cuk & Wanner (1994) marks, in my opinion, the state of what can be achieved in an approach based on a theory of signs. Both as regards the depth of analysis and the extent of the material that is analysed and as regards the recognition of what the actual problem is, it is an excellent study which is at the same time aware of the limits of its own approach. These limits lie in the premise that a rule-governed descrip tion is possible which excludes all non-rule-governed cases and thus neces sarily requires a limitation of the material. In contrast to that, our pragmatic approach aims to keep these borders open and to address the total lexis, i.e. to include the 'caprices' of language use as an essential element, which Mel'cuk & Wanner have excluded since they regard 'caprices' as exceptions. Lexical meaning is not only based on rule-governed use, but to just as great an extent on 'idiomatic choice', to take up Sinclair's (1991:109-110) distinction between "free and idiomatic choice".
THE VOCABULARY OF EMOTION
47
In this case the term 'idiomatic choice' goes beyond traditional idioms and phraseological expressions and includes multi-word expressions in general whose combination is subject to certain restrictions. 2.
The universal structure According to our model we correlate universal structure with ways-ofuse. In trying to structure the functional universal part, we have first to delimit the field of what we call EMOTIONS. The literature does not present a generally accepted definition. Should emotions be distinguished from moods such as irritation, stress, or aggressivity? If you assume an emotion of tension opposed to another of relaxation, even moods can be considered as a type of emotion. We therefore start from a large concept of EMOTION which includes feelings, moods and urges. Certain cognitive attitudes such as cynicism, sarcasm, and also presumably, 'feelings of guilt' and 'trust' which are taken to belong to the domain of RATIONALITY, and qualities such as friendliness, helpfulness, brotherliness, which are assigned to the predicative domain of SOCIAL BEHAVIOUR are excluded. Expressions like feeling of coldness, feeling of dampness are not part of the domain of emotions, but of the domain of AWARENESS. In general, we define emotions as cognitive phenomena that emerge neither by convention nor by intention but causally as a reaction to inner physiological processes or as a possible reaction to exterior constellations of happenings, a reaction which might be rationally explained even if at the moment we are unable to do it. Second, we have to structure the whole field of EMOTION according to partial fields of basic EMOTIONS. We assume that the multiple expressions for emotions in individual languages can be grouped together as expressions for specific underlying emotions which we have to establish. Most of these underlying emotions seem to be structured according to antonyms, for instance, HAPPINESS/UNHAPPINESS, JOY/SADNESS, FEAR/COURAGE, ADMIRA TION/DISGUST, or LOVE/HATRED. But there are others that apparently lack a clear antonym, for instance, ANGER, SURPRISE, or PAIN. Nevertheless, the problem of finding out all the basic emotions does not in principle seem very difficult or insoluble. We simply have to check the whole vocabulary and structure it according to central concepts. Empirical analysis, however, has told us that the assumption of such a simple procedure has to be understood as a maxim of hard logical work. There are many, many expressions that have to be taken into consideration and brought into the structure of basic emotions. Such a structure has to be
48
EDDA WEIGAND
seen as a heuristic one for analysing vocabulary without any ontological biological claim. Basic emotions have to be understood as specific possible reactions to specific happenings. In this way, we could define ANGER as basic emotion which might result from a constellation of events which we do not accept, and JOY as basic emotion which might result from happenings we do accept. Here the problems begin already. We might be tempted to add a farther feature but it is difficult to explicate it. Can JOY be characterized as an emotion that results from happenings we do not expect? The following schema of deriving partial fields of EMOTION which are defined as specific reactions therefore has to be seen as very preliminary and tentative: EMOTION [causal reaction to inner physiological processes or exterior events]
Fig. 1: Classifying basic emotions
When we try to distinguish specific basic emotions by criteria, the complexity of the whole attempt becomes immediately evident. We discover, for instance, as antonyms not JOY and SADNESS, but JOY and DISAPPOINT MENT. The criteria represent only a loose heuristic network which is intended to give structure to the field of emotion, nothing more. But perhaps the problems indicate that we should not use criteria at all but should consider basic emotions as concepts a priori. In any case, it does not seem necessary and perhaps not even adequate to try to structure the whole field of emotions systematically. Emotions in the end are not rational, clearly structured phe nomena. Another question has also to be clarified: how compatible with their universal nature is the idea that emotions are shaped by cultural and historical values? In order to draw up a list of basic emotions which can serve as a heuristic foundation for empirical studies, it is necessary to run through the whole vocabulary using a representative dictionary and to structure the whole vocabulary of feeling according to basic emotions by continually checking between a hypotheticallly drawn up list of basic emotions and vocabulary.
THE VOCABULARY OF EMOTION
49
Each basic emotion represents a predicative sector. We will assume the following are candidates for basic emotions: ANGER JOY HAPPINESS
SADNESS UNHAPPINESS
HOPE
DISAPPOINTMENT DESPAIR
LOVE TENDERNESS SYMPATHY CARE
HATRED
ADMIRATION
DISGUST CONTEMPT
FEAR
COURAGE
PRIDE SELF-CONFIDENCE
SHAME EMBARRASSMENT
INTEREST ENTHUSIASM
INDIFFERENCE BORINGNESS
TENSION DISSATISFACTION
RELAXATION SERENITY SATISFACTION
DESIRE
REJECTION
JEALOUSY ENVY
MAGNANIMITY
ANTIPATHY
SERIOUSNESS COMPASSION PAIN SURPRISE RESPECT REMORSE
Third, we are faced with the problem that in different languages there are different expressions for the same basic emotion and not at all always the same number of expressions. We admit that every expression might be characterized by a difference in meaning. However, trying to give structure
50
EDDA WEIGAND
to the vocabulary in different languages implies that we form groups of expressions on the basis of our principle of meaning equivalence. These groups correspond to basic emotions. Their members can be considered as equivalent for expressing the same basic emotion. Thus, for instance, the set of German expressions for the basic emotion of ANGER contains Ärger, Zorn, Wut, Grimm, etc., the set of Italian expressions rabbia, ira, furia, collera, etc., the set of French expressions rage, colère, dépit, ennui, and of English expressions anger, fury, rage, resentment, etc. It is not the case that Ärger always correlates with rabbia, Zorn with furia and Wut with collera, nor is it always possible to indicate precisely the difference between the use of furia and the use of collera. It all depends on the ways-of-use. Forming groups of expressions for basic emotions seems necessary because otherwise you will have innumerable single expressions for single emotions which would have to be all analysed individually. The list of 40 emotion nouns, for instance, in Mel'cuk & Wanner (1994) contains Zorn, Wut, Ärger, Empörung, Groll, Verärgerung, Verdruß as independent entries all of which we assign to the basic emotion ANGER. We can structure the sets of expressions in different languages according to partly the same criteria. These criteria seem to be for emotions in general the criteria [intensification] and [minimization] to which other more specific criteria have to be added, in the case of ANGER a moral aspect and the criterion of being suppressed or curbed. In this way we can structure the sets of expressions for ANGER in different languages according to subsets. Within the subsets however there is no clear correspondance between individual words of different languages as already mentioned above referring to the example Ärger, Wut, Zorn and their Italian counterparts. The following schemata for the language pairs German/English and German/Italian are to be understood in this way, i.e. correlating subsets and not single words:
51
THE VOCABULARY OF EMOTION
ANGER
Emporung Entriistung
AnstoB
indignation outrage offence
~
Raserei Wut Zorn
fury rage anger, wrath, ire (lit.)
/
[INTENSIFICATION]
[MORAL]
~
[CURBED]
/
Arger annoyance, trouble Argerlichkeit anger, crossness Verargerung
~
/
[MINIMIZA nON]
~
/
resentment, grudge, Groll pique Grimm Verbitterung bitterness, embitterment, rancour
VerdruB annoyance, frustration Unwille displeasure, irritation, unwillingness, moroseness, sullenness, exasperation, indignation
Fig. 2.1: German/English ANGER
Emporung Entriistung
indignazione sdegno
Raserei furia, escandescenza, sfuriata Wut collera Zorn ira
AnstoB
/
[MORAL]
[INTENSIFICAnON]
~
/
Arger rabbia Argerlichkeit arrabbiatura Verargerung irritazione, nervoso
~
[CURBED]
~
rancore, astio Groll risentimento Grimm Verbitterung amarezza
~
[MINIMIZATION]
~
VerdruB fastidio, dispetto Unwille stizza Fig. 2.2: German/Italian
52
EDDA WEIGAND
We do not claim that we have included all possible sub-types of ANGER but consider this structure to be a working basis at the present stage. A similar structure can be established for JOY, distinguishing joy/Freude/gioia from delight/Entzücken/estasi by the criterion [intensification], from serenity/ Heiterkeit/serenità by the criterion [minimization] to which the criteria [deep] versus [superficial] might be added, distinguishing, for instance, bliss/Won ne/delizia from amusement/Vergnügen/divertimento. True to our principle that it is not concepts alone that play a role in language use, we now have to pose the essential question of the semantic constellations in which these nouns are used. We assume that in principle these constellations are the same for all basic emotions. It is the basic predicating positions of BE, BECOME versus LOSE, and CAUSE which mainly structure the field of emotion. Besides, emotions play a role as a motivating concept, for instance, in the expressions with anger, in fury/aus Wut, vor Wut which cannot be changed freely but are partly bound to specific ways-ofuse. Moreover, we have to include the process of grading which in general represents a very specific, rather idiomatic process in the individual lan guages. Also emotions take specific adjectives in the NP and specific, not freely choosable adverbs in the VP. Thus the basic universal structure for emotions looks like: EMOTION + BE/BECOME, LOSE/CAUSE EMOTION AS MOTIVATING CONCEPT GRADING EMOTIONS
This universal structure has to be considered as a working hypothesis. It will be changed, corrected, and completed by ongoing empirical work. Especially the questions whether more predicating positions should be used and whether complex predicates, such as CAUSE + {BE, BECOME, LOSE}, should be introduced will have to be clarified. There is also the interesting question whether utterances like (1)
The girl's anger could clearly be seen. Die Verärgerung des Mädchens war deutlich zu sehen.
necessitate the introduction of a further type of constellation in universal structure. We assume that in this case the noun can be derived from ANGER + BE as can be seen by the paraphrases
THE VOCABULARY OF EMOTION
(2)
53
The girl was angry, which could be clearly seen. It could be clearly seen that the girl was angry. Das Mädchen war verärgert, und man konnte dies deutlich sehen. Daß das Mädchen verärgert war, konnte man deutlich sehen.
Similar structures can be established for other predicating fields, such as the field of AWARENESS, in which partly the same meaning positions and constellations recur, for instance AWARENESS + BE/BECOME, LOSE/CAUSE. 3.
The ways-of-use When we correlate universal structure with ways-of-use, we achieve the following working structure: universal structure
↔
EMOTION+BE/BECOME,LOSE/CAUSE EMOTION AS MOTIVATING CONCEPT GRADING EMOTIONS
ways-of-use VP, S VP[Adv/PP...] NP, VP[Adv...]
Fig. 3: Ways-of-use
By classifying ways-of-use according to their syntactic function, syntax finds its place in the model of language use. Constellations of meaning positions are expressed by syntactic phrases which have their defin ed place in the construction of the utterance. As mentioned in the article on "Contrastive Lexical Semantics" (above), syntactic phrases such as the NP also contain referential elements. On the one hand, the phenomenon reference has also turned out to be a very complex phenomenon that, in a pragmatic framework, does not only depend on grammatical rules and expressions like determiners. On the other hand, these grammatical expressions do not always have a specific referential function but belong to the complex lexical units as can be seen by examples like the English and German to suppress one's anger, seinen Ärger unterdrücken against the Italian reprimere la rabbia or by the English and German to go for a run, einen Lauf machen against fare la corsa. We therefore, in general, consider the NP as complex lexical unit, not only the reduced configuration of N + Adj. On the basis of such a theoretical-methodological background we can analyse the vocabulary of ANGER in German, English, Italian and other languages. We take as our starting point the universal structure of semantic
54
EDDA WEIGAND
emotional constellations, distinguish the sub-types of the basic emotion ANGER and correlate them to ways-of-use in the individual languages. Similar to the English Cobuild Programme by Sinclair (1987) and the French Trogramme Eurolexique' by Gruaz (1996) and Honvault (1993), we consider the main ways-of-use to be the frequent ways-of-use, i.e. those which are necessary for the level of the advanced learner. In a first step, and this is the only step represented in the analysis given below, we find these ways-of-use by our own linguistic competence, by native speakers and by dictionaries. In a second step it is necessary to consult large representative corpora of the individual languages, check all the ways-of-use already established and their frequencies and complete them if necessary. Perhaps also new positions in the universal structure might have to be introduced. In a last step we list all the words referred to in an index. This vocab ulary list is structured according to single words and has the function of making the ways-of-use easily accessible. Thus dictionaries of ways-of-use can be compiled after having structured and analysed all semantic fields. These dictionaries will keep to the alphabetical order of single words but will list all their relevant ways-of-use. As can be seen from the analyses, only very general meaning categories are used in universal structure. When we look at traditional contrastive dictionaries which introduce meaning categories only by means of synonyms for the starting entry, we might conclude that it is not the aim of contrastive dictionaries to indicate precisely all meaning differences. The main part is left to the assumption of equivalence between ways-of-use in different lan guages. Naturally, we might refine our analysis and add more differentiated meaning categories. However, at the moment we are not sure whether this is desirable for the level of the advanced learner. 4.
Contrastive analyses of ANGER The contrastive analyses German/English and German/Italian represent the present state of our work. They are by no means meant to be finished or complete. Metaphorical phrases such as die Wände hochgehen/to go up the wall, aus der Haut fahren/to go through the roof are only occasionally included. In a few cases there seems to be no specific expression in one language; a paraphrase therefore would have to be found. As already mentioned the analyses are made mainly on the basis of native competence and dictionaries. The verification and completion by representative corpora has still to be done.
THE VOCABULARY OF EMOTION
55
4.1 German/English Ärger, Ärgerlichkeit, Verärgerung
annoyance, anger, trouble, crossness
BE ärgerlich/verärgert sein über sich ärgern Ärger haben Ärger empfinden viel Ärger erleben es gibt Ärger ärgerliche Töne ärgerliche Briefe seine Ärgerlichkeit die Ärgerlichkeit seines Tons seine Ärgerlichkeit auslassen gegen/an seinem Ärger Luft machen seinen Ärger herausschreien seinen Ärger in sich hineinfressen
to have trouble to feel annoyed/resentful about sth to have a lot of trouble there's trouble, there'll be trouble annoyed/cross tone/manner/voice angry letters his annoyance/crossness the annoyance in/crossness of his voice to give vent to his anger about sth towards sb
(entzündete Wunde (finstere Wolken (aufgewühlte See
angry wound) angry clouds) angry sea [lit.])
BECOME ärgerlich werden in Ärger geraten Ärgernis nehmen an Ärger überfällt/überkommt/packt j-n Ärger kommt auf mein Ärger wächst Ärger bekommen
to become/get annoyed/cross/angry to fly into a rage to be offended by anger comes over/overcomes/seizes annoyance wells up I'm getting more and more angry to get into trouble
LOSE der Ärger verfliegt der Ärger vergeht der Ärger läßt nach sich Ärger ersparen seinen Ärger unterdrücken seinen/den Ärger herunterschlucken seinen Ärger abreagieren
anger blows over anger passes anger diminishes to save o.s. the trouble to suppress/curb one's anger to swallow one's anger to work off/get rid of one's anger
CAUSE ärgern/verärgern j-n ärgerlich machen j-n ein Ärgernis sein Ärger/Ärgernis erregen bei j-m Ärgerlichkeiten bereiten
to anger/annoy/enrage to make sb annoyed to be a nuisance to cause sb trouble, to cause annoyance to sb to make trouble for sb
to be annoyed/cross/angry with sb about sth
to shout out one's anger to suppress one's anger
56
EDDA WEIGAND
j-η in Ärger versetzen ärgerliche Situation das Ärgerliche an der Sache das ist ärgerlich ärgerlich langsam ärgerliche Briefe
to anger, make sb angry, make sb feel angry annoying situation the annoying thing about the matter that's annoying annoyingly slow(ly) annoying letters
MOTIVATING CONCEPT außer sich sein vor Ärger grün und gelb vor Ärger aus Ärger etwas tun zu meinem Ärger zum Ärgernis aller bei aller Ärgerlichkeit voller Ärger zum Ärger (der Anwesenden)
to be beside o.s. with anger to be livid with rage to do sth out of annoyance to my annoyance to the annoyance of everyone in spite of all the trouble full of anger to the annoyance (of those present)
GRADING leiser Ärger heftiger Ärger großer Ärger viel Ärger/wenig Ärger sich sehr ärgern sich schwarz/grün und blau/ gelb und grün ärgern sich maßlos/unglaublich ärgern sich furchtbar ärgern sich zu Tode/krank/ins Grab ärgern sich über die Fliege an der Wand ärgern
quiet anger violent anger great anger/annoyance a lot of trouble, not much trouble to get extremely annoyed, very angry to get hopping mad (informal) to get exceedingly/incredibly angry to get terribly angry to make o.s. sick with anger to get irritated over nothing, to make a fuss about nothing
INTENSIFIED Zorn, Wut, Raserei
anger, fury, rage, wrath, ire (lit.)
BE zürnen, zornig/erzürnt/wütend/wutentbrannt/ aufgebracht/aufbrausend sein Zorn haben auf j . zürnt mir von Zorn/Wut erfüllt sein von Zorn ergriffen sein sein Zorn richtet sich gegen der Zorn der Götter trifft ihn dem Schicksal zürnen Wut im Bauch haben wütende/zornige Stimme
to be angry/enraged/furious/outraged/in censed/irascible/infuriated about to be furious/enraged with sb is angry with me to be filled with anger/rage, to be enraged to be seized with anger his anger is directed at the wrath of the gods strikes him to rage against fate to be seething (with anger) furious/angry voice
THE VOCABULARY OF EMOTION
wütende Gefühle
raging feelings
(wütender Durst (wütende See (draußen wütete ein Sturm (Feuer wütete in der Gegend (die Debatte tobte den ganzen Tag (die jetzige Mode (ein großer Ansturm auf (der letzte Schrei
raging thirst) raging see) a storm raged outside) fires raged throughout the area) the debate raged throughout the whole day) the current rage) a great rage for) to be all the rage)
BECOME sich erzürnen, aufbrausen in Zorn geraten/in Wut kommen der Zorn packt/überkommt j-n Zorn steigt auf dieWände hochgehen aus der Haut fahren
to become/grow/get angry, flare up to fly into a rage anger seizes/overcomes sb anger wells up to go up the wall to go through the roof
LOSE Zorn/Wut legt sich, läßt nach
anger/rage wears off, diminishes
CAUSE j-s Zorn/Wut erregen zornig/wütend machen
57
erzürnen
to excite/cause sb's anger to make sb angry/furious, to make sb enraged, to enrage sb to anger/incense/infuriate/enrage sb
MOTIVATING CONCEPT rot vor Zorn voll Zorn, voller Wut schäumend vor Wut aufbrausend vor Zorn außer sich vor Wut vor Wut kochen aus Wut
red with anger full of anger foaming at the mouth flaring up with anger to be beside o.s. with anger to be hopping mad, seething with anger/rage in fury
GRADING aufflammender Zorn, Raserei heller, lodernder leidenschaftlicher heiliger gerechter dumpfe Wut wilde Wut Zornanfall, Wutanfall
flaring up anger/rage/frenzy/fury blazing passionate holy righteous dull anger/rage fierce anger fit of anger/rage
(Theater machen
to make a [great] fuss)
58
EDDA WEIGAND
WEAKENED Verdruß, Unwille
BE Verdruß haben/empfinden verdrossen/ungehalten sein mürrisch/unwirsch sein Unwillen erregendes Benehmen BECOME unwillig werden Verdruß kommt auf CAUSE Verdruß machen/bringen/bereiten verdrießen
annoyance, displeasure, moroseness, frus tration, sullenness, irritation, exasperation, indignation, unwillingness to be/feel annoyed/frustrated/indignant surly/gruff/sullen/morose/grumpy behaviour which causes displeasure/indigna tion to get annoyed annoyance/frustration arises to cause annoyance/frustration/irritation/vex ation to irritate/annoy/vex
eine Frau belästigen to annoy/bother/pester a woman MOTIVATING CONCEPT zum Verdruß to sb's annoyance
SUPPRESSED/CURBED Groll, Grimm Verbitterung
resentment, grudge, pique bitterness, embitterment, rancour
BE Groll/Grimm empfinden, grollen Groll hegen Bitterkeit/Bitternis/Verbitterung empfinden verbittert sein etwas wurmt j-n
to feel anger, to be filled with anger to harbour a grudge to feel bitterness/embitterment/rancour to be embittered sth rankles with s.o.
CAUSE j-s Groll erregen
to make s.o. angry/resentful
MORAL Entrüstung, Empörung Anstoß
indignation, outrage offence
BE empört/entrüstet sein über E. erfüllt j-η, voller E. sein
to be indignant/outraged about to be full of indignation/outrage
THE VOCABULARY OF EMOTION
E. empfinden Anstoß nehmen seiner E. Luft machen ein entrüstetes Gesicht
to feel indignation/outrage to take offence to give vent to one's indignation an outraged face
BECOME sich entrüsten, sich empören ein Sturm der Entrüstung bricht los
to become indignant/outraged a wave/storm of outrage breaks out
CAUSE etwas entrüstet/empört j-n Empörung/Entrüstung erregen ein empörendes Benehmen
sth outrages, fills with indignation to excite indignation/outrage outrageous behaviour
GRADING gerechte Entrüstung
righteous indignation
59
4.2 German/Italian (q.c. = qualche cosa, qc. = qualcuno) Ärger, Ärgerlichkeit, Verärgerung
rabbia, arrabbiatura, irritazione, nervoso
BE
seinen Ärger herausschreien seinen Ärger in sich hineinfressen
essere arrabbiato/irato/nervoso per/a causa di q.c. con qc. essere in preda alla rabbia, essere preda della rabbia provare rabbia/ira, avere/sentire una rabbia, avere il nervoso avere molta rabbia tira aria di rabbia, c'è aria di rabbia toni arrabbiati lettere irritate la sua arrabbiatura l'arrabbiatura nel suo parlare, la rabbia nei toni lasciare esplodere la propria rabbia contro, lasciare cadere la propria rabbia su, rovesciare la propria rabbia su, investire qc. con/per la propria rabbia lasciare esplodere la propria rabbia contro, sfogare la sua rabbia urlare la sua rabbia mangiare la rabbia, mangiarsi la rabbia
(angry wound (angry clouds
una ferita infiammata, entzündete Wunde) nubi minacciose, finstere Wolken)
ärgerlich/verärgert sein über sich ärgern Ärger haben/empfinden viel Ärger erleben es gibt Ärger, es riecht nach Ärger ärgerliche Töne ärgerliche Briefe seine Ärgerlichkeit die Ärgerlichkeit seines Tons s. Ärger auslassen gegen/an
seinem Ärger Luft machen
60
EDDA WEIGAND
(angry sea BECOME ärgerlich werden
mare tempestoso, aufgewühlte See)
Ärger kommt auf mein Ärger wächst wenn ... dann gibt es Ärger dann bekommst du Ärger
diventare arrabbiato, arrabbiarsi, farsi venire il nervoso, farsi prendere dal nervoso venire preso dalla rabbia, tirare fuori la rabbia irritarsi, prendersi un'arrabbiatura diventare preda della rabbia, avere uno scatto di rabbia la rabbia aumenta/cresce mi monta la rabbia/una rabbia se ... ti faccio vedere (io), vedrai le sentirai, ne sentirai quattro
LOSE der Ärger verfliegt der Ärger vergeht der Ärger läßt nach sich Ärger ersparen seinen Ärger unterdrücken seinen/den Ärger herunterschlucken seinen Ärger abreagieren
la rabbia svanisce/sparisce la rabbia passa la rabbia diminuisce risparmiarsi la rabbia reprimere la rabbia ingoiare la rabbia dover lasciar passare/raffreddare la rabbia
in Ärger geraten Ärgernis nehmen an Ärger überfällt/packt/überkommt j-n
CAUSE ärgern, verärgern j-n ärgerlich machen j-n ein Ärgernis sein für j-n Ärger, -nis erregen bei j-m/j-m bereiten j-m Ärgerlichkeiten bereiten j-n in Ärger versetzen ärgerliche Situation für j-n das Ärgerliche an der Sache das ist ärgerlich ärgerlich langsam ärgerliche Briefe MOTIVATING CONCEPT außer sich sein vor Ärger grün und gelb vor Ärger aus Ärger etwas tun zu meinem Ärger zum Ärgernis aller
far arrabbiare qc, seccare qc. con q.c. rendere rabbioso/nervoso qc, far venire il nervoso a qc. essere causa/motivo di un'arrabbiatura causare rabbia/un'arrabbiatura a qc. fare un dispetto a qc. indurre alla rabbia, spingere alla rabbia qc. una situazione irritante per qc. l'irritante della situazione questo è irritante/seccante fastidiosamente lento, in modo fastidiosamen te lento lettere irritanti essere fuori di sé per la rabbia/a causa della rabbia essere nero/verde dalla rabbia (giallo/rosso/ bianco/blu) fare q.c. per/a causa della rabbia per/a causa della mia rabbia facendo arrabbiare tutti
THE VOCABULARY OF EMOTION
bei aller Ärgerlichkeit voller Ärger zum Ärger (der Anwesenden) GRADING leiser Ärger heftiger Ärger großer Ärger viel/wenig Ärger sich sehr ärgern sich schwarz/grün und blau/ gelb und grün ärgern sich maßlos/unglaublich ärgern sich furchtbar ärgern sich zu Tode/krank/ins Grab ärgern sich über die Fliege an der Wand ärgern
61
nonostante la rabbia pieno di rabbia per dispetto (contro i presenti) rabbia soffusa/repressa rabbia violenta grande rabbia molta/poca rabbia arrabbiarsi molto essere nero/verde dalla rabbia arrabbiarsi arrabbiarsi arrabbiarsi arrabbiarsi
da matti, incredibilmente terribilmente a morte per un nonnulla/una banalità
INTENSIFIED Zorn, Wut, Raserei BE zürnen, zornig/erzürnt/wütend/wutentbrannt/ aufgebracht/aufbrausend sein
ira, collera, furia, sfuriata, escandescenza
wütende Gefühle
essere in collera con qc., essere irato/collerico/inviperito/infuriato con/contro qc. per q.c. avere in ira qc. per q.c. sono in ira a qc. essere pieno di essere preso dall'ira portare ira a qc, la sua ira è indirizzata a l'ira degli dei lo colpisce essere inviperito con il fato avere rabbia dentro voce collerica (#persona collerica, tipo collerico) sentimenti di rabbia
(wütender Durst (wütende See (draußen wütete ein Sturm (Feuer wütete in der Gegend (die Debatte tobte den ganzen Tag (die jetzige Mode (ein großer Ansturm auf (der letzte Schrei
[sete ardente], sete da morire/da pazzi) mare tempestoso) fuori infuriava una tempesta) il fuoco infuriava nella/sulla regione) il debattito infuriò per tutto il giorno) la moda corrente, the current rage) una corsa a, a great rage for) una cosa all'ultimo grido, to be all the rage)
Zorn haben auf j . zürnt mir von Zorn/Wut erfüllt sein von Zorn ergriffen sein sein Zorn richtet sich gegen der Zorn der Götter trifft ihn dem Schicksal zürnen Wut im Bauch haben wütende/zornige Stimme
62
EDDA WEIGAND
BECOME sich erzürnen, aufbrausen in Zorn geraten, in Wut kommen der Zorn packt/überkommt j-n Zorn steigt auf die Wände hochgehen aus der Haut fahren
adirarsi, accendersi d'ira, infuriarsi andare in collera dare in escandescenze, inviperirsi la collera aumenta, cresce andare su tutte le furie, andare fuori (di) testa esplodere, uscire di testa
LOSE Zorn/Wut legt sich/läßt nach
la collera/l'ira passa/svanisce/sparisce
CAUSE j-s Zorn/Wut erregen zornig/wütend machen, erzürnen
far andare in collera qc. far adirare qc.
MOTIVATING CONCEPT rot vor Zorn voll Zorn, voller Wut schäumend vor Wut aufbrausend vor Zorn außer sich vor Wut vor Wut kochen aus Wut (etwas zerbrechen)
rosso di rabbia/per la rabbia pieno d'ira traboccante/schiumante di collera/ira/rabbia furibondo dall'ira fuori di sé per la collera cuocere dalla rabbia per la rabbia/collera (distruggere q.c.)
GRADING aufflammender Zorn, Raserei heller, lodernder leidenschaftlicher heiliger gerechter dumpfe Wut wilde Zornanfall, Wutanfall (Theater machen)
ira/collera/furia accesa ira/collera/rabbia furiosa/furibonda/furente ira/collera passionale ira santa/sacra ira giusta ira/collera cupa collera furiosa impeto di ira, collera fare un'ira di Dio, succedere l'ira di Dio, fare una sfuriata
WEAKENED Verdruß, Unwille BE Verdruß haben/empfinden verdrossen/ungehalten sein mürrisch/unwirsch sein
fastidio, dispetto, stizza avere un fastidio, provare fastidio/stizza, mi viene a fastidio essere infastidito da, provare fastidio/ stizza, essere seccato di/per/a causa di, essere cupo per/a causa di essere burbero/un brontolone/arcigno,
THE VOCABULARY OF EMOTION
63
Unwillen erregendes Benehmen
brontolare comportamento fastidioso
BECOME unwillig werden Verdruß kommt auf
indignarsi per, diventare burbero emerge fastidio
CAUSE Verdruß machen/bringen/bereiten verdrießen
dare/causare fastidio a q c , far (un) dispetto a stizzire/infastidire qc.
(eine Frau belästigen molestare/importunare una donna) MOTIVATING CONCEPT zum Verdruß per dispetto/fastidio SUPPRESSED/CURBED Groll, Grimm Verbitterung BE Groll/Grimm empfinden, grollen
rancore, astio, risentimento amarezza
Groll hegen Bitterkeit/Bitternis/Verbitterung empfinden verbittert sein etwas wurmt j-n
provare/avere rancore/astio contro/per qc. per q.c, serbare rancore/astio contro/per/a qc. per q.c. gli serba rancore, serba rancore per lui, essere risentito nutrire rancore nei confronti di/contro provare amarezza essere amareggiato si rode (il fegato)
CAUSE j-s Groll erregen
creare astio/rancore/risentimento
er empfindet Groll für ihn
MORAL Entrüstung, Empörung BE empört, entrüstet sein über E. erfüllt j-η, voller E. sein E. empfinden Anstoß nehmen seiner E. Luft machen ein entrüstetes Gesicht
indignazione, sdegno essere indignato, preso dall'indignazione, es sere mosso a sdegno, restare/rimanere sde gnato di/per/a causa di essere pieno di sdegno provare/sentire indignazione/sdegno scandalizzarsi far esplodere l'indignazione un viso sdegnato
64
EDDA WEIGAND
BECOME sich entrüsten, sich empören ein Sturm der Entrüstung bricht los
indignarsi/sdegnarsi per un'ondata di sdegno scoppia
CAUSE etwas entrüstet/empört Entrüstung/Empörung erregen ein empörendes Benehmen
q.c. indigna/sdegna qc., muovere a sdegno suscitare l'indignazione/lo sdegno di comportamento sdegnoso
GRADING gerechte Entrüstung
5. -
-
j-n
giusta indignazione
Conclusion From the analysis, the following points can be drawn as a conclusion: Language use is based on a network of ways-of-use. To learn a lan guage means to know how words are used and what utterances are used in specific situations. Semantics can no longer be left to intuition. We have to control what we consider to be semantic conventions by large text corpora. Only thus can we achieve "hard, measurable evidence" to use Sinclair's ex pression (Sinclair 1987:XV). Working with vocabulary makes us aware that languages cannot be totally explicated. It is an illusion to think that one might achieve a complete description of vocabulary. Ways-of-use in different languages are not absolutely equivalent nor can the meaning of a specific way-of-use always be indicated explicitly. It does not seem very helpful for language use to differentiate semantically single words for ANGER in a finer way than is done in figure 2. You may indicate, for instance, in Italian for ira [literary use]. How ever, sometimes ira is used nearly in the same sense as rabbia, provare ira/provare rabbia, sometimes nearly in the same sense as collera, ira cupa/collera cupa. Moreover, it seems impossible to indicate all the subtle conventional differences on the expression side. For instance, as already mentioned, you have to say seinen Ärger unterdrücken/ reprimere la rabbia with a reflexive pronoun in German, whereas you can use the definite article in den Ärger herunterschlucken/ingoiare la rabbia. These complexities are an inherent characteristic of natural lan guages and are a result of their being based on conventions and only partly on rules. This is the reason why foreign languages cannot be
THE VOCABULARY OF EMOTION
65
completely taught by books, even with a description based on use, but have to be learned by living in the country in which they are spoken and by becoming slowly acquainted with the foreign conventions. Progress has been achieved in the past in the field of monolingual semantics and dictionaries such as the Cobuild dictionary. Progress remains to be made in the case of contrastive pragmatic semantics and corresponding dictionaries which will be most important for the future of our societies. Globalization and internationalization will be essentially based on dialogic multi-lingual competence. Acknowledgement I would like to thank David Beal and Luigi Torchio for helping me in the English and Italian analyses, respectively.
References Bensa, Stéphanie 1998. "Définition sémantique des fonctions d'arguments pour les prédicats du domaine des emotions en allemand". Bresson & Kubczak 1998. 145-162. Bergenholtz, Henning 1980. Das Wortfeld Angst. Eine lexikographische Untersuchung. Stuttgart: Klett. Bresson, Daniel & Dmitrij Dobrovol'skij 1998. "Semantik und Syntax der 'ANGST'Ausdrücke. Versuch einer integrativen mehrsprachlichen Analyse". Bresson & Kubczak 1998. 163-186. Bresson, Daniel & Jacqueline Kubczak, eds. 1998. Abstrakte Nomina. Vorarbeiten zu ihrer Erfassung in einem zweisprachigen syntagmatischen Wörterbuch. Tübingen: Narr. Gruaz, Claude 1996. "Lexicologie contrastive. Forme et sens dans la synchronie contemporaine". Weigand & Hundsnurscher 1994. I. 29-38. Honvault, Renée 1993. "Le dictionnaire morphosémantique des familles synchroniques des mots français DISFA". Actes de XXe Congrès International de Linguistique et Philologie Romanes, Zürich 1992 ed. by Gerold Hilty, Tome IV, 147-160. Tübingen & Basel: Francke. Kövecses, Zoltán 1989. Emotion Concepts. New York, Berlin & Heidelberg: Springer. Mel'cuk, Igor A. & Leo Wanner 1994. "Lexical Co-occurrence and Lexical Inheritance. Emotion lexemes in German: a lexicographic case study". Lexikos 4.88-160 (AfrilexReeks/Series 4). Stellenbosch/Republik Südafrika. Schwarz, Manuela & Wolfram Ziegler 1996. "Emotionen in Neurolinguistik und Lexiko logie. Ein Forschungsüberblick". Lexicology 2.34-62. Sinclair, John 1987. "Introduction" to: Collins COBUILD English Language Dictionary ed. by John Sinclair et al., xv-xxI. London, Glasgow & Stuttgart: Collins & Klett. 1991. Corpus, Concordance, Collocation. (= Describing English Language). Oxford: Oxford University Press. Weigand, Edda 1996. "Words and their Role in Language Use". Weigand & Hundsnurscher 1996.I. 151-167.
66
EDDA WEIGAND
1997. "Semantik und Pragmatik in der lexikalischen Beschreibung". Kaleidoskop der Lexikalischen Semantik ed. by Ulrich Hoinkes & Wolf Dietrich, 129-145. (=Tübinger Beiträge zur Linguistik, 428.) Tübingen: Narr. & Franz Hundsnurscher, eds. 1996. Lexical Structures and Language Use. Proceedings of the International Conference on Lexicology and Lexical Semantics, Münster, September 13-15, 1994, 2 vols. (= Beiträge zur Dialogforschung, 9/10.) Tübingen: Niemeyer. Wierzbicka, Anna 1972. Semantic Primitives. (= Linguistische Forschungen, 22.) Frank furt/M.: Athenäum. —— 1992. Semantics, Culture, and Cognition. Universal human concepts in culture-specific configurations. New York & Oxford: Oxford University Press.
SPRACHVERGLEICH ALS TEXTVERGLEICH CHRISTIAN SCHMITT Universität Bonn Summary The comparison of language pairs as well as the comparison of genealogically related and unrelated languages has - within the framework of language typology and contrastive linguistics - seldom presented results beyond the establishment of structural comparisons; the reason for the evident deficits of such studies are to be found primarily in the fact that the contrasted elements and language rules have not been analysed with pragmatic aspects in mind. The lack of action-oriented descriptions of the language sign has been particularly obvious in the field of semantics where componential analysis is still the predominant method of description since it neglects the representation of conceptual fields in actual texts and puts too much weight on a linguist's competence or intuition, sometimes even resorting to etymo logical criteria. If the description of language signs is to make language actions possible or at least improve them, their description must originate from the function of words. However, the function of words does not consist in making objective and ontological statements about the world but in applying words in meaningful propositions; this is why first of all we have to determine the purpose of expressions and the role linguistic entities play in concrete speech acts. The method that is proposed and defended here is exemplified on the basis of the conceptual field of the German word for 'Ärger' (engl. ANGER) and its partners in modern French and Spanish, which is very badly described in bilingual dictionaries. Proceeding from lexicographic principles it is shown how the problem can be solved and which postulates need to be considered when describing the vocabulary of emotions in a contrastive analysis. Establishing the French equivalents for German Scheiß-, Drecks- and Mist- as well as finding the Spanish equivalents for German -haltig is intended to exemplify the proposed method: on the basis of morphemes in texts it is shown how the theoretical procedure is applied to general and specialized texts.
68 1.
CHRISTIAN SCHMITT
Vorbemerkungen Sprachen vergleichen wir, um Einsichten in die vielfältige Struktu rierung und das unterschiedliche Funktionieren von Kommunikations systemen zu gewinnen, die genealogisch verwandt oder auch nicht verwandt sein können. In diesem Zusammenhang dient der Sprachvergleich in der Tradition von Wilhelm von Humboldts Schrift "Über die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluß auf die geistige Entwickelung des Menschengeschlechts" von 1835 dem Erkenntnisgewinn und sicher auch dem Verständnis für die je gegebene Andersartigkeit sozialer Gruppen im sprachlichen Handeln (ed. 1907). Wir vergleichen aber auch Sprachen, um eine Antwort auf konkrete Fragen wie z.B. das Verhältnis einer Ausgangs sprache zu einer Zielsprache und die Vermittelbarkeit der Gemeinsamkeiten und vor allem der Unterschiede geben zu können und damit das Übersetzen oder die Lehre einer Fremdsprache zu erleichtern und die Gestaltung von Lehrbüchern zu begründen. In diesem Fall dient der Sprachvergleich im Rahmen der Angewandten Sprachwissenschaft der Sprachendidaktik, die aus geht von der Tatsache, daß es rekurrente Schwierigkeiten beim sprachlichen Handeln in der L2-Sprache und immer wieder auftretende, durch die Grundsprache bedingte Interferenzen gibt, die zu vermeiden allein das gezielte Einstellen der Lernenden und das Wiederholen in gut begründeten Übungstexten beitragen können (Rein 1983:5ff.; Schmitt 1997:9ff.). Beide Fragestellungen sind uneingeschränkt gerechtfertigt, und es ist sicher aus der je gegebenen Funktion herleitbar, wenn man die These aufstellt, daß sich beide Sichtweisen ergänzen: Die Sprachtypologie leistet die wissenschaftliche Beschreibung der Humansprachen (Müller 1971), die Kontrastive Linguistik hingegen hat die Aufgabe, aus wissenschaftlich ermittelten Fakten und Regeln die fachdidaktischen Folgerungen zu ziehen und die Erkenntnisse für eine verbesserte Sprachlehre umzusetzen. Dabei besteht heute ein Grundkonsens hinsichtlich der Verwirklichung kontrastiver, der Sprachlehre förderlicher Studien, wie dies nicht zuletzt auch das DUDEN-Programm, dem wir für die romanischen Sprachen die kontrastiven Grammatiken zum Französischen (Zemb 1978/1984) und Spani schen (Cartagena & Gauger 1989) verdanken, uns deutlich vor Augen geführt hat: Unsere Grammatiken - und hier spreche ich nicht nur für die Romanistik - sind so heterogen und disparat, teilweise auch nach so verschiedenen theoretischen Ansätzen konzipiert, daß der Vergleich einer französischen Grammatik mit einer spanischen auf keinen Fall eine kontrastive deutsch spanische Grammatik zur Folge hat; gleiches gilt auch für die Wörterbücher, denn hier liegen zum einen stets explizit oder implizit divergierende Normen
SPRACHVERGLEICH ALS TEXTVERGLEICH
69
zugrunde und, was wohl noch viel schwerer wiegt, die Makrostrukturen und die Mikrostrukturen sind so verschieden, daß es z.B. keine zwei einander völlig entsprechenden Wörterbücher für das Neuspanische gibt und die Aussicht auf zwei auch nur annähernd einander entsprechende Wörterbücher für die Sprachenpaare Spanisch/Italienisch oder Spanisch/Französisch eine noch größere Illusion darstellt. Hinzu kommt, daß die kontrastiv-pragma tische Analyse, die zu Recht bei Weigand (1993; 1996) im Vordergrund steht, nur dann überhaupt vorgenommen werden kann, wenn die beiden contrastanda in gleichem Maße Aspekte des Zeichengebrauchs berücksichtigt haben. Im konkreten Fall heißt dies, daß beim Sprachenpaar Deutsch/Eng lisch bezüglich he is a bit thick "er ist schwer von Begriff' und he is as thick as two short planks "er ist strohdumm" (Weigand 1996:161; 1997:131-135) nur dann von einem Unterschied zwischen Ausgangs- und Zielsprache gesprochen werden kann, wenn gesichert ist, daß beide Wörterbücher in ihren Einträgen alle Bedeutungspositionen gleichermaßen erfaßt und vollstän dig aufgelistet haben. Aufgrund der eigenen Sprachkompetenz wird kein Sprachwissenschaftler eine solche Aussage für die Sprache wagen, für die er muttersprachliche Kompetenz besitzt; für die Zweitsprache wäre eine der artige Assertion natürlich noch risikoreicher, denn bei der Beschreibung von Semantik und Pragmatik in der L2-Sprache ist er grundsätzlich auf Garanten angewiesen, mit denen zusätzlich bei der Evaluation sprachlicher Zeichen und Regeln hermeneutische Probleme bestehen (können). 2.
Zur Konstituierung von Wortfamilien Die Beschreibung der lexikalischen Einheiten wird immer noch von zwei Defiziten bestimmt: der unzureichenden Erklärung und Konzeption des sprachlichen Zeichens und der defizitären Definition des pragmatischen Aspekts. Man könnte auch behaupten, daß dem Verhältnis zwischen den semantischen und pragmatischen Werten eines sprachlichen Zeichens bisher nicht die gebotene Aufmerksamkeit geschenkt wurde (Schmitt 1997b). Trotzdem, wird man auf diese Aussage entgegnen, funktioniert in Frankreich wie in Deutschland die ein- und mehrsprachige Lexikographie, denn nie wurden mehr Wörterbücher produziert als heute. Bisweilen wird gar unterstellt, hier folge die Theorie der Praxis, und die Methodologie lasse sich aus dem erstellten Werk ableiten bzw. die Analyse der Werke zeige meist ein erschreckendes Theoriedefizit. Es scheint mir geboten, sich die Frage zu stellen, ob sich die Ziele der Pragmatik des sprachlichen Zeichens und der Semantik als Beschreibung des Begriffssystems nicht verbinden lassen oder ob eine Inkompatibilität
70
CHRISTIAN SCHMITT
zwischen beiden besteht, da bei der Definition eines sprachlichen Zeichens einer Wortfamilie oder des Gesamtwortschatzes der Semantiker von einem bestimmten Wert des oder der Zeichen ausgeht, der Pragmatiker hingegen das Zeichen immer als Teil einer Sprechhandlung versteht. Daraus resultiert, daß für den Semantiker die Einheit Lexem oder Wort mit einem Sinn verbunden bleibt, der sich relativ präzise beschreiben läßt, während für den Pragmatiker diese genaue Bedeutung nie vorgegeben ist, denn er geht immer vom Wort-im-Text oder vom Zeichen-in-Funktion aus; für ihn bleibt entscheidend, zu ermitteln, welche Rolle ein Zeichen oder Wort in einem bestimmten Text oder in einer bestimmten Handlung spielen kann. Es erübrigt sich, den Beweis dafür anzutreten, daß die heutigen Wörter bücher den pragmalinguistischen Anforderungen nicht gerecht werden, da vielfach pragmatische Ansätze nicht einmal rezipiert wurden. Auf die defi zitäre Berücksichtigung pragmatischer Markierungen verweisen auch - in Anlehnung an Wolski (1980), Hölker (1981), Wiegand (1985) und Harras (1985) - Jäger & Plum (1989:851b), wenn sie resümierend folgende Fest stellung wagen: Auf jeden Fall läßt sich - und dies hat die jüngere Diskussion zur Form von Bedeutungsangaben auch recht einhellig und deutlich herausgestellt - der Wortschatz einer Sprache zu großen Teilen nicht nach dem Muster der ana lytischen Definition beschreiben.
und es ist ihnen zuzustimmen, wenn sie das Postulat aufstellen (1989:851a): Wertende Bedeutungselemente oder situationelle und kontextuelle Restriktionen der Wortverwendung müssen [...] zusätzlich in pragmatischen Markierungen angegeben werden, was aber nur dann möglich ist, wenn eine entsprechende Varietätendimension auszumachen ist.
Für die französische Sprache darf man feststellen, daß die Lexiko graphie (und, zumindest teilweise, die Lexikologie) in einer verkrusteten Tradition verhaftet bleibt, die von Estienne (1532) bis zu den RobertWörterbüchern (31993, u.ö.) reicht. Mit derselben Dokumentation werden Sprachwörterbücher hergestellt, die begriffssystematischen Anforderungen genügen sollen. Die Wörterbücher sollen den unterschiedlichsten Ansprüchen gerecht werden: Der Grand Robert wurde für den Universalisten und den Spezialisten gemacht, der Petit Robert ist für den Lehrer und die Studieren den konzipiert, der Robert Quotidien dient dem gebildeten 'Otto Normalver braucher', während der Micro Robert den weniger gebildeten Benutzer im Visier hat. Die Lemmata entsprechen einander cum grano salis; was jedoch die Bedeutungsangaben betrifft - ich zitiere hier den Robert Quotidien,
SPRACHVERGLEICH ALS TEXTVERGLEICH
71
exempli gratia - so steht fest, daß dieses Wörterbuch "offre des définitions, une organisation des sens (polysémie), la mention des synonymes ou des contraires. L'ordre de sens est un ordre historique qui permet de comprendre l'évolution du mot" (1996, VII), so als ob der Benutzer des Wörterbuchs sich in erster Linie über diachronische Fragestellungen informieren wollte und der Gebrauch der sprachlichen Zeichen nur von untergeordnetem Interesse wäre. Seit Bréal erkennt man im sprachlichen Zeichen eine aus Lautkörper und Bedeutung bestehende Einheit und seit de Saussure - man könnte auch sagen, seit Charles de Bovelles (1533; vgl. Schmitt 1982:261f.) - spricht man von der Arbitrarität des sprachlichen Zeichens, was heißt, daß mit Ausnahme der Onomatopoeien, wo eine Imitation des Objekts durch den Lautkörper besteht, es zwischen Signifikanten und Signifikat keine direkten Beziehungen gibt. Für den Semantiker besitzt der Signifikant einen Inhalt, der einen Reflex in der Realienwelt besitzen oder auch nur ein Abstraktum bezeichnen kann. Bei der Analyse ergeben sich zahlreiche Probleme, wie z.B. die Polysemie (die vorliegt, wenn ein Zeichen mehrere Inhalte kennt), die Synonymie (die besteht, wenn ein Zeichen oder eine Reihe von Zeichen in einem bestimmten Zusammenhang mit einem anderen oder einer Reihe von anderen Zeichen bedeutungsgleich ist) und vor allem die kombinatorischen Regeln, die idiomatische Einheiten betreffen, deren Bedeutung nicht aus den konstitutiven Einheiten deduziert werden kann oder deren Inhalt einen suprasummativen Wert insofern bildet, als die Einheiten der Kombination nicht aus den Einzelteilen erschlossen werden können. Die pragmatische Analyse hingegen behandelt keine isolierten oder segmentierten Einheiten als solche, sondern bezieht sich jeweils auf den aktualisierten Gebrauch, sie beschreibt die sprachlichen Zeichen aus der Sprechhandlung heraus, denn allein in der Textverwendung entwickeln Zeichen ihren Sinn: Sie dienen dazu, Propositionen zu bilden und mittels eines Prädikats etwas über ein Subjekt auszusagen. Der Sprecher gebraucht sie, um die Modalitäten der Prädikation zu bilden, die Zeichen dienen ihm bei der Beschreibung der Welt, wie er sie sieht, bei Aussagen oder bei Ausrufen über das, was er wahrnimmt. Der Sinn einer Proposition hängt von der Pragmatik des Sprechers ab, die Textbedeutung bleibt stets eng an die Sprechhandlung gebunden und kann nicht von einem vorgegebenen Begriffssystem aus erklärt werden, wie es z.B. die strukturalistische Semantik oder die generative Semantik in Mißachtung der sprachlichen Wirklichkeit im Modell wie in zahlreichen Einzeluntersuchungen aufgebaut haben. Was die Konstitution und die Analyse von Wortfamilien betrifft, so möchte ich hier nicht die Auffassung der deutschen Germanistik, die
72
CHRISTIAN SCHMITT
Weigand (1998) bereits ausführlich gewürdigt hat, ein zweites Mal vortragen. Eines ist klar: Weder das etymologisch ausgerichtete Modell von Walde & Hofmann (31938-1954), noch der auf die Synonymik rekurrierende Ansatz von Hundsnurscher & Splett (1982) oder der strukturalistisch ausgerichtete Zugriff von Splett (1990) können als pragmatische Methoden gewertet werden: die Argumente liegen auf der Hand. Auch zu dem Ansatz von Gruaz und Honvault (vgl. Gruaz' Beitrag in diesem Band) sei nur soviel gesagt: I II III
Semantische Beschreibung darf sich nicht der Intuition oder der Subjektivität des beschreibenden Linguisten bedienen, um Vergleiche und Klassifikationen zu rechtfertigen. Das bei der Konstituierung von Wortfamilien herangezogene Kriterium der Etymologie bleibt nur begrenzt tauglich und ist daher zu meiden. Die Annäherung von Ausdruck und Sinn verhindert die Beachtung pragmatischer Regeln und erschwert zumindest den Sprachenpaarbezogenen Vergleich.
Meine Kritik möchte ich mit einem französischen Beispiel verdeut lichen: Für das Verb saupoudrer (PRob 31993:2041) ist die semantische Beschreibung dann schwierig, wenn es um die im Text übliche Verbindung von saupoudrer + NOMEN (das ein Gericht bzw. eine Mahlzeit bezeichnet) geht. Wie sollte man auch seinen Gebrauch in einem Satz wie les pouvoirs publics saupoudrent les crédits beschreiben. Es ist sicher wenig ratsam, Gruaz und Honvault folgend, die Metapher ausgehend von der kulinarischen Verwendung zu beschreiben, und sicher vorteilhaft, saupoudrer mit den usuellen Verben, die mit crédit in Kollokation stehen, zu determinieren, da nur dann die Bedeutung "attribuer à de très nombreux bénéficiaires (des crédits minimes ou des moyens faibles) au lieu d'affecter le budget à quelques postes prioritaires" (PRob 31993) nachvollziehbar wird. Wer sich von etymologischen Kriterien leiten läßt, wird von den Konstituenten des Verbs ausgehen, frz. sel (bzw. sein gelehrtes Allomorph sal) und poudrer, und sich evtl. auf die Seite der Puristen schlagen, die unter Berufung auf die Etymologie Kollokationen wie saupoudrer de sucre, de canelle etc. kritisieren. Eine derartige semantische Beschreibung wird bereits vom RobQ (1996:1732) abgelehnt, der betont, es gebe "aucun bon sens dans cette interdiction puisque saupoudrer ne s'emploie pas pour saler". Hier wird beispielhaft deutlich, welche Folgerungen eine etymologische Klassifikation bedingen kann und wie dadurch falsche Einschätzungen bewirkt werden können.
SPRACHVERGLEICH ALS TEXTVERGLEICH
3.
73
Kontrastiver Vergleich von Teilwortschätzen Was ein kontrastiver Vergleich leistet bzw. zu leisten vermag, soll am Beispiel des prädikativen Feldes dt. ÄRGER gezeigt werden, das Edda Weigand für den deutsch-englischen und deutsch-italienischen Sprachver gleich abgehandelt hat (in diesem Band); im Mittelpunkt meiner Aus führungen soll jedoch das Sprachenpaar Deutsch-Französisch stehen, wobei die verbreitetsten ein- und zweisprachigen Wörterbücher herangezogen werden sollen. Was die zu beachtende Universalstruktur betrifft, so hat Weigand bereits deutlich gemacht, daß es sich hierbei um ein Teilfeld eines noch viel weiteren, aus schweren Wörtern und brisanten Lexemen (Jäger & Plum 1989:851b-852c) bestehenden Feldes handelt, das der EMOTIONEN, das angenehme wie unangenehme Empfindungen aus einer affektiven Sicht umfaßt. Man muß daher streng genommen auch die Bezeichnungen für den Ausdruck der Freude und des Gefallens wie für Schmerz und Wut mitberück sichtigen, d.h. eine etwas weiter ausholende, wohl angemessenere Studie müßte auch alle Gründe und Reaktionen auf Wirkungen umfassen (Weis 1998), und in diesem Sinne wäre Ärger eine Art Reaktion auf etwas, das sich gegen unsere Erwartung oder unseren Willen ereignet hat: Deshalb haben wir ÄRGER oder ÄRGERN wir uns. Niemand wird erwarten, daß zwischen beiden Sprachen ein Äquivalenz verhältnis besteht; ein solches gibt es hier ebenso wenig wie zwischen dem Deutschen und dem Spanischen (Faber & Jiménez 1997). Ein Psychologe hätte sicher keine Schwierigkeiten, den Nachweis dafür zu erbringen, daß die unendliche Anzahl möglicher Reaktionen auf ein unangenehmes Ereignis keine adäquaten Entsprechungen in den Ausdrucksmitteln unserer Sprachen findet, weshalb wir auch vom Topos der sprachlichen Unzulänglichkeit im Zusammenhang mit den menschlichen Gefühlen sprechen (Müller 1966; Schmitt 1979). Bei jedem Sprecher besteht mehr oder weniger der Eindruck, daß die Gefühle weiter reichen und sich differenzierter präsentieren als die sprachlichen Möglichkeiten. Auch hier gibt es zunächst einige Punkte zu klären: Klassifiziert Weigand ÄRGER, ZORN, WUT, GRIMM als Grundelemente, weil sie germanophon ist, oder handelt es sich dabei wirklich um eine Grundstruktur? Nach intensiver Beschäftigung mit Sprach- und Fachwörterbüchern muß ich gestehen, daß ich noch nicht in der Lage bin, die konstitutiven Elemente dieses Feldes eindeutig zu benennen. Für das Spanische bieten Slabý, Grossmann & Illig (41994) folgende konstitutive Einheiten: disgusto, enfado, enojo, rabia, furia, indignación, fastidio-, auch hier muß ich einräumen, daß diese Klassifikation für mich keine inhärente Systematik besitzt. Welches
74
CHRISTIAN SCHMITT
spanische Wort steht in usueller Korrelation mit dt. Ärger? Ist dies disgusto oder enfado, wie dies das Wörterbuch nahelegt, oder eher doch enojo, eine Bezeichnung, die gewöhnlich mit Zorn korreliert? Kann man behaupten, sp. furia entspreche eher Zorn oder Wut? Immerhin scheint außer Zweifel zu stehen, daß rabia stärker ist als enojo und daß furia eine Emotion ausdrückt, die stärker ist als enojo und rabia; doch ist die dt. Redewendung seinen Zorn an jemandem auslassen eher korreliert mit dem im Spanischen recht usuellen desfogar su enojo en alguien. Die Wörterbücher können die bestehende Verlegenheit nicht beseitigen, und dieselben Schwierigkeiten bestehen natür lich auch, wenn es gilt, die sogenannten Grundemotionen zu definieren (vgl. Faber & Jiménez 1997:271f.). Die Haltung, die Menschen einnehmen kön nen, ihre Reaktionen auf eine Sache, einen Zustand, eine Gegebenheit und ein Ereignis, das sie nicht akzeptieren möchten, können vielfältiger Art sein. Sicher ist es notwendig, ausgehend von einem Grundkonzept die Emotionen zu klassifizieren, die elementaren Gefühle zu definieren, doch glaube ich nicht, daß es sich dabei um eine leichte Aufgabe handelt und daß man nur das Wörterbuch durchgehen muß und dann die begrifflichen Bereiche zu strukturieren hat. Nichts anderes haben mit der systematischen Ausarbeitung des Leipziger Enquète-Fragebogens z.B. Hallig & von Wartburg (21963) für das Französische gemacht, und es braucht nicht weiter begründet zu werden, wie problematisch ihre begriffliche Kategorisierung ist. Ohne weiterführende Textuntersuchungen ist eine derartige Strukturie rung für das heutige Spanisch noch nicht möglich, denn die einsprachige Le xikographie des Spanischen reicht dafür nicht aus und die zweisprachige, auf das Sprachenpaar Deutsch-Spanisch ausgerichtete ist immer noch viel zu defi zitär. Würde man auf der Basis der aktuellen Wörterbücher eine solche Gra duierung vornehmen, sähe die recht rudimentäre Gliederung wie folgt aus: indignación - enfado - disgusto -fastidio - enojo - rabia -furia Da aber diese Reihenfolge auf keiner kontrastiven Studie aufbaut, sollte man bei der Auslegung dieser graduellen Abstufung Vorsicht walten lassen. Dasselbe gilt natürlich auch für die Derivationsadjektive: eine Sache kann enojoso/-a ein Ereignis fastidioso/-a, molesto/-a, desagradable eine Person enfadado/-a, indignado/-a, de mal humor ein Geschehnis escandaloso/-a sein, wenn es gewisse Reaktionen hervorruft.
SPRACHVERGLEICH ALS TEXTVERGLEICH
75
Was den Zustand des Ärgerlich-Seins betrifft, so besitzen wir für das Neuspanische eine Reihe von Bezeichnungsmöglichkeiten, die ebenfalls in den Wörterbüchern nur undeutlich auseinandergehalten werden: für die Bezeichnung des Zustands werden estar enfadado con, estar resentido con gebraucht; der Wechsel von einem Zustand der weitgehenden Emotionslosigkeit zum Zustand des Ärgers wird durch enfadar a, enojar, irritar, escandalizar, vulg. joder (Beispiele: me fastidia, me molesta, etc.) ausgedrückt, doch damit sind die Ausdrucksmöglichkeiten noch längst nicht erfaßt; beim ätiologischen Aspekt ist zu unterscheiden zwischen (a) dem, was eine Wirkung hervorruft, und (b) dem, was den Anlaß eines Ärger nisses bilden kann. Hier verfügen wir nach Auskunft der Lexika über folgende Möglichkeiten: (a) (b)
enfadarse, enojarse, irritarse hacer rabiar, sofocar a alguien,
aber es weiß jeder, der auf spanisch ein emotionales Gespräch geführt hat, daß damit die Möglichkeiten noch längst nicht erschöpft sind. Doch sei mit diesem Inventar der generelle Hinweis verbunden, daß ohne Textuntersuchungen sowohl die Liste wie auch die semantische Beschreibung vorläufig bleiben. Was das Sprachenpaar Deutsch-Französisch betrifft, so bestehen hier dieselben Schwierigkeiten, obwohl wir von einem besseren Stand der zweisprachigen Lexikographie und einem deutlich besseren Niveau der einsprachigen Lexikographie ausgehen dürfen. Ich möchte das gegebene Korrelationsverhältnis ausgehend von Weigands Darstellung auf der Grund lage von Sachs & Villatte (4896; 21979) und Weis & Mattutat (1967) sowie den verschiedenen Robert-Versionen, speziell dem Grand Robert (GRob 2 1985) und dem Petit Robert (PRob 31993), darstellen. Auf der Basis dieser Wörterbücher läßt sich für die deutsche Ausgangs- und die französische Zielsprache folgende Wortliste ermitteln: Ärger, Ärgerlichkeit, Verärgerung Empörung, Entrüstung Grimm, Groll Raserei Unwille, Verdruß, Wut, Zorn
dépit, colère, irritation, ennui, scandale, outrage, contrariété indignation, révolte colère, humeur massacrante, courroux rage, colère folle, frénésie, fureur indignation, dépit, colère, chagrin, fureur, furie, rage
76
CHRISTIAN SCHMITT
ärgerlich, aufbrausend, aufgebracht empörend, empört, entrüstet, erzürnt
irrité, fâche, contrarié, ennuyé indignant, révoltant, révolté, outré, scandalisé, scandaleux /unwirsch* bourru, rébarbatif, hargneux, renfrogné grimmig, verärgert furieux, furibond, rigoureux, scandalisé verdrießlich, verdrossen, wütend, zornig; ennuyeux, fastidieux, désagréable, fâcheux, maussade, chagrin, dégouté, dépité, furibond, las/se ärgern, aufbrausen, empören, entrüsten, être/mettre en colère, irriter, agacer, rager de erzürnen, grollen, verdrießen, wurmen, zürnen voir/penser, cela contrarie, s'irriter, se fâcher, se mettre en colère, (s')indigner, être indigné/outré/révolté/ scandalisé, garder rancune à, en vouloir à, gronder, contrarier, tracasser, tourmenter, ronger, chiffonner, turlupiner
Daraus läßt sich - stets unter dem Vorbehalt, daß die lexikographischen Daten abgesichert sind - folgende Gradation bilden: Empörung Entrüstung
indignation révolte
[moral] [retained] Groll Grimm
↑ ↓
Raserei Wut Zorn
rage, frénésie fureur, furie colère
Ärger, Ärgerlichkeit, Verärgerung ↑ [intensified] dépit, colère, irritation ↓ [weakened]
courroux rancœur
Verdruß Unwille
dépit chagrin
Figur 1: Gradation ÄRGER
Betrachtet man Arger, Àrgerlichkeit, Verärgerung unter dem Aspekt der Befindlichkeit, legen die Wörterbücher folgende Äquivalenzen nahe: BEFINDLICHKEIT (BE) Ärger, Ärgerlichkeit, Verärgerung ärgerlich/verärgert sein über sich ärgern über Ärger haben, empfinden viel Ärger erleben es gibt Ärger ärgerliche Töne seinen Ärger auslassen ein verärgertes Gesicht machen seinem Ärger Luft machen
dépit, colère, irritation, ennui être irrité, fâché, contrarié cela m'irrite, me met en colère avoir des ennuis, des désagréments, des embêtements avoir beaucoup de tracas, tintouin il y a des ennuis d'un ton irrité, fâché déverser, décharger sa bile/colère avoir l'air irrité se fâcher tout rouge
SPRACHVERGLEICH ALS TEXTVERGLEICH
77
Wenn der inkohative Aspekt im Vordergrund steht, ergeben sich die folgen den Entsprechungen: WERDEN (BECOME) ärgerlich werden in Ärger geraten Ärger bekommen
s'irriter, se fâcher se mettre en colère, cela m'irrite, s'offusquer, se scandaliser de avoir des ennuis
Das Abnehmen des Ärgers wird wie folgt ausgedrückt: ABSCHWÄCHUNG (LOSE) der Ärger verfliegt, vergeht, läßt nach sich Ärger ersparen seinen Ärger unterdrücken den Ärger herunterschlucken
le dépit, l'irritation, la colère disparaît, a disparu s'éviter des ennuis réprimer son dépit ravaler son dépit, sa colère
Die Verursachung kennt folgende Korrelate: GRUND (CAUSE) ärgern, verärgern ärgerlich machen ein Ärgernis sein Ärgernis bereiten/erregen in Ärger versetzen ärgerliche Situation das Ärgerliche an der Sache ist ärgerlich groß (um j-n ärgerlich zu machen)
mettre qn. En colère, irriter qn., agacer qn. mettre en colère, en rage être un scandale, un outrage faire scandale fâcher, agacer incident fâcheux, situation fâcheuse/em bêtante/ennuyeuse, c'est bête le fâcheux de l'affaire c'est douloureusement important/grand pour faire des ennuis
Ärger als Motiv für Handlung entsprechen die folgenden Äquivalente: HANDLUNG AUS ÄRGER (EMOTION AS MOTIVATING CONCEPT) außer sich vor Ärger rager de grün und gelb vor Ärger se fâcher tout rouge, en attraper une jaunisse aus Ärger etwas tun faire qc par colère zum Ärger aller à l'irritation de tout le monde bei aller Ärgerlichkeit malgré toutes les difficultés/tous les inconvénients voller Ärger plein de colère, d'irritation
Für die Graduierung lassen sich folgende Zuordnungen vornehmen, wobei wiederum von Weigands ("The Vocabulary of Emotion", in diesem Band) Konzept ausgegangen wird:
78
CHRISTIAN SCHMITT
GRADUIERUNG (GRADING EMOTIONS) leiser Ärger heftiger Ärger großer Ärger viel Ärger/wenig Ärger · sich sehr ärgern sich schwarz/grün und blau/gelb und grün ärgern sich maßlos/unglaublich ärgern sich furchtbar ärgern sich zu Tode/krank/ins Grab ärgern sich über die Fliege an der Wand ärgern
irritation légère colère violente grande colère beaucoup d'irritation/peu d'irritation en prendre une grande colère se fâcher tout rouge, en attraper une jaunisse se mettre en colère outre mesure, démesurément se fâcher/se mettre en colère terriblement, rager d'avoir fait je pourrais en faire une maladie, souffrir à mort de qc se mettre en colère pour un rien
INTENSIFIED Zorn, Wut, Raserei GRADUIERUNG (GRADING) aufflammender Zorn heller, lodernder Zorn leidenschaftlicher Zorn heiliger Zorn gerechter Zorn dumpfe Wut wilde Wut Zornanfall aufbrausend vor Zorn (Theater machen)
colère, fureur, furie, rage, frénésie accès de colère, colère naissante, colère qui s'éveille/point, emportement bouillant de colère colère passionnée sainte colère juste colère colère sourde colère sauvage accès de colère emporté par la colère faire du cinéma
Für den Zustand des Zorns und der Wut lassen sich folgende Entsprechungen nachweisen: BEFINDLICHKEIT (BE) zürnen, zornig/erzürnt/wütend/ wutentbrannt/aufgebracht/ aufbrausend sein Zorn haben auf von Zorn/Wut erfüllt sein von Zorn ergriffen sein sein Zorn richtet sich gegen
être en colère, se laisser aller à la colère, s'emporter, s'enrager, avoir l'accès d'orage, bouillir de colère être enragé contre, avoir une colère envers être plein de colère, être emporté par la colère, être envahi par la colère, se laisser dominer par la colère être saisi par la colère sa colère est portée sur, se porte contre/sur
SPRACHVERGLEICH ALS TEXTVERGLEICH
der Zorn der Götter trifft ihn Wut im Bauch haben wütende, zornige Stimme
79
la colère des Dieuxl'touché/atteint, féru avoir la colère aux tripes voix pleine de colère
Entsprechendes gilt für den inkohativen Aspekt, die Abschwächung und die Verursachung: WERDEN (BECOME) sich erzürnen in Zorn geraten, in Wut kommen der Zorn packt j-n /leicht in Zorn geraten*
se mettre en colère, piquer une colère se laisser emporter par la colère, se laisser aller à la colère la colère prend possession, envahit, saisit qn avoir la tête près du bonnet
ABSCHWÄCHUNG (LOSE) Zorn/Wut legt sich/läßt nach
la colère s'apaise
GRUND (CAUSE) jms Zorn/Wut erregen zornig/wütend machen, erzürnen
provoquer la colère de qn., mettre en colère faire enrager qn, faire monter la moutarde au nez de qn
HANDLUNG AUS ÄRGER (MOTIVATING CONCEPT) rot vor Zorn rouge de colère voll Zorn, voller Wut plein de colère schäumend vor Wut bouillant de colère außer sich vor Wut emporté par la colère vor Wut kochen bouillir de colère
Bescheidener präsentiert sich das prädikative Feld bei den abge schwächten und zurückgehaltenen Formen von ÄRGER wie Verdruß, Unwille und Groll, Grimm: WEAKENED Verdruß, Unwille BEFINDLICHKEIT (BE) Verdruß haben verdrossen, ungehalten sein Unwillen erregendes Benehmen
dépit, chagrin, ennui, contrariété, déplaisir, mécontentement, déboire être contrarié, être fâché, être vexé, être ennuyé être dépité, fâché, maussade, renfrogné un comportement désagréable, inconvenant
GRUND (CAUSE) Verdruß machen, bringen, bereiten, verdrießen causer des ennuis, chagriner
80
CHRISTIAN SCHMITT
HANDLUNG AUS ÄRGER (MOTIVATING CONCEPT) zum Verdruß en dépit de qn
SUPPRESSED/CURBED Groll, Grimm Groll empfinden etwas wurmt GRUND (CAUSE) j-s Groll erregen
colère (persistante), humeur massacrante, courroux être de mauvaise humeur, éprouver des courroux qc tracasse, tourmente, ronge, cela me ronge, me tracasse provoquer des courroux de qn, provoquer la rancœur de qn
MORAL Entrüstung, Empörung ≠ Empörung, Aufstand BEFINDLICHKEIT (BE) empört, entrüstet sein über Empörung erfüllt j-η, voller Empörung sein seiner Empörung Luft machen ein entrüstetes Gesicht WERDEN (BECOME) sich entrüsten, sich empören ein Sturm der Entrüstung bricht los GRUND (CAUSE) etwas entrüstet/empört j-n Entrüstung, Empörung erregen ein empörendes Benehmen
indignation, révolte
être indigné l'indignation s'empare de qn, plein d'indignation laisser sortir son indignation, donner libre cours à son indignation un visage indigné s'indigner une indignation générale se lève, une vague d'indignation qc indigne, scandalise provoquer l'indignation, un scandale un comportement scandaleux
Diese Zusammenstellungen sind eher geeignet, eine Lösung schwieriger lexikographischer Fragen vorzugaukeln, als tatsächliche Hilfen für das sprachliche Handeln zu bieten. Wer sich entsprechend den Einsichten der Korpuslinguistik daran macht und etwa mit der CD-ROM des GRob über prüft, wie vollständig bzw. wie brauchbar diese Auflistung ist, wird schnell ernüchtert durch die Tatsache, daß hier fiir das Französische ca. 300 Einheiten fehlen wie basta, grognon, hérisser, houspiller, importuner,
SPRACHVERGLEICH ALS TEXTVERGLEICH
81
incommodité, jacqerie, marri, etc., die auch nicht in den deutsch französischen Korrelaten auftauchen. Dieselben Defizite betreffen auch das Deutsche, so daß man gut beraten ist, keine Auskünfte über die semantische Feinabstimmung und erst recht nicht über die Pragmatik in diesen Wörterbüchern zu suchen und auch nicht das Französische neben das Deutsche oder das Deutsche neben das Französische zu stellen, es sei denn, man wollte eine Wörterbuchkritik leisten, um die es hier aber nicht geht, da nicht Leistungen bewertet, sondern Sprachen kontrastiert werden sollen (Westheide 1997). 4.
Äquivalente Textsegmente als Grundlage für den Sprachvergleich Solange Wörterbücher nur nach begrifflichen Systemen aufgeschlüsselt und zum sprachlichen Handeln in einer Zielsprache kontrastiv nebeneinander gestellt und korreliert werden, werden in erster Linie Inventare verglichen und keine Äquivalenzregeln mit pragmatischer Zielsetzung erarbeitet. Die pragmatische Analyse darf nicht, ja kann nicht aus Wörterbüchern gewonnen werden, deren Konstituierung weder handlungstheoretischen Anforderungen gefolgt ist noch in ihrer Ausrichtung solchen Ansprüchen gerecht zu werden sucht (Westheide 1997). Daher kann ein Sprachvergleich, wie er oben ansatzweise für das Spanische in bezug auf die wichtigsten Wörterbücher für das Französische geleistet wurde, auch nie Wörterbücher zum Ausgangspunkt wählen, sondern muß sich auf Texte beziehen, aus denen sich dann die der jeweiligen Verwendung zugrundeliegenden pragmatischen Regeln erhellen. Dabei muß - bei einer sprachenpaarbezogenen Analyse - jede der beiden Sprachen sowohl als Ausgangs- wie als Zielsprache figurieren. Grundlage darf nicht die Intuition des Sprachwissenschaftlers sein, vielmehr müssen sich die Analysen auf pragmatisch äquivalente Paralleltexte oder - je nach Fall auf pragmatisch äquivalente Paralleltextsegmente beziehen, worunter ich professionelle Übersetzungen zähle, wenn diese nicht für den Sprachvergleich hergestellt wurden. So lassen sich die usuellen Entsprechungen ermitteln, die dann - unter vergleichbaren Vorgaben - durchaus noch an Primärtexten ohne translatorisches Korrelat überprüft werden können. Durch eine solche Studie, die sicher aufwendiger ist als der Lexikonvergleich, lassen sich neben den erwartbaren Äquivalenzen auch Aussagen zur Frequenz machen, die nicht auf zufälligen Korrelaten in den Usus kaum repräsentierenden Wörterbüchern basieren. Hier werden auch die üblichen Kollokationen, die frequentesten Syntagmen und die pragmatisch einander am meisten entsprechenden Einheiten auftreten, da der Zieltext bei versierten Übersetzern, die ohnehin
82
CHRISTIAN SCHMITT
nur in die jeweilige Muttersprache übersetzen, jeweils in der Hermeneutik ganz oben steht und dem materiellen Aspekt, i.e. der Entsprechung oder Übertragung von Formen und Formelementen, jeweils nur eine nachgeord nete Bedeutung zukommt. Dabei ist grundsätzlich vom Text auszugehen, d.h. eine pragmatische Konzeption vom Begriffsfeld muß die Analyse der Texte bestimmen, wobei der analysierende Sprachwissenschaftler Interferenzformen - sofern diese überhaupt auftreten - ausscheiden sollte. Die wichtigste Frage wird dabei nicht lauten, ob die in den Wörterbüchern auftretenden Korrela tionen durch das Paralleltextstudium bestätigt werden können, sondern wie ein sprachlich handelndes Individuum mit deutscher, französischer oder spanischer Muttersprache in identischen oder zumindest vergleichbaren Situationen agiert bzw. sprachlich handeln könnte. Was das bereits behan delte Wortfeld Ärger betrifft, so ist durchaus eine Handlungsäquivalenz (Albrecht 1990) von dt. ärgerlich! und frz. basta! denkbar, obwohl diese in den analysierten Wörterbüchern überhaupt nicht auftritt. Was wir aus dem Textvergleich gewinnen, ist, wie ich an anderer Stelle bereits ausgeführt habe, zunächst einmal pragmatisch äquivalente 'parole', vergleichbar dem einsprachigen Beleg, der auch jeweils zunächst nur als Textsegment und 'parole'-Phänomen zu bewerten und zu beschreiben ist (Schmitt 1991a; 1995; 1997a); hier ist, wie auch Rein ausführt, Kontrastive Linguistik (KL) eine von der Sprachverwendung auf die Struktur und die Pragmatik ausholende deskriptive Grammatik oder Lexikographie, und er hat Recht, wenn er betont, daß "es sich von der Methode her um 'deskriptive' Sprachwissenschaft, d.h. um beschreibende und damit im modernen, strukturalistischen Sinne auch schon erklärende Linguistik" (1983:2) handelt. Mit der Berücksichtigung der pragmatischen Dimension verlassen wir aber nicht die bewährte Methode der (beschreibenden) KL, sondern erweitern nur mit der sprachenpaarbezogenen Analyse die Anzahl der Fragestellungen, die wir mit dem Textbeftmd verbinden, bzw. versuchen, aus dem Textbefund nicht nur strukturale, sondern auch handlungstheoretische Aussagen abzuleiten, die sich selbstverständlich nicht auf einen einzigen 'parole'-Akt beziehen dürfen, sondern - parallel zu de Saussures aus der Summe von 'parole'-Phänomenen gewonnener Abstraktion der 'langue' - auf einer aus der Summe von Paral leltextsegmenten gewonnenen sprachenpaarbezogenen ' langue ' -Konzeption basieren. Dabei geht es allerdings nicht um Okkurrenzen, sondern in erster Linie um die Pragmatik systemhaft auftretender Äquivalenzverhältnisse und um die generelle Zuordnung pragmatisch gleichwertiger Textsegmente, die nur auf der Grundlage möglichst vieler Translate ermittelt werden können. Es kann kein Zweifel bestehen, daß der Weg über Paralleltextsegmente,
SPRACHVERGLEICH ALS TEXTVERGLEICH
83
ausgehend von der 'parole'-Analyse in Richtung (Teil-)Systemvergleich, trotz aller Schwierigkeiten, die die Materialerfassung bereiten kann, die beste Methode des Sprachvergleiches bietet, auch und gerade beim Wortschatz. Denn für die kontrastive Lexikographie gilt, daß ein Äquivalent zu einem ausgangssprachlichen Zeichen dann gegeben ist, wenn sich in einer Reihe von gleichwertigen sprachlichen Handlungen jeweils ein sprachliches Zeichen rekurrent in der Ausgangs- wie in der Zielsprache auf ein identisches Welt stück bezieht. Gleiches gilt natürlich auch für Abstrakta oder die Einstellung markierende Partikeln (Masi 1996) und redebewertende und redesteuernde Elemente im Text (Beerbom 1992), und selbstverständlich folgt der Grammatikvergleich, der hier nicht Gegenstand ist, denselben methodischen Vorgaben, wie dies z.B. Cartagena & Gauger (1989) gezeigt haben; hier wie beim Wortschatz - müssen die zu vergleichenden einzelsprachlichen Phänomene "zunächst unabhängig voneinander nach der gleichen Methode, vor dem gleichen theoretischen und terminologischen Hintergrund beschrie ben werden, bevor ein sinnvoller Sprachvergleich erfolgen" (Beerbom 1992:90) kann. Für den Wortfeldvergleich bedeutet dies konkret, daß bei einer kontrastiven Analyse vorzugehen ist wie bei jedem anderen gramma tischen Vergleich: der auf möglichst vielen Einzelbeispielen basierenden 'parole'-Beschreibung folgt die idealiter auf der Summen-'parole' beruhende Regelfestlegung, und dieser erst kann sich ein begründeter Sprachenvergleich anschließen, wobei grundsätzlich zu postulieren ist, daß die Ebenen des Systems stets aus der normativen Realisierung evtl. unter zusätzlicher Zuhilfenahme der eigenen (muttersprachlichen) Sprachkompetenz abstrahiert werden (Schmitt 1991b). Äquivalenz gibt es bekanntlich nur in je gegebenen Texten; Äquivalenz ist ohne Einschränkung eine textlinguistische Größe. Die Probabilität ausgangs- und zielsprachlicher Äquivalenz von sprachlichen Zeichen nimmt bei kommunikativ gleichwertigen Ausgangs- und Zieltexten (Schmitt 1991b:50) zu, da hier die berechtigte Annahme besteht, daß eine recht hohe Zahl von bedeutungsäquivalenten Zeichen und Zeichenkom binationen auftritt, die jeweils derselben Funktion dienen, weil sie mit derselben kommunikativen Absicht aktualisiert wurden (Schmitt 199la:232233). Für den Wortfeldvergleich dt. GEFÜHLE/sp. EMOCIONES/frz. EMOTIONS, der rudimentär auf der Basis ein- und zweisprachiger Wörterbücher andis kutiert wurde, würde angesichts der Differenziertheit des Feldes allein die Materialerfassung mehrere Jahre in Anspruch nehmen; auch die Kontra stierung eines Teilfeldes - etwa das bereits behandelte dt. ÄRGER/engl. ANGER - kann nur mit hohem Textaufwand geleistet werden. Eine solche
84
CHRISTIAN SCHMITT
Kontrastierung kann daher - materiell wie zeitlich - auch nicht ansatzweise in einem Referat geleistet werden. Zur Demonstration der Methode wähle ich daher für das Französische wie das Spanische einen Bereich mit geschlos sener Liste aus: den Morphemvergleich (vgl. auch Schmitt 1995). Dabei bin ich davon überzeugt, daß die gewählte Vorgehensweise sich auf Wortfelder wie auf Teilfelder problemlos übertragen läßt. Entscheidend ist nicht die Durchführung bei einem möglichst weiten Wortfeld, sondern die Grund überzeugung, von der ich mich leiten lasse: Sprachvergleich ist Textver gleich. Wegen des oft vulgären Sprachniveaus und immer noch in Wörterbüchern vorherrschender tabuistischer Einschränkungsregeln sucht man nach französischen Äquivalenzen für die deutschen intensivierenden Präfixoide Scheiß-, Dreck(s)- und Mist- in zweisprachigen Wörterbüchern meist vergebens; wo solche Informationen für das gewählte Sprachenpaar vorhanden sind, bleiben sie meist bescheiden und beschränken sich auf wenige lexikalisierte Beispiele. Die in meinem Korpus nachgewiesenen Präfixoide (vgl. Erben 31993:130ff.) ersetzen grundsätzlich negativ-wertende Adjektive einer neutralen Aussage. Der DUDEN definiert das Erstglied Scheiß- als präfixoides Bestimmungsglied in substantivischen und adjektivischen Bildungen, "das auf derb emotionale Weise Abneigung, Ärger o.ä. ausdrückt" (1976-81, 1scheiß-, Scheiß-), so z.B. in den Bildungen Scheißarbeit, -beruf. Der Bezug zum Substantiv Scheiße bestimmt die Restriktion des Gebrauchs auf die derbe Umgangssprache, erklärt aber nicht mehr die semantische Struktur der Bildungen (Kühnhold 1978:206); das Erstglied hat eine Bedeutungsver schiebung erfahren, es hat die Funktion der Intensivierung übernommen: Das "Kraftwort verleiht dem Ganzen eine sehr verächtliche Bedeutung" (Küpper 1982-4:s.v. Scheiß-, scheiß-). Das offenbar in der Jugendsprache frequente Morphem läßt sich in erster Linie in der gesprochenen Sprache belegen, doch gestattet die besondere Frequenz des 1981 ins Französische (MCF 1981) übersetzten Christiane F., Wir Kinder vom Bahnhof Zoo (CFZ 1979) zu dokumentieren, wie hier methodisch vorzugehen ist, wobei wohl nicht betont werden muß, daß die Studie sich über möglichst viele Textpaare und Textsorten erstrecken sollte. Hier tritt das Präfixoid in Personenbezeich nungen (z.B. Spießer), Sachbezeichnungen (z.B. Droge) und Bezeichnungen für Institutionen (z.B. Gesellschaft) auf: Nur daß man in ihren Gesichtern jetzt noch viel deutlicher sah, was für ekelhafte Spießer das waren. Ich stellte mir vor, daß diese fetten
SPRACHVERGLEICH ALS TEXTVERGLEICH
85
Spießer jetzt aus irgendeiner Scheißkneipe oder von irgendeiner Scheißarbeit kamen. Dann gingen diese Schweinsgesichter ins Bett und dann wieder zur Arbeit und dann sahen sie fern. Ich dachte: "Du kannst froh sein, daß du anders bist. [...] Daß du jetzt auf Pille bist und den Durchblick hast und siehst, was für Scheißspießer in der U-Bahn sind" (CFZ 1979:53). C'est juste que je les vois mieux, que je me rends mieux compte à quel point ils ont l'air dégueulasse, ces bourgeois. Ils doivent rentrer de leur saleté de travail. Après, ils vont regarder la télé, aller au pieu, et remettre ça, métro-boulot-dodo. Je pense: "Tu as bien de la chance de ne pas être comme eux. [...] D'avoir pris ce truc qui te montre ce qui se passe dans le métro. Quelles têtes ils ont!" (MCF 1981:63). In diesen drei Belegen liegt intensivierende Bedeutung vor, wobei das Präfixoid negativ-wertende Adjektive ersetzt: es ließe sich auch herunter gekommene Kneipe, mühselige Arbeit oder verhaßte Spießer sagen, doch leistet das bewußt gewählte Präfixoid, daß die subjektiven Gefühle der Protagonistin in derber Form Ablehnung und Verachtung zum Ausdruck bringen. Beim Translat fällt auf, daß Scheißkneipe keine Entsprechung hat, bei saleté de travail bereits das affektgeladene saleté die Äquivalenz darstellt, während bei Scheißspießer ein Ausrufesatz vorliegt (mit bourgeois als pronominalisiertem Subjekt). Ich hatte keine Angst. Ich wußte, daß sie eine Vierzehnjährige nicht in den Knast stecken konnten. Ich war nur voll sauer auf die ScheißZivilbullen (CFZ 1979:200). Je n'ai pas peur. Je sais bien qu'ils ne peuvent pas mettre en taule une gamine de quatorze ans. Mais quels salauds, ces flics en civili (MCF 1981:196). Salopp abwertendes Bulle (DUDEN 1976-81:s.ν.) wird hier emotional verstärkt, wobei der stark umgangssprachlich gefärbte Kontext noch zur Intensivierung beiträgt; da bei flic gilt, daß "le terme est aujourd'hui employé par les policiers eux-mêmes" (GRob 1985:s.v. flic), wird die Äquivalenz durch die starke Beschimpfung salaud geleistet. Mich störte das nicht mehr weiter, denn ich wollte ja nur von meiner Clique anerkannt werden, in der es eben den Scheißkampf und Scheißkrampf nicht gab (CFZ 1979:63). Au fond ça ne me dérange plus guère, car tout ce qui m'importe désormais c'est d'être reconnue par ceux de la bande. Et dans la bande, toute cette merde, la compétition, les stress, etc., ça n'existe plus (MCF 1981:76).
86
CHRISTIAN SCHMITT
Das auf derbe Weise ablehnende, mißbilligende Haltung ausdrückende Präfixoid kennt als Äquivalenz das ebenfalls vulgäre, kataphorisch gebrauchte merde; auch hier liegt eine gelungene Wiedergabe vor, die mehr als nur ein parole-Phänomen darstellt. "Willst du denn nicht damit aufhören?" Ich sagte: "Mutti, ich täte nichts lieber als das. Ehrlich. [...] Ich will wirklich von diesem Scheißdreck weg" (CFZ 1979:143). "Tu ne veux pas arrêter?" Je lui réponds: "Maman, c'est mon plus cher désir. Sincèrement. [...] Je veux vraiment me sortir de cette merde" (MCF 1981:166). Über meinem Bett hing ein Poster. Da war eine Skeletthand mit einer Spritze drauf. [...] Stundenlang starrte ich auf dieses Scheiß-Poster und wurde beinahe wahnsinnig (CFZ 1979:204). Au-dessus de mon lit il y a un poster: une main de squelette barrée d'une seringue. [...] Je passe des heures à fixer cette saleté de poster, j'en deviens presque dingue (MCF 1981:200). Sie redeten auch über H, über Heroin. Alle waren sich einig, daß das eine Scheißdroge sei, daß man sich auch gleich eine Kugel in den Kopf schießen könne [...] (CFZ 1979:68). Ils parlent aussi de l'Η, de l'héroïne. Ils sont tous d'accord pour dire que c'est une saloperie, qu'on ferait mieux de se tirer une balle dans la tête [...] (MCF 1981:81). Das fängt doch erst mal damit an, daß die Menschen lernen, miteinander umzugehen. Das sollten wir an dieser Scheißschule erst mal lernen. Daß der eine irgendein Interesse für den andern hat (CFZ 1979:62). C'est d'abord apprendre aux gens à vivre avec les autres. Voilà ce qu'on devrait nous apprendre dans cette foutue école. A s'intéresser les uns aux autres (MCF 1981:75). Als ich noch nicht auf H gewesen war, hatte ich vor allem Angst gehabt. Vor meinem Vater, [...], vor der Scheiß-Schule und den Lehrern, vor den Hauswarten, [...] (CFZ 1979:195). Avant, j'avais peur de tout. De mon père, [...], de cette saloperie d'école et des profs, des gardiens d'immeubles, [...] (MCF 1981:190). Aber ich dachte dann, die von der RAF hätten vielleicht doch den Durchblick. Man könne diese Scheißgesellschaft nur mit Gewalt ändern (CFZ 1979:275).
SPRACHVERGLEICH ALS TEXTVERGLEICH
87
Je me disais cependant que ces gens de la bande à Baader faisaient peutêtre la bonne analyse de la situation. Qu'on ne peut changer cette société pourrie que par la violence (MCF 1981:277). Ich bekam auf einmal eine unbändige Wut auf die Polizei und unsere Regierung. Ich kam mir völlig alleingelassen vor. [...] Plötzlich hörte ich mich laut "Scheiß-Staat" sagen (CFZ 1979:245). Je me suis tout à coup sentie abandonnée de tous, envahie d'une rage folle contre la police et le gouvernement. [...] Tout à coup je me suis entendue prononcer, à haute voix, "les salauds" (MCF 1981:240f.). In der U-Bahn hätte ich heulen können vor Wut, daß ich mich mit Kakao und Kuchen und ekelhafter Scheißfreundlichkeit von einer Bullenbraut hatte ablinken lassen (CFZ 1979:221). Je me suis fait avoir par cette flicaille, son chocolat, son gâteau et ses sourires. J'en pleurerais de rage (MCF 1981:215). Mit diesen Beispielen bildet sich bereits - obwohl alle aus einem einzigen Translat stammen - ein gewisser Schwerpunkt bei den zielsprachlichen Äquivalenzen aus: die Familie von salaud, sale, saloperie, während merde und foutu nur in zweiter Linie ins Spiel kommen und bei sourire keine Äquivalenz vorliegt, so daß sich der Übersetzer gezwungen sieht, mit dem Pejorativ von flicaille und dem umgangssprachlichen se faire avoir wenig stens für einen konnotierten Kontext zu sorgen. Abgesehen von seiner Verwendung in bezug auf ein kleines Kind (Dreckspatz), wo noch etwas Zärtlichkeit zum Ausdruck kommen kann (Küppers 1982-4:s.v.), bewirkt das in zweisprachigen Lexika ebenfalls vernachlässigte, umgangssprachlich wie salopp abwertend gebrauchte Dreck(s)-, das in zahlreichen Kraftworten und umgangssprachlichen Redens arten Bestandteil bildet, eine negativ-wertende, von dem auch als freies Morphem auftretenden Erstglied nicht ganz losgelöste Bedeutungsverände rung. Auch hier seien demselben Text zwei Beispiele entnommen: Ich sah, wie sie von den Türken in eine Pension abgeschleppt wurden, und wartete draußen stundenlang, bis sie endlich mit der Drecksarbeit fertig waren (CFZ 1979:293). J'ai vu les Turcs les traîner dans un hôtel. On attend de les voir ressortir après leur sale boulot; ça dure des heures (MCF 1981:297). Als ich das hörte, rastete ich aus: "Du bist doch ein ganz mieser Dreckskerl. [...] Aber du bist eben doch nur eine miese schwule Sau" (CFZ 1979:299).
88
CHRISTIAN SCHMITT
En entendant ça, j'éclate: "Tu es vraiment un salaud, un dégueulasse. [...] Mais tu n'es qu'un sale pédé" (MCF 1981:303-304). Die Translate zeigen, daß Dreck(s)- etwas weniger derb ist, dokumentieren jedoch durch dégueulasse und sale, daß zwischen Scheiß- und Dreck(s)offene Übergänge bestehen. Ähnlich präsentiert sich auch die Suche nach Äquivalenzen zum weniger derben, eher intensivierenden Präfixoid Mist-, das zur abwertenden Bezeichnung för Wertloses, Nichtiges, Unsinniges (Küppers 1982-4:s.v.) verwendet wird und nur noch eine lose Verbindung zum homonymen freien Morphem kennt: Ich sagte: "Du bis das letzte Miststück. Ich will nie wieder was mit dir zu tun haben" (CFZ 1979:294). Je lui lance: "Tu es une salope. Je ne t'adresse plus la parole" (MCF 1981:298). Ich sagte mir: "Die können dich mal am Arsch lecken. Dein ganzes Leben haben sie sich nicht um dich gekümmert. ... Und diese Mist stücke von Eltern glauben nun plötzlich zu wissen, was für dich gut ist" (CFZ 1979:240). Je me dis: "Eh bien dorénavant tu les emmerdes. Ils ne se sont jamais occupés de toi, ... Tes salauds de parents n'ont pas levé le petit doigt, et tout à coup ils s'imaginent savoir ce qui est bon pour toi" (MCF 1981:234f.). Wir mußten uns gegen einen VW lehnen, die Arme hoch und wurden nach Waffen durchsucht, obwohl keiner von uns älter als sechzehn war. Dieser Mistkerl von Bulle ging mir dabei auch noch an die Titten (CFZ 1979:286). On nous fait mettre bras levés contre une Volkswagen pour nous fouiller - des fois qu'on serait armés. Pourtant, le plus vieux d'entre nous n'a pas seize ans. Le salaud de flic en profite pour me peloter les nichons (MCF 1981:290). Auch hier bestehen Überlappungen, die nur aus dem Text heraus verständlich sind. Es zeigt sich, daß eine adäquate Beschreibung der negativen Wertungs richtung nur nach vorangehender Textanalyse möglich wird, wobei einem Wörterbucheintrag natürlich die Untersuchung zahlreicher Paralleltextseg mente der unterschiedlichsten Textsorten vorangehen muß. Als Beispiel für den deutsch-spanischen Sprachvergleich habe ich das ausschließlich Teil-Ganzes-Beziehungen realisierende Suffixoid -haltig ge wählt, das nach Kühnhold (1978:441) durch enthalten paraphrasiert werden
SPRACHVERGLEICH ALS TEXTVERGLEICH
89
kann und das Enthaltensein des physikalischen oder chemischen Basisinhalts beim Bezugsnomen bezeichnet, das in der Regel eine Stoffbezeichnung, seltener eine Individualbezeichnung darstellt. Es tritt mit besonderer Fre quenz auf allen Stufen der fachsprachlichen Kommunikation auf, wo ich auch elf basisverschiedene -haltig-Belege mit Paralleltextentsprechungen ermitteln konnte, bei insgesamt 14 spanischen Äquivalenzen, und hat hier die Funk tion, dem Bedarf an adjektivischen Qualitäts- und Relationsbezeichnungen v.a. in Fachsprachen (Wilss 1986:111) gerecht zu werden. Mit einer Ausnahme wurden die -haltig-Derivate in Gebrauchsanwei sungen und Produktbeschreibungen ermittelt; dabei wird zum Ausdruck gebracht, daß - aus bestimmten, meist explizit genannten Gründen - das Vorhandensein des Basisinhalts der -haltig-Bildung vermieden werden soll bzw. daß es sich als gefährlich erweist oder auch schädlich sein kann. Wer mit spanischen Fachsprachen etwas vertraut ist, wird kaum zögern, ausgehend von seiner Leseerfahrung in -haltig die Äquivalenz schlechthin zu sp. -(í)fero/-a (vgl. oleífero, petrolífero, caulífero, pomífero, etc.) zu er kennen; doch lassen wir die Beispiele sprechen: Keine säure- oder bleihaltigen Partikel in die Augen, auf die Haut oder an die Kleidung kommen lassen. Evítese que el ácido o las partículas de contenido de plomo entren en contacto con los ojos, la piel o las prendas de vestir (BMW AG, BMW 850, pp. 80a/b). In Industriegebieten führen Ablagerungen von Flugrost, Kalk, ölhaltigem Ruß, schwefeldioxydhaltigen Niederschlägen (saurer Regen) [...] zwangsläufig zu Lackschäden - meist beschränkt auf die horizon talen Außenflächen. En zonas industriales suelen concentrarse en las superficies exteriores horizontales las deposiciones de óxido volante, cal, hollín oleoso, precipitaciones sulfurosas (lluvia ácida) [...], que forzosamente conducen a daños de la pintura, si no se le otorgan los suficientes cuidados (BMW AG, BMW 850i, pp. 113a/b). Spiegelgläser nicht mit quarzhaltigen Reinigungsmitteln wie Polierpasten in Berührung bringen! No hay que limpiar los espejos con productos de limpieza que contengan cuarzo, como son las pastas de pulir y similares (BMW AG, BMW 850i, p. 115). Deshalb: Vermeiden Sie schadstoffhaltige Produkte! Por esto: Eviten Ustedes productos que contienen sustancias conta minantes (Stadt Bonn, Abfallplaner).
90
CHRISTIAN SCHMITT
Flüssigkeiten in Flaschen und Dosen, im besonderen kohlensäurehaltige Getränke, nicht im Gefrierraum lagern. No meter en el aparato botellas que contengan líquidos que puedan congelarse (Bosch, Tiefkühlgerät, p. 4 dt., p. 90 sp.). Ungeeignet sind sand- oder säurehaltige Putzmittel bzw. chemische Lösungsmittel. No usar nunca detergentes a base de arenilla, ácidos o productos químicos disolventes (Bosch, Tiefkühlgerät, 2 Belege, p. 11a dt., p. 97a sp.). Eventuelle Rostflecken mit einem nicht chlorhaltigen Putzmittel beseitigen. Eliminar las manchas de óxido eventualmente existentes con un producto de limpieza que no contenga cloro (Bosch, Waschmaschine, p. 53). Asbesthaltige Materialien dürfen nicht bearbeitet werden. No se deben esmerilar materiales que contengan amianto (Bosch, Winkelschleifer, p. 1). Durch Verwendung von sandhaltigen Putzmitteln beschädigen Sie den Kunststoff. Si utiliza Vd. artículos de limpieza que contengan arena, deterioraría Vd. el material de que está construida la máquina (Krups, 3 Mix, Gebrauchsanweisung). Die Verwendung von bleihaltigem Funktion der Abgasreinigung erheblich El uso de gasolina con plomo perjudica depuradora de gases de escape [...] Passat, p. 94).
Kraftstoff beeinträchtigt die [...] considerablemente la instalación (Volkswagen AG, Volkswagen
Keine bleihaltigen Lacke [...] Lösemittelfreie Transportkonservierung. Sin pinturas que contengan plomo [...] Conservación para el transporte sin disolventes (Volkswagen AG, Volkswagen Vento, p. 138). Diese Beispiele zeigen deutlich, daß die von den Wörterbüchern bevorzugten lexikalischen Möglichkeiten, -haltig-Derivate im Spanischen wiederzugeben, im Grund nicht auftauchen. Solche Regeln liefert allein der auf dem Usus basierende, Paralleltextsegmente systematisch berücksichtigende Sprachen vergleich, der nicht vom System und den gegebenen Strukturen, sondern vom Textvergleich ausgeht und damit auch wieder zur Textproduktion führt.
SPRACHVERGLEICH ALS TEXTVERGLEICH
91
5.
Ergebnisse und Perspektiven Bei der Definition komplexer wie einfacherer lexikalischer Einheiten führt damit kein Weg vorbei an der syntaktischen wie textlinguistischen Definition ihrer Verwendungsweisen; ob Gebrauchsnormen und Verwen dungsweisen usuell und konventionell sind, muß jeweils an repräsentativen Korpora in bezug auf die Ausgangs- wie die Zielsprache überprüft werden (vgl. Harras, Haß & Strauß 1991). Zweisprachige Wörterbücher, die auch Verwendung finden wollen und beim sprachlichen Handeln eine effiziente Hilfe darstellen, müssen korpus basiert, pragmatisch und kontrastiv erstellt werden; wer das Einzellexem isoliert, kann keine adäquate semantische Beschreibung geben, wird im Sprachunterricht wie beim Übersetzen unbrauchbare Wörterbücher liefern, da der Usus einer Sprache sich nach dem Wort-im-Text und nicht nach dem isolierten Wort-im-Wörterbuch orientiert. Mir ist bewußt, daß der hier vorgeschlagene Weg ein mühsamer ist, da er riesige Investitionen in die Textarbeit voraussetzt; aber die mehrsprachige Handlungskompetenz, die als Ziel des Spracherwerbs heute unumstritten ist, verlangt nach einem neuen Wörterbuchtyp, da die auf der traditionellen Komponentenanalyse basierenden Wörterbücher noch nicht einmal eine be grenzt brauchbare Grundlage für das sprachliche Handeln bieten, denn sie eignen sich eigentlich nur für den Pattern-Drill, der mit sprachlichem Handeln auf keinen Fall gleichzusetzen ist. Nachdem die Defizite des bisherigen Vorgehens erkannt sind, sollte die Zeit dafür reif sein, mit neuen, textorientierten Projekten und einer Umorientierung in der Grundlagenforschung die Voraussetzungen und Vorar beiten für derartige Wörterbücher zu schaffen, die von der sprachlichen Pragmatik bestimmt sind. Literatur Albrecht, Jörn 1990. "Invarianz, Äquivalenz, Adäquatheit". Übersetzungswissenschaft. Ergebnisse und Perspektiven. Festschrift für Wolfram Wills hg. von Reiner Arntz & Gisela Thoma, 71-81. Tübingen: Narr. Beerbom, Christiane 1992. Modalpartikeln als Übersetzungsproblem. Eine kontrastive Studie zum Sprachenpaar Deutsch-Spanisch. Frankfurt/M.: Lang. Bierwisch, Manfred 1983. "Semantische und konzeptuelle Repräsentation lexikalischer Einheiten". Untersuchungen zur Semantik hg. von Rudolf Rủžička & Wolfgang Motsch, 61-99. Berlin: Akademie-Verlag. Cartagena, Nelson & Hans-Martin Gauger 1989. Vergleichende Grammatik SpanischDeutsch. Mannheim, Wien & Zürich: Bibliographisches Institut. Christiane F., Wir Kinder vom Bahnhof Zoo. Nach Tonbandprotokollen aufgeschrieben von Kai Hermann und Horst Rieck, mit einem Vorwort von Horst E. Richter, Hamburg
92
CHRISTIAN SCHMITT
5
1979 [ = CFZ 1979]; Moi, Christiane F., 13 ans, droguée, prostituée ... Témoignages recueillis par Kai Hermann et Horst Rieck, préface de Horst-Eberhard Richter, traduit de l'allemand par Léa Marcou, Paris 1981 [= MCF 1981]. Duden 1976-1981. Das große Wörterbuch der deutschen Sprache in 6 Bänden, hg. und bearbeitet vom Wissenschaftlichen Rat und den Mitarbeitern der Dudenredaktion unter Leitung von Günther Drosdowski. Mannheim, Wien & Zürich: Dudenverlag. Erben, Johannes 31993. Einführung in die deutsche Wortbildungslehre. Berlin: Schmidt. Faber, Pamela & Catalina Jiménez 1997. "Análisis lexicológico del léxico verbal contrastado como base para la elaboración de un diccionario para traductores. Un ejemplo en los verbos de sentimiento". Wotjak 1997. 265-286. GRob 21985. Dictionnaire alphabétique et analogique de la langue française, par Paul Robert. 9 Bde. Paris: Dictionnaires Le Robert [4951-1970, 6 Bde + 1 Suppl.]. Hallig, Rudolf & Walther von Wartburg 21963. Begriffssystem als Grundlage ßr die Lexikographie. Versuch eines Ordnungsschemas. Berlin: Akademie-Verlag. Harras, Gisela 1985. "Bedeutungsangaben im Wörterbuch. Scholastische Übungen für Lingu isten oder Verwendungsregeln für Benutzer?" Textlinguistik contra Stilistik? - Wort schatz und Wörterbuch - Grammatische oder pragmatische Organisation der Rede? hg. von Walter Weiss, Herbert Ernst Wiegand & Marga Reis, 134-143. Tübingen: Niemeyer. Harras, Gisela, Ulrike Haß & Gerhard Strauß 1991. Wortbedeutungen und ihre Darstellung im Wörterbuch. Berlin & New York: de Gruyter. Hoinkes, Ulrich & Wolf Dietrich, eds. 1997. Kaleidoskop der Lexikalischen Semantik. Tübingen: Narr. Hölker, Klaus 1981. "Lexikondefinitionen und analytische Definitionen". Lexikoneinträge. Grundelemente der semantischen Struktur von Texten hg. von Wolfgang Heydrich, 1673. Hamburg: Buske. Humboldt, Wilhelm von 1907. Über die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluß auf die geistige Entwicklung des Menschengeschlechts [1830-1835]. (=Akademie-Ausgabe.) Berlin: Akademie-Verlag. Hundsnurscher, Franz & Jochen Splett 1982. Semantik der Adjektive des Deutschen. Analyse der semantischen Relationen. Opladen: Westdeutscher Verlag. Jäger, Ludwig & Sabine Plum 1989. "Probleme der Beschreibung von Gefühlswörtern im allgemeinen einsprachigen Wörterbuch". Wörterbücher. Ein internationales Handbuch zur Lexikographie hg. von Franz Josef Hausmann, Oskar Reichmann, Herbert Ernst Wiegand & Ladislav Zgusta, vol. I, 849-855. Berlin & New York: de Gruyter. Kövecses, Zoltán 1989. Emotion Concepts. New York, Berlin & Heidelberg: Springer. Kühnhold, Ingeburg 1978. Das Adjektiv. (= Schriften des Instituts ßr deutsche Sprache: Deutsche Wortbildung. Typen und Tendenzen in der Gegenwartssprache, Hauptteil 3.) Düsseldorf: Schwann. Küpper, Heinz 1982-1984. Illustriertes Lexikon der deutschen Umgangssprache (in 8 Bänden). Stuttgart: Klett. Masi, Stefania 1996. Deutsche Modalpartikeln und ihre Entsprechungen im Italienischen. Äquivalente ßr 'doch', 'ja', 'denn', 'schon' und 'wohl'. Frankfurt/M.: Lang. Müller, Bodo 1966. "Der Verlust der Sprache. Zur linguistischen Krise in der Literatur". Germanisch-Romanische Monatsschrift 47.225-243.
SPRACHVERGLEICH ALS TEXTVERGLEICH
93
1971. "Die typologische Klassifikation der romanischen Sprachen. Methode und Entwurf". Interlinguistica, Festschrift für Mario Wandruszka hg. von Karl-Richard Bausch und Hans-Martin Gauger, 242-253. Tübingen: Niemeyer. RobQ 1996. Le Robert Quotidien. Dictionnaire pratique de la langue française, direction éditoriale de Michel Legrain, conception et direction de la rédaction de Josette ReyDebove. Paris: Dictionnaires Le Robert. PRob 31993. Le Nouveau Petit Robert. Dictionnaire alphabétique et analogique de la langue française, nouvelle édition remaniée et amplifiée sous la direction de Josette ReyDebove et Alain Rey du Petit Robert, par Paul Robert. Paris: Dictionnaires Le Robert. Rein, Kurt 1983. Einführung in die Kontrastive Linguistik. Darmstadt: Wissenschaftliche Buchgesellschaft. Sachs, Karl & Cesaire Villatte 1896. Encyclopädisches Wörterbuch derfranzösischenund deutschen Sprache. Berlin: Langenscheidt. Schmitt, Christian 1979. "Mensch und Sprache. Zur Darstellung des Sprachproblems bei Jean-Paul Sartre". Romanistisches Jahrbuch 30.17-42. 1982. "Bovelles, linguiste". Charles de Bovelles en son cinquième centenaire 1479-1979 hg. von Jean-Claude Margolin, 247-263. Paris: Trédanel. 1991a. "Kontrastive Linguistik als Grundlage der Übersetzungswissenschaft. Prole gomena zu einer Übersetzungsgrammatik für das Sprachenpaar Deutsch/Französisch". Zeitschrift für französische Sprache und Literatur 101.227-241. 1991b. "Übersetzen und Kontrastive Linguistik". Neue Methoden der Sprachmittlung hg. von Christian Schmitt, 49-83. Wilhelmsfeld: Gottfried Egert Verlag. 1995. "Distanz und Nähe romanischer Sprachen: zum Beitrag des Übersetzungsver gleichs, dargestellt an den Sprachenpaaren Deutsch-Französisch/ Spanisch". Konvergenz und Divergenz in den romanischen Sprachen hg. von Wolfgang Dahmen et al., 349-380. Tübingen: Narr. 1997a. "Prinzipien, Methoden und empirische Anwendung der kontrastiven Linguistik für das Sprachenpaar Deutsch/Spanisch". Wotjak 1997. 9-30. 1997b. "Bedeutung, Pragmatik und Lexikographie". Hoinkes & Dietrich 1997. 115127. Schwarz, Manuela & Wolfram Ziegler 1996. "Emotionen in Neurolinguistik und Lexiko logie. Ein Forschungsüberblick". Lexicology 2.34-62. Schwarze, Christoph & Dieter Wunderlich 1985. Handbuch der Lexikologie. Königstein/Ts. Sinclair, John M. 1987. "Introduction". Collins COBUILD English Language Dictionary, XV-XXI. London, Glasgow & Stuttgart: Collins & Klett. Slaby, Rudolf, J., Rudolf Grossmann & Carlos Illig 41994. Wörterbuch der spanischen und deutschen Sprache. Wiesbaden: Brandstetter. Splett, Jochen 1990. "Zur Strukturierung des Wortschatzes im Rahmen eines althoch deutschen Wortfamilienwörterbuchs". Historical Lexicography of the German Language hg. von Ulrich Goebel, vol. I, 81-105. Lewiston, NY: Mellen. Walde, Alois & Johann Baptist Hofmann 31938-1954. Lateinisches Etymologisches Wörterbuch, 1. Band A-L (1938), 2. Band M-Z (1954). Heidelberg: Carl Winter's Universitätsbuchhandlung. Weigand, Edda 1993. "Word Meaning and Utterance Meaning". Journal of Pragmatics 20.253-268.
94
CHRISTIAN SCHMITT
1996. "Words and their Role in Language Use". Lexical Structures and Language Use hg. von Edda Weigand & Franz Hundsnurscher, vol. I, 151-167. Tübingen: Niemeyer. 1997. "Semantik und Pragmatik in der lexikalischen Beschreibung". Hoinkes & Dietrich 1997. 129-145. 1998. "Wortfamilien und Semantik". Lingua Germanica. Studien zur deutschen Philologie. Jochen Splett zum 60. Geburtstag hg. von Eva Schmitsdorf, Nina Hartl & Barbara Meurer, 342-350. Münster, New York etc.: Waxmann. Weis, Elisabeth 1998 (im Druck). Der Sinnbereich FREUDE/TRAURIGKEIT im Sprachen paar Deutsch-Französisch. Eine kontrastive Studie zur Textsemantik. Frankfurt/M.: Lang. Weis, Erich & Heinrich Mattutat 1967. Wörterbuch der französischen und deutschen Sprache. Deutsch-Französisch, bearbeitet von Heinrich Mattutat. Stuttgart: Klett. Westheide, Henning 1997. Contrastief idioomboek Duits. Handleiding voor doelmatig taalgebruik. Bussum: Coutinho. Wiegand, Herbert Ernst 1985. "Eine neue Auffassung der sog. lexikographischen Definition". Symposium on Lexicography II: Proceedings of the Second International Symposium on Lexicography at the University of Copenhagen hg. von Arne Zettersten & Karl Hyldgaard-Jensen, 15-100. Tübingen: Niemeyer. Wierzbicka, Anna 1972. Semantic primitives. Frankfurt: Athenäum. 1992. Semantics, Culture and Cognition. Universal human concepts in culture-specific configurations. New York & Oxford: Oxford University Press. Wilss, Wolfram 1986. Wortbildungstendenzen in der deutschen Gegenwartssprache. Theoretische Grundlagen - Beschreibung - Anwendung. Tübingen: Narr. Wolski, Werner 1980. Schlechtbestimmtheit und Vagheit - Tendenzen und Perspektiven. Methodologische Untersuchungen zur Semantik. Tübingen: Niemeyer. Wotjak, Gerd, ed. 1997. Studien zum romanisch-deutschen und innerromanischen Sprachvergleich. Frankfurt/M. etc.: Lang. Zemb, Jean M. 1978/1984. Vergleichende Grammatik Französisch-Deutsch, Teil 1 Comparaison de deux systèmes, Teil 2 L'économie de la langue et le jeu de la parole, 2 vols. Mannheim, Wien & Zürich: Bibliograhisches Institut.
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN SEMANTISCHE 'ÄRGERDÖRFER' EVI RUSSISCHEN UND IM DEUTSCHEN
VALERIJ DEM'JANKOV Moscow State University Summary A contrastive-semantic analysis of lexical items denoting ANGER and similar emotions in German and in Russian is given in the framework of a model of lexical description in which syntactic, semantic, and pragmatic properties of lexical items are exploited. Semantic village as a metalinguistic notion is a logical continuation of more traditional notions of 'word family' and 'semantic field'. Lexical items, like people, 'dwell' in certain places and have 'family resemblances' to other items. They also have 'occupations' or 'jobs', which correspond to pragmatics or word use in actual speech. Provenance of a lexical item corresponds to (lexical) derivation. Thus, a lexical item may be 'rooted' in one village, where its root(s) semantically belong(s), but dwell in another, being an emigrant, if its lexical meaning belongs to a different semantic village. Villagers are grouped in 'collective farms', which correspond to (pre)theoretical taxonomies found mostly outside linguistics, in our case, in folk-psychological and/or properly psychological theories of emotions. These organizational, or institutional, taxono mies do not necessarily coincide with linguistic classification based on actual linguistic behaviour, or.'jobs', of lexical items. When the language in question and the extra-linguistic theoretical background of the 'lexical institutions' change, both evolution and revolutions are possible, e.g., when a relatively more democratic 'lexical organization', à la German Genos senschaft, is superceded by a totalitarian system like a Soviet kolkhoz, with its utter centralization, or vice versa. In these and similar terms, semantics may be said to be an analogue of dwelling place, and pragmatics, as opposed to semantics, corresponds to both, - travelling, i.e. change of actual use of a lexical item which either 'lives' outside its usual dwelling place (analogously to what happened during the Cultural Revolution in China when intellectuals were sent to villages, mostly to be occupied in new, far less intellectual activities) or is temporarily 'occupied' outside its usual collective farm, and - relations between an actual lexical meaning of a word and its 'inner form', or between emigrants and their relatives.
96 1.
VALERIJ DEM'JANKOV
Die Metapher des Feldes in der semantischen Metasprache Metaphern werden in der linguistischen Metasprache sehr häufig gebraucht. In der iexikologischen Terminologie finden wir z.B. solche Metaphern wie 'Feld' ('semantisches Feld', 'Wortfeld' usw.), 'Familie' ('Wortfamilie', im Russischen 'Wortverwandschaft5 - 'rodstvennye slova', besonders in der Schulgrammatik), 'domain', 'province', etc. Der Begriff 'Wortfeld (Ipsen 1924, Jolies 1934, Porzig 1934, Trier 1934), obschon nicht der Terminus selbst, wurde, wie Geckeier (1971:86) zeigt, schon Mitte vorigen Jahrhunderts gebraucht und kann auf Ideen von Humboldt (1836) und Herder (1772) zurückgeführt werden. Interessante historische Angaben darüber enthalten auch Kronasser (1952) und Lyons (1977). Ähnliche Begriffe sind auch in der Geschichte der russischen Sprachwissenschaft bekannt (vgl. Pokrovskij 1895). Der Streit um die Angemessenheit der genannten Metaphern entsteht besonders oft in der lexikalischen Semantik. Als Argumente für oder gegen den Gebrauch einer Metapher gelten die Analogien (bzw. das Fehlen von Analogien) zwischen dem außermetasprachlichen und dem metasprachlichen Bereich. So handelt es sich bei der Metapher 'Feld' z.B. um ein physikalisches Bild - das Bild des Magnetfeldes, in dem Kräfte wirken und Eisenkörper angezogen werden. Entsprechend zieht das Wortfeld die Wörter zusammen. Feld bezeichnet auch eine Landschaft, in der verschiedene Pflanzen wachsen. Man kritisiert diesen Begriff (eine Systematisierung der Kritikrichtungen der Wortfeldtheorie ist in Chu [1990:11] enthalten), weil: "die semantische Facettenvielfalt des Einzelwortes in den Hintergrund trat; man wandte sich vor allem bestimmten Beziehungen zwischen den ganzheitlich konzipierten Einheiten in ausgewählten Wortverbänden zu und arbeitete vorzugsweise die halbwegs generalisierbaren Aspekte des Wortschatzes heraus" (Hundsnurscher 1995:347); diese Metapher kaum vernünftige Modifikationen gestattet, die das Ausgangsbild nicht zerstören (Herbermann 1995:288); der Begriff 'Wortfeld' es nicht gestattet, chaotisch aussehende Gruppen von Lexemen zu charakterisieren: die Metapher ist nur für gut struktu rierbare Lexemgruppen hilfreich, einerseits, und andererseits kann man nicht hoffen, eine einheitliche Darstellung des Lexikons im Ganzen als eine Organisation von Feldern zu erreichen (Ullmann 1953:227).
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
97
Eine Modernisierung des Begriffs 'Wortfeld' wurde im späteren Werk Weisgerbers angeboten (Weisgerber 1956/57:70), und zwar der Begriff 'Sinnbezirk': ein relativ selbständig erscheinender Ausschnitt aus dem sprachlichen Weltbild, wobei außer- und innersprachliche Bedingungen bei der Umgrenzung zusam menwirken. ... in jedem Sinnbezirk Ausprägungen aller genannten Formen sprachlichen Zugriffs, grammatisch gesprochen also inhaltlicher Bestimmtheit, antreffen.
2.
'Semantisches Dorf in der lexikalischen Semantik Unter anderem fehlt der Metapher des Feldes ein menschliches Gesicht. Im Zeitalter der physikalischen Erklärungen (etwa bis Anfang der achtziger Jahre) versuchte man, von dem menschlichen Faktor zu abstrahieren. Alles in der linguistischen Beschreibung mußte als etwas automatisch vor sich Gehen des ausgelegt werden. Gelang dies nicht, mußte man eine alternative Erklä rung finden. Menschen - wie auch alles, was mit ihrer Intentionalität ver bunden ist - unterscheiden sich jedoch vom rein Physikalischen durch den Willen, die Absichten. In der Wortfeldtheorie dürfen Elemente eines lexikali schen Feldes den Magnetkräften keinen oder nur geringen Widerstand lei sten. Die Menschen hingegen sind nicht so: Sie haben eigene Prädisposi tionen, auch von ihren Vorvätern geerbte Bräuche und Sitten, sogar rein indi viduelle, unikale Eigenschaften. In diesem Sinn ist der Begriff 'Familie' etwas menschlicher: Familienmitglieder sind letzten Endes menschliche Wesen, und die Metapher entspricht etwas besser dem heutigen Akzent auf dem menschlichen Handeln in allen Sphären der menschlichen Existenz. Und so dürfen auch die Wörter schon etwas menschlicher werden: sie dürfen auch individuelle und unikale Eigenschaften besitzen, was jedoch ihre Zugehörig keit zu einer Gruppe nichtverwandter Wörter nicht ausschließt. In unserem Fall geht es um eine kontrastiv-semantische Analyse von Bezeichnungen von Emotionen. Nietzsche kritisierte die Vergöttlichung der Emotionen als naiv: In derselben Weise, in der jetzt noch der ungebildete Mensch daran glaubt, der Zorn sei die Ursache davon, daß er zürnt, der Geist davon, daß er denkt, die Seele davon, daß er fühlt, kurz, so wie auch jetzt noch unbedenklich eine Masse von psychologischen Entitäten angesetzt wird, welche Ursachen sein sollen: so hat der Mensch auf einer noch naiveren Stufe eben dieselben Erschei nungen mit Hilfe von psychologischen Personal-Entitäten erklärt. Die Zustän de, die ihm fremd, hinreißend, überwältigend schienen, legte er sich als Ob-
98
VALERIJ DEM'JANKOV
session und Verzauberung unter der Macht einer Person zurecht. ... Man konkretisiert einen Zustand in eine Person: und behauptet, dieser Zustand, wenn er an uns auftritt, sei die Wirkung jener Person. Mit andern Worten: in der psychologischen Gottbildung wird ein Zustand, um Wirkung zu sein, als Ursache personifiziert (Nietzsche 1887:99-100).
Aber genau diese naive Metapher, diese Vergöttlichung oder eher die Vermenschlichung der Emotion, halte ich für produktiv bei der kontrastiv semantischen Analyse. Ich möchte etwas weiter gehen und gebrauche den Begriff 'semanti sches Dorf, um die Bilder vom 'Sinnbezirk' (oder 'conceptual domain') einerseits und von der 'Wortfamilie' andererseits zu vereinheitlichen. Dabei bleibt auch der Begriff der Wortfamilie für uns relevant, und zwar in dem Sinn, daß die Mehrzahl der Wörter in morphologischer Hinsicht komplex sind und daß die zwischen den Wortkonstituenten bestehenden Relationen sowie die darauf beruhenden Beziehungen zwischen den entsprechenden Wörtern wesent liche Strukturelemente darstellen (Splett 1993:xiii).
Die Metapher des semantischen Dorfes hilft uns, die Prozesse zu veranschaulichen, bei denen Wortderivation (teilweise auch Etymologie oder 'innere Wortform'), Semantik und Pragmatik (Wortgebrauch) in einer kontrastiven lexikalischen Beschreibung zusammenwirken. Es handelt sich hier nicht um eine neue Theorie, sondern eher um eine Phänomenologie, ein Instrumentarium zur Beschreibung sprachlicher Phänomene. Dieses Instrumentarium kann im Rahmen ganz verschiedener semantischer und/oder pragmatischer Theorien gebraucht werden, je nachdem, wie man die meta sprachlichen Metaphern interpretiert. 3.
Deutsches 'Ärgerdorf', russisches 'gnevlivka' und ihre Einwohner Als illustratives Material gebrauche ich Lexeme, die Emotionen, und zwar Ärger, Zorn, Wut usw., im Deutschen und im Russischen bezeichnen: Substantive Ärger, Ärgerlichkeit, Ärgernis, Empörung, besenstvo, dosada, gnev, isstuplenie, jarost', Entrüstung, Grimm, Groll, Raserei, Unwille, nedovol'stvo, negodovanie, neistovstvo, Verdruß, Wut, Zorn neudovletvorenie, neudovletvorennost', neudovol'stvie, ozloblenie, razdraženie, serditost', vozmuscenie, zloba, zlost'
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
99
Epitheta ärgerlich, aufbrausend, aufgebracht, empörend, empört, entrüstet, erzürnt, grimmig, verärgert, verdrießlich, verdrossen, wütend, zornig
besenyj, dosadnyj, dosadujuscij, gnevnyj, gnevlivyj, isstuplennyj, jarostnyj, nedovol'nyj, negodujuscij, neistovyj, neistovstvujuščij, neudovletvorennyj, (neudovletvoritel'nyj,) ozloblennyj, razdražennyj, raz'jarennyj, rasseržennyj, serdityj, vzbesennyj, vozmuscennyj, vozmutitel'nyj, zlobnyj, zlobstvujuščij, zloj
Prädikate ärgern, aufbrausen, empören, entrüsten, erzürnen, grollen, verdrießen, wurmen, zürnen
besit', besit'sja, besnovat'sja, dosadovat', dosaždat', gnevat'sja, gnevit', negodovat', neistovstvovat', ozlobljat', rasserdit'sja, razdrazat', razdražat'sja, razgnevat'sja, razjarit', raz'jarit'sja, serdit', serdit'sja, razozlit'sja, vozmuscat', vozmuščat'sja, voznegodovat', vzbesit'sja, zlit', zlit'sja, zlobstvovat' Tabelle 1: 'Dorfbewohner'
In einem gegebenen semantischen Dorf, in unserem Beispiel im 'Ärger dorf , gibt es recht viele Einwohner, die aus verschiedenen Wortfamilien stammen. Dabei gibt es solche Dorfbewohner, die in diesem Dorf geboren sind (Wut, Zorn, Groll, Grimm). Es gibt aber auch Aussiedler, deren 'Wur zeln' (im linguistischen Sinn des Wortes "Wurzel") in einem anderen Dorf (nicht unbedingt in einem benachbarten Dorf) zu finden sind, z.B.: Raserei, Ärger, Ärgerlichkeit, Ärgernis, Unwille. Im Unterschied zum Deutschen wohnen im russischen Ärgerdorf fast ausschließlich Aussiedler. Nur jarost' kann - rein historisch gesehen - als alter Bewohner dieses Dorfes angesehen werden: jarost': vom altruss. jar', jaryj "zornig", "unbarmherzig". Dieselbe Bedeutung hatten auch: jar'kyj, jaro "unbarmherzig"; jar', jarost' "Zorn" (im IE. bedeutete jôros wohl "brennend", "rasend", "rennend"). Ein anderer in der modernen Sprache sehr gebräuchlicher und sogar prototypischer Dorfbewohner gnev bedeutete im Altrussischen eigentlich "Eiter", "Verfaul tes", "Vermodertes" (vgl. im modernen Russischen gnil', gnoj). Die ältere Bedeutung war wahrscheinlich mit dem Zustand eines Kranken verbunden, der mit Schorf, Grind, eiternden Wunden bedeckt war. Daher die alte Bedeutung gnoj ("Eiter") im "Paremejnik" 1271. Bezüglich der Herkunft der anderen russischen Ärgerdörfer sei noch folgendes bemerkt.
100
VALERIJ DEM'JANKOV
negodovanie: synonym krajnee nedovol'stvo, vozmuscenie "Entrüs tung", "Unwille", etymologisch von negodovaV (im Altruss. seit dem 11. Jh. negodovati) "unzufrieden sein", "Unzufriedenheit äußern", auch "schwer vertragen", "leiden", "abschlagen (eine Bitte)", "verweigern"; vgl. godovati "gestatten" (Stamm god wie in ugozhdat', ugodit', godnyj). Die Bedeutung "gestatten" entwickelte sich wohl aus der Bedeutung "Vergnügen machen" (ugozdat'). Andere Bedeutungen erschienen erst später. In negodovaV ent stand die Bedeutung "Empörung empfinden" wahrscheinlich aus der Bed eutung ne ugozdat' "nicht das Vergnügen machen", woher auch "nicht gestatten" kommt. vozmuscenie: von mutit' "undurchsichtig machen", vgl. altruss. mut', mut'v' (im heutigen Russisch: volnenie, smjatenie). neistovstvo: Verneinung von istyj "wirklich", "echt", "eifrig", vgl. altruss. isto "Kapital". isstuplenie: der Zustand, in dem man "außer sich, außer seiner Existenz" ist. bešenstvo:von bes "Teufel". razdrazenie "Verdruß": von draznit' (mundartlich: dražnit') "absicht lich jemanden wegen Kleinigkeiten aus der Fassung bringen"; Epitheta: bessiVnoe "kraftlos", boleznennoe "krankhaft", bujnoe "tobend", burnoe "rasend", gluxoe "taub", zguchee "verbrennend", zelchnoe "gallig", zlobnoe "boshaft", zloe "böse", zloradnoe "schadenfroh", molcalivoe "schweigend", muäteVnoe "quälend", nevoVnoe "unwillkürlich", neizlitoe "nicht ausgegos sen", "stumm", neponjatnoe "unverständlich", nervnoelnervoznoe "nervös". nedovol'stvo, neudovol'stvie: Verneinung von dovoVnyj "zufrieden", altruss. (seit dem 11. Jh.) dov'lnyj und dovoVnyj, im modernen Russisch: dovoVnyj, dostatočnyj, udovletvorjajuščij razmerami "genügend", "von ge nügender Größe". neudovletvorenie: Verneinung von dovleV "genug sein, ausreichen"; Derivat von vol (ja) "Wille". zloba, zlost', ozloblenie "Groll", "Grimm": im Altruss. "Sünde", "Verbrechen", "das Böse", "Feindseligkeit", von zloj' (z'l' - z'lyj) "schlecht", "böse", "gemein". dosada: (im Altruss. "iniuria") von do + sadiV (vgl. "jemandem zusetzen" pristavaV, dosazdat' k.-l.) "Verdruß"; Epitheta: velikaja "groß", vidimaja "bemerkbar", vnutrennjaja "inner", gluxaja "taub", gnevnaja "zor nig", gor'kaja "bitter", grustnaja "traurig", zgucaja "verbrennend",
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
101
zestokaja "unbarmherzig", zlobstvujuscaja "wütend", izlišnjaja "überflüs sig", koljucaja "stechend", krajnjaja "extrem", ljutaja "böse", mimoletnaja "flüchtig", minutnaja "minutenlang", nevol'naja "unwillkürlich", otkrytaja "offen", revnivaja "eifersüchtig", sil'naja "stark", skrytaja "verborgen", smutnaja "unklar", strasnaja "unheimlich", tajnaja "geheim", tupaja "stumpf", uzasnaja "schrecklich". 4.
Semantische 'Beschäftigungen' Die Aussiedler können die von ihren Eltern geerbten Bräuche und Sitten mitbringen, z.T. auch 'Kleider' (Epiteta) usw. Für das Russische gilt dies aber nur begrenzt, und zwar für gnev: izlit' gnev (wörtl.) "den Zorn ergießen" kann darauf zurückgeführt werden, daß gnev früher eine Flüssig keit bedeutete ("Eiter"), die sich eines schönen Tages aus der Wunde ergießt (das russische ego prorvalo [wörtl.] "er hat es ausgegossen" kann unter ande rem auch einen plötzlichen Zornausbruch schildern). Auf diese Weise sind verschiedene Verhältnisse zwischen der Deriva tion des Wortes einerseits und seinem 'Wohnorf andererseits auszulegen. Was hier unter Wohnort gemeint ist, kann in einigen modernen linguistischen Konzeptionen als Analogie für semantische Beschaffenheiten des Wortes betrachtet werden. Ein 'Aussiedler', d.h. ein Familienmitglied, das nicht in seinem Heimatdorf, also von seinen Eltern weit entfernt wohnt, verliert aber zu anderen Familienmitgliedern die Beziehungen nicht. Seine Eltern können ihn anrufen oder einen Brief schreiben, in dem sie ihn davor warnen, sich zu weit von den angeborenen Eigenschaften zu entfernen. Dadurch daß neben den Stammbewohnern in einem Dorf auch Aussiedler wohnen, entstehen auch Familienähnlichkeiten zwischen den etymologisch sehr verschiedenen Dorfbewohnern. Aber außer der Semantik (d.h. außer dem 'Wohnort') haben Lexeme auch pragmatische Aspekte: das, was das Wort im Sprachgebrauch leistet. (NB: in dieser Redeweise geht es offensichtlich um eine Übertragung: denn es ist so, daß die Menschen die Wörter gebrauchen und etwas leisten, und nicht die Wörter selbst.) Diesen Aspekt nenne ich 'Dorfarbeiten'. Die Dorfarbeiten bezeichnen die Funktionen, die ein Dorfbewohner (z.B. Ärger, Wut, Zorn usw.) im Rahmen einer größeren Einheit hat. Eine der Arbeiten kann man mit dem Prädikat BE bezeichnen, und zwar, wenn man konstatiert, daß ein bestimmter emotionaler Zustand vorherrscht (vgl. ärgerlich/verärgert sein - byt' nedovol'nym), eine andere als BECOME (ärger lich werden - razdražat'sja/rasserdit'sja) usw. Die Zahl und der Bestand der
102
VALERIJ DEM'JANKOV
Dorfarbeiten variieren von Dorf zu Dorf und entsprechen dem, was man 'Bedeutungsposition' (Hundsnurscher & Splett 1982) oder 'meaning position' (Weigand 1996) nennt. Man kann für diese Positionen universale metasprachliche Bezeichnungen finden; diese sind es, die Weigand 'Bedeu tungsposition' nennt, d.h. universale Konzepte, die unabhängig vom Aus druck in konkreten Sprachen sind (Weigand 1996:157). Um Mißverständ nissen vorzubeugen, könnte man 'realisierte Bedeutungspositionen' und 'abstrakte Bedeutungspositionen' unterscheiden oder, um die Metapher des Dorfes zu gebrauchen, 'Dorfarbeiten' und 'Dorfbeschäftigungeri, Die Psychologen, die Emotionen definieren und voneinander abgrenzen, haben mit 'Beschäftigungen' in unserem Sinn zu tun. Bei einer empirischen kontrastiv-semantischen Analyse kann man aufgrund des vorhandenen sprachlichen Materials der zu vergleichenden Sprachen zuerst 'Dorfarbeiten' auflisten, dann aber eine Liste der 'Beschäftigungen' erstellen, als Ergebnis des Vergleichs von zwei oder mehreren Listen der 'Arbeiten'. Sobald die Listen aufgestellt sind, kann man wieder von 'Dorfarbeit' sprechen, aber dann bedeutet dieser Terminus einen Begriff, im Rahmen dessen die Fakten aller zu vergleichenden Sprachen (und nicht nur einer Sprache) beschreibbar sind. 5.
Konzept und Bedeutung im Rahmen des semantischen Dorfes Um Dorfarbeiten zu leisten, vereinigen sich die Dorfbewohner in 'Genossenschaften', die vortheoretische und theoretische (d.h. psychologi sche und philosophische) Ansichten des Forschers widerspiegeln. Unter anderem spiegeln die Genossenschaften auch verschiedene Definitionen und Klassifikationen der Emotionen wider. 'Emotionen' definiert man z.B. im Unterschied zu 'Gefühlen' als "ausschließlich dem Menschen verfügbare, mittels Sprache systematisch ausdrückbare spezifische Bedeutungen", d.h. als "theoretische Beschreibungsgrößen der Linguistik" (Wallner & Costazza 1995:40). Ryle glaubte, daß Emotion mindestens folgendes bezeichnet: 'inclinations' (oder 'motives'), 'moods', 'agitations' (oder 'commotions') und 'feelings' (Ryle 1949:83). Auf der Basis von Gesichtsausdrücken, die auf Photos abgebildet sind, unterscheidet man sechs Emotionskategorien: Freude "happiness", Überraschung "surprise", Angst "fear", Ärger "anger", Trauer "sadness" und Ekel/Verachtung "disgust/contempt" (Ekman, Friesen & Ellsworth 1982; vgl. Bänninger-Huber 1996:40).
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
103
Bei der Analyse von deutschen und russischen 'Ärgerdörfern' kommt man nach dem Modell von Weigand ("The Vocabulary of Emotion", in die sem Band) zu folgenden fünf 'Genossenschaften': Empörung, Entrüstung negodovanie, vozmuščenie
Raserei, Wut, Zorn gnev, neistovstvo, jarost', isstuplenie, besenstvo
Ärger, Ärgerlichkeit razdraženie, neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost', serditost' Groll, Grimm zloba, zlost', ozloblenie
Verdruß, Unwille dosada Figur 1: 'Genossenschaften'
Diese und ähnliche mit Emotionen verbundene metapsychologische Nomenklaturen werden weiter auch linguistisch analysiert. Oft vertritt man die Meinung, daß "emotion concepts encoded in the English lexicon constitute a coherent and reasonably self-contained (though not sharply delimited) cognitive domain, with a characteristic and specifiable type of semantic structure" (Wierzbicka 1996:178). Wenn man unsere vermenschlichende Metapher annimmt, muß man auch postulieren, daß die Emotionen in demselben Sinn kohärent (oder unkohärent) sein können, wie die Beziehungen zwischen den Leuten. Verschiedene wissenschaftliche Definitionen und Aufteilungen ent sprechen auch verschiedenen Versionen von Genossenschaften. Die Ent wicklung einer (in unserem Fall psychologischen) Konzeption entspricht einer Umgestaltung von Genossenschaften, die selbst unterschiedlich flexibel, 'demokratisch' (oder umgekehrt, 'totalitär', à la sowjetische Kolchose) sein können. Zur Zeit gibt es sehr viele Theorien der Emotion (zur Einteilung der Emotionen und insbesondere des 'Ärgerdorfes' in neuerer Literatur s. auch: Eckensberger 1996, Konstantinidou 1997, Dickie 1996, Lycan 1996, Magai & McFadden 1995, Runggaldier 1996, Harré 1994). Unter ihnen sind folgen de Ansichten für uns von besonderem Interesse (Aylwin 1985:130-133): I
Emotionen haben eine wichtige kognitive Komponente: Sie enthalten Bewertungen und sind im Grunde genommen Mittel zur Repräsentation von Objekten, Ereignissen oder Ideen als Träger von Werten oder hedonistischen Tönen (Aquin, James, etc.);
104
II
III IV
VALERIJ DEM'JANKOV
Objektiv dieselben Emotionen können auf verschiedene Weise interpretiert und sogar empfunden werden, je nachdem, welche Be nennung sie haben, welche Ursache (oder Vorgeschichte) man ihnen zuschreibt; Emotion bezieht sich auf das Bewußtsein: es geht bei den Emotionen nicht bloß um eine Motivation, Handlung, Gesichts ausdrücke usw., sondern um das Bewußtsein ("awareness") von ihnen. Deswegen haben die Emotionswörter eine Referenz zu Empfindungen ("experiences"); Emotionen lassen nichts ungeändert, sie können die ganze Welt transformieren (Heidegger, Sartre). Es gibt emotionale Primitive ("emotional primitives"). Obschon es eine große Zahl von Emotionswörtern in der Alltagssprache gibt, glaubt man, daß sie theoretisch auf eine recht kleine Anzahl reduziert werden können, und zwar: eine kleine Zahl von verschiedenen fundamentalen Emotionen, wie Freude, Ärger und Angst (Izard 1977) oder eine kleine Zahl von bipolaren fundamentalen Dimensionen, z.B. angenehm - unangenehm, Ruhe - Reiz usw. (Wundt 1902-1903) oder strukturelle Kombinationen von beiden Arten von primitiven Ele menten, z.B. Freude und Kummer sind gegenübergestellt und ver binden sich miteinander nicht, aber Liebe und Sicherheit präsupponieren einander oder einige komplexe Emotionen (wie Depression) sind Kombinationen entweder von fundamentalen Emotionen oder von Emotionen mit anderen Prozessen (Izard 1977). Emotionen tragen sozial relevante Informationen. Obschon sie komplex sind, sind die Emotionen aber zur selben Zeit einheitliche Phänomene: sie sind als 'Gestalten' organisiert, in denen physiologische, kognitive und andere Merkmale ein integrales Ganzes bilden (Izard 1977); die Beziehungen zwischen den Konstituenten in diesem Ganzen können aber auch sehr locker und individuell variabel sein, von emotionalen Systemen verschiedener Personen abhängig.
Daraus folgt, daß die 'Emotionsdörfer' unter sich in bestimmten Be ziehungen stehen und daß verschiedene menschliche Werte (angenehm, gut vs. unangenehm, schlecht) analog den Postboten die Dörfer verbinden. Verschiedene Emotionen sind (Parkinson 1995:ix-x) mit verschiedenen evaluativen Bewertungsdimensionen verbunden: so impliziert Zorn, daß es etwas gibt, was verurteilt werden muß. Das heißt, 'Emotion' ist mehr als eine Redeweise über menschliches Handeln.
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
105
Daraus folgt zum einen, daß wir uns bei der linguistischen Beschrei bung von Emotionen auf einem metapsychologischen Niveau befinden, zum anderen aber auch, daß die psychologischen Theorien mit bestimmten semantischen Positionen (d.h. 'Beschäftigungen' im obengenannten Sinn) verbunden sind: sogar ein Neuerer in der Psychologie darf beim Gebrauch der Terminologie (die in der Regel mit den Wörtern der Alltagssprache zusammenfällt) die semantischen Beschäftigungen nicht vernachlässigen. Sonst wird er von anderen nicht (oder nicht immer) richtig verstanden. Im 'Ärgerdorf' definiert man den Zustand der Wut wie folgt (vgl. Kernberg 1992/97:36): der grundlegende Affektzustand, der die Aktivierung von Aggression in der Übertragung kennzeichnet. Gereiztheit ist ein leichter aggressiver Affekt, der das Potential für Wutreaktionen anzeigt und als chronische Gestimmtheit in Form von Reizbarkeit auftritt. Ärger ist ein intensiverer Affekt als Gereiztheit und gewöhnlich differenzierter in seinem kognitiven Inhalt und in der Art der dabei aktivierten Objektbeziehung. Eine voll ausgeprägte Wutreaktion mit ihrem überwältigenden Charakter, ihrer diffusen Art und dem "Verschwim men" spezifischer kognitiver Inhalte und entsprechender Objektbeziehungen kann irrtümlicherweise zu der Vorstellung verleiten, es handele sich dabei um einen "reinen" primitiven Affekt. In der Praxis zeigt die Analyse der Wut reaktion aber - genauso wie die anderer intensiver Affektzustände - eine dahinterstehende bewußte oder unbewußte Phantasie mit einer spezifischen Beziehung zwischen einem Aspekt des Selbst und einem Aspekt eines bedeutsamen Anderen.
Man darf nur begrenzt die Meinung akzeptieren, daß Emotionen in erster Linie nonverbal ausgedrückt werden (vgl. Thomas 1996:43): The affective component is displayed primarily through non-verbal behavior, which is an extremely rich notion that ranges over voice (quality, intonation, and flow of speech), facial expression (including eye movement), body posture, and timing of behavior (words and deeds). Non-verbal behavior among human beings is ineluctably connected with the affective component, in that if human beings either lacked an affective component or it were radically different, the non-verbal behavior of human beings would also be substantially different from what it is presently.
Die Sachlage ist komplizierter und differenzierter (vgl. Lazarus-Mainka & Siebeneick 1997:10): Emotionale Inhalte sind rein privater Natur. Sie sind subjektiv und werden auch wenn man über die Art ihrer Repräsentation keine endgültigen Aussagen machen kann - mittels Sprache veröffentlicht und damit für andere nachvoll-
106
VALERIJ DEM'JANKOV
ziehbar gemacht. Sie werden in ihrer Bedeutung und Intensität in Relation zu anderem Erlebten beschrieben, das im Gedächtnis in Wissensstrukturen abge speichert ist. Mit dem Gebrauch der Sprache wird ein differenziert struk turiertes Regelsystem aktiviert, das auch die Beschreibung nicht primär sprach licher, so z.B. emotionaler Inhalte mit determiniert.
Auf die 'Codierung von Emotionen' scheinen verschiedene sprachliche Mittel spezialisiert zu sein (Grayshon 1975:27, Fries 1996:37-38); aber besonders oft werden Emotionen durch periphere Systeme ausgedrückt: prosodische Merkmale, Vokalquantität und Gebärden. Viel weniger effizient werden Emotionen mit Hilfe einer Proposition ausgedrückt, z.B. I am moderately angry with you, oder I'm entirely committed to this line of action. In der Alltagsrede findet man überhaupt kaum direkte Beschreibungen des aktuellen emotionalen Zustands in der ersten Person: Sätze wie Ich bin blaß vor Wut sind im Alltagsdialog kaum zu erwarten. Aber nicht so im Theater, besonders in den Epochen, in denen der Zuschauer aus technischen Gründen das Gesicht des Schauspielers nicht genau sehen konnte (z.B. trugen die Schauspieler Masken oder waren zu weit von den Zuschauern entfernt). Die 'metapsychologischen Bemerkungen' der handelnden Personen halfen den Zuschauern, die Handlung besser zu verstehen. Wenn man von eigenen Emotionen spricht und nicht bloß paralinguistische Mittel gebraucht, so weigert sich der Zuhörer, solche Äußerungen allzu ernst zu nehmen (Labov 1984:43), und interpretiert sie als Theater (auch im wörtlichen Sinn von 'Theater'). Das gilt besonders für Ärger/ANGER (vgl. das etwas unnatürliche Ich bin außer mir vor Wut), viel weniger für den Ausdruck der Gekränktheit (vgl. Ich bin auf Sie böse) und noch weniger für Liebeserklärungen {Ich liebe dich klingt ganz natürlich und wird in Liebeserklärungen sogar quasi erwartet: ein einfaches Seufzen oder ein Bericht über die Wettervorhersage sind dann weniger natürlich). Hier sei bemerkt, daß dem Ärger (Zorn etc.) keine illokutionären Verben entsprechen (Morin & O'Malley 1969), obschon die parenthetischen Verben, im Sinne von Urmson (1952) - regret, assume, believe, suppose, wonder, like to know, etc. - eine emotionale Wichtigkeit denotieren können. Eine kontrastiv-linguistische und sogar nicht-linguistische Analyse (z.B. in der Ethnographie) besteht nach Röttger-Rössler (1997:201) in "der Auf deckung der kulturspezifischen Besonderheiten des emotionalen Verhaltens in bezug auf die Fragestellung, inwieweit Emotionen kulturgebunden oder universal sind".
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
107
Eine kontrastiv-semantische Analyse von Emotionen basiert auf der (vor)theoretischen Taxonomie der Begriffe (also auf einer Organisation der 'Genossenschaften'), die meistens mit nicht-linguistischen Methoden entdeckt werden. Aber: "Den direktesten und einfachsten Zugang zu den Emotions konzeptionen einer anderen Kultur scheint die Sprache zu ermöglichen, d.h., die Erhebung des jeweiligen Emotionsvokabulars" (Röttger-Rössler 1997: 201). Die wichtigsten Komponenten, auf deren Basis die einzelnen Termini voneinander unterschieden werden, sind nach Röttger-Rössler (1997:203204): die soziale Position desjenigen Individuums, das die Emotion fühlt/ zeigt; der Anlaß, der Grund der Emotion, d.h. die Frage, ob die Wut berech tigt oder unberechtigt ist; die Ausdrucksformen der Wut, u.a. angemessene und unangemessene Ausdrucksformen. Ein wichtiger Aspekt ist des Weiteren, ob die Wut verbalisiert wird oder nicht. 6.
'Dorfarbeiten ' und Wortgebrauch Wie werden nun Dorfarbeiten geleistet? Es gibt folgende Möglich keiten: I
II
Die Lexeme, die zu einem semantischen Dorf gehören, können in bestimmten Wortverbindungen in einer Bedeutung gebraucht werden, die nicht die Grenzen des ständigen 'Wohnortes' überschreiten. Dies ist der Normalfall. Die Lexeme können in bestimmten Wortverbindungen in einer Bedeutung gebraucht werden, die außerhalb ihres ständigen 'Wohnortes' liegt. Diesen Fall kann man 'Strafarbeiten' nennen analog dazu, was während der Kulturrevolution in China oder in Rußland passierte: Als Strafe verbannte man Intellektuelle in ein fremdes Dorf, um sie für ungewöhnliche Arbeiten einzusetzen. Bei unserer Metapher kann diese 'Verbannung' aber keine lebenslange Strafe sein. Dabei gibt es zwei Möglichkeiten: ein Bewohner des Ärgerdorfes wird in irgendein anderes Dorf verbannt; vgl. Zornader, Grimmdarm, Grimmen,.. grimmige Schmerzen, wo Zorn und Grimm nach außerhalb des Ärgerdorfes verbannt werden; ein Bewohner eines anderen semantischen Dorfes wird nach Ärgerdorf verbannt; z.B.: j-η böse machen ('Strafarbeit' für das Lexem böse), igrat' na nervax kogo "auf die Nerven gehen"
108
III
VALERIJ DEM'JANKOV
(Strafarbeit für das Lexem Nerv), polezt' na stenu "die Wand/ Wände hochgehen" (Strafarbeit für Wand). 'Dienstreise': ein Lexem aus dem Ärgerdorf wird mit einer Dorfarbeit beschäftigt, die eine im Rahmen des Ärgerdorfes liegende Bedeutung ergibt: ein Dorfbewohner hilft seinen Genossenschaftskollegen, vgl. Zorneswut, wo Zorn und Wut einander helfen und zu derselben Genossenschaft gehören, - oder ein Dorfbewohner beteiligt sich bei den Arbeiten einer anderen Genossenschaft, z.B. nichts als Ärger und Verdruß haben.
Bei linguistischer Analyse, d.h. bei Ermittlung der in einem gegebenen semantischen Dorf einer konkreten Sprache existierenden Genossenschaften und Dorfarbeiten muß ein Forscher einen Muttersprachler fragen, ob eine gegebene Wortverbindung möglich ist. Es gibt folgende Möglichkeiten, die verschiedenen Zuständen des inneren Lexikons entsprechen: Ja, sicher, ganz bestimmt - d.h. eine Position, ein Prädikatenrahmen (in anderen Terminologien Valenz usw.) ist positiv angegeben: das Lexem paßt bestimmt zum Prädikat. Diese Möglichkeit liegt z.B. vor, wenn man jemanden über den Status einer Wortverbindung wie j-η in Raserei (Wut, Zorn) versetzen fragt. Nein, ganz bestimmt nicht - eine Position, ein Prädikatenrahmen usw. ist negativ angegeben: das Lexem paßt nicht zum Prädikat, z.B. j-η in *Empörung (*Entrüstung) bringen. Vielleicht ja, aber ich bin nicht ganz sicher; viel besser wäre es mit dem Wort X, wobei gewöhnlich ein Nachbar im Dorf genannt wird; wie wenn man fragt: Wie komme ich bitte in die Hammerstraße? und man antwortet: Irgendwo dort, aber ich bin hier selbst fremd, fragen Sie noch jemanden. Im mentalen Lexikon finden wir keine direkte Angabe; man muß bei den nächsten Nachbarn fragen: zuerst bei den Genossen schaftskollegen, und wenn diese auch keine Antwort geben, dann muß man sich an andere Mitbewohner wenden. Z.B.: in *Raserei (Wut, *Zorn) kommen. Das heißt, man ist ganz sicher, daß *in Raserei/Zorn kommen unmöglich ist, aber nicht so sicher im Falle von in Wut kommen. Die ersten zwei Fälle sind in den existierenden Theorien des Lexikons verhältnismäßig leicht behandelt, meistens durch eine Markierung in der Beschreibung einer lexikalischen Einheit ('lexical entry') und durch lexikalische Redundanzregeln ('lexical redundancy rules' oder 'rules of lexicon', à la Jackendoff 1972 oder Aronoff 1976). In unserer Terminologie
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
109
geht es dabei um ein reserviertes 'Instrument', das ein Dorfarbeiter entweder geerbt und bei der Einsiedlung mitbringt (s. z.B. gnev) oder das organisch mit einer gegebenen Genossenschaft (un)verträglich ist, vgl. j-η in *Empörung (*Entrüstung) bringen und bei j-m Empörung (Entrüstung) her vorrufen. Der dritte von den genannten Fällen ist besonders interessant. Bei einem Zweifel in solchen Fällen (was übrigens eine kolossale intellektuelle Arbeit für den Informanten bedeutet) helfen einem nicht die gespeicherten Daten über die Genossenschaften, Instrumente, Dienstreisen und Strafar beiten eines gegebenen Dorfes, sondern andere Weltkenntnisse, das 'Welt bild'. Die Konstatierung des Zweifels selbst bei Beantwortung solcher Fragen, besonders für die kontrastive Lexikologie, ist sehr wichtig. Solche Stellen markieren für den Lernenden gefährliche Zonen, die jeweiligen Grenzen der semantischen Dörfer, die der Studierende möglichst nicht über schreiten sollte. 7.
Semantik und Pragmatik im semantischen Dorf Im Rahmen der vermenschlichenden Metapher des semantischen Dorfes, wobei der Semantik der 'Wohnort' eines Wortes entspricht, können wir sehen, daß Pragmatik kein einheitlicher Begriff ist. Zum einen, wenn man unter Pragmatik eine 'nicht-wörtliche' (u.a. auch etymologisch nicht ursprüngliche) Bedeutung versteht, geht es bei uns um Wohnortswechsel, und man kann sagen, daß die 'Aussiedler' pragmatisch motiviert sind; zum anderen entspricht die Pragmatik eines Wortes auch dem, was das Wort 'leistet', seiner 'Dorfarbeit' (konkret im gegebenen Dorf) oder 'Beschäf tigung' (universal gesehen). Unsere Metapher gestattet es unter anderem, den Zusammenhang zwischen beiden Pragmatiken anschaulich zu machen. Die sprachlichen Einheiten suchen wie Menschen (und können oft auch finden) einen Wohnort, eine Genossenschaft usw., die ihren Leistungsfähigkeiten besonders gut entspricht. Und umgekehrt setzt ein Wohnort eine bestimmte Menge von Beschäftigungen für die Einwohner voraus. Der Produktivität im Sprachgebrauch und im Sprachsystem entspricht der Fall, wo die Beziehungen zwischen Mitgliedern einer Wortfamilie nicht zerstört sind, auch wenn ein Familienmitglied aus seinem Heimatdorf auswandert. 'Verlorene Söhne' (und sogar 'verlorene Väter') dagegen sind diejenigen ausgewanderten linguistischen Einheiten, die im Laufe der Sprachent wicklung die Bedeutung ihrer Vorväter (bzw. Söhne, wenn die letzteren das Heimatdorf nicht verlassen haben) vollkommen verloren haben.
110
VALERIJ DEM'JANKOV
Diese und ähnliche Analogien scheinen mir nicht weniger berechtigt und einleuchtend zu sein als andere Metaphern, die in der modernen linguistischen Metasprache üblich sind. 8.
Vergleichende Tabellen (nach Weigand: "The Vocabulary of Emotion", in diesem Band) deutsches 'Ärgerdorf - russisches 'gnevlivka'
I.
Genossenschaft NEUTRAL
Ärger, Ärgerlichkeit
razdraženie, neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost', serditost'
Dorfarbeiten: BE ärgerlich/verärgert sein über sich ärgern Ärger haben/empfinden viel Ärger erleben
wenn ... - dann gibt es Arger (es riecht nach Ärger) ärgerliche Töne die Ärgerlichkeit seines Tons s. Ärger auslassen gegen/an seinem Ärger Luft machen BECOME ärgerlich werden in Ärger geraten Ärgernis nehmen an Ärger überfällt/packt/überkommt j-n Ärger bekommen LOSE der Ärger verfliegt
byt' nedovol'nym + Instr. serdit'sja/byt' serditym: nakogo-L, na cto-1. za cto-1. ispytyvat' neudovol'stvie 4- Instr. ispytyvat'/pokazyvat'/vykazyvat'/ispytyvat' : razdrazenie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost' ) razdraženie (nedovol'stvo, neudovletvorenie, neudovletvorennost') tornilo ego esli ... to ty u menja polučiš' delo paxnet neprijatnostjami razdražennyj ton razdrazennost' ego tona vymestit' svoe razdraženie, neudovol'stvie, nedovol'stvo na kom-1. vypustit' par razdražat'sja/rasserdit'sja iz-za chego-L, na chto-1. vpast' ν razdrazenie; razdražat'sja rasserdit'sja na + Akk. razdraženie ovladevaet kern (Instr.) oscutit' nedovol'stvo razdraženie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost', serditost') uletucivaetsja
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
der Ärger vergeht
der Ärger läßt nach sich Ärger ersparen seinen Ärger unterdrücken seinen/den Ärger herunterschlucken CAUSE j-η ärgern; verärgern
j-η ärgerlich machen ein Ärgernis sein für j-n Ärger, -nis erregen bei j-m/j-m Ärger bereiten j-η in Ärger versetzen ärgerliche Situation für j-n das Ärgerliche an der Sache MOTIVATING CONCEPT außer sich vor Ärger grün und gelb vor Ärger aus Ärger etwas tun zu meinem Ärger zum Ärgernis aller bei aller Ärgerlichkeit voller Ärger
zum Ärger GRADING leiser Ärger heftiger Ärger großer Ärger viel Ärger/wenig Ärger
111
razdraženie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost', serditost') proxodit razdraženie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost', serditost') spadaet ne brat' ν golovu podavit' svoe razdraženie prikusit' gubu; smirit'sja; zabyt' svoe razdraženie vyzyvat'/vozbuzdat': razdraženie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost ') razdražat' + Akk. (nur ohne Absicht!); serdit' 4- Akk. byt' pricinoj neudovol'stvija + Gen. vyzyvat' razdraženie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost', serditost') privesti ν razdražennoe sostojanie neprijatnaja/scekotlivaja situacija dlja + Gen. neprijatnost' ν tom, cto ... vne sebja ot razdrazenija (neudovol'stvija, nedovol'stva, neudovletvorennosti) zelenyj ot razdraženija sdelat' ot razdraženija moemu neudovol'stviju ko vseobscemu neudovol'stviju vopreki vsemu s razdrazeniem (neudovol'stviem, nedovol'stvom, neudovletvoreniem, neudovletvorennost'ju) nazlo legkoe neudovol'stvie, razdraženie sil'noe razdraženie (neudovol'stvie, nedovol'stvo, neudovletvorenie, neudovletvorennost' ) velikoe neudovol'stvie izrjadnoe neudovol'stvie/nekotoroe neudovol'stvie
112
VALERIJ DEM'JANKOV
sich sehr ärgern sich schwarz/grün und blau/gelb und grün ärgern sich maßlos/unglaublich ärgern sich furchtbar ärgern sich zu Tode/krank/ins Grab ärgern sich über die Fliege an der Wand ärgern
II.
ocen' serdit'sja pozelenet' ot neudovol'stvija sil'no/črezvyčajno serdit'sja strasno rasserdit'sja ot neudovol'stvija sleč' (v postel') serdit'sja po pustjakam
Genossenschaft INTENSIFIED
Zorn, Wut, Raserei
gnev, neistovstvo, jarost', isstuplenie, besenstvo
Dorfarbeiten: BE zürnen, zornig/erzürnt/wütend sein
wutentbrannt/aufgebracht/aufbrausend sein Zorn haben auf j . zürnt mir von Zorn/Wut erfüllt sein von Zorn ergriffen sein sein Zorn richtet sich gegen der Zorn der Götter trifft ihn Wut im Bauch haben wütende/zornige Stimme
BECOME sich erzürnen, aufbrausen in Zorn geraten/in Wut kommen der Zorn packt/überkommt j-n
gnevat'sja, neistovstvovat'; byt' raz'jarennym, isstuplennym, vzbesennym ispytyvat' gnev (neistovstvo, jarost', isstuplenie, besenstvo) dat' volju svoemu gnevu (neistovstvu, jarosti, besenstvu) neistovstvovat' raz'jarennyj; jarostnyj besnovat'sja; vzbesennyj polnyj/pripadok/vne sebja ot: gneva (neistovstva, jarosti, isstuplenija, besenstva); pylat' gnevom (jarost'ju) tocit' zub na + Akk. nekto besit menja ispolnennyj gneva/jarosti/neistovstva/ bešenstva byt' oxvačennym gnevom ego gnev/jarost'/besenstvo napravlen na + Akk. kara gospodnja na nego pala kipet' ot gneva (jarosti) gnevnyj, razdražennyj, golos jarostnyj poryv neistovyj, gnevlivyj (po xarakteru) celovek razgnevat'sja, raz'jarit'sja, vzbesit'sja prixodit'/vpast': ν jarost' (isstuplenie, besenstvo) gnev (neistovstvo, jarost', besenstvo) oxvatil ego
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
die Wände hochgehen aus der Haut fahren
vspyska gneva (jarosti, besenstva) polezt' na stenu; polezt' s kulakami vyjti iz sebja
LOSE Zorn/Wut legt sich, läßt nach
smenit' gnev na milost'
CAUSE j-s Zorn/Wut erregen
zornig/wütend machen, erzürnen MOTIVATING CONCEPT rot vor Zorn voll Zorn, voller Wut schäumend vor Wut aufbrausend vor Zorn außer sich vor Wut vor Wut kochen
GRADING aufflammender Zorn, Raserei heller, lodernder Zorn leidenschaftlicher Zorn heiliger Zorn gerechter Zorn dumpfe Wut wilde Wut Zornanfall, Wutanfall (Theater machen)
113
navlec' na sebja gnev (jarost') kogo-1. privodit' ν neistovstvo (jarost', isstuplenie, besenstvo) + Akk.;vyzyvat' gnev (neistovstvo, jarost', isstuplenie, besenstvo) ν kom-1.; dovodit' do neistovstva (isstuplenija, besenstva) kogo-1. gnevit', besit' kogo-1. krasnyj ot gneva; blednyj ot jarosti (besenstva) ν gneve (neistovstve, jarosti, isstuplenii, besenstve) zadyxat'sja ot gneva (jarosti, isstuplenija, besenstva) vspyxnut' ot gneva, jarosti vne sebja ot gneva, jarosti kipet' ot jarosti ne pomnit' sebja ot gneva (jarosti, isstuplenija, besenstva) s gnevom (neistovstvom, jarost'ju, isstupleniem) kogo-1. dusit gnev (jarost', besenstvo) kraska gneva
jarost' kipet' ot gneva, jarosti isstuplennyj gnev svjatoj gnev gnev pravednyj gluxoe isstuplenie dikij gnev pripadok, vspyska gneva, jarosti ustraivat' sceny
114 III.
VALERIJ DEM'JANKOV
Genossenschaft WEAKENED
Verdruß, Unwille
dosada
Dorfarbeiten: BE Verdruß haben, empfinden verdrossen, ungehalten sein Unwillen erregendes Benehmen (Unwillen erregendes Erreignis)
pokazyvat'/vykazyvat': dosadu dosadovat' na cto-l.;byt' razdosadovannym čem-1. ν dosade, dosadujuscij povedenie, vyzyvajuscee dosadu dosadnoe proissestvie
BECOME Verdruß kommt auf
dosada voznikaet
LOSE Verdruß vergeht
dosada proxodit
CAUSE Verdruß machen, bringen, bereiten verdrießen
vyzyvat' dosadu ν kom-1. dosazdat'
MOTIVATING CONCEPT zum Verdruß
(ego) dosade
IV. Genossenschaft SUPPRESSED/CURBED Groll, Grimm
zloba, zlost', ozloblenie
Dorfarbeiten: BE Groll/Grimm empfinden, grollen
etwas wurmt j-n BECOME Groll/Grimm kommt auf
kipetVdysat': zloboj (zlost'ju, ozlobleniem) pitat' (besenuju) zlobu (zlost', ozloblenie) komu-1. byt' vne sebja/pozelenet'/poblednet'/ lopat'sja/sebja ne pomnit'/zadyxat'sja/ drožat': ot zloby (zlosti,ozloblenija) kipit zloba (zlost', ozloblenie) v kom dat' volju svoej zlobe (zlosti, ozlobleniju) so zloboj (so zlost'ju, s ozlobleniem) ozloblennyj; zlit'sja (na kogo-L, na čto-1.) kogo-1. (Akk.) gložet zloba, zlost' zloba (zlost') vspyxivaet zlost' menja vzjala
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
115
vpadat' ν ozloblenie razozlit'sja (na kogo-1.) LOSE Groll/Grimm vergeht CAUSE j-s Groll erregen
V.
zloba (zlost', ozloblenie) utixaet, proxodit vyzyvat' zlobu (zlost', ozloblenie) ν kom-1. zlit'; ozlobljat' kogo-1.
Genossenschaft MORAL
Entrüstung, Empörung
negodovanie, vozmuschenie
Dorfarbeiten: BE empört, entrüstet sein über
E. erfüllt j-η, voller E. sein E. empfinden seiner E. Luft machen ein entrüstetes Gesicht (empörte Ausrufe) BECOME sich entrüsten, sich empören ein Sturm der Entrüstung bricht los CAUSE etwas entrüstet/empört j-n j-η verärgern Entrüstung/Empörung erregen ein empörendes Benehmen
negodovanie (vozmuscenie) kipit ν kom-1. poblednet' ot negodovanija (vozmuscenija) ν negodovanii (vozmuscenii) negodovat' (na kogo-1., na čto-1.) vozmuscat'sja (kem-1., čem-1.) oxvacennyj negodovaniem (vozmusceniem) ispytyvat' negodovanie (vozmuscenie) dat' volju svoemu negodovaniju, vozmusceniju vozmuscennoe lico negodujuscie vosklicanija prixodit' ν negodovanie voznegodovat'; vozmutit'sja burja gneva (vozmuscenija) razrazilas' privodit' ν negodovanie (vozmuscenie) kogo-1. vozmuscat' kogo-1. vyzyvat' negodovanie (vozmuscenie) ν kom-1. vozmutitel'noe povedenie
116
VALERIJ DEM'JANKOV
Literatur Aronoff, Mark H. 1976. Word Formation in Generative Grammar. Cambridge, Mass.: MIT. Aylwin, S. 1985. Structure in Thought and Feeling. London & New York: Methuen. Bänninger-Huber, Eva 1996. Mimik - Übertragung - Interaktion: Die Untersuchung affektiver Prozesse in der Psychotherapie. Bern etc.: Huber. Chu, Run 1990. Maschinenunterstützte semantische Analyse bilingualer Wortfelder. Diss. Bonn: Philos. Fakultät der Rheinischen Friedrich-Wilhelms-Universität. Dickie, George 1996. The Century of Taste: The philosophical Odyssey of taste in the eighteenth century. New York & Oxford: Oxford University Press. Eckensberger, Lutz H. 1996. "Agency, Action and Culture: Three basic concepts for crosscultural psychology". Asian Contributions to Cross-cultural Psychology hg. von J. Pandey et al., 72-102. London, Newbury Park & New Delhi: Sage. Ekman, Paul, W. V. Friesen & P. Ellsworth 1982. "What Emotion Categories or Dimensions can Observers Judge from Facial Behavior?" Emotion in the Human Face, 2nd ed. hg. von Paul Ekman, 40-55. Cambridge: Cambridge University Press. Fries, Norbert 1996. "Grammatik und Emotionen". Sprache und Subjektivität I hg. von Wolfgang Klein & Brigitte Schlieben-Lange, 37-69. Stuttgart: Metzler. Geckeier, Horst 1971. Strukturelle Semantik und Wortfeldtheorie. München: Fink. Grayshon, M. 1975. "Some Aspects of Social Grammar Features of one Type of Question in English and Yoruba". Language in Society 4:1.17-29. Harré, Rom 1994. "Emotion and Memory: The second cognitive revolution". Philosophy, Psychology and Psychiatry hg. von A. Ph. Griffiths, 25-40. Cambridge etc.: Cambridge University Press. Herbermann, Clemens Peter 1995. "Felder und Wörter". Hoinkes 1995. 263-291. Herder, Johann Gottfried 1772. "Über den Ursprung der Sprache". Johann Gottfried Herder. Werke in fünf Bänden, Bd. 2, 89-200. Berlin & Weimar: Aufbau. Hoinkes, Ulrich, ed. 1995. Panorama der lexikalischen Semantik: Thematische Festschrift aus Anlaß des 60. Geburtstags von Horst Geckeier. Tübingen: Narr. Humboldt, Wilhelm von 1949[1836]. Über die Verschiedenheit des menschlichen Sprach baues und ihren Einfluß auf die geistige Entwicklung des Menschengeschlechts hg. von Johann Carl Eduard Buschmann. Berlin: Dümmler. (Repr. Darmstadt: Claasen & Roether.) Hundsnurscher, Franz 1995. "Das Gebrauchsprofil der Wörter: Überlegungen zur Metho dologie der wortsemantischen Beschreibung". Hoinkes 1995. 347-360. Hundsnurscher, Franz & Jochen Splett 1982. Semantik der Adjektive des Deutschen: Analyse der semantischen Relationen. Opladen: Westdeutscher Verlag. Ipsen, Gunther 1924. "Der alte Orient und die Indogermanen". Stand und Aufgaben der Sprachwissenschaft: Festschrift für W. Streitberg hg. von J. Friedrich et al., 200-237. Heidelberg: Winter. Izard, Carroll 1977. Human Emotions. New York & London: Plenum. Jackendoff, Ray S. 1972. Semantic Interpretation in Generative Grammar. Cambridge, Mass. & London: MIT. Jolles, Andre 1934. "Antike Bedeutungsfelder". Beiträge zur Geschichte der deutschen Sprache und Literatur 58.97-109.
ZUR KONTRASTIV-SEMANTISCHEN ANALYSE VON EMOTIONEN
117
Kernberg, Otto F. 1992/97. Wut und Haß: Über die Bedeutung von Aggression bei Persönlichkeitsstörungen und sexuellen Perversionen, aus dem Amerikanischen übers. Stuttgart: Klett-Cotta. Konstantinidou, Magdalene 1997. Sprache und Gefühl. Hamburg: Buske. Kronasser, Heinz 1952. Handbuch der Semasiologie: Kurze Einführung in die Geschichte, Problematik und Terminologie der Bedeutungslehre. Heidelberg: Winter. Labov, William 1984. "Intensity". Meaning, Form, and Use in Context: Linguistic applications hg. von Deborah Schiffrin, 43-70. Washington: Georgetown University Press. Lazarus-Mainka, Gerda & Stefanie Siebeneick 1997. Ängstlichkeit als Selbstkonzept. Göttingen: Hogrefe. Lycan, William G. 1996. "Philosophy of Mind". The Blackwell Companion to Philosophy hg. von Nicholas Bunnin & E.P. Tsui-James, 167-197. Oxford: Blackwell. Lyons, John 1977. Semantics, 2 voll. Cambridge, New York & Melbourne: Cambridge University Press. Magai, Carol & Susan H. McFadden 1995. The Role of Emotions in Social and Personality Development: History, theory, and research. New York & London: Plenum. Morin, Yves-Charles & Micheal-H. O'Malley 1969. "Multi-rooted Vines in Semantic Representation". Papers from the Fifth Regional Meeting of the Chicago Linguistic Society, April 18-19, 1969 hg. von Robert I. Binnick, Alice Davison, Georgia M. Green & Jerry L. Morgan, 178-185. Chicago: University of Chicago. Nietzsche, Friedrich 1930[1887]. Der Wille zur Macht: Versuch einer Umwertung aller Werte. Leipzig: Kröner. Parkinson, Brian 1995. Ideas and Realities of Emotion. London & New York: Routledge. Pokrovskij, Michail M. 1959[1895]. Semasiologicheskie issledovanija ν oblasti drevnix jazykov. Magisterdiss. Moskau. (Repr. Izbrannye raboty po jazykoznaniju. 61-153. Moskau: Izd-vo AN SSSR.) Porzig, Walter 1934. "Wesenhafte Bedeutungsbeziehungen". Beiträge zur Geschichte der deutschen Sprache und Literatur 58.70-97. Runggaldier, Edmund 1996. Was sind Handlungen? Eine philosophische Auseinandersetzung mit dem Naturalismus. Stuttgart etc.: Kohlhammer. Röttger-Rössler, Birgitt 1997. "Die Wortlosigkeit des Ethnologen: Zum Problem der Übersetzung zwischen den Kulturen am Beispiel indonesischer Gefühlstermini". Übersetzung als Repräsentation fremder Kulturen hg. von Doris Bachmann-Medick, 199-213. Berlin: Schmidt. Ryle, Gilbert 1949. The Concept of Mind. London: Hutchinson. Splett, Jochen 1993. Althochdeutsches Wörterbuch: Analyse der Wortfamilienstrukturen des Althochdeutschen, zugleich Grundlegung einer zukünftigen Strukturgeschichte des deutschen Wortschatzes, vol. I,1. Berlin & New York: de Gruyter. Thomas, L. 1996. "Evil and the Concept of Human Person". Moral Concepts hg. von P. A. French et al., 36-58. Notre Dame/Indiana: University of Notre Dame. Trier, Jost 1934. "Das sprachliche Feld: Eine Auseinandersetzung". Neue Jahrbücher ßr Wissenschaft und Jugendbildung 10.428-449. Ulimann, Stephen. D. 1953. "Descriptive Semantics and Linguistic Typology". Word 9:3.225-240.
118
VALERIJ DEM'JANKOV
Urmson, James O. 1963[1952]. "Parenthetical Verbs". Mind 61.480-496. (Repr. Philosophy and Ordinary Language hg. von Charles E. Caton, 220-240. Urbana, I11.: University of Illinois.) Wallner, F. G. & M. Costazza 1995. "How Ludwig Wittgenstein would have Reacted to Recent Changes in Psychology". Wittgenstein: Mind and language hg. von Rosaria Egidi, 279-288. Dordrecht: Kluwer. Weigand, Edda 1996. "Words and their Role in Language Use". Lexical Structures and Language Use: Proceedings of the International Conference on Lexicology and Lexical Semantics, Münster, September 13-15, 1994 hg. von Edda Weigand & Franz Hundsnurscher, vol. I, 151-167. Tübingen: Niemeyer. Weisgerber, Leo 1973 [1956/57]. "Die Erforschung der Sprach'zugriffe': I. Grundlinien einer inhaltbezogenen Grammatik". Wirkendes Wort 7.65-73. (Repr. Das Ringen um eine neue deutsche Grammatik: Aufsätze aus drei Jahrzehnten (1929-1959), 3., durchges. Aufl. hg, von Hugo Moser, 21-35. Darmstadt: Wissenschaftliche Buch gesellschaft.) Wierzbicka, Anna 1996. Semantics: Primes and Universals. Oxford & New York: Oxford University Press. Wundt, Wilhelm 1902-1903. Grundzüge der physiologischen Psychologie, Bd. 1-3. Leipzig: Engelmann.
EQUIVALENCE IN CONTRASTIVE SEMANTICS THE EFFECT OF CULTURAL DIFFERENCES
HENNING WESTHEIDE Rijksuniversiteit Leiden 1.
Preliminary remarks In modern translation theory the benefits of contrastive semantics are not always appreciated. Scholars as Hönig & Kußmaul (1982:14), SnellHornby (1988:22), and Van Leeuwen-Zwart (1992:115) claim that the predicate 'equivalent' can only be used in respect to texts and not to single words or expressions. In Westheide (1995) I discussed these reservations about this notion. My conclusion then was that we can't do without this notion. As Wunderlich (1980:17) pointed out, if we want to prove texts to be globally equivalent we first have to ensure that utterances, expressions and lexical units are equivalent in relation to the rest of the texts we translate. This had already been the conclusion of Catford (1965) who developed the technique of 'inter-system shift'. Of course in contrastive semantics, as in the practice of translation, we cannot do without the competence of bilingual speakers. Only they can judge if two expressions are equivalent in a special context. But as Weigand ("Contrastive Lexical Semantics", in this volume) emphasizes, it is the task of comparative semantics to make competence transparent and explicit. In bilingual dictionaries as well as in computerized translation assistance we cannot do without this knowledge. The comparative description of the conventions of language usage remains, therefore, one of the prime fundamentals for translation in practice and for the rational testing of the equivalence of texts. Another field which is just now starting to boom is CALL, computer assisted foreign language learning. In the Netherlands many foreign language departments have their own CALL-projects, often in cooperation with other
120
HENNING WESTHEIDE
universities. We in Leiden are working together with Groningen and Nijmegen using a program named HOLOGRAM. My project therein is called LEXIGRAM Duits. It tries to employ the computer for the acquisition of the vocabulary in combination with lexical bound syntactic rules. I will not deal with the benefits and the disadvantages of computer assisted foreign language learning now. But I do want to emphasize that we need comparative seman tics if we don't want to make programs which deliver only some morphologi cal and syntactic drills. All these CALL-programs have in common that they want to instruct the student to use the right linguistic forms in special standardized situations. In feed backs and hypertexts we want to provide the knowledge which leads to an adequate language use. In the following I want to comment on my description of the DutchGerman meaning area of ANGER with respect to cultural differences as re striction to equivalence. In order to extend the basis of discussion I will add a contrastive analysis of the Dutch-German meaning area of JOY. In connection with this discussion I shall report on my learner's dictionary which is also the basis for my CALL-project 2.
Some difficulties in correlating equivalent vocabulary It is very difficult to make up the equivalent expressions of a second language starting with words or groups of words that are being considered as one unit. Even if we categorize these forms according to overall accepted basic universal categories on the semantic level, we are confronted with properties of these forms on the pragmatic and metapragmatic level respec tively. Stylistic values and contextual restraints form an environment which makes it very difficult to judge the equivalence on the level of universal concepts. According to a sociolinguistic tradition we distinguish between 'ways of speaking' (Hymes 1974) with respect to different 'speech communities'. Gumperz (1982) emphasizes the importance of the interactional entextualization by introducing the term 'contextualization cues'. As part of an ethnolinguistic model Silverstein (1993) has worked out his concept of metapragmatics which refers to the wide realm of mostly unconsciously applied rules according to which text-meaning is constituted on the basis of semantic and pragmatic values of linguistic forms. If we try to make judgements about the semantic equivalence of linguistic forms of two different languages we must be aware of the fact that we can only gain access
EQUIVALENCE IN CONTRASTIVE SEMANTICS
121
to semantic values through our knowledge of the way they are used in the society in question. Thus we have to take into account that the pragmatic and metapragmatic rules which are bound to these forms, by the way they are used, form an important part of the quality by which we judge them being equivalent or not. If we discuss the question whether cultural differences have an effect on equivalence or not, we have to distinguish at least three different causes of these effects. First, we could assume that the realm of universal concepts is far smaller than required. It is a stereotype of the Dutch stated over and over again that they are less guided by emotions, that they rather act rationally whereas Germans are supposed to be emotional and rather act irrationally. Thus we could assume that the vocabulary in this field would be different: richer in German, less rich in Dutch; the meaning of the vocabulary could be different, referring to different mental states. Second, if the universal semantic concepts are as common as believed to be there could be a different structure of the vocabulary, for example, caused by different syntactic rules. These differences - for instance, the difference in the case system of Dutch and German - have led to different ways of expressing emotions. In German all verbs in the field of emotions can be used according to the same syntactic patterns: (1) (2) (3) (4) (5) (6) (7)
es ärgert mich, daß ... ich ärgere ihn mit etwas ich ärgere mich über etwas ich bin über etwas verärgert ich bin vonj-m geärgert worden
es freut mich, daß... es freut mich für j-n ich erfreue ihn mit etwas ich freue mich über/an etwas ich freue mich auf etwas ich bin über etwas erfreut
In Dutch the use of equivalent verbs of this field is restricted to some relatively simple patterns such as zich aan/over iets ergeren "to be cross with sb/to be annoyed about sth" in the case of ergeren. Equivalent forms as Dutch geïrriteerd zijn over "to be irritated by sth" or iemand pesten met "to tease sb" could be seen as part of the Dutch culture and be mistaken as being caused by a different way of experiencing emotions. In any case, many Dutch
122
HENNING WESTHEIDE
people are convinced that this complicated structure of the German language expresses the pedantic character of the people.1 Third, even if the concepts are universal and the lexical units equivalent there could be a difference in the rules according to which they are used. This difference caused by culturally bound metapragmatic rules is apparent in the field of politeness: Whereas in the Netherlands it is in accordance with the rules of politeness to play down the fact that you want someone to do something, in Germany you have to play up the fact that you are bothering someone by asking things. I use to demonstrate the pecularities of everyday language with a transcribed interview in which a German woman says on being asked about her relations to her neighbours: Da habn we viel Ärger gehabt..."We did have a lot of trouble...". In a comparable situation in the Netherlands you would say: We hebben er wel es problemen gehad ... The fact that you use viel Ärger haben mit j-m in German, and that you rather would use er "there" wel eens "sometimes" problemen hebben in Dutch could be a result of the rules which constitute meaning in texts. The troubles caused by the neighbours would be the same in both cases, and yet, the way of speaking is very different and can be easily misunderstood. 3.
Commentary on the German-Dutch meaning area of ANGER Thus looking for equivalents to the expressions of Weigand's Eng lish/German material (see "The Vocabulary of Emotion", in this volume) gives rise to a lot of problems, especially in the case of expressions with English anger/annoyance and German Ärger. Dictionaries, even bilingual ones, are of little help. In our case they very often use words which do not appear in everyday language or words which refer rather to the concrete reason for anger and not to the emotion. Native speakers of Dutch who know German . well do not accept words as Dutch misnoegen, wrevel or ontstemming in equivalent expressions. The practice of the Van Dale (1983:114) bilingual dictionary to refer to German beruflicher Ärger "trouble at work" as Dutch narigheden en ellende in het beroep and to German seinen Eltern Ärger bereiten "to annoy one's parents" as Dutch zijn ouders narigheid, last bezorgen may contextually be adequate, but it leaves the emotional part of the message to the imagination of the receiver. I therefore decided to rely on the competence of advanced learners of German and bilingual colleagues. The problem of this method is that you get 1 The dominant role of differences between the Dutch and the German language as reason for national stereotypes is worked out in Westheide (1997b).
EQUIVALENCE IN CONTRASTIVE SEMANTICS
123
a lot of suggestions but you don't know what "context-of-occurrence" (Silverstein 1993:36) these forms presuppose for the informants. Thus we have many more expressions on the Dutch side than on the English/German side. But perhaps this is the result of a more profound problem, because in the case of German Entrüstung/sich entrüsten "indignation/to become indig nant", as used in a moral sense in the meaning area of ANGER, we have one overall accepted equivalent in Dutch verontwaardiging/verontwaardigen. And even in the case of German Zorn, Wut "anger, fury", where Dutch toorn is a purely literary word, Dutch woede is in most expressions equivalent as well to German Zorn as to Wut. In Dutch erger is the comparative form of Dutch erg, equivalent Ger man schlimm/sehr, English bad/very and etymologically related to German antiquated arg (German arge Not - English bad misery). In most cases, namely 18 times, German Ärger is translated with Dutch woede, 14 times Dutch problemen is supposed to be equivalent. 12 times Dutch irritatie "irritation", irritant or geïrriteerd is used, and 11 times Dutch ergernis "annoyance". Dutch woede is also the most chosen equivalent to German Wut and Zorn, while Dutch toorn is not used in everyday language, as already mentioned above. Woede therefore belongs to the 'intensified' part of the meaning area of ANGER. Dutch irritatie is the most chosen equivalent to German Verdruß and Unwille and therefore belongs to the 'weakened' part of the meaning area of ANGER.
Dutch ergernis, in the first place, is something which causes trouble, seldom it is used to refer to an inner state. The fact that my informants so often chose Dutch problemen, which is equivalent to English problems and has little to do with emotions, indicates that the absence of a direct equivalent to German Ärger and English anger is not arbitrary. The equivalence of German Ärger and Dutch woede, irritatie, and problemen is not without doubt. On the one hand, the emotion expressed by the usage of woede is too intense and bound to a short time, on the other irritatie and problemen can be reduced to concrete causes. Thus the inner state is rationalized. As we have seen the equivalents for German Arger are either inten sified or weakened or they refer to the reason of the emotion. There are only scarcely used words such as Dutch misnoegen, wrevel or ontstemming "anger" which refer to emotion; but why are they so unpopular?
124
HENNING WESTHEIDE
This is the point where we cannot ignore any longer that speaking about emotions is socially bound to a specific way of life and that thus the differ ences in the use of the Dutch expressions in the field of ANGER could be the result of the way of speaking ('Sprachkultur'). It is one of the most striking differences between the Dutch and the German way of living that in Dutch social life it is very important to avoid conflict.2 In economy we speak of the German model of decision making as a model of confrontation whereas the Dutch is spoken of as a model of harmony. In Dutch conferences (Dutch vergaderingen) as in the parliament the argumentation must take place without confrontation. The greatest taboo is showing your anger. If you cannot stand the situation you have to leave. It is the same in daily life; Dutch social life is considered to be so peaceful, because there is no open confrontation through scolding or complaining, in some regions cursing is even prohibited by the local authorities. This impression of peacefulness contrasts with the fact that minor criminality is more widespread in the Netherlands than in most of the other European countries. So you have no equivalent utterances to German Das find ich aber ärgerlich! "But I do find it annoying." or Da hab' ich mich vielleicht geärgert! "That really annoyed me!" and Endlich hat er mal seinem Ärger Luft gemacht. "In the end he was annoyed." In our investigation into the habitual irritations of our Erasmus-students studying in Germany for one semester, this way of criticizing one another was the most frequent reason for complaints.3 The foreign language student must learn that the language he learns is not a code for universal communicative needs, but a communicative system which is bound to the cultural environment. This last point brings me to the second subject of my contribution, namely some information about my learner's dictionary of German for advanced Dutch learners. I will start with a discussion of German ärgern and its position in this book. 4.
A contrastive learner's dictionary of German for Dutch students It is apparent to everybody who comes into contact with Dutch and German that these two languages are closely related. Especially the common Germanic vocabulary opens up all sorts of possibilities for language transfer. At the same time this relation is also the reason for interference, as we call 2
For Dutch-German cultural differences in language behaviour I refer to the chapter: "Die Sprachen als Brücke und Hindernis" in Westheide (1997b). 3 The results of this research are written down in Hoogduin-Pitzer (1995).
EQUIVALENCE IN CONTRASTIVE SEMANTICS
125
negative transfer. Usually 'faux amies/false friends' are referred to as inter ference traps. But in the case of Dutch and German the situation is much more subtle, because very often even a native speaker does not notice that interference is the cause of a special language behaviour of a person using the hearer's language as a foreign language. Thus the utterance in question is taken for what it normally would be in the given situation, and just this is the origin of misunderstanding. The typological differences of the two languages4 and the development of Dutch and German culture which at times seem to be even quite opposite, are the reasons for a different use of nearly the same lexical material. One problem of German as a second language for Dutch people is well known to nearly all students of German, namely the difficulty of the case system and the possibilities it opens up to the construction of phrases. In my learner's dictionary the case system and the different use of the similar vocabulary are the most prominent learning problems I want to solve. I am convinced that in the case of German for Dutch students it is of little use to separate sentence syntax from semantics. Patterning and conventions of language use are two sides of the same coin. Perhaps it is because I chose verbs as a starting point that, in this respect, I differ from Weigand ("Contrastive Lexical Semantics", in this volume). If we look for the mini mal functional units then we have to differentiate between the different classes of parts of speech. It lies in the nature of nouns that their semantic value is much more independent than that of verbs. Let me demonstrate how in the case of verbs syntactic patterning and ways of use form an integrated whole by some examples of my learner's dictionary. 5.
The description of ärgern in a learner's dictionary of German for Dutch students In my "Contrastief idioomboek Duits" (Westheide 1997a) ärgern is dealt with in the very last chapter, which is part of the macro-field "Innermenschliche Prozesse und Zustände" (processes and states of the inner human being). This I thought to be the most complicated field because it is not accessible to objective judgement. The referring value therefore is ex tremely vague. The last meso-field which includes verbs such as interessieren "interest", erregen "excite/stimulate", aufregen "upset", stören "disturb", ärgern "anger", erschrecken "frighten", wundern "surprise", freuen "enjoy" 4
For the description of typological differences of the Dutch and the German language see Abraham (1995).
126
HENNING WESTHEIDE
and gefallen "please" refers to those processes and states which react against impulses from the outside world ("Reaktion auf Anstöße von außen"). Most of these verbs have to do with emotions and they all have a comparable display of usage: (8)
ich ärgere ihn mit etwas (9) ich ärgere mich über etwas (10) es ärgert mich (11) ich bin verärgert
ich rege ihn mit ich störe ihn bei etwas auf etwas ich rege mich über ich störe mich an etwas auf etwas es regt mich auf es stört mich ich bin aufgeregt ich bin (geistig) ge stört
There are some variations in the patterning, but generally these verbs which I found by contrasting Dutch and German verbal usage and which are brought together only through my onomasiological approach, have nearly the same patterning. As the usage of these verbs cannot be predicted from the usage of any Dutch equivalent, they are most economically learned together. In my comparative learner's dictionary as well as in the hypertexts of my computer program LEXIGRAM, I use commentary texts in those cases where metapragmatic rules are of importance and where the differences might cause prejudices. In this case I inform my Dutch students that it is overall accepted in Germany that someone expresses his anger and that this is done by openly criticizing, by schimpfen "scolding" and meckern "grum bling". On the micro-level ärgern is put together with the semantically related verbs erregen, aufregen, stören. With categories as 'reaction', 'intensity', 'negative value', and 'social disapproval' I try to differentiate the semantic value of these verbs (cf. Westheide 1997a:99; my translation, H.W.): Erregen and aufregen refer to intensive reactions. Erregen can be of positive or negative value: jemandes Interesse før etwas erregen "to attract someone's attention" and mit etwas Ärger erregen "to cause offence by doing something". Aufregen is negative: Der Lärm regt ihn auf/er regt sich über den Lärm auf "the noise annoys him". Ärgern is even more negative than aufregen. The mental state causes that one is verärgert: Er ist durch den ständigen Lärm während der Arbeit sehr verärgert "he is very annoyed by the constant noise during his work". Stören is used as a verb of social action: Er stört mich bei der Arbeit "he interrupts my work", but also as a mental state of the experiencer: Er stört mich bei der Arbeit "he bothers me while working" and
EQUIVALENCE IN CONTRASTIVE SEMANTICS
127
also with theme-rising: Ich störe mich an seinem Verhalten "I don't like his behaviour". But how to indicate the rules according to which ärgern is used in the German language? If you take the Dutch language as a starting point, you wouldn't use it in situations where Germans would. And how to prevent the misunderstanding of the usage of ärgern by the Dutch recipient, due to his use of Dutch ergeren? In the dictionary I illustrate the semantic value of use of each of the verbs together with their Dutch equivalents (Westheide 1997a: 100): Du regst mich auf mit deinem ewigen Meckern. Er regt sich darüber auf, daß wir die Chance verpaßt haben. Es regt mich auf, daß er nicht kommt. Der Lärm regt mich auf. Aufmerksamkeit erregen Anstoß erregen eine erregte Diskussion Das Kindergeschrei stört mich bei der Arbeit. Darf ich Sie einen Augenblick stören? Das grelle Licht stört mich. Du sollst nicht immer deine kleinen Ge schwister ärgern. Ich ärgere mich schon lange über sein Verhalten. Es ärgerte ihn besonders, daß sie immer wieder seine Garage beschmierten. Sie können ihm ruhig sagen, daß ich über sein Verhalten sehr verärgert bin.
Ik wind me op over je voortdurend ge zeur. Hij maakt er zich druk over dat wij die kans hebben gemist. Ik wind me erover op dat hij niet komt. Ik word zenuwachtig van het lawaai. aandacht trekken aanstoot geven een opgewonden discussie Het geschreeuw van de kinderen stoort me bij mijn werk. Mag ik u even storen? Het felle licht hindert me. Je moet niet altijd de kleintjes pesten. Zijn gedrag irriteert mij al een geruime tijd. Hij was er bijzonder boos over, dat zij steeds weer zijn garage bekladden. U kunt rustig tegen hem zeggen dat zijn gedrag mij sterk irriteert.
There are nearly as many equivalent Dutch expressions as there are forms of use of the German verbs. Through exercises such as the following I try to build up the competence of the learner, so that he will be able to use the appropriate pattern in a given context (Westheide 1997a: 142): Exercise D In this group of verbs the 'theme' often is realized through a prepositional object. The special value of this usage is marked by the preposition. Choose the appropriate preposition with the correct 'Rektion'.
128
HENNING WESTHEIDE
Example: 1) Es war offensichtlich, daß er d... Zwischenfall sehr erregt war. Answer: Es war offensichtlich, daß er durch den Zwischenfall sehr erregt war. 2) Es freut uns, daß Sie d... Erwerb des Grundstückes interessiert sind. 3) Auch d... ander... Objekt interessieren wir uns. 4) Keiner stört sich mehr d... viel... Paar..., die unverheiratet zusammen leben. 5) Regen Sie sich doch nicht solch... Lappalien auf!
In my computer course LEXIGRAM wrong answers lead to feed backs and hypertexts which help the student to find the right ones. 6.
The description of (sich) freuen in a learner's dictionary of German for Dutch students My learner's dictionary is based on a corpus of such usages ('Ge brauchsweisen') of the most frequent German verbs which contrast with their Dutch equivalents. The macro- and the meso-structure of the dictionary cor respond to onomasiological principles, the micro-structure is based on trans lation equivalence of the usage of the verbs in both languages. As pointed out above, the meaning areas of ANGER and JOY belong to the macro-field "Innermenschliche Prozesse und Zustände" (Processes and states of the inner human being) and the meso-field "Reaktion auf Anstösse von außen" (Reaction to impulses from the outside world). The description of the microstructure of the entry freuen is as follows (Westheide 1997a: 101-102; my translation, H.W.): It is part of the German culture that you show your emotions more easily than in the Netherlands. Thus you very often are confronted with the fact that someone meckert "grumbles", kritisiert "criticizes" and schimpft "scolds", but also that he more easily would say that he sich freut "enjoys himself" and etwas genießt "enjoys something". Ich freue mich über ... "I am happy with" or Es freut mich, daß ..."I am happy that" and Ich bin erfreut darüber, daß ..."It makes me happy that" are part of many routine formulas which you use to say "thank you" or just to make some nice remark. Sich freuen with object in the genitive refers to something which you like: Freut euch des Lebens ... "Enjoy life". In the case of genießen "relish" with direct object: Er genießt es, überall bewundert zu werden "He loves being admired by everybody" in comparison with Dutch genieten van the emotional attitude is expressed more directly. Freude "joy" and Genuß "with relish" are shown at the many Volksfeste "public festivals" which are very popular in most parts of Germany and scarce in the 'ranstad' (the heart of Holland).
EQUIVALENCE IN CONTRASTIVE SEMANTICS
129
The value of usage has again to be illustrated by examples (Westheide 1997a: 102): Dein Erfolg freut uns.
Wij verheugen ons over jouw succes.
Erfreuliche Berichte Es freut mich, dich glücklich zu sehen. Er freut sich des Lebens/genießt das Leben. Die Kinder freuen sich auf die Ferien.
verheugende berichten Ik' ben blij jou gelukkig te zien. Hij geniet van het leven.
Karola freut sich über ihren Erfolg. Ich habe mich sehr über Ihr Geschenk gefreut. Wir haben uns sehr gefreut, daß Sie unser Angebot angenommen haben. Wir haben uns schon Monate auf dieses Fest gefreut.
De kinderen verheugen zich op de vakan tie. Karola is blij met haar succes. Ik ben erg blij met uw kado. Het verheugt ons zeer dat u ons aanbod heeft aangenomen. We hebben maanden naar dit feest toegeleefd.
We see that the usages of freuen marked by prepositions correspond with different Dutch equivalents: (12) sich freuen auf (13) sich freuen über -
zich verheugen op blij zijn met
If we use other patterns in German we have to decide whether to choose the first or the second: (14) etwas freut uns (15) es freut mich, . . . -
wij verheugen ons over iets ik ben blij ...
But in the following example we see that aspects of the wider context guided my informants looking for equivalents. In the following usage we have a fairly formal routine formula from business correspondence: (16) Wir haben uns sehr gefreut, daß Sie unser Angebot angenommen haben. - Het verheugt ons zeer dat ons aanbod heeft aangenomen. This could be an indication that blij zijn mee "to be happy with" even for the native speaker of Dutch is less intensive. But this could be an inter pretation based on the German meaning of froh sein über which expresses
130
HENNING WESTHEIDE
less joy than sich über etwas freuen. This is an important aspect of the affinity of the Dutch and German language: the German hearer interprets the Dutch expression on the basis of the semantic value of 'natural' German equivalents. With 'natural equivalents' we mean those equivalents which can be derived from each other by using simple phonological rules or by word for word translation. In the case of German das Leben genießen "to enjoy life" - Dutch van het leven genieten this leads to the impression of the German hearer as if the Dutch way of saying that you enjoy life is less intense, you enjoy parts, not the whole of it. But who knows, is it the structure of the language or the difference in culture? 7.
First efforts at a contrastive description of the Dutch-German meaning area of JOY In contrast to German Ärger there is a Dutch noun as a 'natural equivalent' of German Freude "joy", but this doesn't make things easier. In the German-Dutch bilingual Van Dale (1983:485-486) the Dutch equivalents for German Freude are: vreugde as major translation with the variants plezier, genoegen and blijdschap. If you look up the contexts which are pre sented for the use of these words then you get for vreugde literary, formal or antiquate ways of use; genoegen is used in formal speech formulas where it can be replaced by plezier in informal use. Plezier which is not mentioned in the monolingual Van Dale (1995) as a synonym for vreugde is used in most of the translations of German usages of Freude, especially if it is about joy in daily situations. You have plezier in je werk (Freude an der Arbeit "pleasure in your work"), you do iemand een plezier (j-m eine Freude bereiten "to please sb") and you do something met groot plezier (etwas mit großer Freude tun "to do sth with keen interest"). The very formal word blijdschap does not appear in any of the examples for the translation of German Freude. It is apparently taken over from Van Dale's monolingual dictionary which was the model for the bilingual ones. There blijdschap is a synonym of vreugde. In German Pläsier is morally marked and scarcely used, its value of use is somewhat pejorative. The Dutch plezier, on the other hand, is the most fre quent word in the field and is in addition to Freude also equivalent to Spaß and Vergnügen "fun/pleasure". Thus, in the field of JOY the contrast between the Dutch and the German vocabulary is even more complicated than in the field of ANGER. The most frequently used equivalent of German Freude is also the equivalent of Spaß and Vergnügen, emotions which are nearer to Genuß "with relish" than
EQUIVALENCE IN CONTRASTIVE SEMANTICS
131
to Freude. In addition to the impression which is caused by the natural equivalence of Dutch plezier and German Pläsier, the nearness of Dutch plezier to German Genuß and Spaß which both have something of worldly pleasures extends the impression on the side of the German recipient that emotions such as joy are looked at as those of worldly pleasures. Many examples of Dutch vreugde in Van Dale (1995:3480) are taken from a religious context. Thus you could conclude that the way of use of the Dutch words in the meaning area of JOY is influenced by the Calvinistic view on life.5 But if this influence is still of importance for the description of this field of EMOTION or if it is only the imagination seen through the glasses of a native speaker of German, we don't dare to answer. 8.
Instead of a conclusion The problems we meet looking for equivalences in fields as this one of EMOTION have always bothered translators. For the Dutch-German case we have a lively report in the correspondence of the great Dutch historian Huizinga and his equally famous Dutch friend André Jolles.6 Huizinga consults Jolles, teaching Dutch language and literature in Leipzig, in the case of the translation of Dutch felheid in the title of the first chapter of his masterpiece "Herfsttij der Middeleeuwen": "'s Levens felheid". Jolles' former wife, who did the translation, had been proposing Grellheit, Huizinga himself thought Heftigkeit to be more adequate. Jolles isn't pleased with both and comments: "Het zijn nuances - maar nuances die mij altijd weer bij vertalingen een griezelijk gevoel geven." ("These are nuances - but nuances which again and again impart an unpleasant feeling on me while trans lating.") In my edition of the book (Huizinga 1939) the translation is: "Die Spannung des Lebens". While Jolles helps Huizinga with the translation of his texts into German, Huizinga, on the other hand, helps Jolles to clean his Dutch texts from German influences, for in his academic texts he writes "a more and more German-coloured Dutch", as is said in a Dutch comment (Hanssen 1995:226, note 26). Jolles is aware of the contrasts between the two languages: "De psychologische afgrond tusschen de twee talen, die ons misschien uit een zekere overgevoeligheid - zowel in de woordbetekenis als in de syntaxis opvalt, is niet te dempen." ("The psychological abyss between the two languages which strikes us - perhaps in an excess of sensibility - as 5 6
The cultural background of the creation of the Dutch nation is subject of Schama (1988). For more information I refer to the excellent article of Harmsen (1995).
132
HENNING WESTHEIDE
well in word meaning as in syntax cannot be reduced.") In another com mentary he defines this 'deceptive similarity' even sharper (cf. Hanssen 1995:228): Je hebt gelijk - het wordt voortdurend wonderlijker hoe dichter men die twee talen naast elkaar zet - die twee volken kunnen niet overweg - vlak bij en toch eigenlijk ieder woord met een andere psychologie. "You are right - it becomes more and more miraculous the nearer you place the two languages to one another - the two people cannot get along - near to one another but actually each word with a different psychological background. "
References Abraham, Werner 1995. "German: Cold and Male - Dutch: Warm and Female. Reflections on a realistic typology". The Berkeley Conference on Dutch Linguistics 1993 ed. by Thomas F. Shannon & Johan P. Snapper. Lanham, MD: University Press of America, 171-210. Catford, John Cunnison 1965. A Linguistic Theory in Translation. An essay in applied linguistics. Oxford: Oxford University Press. Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge University Press. Hanssen, Léon 1995. "Ein deutscher 'Herbst'. Zur Geschichte und Deutung von Huizingas Meisterwerk". Duitse kroniek 45.219-237. Hoogduin-Pitzer, G. 1995. Sprache als soziales Handeln: Höflichkeit in der interkulturellen deutsch-niederländischen Kommunikation. Ms. Leiden. Hönig, Hans G. & Paul Kußmaul 1982. Strategie der Übersetzung. Ein Lehr- und Arbeitsbuch. Tübingen: Narr. Hymes, Dell 1974. "Ways of speaking". Explorations in the Ethnography of Speaking ed. by R. Baumann & J. Sherzer. London: Cambridge University Press. Leuven-Zwart, Kitty M. van 1992. Vertaalwetenschap: ontwikkelingen en perspectieven. Muiderberg: Coutinho. Schama, Simon 1988. The Ambarrassment of Riches: an Interpretation of Dutch Culture in the Golden Age. London: Fontata. Silverstein, Michael 1993. "Metapragmatic Discourse and Metapragmatic Function". Reflexive Language: Reported speech and metapragmatics ed. by John A. Lucy, 33-58. Cambridge: Cambridge University Press. Snell-Hornby, Mary 1988. Translation Studies. An integrated approach. Amsterdam & Philadelphia: Benjamins. Van Dale 1983. Groot woordenboek Duits-Nederlands. Utrecht & Antwerpen: Van Dale Lexicografie. Van Dale 1995. Groot woordenboek van de Nederlandse taal. Utrecht & Antwerpen: Van Dale Lexicografie. Wunderlich, Dieter 1980. Arbeitsbuch Semantik. Königstein/ Ts.: Athenäum. Westheide, Henning 1995. "Übersetzungsäquivalenz - Ein zentraler Begriff für Lexikogra phie und Übersetzungswissenschaft". Duitse kroniek 45.200-218.
EQUIVALENCE IN CONTRASTIVE SEMANTICS
133
1997a. Contrastief Idioomboek Duits. Handleiding voor doelmatig taalgebruik. Bussum: Coutinho. 1997b. Trügerische Nähe. Niederländisch-deutsche Beziehungen in Geschichte, Sprache und Kultur. Münster: LIT Verlag.
Appendix Ärger, Ärgerlichkeit, Ärgernis BE ärgerlich, verärgert sein sich ärgern Ärger haben
Ärger empfinden viel Ärger erleben es gibt Ärger ärgerliche Töne seine Ärgerlichkeit die Ärgerlichkeit seines Tones seine Ärgerlichkeit auslassen gegen/an
7
woede, irritatie7, problemen, ergernis geïrriteerd zijn over, geërgerd zijn zich boos maken, zich ergeren aan/over, ergens aanstoot aan nemen zich kwaad maken over iem./over iets, ruzie hebben, last hebben van, proble men/moeilijkheden hebben met zich geïrriteerd voelen, geïrriteerd zijn, ontstemd zijn, zich storen aan veel problemen tegenkomen, veel narigheid meemaken, veel ellende/tegenslagen hebben/beleven daar komen problemen van, daar krijgen we last mee, er zijn moeilijkheden, er is stront aan de knikker akelige, irritante geluiden, op een vervelende toon, geërgerde stem/toon zijn irritatie, zijn woede, zijn erger nis/geïrriteerdheid (zijn geatiteerdheid8) de irritatie/ergernis in zijn stem (de ergernis klinkt door in zijn stem) zijn boosheid laten blijken over iets/tegenover iem., zijn woede de vrije loop laten, zijn woede bekoelen op, zijn ergernis afreageren op
Van Dale (1995:1301) irritatie 1 het irriteren, zijn prikkeling 2 het geïrriteerd-zijn irriteren 1 prikkelen in eig. zin 2 prikkelen in psychische zin, boos of ongeduldig maken, syn. ergeren irritant irritatie opwekkend, syn. ergerniswekkend 8 Van Dale (1995:149) agitatie 1 opgewondenheid (Aufgeregtheit), zenuwachtige gemoedsbeweging (aufgeregte Gemütsbewegung) 2 onrust (Unruhe), gisting (Aufgehen als Folge von Hefe), spanning (Spannung), woeling (Aufgewühltheit) 3 het agiteren (das Agitieren), actie (Aktion), het beïnvloeden van publieke gevoelens (das Beeinflussen öffentlicher Gefühle), politieke of sociale strijd (politischer oder sozialer Kampf)
134
HENNING WESTHEIDE
seinem Ärger Luft machen BECOME ärgerlich werden in Ärger geraten Ärger überfällt/überkommt/packt j-n
Ärger bekommen LOSE der Ärger verfliegt der Ärger vergeht der Ärger läßt nach sich Ärger ersparen seinen Ärger unterdrücken seinen/den Ärger herunterschlucken ("Er wird sich schon beruhigen.")
z'n boosheid ventileren geïrriteerd raken, boos/kwaad worden in problemen komen, kwaad worden, in woede raken door woede overvallen/gepakt/overmand worden, de woede maakt zich van iem. meester, kwaad worden problemen/ruzie krijgen, in problemen ra ken/komen de problemen gaan over/voorbij, de woede waait over, de bui drijft over de woede/ boosheid/ irritatie neemt af, bekoelen, de woede gaat voorbij/ verdwijnt de woede zakt/neemt af/vermindert problemen voorkomen, zich problemen besparen zijn woede onderdrukken zijn woede verbij ten/inslikken, tot tien tellen ("Hij koelt wel af!")
CAUSE ärgern, verärgern j-n ärgerlich machen ein Ärgernis sein Ärger, Ärgernis erregen bei j-m Ärgerlichkeiten bereiten j-n in Ärger versetzen ärgerliche Situation das Ärgerliche an der Sache ärgerlich groß MOTIVATING CONCEPT außer sich sein vor Ärger grün und gelb vor Ärger
iem. treiteren/pesten, iem. ergeren/ irriteren iem. irriteren, iem. geërgerd/geïrriteerd maken, iem. boos maken een plaag/lastpost zijn, een steen des aanstoots zijn iem. irriteren; problemen/ergernis ver oor zaken, ergernis wekken problemen veroorzaken/iem. problemen berei den iem. kwaad/boos maken, iem. in woede brengen een uiterst onaangename/vervelende/ ergerlijke situatie het irritante/ergerlijke aan de zaak, het vervelende aan de situatie vervelend groot door woede overmand zijn, buiten zinnen zijn/zich zelf van woede zijn je groen en geel ergeren, witheet van woede
EQUIVALENCE IN CONTRASTIVE SEMANTICS
aus Ärger etwas tun zu meinem Ärger zum Ärgernis aller bei aller Ärgerlichkeit voller Ärger zum Ärger GRADING ANGER leiser Ärger heftiger Ärger großer Ärger viel Ärger/wenig Ärger sich sehr ärgern sich schwarz/grün und blau/... ärgern sich maßlos/unglaublich ärgern sich furchtbar ärgern sich zu Tode/krank/ins Grab ärgern sich über die Fliege an der Wand ärgern
van woede/vanuit je woede/boosheid iets doen tot mijn ergernis/ontsteltenis tot een ieders ergernis/woede ondanks alle problemen vol woede, witheet
lichte irritatie grote woede heleboel problemen/weinig problemen bijzonder boos/enorm geïrriteerd zijn zich groen en geel ergeren zich mateloos ergeren zich mateloos ergeren zich dood ergeren zich over alles en nog wat ergeren
INTENSIFIED Zorn, Wut, Raserei BE zürnen, zornig/erzürnt/wütend/ wutent brannt/aufgebracht/ aufbrausend sein
woede, razernij
Zorn haben auf j . zürnt mir von Zorn/Wut erfüllt/ergriffen sein sein Zorn richtet sich gegen der Zorn der Götter trifft ihn Wut im Bauch haben wütende/zornige Stimme wütende Gefühle (wütender Durst (a monsoon raged outside (the debate raged throughout the whole day (fires raged throughout the area
boos/kwaad zijn op iem., toornig/driftig/ woedend/ziedend/furieus/waaiend/ opvliegend zijn, over de rooie zijn woedend zijn op, boos/kwaad zijn op iem. is boos/kwaad op mij van woede vervuld/vol zijn zijn woede richt zich op de woede/toorn van de goden treft hem heel woedend zijn op woedende/driftige toon/stem woedende gevoelens erge/enorme dorst) buiten woedde [tobte] een moesson) het debat woedde [tobte] de hele dag) branden teisterden [wüteten] het gebied)
BECOME sich erzürnen, aufbrausen in Zorn geraten/ in Wut kommen der Zorn packt/überkommt j-n
woedend/kwaad worden driftig/kwaad worden/ woedend raken de woede maakt zich meester van iem.
LOSE Zorn/Wut legt sich/läßt nach
135
136
HENNING WESTHEIDE
CAUSE j-s Zorn/Wut erregen zornig/wütend machen erzürnen
bij iem. woede veroorzaken, iemds woede opwekken, iem. tot razernij bewegen kwaad/woedend maken boos/woedend/kwaad maken
MOTIVATING CONCEPT rot vor Zorn voll Zorn, voller Wut schäumend vor Wut aufbrausend vor Zorn außer sich vor Wut vor Wut kochend aus Wut
rood van woede vol woede schuimbekkend van woede kokend van woede buiten zich zelf van woede koken van woede uit/van woede
GRADING aufflammender Zorn, Raserei heller, lodernder leidenschaftlicher heiliger gerechter dumpfe Wut wilde Wut Zornanfall, Wutanfall (Theater machen
uitbarstende/oplaaiende woede, razernij heftige woede gepassioneerde/hartstochtelijke woede heilige woede oprechte woede felle woede woedeaanval theater maken, veel ophef maken)
WEAKENED Verdruß, Unwille BE Verdruß haben, empfinden verdrossen, ungehalten sein Unwillen erregendes Benehmen BECOME to get annoyed CAUSE Verdruß machen, bringen, bereiten
irritatie, ergernis irritatie voelen, narigheid hebben/voelen knorrig/ontstemd, boos/kwaad zijn aanstoot gevend/wrevel/irritatie opwekkend gedrag geërgerd/geïrriteerd raken
(eine Frau belästigen
irritatie/ergernis opwekken, verwekken/ver oorzaken, irriteren een vrouw lastig vallen)
MOTIVATING CONCEPT zum Verdruß
tot zijn verdriet/ergernis
EQUIVALENCE IN CONTRASTIVE SEMANTICS
137
SUPPRESSED/CURBED Groll, Grimm
wrok
BE Groll/Grimm empfinden, grollen etwas wurmt j-η
wrok koesteren, woedend zijn iets-knaagt aan iem./zit iem. niet lekker
CAUSE j-s Groll erregen
iemands wrok veroorzaken, boosheid op wekken, iem. kwaad maken
MORAL Entrüstung, Empörung BE empört, entrüstet sein über Empörung erfüllt j-η, voller Empörung sein E. empfinden seiner E. Luft machen ein entrüstetes Gesicht BECOME sich entrüsten, sich empören ein Sturm von Entrüstung bricht los CAUSE etwas entrüstet/empört ihn Empörung/Entrüstung erregen ein empörendes Benehmen
verontwaardiging, aanstoot verontwaardigd zijn over vol van woede, over de rooie zijn, diep verontwaardigd zijn zich verontwaardigd voelen/boosheid voelen zijn verontwaardiging/boosheid/woede ventileren een verontwaardigde blik, een verontwaardigd gezicht zich boos maken over, over de rooie gaan, door het lint gaan een golf/storm van ontwaardiging breekt los iets verontwaardigd iem. woede/verontwaardiging opwekken
SPRACHSYSTEM UND SPRACHGEBRAUCH IN DER KONTRASTIVEN LEXIKALISCHEN SEMANTIK ECKHARD HAUENHERM Universität Münster Summary Starting from the statement that the aims of linguistics are (i.) to describe how people talk, and (ii.) to explain why they talk the way they do, I will compare two approaches in contrastive lexical semantics: on the one hand, the lexical field theory in Eugenio Coseriu's approach and, on the other hand, a pragmatic approach which makes use of the notion of 'meaning position'. By examining the power of these two approaches in describing and explaining the different meanings and uses of the German verb stehen in contrast with Italian stare, I will show that lexical field theory with its many restrictions (single-word lexical item, limiting a lexical field to only one part of speech, etc.) does not describe the whole of language use, but only a small part of it, namely the so-called regular uses of both verbs. In addition to this regular usage, speakers of German or Italian use the verbs in many further constructions and utterance forms which cannot be explained with reference to a 'coremeaning' of the verbs, but have to be described as singular uses. A pragmatic approach, which starts from considering the possibilities speakers have of acting verbally in a specified field, brings up a unified description of all uses of a word. It is necessary for this approach to go beyond the limitations of lexical field theory. First, it has to leave the single-word lexical item in favour of a multi-word lexical unit, and second, it has to compare different parts of speech which can be used to express the "same" meaning. This approach does not describe the meaning of single words in opposition to other words but ways-of-use as lexical units. Ways-of-use are phrasal units that are used in predicating function. Their meaning is given by their position in a predicating field. Therefore the global predicating fields have to be structured first according to subfields and second according to meaning positions, which are considered to be the minimal units of lexical meaning. The resulting structure represents a universal basis of comparison and gives us the possibility of a unified description of different ways-of-use in the compared languages. To some extent there are links between the two approaches. The functioning of a lexical item in a specified lexical field has to be analysed as a special way-of-use in the pragmatic approach, the functioning in more than one lexical field therefore has to be classified as multiple ways-of-use of the lexical item.
140
1.
ECKHARD HAUENHERM
Sprachsystem und Sprachgebrauch in der Semantik Zwei Aufgaben bilden meines Erachtens das Arbeitsgebiet der Lingu
istik: Erstens will sie beschreiben, wie die Leute reden, zweitens erklären, warum sie so reden, wie sie reden. Diese etwas saloppe Formulierung, die sich natürlich an Chomskys Adäquatheitskriterien, der Beschreibungsadäquatheit und der Erklärungsadäquatheit orientiert, verdeckt die Tatsache, daß die Lösung der ersten Aufgabe von der Art abhängt, in der die Lösung der zweiten gesucht wird. Sie scheint mir aber die grundlegenden Ziele jeder linguistischen Theorie zu charakterisieren. Die Unterschiede zwischen den einzelnen Theorien be treffen meines Erachtens nur die Frage, was als Erklärung des Sprachge brauchs angesehen wird, und welche Methodik der Beschreibung sich daraus ableitet (vgl. Keller 1991:117). Auch die Strukturalistische Wortfeldtheorie in der Tradition Coserius, die hier als Vertreter eines zeichenorientierten Ansatzes betrachtet werden soll, hat letzten Endes das Ziel, diese Aufgaben zu bewältigen. Sie geht davon aus, daß der Sprachgebrauch auf der Grund lage eines Sprachsystems geschieht (oder zumindest beschrieben werden kann), das in seinem Kernbereich ein System von Zeichen darstellt, wobei grundsätzlich eine eineindeutige Zuordnung von Ausdrucks- zu Inhaltsseite eines Zeichens besteht. Ein pragmatischer Ansatz versucht dagegen, den Sprachgebrauch mit Bezug auf Konventionen und Kommunikations bedürfnisse zu erklären. Gegen einen rein zeichenorientierten Ansatz wie die Wortfeldtheorie führt er ins Feld, daß jener nur einen geringen Teil des tat sächlichen Sprachgebrauchs erklären kann und einen Großteil der Verwen dungsweisen als singulär beschreiben muß.1 Wollen wir die einleitenden Fragen auf unser Thema, die lexikalische Semantik, hin präzisieren, so müßten sie ungefähr folgendermaßen lauten: Welche Wörter benutzen die Sprecher? und Warum benutzen sie gerade diese Wörter? Der vortheoretische Begriff 'Wort' soll hier soviel bedeuten wie der Begriff 'lexikalische Einheit', wobei auch dessen Definition theorieabhängig ist. Für eine Theorie, die Sprache als ein System von Zeichen auffaßt, steht die lexikalische Semantik von vornherein im Kern des Interesses. Für eine Gebrauchstheorie hingegen ist es erforderlich, die lexikalische Semantik aus 1 Coseriu (1978:105) schränkt die Reichweite seines Ansatzes aber selbst ein, indem er davon ausgeht, „daß der Wortschatz nicht im Ganzen, sondern nur partiell strukturiert ist".
SPRACHSYSTEM UND SPRACHGEBRAUCH
141
dem Sprachgebrauch, und das heißt, aus der Bedeutung von Sprechakten abzuleiten (vgl. Hundsnurscher 1996). Eine der größten Schwierigkeiten für einen systemorientierten Ansatz ist die oben schon erwähnte Eineindeutigkeit der Zeichenrelation.2 In einem System kann ein Zeichen nur eine Bedeutung haben, verschiedene Verwen dungsweisen werden auf eine Grundbedeutung, die durchaus sehr abstrakt sein kann, zurückzuführen versucht. Die aktuelle Redebedeutung erschließt sich dann aus der Systembedeutung, zzgl. der vom Kontext beigetragenen Einschränkungen bzw. Konkretisierungen. In vielen Fällen ist aber die Ver wendung einer lexikalischen Einheit durch Routinen bestimmt, die allerhöchstens diachron aus ursprünglichen Bedeutungen herleitbar sind. Für die Beschreibung dieser Phänomene hat Coseriu (1978:44) die Ebene der Norm eingeführt, die er als Summe aller tatsächlichen Realisierungen der vom System vorgegebenen Möglichkeiten beschreibt. Hierzu gehört all das, was nicht systematisch-funktioneil, sondern gesellschaftlich-zufällig ist (vgl. Coseriu 1978:46), z.B. verfestigte Gebrauchsweisen, soweit sie nicht durch die Vorunterscheidung zwischen 'Sprachtechnik und Gesprochenem' aus der Beschreibung ausgeschlossen werden (vgl. Coseriu 1978:35-36). Im Rahmen der lexikalischen Semantik heißt das, daß Verwendungsweisen, die sich nicht aus der systematischen Oppositionalität der Systembedeutungen herleiten lassen, als Normbedeutungen beschrieben werden, die eben nicht Teil der Systembedeutung einer lexikalischen Einheit sind. Demgegenüber stellt ein am Sprachgebrauch orientierter Ansatz ein Verwendungsprofil der Wörter auf, das als Bedeutung derselben angesehen wird. Dieses Verwendungsprofil umfaßt sowohl die 'regelgerechten-systematischen' Bedeutungen als auch die nur routinemäßig verfestigten Gebrauchsweisen, wobei später noch anzuge ben ist, was unter einem 'Verwendungsprofil' verstanden wird.3 Sicherlich muß dabei das Einzelwort, an dem sich die Wortfeldtheorie orientiert, in größeren Zusammenhängen betrachtet werden. Ob es dann noch als lexika lische Einheit taugt, soll später diskutiert werden. Da das System aber auch für den Wortfeldtheoretiker nicht offen vor ihm liegt, sondern nur über den Sprachgebrauch der Sprecher zugänglich ist, 2 Diese wird zwar nicht immer explizit vertreten, ist aber im Zeichenbegriff implizit vorhanden, da einer Ausdruckseinheit eine bestimmte Inhaltseinheit zugeordnet wird und die Verbindung dieser beiden das Zeichen bildet. Sobald eine der Seiten geändert wird, liegt ein neues Zeichen vor. Insofern stellt auch Polysemie eine Beziehung zwischen verschiedenen Zeichen dar, wie Herbermann (1995:170-171) treffend darstellt (vgl. auch Cruse 1986:76). 3 Auch bei Hundsnurscher (1995) wird nicht ganz klar, was mit diesem Begriff gemeint ist. Vgl. die Kritik daran bei Herbermann (1995).
142
ECKHARD HAUENHERM
unabhängig davon, ob man es als reine Abstraktion auffaßt oder ihm psychi sche Realität zugesteht, muß auch ein systemtheoretischer Ansatz seinen Ausgangspunkt in der Beschreibung des Sprachgebrauchs nehmen. Es ist also in jedem Fall vom Sprachgebrauch auszugehen. D.h., wir müssen uns auf ein Korpus, eine Sammlung authentischer oder zumindest quasi-authentischer Äußerungen4 beziehen, das wir beschreiben wollen, und das dem tatsächlichen Sprachgebrauch zumindest einer relevanten Sprecher gruppe annähernd entspricht. 2.
Ein kontrastives Beispiel: stehen vs. stare Nehmen wir an, wir wollten eine kontrastive Beschreibung der Semantik des deutschen Verbs stehen und seiner italienischen 'Entsprechung' stare geben. In unseren Korpora finden wir u.a. folgende Belege (zur Verdeutlichung und zur Abschätzung der Wortfeldstruktur sind auch syn onyme und antonyme Verben bzw. Paraphrasen und mögliche Varianten der Ergänzungen der jeweiligen Äußerungsformen angegeben, abweichende Ausdrücke in der Zielsprache sind unterstrichen): 5 (1) (2) (3) (4) (5)
4
Hinter dem Haus steht eine Leiter:liegt eine Leiter. Dietro la casa c'è.'giace una scala. ...sta una scala (appoggiata al muro). Das Haus steht/liegt in der Stadt. ... liegt/?steht am Meer. La casa si trova in città/al mare. Ich steh'/warte schon eine halbe Stunde an der Bushaltestelle. Aspetto già da una mezz'ora alla fermata. Wir mußten eine halbe Stunde stehen:sitzen:liegen:laufen. Dovevamo stare in piedi.stare seduti:giacere:correre per mezz' ora. Weißt du, wieviel Sternlein stehen/*liegenl* sitzen/Haufen? Sai, quante stelle ci sono in cielo ?
Für die vorliegenden Zwecke genügt es, ein Korpus aus der eigenen Kompetenz heraus zusammenzustellen und die Belege an einem 'richtigen' Korpus oder von mehreren Mutter sprachlern überprüfen zu lassen. Man kommt damit zumindest zu einer Annäherung an den tatsächlichen Gebrauch (vgl. Bergenholtz & Mugdan 1989:141). 5 Diese Belege sind zwar keine authentischen Beispiele; daß die deutschen Verwendungs weisen aber möglich sind und konventionelle Ausdrücke des Deutschen darstellen, habe ich an einem Korpus der deutschen Wochenzeitschrift "Die Zeit" überprüft. Dort kamen allen angegebenen Verwendungen entsprechende Beispiele vor. Für das Italienische stand mir kein Korpus zur Verfügung, alle Übersetzungen und Beispiele wurden aber von Muttersprachlern bestätigt.
SPRACHSYSTEM UND SPRACHGEBRAUCH
(6) (7) (8)
(9)
(10) (11) (12) (13) (14) (15)
(16) (17) ( 18) (19) (20)
143
Die Sonne steht/*sitzt/*liegt/*läuft hoch am Himmel. Il sole è alto (in cielo). Die Teller stehen/*fliegen auf dem Tisch/. ..liegen auf dem Boden. I piatti stanno/sono sul tavolo/...sono sul pavimento. Meine Uhr steht:... läuft/geht (wieder). II mio orologio si è fermato/non funziona. Mi si è fermato l'oro logio. Il mio orologio rifunziona/si è rimesso in funzione. In der Küche steht".Häuft die Standuhr schlecht. Nella cucina I'orologio a pendolo non sta bene. (vgl. : ... non va bene = ... geht falsch) Die Uhr in der Küche steht .läuft. L'orologio nella cucina è fermo iva/funziona. Die Uhr steht/hängt/liegt/Häuft in der Küche. L'orologio sta/è appeso/è nella/in cucina. Die Tür steht offen. La porta è aperta. Das Bad steht/*liegt schon wieder unter Wasser. Il bagno è di nuovo tutto bagnato. Wie hoch steht/steigt das Wasser? Che livello ha l'acqua? Was steht/*sitzt/Hiegt/Häuft heute in der Zeitung? Cosa c'è sul/scrive/dice il giornale? Cosa si dice sul/nel gior nale ? Wie steht's/läuft es in Schalke? (nicht antonym) Come sta/va la partita a Schalke? A quanto stanno a Schalke? Wie steht es um deine Mutter? Wie geht es deiner Mutter? (n.a.) Come sta tua madre? Wie stehst du dazu ? Che cosa (ne) pensi tu? Deine Hose steht :Hiegt ja vor Dreck-.Häuft ja vor Sauberkeit! I tuoi pantaloni sono tanto sporchi che stanno in piedi da soli. Der Anzug stehtI*sitzt Ihnen. ... sitzt/*steht perfekt. L'abito gli sta benissimo.
Vom Italienischen ausgehend, können noch folgende Verwendungswei sen des italienischen Verbs stare gefunden werden, die in die Betrachtung mit einzubeziehen sind:
144
ECKHARD HAUENHERM
(21) Ci starò poco tempo. Ich bleibe nur kurz. (22) Stai bene? Geht es dir gut? (23) Stai fermo! Fermo! Fermati! Bleib ' stehen! Stehenbleiben! Hör auf! (24) Dove stai adesso? - St/abito in via Roma. Wo wohnst du jetzt? - Ich wohne in der Via Roma. (25) Adesso sto tranquillo! Jetzt bin ich ruhiger/beruhigt. (26) Quanto ne sta dentro in questa bottiglia? Wieviel geht in diese/ist in dieser Flasche? (27) Non sta a noi pensarci. Es ist nicht unsere Aufgabe/es ist nicht an uns, darüber nachzu denken/daran zu denken. (28) Sto leggendo. Ich lese gerade. (29) Sto per uscire. Ich bin dabei auszugehen. (30) Io ci sto! Ich bin dabei! Neben diesen regulären Verwendungsweisen existiert im Deutschen noch eine große Anzahl von mit stehen gebildeten Funktionsverbgefiigen, wie z.B. unter Verdacht stehen, in dem Ruf stehen, unter dem Vorzeichen stehen, zu Gebote stehen, zur Debatte/Diskussion/Disposition stehen etc. bis hin zu sog. metaphorischen Verwendungsweisen wie in den Sternen stehen, ins Ge sicht geschrieben stehen u.a., die sich in konventionellen Texten zuhauf fin den, wie ich am Korpus der "Zeit" überprüfen konnte. Sie sind ähnlich zu analysieren wie die italienischen Verwendungen von stare als 'verbo fraseologico' (vgl. die Beispiele [4], [28] und [29]). 2.1 Ein wortfeldtheoretischer Beschreibungsversuch Im Rahmen einer wortfeldtheoretischen Analyse nach Coserius Modell müßte man versuchen, diese verschiedenen Verwendungsweisen des Verbs stehen möglichst auf eine Grundbedeutung des Verbs zurückzuführen. Diese müßte dann in einem Wortfeld einer bestimmten Art von Verben, etwa der Verben der Fortbewegung oder der Positionsangabe, durch seine Opposi-
SPRACHSYSTEM UND SPRACHGEBRAUCH
145
tionen bestimmt werden. Versuchen wir uns zuerst den Gebrauch in den einzelnen Belegen zu verdeutlichen und ansatzweise im Rahmen eines Wortfeldmodells zu erklären. In (1) wird von einer Leiter ausgesagt, daß sie sich hinter einem Haus befindet. Das Verb bringt aber noch einen weiteren Aspekt in die Satz bedeutung mit ein, nämlich daß die Leiter sich in aufrechter Lage befindet (vgl. die abweichende Bedeutung von Hinter dem Haus liegt eine Leiter.) Offensichtlich funktioniert stehen hier im Wortfeld der Verben der Positions angabe. Es steht damit in Opposition zu sitzen, liegen, hängen etc. (vgl. auch Schmidt 1963:57). Das Italienische verwendet für diesen Fall normalerweise esserci oder, auf scala bezogen, die komplexe Verwendungsweise stare appoggiata a. Diese Semantik trifft auch auf (2) zu, jedoch ist der kontrastive Gebrauch in Das Haus liegt am See, der keine Bedeutungsänderung bewirkt, damit nicht erklärbar. Hierbei handelt es sich offenbar um eine jener Kollokationen, die für einen Wortfeldtheoretiker in den Bereich der Norm fallen, was sich auch an der Verwendung von trovarsi im Italienischen bestätigen läßt.6 Auch in (3) und (4) bezieht die Bedeutung von stehen sich in erster Linie auf die Position, wie die Bedeutungsänderungen beim Einsetzen der oppositiven Lexeme zeigen. In (3) wird aber deutlich, wie diffizil die Zuord nung äquivalenter Äußerungen auf der Handlungsebene ist. Im Italienischen würde der Situation die Verwendung von aspettare bzw. essere entsprechen, wobei aber zu fragen ist, ob der propositionale Gehalt dem der deutschen Verwendungs weise entspricht. Im fünften Beispiel, das von Coseriu anhand seiner Vorunterscheidung "Technik der Rede vs. Wiederholte Rede" (vgl. Coseriu 1978:35ff.) aus dem Korpus und damit von der Beschreibung ausgeschlossen würde, scheint sich die Bedeutung auf 'sich befinden (an einem bestimmten Ort)' zu reduzieren. Stehen nimmt hier die Position des Archilexems des Wortfeldes ein, eine Position, an der eher sich befinden anzusetzen wäre. Wie aber (2) zeigt, kann offensichtlich auch liegen eine ähnlich allgemeine Bedeutung annehmen. (6) zeigt meines Erachtens, daß (5) nicht aus dem Korpus auszuschließen ist, da die Verwendungsweise Himmelskörper + stehen eine relativ freie Verwen6 Trotz Befragung mehrerer Sprecher des Deutschen war nicht endgültig auszumachen, ob die Verwendung von stehen oder liegen in diesem Zusammenhang geläufiger ist und ob der Kontrast zwischen ...steht in der Stadt und ?... liegt in der Stadt bzw. ? ....steht am See und ...liegt am See für alle Sprecher gleichermaßen besteht. Auch das überprüfte Korpus der "Zeit" bietet für beide Verwendungs weisen Beispiele.
146
ECKHARD HAUENHERM
dung von stehen darstellt. Die italienischen Entsprechungen zeigen deutlich die Allgemeinheit der Bedeutung. Das Italienische verwendet hier essere 'sein'. (7) scheint sich ebenfalls mit der Bedeutung 'sich befinden (an einem bestimmten Ort)' paraphrasieren zu lassen, man muß sich aber fragen, in welcher Beziehung dieses Beispiel zu Verwendungsweisen anderer Verben steht, wie z.B. Die Teller stehen an der Wand, oder Die Teller liegen auf dem Boden. Offensichtlich bringt stehen etwas in die Semantik des Satzes mit ein, das man als 'sich in geordneter Weise befinden' umschreiben kann, denn die Lage allein spricht nicht für den Bedeutungsunterschied zwischen Die Teller stehen auf dem Tisch, und Die Teller liegen auf dem Boden. Mit Schmidt (1963) könnte davon ausgegangen werden, daß in solchen Zusam menhängen die Art der Ortsergänzung für den zusätzlichen Bedeutungsunterschied verantwortlich ist. Dann müßte aber der Satz Die Teller liegen auf dem Tisch, dieselbe Bedeutung haben wie derjenige mit stehen. Das scheint mir nicht zuzutreffen. Die Verwendung von essere in der italieni schen Entsprechung weist ebenfalls auf einen Bedeutungsunterschied hin. Doch wie ist das mit den Beispielen (8) bis (11) zu verbinden? Wie der Kontrast in (8) zeigt, bildet stehen hier ganz andere Oppositionen. Es gehört zu den Verben der Bewegung und drückt die Abwesenheit von Bewegung aus. Daß es sich tatsächlich um eine weitere lexikalische Einheit handelt, beweist auch Beispiel (10). Als Verb der Positionsangabe erfordert stehen eine von ihm abhängige Ortsangabe, weshalb (10) nicht als Bestimmung des Ortes der Uhr gelesen werden kann, da die Ortsangabe hier Teil der Subjekts-NP ist. In Beispiel (11) führt Topikalisierung der Präpositionalphrase zur Mehrdeutigkeit des deutschen Satzes. Bis hierhin können wir also festhalten, daß stehen mindestens in zwei Wortfeldern funktioniert, einmal als Verb der Bewegung, andererseits als Verb der Positionsangabe, eine Tatsache, auf die auch schon Wilhelm Schmidt (1963:57) hinweist, wenn er stehen als Gegensatzwort sowohl zu gehen als auch sitzen, hängen, liegen etc. auffaßt. Hier zeigt sich deutlich, daß wir es bei stehen im Sinne der Wortfeldtheorie mit verschiedenen Einheiten zu tun haben. Die italienischen Entsprechungen unterstützen diese Annahme. Anstelle von stare tritt hier fermarsi auf, was aber in ähnlichen Oppositionen steht wie stehen, vgl.: funzionare -funktionieren, andare - laufen. (12) kann tatsächlich in der Analyse übergangen werden, wenn offen stehen als eigene lexikalische Einheit betrachtet wird. Dafür spricht, daß die Opposition geschlossen stehen in diesem Zusammenhang nicht existiert.
SPRACHSYSTEM UND SPRACHGEBRAUCH
147
Ebenso sind weitere Wortbildungen wie vorstehen, anstehen und erst recht verstehen, gestehen etc. zu behandeln, auch wenn die Kriterien für die Aus bzw. Eingliederung solcher Einheiten im Einzelfall nicht immer problemlos anzugeben sind. In den folgenden beiden Beispielen scheint mir aber keine der bisherigen beiden Bedeutungen die Verwendungsweisen erklären zu können. In (13) wäre es sicherlich übertrieben zu sagen, das Bad "befindet sich" unter Wasser, im Sinne einer Ortsangabe. Die Semantik von stehen betrifft sicherlich nicht die räumliche Lage des Badezimmers, auch wenn die Ortsangabe eine solche Lesart zulassen würde. Vielmehr scheint diese Ver wendungsweise mit der im nächsten Beispiel zusammenzuhängen. Auch hier kann man nicht sagen: Das Wasser befindet sich so und so hoch. Es wird zwar auf eine Senkrechte abgezielt, allerdings als Höhen-, nicht als Ortsangabe. Auch die zweite Bedeutung als Verb der Bewegung scheint den Verwendungsweisen nicht gerecht zu werden. Die italienischen Entspre chungen weisen darauf hin, daß es sich bei diesen Beispielen um zwei unterschiedliche Verwendungsweisen handelt, die beide von den bisherigen Bedeutungsangaben nicht abgedeckt werden. Auch für (15) bedarf es schon einiger Abstraktionskunst, um es mit einer der bisherigen Bedeutungen von stehen zu interpretieren. "Befindet sich" der Inhalt in der Zeitung? Ich denke, hier ist es erforderlich, minde stens noch eine weitere Bedeutung von stehen anzunehmen, zumal es sich dabei um eine sehr freie Verwendungsweise handelt, man denke an Sätze wie Was steht in dem Buch? Auf der Anschlagtafel steht, daß der Zug fünf Minuten Verspätung hat. etc. Das Italienische kennt für stare offenbar dieselbe Verwendungsweise. Für (16) und (17) ließe sich behaupten, es handelt sich um Verwen dungsweisen, die vom Nomen Zustand abgeleitet sind und nicht mit dem Verb stehen in seiner eigentlichen Bedeutung zusammenhängen. Doch was heißt das? Zustand ist von Stand abgeleitet und dieses von stehen, kann eingewendet werden. Der Bezug zu einem speziellen Nomen zeigt aber, daß wiederum von einer neuen Verwendungsweise auszugehen ist. Hier ist ein Wortfeld von 'Zustandsverben' anzusetzen, zu dem evtl. auch laufen, sicher lich aber gehen in den angegebenen Verwendungen zu zählen sind. Das Italienische verhält sich ähnlich. Demgegenüber zeigt (18), das auf den ersten Blick dieser Verwen dungsweise zu entsprechen scheint, wieder eine anderen Bedeutung. Die nominale Ableitung ist Stellung, was für die Annahme zweier verschiedener
148
ECKHARD HAUENHERM
Einheiten in (17) und (18) spricht. In (18) ist der syntaktische Rahmen mit zu berücksichtigen, da diese Verwendungsweise immer mit stehen zu etw. realisiert wird. Die italienische Entsprechung macht deutlich, daß es sich um eine spezielle Verwendungsweise handelt, die den Einstellungsverben zuzu ordnen ist. (19) - das eigentlich metaphorisch ist und somit aus dem Unter suchungsgebiet der strukturellen Semantik herausfällt - korreliert wiederum sehr gut mit der ersten Bedeutung von stehen, 'sich aufrecht befinden'. Auch die italienische Entsprechung, die hier allerdings dasselbe Verb verwendet wie bei personalen Subjekten, bestätigt den Zusammenhang. Sie zeigt eine ähnliche Metaphorik wie das Deutsche. (20) eröffnet völlig neue Bedeutungsperspektiven, da es nicht nur nicht zu den bisherigen Bedeutungen paßt, sondern auch neue syntaktische und kontextuelle Muster bietet. Auch hier ist ein neues Wortfeld anzusetzen (zu dem erstaunlicherweise auch sitzen gehört, das aber nicht an die Kollokation mit "Kleidungsstück" gebunden ist; vgl. Der Anzug sitzt perfekt, vs. Die Schraube sitzt fest.). Stare funktioniert in ähnlichen Kontexten. Gehen wir von diesem aus, fallen weitere Verwendungsweisen auf, von denen einige in den letzten zehn Beispielen wiedergegeben sind. In (21) liegt eine typische Verwendung des italienischen Verbs vor, die dem deutschen bleiben entspricht. (22) gleicht der deutschen Verwendungs weise in (16) und (17). Die deutsche Entsprechung dieses Beispiels zeigt aber, daß sich beide Verwen dungsweisen nicht vollständig decken, da die Fragen Stehst du gut? bzw. Steht es dir gut? nicht die hier erforderliche Bedeutung repräsentieren. (23) ist sicherlich mit (21) zusammenzufassen, wobei aber, wenn vom Wort als lexikalischer Einheit ausgegangen wird, die Entsprechungen nicht eindeutig zu klären sind, da auch fermarsi und rimanere mit bleiben wiedergegeben werden können, vgl.: Quanti giorni vi fermate?/Quanto rimanete? - Wie lange bleiben Sie? (24) bietet eine Verwendungs weise von stare, die im Deutschen nur mit wohnen wiedergegeben werden kann. Der Unterschied zwischen stare und abitare ist im Italienischen wahrscheinlich eher funktionaler, als lexikalischer Art.7 7 Diaphasische Unterschiede, bzw. sprachliche Stilebenen, ordnet Coseriu (1978:39) verschiedenen funktionalen Sprachen zu. Die Beschreibung lexikalischer Oppositionen schränkt er (1978:40) auf eine funktionale Sprache ein. Dabei scheint mir aber erstens die Abgrenzung verschiedener 'funktionaler' Sprachen problematisch zu sein (es könnte sich um ein zirkuläres Verfahren handeln, da sich die Abgrenzung evtl. nur an der unterschiedlichen
SPRACHSYSTEM UND SPRACHGEBRAUCH
149
(25) scheint wieder eine neue Verwendungsweise darzustellen. Hier liegt ein zu Beispiel (5) entgegengesetzt gelagerter Fall vor. Während das Italienische ein spezielles Verb hat, steht im Deutschen sein als 'Zustandsverb' zur Angabe eines emotionalen Zustandes zur Verfügung. Ähnlich ist Beispiel (26) zu interpretieren, auch wenn es sich nicht auf emotionale, sondern auf Füllungszustände bezieht. Einen besonderen Fall stellt (27) dar. Dem italienischen stare a entspricht im Deutschen ein Syntagma. Hier werden die Probleme einer wortbezogenen Beschreibung deutlich. Es stellt sich die Frage, in welches Wortfeld stare einzuordnen ist. Auch (28) und (29) zeitigen dieselben Probleme. Stare kann kaum noch lexikalisch erfaßt werden. Es funktioniert, ähnlich einem Modalverb, als 'Aspektverb' ('verbo fraseologico'). Als Be deutungseinheit ist mindestens die jeweilige syntaktische Konstruktion anzusehen, wie stare + Gerundio in (28), stare per in (29). Für (30), das ebenfalls eine eigene Verwendungsweise darstellt und in dieser Bedeutung an bestimmte diskursive Kontexte wie Zustimmungen gebunden zu sein scheint, ist starci anzusetzen. Für solche Einheiten sind aber meines Wissens noch keine Wortfelder beschreiben worden, vielmehr werden sie häufig als grammatische Phänomene aus der lexikalischen Semantik ausgeschlossen. 2.2 Probleme der wortfeldtheoretischen Beschreibung Es scheint mir offensichtlich, daß Verben wie stehen und stare erstens in verschiedenen Wortfeldern funktionieren, und zweitens innerhalb dieser Wortfelder nicht in gleichbleibenden Oppositionen stehen. Dies heißt aller dings noch nicht, daß sie nicht auch innerhalb eines Strukturalistischen Ansatzes beschreibbar sind. Ein solcher Fall ließe sich evtl. unter den Stichworten 'Neutralisierung' und 'Synkretismus' behandeln. Trotzdem bleiben aber Verwendungsweisen übrig, die damit nicht erklärt werden können, aber durchaus als konventionell einzustufen sind, wie z.B. (14) und (15), oder auch (28) bis (30). Es stellt sich die Frage, ob ein abstraktes System von Zeichen tatsächlich in der Lage ist, den Gebrauch der Wörter zu erklären. Die Annahme eines Sprachsystem hat nur dann einen Wert, wenn sie dazu beiträgt, die Regularitäten, die hinter den auf den ersten Blick chaotischen Verwendungsweisen stehen, offenzulegen. Verwendung lexikalischer Einheiten zeigt), zweitens die Gefahr zu bestehen, daß nicht eine natürliche Sprache, sondern nur ein künstlich hergestellter Ausschnitt derselben beschrieben wird.
150
ECKHARD HAUENHERM
Für viele Verwendungsweisen ist eben dies nicht der Fall, da sie sich nicht nur durch die Systembedeutung der Worte ergeben, sondern aufgrund mehr oder weniger fester Gewohnheiten sprachlicher Gemeinschaften existieren, die aber als zur Einzelsprache gehörig angesehen werden müssen. Dies ist der Fall bei besonderen Kollokationen, wie auch bei einem Großteil der sog. übertragenen Bedeutungen, deren Übertragung in Zeiten entstand, zu denen die Wörter noch andere Gebräuche hatten.8 Für diese Fälle hält sich Coseriu (1978) u.a. die Ebene der 'Norm' bereit, die er vom Sprachsystem unterscheidet, und in der das untergebracht wird, was sich eben nicht 'systemgerecht'9 erklären läßt. Um eine Einzelsprache als funktionales System, und im Rahmen der strukturellen lexikalischen Semantik heißt das, als System lexikalischer Oppositionen beschreiben zu können, muß entweder ein großer Teil des Sprachgebrauchs außer Acht gelassen werden (übertragene Verwendungs weisen, feststehende Wendungen etc.) oder aber versucht werden, mit Hilfe der Annahme sehr abstrakter Systembedeutungen möglichst viel im System zu begründen. Die Schwierigkeit der ersten Vorgehensweise liegt darin, daß das System damit auch nur einen geringen Teil des wirklichen Sprach gebrauchs erklären kann, die zweite Vorgehensweise hat in erster Linie das Problem, daß bei der Annahme sehr abstrakter und auch allgemeiner Bedeutungen nicht jede Verwendungs weise herleitbar ist, bzw. auch andere als die konventionalisierten. So könnte man für stehen z.B. eine Bedeutung wie 'sich an einem Ort in Ruhe befinden, evtl. in aufrechter Lage' annehmen. Schon das "evtl." macht deutlich, daß für viele Verwendungs weisen spezielle Kriterien angenommen werden müßten, die vielfach eben doch von Kollokationen abhängen und damit wieder zu einer nicht eineindeutigen signifiant-signifié Zuordnung fuhren, so daß zumindest von mehreren Einheiten auszugehen ist. Außerdem kann die Annahme einer Bedeutung, die allen Verwendungs weisen gerecht zu werden im Stande ist, dazu führen, daß sie nur noch eine Abstraktion darstellt, der selbst keine 8 Einen Ansatz zur Beschreibung der Entstehung 'übertragener Bedeutung' aus der Sicht einer handlungstheoretischen Semantik bietet Fritz (1995). 9 Coserius Definition der Norm erscheint mir widersprüchlich. Einerseits sagt er, sie enthielte das, "was unabhängig vom funktionalen Bereich traditionell fixiert ist" (1978:41), andererseits heißt es: "In der Norm findet sich das, was auf Grund des Systems bereits realisiert wurde." (1978:44). Der Unterschied zwischen "traditionell fixiert" und "auf Grund des Systems realisiert" ist aber von eminenter Bedeutung, da sich daraus ergibt, ob bestimmte Kollokationen als Teil der Norm die Systembedeutung beeinflussen, d.h. in sie einfließen müssen, oder ob sie als nicht-systematische Bestandteile des Sprachgebrauchs aus der Untersuchung ausgeschlossen werden können.
SPRACHSYSTEM UND SPRACHGEBRAUCH
151
Verwendungsweise entspricht. Dies widerspricht m.E. aber dem Bedeutungswissen der Sprecher, da diese in den meisten Fällen eine Bedeutung als konkrete Verwendungsmöglichkeit einer Einheit in einer Äußerung angeben können, so daß davon auszugehen ist, daß Bedeutungen nicht als abstrakte Einheiten gespeichert, sind, sondern als Verwendungsmöglichkeiten. (Dieses Problem versucht die Prototypensemantik zu umgehen, indem ein Prototyp nicht als eine Abstraktion aufgefaßt wird, sondern als eine konkrete Einheit, die in den relevanten Merkmalen für andere Verwendungsmöglichkeiten maßgeblich ist.) Vor ähnlichen Problemen steht auch ein dekompositioneller Ansatz, da hier ebenfalls die Merkmale eine eineindeutige Zuordnung der beiden Zeichenseiten ermöglichen sollen.10 2.3 Eine Alternative ? Diese Schwierigkeit vermeidet teilweise der frühe Ansatz Wilhelm Schmidts (1963), der davon ausgeht, daß stehen in seiner Hauptbedeutung mindestens drei Bedeutungsaspekte aufweist, 'auf die Füße gestützt', 'aufrecht' und 'an einer Stelle verharren', wobei als Grundkontext "von Menschen und Tieren gebraucht" angenommen wird (1963:56). Im Gegen satz zur Merkmalsemantik wird damit aber nicht die Bedeutung von stehen definiert, sondern es werden Aspekte genannt, die in der, wie Schmidt es nennt, jeweiligen "aktuellen Bedeutung",11 also der Verwendung, in unter schiedlichem Maße relevant sein können. Damit ergibt sich die Möglichkeit, den Einfluß des syntaktischen und semantischen Kontextes bei der Konsti tuierung der aktuellen Bedeutung zu erfassen. Aber auch hier besteht das Problem, daß die verschiedenen Bedeutungen unter einer allgemeinen, der sog. lexikalischen Bedeutung subsumiert werden sollen (vgl. S. 32). Die Probleme, denen wir in den obigen Beispielen begegnet sind, treffen somit auch auf diesen Ansatz zu, jedenfalls soweit versucht wird, die lexikalische Bedeutung am Einzelwort festzumachen. So lassen sich die Verwendungen von stehen in den Beispielen (12) bis (15) nur schwerlich erfassen. Auch Schmidt nimmt daher neben den durch den lexikalischen Kontext bestimmten Bedeutungen auch sog. phraseologische an, bei denen die Verwendung des 10 Ein weiteres Problem dieses Ansatzes, darauf weist Lutzeier (1985:112) hin, besteht darin, daß die Bedeutung zerlegt wird, es aber offensichtlich Evidenzen dafür gibt, daß nicht Merkmale, sondern ganze Bedeutungen im Gehirn mit einem Lautbild verbunden sind. 11 Seine Unterscheidung zwischen lexikalischer und aktueller Bedeutung entspricht in etwa derjenigen Pauls in "usuelle" und "okkasionelle Bedeutung" (vgl. Paul 1995:75).
152
ECKHARD HAUENHERM
Verbs und seiner Begleiter einen eigenständigen Sinn repräsentiert. Es handelt sich hierbei zwar um feste Fügungen, in denen die einzelnen Bestandteile nicht austauschbar sind, aber noch nicht um metaphorische Redeweisen. Neben unter Wasser stehen können hier nach Schmidt (1963:74) auch Ausdrücke wie in dem Ruf stehen, seinen Mann stehen und ähnliches angeführt werden, die sich dadurch auszeichnen, daß im nominalen Referenz ausdruck das Objekt explizit genannt wird. Auch bei Schmidt zeigt sich daher, daß die sog. lexikalische Bedeutung tatsächlich nur einen kleinen Teil der vorkommenden Verwendungsweisen erklären kann. Dieses Problem verschärft sich noch, bezieht man den Sprachvergleich mit ein. Dann nämlich stellt sich die Frage, was im Rahmen einer solchen Beschreibung in verschiedenen Sprachen verglichen werden soll. Beschränkt man sich auf die 'System-' oder 'lexikalische Bedeutung' einer Einheit, wird man in den meisten Fällen nur feststellen können, daß diese sich, je weniger Verwendungsweisen berücksichtigt werden, desto eher entsprechen. Das, was aus der Erklärung der lexikalischen Bedeutung herausfallt, kann dabei nur als Einzelfall gegenübergestellt werden, wobei die Frage ist, ob die entspre chende Äußerungsform auch in der Vergleichssprache einen solchen singulären Fall darstellt. Daraus folgt die Schwierigkeit, daß sich je nach Vergleichsrichtung unterschiedliche Beschreibungen ergeben. Vergleicht man die deutschen Verwendungsweisen von stehen mit ihren italienischen Entsprechungen bzw. die italienischen Verwendungsweisen von stare mit ihren deutschen Entsprechungen, wird evident, wie schwierig eine Zuordnung von Übersetzungsäquivalenten auf der Wortebene ist. Diese Schwierigkeit zeigt einerseits die Unterschiede im Bau der Wortfelder beider Sprachen als auch das unterschiedliche Verwendungsprofil beider Einheiten. In relevanten syntaktischen und semantischen Kontexten haben beide Sprachen unterschiedliche Verwendungsweisen ausgebildet, die in der jeweils anderen Sprache mit anderen Verben oder anderen Konstruktionen wieder gegeben werden müssen. (23) zeigt, daß, obwohl die Verwendung des ent sprechenden Verbs auf eine Bedeutungsäquivalenz hinweist, das Auftreten einer entsprechenden Einheit in vergleichbaren Kontexten keinen Beleg für eine Funktionsgleichheit darstellt. Das Verb erfüllt im italienischen Beispiel eine andere Funktion als im deutschen, da die Betonung im Italienischen auf fermo liegt (z.B. genügt dieser Ausdruck, um die entsprechende Aufforde rung zu vollziehen Fermo!), im Deutschen aber auf stehen (hier muß die Kurzform Stehenbleiben! heißen). Für eine kontrastive Analyse reicht es daher nicht aus, nur das Einzelwort zu betrachten. Die Einbeziehung der
SPRACHSYSTEM UND SPRACHGEBRAUCH
153
syntaktischen Konstraktion sowie semantischer Kollokationen ist unmittelbar erforderlich, um überhaupt eine Vergleichsmöglichkeit herzustellen. Darüber hinaus muß bei der Analyse der ganze Bereich der Verwendungsweisen erfaßt werden, um zeigen zu können, in welchen Teilbereichen sich die Verwendung von Einheiten zweier Sprachen decken und in welchen nicht. Es gibt keine Möglichkeit, von den Verwendungsmöglichkeiten in einer Sprache auf die Verwendungsmöglichkeiten in einer anderen zu schließen, noch auf weitere Verwendungsweisen innerhalb derselben Sprache. Jede Verwendungsweise ist daher erst daraufhin zu überprüfen, ob sie eine eigene Bedeutung repräsentiert oder ob sie sich mit anderen Verwendungsweisen deckt. 2.4 Eine pragmatische Alternative Soll daher der ganze Bereich der Verwendungsweisen erfaßt werden, scheint mir ein Beschreibungsansatz, der seinen Ausgangspunkt vom Sprachgebrauch nimmt, im Vorteil zu sein. Vom Sprachgebrauch seinen Ausgang zu nehmen, heißt auch, vom Handlungscharakter der Sprache auszugehen. Wir gebrauchen Sprache zum Vollzug kommunikativer Handlungen, die als Sprechakte beschreibbar sind. Zum Vollzug solcher Akte verwenden wir nicht einzelne Wörter, sondern Äußerungen, die nicht notwendig mit Sätzen zusammenfallen müssen. Um also auf die Ebene der lexikalischen Semantik zu gelangen, müssen wir uns vorab fragen, welchen Beitrag die Wörter in Sprechakten zum Vollzug solcher Handlungen leisten. Dieser Beitrag, der in der lexikalischen Semantik beschrieben werden soll, liegt in ihrer Funktion im propositionalen Akt.12 Mit Weigand (1993:264; 1996:156-157) kann die Prädikation als die grundlegende Funktion lexika lischer Einheiten angenommen werden. Auf der Äußerungsseite wird diese Funktion nicht durch Einzelwörter, sondern durch sog. Verwendungsweisen ausgedrückt, die nach Weigand als komplexe Einheiten syntaktisch definiert sind. Die Bedeutung dieser Verwendungsweisen wird als Konfiguration kleinster Bedeutungseinheiten oder Bedeutungspositionen repräsentiert (vgl. Weigand: "Contrastive Lexical Semantics", in diesem Band).
12 Obwohl dies genauer analysiert werden müßte, da es durchaus auch an der Wahl der Lexeme liegen kann, ob eine Äußerung z.B. als Frage oder als Vorwurf verstanden werden kann, z.B. stinken vs. riechen oder Mistkerl vs. Kumpel etc. Eine direkte Zuordnung der Lexeme zur kommunikativen Funktion einer Äußerung scheint aber ein Irrweg zu sein (vgl. Rolf 1992). Eine Lösung könnten die semantischen Muster bieten, wie sie von Hindelang (1978) eingeführt wurden.
154
ECKHARD HAUENHERM
Der handlungstheoretische Semantiker steht also vor der Aufgabe, die im Korpus belegten Äußerungen zu bestimmten Mustern, den Verwendungs weisen, zusammenzufassen, und diese dann in einem zweiten Schritt Bedeu tungspositionen im Rahmen der Prädikationsfelder zuzuordnen. Die so gewonnene Strukturierung der Verwendungsweisen umfaßt Einheiten unter schiedlichen Status, einerseits die freien Verwendungen eines Wortes, andererseits die festen Kollokationen und syntaktisch beschränkten Aus drücke. Verwendungsweisen beziehen sich demnach nicht auf einzelne Wörter, sondern auf syntaktisch strukturierte Einheiten, deren Bandbreite von unterschiedlich engen Kollokationen über Funktionsverbgefüge bis hin zu ganzen Äußerungsformen (Sätzen) reicht. Die Frage ist nun, wie hier der kontrastive Vergleich ansetzen kann. Es ist unbestritten, daß jede Sprache nicht nur die außersprachliche Wirklichkeit anders wiedergibt, sondern auch andere Arten der Gliederung aufweist (vgl. Coseriu 1978:13). Gerade fur eine kontrastive Semantik ist daher die Frage der Vergleichsgrundlage von entscheidender Bedeutung. Die Wortfelder der strukturellen Semantik stellen Coseriu (1978:55) zufolge keine Sachbereiche dar, sondern sprachliche Bereiche von Einheiten 'ähnlicher' Bedeutung, die durch funktionale Oppositionen der in ihnen funktionierenden Lexeme gegliedert sind. Aus der Sicht der Gebrauchssemantik hat diese Theorie zwei Nachteile. Erstens sind Wortfelder an eine Wortart gebunden, was, wie oben deutlich geworden sein sollte, dem tatsächlichen Sprachgebrauch widerspricht, da wir in vielen Fällen nicht nur innerhalb derselben Wortklasse, sondern zwischen Nominalphrasen, Verben, Nomen und phraseologischen Konstruktionen wählen können (vgl. Beispiel [4] im Italienischen, in dem die Wahl zwischen stehen und sitzen lautet: stare in piedi vs. stare seduti). Zweitens können mit funktionalen Oppositionen viele Routinen in den Verwendungsweisen nicht erklärt werden (was im Rahmen der strukturellen Semantik aber auch nicht beabsichtigt wird, da Routinen nicht als zum System, sondern allenfalls als zur Norm gehörig betrachtet werden). Auch die Frage nach der Struktur aller Wortfelder einer Sprache ist bisher in der strukturellen Semantik noch nicht geklärt worden (vgl. Anmerkung 1). Da die Gliederung der Wortfelder rein sprachlich ist, scheint mir das auch nicht ohne weiteres möglich, da je nach Art der Gliederung die Menge der Wortfelder variiert. Vor einem ähnlichen Problem steht auch die Merkmalsemantik, da die Anzahl der Merkmale, die zur Beschreibung aller Lexeme notwendig sind, nicht vollständig angegeben werden können (vgl. a. Coulmas 1977:38).
SPRACHSYSTEM UND SPRACHGEBRAUCH
155
Der gravierendste Nachteil für eine kontrastive lexikalische Semantik von der Wortfeldtheorie her scheint mir aber darin zu liegen, daß diese zwar in der Lage ist, die einzelsprachliche Strukturierung der Wortfelder zu beschreiben, sich aber mit den obigen Beschränkungen gerade der Ver gleichsmöglichkeiten beraubt. Indem sie nämlich erstens nur auf die jeweilige Wortart beschränkt bleibt13 und sich zweitens die einzelsprachliche Struk turierung als Ziel setzt, werden die Möglichkeiten, der sich die Sprecher einer Sprache bedienen können, nicht als Wahlmöglichkeiten sichtbar. Damit fehlt ein Mittel, die Handlungsmöglichkeiten, die ein Sprecher einer anderen Sprache hat, um dasselbe Ziel zu erreichen, diesen gegenüberzustellen. Man kann zwar die Strukturen der Wortfelder, z.B. der Verben der Bewegung des Italienischen und des Deutschen, vergleichen, damit ist aber nichts darüber gesagt, daß der Italiener stare in piedi verwenden muß, wo der Deutsche nur stehen gebrauchen kann, und warum im Deutschen Uhren auf zweierlei Art stehen können, im Italienischen aber nur orologi a pendolo stanno ("ste hen"). Schon die vorliegenden Beispiele zeigen, daß sich die Verwendungs weisen von stehen und stare nur zu einem geringen Teil decken, womit ein Problem offengelegt ist, das in der Übersetzungswissenschaft eine Binsen wahrheit ist: die Unmöglichkeit der wörtlichen Übersetzung. Solange man in einer an Verwendungsweisen orientierten Beschreibung diese auch nur im Rahmen der Verwendungsprofile einzelner Einheiten kontrastiert, werden die Probleme nicht gelöst. Auch dann zeigt sich nur, daß in der einen Sprache ein Wort oder eine Konstruktion in der einen Weise verwendet wird, in der anderen Sprache eine entsprechende Einheit aber in anderer Weise. Der Vergleich würde sich ebenso wie in der Wortfeldtheorie auf das Gegenüber stellen der untersuchten Einheiten beschränken. Für eine kontrastive Gebrauchssemantik ist daher das Prinzip der Sinn bereiche hilfreich und aus der Wortfeldtheorie übernehmbar. Diese können als Prädikationsbereiche aufgefaßt und somit als Bedeutungsbereiche be schrieben werden. Bereits die von Hundsnurscher & Splett (1982) angeführ ten Großgruppen für Adjektive stellen in diesem Sinn Bedeutungsbereiche dar. Ähnlich ist der Versuch Ballmers & Brennenstuhls (1986) einzuordnen, den VerbWortschatz nach Verbkategorien und Modellen bzw. Modellgruppen zu ordnen (vgl. auch Ballmer 1990:179). Für eine Beschreibung des ganzen 13
Coseriu (1978:91) schränkt die Oppositionalität innerhalb der Wortfelder weiterhin auf "gleich strukturierte" Einheiten ein, d.h. nicht nur auf die Wortart, sondern auch auf den gleichen morphologischen Status innerhalb der sekundären Strukturen. Demgegenüber betont schon Greimas (1971:63) das "Prinzip der Äquivalenz ungleicher Einheiten beim Vergleich natürlicher Sprachen".
156
ECKHARD HAUENHERM
Wortschatzes muß aber die Beschränkung auf die Wortart aufgegeben werden, da sie dort dieselben Probleme verursacht wie in der auf Wort klassen beschränkten Wortfeldtheorie. Die Prädikationsbereiche sind daher zu Verwendungsweisen in Beziehung zu setzen, die unterschiedliche Wort arten und Konstruktionsmöglichkeiten enthalten, wie Weigand (1995) es for dert, gerade auch im Hinblick auf den Sprachvergleich, da hier natürlicher weise keine Deckungsgleichheit in der Extension der Wortarten vorliegt. Prädikationsbereiche und Bedeutungspositionen sind von vornherein als übereinzelsprachliche Einheiten anzusehen, denen die einzelnen Verwen dungsweisen einer Einzelsprache zugeordnet werden können« Wichtig ist, im Auge zu behalten, daß Prädikationsbereiche und Bedeutungspositionen heuri stische Mittel zur Beschreibung der Bedeutung sind. Mit ihnen wird keinerlei Aussage über die Wirklichkeit selbst getroffen. Sie dienen der Vermeidung eines Sprachrelativismus, der die Vergleichbarkeit der Sprachen auf dieser Ebene von vornherein ausschlösse. Man kann sich das folgendermaßen verdeutlichen: Sprachen sind vergleichbar, und zwar auf jeder Ebene, weil sie dem Menschen zur Bewältigung kommunikativer Aufgaben dienen. Die Aufgaben selbst sind für alle Menschen weitgehend dieselben, die Art der Aufgabenbewältigung unterscheidet sich jedoch von Sprache zu Sprache. Sprachen vergleichen heißt daher, die Unterschiede in der Bewältigung kom munikativer Aufgaben aufzeigen. Für die kontrastive lexikalische Semantik folgt daraus, daß sie die Unterschiede in der Art, wie über die Dinge geredet wird, untersucht. Dies geschieht, indem die Verwendungsweisen bestimmten Prädikationsbereichen zugeordnet werden, die nach Bedeutungspositionen strukturiert werden können. Unter der Annahme einer quasi-universellen Bedeutungsstruktur ergeben sich aus der Art und den Unterschieden der Zu ordnung innerhalb verschiedener Sprachen die Vergieichsmöglichkeiten. Im Gegensatz zur Wortfeldtheorie ist zu betonen, daß nicht einzelne Wörter zugeordnet werden, sondern Verwendungsweisen. Damit umgeht man das Problem, das sich ergibt, wenn in einer Sprache ein syntaktisch strukturierter Ausdruck verwendet wird, dem in der anderen eine morphologische Einheit gegenübersteht. Nach der Wortfeldtheorie sind solche Verwendungsweisen nicht vergleichbar, aus pragmatischer Sicht zeigen sie aber gerade die unterschiedliche Kompetenz der Sprecher verschiedener Sprachen auf. Als erstes ist zu fragen, wie der gesamte Prädikationsbereich sinnvoll zu gliedern ist. Einen Versuch hierzu stellt die Unterteilung in sechs Prädika tionsfelder dar, die Weigand ("Contrastive Lexical Semantics", in diesem Band) aufstellt. Die sechs Bereiche, die sich an den grundlegenden
SPRACHSYSTEM UND SPRACHGEBRAUCH
157
Fähigkeiten des Menschen orientieren, sind: AWARENESS, ACTION, MOTION, EMOTION, SOCIAL BEHAVIOUR und RATIONALITY. Diese bilden eine oberste Gliederungsebene, der weitere Subkategorisierungen folgen müssen. So ist der Bereich EMOTION Z.B. nach den verschiedenen Basisemotionen wie ANGER, LOVE etc. und möglichen Bedeutungspositionen wie BE, BECOME, LOSE, etc. zu untergliedern (vgl. auch Weigand 1998). Ähnliche Unterteilun gen müssen für die anderen Großbereiche noch erarbeitet werden. Gehen wir davon aus, daß das Funktionieren des Verbs stehen in verschiedenen Wort feldern ein Hinweis auf die Zugehörigkeit der Verwendungsweisen zu ver schiedenen Prädikationsfeldern ist, so können wir in einer ersten Annäherung folgende Zuordnungen treffen. In den Beispielen, in denen wir stehen als Verb der Positionsangabe beschrieben haben, gehört seine Verwendungs weise sicherlich in den Bereich AWARENESS, dort, wo es als Verb der Bewegung beschrieben wurde, zu MOTION bzw. ACTION. Die jeweiligen funktional äquivalenten italienischen Verwendungsweisen, sei es von stare oder eines anderen Wortes, sind per definitionem demselben Bereich zuzuordnen. 2.5
Ein pragmatischer Beschreibungsversuch Gehen wir die Beispiele im einzelnen durch. Da noch nicht alle Prädikationsfelder nach Bedeutungspositionen strukturiert sind, kann hier nur eine erste Annäherung versucht werden. Es soll aber im Einzelfall diskutiert werden, welche Bedeutungspositionen sinnvollerweise anzunehmen sind. Beispiel (1) repräsentiert sicherlich eine Verwendungsweise, die dem Bereich AWARENESS zuzuordnen ist. Es ist schon hier auf die Wahl der lexikalischen Einheit zu achten. Wie erwähnt, geht ein pragmatischer Ansatz, wie der hier vorgeschlagene, davon aus, daß nicht das Einzelwort als lexikalische Einheit angenommen wird, sondern, je nachdem, was als mini male funktionierende Einheit festgestellt werden kann, syntaktisch organi sierte Einheiten, bis hin zu speziellen Kollokationen oder der Einbeziehung besonderer Kontexte. Die lexikalische Einheit hat daher immer den Status einer Phrase, selbst wenn sie durch ein einzelnes Wort ausgedrückt ist, wie im Fall einer VP, die durch ein Verb realisiert ist. In diesem ersten Beispiel scheint die Verwendung vom Bezug zu einer Ortsangabe abzuhängen. Man vergleiche etwa die abweichende Bedeutung in Die Leiter steht = ... wackelt nicht im Gegensatz zu Die Leiter steht hinter dem Haus. Als Verwendungs weise ist daher stehen + Ortsangabe (PP) anzusetzen. Als Bedeutungspara phrase kann 'sich aufrecht an einem bestimmten Ort befinden' angegeben
158
ECKHARD HAUENHERM
werden. Wie das Beispiel zeigt, kann auf den Bestandteil Ortsangabe' nicht verzichtet werden, wenn auch dessen explizite Erwähnung in der Paraphrase dies nahezulegen scheint. Stehen ohne Ortsangabe ist eine andere Verwen dungsweise und hat folglich eine andere Bedeutung. Eine anzunehmende Subkategorie von AWARENESS wäre dann z.B. POSITION, eine weitere könnte PLACE sein. Unter die erste fallen die Verwendungsweisen, die zum Ausdruck der räumlichen Lage eines Gegen standes dienen, unter die zweite jene, die den Ort eines Gegenstandes im Raum angeben. Hierzu gehört die Verwendung von stehen im zweiten Beispiel. Die Differenz beider Verwendungsweisen ist auch aus pragma tischer Sicht schwer zu fassen, da beide als stehen + Ortsangabe (PP) ange geben werden können. Der fehlende Bedeutungsunterschied zwischen der Verwendung von stehen und liegen im zweiten Beispiel scheint aber die Annahme zweier Verwendungsweisen zu rechtfertigen. Die Unterschiede müssen außerhalb der syntaktischen Konstruktion, evtl. im Bezugsnomen (bei Gegenständen, die sich normalerweise in aufrechter Lage finden, gilt Ver wendungsweise [2], sonst Verwendungsweise [1], Häuser liegen und stehen im Wald, stehen in der Stadt, Bäume stehen im Wald und in der Stadt) oder sogar im diskursiven Kontext gesehen werden (etwa: welche Angabe ist im gegenwärtigen Kontext fraglich, Lage oder Ort eines Gegenstandes). (3) und (4) sind besonders in kontrastiver Perspektive interessant. Die Verwendungsweise (1) von stehen, die die aufrechte Position eines Gegen standes kennzeichnet, muß offensichtlich noch unterteilt werden, je nach Subjekt. Während für unbelebte Gegenstände das Italienische stare bereithält, muß bei Lebewesen stare in piedi verwendet werden. Das entspricht der Tatsache, daß der Kontrast zu sitzen im Deutschen für Gegenstände nicht gegeben ist. Wir müssen also eine Verwendungsweise für die Körperposition eines Menschen ansetzen. In der Wortfeldtheorie wäre zu überlegen, ob dieser Kontextunterschied nicht die Annahme eines weiteren Wortfeldes rechtfertigt, das sich von dem Wortfeld 'Positionsangabe' durch die Existenz der Einheit sitzen unterscheidet. (5) und (6) machen den Unterschied zwischen beiden Sprachen noch deutlicher. Während das Deutsche den Standort von Himmelskörpern mit stehen angeben kann, verwendet das Italienische hier esserci bzw. essere. Das zeigt, daß stare keineswegs eine allgemeinere Bedeutung hat als stehen, auch wenn seine Verwendungen als 'verbo fraseologico' (vgl. Bsp. [28] u. [29]) dies zuerst vermuten lassen. Die vorliegende Verwendungsweise ist abhängig von der Art des Subjekts. Sie müßte daher als "Himmelskörper +
SPRACHSYSTEM UND SPRACHGEBRAUCH
159
stehen" angegeben und ebenfalls im Bereich AWARENESS untergebracht werden. Auch (7) repräsentiert eine Verwendungsweise, die dem Bereich AWARENESS zuzuordnen ist. Als lexikalische Einheit muß dabei aber die Kollokation mit "Geschirr" angenommen werden, da nur dadurch der Gebrauch von stehen in Bezug auf die normale Vorkommensweise von Tellern (und auch z.B. Kuchenplatten u.a.) erklärt werden kann. Die Verwendungsweise in (8)-(10), die durch die Kollokationen mit technischen Geräten gekennzeichnet ist, gehört dahingegen, gemeinsam mit den entsprechenden Verwendungsweisen von laufen und funktionieren, ebenso wie das italienische fermarsi und funzionare in den Bereich der RATIONALITY. Es hat mit dem Wissen um die Funktionsweise einer Maschine zu tun, ihr Nichtfunktionieren feststellen zu können. Wie die Ortsangabe verdeutlicht, repräsentiert (11) dagegen dieselbe Verwendungsweise wie (1). (12) hat als lexikalische Einheit offenstehen und gehört in den Bereich AWARENESS. Es hat zusätzlich eine Kollokation "Gebäudeöffnungen" (Tür, Tor, Fenster u.a.) und Gebäude (Garage, Scheune etc.). (13) und (14) wird man als jeweils eigene Verwendungsweise berück sichtigen müssen. (13) hat sicherlich unter Wasser stehen als komplexe Einheit. Für (14) scheint 'stehen + Flüssigkeit' ausreichend. Beide Verwen dungsweisen gehören zu AWARENESS in den Bereich der Wahrnehmung von Zuständen (STATE).
Die Verwendungsweise in (15) stellt eine großen Teil der Vorkommen dieses Verbs. Sie bildet Kollokationen mit unterschiedlichen Präpositionalphrasen wie in der Zeitung, in einem Buch, an der Tafel, auf der Bahnhofsmauer etc. stehen. Da diese Verwendungsweise auf den kognitiven Inhalt eines Textes zielt, ist sie sicherlich dem Bereich der RATIONALITY zuzu ordnen. Auch die Verwendungsweisen in (16) und (17) sind in diesen Bereich einzuordnen. Stehen wird hier in Bezug auf Zustände, die wir mit Hilfe un seres Verstandes erkennen, verwendet. (18) hingegen gehört eher zu SOCIAL BEHAVIOUR, da zu etwas stehen bzw. stehen zu, das hier als lexikalische Einheit zu betrachten ist, in erster Linie eine Festlegung auf ein Verhalten darstellt, nämlich zu etwas in bestimmter Weise "Stellung zu nehmen bzw. beziehen". (19) und (20) scheinen ebenfalls zu AWARENESS ZU gehören, da sie etwas beschreiben, das durch den Gesichtssinn wahrgenommen wird. Sie
160
ECKHARD HAUENHERM
könnten aber mit einigem Recht dem SOCIAL BEHAVIOUR zugeordnet werden, weil sie sich auch auf gesellschaftliche Normen beziehen. (21) stellt ebenso wie (23) eindeutig eine Verwendungsweise aus dem Bereich ACTION dar, bleiben ist ebenso dem Bereich der menschlichen Hand lungen zuzuordnen. (22) ist ein Parallelbeispiel zu (17), gehört also zu RATIONALITY.
Solange man bereit ist, wohnen als Handlung aufzufassen, ist (24) dem Bereich ACTION zuzuordnen, es könnte aber ähnlich wie (19) und (20) SOCIAL BEHAVIOUR zugeordnet werden, da es einen gesellschaftlich normierten Zu stand beschreibt (man denke etwa an Nomadenvölker oder Obdachlose). (25) hingegen beschreibt einen emotionalen Zustand, seine Verwen dungsweise ist daher im Bereich EMOTION anzusiedeln. Von den restlichen Beispielen ist (26) sicherlich dem Bereich AWARENESS zuzuordnen, (27) und (30) SOCIAL BEHAVIOUR und (28) und (29) dem Bereich ACTION. 3.
Fazit Diese kurze Beschreibung kann keinen Anspruch irgendeiner Voll ständigkeit erheben. Sie zeigt aber die Vorgehensweise eines pragmatisch orientierten Ansatzes in der lexikalischen Semantik. Es sollte deutlich ge worden sein, wie durch den Rückgriff auf die Prädikationsbereiche und die Überschreitung der Wortgrenze für lexikalische Einheiten auch mannigfaltig polyseme Einheiten wie stehen und stare einheitlich beschrieben werden können. Das Problem der vielfältigen Verwendung stellt sich in einem prag matischen Ansatz nicht mehr, da hier die komplexe Verwendungsweise als lexikalische Einheit betrachtet wird, die normalerweise nicht mehr polysem ist. Ziel der Analyse ist die Erfassung der kommunikativen Mittel eines bestimmten Kommunikationsbereichs, die den in einer anderen Sprache zur Verfügung stehenden Mitteln gegenübergestellt werden. Im Rahmen der Wortfeldtheorie müssen dahingegen Phänomene der unterschiedlichen Verwendung von Wörtern entweder, streng dem Saussure schen Zeichenmodell folgend, als Polysemie und kontextuelle Disambiguierung von Einzeleinheiten betrachtet werden oder aber, dem Prinzip der Einheit der Äußerungsform folgend, eine Bedeutung als Kernbedeutung vorausgesetzt und die Verwendung der Ausdrücke in anderen Sinnbereichen als abgeleitete oder übertragene Verwendungsweisen beschrieben werden. Letzteres bringt das Problem mit sich, begründen zu müssen, welche Ver wendungsweise als 'Kern' angenommen wird (vgl. z.B. Hundsnurscher 1988). Außerdem würden die abgeleiteten Verwendungsweisen nicht als
SPRACHSYSTEM UND SPRACHGEBRAUCH
161
originärer Bestandteil der jeweiligen Wortfelder angesehen, was dem Sprach gebrauch insofern nicht gerecht wird, als sie innerhalb dieser Felder genauso funktionieren wie deren 'originäre' Bestandteile (vgl. hierzu Geckelers [1970:199] Analyse von frz. enfant im Bereich der Altersstufen und der Verwandtschaftsbezeichnungen). Es gibt aber auch einige mögliche Verknüpfungspunkte zwischen beiden Arbeitsweisen. Die Zuordnung eines Wortes zu einem Wortfeld be schreibt schon eine Verwendungsweise. Sein Funktionieren in mehreren Wortfeldern ist dann als ein Hinweis auf verschiedene Verwendungsweisen zu deuten. Wo die Wortfeldtheorie dabei auf den Kontext zurückgreift und diesen als Bedingung für die Herleitung bestimmter spezifischer Bedeutungen benutzt, muß ein pragmatischer Ansatz, wie der hier vorgestellte, die Wortgrenze überschreiten und die Mehrwort-Einheit in Betracht ziehen und deren Verwendungsbedingungen miterfassen. Literatur Ballmer, Thomas T. 1990. "Neurobiologie und Lexikon". Sprache und Gehirn. Roman Jakobson zu Ehren hg. v. Helmut Schnelle, 2. Aufl. Frankfurt: Suhrkamp. & Waltraud Brennenstuhl 1986. Deutsche Verben. Eine sprachanalytische Untersuchung des deutschen Verbwortschatzes. ( = Ergebnisse und Methoden moderner Sprachwissen schaft, 19.) Tübingen: Narr. Bergenholtz, Henning & Joachim Mugdan 1989. "Korpusproblematik in der Computer linguistik: Konstruktionsprinzipien und Repräsentativität". Computational Linguistics Computerlinguistik. Ein internationales Handbuch zur computergestützten Sprach forschung und ihrer Anwendungen hg. v. István S. Bátori, Winfried Lenders & Wolfgang Puschke, 141-149. Berlin, New York: de Gruyter. Coseriu, Eugenio 1978. Probleme der strukturellen Semantik. Vorlesung gehalten im Wintersemester 1965/66 an der Universität Tübingen. Autorisierte und bearbeitete Nachschrift von Dieter Kastovsky. Tübingen: Narr. Coulmas, Florian 1977. Rezeptives Sprachverhalten. Eine theoretische Studie über Faktoren des sprachlichen VerStehensprozesses. (= Rapiere zur Textlinguistik, 15.) Hamburg: Buske. Cruse, D. Α. 1986. Lexical Semantics. Cambridge: Cambridge University Press. Fritz, Gerd 1995. "Metonymische Muster und Metaphernfamilien. Anmerkungen zur Struktur und Geschichte der Verwendungsweisen von scharf". Hindelang, Rolf & Zillig 1995. 77-107. Geckeier, Horst 1970. Zur Wortfelddiskussion. Untersuchungen zur Gliederung des Wort feldes "alt - jung - neu" im heutigen Französisch. (= Internationale Bibliothek fir Allgemeine Linguistik, 7.) München: Fink. Greimas, Algirdas Julien 1971. Strukturale Semantik. Methodologische Untersuchungen. Autorisierte Übersetzung aus dem Französischen von Jens Ihwe. (= Wissenschaftstheorie, Wissenschaft und Philosophie, 4.) Braunschweig: Vieweg.
162
ECKHARD HAUENHERM
Herbermann, Clemens Peter 1995. "Gebrauchsvielfalt, Mehrdeutigkeit und Bedeutungszu sammenhang bei lexikalischen Einheiten". Hindelang, Rolf & Zillig 1995. 147-175. Hindelang, Götz 1978. Auffordern. Die Untertypen des Aufforderns und ihre sprachlichen Realisierungsformen. Göppingen: Kümmerle. , Eckard Rolf & Werner Zillig, Hgg. 1995. Der Gebrauch der Sprache. Festschrift für Franz Hundsnurscher zum 60. Geburtstag. Münster: LIT. Hoinkes, Ulrich, Hg. 1995. Panorama der lexikalischen Semantik. Thematische Festschrift aus Anlaß des 60. Geburtstags Horst Geckelers. Tübingen: Narr. Hundsnurscher, Franz 1988. "Über den Zusammenhang des Gebrauchs der Wörter. Eine methodologische Untersuchung anhand des deutschen Adjektivs GRÜN". Poetica (Tokyo) 28.75-103. 1995. "Das Gebrauchsprofil der Wörter. Überlegungen zur Methodologie der wort semantischen Beschreibung". Hoinkes 1995. 347-360. 1996. "Wortsemantik aus der Sicht einer Satzsemantik". Weigand & Hundsnurscher 1996. 39-51. & Jochen Splett 1982. Semantik der Adjektive des Deutschen. Analyse der semantischen Relationen. (= Forschungsberichte des Landes Nordrhein-Westfalen, Nr. 3137, Fachgruppe Geisteswissenschaften.) Opladen: Westdeutscher Verlag. Keller, Rudi 1991. "Erklärungsadäquatheit in Sprachtheorie und Sprachgeschichtsschrei bung". Diachrone Semantik und Pragmatik: Untersuchungen zur Erklärung und Beschreibung des Sprachwandels hg. v. Dietrich Busse, 117-138. (= Reihe germanistische Linguistik, 113.) Tübingen: Niemeyer. Lutzeier, Peter Rolf 1985. Linguistische Semantik. (= Sammlung Metzler, 219.) Stuttgart: Metzler. Paul, Hermann 1995. Prinzipien der Sprachgeschichte, 10. Aufl. Tübingen: Niemeyer. Rolf, Eckhard 1992. "Gibt es so etwas wie eine 'Gebrauchstheorie der Bedeutung'?" Münstersches Logbuch zur Linguistik 1.49-64. Schmidt, Wilhelm 1963. Lexikalische und aktuelle Bedeutung. Ein Beitrag zur Theorie der Wortbedeutung. (= Schriften zur Phonetik, Sprachwissenschaft und Kommunikationsforschung, 7.) Berlin: Akademie-Verlag. Weigand, Edda 1993. "Word Meaning and Utterance Meaning." Journal of Pragmatics 20.253-268. 1995. "Grundfragen einer lexikalischen Semantik auf handlungstheoretischer Basis". Hoinkes 1995. 697-714. 1996. "Words and their Role in Language Use". Weigand & Hundsnurscher 1996. 151167. 1998. "Wortfamilien und Semantik". Lingua Germanica. Studien zur deutschen Philologie. Jochen Splett zum 60. Geburtstag hg. von Eva Schmitsdorf, Nina Hartl & Barbara Meurer, 342-350. Münster, New York, etc.: Waxmann. -& Franz Hundsnurscher, Hgg. 1996. Lexical Structures and Language Use. Proceedings of the International Conference on Lexicology and Lexical Semantics, Münster, September 13-15, 1994, vol. I. (= Beiträge zur Dialogforschung, 9.) Tübingen: Niemeyer.
COMPOSITION PRINCIPLES WITHIN THE WORD AND WITHIN WAYS OF USE OF WORDS CLAUDE GRUAZ CNRS et Sorbonne Paris III 1.
Theoretical position The notion of utterance meaning is essential to any linguistic analysis dealing with semantics in so far as the word, just like any linguistic unit, is basically a unit which takes on its true value in the process of communi cation. As Weigand asserts, "the general question of the relationship between word meaning and utterance meaning represents a problem which has to be settled in every linguistic theory" (1993:253). Utterance meaning is closely linked to context. To my mind a distinction has to be made between two types of contexts. The first one is the situational context, depending on the situation of communication, for instance, the word maison, even in the same sentence, will either refer to the building or to its inhabitants. The second one is the linguistic context and deals with what Weigand defines as "the rules and routines of use" (1992:331); as a consequence, "what is primary are the conventional ways of use that determine how we may express cognitive concepts, e.g. with great surprise, not *with deep surprise" (1996:154). Extending Searle's (1969) formula F(p) Weigand considers that "a pragmatic theory of language has to describe how [these] functional structures are assigned to utterance structures" (1992:327):
164
CLAUDE GRUAZ
From that point of view, there is no one-to-one relation between form and meaning on the syntactico-lexical level. My purpose is to attempt to show that this is also true on the morphemo-lexical level.1 Such a similarity between different levels is consistent with the principle of homology which I developed in previous works (Gruaz 1987, 1988, 1990) and which reads as follows: "there exists in the graphematic, morphemic and lexematic domains a homological hierarchy of regular, marked and non-functional units proper to each of these domains" (adapted from Gruaz 1993). That hierarchy determines X-emes (graphemes, mor phemes, lexemes), marked X-emes (marked graphemes, marked morphemes, marked lexemes), X-ons (graphons, morphons, lexons). 2.
The 'meaning position': from word to morpheme. The homological connection between the lexical and the morphemic approaches allows me to re-use one of Weigand's assertions (1992:334), just changing the grammatical categories (here underlined, W. for Weigand): the unit of description is not the morpheme (W.: word) but the 'meaning position'/reading, i.e. a specific way of using the morpheme (W.: word) which is determined by means of different relationships, especially the quasisynonymy relationship. But the notion of use in this sense, related to the different ways morphemes (W.: words) are used in words (W.: phrases), seems at least partially, in the final analysis, to be a notion of the expression side of language or of grammar in the traditional sense since the lexical (W.: syntactic) combination of morphemes (W.: words) ... are classified according to the structural criterion of meaning identity.
An example of my position is the use of different morphemes to translate the content 'le petit de' in French: -eau in renardeau, -on in chaton, -at in cignat. 3.
Aspects of 'content' To make things clearer, I will first propose a distinction among a certain number of terms related to the notion of content (cf. Gruaz 1998): the 'referent' is the content in its cognitive aspect, e.g., renardeau 'un petit animal qui ressemble à un renard', the referent may well be represented by a picture;
1
I wish to thank Susan Baddeley for re-reading this article.
COMPOSITION PRINCIPLES WITHIN THE WORD
165
the 'sense' is the content in its global and specific aspect, expressed by means of a linguistic expression, e.g., the word chameau 'mammifère ruminant à deux bosses'; the 'meaning' is the content from the point of view of linguistic use in its relational pragmatic aspect, expressed by means of a linguistic expression, e.g., the three morphemes -eau 1 (in renardeau), -on 1 (in chaton) and -elet (in coquelet) have the same meaning 'le petit de'; the 'signification' is the content from the point of view of synchronic family relationship, that is in its relational paradigmatic aspect, expressed by means of linguistic expressions, e.g., per- in père, paternin paternel, parr- in parrain, all refer to the notion of 'father'. 4.
Motivation and synchronic word families Motivation is based upon signification. It can be defined as the formal and semantic relation which expresses the connection between the content of root words and that of derived or inflected words. The synchronic word family stands then as a network of motivated terms. The "Synchronic dictionary of derivational families of French words" (DISFA, cf. Gruaz 1998, 1990, 1993, 1996, etc.; Honvault 1993) contains synchronic families defined for the most part as a set of words linked by etymology and meaning. For instance, pouvoir, possible, puissant, impotent are semantically linked with the notion of 'power' and thus belong to the same synchronic family, while ventripotent does not since its meaning is not associated with the notion of 'power'. The criterion of form cannot be re tained to define a synchronic word family since there are no general rules accounting for all the formal variations encountered in synchronic families (cf. Gruaz 1988). Derivatives are based on two types of motivation which are closely linked with the hierarchy existing within synchronic families (this hierarchy rests on the distinction between macrofamilies, families proper, subfamilies and microfamilies presented at length in Gruaz (1998), a distinction which needs not be developed here). On the one hand, close motivation connects morpholexical derivatives within families in the strict sense of the word, that is to say words which are synchronically derived from one root word (at different levels), e.g., chemin (degree 0) → cheminer (degree 1) → cheminement (degree 2). The derived terms which contain marked morphemes are also considered as closely motivated. For example, if écriture belongs to the synchronic family whose head is écrire since its components écri(t)- and -ure are transparent, the case is different with transcrire "reproduire en recopiant sur un support ou avec
166
CLAUDE GRUAZ
des caractères différents" which is the head of a subfamily within this family since trans- does not mean "through" and is therefore marked. On the other hand, loose motivation associates morphosemantic derivatives within macrofamilies. Those morphosemantic derivatives are linked by etymology and meaning but they are lexicalized, that is to say they are not derived from the head of the family by synchronic derivational rules. Such is the case of véhicule and wagon which are heads of families within the macrofamily VOIE.
The duality of the derivative process appears not only through the two types of motivation which determine two types of derivatives but also through two types of derivation. Taking into account notions exposed by Guilbert (1975:177-178) et Martin (1983:63), I will distinguish successive and associative derivations. The former generates chains of derivatives, each having the degree of its source + 1 (cf. successive accumulations of Guil bert, polysemy of acception of Martin), e.g., habiter (degree 0) → habitable (degree 1) → habitabilité (degree 2), or renard (degree 0) → renardeau (degree 1). The latter produces derivatives which have the same degree (cf. fan-shaped derivatives of Guilbert, polysemy of sense of Martin), e.g., renard (degree 0) → renardeau (degree 1), renardière (degree 1), renarder (degree 1). A more detailed presentation of the part played by association and succession in synchronic word families is to be found in Gruaz (1998). 4.1
Aspects offormal motivation From the formal point of view, the motivation principle which governs derivation is linked to the three main categories of the homological grammar since a motivated term may be: 'regular': in renardeau, rossignolet, the root is integrally maintained; 'marked': in lapereau, the root lapin is formally marked (the mor phemic intermediary element -er- takes the place of the final -in, the expected derivative being *lapineau); 'non-existent': in agneau, the root, whether it be mouton or brebis, is absent. 4.2
Aspects of semantic motivation From the semantic point of view, motivation may appear as: 'partial motivation': some components are not transparent (morphons), e.g., in- is opaque in ingénieur,
COMPOSITION PRINCIPLES WITHIN THE WORD
167
'remotivation': in hydrocarboné, hydro- means 'hydrogen' and not 'water' (markedness); 'new motivation': - means 'two' in bikini 'maillot de bain féminin comportant deux pièces' (/ monokini); 'overmotivation': in monoplan, mono- is marked since it means 'only one', so are also plan 'plan de sustentation' and the whole word monoplan which refers to an aeroplane; 'demotivation': the stem -dic- in indication is etymologically but not synchronically linked to di- as in dire. Such semantic specificities may be reinterpreted by means of the homological categories: taking the case of the content 'petit', we notice that it takes on several meanings and is translated by different morphemic elements which can be regular, marked or morphonic. Several suffixal units are used to express the content 'petit', such as EAU (e.g., renardeau), ET (e.g., coffret), ON (e.g., aiglon), (e.g., bardot), AT (e.g., cicognat), ART, ARD (e.g., bouvart, bouvard), IN (e.g., poussin), ou (e.g., minou), etc. The notions covered are size (e.g., jambonneau 'petit jambon'), kinship (e.g., renardeau 'petit du renard'), age (e.g., lapereau 'jeune lapin'), pejorative value (e.g., poètereau 'poète mineur'), familiarity (e.g., minet 'terme d'amitié'). These units are regular morphemes, marked morphemes or morphons. Regular morphemes express: size: EAU in jambonneau, ET in coffret, ON in aiglon, in bardot; kinship: EAU in renardeau, ET in biquet, ON in aiglon, in bicot, AT in cicognat, age: EAU in lapereau, ET in coquelet, ON in chaton, ART, ART in 'mpoussin. bouvart, bouvard 'jeune taureau', IN Marked morphemes are ranked in two categories: 1. The suffix contains either an additional piece of information or another type of information: size: EAU in ruelle 'petite rue étroite', ET in bidet 'petit cheval de selle', in mulot 'petit rat des bois et des champs', etc.; kinship: let us notice that the notion of kinship refers to the male (but presumably in a generic sense) or to the female, e.g., renardeau 'petit du renard', agneau 'petit de la brebis'. age: EAU in perdreau 'jeune perdrix de l'année', ET in daguet 'jeune cerf de moins de deux ans', etc.; pejorative value: EAU in poètereau 'mauvais poète / poète mineur', etc.; familiarity: ET in minet 'terme d'amitié', etc.
168
CLAUDE GRUAZ
2. The definition refers to: a whole species, e.g., levrette 'variété petite du lévrier d'Italie'; a stage of development: e.g., chenille 'larve du papillon'; a superlative: e.g., carpillon 'très petite carpe'; 'petit petit': e.g., agnelet 'petit agneau' (= 'petit petit de la brebis'). Markedness is linked to lexical polysemy when the words belong to the same synchronic family: e.g., chaînon 1 'chacun des anneaux d'une chaîne', chaînon 2 'chacun des éléments (personnes ou choses) d'une suite', chaînon 3 'ensemble montagneux formant une chaîne secondaire'. When the words belong to different synchronic word families, they come under the province of lexical or morphemic homonymy: e.g., 1: lexical homonymy in bardeau 1 'planche mince' I bardeau 2 'petit mulet'; e.g., 2: morphemic homonymy of the stem in mulet 'animal issu de l'accouplement d'un âne et d'une jument' / mulot 'petit rat des bois et des champs'. In the case of morphons, the 'suffixal' element has no meaning, e.g., eau in chameau, ot in escargot. From the examples listed above, it is clear that the correspondence one unit - one meaning is delusive: one unit may correspond to several meanings (e.g., eau: size, kinship, age, pejorative value), and one meaning may corre spond to several units (e.g., size: eau, et, on, ot). There must be added cases of divergences of meaning according to the dictionaries: if they agree on the referent, they often disagree on the meaning: oisillon is 'jeune oiseau' for "Lexis", 'petit oiseau; jeune oiseau (surtout en parlant des espèces de petite taille)' for "Petit Robert"; carpillon is 'très petite carpe' for "Lexis", 'très petite carpe; petit de la carpe' for "Petit Robert". 5.
Morphemic composition and speech acts Weigand refers to the fact that a speech act can be realized by different utterances (1993:262):
It seems to me that one could refer in the same way to different morphemo-lexical utterances which all contain the meaning position 'petit' and represent them as follows:
COMPOSITION PRINCIPLES WITHIN THE WORD
169
It is worth noticing that some suffixes are more often used than others in neological French expressions. In particular, the suffix -ette seems to be frequently used. Let us mention berlinette 'petite voiture familiale', cymbalette, grippette, mini-jupette, sleeperette, fumette ('les premières fumettes commencent dès l'enfance' "Le Monde" 2/3/1993, p. 11). 6.
Contrastive approach The composition principles vary from one language to another. Here are some correspondences between the French derivative construction and some English equivalents: Fr. derivative ↔ Engl. derivative (= with a suffixal morpheme): coffret ↔ casket; livret ↔ booklet; oison ↔ gosling; chaton ↔ kitten; aiglon ↔ eaglet; Fr. derivative ↔ Engl. compound: passerelle ↔ footbridge, gangway; puceron ↔ greenfly; ruelle ↔ alley (-way); renardeau ↔ fox cub; ourson ↔ bear cub; baleineau ↔ whale calf; dindonneau ↔ turkey poult; Fr. derivative ↔ Engl. lexeme: bardeau 1 ↔ shingle; agneau ↔ lamb; chaînon 2 ↔ link; allumette ↔ match; veau↔calf; Fr. derivative ↔ Engl. (adjective) + derived lexeme: barbillon ↔ (small) barbel; Fr. derivative ↔ Engl. (adjective) + noun: souriceau ↔ young mouse; chevrette ↔ young she-goat; fillette ↔ (little) girl; chainette ↔ (small)
170
CLAUDE GRUAZ
chain; raton↔young rat; perdreau ↔ (young) partridge; moinillon ↔ little monk; Fr. derivative ↔ Engl. derivative or (adj.) + noun: oisillon ↔ fledgling/young bird; Fr. derivative ↔ Engl. possessive case: ânon ↔ ass's foal. There must be added some particular cases: one ↔ several / several ↔ one: daguet: young stag, brocket; goret, cochonnet, porcelet: piglet; chevreau, biquet ↔ kid; caneton diff. from canardeau↔duckling; polysemy: chevalet (joinery)↔sawhorse I chevalet (painter)↔easel; distinctive suffix: poulet ↔ chicken I poussin ↔ chick. 7.
Conclusion There is little doubt that a certain homology governs the ways of use of words in phrases and the ways of use of morphemes in words. Yet a distinction must be made between them. If they have in common the fact that both are imposed by use, one speaker has often a greater choice in the former than in the second, which is but another evidence of the greater scale of freedom on the syntactic level compared to the morpholexical level. But it remains that both testify to the inadequacy of one word/morpheme - one meaning theories from a pragmatic point of view. References Gruaz, Claude 1987. Le mot français, cet inconnu. Rouen: Presses Universitaires de Rouen. 1988. La dérivation suffixale en français contemporain. Rouen: Presses Universitaires de Rouen. 1990. Du signe au sens. Rouen: Presses Universitaires de Rouen. 1993. "Le Programme EUROLEXIQUE: dictionnaire morphosémantique du français et d'autres langues; Dictionnaire structurel du lexique français. Principes, méthodologie et contenu". Hilty 1993. 119-131. 1995. "Les variations sémantique et formelle du mot français et d'autres langues romanes: les familles synchroniques de mots". Atti del XXI Congresso Internazionale di Linguistica e Filologia Romanza, Palermo, 18-24 settembre 1995 ed. by Giovanni Rufino, voi. III. Tübingen: Niemeyer. 1996. "Lexicologie contrastive: forme et sens dans la synchronie contemporaine". Weigand & Hundsnurscher 1996. I. 29-38. 1997a. "Nouveaux aspects théoriques du Dictionnaire synchroniques de familles dérivationelles du français". Actes du Colloque "La journée des dictionnaires", Université de Cergy-Pontoise.
COMPOSITION PRINCIPLES WITHIN THE WORD
171
1997b. "Catégories lexémiques et catégories morphémiques dans les familles synchroniques dérivationelles du français". Recherches linguistiques de Vincennes 26.39-59. 1998. "Structure homologique et traitement de la polysémie et de l'homonymie dans le Dictionnaire synchronique des familles dérivationelles de motsfrançaisDISFA". Acts of the XVIth International Congress of Linguists, Paris 1997. Guilbert, L. 1975. La créativité lexicale. ( = Coll. Langue et langage.) Paris: Larousse Université. Hilty, Gerold, ed. 1993. Actes du XXe Congrès International de Linguistique et Philologie Romanes, Zürich 1992. Tome IV. Tübingen & Basel: Francke. Honvault, René 1993. "Le dictionnaire morphosémantique des familles synchroniques de mots français DISFA". Hilty 1993. 147-160. Martin, Robert 1983. Pour une logique du sens. Paris: Presses Universitaires de France. Searle, John R. 1969. Speech Acts. An essay in the philosophy of language. Cambridge: Cambridge University Press. Weigand, Edda 1992. "Semantic Methodologies". Zeitschrift fir Phonetik, Sprachwissen schaft und Kommunikationsforschung 45.321-340. 1993. "Word Meaning and Utterance Meaning". Journal of Pragmatics 20.253-268. 1996. "Words and their Role in Language Use". Weigand & Hundsnurscher 1996. I. 151-167. Weigand, Edda & Franz Hundsnurscher, eds. 1996. Lexical Structures and Language Use. Proceedings of the International Conference on Lexicology and Semantics, Münster 1994, 2 vols. (= Beiträge zur Dialogforschung, 9/10.) Tübingen: Niemeyer.
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY IN ENGLISH AND GERMAN JÜRGEN ESSER Universität Bonn 1.
The lexeme as an abstract unit Textbooks on semantics and lexicology stress that the lexeme is an abstract unit of the language system which constitutes a class of word-forms (cf. Lyons 1977, vol. 1:19, 22 and Lipka 1992:73). Thus find ma found are different word-forms (among others) of the lexeme FIND. The word-forms of a lexeme are usually conceived of as written or orthographic word-forms (cf. Lyons 1977:18).1 While Jackson (1988:8) and Lipka (1992:72, 76) allow also for the possibility of 'phonological words', these two authors are in fact concerned only with written or orthographic word-forms. In the past, Lamb (1964), for example, related lexemes to phonological forms while Halliday, Mcintosh & Strevens (1964:51) only allow for an 'orthographic word' in relation to the 'lexical item'2 and not a phonological word. Many lexicologists agree in distinguishing a third unit, besides the word-form and the lexeme, namely the 'grammatical' word-form (cf., for example, Halliday, Mcintosh & Strevens 1964:51, Matthews 1974:20-26, Jackson 1982:55, and Lipka 1992:72). This notion is needed to differentiate between different grammatical senses of one orthographic word-form, indicated by < > below, or one phonological word-form, indicated by / /, for example:3 1
The qualification 'usually' is needed because we also find more general statements like this: "Words have both a spoken form and a conventionally accepted written form." (Lyons 1995:23). 2 Like Lipka (1992:73) I use the terms 'lexical item' and 'lexeme' in a synonymous way. 3 Tag meanings: VB: base form of lexical verb, NN: singular common noun.
174
JÜRGEN ESSER
(
1
)
b
a
s
e
form of lexical verb singular common noun
Grammatical word-forms separate homographs and homophones on the basis of their grammatical status, which is annotated in corpus linguistics by means of grammatical tags, taken here from the LOB tag set (cf. Johansson, Garside, Hofland & Leech 1986). Grammatical word-forms of the same word-class label can then be grouped together as lexical items, for example:4
(2)
We are faced, then, with three layers of abstraction that focus on three senses of the term 'word' that are normally distinguished: (3)
medium-bound (orthographic or phonological) word-form
e.g., <shows> /∫euz/
grammatical word-form
e.g., shows_VBZ
lexical item / lexeme
e.g., SHOW2
As the examples show, different notational conventions keep the three senses of word apart in metalinguistic discourse.
4
It should be noted that grammatical word-form and lexeme are not logically independent. Thus the distinction between SHOWj and SHOW2 is already implicit in (1) due to the dif ferent word-class labels. NNS: plural common noun, NNS$: plural common noun + genitive, NN$: singular common noun + genitive, VBZ: 3rd person singular of verb, VBD: past tense of lexical verb, VBN: past participle, VBG: present participle, gerund.
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY
175
2.
Realization of lexemes in two media As I have indicated above, studies about lexical items (or lexemes) tend to take the written realization as the normal case, which is quite under standable since linguistics as a discipline is communicated chiefly in the written mode. This tendency to take the written form as the normal case is strengthened by the fact that we often associate lexical items with entries in printed dictionaries (cf. Jackson 1988:8). From a methodological point of view this situation is somewhat arbitrary if one compares morphology with lexicology. Traditionally, the study of morphemes has been more concerned with (phonological) form whereas the study of lexemes has been more con cerned with meaning aspects,5 for example questions of homonymy, poly semy, semantic features and semantic transfer (cf. Lyons 1977:20). One reason for this situation lies in the fact that linguists do not usually make a systematic distinction between the spoken and the written medium and their respective units and structures. Two exceptions in the past are note worthy in this connection. The first is expressed programmatically in the title of an article by Bolinger from 1946: "Visual morphemes". Visual mor phemes give homophonic expressions a linguistic analysis in the graphic medium, for example:
Secondly, there is the model by Halliday, Mcintosh & Strevens (1964:51), which distinguishes between medium-dependent units and struc tures and abstract form:
5 This incongruence is sometimes noticed by students in introductory courses who find it odd that morphemes are presented in phonetic transcription but lexemes in normal orthography.
176
JÜRGEN ESSER
GRAPHOLOGY Paragraph Orthographic sentence Sub-sentence
PHONOLOGY
Tone group
FORM (Grammar)
FORM (Lexis)
Sentence Clause Group
Foot Orthographic word Letter
Syllable Phoneme
Word Morpheme
Lexical item
Table 1 : Medium-dependent and medium-independent units and structures
In this table, the columns GRAPHOLOGY and PHONOLOGY list units and structures that are medium-dependent, whereas the two FORM columns list units and structures that are medium-independent. That is, a sentence is a sentence both in the written medium and in its spoken equivalent, and vice versa. The same holds for the clause, the group (phrase), the (grammatical) word(-form) and for the morpheme.6 The distinction between medium-dependent and medium-independent units and structures can be related to Lyons' (1972:65; 1977:87; 1981:11, 18, 60; 1995:36) concept of 'medium-transferability'. This notion accounts for the fact that (1972:65): people can learn, fairly easily and successfully for the most part, to transfer from one medium to the other, holding invariant much of the verbal part of language.
The property of medium-transferabilty means therefore that large parts of the language system exist independently of a substance or medium and that there must be orthographical and phonological subsystems to accommodate medium-dependent units and structures. It is the achievement of Halliday, Mcintosh & Strevens to have made proposals in this direction although their classification and further implications have been largely ignored in the development of linguistics.
6
There is no one-to-one horizontal correspondence between the entries in Table 1. Some aspects are of course debatable. Not all researchers would keep the foot as a unit of phonology and some would like to include the paratone (intonational paragraph) above the tone group (cf. Esser 1988).
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY
177
3.
Medium-independent word-forms In what follows I want to argue that, for a complete description of lexemes, it is necessary (i) to include systematically both orthographic and phonological word-forms as alternative realizations of a lexeme7 and (ii) to distinguish a fourth level of description, i.e. the level of medium-independent word-form. This is graphically already implied in (1) above, where the node of the two braces points to a kind of word-form that is medium-independent but not yet grammatically specified. A more informative representation would therefore be the following: medium-bound word-forms medium-independent word-form grammatical word-form lexical items The concept of a medium-independent word-form, marked > < , follows from the abstract nature of the lexeme and its property of being susceptible to medium-transferability. Just as the two orthographic word forms
and < found> are abstracted to the lexeme FIND, the ortho graphic word-form < show > and the phonological word-form /∫eu/ can be abstracted to the medium-independent word-form > s h o w < . 8 The symbolization between the inward pointing angles is arbitrary and could follow the orthographic, the phonological or a purely abstract symbolization. (For ease of notation I use the orthographic form.) The identity results from the possibilities of realization which are in an OR-relation. Figure (6) shows a classical instance of true homonymy of two lexemes, i.e. homophony plus homography. 7
This is done to some extent in dictionaries. But the lists given there are, as a rule, far from complete. We usually find only differing stem pronunciations (i.e. not the pronunciations of all word-forms), differences between full and weak forms with function words and differences between British and American English with content words. More explicit than most dictionaries is the COBUILD dictionary (Sinclair 1995), which additionally lists the orthographic inflected forms. 8 That is, in terms of the distinctions made in Table 1 we need an intermediate level of abstraction between the units of abstract form (the two columns on the right) and the units of realization (the two columns on the left).
178
JÜRGEN ESSER
The Situation of homophony which led Bolinger to postulate visual morphemes can now be represented like this:
It is clear that in these cases of homophony the transfer from one medium to the other does not hold (cf. the quotation by Lyons above) because the written forms have a one-to-one relation to the grammatical word-forms and can therefore be regarded as indicators of additional meanings. This applies equally to German where the orthographic capitalization at the beginning of a word is a marker of nouns. This is playfully exploited in the following headline from the Frankfurter Allgemeine Zeitung of 5 August 1997: "Wir haben L/liebe g/Genossen in Berlin". The analysis is as follows:9
(9)
9
JJ: adjective.
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY
179
The medium-independent word-form > Genossen < can function as different kinds of grammatical word in the plural: nominative, genitive, dative and accusative; the form > genossen< can also function in different ways as 1st and 3rd person preterite plural and as past participle. Capitalization also plays a role in English where this orthographic device is used to indicate proper names, cf. Do you like S'/snow? It is important to realize that it is not the spelling that makes < Genossen> or < Liebe > nouns and <dogs> a plural.10 These categories are part of the sentence structure, which is medium-independent. In these cases the implementation of the orthographic conventions merely adds infor mation in the process of encoding, which in turn facilitates decoding. It can therefore be regarded as a metacommunicative activity (cf. Wunderlich 1970:19). The notion of medium-independent word-form is not only useful in the description of homonyms and homophones. It is also indispensible for the comparison of written and spoken corpora (e.g., Biber 1988), and for the analysis of composite corpora which consist of written and spoken compo nents, e.g., the British National Corpus (cf. Burnard 1995). Statistical com parisons of word-forms or lexemes in the written and the spoken mode are only possible if one assumes something like medium-independent wordforms. The orthographic make-up of most spoken corpora often conceals the fact that the representation of a word in ASCII-code is not co-extensive with the lexeme or a medium-independent word-form but is only an orthographic representation (cf. Esser 1994). It is only in special cases that a full 10 Note that in English and German unskilled writers sometimes use < ' s > to mark the plural. Homophony is not only a problem for second-language learners but also for native speakers/writers who learn the written form as a cultural skill after their oral language acquisition is more or less completed. For the learner of a second language the problem is different because for him or her the written form is usually there right from the beginning, i.e., they learn the orthographic and the phonological word-forms simultaneously.
180
JÜRGEN ESSER
differentiation is needed. These include special orthographic conventions for shandi-forms in speech like < 's > , < 've > , < '11 > , < don't> , or < dy ' wanna > . The question is whether these and the normal orthographic word-forms < has > , < have >, < shall >, < do not > and < do you want to > are representations of the same respective medium-independent wordforms and lexemes, as in, for example:11
If one differentiates as suggested in (11), (12) and (13), it becomes possible to answer questions like this: How often do the medium-independent 11 HVZ: 3rd person singular of have, HV: finite base form of have, HVI: infinitive form of have. The distinction between finite base form and infinitive is made in the BNC tagset (cf. Burnard 1995:94). As a comparison of (12) and (13) shows, the medium-bound forms are different.
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY
181
word-forms > has < and > have < occur in a composite corpus consisting of written and spoken texts? That is, besides the lexeme we need the mediumindependent word-form as a unit for stylistic comparison across spoken and written corpora. Further areas which demand recognition of medium-independent wordforms are (i) orthographic variance due to different national standards, (ii) phonetic variance due to national, regional, social or personal norms (diaphones) and (iii) phonetic variance due to connected speech. It is well-known that there are spelling and pronunciation differences of lexemes in American and British English, for example:
All these medium-dependent word-forms must be related to a mediumindependent word-form > colour < which makes it possible to compare word-frequencies in varieties of American and British English both in written and spoken corpora. A more unusual case is the following:
These are possible realizations of the medium-independent word-form > jail < which has a British-only orthographic variant. One area that is not often discussed in connection with the realization of lexemes is the allomorphs of connected speech. It is well-known that many function words have one strong and one or more weak forms, for example: (16)
The strong forms are used in stressed sentence positions (and when the words are spoken in isolation), the weak forms are used in unstressed sentence positions. But not only function words have special, contextualized pronun ciations. Almost all lexemes may undergo phonetic sentence processes of liaison, assimilation and elision, for example:12
12
Usually only the linking-r is given as a possible alternation in dictionaries.
182
JÜRGEN ESSER
(17) more and more in ten minutes empty
/ mo:/ / i n tem / ' emti/
'
liaison assimilation elision
This means, in effect, that the medium-independent word-form >empty < , for example, has the following medium-bound word-forms: (18) < empty > OR /empti/ OR /emti/ The observations made so far lead us to the conclusion that the simple concept of de Saussure's (31965:99) linguistic sign as a combination of a "concept" and an "image acoustique" is just a special case focussing on the medium-bound phonological word-form. The formal side of a linguistic sign should include, at two levels, the medium-independent word-form and two or more medium-bound word-forms of which at least one is orthographic and one phonological.13 4.
Remarks on the comparison of English and German The distinctions made so far also have a bearing on contrastive lexicol ogy. It is well-known that in this area one can distinguish three relations of lexical items: divergence, convergence and false friends (cf. Esser 1980). The relations of divergence and convergence hold if two or more meanings (senses) of one form in LI are represented as two or more distinct forms in L2. For example:
13 . There are some lexemes which in principle can undergo medium-transferability but which have no conventionalized orthographic forms, i.e. filled pauses like [] and [:], which may be rendered as < urn > and < erm > . Many of these (sometimes ad hoc) orthographic forms are not institutionalized in dictionaries.
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY
183
For a German learner of English there is a situation of divergence with regard to the lexical item SEIT in German and SINCE and FOR in English. For an English learner of German there is the relation of convergence (which means facilitation in learning). For an English learner of German there is a situation of divergence with regard to the lexical item COUSIN in English and VETTER and COUSINE in German. Again, for a German learner of English this constitutes a relation of convergence. Additionally, for an English learner of German, the German lexeme COUSINE constitutes a kind of false friend with regard to the English lexeme COUSIN.14 False friends (or faux amis) are defined by Rein (1983:64) as follows: Dies sind Wörter gleicher oder doch nah verwandter Sprachform (meist vom gleichen germanischen oder lateinischen Wortstamm), die aufgrund verschie dener historischer und kultureller Entwicklungen ihrer Sprecher (-gruppen) heute in den einzelnen Kultursprachen z.T. stark abweichende Bedeutungen angenommen haben.
It is interesting to note that the concept of false friends also relies on an abstract word-form that can be seen as an equivalent to the mediumindependent word-form introduced above since no reference is made to spelling or pronunciation. On the other hand, it is similarities in orthographic
14 I use the qualification "kind of" here because there is still a considerable equivalence of meaning, i.e. 'child of the parents' siblings'. How much difference of meaning there must be to establish a true false friend is an open matter. Very often false friends are not as unrelated as the most prominent senses of a lexeme suggest, cf. the comparison of German DUMM vs. English DUMB below.
184
JÜRGEN ESSER
and/or phonological form (inventory of graphemes and/or phonemes) that make the formal basis of false friends. False friends can be regarded as two linguistic signs in two languages which have similar properties of medium-bound word-forms but differing meanings. In this connection it is important to realize that, in the last resort, false friend relations do not exist between lexemes (or lexical items) but between lexical units. According to Cruse (1986:76) and Lipka (1992:131) a lexical unit is defined as a "union of a lexical form and a single sense". Accordingly we have to make the following distinction in the description of the false friends German DUMM and English DUMB: 15 (21)
1 'nicht klug' 2 'in ärgerlicher Weise unangenehm' (cf. ein dummer Zufall) 3 (coll.) 'benommen'
1 'permanently unable to speak' 2 (informal, especially AmE) 'stupid' 3 'unable to speak because you are angry, surprised, shocked, etc.'
The sense definitions show that the German sense 1 and the English sense 2 are equivalent, although not stylistically, and that the other senses differ. 5.
Interlingual medium-shift One of the observations of contrastive analysis is that between language pairs there may be cases of interlingual level-shift. This describes the situation when equivalent expressions belong to different organizational structures. For example, the clause I am sorry to say may be rendered by the word leider in German. Besides cases like these there may also be differences between the media. In a famous study Schubiger (1965) argued that German modal particles, e.g., doch, have intonational equivalents in English, for example:
15 The sense definitions are taken from the Duden Deutsches Universalwörterbuch (1996) and the Longman Dictionary of Contemporary English, Summers (1995).
LEXICAL ITEMS AND MEDIUM-TRANSFERABILITY
185
(22) That's just what I ' said "Das habe ich doch eben gesagt" This means that a medium-independent lexeme in one language is rendered by a medium-dependent intonation form (here a high fall) in another. As I have shown (Esser 1978), this observation does not give the full picture since intonation can also play a role in German. Thus we have to allow for a situation of diversification because the German language possesses one additional device, i.e. the modal particle, which English does not have. 6.
Conclusion I hope to have shown that a systematic inclusion of both orthographic and phonological word-forms in the full description of lexemes and their word-forms together with Lyons' concept of medium-transferability, leads to new insights. We need a fourth systematic meaning level of the term 'word', i.e. the medium-independent word-form, that enables us to describe more precisely such different areas as homonymy, homophony, comparison of spoken and written corpora and corpora of different national standards (e.g., BrE and AmE), false friends and interlingual medium shift. References Biber, Douglas 1988. Variation Across Speech and Writing. Cambridge: Cambridge University Press. Bolinger, Dwight L. 1946. "Visual Morphemes". Language 22.333-340. Burnard, Lou 1995. Users Reference Guide for the British National Corpus. Oxford: Oxford University Computing Services. Cruse, D. A. 1986. Lexical Semantics. Cambridge: Cambridge University Press. Dudenredaktion 1996. Duden Deutsches Universalwörterbuch. Mannheim: Bibliographisches Institut. Esser, Jürgen 1978. "Contrastive Intonation of German and English: Problems and some results". Phonetica 35.41-55. 1980. "Contrastive Analysis at the Crossroads of Linguistics and Foreign Language Teaching". IRAL 18.181-191. 1988. Comparing Reading and Speaking Intonation. Amsterdam: Rodopi. 1994. "Medium-transferability and Corpora: Remarks from the consumer-end of corpus linguistics". Hermes Journal of Linguistics 13.45-53. Halliday, M. A. K., Angus Mcintosh & Peter Strevens 1964. The Linguistic Sciences and Language Teaching. London: Longmans. Jackson, Howard 1982. Analyzing English: An introduction to descriptive linguistics, 2nd ed. Oxford: Pergamon Press. 1988. Words and their Meaning. London & New York: Longman.
186
JÜRGEN ESSER
Johansson, Stig, Roger Garside, Knut Hofland & Geoffrey Leech 1986. The Tagged LOB Corpus: Vertical/horizontal version. Bergen: Norwegian Computing Centre for the Humanities. Lamb, Sydney M. 1964. "The Sememic Approach to Structural Semantics". American Anthropologist 66.57-78. Lipka, Leonhard 1992. An Outline of English Lexicology: Lexical structure, word semantics, and word-formation, 2nd ed. Tübingen: Niemeyer. Lyons, John 1972. "Human Language". Non-Verbal Communication ed. by R. A. Hinde, 4985. Cambridge: Cambridge University Press. 1977. Semantics, 2 vols. Cambridge: Cambridge University Press. 1981. Language and Linguistics: An introduction. Cambridge: Cambridge University Press. 1995. Linguistic Semantics: An introduction. Cambridge: Cambridge University Press. Matthews, P. H. 1974. Morphology: An introduction to the theory of word-structure. Cambridge: Cambridge University Press. Rein, Kurt 1983. Einßhrung in die kontrastive Linguistik. Darmstadt: Wissenschaftliche Buchgesellschaft. Saussure, Ferdinand de 1965. Cours de linguistique générale, 3rd ed. Paris: Payot. Schubiger, Maria 1965. "English Intonation and German Modal Particles: A comparative study". Phonetica 12.65-84. Sinclair, John M., ed. 1995. Collins Cobuild English Dictionary. London: HarperCollins. Summers, Della, ed. 1995. Longman Dictionary of Contemporary English, 3rd ed. Harlow: Longman. Wunderlich, Dieter 1970. "Die Rolle der Pragmatik in der Linguistik". Der Deutsch unterricht 22.5-41.
TYPES OF LEXICAL VARIATION CHRISTOPH SCHWARZE Universität Konstanz 1.
Introduction Lexical variation1, that is, variation of lexical meaning and of syntactic properties, has been an argument for those who do not believe in mental grammar, and a challenge for those who do. Whereas the former bring forward phenomena of variation to justify their skepticism and their belief that, instead of grammar, there are only repertoires, the latter try to show that variation can be accounted for by assuming that, in addition to invariable representations, there are principles or rules of alternation, which derive the observable variants. But the work which is being done in this direction does not only aim at justifying the assumption that the lexicon has linguistic structure. It also promises to give a more complete picture of the language faculty (which it is currently fashionable to call Universal Grammar): indeed, the lexicon turns out to be the locus of elementary idiosyncratic items, plus a set of generative devices which operate on these items. This novel conception will also have positive consequences for a better understanding of language change, a topic I will not go into here. It will also have various applications, namely in lexicography, including the writing of bilingual dictionaries, and in contrastive linguistics, the topic of this volume. In fact, if it is possible to state rules of lexical variation, then it will be possible to reduce the amount of idiosyncratic information in lexical entries. And, regarding contrastive linguistics, it will become possible to formulate 1 This is a revised version of a paper which has been presented as a plenary address on the Klagenfurt conference of the Deutscher Italianistenverband in 1995 and published in German as Schwarze (1997). The revisions are mainly motivated by the difference of the audience: scholars of Italian language and literature in Klagenfurt, linguists in Münster. - I want to thank Peter Roth and Vieri Samek Lodovici for checking my English.
188
CHRISTOPH SCHWARZE
contrasts with respect to virtually universal rules of variation, their presence, their restrictions and their pervasiveness in the languages which are being compared. This is most obvious in phonology, where languages can be compared with respect to alternations of lexical representations, such as word final consonant devoicing or vowel raising. Variation on the level of lexical syntax and lexical semantics certainly is more complex, and there still is much work to be done. In this paper, I want to make some remarks on what kinds of phenomena we have to deal with on the level of lexical semantics. But before we begin, two elementary distinctions must be recalled. The first distinction is between meaning and interpretation. This distinction is meant to express the idea that a word form which is stored in memory is associated with a potential of reference, some kind of represen tation, primitive or derived, mostly linked to a concept, which is used in discourse in such a way that reference is made to something we want to speak about. This potential is called the meaning of the word. Although the meaning of a word may change by frequent use or by deliberate definition, it is relatively stable. The interpretation of a word is not a representation, but a process (or a procedure) which takes place in discourse; it aims at establishing actual reference. In the most simple cases this process consists in looking for something present in the situation which matches with the word's meaning. The result of these processes, i.e. the reference of a text used in actual discourse, may of course be represented in the same format as the meanings, but the representations which are produced by interpretation are essentially unstable, since they arise from language use in changing situations. Variation of interpretations which only depend on situations are not to be accounted for by grammar. The line between what only depends on situations and what does not, may be hard to draw in a general way, but it can certainly be made intuitively clear with respect to singular cases. Consider Pustejovsky's (1995:110-111) examples with the direct object of to want: There are paraphrases in which the direct object of to want is governed by various verbs according to the meaning of the direct object, cf. she wants a beer she wants to drink a beer, *she wants to drink a cigarette - she wants to smoke a cigarette. Now the choice of that verb entirely depends on what you know about situations involving beer or cigarettes, i.e., it is not a matter of lexical semantics. (On the other hand, the fact that to want may take, instead of a direct object, a verbal complement, can be derived from its meaning.)
TYPES OF LEXICAL VARIATION
189
The second distinction is between non lexicalized and lexicalized mean ing. It only concerns derived meanings, since primitive association of meaning to word forms is lexicalized by definition. Non lexicalized derived meanings may be generated by rules of alternation or by lexical morphology. They may also arise from principles of discourse organisation. They are stored in the temporary lexicon, from where, during some time, they may be taken for use, instead of being re-created by rule application. If this happens frequently, they gradually pass into the permanent lexicon. In other words, they are lexicalized. Now, lexicalized word meanings may remain transparent with respect to their original derivation, but they also may undergo further alternations, the results of which may be lexicalized, thus giving rise to polysemy, that is, lexicalized variation of meaning. In the study of lexical variation, lexicalized variation of meaning is much easier accessible, regarding the data, than non-lexical variation, and contrastive lexical analysis usually relies exclusively on data of this kind. But it should be kept in mind that if the aim of contrastive lexical analysis is meant to be more than collecting singular cases of lexical mismatching, it should look out for the rules and principles which create variation. Now, what is the nature of these rules and principles? Which kinds of synchronical variation take place in the lexicon and which are the conditions which restrict them? In recent years, there has been an emphasis on lexical variation linked to the alternation of valency and argument structure of verbs, as, e.g., in Levin (1993). The study of this kind of variation can proceed on relatively safe ground, since only those phenomena are taken into consideration which have a counterpart on the level of syntactic properties. Other kinds of variations have been recognized a long time ago and formulated in terms of traditional categories stemming from rhetoric and from classical philosophy of language. These are categories of semantic ex tension (widening and narrowing of meaning) and of adorned speech (tropes and figures, the most important of which are metaphor and metonymy). Their broad acceptance, however, hinges rather on plausibility than on empirical study and explicit formulation. They need to be studied with the more elaborate methods of lexical semantics which are nowadays available, espe cially the method of lexical decomposition and the representation format of unification grammars. But there certainly are more kinds of lexical variation still to be discovered. In this perspective, I want to discuss two cases of lexical variation which may represent such types. They are taken from Italian, but I
190
CHRISTOPH SCHWARZE
think that they originate from principles of variation which may be observed cross-linguistically. 2.
Numerals in non-numeric interpretation My first case concerns the interpretation of numerals. (I use the term 'numeral' for a class of simple or complex words, such as one, two, three, etc.) Usually numerals refer to numbers: Engl, sixteen or It. sedici refer to the number "16"; the numerals Engl, two hundred forty eight and It. duecentoquarant'otto refer to the number "248". Obviously, the reference of a numeral comes about exclusively by virtue of the meaning. But certain numerals, in addition to their numeric meaning, receive other interpretations. In It., these numerals are due "two", tre "three", quattro "four", cento "(one) hundred", mille "(one) thousand" and centomila "(one) hundred thousand"; cf. e.g.: (1)
La stazione è a due passi. "The station is a few (lit. two) steps from here." Abbiamo fatto due chiacchiere. "We had a little chat." (lit. "We made two chats.") Scrivigli due righe. "Write him a line." (lit. "two lines") Li ha venduti per due soldi. "He sold them for a ridiculous price." (lit. "for two pennies") Me la caverò in due minuti. "I will be done in a minute."
(2)
Facciamo quattro passi. "Let's go for a little walk." (lit. "Let's do four steps.") Abbiamo fatto quattro chiacchiere. "We had a little chat." (lit. "We made four chats.") Per quel lavoro gli danno quattro soldi. "For this work they give him a ridiculous wage." (lit. "four pen nies") Magari faremo quattro salti. "We might even have a little dance." (lit. "do four jumps") C'erano i soliti quattro gatti. "There were only very few people (lit. four cats), as usual."
(3)
Te I'ho detto cento volte. "I told you so a hundred times."
TYPES OF LEXICAL VARIATION
(4)
Avevo mille cose da fare. "I had a thousand things to do." Mille auguri! "Many (lit. [one] thousand) good wishes!" Grazie mille! "Thanks a lot!" (lit. "[One] thousand thanks!)" Inventava mille scuse. "He invented all kinds of (lit. [one] thousand) excuses." Diventò di mille colori. "He blushed." (lit. "He became of [a] thousand colors.") millepiedi, millefoglie " millipede ", " yarrow "
(5)
Te l'ho detto centomila volte. "I told you so a hundred thousand times." Ti darò centomila baci. "I will give you one hundred thousand kisses."
191
In these examples the numeral does not refer to a number, but to an indefinite set. Accordingly, the numeral can often be replaced with words the meaning of which are not numbers, but types of indefinite quantification, such as poco "little, few", molto "much, many" etc.; cf. e.g.: (6)
La stazione è a pochi passi. "The station is a few steps from here." Abbiamo chiacchierato un po '. "We chatted a little bit." Scrivigli una breve lettera. "Write him a little letter." Li ha venduti a un prezzo molto basso. "He sold them at a very low price." Me la caverò in pochissimo tempo. "I will have finished in very little time." Camminiamo un po '. "Let's walk a little bit." Abbiamo chiacchierato un po '. "We chatted a little bit." Per quel lavoro gli danno pochissimo. "For this work, they pay him very little." Magari balleremo un po '. "Maybe we will even have a little dance."
(7)
Te l'ho detto molte volte. "I told you so many times."
192
CHRISTOPH SCHWARZE
Avevo molte cose da fare. "I had many things to do." Tanti auguri! "Many good wishes!" Grazie tante! "Many thanks!" Inventava un sacco di scuse. "He invented lots (lit. a bag full) of excuses." These non-numeric, indefinite readings of numerals are obviously lexicalized, i.e. stored in the memory. They are available only for a small set of numerals; cf. the following examples, which are very odd, and usually don't receive the indefinite interpretation: (8)
??
La stazione è a sei passi. "The station is six steps from here." ?? Abbiamo fatto cinque chiacchiere. "We had five chats." ?? Te l'ho detto due milioni di volte. "I told you so two million times."
We can observe the same reinterpretations cross-linguistically (9), but not necessarily with respect to identical numbers (10). Thus German frequently uses the numerals hundert and tausend in the non-numerical interpretation, but this does not hold for zwei, vier and hunderttausend; cf. : (9)
Te l'ho detto cento volte. Ich habe dir das hundert Mal gesagt.
(10) La stazione è a due passi. Der Bahnhof ist ganz nah. (11) Abbiamo fatto due chiacchiere. Wir haben ein bißchen geplaudert. (12) Scrivigli due righe. Schreib ihm ein paar Zeilen. (13) Li ha venduti per due soldi. Er hat sie für einen Spottpreis verkauft. (14) Me la caverò in due minuti. Ich bin im Nu damit fertig.
TYPES OF LEXICAL VARIATION
193
(15) Facciamo quattro passi. Gehen wir ein Stückchen. (16) Abbiamo fatto quattro chiacchiere. Wir haben uns über dieses und jenes unterhalten. (17) Per quel lavoro gli danno quattro soldi. Für diese Arbeit gibt man ihm nur ein paar Pfennige. (18) Magari faremo quattro salti. Vielleicht tanzen wir auch ein bißchen. (19) L'ho detto centomila volte. Ich habe es tausend Mal gesagt. Now this lexicalized variation is in a close relationship with another variation, which is not lexicalized and does not need to be learned: In certain situations numerals can be used in order to indicate, not a number, but a fuzzy numeric domain. Normally, we have to use expressions like about, approximately, almost, or more than in order to refer to these fuzzy numeric domains. But when, e.g., a friend asks me how much I paid for my new bike, and my reply is "three thousand Marks", this answer will be accepted as correct, even if my new bike cost me a few Marks more or less. This is a variation the conditions of which must be defined on the level of discourse pragmatics. (One of the conditions certainly is that only an approximate information is required.) Now I claimed that this non-lexicalized reinterpretation is closely related to the lexicalized reinterpretation which I described in the beginning. The common feature which links both is fuzziness: in the bicycle example, the numeral three thousand refers to a fuzzy numeric domain, and in the Italian examples with due, quattro, etc. to an indefinite quantity, which is fuzzy by definition. The variation can thus be described as the loss of the feature 'numeric', and its condition as fuzziness. We can thus postulate a link between the ordinary meaning of the numeral and its non-numeric interpretation: the two meanings of quattro passi, "four steps" and "a few steps", are linked by a hypothetical "four steps approximately". There is a second similarity between the two kinds of variation: both only occur with selected numerals, that is, with numerals the meanings of which are so-called small and round numbers. In other words, the numbers must be salient. Now how can numbers be salient? Numbers play different roles in mathematical systems, e.g., in arithmetics, but this is not what
194
CHRISTOPH SCHWARZE
counts here. Salience of numbers derives from two sources: numbers are salient if they are cognitively easily accessed, or if they are landmarks in counting systems (cf. Menninger 1979:33). (The easiness of cognitive access is universal, whereas counting systems are not, a fact which may explain cross-linguistic variation.) Now there seems to be a paradox regarding the indefinite readings of Italian numerals: in the examples given above, It. due and quattro are interpreted as "a small, indefinite quantity". But due has a second, nonnumeric reading, as shown in: (20) Prima di fare una cosa del genere, ci penserei due volte. "Before doing such a thing, I would give it a second thought." (lit. "I would think about it two times.") The expression due volte could be paraphrased by bene "well", or seriamente "seriously"; cf.: (21) Prima di fare seriamente).
una cosa del genere,
ci penserei
{bene,
It is clear that this reading cannot be derived from the fact that "2" is a small number. A similar situation can be found in German; in fact, we can translate (21) by (22), where due "2" translates as drei " 3 " : (22) Das würde ich mir dreimal überlegen. So can due mean "few" as well as "many"? The paradox disappears if we consider one more discourse pragmatic variation in the processing of numerals. When, as in the bicycle case discussed above, a price is given by a round number, the fuzziness is such that a numeric domain is determined by reference to a number at the centre of a local domain on the numerical axis, as illustrated by the following graph:
(23) A different reading results if the number, with respect to the domain, is a threshold value. Thus I can say, to someone who wants to buy my bicycle, that my price is two hundred Marks, meaning that I want at least 200 Marks, since I would not object to a higher price if I could get it. This threshold reading ist illustrated by Fig. (24):
TYPES OF LEXICAL VARIATION
195
(24) The interpretation of due in ci penserei due volte (and of Germ, drei in das würde ich mir dreimal überlegen) is a consequence of the lexicalization of the threshold reading, combined with the loss of the numeric value. The thres hold reading defines a numeric domain which is precisely bounded on the minus side, but open on the plus side. When the numeric value is dropped, and the threshold structure remains, then due does not mean "few", but "more than few", or more precisely, "more than is usually needed". Let me now come back to the three classical distinctions mentioned in the beginning. Which has been their role with respect to the analysis of nonnumeric readings of numerals? Here is the answer: The distinction between meaning and interpretation made it possible to recognize discourse pragmatic variation as a phenomenon of interpretation. As I claimed above, fuzzy, but numerical readings of numerals do not need to be learned. They are a case of a very general principle, which may be called the principle of referential tolerance. On the other hand, the nonnumeric readings of numerals, which vary cross-linguistically, are lexical meanings, just like the ordinary numeric readings. The distinction between proper and derived meanings accounts for the fact that the numeric and non-numeric meanings are related, and that these relationships are unidirectional: we can explain how, e.g., "few" derives from "two" and from "four", but we could not derive "two" and "four" from "few" in a plausible way. I think that the example of numerals very clearly shows the usefulness of the notion of derived meaning, more clearly than in the case of prepositions, which I will discuss later. As to the distinction between simple and complex expressions, it has not played a role up to this point. However, we will now take into consideration some observations, which lead to the conclusion that non-numeric meanings of numerals only occur in complex expressions. The first observation is, that non-numeric readings only occur in very restricted contexts. In their basic meanings, numerals can occur with any count noun, in their derived meanings they only occur with a small set of nouns: passo "step", chiacchiera "chat", minuto "minute", soldo "penny", salto "jump", volta "time", and maybe a few more. If the noun does not figure in this list, as in (25) and (26), then the numeral can only have its basic meaning:
196
CHRISTOPH SCHWARZE
(25) L'aula ha due ingressi. "The lecture-hall has two entries." (26) Nel cortile erano parcheggiate quattro automobili. "In the yard four cars were parked." The second observation is, that the numeral not only has to relate semantically to a noun belonging to that set; it must also be in the same noun-phrase as the noun. If the numeral is outside that noun phrase, then the derived meaning is excluded; cf.: (27) Di passi ne ha fatti due. "He made two steps." (lit. "Steps, he made two of them.") (28) Quanti passi ha fatto ? - Due. "How many steps did he make?" - "Two." It can thus be said that the derived meanings of numerals can only occur in restricted, phraseological contexts. This means that the derived meanings are properties of complex expressions, which are stored and can only be processed as such. This fact is a further confirmation of the claim that the non-numeric meaning is a derived variant. We can now tentatively formulate some principles of semantic alternation: I II III IV
3.
Loss of precision: 'n' is replaced with 'the range of numbers within the local domain of n'. Conditions: a. n is a salient number; b. precision is not required by the situation. Denumerification: 'n' is replaced with 'relative quantity'. If η is a small number (n < 5?) the derived meaning is "few", if η is not small (n > 5?), the derived meaning is "many". Condition: η is a salient number. Thresholding: 'n' is replaced with 'n or n + i'. Condition: a. n is a salient number; b. η is a small number. Thresholding and denumerification: 'n' is replaced with 'more than what is usually needed'. Conditions: a. n is a salient number; b. η is a small number.
The polysemy of spatial prepositions: the case of It sopra/su and sotto Our second case of lexical variation are those prepositions of Italian, which we use to refer to vertical relations in space: sopra and su "on, above"
TYPES OF LEXICAL VARIATION
197
(which we will treat here as synonyms, in spite of some slight differences in usage) and sotto "under, below". The semantics of local prepositions has thoroughly been discussed since the beginning of the eighties; the more general questions were the following: I II III
Which is the general semantic form of spatial prepositions? What kind of concepts are lexicalized in spatial prepositions? How can the polysemy of spatial prepositions be accounted for?
The last of these questions is most central to the topic of this paper. In fact, the interpretation of spatial prepositions varies quite impressively in many languages, including Italian. And it is not easy to pass from the description of the data to a plausible and coherent analysis. The discussion has ebbed down, but no general agreement has been achieved regarding any of these questions. There are mainly two positions which have been held in the literature; they may be characterized as follows: One position (A) puts the emphasis on invariable structures (e.g., Wunderlich 1982, Bierwisch 1988); the other one (B) tends to ignore such structures or explicitly denies them (e.g., Brugman 1983, Herskovits 1986, Vandeloise 1986). These positions are closely connected with the answers given to questions I and II. With respect to the question of the general semantic form of spatial prepositions (question I above) position A may be summarized as follows: A PP with a spatial preposition refers to a segment of space, inside which an object or a state of affairs is located. This segment of space is identified by some material object, called the localizing object, with respect to which it is a region. Regions are mentally constructed with respect to parts of the localizing object. The preposition selects one of the possible regions of the localizing object: in selects the inner region of the object, on the region which is delimited by the upper side of the object, beside a region which is delimited by one of the lateral regions, etc. Position B, if it considers invariant structures at all, assumes simpler relationships: spatial prepositions refer to immediate relations between two objects. As to the question of the nature of the underlying concepts (question II), supporters of position A believe that only strictly spatial concepts (dimensions, regions, paths) are relevant for the lexical system. Supporters of position claim that, besides spatial concepts, several other kinds of concepts are relevant, in particular interactive concepts, such as exposure or support, and they tend to put them all on the same level.
198
CHRISTOPH SCHWARZE
The contrast between the two positions is most apparent with respect to the question of polysemy (question III). The defenders of lexical structure tend to consider prepositions as monosemous as far as the semantic system is concerned. They do not deny that a given preposition often has many interpretations, but they claim that these interpretations result from processes which take place on a presumably independent level of conceptual processing. The opposite position, in its most radical version (Herskovits 1986), represents all readings of a given preposition as a mere list, or if it admits that polysemy has some more elaborate structure, it is not a derivational structure, but rather a family resemblance (Vandeloise 1986). Now I want to plead for a position which accounts for variation without abandoning the idea that there is some invariant structure. I claim that, in spite of certain overlappings, spatial prepositions are organized in semantic fields ('Wortfelder') and that this organization completely hinges on spatial concepts in the narrow sense, and will claim and try to show that non-spatial readings are derived from basic spatial meanings. 3.1
The descriptive facts I will now give an outline of the descriptive facts, using the data and also some of the analyses in Palazzi/Folena (1995) and Serianni (1988). 3.1.1 sopra/su. For sopra and/or su in expressions of the general form 'x V su/sopra y', the following readings can be distinguished: [sopra/su 1] χ is included in the ON-region of y. The ON-region is defined by the positive vertical projection of y. The projection may be strictly vertical (29) or enlarging (30): (29) I fiori sono sopra il tavolo. "The flowers are on the table." (30) Le nuvole sopra di noi diventavano sempre più scure. "The clouds above us became darker and darker." [sopra/su 2] χ is not contained in the ON-region, but in the PROX(IMITY)region of y, and it is in a higher position than y. (The PROX-region of y is any region outside y.) An example is: (31) La rocca è sopra il paese. "The castle is above the village."
TYPES OF LEXICAL VARIATION
199
[sopra/su 3]x is a layer which covers y; the relation is not necessarily a vertical one: (32) Sul maglione vestiva una giacca. "Over the sweater he wore a jacket." [sopra 4] χ is north of y: (33) Prato è sopra Firenze. "Prato is north of Florence." [su 5] y is the goal of an aggressive motion performed by x: (34) Napoleone marciava su Mosca. "Napoleon marched towards Moscow." Ilfinanzieresparò sui contrabbandieri. "The customs officer shot at the smugglers." [su 6] y is the material support of a semiotic object: (35) un olio su tela "an oil-painting on canvas" leggere qualcosa sul giornale "to read something in the newspaper" [su 7] χ has an impact on y: (36) tamburellare sul tavolo "to drum on the table" Γalluvione sulla Toscana (Serianni) "the flood on Tuscany" utilizzare il laser sui pazienti (Serianni) "to use the laser on the patients" [su 8] y is the content of a semiotic object: (37) un saggio su Dante "an essay on Dante" una polemica sul servizio di leva (Serianni) "a controversy on the draft" [su 9] y is a motive or a cause for some action of x:
200
CHRISTOPH SCHWARZE
(38) fermata su richiesta "stop on demand" un vestito fatto su misura "a dress made on measure" Furthermore, there are two readings, in which su is only applied to numbers, that is: [su 10] n1 and n2 are numbers; 'n1 su n2' refers to a proportion: (39) Le possibilità sono di una su un milione. "Chances are one to a million." 24 ore su 24 "around the clock" [su 11] su, unlike what is the case in all other readings, does not express a two-lace relation, but characterizes η as an approximation: (40) Ha sui trent'anni. "He is about 30 years old." un tragitto sui cento chilometri "a route of about 100 km" 3.1.2 sotto. For sotto the list of readings is the following: [sotto 1] χ is inside the UNDER-region of y. (The UNDER-region is defined by the positive vertical projection of y.) The projection may be strictly vertical (41) or enlarging (42): (41) Piove, vieni sotto la tettoia. "It is raining, come under the roof." (42) Metti una bacinella sotto il rubinetto. "Put a bowl under the tap. " [sotto 2] χ is in the PROX-region of and in a lower position than : (43) Il paese è sotto la rocca. "The village is below the castle." [sotto 3] is a layer covering x; verticality is not implied: (44) Sotto la giacca vestiva un maglione. "Under the jacket he wore a sweater."
TYPES OF LEXICAL VARIATION
201
[sotto 4] χ is south of y: (45) Napoli è sotto Roma. "Naples is below Rome." [sotto 5] χ and y are values on a scale, and χ smaller than y: (46) La temperatura è sotto lo zero. "The temperature is below zero." una donna sotto la trentina "a woman under thirty" [sotto 6] y offers a shelter for x: (47) dormire sotto la tenda "to sleep under a tent" presentarsi sotto falso nome "to appear under a false name" ridere sotto i baffi "to laugh in one's sleeve" (lit. "under one's moustache") sotto l'alto patronato del presidente della Repubblica "under the high patronage of the president of the Republic" [sotto 7] χ is exposed to a negative effect of y: (48) sotto la grandine "under the hail" sotto i bombardamenti "under the bombings" sotto un governo tirannico "under a tyrannical government" [sotto 8] y is a ruler or a regime, and x is located in time x with respect to the rule of y: (49) sotto Augusto "under Augustus" sotto la Russia zarista "under czarist Russia" 3.2
Semantic structure and variation of sopra/su and sotto Let me summarize the descriptive outline:
202
I II III
CHRISTOPH SCHWARZE
In some readings, especially the spatial ones, sopra/su and sotto are converses of each other, in others this is not the case. sopra/su has more readings than sotto. The readings of sotto are not free from contradictions: 'x sotto y' may imply the idea of (positive) protection as well as of (negative) exposure.
Cross-linguistically we can observe that the polysemy of sopra/su and sotto largely corresponds to what we find in other languages, e.g., in German. (But it should be mentioned that It. sopra/su and sotto, unlike Engl. on vs. above and under vs. below or Germ, auf vs. über does not have a clear opposition with respect to the feature 'contact'. Therefore the readings of the It. prepositions treated here as it were cumulate readings which in other languages may be distributed among larger sets of items.) But the two quantitative readings [10] and [11] may be less familiar to speakers of other languages; cf.: (50) Le possibilità sono di una su un milione. Germ. Die Chancen sind eins zu eine Million. lit. "... one to one million." (51 ) Ha sui trent 'anni. Er ist um die Dreißig. lit. "... round the thirty." The similarities we can observe across various languages justify the assumption that the semantic variation of sopra/su and sotto is very strongly motivated by concepts, regardless of its storage in the lexical memory. But the question is, of course: does this variation have some structure or is there just a random variation, constrained only by similarity and contiguity? Are the spatial readings derived or are they primary? The position I am going to plead for is the following: Spatial readings constitute the basic semantic structures, the other readings being derived variants, which rest upon the following principles: referential tolerance, transfer from spatial orientation to frames of analogous, non-spatial orientation, devisualization of spatial concepts, and amalgamation of spatial and interactive concepts. 3..2.1 Referential tolerance, transfer, devisualization. Referential tolerance, transfer and devisualization are well-known principles of semantic variation. I will just briefly show how they function with regard to our descriptive facts.
TYPES OF LEXICAL VARIATION
203
We have already seen a case of referential tolerance when examining the variation of numerals. With respect to sopra/su and sotto it explains the following facts: variation between strictly and enlarging vertical projection (reading [1]); variation between the ON-region and the PROX-region, as well as the parallel variation between the UNDER-region and the PROX-region (reading [1] and reading [2]); variation from 'ruler' {sotto Augusto) to historical period {sotto la Russia zarista). Referential tolerance rests upon the fact that, in discourse, context and situation help us to understand utterances which would be rejected as obscure when isolated from context and situation. Transfer takes place when spatial concepts which are given by our body and by perception are analogically extended to other systems of orientation (the cardinal points, scales of temperature, etc.) Devisualization happens when the speech community so to speak forgets the spatial relation which was constitutive in the origin. There are two nice examples which have been pointed out by Serianni: the idiom giurare su qualcosa "to swear on something" goes back to a practice where an oath was confirmed by putting one's hand on the bible, and the proportional reading of su originates from the writing of fractions, where a number was written above another number. 3.2.2Amalgamation of spatial and interactive concepts. The most efficient principle of lexical variation in spatial prepositions is doubtlessly given by the fact that our perception of spatial relations combines with our experience about what practical consequences these relationships may have for us. In other words, we amalgamate spatial and interactive concepts. Thus, in an example like (52) the spatial concept of localization in the ON-region is explicitly amalgamated with an interactive concept: (52) riposare su, appoggiare su "to rest on", "to base on" Semantic variation, then, originates from the fact that the concept of verticality may be taken out of focus, like in (53), or be completely removed, as in (54): (53) tamburellare sul tavolo, sulla vetrina "to drum on the table, the window pane"
204
CHRISTOPH SCHWARZE
(54) fondare un giudizio su qualcosa "to found a judgment on" The concept of localization within the ON-region can also be amalgamated with a second interactive concept, namely the concept of putting a layer on something (55). Verticality, then, can again be taken out of focus (56), and non-spatial readings can originate, such as the accumulative reading of su, cf. e.g.: (55) la polvere sui mobili "the dust on the furniture" (56) l'intonaco sui muri "the paint on the walls" (57) fare errore su errore "to make mistake upon mistake" We can now understand why there is an asymmetry between the non spatial readings of sopra/su and sotto: the spatial relationship 'x on/above y' does not have the same consequences for χ as the converse relation 'x under/below y'. Amalgamation also explains why a given preposition may have contradictory readings. In fact, the consequences which 'x under/below y' has for χ also depend on the nature of y and vary accordingly. If y is a heavy object, then it causes pressure, which may be painful for x; if y is an enemy in a fight, then χ is in a position of disadvantage which may be dangerous; if y is a meteorological event or some other force of nature, then χ is under the impact of y ; if y is a conscious subject, then χ can be better perceived and controlled by y. In all these cases the interactive concept of exposure (reading [7]) is involved. If, on the other hand, y is a tent, a roof, a protec tor, etc., then the interactive concept of protection (reading [6]) is activated. Now what does this mean with respect to our more general question of which kinds of concepts are basic in the meaning of spatial prepositions? The existence of contradictory readings is a strong argument in favour of strictly spatial concepts being more basic. In fact, we can plausibly derive interactive readings from spatial concepts. But the opposite derivation is less obvious. It is true that one could also start from interactive concepts, such as protection, and postulate that these are amalgamated with spatial concepts. But then, "behind" and "in" would be equally good candidates, and even "on" could be connected with the idea of protection, if you think of the safety which
TYPES OF LEXICAL VARIATION
205
climbing on a tree may provide when you are threatened by a bear or a flood. But the descriptive facts do not seem to justify this analysis. And there is a second argument: If one looks at spatial prepositions as a semantic field, that is as a group of words which form a semantic paradigm, then such a field only becomes visible when we take into account the spatial readings. The situation is similar to the one we encounter with the names of body-parts: words like hand, head, heart have a large range of meanings, but the mutual semantic relations between these words can best be stated if we take the body-part readings as basic. 3.3
Conceptual development and the inventory of word forms Our analysis, until now, is still incomplete in an important point: We have said nothing about how the approximation reading of su (as in sulle mille pagine) is related to the other readings of that preposition. We have already seen that the approximation reading is quite peculiar: unlike all other readings of su it shows a one-place predicate, and, in a cross-linguistic perspective, it is less common than the other readings. Now in accordance with these peculiarities, it is difficult to derive the approximation reading from the basic meaning, according to which 'x su y' means 'χ is within the ON-region of y'. One might think, it is true, that the approximation reading is derived from another derived meaning, that is, from 'x moves towards y in a hostile intention' {marciare su una città, sparare su qualcuno): the motion here does have a goal, but the achievement is not implied - Napoleon marched towards Moscow but didn't conquer it - an approximation may be reconstructed as a motion of this kind. But the question remains why Italian did not use a more straightforward derivation, e.g., from "around" (like Germ, um, It. circa) or from "in" (like French dans les mille pages). In order to find an answer to this question, we must remember that language is, above all, a means of communication. Language uses conceptual structure, and therefore must elaborate it to some extent. But it would be wrong to think that concepts determine lexical structure, and consequently, we should not try to derive every single reading from conceptual relationships. Since communicative efficiency is the superordinate principle, conceptual cohesion has to compete with other factors of efficiency. In order to see what these factors may be like, let me formulate a tentative explanation of how the approximation reading of su may have originated. Let us suppose that there was a need, but no linguistic symbol, for an operator expressing approximation. There already was a successful model for creating similar
206
CHRISTOPH SCHWARZE
means of expression, namely the reinterpretation of spatial prepositions. Now, several prepositions were available for that purpose: "around", "near", "toward", and maybe others more. But none of these had such an obvious affinity with the concept of approximation that it could rule out all others. At that point, other criteria than conceptual coherence may become relevant; among these a criterion which we may call 'handiness'. In the present case handiness equals shortness (su is monosyllabic, like all typical prepositions, intorno a has four syllables), morphological simplicity (su is simple, intorno a is complex) and familiarity (su is native, circa is a latinism). There is a second (and last) remark I would like to make regarding the limitations of conceptual affinity as a factor of lexical variation. One might wonder, to come back to our data, why the preposition sotto "under" is not used with a verb like soffrire "to suffer". To a speaker of German such a solution would seem most natural, not only because we use that preposition (58), and even if you don't think of German, you might easily associate that meaning to our reading of exposure (reading [7]) above. And why does one use, in Italian, the same preposition di for a topic of a talk or a conversation, while German uses the ON-preposition; cf. : (58) soffrire di emicrania unter Migräne leiden "to suffer from migraine" (59) parlare di politica über Politik sprechen "to talk about politics" The answer will again be found if we do not exclusively consider conceptual affinities, but also formal aspects of the language: Italian has a rather neutral preposition di, which leaves the kind of relation to be concluded from the context. The second arguments of verbs often have a very specific defaultrole; with soffrire it is what you suffer from, and with parlare it is what you talk about. Therefore, no more specific preposition is needed to express the relationship. 4.
Conclusion I hope to have shown a few examples of how the lexicon meets the challenge to be, on the one hand, constant and limited, on the other hand, variable and open, and, more crucially, learnable. The key to understanding this achievement seems to be that there is an interplay of structure and
TYPES OF LEXICAL VARIATION
207
variation, and that variation is not unconstrained. To find out more about these constraints is a task for future research. References Bierwisch, Manfred 1988. "On the Grammar of Local Prepositions". Syntax, Semantik und Lexikon ed. by Manfred Bierwisch et al., 1-65. Berlin: Akademie-Verlag. Brugmann, Claudia 1983. The Story of over. The University of California, Berkeley. M.A. Thesis. Reproduced by the Linguistic Agency University of Trier. Herskovits, Annette 1986. Language and Spatial Cognition. An inter-disciplinary study of the prepositions in English. Cambridge: Cambridge University Press. Levin, Beth 1993. English Verb Classes and Alternations. A preliminary investigation. Chicago & London: The University of Chicago Press. Menninger, Karl 31979. Zahlwort und Ziffer. Göttingen: Vandenhoek und Ruprecht. Palazzi, Fernando & Gianfranco Folena 1995. Dizionario della lingua italiana, con la collaborazione di Carla Marella, Diego Marconi e Michele Cortelazzo. Torino: Loescher. Pustejovsky, James 1995. The Generative Lexicon. Cambridge & London: The MIT Press. Schwarze, Christoph 1997. "Struktur und Variation im Lexikon". Italienische Studien 18.4765. Serianni, Luca 1988. Grammatica italiana. Italiano comune e lingua letteraria. Torino: UTET (Unione tipografica editoriale torinese). Vandeloise, Claude 1986. L'espace en français. Sémantique des prépositions spatiales. Paris: Éditions du Seuil. Wunderlich, Dieter 1982. "Sprache und Raum". Studium Linguistik 12.1-19 and 13.37-59.
NON-OVERT CATEGORIES IN RUSSIAN PARTITIVE AND PSEUDO-PARTITIVE CONSTRUCTIONS ANITA STEUBE & ANDREAS SPÄTH Universität Leipzig 1.
Introduction The theoretical basis for this article is Chomskian grammar enriched with a semantic component (i.e. a 'Semantic Form' [SF]) in the sense of Bierwisch (1987). The lexicon contains lexical as well as grammatical entries. The latter appear in the form of feature combinations. This paper is written in defence of lexical entries (LE) without a 'Phonological Form' (PF). We will come to the conclusion that this kind of defective lexical entries can be justified in a local grammatical context only. We will deal with Russian noun phrases, to be more precise, with the sortai differences of nouns in the context of quantifiers like malo, malye "little"; nemnozko, nemnogie "some"; mnogo, mnogie "much, many", whereby the first form in each of the three Russian pairs is a so-called short, i.e. morphologically uninflected form, the second form is in the nominative case plural. When partitive constructions contain measure phrases (MP), the quantifiers modified by MPs have no overt realization. The overtly or nonovertly quantified head nouns of partitive constructions have no PF realization either. Last but not least, Russian noun phrases have no phonologically realized articles. But there is reason enough to construe a Russian noun phrase as a determiner phrase (DP) with the functional category D° filled by the semantic content of a definite or of an indefinite article like in German (cf. Birkenmaier 1979, Vater 1996). The non-overt head noun and the non-overt quantifier on the other hand do not mirror typological differences between languages but are based on theoretical assumptions which hold across languages.
210
ANITA STEUBE & ANDREAS SPÄTH
2.
Pseudo-partitive constructions Russian pseudo-partitive constructions are restricted to DPs in non structural case positions and to quantifier phrases (QPs) consisting of quantifiers in the plural, not modified by measure phrases, like (1)
On pomogal nemnogim dorogim druz 'jam. "He helped some dear friends."
with the object-DP in the dative plural (cf. Franks 1995).
[nemnogim dorogim druz'jam]dat.pl. to some dear friends Fig. 1: Logicai form (LF) structure of Russian pseudo-partitive DPs
In German all the pseudo-partitive constructions in figure 2 are paralleled by partitive constructions, independent of the syntactic function and independent of the morphological features these DPs have. Russian, on the contrary, is restricted to the partitive equivalents of figure 2 (cf. Isacenko 1962).
NON-OVERT CATEGORIES IN RUSSIAN
Pseudo-partitives viel Tee "much tea" viel starker Tee "much strong tea"
211
Partitives viel von dem Tee "much of the tea" viel von dem starken Tee "much of the strong tea" drei Tassen von dem starken Tee "three cups of the strong tea" diese drei Tassen von dem starken Tee "these three cups of the strong tea"
drei Tassen starker Tee "three cups of strong tea" diese drei Tassen starker Tee "these three cups of strong tea"
Fig. 2: German pseudo-partitives and partitives
der
viele much
starke Tee strong tea
viele many
rote read
0
Äpfel pl. apples
die drei Tassen VIEL starker Tee the three pots MUCH of strong tea
0
Fig. 3; LF structure of German pseudo-partitive constructions
German quantifiers like wenig/wenige "little, few", etwas/einige "some", viel/viele "much, many" are modifiers of noun phrases. They select mass nouns when used in their singular form, and bare plurals when used in their plural form. As already mentioned, Russian only allows quantifiers in the plural in pseudo-partitive constructions. Mass nouns denote substances, bare plurals denote a disindividuated unspecified mass. The quantifiers in the respective languages turn both these denotata (cf. wenig Geld "little money",
212
ANITA STEUBE & ANDREAS SPÄTH
wenige Äpfel "few apples", nemnogie druz'ja "some friends") into individuals having a locally limited Gestalt (cf. Krifka 1989, Link 1991). In addition, German quantifiers may project to quantifier phrases (QP) containing a modifying measure phrase (cf. drei Tassen VIEL "three pots MUCH"). If quantification is carried out by means of a measure phrase, the head of the QP (VIEL) must be phonologically empty. Bierwisch (1987) used this descriptive means to show the parallel between QPs and dimensional phrases: cf. three cups MUCH - four feet wide, three miles long, etc. The non-overt quantifier VIEL/MUCH expresses that what is measured is the quantity only. When a quantifier like German viel, viele or Russian mnogie (pi.) can change the sortai characterization of a noun we may conclude that the lexical entries of the nouns should be underspecified when they leave the lexicon. Grammatically induced sortai specification in German is carried out by quantifiers, by number and by the articles. Sortai specification in Russian depends on grammatical contexts such as number, case, quantifiers, verbal aspect, the position of the DP inside or outside the focus projection, etc. This paper will deal with the contextual factors number, case and quantification in Russian. Let us begin with the exemplification of the sortai specification in Russian pseudo-partitive constructions. The Semantic Form is mapped onto the Logical Form of the sentence in which the number features [although these bound morphemes are generated in the lexicon, too, and are fixed to the stems of the lexical X°-elements occurring in the agreement chain] appear as sister categories not of № but of N'. This is so because modification of a noun by an adjective results in the denotation of a subset of the denotatum of the noun, and it is the quantity of this subset that is characterized by the meaning of number. The precondition for the selection of number features is the sortai specification of the noun as an individual-noun, because only individuals can be counted. Mass-nouns cannot be specified by number. Figure 3 makes use of the plural morpheme and a zero-ending 0 . Wenn 0 is selected, the sortally underspecified nominal lexical entry is not specified for number but combined with the substance-template instead and thus undergoes the specification of a mass-noun. The characteristic of a mass is that all its subparts have the same properties as the contextually specified maximal mass. The German quantifiers wenig, etwas, viel specify the quantity of the contextually specified or maximal mass. The plural expressions wenige, einige, viele or their Russian equivalents malye, nemnogie, mnogie, however,
NON-OVERT CATEGORIES IN RUSSIAN
213
do not characterize the total sum formed by means of the plural. They refer to the individuals in the set and specify their number as small, great, etc. We will now list the SF representations of the lexical entries needed in the Russian pseudo-partitive construction mnogim doro girti druz'jam "many dear friends" and in the German pseudo-partitive construction viel Tee "much tea".
Fig. 4: Semantic Form representations
The SF representation of the German overt quantifier viel, viele and the Russian quantifiers mnogo, mnogie or their non-overt variant MUCH has to be explained further. REL stands for the relation ⊆ holding between the homogeneous subparts of a mass and the corresponding homogeneous contextually specified or maximal mass as well as for the relation e holding between the individuals of a sum (cf. Krifka 1989). The second conjunct of the SF representation excludes individual-nouns in the singular. The features marked (...)α are optional but they must either all be selected or all be left off. When the (...)a-features are selected, they characterize noun phrases in the plural. When they are not selected, mass-nouns are characterized. This way the SF representation of the quantifiers remains underspecified in the lexicon (cf. Dölling 1992, 1995) and can be amalgamated with the SF repres entation of mass-nouns as well as of bare plurals. That the SF representation of quantifiers must be underspecified even further in order to allow for modification by measure phrases is demonstrated in Bierwisch (1987:193196). Now let us construct the SF-representation of the DP mnogim dorogim druz 'jam on the basis of the LF representation in figure 1. Figur 1 ' shows the SF representations of the relevant lexical entries and the representations of
214
ANITA STEUBE & ANDREAS SPÄTH
the intermediate steps in the process of amalgamation of the typed lambdacategorial structures.
Fig. 1: Compositional construction of the SF of a Russian pseudo-partitive DP
The single steps in the process of amalgamation will be explained in detail. First, the SF representations of dorogoj and drug are amalgamated:
This semantic predicate replaces Ρ in the SF representation of the plural:
In the process of amalgamation between this predicate and the SF of mnogim, the features marked (...)α must be selected since dorogie druz'ja is a DP in the plural:
The characterization can be left out because it is redundant. Finally, the DP receives the default interpretation of its referential state, whereby D°is semantically realized as an existential quantifier. As soon as the complete sentence is considered, the grammatical function of the
NON-OVERT CATEGORIES IN RUSSIAN
215
pseudo-partitive construction and its position in the information structure of the sentence help to predict whether it is the SF of the definite or of the indefinite article that should be contextually specified. Topics tend to be definite, while constituents realizing the focus accent tend to be indefinite. This paper does not consider these contexts, however (but cf. Steube & Spath forthcoming). Therefore the SF representation of the indefinite article is used as the default: Hi ::Ic ([[[[FRIEND, xJ:[DEAR, xJ]:[QUANT Xi ~ 2]]:::Ix [(xXi EO x]]: ::Ic [[QUANT x i ~ [N +c]]]) read: there is an Xi (at least two dear friends Xi' and all Xi are individuals of a sum x) and there is a surplus-value c, such that the quantity of Xi excedes the quantity of the norm N by c. The German pseudo-partitive DP viel Tee has no Russian equivalent. The noun Tee leaves the lexicon with the sortally underspecified SF representation AX [TEA, x]. When left unspecified by number features, the mass~template builds up the sortal specification of the mass-noun Tee:
Ax [TEA, (xXi [Xi
x]]
k
Mass-nouns can be modified by viet.
Ax ([TEA, (xXi [Xi
k
x]]:::Ic [[QUANT X ~ [N +c]]:[ (XxiREL x]])
Again, the characterization ::Ix [(Xx i REL x] can be left out because of its redundancy. Viet selects a quantity of the mass-denotatum. The denotatum of the quantificationally modified NP is now an individual although it contains m,ither an individual-noun like in der Tee aUf dem Tisch "the tea on the table" nor is it specified by number features. The denotatum of an individual has been formed by means of quantification. The referential thetarole of such an NP can only be specified by the SF of the indefinite article. EX ([TEA, (xXi [Xi
k
X]] : ::Ic [[QUANT X ~ [N +c]]).
To check the semantic status of an individual-NP let us test the pseudopartitive DP in a sentence modified by adverbials of duration:
(2)
Er trank stundentang Wodka. (Wodka is a mass-noun.) "He drank vodka for hours."
216
ANITA STEUBE & ANDREAS SPATH
(3)
*Er trank stundenlang etwas Wodka, (etwas Wodka is an individual-DP) *"He drank some vodka for hours."
(3) could have an acceptable interpretation only if the repetition of the event of drinking some Vodka is expressed. 3.
Partitive Constructions German and Russian partitive constructions can be described by the same syntactic structure. The syntactic structure of partitive constructions is more elaborate than that of pseudo-partitives. In addition to the modifying QP (cf. Zimmermann 1991), the nominal projection in Figure 3 has a right modifier, too, which is a partitive PP. In Bhatt (1990) the partitive PP is analyzed as a complement of the head noun of the partitive construction. But these head nouns have no position of an internal argument in their semantic representations. Therefore, the status of the PP as a complement is not accep table. In our as well as in Bhatt's analysis, however, the head noun of partitive constructions itself has no overt realization. The semantic content of the empty noun is that of a mass-noun or of a bare plural. The prepositional heads of the PP are German von and Russian iz. Only in poetic style or in archaic phrases can the German preposition von be replaced by genitive case. In Russian, however, the realization by case forms is normal. In modern Russian (especially in everyday language) genitive case is used, whereas partitive case as the older alternative is restricted to individual lexical items in the singular, and it is stilistically marked. Both the analytic realization of P° by a preposition and the synthetic realization of P° by case forms will be referred to as partitive case here. (4) (5) (6)
drei Tassen von dem starken Tee drei Tassen [starken Tees] gen.sg. tri caski [krepkogo caja/u] gen./part.sg. "three pots of the strong tea"
(7) (8)
drei von diesen Büchern tri iz [äix £/Hg]gen.pl. "three of these books"
217
NON-OVERT CATEGORIES IN RUSSIAN
die drei die drei
Liter Liter
SÜSSER WEIN SÜSSER WEIN SÜSSER WEIN SÜSSER WEIN SÜSSER WEIN
von Part. von Part. Part.
éti tri
kiloMUCH APPLES gramma stakan MUCH STRONG TEA nemnozko STRONG TEA nemnozko TEA MUCH STRONG TEA MUCH STRONG TEA MUCH TEA
Part.
odin
VIEL VIEL viel viel viel
dem des dem des
süßen süßen süßen süßen süßen
Wein Weins Wein Weins Weins
+ + + + +
sg. sg. sg. sg. sg.
Jablok + pl.
Part, kakogo-to krepkogo caju/-a+sg. Part, etogo krepkogo caju/-a+sg. Part. [+/-def.] caju/-a+sg. Part, etogo krepkogo caju/-a+sg. Part. [+/-def.] krepkogo caju/-a+sg. Part. [+/-def.] caju/-a+sg.
Fig. 5: LF representation of German and Russian partitive constructions
Whereas pseudo-partitive constructions denote an unspecified number of individuals or an unspecified quantity of a substance, partitive construc tions either express an unspecified subset of a contextually specified or maximal sum or an unspecified subpart of a contextually specified or maximal substance. Maximal sets or maximal quantities of substances are individuals. They are the denotata of the partitive PPs. In German, the definite or indefinite articles are used to specify the denotata of their argument-noun phrases as individuals; cf. the definite DP-arguments of the
218
ANITA STEUBE & ANDREAS SPATH
partitive prepositions von dem Weins in the German examples above. What grammatical means does Russian make use of to specify these partitive DPs? Partitive constructions in Russian are restricted to argument positions of verbs which select DPs denoting quantities. Furthermore, the verbs must have perfective aspect. The following are examples of partitive DPs in object position. (9)
Anna vypila perf. stakan caja/u gen./part.sg. a glass of (the) tea. " "Anna drank
(10) Daj| perf. mne xleba gen.sg. "Give me some (of the) bread. " (11) On priné sperf. papiros gen.pl. "He brought some (of the) cigarettes." (12) Oleg kupil perf. saxaru part.sg. "Oleg bought some (of the) sugar. " These examples and their English translations make it clear that the quantifier need not be overtly realized but is always semantically present. If, however, an object-PP of this kind of a perfective verb is realized in a normal structural case, the DP does not denote a nonspecific subquantity of an individual but the whole individual: (13) Oleg vypil perf. vodku acc.sg. "Oleg drank the vodka." (14) On prinës perf. papirosy acc.pl. the cigarettes. " " He brought Vodka and papirosy denote discourse objects. Finally we give some examples of partitive constructions in subject position of unaccusative verbs. (15) Gostej gen.pl. priexalo perf./ "What a lot of guests came!" (16) Ptic gen.pl. natetelo perf. nam. "Many birds came flying to us."
NON-OVERT CATEGORIES IN RUSSIAN
219
(17) Nagotovleno perf. zapasov gen.pl. " Collected were supplies. " Again the corresponding noun phrases in the nominative case denote discourse objects; cf. (18) Gosti nom.pl. priexali perf. "The guests came." Perfective verbs in Russian are used to express events, i.e. terminated states of affairs; cf. Oleg vypil vodku. "Oleg drank the/a vodka." (Vodka is a mass-noun.) If the process of drinking or an unlimited iteration of drinkingevents is expressed, the imperfecţive counterpart of the verb must be used: Oleg pil vodku. "Oleg was drinking vodka/Oleg drank vodka by habit." (Vodka is a mass-noun) Where German makes use of the articles to denote individuals, cf. der Tee/ein Tee - Tee, Russian uses verbal aspect, cf. vypit' pit'. This means that German specifies the denotata of the DPs involved in the states of affairs via the articles whereas Russian specifies the denotata of verbs or better: of sentences via aspect (cf. Steube 1997). German Oleg trank Vodka. "Oleg drank Vodka." Oleg trank den/einen Vodka. "Oleg drank the/a Vodka." Oleg trank etwas Vodka, (pseudo-partitive) "Oleg drank some Vodka." Oleg trank etwas von dem Vodka. (partitive) "Oleg drank some of the Vodka. "
Russian Oleg pil (imp&n.) vodku (acc.sg.) "Oleg drank Vodka." Oleg vypil (perf.) vodku. (acc.sg.) "Oleg drank the/a Vodka." Oleg vypil (perf.) vodki (gen.sg., partitive) "Oleg drank some of the Vodka."
Fig. 6: Local grammatical contexts
Perfective aspect expresses the terminativity of the state of affairs, i.e., it is an overt sign of eventhood. If partitive constructions demand perfective verbs in Russian, the partitive constructions must denote individuals: an unspecified subquantity of a contextually specified or maximal substance or an unspecified number of a contextually specified or maximal sum of individuals. German sentences express the terminativity of the state of affairs not on the semantic but on the conceptual level, by interpretation: If the number of individuals or the quantity of a mass involved in the state of affairs is finite, the state of affairs must be finite as well; cf.
220
ANITA STEUBE & ANDREAS SPÄTH
(19) *Er trank stundenlang etwas von dem Wodka. *"Er drank for hours some of the vodka." A second test for the individual-status of partitive constructions are floating quantifiers. They are used to topicalize the PP. (20) Caj part. sg. on vypil stakan. "Of the tea he drank a glass." (21) [Interesnyx knig]gen.pl. on "Of the interesting books he
proätal toVko odnu. read only one."
(22) Mal'ăkov gen.pl. ostalos' sem'. "Of the boys stayed seven. " What is topicalized in these sentences are the modifying PPs on the right side of the empty head noun in figure 5. The topics caju, interesnyx knig, mal'ăkov express contextually definite individuals, maximal substances or maximal sums, and the whole sentence denotes an event. Again German and English use the articles to express the sortai specification of the noun phrases. Russian can achieve the same specification by means of aspect and by the use of genitive or partitive case. Having argued on the basis of the meaning of partitive constructions, we will turn to their Semantic Form. Figure 5 shows that the quantifier is non-overt when measure phrases are used. Its Semantic Form remains intact, however. There is no semantic difference between the corresponding quanti fiers in pseudo-partitive or in partitive constructions either, although partitive constructions might be thought to realize what is called the proportional reading of the quantifiers. Therefore we need only one semantically underspecified lexical entry for the quantifiers under consideration. The list of lexical entries in figure 4 must be enlarged by the SF representation of partitive case only. It expresses the part-whole relation between the quanti fied submass/partial sum expressed by the quantified non-overt head noun and the contextually specified or maximal mass/sum expressed by the modifying PP. As already mentioned partitive case may be realized analyti cally by a preposition or synthetically by case endings. SF of partitive case: First we would like to compositionally construct the SF representation of the partitive construction mnogo Its syntactic representation is
NON-OVERT CATEGORIES IN RUSSIAN
221
Mnogo combines with a mass-noun only. Therefore the non-overt head noun TEA cannot be specified as to number but must be amalgamated with the mass-template instead. The SF of the DP in the modifying PP is obtained in the following way: The sortally underspecified lexical entry is marked by singular, and its referential theta-role is specified by the definite article, because the NP in the partitive PP denotes a contextually specified or maximal mass which is an ontological object. The grammatical context justifying this specification of the DP is partitive case.
This expression of category N replaces ζ in the SF of partitive case and thereby forms the SF of PP:
The next step will be to amalgamate the SF of QP and the SF of the nonovert mass-noun by modification. read: a mass w which is a subquantity of a discourse object χ The quantifier finally specifies the denotatum of the sub-mass of a contextually specified or maximal mass as having a quantity higher than the norm.
To complete the partitive DP, the referential theta-role of the partitive construction λw must be fixed. We will specify this isolated partitive construction by the default article, since we know neither the grammatical function of this partitive construction nor the position in the information structure of the sentence. The missing information constitutes the relevant local contexts for further referential specification of the determiner phrase.
222
4. I II III
ANITA STEUBE & ANDREAS SPÄTH
Arguments in favour of the non-overt categories in partitive constructions The non-overt nominal head noun separates the two agreement chains as they should be; the Semantic Form representations represent the meaning of the construction correctly; the lexical entries of the quantifiers in pseudo-partitives and in partitives remain the same. The latter argumentation counters Diesing (1992) who proposed two lexical entries for weak quantifiers.
Let us begin with the morphological characteristics of the phrases in Russian partitive constructions.
The quantifier nemnozko modifies the non-overt head noun TEA after it has been modified by the PP caja/-u. There cannot be an agreement chain containing a non-overt category and its modifier. The overt quantifier nemnozko therefore is in its short form. It has no influence on the modifying PP to the right of the non-overt head noun, because Russian agreement chains are restricted to NPs and PPs. The empty head noun prevents us from constructing a direct phrasal connection between the overt quantifier and the partitive PP. The non-empty head noun however marks the PP on its right side as its modifier either via genetive or partitive case, or - in rare cases via the overt partitive preposition iz "of". All the categories in the PP agree in case. Partitive case is realized as genitive case by the pronouns, the adjectives and by all the plural nouns. Mass-nouns like caj "tea" which have an old partitive form in its paradigm 'may optionally use this instead of genitive case. (25)
More complex partitive constructions show more clearly that there is no agreement chain ranging from the QP-modifier to the PP-modifier and that
NON-OVERT CATEGORIES IN RUSSIAN
223
partitive case is neither imposed by the overt quantifier nemno žko "some" nor by the noun in the MP dva malen'kix stakana "two little glasses"; cf. (26)
Only dva realizes the oblique case imposed by the verb vypit'. Numbers in measure phrases impose case themselves. Dva imposes genetive plural on the adjective malen'kix and genetive singular on the noun stakana in MP. Here the agreement chain ends. Čajal-u realize partitive case imposed by the non-overt head noun as seen above. Only partitive case has semantic mean ing, whereas the genitive imposed by dva is an agreement feature only. If the non-overt head noun was not considered in the LF structure of partitive constructions, the complex morphological relations would be too difficult to explain; cf. (27)
ja kupila tol'ko odnu acc.sg. MANY BOOKS. "Of the interesting books I bought only one."
Number odnu "one" imposes nominative sg. on its agreement-partners in an MP. So the case form of the partitive PP interesnyx knig cannot be imposed by any overt category in this sentence. Traditional grammars have to make use of non-overt categories as well. When they make stakan "glass" responsible for partitive case in example (26), a non-overt measure-noun like PIECE would have to be introduced into the analysis of this example. But a non-overt category like this could not explain the semantic relations in partitive constructions. Russian partitive constructions are restricted to the theta-positions of transitive and unaccusative verbs. The case imposed by the verb should be visible on the NP containing the head noun. Russian modifiers are included in the agreement chain of the head noun (as long as no other categories like, e.g., numbers impose case on neighbouring categories in the NP themselves). When the head noun is phonologically empty, the case imposed by the verb can be visible on modifiers only. Our explanation makes it clear that in the difficult example Interesnyx knig ja kupila tol'ko odnu quoted as (27) the partitive construction is in a structural case position of the verb kupit' "to
224
ANITA STEUBE & ANDREAS SPÄTH
buy". The non-overt head noun does not realize morphological features. They are realized by its modifiers (cf. odnu) in accord with the agreement principles of the language. This leads us to the conclusion that Russian partitive case unlike Finnish partitive case cannot be a case imposed by verbs (cf. Kiparski 1996 & 1997). Russian partitive case is a case imposed by nonovert head nouns of partitive constructions. Anyone who would argue that non-overt categories do not impose case would have even more general difficulties explaining that a modifying quantifier (either Q° or the noun in MP) imposes case, since modifiers are not case assigners. We will now turn to the Semantic Form of partitive constructions. First, we will consider the selection restrictions between the non-overt head noun and the noun in the partitive PP between which partitive case esta blishes a part-whole relation. Such relations exist only between masses of different kinds or between sums of different kinds, cf. "MUCH TEA of the tea"; "MANY SWEET APPLES of the sweet apples". On the other hand, the identity of the corresponding noun phrases does not include the details of (the process of) their sortai specification in their relative contexts. If our description did not make use of non-overt categories these selectionai restric tions would have to be explained in the conceptual system although the context necessary for the explanation is an overt grammatical context. Let us go on to the non-overt quantifiers. A look at German shows that the overt or non-overt realization of the qantifier is independent of pseudopartitive or partitive constructions. Figures 3 and 5 show that the quantifier is non-overt when measure phrases are used. The argumentation in support of non-overt quantifiers in QPs containing measure phrases is taken from Bierwisch (1987:193ff.). We will have to add arguments in support of only one underspecified SF representation of quantifiers for pseudo-partitive and partitive constructions, although partitive constructions might be thought to realize what is called the proportional reading of the quantifiers. The precon dition for only one SF representation of the quantifier is the presence of the non-overt head noun. It renders explicit the fact that partitive constructions are enlarged pseudo-partitive constructions. The quantifier has the same function in both the constructions. It quantifies a bare plural or mass-noun and turns the denotatum of the resulting NP into an individual. Partitive con structions are more complex only in so far as the quantified mass-noun is more complex, denoting a partial mass of a contextually specified or maximal mass. This complexity is independent of the lexical entry of the quantifier. If the non-overt head noun was not used in the construction, the so-called
NON-OVERT CATEGORIES IN RUSSIAN
225
proportional reading of the quantifier would have to account for the meaning of the non-overt head noun and its selection restrictions on the partitive PP in addition to its own meaning in order to establish the correct semantic representation of the part-whole relation. That the so-called proportional reading of the quantifier is brought about in local contexts can be tested by further examples; cf. (28) Many ANTS were on the plate, (weak quantifier) (29) MANY ants were on the plate. (weak quantifier with strong/proportional reading) In (29) MANY realizes contrastive stress and is focussed, ants belongs to background information. Background information is conceptually defined by the common knowledge already activated by the speaker and the hearer. MANY corrects former assumptions about the quantity of the ants and asserts that the quantity is MANY. The definiteness effect bound to the proportional reading supports our semantic analysis of the PP of partitive constructions as discourse objects. In (29) the same reading is achieved in a different but again in a local grammatical context: the information-structural relations between syntax and suprasegmental phonology (cf. Steube & Späth, forth coming). We use (28) and (29) in support for our argument that there is only one lexical entry for the quantifiers under consideration, their so-called weak reading. To sum up, our description argues in support of the ease of mapping SF representations onto syntactic representations by the use of underspecified lexical entries. In the process of the compositional construction of the SF representations of phrases and sentences, the interpretations which can be achieved in strictly local grammatical contexts are carried out and become part of the SF representations of the phrases and sentences of which they are a part. References Bhatt, Christa 1990. Die syntaktische Struktur der Nominalphrase im Deutschen. ( = Studien zur deutschen Grammatik, 38). Tübingen: Narr, Bierwisch, Manfred 1987. "Semantik der Graduierung". Grammatische und konzeptuelle As pekte von Dimensionsadjektiven ed. by Manfred Bierwisch & Ewald Lang, 91-286 (=studia grammatica, 26/27.) Berlin: Akademie-Verlag.
226
ANITA STEUBE & ANDREAS SPÄTH
Birkenmaier, Willy 1979. Artikelfunktionen in einer artikellosen Sprache. Studien zur nominalen Determination im Russischen. (=Forum Slavicum, 34.) München: Fink. Dölling, Johannes 1992. "Flexible Interpretationen durch Sortenverschiebung". Zimmermann & Strigin 1992. 23-62. 1995. "Ontological Domains, Semantic Sorts and Systematic Ambiguity". International Journal of Human-Computer Studies 43.785-808. Franks, Steven 1995. Parameters of Slavic Morphosyntax. Oxford, New York, etc.: Oxford University Press. Isacenko, Aleksandr V. 1962. Die russische Sprache der Gegenwart. Teil 1. Formenlehre. Halle/S.: Niemeyer. Junghanns, Uwe & Gerhild Zybatow, eds. 1997. Formale Slavistik {Papers of the First European Conference on Formal Description of Slavic Languages. Leipzig, 30. 11.-2. 12. 1995). Frankfurt/M.: Vervuert. Kiparsky, Paul 1996 & 1997. Partitive Case and Aspect. Ms. Krifka, Manfred 1989. Nominalreferenz und Zeitkonstitution. Zur Semantik von Massentermen, Pluraltermen und Aspektklassen. München: Fink. Link, Godehard 1991. "Plural". Semantik. Ein internationales Handbuch der zeitgenös sischen Forschung ed. by Arnim von Stechow & Dieter Wunderlich, 418-440. Berlin & New York: de Gruyter. Steube, Anita 1997. "Der russische Aspekt und die Ereignisrolle des Verbs". Junghanns & Zybatow 1997. 213-227. & Andreas Späth (1998). "DP-Semantik und Informationsstrukturierung im Russischen (auf der Basis eines Vergleichs mit dem Deutschen)". Sprache und Pragmatik 46 ed. by Inger Rosengren. Lund: Germanistisches Institut der Universität Lund. & Andreas Späth (forthcoming). "Determination in German and Russian". Travaux du Cercle Linguistique de Prague (nouvelle série) 3. Vater, Heinz 1996. "Die funktionale Kategorie D". Die Struktur der Nominalphrase ed. by Horst Thilo Tappe & Elisabeth Löbel, 170-188. ( = Wuppertaler Papiere zur Sprach wissenschaft, 12.) Wuppertal: Bergische Universität-Gesamthochschule Wuppertal. Zimmermann, Ilse 1991a. "Die Syntax der Substantivgruppe. Weiterentwicklungen der Xbar-Theorie". Zimmermann 1991b. 1-26. , ed. 1991b. Syntax und Semantik der Substantivgruppe. (=studia grammatica, 33.) Berlin: Akademie-Verlag. & Anatoli Strigin, eds. 1992. Fügungspotenzen, (^studia grammatica, 34.) Berlin: Akademie-Verlag. Zybatow, Gerhild & Uwe Junghanns 1996. "Topiks im Russischen". Vortrag auf dem Treffen des Netzwerkes "Sprache und Pragmatik", 7. 10.—11. 10. 1996 in Rendsburg.
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE
DMITRIJ DOBROVOL'SKIJ Russian Academy of Sciences
1.
Aims and goals The aim of this paper is to consider some general problems of con trastive idiom analysis. Using empirical data from German and Russian, I want to discuss relevant issues of cross-linguistic equivalence and, in parti cular, to focus on factors which have to be taken into account while looking for a L2-equivalent of a given L1-idiom. 2.
On the notion of contrastive idiom analysis In the linguistic literature, different interpretations of the term 'contrastive' can be encountered. First, the contrastive idiom analysis can be understood in its most general sense, i.e., any kind of comparison of idioms taken from different languages can be considered to be contrastive. It does not matter how many languages are involved in the cross-linguistic investigation, which aspects are focused on, or which methods of analysis are made use of. The term 'contrastive' used in this sense embraces, inter alia, linguistic typology, language universals research, translation analysis and cross-cultural aspects of idiom comparison. That is, the terms 'contrastive' and 'cross-linguistic' are used as synonyms (cf. Burger, Buhofer & Sialm 1982:274ff., Palm 1995:113ff.). In some studies, the term 'confrontative' (German 'konfron tativ', Russian 'konfrontativnyj') is used in a similar sense (cf. Eckert 1979:75, Fleischer 1982:30-31, Wotjak 1992:197ff.). Second, contrastive idiom analysis can be regarded as a special type of language comparison which differs from all other kinds of cross-linguistic research. This reading of the term 'contrastive', which goes back to
228
DMITRIJ DOBROVOL'SKIJ
Rojzenzon & Avaliani (1967), implies that the term is understood as a hyponym of the term 'cross-linguistic'. The difference between contrastive idiom analysis and all other cross-linguistic approaches can be explained through the following features: (a) the subject of comparison are idioms of two languages L1 and L2; (b) contrastive analysis is synchronically oriented and aims at re covering both differences and similarities between idioms of L1 and L2; (c) factors, such as genetic relationships, typological parallels, geographic contacts, do not play any role in the framework of this approach; (d) the aims of contrastive analysis are mainly of a practical nature. These criteria allow us to distinguish the contrastive approach to linguistic facts from other cross-linguistic approaches (for further details with regard to phraseology cf., e.g., Dobrovol'skij 1988:62ff., Abraham 1989, Dobrovol'skij & Piirainen 1997). Third, a very narrow reading of the term 'contrastive' is encountered in linguistic studies. According to this view, contrastive idiom analysis is engaged in the identification and description of cross-linguistic differences only. That is to say, all cases of cross-linguistic parallels remain beyond the scope of contrastive analysis understood along these lines. This reading is mentioned, for example, in Rajxstejn (1980:8), Durco (1994:24), and Foldes (1996:15). It implies that the term 'contrastive' is interpreted as a hyponym of the term 'confrontative', which is, in its turn, a hyponym of the term 'cross-linguistic'. 3.
Problems of cross-linguistic equivalence In what follows, only the contrastive approach proper will be discussed, i.e., all other kinds of cross-linguistic idiom research will be put aside. Among the three different readings of the notion of contrastive analysis, the second one is favoured here. This approach is not only widespread, but also the one that is encountered most often in today's literature on phraseology. Furthermore, this approach is the most effective one in terms of the goals pursued here - that is, a detailed comparison of some German and Russian idioms which will show that most idioms traditionally regarded as absolute parallels, in reality, are far from being equivalent. Let us, first, illustrate this by an example.
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 229
The German idiom Eulen nach Athen tragen lit. "to carry owls to Athens" is traditionally described as an equivalent of the Russian idiom ezdit' ν Tulu so svoim samovar lit. "to go to Tula with one's own samovar" (cf. also the English idiom to carry coals to Newcastle). Although the cultural specifics of the Russian idiom (due to the constituents Tula and samovar) as well as that of the English idiom (due to the constituent Newcastle) have often been pointed out, the semantic equivalence of these idioms has never been questioned (cf. for German and Russian Binovic & Grisin [1975:160], and for Russian and English Lubensky [1995:716]).1 However, this assumption turns out to be an erroneous one, which can be easily proved by taking into consideration the contextual embedding of these idioms (cf. examples (1-6)). (1)
(2)
(3)
(4)
"Ich trage eine Eule nach Athen" heißt eine der schönsten Geschichten von Wolfgang Hildesheimer. ... An diese meister liche Erzählung mag sich erinnern, wer ... Günther Christiansen zu seinem 70. Geburtstag gratulieren will. Welches Loblied auch immer er auf den Jubilar singen möchte: Irgendeiner war immer schon vor ihm da. (Börsenblatt für den deutschen Buchhandel, 26.11.1996) "'I am carrying an owl to Athens', that is the title of one of the most beautiful stories by Wolfgang Hildesheimer. ... This brilliant story should be kept in mind by everyone who ... wants to congratulate Günther Christiansen on his 70th birthday. Whatever hymn one would sing to glorify the celebrated one, another one has already sung it. " Sie tragen, was den Neuigkeitswert ihrer Mitteilungen betrifft, Eulen nach Athen. (Mannheimer Morgen, 5.1.1989) "They carry, as far as the informative value of their messages is concerned, owls to Athens." Über "Musikprofessor" Werner Beidinger Worte verlieren hieße Eulen nach Athen tragen. (Mannheimer Morgen, 6.2.1996) "To talk about 'music-professor' Werner Beidinger would mean to carry owls to Athens." V Tulu so svoim samovarom, izvestno, exat' bessmyslenno. A v Japoniju so svoim videomagnitofonom? Okazyvaetsja, vovse net. Japoncy, priznannye mirovye lidery v oblasti bytovoj élektroniki, prodajut vse men'se videoapparatury za rubez i, naoborot,
1 Compare a typical description of these idioms as "eine Variante von Internationalismen im phraseologischen Bereich" (Stojanova 1997:346). "Die Parallelität beim Aufbau und beim Gebrauch ist nicht zu übersehen - etwas Sinnloses wird durch den Phraseologismus wieder gegeben, um die Sinnlosigkeit einer anderen Handlung, eines Geschehens u.a. darzustellen und zwar im Kontext nicht neutral, sondern humorvoll über ironisch bis satirisch" (p. 347).
230
DMITRIJ DOBROVOL'SKIJ
(5)
(6)
vse bol'se vvozjat iz drugix stran. (Svobodnaja Gruzija, 14.9.1996) "It is a known fact that it is not reasonable to go to Tula with one's own samovar. And to Japan with his own video recorder? Obviously not at all. For the Japanese, who are leaders in the field of electronics worldwide, sell fewer and fewer video recorders abroad; on the contrary, they import them more and more." So storony zelanie Rossii vnedrit'sja na "lesnoj" rynok SSA cemto napominaet poezdku ν Tulu so svoim samovarom. Étot rynok kontroliruetsja severoamerikanskimi proizvoditeljami. (Segodnja, 13.3.1996) "Russia's wish to participate in the U.S. 'wood market' somehow reminds one of going to Tula with one's own samovar. This market is controlled by North-American producers." Proslo esce dva goda. My poexali na more vdvoem. ... sutlivo vozmuscalsja: - I zacem ja edu ν Tulu so svoim samovarom? Xoroso, - podygrala ja, - esli xoces', davaj otdoxnem otdel'no. (Novocerkasskie vedomosti, 13.2.1997) "Two years more passed. We went to the sea, two of us. ... You had joked: 'What for should I go to Tula with my own samovar?' Okay', I joined in play, 'we can spend our vacation separately if you like.'"
As contexts (4-6) show, the Russian idiom ezdit' ν Tulu so svoim samovarom lit. "to go to Tula with one's own samovar" means something like "to bring something to a place that already has an abundance of it" or "to bring somebody to a place where there are many persons of that kind". Hence, the possible objects of this idiom are physical entities (goods, com modities, people). On the other hand, the German idiom Eulen nach Athen tragen lit. "to carry owls to Athens" is mainly used to point out that it is not reasonable to present certain mental entities (ideas, artistic achievements, etc.) as being new if they are already well-known at a given place (cf. examples (1-3)).2 These differences seem to be connected with the image component of the idiom's semantic structure: samovar as an artefact predisposes the usage conditions to a certain extent. It is also significant that the noun samovar is used here in the singular form, combined with a possessive. This impedes the interpretation of the Russian idiom in the abstract sense (cf., on the contrary, 2
According to the analysed newspaper-corpora ("The Times" for 1992 and "The Guardian" for 1995), the English idiom to carry coals to Newcastle seems to be used in both readings with nearly the same frequency.
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 231
Eulen "owls" as a well-known symbol of wisdom, used in the plural form without determiner). This example also shows the role of the pragmatic component of the idiom's plane of content. Even in those contexts in which the semantic differences mentioned above are neutralized, it is not possible to translate the idiom ezdit' ν Tulu so svoim samovarom into Eulen nach Athen tragen (or into to carry coals to Newcastle) because of significant differences in the cultural specifics. 4.
Combinatorial properties of idioms from a cross-linguistic point of view Cases like this give sufficient substance to the claim that absolute crosslinguistic equivalents are, as a rule, lexicographic fictions which result from inaccurate semantic analysis. Real equivalents can mostly be found in cases in which both Russian and German idioms go back to the same source, for instance, Russian kamen' pretknovenija and German Stein des Anstoßes "stumbling-block". For this phenomenon some explanations can be put forward. First of all, the more progress in the semantic analysis that is achieved, the more items of the lexicon turn out to be language-specific. Many semanticists point to this fact, e.g., Wierzbicka (1996:15): As all translators know to their cost, every language has words which have no semantic equivalents in other languages, and every language draws semantic distinctions which other languages do not. [...] even comparing languages which are genetically, geographically, and culturally very close, for instance French and English, one constantly encounters examples of profound lexical differences.
Parallels in the 'core meaning' do not necessarily mean perfect equi valence in language use. This accounts especially for lexical units denoting non-physical entities (e.g., emotion concepts or mental predicates), which tend to depend much more on linguistic factors than, for example, words denoting physical entities (e.g., artefacts or natural kinds), because the latter have a different ontological status than the former. Since most idioms are predicates with regard to their semantics (i.e. lexical items denoting properties, states, actions, events and alike), they display a high degree of idiosyncrasy. Another factor of the tendency towards idiosyncratic configurations in the semantic structure of idioms has to be mentioned: the image component
232
DMITRIJ DOBROVOL'SKIJ
of idioms' plane of content. As an essential part of the conceptual structure (cf. Lakoff 1987:450ff., Gibbs 1990, 1993), mental images evoked by idioms influence their actual meanings. Furthermore, as stressed in section 3, the image component is often responsible for relevant restrictions in the usage of idioms (for details cf. Dobrovol'skij 1996, 1997, Baranoν & Dobrovol'skij in press). On the other hand, it should be emphasized that it is not possible to predict the actual meaning of an idiom on the basis of its image, i.e. its literal meaning. The same image can often be traced back to different conceptual metaphors (in the sense of Lakoff), that is to say, the lexical structure of an idiom does not tell us which conceptual metaphor has to be taken as the framework within which this idiom has to be interpreted. The result is that - both within the same language and cross-linguis tically - semantic differences between idioms cannot be predicted. To reveal those differences is a purely empirical endeavor, i.e., they can be discovered only by an investigation of their range of use. That is why the use of textual corpora and computational tools has become an indispensable part of semantic analysis in the recent years (for Russian idiom research, see Baranov & Dobrovol'skij 1995a). Hence, given an Ll-idiom and some I r idiums which are considered near-equivalents to the Ll-idiom, the only way to find out possible differences between them is to inquire into their combinatorial properties. 4.1
A case of non-one-to-one equivalence To take an example, let us consider the Russian idiom nakryt'sja mednym tazom lit. "to cover oneself with a copper basin" with its general meaning of "failure, flop, loss, defect, break down". In principle, this idiom can be translated into German by at least the following idioms: (7)
in die Hose gehen to go in the trousers = "to be a compete flop/washout; to come to nothing; to go to pot" 3 im Eimer sein to be in the bucket
3 Explanations marked with "=" are not semantic definitions, but only very rough English paraphrases (sometimes in combination with English near-equivalent units of figurative language, i.e. idioms or metaphors) which are supposed to provide the reader with a first impression of the meaning of the discussed German idioms, and not to cover the whole range of use. In what follows, I will show that in most cases the translation of an idiom into another language depends on context.
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 233
= "to be broken/wrecked/ruined; to stop working; to be down the drain; to go to pot; to be up the spout, to be bust " im Arsch sein to be in the ass = "to be broken/wrecked/ruined; to stop working; to go to pot; to be bust; to be fucked up" baden gehen (mit etwas) to go bathing (with something) = "to fail to accomplish something; to make a mistake; to goof up" den Geist aufgeben to give up the mind/spirit = "to stop working; to give up the ghost" in die Brüche gehen to go into the breakings = "to come to a rupture (of relations); to be dissolved (as in Their marriage was dissolved); to be ruined" ins Wasser fallen to fall into the water = "to fail to take place" es ist Essig (mit etwas) it is vinegar (with something) = "it is all up; it's down the drain; it did not take place" in die Binsen gehen to go into the rush = "to get lost; to be ruined; to fail to take place" j-m durch die Lappen gehen to go through the rags from someone (etymologically, this idiom goes back to the coloured "rags" used in hunting) = "to be lost for someone; to be not possible for someone to get something desired; to slip through someone's fingers" etwas in den Wind schreiben to write something into the wind = "to regared something as lost; to write something off as a dead loss" etwas in den Schornstein schreiben
234
DMITRIJ DOBROVOL'SKIJ
to write something into the chimney = " to regared something as lost; to write something off as a dead loss" In spite of the fact that in certain contexts all of the German idioms listed above can appear as translational parallels of the Russian idiom in question, it would be wrong to consider them equivalents on the level of a systemic comparison of these two languages. That is to say, a dictionary which has to provide the user with the knowledge of relevant linguistic properties of a given idiom should not just list all possible equivalents, but explain under which contextual conditions each one can be used as such. To be able to do it, we have, first, to ask the following questions: Nouns of which semantic classes (such as persons, artefacts, mental entities, events, and so on) can appear in combination with the Russian idiom in the case role of patient,4 i.e., what entities can "cover them selves with a copper basin"? In other words, combinatorial clusters have to be determined. The same accounts for every German idiom in question. In which way does the choice of the patient (its semantic class) modify the semantic structure of the Russian idiom? That is, which readings does this idiom have in different combinations? The same question is asked with regard to every German idiom under consideration. Which readings of given German idioms correlate with the readings of the Russian idiom nakryt'sja mednym tazom within corresponding combinatorial clusters? Answering this question allows us to approach the issue of cross-linguistic semantic relations of equivalence. Given some parallel readings which fulfil the conditions of crosslinguistic semantic equivalence, how do they differ with regard to the syntactic embedding of the idiom? That is to say, does the translation of a Russian sentence containing the idiom nakryt'sja mednym tazom into German imply any morphosyntactic transformations, and, if so, which ones? Finally, all the relevant pragmatic features of the Russian-German equivalents have to be compared. If, for example, one of the German idioms listed above turns out to be an absolute equivalent with one of 4
It is interesting to note that the semantic case of the noun in question (in the sense of Fillmore) is not as obvious as in standard examples mainly discussed in the literature on the theory of deep cases. While it looks like the patient in sentences in which the reading of 'being broken' is realized, it looks rather like the objective in sentences presupposing the reading of 'getting lost'. For reasons of space, these questions cannot be discussed here in more detail. In what follows, the term 'patient' will be used to refer to this deep case (see also Baranov & Dobrovol'skij 1995b:xvii). Compare in this connection Givón's (1990:565) description of the patient as "an inactive, non-volitional, non-controlling" situation partici pant, "one that registers the changes-of-state associated with the event".
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 235
the Russian idiom's readings from a semantic point of view, it is still possible that it cannot be used as a correspondence in translation of real texts, because, for instance, its usage is limited to certain text-types. In what follows, I will try to describe linguistic properties of the idioms listed above by answering these questions, in order to find out the whole range of cross-linguistic correspondences between them. For reasons of space, I will group the discussed idioms according to their combinatorial settings and semantic readings, instead of describing every idiom succes sively. 4.2
The role of context The Russian idiom nakryt'sja mednym tazom covers rather a wide range of combinatorial possibilities. The nouns filling in its semantic patientvalency (which corresponds in this case to the syntactic subject-valency, but does not need to do so in other cases, as we will see below) denote the following classes of entities: (a) (b) (c) (d) (e) (f)
physical objects (especially artefacts) intellectual products, results of human creativity possession and valuable commodities social events intentions and expectations human relations and unions
Let us consider these combinatorial settings and corresponding semantic readings in some more detail. (a) physical objects (especially artefacts): such as car, computer, TV-set, cup (8)
Kompjuter nakrylsja mednym tazom. The computer covered itself with a copper basin. "The computer does not work any more."
For an accurate semantic analysis different kinds of artefacts should be also distinguished, because sentences like (9) require a slightly different semantic description than (8). (9)
Cajnik nakrylsja mednym tazom. The teapot covered itself with a copper basin. "The teapot has been broken."
236
DMITRIJ DOBROVOL' SKIJ
From the list of German possible correspondences (7), only den Geist aufgeben, im Eimer sein, im Arsch sein5 can be used in combination with (a)nouns having the same semantic reading. The idiom j-m durch die Lappen gehen, which, in principle, allows the combination with physical objects, means something like "not to be able to get something desired" (cf. 10) and is, in this reading, synonymous to the expression j-m vor der Nase wegge schnappt werden (the German idiom j-m etwas vor der Nase wegschnappen means "just to beat somebody for something"). (10) Der Computer/die Teekanne ist mir durch die Lappen gegangen. The computer/the teapot has gone through the rags from me. "I am/was not able to get the computer/the teapot." (b) intellectual products, results of human creativity: such as papers, books, paintings (11) Moja stat ja nakrylas ' mednym tazom. My paper covered itself with a copper basin. "My paper has not been accepted." The comparison of (a) and (b) shows that the choice of the patient (its semantic class) modifies the semantic structure of the idiom in question (cf. examples (8) and (9) on the one hand, and (11) on the other hand). Possible German equivalents with respect to (b) are in die Hose gehen and baden gehen (mit etwas). In the last case, the translation of a Russian sentence containing the idiom nakryt'sja mednym tazom into German would require some morphosyntactic transformations. Compare sentences (11) vs. (12) in which the Russian syntactic subject moja stat ja "my paper" semantically corresponds with the German indirect object mit meinem Artikel "with my paper". (12) Ich bin mit meinem Artikel baden gegangen. I have gone bathing with my paper. "My paper has not been accepted." The acceptance of the idioms in die Binsen gehen, es ist Essig (mit etwas), etwas in den Wind schreiben, and etwas in den Schornstein schreiben used in combination with nouns denoting intellectual products is 5
Although all the relevant pragmatic features of the Russian-German equivalents cannot be discussed here in detail for reasons of space, it should be pointed out at least that the idiom im Arsch sein differs from other idioms in question stylistically, and its usage is limited to very informal, even vulgar, forms of communication.
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 237
questionable, (cf. sentences (13) to (16)). Note also the differences with regard to the syntactic embedding of these idioms. (13) ?Μeίn Artikel ist in die Binsen gegangen. lit. "My paper has gone into the rush." (14) ?ΕS ist Essig mit meinem Artikel. lit. "It is vinegar with my paper." (15) ?Ich kann meinen Artikel in den Wind schreiben. lit. "I have to write my paper into the wind." (16) ?Ich kann meinen Artikel in den Schornstein schreiben. lit. "I have to write my paper into the chimney." (c) aid
possession and valuable commodities: e.g., money, savings, financial (17) Moi sberezenija nakrylis' mednym tazom. My savings covered themselves with a copper basin. "My savings are lost."
Here again we observe a significant shift in the semantic structure of the idiom which puts the reading "to get lost" into focus. From list (7) the following German idioms, embedded in contexts (18) to (23) - all of them meaning "My savings are lost" - can be considered possible equivalents. (18) Ich kann meine Ersparnisse in den Wind schreiben. lit. "I have to write my savings into the wind." (19) Ich kann meine Ersparnisse in den Schornstein schreiben. lit. "I have to write my savings into the chimney." (20) Meine Ersparnisse sind im Eimer. lit. "My savings are in the bucket." (21) Meine Ersparnisse sind im Arsch. lit. "My savings are in the ass." (22) Es ist Essig mit meinen Ersparnissen. lit. "It is vinegar with my savings." (23) Meine Ersparnisse sind mir durch die Lappen gegangen. lit. "My savings have gone through the rags from me." In the last case (23), we deal with an important specification of the sentence meaning. In order to be understood properly, sentence (23) requires an additional presupposition, namely there had to be a good chance for protecting one's savings in a risky situation, but this chance has been lost. Therefore, sentence (17) can be translated into German by means of (23) only
238
DMITRIJ DOBROVOL'SKIJ
if this additional presupposition is a part of the situation standing behind sentence (17). This can also be seen in the context of example (23'): (23') So hat man nichts gespart, sondern sich einige hunderttausend Mark Eintrittsgelder durch die Lappen gehen lassen. (Mann heimer Morgen, 25.8.1989) "So, they saved nothing, but let some hundred thousand marks of entrance-fee go from them through the rags." According to the results of my questioning of native informants, the idiom in die Binsen gehen, though possible in this combinatorial setting in principle, still does not match the contextual conditions perfectly. (d) social events: such as concerts, performances, plays, parties, meetings, congresses (24) Koncert nakrylsja mednym tazom. The concert covered itself with a copper basin. "The concert did not take place." Considering the German idioms which combine with (d)-nouns, it is crucial to pay attention to the meaning they get in this combination. Whereas the idioms ins Wasser fallen and es ist Essig (mit etwas) in sentences (25) and (26) are understood in the same sense as the discussed Russian idiom in (24), the idioms in die Hose gehen, im Eimer sein, im Arsch sein, baden gehen (mit etwas), used in analogous sentences, mean "to be very bad", (cf., for example, (27)). (25) Das Konzert ist ins Wasser gefallen. The concert has fallen into the water. "The concert did not take place." (26) Es ist Essig mit dem Konzert. It is vinegar with the concert. "The concert did not take place." (27) Das Konzert ist in die Hose gegangen. The concert has gone in the trousers. "The concert was very bad." The idiom in die Binsen gehen, which is in general less familiar than the above mentioned expressions, can be interpreted in both these readings, compare (28):
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 239
(28) Das Konzert ist in die Binsen gegangen. The concert has gone into the rush. "The concert was very bad." or "The concert did not take place." The idioms j-m durch die Lappen gehen, etwas in den Wind schreiben, etwas in den Schornstein schreiben, taken in combination with (d)-nouns, get a completely different reading, cf. (29) to (31): (29) Das Konzert ist mir durch die Lappen gegangen. The concert has gone through the rags from me. "I missed the concert." (30) Ich kann das Konzert in den Wind schreiben. I have to write the concert into the wind. "I have no chance to attend the concert." (31) Ich k das Konzert in den Schornstein schreiben. Í have to write the concert into the chimney. "I have no chance to attend the concert." These examples show that a mere list of equivalents (and near-equi valents) as can be found in every traditional bilingual dictionary, from the point of view of language use, is not only insufficient, but even wrong and misleading. Without explicit comments on meaning shifts that every idiom undergoes in a given combinatorial setting, it is not possible to choose an ap propriate equivalent. If a L2-idiom combines with nouns of a given semantic class in the same way in which the discussed LI-idiom does, and if this L2idiom is equivalent with the LI-idiom in other combinatorial settings, there is still no guarantee that the semantic readings of both idioms are identical also in this case. So sentence (24) is not equivalent with sentences (27) to (31), though the idioms used in (27) to (31), while appearing in other sentences (e.g., (18), (19), (23)) are equivalent with the idiom used in (24). (e)
intentions and expectations: for instance, plans, purposes, hopes (32) Moi plany nakrylis ' mednym tazom. My plans covered themselves with a copper basin. "My plans could not be realized."
Possible German equivalents with respect to (e) are im Eimer sein, im Arsch sein, baden gehen (mit etwas), es ist Essig (mit etwas). In the last two cases, the corresponding morphosyntactic transformations (see above) are necessary.
240
DMITRIJ DOBROVOL'SKIJ
It is questionable whether the German idioms in die Hose gehen, in die Brüche gehen, etwas in den Wind schreiben, etwas in den Schornstein schreiben can be regarded as equivalents. The idiom ins Wasser fallen combines with (e)-nouns in principle. However, these nouns require modifiers which let them appear rather as future social events than as intentions and expectations, cf. (33): (33) Meine Pläneflirheute abend sind ins Wasser gefallen. My plans for this evening have fallen into the water. "My plans for this evening could not be realized." This corresponds with the core semantics of the idiom ins Wasser fallen, namely "to fail to take place". See (25) under (d). Compare also the blending of (d) and (e) in (34): (34) Die geplante Fahrt ist ins Wasser gefallen. The planned journey has fallen into the water. "The planned journey did not take place." (f)
human relations and unions: such as friendship, marriage, engagement (35) Ix brak nakrylsja mednym tazom. Their marriage covered itself with a copper basin. "Their marriage was dissolved."
German equivalents are in die Brüche gehen, im Eimer sein, im Arsch sein, es ist Essig (mit etwas), in die Binsen gehen. The idiom in die Hose gehen was judged by informants as questionable in this combinatorial setting, though a few examples of this kind could be found in textual corpora, cf. (36): (36) BAP ist 'ne politische Band. Daß der Anteil an politischen Stücken geringer war, hängt einfach damit zusammen, daß ich die Texte schreibe, und wenn bei mir soeben 'ne Ehe in die Hose geht, dann ist mein Kopf mit ein paar anderen Dingen voll. (Mannheimer Morgen, 3.5.1991) "BAP is a political band. That the number of political pieces were fewer goes back to the fact that it is me who writes the texts, and when a marriage of mine just goes in the trousers, then my head is full of a few other things."
RUSSIAN AND GERMAN IDIOMS FROM A CONTRASTIVE PERSPECTIVE 241
4.3
Concluding remarks: combinatorial settings and semantic readings Our analysis of the contexts taken from textual corpora (first of all, from "Korpus der geschriebenen Sprache" in Mannheim) and classified according to the noun-classes (a) to (f), shows that some readings have their 'favourite' idioms. Compare, for example, ins Wasser fallen for the reading "to fail to take place" in combination with (d)-nouns, or etwas in den Wind schreiben and etwas in den Schornstein schreiben for the reading "to regard something as lost" in combination with (c)-nouns. Combinatorial settings (a) to (f) are clearly connected to different semantic readings of the discussed Russian idiom, such as "to be broken", "to get lost", "to fail to take place", "to come to nothing". Readings of this kind, which have to be distinguished from each other in order to discover cross-linguistic equivalents accounting for the real language use, should not necessarily be regarded as different meanings. For discussion of related problems of polysemy vs. contextually based differences see, among others, Wierzbicka (1996: 242ff.). In the end, for a profound cross-linguistic analysis, splitting the whole range of use into different readings according to relevant combinatorial settings seems to be a necessary prerequisite. References Abraham, Werner (1989). "Idioms in Contrastive and in Universally Based Typological Research: Toward distinctions of relevance". Proceedings of the First Tilburg Workshop on Idioms ed. by Martin Everaert & Erik-Jan van der Linden, 1-22. Tilburg: Inst, voor Taal- en Kennistechnologie. Baranov, Anatolij Ν. & Dmitri] . Dobrovol'skij 1995a. "Sovremennaja russkaja idiomatika (proekt slovarja) ". Rusistika segodnja 4.99-115. Baranov, Anatolij N. & Dmitrij O. Dobrovol'skij 1995b. "Ustrojstvo i material slovarja". Russkaja zavetnaja idiomatika. Vesèîyj slovar' krylatyx vyrazenij compiled by Vasilij Buj, viii-xx. Moskva: Pomowski & Partner. Baranov, Anatolij N. & Dmitrij 0. Dobrovol'skij In press. "Vnutrennjaja forma idiomy i problema tolkovanija". Izvestija RAN, serija literatury ijazyka. Binovic, Leonid E. & Nikolaj N. Gris in 1975. Nemecko-russkij frazeologi ceskij slovar' = Deutsch-russisches phraseologisches Wörterbuch. Moskva: Russkij jazyk. Burger, Harald, Annelies Buhofer & Ambros Sialm 1982. Handbuch der Phraseologie. Berlin & New York: Walter de Gruyter. Dobrovol'skij, Dmitrij 1988. Phraseologie als Objekt der Universalienlinguistik. Leipzig: Enzyklopädie Verlag. 1996. "Obraznaja sostavljajuscaja v semantike idiom". Voprosy jazykoznanija 1.71-93. 1997. Idiome im mentalen Lexikon: Ziele und Methoden der kognitiv basierten Phraseologieforschung. Trier: Wissenschaftlicher Verlag Trier. & Elisabeth Piirainen 1997. Symbole in Sprache und Kultur: Studien zur Phraseologie aus kultursemiotiseher Perspektive. Bochum: Brockmeyer.
242
DMITRIJ DOBROVOL'SKIJ
burco, Peter 1994. Probleme der allgemeinen und kontrastiven Phraseologie: Am Beispiel Deutsch und Slowakisch. Heidelberg: Groos. Eckert, Rainer 1979. "Aspekte der konfrontativen Phraseologie". Linguistische Studien, Reihe A 56.74-80. Fleischer, Wolfgang 1982. Phraseologie der deutschen Gegenwartssprache. Leipzig: Biblio graphisches Institut. Földes, Csaba 1996. Deutsche Phraseologie kontrastiv: intra- und interlinguale Zugänge. Heidelberg: Groos. Gibbs, Raymond W. 1990. " Psycholinguistic Studies on the Conceptual Basis of Idiomaticity". Cognitive Linguistics 1:4.417-451. 1993. "Why Idioms are not Dead Metaphors". Idioms: Processing, structure, and interpretation ed. by Cristina Cacciari & Patrizia Tabossi, 57-77. Hillsdale, NJ: Lawrence Erlbaum. Givón, Talmy 1990. Syntax: A functional-typological introduction, vol. II. Amsterdam: Benjamins. Lakoff, George 1987. Women, Fire, and Dangerous Things: What categories reveal about the mind. Chicago & London: The University of Chicago Press. Lubensky, Sophia 1995. Random House Russian-English Dictionary of Idioms. New York: Random House. Palm, Christine. 1995. Phraseologie. Eine Einführung. Tübingen: Narr. Rajxstejn, Aleksandr D. 1980. Sopostavitel'nyj analii nemeckoj i russkoj frazeologii. Moskva: Vyssaja skola. Rojzenzon, Leonid I. & Julia J. Avaliani 1967. "Sovremennye aspekty izucenija frazeologii". Problemy frazeologii i zadaci izucenija v vyssej i srednej skole, 68-81. Vologda: VGPI. Stojanova, Stanka 1997. "Zur Semantik des Verbs in Phraseologismen (anhand von deutschem, bulgarischem, russischem und englischem Sprachmaterial)". Kaleidoskop der lexikalischen Semantik ed. by Ulrich Hoinkes & Wolf Dietrich, 341-348. Tübingen: Narr. Wierzbicka, Anna 1996. Semantics: Primes and universals. Oxford & New York: Oxford University Press. Wotjak, Barbara 1992. "Mehr Fragen als Antworten? Problemskizze - (nicht nur) zur kon frontativen Phraseologie". Deutsche Phraseologie in Sprachsystem und Sprachver wendung ed. by Csaba Földes, 197-217. Wien: Praesens.
WORD SENSE DISAMBIGUATION AN EXPERIMENTAL STUDY FOR GERMAN
WOLF PAPROTTE Universität Münster 1.
Sense tagging Sense tagging is the process of marking each word form token1 in a text with a label identifying a discrete sense, appropriate for the local context. Word sense disambiguation (WSD) solves the problem of assigning a sense to a polysemous word, using its immediate local context. In general, contextual senses correspond to one of the senses given in a dictionary entry. For exam ple the German Haushalt has two senses: (1) Haushalt = 1. economic unit of a family, "household" 2. budget of a political or social body, "budget" The system should be capable of distinguishing between these two senses given the context (2)
Bitte benennen Sie die zu Ihrem Haushalt gehörenden Familien mitglieder.
and adding a correct tag to the target word Haushalt denoting the actual meaning. Sense tagging (and WSD) is a difficult process even if it is done manually by an experienced lexicographer. A comparison of dictionaries shows that there is no total agreement on the number and kind of senses for a polysemous word; ultimately there is no easy consensus on the kind of tag given a specific context. Tradition has it that when faced with decisions about 1
We distinguish between wordform token (e.g., untersucht, untersucht, UNTERSUCHT), wordform type (e.g., untersucht), and word (type) (e.g., lemma word untersuchen).
244
WOLF PAPROTTE
discrete sense of polysemous items, lexicographers divide into 'lumpers' or meaning minimalists, and 'splitters' or meaning maximalists. Further diffi culties relate to judgements of cross context similarity of senses of a word both for expert and trained taggers (Fellbaum, Grabowski & Landes 1997) and to the problem of systematic polysemy due to metonymic, partonymic and metaphoric lexical relations between senses (Buitelaar 1997). Sense annotations are extremely useful information units for many applications of natural language processing, such as choosing the correct translation equivalent in machine translation. In the above example, the correct translation equivalent would be household rather than budget. The frequent failure of machine translation to provide contextually adequate equivalents as well as shortcomings in precision and recall for information retrieval systems can be blamed on the lack of sense disambiguation, or to put it differently, on the lack of an adequate semantic representation of the local context given the homonymous or the polysemous word in question. Dagan & Itai (1994) and Schütze & Pedersen (1995) investigate the practical utility of WSD for machine translation and information retrieval tasks. Krovetz (1997) notes that lexical ambiguity causes documents to be retrieved which are not relevant to the query: The query aids, for example, will retrieve documents on AIDS (med.) as well as on hearing aids, the verb form aids etc.. At the same time relevant documents which do not use the query term but its semantic (near) equivalents are not retrieved. In recent years, large text corpora have been the main source for WSD research. In most studies, the disambiguation process is viewed as similar to the morphosyntactic annotation/part of speech discrimination tasks, where the part-of-speech category of a word form is determined under the condition of co-occurring tuples. Similarly, given η lexical items, each with m senses, heuristics and/or statistical measures can be applied to pick the most likely sense for each lexical item in the given context (Buitelaar 1997; Wilks & Stevenson 1997). A typical corpus based algorithm uses a training set of many contexts of a syntactically ambiguous, but tagged word w in order to build up a statistical model and to learn classifiers which map instances of w and their contexts into a morphosyntactic category. This approach depends critically on manual tagging which is a costly and timeconsuming process. However, whereas syntactic tagging decisions are mostly based on operational and distributional criteria, a similar straightforward basis for disambiguation decisions does not exist in semantics.
WORD SENSE DISAMBIGUATION
245
Statistical sense disambiguation methods fall into three kinds (Leacock, Towell & Voorhees 1996), studies based on 1. Bayesian decision theory (e.g., Gale, Church & Yarowski 1992), 2. studies using content vectors (e.g., Salton, Wong & Yang 1975), 3. studies employing neural networks which will not be considered here (Cottrell 1989; Rumelhart, Hinton & James 1986). Studies using the Bayesian decision theory need a manually crafted corpus for training purposes and a method for estimating the required probabilities using sparse training data. The main element is a character string token, mostly a content word, with a frequency much lower than the stop word list of hyperfrequent function words and a few very frequent content words. Each string token has a set of saliences for each of its senses which is calculated from the training data. The salience is (3)
Pr(token | sense) / Pr(token).
The weight of a token for a given sense is the logarithm of its salience. To select the sense of the target word in a test context, the classifier computes the sum of the tokens' weights over all tokens in the context for each sense, and selects the sense with the largest sum (Leacock, Towell & Vorhees 1996). Vector-space models which are mainly used in information retrieval also need manually crafted corpora for training purposes. They build on word concepts rather than senses. Concepts are defined as word types, i.e. sets of (content) word form tokens minus their inflectional and derivational affixes. Each word concept defines an axis of the vector space, and a text in the corpus is represented as a point in this space. Each concept is represented by a single vector constructed from its training contexts. A vector in the space defined by the training contexts is also constructed for each test context. To select a concept for a test context, the product between its vector and each of the sense vectors is computed, and the sense with the largest product is chosen. The similarity between two texts is computed as a function of the vectors representing the two texts. Concepts which occur frequently in exactly one sense (and few texts) are favoured in that their vectors are weighted to reflect the relative importance of the concepts in the text (Leacock, Towell & Vorhees 1996). Because both methodologies follow a knowledge rich paradigm and depend critically on the availability of manually crafted training corpora, a method has to be found to avoid the time consuming, very costly process of
246
WOLF PAPROTTE
manually disambiguating every word in the corpus, for diverse topics of discourse. One possible approach is to rely on other corpus-external sources of knowledge, such as lexical or linguistic resources of machine readable dictionaries (LDOCE; Wilks & Stevenson 1997), thesauri (WordNet; Roget's, used in a large number of studies), new types of semantic lexicon (CORELEX; cf. Buitelaar 1997), or valence frames (Lin 1997). Even bilingual materials have been used for purposes of WSD (Dagan & Itai 1991; 1994). 2.
Approaches and options Knowledge rich approaches have a number of drawbacks. Semantically tagged corpora are practically non-existent; hand crafting of such semantic training corpora is extremely expensive and time consuming as large numbers of training examples are necessary before good disambiguating classifiers will have been learnt. Large numbers of training examples will not be suf ficient unless they come from diverse subject domains - an illusory hope. Thus, learning would have to take place for a large vocabulary and a large number of text domains and this would then favour an automatic procedure. However, even if word sense discriminators have been learnt, they will not disambiguate new senses for which classifiers have not been trained or, in case they have been trained, will not work correctly for diverse topic and discourse domains. Lexical and thesaurus resources of the desired kind, especially dictionary definitions, domain codes, and concept hierarchies are non-existent for German. The same holds true for subcategorisation information, includ ing selecţionai restrictions and/or thematic roles for a sizeable number of verbs such that a sense disambiguator could utilise several types of information. However, information on collocates (semantic associations) has been computed for hyperfrequent content words at the Arbeitsbereich Lingui stik, and part-of-speech tagging can be performed with the Münster tagger and tool set xlex.2 Still, a knowledge rich approach for a German WSD project can not be implemented because neither semantic training corpora nor machine readable dictionary resources are available. To tackle this critical NLP task of WSD and to resolve word senses on a level that deals with homonyms and 'lumped', i.e. broad, sense distinctions, we suggest to follow 2
The Münster tagger was trained on the MTP ( = Münster Tagging Project) training corpus of 500 000 manually corrected POS tags. It achieves an accuracy of 96.5%.
WORD SENSE DISAMBIGUATION
247
a knowledge poor paradigm of unsupervised learning, where information is gathered from raw corpus text (cf. Yarowsky 1995). Wilks & Stevenson (1997) note that a knowledge poor approach of unsupervised training has the drawback of not being relative to any welldefined set of senses as in a given lexicon. Their argument depends on the availability, the size and the quality of machine readable dictionaries which where available at all - at best reflect inconsistent lexicographic practice and follow diverse convictions as to the granularity of senses. For the moment, an unsupervised training model has to be the preferred paradigm for German and most languages other than English. The strategy to be followed is to make use of all information that is available in order to arrive at satisfactory results with acceptable effort. 3.
The larger context of meaning acquisition and disambiguation of acquired ambiguities Lexical acquisition of human learners generally begins with an input of raw text, spoken or written, presented in a real world situation which allows for additional information for every sensory channel. The learner's goal is to arrive at a representation which among other things correctly describes the semantic attributes of linguistic signs given specific contexts. Text in its raw form is the primary resource of the learner. It does not come ready with morphosyntactic POS annotation, nor is it given in any other pre-processed form. For the human learner, the objective is to acquire morphology, syntax, semantics all at once, constructing representations from unrestricted, general ly hardly didacticised, raw input and acquiring form-meaning pairs from many single examples, it is plausible to assume that a human learner will use all available information during comprehension and learning to constrain the constructive process. The larger his/her existing knowledge base the easier the representation task for new entries: knowledge of the situation of dis course, syntactic factors, existing semantic knowledge, semantic relations be tween words, observed co-occurrence restrictions and encyclopedic knowl edge all constrain the inscription of meanings. There is no multimodal information for a machine learner and no knowledge base is available. Yet, the machine task of learning form-meaning pairs given a context of use should utilise as much information as possible. The maximum of information is the context itself. It could be used as an inscription characterising the sense of the word token in question. Without a generalisation about the characteristics of the context and a notion of similar
248
WOLF PAPROTTE
contexts, a meaning maximalism will obtain (Pustejovsky 1995): "[·..] words can assume a potentially infinite number of senses in context." From various angles (figurative usage, and motivated polysemy such as metaphor, nonce formations, innovations, etc.) Pustejovsky's position seems correct. Psycholinguistic studies of the human disambiguation process in a sentential process have led to two conflicting views (Tabossi 1988). One position holds that lexical access to an ambiguous word is an exhaustive, context-independent sub-process of the comprehension process. Context be comes operative only at a post-access stage, and only then guides the selec tion of the contextually relevant meaning of the ambiguous word. A second hypothesis challenges the context-insensitive model of lexical access. Lexical access to ambiguous target words is not always exhaustive and not insensitive to prior context. Prior context supplies necessary constraining information for the selective lexical activation effects. The constraining influence of the preceding sentential context on lexical access was experimentally obtained and context primed a very characteristic "salient" aspect of the dominant meaning of an ambiguous lexical item. (4)
Primes: SAFE, RED; target ambiguity: PORT (= safe harbour vel red wine)
The sentential context (5)
The violent hurricane did not damage the ships which were in the port, one of the best equipped along the coast.
biased the dominant meaning of port by priming its being safe, whereas the sentence (6)
Deceived by the identical colour, the host took a bottle of Barolo, instead of one of port, and offered it to his guests.
primes the fact that port is a red wine. Tabossi (1988:335) claims that priming a relevant aspect of the secondary meaning (port = wine) resulted in exhaustive access which was observed in the data. She notes, however, that it is unclear how these feature priming effects function in normal day-to-day, non-experimental contexts. Psycholinguistic studies thus provide no guidance for the question which contextual characteristics should be considered and modelled in automatic WSD.
WORD SENSE DISAMBIGUATION
249
Vis-à-vis Pustejovsky's position and the inconclusive experimental results of psycholinguistics, a common sense view seems in order. It would take into account memory and related processes which prune fine meaning distinctions that result from on the spot processing of input and lead to less specific, more general 'robust' semantic distinctions. These function well for large sets of contexts. Default assumptions, routine and similar 'performance' factors also speak against a meaning maximalist position as regards the semantic representation in the lexicon. Due to many 'situational' factors, the interpretive meaning of a piece of discourse can still differ widely from what has been inscribed in the lexicon. Given a polysemous target word w or a homonym and its context in a large corpus which has not been pre-processed, the aim of a word sense dis ambiguation system is to determine the current unambiguous ("robust") sense ('lump') of w in the given context, and to achieve this with unsupervised learning building mainly on the local context and on previous other uses of the same word. The basic hypothesis of WSD is that with a high probability, a target word will have only one meaning in each specific context of use. This hypo thesis predicts that few sentences will occur in a corpus which cannot readily be disambiguated by a human processor and which will allow ambiguous readings. (Jokes and puns do not constitute the representative text sort in any corpus.) The stronger prediction about the unambiguous occurrences of w in texts ("one sense per discourse"; Yarowsky 1995:189) does not seem war ranted from our experience. Depending on text type and text length, the se mantic coherence of the text or discourse will vary and with it the probability of unambiguous uses of w: Thus, a characteristic feature of short, technical texts will be their precise terminology rather than the use of terms in more than one reading. Rather than assume the unambiguous occurrence of w in texts, we prefer to assume the unambiguous occurrence of w in a given local context. The second hypothesis of this paper is that two occurrences of a word w have identical senses if they occur in same or similar local contexts. If the (local) contexts of w can be categorized into a similarity set on the basis of features relevant for sense discrimination, distinct senses of w will be recog nized. We define these features as all relevance weighted content words which provide co-occurrence information. The local context of a target word w is defined as the current sentence containing w, or as a window of size η words to the left and to the right of w. Elements of the stop word list (pre-
250
WOLF PAPROTTE
positions, particles, conjunctions and in general all the minor part of speech elements) will only be considered relevant as parts of 'multi-word units' (phrasal verbs, idioms, etc.). In general, the content words which make up the current local contexts have a frequency considerably smaller than that of the hyperfrequent elements of the stop word list. Our approach to context similarity based WSD rests on a method which allows us to define similar contexts without prior knowledge of the current sense of w. The system learns to disambiguate occurrences of w with diverse senses sl s2, s 3 ... sn by using other occurrences of w in the semantically un tagged corpus as training set. The training set is a set of similar sentences in which w occurs. Given this assumption, similar contexts to the current one containing w have to be found to enable learning. To avoid manual tagging of a similarity set, similar contexts are defined as containing at least one highly significant, relevant collocator of w in the current sense s¡. Thus, we assume that local contexts are similar if a number of 'same' or 'similar' ele ments, defined as topically related content words, occur in them as relevant collocators. Similar element does not mean synonym; similar means signifi cant co-occurrence (loosely speaking in the sense of a 'Wortfeld') or 'associ ation' ratios or salience values must obtain to w; e.g., the nouns Haushalt, Familie, and Einkommen are considered similar because they frequently occur in the same context, and we consider Familie, and Einkommen as strong indicators of one specific sense of w — Haushalt. Thus, similarity and relevance are computed on the basis of a co occurrence or salience metric. A context is represented by the list of content words co-occurring with w; the similarity set of contexts is the sum/union of all contexts that contain the same highly significant, relevant collocator found in the set of concordance sentences containing w. It may be noted that 'collo cator' is being used solely in the sense of word in a sentence which has a significantly high frequency of co-occurrence or salience and weight with respect to w. Pragmatic notions of collocation such as idioms, phrases etc. are not intended. Another feature determining the relevance of a collocator is its proximi ty to w. The smaller the distance between w and a collocator the larger the likelihood that the collocator is a good classifier and the larger his weight. The precedence relation between w and a content word is another important feature determining relevance. Precedence or succession of a collocator with respect to w may affect a collocator's potential effect on classification. The content word may appear left- or right-adjacent to w with frequently substan-
WORD SENSE DISAMBIGUATION
251
tially different classification results, cf. pesticide plant vs. plant pesticide (Yarowsky 1995:191, fn 9). Bigrams or trigrams preceding or following the ambiguous term may also be relevant features of the local contexts. In future work, additional features such as syntactic relationships (verbobject- and subject-verb relations) will be studied and incorporated into the decision on the weight of individual collocators in the metric for classifying similar/dissimilar contexts (Ng & Lee 1996). The range of potential features whose contributions to WSD has to be studied also includes the -
-
distinction between mass nouns and count nouns subcategorisation information and semantic restrictions on arguments of a verb as in: throw (NP subject: (+ anim); NP direct object (physical object; +movable)) subject area classification: AIDS (medical) global context markers, for example, taken from news wire code: min ister (politics): foreign minister, prime minister, etc. A local context (subject code: religion) may well override the restrictions of the global context: Protestant minister terminology effects versus domain specific senses of general vocabu lary: Internet, digitised, stochastic, database vs. address, parallel, loop, message passing, closed set
Contexts are defined as similar to the extent that they contain similar collocators; collocators are defined as similar to the extent that they occur in similar contexts. The circularity of the definition can be remedied in an itera tive procedure, whereby a start set of similar words is determined and de fines similar sentences whose analyses help to generate further similar words that appear in a number of new and now similar sentences (cf. Karov & Edelman 1996). 4.
Methodology The lack of resources other than raw corpus data and the Münster POS tagged corpus for German precludes all knowledge rich methodological ap proaches which have been reported in the literature. As a consequence, we take an approach of unsupervised learning as suggested by Yarowsky (1995) as point of departure. We also consider the work done by Karov & Edelman (1996) who rely on a machine readable dictionary to build up feedback sets for training. Our present aim is to test a number of available tools for texthandling and corpus analysis in their potential usefulness for tasks of WSD. It is hoped that in an iterative application of these tools, we will get WSD
252
WOLF PAPROTTE
results not worse than those reported in the literature. We also aim at acquiring expertise in the difficulties we are likely to encounter in developing an automatic WSD system. Experimental results will guide decisions on future methodology and system architecture. They should also help to define an acceptable quality level of the results. The following resources and tools have been used: -
a 20 million word corpus of German (ZEIT corpus) a tagged 500 000 word training corpus (MTP corpus) a concordancing tool that lists all occurrences of the target word and its inflected forms (part of xlex) a POS tagger for German that has been trained on the MTP corpus (part of xlex) a German morphology to lemmatise target word and contexts (part of xlex) a number of statistical tools (part of xlex)
The experiment is intended to test this combination of data and tools before implementing a more refined and efficient automatic procedure such as using seed discriminators and a decision list algorithm (Yarowsky 1994, 1995) The test set of polysemous words in this study includes Haushalt, Grenze, Maus, Kapital, Lager, verfallen, einsteigen. Examples will only be given for Haushalt. We followed a procedure that consisted of a pre-processing phase, an iterative process of building up relevant classifiers for senses, and a result phase of tagging the polysemous words in novel contexts. Pre-processing Step 1: A concordance is generated for each of the polysemous target word types (including all inflected forms and orthographic variants). The context of the target words is the whole sentence. In order to guarantee a minimal window size of 5 words to the left and right of the target word, the previous or the following corpus sentence is made to fill the gap in case the target word occurs before the fifth word or less than five words before the end of the sentence. 1013 concordance sentences were retrieved (cf. figure 1). Damit steigt das Volumen des schon im vergangenen Herbst für den Westteil verabschiedeten Haushalts von 27,3 auf etwa 40 Milliarden Mark. Als Ursprung und Zentrum des Haushalts, Indikator oder Motor sozialer Veränderun gen und technischer Innovation bilden das Herdfeuer und seine Nachfolger den Aus gangspunkt dieser faszinierenden bunten Lebensart-Schau.
WORD SENSE DISAMBIGUATION
253
Er würde allerdings sehr gedämpft ausfallen, wenn die Sanierung des Haushalts mit neuen Steuern auf Vermögen und Immobilien oder kräftigeren Erhöhungen der Ein kommensteuer verbunden wäre. Die OECD stellt bei der Bundesregierung "Entschlossenheit" in der Konsolidierung des Haushalts trotz der hohen Ausgaben fest. In ihrer Stellungnahme zu der von Berlitz unterschriebenen Magistratsvorlage erinnert die Kämmerin daran, daß der Innenminister als Aufsichtsbehörde bei der Genehmi gung des städtischen Haushalts 1992/93 den starken Anstieg der Personalkosten erwähnt und deutliche Zurückhaltung bei derartigen Ausgaben gefordert habe. Wenngleich die Stadt am Main aufgrund ihrer Steuer- und steuerähnlichen Einnahmen nach wie vor einen Spitzenplatz in Deutschland behauptet, ist diese Seite des Haus halts doch mit einem großen Unsicherheitsfaktor belastet: der von der wirtschaftlichen Lage abhängigen Gewerbesteuer, die immerhin zwei Drittel der gesamten Steuerein nahmen ausmacht und 1992 1,671 Milliarden Mark einbringen soll. Der Haushaltsausschuß des Bundestages hat in der sogenannten Bereinigungssitzung, die der zweiten und dritten Lesung des Haushalts in der kommenden Woche vorangeht, die Mittel für die Bundesakademie für Sicherheitspolitik in Höhe von 1,716 Millionen Mark gestrichen. Die Kämmerei führt schließlich noch an, daß die rot-grüne Koalition das Verfahren bei der Aufstellung der Haushalts "stark intensiviert" und damit den Arbeitsanfall in der Stadtkämmerei erheblich gesteigert habe. Michael Siebel, Reuters persönlicher Referent, sieht es als Erfolg an, daß die Filteranlage im Tambourbad saniert werden konnte, obwohl die dafür vorgesehen Mittel wegen der NichtVerabschiedung des Nachtragshaushalts und des Haushalts für dieses Jahr gefehlt hätten. Außerdem sei es Aufgabe der Stadt, durch eine "gesunde" Wirtschaftsförderungspolitik dazu beizutragen, daß besonders die Gewerbesteuereinnahmen so reichlich flössen, daß die Finanzierung des Haushalts und der für die Stadt lebensnotwendigen Projekte gesichert sei. Fig. 1: Concordance data Haushalt
The concordance is POS tagged. The Münster tagger achieves a cor rectness rate of close to 96.5% including lexical gaps (mostly proper names), and novel compounds. Step 2: A filter deletes all hyperfrequent elements of the stop word list. The residual contexts contain only content word forms. All word form types have been lemmatized so that each context around the target word is represented as a list of word types (i.e. the canonical lexical citation forms). A co-occur rence metric is computed on the basis of the maximum likelihood ratio (Dunning 1993, Lemnitzer 1997). Figure 2 gives an impression of the results (lemmatized, POS tagged, co-occurrence quantified concordance).
254
WOLF PAPROTTE
Wort y Prozent öffentlich Verabschiedung Sanierung privat städtisch Konsolidierung Mark Jahr Million Ausgleich Finanzierung Ausgabe ! Genehmigung Last Milliarde Volumen Defizit Einkommen Lesung Einbringung 1 Drittel Entwurf Vorlage 1 defizitär Bundesrepublik Aufstellung Gesamtvolumen
ty 80043 9586 891 3148 9264 3680. 604 46274 60523 77717 2325 3826 2800 1502 1694 33287 1353 1492 1667 1965 118 5364 3484 1752 97 11429 780 270
Ixy 66 43 31 31 29 19 14 28 28 29 14 13 11 10 10 17 9 9 9 9 6 10 9 8 5 10 6 5
erg +520.16821 +482.20941 +474.96533 + 395.99927 +233.56250 +217.87595 +202.69150 +202.53033 + 187.96259 + 182.73129 + 164.76523 + 138.12549 + 120.03459 + 119.65347 + 117.24644 + 117.16433 + 107.66180 + 105.90059 + 103.90442 + 100.94664 +96.40498 + 94.27687 + 90.67130 + 89.67247 + 80.47182 +79.34183 +73.49988 +70.07144
pattern 0 / 3 / 1 2 / 2 7 / Ó / X / 0 / 0 / 11 / 7 / 6 / 0/0/2/0/40/X /0/0/0/0/1/ 0/0/1/30/0/X /0/0/0/0/0/ 0/ 1/12/18/0/X / 0 / 0 / 0 / 0 / 0 / 0/0/0/0/29/X/ 0/0/0/0/0/ 0/0/0/0/19/X/ 0/0/0/0/0/ 0 / 0 / 5 / 9 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 1 / 2 / 6 / 6 / 0 / X/ 0 / 0 / 3 / 6 / 4 / 4 / 0 / 2 / 0 / 0 / X / 0 / 1 / 10 / 10 / 1 1/3/0/6 / 5 / X/ 0 / 0 / 3 / 3 / 8 / 0/0/1/13/0/ X/ 0 / 0 / 0 / 0 / 0 / 0/ 1 / 5 / 7 / 0 / X / 0 / 0 / 0 / 0 / 0 / 1 / 2 / 0 / 3 / 0 / X / 1/ 1/ 1/ 1/ 1 / 0 / 0 / 2 / 8 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 0 / 1 / 2 / 5 / 0 / X/ 0 / 0 / 0 / 1 / 1 / 2 / 2 / 1 / 0 / 0 / X/ 0 / 1 / 3 / 3 / 5 / 0 / 0 / 0 / 9 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 0 / 1 / 1 / 4 / 0 / X/ 0 / 2 / 0 / 0 / 1 / 0 / 0 / 3 / 1 / 0 / X/ 0 / 0 / 4 / 0 / 1 / 0 / 1 / 0 / 8 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 0 / 0 / 0 / 6 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 1 / 1/ i / 6 / 0 / X / 0 / 1 / 0 / 0 / 0 / 0 / 0 / 3 / 6 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 2 / 0 / 3 / 3 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 0 / 0 / 0 / 1 / 4 / X/ 0 / 0 / 0 / 0 / 0 / 3 / 1 / 0 / 0 / 0 / X/ 0 / 3 / 1 / 2 / 0 / 0 / 1 / 1 / 4 / 0 / X/ 0 / 0 / 0 / 0 / 0 / 0 / 0 / 1 / 4 / 0 / X/ 0 / 0 / 0 / 0 / 0 /
] 1
/
1
1
Fig. 2: Pre-processed concordance data; X marks the position of w Haushalt Iterative processing phase Step 3: A small number of collocators which intuitively and statistically appear to represent significant and good classifiers of each sense are selected manually from the list that resulted from the co-occurrence metric. Some of these collocators also appear in the dictionary definitions of the target words or in usage citations or phrases occurring with the entries. As in Yarowsky (1995), in a first step, a single defining classifier is chosen for each sense and used as a tag for the target word (Figure 3). Quantitatively and intuitively, the adjectives öffentlich und privat are salient discriminators. They mark and characterise the context of the target word and clearly distinguish the possible senses: öffentlich "budget", pri vat "household". Statistikliebhaber weisen oft anklagend darauf hin, daß Städte wie Frankfurt oder Köln sieben oder gar elf Prozent ihres öffentlichen Haushalts für Kultur ausgeben. Explosiv würde die Lage, wenn ein starker Nachfrageschub aus der DDR mit einem durch Schulden aufgeblähten öffentlichen Haushalt zusammentreffen würde.
WORD SENSE DISAMBIGUATION
255
Das hilft uns bei dringenden Investitionen, die im Haushalt nicht unterzubringen sind, also für öffentliche Gebäude, Schulen, Krankenhäuser, meint Finanzstaatssekretär Werner Heubaum. Das fehlt nun, und für den West-Geschmack muß das Sortiment kräftig verändert werden - die Einsparrate im öffentlichen Haushalt beträgt jetzt sieben Prozent. Für den öffentlichen Haushalt war der Energiesektor bisher ein Faß ohne Boden. Jeder private Haushalt in der alten Bundesrepublik hat nach den Angaben der Deutschen Bundesbank im Durchschnitt rund 100 000 Mark auf der hohen Kante, und Jahr für Jahr kommen über 7000 Mark hinzu. In der Bundesrepublik verfügen dreißig Millionen private Haushalte über ein Gesamtvermögen von über 1,8 Billionen Mark, das entspricht der stattlichen Summe von etwa 60 000 Mark pro Haushalt. Jeder private Haushalt muß seinen Stromverbrauch um zehn Prozent drosseln. Fig. 3: Similarity sets for Haushalt sensej and sense2
The disambiguated concordance lines containing the "inscribed tag" could eventually function as a training set for the classifier system, provided their number is large enough. The remaining concordance lines constitute the untagged residual. Manual inspection of the intermediate results shows that the number of disambiguated concordance lines is relatively small and that the seed öffentlich malfunctions as disambiguator in 8% of the cases (5 out of 65 co-occurrences) and privat similarly gives 8% wrong results (4 out of 50 cases) (cf. figure 4). Manche Sozialdemokraten erfinden regelmäßig neue Steuerarten für Besserver dienende, mit dem Ziel, die öffentliche Verschuldung zu bremsen; so als ob der kon solidierte Staat für die Bürger ein Wert an sich sei, wenn Ordnung in die Staats finanzen gebracht werden soll um den Preis der Dekonsolidierung privater Haus halte. Derart in die Enge getrieben, schüttet die Kommune dann gleich noch das Kind mit dem Bade aus und will sogar die öffentlichen Wannenbäder im Volksbad dichtmachen - in einer Stadt, in der es angeblich 6500 Haushalte ohne eigenes Bad gibt. Nicht zwei Stadthälften kommen zusammen, sondern vierzig Jahre Leben in unter schiedlichen Gesellschaftssystemen, mit unterschiedlichen öffentlichen und privaten Haushalten, unterschiedlichen Eigentumsverhältnissen, unterschiedlichen Planungs systemen, Infrastrukturen, Bildungssystemen, Kulturen, Alltagsverständnissen und nicht zuletzt - einer gemeinsamen deutschen, gleichwohl unterschiedlichen Sprache. Das System soll nicht nur in Haushalten, sondern auch für professionelle Präsentationen oder öffentliche Informationsterminals genutzt werden. Seit die beiden großen Privatsender RTLplus und Sat 1 in den Haushalten mit Kabel anschluß bei den Einschaltquoten mit den beiden öffentlich-rechtlichen Fernseh anstalten gleichgezogen haben und die zwei kleineren Sender Tele 5 und Pro 7 sich als Alternativen zu profilieren beginnen, ist der Kampf um Zuschauer und Werbespots gleichermaßen hart geworden. Fig. 4: Infelicitous disambiguations
256
WOLF PAPROTTE
Iteration of the procedure and intermediate results Step 4: An elegant procedure appeared to be the following. Disregarding the infelicitous cases, it seemed possible to find additional seed collocators by applying the log likelihood metric to the disambiguated and tagged concor dance data. The content words of the disambiguated sentences would be scored under the condition of the defining classifier minus the score of the classifier of contexts under the condition of all classifiers in the concordance. Instead, we simply repeated the application of the co-occurrence metric to the successfully disambiguated sets of öffentlicher Haushalt vs. privater Haushalt. The significant best ratios of co-occurrence with the target words and new collocators were selected. For both, the collocators and the newly found classifiers a ranking by the log likelihood ratio was computed and compared with the co-occurrence metric of step 3. In the case of öffentlicher Haushalt the following terms were found: Konsolidierung, Steuererhöhung, Verschuldung, Kürzung, Fehlbeträge; for privater Haushalt we found Einkommen, Großverbraucher, Gesamtvermö gen, Ersparnisse, verfügbar, verfügen, Energie. For the human user, this set of disambiguating collocators is intuitively convincing and seems promising. However, when the resulting new collocates were applied to the re sidual concordance, a problem arose due to the sparseness of the occurrence of the new seed words: Only in 22 cases did the application of the new seed words disambiguate senses correctly, and the application of the same metric did not significantly reduce the residual set of disambiguated concordance lines. The residual set was unaffected for approximately 85% of all concor dance lines. Iterating the process of generating seed collocators, therefore, does not appear to be a good procedure for finding new similarity sets. This preliminary result is clearly unsatisfying. A further problem may be noted. In the case of Haushalt, the two senses have similar frequencies. In all cases in which one or several of the senses of a term display marked differences in their frequencies, the attempted procedure of iterating the extraction of best collocators from the disambiguated set is bound to fail in the absence of solid statistical measures and to run into the sparse data problem. A simplistic strategy to reduce the size of the problem A new step 4: The above collocation matrix provides quantitatively important information on content words frequently co-occurring with the polysemous Haushalt, without, however, specifying their disambiguation potential with
WORD SENSE DISAMBIGUATION
257
respect to either of the two senses. Most of the residual ambiguous concordance sentences could be disambiguated if the content word collocators of figure 2 would unambiguously occur only with either of the two senses, as is assumed by our hypothesis 2. The heuristic assumption (hypothesis 2) was therefore extended: if any content word which occurred in the concordance sentences can be shown to only occur in one of the sets of already disambi guated concordance lines, and never in the other similarity set, it is con sidered to contribute significantly to the similarity set features indicative of the one sense and may be considered a relevant disambiguator in its own right for the residual set. Therefore, all the content words co-occurring with the two senses in the already disambiguated sets were listed. Those nouns, adjectives and verbs which only occurred with one of the senses were extracted from this list but only nouns were used as additional potential disambiguators in a differential list. Thus -
Asylbewerber, Defizit, West-Berlin, etc. were found to co-occur only with öffentlicher Haushalt (altogether 231 nouns were used as new disambiguators); Familie, Frau, Einkommen, Stromverbrauch, etc. were found to cooccur only with privater Haushalt (altogether 196 nouns were used as new disambiguators).
Results of the improved procedures These new discriminators were used on the residual set. The 231 nouns which were assumed to contribute to disambiguating sense1 reduced the residual set by 453 examples; the 196 nouns found to co-occur solely with sense2 generated another 395 disambiguated examples. Seed sets and the sets created by the application of the potential disambiguators covered 95 % of all sentences, which is a satisfying result from a quantitative point of view. Qualitatively, the results of this simple procedure were found to be also satisfying in that only 12% were incorrectly disambiguated for sense! öffent licher Haushalt, and only 11% were incorrectly disambiguated for sense2 privater Haushalt (cf. figure 5 for some good and some dubious examples): These figures are only slightly worse than those in other experiments. Bevor der japanische Steuerzahler in die Tasche greift, möchte er wissen, warum die Vereinigten Staaten ihren Haushalt nicht ebenfalls durch kräftige Steuererhöhungen in Ordnung bringen, will die Asahi Shimbun wissen. Wenn die Kosten der deutschen Einheit ohne Steuererhöhungen verkraftet werden sollen, so heißt es jetzt, dann würde eine umfangreiche Steuersenkung zugunsten der
258
WOLF PAPROTTE
Unternehmen ein beträchtliches Loch in die jetzt schon stark strapazierten Haushalte von Bund, Ländern und Gemeinden reißen. Sein Haushalt in Schleswig-Holstein hat die Belastungsgrenze aufgrund der Vereini gungskosten längst überschritten. 25,5 Milliarden Mark beträgt der Haushalt von West-Berlin, 13 davon sind Bundes hilfe. Die Inflationsrate ist gesunken, das Bruttosozialprodukt wächst, der Haushalt weist einen Überschuß auf. Trotz Wirtschaftswachstum und ausgeglichener Haushalte Rückgang der Inflation und Zufluß von Fremdkapital ist das Pro-Kopf-Einkommen nur in Kolumbien, Chile, auf Barbados und in Jamaica 1991 höher gewesen als vor zehn Jahren. Ihr Anteil am Haushalt steigt schneller als die Inflation. Mit mehr als 33 Milliarden Mark verschuldet sich der Bund schon in diesem Jahr, die Tilgungslasten blockieren einen immer größeren Teil des Haushalts; jede zusätzliche Schuldenmark vergrößert die Inflationsrisiken . Louven hofft, daß der Haushalt bis dahin saniert ist und einer der Nachfolger von Bundesfinanzminister Theo Waigel das Defizit übernimmt. Um ein Ansteigen der Arbeitslosigkeit auf über zehn Prozent zu vermeiden, stellte Papandreou marode und unproduktive Unternehmen kurzerhand unter staatliche Kon trolle und finanzierte deren Defizite aus dem Haushalt Im Frühjahr 1988 rang sich der Europäische Rat zu dieser grundlegenden Reform der Agrarpolitik durch, um Überschüsse bei der Produktion und Defizite im Haushalt drastisch zu begrenzen. Wir werden eine Ausgangssituation haben, in der die Bundesrepublik ein bestimmtes Defizit im Haushalt decken wird . In den Grabenkämpfen um den Haushalt kann sich nun als Vorteil erweisen, was manche als sein Defizit sehen: *** Sie brauchen keine Verfeinerung der Kläranlage auf 98 Prozent, während wir hier noch 57 Prozent der Haushalte ohne Anschluß an eine Kläranlage haben, wenn der gleiche Betrag, der im Westen diese Steigerung auf 98 Prozent ermöglicht, die Abwasseraufbereitung im Osten auf 80 Prozent anheben könnte. Fig. 5: Disambiguated sentences using the differential list; *** wrong classification
5.
Discussion The outlined method allows to generate disambiguated contexts of polysemous words and to generate training contexts for automatic WSD. Seed disambiguators and elements of a differential list have been used as "signatures" (Zernik 1991a:93ff.) indicative of a disambiguated sense. The results are overall satisfying with respect to the quantity of sentences disam biguated from the whole concordance set (approx. 85%) and with respect to quality. However, it must be noted that the differentiating potential of öffent lich and privat as seed disambiguator also rested on an implicit syntactic effect: both seeds mainly occur as modifiers in an NP with Haushalt. The majority of infelicitous disambiguations in the seed sets were mostly due to the seeds occurring outside the NP phrase or both occurring in the same sentence. Nevertheless, the remaining error rate of approx. 12% infelicitous disambiguations seems acceptable. The above method is interactive and relies
WORD SENSE DISAMBIGUATION
259
also on manual selection of seed disambiguators which serve as classifiers for WSD. It is still time consuming. The next steps of our research are directed towards fully automating WSD. The attempt of tagging a polysemous target word by lists or clusters of signatures marks a new approach towards generalising over the notion of similar contexts (even though in a still simplistic way). A more satisfying notion of similarity between two local contexts would rest on the availability of a thesaurus which points to the 'common' information units, on fully using additional syntactic information, for example on modifier - head or verb argument structures (cf. Lin 1997); and more refined statistical pattern recog nition as well as on methods of dealing with sparse data. It also appears necessary to enlarge the local context to chunks of cotexts. Co-texts which include a larger number of classifying collocates are most probably better examples for feature sets indicative of a specific sense of a homonym or polysemous word. Classifiers in a signature may also have multiple senses so that increasing the number of potential disambiguators will constrain multiple readings. The differential list has been found an astoni shingly effective means for WSD. Acknowledgement The author would like to thank Lothar Lemnitzer for valuable comments, suggestions and help with the preparation of the data. A joint follow-up project is under preparation.
References Boguraev, Branimir & James Pustejovsky, eds. 1996. Corpus Processing for Lexical Acquisition. Cambridge, Mass. & London: MIT Press. Brown, Peter F., John Cocke, Stephen A. Della Pietra, Vincent J. Della Pietra & Robert Mercer 1991. "Word Sense Disambiguation using Statistical Methods". 29th Annual Meeting of the Association for Computational Linguistics : Proceedings of the Conference 18-21 June, 264-270. University of California, Berkeley: Association for Computational Linguistics. Bruce, R. & Janyce Wiebe 1994. "Word Sense Disambiguation using Decomposable Models." 32th Annual Meeting of the Association for Computational Linguistics: Proceedings of the Conference 27-30 June, 139-145. New Mexico State University, Las Cruces: Association for Computational Linguistics. Buitelaar, Paul 1997. "A Lexicon for Underspecified Semantic Tagging". Dept. of Computer Science, Brandeis University, Waltham, Mass. 02254-9110, USA. Available from Internet (Spring 1998). Choueka, Yaacov & Serge Lusignan 1995. "Disambiguation by Short Contexts". Computer and the Humanities 19.147-157.
260
WOLF PAPROTTE
Cottrell, Garrison W. 1989. A Connectionist Approach to Word Sense Disambiguation. San Mateo, Calif.: Morgan Kaufmann. Dagan, Ido & Alon Itai 1994. "Word Sense Disambiguation using a Second Language Monolingual Corpus". Computational Linguistics 20.563-596. & Ulrike Schwall 1991. "Two Languages are More Informative than One". 29th Annual Meeting of the Association for Computational Linguistics: Proceedings of the Conference 18-21 June, 264-270. University of California, Berkeley: Association for Computational Linguistics. Duda, Richard O. & Peter E. Hart 1973. Pattern Classification and Scene Analysis. New York: Wiley. Dunning, Ted 1993. "Accurate Methods for the Statistics of Surprise and Coincidence". Computational Linguistics 19.61 -74. Fellbaum, Christiane, Joachim Grabowski & Shari Landes 1997. "Analysis of a HandTagging Task". Proceedings of the Workshop "Tagging Text with Lexical Semantics. Why, What, and How?", 34-40. Washington, D. C : Association for Computational Linguistics. Gale, William Α., Kenneth W. Church & David Yarowski 1992. "A Method for Disam biguating Word Senses in a Large Corpus". Computer and the Humanities 26.415-439. Hearst, Marti A. & Hinrich Schütze 1996. "Customizing a Lexicon to Better Suit a Compu tational Task". Boguraev & Pustejovsky 1996. 77-96. James, M. 1985. Classification Algorithms. New York: Wiley-Interscience. Karov, Yael & Shimon Edelman 1996. "Learning Similarity-Based Word Sense Disambigua tion from Sparse Data". Proceedings of the Forth Workshop on Very Large Corpora, Copenhagen. Available from Internet (Spring 1998). Krovetz, Robert 1997. "Homonymy and Polysemy in Information Retrieval". 35th Annual Meeting of the Association for Computational Linguistics. Proceedings of the Conference 7-12 July, 72-79. Universidad Nacional de Educación a Distancia, Madrid: Association for Compuational Linguistics. Leacock, Claudia, G. Towell & Ε. Μ. Voorhees 1996. "Towards Building Contextual Representations of Word Senses Using Statistical Models". Boguraev & Pustejovsky 1996. 97-113. Lemnitzer, Lothar 1997. Akquisition komplexer Lexeme aus Textkorpora. Tübingen: Niemeyer. Lin, D. 1997. "Using Syntactic Dependency as Local Context to Resolve Word Sense Ambiguity." 35th Annual Meeting of the Association for Computational Linguistics. Proceedings of the Conference 7-12 July, 64-71. Universidad Nacional de Educación a Distancia, Madrid: Association for Compuational Linguistics. Ng, Hwee u & Hian Beng Lee 1996. "Integrating Multiple Knowledge Sources to Disam biguate Word Sense: An exemplar-based approach". 34th Annual Meeting of the Association for Computational Linguistics. Proceedings of the Conference 23-28 June, 40-47. University of California, Santa Cruz: Association for Computational Linguistics. Pustejovsky, James 1995. Th Generative Lexicon. Cambridge, Mass.: MIT Press. Rivest, R. L. 1987. "Learning Decision Lists". Machine Learning 2.229-246. Rumelhart, David E., Geoffrey E. Hinton & R. J. Williams 1986. "Learning Internal Repre sentations by Error Propagation". Parallel Distributed Processing: Explorations in the
WORD SENSE DISAMBIGUATION
261
Microstructure of Cognition ed. by David E. Rumelhart & James L. McClelland, vol. 1: Foundations, 318-363. Cambridge, Mass.: MIT Press. Salton, Gerard, A. Wong & C. S. Yang 1975. "A Vector Space Model for Automatic Indexing". Communications of the ACM 18(11).613-620. Salton, Gerard 1989. Automatic Text Processing. Reading, Mass.: Addison Wesley. Schütze, Hinrich 1992. "Dimensions of Meaning". Proceedings of Supercomputing '92. 787796. Los Alamitos, Calif.: IEEE Computer Society Press. & Jan O. Pedersen 1995. "Information Retrieval Based on Word Senses". Fourth Annual Symposium on Document Analysis and Information Retrieval, 161-175. Las Vegas, Nv. Small, Steven I, Garrison W. Cottrell & Michael K. Tanenhaus, eds. 1988. Lexical Ambiguity Resolution. San Mateo, Calif.: Morgan Kaufmann. Tabossi, Patrizia 1988. "Sentential Context and Lexical Access". Small, Cottrell & Tanenhaus 1988. 331-342. Wilks, Yorick & Mark Stevenson 1997. "Sense Tagging: Semantic tagging with a lexicon". Proceedings of the SIGLEX Workshop "Tagging Text with Lexical Semantics". Available from Internet . Yarowsky, David 1994. "Decision Lists for Lexical Ambiguity Resolution: Application to accent restoration in Spanish and French". 32th Annual Meeting of the Association for Computational Linguistics: Proceedings of the Conference 27-30 June. Association for Computational Linguistics. New Mexico State University, Las Cruces. Available from Internet < http : //xxx. lanl. gov/abs/cmp-lg/9406034 > . 1995. "Unsupervised Word Sense Disambiguation Rivaling Supervised Methods". 33th Annual Meeting of the Association for Computational Linguistics: Proceedings of the Conference 26-30 June, 189 -196. Massachusetts Institute of Technology, Cambridge: Association for Computational Linguistics. Zernik, Uri, ed. 1991. Lexical Acquisition. Exploiting on-line resources to build a lexicon. Hillsdale, New Jersey, Hove & London: Lawrence Erlbaum. 1991a. "Traini vs. Train2: Tagging word senses in corpus". Zernik 1991. 91-112.
GENERAL INDEX A
- begriffssystematisch 70
ACTION 32
- action function 28 - action game 25→ dialogic action game Adäquatheitskriterium 140 adjective 36 advanced learner 5, 38 allomorph of connected speech 181 amalgamation 214 ambiguity 9 - ambiguous word 248 ANGER 50-51, 54 -+ Ärger angewandte Sprachwissenschaft 68 Äquivalenz 81-83, 87 → equivalence Archilexem 145 Ärger 73 → ANGER assimilation 182 AWARENESS 32 bare plural 224 Bayesian decision theory 245 Bedeutung → meaning - aktuelle 151 - lexikalische 151-152 - okkasionelle 151 - übertragene 150 - usuelle 151 - Bedeutungsäquivalenz 152 → equivalence - Bedeutungsposition 69, 153, 156 → meaning position - Grundbedeutung 141 - Systembedeutung 141, 150 - Textbedeutung 71 begrifflich 74, 81
CALL, computer assisted foreign language learning 119-120 - CALL-program 120 cognitive basis 30, 32 - cognitively based theory of use 37 colligation 15 collocation 7, 15, 39 → Kollokation - collocate 246, 256, 259 - collocator 250-251, 254, 256-257 coming to an understanding 25 competence - communicative competence 29 - linguistic competence 54 comprehension process 248 computational lexicology 1 - computerlinguistic model viii computer course 128 - computer program 126 concept hierarchy 246 - word concept 245 → word - conceptual metaphor 232 concordance data 254 content vector 245 context 9 → local context, → word and context - context similarity 250 - contextually relevant meaning 248 contrastive vii → kontrastiv - contrastive idiom analysis viii, 227 → idiom - contrastive pragmatic analysis 42 - contrastive semantics 34, 119
264
GENERAL INDEX
convention 25, 27 convergence 182 co-occurrence metric 253, 256 - co-occurrence restriction 247 core 15 corpus 3, 28, 54→ Korpus, → written and spoken corpus - corpus-based vii - corpus-oriented approach viii cotext 3, 9 context - local context 249-251 cross-linguistic equivalence 227-228 culture 121, 125, 128 - cultural 124 - cultural difference 120 - cultural specifics 229, 231 - culturally 122 D Deutsch 55-64, 75-80, 97-101, 133-137, 142-144 → German dialogic action game 27 → action game - dialogic interaction 25 dictionary 3, 28, 54, 119, 122, 127-128 → Wörterbuch, → machine readable dictionary - bilingual 42 - dictionary definition 246 disambiguation process 244 - human disambiguation process 248 divergence 182 domain code 246 DUDEN-Programm 68 Dutch 120-131, 133-137 E elision 182 EMOTION 32, 45, 47, 52, 121, 123-124, 126, 128, 131 - basic emotion 46-48 -Emotionen 73, 103-107 encyclopedic knowledge 247 English 45, 51, 55-59, 169, 173, 190 equivalence 33, 37, 119-121, 123, 128, 131 → Äquivalenz,
→ Bedeutungsäquivalenz - equivalent 119-123, 127-128, 130-131 - natural equivalence 131 - natural equivalent 130 Etymologie 72 F Fachsprache 89 - Fachwörterbuch 73 false friends 182 feature priming effect 248 figurative 22 form-meaning pair 247 French 169 → Französisch frequency 27, 54 → Frequenz - frequent 42 Frequenz 81 → frequency Französisch 75-80 → French funktionale Opposition 154 Funktions verbgefüge 144, 154 generative Semantik 71→ Semantik genitive case 216 G German 45, 51, 120-128, 130-131 139, 173, 178, 184, 192 → Deutsch Grammatik 68 - grammatical word-form 174, 177 graphematic domain 164 Grundbedeutung 141 → Bedeutung H homograph 174 - homography 177 homonym 249, 259 - homonymy 177 homophone 174 - homophony 177-178 human learner 247 I idiom 2 → contrastive idiom analysis - combinatorial property of idioms 231 idiosyncratic configuration 231 image 232 - image component 231-232
GENERAL INDEX
imperfective 219 information retrieval 244-245 interference 124-125 interlingual level-shift 184 intonation 185 intuition 27, 31, 64→Intuition Intuition 72, 81→ intuition irony 6 Italian 45, 51, 59-64, 139, 189 → Italienisch Italienisch 142-144 → Italian item-environment 8 knowledge base 247 Kollokation 81, 150→collocation Kontext 141 kontrastiv 91→· contrastive - kontrastive Grammatik 68 - kontrastive Lexikographie 83 - kontrastive Linguistik 68, 82 - kontrastiver Vergleich 73 - kontrastiv-pragmatisch 69 - kontrastiv-semantische Analyse von Emotionen 55-64, 75-80, 133-137, 97, 107, 110-115 →Semantik Korpus 91→corpus - korpusbasiert 91 - Korpuslinguistik 80 L lambda-categorial 214 language use 25 learner's dictionary 120, 124, 126, 128 lexematic domain 164 lexeme 173-174 lexical - lexical access 248 - lexical acquisition 247 - lexical ambiguity 244 - lexical entry 209 - defective lexical entry 209 -lexical item 1, 173-174, 177 - lexical relation 244 - lexical unit viii, 1, 34, 184→ lexikalische Einheit
265
lexicon 3→Lexikon Lexikographie 69, 75 Lexikon 108→lexicon - lexikalische Einheit 91→lexical unit lexis-in-use 41 liaison 182 literal 232→literal meaning lumper 244 M machine learner 247 - machine readable dictionary 246-247, 251→dictionary - machine translation 244 manual tagging 244→·tagging mass noun 211, 224 mass-template 215 maximum likelihood ratio 253 meaning 4, 165→Bedeutung - meaning acquisition 247 - meaning equivalence 30 - literal meaning 232→literal - meaning maximalist 244, 249 - meaning minimalist 244 - meaning position 33, 53 → Bedeutungsposition - type of meaning 28 measure phrase 209 medium → spoken medium, → written medium - medium-bound 174 - medium-bound word-form 177 - medium-dependent unit 176 - medium-independent unit 176 - medium-independent word-form 177, 181 - medium-transferability 176 mehrsprachige Handlungskompetenz 91→multi-lingual action competence Mehrwort-Einheit 161→multi-word unit Merkmalsemantik 154 metacommunicative activity 179 Metapher - in linguistischer Metasprache 96 - metaphoric 244
266
GENERAL INDEX
- metaphorical phrase 54 metapragmatic 120-122, 126 metonymie 244 modal particles 184 model-oriented 39 - model-oriented approach viii morphemic domain 164 Morphemvergleich 84 morphosyntactic annotation 244 - morphosyntactic transformation 236, 239 MOTION 32
motivation 165 MTP ( = Münster Tagging Project) - MTP corpus 252 - Münster tagger 246, 253 multi-lingual action competence 42 → mehrsprachige Handlungskompetenz multi-word unit 41→ Mehrwort-Einheit - multi-word-expression 35 N native speaker 28, 54 neural network 245 Neutralisierung 149 NLP 11 non-overt quantifier 209 Normbedeutung 141 noun 36 numeral 190 object 36 object-oriented approach 40 Opposition 146 - Oppositionalität 141 orthographic word-form 173 Ρ paradigmatic 13 Paralleltext 81 - Paralleltextsegment 81 part of speech 36 - part of speech discrimination 244 - part-of-speech tagging 246 partitive case 216 - partitive construction 209
- pseudo-partitive construction 210 partonymic 244 pattern 121, 127 - Pattern-Drill 91 - patterning 125-126 perfective aspect 219 phonological word 173 phraseologisch 154 Polysemie 71→polysemy polysemy 248→Polysemie - polysemous word 243, 252, 258, 259 - polysemous target word 249 - systematic polysemy 244 Prädikationsbereich 156 - Prädikationsfeld 154 - prädikatives Feld 73 pragmatic vii, 165, 120-121 → Pragmatik - pragmatic model viii - pragmatic turn 28 Pragmatik 69→pragmatic - Semantik vs Pragmatik 98, 109 → Semantik - Pragmatiker 70 -pragmatisch 71, 81, 91 predicate 29 - complex predicate 52 - predicating field 32 - predicating position 33 - predicative function 28 preposition 196 priming 248 Principle 27 - Action Principle 27 - Dialogic Principle 27 - principle in social interaction 27 proportional reading 220, 225 Prototypensemantik 151 Q quantifier 209 quasi-synonymy 33→Synonymie quasi-universelle Bedeutungsstruktur 156→universal structure
GENERAL INDEX
R rank scale 9 RATIONALITY 32
Rede - Redebedeutung 141 - Technik der Rede 145 - wiederholte Rede 145 referent 164 - referential function 28 reversal 6 rule 26 - rule-governed way 39 Russian 209→Russisch Russisch 95→Russian S salience 245 seed collocator 256 - seed disambiguator 258 - seed discriminator 252 selecţionai restriction 246 semantic -> Semantik - semantic coherence 249 - semantic constellation 52 - semantic field 95 - Semantic Form 209 - semantic knowledge 247 - semantic preference 15 - semantic prosody 15 - semantic reversal 6 semantics proper 28 Semantik 69→generative Semantik - Semantik vs Pragmatik 98, 109 → Pragmatik - strukturalistische Semantik 71 - Semantiker 70-71 - 'semantisches Dorf 97-98, 102, 109 sense 165 - sense annotation 244 - sense discrimination 249 - sense tagging 243 - sense vector 245 shandi-form 180 sign → Zeichen - linguistic sign 182 signification 165
similarity set 250, 257 SOCIAL BEHAVIOUR 32
sortai specification 212 Spanisch 74 - spanisch 88 sparse data 259 - sparse data problem 256 - sparse training data 245 spatial concept 198 speech formula 130 splitter 244 spoken medium 175→medium Sprachendidaktik 68 Sprachenpaar 69 Sprachenvergleich 90 - Sprachvergleich 67-68 Sprachtechnik 141 Sprechhandlung 71 - sprachliches Handeln 91 Sprachtypologie 68 Statistical model 244 structural model viii subcategorisation information 246 synchronic word family 165 Synkretismus 149 Synonymie 71→quasi-synonymy - synonym 250 syntactic relationship 251 - syntagmatic 13 - syntax 37 Τ tagging 243→manual tagging Teilwortschatz 73 text 5→Text - text corpus 27, 64 - textual corpus 232, 240-241 Text→text - Textbedeutung 71→Bedeutung - Textvergleich 67, 82, 90 thematic role 246 thesaurus 246 Topikalisierung 146 transfer 124 transformation 12 two-level model viii, 30
267
268
GENERAL INDEX
U Übersetzung 81 unaccusative verbs 218 universal structure 32, 47, 52→quasi universelle Bedeutungsstruktur, → Universalstruktur Universalstruktur 73→universal structure unsupervised learning 247, 249, 251 - unsupervised training 247 V valence frame 246 vector-space model 245 verbal aspect 219 verbo fraseologico 144 verification 25 Verwendungsprofil 141, 155 Verwendungsweise 91, 153→way-of-use visual morpheme 175, 178 vocabulary 120-121, 124-125, 131 W way of speaking 120 way-of-use 32, 34, 50, 53, 64
→ Verwendungsweise weak quantifier 222 weight 245 word - word and context 39→context word concept 245→concept - word-form 173-174 - word-in-use 34 - word sense disambiguation 243 Wortfamilie 69, 71, 98 Wortfeld 96, 154-155 - Wortfeldtheorie 140, 155, 161 - Wortfeldvergleich 83 Wörterbuch 68, 70, 91→dictionary Wort-im-Text 70, 91 Wort-im-Wörterbuch 91 written and spoken corpus 179→corpus written medium 175→medium X xlex 246 Ζ Zeichen 71→sign
LIST OF CONTRIBUTORS
Prof. Dr. Valerij Dem'jankov, Nauchno-Issledovatel'skij Vychislitel'nyj Centr, Moscow Lomonosov University, Vorob'evy Gory, RUS-119899 Moskva, vdemiank@vdemiankov. msk.ru Prof. Dr. Dmitrij O. Dobrovol'skij, Russian Academy of Sciences, Institute of Russian language, Volkhonka 18/2, RUS-121019 Moskva, dobrov© irlex. msk. ru Prof. Dr. Claude Gruaz, CNRS and Sorbonne Paris III, 5, rue aux Boulangers, F-27240 Avrilly, [email protected] Eckhard Hauenherm M.A., Westfälische Wilhelms-Universität Münster, Fachbereich 11: Philologie, Sprachwissenschaft, Bispinghof 2B, D-48143 Münster, [email protected] Prof. Dr. Wolf Paprotte, Westfälische Wilhelms-Universität Münster, Fachbereich 11: Philologie, Englisches Seminar, Arbeitsbereich Linguistik, Hüfferstr. 27, D-48149 Münster, [email protected] Prof. Dr. Christian Schmitt, Universität Bonn, Romanisches Seminar, Am Hof l,D-53113 Bonn. Prof. Dr. Christoph Schwarze, Universität Konstanz, Fachgruppe Sprachwissenschaft, Postfach 5560, D-78434 Konstanz, Christoph.schwarze© bluewin.ch
270
LIST OF CONTRIBUTORS
Prof. Dr. John Sinclair, Tuscan Word Centre, Azienda Casanova, 409, Vellano, 1-51010 Pescia (PT), [email protected] Dr. Andreas Späth, Universität Leipzig, Zentrum für Kognitionswissenschaften am Zentrum für Höhere Studien, Augustusplatz 10/11, D-04109 Leipzig, [email protected] Prof. Dr. Anita Steube, Universität Leipzig, Institut für Sprach- und Überset zungswissenschaft, Augustusplatz 9, D-04109 Leipzig, [email protected] Prof. Dr. Edda Weigand, Westfälische Wilhelms-Universität Münster, Fach bereich 11: Philologie, Sprachwissenschaft, Bispinghof 2B, D-48143 Münster, weigand@uni-muenster. de Dr. Henning Westheide, Rijksuniversiteit Leiden, Vakgroup Duitse Taal- en Letterkunde, Postbus 9515, NL-2300 RA Leiden, westheide@rullet. leidenuniv.nl