Papers on functional sentence perspective [Reprint 2015 ed.] 9783111676524, 9783111291321

Citation preview

JANUA LINGUARUM STUDIA MEMORIAE NICOLAI VAN WIJK DEDICATA edenda curat C. H. VAN SCHOONEVELD Indiana University

Series Minor,

147

PAPERS ON FUNCTIONAL SENTENCE PERSPECTIVE

edited by

F. D A N E S

1974 ACADEMIA P U B L I S H I N G HOUSE OF THE CZECHOSLOVAK ACADEMY OF SCIENCES PRAGUE

MOUTON THE

HAGUE

·

PARIS

Scientific

Editor:

Dr. Milos Dokulil, DrSc. Scientific

Adviser:

Prof. Dr. Jozef Mistrik, DrSc.

•© Academia, Publishing House of the Czechoslovak Academy of Sciences 1974

All rights reserved No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means without written permission from the publishers Printed in Czechoslovakia

CONTENTS

Foreword JAN FIRBAS,

9 Some aspects of the Czechoslovak approach to

problems of functional sentence perspective ALES SVOBODA,

On two communicative dynamisms

11 38

The place of "functional sentence perspective" in the system of linguistic description

43

Zur Stellung der Thema-Rhema-Gliederung in der Sprachbeschreibung

54

Topic - comment structure in a generative grammar with a semantic base

75

MICHAEL A . K . HALLIDAY, PETR SGALL,

ÖSTEN DAHL,

LÁSZLÓ DEZSÔ — GYÖRGYE SZÉPE,

Two problems of topic-

comment WOLFGANG DRESLER,

81 Funktionelle Satzperspektive und Text-

theorie

87

Functional sentence perspective and the organization of the text

106

Zur Frage der funktionalen Satzperspektive im dramatischen Text

129

FRANTISEK DANES,

JOSEF FILIPEC,

HPHHA H. KOBTYHOBA, AKTyajibHoe HJieHemie H CHCTeMa jotnca (Ha MaTepnane pyccKoro a3biica)

142

HeKOToptie ΠΟΗΗΤΗΗ Teopun aKTyajitΗΟΓΟ HJieHeHH« npHMeHHTejIbHO Κ H3yHeHHK) BblCKa3bIBaHHH Β pa3r0B0pH0H penn 152

OJLTRA A . J I A N T E B A ,

6 OJïbrA Β. CtíPOTHHHHA, 0 6 AKTYAJIBHOM «UIEHEHHH Β pa3rOBOpHOH pena

172

PAVEL NOVÁK, Remarks on devices of functional sentence perspective EJIEHA ΓΕΟΡΓΗΕΒΑ,

175

CpeflCTBa aKTyajn>Horo HJIEHEMW

Β öojirapcKOM H3biKe

179

JANKO BACVAROV, Concerning the peculiarities of the Bul-

garian grammatical system from the point of view of functional sentence perspective ILPJKEMMGJI A/JAMEIJ,

AKTyajiBHoe HJieHeHHe, rjiyÔHHHwe

CTpyKTypti Η πβρΗφρ33Η

KAREL PALA, Semantic classes of verbs and FSP LUDMILA UHLÍftovÁ, On the role of statistics in the investi-

gafiön of FSP

185 189

196 208

et al., Zur Terminologie der funktionalen Satzperspektive 217

FRANTISEK DANES

LIST OF ABBREVIATIONS

AO

Archiv orientální

AUC

Acta Universitatis Carolinae

AUP

Acta Universitatis Palackianae

BSE

Brno Studies in English

CM F

Casopis pro moderni filologi!

CL

Ceská literatura

FL

Foundations of Language

IRAL

International Review of Applied Linguistics and Language Teaching

JL

Journal of Linguistics

KZ

Zeitschrift für vergleichende Sprachforschungen auf dem Gebiete der indogermanischen Sprachen, gegründet von A. Kuhn

LF

Listy filologické

Lg

Language

PBML

The Prague Bulletin of Mathematical Linguistics

PMLA

Publications of the Modern Language Association

PSML

Prague Studies in Mathematical Linguistics

PhP

Philologica Pragensia

RRL

Revue roumaine de linguistique

SaS Slovo a slovesnost SPFFBU Sbornik praci filosofické fakulty bmënské university

8

STZ StudGen TCLP TLP

Sprache im technischen Zeitalter Studium Generale Travaux du Cercle linguistique de Prague Travaux linguistiques de Prague

VJa ZRP ZS1

Voprosy jazykoznanija Zeitschrift für romanische Philologie Zeitschrift für Slawistik

FOREWORD

This volume represents the revised papers of the first international symposium on Functional Sentence Perspective* at Mariánské Làznë (Marienbad) in the autumn of 1970, organized by the Institute of Czech Language of the Czechoslovak Academy of Sciences. The purpose of the meeting was to discuss in a roundtable manner certain aspects of the communicative function of the sentence, together with questions of the organization of the text. It was evident (and the most recent developments of our science confirmed this fact very clearly) that these aspects of the study of language have not always been given due consideration in contemporary linguistics, in spite of the fact that the phenomena in question, referred to as, e.g., "point de départ" and "but du discours", "psychological subject" and "psychological predicate", "thème" and "propos", "theme" and "rheme", "topic" and "comment", "presupposition" and "focus" have been the concern of scholars at least since the first half of the nineteenth century. Any linguistic theory with a claim to adequacy should account for FSP phenomena in a consistent way within its framework. The symposium was supposed to deal with the following basic topics: 1. The FSP in the theory and practice of the Prague Lin* Referred to also as "aktuální ilenëni vëtné", "aktualnoje ölenenije predlozenija", "aktuelle Satzgliederung", "funktionale Satzperspektive", „Mitteilungsperspektive", "kontextové ilenëni vëty", "contextual segmentation of the sentence", "vyznamová vystavba vypovëdi", "the thematic organization of the utterance", "smyslovoje ilenenije predlozenija", "ThemaRhema-Gliederung", "topic-comment structure".

10

FOREWORD

guistic School, and of other currents of linguistic research. 2. The place of FSP in the system of linguistic description. 3. The FSP and the structure of the text. 4. The ways of expressing the FSP in various languages. 5. The methods and procedures of the examination of the FSP. The papers in the present volume are arranged according to these topics; and are supplemented by some of the contributions made by the participants during the discussions. In the concluding session participating linguists from ten European countries discussed the results of the symposium and agreed on the text of a Memorandum (that was afterward published in a number of linguistic journals in the world). The Memorandum states: "While admitting that there were different possible approaches to the problems under discussion it was conceded that a more unified and coordinated approach can only be achieved by pooling of interests and reappraisal of individual approaches in the light of the insights gained into other approaches in the course of discussion. This aim would certainly be promoted by an extensive bibliography**, and a glossary of the most frequent terms used by the FSP theorists." "Further investigation into FSP, duly employing also the methods of contrastive analysis, should be concerned with an explicit formulation of the FSP phenomena within a theory of grammar aiming at completeness, as well as with the relevance of FSP to various fields of applied linguistics (language teaching, stylistics, practical and theoretical problems of translation, automatic processing of linguistic data)." The editor of the present volume has nothing to add, but his wish that this publication of the papers should promote the further study of the phenomena of FSP in the same spirit of international collaboration that characterized the discussions of the Symposium.

** Cf. also Z. TYL (ed.), A tentative bibliography of studies in functional sentence perspective 1900—1970 (mimeo), Prague 1970.

SOME ASPECTS OF THE CZECHOSLOVAK APPROACH TO PROBLEMS OF FUNCTIONAL SENTENCE PERSPECTIVE Jan Firbas (Brno)

For some time Czechoslovak linguists have devoted considerable attention to problems of what has been called in Czech "aktuální clenëni vëtné" and has usually come to be termed in English "functional sentence perspective" (=FSP), "contextual sentence (utterance) organization", "theme-rheme" or "topic-comment structure". 1 It is Vilém Mathesius's (1882-1945) merit that in the period between the two wars these problems began to be intensively studied in Czechoslovakia. In the last quarter of the 19th century, they were dealt with by German scholars, for instance, by G. von der Gabelentz, H. Paul and Ph. Wegener. Most of them used the terms "psychological subject" and "psychological predicate", which according to Mathesius unduly tended to relegate the problems of FSP to spheres not treated by current linguistic research.2 A true pioneer work in the field of FSP, however, is the monograph De Vordre des mots dans les langues anciennes comparées aux langues modernes by Henri Weil, a French classical scholar. It was published as early as 1844.3 Two practically unchanged later editions followed in 1869 and 1879. An English translation, entitled The Order of Words in the ancient languages compared with 1 D A N E S , F., Prispëvek k novëjgi syntaktické terminologii [ A Contribution to an Inquiry into Recent Syntactic Terminology], Slavjanskaja lingvistiêna terminologija [Slavonic Linguistic Terminology], Sofìa 1963, 46—52. 2 MATHESIUS, V . , Cestina a obecny jazykozpyt [The Czech Language and General Linguistics], Prague 1947, 235. 3 Not 1 8 5 5 as stated by V . MATHESIUS: In Cestina... (quoted here in note 2 ), 234, and elsewhere.

12

J. FIRBAS

that of the modern languages, appeared in Boston in 1878. It is evidently chiefly the ideas of H. Weil that inspired Mathesius and were further developed by him. It will not be without interest to recall some of Weil's ideas here. Weil distinguishes between the movement of ideas and the syntactical movement. The former is expressed by the order of words, the latter by terminations (cf. p. 36 of the English translation). Weil endeavours to show that men think and express themselves in the same order whether they speak a modern langauage or use one of the ancient languages (37). Romance and Germanic languages represent the former, Greek and Latin represent the latter. (No mention is made of Slavonic languages.) A sentence contains a point of departure (an initial notion) and a goal of discourse. The point of departure is equally present to the speaker and to the hearer; it is their rallying point, the ground on which they meet (29). The goal of discourse presents the very information that is to be imparted to the hearer (30). Weil claims that the movement from the initial notion to the goal of discourse reveals the movement of the mind itself (43). Languages may use different syntactical constructions, but the order of ideas remains basically the same (35). Weil finds that modern languages tend to make the grammatical subject express the point of departure (37). The reverse order, which puts the goal of discourse first and the initial notion last, also occurs; Weil refers to it as the pathetic order and looks upon it as a vehicle of emotion (43). In his researches Mathesius has corroborated Weil's idea that a more rigid order of words increases the frequency of thematic subjects. Comparing English with Czech and also with German, he has established the marked English tendency to make the subject express the theme.4 He has also inquired into the constructions 4

See his Nëkolik poznámek o funkci podmëtu ν moderni angliétinè [Some Notes on the Function of the Subject in Modern English], Casopis pro moderni filologa 10,1924, 1 — 6; On Linguistic Characterology, Actes du Premier Congrès International de Linguistes à La Haye, Leiden 1928, 56—63, republished in Prague School Reader in Linguistics, compiled by J. VACHEK, Bloomington 1964, 59—67; Zur Satzperspektive im modernen Englisch,

THE CZECHOSLOVAK A P P R O A C H TO FSP

13

English may employ to make the subject-predicate order conform to the theme-rheme sequence. Thus apart from the passive construction with a subject directly affected by the action (The book is being read), he treats the passive construction with a subject indirectly affected {I have been given a piece of advice), the so-called possessive passive (Everywhere he had crowds hanging on his lips), or the so-called perceptive passive (/ found a certain boldness of temper growing in me), etc.5 One of Mathesius's most significant contributions to the theory of FSP is his inquiry into the role played by FSP in determining the order of words. In his view word order phenomena constitute a system characterized by a hierarchy of word order principles. The hierarchy is determined by the extent to, and the manner in, which the principles operate. 6 Within the Czech system of word order, the leading principle is that of FSP: the theme-transition-rheme sequence renders the word order non-emotive, unmarked, whereas the rheme-transition-theme sequence renders it emotive, marked. (By transition Mathesius understands elements that actually belong to the rheme, but occur at its periphery and in this way intermediate between theme and rheme.)7 It appears that Weil's observations on the ordinary and the pathetic order of words are fully applicable to Czech. Apart from the word-order principle of FSP other word order principles are in play. In fact, every word order is the outcome of an interplay of principles. In regard to Czech, however, the other principles, though by no means unimportant, merely play a secondary role. They are mainly the rhythmical principle, the grammatical principle and the principle of coherence of sentence elements. Mathesius's Archiv für das Studium der neueren Sprachen und Literaturen 84. Jahrg., 155. Band (der neueren Serie 55. Band), 1929, 202—210. 5 Cf. MATHESIUS, V., O pasívu ν moderni angliitinë [On the Passive Voice in Modem English], Sborník filologicky 5, 1915, 198-220. 6 See Centina... (quoted here in note 2 ) 327—352, and Ze srovnávacích studií slovoslednych [From Comparative Word Order Studies], CMF 28, 1942, 181-190, 302-307. 7 See his Cestina... (quoted here in note 2 ), 340.

14

J. FIRBAS

interpretation of the Czech system of word order is widely accepted by Bohemicists. As can be gathered from the works of J. Horecky, J. Mistrik, E. Pauliny and J. Ruzicka, Mathesius's interpretation is also applicable to Slovak. Before Mathesius advanced his conception of Czech word order, three other Czech scholars had proved to be aware of the relevance to word order of what may be termed FSP phenomena. These scholars were J. Zubaty,8 V. Erti 9 and F. Trávnícek.10 Each of them spoke of the psychological subject and/or the psychological predicate. In appreciating their approaches, 11 Mathesius points out that Zubaty did not offer any finished theory of word order, but realized the importance of FSP, though taking into account only the psychological subject and paying no attention to its counterpart, the psychological predicate. Erti offers a keen-sighted analysis of Czech word order, but falls short of detecting the leading role of FSP. Trávnícek's thorough and able analysis establishes two basic word order principles: the semantic and the rhythmical principles. Distributing the sentence elements according to their communicative importance, the semantic principle is evidently an FSP factor. It is, however, put on the same hierarchical level as the rhythmical principle. Mathesius holds that in communication the lexical and grammatical means of language are made to serve a special purpose imposed on them by the speaker at the moment of utterance, i.e. in the very act of communication. In accordance with the requirements of the 8

See his Die Wortfolge in den slavischen Sprachen von Dr. Erich Berneker, Lis Iy fiìologické 28, 1901, 129—134. 9 See his O postavení podmëtu po ölenech úvodních [On the Position of the Subject after Introductory Sentence Elements], Nase fee 1,1917,33 — 38, 7 5 - 7 9 , 1 0 9 - 1 1 4 , 1 3 6 - 1 4 0 , 1 7 2 - 1 7 7 , 2 0 0 - 2 1 0 ; a n d GEBAUER, J . - E R T L , V . ,

Mluvnice ceská II (Skladba) [A Grammar of Czech II (Syntax)],

1926 9 ,

55-63. 10 See his Základy òeskoslovenského slovosledu [The Foundations of Czechoslovak Word Order], SaS, 1927, 78—86; Slovosled pri durazu [Emphatic Word Order], SaS 5, 1939, 131 — 144; Mluvnice spisovné âestiny II [A Grammar of Standard Czech II], 1951 2 , 8 6 2 - 1 0 0 4 . 11 In Cestina... (quoted here in note 2 ), 327—352 passim.

THE CZECHOSLOVAK APPROACH TO FSP

15

context, the lexical units acquire specific meanings, and the sentence, which grammatically speaking consists of a subject and a predicate, splits up into a theme and a rheme. 12 The lexical and grammatical means have been made to function in a definite kind of perspective; they show a certain kind of contextual organization. It is due especially to F. Danes and M. Dokulil that these thoughts have been developed into what has come to be called the three-level approach to syntax. This approach distinguishes between three levels: the semantic level, the grammatical level and the level of FSP (or contextual organization). 13 In this connexion it is not without interest to recall Danes's observation that the forerunner of this approach is V. Erti, 14 who in 1926 distinguished between the grammatical, logical and psychological subject. Expressing, for instance, the bearer of a quality or the agent, Ertl's logical subject is a phenomenon of the semantic level. I hope to be able to adduce ample proof of the fruitfulness of the three-level approach. Regarding the semantic and grammatical sentence structure as a means that can function in different contexts and consequently display different perspectives, Mathesius distinguishes between the sentence as a pattern belonging to the language system and the sentence as part of the context, i.e. an utterance (a component of the discourse).15 This might suggest the conclusion that non12

See Cestina... (quoted here in note 2 ), 294. Cf. also DANES, F. Intonace a véta ve spisovné cestini [Sentence Intonation in Present-Day Czech], Prague 1957, 55. 13 Cf. DANES, F., op. cit., 56; DANES, F.,Vedlejsi vëty ùôinkovë prirovnávací se spojkou "nez aby" [Consecutively Coloured Comparative Subclauses with the Conjunction "nez aby"], Nase ree 37, 1957, 12—22; DANES, F. DOKULIL, M., Κ tzv. vyznamové a mluvnické stavbë vëty [On the so-called Semantic and Grammatical Sentence Structures], O vêdeckém ροζηάηί soudobych jazyku [On a Scientific Inquiry into Contemporary Languages], Prague 1958, 231 — 246; DANES, F., A Three-Level Approach to Syntax, TLP 1, 1964, 2 2 5 - 2 4 0 . 14 See DANES, F., Intonace... (quoted here in note 1 2 ), 56; GEBAUER, J. ERTL, V., Mluvnice... (quoted here in note 9 ), 5—6. 15 Cf. MATHESIUS, V., fteC a sloh [Speech and Style], ¿teñí o jazyce a poezii [Readings about Language and Poetry], Prague 1942, 6.

16

J. FIRBAS

utterance phenomena belong to the sphere of "langue", whereas utterance phenomena belong entirely to that of "parole". But Danes is right in pointing out that even utterance phenomena are subject to generalization.16 They display patterns that have become the object of study of the theory of FSP. In accordance with the three level approach it is therefore possible to distinguish between the Semantic Sentence Pattern (=SSP), 1 7 the Grammatical Sentence Pattern (=GSP) 1 7 and the Utterance or FSP or Communicative Sentence Pattern (=CSP). It would certainly be possible to imagine a context in which the semantic and grammatical structure John wrote a poem would function as an utterance event following the Agent-Action-Goal SSP, the Subject-Verb-Object GSP and the Theme-Transition-Rheme CSP. It follows that the resultant complex sentence pattern constituted by the semantic sentence pattern, the grammatical sentence pattern and the communicative sentence pattern is part of "langue". This view would evidently be subscribed to also by P. Sgall18 and others. As has already been pointed out, in a number of papers Mathesius has endeavoured to establish the means English has recourse to in order to reconcile the requirements of FSP with those of grammatical structure. He regards the existence of such means as proof of the powerful formative character of FSP. In regard to German, 16

Cf. DANES, F., Syntakticky model a syntakticky vzorec [The Syntactic Model and the Syntactic Pattern], Ceskoslovenské prednásky pro V. mezinárodní sjezd slavistû υ Sofii [Czechoslovak Papers Prepared for the Fifth International Congress of Slavists in Sofia], Prague 1963, 116—117. 17 See DANES, F., Some Thoughts on the Semantic Structure of the Sentence, Lingua 21, 1968, 5 5 - 6 9 . 18 Cf. his Functional Sentence Perspective in a Generative Description, PSML 2, Prague 1967, p. 206. Cf. also BENESOVÁ, Ε., O sémantickém charakteru Ceského slovosledu [On the Semantic Character of Czech Word Order], SaS 29, 1968, 34. A frequency count of CSPs based on Czech scientific prose has been offered by UHIÄOVÄ, L. in Part Four of her unpublished ScC. dissertation Kvantitatwni rozbor vëty a vypovëdi υ éestinê [A Quantitative Analysis of the Sentence and the Utteracne in Czech]. [For a summarizing account of the analysis, see her On the Quantitative Analysis of Clause and Utterance in Czech, PSML 4, 1972, 107—128.]

THE CZECHOSLOVAK APPROACH TO FSP

17

similar researches have been carried out by E. B e n e s . 1 9 In his last paper o n English w o r d order, 2 0 however, Mathesius comes to the conclusion that there are not e n o u g h means to put all the n o n emotive English sentences in the theme-transition-rheme

order

(op. cit., p. 187). In determining the hierarchy of English word order principles, he allots the leading role t o the grammatical principle, ranking F S P only with factors o f secondary importance. H e even goes t o the length o f concluding that "English differs from Czech in being so little susceptible to the requirements of F S P as to frequently disregard them altogether". The mentioned wording has proved to be a real incentive to an inquiry into the position o f F S P within the structure of English. The inquiry has s h o w n that besides word order and intonation, semantic structure operates as a n effective means of FSP. This applies not only t o English, 2 1 but to other languages as well, f o r instance, C z e c h , 2 2 R u s s i a n , 2 3 G e r m a n , 2 4 Spanish, 2 5 Albanian; 2 6 19 Cf. his Die Verbstellung im Deutschen, von der Mitteilungsperspektive her betrachtet,/Ά.Ρ 5, 1962, 6—19; Die Ausklammerung im Deutschen als grammatische Norm und als stilistischer Effekt, Muttersprache 78, 1968, 289—302; Die funktionale Satzperspektive im Deutschen im Vergleich mit dem Tschechischen, Deutsch-tschechische Beziehungen im Bereich der Sprache und Kultur, Aufsätze und Studien II, Abhandlungen der Sächsischen Akademie der Wissenschaften, Berlin 1968, Philologisch-historische Klasse, Band 59, Heft 2, 5 7 - 6 9 . 20 MATHESIUS, V., Ze srovnávacích studií slovoslednych [From Comparative Word Order Studies], CMF2S, 1942,181-190, 302-307. Cf. FIRBAS, J., From Comparative Word Order Studies, BSE 4, 1964, 111—128. 21 Cf., e.g., FIRBAS, J., Κ otázce nezákladovych podmëtû ν souCasné angliôtinë [On the Problem of Non-Thematic Subjects in Contemporary English], ¿ M F 39,1957,22—42, 165—173; Non-Thematic Subjects in Contemporary English, TLP 2, 1966, 239—256. 22 Cf.,e.g.,DANES, F., Prispëvek k rozboru vyznamové vystavby vypovëdi [A Contribution towards the Analysis of the Contextual Organization of Utterance ( = Functional Sentence Perspective)], Studie a präce lingvistické I [Papers and Monographs in Linguistics I], Prague 1954, 263—274. 23 Cf., e.g., ADAMEC, P., Κ úloze sémantiky ve slovosledu [On the Function of Semantic Structure in Word Order], A UC, Slavica Pragensia 4, Prague 1962, 297—300; the same, Porjadok slov ν sovremennom russkom jazyke [Contemporary Russian Word Order], Prague 1966.

18

J. FIRBAS

semantic structure operating as a means of FSP seems to be at least an Indo-European universal. Let me adduce an example in illustration. In Mathesius's terms the non-emotive structure A girl came into the room would have to be looked upon as insusceptible to FSP, because in its most natural use it does not display the themetransition-rheme, but the rheme-transition-theme sequence. The semantic structure of the example is the following. The verb came expresses the notion of appearance on the scene, the adverbial into the room expresses the scene and the subject a girl a newcomer appearing on it. Under the circumstances, i.e. in the case of the most natural use of the structure, the adverbial element carries known information (derivable,27 recoverable 27 from the preceding context) and is in consequence contextually dependent. The offered semantic and contextual interpretation equally applies to the less frequent, but not impossible order Into the room came a girl, which would most naturally be interpreted as a theme-transitionrheme sequence. In either case, being contextually dependent, the adverbial element contributes least to the further development of the communication. Of the two remaining elements, the contextually independent subject announcing the person appearing on the scene is communicatively more important than the contextually independent verb merely expressing the notion of appearance. The contextual independence of the subject is sufficiently signalled by 24 Cf., e.g.,BENES, E., Die Ausklammerung... (quoted here in note 1 9 ); FIRBAS, J., Thoughts on the Communicative Function of the Verb in English, German and Czech, BSE 1, 1959, 39—68. 25 Cf., e.g., DUBSKÍ, J., L'inversion en espagnol, SPFFBU A 8, 1960, 111 —115; FIRBAS, J., Notes on the Function of the Sentence in the Act of Communication, ibid, A 10, 1962, 133—148. 26 NOVÁK, P., Κ zdvojování predmëtu Ν albânStinë [The Doubling of the Object in Albanian], Sbornik slavistickych praci vënovanych IV. mezinárodnimu sjezdu slavisti ν Moskvë [A Collection of Papers Prepared for the Fourth International Congress of Slavists in Moscow], Universitas Carolina, Prague 1958, 2 7 - 3 2 . 27 Terms coined by Μ. A. K. HALLIDAY; see, e.g., his Notes on Transitivity and Theme in English II, JL 3, 1967, Part 4. 3.

THE CZECHOSLOVAK APPROACH TO FSP

19

the non-generic indefinite article. In terms of degrees of communicative dynamism (=CD), the subject carries the highest, the adverbial element the lowest degree of CD, the verb ranking between them. By a degree of CD carried by a linguistic element, I mean the extent to which the element contributes towards the development of the communication. 28 The comments have indicated that irrespective of sentence position, context and semantic structure are capable of signalling degrees of CD. I have attempted a more detailed discussion of the important roles played by the verbs of existence and appearance on the scene and of the articles in FSP in another place. 21 The importance of these means of FSP has also been duly appreciated by P. Adamec 23 and P. Novák. 26 ' 29 A. G. Hatcher's work should not be left unmentioned either in this connexion. 30 Let me adduce some further examples of the operation of semantic structure on the level of FSP and let me do so again in terms of CD. I believe I am right in claiming that it follows both from Adamec's 23 and my own researches24·'31 that a contextually independent object will carry a higher degree of CD than the verb. A verb cannot carry the highest degree of CD within a sentence when preceded or followed by such an object. If the semantic agentaction-goal pattern expressed by means of the grammatical subjectverb-object pattern is contextually independent in its entirety or 28 For a more detailed discussion of the concept of CD, see my On the Concept of Communicative Dynamism in the Theory of Functional Sentence Perspective, summarizing my previous observations; mimeographed for the Seminar on the Construction of Complex Grammars, Cambridge, Mass., June 1970 (held by Thomas J. Watson Research Center, Yorktown Heights, New York 10598) [and published in SPFFBU A 19, 1971, 135-144], 29 Novák, P., O prostfedcích aktuálního ëlenëni [On the Means of Functional Sentence Perspective], AUC 1959, Philologica I, 9—15. 30 Syntax and the Sentence, Word 12, 1956, 234—250; Theme and Underlying Question, Two Studies in Spanish Word Order, Supplement to Word 12, 1956. 31 See also Firbas, J., On the Prosodie Features of the Modern English Finite Verb-Object Combination as Means of Functional Sentence Perspective, BSE 8, 4 9 - 59.

20

J. FIRBAS

contextually dependent merely through its agent-subject element (A I The girl broke a vase), the following interpretation applies. The verb will carry a lower degree of CD than the object, but a higher degree of CD than the subject. This is because a known or unknown agent appears to be communicatively less important than an unknown action and its unknown effect or result. I will come back to this interpretation in the last section of my paper. It holds good for all the subtypes of the described complex sentence pattern as they have been discussed by F. Danes. 17 With due alterations, what has been said about the contextually independent object applies to a contextually independent adverbial element of place expressing the direction or goal of a motion (Ich bin gestern nach Prag gefahren). Both the object and the adverbial element express an absolutely essential semantic amplification of the verb. Provided they are contextually independent, the former will most naturally carry a higher degree than the verb. 32 It is especially F. Kopecny's merit to have drawn attention to the syntactic importance of such amplifications.33 Their important role in word order has recently been stressed again by L. Uhlírová. 34 Special mention should be made of elements that may be referred to as rhematizers. For instance, Cz. i, Engl, even, G. sogar, R. daze would come under this heading. They render the element they accompany rhematic. On the other hand, there are elements that because of their semantic content most naturally function in the theme. Personal pronouns are the most typical representatives of this group. This sketchy note has only superficially touched upon a sphere of problems discussed especially by F. Danes, 22 J. Mistrik 35 and most recently by B. Palek, whose monograph on 32

Cf. FIRBAS, J., Thoughts... (quoted here in note 2 4 ), 49. KOPECNY, F., Základy ceské skladby [Foundations of Czech Syntax], Prague 1958, 2 9 - 3 4 . 34 UHLÍftovÁ, L., Vztah syntaktické funkce vëtného òlenu a jeho mista ve vëtë [Relationship between Syntactic Function and Linear Position of Sentence Elements], SaS 30, 1969, 365. 35 See his Κ realizácii aktualneho Clenenia [On the Implementation of Functional Sentence Perspective], Slovenská rei 24, 1959, 193—212; S Ιουo33

T H E CZECHOSLOVAK A P P R O A C H TO F S P

21

Cross-Reference 36 — though not explicitly referring to the theory of FSP - is highly relevant to it. Not all types of semantic content, however, are capable of signalling degrees of CD in the way indicated above. This is borne out by the following examples, each of which contains a contextually independent indirect and a contextually independent direct object: He gave a boy an apple. He gave an apple to a boy. Of the two objects, the one occurring later evidently carries a higher degree of CD. Similarly, a contextually independent infinitive of purpose will carry a lower degree of CD when occurring initially than when occurring finally: In order to meet his friend, he went to Prague. He went to Prague in order to meet his friend. In all these cases it is the linear arrangement that decides the degree of CD. 3 7 I believe it is not an exaggeration to say that the recognition of the important role played by the semantic structure in determining FSP and word order is one of the most essential features of the post-war development of Mathesius's ideas in Czechoslovakia. Much of course still has to be done. It remains to be seen whether K. Pala's 3 8 plausible suggestion that the number of relevant semantic patterns may be final proves to be correct. As I see it, the distribution of the degrees of CD over the sentence elements, which makes the semantic and grammatical structure function in a definite kind of perspective, is an outcome of a tension between, or rather interplay 39 of, the tendency towards the basic sled a vetosled ν sloveníine [Word Order and Clause Order in Slovak], Bratislava 1966. 36 Cross-Reference, A study from Hyper-Syntax, AUC, Prague 1968, Philologica, Monographie 21. 37 GOLKOVÁ, E., On the English Infinitive of Purpose in Functional Sentence Perspective, BSE 7, Brno 1968, 119—128. 38 See his Otnoäenije mezdu porjadkom slov i aktual'nym ölenenijem ν öeäskom jazyke [The Relation between Word Order and Functional Sentence Perspective in Czech], PSML 2, 1967, 58. Cf. also BENESOVÁ, E., O sémantickém charaketru ¿eského slovosledu [On the Semantic Character of Czech Word Order], SaS 29, 1968, 3 4 - 4 1 . 39 Cf., e.g., On the Interplay of Prosodie and Non-Prosodic Means of Functional Sentence Perspective, to be published in The Prague School of

22

J. FIRBAS

distribution of CD on the one hand, and the context and the semantic structure on the other. The basic distribution of CD is implemented by a series of elements opening with the element carrying the very lowest and gradually passing on to the element carrying the very highest degree of CD. Context and semantic structure operate either in the same direction as or counter to the basic distribution of CD. The basic distribution of CD would reflect what H. Weil has called the "movement of the mind". Insusceptibility to FSP displayed by word order is then no unmistakable proof of the insusceptibility to FSP of the entire language system. In my opinion, insusceptibility to FSP could be spoken of when in the very act of communication, written or spoken, a structure permits of more than one interpretation of its functional perspective. Such cases of multifunctionality or multivalence will certainly be far more frequent in written than in spoken language. It must, of course, be borne in mind that at the present stage of research it may be the investigator's insufficient knowledge of the interplay of means that induces him to interpret a structure as multifunctional. 40 On the other hand, as J. Vachek 41 and other scholars of the Prague group have pointed out, language is not a closed and perfectly balanced system. Multifunctionality on the level of FSP would have to be accepted as a peripheral phenomenon of this system.42 The concept of the interplay of means of FSP accommodates various types of word order hierarchies. In the light of the researches into the interplay of means of FSP, Mathesius's concluLinguistics and Language Teaching (vol. 27 of the Oxford University Press series Language and Language Learning), London. [The volume appeared in 1972; for the paper referred to, see 77—94.] 40 On multifunctionality in FSP, see FIRBAS, J., Non-Thematic Subjects (quoted here in note 21 ), 249—251. 41 See, e.g., his Some Notes on the Development of Language Seen as a System of Systems, Proceedings of the Eighth International Congress of Linguists, Oslo University Press 1958, 418—419. 42 Cf. VACHEK, J., On Peripheral Phonemes of Modern English, BSE 4, Brno 1964, 7—110; DANES, F., The Relation of Centre and Periphery as a Language Universal, TLP2, 9—21.

THE CZECHOSLOVAK APPROACH TO FSP

23

sion to regard the grammatical principle as the leading one within the English system of word order is undoubtedly correct. He is, however, not right in making the FSP principle responsible for emotive (marked) word orders. It has been shown that the common feature of such word orders in English is not their deviation from the basic distribution of CD, but from the grammatical sequence SVOMPT (These men we trust that we know to prize).43 On the other hand, researches into the interplay of means of FSP fully corroborate Mathesius's view allotting the rhythmical principle a merely secondary status. Because of their semantic weakness, rhythmically light words most naturally function in the theme. They are mainly responsible for the practically clashless co-operation of the rhythmical factor with the other word order principles. Entering into groupings of rhythmically light and heavy words, they appear in different positions, but remain thematic and entail no change in the relationship of degrees of CD. In this way they enable the rhythmical factor to assert itself within the limits set by the requirements of other word order principles.44 For certain purposes it would certainly be sufficient to establish just the mutual relations of elements in regard to the degrees of CD carried by them. It has, however, proved to be expedient to distinguish between theme and rheme, or between theme, transition, and rheme. These concepts require some comment. F. Danes 45 has pointed out that Mathesius in fact offers two conceptions of the theme. Roughly speaking, according to one of them the theme expresses something that is spoken about; according to the other it expresses something that is known or at least obvious in the given situation. Mathesius has not elaborated these two conceptions. But the two aspects presented by them need not coincide. Thus in the sentence An unknown man has asked him the 43

See FIRBAS, J., From Comparative Word Order Studies (quoted here in note 20 ), 117-120. 44 Cf. also ibid., 121. 45 Cf. his Intonace... (quoted here in note 12 ), 56; and his Téma || základ II vychodisko vypovëdi [Theme || Basis || Initial Point of Utterance], SaS 25, 1964. 148-149.

24

J. FIRBAS

way to the railway station, An unknown man expresses the person talked about, but conveys neither known nor obvious information. F. Trávnííek 46 is right in refusing to regard the criterion of known or unknown information as accommodating the most general feature of the theme. According to this criterion, some types of sentences opening a message (A girl broke a vase) would have to be interpreted as themeless. Trávníéek, however, accounts for the theme on psychological grounds and links it up invariably with the beginning of the sentence. This makes it impossible to appreciate the interplay of means of FSP, which does not permit of invariably linking up a sentence position with one definite degree of CD. As I see it, within the theory of FSP the theme is best defined as constituted by an element or elements carrying the lowest degree(s) of CD within a sentence. According to this definition both An unknown man and him would be thematic, him functioning as theme proper because carrying the very lowest degree of CD. It must, of course, be remembered that inquiries into the ways sentences open in various languages may reveal considerable differences. This view is borne out by the researches of E. Benes and M. A. K. Halliday. Benes distinguishes between a basis and a theme. While applying the term "theme" to the phenomenon defined by me here as the element(s) carrying the lowest degree(s) of CD, he understands by the "basis" the phenomenon that "as the opening element of the sentence links up the utterance with the context and the situation, selecting from several possible connexions one that becomes the starting point, from which the entire further utterance unfolds and in regard to which it is orientated". 47 46 See his O tak zvaném aktuálním dlenëni vëtném [On the So-Called Functional Sentence Perspective], SaS 22,1961,163 — 171. For a discussion of V. Mathesius's and F. TrávníCek's views of the "theme", see FIRBAS, J., On Defining the Theme in Functional Sentence Analysis, TLP 1, 1964, 267 to

280. 47

BENES,E.,Zaíátek nëmecké vëty Ζ hlediska aktuálního ilenëni vëtného [The Beginning of the German Sentence from the Point of View of Functional Sentence Perspective], C M P 41, 1959, 216.

THE CZECHOSLOVAK APPROACH TO FSP

25

Halliday's conception of theme comes near to Benes's basis, but differs from it in one important point. Halliday totally abstracts from the preceding context. 48 The problem of the theme is naturally related to that of the segmentation of the sentence on the level of FSP. Is there a bipartition of theme and rheme, or a tripartition of theme, transition and rheme? It is possible to proceed even further. Assuming that any linguistic element (i.e. even a morpheme or a sub-morphemic phenomenon) expressing some meaning is capable of carrying a degree of CD, we arrive at a gamut of degrees of CD constituted by theme proper, rest of rheme, transition proper, rest of transition, rheme to the exclusion of rheme proper, rheme proper. Needless to say, not all these degrees must necessarily be present in a sentence. The delicacy of segmentation depends on the purpose of the investigation. A special problem, however, seems to be presented by the concept of transition. Some scholars (e.g., V. Mathesius, F. Danes, J. Firbas, E. Golková, F. Kopecny, P. Mistrík, E. Pauliny, P. Sgall, A. Svoboda, L. Uhlírová 18 ) employ it, others (especially P. Adamec) do not. I believe that the concept of transition can be used to good purpose. The scope of the present paper does not permit repeating the detailed argument developed elsewhere in favour of this view.49 Let me therefore at least state that it leads to the following interpretation of the function of the temporal and modal exponents of the finite verb on the level of FSP. In their non-marked use, they mediate between the thematic and the nonthematic section of the sentence, carrying the lowest degree of CD 48

See HALLIDAY, M. A. K., Notes on Transitivity... (quoted here in note 2 7 ), 212. 49 See FIRBAS, J., A Note on Transition Proper in Functional Sentence Analysis, PhP 8, 1965, 170—176; the same, On the Prosodie Features of the Modern English Finite Verb as Means of Functional Sentence Perspective (More Thoughts on Transition Proper), BSE7, 1968,11—48. The conclusions arrived at in these papers have recently been corroborated by L. UHLÍftovÁ's quantitative analysis of Czech communicative sentence patterns (see here note 1 8 ).

26

J. FIRBAS

within the latter and constituting what might be termed transition proper (He [th] has [tr pr] fall- [tr] en [tr pr] ill [rh]).50 In their marked use, i.e. when occurring in a sentence one element of which has been singled out for sharp, ad hoc contrast, they are either part of an extensive theme proper (HE [rh pr] has fallen ill [tr pr]) or constitute the rheme proper (He [th pr] HAS [rh pr] fallen ill [th pr]). If this interpretation is correct, the temporal and modal exponents of the finite verb would constitute a boundary between the thematic and the non-thematic section51 of sentences not occurring in sharp, ad hoc contrast on account of one of their elements. Such a boundary is no doubt desirable in an inquiry into the structure of the thematic layer within a paragraph, constituted by all the thematic elements occurring within the sentences of this subdivision of discourse. Mathesius has drawn attention to the fact that in the course of the communication English tends to change the theme less often than, for instance, Czech does. The structure of the thematic layer is an important problem that has recently been taken up by F. Danes, 52 who speaks of thematic progression. On other occasions, the concept of transition makes it possible to distinguish between the rhematic and the non-rhematic section. Transition may be delimited from rheme by means of a conspicuous rise in CD, which brings the rheme into distinct relief. (There [th] was [tr] a cold bitter taste [rh] in the air [th].) This does not only effect a clear delimitation of the transition itself, but also a clear, distinct delimitation between the non-rhematic (the thematic plus 50 The abbreviations "th", "tr", "rh", "pr" respectively stand for "theme", "transition", "rheme", "proper". 51 "Section" does not necessarily denote here an uninterrupted sequence of elements, but all elements of a type within a sentence, irrespective of their positions. Thus the "thematic section" is constituted by all the thematic elements within the sentence no matter where they occur. The term "boundary" is to be interpreted accordingly. 52 See his Typy tematickych posloupnosti ν textu (na materiále òeského textu odborného) [Types of "Thematic Progression" in Text (Based on an analysis of Czech technical texts)], SaS 29,1968,125—141. Cf. also his paper in the present volume.

THE CZECHOSLOVAK APPROACH TO FSP

27

the transitional) and the rhematic section. It has been shown 53 that the degree of distinctness is here in direct proportion to that of the shift towards non-verbal expression (in other words, towards an all-round weakening of the communicative value of the finite verb). It appears that the rhematic layer of an English paragraph tends to contain fewer verbal elements than the rhematic layer of a Czech paragraph. 54 Speaking of the distinctness of the boundaries delimiting the transition, I have in mind especially the boundary between transition and rheme. It is at this point that the delimitation may become less distinct or perhaps even difficult to determine. This, however, may not be at variance with the basic character of the transitional elements. Is it not their very function to be transitional? In connexion with the question of the delicacy of segmentation, mention should be made of the hierarchy of segmentable sections. A. Svoboda 55 has elaborated the idea that sections, or fields, within which the distribution of CD can take place are provided by explicit or hidden predication. An example of a distributional or communicative 56 field of the former type would be a complex sentence, a simple sentence or a subordinate clause, an example of the latter type an attributive construction. A communicative field of lower rank, for instance, that of a subordinate clause, functions as a communicative unit within a field of higher rank. It follows that besides the functional perspective of a sentence, simple or complex, there are also the functional perspectives of the clause and the attributive construction. There are, of course, distributional fields of higher rank than that provided by the sentence, for 53 In F i r b a s , J., On the Communicative Value of the Modern English Finite Verb, BSE 3, 79—104. The paper also discusses the notions of conspicuous rise in C D and distinct relief. Cf. also H l a d k t ? , J., A n Attempt at a Quantitative Expression of the Communicative Value of the Verb in English and Czech, BSE 7, Brno 1968, 103—118. 5 4

55

S e e F i r b a s , J., o p . cit. 95.

See his The Hierarchy of Communicative Units and Fields as Illustrated by English Attributive Constructions, BSE 7, 1968, 49—101. 56 A. Svoboda's term.

28

J. FIRBAS

instance, those of the paragraph, the chapter, etc. Consequently it is possible to speak also of their functional perspectives. The functional prespective of Czech or Slovak complex sentences has been the object of special studies produced by F. Danes, 5 7 J. Mistrik 58 and S. 2aza. 5 9 Speaking of the hierarchy of segmentable sections, mention should be made of A. Svoboda's 55 communicative units, i.e. elements making up the communicative (distributional) fields. A special problem is presented by the range constituted by all the possible perspectives in which semantic and grammatical structures can function. This range unmistakably shows two ends. One is characterized by total contextual independence, the other by the greatest possible contextual dependence. If occurring at the opening of a narrative or forming a single, isolated message, the structure A girl broke a vase functions at the former end. The same structure occurs at the latter end if appearing in sharp, ad hoc contrast on account of a semantic item conveyed by one of its elements, e.g., A girl BROKE a vase (she did not paint it). Cases of the latter type have been referred to by D. L. Bolinger 60 as belonging to second instance. Let me add that on second instance level any element within a sentence may come to express rheme proper, and any possible distributional subfield become eliminated (e.g. I didn't tell YOU that a girl broke a vase, with you functioning as rheme proper 57 See his Vedlejsi vëty úCinkové prirovnávací... (quoted here in note 1 3 ) and Konfrontaüní souvëti se spojkami "jestlize", "zatímco", "aby", "kdyz" [Complex Sentences of Confrontation Introduced by "jestlize", "zatímco", "aby", "kdyz"], Nase ree 46, 1963, 113 — 130. 58 See his Slovosled a vetosled... (quoted here in note 3 5 ). 59 See his Κ problematice vëtosledu a aktuálního élenëni podradného souvëti ν ruStinë [On Problems of Clause Order and the Functional Perspective of the Russian Complex Sentence], Ceskoslovenské prednáSky pro VI. mezinârodni sjezd slavistû [Czechoslovak Papers Prepared for the Sixth International Congress of Slavists], Prague 1968, 175—179. 60 In his Linear Modification, PMLA 1952,1117-1144. Reprinted in his Forms of English, Cambridge, Mass.—Tokyo 1965, 279—307. See also FIRBAS, J., On the Prosodie Features... (quoted here in note 4 9 ), esp. 12—18 and 36.

THE CZECHOSLOVAK APPROACH TO FSP

29

and the rest as an extensive theme proper). It is important to note that second instance cases still convey some new information. Though showing an unusually high degree of contextual dependence, they never become contextually dependent in their entirety. Cases of total contextual independence may be interpreted as occurring on the basic instance level within the sphere of first instance. A final classification of the intermediary instance levels ranging between the two ends will have to await further inquiries, which should not disregard the highly relevant researches of A. G. Hatcher 30 and P. Adamec. 61 I have also attempted to throw some light on these problems. 62 Last but not least, E. Benes's approach should be mentioned in this connexion. 63 Not every structure can operate on all instance levels. The range of perspectives in which a particular structure can function may be denoted as the contextual applicability of the structure. Not only clausal structures, but even non-clausal structures, including words and morphemes, should be studied in regard to their contextual applicability. L. Uhlírová is right in emphasizing that efficient generative rules should be capable of coping with the contextual applicability of a sentence structure. 64 Earlier in this paper I touched upon the question of congruence65 61

See his Porjadok slov... (quoted here in note 2 3 ), esp. 26—30. See Thoughts... (quoted here in note 2 4 ), esp. 51—53; and Notes... (quoted here in note 2 5 ) , esp. 141—143. 63 See his Two Aspects of Functional Sentence Perspective, TLP 3, 1968, 267— 274. He distinguishes between two types of context: left-hand and right-hand (whereas the above observations of mine have been made solely in regard to left-hand context). As I see it, Beneä's contextually independent sentences — viewed by him as free utterances affected neither by context (left-hand or right-hand) nor by situation — and his contextually semidependent sentences that occur without any left-hand context (functioning as introductory sentences of discourses) virtually correspond to the basic instance level cases as described above. 64 See her Some Aspects of Word Order in Categorial and Transformational Grammars, PSML 2, 1966, 164. 65 The problem of congruence (coincidence) has also been raised by TRNKA, B. in On the Linguistic Sign and the Multilingual Organization of Language, TLP 1, Prague 1964, 38. 62

30

J. FIRBAS

between the semantic, the grammatical and the FSP level. It has perhaps now become evident that the degree of congruence must necessarily be determined in regard to the instance levels. As has already been pointed out, a case of high degree of congruence is revealed by the temporal and modal exponents of the finite verb. On first instance levels they function as transition proper; within second instance they either constitute rheme proper (They [th pf] HA VE [rh pr] already been to London [th pr]) or become part of an extensive theme proper ( T H E Y [rh pr] have already been to London [th pr]). The question of congruence also arises in regard to prosodie and non-prosodic means of FSP. F. Danes's extensive and important researches into Czech intonation and into problems of intonation in general 66 offer ample proof of the view that the rheme is signalled by the intonation centre. P. Adamec's 23 examination of the relation of Russian word order to intonation substantiates this view. A disapproval of its general validity, however, has come from H. Krízková. 67 In interpreting pronominal (special) and verbal (yes/no) Czech questions, she regards the interrogative word (pronoun or adverb) and the finite verb of the pronominal and the verbal question respectively as rhematic. In either case she interprets the rest of the interrogative sentence as thematic. This is because in the respective question types the interrogative word and the finite verb point to the rhemes to be disclosed in the replies. They represent the only elements that are unknown to the speaker (inquirer). In offering this interpretation, Krízková in fact follows Mathesius, 68 her approach to the problem being essentially the 66

Apart from his Intonaca... (quoted here in note 1 2 ), Prispëvek ... (quoted here in note 2 2 ), see also his Sentence Intonation from a Functional Point of View, Word 16, 34—54, and his Order of Elements and Sentence Intonation, To Honor Roman Jakobson, The Hague—Paris 1967, 500—512. 67 See herTázací vëta a nëkteré problémy tzv. aktuálního (kontextového) ¿lenëni [The Interrogative Sentence and Some Problems of the so-called Functional Sentence Perspective (Contextual Organization of the Sentence)], Naie reS 4, 1968, 2 0 0 - 2 1 0 . 68 Cf. his Cestina... (quoted here in note 2 ), 3 3 6 - 3 3 7 .

THE CZECHOSLOVAK APPROACH TO FSP

31

same as his. As to the fact that the intonation centre does not normally occur on the interrogative word (interpreted by her as rhematic), but on some non-rhematic element within the question (interpreted by her as thematic), Krízková accounts for it on rhythmical grounds. F. Danes 69 has doubted the full validity of Mathesius's approach. Further developing his observations,70 I have come to the conclusion that not only the speaker's point of view, but also that of the listener (the prospective informant) must be taken into account. Viewed in this light, the question performs a double function: (i) it indicates the want of knowledge on the part of the inquirer and appeals to the informant to satisfy this want; (ii) it imparts knowledge to the informant in that it informs him what the inquirer is interested in (what is on his mind) and from what particular angle the intimated want of knowledge is to be satisfied. It is the second function that determines the rheme proper of the question. In accordance with the requirements of the context, any element within the question may become rheme proper. The offered solution establishes a coincidence between the bearer of the intonation centre (the most important prosodie feature) and the bearer of rheme proper. It is fully borne out by M. A. K. Halliday's recent interpretation of the function of the question. 71 It should 69

DANES, F., Intonace otázky [The Intonation of Interrogative Sentences], Nase fee 33, 1949, 6 2 - 68. 70 In Some Thoughts on the Function of Word Order in Old English and Modern English, SPFFBU A 5, 1957, 91 — 92. See also my Note on the Intonation of Questions from the Point of View of the Theory of Functional Sentence Perspective, to be published in the proceedings of the Symposium on Intonology, Prague 1970. [The proceedings constitute Acta Universitatis Carolinae, Philologica 1, 1972 (Phonetica Pragensia III); the relevant pages are 91 — 94.] The Paper is based on my On the Function of the Question in the Act of Communication, mimeographed for the Seminar on the Construction of Complex Grammars, Cambridge, Mass., June 1970 (quoted here in note 2 8 ). [For a Russian version, Funkciji voprosa ν processe kommunikaciji, see VJa 1972, 2, 55—65; the original English version is to appear in BSE 12.] 71

Cf. his N o t e s . . . (quoted here in note 4 8 ).

32

J. FIRBAS

be added that also in M. Grepl's 72 and J. Mistrik's 73 views the bearer of the intonation centre of a question is identical with the bearer of its rheme proper. Inquiries into the degree of congruence of non-prosodic and prosodie means are of primary importance. A special aspect of such inquiries is the question of congruence of the gamut of degrees of CD and the gamut of functional weight carried by the prosodie features. 74 To a considerable extent, these inquiries may be looked upon as a touchstone of interpretations of degrees of CD offered on the non-prosodic level. So far, in regard to English, they have borne out the non-prosodic interpretations of the functions performed in FSP by the temporal and modal exponents of the finite verb 49 and by the verb-object phrase. 31 As FSP plays a decisive role in controlling the intensity of prosodie features, and above all in determining the location of the intonation centre, the relation of grammar to intonation cannot be established without due regard to this important phenomenon. One of my primary concerns has been to show that FSP is an outcome of an interplay of means. The fact that susceptibility to FSP is by no means dependent on word order has far reaching consequences. It means that the relations in regard to CD between the elements of a distributional field may remain the same in spite of differences or changes in word order. Synchronically speaking, the laws of the interplay of means of FSP are flexible enough to make room for differences between separate languages (English, German, Russian and Czech, for instance) or for differences between different standards or substandards of one language, or for stylistic differences within one of such standards or sub-standards. Diachronically speaking, the laws of the interplay are flexible enough to make room for changes in word order in the course of 72

See his Emocionâlnë motivované aktualizace ν syntaktické strukture vypovëdi [Emotionally Motivated Actualizations in the Syntactic Structure of Utterance], Opera Universitatis Purkynianae Brunensis, Facultas Philosophies 113, Brno 1967, 41. 73 See his Slovosled... (quoted here in note 3 5 ), 97—98. 74 See On the Prosodie Features... (quoted here in note 4 6 ), 21.

THE CZECHOSLOVAK APPROACH TO FSP

33

historical development. 7 5 Thanks to the adaptability of means of FSP, demands o n the system of word order raised by other systems of language in the course o f historical development may be complied with without impairing the purpose o f the communication. This is quite in keeping with J. Vachek's conception of language as a system of systems. 4 1 Quite new problems have arisen owing to the recent interest in FSP by scholars aiming at a generative description of language. One problem deserves particular attention in this connection: that o f the basic word order from which all the other orders may be derived. P. Sgall is inclined to regard the basic distribution of C D as a suitable point o f departure. 7 6 Some other authors, o n the other hand, endeavour to find a basic order established quite independently of any context. Let me offer some comment o n this highly important problem. One o f the authors 7 7 speaks of two systems of 75

Non-standard Czech word order has been examined, e.g., by MATHESIUS, V. (K porádku slov ν hovorové ¿eátiné [On Word Order in Colloquial Czech], Nase ree 14, 1930, 117—121), CHLOUPEK, J. (Porádek slov ν náreóí, zvl. vychodomoravském [On Dialectal Word Order, with special regard to that of the East Moravian Dialect], SaS 19, 1958, 260 to 265); word order in modern Czech poetry has been investigated by JELÍNEK, M. (Slovosled ν dneäni ¿eské poezii [Word Order in Present-Day Czech Poetry], Theorie verse I [The Theory of Verse I], Brno 1966, 33—46). Word order of older stages of Czech has been investigated, e.g., by JELÍNEK, M. (Postavení atributu ν obrozenské odborné literature [Sentence Position of the Attribute in the Technical Literature of the Period of National Revival], SPFFBU A 15,1967, 69— 84) and GREPL, M. (Vyvoj slovosledu Ν Tylovë próze [The Development of Word Order in Tyl's Prose], SaS 20,1959,247— 261 ). The development of English word order has been studied, e.g., by §IMKO, J., ( Word Order in the Winchester Manuscript and in William Caxton's Edition of Thomas Malory's Morte Darthur (1485) — A Comparison, Halle a. Saale 1957). See also FIRBAS, J., Some thoughts... (quoted here in note 7 0 ). 76 Cf. his Porjadok slov i aktual'noje Clenenije predlozenija ν generativnom opisaniji slavjanskych jazykov [Word Order and Functional Sentence Perspective in a Generative Description of Slavonic Languages], Ceskoslovenské prednásky pro VI. mezinárodni sjezd slaoistä ν Praze, Prague 1968, 61-65. 77

See his papers O gramatiöeskom porjadke slov [On the Grammatical Order of Words], VJa 15, 1966, 6, 27—34; Frazovoje udarenije i porjadok

34

J. FIRBAS

word order rules. 78 One system operates in the act of communication, determines the order of words in regard to context and is to be accounted for by the theory of FSP. The other system is to be looked for outside any context. The order it produces is to be regarded as primary and grammatical. It may be quite fictitious, though in Russian it can be established on the basis of linguistic material. Hierarchically, the latter system is superior to the former. Further elaboration of the theory of FSP totally depends on the establishment of the primary grammatical order. 79 For Russian the author finds the grammatical word order in such sequences as in all possible contexts appear to be unchangeable and fixed.80 Thus the matrix sentence Kupiv bilet, passazir uspokojilsja contains a transform of Passazir kupil bilet. The impossibility of * Bilet kupiv,passazir uspokojilsja points to the grammatical V—Oacc. The fact that outside context the sentence Doc' ljubit mat' is interpreted as S-V-0 by Russian speakers bears out the grammatical character of the order. In the end, the author tentatively establishes the following grammatical word order for Russian: S Adv modi V O dat Oacc Adv dir . The adduced grammatical word order pattern contains the order S-V-O, which in English has come to be established as the grammaticalized core of the word order system. A note on this grammaticalized word order may not therefore be out of place here. It is revealed, for instance, by the following sentence structures: A boy liked a girl, A lion killed a hunter, A dog bit a wolf. Owing to the presence of the non-generic indefinite articles, they could open narratives or constitute complete single messages.63 It is slov [Phrase Accent and Word Order], To Honor Roman Jakobson, The Hague—Paris 1967, 967—976; Porjadok slov ν porozdajuäCej modeli jazyka [Word Order in a Generative Model of Language], Ceskoslovenské prednásky pro VI. mezinárodni sjezd slavistù ν Praze [Czechoslovak Papers Prepared for the Sixth International Congress of Slavists in Prague], Prague 1968, 5 1 - 5 9 . 78 See his Porjadok slov... (quoted here in note 7 7 ), 51. 79 Ibid., 58. 80 Ibid., 52.

THE CZECHOSLOVAK APPROACH TO FSP

35

certainly remarkable that a reader or hearer will most naturally interpret them as actor-action-goal, subject-verb-object, themetransition-rheme sequences on the semantic, the grammatical and and the F S P levels, respectively. It is evident that it is first and foremost sentence linearity that, particularly in regard to the nouns, signals the mentioned orders of functions. But what is of particular interest is the question why, in the absence of semantic and grammatical signals and any dependence on preceding context, sentence linearity signals just the mentioned orders. The explanation seems to be quite simple. Being a very primitive (though efficient) means, sentence linearity cannot — under the circumstances — but reflect the normal and natural order of phenomena as occurring in the extra-linguistic reality. Initiating an action, the actor necessarily exists before it. Only after it has started, can the action reach or affect its goal or produce some altogether new object (A potter made a vessel). The communication develops along the same line. The degrees of C D rise accordingly and the intonation centre falls on the object, expressing the goal of the action. The sequence displays the basic distribution of C D . The way the grammaticalized core of Modern English word order has become established is certainly not at variance with the nature and requirements of FSP. The conclusion suggests itself that it could not be otherwise, for F S P serves the communicative purpose of the sentence. On the basic instance level, i.e. on the level of complete contextual independence, the grammaticalized order S - V - 0 is in full harmony with the basic distribution of C D . Language is, of course, a pliant tool. In the presence of semantic and/or grammatical signals linearity may become inoperative and deviations from the basic distribution of C D may take place. Moreover, context may intervene and some of the items, for instance, the object, become contextually dependent and the extralinguistic reality in consequence viewed and presented from a different angle, the degrees of C D changing accordingly. Language cannot be severed from its primary function, i.e. that of communication, if its essential character is not to be overlooked. Communication involves a communicative purpose. A communica-

36

J. FIRBAS

tive purpose is always prompted by a situation, a context, be it one in which a semantic and grammatical structure appears at the beginning of a message or itself constitutes a complete isolated single message. The offered notes substantiate the view that the basic distribution of CD is a factor actually respected by language. In actual utterances, or utterance types, it tends to be either fully observed or to have the deviations duly signalled. As I have attempted to show in the present paper and elsewhere, this is implemented by an interplay of means of FSP. The ways this interplay is implemented will naturally vary from language to language and lead to different hierarchies of word order principles. The ultimate point of departure, however, will remain the basic distribution of CD. If this argument is correct, it shows that it is not on the level of grammatical structure, but on the level of FSP that the communicative purpose of an utterance is decided. Grammatical structure can become the leading factor within the hierarchy of word order principles, but only on condition it duly serves the communicative purpose of the utterance, not infringing on the requirements of FSP. Grammatical structure certainly is capable of complying with these requirements, for in actual fact it is not severed from lexical and ultimately from cognitive meaning. As A. Reichling has pointed out and F. Danes recently emphasized, grammatical structure does not merely combine forms as such, but with the aid of formal relations effects a semantic connexion, i.e. a connexion of meanings. 81 "These meanings might be called syntactic meanings and characterized as the generalization of lexical meanings contained in the sentence, the generalization being accomplished by the relational structure of the ... grammatical sentence pattern." 8 2 In the light of what has just been put forth, the basic distribution of CD seems to be a more suitable starting point for generating word orders than a primary grammatical sentence pattern. 81 REICHLING, Α., Principles and Methods of Syntax: Cryptoanalytical Formalism, Lingua 10, 1961, 1. Quoted after DANES, F., Some Thoughts... {quoted here in note 1 7 ) , 55. 82 DANES, F., op. cit., 56.

THE CZECHOSLOVAK A P P R O A C H TO FSP

37

I have only been able to touch upon some aspects of the theory of FSP. My notes have by no means offered an exhaustive account. Moreover, I have focussed my attention on the Czechoslovak approach. Nevertheless, I believe I am right in saying that the problems raised by H. Weil more than a hundred years ago have opened vistas of research that might bring us a little nearer to a better understanding of language as a tool of communication.

ON TWO COMMUNICATIVE DYNAMISMS Ales Svoboda(Brno)

When dealing with the theory of functional sentence perspective (FSP), some scholars have recently pointed out that some concepts of FSP are insufficiently defined. On the one hand, this needn't be a stumbling block to the analyst, since there are many examples of insufficiently defined basic concepts (even such as "word", "sentence", "deep structure", etc.) in both older and present-day linguistic theories; nevertheless these theories have been successfully developed. On the other hand, of course, concepts should and must be defined as precisely as possible to make understanding easier and research more effective. Let us turn our attention to one of the basic concepts of FSP, the concept of communicative dynamism (CD). According to Jan Firbas, CD is a quality displayed by communication in its development (unfolding) of the information to be conveyed and consisting of advancing this development (FIRBAS, 1965), or, in other words, CD is a quality expressible in degrees of the contribution to the development of the discourse (cf. FIRBAS, 1961, and other papers by the same author). If these statements are taken in isolation, they may seem insufficient and may lead to the following questions: What are the rules for ascertaining the degrees of the contribution to the development of the discourse? If the rules have been discovered, are they correct? As I see it, however, I regard the entire work of J. Firbas as an extensive definition of CD, presenting an open system of rules and at the same time giving the reasons for their existence. Another point is to prove the relevance of the reasons and the

ON TWO COMMUNICATIVE DYNAMISMS

39

validity of the rules, not only on the grounds of logic and common sense, but also by means of measurement or testing. In my opinion, the concept of CD as defined above is closely connected with two things: (i) with the general way of conveying information by means of a language and therefore with the objective information-amount carried by certain language elements, (ii) with the personal approach of a language user to the communication, which consists of increasing or decreasing the information-amount of certain language elements according to the intention of a language user and may be termed as subjective information. (This concept is close to R. Well's "subjective information" — cf. Wells, 1961 — provided that not only one sentence but also one element may be more informative than the other according to the intention of a language user.) On proving the validity of the theory of FSP, the ideal case would be to measure either the relative or the absolute amounts of both objective and subjective information carried by language elements. I think, however, that to the present stage of the development of the information theory the probability of proving the theory of FSP in this way is very low. That is why I will try to approach the problem from a different angle and will make an attempt to define CD by means of contextual dependence, which seems to be less resistant to measuring or testing than the informationamount is. By contextual dependence I understand two different kinds of dependence: (i) horizontal (or linear) dependence, constituted by the sequence of elements, (ii) vertical dependence (or crossreference), constituted by the occurrence of identical and/or closely related elements. According to the three-level approach to syntax (cf. D A N E S , 1 9 6 4 ) , each of the two kinds of contextual dependence is to be ascertained and measured on all the three levels with certain modifications reflecting the specific character of the respective level. On the grammatical level, contextual dependence is constituted by (i) the sequence of formal elements, (ii) the occurrence of identical formal elements and their substitutes. Both in regard to (i) and (ii), formal contextual dependence would be measured in terms of

40

A. SVOBODA

degrees of formal distance existing between the elements within a certain formal textual segment. On the semantic level, contextual dependence is constituted by (i) the sequence of semantic elements, (ii) the occurrence of identical or closely related semantic elements. Both in regard to (i) and (ii), semantic contextual dependence would be measured in terms of degrees of semantic affinity (or semantic distance) of the elements within a certain semantic textual segment. On what I have elsewhere termed functional level (SVOBODA, 1 9 6 8 ) , contextual dependence is constituted by (i) the sequence of communicative elements, each of them reflecting a definite part of the extra-lingual reality in the very act of communication, (ii) the occurrence of a communicative element or elements closely related to the part of the extra-lingual reality to which a language user is taking a standpoint. (This formulation is based on J. Vachek's functional definition of a sentence: The sentence is an elementary verbal act of taking a standpoint towards some reality. Cf. VACHEK, 1 9 6 2 . ) Contextual dependence on the functional level would in both cases be measured in terms of degrees of correspondence between the parts of the extra-lingual reality expressed by communicative elements and that which is being taken a standpoint to within a certain communicative textual segment reflecting the situation in the extra-lingual reality at a given moment of communication. Before adducing the definition of CD, I should like to emphasize the fact that it is only a working definition, its aim being to make CD verifiable not only in languages with the distribution of CD representing the leading word-order principle, but also in those where other word-order principles play the dominant role. Communicative dynamism is an abstraction from and generalization of the reciprocal of contextual dependence. It is evident from what has just been stated that, on the one hand, CD and contextual dependence are regarded as complementary phenomena, but, on the other hand, CD is a broader concept than that of contextual dependence, since the former is an abstraction from and generalization of the reciprocal of the latter. How to understand the terms "abstraction" and "generalization"? Suppose we have symbolic marks denoting the values attached

O N TWO COMMUNICATIVE DYNAMISMS

41

to the above elements on the grounds of measuring or testing both kinds of contextual dependence on all the three syntactic levels. If we abstract from the actual text and replace the elements by the above symbols (denoting the degrees of contextual dependence) in a large number of utterances, we arrive at a set of various patterns displaying the distribution of contextual dependence in a language. These patterns reveal certain laws according to which the utterances are organized. As it seems, however, that the laws hold good not only for utterances within a certain context, but also for those standing in isolation or in the initial position of the discourse, we generalize the above laws of utterance organization and apply them also to utterances where we can hardly speak of any context or contextual dependence. From this point of view, the concept CD is apparently more suitable than that of contextual dependence, since it may be employed in all the possible kinds of utterances. As we are mainly interested not in the absolute but in the relative amounts of CD in a given pattern of utterance organization, we are to a certain extent free in choosing the criteria of measuring contextual dependence, the results being approximately the same. If we take this into account, we are also able to understand why the employment of different criteria of contextual dependence in different languages needn't be any obstacle to comparing their patterns of utterance organization by means of the concept of CD. Though measuring contextual dependence may appear to be no easy task, in any case it seems to be soluble at the present stage of linguistic research and may, at least, prove the validity or invalidity of such part of the theory of FSP as deals with utterances within a context.

REFERENCES DANES, F., A three-level approach to syntax. TLP 1, 1964, 225— 240. FIRBAS, J., On the communicative value of the modern English finite verb. BSE 3, 1961, 79-104. FIRBAS, J., A note on transition proper in functional sentence analysis. PhP 8, 1965, 170-176.

42

A. SVOBODA

SvOBODA. Α., The hierarchy of communicative units and fields as illustrated by English attributive constructions, BSE 7, 1968, 49—101. VACHEK, J. - FIRBAS, J., Lingvistickä Charakteristika souëasné angliítiny [Linguistic characterology of contemporary English]. Prague (multiplicated), 1962. WELLS, R., Measure of subjective information. Structure of language and its mathematical aspects. In: Proceedings of Symposia of applied mathematics 12, 1961, 237-244.

THE PLACE OF "FUNCTIONAL SENTENCE PERSPECTIVE" IN THE SYSTEM O F LINGUISTIC DESCRIPTION Μ. A. K. Halliday (London)

1. 2. 3. 4. 5. 6.

"Function" in FSP "Function" in functional theories of language The functions of language Manifestation of these functions in the language system Why is language as it is? Summary of the place of FSP

1. "FUNCTION" I N FSP

The concept of "function" has always been centrally associated with Prague linguistic scholarship. It is not accidental, I think, that the term appears in both the expressions "functional theory of language" and "functional sentence perspective". The notion of functional sentence perspective is very closely bound up with that of functional theories of language. I suspect, however, that the link between them is actually rather obscured by the use of "functional" in both these expressions, since the meaning of the term is somewhat different in the two places. The notion of a functional theory of language is itself capable of more than one interpretation, and since it is important to the understanding of the place of FSP in linguistic description I shall return to it below. Here it will suffice to note that "functional", in the context "functional theory", refers to the functions of language. In "functional sentence perspective", on the other hand, "functional" relates to the analysis of the sentence into parts having a function in the total communication process.

44

Μ . Α. Κ . H A L L I D A Y

I will assume here that FSP is a universal phenomenon. There were at one time suggestions that it might be specific to certain languages, at least in its fully developed form; but these were founded, in my opinion, on a confusion between FSP as the organization of meaning, on the one hand, and the means by which it is realized on the other. I do not think we need take very seriously the notion that there are languages without FSP; in fact I would define FSP in such a way that this would be a theoretical impossibility — a semiotic system without FSP would not be a language. This is not to say that FSP is irrelevant to linguistic characterology; there is significant variation as regards the choices available in different languages, as well as in where and how these choices are made. But all languages have an FSP component. FSP is concerned with the organization of the sentence as a message: with "how the grammatical and semantic structures function in the very act of communication" (Danes, in TLP 1, p. 227). This is not simply a matter of relating the sentence to the context in which it occurs. In FSP, as in any other component of linguistic organization, it is the speaker's meaning potential that is being represented. He has, among other things, the option of deciding not to relate what he is saying to the context of utterance. He makes his own choices. What he does is to express a particular pattern of information, which represents his selection from the complex of systematic options that are available for the creation of text. I would define FSP as the "text-creating" component of language. This puts a somewhat different emphasis on the "F". I think the term "functional" is most appropriate here if we interpret FSP as being the expression of one of the basic functions of language, namely the function of creating text. This is using the term in the general sense in which it is used in "functional theory of language". A "functional" theory is one which is concerned with the functions of language itself; and the text-creating function is one of them.

T H E PLACE O F FSP I N LINGUISTIC DESCRIPTION

45

2. "FUNCTION" IN FUNCTIONAL THEORIES OF LANGUAGE

The limitation of functional theories of language, from the point of view of linguistics, is that they have usually been purely extrinsic. They have been attempts to incorporate language into a psychological or ethnographic theory. This is entirely justifiable and desirable, but it means that the categorization of language functions which is offered in such a theory is inadequate to explain the nature of language. That is not what it is designed to do. Let me make my own position clear on this point. In attempting to understand the organization of language I find it most helpful to work with Firth's concepts of "system" and "structure" — with "system-structure theory", as it has been called. In particular, in investigating language in social contexts and settings I like to take the "system" as the fundamental concept. A system is a set of options in a stated environment; in other words, a choice, together with a condition of entry. It seems to me that the system-structure framework needs to be supplemented in two ways. First, it is insufficiently explicit, and needs to be underpinned by a formalized generative model. Lamb's stratification theory, with its Hjelmslevian basis, provides this, and gives a systematic account of linguistic levels that is entirely compatible with the general theory of levels developed in European linguistics. Secondly, it is insufficiently explanatory in the sense of explaining why language has the particular form and shape it has. We cannot give anything approaching a definitive answer to the question why the human semiotic should have taken precisely this form and no other; but we can begin to look into it: and for this purpose we need a theory of linguistic functions. In principle, language is as it is because of the functions it has evolved to serve, and Prague linguistics is unique in attempting to incorporate a theory of what these functions are. Here I venture to disagree with the observation of Novák and Sgall, that "the functions concerned should be ascribed not to the language but to utterances" (TLP 3, p. 292). In my opinion a fune-

46

Μ. Α. Κ. HALLIDAY

tional theory is a theory of language, and is an essential aspect of any theory that attempts to explain the nature of language. And this is how I would approach the question of the place of FSP. The place of FSP in the system of linguistic description is determined by the fact that it is (or at least it is an aspect of) one of a small number of functional components of the language system. In his well-known article in TLP 1, Danes distinguished three "levels" within syntax: (1) the level of the semantic structure of the sentence (2) the level of the grammatical structure of the sentence (3) the level of the organization of the utterance Svoboda (BSE 7, p. 56) regards these as three "systems", semantic, grammatical and functional, each with its own syntactic elements and relations. These are important and fundamental categories. But they are not just accidentally co-existent systems or levels of structure; they are functional components of the grammar. They are the manifestation, in the language system, of the functions of language, in the general sense in which the term has been used from the work of Bühler onwards. This being the case, however, we need to show how Biihler's categories, or else some modified version of them, are represented in the Danes-Svoboda framework of analysis.

3. THE FUNCTIONS OF LANGUAGE

We have no difficulty in identifying Danes' "semantic level" with Biihler's "representational function". If one of the functions of language is to express our experience of the world that is around us and inside us, it is natural that this should be reflected in the linguistic system. At first sight, the equivalence might seem to go no further than this; but it does. The most problematical of Danes' categories is that of the "level of grammatical structure". This seems rather circular. Why should language have a level of structure whose only function is to be a level of structure? Let us look back at Bühler. He has a con-

T H E PLACE O F FSP I N L I N G U I S T I C DESCRIPTION

47

ative function and an expressive function. The difference between these two is significant psychologically, but linguistically it is very tenuous: is an interrogative, for example, a demand to be given information (conative), or an expression of a desire for knowledge (expressive)? It is not surprising to find that expressive and conative are not really distinct in the language system. They are combined into a single "personal" function — or, as I would prefer to call it, to bring out its social nature, an "interpersonal" function. How is this represented in linguistic structure? Actually, I think, it is Danes' "level of grammatical structure", the "level" whose elements are subject, predicate and the like. This is not very well named, if with Svoboda we understand "grammatical" to mean "purely formal"; it is a functional component like the first one — and, we may add, is equally "semantic". It has often been noted, for example, that the subject in English is essential to the expression of mood; and in fact the organization of clauses into some form of predicative structure has in many languages a modal function, expressing the speaker's participation in, or intrusion into, the speech event: his choice of speech role (mood) and his assessment of the validity of what he is saying (modality). This in turn is part of a more general component of meaning which includes his attitudes and comments, assertions of familiarity and distance, and the like. Let us postulate, then, an "experiential" component ( = Biihler's "representational", Danes' "semantic") and an "interpersonal" component ( = Biihler's "conative" and "expressive", Danes' "grammatical"), and insist that both are represented on equal terms in the description of language. In other words, each component has both semantic and lexicogrammatical connotations. When it comes to the third component, Bühler has nothing to say — naturally, since he was not primarily concerned with the nature of the linguistic system. But this is also a functional component, provided we accept the notion of an enabling function that is intrinsic to language: this is what we referred to above as the

48

Μ . Α. Κ. HALLIDAY

function of creating text. It is this that enables language to be operational; "text" is language in use. The speaker's command of his langue includes an awareness of the difference between text and non-text — lists of words, or random sets of sentences. Normally he will assume that what he hears or reads is text, and he will go to great lengths to justify his assumption and ensure that communication is taking place. This assumption is a functional one; it rests not so much on recognizing words and structures as on recognizing the role that language is playing in the situation. And the language will be recognized as playing some role only if it is acceptable as text. Let us refer to this as the "textual" function of language, since it is the function of creating text — or "texture", to use a stylistic term, which is really the same notion. This function is not in Bühler's scheme; but it is Danes' "level of organization of utterance", whose elements are Svoboda's "communicative units". Although the textual function differs from the other two in that it is intrinsic to language, and thus instrumental and not autonomous, I do not think it should be regarded as restricted to parole, or to the utterance. It is an integral component of the language system, and represents a part of the meaning potential of this system. FSP can be defined, in this way, as the "textual" component in the grammar of the sentence. The study of FSP was at first directed just to the structure of sentence and clause. Subsequently, it has been extended to other units having a "communicative" element in their structure: to various classes of the phrase — deixis in the noun phrase is an example of a text-creating element — as well as to units which, as Danes pointed out (Word 1960), may have no equivalent in the grammatical hierarchy (e.g. the unit realized as one tone group in English). References to work on FSP are too numerous, and too well-known, to be cited here; there have been detailed and perspicuous descriptions, especially of English and Czech, by Czechoslovak linguists such as Trávnícek, Vachek, Danes, Poldauf, Firbas, Benes, Svoboda and others, which have given a fundamental insight into this important and otherwise rather neglected area of linguistic pattern.

THE PLACE OF FSP IN LINGUISTIC DESCRIPTION

49

4. MANIFESTATION OF THESE FUNCTIONS IN THE LANGUAGE SYSTEM

Where are these functions of language manifested in the language system? Obviously, in the grammatical structure. A clause such as the sun was shining on the sea shows (at least) these three dimensions of grammatical structure: the sun experiential:

Actor

interpersonal:

Modal

textual:

Theme

was

shining

on the sea Locative

Process Propositional Rheme

(I take "modal propositional" as the more general structure from which categories like subject, predicate, object etc. can be derived.)

But since these structures are the means of expression of the basic functions of language, they relate to particular functionally defined areas of meaning. Grammatical structure may be regarded, in fact, as the means whereby the various components of meaning, deriving from the different functions of language, are integrated together. We can see that each component makes its contribution to the total structural complex. The different functions are, quite evidently, simultaneous and compatible. We should not be misled into equating "function of language" with "use of language". There are indefinitely many uses of language, which no linguistic theory has attempted to systematize; but the fact that language can serve such a variety of purposes is precisely because the language system is organized into this small set of highly generalized functional components. Whatever we are using language for, we need to make some reference to the categories of our experience; we need to take on some role in the interpersonal situation; and we need to embody these in the form of

50

Μ. Α. Κ. HALLIDAY

text. (I think there may also be a "logical" component to be brought in, but this need not concern us here.) We draw on all these areas of linguistic potential at the same time. If we describe the grammar in terms of paradigmatic sets of features, or "systems" in the Firthian sense, we find a clear indication that the grammatical system itself has a functional basis. In the English clause, for example, there is one self-contained set of interrelated systems concerned with transitivity, and another concerned with mood. But transitivity and mood are nothing other than functional components in the meaning potential of the clause. Transitivity is the grammar of processes — of actions, mental processes, relations — and the participants in these processes, and the attendant circumstances. This is the experiential component in the clause, Danes' "semantic". Mood is the grammar of speech functions — the roles adopted (and those imposed on the hearer) by the speaker, and his associated attitudes. This is the interpersonal component, Danes' "grammatical". And we find a third set of systems (which I referred to elsewhere under the general heading of "theme") concerned with the grammar of messages — the status of the clause and its parts as "units of communication". This is the textual component of clause structure, or FSP. FSP shows up very clearly as a distinct, functionally determined set of options in the underlying grammar of the clause.

5. WHY IS LANGUAGE AS IT IS?

We are interested in FSP because it is an integral part of the system of language, and therefore essential to the understanding of the processes of speaking, listening, reading and writing. We take it for granted that language is a multiple coding system, organized into levels, or "strata". This can be explained, it is true, in terms of the problem of reduction. To oversimplify the matter, a large number of complex meanings is to be encoded in a small number of simple sounds, and this cannot be achieved without intervening levels (i.e. grammar and phonology). But this stratal

T H E PLACE O F FSP I N L I N G U I S T I C DESCRIPTION

51

structure has evolved in the context of the demands that are made on language, and the nature and organization of these intermediate levels — the nature of linguistic form, in Hjelmslev's sense — reflects the role of language in the life of man. With the very young child, the uses of language seem to be rather discrete; and each has its own "grammar", or "proto-grammar" since it has no stratal organization. With my eleven-month-old son, for example, I can recognize four uses of language, with just two or three options in each. But adult use of language is such that, with minor exceptions, each utterance has to be multifunctional — while at the same time having an integrated structure. There must therefore be a level of organization of meaning: a semantic level, or in Lamb's terms "semological stratum". In Hjelmslevian terms, the "content purport" has to be separated from, and organized into, a "content substance" as a precondition of its encoding in "content form". What we are calling the functions of language may be regarded as the generalized categories of "content substance" that the adult use of language requires. An utterance must be about something; it must express the speaker's stake in the matter; and it must be operational in its own context, either in the "here and now" or in some second-order context created by the language. These conditions would seem to determine a significant part of the properties of the language system. Specifically, the functional orientation of the system determines the kind of interdependence that exists within the meaning potential. Certain options are dependent on others; for example both modality and "key" are largely dependent on mood, and all these are within the general "interpersonal" domain. Likewise there is considerable interdependence among the options within FSP, although these are largely independent of options in the other components. Also, the functional basis of language is reflected in the nature of constituent structure, which has not merely to serve in the realization of meaning but to accommodate in a single structural realization configurations of elements deriving from different functional points of origin.

52

Μ. Α. Κ. HALLIDAY

6. SUMMARY OF THE PLACE OF FSP

Since FSP is readily incorporated into what Jakobson called a "means-ends" model of language, this enables us to define its place in the description of language. First, it is part of general linguistic theory. There is a "textual" component in every language (clearly, if we claim that it is essential to the functioning of language, we are claiming that it is universal). Hence it enters into the description of every language. Secondly, the textual component is not a "level" in the usual (stratal) sense of the term; it is a "vertical" division within the content plane. There is no suggestion of one component being "deeper" or "more surface" than another. Thirdly, the textual component differs from the others in that, while they are directly relatable to the conditions of language use, the textual component is related only indirectly, through its function of creating text. Fourthly, there is the place of FSP within the textual component. We have kept these two terms distinct because the latter is expressly related to a theory of language function, and is a broader category. FSP has been used mainly to refer to structural relations within the sentence; but the textual component specifies both intra-sentence and inter-sentence relations, including non-structural relations of presupposition. Let me try and summarize the domain of the textual component, as I see it. The textual component includes: 1. relations of presupposition (i.e. reference, substitution, conjunction and lexical presupposition) (a) verbal (i.e. anaphora and cataphora) (i) between sentences (my "Cohesion") (ii) within sentences (b) situational 2. structural relations (i.e. FSP) (a) in syntactic units (i) sentence and clause (ii) phrase ("group") (b) in communicative units (my "information structure")

T H E PLACE O F FSP IN LINGUISTIC DESCRIPTION

53

As far as English is concerned, the structural relations in the sentence and clause are those of theme and rheme, together with various patterns of identification and predication (the thing t h a t . . . is ...; it's ... that ...; etc.). I myself take "theme" in Trávnícek's sense: it is the FSP element that is realized by first position, and has nothing to do with previous mention. The relations of information structure — that is, those within the unit realized by the tone group; Firbas' "prosodie means of FSP" — are those of "given" and "new", as I see it, which do relate to the recoverability of information from situation or preceding text. In the nominal and verbal group, the structures involved in FSP are those relating to the here and now, such as the TMEs in Firbas' work: the "deictic" elements in the widest sense of the term. It may be that, if the general notion of the textual component (or "communicative component" — but I am less happy with this, because it suggests the personal interaction of speaker and hearer, which is treated as a different function in the language system) is accepted, the term FSP would be regarded as covering all the relations which derive from this function of language. Be that as it may, the two concepts, that of the text-creating function in a functional theory of language, and that of functional sentence perspective, clearly belong together. The emphasis on functional theories, in Czechoslovak linguistic scholarship has thrown light on a general property of language, one that helps to explain the nature of the linguistic system; while work on FSP has thrown light on the specific text-creating function within the linguistic system, that whereby language is in turn enabled to serve the variety of ends in the context of which it has evolved.

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG IN DER SPRACHBESCHREIBUNG Petr Sgall (Prag)

Wir möchten zeigen, daß die Thema-Rhema-Gliederung als eine Hierarchie 'der Elemente der semantischen Satzstruktur in jede Sprachbeschreibung eingegliedert werden sollte, die das Sprachsystem als ein Ganzes (samt semantischen und prosodischen Phenomena) erfassen will. In einer formal gefaßten Sprachbeschreibung kann vielleicht die Thema-Rhema-Gliederung (TRG) als eine Anordnung der Elemente der Satzrepräsentation an der gesagten Ebene beschrieben werden, wobei diese Anordnung der Wortstellungsrelation der "Oberflächenstruktur" nahe steht und mit der syntaktischen Hierarchie eng verbunden ist. (Wenn man mit einer Abhängigkeitsgrammatik arbeitet, ist die enge Beziehung zwischen der TRG und der syntaktischen Struktur durch die Projektivität des Satzes — und auch durch Abweichungen von ihr — wiedergegeben; wird eine Phrasengrammatik als Basis der Sprachbeschreibung gewählt, dann spiegelt sich diese enge Beziehung darin ab, daß die TRG in der Regel die Grenzen der Konstituenten nicht überschreitet.)* 1. Wenn wir die Stellung der TRG in einer theoretischen Sprachbeschreibung beurteilen wollen, ist es notwendig, zuerst zu wissen, was unter dem Termin „TRG" (oder „aktuelle Satzgliederung", „funktionelle Satzperspektive") zu verstehen ist. Es gibt Linguisten, die die TRG als etwas durch den Kontext (oder genauer durch * Neuere Forschungsergebnisse und eine ausführlichere Fassung der ganzen Konzeption sind jetzt im Buch P. Sgall, E. Hajiôovà, E. Benesová, Topic, Focus and Generative Semantics, Kronberg/Taunus 1973, enthalten.

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG

55

die Konsituation) Bestimmtes ansehen, als eine Gliederung des Satzes in das Bekannte, schon vorher Gegebene, und in Elemente, die neue Information mit sich bringen. D A H L (1969, §2) lehnt diese Ansichten wohl mit Recht ab, ungenau rechnet er aber Adamec zu ihren Anhängern. Die meisten Forscher wollen zwischen diesen Kontextbedingungen und der eigentlichen T R G unterscheiden, u. a. deshalb, weil auch das Rhema aus der Konsituation bekannt sein kann. Allerdings, wenn man zeigen will, daß die T R G im allgemeinen nicht durch die Konsituation eindeutig bestimmt wird, ist es vielleicht noch wichtiger, daß Beispiele vorhanden sind, wo bei derselben Konsituation verschiedene Sätze (Aussagen) mit demselben Inhalt vorkommen können, die sich eben durch ihre T R G unterscheiden. Nehmen wir ζ. B. als den vorhergehenden Kontext den Satz „Im heutigen Match der Hockey-Weltmeisterschaft spielte Schweden gegen Finnland." Der nächste Satz kann dann lauten „Den Match hat Finnland gewonnen", aber auch „Finnland hat den Match gewonnen"; offenbar ist im ersten Fall das Wort Finnland der Kern des Rhemas, während es im zweiten Fall (wenn der Satz normale Intonation hat, ohne Satzakzent am Anfang) als das Thema auftritt. Anders gesagt: in der gegebenen Konsituation kann man dieselbe Tatsache verschieden formulieren — entweder spricht man über (die Mannschaft von) Finnland und man konstatiert, wie der Match für diese Mannschaft endete, oder man spricht über den Match und stellt fest, wer ihn gewann. Oder („Am Weihnachtsabend erwarteten wir die Verwandten.") „Onkel Ernst ist als erster gekommen." „Als erster ist Onkel Ernst gekommen." (Entweder spricht man nach dem gegebenen Kontext vom Onkel Ernst, und man konstatiert, wann er gekommen ist, oder man spricht vom Ersten, der gekommen ist, und man stellt fest, daß es der Onkel war.) Man wird also zugeben, daß die einfachste und oft benutzte Charakteristik der T R G — „worüber man spricht, ist das Thema, was man darüber aussagt, ist das Rhema" — im Grunde richtig ist. (Man kann diese Formulierung ζ. B. bei MATHESIUS, 1961, 91, finden, ebenso wie bei HALLIDAY, 1967, 212, Β REKLE, 1970, 72, und bei anderen.) Diese einfache Charakteristik genügt aber nicht, um

56

P. SGALL

alle in Betracht kommenden Erscheinungen zu erfassen. Mindestens in zwei Richtungen muß sie weiter spezifiziert werden. Erstens handelt es sich nicht nur um eine Dichotomie, sondern um eine Skala, oder genauer um eine Hierarchie, die rekursive Eigenschaften ausweist; in den Arbeiten von Firbas und seinen Nachfolgern (s. jetzt vor allem SVOBODA, 1967) wurde sie als die Hierarchie des kommunikativen Dynamismus vielseitig charakterisiert. (Mit einer solchen Hierarchie arbeitet auch z. B . UHLÍROVÁ, vorb., die aus DANES'S Konzeption ausgeht.) Zweitens ist nicht immer ein einziges nominales Satzglied (Subjekt, Objekt) das Thema des Satzes, sondern man kann in dieser Position auch zwei solche Satzglieder, oder auch Adverbialien finden, die oft als eine „Situationskulisse" aufgefaßt werden (z. B. „local setting" bei FIRBAS, 1964, 271). Wie ADAMEC (1966, 21f) mit Recht bemerkt, ist es nicht einfach, eine Grenze zwischen dem Thema im eigenen Sinn und dieser Situationskulisse zu ziehen; auch das zweite von unseren oben angeführten Beispielen zeigt, daß die Position des Themas nicht immer ganz klar ist; eine Paraphrase wie „Von dem, der als erster kam, sage ich, daß es Onkel Ernst war" klingt mindestens unnatürlich. Ähnliche Schwierigkeiten gibt es auch dort, wo ein verbales Thema auftritt (vgl. unten, § 2.21). Der Unterschied zwischen dem „worüber man spricht" und „was man darüber sagt" kann also beim Studium der TRG nützlich sein, er reicht jedoch nicht dazu aus, um die TRG als Objekt einer Theorie zu charakterisieren. In Carnaps Terminologie können wir sagen, daß wir diesen Unterschied als ein Explicandum (den präsystematischen Begriff, der durch die Theorie erst präzisiert werden soll) nehmen dürfen, daß es aber notwendig ist, das entsprechende Explicatum zu suchen, d. h. einen explizit formulierten Begriff, der gut definiert ist und dem Explicandum so gut wie möglich entspricht. Die meisten theoretischen Arbeiten über die TRG bieten leider keine klare Charakteristik ihres eigenen Objekts, die man als einen Ausgangspukt für diese Explikation betrachten könnte. Eine der klarsten Charakteristiken ist bei D A N E § (1968, 127) zu finden; falls wir gut verstehen, da könnte nach ihm (für das Tschechische, also

Z U R STELLUNG D E R T H E M A - R H E M A - G L I E D E R U N G

57

für eine Sprache, wo die TRG mehr als anderswo mit der Wortstellung verbunden ist) die TRG mittels der Unterschiede zwischen Sätzen (Aussagen) charakterisiert werden, die verschiedene Wortfolge (oder Intonation), aber eine übereinstimmende grammatische (und lexikalische) Struktur aufweisen. Das ist freilich nur dort ausreichend, wo grammatische Mittel der TRG nicht in Betracht kommen. Es wird wahrscheinlich möglich sein, eine festere Grundlage für die Theorie der TRG auf Kriterien aufzubauen, die dadurch gegeben sind, daß man feststellen kann, auf welche Frage der gegebene Satz eine Antwort sein könnte (und welche Fragen ihm in einem Dialog folgen können); vgl. schon H A T C H E R (1956), aus den neueren Arbeiten vor allem D A N E S (1968,1968a), STAAL (1967, 77f). Bis jetzt sind aber diese Kriterien nicht systematisch genug formuliert worden. Um eine explizitere Charakteristik der TRG zu erreichen, die einem formalen Explicatum näher käme, müssen wir die Stellung der TRG im Ganzen der Sprachbeschreibung nachprüfen, denn die formale Deutung muß Begriffe benutzen, die in der Sprachbeschreibung schon ihren festen Platz, eine befriedigende Definition und klare Beziehungen untereinander haben. So kommen wir zum eigenen Thema unserer Ausführungen. In der gegebenen Sachlage können wir allerdings eher Fragen stellen als lösen. Wir versuchen es, einige Hypothesen aufzustellen und hie und da auf Beziehungen zwischen verschiedenen Gesichtspunkten aufmerksam zu machen. 2.1 Es gibt viele Konzeptionen der theoretischen Sprachbeschreibung (der Analyse des Sprachsystems), die meistens nicht explizit formuliert sind und oft die TRG nicht systematisch berücksichtigen. Es fragt sich, ob man unter diesen Umständen etwas allgemein gültiges über die Stellung der TRG in einer Sprachbeschreibung sagen kann. Eine solche Möglichkeit muß vor allem dort gesucht werden, wo die verschiedensten Typen der Sprachbeschreibung gemeinsame Voraussetzungen aufweisen. Wir glauben, daß man solche Voraussetzungen doch finden kann, wenn auch in bescheidenem Maße. Man kann wohl behaupten, daß jede Sprachbeschreibung mindestens mit drei Dimensionen arbeiten muß.

58

P. SGALL

1. „von oben nach unten" (oder „von der.Tiefe zur Oberfläche"), d. h. die Dimension, die dem Verhältnis zwischen Inhalt und Ausdruck (in der glossematischen Terminologie) entspricht und die für die Sprache als Zeichensystem besonders wichtig ist; 2. „von außen nach innen", oder von den einzelnen Erscheinungen zu ihrer Klassifikation, bei der man vom Einzelnen abstrahiert, um mehr oder weniger allgemeine Klassen von Erscheinungen, Einheiten und ihren Arten zu erkennen; 3. „von links nach rechts", oder die Dimension der Zeitachse (es ist allerdings eine Dimension im System, nicht im Text gemeint). Wir möchten jetzt die Stellung der TRG in Hinsicht auf diese drei Dimensionen untersuchen. 2.11 Was die erste Dimension betrifft, handelt es sich vor allem um das Verhältnis der TRG zur Semantik. Es gibt hier zwei Gesichtspunkte, die näher nachgeprüft werden sollten. Einerseits gibt es — in der Prager Schule, wo man mit mehreren Ebenen des Sprachsystems arbeitet — Konzeptionen, in denen man mit einer spezifischen Ebene der TRG rechnet. Bei T R N K A (1964) und HOREJSÍ (1961) ist es die Ebene der Aussage, bei D O K U L I L und D A N E S (1958; D A N E S 1968a) ist es eine der Ebenen der Satzstruktur (neben der semantischen und der grammatischen Satzstruktur), ähnlich auch bei FIRBAS (1970a). Es ist jedoch nicht immer klar, wie die Beziehung zwischen den verschiedenen Ebenen, und besonderes die Beziehung zwischen der TRG und der semantischen Satzstruktur zu verstehen ist. Bei DANES (1968a, 68f) wird diese Beziehung durch Beispiele illustriert, die eben von diesem Gesichtspunkt eine Erklärung brauchen: Sätze wie „France borders on Switzerland" und „Switzerland borders on France" unterscheiden sich dadurch, daß in dem ersten die „dominierende Position" mit dem Wort „France" besetzt ist, in dem zweiten mit „Switzerland". Das ist bestimmt so, aber es bleibt noch zu zeigen, ob und wodurch die „dominierende Position" vom Thema abgehoben werden kann, und weiter, ob und womit dieser Begriff wirklich der Ebene der semantischen, nicht nur der grammatischen (syntaktischen) Satzstruktur gehört. Aus der Literatur zu diesen Fragen möchten wir außer JAKOBSON (1936, 251) auch HORÁLEKS (1967, 117) Ansicht

ZUR STELLUNG D E R THEMA-RHEMA-GLIEDERUNG

59

erwähnen, der hier „eine Hierarchie vom Standpunkt des Sprechenden oder des Kontextes" sieht. Jetzt bringt besonders Fillmore (1970) neue Erwägungen, die für diesen Themenkreis relevant sind (und zeigen, wie tiefgreifende Einzelforschungen notwendig sind, um hier mindestens einige feste Kriterien zu erhalten). — Warum wir mit dem Unterschied zwischen den Begriffen „Satz" und „Aussage" (utterance) nicht in dem Sinne arbeiten, wie es D a n e s — mit Hinsicht auf die T R G — tut (wonach verschiedene TRGYarianten Aussagen sind, die als Varianten desselben Satzes gelten können), wurde anderswo gezeigt ( S g a l l , 1967, 205f und 209). Der andere Gesichtspunkt (der in der letztgenannten Arbeit zur Diskussion gestellt wurde) kann vielleicht so formuliert werden, daß die T R G als eine Organisation oder Hierarchie der Einheiten der semantischen Ebene gedeutet wird. Wie aus den Arbeiten von Firbas — aber auch Danes und anderen — gut bekannt ist, muß man mindestens mit drei Schichten der T R G rechnen. Die erste von ihnen (die Grundschicht, wo die hierarchische Skala des kommunikativen Dynamismus eine merkmallose Form hat) besteht aus Fällen, in denen die hierarchische Skala des kommunikativen Dynamismus unmittelbar von den Beziehungen zwischen den Einheiten der semantischen Satzstruktur bestimmt ist: wo die Konsituation diese Beziehungen nicht anders beeinflußt, steht der Agens in dieser Skala vor der Aktion (d.h. er ist mehr „thematisch" als sie, trägt weniger kommunikativen Dynamismus), die Aktion steht vor dem Patiens, ein Existenzverb steht vor dem Existierenden (in Sätzen wie russ. „Nastupila vesna"), eine Situationskulisse steht vor der Aktion, eine spezifizierte Adverbialbestimmung nach ihr, usw. Schon diese Ausdrucksweise zeigt, daß es eine enge Beziehung zwischen dieser Skala und der Wortfolge gibt; wie bekannt, gilt das zu einem gewissen Grad selbst für Sprachen mit einer grammatisch gebundenen Wortfolge. Diese Beziehung wollen wir unten, im •§2.2, ausnützen. Eine Grundschicht, die durch die semantische Satzstruktur bedingt ist, wird von Forschern, die verschiedene Sprachen analysieren und verschiedene Konzeptionen anwenden, erschlossen; vgl. jetzt besonders auch Z e m b ( 1 9 6 8 ) für das Deutsche. Zur zweiten Schicht gehören die Fälle, wo die Konsituation (der

60

P . SGALL

vorhergehende Kontext usw.) zu einer merkmalhaften Variante der erwähnten Skala führt. (Wo z. B. der Patiens aus der Konsituation bekannt ist, geht er in einer solchen in den Kontext eingegliederten Variante dem Aktionsverb voran: „Deine Uhr habe ich schon gefunden.") — Weniger wichtig ist für uns hier die dritte Schicht der TRG, die sog. „zweite Instanz", d. h. Varianten, die durch einen Kontrast, eine nachdrückliche Wiederholung eines (Teils des) mißverstandenen Satzes u. ähnl. beeinflußt sind. Es ist zu bemerken, daß in der Analyse der semantisch- und kontextbedingten Varianten z. B. HEIDOLPH (1966) ganz unabhängig und von einer anderen Konzeption ausgehend zu ähnlichen Ergebnissen gekommen ist; für die Charakterisierung der TRG als einer Organisation der semantischen Repräsentation des Satzes (also nicht nur durch Kontext bedingt) haben — in verschiedenen Hinsichten - die Arbeiten von HALLIDAY (1967), BENES (1968; 1970) und BREKLE (1970) viel neues gebracht; auch z. B. ZATORSKI (1970) und unter den Transformationalisten LAKOFF (1969) sind sich dessen klar bewußt, daß die TRG (oder einzelne Erscheinungen, die in unserer Terminologie zu ihr gehören) semantisch relevant sind, daß sie in der Repräsentation eines Satzes auf der semantischen Ebene figurieren müssen. Sätze wie „Jeder dieser Beschreibungstypen ist für manche Sprachen adäquat" und „Für manche Sprachen ist jeder dieser Beschreibungstypen adäquat" zeigen, daß ein semantischer Unterschied (sogar ein Unterschied des ontologischen [kognitiven] Inhalts, der Wahrheitsbedingungen) auch dort bestehen kann, wo wir keine grammatischen oder lexikalischen Unterschiede finden: wenn keine Sprache existiert, für die alle erwähnten Beschreibungstypen adäquat seien, kann der erste, nicht aber der zweite Satz wahr sein. Es ist dann allerdings nicht nützlich, hier von zwei bloßen Varianten desselben Satzes zu sprechen. Man kann einwenden, daß es sich in solchen Fällen um Qualifikation handelt, und daraus schließen, daß es entweder ein Sonderfall ist, der eine spezifische Deutung braucht, oder daß der gegebene Unterschied nicht zur TRG gehört. Wir können aber fragen, ob ein Sonderfall unter die allgemeine Regel nicht fallen soll (wenn sie wirklich allgemein ist),

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG

61

und was für Unterschiede das sind, wenn nicht die in der T R G (sind sie grammatischer Art, so müßte man sie in einer Grammatik beschreiben; das hat aber wohl vor den Transformationalisten — dazu s. unten, § 2.21 — niemand getan). Wie man den folgenden Beispielen entnehmen kann, ist der Sonderfall gar nicht so spezifisch, und wenn man alle solche Fälle Qualifikation nennen will, dann müßte die Lehre über die Qualifikation in der Semantik natürlicher Sprachen zu einem neuen Verständnis der ganzen Semantik vieles beitragen: „In Österreich spricht man deutsch" — „Deutsch spricht man in Österreich" (vgl. SGALL, 1967, 209f); „Kurit' zdes'" — „Zdes' kurit'" (ein Beispiel von Martemjanov); „Kurit' vredno" — „Vredno kurit'" (GLADNEY, vorb.); „Kusajte za stolom" — „Za stolom kusajte" (SGALL, 1969). Es wäre nicht einfach, zu sagen, was hier quantifiziert wird und wo eigentlich die Grenzen der Qualifikation im Sprachsystem sind. Die hier illustrierten Fragen, die das Verhältnis der T R G zur Qualifikation betreffen, müssen in der Sprachbeschreibung auf der semantischen Ebene respektiert werden. Wenn die von DAHL (1969) gefundene Möglichkeit, die semantische Satzstruktur mittels der Beziehung der Implikation zu beschreiben, bestätigt wird (und wenn auch z. B. die Schwierigkeiten beseitigt werden, die damit verbunden sind, daß eine mehrfache Kombination von Implikationen mit einer Verbindung von Konjunktionen und Implikationen ontologisch gleichbedeutend, aber nicht linguistisch synonym ist), dann hätte man einen festen Stützpunkt für die Deutung der semantischen Relevanz der TRG. Die Skala des kommunikativen Dynamismus wird dann im wesentlichen mit dem Unterschied der Positionen vor und nach dem Implikationssymbol beschrieben werden; vor dem Implikationssymbol steht — im typischen Fall — als Thema eine generische (s. DAHL, 1969, § 3), oder eine bestimmte (vgl. BENESOVÁ, vorb.) Nominalphrase. Man muß dann die Frage stellen, welche Beziehungen (auf der semantischen Ebene) als Implikation interpretiert werden können, und welche nicht. Es handelt sich hier einerseits um verschiedene Stufen der Deprädikation (backgrounding), vgl. weiter unten, §2.13, andererseits um die Quantifikation; die Position der Quan-

62

P. SGALL

toren (und auch anderer Operatoren, wie der Negation, der unbestimmten Pronomina und mancher Typen der Adverbialbestimmung, vgl. SEUREN, 1 9 6 9 ) in der semantischen Repräsentation (und in der Hierarchie des kommunikativen Dynamismus) ist dann ebenso wichtig wie die Position der Argumente der Implikation. Es ist möglich, hier eine Hypothese auszusprechen, deren empirische Verifikation im Gange ist (s. HAJICOVÁ, vorb., und vgl. dazu die im § 2,21 erwähnten neuren Arbeiten der Transformationalisten): der Bereich (scope) eines Operators ist in der semantischen Repräsentation des Satzes dadurch gekennzeichnet, daß er links durch den Operator, rechts druch das Ende des (Matrix-, Konstituenten-) Satzes, der den Operatoren enthält, begrenzt ist. Anders gesagt, ein Quantor oder ein anderer Operator steht — in der semantischen Repräsentation, in der Hierarchie des kommunikativen Dynamismus — links von seinem ganzen Bereich. In der Oberflächenstruktur kann diese Anordnung allerdings anders sein (sei es mit einer merkmalhaften Intonation, oder mit bestimmten grammatischen Bedingungen verbunden). 2.12 Zu den Fragen der zweiten Dimension (die die Klassifikation der Einheiten einzelner Ebenen betrifft und z. B. durch den Unterschied der terminalen und nonterminalen Symbole in der Transformationsgrammatik, oder zwischen einzelnen Wörtern und Wortarten, Satzgliedern usw. in anderen Konzeptionen illustriert werden kann) möchten wir hier nur kurz folgendes bemerken. Nach dem, was im § 2.11 gesagt wurde, ist es notwendig, auf der semantischen Ebene Klassen von Einheiten (d. h. disjunkte Mengen komplexer Symbole) zu finden, die die einzelnen Distinktionen im Funktionieren dieser Einheiten in Hinsicht auf die T R G wiederspiegeln. Einen höchst nützlichen Schritt in dieser Richtung bilden z. B. die Gruppen von Beispielen (und ihre Klassifikation) bei DANES (1968a). Man sollte eine solche Trennungslinie zwischen den Verben finden, die einer Aktion entsprechen, und denen, die die Existenz wiedergeben, — oder mindestens zwischen den beiden Typen von Einheiten, die primär durch das Subjekt dieser oder jener Verben ausgedrückt werden. Die Beispiele, die BENESOVÁ (vorb.) bringt, zeigen, daß die erste dieser beiden Möglichkeiten

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG

63

mindestens für Sprachen wie das Tschechische gar nicht gesichert ist. Ein Verb wie „fahren" gilt in vielen Sätzen als ein Aktionsverb, es gibt aber Sätze wie „In Prag fahren jetzt Autobusse statt Troleybusse" (im Sinne von „...gibt es..."), wo es sich um eine bloße Existenz handelt. Der Unterschied kann vielleicht besser mit Hilfe von Subjekten charakterisiert werden, denn die Autorin weist darauf hin, daß in diesem Fall das (rhematische — mindestens in einem merkmallosen Fall) Subjekt unbestimmt ist, während dort, wo das Subjekt einer bestimmten Menge (die generische Bedeutung eingerechnet) entspricht, handelt es sich um ein Subjekt, das primär thematisch ist und auf der semantischen Ebene als Agens oder ähnlich gedeutet werden kann. Wie wir schon bemerkt haben, scheint dieses Ergebnis mit Dahls Resultaten übereinzustimmen..— Ähnliche Probleme, die eine systematische Analyse brauchen, gibt es allerdings auch bei den „Situationskulissen" (thematische Adverbialia) und rhematischen Adverbialien. 2.13 Das Verhältnis der T R G zur dritten Dimension (zur „Größe" der Einheiten) bietet ebenso wichtige und zahlreiche Probleme, wie bei der ersten Dimension. Hier können wir aber nichts mehr tun als die Probleme registrieren und zu einem gewissen Maß klassifizieren. Die wichtigste Einheit, die hier in Betracht kommt, ist der Satz (Aussage in der Terminologie von Dokulil und Danes) und die vorhergehenden Erwägungen wurden auch vom Standpunkt des Verhältnisses der T R G zum Satz formuliert. Hier muß man jedoch auch „kleinere" und „größere Einheiten" — und ihre Beziehungen zur T R G — analysieren. Beginnen wir mit den „kleineren", mit den Bestandteilen des Satzes, müssen wir zuerst Fragen der Hierarchie des kommunikativen Dynamismus in Betracht ziehen (vgl. oben über die Arbeiten von Firbas und Svoboda). Zu den wichtigsten Fragen gehört hier das Verhältnis zwischen dieser Hierarchie einerseits und der syntaktischen Struktur der semantischen Repräsentation des Satzes andererseits. Die erste Hypothese, die wir aufstellen können, besteht darin, daß die Grundschicht der T R G mit der syntaktischen Struktur insofern übereinstimmt, daß jedes Kommunikationsfeld (in der Terminologie Svobodas) einer syntaktischen Verbindung

64

P. SGALL

zweier Satzglieder entspricht (in der Terminologie der Abhängigkeitssyntax), oder einer Konstituente des Satzes (in einer Phrasenstrukturgrammatik). Es sollte systematisch nachgeprüft werden, was das „oder" hier bedeutet, denn wahrscheinlich sind die beiden erwähnten Typen von Grammatiken in dieser Hinsicht nicht völlig äquivalent, und ihre Konfrontation könnte vieles nicht nur zur Erkenntnis der TRG, sondern auch zur Bewertung dieser Grammatiken beitragen. Die weitere Frage ist dann, ob auch für die zweite Schicht der TRG (für die der Einfluß der Konsituation charakteristisch ist) ein solches Verhältnis zwischen der Hierarchie der TRG und der Syntax der semantischen Ebene gilt. Eine Hypothese, die mit diesen Erwägungen eng zusammenhängt, ist, daß die Wirkung der Konsituation in dieser Schicht der TRG die Projektivität des Satzes nicht stört. (Der Begriff der Projektivität ist aus der Abhängigkeitssyntax gut bekannt, vgl. z. B. MARCUS, 1965; vom mathematischen Standpunkt s. jetzt L . NEBESKÍ, 1 9 6 9 ; für die Phrasenstrukturgrammatik wäre eine mehr oder weniger parallele Frage, ob die Wirkung der Konsituation zu diskontinuierlichen Konstituenten führt.) In vollem Umfang wird diese Hypothese wahrscheinlich nicht gelten, denn z. B. tschechische Sätze wie „Sportovec je dobry" könnten nur schwierig zur dritten Schicht der TRG zugeordnet werden. Man muß jedoch nachprüfen, ob die Stellung des Verbs in der Oberflächenstruktur hier seiner Stellung in der Hierarchie des kommunikativen Dynamismus wirklich entspricht. Jedenfalls scheinen solche Konstruktionen durch eng begrenzte grammatische Bedingungen bestimmt zu sein, sodaß sie die Annahme einer projektiven Struktur der semantischen Repräsentation der Sätze nicht ausschließen. Ein anderer Fragenkreis, der hierher gehört, betrifft Begriffe wie Deprädikation, backgrounding, condensation (vgl. WEINREICH, 1 9 6 3 , auch die bei SGALL, 1 9 6 9 , 2 3 5 angeführte Literatur). Wenn GRUBER ( 1 9 6 7 , 6 5 ) gefunden hat, daß die TRG im gewissen Sinn in der Entwicklung der Sprache bei einem Kind der SubjektPrädikat-Struktur des Satzes vorausgeht, sollte man systematisch nachprüfen, ob sich die TRG (samt der Hierarchie des kommunikativen Dynamismus) als die Grundlage aller erwähnten syn-

ZUR STELLUNG D E R THEMA-RHEMA-GLIEDERUNG

65

taktischen Beziehungen nicht auffassen ließe (vgl. D A H L , 1969, 28f). In der Terminologie der neueren Diskussion innerhalb der Chomsky-Schule spricht man in einem ähnlichen Sinn vom Unterschied zwischen der Bedeutung des Satzes (die mittels der Satznegation negiert wird) und den entsprechenden Präsuppositionen (die durch die Satznegation semanitsch unberührt bleiben), vgl. Sätze wie „Er war nicht in Prag in der Zeit, als unser Haus gebaut wurde" (es wird nicht negiert, daß das Haus gebaut wurde, sondern nur, daß er damals in Prag war). Es ist notwendig, die verschiedensten Satztypen und den Effekt der Kondensation (backgrounding) mit Rücksicht auf die Semantik der Negation — und auch anderer Operatoren — zu analysieren, um die semantische Relevanz der T R G zu präzisieren. Wenn wir von der oben erwähnten Formulierung ausgehen, daß das Rhema eines Aussagesatzes eben das ist, was von dem Thema behauptet (ausgesagt) wird, ist es nicht überraschend, daß das Thema und das Rhema in der Bedeutung des Satzes als Ganze auftreten, und daß die Beziehungen innerhalb des Themas und des Rhemas zu den Präsuppositionen gehören, wie im Satz „Emmas Bruder ist mein bester Freund" (dessen Negation die Präsuppositionen, daß Emma einen Bruder und der Sprechende einen besten Freund hat, unberührt läßt). Wie HAJIÖOVA (1971; 1974; vorb.) überzeugend gezeigt hat, gibt es auch Fälle, wo eine Beziehung innerhalb des Rhemas im negierten Satz zwar nicht negiert wird (also nicht zur eigenen Bedeutung des Satzes gehört), aber auch nicht unberührt bleibt (wie eine Präsupposition); sie wird durch die Negation vom behaupteten zu einem nur potentiellen Objekt: „Karl wurde nicht dadurch aufgehalten, daß er einen Freund erwartete". Dieser Satz kann in den beiden Fällen wahr sein, wenn Karl einen Freund erwartete, oder nicht. Hier handelt es sich also um einen „dritten Typ" semantischer Einheiten (neben der Bedeutung und der Präsupposition). Wenn jedoch der Nebensatz im positiven Satzgefüge thematisiert wird, hat die Negation einen anderen Effekt: „Dadurch, daß Karl einen Freund erwartete, wurde er nicht aufgehalten." In diesem Satz wird — auch mit Negation —

66

P. SGALL

vorausgesetzt, daß Karl einen Freund erwartete, es handelt sich also um eine Präsupposition. Ob der Unterschied zwischen Präsuppositionen und dem „dritten Typ" im Rhema durch die Art der Kondensation bedingt wird, oder durch die Art der Präsupposition, diese und viele andere Fragen können bei der Analyse der semantischen Struktur des Satzes nicht außer Acht gelassen werden. Probleme einer anderen Art, auf die wir hier nicht eingehen können, sind mit der „Emphase" im weitesten Sinn verbunden (vgl. den Begriff „podöerkivanije", mit dem Apresjan und andere arbeiten, weiter besonders WORTH, 1964, KIEFER, 1967, DAHL 1969, § 6). Andere Probleme bietet die Struktur des Fragesatzes, in dem die TRG in einer mehr nüancierten Beziehung zur semantischen Struktur steht (vgl. z. B. KRÍZKOVÁ, 1968; jetzt vor allem FIRBAS, 1970). Der andere Themenkreis, der die Stellung der TRG auf der Achse der dritten Dimension betrifft, umfaßt, wie oben gesagt, die Beziehungen der TRG zu „größeren" Einheiten als der des Satzes, also zur „Hypersyntax". Für die Verbindung der Lehre über die TRG mit der Textsyntax (vgl. DRESSLER, 1970) hat unter den neueren Arbeiten die Klassifikation der „thematischen Reihen", die DANE§ (1968, 1970) formuliert hat, eine große Bedeutung. Wir möchten hier nur zwei Bemerkungen hinzufügen, die verwandte Fragen betreifen. Wenn man die Synonymie zwischen einem Satz und einer Kette von Sätzen (vgl. SGALL, 1967, 212f) untersucht (samt der Funktion der Wörter wie deshalb, also, trotzdem, des Strichpunktes usw.), sollte man auch die TRG der betreffenden Sätze in Betracht ziehen (in Verbindung mit den „thematischen Reihen"). Das ist notwendig, wenn wir Kriterien finden wollen, die darüber entscheiden, ob und wie die gegebene Kette von Sätzen durch einen einzigen Satz paraphrasiert werden kann (und umgekehrt); eine solche Analyse wird für die verschiedensten Fragen der semantischen Struktur des Textes von Bedeutung sein. Einen anderen Ausgangspunkt für das Studium der Beziehungen zwischen der TRG und der Textstruktur bietet der oben (§2.11) charakterisierte Unterschied zwischen den beiden ersten Schichten der TRG. Während die Grundschicht keine Elemente der Konsitua-

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG

67

tion voraussetzt, ist die zweite Schicht eben durch solche Elemente bedingt. In der neueren transformationalistischen Terminologie kann man sagen, daß ein Satz, dessen TRG zur Grundschicht gehört, eine gewisse Menge von Präsuppositionen enthält, die durch die lexikalischen Elemente und durch die eigene Struktur des Satzes bestimmt werden. Daneben besteht jedoch noch eine dritte Art von Präsuppositionen (vgl. SGALL-HAJIÈOVA, 1970, § 5), die

durch die TRG des Satzes bestimmt werden, nämlich durch die von ihren Elementen, die zur zweiten Schicht gehören. Ein Satz, der zur zweiten Schicht gehört, unterscheidet sich also vom entsprechenden Satz der Grundschicht in semantischer Hinsicht dadurch, daß er zusätzliche Präsuppositionen aufweist; ζ. B. der Satz „Einen Film von Forman habe ich gestern gesehen" unterscheidet sich vom Satz „Gestern habe ich einen Film von Forman gesehen" (wenn sie beide mit normaler Intonation gesprochen werden) dadurch, daß er voraussetzt, daß der semantische Komplex „ein Film von Forman" im „pool of presuppositions" (DAHL, 1970) enthalten ist. 2.2 Mehr Konkretes kann man über die Stellung der TRG im System einer Sprachbeschreibung sagen, wenn man über Typen von Sprachbeschreibungen spricht, die explizit formuliert sind, wo also die Beziehungen zwischen den einzelnen Bestandteilen (Ebenen, Regelmengen usw.) klar ausgeprägt sind und strenger kontrolliert werden können. Wo diese Bedingungen erfüllt sind, ist es schon möglich, eine formale Explikation einzelner syntaktischer und semantischer Begriffe anzustreben. Wir möchten hier ganz kurz zwei von solchen Beschreibungstypen erörtern, nämlich die transformationelle und die stratifikative (funktionelle) Beschreibung. 2.21 In der Transformationsgrammatik wurde der Begriff des Themas (topic) wohl zuerst in Chomskys Aspects of the Theory of Syntax erwähnt, die Unadäquatheit dieser Bemerkung wurde aber bald bemerkt (STAAL, 1967; DAHL, 1969, 1 2 - 1 5 ) . CHOMSKY (1968) und einige seiner Schüler (LAKOFF, 1965, 1969, FILLMORE, 1970, JACKENDOFF, 1969) haben dann neue Formulierungen gesucht. Daneben stehen wichtige Arbeiten, die mit der eigenen Chomsky

68

P. SGALL

Schule nicht so eng zusammenhängen: wir haben schon einige von ihnen erwähnt. Es wird da allgemein anerkannt — wenn auch in verschiedener Terminologie — daß Erscheinungen, die wir zur T R G rechnen, in der semantischen Repräsentation des Satzes wiedergegeben werden sollen. Die Fragen der Operatoren (Quantifikation, Negation) werden oft von denen der eigenen T R G getrennt; wir haben in SGALL - HAJICOVÁ, 1970, § 5, versucht zu zeigen, daß man zu einer einfacheren Beschreibung kommen kann, wenn man die Operatoren mit der T R G verknüpft und ihren Bereich eben durch ihre Stellung in der Hierarchie des kommunikativen Dynamismus beschreibt. Die Anordnung der lexikalischen Elemente in der semantischen Repräsentation des Satzes (oder in der Tiefenstruktur) kann man als Grundlage für das formale Explicatum der Hierarchie des kommunikativen Dynamismus in der Grundschicht der T R G halten, die zweite Schicht der TRG kann mittels Regeln der Thematisierung (topicalization) formal gefaßt (expliziert) werden, und wahrscheinlich sollte man auch mit Regeln der Rhematisierung arbeiten (mindestens für die dritte Schicht der TRG). Es sollte sich jedoch nicht um Transformationsregeln, deren Applikation doch wohl keine semantischen Effekte haben soll (dazu s. auch KATZ, 1970, 222), handeln, sondern um Formationsregeln innerhalb der semantischen Komponente (oder der Basis); die jetzige Form der Phrasenstrukturregeln ist allerdings für diese Regeln nicht geeignet, denn ihre Applikation muß vom Bestand des „pool of presuppositions" abhängen (vgl. hier den letzten Absatz des § 2.13). Außerhalb der Hierarchie des kommunikativen Dynamismus sollte wahrscheinlich die Gliederung des Satzes in Thema und Rhema formal expliziert werden (um nicht nur die Skala, sondern auch die Dichotomie beschreiben zu können). Zu diesem Zweck könnte man die Hypothese von Ross und Sadock benützen, nach der jeder Satz in seiner semantischen Repräsentierung einen „performativen" Matrixsatz von der Form „Ich sage dir S" enthält. Wenn wir auf die enge Beziehung zwischen der T R G und der Syntax der Verba dicendi zurückgreifen (vgl. § 1 und die Formulierungen wie „worüber man spricht = Thema", „was man darüber sagt =

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG

69

= Rhema"), könnten wir eine einfache formale Explikation der Dichotomie darin sehen, daß der Matrixsatz die Form „Ich sage dir über NP, daß S" haben sollte (wobei S ein Element enthalten muß, das zum NP referiert). Es sind jedoch mit dieser Annahme einige Schwierigkeiten verbunden, auf die wir hier nicht mehr ausführlicher eingehen können: es müßte zuerst die Beziehung dieses Matrixsatzes zu den Regeln der Thematisierung (also auch zu Sätzen wie „Daß er sich gut benimmt, habe ich über Karl nicht gesagt") analysiert werden; auch die Deutung der Existenzverba und der Situationskulisse wird hier an Schwierigkeiten stoßen (für den ersten diesen beiden Typen könnte man zwar mit einem „Nullthema" arbeiten, vgl. ζ. B. ADAMEC, 1966, §22, für den zweiten wäre aber auch das keine Lösung). Die Stellung dieses Matrixsatzes in verschiedenen Typen von Satzgefügen ist noch nicht ganz klar (vgl. STAAL, 1970, über Sätze, die mehrere performative Matrixsätze an verschiedenen Niveaus der Einbettung enthalten). 2.22 Was die funktionelle generative Sprachbeschreibung betrifft, in der die Ebenen (ähnlich wie im stratifikativen System von S. M. Lamb) von der semantischen bis zur phonetischen linear geordnet sind, wurde eine Möglichkeit der Beschreibung der T R G zuerst in SGALL ( 1 9 6 7 ) versucht (vgl. jetzt auch SGALL - HAJICOVÁ, 1970). Wir können dazu folgendes hinzufügen: Die Grundschicht der T R G kann in einem solchen System dadurch beschrieben werden, daß drei Teilmengen von Regeln der generativen Komponente unterschieden werden (d. h. von Regeln, die die semantische Struktur der Sätze spezifizieren): (a) Regeln, die Syntagmen mit der Anordnung Regens post Rectum entsprechen (z. B. Agens—Aktion, oder Situationskulisse — Aktion); diese Anordnung wird so interpretiert, daß das regierende Wort eine größere Ladung des kommunikativen Dynamismus trägt als das abhängige Wort im Syntagma; (b) Regeln für Syntagmen mit der Anordnung Rectum post Regens (für Syntagmen wie z. B. „Nastupila vesna", oder „citat' knigu"); (c) Regeln, bei denen beide Anordnungen in der Grundschicht der T R G gleichberechtigt sind.

70

P. SGALL

Es müssen auch Bedingungen hinzugefügt werden (z. B. zur Definition eines Syntagmas), die eine Anordnung der von demselben Wort abhängigen Satzglieder ausschließen, die in der Grundschicht der T R G nicht vorkommt. Die Thematisierung (in der zweiten Schicht der TRG) würde dann dadurch in die Beschreibung eingegliedert, daß — nach einer allgemeinen Regel — die Stellung der Wörter im Typ (a) und (b) umgetauscht werden kann, wenn nur das Wort, das in der Grundschicht nach dem anderen stehen sollte, zu einem Element referiert, das im „pool of presuppositions" enthalten ist. Eine weitere Frage betrifft die Fälle, wo Elemente, die in der Oberflächenstruktur ihre eigene Rolle im Verhältnis zur T R G spielen (die in der Wortfolge oder Intonation verschiedene Positionen haben können), in der semantischen Repräsentation dagegen nur als Bestandteile von komplexen Symbolen auftreten. Für diese Fälle ist es nicht ausreichend, die Hierarchie des kommunikativen Dynamismus mittels der Anordnung der Symbole der semantischen Repräsentation zu beschreiben. Wie BENESOVÁ (vorb.) am Beispiel der modalen Verben zeigt, kann hier die Thematisierung und die Rhematisierung solcher Elemente mit Hilfe von spezieller Symbole bezeichnet werden, die dann beim Übergang zur Oberflächenstruktur (morphematischen Ebene) die Stellung der betreffenden Ausdrucksmittel in der Wortfolge und (oder) ihre Intonation bestimmen. Eine solche Beschreibung sollte vielleicht für die Rhematisierung auch in anderen Fällen (auch bei semantisch selbständigen Wörtern) angewendet werden. (Was die Thematisierung betrifft, ist es noch nachzuprüfen, ob sie die unselbständigen Elemente wirklich betreffen kann.) Auf diese Weise könnte man viele unprojektive Wortfolgetypen in die Beschreibung ohne grundsätzliche Schwierigkeiten eingliedern (denn die Abweichungen von der Projektivität betreffen dann nicht die höheren Ebenen, und auf der morphematischen Ebene arbeitet man mit keinem Abhängigkeitsbaum, sodaß hier die Frage der Projektivität irrelevant ist). Diese Beschreibungsweise kann dann adäquat sein, wenn die am Ende des § 2.11 formulierte Hypothese gilt. Ob man in einer Sprachbeschreibung, die mit der Abhängigkeitssyntax arbeitet,

ZUR STELLUNG DER THEMA-RHEMA-GLIEDERUNG

71

auch mit einem performativen Matrixsatz rechnen kann, ist noch nachzuprüfen. Auch die Darstellung der Präsuppositionen wurde bisher nicht formal erfaßt. Für jede Sprachbeschreibung gibt es mehrere Fragen, die erst empirisch gelöst werden müssen (vgl. besonders die im § 2.11 und 2.13 erwähnten Probleme), d. h. in denen man das Niveau der „deskriptiven Adäquatheit" noch nicht erreicht hat. In der Bewertung der einzelnen Beschreibungstypen muß man also heute auch das Maß berücksichtigen, in dem jeder Beschreibungstyp dazu verhilft, für die empirische Forschung relevante Fragen klar zu stellen. 3. Zum Abschluß können wir unsere vorläufigen Entwürfe folgendermaßen zusammenfassen: Als das Explicandum (den präsystematischen Inhalt des Termins „Thema-Rhema-Gliederung") nehmen wir den Unterschied zwischen dem, wovon man spricht (dem Thema, bzw. auch der Situationskulisse), und dem, was man darüber sagt (Rhema). Dieser Unterschied ist vielleicht jeder (zweigliedrigen) Prädikation eigen; bei den mehr oder weniger deprädikationsartigen (kondensierten) Konstruktionen und Formen muß man jedoch daneben auch weniger auffällige Stufen des kommunikativen Dynamismus in Betracht ziehen. Als eine Grundlage für das entsprechende formale Explicatum schlagen wir zwei Beziehungen vor: einerseits die Anordnung, die man „Wortfolge der Tiefenstruktur" (oder genauer die Folge der Elemente der semantischen Satzstruktur) nennen könnte, andererseits den Unterschied zwischen den beiden Objekten des performativen Matrixsatzes: „Über Obj t sage ich dir (hier jetzt), daß Obj 2 " (wo Obj 2 ein Satz ist, der als eines seiner Elemente eine Kopie von Obj t enthält); dabei entspricht Obj t dem Thema, Obj 2 dem Rhema und die erwähnte Anordnung der hierarchischen Skala des kommunikativen Dynamismus. Es handelt sich allerdings nur um Vorschläge, deren Berechtigung in mancher Hinsicht noch nachgeprüft werden muß und deren explizite Formulierung noch nicht vorhanden liegt.

72

P. SGALL

LITERATURVERZEICHNIS ADAMEC, P., Porjadok slov ν sovremennom russkom jazyke. Praha 1966. BENES, E., On Two Aspects of Functional Sentence Perspective, TLP 3, 1968, 267-274. BENES, E., Über zwei Aspekte der funktionalen Satzperspektive. In: Actes du Xe Congrès international des linguistes II, Bucarest 1970, 1021 — 1062.

BENESOVÂ, E., (vorb.). Nëkteré otázky aktuálního ôlenëni a slovosledu ν ôeâtinë (Einige Fragen der aktuellen Satzgliederung und der Wortfolge im Tschechischen], AUC, Slavica Pragensia 13, 1971, 179—198. BREKLE, H. E., Generative Satzsemantik und transformationeile Syntax im System der englischen Nominalkomposition. München 1970. CHOMSKY, N . , Deep Structure, Surface Structure and Semantic Interpretation. (Vervielfältigt.) 1968, jetzt in CHOMSKY (1972), 62—119. D A H L , Ö . , Topic and Comment: A Study in Russian and General Transformational Grammar. Göteborg 1969. D A H L , Ö., Some Presuppositions about Presuppositions. (Vervielf.) 1970, DANES, F., Order of Elements and Sentence Intonation. In: To Honor Roman Jakobson, The Hague 1967, 499—512. DANES, F., Typy tematickych posloupnosti Ν textu [Types of Thematic Progressions in Texts], SaS 29, 1968, 125-141. DANE§, F., Some Thoughts on the Semantic Structure of the Sentence, Lingua 21, 1968a, 5 5 - 6 9 . DANES, F . , Zur linguistischen Analyse der Textstruktur, Folia Linguistica 4 , 1970, 4 9 - 5 6 .

tzv. vyznamové a mluvnické stavbë vëty [Zu der sogenannten semantischen und grammatischen Satzstruktur]. In: O vëdeckém poznáni soudobych jazykû, Prag 1958, 231 — 246. DRESSLER, W., Modelle und Methoden der Textsyntax, Folia linguistica 4, 1970. FILLMORE, J . C . , Subjects, Speakers and Roles. In: Working Papers in Linguistics 4, 1970, 31 — 63, Ohio State University, jetzt in Synthese 21, 1970, 251-274. FIRBAS, J., On Defining the Theme in Functional Sentence Analysis, TLP 1, 1964, 267-280. FIRBAS, J., On the Function of the Question in the Act of Communication, 1970 (vervielfältigt für das Seminar über komplexe grammatische Systeme, Cambridge, Mass.), jetzt russisch in Voprosy jazykoznanija 1972, No. 2, 5 5 - 6 5 . FIRBAS, J., On the Interplay of Means of Functional Sentence Perspective. In: Actes du Xe Congrès internat, des linguistes II, Bucarest 1970a, 741-745. DOKULIL, M . - DANES, F . , Κ

ZUR STELLUNG DER

THEMA-RHEMA-GLIEDERUNG

73

F. M., (vorb.), handschriftliche Arbeit über Fragen der Intonation im Russischen. G R U B E R , J . S . , Topicalization in Child Language, FL 3 , 1 9 6 7 , 3 7 — 5 6 . H A J I Í O V Á , E., 1971. Some Remarks on Presupposition, vorgetr. an der Konferenz "Computational Linguistics", Debrecen; auch in PBML 17, 1972, 11-23. HAJICOVÁ, E., 1974. Meaning, Presupposition, and Allegation, PhP 17, No. 1. HAJICOVÁ, E. (vorb.), Negace a presupozice ve vyznamové stavbë vëty (Negation und Präsupposition in der Satzsemantik), Academia, Prag. H A L L I D A Y , M . A . K . , Notes on Transitivity and Theme in English, JL 3 ,

GLADNEY,

1967, 3 7 - 8 1 , 1 9 9 - 2 4 4 ; 4, 1968,

179-215.

A. G., Syntax and the Sentence, Word 12, 1956, 234—250. H E I D O L P H , K. E., Kontextbeziehungen zwischen Sätzen in einer generativen Grammatik, Kybernetika 3, 1966, 274—281. H O R Á L E K , Κ . , Filosofie jazyka. Prag 1967. HoREjäf, V., Les plans linguistiques et la structure de l'énoncé, PhP 4,1961, 193-203. JACKENDOFF, R. S., An Interpretive Theory of Negation, FL 5, 1969, 218-241. JAKOBSON, R., Beitrag zur allgemeinen Kasuslehre, TCLP 6, 1963, 240—288. K A T Z , J . J . , Interpretative Semantics vs. Generative Semantics, FL 6 , 1 9 7 0 , HATCHER,

220-259.

F., On Emphasis and Word Order in Hungarian. Bloomington 1 9 6 7 . Tázací vëta a nëkteré problémy tzv. aktuálního (kontextového) dlenëni [Der Fragesatz und Probleme der aktuellen Gliederung], Nase fee 51, 1968, 2 0 0 - 2 1 0 . LAKOFF, G . , Passives, Adverbs and Quantifiers. In: Mathematical Linguistics and Automatic Translation, 1965, Rep. NSF-16, F-l —F-48. LAKOFF, G., On Derivational Constraints. In: Papers from the 5th Reg. Meeting, Chicago Ling. Society, 1969, 117—139. M A R C U S , S., Sur la notion de projectivité, Zeitschr.f. math. Logik u. Gründl, d. Math. 11, 1965, 1 8 1 - 1 9 2 . MATHESIUS, V . , Obsahovy rozbor soucasné angliètiny na zàkladë obeenë lingvistickém [A Functional Analysis of Present Day English on General Linguistic Basis], herausg. v. J. Vachek, Prag 1961. SEUREN, P. A. M., Operators and Nucleus. Cambridge 1969. SGALL, P., Functional Sentence Prespective in a Generative Description, PSML 2, 1967, 2 0 3 - 2 2 5 . SGALL, P., L'ordre des mots et la sémantique. In: Studies in Syntax and Semantics, Dordrecht 1969, 231 — 240. SGALL, P. - HAJICOVÁ, E., 1970. A "Functional" Generative Description, PBML 14, 3 - 3 8 . STAAL, J. F., Some Semantic Relations between Sentoids, FL 3,1967, 66—88. KIEFER,

KRÍZKOVÁ, H . ,

74

P. SGALL

ST AAL, J. F. .Performatives and Token-Reflexives, Linguistic Inquiry 1,1970. No. 3, 373-381. SVOBODA, Α., The Hierarchy of Communicative Units and Fields as Illustrated by English Attributive Constructions, BSE 7, 1967, 49—101. TRNKA, B.: On the Linguistic Sign and the Multilevel Organization of Language, TLP 1, 33—40. UHLÍftovÁ, L., (vorb.), Kvantitativni rozbor vëty a vypovëdi ν ôeStinë. [Quantitative Analyse des Satzes und der Aussage im Tschechischen], 1970, handschriftliche Dissertation über statistische Bezeihungen der TRG zur syntaktischen Struktur des Satzes (Institut der tschechischen Sprache, Tschechosl. Akad. d. Wiss.). WEINREICH, U., On the Semantic Structure of Language. In: Universals of Language (J. H. Greenberg, ed.), 1963, 114—171, Cambridge, Mass. WORTH, D. S., Suprasyntactics. In: Proceedings of the Ninth International Congress of Linguistics, 1964, 698—704, London—The Hague — Paris. ZATORSKI, R. J., Grammatical Relations and the Structure of the Categorial Aspects-Type Base, Linguistics 55, 1970, 70—81. ZEMB, J. M., Les structures logiques de la propositoin allemande. Paris, 1968.

TOPIC - COMMENT STRUCTURE IN A GENERATIVE GRAMMAR WITH A SEMANTIC BASE Osten Dahl (Göteborg)

During the last few years, the adherents of generative grammar have split into a few major "schools". One of these, headed by linguists such as G. Lakoff, J. McCawley and J. R. Ross, calls itself "Generative Semantics". The most important theoretical assumption that characterizes this school is the identification of the semantic representation of logical form of a sentence and its deepest underlying grammatical structure. What I have tried to do is to integrate the description of topic-comment structure into a theory of this type. The following points sum up the main claims that I want to make: 1. It is necessary to assume that topic-comment structure is to be accounted for in the semantic representation of the sentence and is not merely a surface phenomenon. 2. Topic-comment structure can be regarded as the reflection of some fundamental aspects of the semantic representation or logical form of the sentence, as outlined below: In the semantic representation of a sentence, there are normally two main parts, the topic, where we name or define a set or an individual, and the comment, which is a propositional function "predicated" about this set or individual. Thus, I would argue that the logical forms of sentences are best rendered by a notation similar to that used in set theory, i.e., the sentence (1) My brothers are drunkards might be quasi-formally represented as (2) Let A = {χ I χ is my brother}. Then, for every χ such that χ is in Α, χ is a drunkard.

76

ö . DAHL

Some arguments for these claims are the following: 1. In languages like Russian, where topic-comment structure is one of the major factors for determining intonation and word order, it is quite striking to what extent one can match logical forms and "types of topic-comment structure" (what Adamec 1966 calls aktual'no-sintaksiceskie tipy). To take some simple examples, the sentence KoroV — bogac "The king is a rich man" defines an individual, "the king", and assigns to him the property "to be a rich man", whereas Bogac — korol' defines an individual, "the rich man" and assigns to him the property "to be King". Kniga — na stole "The book is on the table" defines the individual entity "the book" and predicates about it the property "to be on the table", whereas Na stole — kniga defines a place, "on the table", and asserts the existence of a book in this place, Èensciny rabotajut "the women work (are working)", defines a class, "the women", and asserts about this class the propositional function "the members of χ work". Rabotajut zensciny "It's the women who work", on the other hand, defines the class {χ | works} and assigns to the members of this class the property "x is a woman". Noun phrases in topic position are typically definite or generic, which exactly corresponds to the individual and set descriptions we have talked about, whereas noun phrases in comment position are often what I have called "existential indefinites", i.e. assert the existence of the entity denoted by the NP, e.g. knigu in the Russian sentence Ja kupil knigu " I bought a book". What have been called "topicless sentences" are mostly sentences having the logical form of an existential proposition, e.g. Suscestvujut belje medvedi "There exist white bears", which do not contain set descriptions of the type talked about above. 2. Consider the following sentences: (3) (a) Only I love my wife. (b) My wife is loved by me only. (4) (a) Zena rugaet Ivanova — Menja toze. "Ivanov's wife scolds him — Me too" (b) Ivanova rugaet zena — Menja toze. "Ivanov is scolded by his wife — I too"

T O P I C — C O M M E N T I N A GENERATIVE G R A M M A R

77

The (a) sentences in each pair are ambiguous — e.g. (3a) can mean (5) (a) I am the only person who loves his wife (b) I am the only person who loves my wife The (b) sentences are unambiguous — (3b) has only the interpretations (5b). The explanation I would give for these facts is approximately this: Sentence (3a) expresses a relation between the individual I and a set, either the set "persons who love their wives" or "persons who love my wife". The relation is being the only element of the set in question. Now, the condition for this ambiguity is that the description of the individual that is referred to by the noun phrase my wife be contained within the other set description. If the theory about topic-comment structure that I propose is correct, the noun phrase my wife, if it is topic of the sentence, would necessarily be outside this set description and the semantic representation of (3b) would look like the following: (6) Let A = {χ I χ is my wife}: Then if Β = {y | y love A}, I am the only element of B. As we see, in a sentence of this logical form, the ambiguity would be impossible. 3. Sgall 1970 considers basically sound the old idea that topic is what the sentence is about and that the comment is what is said about the thing named in the topic. I certainly agree with this, though I am sceptical about the representation he proposes, where the topic is an argument of the performative assumed to be the topmost S of every semantic representation, thus, (7) About χ I tell you that S Notice that this implies that the about-relation is a primitive. I would rather like to regard it as derived from the notion of predication, where I take "predication" to denote the types of logical forms illustrated above (e.g. (6)). Notice that taking "about" as a primitive we still have to explain why only definites and generics can be topics. This is not at all clear from meaning of "about", as we can very well have indefinites after it, e.g. (8) I am talking about a man I met last year. I would not exclude, though, that it might be possible to derive the

78

ö. DAHL

constraint in question from the special properties of performative sentences. The idea that the topic is what the sentence is about is however interesting from slightly other points of view. It might be thought that in a sentence with several noun phrases it would be more or less arbitrary to say that a certain of these NPs is what the sentence "is about". However, it is possible to construct tests where the intuitions of speakers are fairly consistent in this respect. One such way of learning which N P in a sentence is what the sentence "is about" is illustrated by the following examples: Consider sentences of the form NP Verb NP, like (9) John dislikes Bill. Logicians usually say that such sentences have the logical form (10) Fab where F is the relation "dislike", and a and b are individual constants referring to John and Bill. If we accept this, there would really be no reason to say that the sentence is about John rather than about Bill. Let us now add the clause ... and the same holds true for Harry to (9). Then, if (9) can be said to "be about" Bill as well as about John, the resulting sentence (11) John dislikes Bill, and the same holds true for Harry would be unambiguous, however, most speakers find it unambiguous. The same in (11), for them, can refer just to the propositional function "— dislikes Bill". Consider now the same sentence with topicalized object: (12) Bill John dislikes, and the same holds true for Harry (12) turns out also to be unambiguous, but this time the same can refer only to the propositional function "John dislikes — ". The conclusions to be made seem to be the following: The logical notation Fab is not adequate to render the meaning of sentences like (9) or (12). The semantic representation of these sentences must in some way make clear that (9) is about John, whereas the first clause of (12) is about Bill. How the assumed semantic representations are to be linked with the respective surface manifestations of the sentences is rather unclear. It is usually considered that we have a basic word order that

TOPIC — COMMENT IN A GENERATIVE GRAMMAR

79

is modified by various transformations like topicalization and passivization, as a result of which different topic-comment structures arise. The hypotheses I have put forward imply, however, that the surface word order is a reflection of the hierarchical organization of the underlying structure. This seems to imply in its turn that the idea of transformations that modify topic-comment structure is wrong. However, it is rather hard to state the rules that we would have instead. Moreover, the transformational character of at least some of the rules assumed in the earlier theory seems to be rather indisputable. I would agree with Sgall's (1970) claim that the problem of representing topic and comment and the problem how to describe quantifiers in a grammar are really the same thing. Lakoff (1969) has proposed a theory of derivational constraints to account for the fact that the surface order of quantifiers normally coincides with the order of quantifiers in the semantic representation. It might be possible to extend this type of solution to the description of topiccomment relations. This would mean that we would have a general constraint on derivations to the effect that a derivation is wellformed only if the order of the surface elements is the same as the order of the elements in the underlying structure. Thus, we might have cases where the original order is destroyed by some rule such as NP-lowering (that is, the assumed rule that moves down the set and individual descriptions into the propositional functions) and then restored by a later rule, e.g. Topicalization. This hypothesis certainly needs further qualification and empirical support. REFERENCES ADAMEC, P., Porjadok slov ν sovremennom russkom jazyke. Rozpravy Ceskoslovenskê akademie vëd, rada spoleòenskych vëd, Praha 1966. D A H L , Ö . , Topic and Comment. A Study in Russian and General Transformational Grammar. Slavica Gothoburgensia 4. Acta Universitatis Gothoburgensis 1969. Distr.: Almkvist and Wikseil, Stockholm. DAHL, ö . , Studies in Russian and General Transformational Grammar, Göteborg 1970. (Thesis in summary form, available from the author.)

80

Ö. DAHL

DAHL, Ö., (in press). On Sets and Propositional Functions in Grammar. To appear in Scando-Slavica 16. Munksgaard, Copenhagen. LAKOFF, G., On Derivational Constraints. In: Papers from the 5th Regional Meeting of the Chicago Linguistic Society (Binnick et al., ed.). Dept. of Linguistics, University of Chicago 1969. SGALL, P., Zur Stellung der Thema-Rhema-Gliederung in der Sprachbeschreibung, 1970, pp. 54—74 in this volume.

TWO PROBLEMS OF TOPIC - COMMENT László Dezsö - Görgy Szépe (Budapest)

1. ON THE UNIVERSAL CHARACTER OF TOPIC-COMMENT RULES

1.1. Let us ask the following question: is the topic-comment relation universal? The universal character of Topic - Comment has been mentioned by Ch. F. Hockett: "Every human language has a common clause type with bipartite structure in which the constituents can reasonably be termed 'topic' and 'comment'. 1 If we accept that the topic-comment is universal we have to ask whether it is a formal or substantive universal. I

Our tentative answer is the following: (i) topic - comment is a substantive universal of semiotical order, i.e. covers an area where language falls into the same class with zoo-semiotical systems and with secondary sign systems (represented by arts, games etc.). The language specific topic - comment relation is thus ipso facto a substantive human language universal, of course, but there are some major open questions in this respect. Inasmuch as none of the topiccomment-like notions (psychological subject, psychological predicate; actual division of sentences, topic - comment) was born within the framework of a grammatical theory, t o p i c - c o m m e n t t h e o r y needs a c o m p l e t e r e f o r m u l a t i o n w i t h i n any d e f i n e d g r a m m a t i c a l t h e o r y (such as generative theory preferred by us). But this requirement is doubleedged, so we claim HOCKETT, CH. F., The problem of universale in language. In: Universali of language (J. H. Greenberg, ed.), Cambridge, Mass. 1963. 1

82

L. DESZÓ - G. SZÉPE

that any g r a m m a t i c a l t h e o r y needs r e f o r m u l a t i o n if it t r i e s to cover the t o p i c - c o m m e n t r e l a t i o n . For the moment let us assume that the deep structure is a well defined subset of rules. Then let us assume another subset of rules, that of the syntactic transformations. Now, if we look closer at the syntactic transformations then we find the following apparently isolated minor characteristics — among other major ones: (i) some rules are order-oriented, i.e. their only effect is upon the order of the elements of the string, and (ii) some rules may be divided into two parts: (a) substitution of symbols; (b) re-arrangement of symbols. The ordering and reordering are language-specific, but their mere existence (with some few features) is universal. Saussurean linguistics stressed the linear character of the verbal sign systems. Now in our framework we must consider that: (i) surface structure is the outcome of a linearization; so s u r f a c e s t r u c t u r e is an a l r e a d y l i n e a r i z e d s y m b o l system; (ii) furthermore linearization is carried by the syntactic transformations; (iii) t h a t p a r t of the g r a m m a t i c a l m e t a - t h e o r y which r e g u l a t e s the general o r d e r i n g a s p e c t of the s y n t a c t i c t r a n s f o r m a t i o n s is f i l l e d by the t o p i c - c o m m e n t rules; (iv) must be somehow connected with more general rules valid for all kinds of sign systems. 1.2. We have to differentiate three groups of rules. 1. Primary ordering rules applied to deep structure and having as an output the primary order of sentence corresponding roughly to the Subject-Predicate division, e.g: Péter (topic) irja a levelet (comment) "Peter is writing the letter." or Péter (topic) levelet ir (comment) "Peter is letter-writing". 2. Context rules adjusting word order to the context, if it is necessary, e.g.: A levelet (topic) Péter irja (comment)

TWO PROBLEMS OF TOPIC — COMMENT

83

In this sentence it must be that the word levél "letter" has been mentioned in the foregoing context. Az utcân âllt (topic) egyfiú

(comment)

"There was a boy standing in the street" (i.e. "In the street there stood a boy") where utca "street" is known anáfiú "boy" is novel; and 3. emphatic rules, which may be applied after both primary and contextual rules; e.g.: Péter

ir ja a

levelet

"It is Peter who is writing the letter" or A levelet

ir ja

Péter

"It's the letter that Peter is writing" where the word levelet "letter" is both mentioned and stressed. These examples may be misleading and insufficiently developed as they stand, but the complexity of the stress, focus and topiccomment rules and the tight interwoven character of the entire system make it impossible to thoroughly specify the working of the rules at this point. 2 Rather we wish to continue on to a study of the ontogenesis of topic - comment. 2. ON THE GENESIS OF TOPIC - COMMENT RULES

2.1. For a proper understanding of the genesis of the topic-comment relation, it would be advisable to study not only the ontogenesis of topic-comment in child language but also the possible sources of this relationship in animal communication. The former will however occupy our present interest. First among the questions to be considered is the relation of the above stated order of derivation of the three sets of topic - comment rules and the sequence of their acquisition in the language learning of children. If we as2 For details see DEZSÖ, L., SZÉPE, GY. Adalékok a topic-comment problémához. Nyelvtudományi Kôzlemények 69 (1967), 356—388, its English version: Contribution to the topic-comment problem. Papers in Text linguistics (Ed. by ö . Dahl) (forthcoming).

84

L. DESZÔ - G . SZÉPE

sume that the three sets of tules apply in the following order during the course of derivation: (1) primary rules, (2) context rules, and (3) emphatic rules, we can pose the following questions: Does it also necessarily follow that they are logically connected to each other in a similar order? Should the primary rules in the language of children attain priority over the context and emphasis rules? Thus do they (i.e. the primary rules) appear earlier in their speech? In the earliest infantile monomorphemic utterances, we find that comment plays a predominant role while topic is either evident in the communicative context or otherwise excluded from linguistic expression. Two examples may be adduced from a monolingual Serbo-Croatian speaking child of 15 months who says: bu-bu (ljubi) "like", i.e. "I like my game", the other element of comment is expressed by gesture or bibe (cipele) "shoe" 3 i.e. "Give me the shoe", the other element of comment is expressed by gesture. At this stage one cannot speak of a system of rules governing the topic-comment relationship, however the child already evidences a system which may be characterized as a set of proto-primary rules, since it is from these rules that the primary rules later should develop. Paralleling the development of the basic rules, but staggerred at later times in terms of a developmental sequence, we find the appearance of context and then emphasis rules. In order to form a proper understanding of these rules we must also take into consideration the characteristics of the speech situation and the elocutionary act. The primary topic - comment rules are to be applied to the structures represented in terms of the components of the deep structure which may also be characterized by such terms as Agent, Locative etc. The exact nature of these components however, in the earliest 3 For a more detailed analysis of these examples see Vlahovió, P. — MikeS, M. - Dezsô, L., Développement des constructions de complément d'objet et de lieu dans le langage des enfantes serbocroate et hongrois. Colloquium paedolinguisticum (Ed. K. Ohnesorg), The Hague 1972, 260—269.

TWO PROBLEMS O F TOPIC — C O M M E N T

85

phases of language acquisition must be reconsidered to some extent, for we cannot assume as an a priori fact that the child embodies the adult deep structure component system in toto. The application of topic-comment rules can be detected in various word order and stress regularities of the surface structure. 2.2. We must also investigate into the nature of the relation between topic - comment rules and the surface structure categories of the sentence "Satzglieder" such as Subject and Object. The primary topic - comment rules seem to occur earlier in the acquisition of language than the formation of such categories of surface structure as Subject and Object. It is possible to speak of underlying Agent and Goal in child language before the acquisition of morphological (or strictly ordered sequential) means of representation of the surface categories Subject and Object. For example: a 16month-old bilingual (Hungarian and Serbo-Croatian speaking) child says apu tiitü ("father water") in the context where the father is to bring the child water. Thus the adult Hungarian translation would be: Apu, hozz vizet "Father bring me some water". 4 In this utterance the underlying deep structure components are not expressed through formal means such as morphological marking in the surface structure. Nonetheless understanding or interpretation of the sentence is only possible through the postulation of deep structure categories of Agent and Goal (or Object). If we rely only upon the information of the surface structure data, we can interpret this sentence in a variety of different ways, for instance "Father is drinking water", "Father is in the water", or "Father is being sprayed by water". However, we are limited by accurate perception of the elocutionary act and of the speech situation in our selection of possible underlying categories. We have focussed our attention only upon the sequence of acquisition of the topic - comment rules and have not fully considered 4

F o r a m o r e detailed analysis see MikeS, M . - D e z s ô , L. - V l a h o v i c , P.,

Sentence-programming span in child language. Colloquium paedolinguisticum (Ed. K. Ohnesorg), The Hague 1972, 166-178.

86

L. DESZÖ - G. SZÉPE

the various means available in the surface structure for the expression of these rules such as word order, case markings, agreement between verb and noun and various larger sentence constructions.

FUNKTIONELLE SATZPERSPEKTIVE UND TEXTTHEORIE Wolfgang Dressler (Wien)

1. Die Theorie der FSP wird von vielen Textlinguisten vernachlässingt.1 Dies muß jeden schmerzlich berühren, der, wie der Berichterstatter,2 durch das Studium der FSP zur Textlinguistik gelangt ist. Dennoch möchte ich diesen Bericht nicht dazu benützen, in erster Linie die verkannte Bedeutung der FSP hervorzuheben, denn erstens haben dies berufenere Linguisten schon oft getan; zweitens darf ich hier eine Vertrautheit mit den Arbeiten zur Rolle der FSP in den übersatzmäßigen Bezügen voraussetzen. M. E. ist es für die Forschung nützlicher, wenn ich die FSP einer möglichst gründlichen und herausfordernden Kritik unterziehe, wobei ich über P. SGALLS (1967,203 - 8; 1969,67f. ) Kritik einigermassen hinauszugehen suche. Der Standpunkt, von dem aus ich kritisiere, ist einerseits der der Prager Schule, andererseits der einer Sprachbetrachtung, die wie die verschiedenen Schulen der generativen Semantik,3 P. SGALLS ( 1 9 6 9 ) Modell, die stratifikationelle Grammatik (cf. SGALL 1969; 4 TABER 1966) und entwickeltere Modelle der Dependenzgrammatik von der Bedeutung ausgehen, ihr die Syntax unterordnen und die Oberflächenstrukturen beobachtbarer Sätze durch Transforma1 Wertvolle Förderung verdanke ich Diskussionen nach textlinguistischen Vorträgen in Bratislava, Wien, Salzburg, Berkeley und im slawistischen Seminar an der University of California, Los Angeles. 2 DRESSLER, W., Studien zur verbalen Pluralität. Wien 1968, 30f. 3 Vgl. die Bibliographie von H . K R E N N - K . MÜLLER, Lingu. Berichte 5 , 1970, 8 5 - 1 0 6 . 4 Cf. HERINGER, J., Neuhochdeutsche Syntax. München 1970.

88

W. D R E S S L E R

tionen aus semantischen Basisstrukturen voraussagen.5 Dementsprechend versuche ich auch in der Terminologie die lingua franca der generativen Grammatik mit den Termini der FSP zu harmonisieren: So werde ich in folgendem die FSP-Termini „Thema" und „Rhema" in derselben Bedeutung verwenden wie M. H A L L I D A Y (1967, 1968) in seinen brillanten „Notes on transitivity and theme in English" „given" ( = Thema) und „new information" ( = Rhema); andere sagen dafür „topic" und „comment". Thema ist das aus dem Text Gegebene, Rhema das Neue, das der Sprecher zur Weiterführung des Textes beisteuert. Andererseits verwende ich „Topik" satzgrammatisch im Sinne der generativen Grammatik; andere Termini dafür sind „Thema" (Halliday, Ammann, Boost, Trávníóek), logisches Subjekt, so E. B E N E S (1968, 269) und nach ihm gelegentlich J. F I R B A S (1964, 273-6). M. E. müssen „Thema" und „Topik" unterschieden werden, wenn man Textlinguistik sprachimmanent betreibt. 2.1. Die bisherige textlinguistische Untersuchung der FSP rankt sich um den Begriff der „Suprasyntax". In der Suprasyntax studiere man die Ebene der Äußerung (utterance) als minimaler Kommunikationseinheit, die über dem Satz stehe.6 Dabei wird jedoch meist nicht klar geschieden, ob mit „über dem Satz" die hierarchisch höhere, ausgedehntere Text-Ebene gemeint ist, oder die suprasegmentale Ebene,7 oder die kommunikative Ebene, oder gar die Ebene der Stilistik.8 Damit hängt innig der bereits von P. S G A L L (1967, 206, 213, Fn. 12) bemerkte Fehler zusammen, suprasyntaktisch als supra- oder extragrammatisch aufzufassen, und weiters auf der Ebene der Suprasyntax nicht zwischen „langue" und „parole" zu unterscheiden.9 Wenn jedoch die Suprasyntax eine sprachliche Ebene darstellt, 5

Vgl. auch

6

TRNKA, B „

1965, 1966a,b. PhPA, 1 9 6 1 , 1 3 0 , cf. TLP 1 , 37f; DANES 1 9 6 4 ; BENES 1 9 6 8 ; J . FIRBAS spricht passim von „act of communication". 7 Nach WORTH, D. S., Proceedings of the Ninth Intern. Congress of Linguists, 698ff. 8 VACHEK, J., In: Symbolae Kurylowicz, 1965, 349; HAUSENBLAS 1964. 9 Explizit TRNKA (Fn. 6), vgl. auch HAUSENBLAS 1964. KOCH

FS Ρ U N D TEXTTHEORIE

89

so ist die Unterscheidung zwischen „langue" und „parole" (oder ähnlich Kompetenz und Performanz) unerlässlich. Soll sie eine kommunikative Ebene darstellen, so muß auch dort Kompetenz und Performanz auseinanderhalten werden, vgl. K. Pike's Begriff des „Behaviorems". Manche Textlinguisten unterscheiden daher konsequent zwischen „Textern" und „Textvorkommen" (KOCH 1966b; HARWEG 1968). Idealziel der Textgrammatik muß es demnach sein, durch textematische Regeln alle möglichen (korrekten) Textvorkommen einer gegebenen Sprache voraussagen zu können. 2.2. Mit der mangelhaften textologischen und grammatikalischen Einordnung der FSP hängt auch die Praxis zusammen, daß alle V e r t r e t e r d e r F S P , a u ß e r DANE§ (1968) 1 0 u n d PALEK (1968),

in der Regel nur isolierte Sätze betrachten. 10 " Ganz im Sinne der Satzgrammatik sprechen sie von der Parole-Einbettung der FSP in wechselnde Kontexte; diese sprachlichen Kontexte werden aber weder untersucht noch auch meist zitiert. Die FSP ist jedoch wenigstens zum Teil textgrammatisch bedingt. Daher müssen sowohl „means" als auch „ends" der FSP im Rahmen des oft zitierten Prager „functional approach" textlinguistisch untersucht werden. Ja selbst Satzgrammatik sollte man heutzutage nur mehr in konstanten Textrahmen analysieren. 3.1. Wenn wir den Beitrag der FSP zur Texttheorie bewerten wollen, so müssen wir fragen: Welche Anforderungen dürfen wir an eine linguistische Theorie stellen? Die kürzeste Antwort mag lauten: Eine linguistische Theorie muss nach theoretischer, empirischer und applikativer Adäquatheit streben. 11 Um das letzte, die applikative Adäquatheit kurz zu streifen, die heute im Zeitalter der „Relevanz" immer wichtiger wird, so kann man m. E. in den Arbeiten zur FSP kaum mehr praktische Anweisungen zur Übersetzung etwa der textbedingten Wortstellung oder Intonation lesen als anderswo. 10 10a

11

Der leider nicht zwischen Thema und Topik unterscheidet. Vgl. schon N . FRANCIS, Lg 42, 1966, 149.

Cf. POPPER, K„ Logik der Forschung3. Tübingen 1966; HABERMAS, J., Philosophische Rundschau, Beiheft 5, 1967.

90

W. DRESSLER

3.2. Nun zur theoretischen Adäquatheit: Abgesehen von P . SGALL

haben sich die Vetreter der FSP kaum darum bewußt gekümmert. Wenn etwa nach DANE§ (1964) zwischen den Ebenen der Suprasyntax, Grammatik und Semantik geschieden wird, so werden keine logisch konsistenten, explizit dargestellten deduktiven Hypothesengebilde anvisiert, in denen die Beziehungen dieser drei Ebenen zu einem kohärenten System zusammengefügt werden. Die textologisch wertvollen Beobachtungen der Vertreter der FSP sind eben weitgehend empiristisch. 3.3. Ist deswegen die Theorie der FSP empirisch adäquat? Eine adäquate Theorie muß intersubjektiv überprüfbar bzw. falsifizierbar sein. Dies sind die Hypothesen der FSP größtenteils noch nicht, was in Anbetracht der theoretisch unzureichend definierten Begriffe auch nicht verwunderlich ist. Schon Sgall 12 hat die mangelnde Abgrenzung zwischen Thema und Rhema kritisiert, aber auch die „transition" (vgl. besonders FIRBAS 1965, 171) ist unscharf umrissen. 4.1. Als Mittel der Zuordnung eines Satzteils zu Thema oder Rhema wird oft das Kriterium der Frage angegeben. 13 D. h. der Linguist könne zu einem gegebenen Satz Fragen stellen; die einen führten zum Thema, die anderen zum Rhema. Etwa auf die Fragen „What happened? What's the matter?" erhielte man das Rhema. Doch 1. versagt diese Frage bei den meisten Sätzen, die die Kopula enthalten, wie (1) Prag ist schön. Prag ist die Hauptstadt der CSSR. 2. erhält man zwar vielleicht oft den rhematischen Pol des Satzes, aber erhält man ihn ganz und nur ihn allein? 3. ist mir nicht klar, wie man sauber die Fragen voneinander trennen kann, die erstens auf das Thema, zweitens auf den Topik (logisches oder psychologisches Subjekt), drittens auf das grammatische Subjekt zielen. 4. Was ist der theoretische Status solcher „inhärenten Fragen", die sich aus der römischen Grammatik bis in die Satzanalyse unserer Tage gerettet haben? Es handelt sich offenbar um eine Entdeckungsproze12

1967, 207f, 223, Fn. 15; 1969, 67.

1 3

C f . FIRBAS 1 9 6 2 m i t L i t . ; D A N E S 1 9 6 4 , 5 0 7 ; 1 9 6 8 ,

158.

FSP U N D TEXTTHEORIE

91

dur des Sprachbetrachters. Entdeckungsprozeduren können aber nur Werkzeuge zur provisorischen Abtastung des Untersuchungsfeldes sein. Zur Überprüfung könnten solche Fragen nur verwendet werden, wenn man sie textgrammatisch verwendet, als Teiläußerungen der Texteinheit Dialog; d. h. man müßte ζ. B. das Sprachverhalten von Versuchspersonen in Antworten auf Vergewisserungsfragen testen, ζ. B.. (2) Prag ist schön. — Wovon ist die Rede? — Von Prag. Dieser Topik-Test ist aber nur sehr begrenzt anwendbar. 14 4.2. Nützlicher ist m. E. in vielen Sprachen ein „responsiver Auslassungstest". Abgesehen von satzgrammatischen Restriktionen kann in Second-instance-Antworten auf Fragen alles bis auf das „Thematisierte" 15 Satzglied ausgelassen werden: (3) Wenn hat Hans gesehen? — Maria {hat Hans gesehen). Die gilt auch für zweizielige, zweifach Thematisierte Fragen: (4) Wer hat wen gesehen? — Hans # Maria. Nehmen wir als nächstes die Frage: (5a) Hat Hans Maria gesehen? — Ja # Maria. Hier bleiben zwei Elemente übrig, denn das Wort „ j a " kann nur dann ausgelassen werden, wenn man es durch eine Geste, ζ. B. Kopfnicken, ersetzt. Dies hat zur Konsequenz, daß schon die neutrale Entscheidungsfrage 5b) Hat Hans Maria gesehen? ein Thematisches Glied enthält, das nicht mit dem Rhema des korrespondierenden Aussagesatzes „Hans hat Maria gesehen" iden14

Z u r M ö g l i c h k e i t des D i a l o g t e s t s vgl. GREENBERG 1969; DRESSLER -

fordert für die FSP die Befragung von Informanten: Wo ist diese Forderung (unter notwendiger Einbeziehung der Statistik usw.) schon erfüllt? 15 D a ich hier keine kohärente Texttheorie entwickeln kann, verwende ich für „Hervorhebung", das Kennzeichen der Second-instance-Sätze, den Terminus der F S P „Rhematisierung" (andere: „focus"); vgl. FIRBAS (1962, 141; 1968, 15) und D A N E S (1967, 509f; cf. WEIL 1887, 101). Eine gute Darstellung im Rahmen der generativen Grammatik bietet POSTAL (1968, 210if; anders CHOMSKY 1969, 18ff), doch macht der Einbau in eine Textgrammatik Schwierigkeiten. Im Begriff „Rhematisierung" meint Rhema leider sowohl den Gegensatz zu Topik als auch zu Thema (vgl. u. 8.2). GABRIEL 1 9 7 0 . BENES ( 1 9 6 8 , 2 6 9 )

92

W . DRESSLER

tisch ist, und das sich in der Oberflächenstruktur vieler Sprachen in der Frage-Intonation (und auch in der Inversion) manifestiert; denn in einem sinnvollen Fragesatz (5b) muß die Existenz von Hans und Maria als bekannt vorausgesetzt (präsupponiert) werden, ferner muß die Möglichkeit vorausgesetzt werden, daß Hans Maria gesehen hat; neu oder rhematisch ist also nur die Frage, ob diese Möglichkeit im aktuellen Fall zutrifft. 16 Eine positive Antwort muß zumindest eine Bejahungspartikel wie „ j a " enthalten, oder ein Verb wie „he did" oder ein Pronomen wie griech. „egö". Übrigens ist in der Basisstruktur selbst von „ j a " wenigstens ein performatives „ich antworte dem Frager" enthalten. Der syntagmatischen Substitution des rhematischen Frageperformativs der Frage durch das rhematische Antwortperformativ der Antwort korrespondiert die Substitution des „du" der Frage durch das „ich" der Antwort (immer in der Basisstruktur). 17 Im Litauischen genügt die Wiederholung des perfektiven Aspektpräfixes „pa-", denn die Angabe des perfektiven Aspekts impliziert die Durchführung der gefragten Handlung, also ζ. B. (6) Ar pa-mâtê Jonas Marijq? — Pa (wörtlich etwa „Hat Hans Maria ge-sehen? — Ge."). 4.3. Die vorangegangene Diskussion spricht gegen die Anschauung, jedem Aussagesatz liege mindestens eine mögliche, inhärente Frage zu Grund ( F I R B A S 1962,133), denn Aussagesätze sind keineswegs mit Antwortsätzen identisch, da diese mindestens ein Thematisiertes Glied enthalten müssen (zur Kategorie der Antwortsätze 16

Kommt dies daher, daß jeder Fragesatz generativ aus zwei Sätzen abgeleitet werden muß, so wie ein Satz mit Kontrastemphase (Rhematisierung)? (vgl. Fn. 15). Dann enthalten alle Satztypen außer unemphatischen Aussagesätzen mindestens eine Hervorhebung oder Rhematisierung, die aus dem Rhema des zusätzlichen (ζ. T. performativen) Basis-Satzes resultiert; ein eventuelles deklaratives Performativ (Ross, J., vgl. DRESSLER 1970, § 4 ) wäre dann unmarkiert. Freilich kann man in der Antwort von (5a) „Maria" auslassen; der Unterschied zu (5b) besteht aber darin, daß dort die Antwort „Ja, Maria" problematisch wäre. Anders HALLIDAY 1967, 212f. 17 Zum Dialog vgl. noch WUNDERLICH, D., STZ 32, 1969, 272—280; CHOMSKY 1968, 18—27, 30f.

FSP U N D TEXTTHEORIE

vgl. noch satz

DRESSLER

(7) Ich gehe weg.

93

1970). Vgl. ζ. B. den u. U. möglichen Antwort-

weg « Ja, ich gehe weg 4= (Aussagesatz) Ich gehe

Die Anschauung, „ja" sei in einer Antwort ein zweiter Satz, ist rein oberflächenstrukturell, da in der Basisstruktur jede Antwort auf mindestens zwei Sätze zurüchzuführen ist (vgl. u. 6.3.). 4.4. Die Kehrseite der Beibehaltung des Rhemas ist das Prinzip der „Thema-Ellipse" oder "elliptischen" Anapher. 18 Daher ist in der Frage-Antwort-Einheit (8) Hast du ihm ein Buch gege—ben?

Ja

auch die indefinite Nominalphrase „ein Buch" ein Teil des Topiks der Frage bzw. des Thema-Satzes der Antwort „x ( ( = ich/duVariable) hat ihm ein Buch gegeben". Dabei ist die Ellipse im Rahmen eines Konstituenten in der Regel rechtsläufig (die Möglichkeit der Ellipse nimmt von links nach rechts zu). Auch dies spricht gegen die Auffassung des Antwortsatzes als einfacher Satz, denn dann sollte das Thema bzw. der Topik in der Regel am Anfang stehen und auslassbar sein. 4.5. Neben dem responsiven Auslassungstest gibt es noch einen anderen Rhema-Test, nämlich eine Kommutationsprobe: Das Rhema kann nach einigen Sprachwissenschaftlern18" nicht durch Anaphorika paradigmatisch (wohl aber syntagmatisch) substituiert werden. Dies gilt jedoch nicht, wenn ein Anaphorikum kataphorisch oder deiktisch verwendet wird (vgl. DRESSLER 1970, § 2.6.) 4.6. Bekannt ist die Bedeutung der FSP für die Prosodie, darunter für die Satzintonation. 19 Auch hier stellt sich die Frage nach der empirischen Überprüfbarkeit, wozu ebenfalls eine textgrammatische Betrachtung notwending zu sein scheint.20 5.1. Das Thema (im Sinn von oben 1.) hat zwei konkurrierende Definitionen erfahren: Bei HALLIDAY ( 1 9 6 7 , 200, 204f) und ge18

18a

DRESSLER 1970 § 2; v g l . LYONS, J . , JL 2 , 1 9 6 6 , 212.

Vgl. ζ. B. To Honor R. Jakobson I, The Hague-Paris 1967, 967-976.

19

DANES 1960; 1967, 508ff; FIRBAS 1961 u n d 1968.

20

GRBENBERG 1969; DRESSLER-GABRŒL 1970; MORGAN 1967, 124.

94

W. DRESSLER

wohnlich in der Prager Schule wird „ T h e m a " als bekannte Information aufgefaßt, „ R h e m a " als neue Information. Dann kann „ T h e m a " kaum mit „Topik" ( = Basis = logisches Subjekt) identisch sein. Hingegen definiert J. F I R B A S 2 1 das Thema als denTräger der geringsten kommunikativen Dynamik (CD) im Rahmen eines kommunikativen Feldes ( S V O B O D A 1968) und gelangt zu einer gewissen Harmonisierung von Thema und Topik: „theme p r o p e r " mit der allergeringsten C D wäre dann vermutlich der Topik-Teil des Themas. Doch die CD wird weder im Rahmen einer linguistischen noch einer kommunikativen Theorie hinreichend definiert, sondern bleibt ein intuitiver, bloß empfindbarer Primitivbegriff. 5.2. Second-instance-Sätze sollen ζ. B. zum Unterschied von nichtemotionellen Sätzen keine ansteigende (oder gar keine) C D besitzen. 22 Doch wenn wir in den acht Sätzen (9) Karl gab das (ein) Buch einem (dem) Freund [mit optionaler Vertauschung von direktem und indirektem Objekt] den Topik „ K a r l " Thematisieren, so ändert sich m. E. die darauffolgende Kontur der CD keineswegs, sie tritt nut „in den Schatten" des neuen wichtigeren Rhemas. Auch das alte Rhema bleibt untergeordnet erhalten (vgl. u. 7.3.), der nicht-rhematisierte Satzteil ist keineswegs ein kompaktes „theme proper", wenn auch sein kommunikatives Feld weniger C D besitzt als das rhematisierte Satzglied. Auch in Second-instance-Sätzen wie (9) beeinflussen Änderungen in Wortstellung und Artikelgebrauch die C D des sog. Thema-Teils. 5.3. Oder wann haben zwei Segmente dieselbe oder eine verschiedene C D ? F I R B A S (1961, 84f, 88f) spricht zwar von „Kompaktheit", wenn „eine Kette von Elementen voneinander relativ wenig im Grad der von ihnen vermittelten CD differieren", doch dies ist noch keine hinreichende Definition. 5.4. Nehmen wir ζ. B. verschiedenartige mit der Konjunktion „ u n d " verbundene Segmente: die idiomatischen Wendungen „Stadt und Land, Katz und Maus" sind wohl kompakt: Ihre Wort1964, 272; 1966, 240f; 1968; SVOBDOA 1968, 54.

2 1

FIRBAS

2 2

FIRBAS 1 9 6 8 , 1 3 f f , vgl. BSE

1 , 1 9 5 9 , 43.

FSP U N D TEXTTHEORIE

95

Stellung ist nur in der Sprachform festgelegt. Anders steht es mit (10) Hans und Maria gingen spazieren ein Satz, der (wenn man nicht „conjunction reduction" annimmt) vielleicht dieselbe Basisstruktur hat wie der Satz (11) Hans ging zusammen mit Maria spazieren (cf. SGALL 1969, 64f). Schon in (10) hat „Maria" mehr CD als der Topik „Hans", auch wenn beide zum Thema gehören könnten. In (11) kann „Maria" sogar Rhema sein, allerdings kaum nach dem vorausgehenden Satz „Gestern traf ich Hans und Maria" (mit kompakten „H. und M."). Nun zur Textgrammatik: In den beiden Oberflächenstrukturen identischen Teilsätzen (clauses) bzw. Sätzen (sentences) (12a) Hans lief und Hans lief (12b) Hans lief. Und Hans lief kann nur dann die korrekte Kontur der CD identisch sein, wenn „und" soviel wie „und dann immer weiter bis" bzw. im Fall der Mehrmaligkeit „ . . . immer wieder bis" bedeutet, ζ. B. vor dem Abschlußsatz (12c) Aber er erreichte sein Ziel nicht mit Thematischer Negation (cf. FIRBAS 1962,142f). Oder der zweite Satz bzw. Teilsatz ist ein Second-instance-Satz: (13) Hans lief und Hans lief (gedehnt, steigend-fallend), wobei „und" sowohl „und zwar" als auch „und dann..." bedeuten kann. In anderen Sprachen leitet „und" nach der Apodosis die Protasis ein („Parahypotaxe") und markiert sozusagen die Apodosis als Thema oder Topik (vgl. DRESSLER, hier, Fn. 3 9 ) . Wie wichtig eine empirisch überprüfbare linguistische oder kommunikative Theorie der CD wäre, zeigt das Problem der Gleichwertigkeit gewisser mit „und" verbundener Satztypen mit Relativsatz-Satzgefügen,23 die sich im Neugriechischen in der Oberflächenstruktur manifestiert: Relative Clauses and Conjunctions. Mimeo, June 1967; 1967; LANGACKER R. W., Mirror Image Rules in Natural Languages. Mimeo, January 1968; LAKOFF, G., Deep and Surface Grammar, 36ff, 2 3

ANNEAR, S . ,

MOORE

96

W. DRESSLER

(14) Eide tà paidià hai (= poü) paizane wörtlich „Er sah die Kinder und sie ( = welche) spielten". 6.1. Noch einmal zur intersubjektiven empirischen Überprüfbarkeit der FSP: Am objektivsten wäre eine Überprüfung durch einen Computer. Ich habe daher versucht, die Behauptung, gewöhnlich sei der definite Artikel thematisch, der indefinite Artikel rhematisch, und das Thema gehe gewöhnlich dem Rhema voraus, durch einen Computer zu überprüfen, und zwar in den Gedichten des portugiesischen Dichters Fernando Pessoa. 24 Der erwartete Normalfall trat aber nur in 3% der Sätze ein, der indefinite Artikel stand in kaum weniger Sätzen (2,5%) vor dem definiten Artikel. Diese lächerlich niedrigen Prozentsätze erklären sich z. T. daraus, daß pluralische definite Artikel nicht gewertet werden konnten, weil der Computer die oppositionellen indefiniten pluralischen Nominalphrasen nicht zählen konnte. Der tiefere Grund, warum eine solche Überprüfung von vornherein zum Scheitern verurteilt war, ist das Faktum, daß die FSP keine Eigenschaft der Oberflächenstruktur ist, wie manche amerikanische Generativisten implizit annehmen, wenn sie die sog. „freie Wortstellung" durch bedeutungslose, oberflächliche „scrambling rules" generieren wollen. Die Annahme der Bedeutungslosigkeit der „freien" Wortstellung ist natürlich für den Linguisten eine recht angenehme Vereinfachung seiner Aufgabe, wenn man etwa bedenkt, daß der russische Satz (15) Ja zavtra utrom pojdu guljaf (Ich werde morgen früh spazieren gehen) 25 mimeo, Indiana Linguistics Club 1968; NICKEL, G., Some Contextual Relations between Sentences in English, Proceedings of the Tenth Intern. Congress of Linguists. — Zur Analyse von „und" vgl. noch PALA, P., PSML 1, 1 9 6 6 , 1 9 5 - 2 1 7 ; FREY, G., StudGen 19,1966, 439f; THÜMMEL, W„ Lingua 20, 1968, 3 8 1 - 4 1 4 ; STAAL, J. F., JL 4, 1968, 7 9 - 8 1 ; VASILIU, E., RRL 14, 1969, 4 3 5 - 4 4 6 . 24 Gespeichert im Statischen Institut der Universität Wien. Das Programm hat liebenswürdigerweise der Mathematiker Norbert Winterleitner ausgearbeitet. 25

FEDOROV, Α . V . - KUZNECOVA, N . N . - MOROZOVA, E. N . - C'IGA-

NOVA, I. Α., Nemecko-Russkie jazykovyeparalleli.

Moskau 1961, 112.

FSP U N D TEXTTHEORIE

97

zumindest rechnerisch 120 Wortstellungsvarianten besitzt. Sie alle textgrammatisch einzuordnen, ist eine entmutigende Aufgabe! 26 6.2. Wenn die FSP also nicht der Oberflächenstruktur angehört, dann muß die Frage, wo sie in der „underlying structure" einzuordnen ist, je nach der Theorie einer Schule ganz verschieden ausfallen. M. E. läßt sich der Begriff Topik an wenigstens drei Stellen nebeneinander anwenden: 1. in der satzgrammatischen Basisstruktur, wenn wir eine semantische Erklärung für folgende Übersetzungsäquivalenzen: 263 (16a) russ.: U Peti byl velosiped = deut.: Peter(chen) hatte ein Fahrrad (16b) russ.: Velosiped u Peti byl = deut.: Das Fahrrad befand sich bei P. D. h. von den semantischen Tiefenkasus des Dependenzschemas wird jeweils ein anderer zum Topik. 2 7 2. brauchen wir in der Derivation den Topik zur sog. Topikalisierungsregel, die in der generativen Grammatik „cleft sentences" 28 erzeugt, wie (17) It was Mathesius, who founded the theory of FSP. Hier liegt zugleich auch eine Rhematisierung vor, sodaß man Fall 2. als Abart von Fall 1. auffassen könnte. 3. können wir den Topik (oder wenigstens eine CD-Hierarchie) in der Tiefenstruktur der Textgrammatik anwenden, um etwa die folgenden drei Äußerungen zu unterscheiden, die oifenbar dieselbe text- und satzsemantische Basisstruktur haben: (18a) Ich frage dich : Gehst du weg? (18b) Ich frage dich, ob du weggehst, (18c) Gehst du weg? 26

Vgl. zum Deutschen SEILER, H., Word 18, 1962, 121ff. Vgl. Cs. pfednäsky pro VI. mezindrodm sjezd slavistä, Praha 1968, 51-59. 27 Vgl. auch HALLIDAY 1968, 205, 208; STAAL, J. F . , FL 3, 1967, 66ff; BREKLE, H . E . , Generative Satzsemantik und transformationelle Syntax im System der englischen Nominalkomposition. München 1970, 78f, 128—135. 26a

2 8

C f . BREKLE, F n . 2 7 ; MOORE 1 9 6 7 ; A n d e r s LAKOFF, G . , O n

Semantics, In: Steinberg-Jakobovits,

Semantics.

Generative

98

W. DRESSLER

Interessanterweise scheint in der Sprachnorm die optionale Frageintonation in (18a) seltener verwendet zu werden als in (18c); in (18b) wäre sie irregulär. Verwandt ist auch die Basisstruktur von (18d) Ich stelle die Frage nach deinem Weggehen. 6.4. Und damit kommen wir zur dornigen Frage der Beziehung von Parataxe und Hypotaxe. Gleichbedeutende parataktische und hypotaktische Äußerungen haben dieselben semantischen Basisstrukturen, unterscheiden sich aber in den gegenseitigen Beziehungen ihrer kommunikativen Felder. 29 Wenn man die Hypotaxe nach alter Grammatikertradition aus der Parataxe ableitet, dann wird der Unterschied zwischen Satz- und Textgrammatik scharf (cf. DRESSLER 1969, 2, 2 1 f ) . W . THÜMMEL (1968) deriviert hingegen

die Parataxe aus der Hypotaxe, und dies muß letzlich auch jede konsequente Textgrammatik tun, wenn sie einen ganzen Text aus einem einzigen Textsymbol ableitet. 30 6.5. Topiks von Textabschnitten haben zwar in der Rhetorik und Kommunikationstherie (Ungeheuer 1968) eine Rolle gespielt, m. W. aber noch kaum in der Linguistik (cf. DRESSLER 1970, § 3.2). A. WHEELER (1967) entdeckte in der kolumbianischen Indianersprache Siona Partikeln, die der Hervorhebung im Text, nicht im Satz dienen. Sie spricht (nach K. Pike) von „focus". Wenn man ihre Bemerkungen und ihren Beispieltext liest, so erhält man den Eindruck, daß hier die Termini Thema, Rhema, Topik, Nicht-Topik („comment") in ihrem gewöhnlichen Sinn nicht passen. 7.1. Ein Thema („given") wird durch semantische Anapher bzw. durch semantische Äquivalenz- oder Kontinuitätsbeziehungen zwischen Sätzen konstituiert, 31 die nach J. Carroll 32 ζ. T. durch Asso2 9

SVOBODA, S . , 1 9 6 8 ; FIRBAS 1 9 6 8 , 1 2 f , 4 2 , F n . 4 1 ; c f . SGALL 1 9 6 7 , 2 1 2 f .

Damit ist auch HEGERS, K . semantische Taxonomie vereinbar (ZRP 8 5 , 1969, 204-206). 30 Vgl. DRESSLER, W., Modelle und Methoden der Textsyntax. Folia linguistica 4, 1970. 3 1 BUTTKE 1959, 557; HARRIS, Z. S., Discourse Analysis Reprints (The Hague 1963); WORTH 1964, 56f; K O C H 1965, 12ff; HARPER 1965, 2 - 9 ; KARTTUNEN 1 9 6 8 , 1 4 , 2 2 — 2 6 ; UNGEHEUER 1 9 7 0 ; D A N E S 1 9 6 8 ;

H., Lingua 23,1969, 285; DRESSLER 1970 § 3.1. 3 2 HILL, Α . , ed.: Linguistics Today. New York 1969, 168f.

BAUMANN,

FSP U N D TEXTTHEORIE

99

ziationsexperimente getestet werden können. Wenn jedoch im Folgesatz mehr als eine semantische Anapher besteht, so ist das Satzglied mit der geringsten C D textgrammatisch nicht einwandfrei vorauszusangen. Dazu ist der satzgrammatische Topik notwendig. Nehmen wir eine Abwandlung einer Firbas'schen (1968, 12f) Satzfolge, so haben wir zwei Varianten, die sich nur durch die Wahl des Topik unterscheiden: (19a) Ein Mann sah einen Jungen. [JAL] g a b [ i ] e i n e n B a l L Die weiteren, Thematisierten Wortstellungsvarianten zeigen eine topikale Bevorzugung des Subjekts: (19b) Einen Ball gab er (er) ihm II er ihm // ihm er // *ihm (ihm) er. 7.2. Die Frage, ob man auch von Thema und Rhema von Textabschnitten sprechen kann, wurde m. W. noch nicht untersucht (doch vgl. W H E E L E R 1967 und DRESSLER Fn. 39). F I R B A S (1968, 39 Fn. 12) nimmt an, daß auch in Textabschnitten die C D graduell ansteigt. Dies trifft zumindest bein Phänomen des Ausklingens nicht zu. 7.3. Vier weitere wichtige Probleme wurden im Rahmen der FSP m. W. noch nicht angeschnitten. Erstens genügt es nicht, den thematischen Charakter von anaphorischen Pronomina fesztustellen; man muß auch untersuchen, wann überhaupt pronominale Anaphora möglich ist, und warauf sie sich bezieht. Darüber gibt es schon eine umfangreiche Literatur, 3 3 ich möchte hier ein neues Beispiel für die topikale Bevorzugung des Subjekts (cf. 7.1.) bzw. den Parallelismus der grammatischen Rolle 3 4 bringen, wobei ich den pronominalen Bezug durch Indices angebe: 33

TABER 1 9 6 6 ,

113; 117; PALEK 1968; HARWEG 1968; POSTAL

1968;

LAKOFF G., Counterparts, or the Problem of Reference in Transformational Grammar. (Mimeo.) Indiana Linguistics Club 1968; KARTTUNEN, L., Discourse Referents. COLING (Sänga-Säby 1969). Usw. 3 4

C f . HARPER 1 9 6 5 , 12; e h e r s t i l i s t i s c h MORGAN 1 9 6 7 , 1 2 7 , v g l , d e n b e -

kannten „parallelismus membrorum".

100

W. DRESSLER

(20a) Die Herrin¡ machte mit ihrer Hündin^ eine Spazierfahrtk. Siei(j7k?) war sehr schön. / Und sie¡ ( j Í ek) wahr sehr schön. Man könnte freilich auch die Faustregel aufstellen: Thema bezieht sich auf Thema, Rhema auf Rhema. Vgl. auch die Satzfolge: (19c) Ein Mann, sah einen Jungen y Er¡ erhielt von ihm¡ einen Ball. I Von ihm¡ erhielt eri einen Ball. Kein Gegenbeispiel ist die parallel Satzfolge (20b) Der Herri machte mit seinem Hund¡ einen Spaziergangk. Er j(k7) war sehr schön. Denn von einem Mann sagt man gewöhnlich nicht, daß er schön ist: Die semantische (oder pragmatische) Sprachnorm ist hier entscheidend. Hingegen bezieht sich anaphorisches „Der" (vgl. sie in (20a)) auf das Rhema, wodurch Second-instance-Sätze mehrdeutig werden können (cf. Fn. 15): (21a) Die Herrin¡ machte mit ihrer Hündin} eine Spazierfahrt^. Diek war sehr schön. (21b) Die Herrini machte eine Spazierfahrtk mit ihrer Hündin-y Diej(kV war... (21c) Eine Spazierfahrtk mit ihrer Hündin¡ machte Herrin ¡. Die kj¡ war ,..35 (21d) Eine Spazierfahrtk machte die Herrin¡ mit ihrer Hündiny Diek j war ...35 Die anaphorische Zuordnung erfolgt in solchen Zweifelsfällen wohl nach Wahrscheinlichkeitsüberlegungen in der Performanz. 35 7.4. Zweitens: Da im Rahmen der F S P (vgl. SGALL 1967) nicht zwischen Oberflächen- und Tiefenstruktur unterschieden wurde, blieb das Phänomen der Thema-Ellipse unbeachtet. Thema-Ellipse ist z. B. eine der häufigsten Ursachen für anaphorische Anfangstellung des Verbums 36 , vgl. etwa die folgende Passage aus dem ersten Kapitel von Caesars „De bello Gallico": 35 Die Zweideutigkeit der Anaphora scheint bei emphatischer Betonung der jeweils letzten Nominalphrase des Erstsatzes stärker zu sein. Dies müßte genauer überprüft werden, ebenso die Frage, ob die Hervorhebung eines bereits rhematischen Element nicht-diskret ist. 36 s. DRESSLER 1969. Weitere Ursachen sind topikale oder rhematisierte

FSP U N D TEXTTHEORIE

101

(22) Eorum una pars ... initium capii a flumine Rhodano; continentur Garumna flumine, Oceano, finibus Belgarum; attingit edam ab Sequanis et Helvetiis flumen Rhenum; vergit ad semptemtriones. 7.5. Drittens ist die Beziehung von pronominaler Referenz und Second instance aufzuklären (cf. 7.1.). So haben A. AKMAJIAN und R. JACKENDOFF37 folgende Satzfolgen mit wechselnder Referenz bemerkt: (23) John¡ hit Bill¡ and then George hit him¡ (him¡) Μ. E. ist hier allerdings zumindest im ersten Fall „George" ebenfalls rhematisch. 7.6. Viertens wurde nicht untersucht, wann Second-instanceSätze textgrammatisch regulär sind (cf. HALLIDA Y 1968, 210). Nehmen wir die texthomonyme Satzfolge (24a) Ein Mann sah einen Jungen. Ihm (dem Jungen) gab er (der Mann) einen Ball. Der war sehr schön. Annehmbar ist hier im Mittelsatz nur eine einzige Rhematisierung (24b) Er gab ihm einen Ball // Einen Ball gab er ihm und auch dies höchstens dann, wenn in dem Land, in dem diese Geschichte spielt, Bälle eine große Mangelware darstellen. Der Satz (24c) Er gab ihm einen Ball wäre allenfalls in einer Welt denkbar, in der Erwachsene grundsätzlich Jugendlichen Spielsachen wegnehmen, aber keine geben. (Unter besonderen Umständen könnten sehr religiöse Menschen „ E r " auf „ G o t t " beziehen). 8.1. Und damit kommen wir zur Eingliederung der Textgrammatik und damit auch der FSP in die Pragmatik, 3 8 die die extralinguiVerwendung des Verbs, vgl. BUTTKE 1959, 554ff. Bei der kataphorischen Anfangstellung des Typs tschech. „Byl jednou jeden král... [Es war einmalein König...]" scheint das Verb Text-Topik zu sein (anders FIRBAS 1964, 268; WEIL 1887, 33). Die Semantik der Verba (BENES 168, 267 mit Lit.) wird m. E. überschätzt. 37 Squib (Mimeo), February 1968. 3 8 MONTAGUE, R . , Pragmatics. In: Contemporary Philosophy (ed. R . Klibansky), Florenz 1 9 6 8 , 1 0 2 — 1 2 1 ; WUNDERLICH, D . , Pragmatik, Sprechsituation und Deixis. Univ. Stuttgart, Lehrstuhl für Linguistik, Papier 9,

102

W. D R E S S L E R

stische, situative Anwendbarkeit der Sprache untersucht. Hier wäre es ζ. B. wichtig, die Beziehung zwischen Topik und Präsuppositionen zu untersuchen. 8.2. Mit der Pragmatik konkurriert die Kommunikationstheorie (UNGEHEUER 1968). Ihre Bedeutung für die FSP erhellt schon aus folgendem hypothetischen Axiom: „Sprachlich muß nur mitgeteilt werden, was dem Empfänger unklar oder unbekannt ist. Was bekannt ist, kann vorausgesetzt, d. h. verschwiegen oder als Topik eingeführt werden". Wenn man die FSP im Rahmen einer kommunikativen Theorie untersucht, so könnte man vielleicht Thema und Topik unter dem Begriff der kommunikativen Voraussetzung zusammenfassen.39 8.3. Oft kritisieren Linguisten die Ausdrücke „psychologisches Subjekt" und „Prädikat" als unlinguistisch. Die Psychologie spielt hier aber eine wichtige Rolle. Denn wie der Psycholinguist S. Ertel (Münster/Heidelberg) in Vorträgen gezeigt hat, kann die Psychologie voraussagen, welche Nominalphrase in isolierten, texthomonymen Einzelsätzen als Topik gewählt wird. 8.4. Auch die Bezeichnungen „logisches Subjekt" und „Prädikat" dürfen nicht ohne weiteres abgelehnt werden, denn die Linguistik braucht eine Beziehung zur Logik, sei es zu der unter den Generativisten der USA und BRD so beliebten Prädikatenlogik oder zur „natural logic" G. LAKOFF'S (zu beidem s. Fn. 3) oder zur Universalgrammatik R. MONTAGUE'S. 4 0 8.5. Verdienst der Prager Schule ist es, selbst in den Zeiten des größten Sprachimmanentismus die extralinguistischen Bezüge der FSP nicht ganz vergessen oder verschwiegen zu haben. Doch geht es heute nicht mehr an, textgrammatische, pragmatische, kommu-

1968. Vgl. die Lehre von den Präsuppositionen (Ζ. B. LAKOFF, R . , Lg 45, 1 9 6 9 , 6 0 8 f f . ; v g l . KRENN-MÜLLNER

F n . 3 ) u n d K . PIKE'S

Sprechsituation.

Zum Begriff des Subkontexts s. SKALIÖKA, V., Slav Ρ 3, 1961, 73—78. 39 DRESSLER, W., am Ende von Grundsätzliches zur Funktion der altanatolischen Satzpartikeln, AO 38, 1970. 40 English as a Formal Language. In: Languages in Society and the Technical World, Milano 1970.

FSP UND TEXTTHEORIE

103

nikative, psychologische und logische Relationen in denselben Sammeltopf „Kontext" z u werfen. W e n n wir dies alles bedenken, s o müssen wir gestehen, daß im R a h m e n der F S P noch äußerst viel geleistet werden muß, und daß die m. W. einzige bisher gedruckte längere textlinguistische Arbeit aus der FSP-Schule, DANES'S (1968) Untersuchung zur thematischen Progression, nur einen allerersten Anfang darstellt. U n d wenn wir ferner bedenken, daß dieser Bericht nur einen kleinen Teil textlinguistischen Problemkreise erwähnt hat (ζ. B. überhaupt nicht Tempus, M o d u s , Aspekt), s o kann sich unser Eindruck nur verstärken, daß die wirklich großen Aufgaben erst bevorstehen!

LITERATURVERZEICHNIS BENES, E., On two aspects of functional sentence perspective, TLP 3, 1968, 267-274. BUTTKE, K., Beziehungen zwischen Wortstellung und sprachlichem Umfeld, ZSl 4, 1959, 551-559. CHOMSKY, N., Deep structure, surface structure, and semantic interpretation. (Mimeo.) Indiana Linguistics Club, 1969. DANES, F., Sentence intonation from a functional point of view, Word 16, 1960, 3 4 - 54. DANES, F., A three-level approach to syntax, TLP 1, 1964, 225—240. DANES, F., Order of elements and sentence intonation. In: To Honor R. Jakobson I, The Hague 1967, 499-512. DANES, F., Typy tematickych posloupnosti Ν textu (na materiále íeského textu odborného) [Thematische Progressionen im Text]. SaS 29, 1968, 125-141. DANES, F., Zur linguistischen Analyse der Textstruktur, Folia linguistica 4, 1970,49-56.

DRESSLER, W., Eine textsyntaktische Regel der idg. Wortstellung. KZ 83, 1969, 1 - 2 5 . DRESSLER, W., Towards a semantic deep structure of discourse grammar. Papers from the Sixth Regional Meeting of the Chicago Linguistic Society 1970.

DRESSLER,, W. - GABRIEL. E., Vorbericht über experimentelle textsyntaktische Untersuchungen zur deutschen Satzintonation, Sprache 16, 1970. FIRBAS, J., On the communicative value of the modern English finite verb, BSE 3, 1961, 7 9 - 1 0 4 .

104

W. DRESSLER

FIRBAS, J., The function of the sentence in the act of communication (Marginalia on two important studies by Anna Granville Hatcher), SPFFBU 11 A 10, 1962, 133-148. FIRBAS J., A note on transition proper in functional sentence analysis, PhP 8, 1965, 170-176. FIRBAS, J., On the prosodie features of the modern English finite verb as means of functional sentence perspective, BSE 7, 1968, 11—48. GREENBERG, S. R . , An experimental study of certain intonation contrasts in American English, Working papers in phonetics 13, Univ. California, Los Angeles 1969. HALLIDAY, M . A . K . , Notes on transitivity and theme in English. JL 3, 1967, 37— 81, 1 9 9 - 244; 4, 1968, 1 7 9 - 2 1 5 . Cr. SSE 8, 1969, 8 1 - 8 8 . HARPER, Κ. E., Studies in inter-sentence connection. Sta. Monica 1965. HARWEG, R . , Pronomina und Textkonstitution. München 1968. HAUSENBLAS, Κ . , On the characterization and classification of discourses, TLP 1, 1964, 67-83. KARTTUNEN, L., What makes definite noun phrases definite? Santa Monica 1968. KOCH, W., Preliminary sketch of a semantic type of discourse analysis, Linguistics 12, 1965, 5 - 30. KOCH, W., Recurrence and a three modal approach to Poetry. The Hague 1966a. KOCH, W . , Einige Probleme der Textanalyse, Lingua 16, 1966b, 383 — 398. MOORE, T . H . , The topic-comment function: A performance constraint on a competence model. Diss.: Univ. of California, Los Angeles 1967. MORGAN, J. O . , English structure above the sentence level. Monograph series in languages and linguistics 20, 1967 (Georgetown Univ.), 123 — 132. PALEK, B., Cross-reference: A contribution to hyper-syntax, TLP 3, 1968, 253-266. POSTAL, P. M., Cross-Over phenomena. In: W. J. Plath, ed.: Specification and utilization of a transformational grammar. Yorktown Heights 1968. SGALL, P., Functional sentence perspective in a generative description. PSML 2, 1967, 203-225. SGALL, P . - NEBESKY, L . - GORALÉÍKOVÁ, A . - HAJICOVÁ, E . ,

A

functional

approach to syntax in generative description of language. New York 1969. SVOBODA, Α . , The hierarchy of communicative units and fields as illustrated by English attributive constructions, BSE 7, 1968, 49—102. TABER, C. R., The Structure of Sango Narrative. Diss., Hartford Studies in Linguistics 17, 166. THÜMMEL, W . , Subordination und Koordination von Sätzen. 3. linguist. Kolloquium über generative Grammatik. Univ. Stuttgart, Lehrstuhl für Linguistik, Paper 8,1968, 143—149.

FSP U N D T E X T T H E O R I E

105

UNGEHEUER, G., Paraphrase und syntaktische Tiefenstruktur. Folia linguistica 3, 1970. WEIL, H . , The order of words in the ancient languages compared with that of the modern languages. (Translated by C. W. Super.) Boston 1887. WHEELER, Α . , Grammatical structure in Siona discourse, Lingua 19, 1 9 6 7 , 60-77.

WORTH, D. S., Ob otobrazenii linejnych otnoäenij ν porozdajuädich modeljach jazyka, VJa 5, 1964, 4 6 - 5 8 .

NACHTRAG Erst nach Fertigstellung des Manuskripts sind dem Verfasser u. a. folgende Arbeiten bekannt geworden, die für Probleme dieses Berichtes von Bedeutung sind: PALEK, B., Cross-Reference. A study from hype'-syntax. Praha 1968/70 (zu § 3.2, 8.5). — TORSUEVA, I. G., Akustiöeskie charakteristiki smyslovogo àlenenija predlozenija. Proceedings of the 6th Internat. Congress of Phonetics. Praha 1970, 917f. — FIRBAS, J . , Note on the prosodie features as means of functional sentence perspective. Ibid. 327—330 (beides zu § 4, 6) — KOCH, W. Α., Phonologie und Textanalyse. Ibid. 493—497 (bezieht ausserdem das Redetempo ein). — BENES, E. Über zwei Aspekte der funktionalen Satzperspektive. Proceedings of the Tenth International Congress of Linguistis 2, 1970, 1021—1025 (zu §1, 3.2, 8.2). — Vgl. auch ibid. die Beiträge von P A D U & V A , E. V. - FIRBAS, J . , - KIEFER, F . , - KIRKWOOD, H . W . , Some systemic means of "functional sentence perspective" in English and German. IRAL 8, 1970, 103-114 (zu § 4.3, 4.6, 5 . 1 - 2 ) .

FUNCTIONAL SENTENCE PERSPECTIVE AND THE ORGANIZATION OF THE TEXT Frantisele Danes (Prague)

The relevance of functional sentence perspective for the organization of discourse (or text) is beyond doubt. We do not claim that the whole linguistic theory or "grammar" of texts should be reduced to F S P (cf. SKALIÖKA 1960), but Halliday's statement that "given the clause as domain, ... theme [= FSP] is the grammar of discourse" (HALLIDAY 1967) holds good, with certain modifications, beyond the domain of the clause as well. I In the works dealing with FSP three aspects of the phenomenon under discussion have been pointed out by various authors: (1) known (given) information — new information; (2) theme (T) — rheme (R); (3) different degrees of communicative dynamism (CD). As I have pointed out elsewhere (cf. DANES 1964) the distinctions (1) and (2) go back to V. MATHESIUS. In his wellknown paper from 1939 he defines the "starting point of the utterance (vychodisko)" as "that which is known or at least obvious in the given situation and from which the speaker proceeds", whereas "the core of the utterance (jádro)" is "what the speaker states about, or in regard to, the starting point of the utterance". The same author defines (in 1942) "the foundation (or the theme) of the utterance (základ, téma)" as something "that is being spoken about in the sentence", and "the core (jádro)" as what the speaker says about this theme. — Distinction (3) has been introduced by J. FIRBAS. By CD he means "the extent to which the sentence element contributes to the de-

FSP A N D T H E T E X T O R G A N I Z A T I O N

107

velopment of the communication" (1964, 270) and at the same time he states that Τ is constituted "by the sentence element(s) carrying the lowest degree(s) of CD within the sentence" (ibid., 272). Τ "need not necessarily convey known information or such as can be gathered from the verbal and situational context" (ibid.). This third aspect of FSP may be viewed as a refined analysis of aspect (2). (In fact, the different degrees of the thematic and rhematic character of sentence elements were mentioned even by Mathesius.) Instead of a strict bipartition of the "information-bearing structure" of the sentence (to use P. Garvin's rendition of Mathesius's Czech term "aktuální clenëni") we arrive at an uneven distribution of CD over the sentence, assigning various degrees of thematicity, or rhematicity to different sentence elements. The two basic aspects of FSP, i.e., the contextual and the thematic ones, have been pointed out by other linguists as well, e.g. E. BENES (1959, 1968), M. A. K. HALLIDAY (1967), P. SGALL (1969), F. DANES (1964, 1970). Most distinctly and consequently this distinction has been pursued by Halliday: in the broad area of "Theme" he distinguishes two simultaneous structures of text: (1) "information focus" (given — new), and (2) "thematization" (T — R). The former determines the organization of text into discourse units, the latter frames each clause into the form of a message about one of its constituents.1 It should be noted, however, that the said distinction is an incomplete dichotomy: the differentiation concerns the first members of the two pairs only (i.e., the known (given) piece of information vs. theme), while the second members are identical, viz. the core of the utterance or the rheme (what the speaker says about the

1

HALUDAY (1967) summarizes this distinction in the following way: "...while 'given' means 'what you were talking about' (or 'what I was talking about before'), 'theme' means 'what I am talking about' (or 'what I am talking about now')". He calls the theme also "the point of departure"; this term (Czech "vychodisko") was used by Mathesius in connection with "known information", however.

108

F. D A N E S

known information, or what he says about the theme).2 After all, what makes the investigators differentiate between "known" and "theme" is the fact that there exist cases where Τ does not convey known information (cf. FIRBAS 1964) or where the ranges of both do not fully coincide. It is true that such cases remain in the minority (cf. Mathesius's statement (1939) that the "starting point" (defined as known information) very often represents the theme of the utterance) and are experienced as special or marked (cf. HALLIDAY,. 1967, 17: "there is in the unmarked case ... and association of the theme with the given"). Nevertheless their existence undoubtedly calls for, and justifies, the said distinction. This being so, we may assume that the connexion of FSP with the text structure proceeds along two lines. The first line, i.e., the opposition between known (given) and new information, clearly involves the textual and situational environment. Halliday (op. cit.) states that it is "closely bound up with the cohesive patterns such as those of substitution and reference" (17) and "does contribute in large measure to the organization of discourse" (16). (The close relationship of the phenomena of anaphora with FSP has been pointed out by B. PALEK, 1968.) 3

From Halliday's statement that "thematization is independent 2 The position of HALLIDAY (1967) is somewhat different, and not quite clear. He defines the rheme in English clauses very indistinctly and indirectly ("the theme is assigned initial position in the clause, and all that follows is the rheme" 17) and his discussion of the Τ — R structure is concentrated on the choice of T; only from the example on p. 22 may we guess that focus and theme principally do not coincide. Cf. also his statement on p. 8: "...in the unmarked case the focus of information will fall on something other than the theme; it will fall at least within the rheme, though not necessarily extending over the whole of it." Roughly speaking, the most discussed problems are the focus (new information) and the theme (what is being talked about), while the other two functions stand rather in background. 3 PALEK (1968) has also suggested a useful distinction of the contextual and the textual approach: the former proceeds from the sentence and takes into account those features of it that are due to its cohesion with neighbouring sentences, while the latter takes as its point of departure the discourse and looks for the network of relations linking together its elements.

FSP A N D T H E TEXT O R G A N I Z A T I O N

109

of what has gone before" (ibid., 17), i.e., of the preceding context, it might follow that this second aspect of FSP is irrelevant in respect to the organization of text. But such a conclusion appears very doubtful in the light of the fact that the choice of the themes of particular utterances can hardly be fortuitous, unmotivated, and without any structural connexion to the text. In fact, even a superficial observation of texts shows that the choice and distribution of themes in the text reveal a certain patterning; this statement also corresponds to our intuitive expectations that the progression of the presentation of subject-matter must necessarily be governed by some regularities, must be patterned. In order to throw more light upon the relationship of the notions "known (given) information" and "theme", let us analyse more deeply the former notion. It is evident that the notion "given (known)" is relative and very broad (if not vague): (1) Given or known is that information which is derivable or recoverable (to use Halliday's wording) from the context, situation and the common knowledge of the speaker and listener. Certainly, there exist individual divergencies between the two, due to differences in their experience, memory, attention, etc. But after all, it is the speaker's evaluation that is the determining factor; this does not exclude, of course, that the speaker takes, more or less, into account the presupposed position of the listener. (2) The communicative feature of "givenness", assigned to particular sentence elements, is a graded property. (3) "Givenness" depends on the length of the portion of preceding text in relation to which the evaluation is being carried out. The upper limit of such a portion should be empirically ascertained. We may tentatively assume, that these portions or "intervals" are in a way correlated with the segmentation of text into paragraphs, groups of paragraphs, chapters, etc. We may even expect a kind of hierarchy or stratification of the feature "given": taking for granted that not only particular utterances but also the sections of text, as paragraphs, etc., and the whole text have "themes" of their own ("hyperthemes"), we can expect that, e.g., the theme of

110

F. DANES

a chapter will be evaluated as "given" throughout the chapter, so that the "interval of givenness" in respect to the information carried by this "hypertheme" will be the whole chapter. (4) The contextual determination of givenness is far from being a simple phenomenon. We might tentatively suggest that as "contextually given" may be regarded such semantic information that has been somehow mentioned in a qualified portion (interval) of the preceding text. It can be mentioned directly, or indirectly. In the first case, it can be mentioned not only with the identical wording, but also with a synonymous expression, or with a paraphrase (cf. Pike's "hypermeaning' or "verbalized concept"). The indirect mentioning is based on semantic inference (or semantic implication, if viewed from the opposite point). Thus, e.g., the expression "illness", occurring in an utterance, might be experienced as conveying a known piece of information if in a preceding sentence (belonging to the same text interval) "health" has been somehow mentioned. The notion of semantic inference (implication) needs a more exact elaboration in terms of distinctive semantic features and their sets. It is clear that, in principle, such semantic relations are involved as those obtaining between a term and its generic terms (hyponymy and hyperonymy), "associative" relations, exemplified by such as "restaurant" — "lunch"; "summer" — "vacations"; "science" — "investigator", etc. (5) The evaluation of (the degree of) contextual givenness depends also on the delicacy (determined by various factors, partly objective — e.g., stylistic — partly subjective) with which the speaker (and listener) evaluates a given expression as semantically implied in a certain preceding expression. (6) Last but not least let us point out the very important fact that the relative character concerns the notion of "new" information as well. Halliday has pointed out that the new piece of information is "new" not in the sense that "it has not been previously mentioned although it is often the case that it has not been, but in the sense that the speaker presents it as not being recoverable from the preceding discourse" (1967, 7/8). The first part of this exposition is obviously true, but the final statement is somewhat obscure, since

FSP A N D THE TEXT ORGANIZATION

111

the author does not explain by what kind of procedure the speaker gets the listener to interpret what, in fact, is recoverable from context, as not being recoverable from it, as being "new". There must exist some objective principle underlying the possibility of presenting something that has been previously mentioned as a "new" piece of information. Let us examine the following example: Sedimentary rocks. (1) Most of historical geology has to do with sedimentary rocks and their contained organic remains. (2) This is accounted for by the fact that events in earth history are recorded mainly in terms of differing kinds of sedimentation...

It is obvious that "sedimentation", representing an essential part of the new information of sentence (2), is fully recoverable from the preceding sentence (in respect to the expression "sedimentary rocks"). But what is new is the connexion of "sediment-" into which it has been put. By "connexion" I mean here not only the rather trivial fact that a word may occur in different collocations or other phrases (i.e., in different multiverbal denominating units), but also, and foremost, the position (or function) of the given element in the communicative structure of the utterance. In other words: the property of being new has two, independent, aspects: (1) "new" in the sense of "not mentioned in the preceding context", (2) in the sense "related as Rheme to a Theme to which it has not yet been related". In the former case, the property "new" is assigned to the expression itself, while in the latter it is the Τ — R nexus that appears as new. This interpretation is justified by the following facts: First, in all cases the new element functions as R (as we have mentioned above, Mathesius did not make a distinction between "new piece of information" and R , and also H a l l i d a y (1967, 8) states that new information "will fall at least within the rheme"). Second, it is not R alone, but its connexion with the given Τ that is communicatively relevant (cf. E . B e n e § 1968, 271). - Thus we may conclude that the information accumulated, at a certain point of a text (or, within a text interval), comprises two kinds of elements (appearing as "known"): denominating units, and Τ — R nexuses.

112

F . DANES

The amount (or the potential) of successively accumulated information is mostly so extensive that the speaker, carrying on the discourse, must necessarily make a choice from this mass. And we may rightly assume that he selects the utterance theme from it (unless he has some special reason to choose something that is not comprised in it). In any case, the portions (elements) of "known" information occurring in an utterance are exactly those elements that are closely connected with the selected Τ (and indirectly with R). Our conception of the utterance theme stands near to E. Benes's characterization of the "point of departure" (Cz. "vychodisko", G. "Basis") as "the opening element of the sentence"· that "links up the utterance with the context and the situation, selecting from several possible connexions one that becomes the starting point, from which the entire further utterance unfolds and in regard to which it is oriented" (1959, 216).4 To put it differently: it is evidently necessary to distinguish between the mass of information accumulated up to a certain point of text, and the portion of this mass contained (occurring) in the particular utterance following this point. This distinction involves a selection from the mass of known information for every utterance. We assume that this selection is determined, directly or indirectly, by the choice of the utterance theme. Thus we must not be content with a statement that certain sentence elements convey the known information (in contrast to others, conveying the new one), but we ought to find out the principles exactly according to which this and not another portion of the mass of known information has been selected. In other words, we have to inquire into the principles underlying thematic choice and thematic progression. Note: In his stimulating article K . HAUSENBLAS ( 1 9 6 9 ) defines the theme as "what has been posited to the fore, into the focus of the field of vision and, at the same time, what presents a foundation to be developed (elaborated) in the BENES (1959) distinguishes between this "point of departure", and what he calls "foundation (základ)" of the utterance. Cf. p. 221 of the present volume. 4

FS Ρ AND THE TEXT ORGANIZATION

113

subsequent discourse" (7). From this statement two functions of the theme may be deduced: (1) the perspective function, consisting in hierarchical graduation of thematic text components (and involving a static point of view, regarding the text as a completed whole), (2) the prospective function, in which the theme serves as a point of departure for the further development of the semantic progression and, at the same time, as a prospect or plan of this development (in which case, the dynamic aspect of the progressive realization of the text is accounted for). The pointing out of the dynamic aspect of text construction is new and undoubtedly deserves further attention. Unfortunately, from Hau'senblas's brief exposition it is not easy to get a clear-cut picture what this aspect really consists in and where to draw the line between the two aspects. Generally speaking, these difficulties probably arise from the lack of an exact model of the dynamic structure of objects, realized in time (real or fictitious); such a model, taking into account their progressive growth, would involve a progressive nexus (relative to the "future" functions of components in the subsequent portion of the text and in the resulting whole), a regressive nexus (relative to possible modifications and transformations of components arising from the backward effect of subsequent components), and a continuous process of cumulation. It might be interesting to reinterpret our notion of "thematic progression" in terms of the two aspects. It is obviously not by chance that the studies o f F S P predominantly concern the problems of theme (and not those o f rheme — cf. the frequent term "thematization" and the rarely Used term "rhematization"), in spite o f the fact that it is just the rheme that represents the core o f the utterance (the message proper) and "pushes the c o m m u n i c a t i o n f o r w a r d " (FIRBAS): f r o m the point o f view o f text organization, it is the theme that plays an important constructional role. The rheme shows its significance as the conveyor of the "new", actual information, while the theme, being informatively insignificant, will be e m p l o y e d as a relevant means o f the construction. (The relation between the rheme and the text will be touched u p o n in section III of our paper.) The inquiry into the thematic organization o f the text is closely connected with the investigation o f the so-called "text coherence" or "text connexity". S o m e scholars even define the text in terms o f this property. (Cf., e.g., H. ISENBERG, 1970, 1: "Wir verstehen unter einem "Text" eine kohärente Folge v o n S ä t z e n . . . " . ) Nevertheless, as w a s

duly

pointed

out by Κ . HAUSENBLAS (1964),

114

F. DANES

(1962) and others, coherence (connexity, continuity) is not a necessary property of texts: they not only display this property to a very uneven degree, but some of them may be characterized exactly as "discontinuous" (HAUSENBLAS, op. cit., 79f.); and TROST (op. cit., 268) calls attention to the very old distinction between the "connected style", tending towards a very close linking up of the sentence with the text (harmonía glaphyra), and the disconnected one (harmonía austera), which tends towards a clean-cut independence of each sentence. In other words, when analysing text coherence (connexity), we should employ the term "coherence" in the neutral (unmarked) sense. The following exposition will be devoted to the way in which FSP contributes to the inner connexity of texts. (It is based on an investigation discussed in DANES 1968, 1970, 1970a.) P . TROST

II Our basic assumption is that text connexity is represented, inter alia, by t h e m a t i c p r o g r e s s i o n (TP). By this term we mean the choice and ordering of utterance themes, their mutual concatenation and hierarchy, as well as their relationship to the hyperthemes of the superior text units (such as the paragraph, chapter, ...), to the whole text, and to the situation. Thematic progression might be viewed as the skeleton of the plot. For our purposes it is necessary to discover an objective criterion for ascertaining the theme (T) of a given utterance. The detailed analysis done by Firbas and others, ascertaining the distribution of different degrees of the communicative dynamism over sentence elements, establishes the communicative microstructure of the utterance. We may content ourselves with establishing the macrostructure, i.e., with a rough determination of the thematic and the rhematic part of the utterance, without specifying the central, peripheral and transitional elements. To this aim we employ a procedure using wA-questions, prompted by the given context and situation, for eliciting the rheme (R) of a given utterance (R-questions). Generally speaking, we assume that

FSP A N D T H E TEXT O R G A N I Z A T I O N

115

it is possible to assign to any sentence (taken as a grammatical unit) a set of wA-questions, representing all possible types of context in which the given sentence is applicable, and consequently, revealing all possible FSP-structures which it can acquire. In this way we are also able to find out, indirectly, the theme of the given utterance. This procedure seems workable, since it is objective, purely linguistic, and involves both the contextual and the thematic aspect of FSP. Before starting with the classification of TP's, let us state that between a simple utterance (i.e., a sentence containing only one Τ — R nexus, with simple Τ and R) and a textual concatenation of grammatically independent sentences, a transitional zone exists, comprising cases of sentence units that reveal a more complicated (condensed or composed) Τ — R structure, i.e., units that, from the point of view of FSP, reveal a textual character, which, however, represent a single grammatical unit, one sentence only. Such sentences are now usually described by grammarians as transforms of a combination of two (or more) underlying simple sentences (different linguists will use different ways of description, but this is irrelevant for our discussion). A similar approach might be employed for the description of the FSP-structure of non-simple utterances as well. The English sentence (describing Wöhler's well known discovery from 1828) (1) Wähler heated ammonium cyanate andfound that it was thereby converted into urea, previously known only as a product of living organisms, evidently revealing a complicated FSP-structure, may be reduced on the following sequence of three sentences (a), (b), (c), each of them having a simple Τ — R structure: (2) (a) Wöhler heated some ammonium cyanate. (b) He found that it was thereby converted into urea. (c) This substance had been previously known only as a product of living organisms. A comparison of (2) with (1) shows the following transformational processes on the level of FSP leading from (2) to (1): 1. (b) has lost its independent status and has been, without the loss of its explicit Τ — R structure, combined with (a) into a multiple

116

F. DANES

utterance, composed of two complete Τ — R nexuses; the connexion between (a) and (b) is supplied by the identity of T a and T b . 2. (c) has lost its independent utterance status, and its Τ — R structure as well; it has been restricted to its rhematic elements and fused with R b into a single complex R{b c) ; or briefly, it has been rhematized. The fusion has been allowed due to the fact that T c is a paraphrase of R b and thus it may be omitted. Thus the FSP-structure of (1) might be symbolized as T a Ra + + T b ( = T a ) -> R(b,c) and described as a multiple utterance with the complex second R. (The grammatical aspect of these processes will not be discussed here.) Generally speaking: In respect to their Τ — R structure, utterances (U's) may be divided into simple U's ,composed U's, and condensed U's. — The composed U's result from composition, by which two (or more) simple utterances are combined in a single sentence frame; if the T's, or R's of the two utterances are the same (from the semantic point of view), they will be mentioned only once. — The condensed U's are based on fusion: If two subsequent «imple Ui and U 2 share a common FSP-element, they may be fused into a single condensed U, either by way of thematization, or rhematization of one of the utterances. The two possibilities depend on the type of the thematic interrelations obtaining between U j and U 2 : (1) If T 2 = Ri, principally both possibilities are available: (a) T 2 will be deleted, and R 2 fused with R x into a complex R* (rhematization of U 2 ). (b) Τ -> R t will be fused into a complex T* (thematization of U t ), T 2 deleted, and R 2 linked with T* as R* of the resulting condensed U*. The choice between (a) and (b) depends on the proportion between respective communicative relevance of RJL and R 2 : if, in the given context, R j appears more relevant than R 2 , then U 2 will be rhematized (i.e., deprived of its utterance status, and thus backgrounded); in the inverse case, R 2 will be brought to the fore by means of thematization of Ui. (2) If T 2 = T t , then T 2 will be deleted and R t fused with Ti

FSP A N D THE TEXT ORGANIZATION

117

into complex T* (i.e., R t will be thematized), to which R 2 will function as R* of the resulting condensed U*. Schematically. 1. Composed U's: a) multiple U: "Goethe wrote the second part of Faust after eighty, and V. Hugo astounded the world with Torquemada at eighty." b) U with a multiple T: "The melting of solid ice and the formation from ice of liquid water exemplify physical changes." c) U with a multiple R: "It is further postulated that the activated amino acids are joined together... and that the long chains are molded in a specific manner..." 2. Condensed U's: a) U with a complex T: "This dark-coloured liquid, known as crude petroleum or crude oil, is obtained from wells of different depth." b) U with a complex R: "The amino acids are required for making proteins, consisting of long chains of these units."

From these elementary types various combinations may be produced (cf., e.g., the above adduced compound sentence (1)). These may be called "complicated" utterances. Note: The processes of composition and fusion on the level of FSP (on the utterance level) are manifested by means of different grammatical devices on the level of the sentence, such as coordination, apposition, some nominalizations, some relative transformations, etc. (This does not mean, however, that the said processes are the only functions of these grammatical means.) Some functions described here in terms of FSP are sometimes referred to as "backgrounding", "complex condensation", etc. (cf., e.g., WEINREICH 1963, VACHEK 1955). But it seems to me that the explanation having recourse to FSP may supply a more exact structural explanation of these somewhat impressionistic notions of a semantic and stylistic character. The grammatical descriptions, especially the transformational ones, have ascertained many synonymous, or nearly synonymous relationships between syntactic constructions. But they tell us nothing or very little about functional differences between such constructions, in spite of the fact that only certain differences in the functional employment of apparent synonymous linguistic means of expression are able to account for their existence in the given language. Assuming that the level of FSP, lying above the other syntactic levels (i.e., the grammatical and the semantic one), represents the domain of the functional employment of sentences, we may try to find out the motivation

118

F. DANES

for the choice between different (semantically) synonymous syntactic forms and transforms exactly in the communicative needs of FSP, to associate different syntactic options with the alternatives of the distribution of the communicative dynamism. Our analysis of Czech scientific and other professional texts, as well as some tentative soundings in the area o f German and English language materials has ascertained the following t h r e e m a i n t y p e s of TP: (1) Simple linear T P (or TP with linear thematization of rhemes):

T2(=Rj)->R2

;

Examples: "Γΐί- ^2) Cz.: V oboru izolâtorù se vënuje velká pozornost tzv. feroelektrikûm. Tyto látky mají schopnost mënit energii elektrickou ν mechanickou a naopak. G.: Eine besondere Klasse stellen tragbare Geräte dar. Diese werden besonders zur Ueberwachung von Strahlungsfeldern verwendet. E.: (a) The first of the antibiotics was discovered by Sir Alexander Flemming in 1928. He was busy at the time investigating a certain species of germ which is responsible for boils and other troubles, (b) The chief organic compound obtained from natural gas is saturated methane. Small quantities of other volatile hydrocarbons are associated with methane. Type (1) represents the most elementary, basic TP. Briefly, R¡ of the utterance U ¡ appears in the next U i + 1 as its T i + 1 , or, in other words, each R becomes the Τ o f the next utterance. » [In the formulae the horizontal arrow -> indicates the Τ — R nexus within an utterance, while the vertical one j, indicates the contextual connection of U's. — In the formula Τ -»• R the order of symbols does not necessarily correspond to the sequence of expressions in a particular sentential utterance based on this formula, since this sequence depends on the interplay of language means employed in FSP.] (2) TP with a continuous (constant) theme: Ψ I

119

FSP A N D T H E TEXT O R G A N I Z A T I O N

Examples: Cz.: [Mezi tzv. ovládací zarízení patri i vëtâina automatickych vyrobních linek.] Tyto linky a jim podobná zarízení vykonávají samoëinnë vSechny operace nutné k zhotovení vyrobku. Tato zarízení vsak nemohou kontrolovat prubëh a vysledky své íinnosti. Nejsou také schopna prizpûsobit se zmënàm vnëjâich podmínek... G.: ["Goethes Erbe in unserer Zeit".] Goethe war überzeugt von dem Fortschritt der menschlichen Entwicklung. Er trat für die Erreichung des Menschengeschlechtes zur friedlicher Entwicklung... Goethes Humanismus ging aus von dem Glauben an das Gute im Menschen... Goethe nannte sich "ein Kind des Friedens". E.: The Rousseauist especially feels an inner kinship with Prometheus and other Titans. He is fascinated by any form of insurgency... He must show an elementary energy in his explosion against the established order and at the same time a boundless sympathy for the victims of it... Further the Rousseauist is ever ready to discover beauty of soul in anyone who is under the reprobation of society.

In this type one and the same Τ appears in a series of utterances (to be sure, in not fully identical wording), to which different R's are linked up. (There are several ways of introducing T, in the utterance, but this is irrelevant here.) (3) TP with derived T's: [ τ ]

τ3 — -

R3

Examples: Cz.: Záákrt (diphteria) je infekëni onemocnëni. Pûsobi je corynobacterium diphteriae. Siri se kapénkovou infekci primym stykem s nemocnym, ëastëji väak bacilonosiôem, nebo neprímo predmëty potrisnënymi hlenem. Inkubaëni doba je 2 az 5 dnû. G.: Die sozialistische Republik Rumänien liegt am Schnittpunkt des 45. Breitenkreises mit 25. Längenkreis. Die Bodenfiäche des Landes beträgt 235 000 Quadratkilometer; seine Bevölkerungszahl ist 19 Milionen Einwohner. Die Staatsgrenze hat eine Gesamtlänge von... Kilometern.

120

E.:

F. DANES

Im Westen hat Rumänien gemeinsame Grenze mit... Im Süden bildet der Fluss Donau die Grenze mit... Die östliche Grenze ist teilweise das Schwarze Meer. New Jersey is flat along the coast and southern portion; the northwestern region is mountainous. The coastal climate is mild, but there is considerable cold in the mountain areas during the winter months. Summers are fairly hot. The leading industrial production includes chemicals, processed food, coal, petroleum, metals and electrical equipment. The most important cities are Newark, Jersey City, Paterson, Trenton, Camden. Vacation districts include Asbury Park, Lakewood, Cape May, and others.

The particular utterancé themes are derived from a "hypertheme" (of a paragraph, or other text section). The choice and sequence of the derived utterance themes will be controlled by various special (mostly extralinguistic) usage of the presentation of subject matter. The types of TP just established may be employed in various combinations. Thus the combination of (1) and (2) is frequent. Some of such combinations, revealing a certain regular pattern, may be considered as TP-types of a higher order, representing a formal frame for the employment of the basic types. The most important of such frames may be called the exposition of a split Rheme: T, — -

R,

Tj

( = RÍ • + R'i )

- Rj

f r; — •

R'5

Examples: Cz.: Na poCátku 17. stol. polozili základ novému rozvoji astronomie dva velcí muzové. Jan Kepler zalozil teoretickou astronomi!. Ukázal, ze je

FSP A N D T H E TEXT O R G A N I Z A T I O N

121

mozno ζ pozorování odvodit... Galileo Galilei zaloiil mechaniku. Svymi pokusy... G.: Die Widerstandsfähigkeit in feuchter und trockener Luft ist bei verschiedenen Arten pathogener Viren sehr unterschiedlich. Poliomyelitisviren sterben in trockener Luft sofort ab, während bei einer Luftfeuchtigkeit von 50% relativ stabil sind. (,..)Bei einem Grippenvirus ist es hingegen umgekehrt; wenn die Luftfeuchtigkeit unter 40% bleibt, so halten sie sich recht gut, sie gehen aber rasch zugründe, wenn die Luftfeuchtigkeit höher steigt. E.: All substances can be divided into two classes: elementary substances and compounds. An elementary substance is a substance which consists of atoms of only one kind... A-compound is a substance which consists of atoms of two or more different kinds...

This type of TP is characterized by the fact that a certain R is explicitly or implicitly doubled (R' + R") or multiple (R' + R" + + R'" + ...), so that it gives rise to a pair (triple,...) of thematic progressions: first R' is expounded and after this progression has been finished, R" becomes Τ of the second TP. (These two (three,...) partial progressions may be of one type only, or they may represent a combination of different types, without having necessarily a parallel structure.) Further, TP's are often complicated by various insertions (supplements, explanatory notes) or asides. They may also occur in an incomplete or somewhat modified form. Let us mention here a typical modification of type (1), namely a TP with an omitted link (or with a thematic jump). Essentially, it consists of the omission of an utterance in a TP. That is to say, the content of such an utterance is to such a degree evident, plainly implied by the context, that it appears redundant, unnecessary, and consequently omissible. Our types of TP are to be considered as abstract principles, models, or constructs. The implementation (manifestation) of these models in particular languages depends on the properties of the given language, especially on different means available for expressing FSP. It should be also mentioned that languages have at their disposal some special means even for the purposes of TP. Thus such expressions as English both ... and, on the one hand — on the other hand·, in the first instance — in the second instance;

122

F. DANES

etc., are often used in connexion with type (4). Every text (mainly in scientific or technical prose) is interwoven with expressions signalling significant points of TP of the text. The distribution of such expressions in a particular text might be termed its network of orientation. The ascertaining of the set óf these devices for each language, and their functional classification seems to be an important as well as interesting task. (Cf. now GÜLICH, 1970.) The study and knowledge of the thematic organization of texts have some practical applications as well, namely in practical stylistics and computational linguistics, especially in information retrieval. As for the former, the central question is how to construct and present (express) the thematic progression. As for the latter, we have to find out how to discover it, and how to make the concept of FSP and TP workable in the non-human conditions of a computer. But considerations of this kind are beyond the frame of the present paper. (Some suggestions will be found in DANES, 1970.)

At the end of section II let us try to find out the systemic correlations existing between the basic types of TP's and the basic types of multiple and condensed utterances (treated as transforms of the former). (1) Simple linear TP involves the following relevant relations: R.J = T 2 , T t =f= T 2 , Ri φ R 2 ; therefore it yields utterances with a complex T, or R; (2) TP with a constant Τ involves the following relevant relations: T j = T 2 , R t Φ T 2 , Ri φ R 2 ; therefore it yields multiple utterances, utterances with a multiple R and utterances with a complex Τ as well. (3) TP with derived T's involves the following relevant relations: T j Φ T 2 , R t φ T 2 ; therefore it yields multiple utterances. — Utterances with a multiple Τ are derivable only from a progression based on the schema T t R t + T 2 -» R 1; in which R's of Uj and U 2 are identical; such a progression may be considered a very rare modification of (3).

123

FSP AND THE TEXT ORGANIZATION Schematically:

U's U's U's U's multiwith with with with ple a multi- a multi- a com- a comU's ple Τ ple R plex Τ plex R

Types of TP's

Tj - Rj + T 2 ( = Rj)

R2

—

T, ^ R j + T 2 ( = TJ) —>• R 2

+

Ύ, ^ R 1

+

(T,

+

T2->R2

Rj + T 2

-

+

+

+

+

_

-

(+)

Rt)

III Linguists pursuing the analysis of intersentential relations in the text mostly interpret these relations in semantic terms, disregarding FSP. (One of the rare exceptions is Κ. E. H E I D O L P H (1966), whose approach implies FSP, without mentioning it.) Now a query arises whether or how the said semantic relations bear upon FSP. We shall content ourselves with raising some questions without trying to supply satisfactory answers. To start with, it will be useful to find out connexions between FSP and the semantic structure of the sentence. In my paper at the Tenth International Congress of Linguists in Bucarest 1967 (cf. D A N E S 1970b, 409) I suggested that the different semantic relations between R and Τ might supply a criterion for a linguistically relevant classification of utterances. A similar idea has been proposed by E. BENES (1968): "This relationship of the rheme to the theme can be regarded as the constituent act of utterance, just as the relationship of subject and predicate as the constituent act of a sentence" (271). He exemplifies his thought by the following utterances: the actual communicative aim or sense of the utterance Prague is the

124 capital

F. D A N E S

of CSSR

is the assignment of a particular quality to its

bearer, while the sense of the other utterance (revealing the same grammatical as well as semantic structure, and the identic lexical filling),

The capital

of CSSR

is Prague,

may be described as the

assignment of a bearer to a quality. Following this line we might dare propose a further generalization: A n y Τ — R nexus actualizes a particular semantic relation contained in the semantic (propositionäl) structure o f the underlying sentence, so that the communicative sense of an utterance ( C U S ) may be defined in terms o f the semantic function o f R-portion in relation to T-portion of the underlying sentence. 5 A German example: (la) Unsere Mutter schreibt ihre Briefe mit der Feder Semantic sentence structure: Ag-Act-Res-Instr Phonological shape: unmarked, centre of intonation (CI) on the terminal word (stress-group) Feder Diagnostic R-question: "Womit schreibt unsere Mutter ihre Briefe?" R: mit der Feder CUS: the assignment of an instrument to an agentive resultative action (or, more generally: indication of instrument) (lb) Mit der Feder schreibt ihre Briefe unsere Mutter Semantic sentence structure: the same as in (la) Phonological shape: unmarked, CI on the terminal word Mutter Diagnostic R-question: "Wer schreibt seine Briefe mit der Feder?" R: unsere Mutter CUS: the assignment of an agent (to an instrument used in a resultative action) (lc) Mit der Feder schreibt unsere Mutter ihre Briefe Semantic sentence structure: the same as in (la) Phonological shape: unmarked CI on the terminal word Briefe Diagnostic R-question: "Was schreibt unsere Mutter mit der Feder?" R: ihre Briefe CUS: the assignment of result (achieved with an instrument in an agentive action) Note: It is evident that the adaptation of a sentence to different contexts (resulting 5

An interesting attempt at a semantic classification of statements conveyed by different sentences has been made by GARVIN, BREWER, and MATHIOT ( 1 9 6 7 ) .

FSP A N D T H E TEXT ORGANIZATION

125

in different utterances with different CUS) merely by means of word order variations and/or of changes in the position of CI is possible in some languages only, and even there this possibility is not without restrictions. But I will not recapitulate here what is known from analyses done by Firbas and others, namely that the means for signalling FSP are various and numerous, word order and sentence intonation being only the most elementary of them (cf. DANES 1 9 6 7 ) .

It is evident that CUS, being defined as a function of R, plays no part in building up thematic progressions. But this does not mean, of course, that CUS plays no part in constructing texts. We may reasonably ask, e.g., what kinds of relations hold between the rhemes of concatenated sentences in a text. Thus in the following sequence of German sentences (2) (a) Dieser Brief kommt nicht von meiner Mutter her. (b) Meine Mutter schreibt ihre Briefe immer nur mit der Feder. CUS of (a) is the contradiction of a presupposed originator (source), and that of (b) is the assignment of an instrument. The sequence is based on a simple linear TP, schematically T a -> R a + + T b ( = R a ) -> R b . The semantic relation of R b to R a is inexplicit, but since (b) may be considered an answer to the question (a') "Warum kommt dieser Brief nicht von meiner Mutter h e r ? " (cf. the possibility of complementing (b) with the particle nähmlich, stating the relation explicitly), we may identify it as the relation of "reason". Schematically: t

I

In terms of CUS we may state that the indication of an instrument (CUS b ) supplies the reason for the contradiction of a presupposed originator (CUS a ). (It may be noticed as well that the semantic sentence function of R a ("originator") switches, when this item becomes T b (its new function being "agent"). But such a functional semantic switch has no structural relevance for the text; it is conditioned by the choice of the grammatical construction.) This being so, we may conclude (1) that (at least some) semantic relationships between concatenated sentences in a text (regarded by

126

F. DANES

many linguists as the text constituting relations — cf. Isenberg's notion of "Vertextungstypen") do not hold, in fact, between the whole sentences, 6 but only between their R's, and (2) that these semantic relations (or "semantic text functions") are of another kind than those involved by CUS (i.e., the intersententional semantic relations): the former necessarily belong to a higher level of abstraction, since they appear as functional implementations of the latter (i.e., CUS's are employed as means of expression of intersententional textual relations). But by far not all intersentential relationships belong to the same type as "reason" does, i.e., to the type of "causal" or "logical" relations (such as cause, consequence, concession,...). Another type is represented by the temporal and local relations; to another kind of abstraction belong such notions as "explication", "enumeration, or again, "adversative relation", "gradation", "conf r o n t a t i o n " , e t c . (cf., e.g., SKALICKA 1960, BECKA 1960).

Unfortunately, the classifications done by various investigators of text structure often fail to differentiate systematically the different types and levels of semantic abstraction (they often content themselves with semantic relations used by "traditional" as well as "modern" grammarians for the classification of clauses within compound and complex sentences, of adverbials, etc.), and what they offer seems to be an (unexhaustive) list of heterogeneous relations, lacking theoretical justification, a hierarchical order and objective classificatory criteria. Another analytical problem is prompted by the fundamental distinction of the "semantics of reference", and the "semantics of meaning" (Quine): It is necessary to differentiate the multiple factual extralinguistic relations existing between the denotata (i.e., objects or events) of concatenated sentential utterances, from that 6

Our interpretation is backed by the fact that in some languages particles and conjunctions explicitly expressing the given relation may be shifted from the beginning of the sentence and placed exactly before R (or R proper). Cf. the Czech version of (3) with the particle totiz (equivalent to G. nähmtich): "Tento dopis není od mé matky. Moje maminka píSe dopisy totiz vzdy jen perem."

FSP A N D THE TEXT ORGANIZATION

127

(those) relation(s) which the speaker has selected for his message and which he is now conveying by means of the specific linguistic meanings (lexical and syntactic) of language units he has chosen to this aim from the overall inventory of the given language system. Especially when the intersentential relation is not explicitly expressed, all the linguist may do is to find out all linguistically possible interpretations, i.e., interpretations prompted (allowed for) by the semantic context (lexical and syntactic meanings) in question. T o s u m u p : In respect to FSP, the generalized structure of a coherent text may be described in terms of an underlying thematic progression (representing the most abstract thematic relationships of several types) and a rhematic sequence of semantic relations obtaining between the particular rhemes. (It is not yet clear whether there exist standardized types of rhematic sequences as well, i.e., whether the rhematic sequences reveal an underlying pattern, as TP's do.)

REFERENCES BECKA, J. V., Základy kompozice jazykovych projevû (Fundamentals of stylistic composition). Acta Unwersitatis Palackianae Olomucensis, Philologica III, Praha 1960. BENES, E., Zaöatek nèmecké vëty Ζ hlediska aktuálního ¿lenëni vëtného [with a German Summary "Der Satzbeginn im Deutschen, von der Mitteilungsperspektive her betrachtet"], CMP 41, 1959, 205—217. BENES, E., On two aspects of functional sentence perspective, TLP 3, 1968, 267-274. DANES, F., Téma //(základ)// vychodisko vypovëdi [Theme //(Foundation)// Starting point of the utterance], SaS 25, 1964,148f. DANES, F., Order of elements and sentence intonation. In: To Honor Roman Jakobson I, The Hague 1967, 4 9 9 - 512. DANES, F., One instance of Prague School methodology: Functional analysis of utterance and text. In: Theory and method in linguistics (P. Garvin, ed.), The Hague 1970, 1 3 2 - 1 4 6 . DANES, F., Zur linguistischen Analyse der Textstruktur, Folia Linguistica 4, 1970a, 7 2 - 7 8 .

128

F. DANES

DANEÍ, F., Semantic considerations in syntax. In: Actes du Xe Congress international des linguistes II., Bucarest 1970b, 407—413. FIRBAS, J., On defining the theme in functional sentence analysis, T L P 1, 1964, 2 6 7 - 2 8 0 . GARVIN, P. - BREWER, J . - MATHIOT, M . , Predication-Typing. A pilot study in semantic analysis. Language Monograph 27, Washington 1967. GÜLICH, E . , Gliederungssignale in der Makrosyntax des gesprochenen Französisch. München 1970. HALUDAY, Μ . Α . Κ . , Notes on transitivity and theme in English. Part I I , JL 3, 1967, 199-247. HAUSENBLAS, K . , On characterization and classification of discourses, TLP I, 1964,67-84. HAUSENBLAS, K . , Krátká úvaha na téma "téma" [ A short paper on the theme "theme"], CL 17, 1969, 3 - 1 0 . HEIDOLPH, K. E., Kontextbeziehungen zwischen Sätzen in einer generativen Grammatik, Kybernetika 2, 1968, 274—281. ISENBERG, H., Ueberlegungen zur Texttheorie, ASG-Bericht Nr. 2, Berlin 1968. (Multiplicated.) ISENBERO, H . , Der Begriff "Text" in der Sprachtheorie, ASG-Bericht Nr. 8, Berlin 1970. (Multiplicated.) KLEMENSIEWICZ, Ζ., O syntaktycznym stosunku nawiazania, Slavia 19, 1949, 13-27. MATHESIUS, V . , O tak zvaném aktuálním ôlenëni vëtném [On the socalled actual bipartition of the sentence], SaS 5, 1939, 171 — 174. MATHESIUS, V . , fteiS a sloh [Language and style]. In: Cteni o jazyce a poezii, Praha 1942, 1 3 - 1 0 2 . PALEK, B . , Cross-Reference. A study from hyper-syntax. A UC, Monographia 21. Praga 1968. SGALL, P . , L'ordre des mots et la sémantique. In: Studies in syntax and semantics (F. Kiefer, ed.), Dordrecht 1969, 231—240. SKALICKA, V., Syntax promluvy (enunciace) [Syntax of the enunciation], S a S 2 1 , 1960, 2 4 1 - 2 4 9 . TROST, P., Subjekt a predikát [Subject and Predicate], Slavica Pragensia 4, 1962, 2 6 7 - 2 7 0 . VACHEK, J . , Some thoughts on the so-called complex condensation in modern English, SPFFBU, 1955, A3, 6 3 - 7 7 . WEINREICH, U., On the semantic structure of language. In: Universals of language (J. H. Greneberg, ed.), Cambridge 1963, 114—171.

ZUR FRAGE DER FUNKTIONALEN SATZPERSPEKTIVE IM DRAMATISCHEN TEXT Josef Filipec (Prag)

1. Bei der Analyse der Texte der schönen Literatur darf die ThemaRhema-Gliederung keineswegs vernachlässigt werden. Natürlich sind die Verhältnisse hier komplizierter als in rein kommunikativen Texten und Fachtexten. Das bedeutet aber nicht, daß man die schwierige Arbeit nicht unternehmen sollte. Ich habe zum Objekt meiner Untersuchung den dramatischen Text (DT) des heutigen tschechischen Prosaschriftstellers und Dramatikers Bohuslav Brezovsky, Nebezpecny vëk ÇDas gefährliche Menschenalter, Praha 1962) gemacht, der trotz aller Stilisierung den mündlichen Charakter der natürlichen Rede aufweist. Das Drama stellt den Kampf der jungen Leute und ihrer Eltern für die Wahrheit ihres Lebens und ihrer gemeinsamen Beziehungen gegen die bürgerlichen Konventionen dar. Die Eigentümlichkeit des mündlichen Textes äußert sich darin, daß dieser situationsgebunden ist. Unter S i t u a t i o n eines DT verstehe ich spezifische Beziehungen der Gegenstände einer Handlung (Personen, Ort, Zeit, Requisiten, Stimmung), also ζ. B. eine Liebesszene. Diese Situation des Textes verweist auf eine Situation des realen Lebens und kann mit ihr konfrontiert werden. Derjenige, der einen DT wahrnimmt, bewegt sich also in drei E b e n e n : Textebene, Szeneebene (charakterisiert durch szenische Anmerkungen) und Ebene der außerhalb des Textes stehenden Realität. Wichtig ist ζ. B. der Unterschied der realen, szenischen und der dramatischen Zeit: was sich zwischen zwei Personen früher ereignet hat, kann im Text erst nachher angeführt sein und der Leser oder Zuschauer erfährt es erst im Verlauf der Handlung.

130

J. FILIPEC

Der DT hat immer den Situationellen Hintergrund, er wird immer im Kontext der Situation1 wahrgenommen. Immer wird die erste Satzaussage von einer Person in einer bestimmten Situation gemacht, der sprachliche und vor allem der Situationskontext ist also immer anwesend.2 Auch bei der Analyse der funktionalen Satzperspektive (FSP) muß man dementsprechend zwei Linien berücksichtigen: die Textfolge I. (Sprachtext) und die Situationsfolge II. (Situationskontext). Die Thema-Rhema-Gliederung der Ebene I. muß durch die Elemente der Ebene II. ergänzt werden, und zwar auf die möglichst einfache und kurze Weise. Zum Beispiel spricht Professor Hába mit seiner Schwiegertochter Olga über seinen Schwiegersohn, der gerade aufgrund einer Amnestie vom Gefängnis entlassen wurde und im Nebenzimmer schläft (S. 64). Seine Ankunft bedeutet für die ganze Familie eine Komplikation. Vor allem die Ehefrau des Entlassenen, Professors Tochter, steht vor der Wahl zwischen zwei Männern und vor der Ehetrennung. Gerade diese Situation gibt der Textfolge einen anderen Sinn, den man rekonstruieren muß, um den Text zu verstehen. Die Textfolge hat also einen gewissen Subtext (s. weiter). Die erfolgreiche und genaue FSP-Analyse setzt also die volls t ä n d i g e R e k o n s t r u i e r u n g der T h - R h - S e q u e n z e n und ihres Zusammenhanges und diese wieder die vollständige syntaktische Analyse und die explizite semantische Interpretation und Analyse der Textaussagen in den beiden angezeigten Ebenen voraus. 2. Daraus ergeben sich gewisse m e t h o d o l o g i s c h e F o l g e r u n gen. Die FSP ist nicht der einzige textbildende Faktor, wie es z. B.

1

FIRTH, J. R., Personality and Language in Society. In: Papers in Linguistics (1934—1951), London—New York—Toronto 1964, 182. 2 Dazu auch SKALIÍKA, V., Text, kontext, subtext. In: AUC — Philologica 3, Slavica Prag. III., Praha 1961,76: „Der Kontext ist immer anwesend, d. h. die Einheiten des Textes stehen immer in einem Verhältnis zueinander und zur Situation".

ZUR FRAGE D E R FSP IM DRAMATISCHEN TEXT

131

aus Hallidays 3 Formulierung fließen möchte. Dieser kommunikative Mechanismus setzt vor allem die Schicht der semantischen und in gesprochenen Texten die Schicht der prosodischen Mittel und Klangqualitäten voraus. 4 Die thematischen Zusammenhänge konstituieren die Komposition der Texte. Die Mittel aller Schichten stehen in gemeinsamen Beziehungen, deren Zusammenspiel eine Gesamtqualität ästhetischer Werte hervortreten läßt, die in der kommunikativen Ebene ihre Grundlage finden. Der Zusammenhang dieser Schichten ist ein semiologischer, denn zum Beispiel die Gesten drücken den semantischen Inhalt aus oder begleiten und unterstreichen den Akzent oder die Intonation in ihrer mimischen Funktion. Zum Beispiel wiederholt der Vater die Worte seines Sohnes, der ihn über seine Absicht, einen Film zu sehen, informiert, mit großer Aufregung, die sich in Rufesätzen mit spezifischer Intonation realisiert. Jan (Vater): Wo gehst du hin? — Honzik: Nur ins Kino. Verkaufte Leben. Große Spannung. Frankreich — — Jan: Ins Kino\ Sol Große Spannung! Frankreich! Also keine Spannung und kein Frankreich! Lernen mußt du! Weißt du, daß du vor der Reifeprüfung stehst - -? (17) In der s e m a n t i s c h e n S c h i c h t eines Dramas sind mehrere Komponenten zu unterscheiden: a) Ebene der dargestellten D e n o t a t e (Personen, Gegenstände, Ort, Zeit, Stimmung) und ihrer Beziehungen, die mit den außerhalb des Textes stehenden realen Denotaten konfrontiert werden können, z. B. im Falle eines zeitlich unpassenden Kostüms oder der veränderten Nacheinanderfolge 3

HALLIDAY, M. A. K., The Place of FSP in the System of Linguistic Description, S. 44, in dieser Sammelschrift: „I would define FSP as the textcreating component of language". 4 Zu den Strukturkomponenten eines Textes vgl. auch FILIPEC, J., Zur sprachlich-stilistischen Analyse der tschechischen Übersetzungen der Dramen Bertolt Brechts. In: Deutsch-tschechische Beziehungen im Bereich der Sprache und Kultur. 2. Bd., Abhandl. der Sächs. Akad. der Wiss. zu Leipzig, Berlin 1968, 17—45. — Vgl. auch Abteil Drama in Knízka o jazyce a stylu soudobé ieskè literatury (Buch über Sprache und Stil der heutigen tschechischen Literatur), Praha 1961, SS. 161 — 195.

132

J. F I L I P E C

der realen Zeit usw.;5 b) Ebene der s i g n i f i k a t i v e n (eigentlich sprachlichen) B e d e u t u n g , die den sprachlichen Inhalt der Satzaussagen, ihren Sinn zum Ausdruck bringt, c) Ebene der p r a g m a t i s c h e n B e d e u t u n g , das heißt Konnotationen und zusätzliche expressive Qualitäten, die die Standpunkte der Personen untereinander, zur Äußerung der Partner und zur dargestellten Realität, ausdrücken, d) Ebene des S u b t e x t e s 6 in der Auffassung der Theatertheoretiker, d. h. des aktuell gemeinten Sinnes einer Satzaussage, der von der Handlung gefordert wird, ihr neue Entwicklungsanregungen erteilt und den eigentlichen Kontext herausbildet. Ad a) Die Repliken (Teiläußerungen der einzelnen Dialogpartner) oder ihre Teile befinden sich in dem semantischen Zusammenhang, der durch die Relation Stimulus — Reaktion 7 charakterisiert werden kann. Der Stimulus ist entweder die Replik eines Partners oder die vorangehende Replik des Sprechers selbst oder eine außersprachliche Äußerung (Lachen, Bewegung) oder Eingreifen einer Person oder eines Gegenstandes (ζ. B. spricht man über ein Lichtbild). Ad c) Dabei sind vor allem die pragmatischen Aspekte relevant. Diese zeigen sich in syntaktischer Hinsicht in der Bevorzugung von expressiven Sprachmitteln,8 die der Dominanz der emotiven Funktion entsprechen. Dazu gehört (1) die relativ hohe Häufigkeit von Rufesätzen, (2) die subjektive Reihenfolge und (3) die emotionell gefärbte Intonation und andere auditive Mittel. Über Interjektionen vgl. weiter. Ad (1) Die Statistik aus den ersten 100 Satzganzen des zweiten Teiles und aus den weiteren 100 Satzganzen anderen 5

Dazu die Ausführungen von R. INGARDEN, Das literarische Kunstwerk, 1931 und O poznawanitt dziela literackiego, Warszawa 1957 2 (übersetzt ins Tschechische, Praha 1967). 6 Vgl. dazu in dem von SKALICKA hier sub 2 zit. Aufsatz, 77f. 7 BOSÁK, C. - CAMUTALIOVÁ, I., Κ vystavbë dialogu (Zum Dialogaufbau), SaS 28, 1967, 2 3 7 - 245. 8 GREPL, M., Emocionàlné motwované aktualizace υ syntaktickê strukture vypovëdi (Emotionell motivierte Aktualisationen in der syntaktischen Aussagenstruktur), Brno 1967.

Z U R F R A G E D E R FSP I M D R A M A T I S C H E N T E X T

Charakters zeigt 50% Rufesätze mit oder ohne

Rufezeichen. 9

Rufesätze Satzganze

Aussagesätze

mit

ohne

Rufezeichen 1 100 S. 4 6 - 4 8

20

21

100 S. 5 3 - 5 6

23*)

24

Zusammen

I

zusammen

133

Fragesätze

Befehl- u. Wunschsätze

29

50

28

2

26

50

25

2

55

100

53

4

¡

1 43

45

i ! I

* Bemerkung: Die relativ höhere Anzahl der Aussagesätze entspricht hier auch dem absichtlich stereotypen Standpunkt der handelnden Person (Honzik: monoton, stereotyp usw.). A d (2) Der mündliche Charakter des D T bevorzugt die s u b j e k t i v e R e i h e n f o l g e in Verbindung mit der Inversion, Wiederholung und Gradation. Zum Beispiel: Jan (Vater): . . . Co zapírám? (Was leugne ich ab?) — Honzik: Podvádís— (Du betrügst —) — U

M

Jan: Coze! (Was denn!) — Honzik: Mámu! Mámu podvádís! w w Mutter! Die Mutter betrügst du! S. 55). Oder Jan: Co vis? w \\ \\ weißt du denn?) — Honzik: Vsechno! Vsechno vim. (Alles! weiß ich. S. 57). Oder Hába ... Jenze to uz neni nie platné, ty M

hovory. Je pozdë! 9

Ulicnika jste ζ nëho vychovali.

(Die (Was u Alles vase

M

Lumpa —

(Nur

Vortreffliche Belehrung über Satztypen findet man bei SMILAUER, V., Novoéeská skladba (Neutschechische Syntax), Praha 19662, hier 34ff. Auch K O P E C N Í , F . , Základy íeské skladby (Grundlagen der tschechischen Syntax), Praha 1958 und BAUER, J . - G R E P L , M . , Skladba spisovné iestiny (Syntax der tschechischen Schriftsprache), Praha 1972.

134

J. FILIPEC

daß sie keine Suppe fett machen, diese Plaudereien. Es ist schon u \\ spät! Einen Lausbuben habt ihr erzogen. Einen Gauner — S. 59) Ad (3) In den angezeigten Fällen wird das Rhema und auch das Intonationszentrum gegen Anfang der Aussage vorgeschoben.10 Natürlich kommen dabei oft auch andere p r o s o d i s c h e Mittel, wie Akzent, Intonationskadenzen, Pausen, Rhythmus (ζ. B. bei den Enklitika) und Timbre, zum Vorschein. Der Akzent dominiert ζ. B. in folgenden Aussagen (in der ersten ist das Rhema vorgeschoben): w Jan: Jestli to mâ nëkdo ν neporádku, tak jsi to ν prvé rade asi ty — \ \

\ v

\ \

Honzík: Já to nezapírám. Jan: A kdo? Snad já — Honzík: Ty! w (iSpricht man von Unordnung, dann ist das vor allem deine Sache —. w

Wörtliche Übersetzung:... dann bist es in erster Reihe Du! — Diese merkmalhaltige Formulierung anstatt des Merkmallosen: Und du hast alles in Ordnung? oder: Aber du hast doch Unordnung! — Honzík: Ich leugne es nicht. — Und wer? Vielleicht ich Du! S. 55) Oder eine andere Stelle: Honzík: Jâ opravdu nevím, kam chce w p o t o m jit — Olga: Jak to p o t o m ? ! Ty vis, kde je ted —? (Ich w weiß wirklich nicht, wohin sie nachher gehen will — Olga: Wie so \\ u nachher?! Du weißt, wo sie jetzt ist — ? S. 70) 3. Man darf DT nicht mit denselben Kriterien beurteilen, die auf Grund der Analyse der Texte anderer Gattungen und nicht mündlichen Charakters gewonnen worden sind. Der D T h a t eigene W e s e n s z ü g e und gehört zum Texttyp mit eigenen Normen. Auch 10

DANES, F., Intonace a vëta ve spisovné cestinë (Intonation und Satz in der tschechischen Schriftsprache), Praha 1957. In unserem Zusammenhang vor allem 55fF. und 75ff. Auf 76 spricht er auch über Theaterstücke. — Derselbe Autor, Order of Elements and Sentence Intonation. In: To Honor R. Jakobson, The Hague 1967, 4 9 9 - 5 1 2 .

135

ZUR FRAGE DER FSP I M DRAMATISCHEN TEXT

die syntaktischen Verhältnisse sind arteigen, wie eine kleine Übersicht aus den 200 früher angeführten Satzaussagenganzen (SAG) des DT zeigt:

Fragment.

1-Glied-Aussagen κ η ε B c e r ^ a . 3 . T e M c a M U M n p n X a p a K T e p H C T H K e Η Η φ θ ρ Μ 3 Τ Η Β Η Μ Χ UeHTpOB B b I C K a 3 b I B a H H H B

Ka-

n e c T B e p e j i e B a H T H b i x B b i c r y n a i o T K O J i H H e c T B e H H b i e n p n 3 H a K H , β icanecTBe HepejieBaHTHbix — KaiiecTBeHHbie. II.

K a K H e ace φ 0 ρ Μ 3 ^ Η 0 - Η 3 Μ Κ 0 Β Β ΐ ε

cpe^cTBa yKa3biBaioT

Ha

HaJIHHHe Β B b I C K a 3 b I B a H H H H e K O T O p b l X Η Η φ ο ρ ί ν ^ Τ Η Β Η Η Χ I i e H T p O B ? Kan

ocymecTBJiaeTCH

opramoaiiiisi

BbicKa3bmaHHH

β

jihhchhom

njiaHe? T l p e a c ^ e B e e r ò , B b i p a a c e H H i o KOMMyHHKaTHBHOH C T p y K T y p b i B b i cKa3biBaHH« c n o c o ô c T B y e T n p H c y m a s ycTHOMy e r o pHTMHHecKaH o p r a H H 3 a u H H . Π ο

cyTH

flejia,

BonjiomeHHio

t o t h o t hhoìì p h t m

B c e r f l a conyTCTByeT y c T H o ñ penH. H a n ö o j i e e xapaicTepHhiH Hañ

—

nepeflOBaHHe y ^ a p H b i x

h

6e3yaapHbix

3BeHbeB

uenH, conpoBOKflaioineeca cooTBeTCTByiomeH jihhchhoh

paccTa-

HOBKOH O T f l e j I b H b l X SJieMeHTOB B b I C K a 3 b I B a H H H (CJIOB). n p H β KanecTBe y a a p H o r o h 6 e 3 y a a p H o r o

cjiy-

peneBoñ

3TOM

3BeHa M o a c e T B b i c T y n a T b

η

coöcTBeHHo φοΗετκιεοκαΗ eflHHHiia (MeHbmaa, neM cjiobo), h o OÖbIHHO, Β yCJIOBHHX TeCHOTO B3aHMOfleHCTBHfl p H T M a H CMblCJia Π ρ Η yCTHOM npOH3HeceHHH, T a K H M 3BeHOM HBJIHeTCa eflHHHija ( c j i o b o ) . Y f l a p H o c T b n e p e B o r o Bcex

cjiyiaax

3BeHbeB CJIOBa. φρ33β

co

MoryT

cjiOBecHbiM

BbinojiHHTb

KojIHHeCTBO

3 B e H a He c o B n a a a e T

y^apeHHeM

h

—

pojib

nojiHoy,ziapHhie

pHTMHKO-CMbICJIOBblX

OÖbIHHO M e H b l U e ,

3HaHHMaa bo

6e3y/iapHwx

3HaMeHaTejibHbie

yflapeHHH

yCTHOH

Β

HeaCejlH KOJIHHeCTBO B X O a « m H X

Β

Hee

c j i o b . O c o 6 a a p o j i b 3flecb npHHafljieacHT s h k j i h t h h c c k h m η π ρ ο KJIHTHHeCKHM 3 J i e M e H T a M . P H T M H K O - C M b l C J I O B a a y f l a p H O C T b OÖbIHHO c o n p o B o a c f l a e T C H η O T ^ e j i t H b i M .nBJDKeHHeM TOHa. B o t pbie xapaKTepHbie npHMepw crpoeHHa

CKa3biBaHHJi:

\ —

\ a cboh

MaTpacmc |

ν j.

ì >

a c

B c e

bh-

*

c K w a J, n p H T a m y j . : \

Π ο Η ^ ε Μ β MaraHHH. — A o h h t m

Ηβκοτο-

noBecTBOBaTejibHoro

\

—

\

3 a K p b i T b i J,; — 3 a B T p a j.

\

Mbi n o n p o c H M Î o h cfleJiaeT { ; — E c j i h η He 6ypy c n a T b , τ ο Ha

Λ

\

\

>

K o f i MHe Η β ρ τ 3 T a HyxcHa G o j i b H H u a . Π ρ κ 3 t o m M o r y T B 0 3 H H K a T b

162

O. A. JIAHTEBA χ \ paBHOueHHbie BapnaHTU, cp. : OneHt flojiro ohh Ghjih β kjihhhkc = Λ X X = ΟΗβΗΒ OHH 6mJIHflOJITOΒ ΚΛΗΗΗΚβ. HHOr^a pHTMHHCCKdfl OpraHH3amis βμοκ33ϊ>ιβ3ηηη oKa3MBaeTCH CBH3aHHoíí c napajuiejira\

MOM CHHTaKCHHeCKOrO CTpOCHHH, Cp.t — AnejIbCHH«fflK XOieTCH \ \ \ Β35ΠΊ» nococan. η hôjiohko; IlojioxcHJi Ha CTOJi, y6paji β Kop3HHy, X X X X B3HJI c C060H aeTen; — Cnipajia, njiaTKna 3a KBaprapy, Bcë (ποcjie^HHH njieoHacTHHecKHH MecTOHMeHHBiH 3JieMeHT Ao6aBjieH paan yKpenneHHH npHHmma napajieJun>HOCTH cTpyKTypti BbicicaX

\

3biBaHna). Π ο t o h ace cxeMe: — Bo ma 6pouiy β οκοπικο, bo rypß. nojieTHT (MajiBHHK HTpaeT β öaHOHKy). 3Ta ace cxeMa MoaceT coömoflaTBCH η πρΗ nepe^aie pa3roBopm>ix HHTOHauHìi Ha iracbMe, \ \ cp.: riepeflo mhoh cjioBapHK jieacHT, 1621 r. h3ashhh, crapule \ IleTpa I KHnaceiica (h3 lacTHoro HHCbMa). HHOrfla npH pHTMHHeCKOH OpraHH3aHHH BbICKa3BIBaHHH 6e3yaapHbie 3Bem>H Moryr HacraiHO hjih nojiHOCTbio oTcyrcTBOBaTb, X X X X cp.: — Bce y Hac l y r ckjiokh Glijih H3-3a Heë. 06ηηηο sto npoHcXOflHT npH npeAeJIbHOM HaCbimeHHH BbICKa3bœaHHH HHfjlOpMaTHBHO 3HaHHMbiMH sjieMeirraMH, cp. β BonpocHTejibHbix pemiHicax: X X X X X X — Mara3HH 11 3flecbf / GyaeT? j.; — flßa j. / KOMy f / ÖHJieTa? | (xoth η 3«ecb B03M0acH0 eme 6ojibuiee ΗΗφορΜ3ΤΚΒΗοε cacaTHe — nocjieflHHe sjieMeHTti MoryT OTcyrcTBOBaTb, npeepaman pemiHKH β flBycnoBHbie — β nepBOM cjiyiae H3-3a cjiyaceÖHoro xapairrepa nocjieAHero ajieMeHTa, bo btopom — H3-3a CHTyaraBHOÖ onpeaejieHHOCTH). OöpaTHM BHHMaHHe Ha HajiHHHe nay3 Meacoy y^apHblMH 3BeHbHMH. fljIÄ CpaBHeHHH ΠρΗΒβΛβΜ BOnpOCHTeJIbHyK) peX

ruimcy c 6e3yziapHbiMH 3BeHbHMH: — A Tenepb eñ npnexaTb 4 X X Tbl 3aHHTa t β 3TO BpeMH 4 Hejn.3«? t CjieflyeT otmcthtb, hto yaapHocTb η 6e3yflapHocn> sjieMeHTa

163

AKTyAJItHOE HJIEHEHHE Β ΡΑ3ΓΟΒΟΡΗΟ« PEHH

Β BHCKa3HBaHHH Ηβ e C T b ΗβΗΤΟ a 6 c O J I K ) T H O e , 3 T H ΠρΗ3Η3ΚΗ Ηβ

oôjiaaaiOT cbohctbom τοηηοη Φιτκοειιηη hmchho

Ha flaHHOM 3 J i e -

MeHTe. C y m e c r a e H H H M OKa3biBaeTc$i cairn,™ n p m n m n

pHTMHHec-

k o h opraHH3an.HH, K O T o p b r i i n a m e B e e r ò n p e / j c T a e T β BHfle n e p e flOBaHZH

y ^ a p H B i x η 6e3y.napHi.ix 3BeHbeB.

3 τ ο τ πρΗΗϋ,ΗΠ 0 K a 3 B i B a e T C H y a o Ô H b i M A J i a B o n j i o m e H H H h j i h -

ΗβΗΗΟΓΟ npeflCTaBJieHHH J Q ^ O p M a T H B H b l X I^eHTpOB. OÖHHHO Β K a necTBe yAapHBix 3Bem>eB φpa3bx BHCTynaiOT HHφopMaτHBHO 3HaHHMBie 3JieMeHTBI, φθρΜΗρ>ΤΟΙΐΧΗε KOMMyHHKaTHBHUH CMBICJ1 BblCKa3bIBaHHH. IlOCKOJIbKy ΗΗφθρΜ3ΤΗΒΗΗΧ HeHTpOB Β sjieMeHTapHOH φρ33β name Bcero öbiBaeT flBa (xoth sto η He o6a3aTejn>Ho), a yaapHOCTb hjih 6e3y^apHOCTb 3BeHa — npn3HaK noflBHHCHbiH, OCHOBHblX yaapHblX 3BeHbeB Β BbICKa3bIBaHHH npH COBHaaeHHH HX c HHφopMaτHBHbIMH ueinpaMH Taicxce oôbiHHO 6bœaeT ΚΒΆ. Π ρ κ 3ΤΟΜ Β pOJIH ID^OpMaTHBHOIO IieHTpa MOryT BbICTynaTb CHHTaKCHHecKH pa3JiHHHHe uieHbi npeAJioxceHHH, 3aHHMaTb MecTO 6e3yflapHoro 3BeHTa T a o c e MoryT c n H T a K C H i e c K n η ΜορφοΛΟΗΓκεοκπ pa3JiHTHbie HJieroi (name r j i a r o j i , MecTOHMemie, H a p e n n e — ohh jierne cTaHOBSTca 3ηκπητηκ3μη h npoKJiHTHicaMH, η sto o6ctohTejiBHO He BCTynaeT β προτΗΒορεπΗβ c pa3MemeHHeM KOMMyHHicaT H B H o i í H a r p y 3 K 0 H a HJieHax B b i c K a 3 b i B a H H a ) . C p . : y

Bac

3flecb Ηετ

f

n a nepBOM STaxce?;

—

— IÜHypKOB j.

TaM hhkto î mhoh

He

\ HHTepecyeTCH?î ; — B b i 6yn;eTe \

HTO nenieTCH n p o c T O r o j i O B a ; KojieHbKa, T a n ™ KapMaHa

jih!; H h

mhtb

r o j i O B y ? — Ά 6y.ny, noTOMy

\

j

t u cboh \

/ He 6 p a j i \

f

\ — Τ η

6pocnji;

\ / MejioHb?

*

c p a 3 y e m e He e a p a n H ,

j,;

— Τ ω y Memi —

3apoäobo

—

οληη

τ β J l a ^ y MHe H e K p a c o B y ;

—

n a j n > T o H3 \

Mbi B i e p a

CBapHTb; — Τ ω

n j i o m a f l K e ? — Ά CHanajia Ha m i o m a ^ K e , a

pac;

h3

\

μοχηο

T a M CHjibHO H a i a n o . n y n . ; Ά e m e

\

οτκροκ ce6e οκοπιεικο;

ποτομ

yace

nrpa-

ctohji β

Ha

BaroHe.

6 y a y noKynanb c e 6 e MaT— Hy

κοηθηηο hx

ceìreac

164

O. A. JIAIITEBA \

\

Ηβτ. OHH TOJIBKO GbraaiOT oceHLio; — ΗΟΓΗ,

Bce

\ ΑΘΦΟΡΜΗΡΟΒΕΗΗΒΊΕ, H TOJIbKO

Y

M e m i o i e H b GojibHbie \

Mory HA ΜΗΓΚΟΜ; — TaM

Λ X 6MJIO neTbipe HJIH ΠΚΊΊ> φορΜ TOJIbKO, a ceñiac ropa3flo HX 6OJIB\ \ me; ^lepeanyp Bcer^a npHMepHbi. H a 3TOM CTOHM; — HaBepHoe, \ \ \ \ Λ MACTB Y^ajiHJiH ΚΗΠΙΟΚ; ΠΟΤΟΚ TaM MEIIIHH — yjimiy He nepeñflenib Ι Λ \ HH 3a HTO; H a ypoBHe pemajiocb 3aM3aBa; — Ί τ ο MM .uejiaeM AJIH \

Toro, HTO6Í>I noBbimaTb ypoacañHocTb, aaTb öojibiue rocyaapcTBy npOflyKUHH? Β πρΗΗυ,Ηπε Ty »ce opraHH3auHio HMeioT BbicKa3biBaHHH, Β koTOpblX HHCJIO y/japHHX pHTMHKO-HHTOHaiIHOHHblX 3BeHbeB yBeJIHHHBAETCH AyMaio? Λ Ά jiyMaio, \ ΛΟ Tpex, cp.: Λ — Tbi 3Haeuib κ I T O \ MTo NHCATB MONCHO MY3HKY coBepmeHHo ΟΠΟΚΟΗΗΟ; MRE xoTb 6bi \ \ \ \ Äpyryio KapTOHicy;... Η 6e3 KOHiia, 6e3 Komja C HHM TaKHe npoHcxo-

\

\

\

\

ΛΗΛΗ HCTopHH; Ά YACE Β Tpex 6biJia Mara3HHax xo3HHCTBeHHbix — \ \ \ HHRFLE najiKH HETY; ...BOT caMbiñ CHHTAETCH JIYQUIHH φκκοοοφ Χ \ STO TameHKO Β Jlemmrpa^e; Ka/ipoe

\ 3flecb KpacHBbix

oneHb

MHoro 6yaeT; O i e H b ΜΗΟΓΟ GtiJio HHTepecHbix Bemeñ; Κρεπκο \ * \ \ OH coacajieji, HTO 6pocHJiH ero Ha xo3añcTBeHHyio pa6oTy. C p . c OAHHM yaapHHM 3BeH0M: — A 3Haeuib, ecTb najiaTKH no τρΗΑ-

iiaTb pyójieñ. — Ά 3Haio, η Τ3ΚΗΧ an^esi ΜΗΟΓΟ najiaTOK. Β κομMyHHHKâTHBHOM njiaHe yBejiHHeHHe nncjia y^apHtix 3BeHbeB 03HanaeT pa3flejibHoe npejjcTaBJieHHe, pacmeiuieHHe MHoroHJieHHoro ΗΗφσρΜ3ΤΗΒΗθ BaacHoro ijeHTpa (name Beerò — Ρ cyxcfleHHa) β ijejiax ero Jiyiuiero npe^cTaBjieHHH Β ΒΜΟΚ&3ΗΒΕΗΗΗ. KaK BHflHO H3 npHMepOB, pHTMHKO-HHTOHaUHOHHaH CXeMa ΗβρβOLOBAHHH YAAPHBIX H 6E3Y^APHBIX 3BEHBEB OKA3BIBAETCA ONEHB

AKTYAJIbHOE HJIEHEHHE Β ΡΑ3ΓΟΒΟΡΗΟΚ PEHH

165

yflOÔHOH AJI5I BapBHpOBaHHÄ Β JIHHeÍÍHOM p a 3 M e m e H H H KOMMyHHKaTHBHO 3HaHHMbIX 3JieMeHT0B. M o a c e T

6bITb, TOJIbKO HapeHHH

noKE3biBaK>T TeHfleHUHio κ HanajibHOMy n o j i o a c e H H i o , o c T a j i b H b i e Μ ο ρ φ ο Λ Ο Γ Η Η ε ο κ κ θ p a 3 p a a b i CJIOB M o r y T 3aHHMaTb ΠΟ3ππ;ΗΙΟ JIK)6Or o y ^ a p H o r o 3BeHa. /JJIH H e K O T o p w x ΤΗΠΟΒ KOHCTpyiciyiH, n p e a c a e B e e r ò fljia TaK Ha3biBaeMbix k o h c t p y k u h h c HMeHHTejibHbiM T e M b i , o c o ß e H H o T e x H3 HHX, KOTopbie npeacTaBJieHbi B o n p o c H T e j i h H b i M H p e i u i H K a M H , MONCHO OTMeTHTb cTpeMJieHHe κ BhiHocy Β H a n a j i b H o e n o j i o x c e H H e HHcfjopMaTHBHo 6 o j i e e 3 H a i H M o r o s n e M e H T a ; OHO XOp o n i o c o r j i a c y e T c a c B03M0acH0CT»MH p a c c M a T p i r e a e M O H

cxeMbi.

Β TO »Ce BpeMH ΠρΗΗΙίΗΠ pHTMHHeCKOH OpraHH3aUHH BbICKa3bIBaHHH MOaCeT BOHTH Β CTOJlKHOBeHHe C He06x0flHM0CTbK>

nOfl-

HepKHyTOrO npOH3HeceHHH HH(J>OpMaTHBHO 3HaHHMbIX 3JieMeHTOB. ΠΟ

HaniHM HaöjiioaeHHHM,

Β 3T0M

c j i y n a e OÔHHHO

no6eac,o;aeT

pHTM. T a K , Β c j i e ^ y i o m H X npHMepax r j i a r o j i , H e c y m n i i ocHOBHyio m^opMauHjo,

3aHHMaeT

MecTO

6e3y^apHoro

3BeHa:

—

Hmca,

\ χ \ \ rpHÔbi n e p e c o x j i H ΤΒΟΗ; — ^ τ ο ace STO, MOH ocTaioT nacbi HT o JIH, fla? B O T T e Ha. T o »ce C HapenneM: — B b i yace n p H u i J i H , ΜΗΟΓΟ Ha\

\

\

p o a y 6biJio, fla?; — O H MHe He HyaceH. — M e c r a M a n o 3aHHMaeT, yaoÔHO

(3^ecb

cbirpaji

CBOIO

pojib

nprnnmn

napajuiejiPOMa).

HHOTAa TpeôoBaHHH cxeMbi 3acTaBJi«ioT aKneHTpapoBaTb Η Η φ ο ρ MaTHBHO He 3HaHHMoe CJIOBO (B c j i e ^ y i o m a x n e r a p e x n p H M e p a x — \

HapeHHe HJIH n a c T H u y ) : — 3 a HHHKH-TO 3 f l e c b n p o Ö H B a i o T HaBepHO Toace, fla?; — T b i acapb BEE: a Be AB B c r a H y Toace 6ypy \ \ X \ Haa neM ceßnac paöoTaeTe?;

...nyTKHH

e c T b ; — Bbi \

OH a B T o p Bcë-TaKH,

a?

JlaÔHjibHOCTb npH3Haica „ y ^ a p H O C T b — 6 e 3 y a a p H O C T b " Β y c j i O B n a x yCTHOrO npOH3HeCeHHH MOaceT npHBeCTH Κ OTCyTCTBHIO ΟΛΗΟΓΟ H3 y^apHblX

3ΒβΗΒβΒ, 3aHHMaeMbIX X

3JieMeHTOM, c p . :

—

KOMMyHHKaTHBHO \

BaaCHbIM

OieHí. y Hac H e n p H j m n > i ñ ¿ m p e r r o p

H3^a-

TejibCTBa 6biJi STO M m i e H T b e B ; — BOT y Hac e m e ΟΛΗΗ nonyTHHK

166

O. A. JIAIITEBA

Λ noflBimciî; — 3HaeTe ito, y Hac 3aBTpa ^οβολβηο 6y.neT TpyçHHH \

fleHB, TaK HTO aaeaHTe; — ^oporaa, TaK βη iijiotho oGcejin... KaK bhaho, HCCMOTpH Ha nponycKflHHaMHnecKoroycmiemia, nopjIflOK CJieflOBaHHH 3JI¿MCHTOB OCTaeTCS TaKHM xe, KEKHM oh 6mji 6hi 6e3 3τογο nponycxa. 3flecb XOTeJIOCb 6bl OCTaHOBHTbCH Ha HeOflHOKpaTHO BbICKa3bIBaBnieMCH HaÓJIIOAeHHH, COrjiaCHO KOTOpOMy Β yCTHOH pCHH ΠΟpazioK cjioB noflHHMeTca npmmnny accoijHaTHBHoro ηεηιϊ3ηβεhhs, cjie,nyeT HenocpeflCTBeHHoS CMeHe npeacTaBJieHHH. 3το 6buio 3aMeieHO He tojibko cneunajiHCTaMH. A. H. Kynpra, HanpHMep, HHcaji: „Ohh (cjioBa - O. JT.) BbiJieTaioT ra ee pfa 6e3 βοικογο nopa^Ka, Tax hto nopoa KaaceTca, i t o mecTHaflHaToe no ciery onepe»aeT TpeTbe" („Cama η £nia"). 3το τοΗκοβ 3aMenaHHe, KaK h Bee HaôjnoAeHHH sToro po/ja, HecoMHeHHO, HMeeT noa co6oh noHBy, OAHaKO geñcTBHe npHHmraa accoipiaTHBHoro npacoe^HHeHHH npH óJiracanmeM paccMOTpeHHH OKa3biBaeTCH He 6e3rpaHHHHWM h ocymecTBJweTCfl bo tapase jiHini. β onpeflejieHHtix ycjioBH5IX. Haaôojiee aicTHBHO stot nprnunn nposBjraeTCH β npe^ejiax 6e3yaapHoro phtmhko = HHTOHaixHOHHoro 3BeHa pa3bi. 3jieMeHTH, cocTaBJifltomne sto 3bcho, Moryr cboöoaho mchhtbch MecraMH (eCJIH HXflBaHJIH HeCKOJIbKO), B3aHM03aMemaTbCH, HTO C03AaeT mHpOKOe B03M03CH0CTH PJIH Β03ΗΗΚΗΒ0βΗΗΗ 3KBHBajieHTOB = BapHaHTOB. Cp.: — AjieKcaH^p IlaBJioBHi, CKaHorre, a Kor^a B a m 6tur nepBbiH peñe Ha ΊΎ-104?; — Eme JiHTepaTypti HaflOTaK \ \ μ η ο γ ο npoHecTB (cp.: — TaK μ η ο γ ο jiHTepaTypw eme Haflo npoHHTaTb); — Η ποτομ oh îkhji β XaMOBHincax, HbiHe r a e \

My3eñ; — Βο-nepBbix, HyacHO tojibko noKynaTb BpamaioiuyioCH ÖpHTBy; ΚοΗβΗΗΟ, 3flecb ΜΗΟΓΟ MOXCeT 6bITb TOJIbKO npeanonoaceHHH, a peajibHbix BbiBO^OB 6y^eT h He TaK μηογο; — npe^cTaBjiHK), KaKHe y Bac TyT 6e3 μθηη Bbipocjm Ha Horax Η0ΓΤΗ (cp. B03M03KHBie 3KBHBajieHTbi: y Bac BbipOCJIH TyT 6e3

AKTyAJILHOE HJIEHEHHE Β ΡΑ3ΓΟΒΟΡΗΟΪΪ PETO 167 \ \ M C H H ; y B a c τ γ τ B b i p o c j i H 6 e 3 M C H H ) ; Ά T e 6 e ΠΟ3ΒΟΗΚ> n e p e 3 \ Λ \ MHHyΤ

AecHTt;

a y MEMBI T O J I I . K O 6 b i J i a

cyMOHKa Β p y i c a x ;

. . . T e n e p b y MCHH n o j n r o a H j r r m i T O J I B K O 6 y a y T ro

HHCJIA;

— Kro-HH6y,zib

3y6bij,

6paJi |

B0ceMHa;maT0-

HHCTHTB

Kpyacicy?

3 H a e T e , a x o T e j i a HTO e e c n p o c H T b ? C a M a H e o ß x o A H M O C T B c o 3 f l a H H x BO 4>pa3e 6 e 3 y A a p H o r o

ρπτ-

MHKO-HHTOHaiXHOHHOrO 3 B e H a B e f l e T Κ CBOÔOflHOMy p a 3 M e m e H H K ) 3JIEMEHT0B c j i y x c e 6 H o r o HA3HAHEHHH, BBO^HBIX CJIOB Η Π Ο Λ , c p . : \ \ ^ — D i e x o T b EAEM, O6I>HBJMJI 6 M ; — Α . , B 0 3 b M H T e CBOK>, n o a c a j i y ñ c T a , n o c T e j i b ; B b i He ΜΟΓΛΗ 6 b i T a M n o x c a j i y ñ c T a OTOPEATB ÔHJieTHKH?; a T a i o K e κ c B o G o f l H O M y BKJiHHHBaHino CJIOB Β c o c T a B φ p a 3 e 0 J I 0 Γ H 3 M a ,

cp.:

— A

6e3yflapHbix

CMMCJI c T O H T b

ecTb

Β o n e p e f l H , Bbi flyMaere?; — ECJIH Η He 6 y ^ y cnaTb, τ ο H a κ ο ί ί \ \ \ M H e π ε ρ τ 3 T a HyacHa G o j i b H i m a ? E e s y q a p H o e 3BeHO M o r y T c o c r a B -

jiHTb H TaKHe cHHTaiccHHecKHe 3JieMeHTBi, pacnojioxceHHe κοτορΗΧ Β yCJIOBHHX IIHCbMeHHOH ΚΟΑΗφϋΙΙΗρΟΒΕΗΗΟΗ ρβΗΗ C T p O I O φίΙΚΟΗpoBaHO, c p . : — H e p H b i ñ κ ο τ TaM aôeflHHHaji. — Η τ ο ? — H e p H b i ñ κ ο τ A E M Y R O B O P I O H T O TAM NÓE^HHIAJI (XOTH OÔHHHO CBOÔOFLA c j i o B o p a c n o j i o a c e H H H Β n p e ^ e j i a x 6 e 3 y a a p H o r o 3 B e H a He H a p y m a e T 3 a a a H H O C T H cHHTaKCHHecKOH M O f l e j i H ) . H a K O H e n , 6 e 3 y ^ a p H o e 3 B e H o M o a c e T p a 3 M e m a T b c a BO ΦPA3E n o NPHHNPNY

flo6aBJieHna,

BO3-

B p a m e H H f l κ H e c K a 3 a H H O M y , c p . : — T b i w r a j i a Î , KaK ΟΑΗΗ ΒΟΛΟjia3 Β „ I Ï 3 B e c T H 5 i x " norpyacajicH |

Β TpiOM TaHKepa

3anoji\

H e H H o r o Η ε φ τ Β ί ο ? Ι ; H e 3 H a i o , i c y n a OHH e r o

fleHyr.

HJIH

οτειι

K y a a n o e a e T B 0 3 b M e T C M a T e p b i o C COÖOH. O c o ö e H H O H a c T o n o n p H H Q m i y ÄOÖaBJieHHa o p r a H r o y r e c H

6e3-

y f l a p H o e 3 B e H o , p a c n o j i a r a e M o e Β K o m j e BbtCKa3MBaHHH Β K a n e c T B e 3JieMeHTa n o a c H H T e j i b H o r o x a p a K T e p a , oTcyTCTBOBaemero Β n e p BOHanaJIbHblX

KOMMyHHKaTHBHblX

ycTaHOBKax

npH

nocTpoeHHH

168

O. A. JLAJITEBA

BBICKA3I>IBAHH5I. C p . : — O H BEßB YE3HCAET BOTFLEBHTORO,C ACTCKHM \

\

c a a o M O T n p a B J i a e M M M e r o ; — A Μ β τ ρ ο / r ^ e 3flecb 6 j i H 3 K a ö m e e ? ; — B b i y aceHmHHH JieHHTecb H e B p o n a T O J i o r a ? ;

— A

rae

\

n a y c T O B C K H H ? — Β p r o s a i c e riaycTOBCKHÜ JIOKHT Η o ; B o c e M H a a u a T a a M H H y r a 3aKaHHHBaeTca

npecnoKoií-

BToporonepnoaa;

— 3 H a e u i b I T O MHe HyacHO? Kaicyio-HHÔyzib C T a p y i o m e T K y .

Moa

\

c j i o M a j i a c b B b i H n i u a T b n b i j i e c o c ; C M e T a H a He HyacHa. T a M e m e C T a p a a e c T b C M e T a H a ; — H y , M U o n e H b MHJIO C Heñ n o r o B o p a n H , TCMy

6 b i j i H o n e H b p a ^ H yneHHKH B c e ΚΟΗΘΜΗΟ.

npOHJIJHOCTpHpOBaHHblË ΠρΗΗΗΗΠ flOÔaBJieHHH I H ^ O p M a T H B H O M a j i o 3 H a H H M o r o 3JieMeHTa x o p o r n o c o r j i a c y e T c a c o CTpeMJieHneM KHanajibHOMy pacnojioaceHmo

Hanôojiee

3HaHHMoro

sjieMenra,

K O T o p o e HaÔJiioflaeTca Β H e K O T o p w x c j i y n a a x Β3&ΗΜΗΟΓΟ p a c n o JioaceHHH y a a p H b i x 3BeHbeB, n p e / i c T a B j i a i o m H X

ΗΗφορΜατΗΒΗΗβ

q e H T p w HJiH HX n a c r a . Π ο ρ Η ^ ο κ HX Β33ΗΜΗΟΓΟ cjie^oBaHHH o K a 3biBaeTCH He6e3pa3OT«mbiM T o r a a , K o r e a ο λ η ο H3 HHX n p e a c T a B j i a e T CJIOBO, H e c y m e e ß o j i e e BaacHyio m H f r o p M a m i K ) H H M e i o m e e ß o i t e e sBCTBeHHoe

flimaMHHecKoe

ycHjiemie n p a

npOH3HeceHHH.

Β 3ΤΟΜ c j i y i a e OH Η Μ β ε τ T e H / i e m m i o κ p a c n o j i o a c e H H i o β H a n a j i b HOÍÍ LACRA BBICKA3BIBAHH5I. C p . :

— M o a c e T 6μτί>, MHe yace KOH-

HHTb e c T b T a K y i o e s y H e c y c B e r a y i o ? 3 H a e T e , H a ^ o e j i a yace o n e t i b ; — Bce? -

P.a.. ^AACE HexojibHHK n p o c j i e f l H j i , HTOÔ y ö p a j i H (HHH-

H H a j i b H o e p a c n o j i o a c e H H e y u a p H o r o a j i e M e n r a B e « e T 3,necb κ ο τ CTynjieHHK) ο τ K O A i ^ i m H p o B a H H b i x H o p M pacnojioaceHHH n o c j i e \

flyiomeH

CHHTaKCHHecKOH r p y n n w ) ; I T o j i H y i o c y M K y H a ô p a j i a 6 a -

p a x j i a ; — B o j i b i i i e KaK 6 y A T O Η ε τ n o e 3 A O B ? — E m e .nonojiHHTejibHblH e c T b CTO COpOK BTOpOH n O e 3 f l ( c p .

B03M0KHbIH

BapHaHT

Β KOHeHHOH n a c r a : = nOe3fl CTO COpOK BTOpOÍÍ); y H a c 3HaHHTeJIb» HO M e H b m e CTAJIO MCJIKHX x y j i n r a H O B ; — A STO HE TE? — H y HTO *

Tbl, Mbi yace .NABHBIM aaBHO OTflanH HXHHÊ Bce B e m n

(cp.

Ba-

169

AKTYAJIbHOE HJIEHEHHE Β ΡΑ3ΓΟΒΟΡΗΟΪΪ PEHH pHaHT:

=

Bee HXHHC B e i i m ) ;

—

A

cBoeo6pa3Hbiñ

η3μκ

o new.;

\

^CajiKo

fleBHOHKH

najibHOMy

ye3acaioT

pacnojioaceHHio

oneHi,.

BnpoieM,

HHφopMaτHBHO

TCHfleHUHS κ

Hanôojiee

n a c r a He HBJiaeTCH a ô c o j i i o T H b i M 3aKOHOM, c p . : n a r a a f l i ^ a T o r o , τ ο He H a a o ?

—

Μ η

—

A

Ha-

3HaiHMOH ecjin nocjie

H a KHHacenicy 3aHHCJum» τ ε -

\ n e p b t o j i l k o n H T H a a u a T o r o 6 y ^ e M HHCJia. ECJIH y q a p H b i e

3Bem>a r e s p . Η Η φ ο ρ ρ ^ τ Η Β Η Μ ε

HaCTH OÔJia^aiOT

npHÔJHÔHTeJIbHO

ixeHTpbi h j i h HX

O/IHHaKOBLIM

ΗΗφθρΜ&ΤΜΒ-

H H M B e c o M , τ ο h x B3aHMHoe p a c n o j i o a c e H H e He p e r j i a M e H r a p y e T c a , c p . : — Τ η 3ΤΟ HaBepHO e m e r y f l 3 H K > c ^ a B a j i a B b i y i H J i a ? — \

Ά

\

Β yHHBepcHTeTe H H i e r o He Π Ο Μ Η Ι Ο h t o y i H j i a ( B a p H a w r : =

Ά

hh-

n e r o He Π Ο Μ Η Ι Ο η τ ο β yHHBepcHTeTe y m n a . C p . , KaK η β B h i u i e n p H Be^eHHbix n p H M e p a x , HeHHOH

HeKoaf^imHpoBaHHoe

CHHTaKCHHeCKOH

rpynnbl);

—

pacnojioaceHHe

liCTOpHHeCKHX

ko-

pOMaHOB

o r p o M H o e KOJiHHecTBO η o i e H b M H o r o x y n o a c e c T B e H H o i i

jimrepa-

Typbi; B e e s t h a r m c t M a T y a a noflniHBa^a (BapnaHT:

Bce

\

\

=

3th

\

\

a T y a a n o f l u i H B a j i a n n c b M a ) ; — 3 τ ο npHHTejitHHiia M o a 6 o j i b i u a a η e r o G o j i b i u a a n p n a T e j i b H r n i a ; I I p H M e p H O Taicaa ace 6hiJia n o r o f l a \

\

h KapTHHa 6 b i j i a T a K a a ace. OcOÔeHHO

ABCTBeHHO 3TO

6e3pa3JIHHHe

npOHBJWeTCH ΠρΗ ΠΟ-

B T O p a x , K o r a a HaJinijo n e p e c n p o c . r o B o p a i i m i i o ö h h h o n p o a B j i a e T Heacejiamie

coxpaHaTb

BbICKa3bIBaHHa Cp.: Hto? jih?; 6biTb,

—

—

nepBOHanaJibHoe

ΒΗΛΗΜΟ,

C

RejIbK)

pacnojioaceHHe 6bITb JiyTOie

hjichob

nOHHTblM.

M a j i b H H u i K H Toace K y A a - H H Ô y f l b XO^HJIH M a j i e H b K H e ? -

—

MajibHHiHKH MoaceT

Maca

6w

6biJia? — H t o ?

MajieHbKHe

6biTb,

cteji

ci>e.ii?;

—

Toace Ky^a-HH6yflb

6bi Maca? Tbi

—

cMOTpejia,

Ητο? icaicaa

—

MoaceT

ocTaHOBKa

— K a i c a a OcTaHOBKa 6 b u i a , t b i C M O T p e j i a ? ; —

BCK> K O j i 6 a c y c b e j i ?

—

Mto?

n j i a H 3TH ΛΟρΟΓΗ HHHHTb? —

—

Kojiôacy

ΗΤΟ? —

—

ΧΟΛΗ-

BCIO e b e n ? ;

—

Tw Ectb

E C T b ÜJiaH, 6 y f l y T HHHHTb

3 T H f l o p o r H ? ; — A c e n n a c ö o j i b H a a npHXOflHJia. — H t o ? — C e f i n a c

170 npHxoflHJiafleByiioca,h

Ο. Α. ΛΑΠΤΕΒΑ

OHa ...; — 3,apaBCTByHTe, 3το HacieT 0 6 -

MeHa. — H t o ? — H a c n e T oßMeHa B a M 3BOHHT; — Ilomjia τβτβω»-

Ka KypHTb. — H t o ? — KypHTb TeTeHtKaa nonuia; — r^e OHa? — Κ 6a6ymKe β tocth noexajia. — T^e? — Κ 6a6yiiiKe noexajia β rocTH. Heacejiairae coxpaHjrn» yace Hcn0Jib30BaHH0e cjioBopacnojioxcerae roBopammi oÔHapyîKHBaeT h πρκ ποΒτορβ 6e3 noßyacfleHHH H3bhc, h πρπ οτΒετε Ha Bonpoc, cp.: — CKOpo Bce CTeKjia ôy^yr β Mopo3e. Β Mopo3e 6yayT Bce cremia; — M a m m a cjie^yeT ro rjiaBHoro Bxo^a BbicraBKH. ΠοΒτορίπο: äo rjiaBHoro Bxofla M a m m a cjieayei·; — TaK h ocTanact PiraccicaH, fla? — PrajccKaa TaK h ocrajiacb. HTaK, aHajiH3 B3aHMOAeöcTBHH miara coflepacaHHH h miaHa BbipaaceHHH H a y p o B H e j i h h c h h o h o p r a H i m i i H H

BbicKa3biBaHHX

Β y C T H O - p a 3 r O B O p H O H p a 3 H O B H A H O C T H C O B p e M C H H O pyCCKOrO JIHT e p a T y p H o r o »3biKa n03B0JiaeT c^ejiaTt cueflyiomae npeflBapnTejiBHBie b b i b o ^ h :

1. OCHOBHblM φορΜίυίΒΗΗΜ CpeflCTBOM BbipaXCeHHH KOMMyHHKaTHBHOH CTpyKTypbl BMCKa3bIBaHH5I ΠρΗ MJieHeHHH ero Ha ΗΗφορMaTHBHbie iteHTpbi cjiyxcHT ρΗΤΜΗΚο-HHTOHauHOHHaji CTpyKTypa BbiCKa3bœaHHH, opramoyiomerocH oôhhho no npaBHjiy nepe^oBaHHH yaapHhix h 6e3yaapHbix 3BeHbeB. Β Kaiecrae TaKHX 3BeHbeB OÖbIHHO BbICTynaiOT CaMOCTOHTeJIbHbie OTfleJIbHO HHTOHHpOBaHHbie cjioBa, ho Moryr BbicTynaTb h ôojiee MejiKHe φοΗετΗΗβοκπβ eAHHHiïbi. Β cjiyiae OTcyrcTBM 6e3y,napHoro 3BeHa Ha CTbiKe yaapHbix 3BeHbeB B03HincaeT nay3a, nrpaiomaa pojih pa3T>e^nHHTejiH, 3KBHBajieHTHoro 6e3yaapHOMy 3BeHy. 2. Π ρ Η Hajioacemra pHTMHKO-HHTOHaUHOHHOrO HJieHCHHH Ha KOMMyHHKaTHBHOe B03MOXCHO KaK nojiHoe coBnaflemie yaapHbix 3BeHbeB h m^opMaTHBHbix ueHTpoB, TaK h hx HenojiHoe coBnafleHHe, a Taoce HecoBnafleHHe. Β nocjieflHHX cjiynaax KOMMyHmcaTHBHO 3HaHHMbiñ 3JieΜβΗΤ MOxceT 0Ka3aTbca npe^cTaBJieHHbiM flpoÔHo 6ojiee mêm yAapHWM οληημ 3bchom, a TaKxce MoaceT 3aHjm> πο3ηηηιο yflapHoro 3BeHa. Β KanecTBe ochobhoio pejieBaHTHoro npH3HaKa bhCTynaeT momcht pa3beaHHeHHH yaapHbix 3BeHbeB nocpeflCTBOM nepeflOBaHHH hx c 6e3yaapHHMH hjih nay3bi. 3. Πορκβοκ cjie^o-

AKTyAJItHOE HJIEHEHHE Β ΡΑ3ΓΟΒΟΡΗΟΗ PEHH

171

BaHHfl ΗΗφορΜατΗΒΗΜΧ πβΗτροΒ .zipyr OTHOCHTejibHO a p y r a Β yCTHO-pa3rOBOpHOM BbICKa3bœaHHH OÔHHHO He φϋΚΟΗρΟΒΕΗ β cjiynae ηχ KOMMymncaTHBHOH paBHoqeHHOcra η oÒHapyacHBaeT TeHaeHijHK) κ HHimiiajibHOMy pacnojioaceHHio KOMMymiKaTHBHO 6ojiee BaacHoro sjieMeHTa β cjiynae hx HepaBHOixeHHOCTH β 3tom OTHOIIieHHH. 4. AcCOIJHaTHBHOe HaHH3bmaHHe 3JieMeHTOB ΠρΗ JIHΗβΗΗΟΗ OpraHH3aUHH BBICKa3BIBaHHH B03HHKa6T OÔblHHO JIHIIIb β npe^ejiax p h t m h t c c k h 6e3y^apHtix 3BeHteB cxeMM. IlpHMeiaHHe: ycjioBHBie o6o3HaHeHHH β npHMepax: noÓHepKnymbiü xypcue: „Hoeoe" („aapo") Kypcue: „flaraoe" („HcxoflHaa lacTb") η HeKBajiHΙΜΗ HO, NOCKOUBKY Β p a 3 r 0 B 0 p H 0 H apKO

penn

MHoro

B b i p a a c e H H O H TEMBI, x a p a K T e p H M M

CPE^CTBAMH,

npeflJioaceHHH

HBJiaeTca He

6e3

HJieHeHne

Β noJiHOM CMHCjie cjiOBa Ha T e M y - p e M y , He HX n p o T H B o n o c T a B J i e H H e , a B w a e j i e H H e COÖCTBCHHO PEMBI Η PA3JIHHNE n o c r e n e H H KOMMyHHKaTHBHOH 3HaHHMOCTH O C T a j i t H L i x 3 j i e M e H T O B n p e ^ J i o x c e H H H . 3 τ θ OÖCTOHTeJIbCTBO 3aCTaBHJIO MeHH nOJIbSOBETbCfl Β CBOHX p a 6 o T a x TepMHHaMH „KOMMyHHKaTHBHaa C T p y K T y p a n p e f l j i o a c e H H a " , „KOMMyHHKaTHBHO 3 H a H H M b I Ì Ì " Η „KOMMyHHKaTHBHO He3HaHHMbIÍÍ HJieH". SKcnepHMeHTajibHbie HccjieaoBaHHa p a 3 r o B o p H o ñ

pe^H

noKa-

3 a j i H , H T o p e M a β H e f t B b i a e n a e r c a H e TOJibKO MejiOAHKOH ( p e 3 K o e CHHaceHHe n a c T O T b i K O J i e ß a H H H OCHOBHOIO T O H a H a y ^ a p H O M c j i o r e y^apHOrO

CJIOBa) HO Η ÖOJIbUieii HHTeHCHBHOCTblO,

CTbio, HanpaxceHHOCTbK) npoH3HoineHHH,

flJIHTeJIbHO-

π ρ κ π ε Μ STH n p H 3 H a K H

p e M b i B3aHM03aMeHHMbI H HHAHBHayajIbHO BapHaÖHJIbHbl (Beflym y K ) p o j i b M o a c e T H r p a T b JIK>6OH h 3 STHX n p H 3 H a K O B HJIH OHH MOr y T H C n 0 J I b 3 0 B a T b C H KOMnJieKCHO), 2 H e r o BHflHMO, H e y H H T b l B a e T O.

A . JlanTeBa,

roBopn

o n a c T o ñ HeB03M03KH0CTH

onpejjejiHTb

a K T y a j i b H o e HjieHeHHe n p e A J i o x c e m u i β p a 3 r o B o p H o i t p e n n . BbwejiaeTca

Bcer.ua coBepuieHHO

OTHCTJIHBO η

PeMa

0£H03HaiH0,

HO

κ ρ ο Μ β p e M b i Β03Μ03ΚΗΗ Η A p y r a e uieHTpbi (Β 3TOM O . A . J l a n T e B a c o B e p m e H H o n p a e a ) — B b i f l e j i a i o m H e c H H a φ ο Η β HHTOHauHOHHbix n p O B a j I O B ( T e p M H H H . H . K o B T y H O B O ñ ) KOMMyHHKaTHBHO 3 H a H H Mbie HjieHbi. 3 τ ο h c o s ^ a e T τ ο HHT0Han;H0HH0e c B o e o 6 p a 3 H e p a 3 ΓΟΒΟΡΗΟΗ p e n n , o KOTOpOM Γ Ο Β ο ρ π τ O . A . J l a n T e B a . HenocpeacTBeHHoe

HHTOHauHOHHoe

BbipaaceHHe

aKTyajibHoro

MjieHeHHH (KOMMyHHKaTHBHOH C T p y K T y p w n p e A n o a c e H H a ) Β p a 3 r o BopHoií p e r a πρπΒΟΛΗΤ κ ocoßbiM H o p M a M p a 3 r o B o p H o r o n o p a ^ K a CJIOB Β pyCCKOM A3bIKe (npen03HI^HH KOMMyHHKaTHBHO 3HaHHMO2

CnpoTHHHHa, O. B., Πορηόοκ cjioe β pyccKOM n3biKe. CapatOB, 1965.

O. Β. CHPOTHHHHA

174 ro),

flejiaeT

H36MTOHHHM H c n 0 j n > 3 0 B a H H e c n e i y i a j i t H B i x

(πρπ 3TOM HX Π03ΗΙΙΗ« flaace n o H p a B H J i o c b H

Β pa3rOBOpHOH

ρβΗΗ

cpe^CTB

Cp.¡ Μπβ κοτορΒίε ιππροκο

CBOÔOflHa.

flaace M u e n o H p a B H j i o c b ) ,

p a c n p o c T p a H e H B i Β p a 3 r o B o p H o ñ p e n e , HO O6MHHO KaK HHAHBHflyajn.Hwe

3

c p e f l C T B a 3MOUHOHaJn>HOCTH p e r a . 3

CHpoTHHHHa, O. B. - Ky3bMHieBa, H . C . - ΟφπβΗκο, JI. H . - ToKapeBa

JI. T. - TpaBKHHa, Η . H . , HeKomopue cmmKacmecKue HOU peuu. PyccKüH pa32oeopnan penJ^CapaTOB, 1970.

ocoSemocmupaitoeop-

REMARKS ON DEVICES OF FUNCTIONAL SENTENCE PERSPECTIVE Pavel Novák (Prague)

Given our present knowledge of the so-called functional sentence perspective (FSP), it is to be expected that anyone wishing to survey the ways of expressing FSP in different languages will meet with serious difficulties of a conceptual and/or empirical nature. As for the empirical ones, suffice it to say that very few languages have been described, from the point of view in question, at all satisfactorily (systematically, completely and in detail). As for the conceptual obstacles of such a survey, the issue is rather involved. It is commonly felt that the main terms of FSP have not been fully clarified, but it remains to be seen what kind of elucidation we are aiming at, in what framework we are to look for it. In this connection two things call for emphasis. On the one hand, at the descriptive, "observational" level it is of no use to employ terms referring to distinctions of meaning that a normal native user of a language cannot be aware of. On the other hand, it is unwise to neglect this very descriptive, "observational" aspect of the terms of FSP by claiming that we are operating on a theoretical level, for, in fact, we are short of a theory which would provide an explanation of the phenomena of FSP. I will give an example to convey what I mean. Since, in the world around us, there are phenomena subsumable under such notions as an asymmetrical relation, we need, in languages, such devices as cases or fixed word order and the like (Sapir 1921, 99). We cannot imagine a language which lacks these devices and yet can be used in the world we live in. It is, however, quite easy to imagine, in a

176

P. N O V Á K

Gedankenexperiment, a language lacking many features that are said to pertain to the FSP. Unfortunately, I do not know of any explanation of the very existence of the phenomena referred to in various conceptions of FSP, but I feel that a theory of language behaviour might be a fitting framework for it, and that a rich mine of ideas for it might lie in the study of cases of misunderstanding in verbal communication where phenomena of FSP might play a role, in the study of the communicative significance of this type of misunderstanding, ways of, and ease in, discovering it, and ways of, and ease in, removing it. I would sum up my first, general point by way of a methodological slogan: let us attempt to explicate our terms but, at the same time, let us try to provide an explanation for the phenomena referred to in these terms (for the distinction between explication and explanation, see, for example, Pap 1963). After all that has been said, it is clear that in order to be able to say anything at all about FSP devices in different languages I must assume some intuitive knowledge of FSP. It is known that, in principle, devices of any kind (not only intonation contour and word order) can serve to express FSP components (theme — rheme) (Novák 1959). Thus, for our present purpose, it would be futile to write up lists of particular FSP devices in various languages, and the establishment of a series of typological regularities involving FSP devices would place us, on the other hand, at the opposite end of the gamut of possibilities of how to treat the topic. My aim here is more modest. I would like to call attention to the pattern of FSP devices in two languages as it seems to be of considerable general interest. 1 The place of the devices of FSP in the language structure varies according to the main typological features of these devices and to the similarity to (dissimilarity from) devices used in the other aspects ("parameters", "coordinates") of overall sentence mean1

Only during this Symposium I learnt that the relevance of the two languages (Yukaghir and Tagalog) for the study of FSP had been recognized, apparently for the first time, by Dahl (1969, 53).

R E M A R K S ON DEVICES O F FSP

177

ing. What I mean by aspect of overall sentence meaning I can explain only in terms of examples. Consider, for example, the sentence "Can you get home?" (with one or another kind of intonation contour). This asks a question rather than giving us a statement or a command, conveys the idea that the question relates to the possibility of doing something rather than to the actual realization of it, circumscribes a certain "cognitive content" of the question (it can be rendered by the neutral phrase "your getting home") etc., etc. (see Novák 1966, 220). Now, in the Slavonic languages the main devices of FSP are "non-grammatical", in contradistinction to the devices of the other aspects of overall sentence meaning. There are languages, however, of a different structure. For instance in Yukaghir, a Paleo-Asiatic language, we have for each transitive verb three series of forms (distinguished from one another by the presence or absence of personal inflexion, by morphological exponents, and by the presence or absence of certain prefixes) which are used according to whether the rheme-component coincides with the subject of the verb, or its object, or the verb itself, respectively. Moreover a suffix is attached to the subject or object under conditions that pertain to the distribution of the rheme as well. In short, in Yukaghir, FSP is expressed obligatorily and using morphological means (Krejnoviö 1958). Another example: From the Indo-European languages we are accustomed that one participant in an action or one term of a relation is singled out from the other participants or terms mostly by the so-called agreement between the verb and the noun denoting this participant. This is commonly called a subject and so far it has resisted a unifying formulation of meaning. By contrast, in Tagalog, an Indonesian language, what is singled out by means of certain particles (articles) and word order is just the theme, whereas the rôle of the participant in question (actor, patiens, beneficiary, place) is indicated by an appropriate form of the verb (Bowen 1965). Again, the FSP devices of Tagalog are of the same typological nature as devices that function otherwise. If my sources are reliable, it is reasonable to repeat an objection

178

P. N O V Á K

made to Danes's three-level approach to syntax (if interpreted as aiming at offering a scheme for a language universal) (Danes 1964, Novák 1966). As is known, in this conception FSP is "detached" from the other aspects of what I call overall sentence meaning and conceived of a belonging to a specific level of syntactic structure ("level of the organization of the utterance"). I cannot see any motivation for this "detachment" of FSP other than the character of the FSP devices in Slavonic languages mentioned earlier.2 The trouble is that this step is not warranted by universal features of language structure (cf. Yukaghir and Tagalog). It seems to be commonly accepted that our conceptions and theories of general language structure can be and are influenced to a considerable extent by the structure of the languages we speak and study (in the sense that special is, tacitly, taken for general). What I have wanted to say is merely that it would perhaps be useful to explore fully the possibilities and realities of this type in the realm of FSP as well. This is my second and last point.

REFERENCES BOWEN, D. J., (ed.). Beginning Tagalog. Berkeley and Los Angeles 1965. DAHL, O., Topic and comment: a study in Russian and general transformational grammar. Slavica Gothoburgensia 4. Götteborg 1969. DANES, F., A three-level approach to syntax, TLP 1, 1964, 225—240. KREJNOVIÈ, JE. Α., Jukagirskij jazyk. Moskva—Leningrad 1958. NOVÂK, P., O prostredcích aktuálního dienení [On the means employed by the functional sentence perspective], AUC Philologica 1, 1959, 9—15. NOVÁK, P., On the three-level approach to syntax, TLP 2, 1966, 219—223. PAP, Α., An introduction to the philosophy of science. London 1963. SAPIR, E . , Language.

2

N e w Y o r k 1921.

By this I would not wish to deny that it is useful to distinguish between the intra-sentential and inter-sentential texturing of various aspects of overall sentence meaning.

CPEACTBA

AKTYAJILHORO

Β EOJIRAPCKOM

^JIEHEHHÜ

Λ3ΜΚΕ

EjieHa TeoprneBa (ΟοφΗπ)

1. 0 6 0C06EHH0CTHX aKTyajii>Horo HJIEHEHNA NPEFLJIOACEHHH 6OJIrapcKoro a3biKa, κροΜβ HCCKOJIBKO cTaTeS πρόφ. C B . MBAHHCBA o poJiH KaTeropHH onpeaejieHHocTH, BbipaaceHHoñ ΜορφθϋθΓΗHECKHMH cpeacTBaMH, Β aKTyajiH3auHH npe^JioaceHHH, a o CHX nop He 6biJio nyôJiHKaijHH. HccjieflOBaHHH Β STOM OTHomeHHH TOJIBKO HTO HaHHHaiOTCH.

1

2. CaMo CO6OH pa3yMeeTCH, ΗΤΟ Β KpaTKOM cooômeHHH Hejn>3H OTpa3HTi> BO Bceñ nojiHOTe Bee MHoroo6pa3He

B03M0acH0creii

öojirapcKoro »3bnca, aHajinranecKoro no CTpyKType, HTO ero o6ocoGjiaeT ο τ Bcex flpyrnx cjiaBHHCKHx JBHKOB, πρκ aKTyajiH3amm NPEAJIOACEHHH. ΠΟ3ΤΟΜΥ BTIABJIEHHE ΕΠΕΊΊΚΦΗΗΕΣΚΗΧ AJM H3BIKA cpeacTB aKTyajibHoro uieHeHHH 6y.neT npeacTaBJiaTb nepBocreneHHbiii HHTepec AJIH τ ε ο ρ κ κ aKTyajibHoro HjieHemm H eë πρηΜεΗβΗΗΗ πρκ KOHKpeTHOM HCCJieflOBaHHH CHHTaKCHCa OTflejIbHblX H3MKOB. 3. AKTyajibHoe HjieHeHne KaK npoixecc ceMaHTHKO-CMHCJIOBOII cerMeHTaiiHH npefljioaceHHH Ha TeMy h peMy (ocHOBa η

χχρο, μ η -

Hoe H HOBoe, H3BecTHoe H HeH3BecTHoe η τ . #.) noKa emë He nojiyHHJio flocTaTOHHO HeTKoro η ΛΗφφερεΗίικροΒαΗΗΟΓΟ onpeflejieHHH no OTHomeHHio κ τοίί ςφερε jrebnca, Β κ ο τ ο ρ ο κ OHO ocymecTB^aετοίΐ. AKTyajibHoe υιεΗεΗΗε paccMaTpHBajiocb 6ojibineií nacTbio KaK H3biK0B0e flBJiemie. Π ο

cyra

flejia

«3biK0Baa

CTpyKTypa

(ocHOBHaa CHHTaKCHiecKaa CTpyKTypa HJIH rpaMMaranecKoe HJieHeHHe) H aKTyajibHoe HJieHeime — ABa Meatfly CO6OH CBJNAHHBIE HBJieHHH, Haxowmyieca Β HepapxHiecKOM p«,ny. H e CHHTaTbCH HMeHHO C 3T0ÎÎ OTHOCHTejlbHOH HepapXHieCKOH CBH3aHHOCTblO Η

180

Ε. ΓΕΟΡΓΗΕΒΑ

B3äHM03a.BHCHM0CTbK) ÄByX OCHOBHMX HJieHeHHH JBbIKa AOBOAHT MACTO AO OTOHMECTBJIEHBÜI n p o ö j i e M a T H K H n o p a A K a CJIOB H a i a y a j i b H o r o MJIEHEHHH. OCOÔEHHOCTH NOPHAKA CJIOB He HCHEPNBIBAIOT npH3HaKH

aKTyajibHoro

HJieHeHHsi.

AKTyajibHoe

HJieHeHiie

—

TOJIbKO OAHa H3 ΦΓΗΚΙΙΚΟΗΊΥΊΒΗΒΙΧ OCOÖeHHOCTeÖ NOP^AKa CJIOB. 3 Τ Ο 03HAHAET, HTO AKTYAJIH3AU,HH C n o M o m t i o NOPAAKA CJIOB aBAasTca

TOJIbKO nacTHbiM

acneKTOM Β M e x a m o M e

aKTyajibHoro

HAEHEHHH. Β 3THX CAynaax n o p a A O K CAOB He HTO HHoe, KAK c p e A CTBO o c y m e c T B j i e H H H a K T y a j i b H o r o HJieHeHiia. 4 . IlOpHAOK CAOB Β ÔOArapCKOM H3bIKe, BOnpeKH HeKOTOpbIM BbicKa3aHHHM yace MHCHHÄM, He HMeeT, β XanecTBe CBoeü OCHOBHOH φγΗκιΐΗΗ, 3HaHeHHH CHrHajiH3aTopa a K T y a j i b H o r o HAeHeHHa, x o T H H H o r A a e ë H c n o j m a e T . E r o ΦΓΗΚΙΙΙΙΚ c e a 3 a H a npeacAe B e e r ò C BblAeAeHHeM OCHOBHOÌÌ CTpyKTypbl CHHTaKCHHeCKOH

eAHHHUbl

Η CHHTaKCHHeCKOH pOAH ΚΟΜΠΟΗβΗΤΟΒ npeAJIOJKeHHH π ρ κ ΠΟΜΟU(H HX JIHHeHHOrO paCnOAOXCeHHa. O A H H M CAOBOM, n o p a A O K CAOB Β 6 0 J i r a p C K 0 M a3bIKe CBH3aH npeHMymeCTBeHHO

C BbWBJieHHeM

r p a M M a T H i e c K o r o HJIÊHCHHJI npeAAoaceHHa H OTpaacaeT ocHOBHyio CHHTaKCHHecKyio C T p y K T y p y , CHHTaKCHHecKyio

cxeMy,

cHHTaKCH-

l e c K y i o M O A e A b npeAJioaceHHa. 4 . 1 . T a K H M 0 6 p a 3 0 M M b i ΠΟΑΧΟΑΗΜ Κ B o n p o c y

onpeACJicHiia

c y m H o c T H a K T y a A b H o r o HJIEHEHIRA Β ΟΤΗΟΙΙΙΘΗΗΗ Κ NOPAAKY CAOB, HHane rOBOpa, Β ΟΤΗΟΗΙΒΗΗΗ Κ OCHOBHOH CHHTaKCHHeCKOH MOAeAH «3biKOBOH eAHHHHbi. I I p H 3 H a B a a r p a M M a T H H e c K o e HAeHeHHe H aKT y a A b H o e HAeHeHHe CAMOCTOATEJIBIIUMH YPOBHAMH H3MKOBOH CHCT e M b i , Mbi T e M c a M b i M

o6a3yeMca

yKa3aTb

HX CBa3aHHOCTb

Η

B3anM03aBHCHM0CTb. A K T y a A b H o e HAeHeHHe e c T b HAeHeHHe n p e a AoaceHHJi, 0 6 y c A 0 B A e H H 0 e r p a M M a r a n e c K H M HAeHemieM, c

κοτο-

p b l M OHO CBH3aHO TaKHM 0 6 p a 3 0 M , HTO AOA5KHO COXpaHHTb ΒΓΟ o c H O B H y i o c x e M y . B e e A o n o A H H T e A b H b i e npnÖMbi, B X O f l a i i m e Β H a 6 o p cpeACTB, K O T o p b i e H c n o A b 3 y i o T c a π ρ κ aicTyaAbHOM HJieHemm, H i p a i O T p o j l b OTHpaBHOH ΤΟΗΚΗ A ™ yKa3aHHH COXpaHeHHH, H a p y HieHHH HAH Η3ΜΒΗΕΗΗΛ OCHOBHOH CXeMbl IipeAAOvKeHHH ΠΟ OTHOUieHHK) HOpM CHHTaKCHHeCKOH C T p y K T y p b l Η nOpHAKa CAOB. 5. H a p y m e H H e

HAH ΜΟΑΗΦΗΜΗΗΗ OCHOBHOÍÍ CHHTAKCHHECKOÖ

MOAeAH 33bIKa npOH3BOAHTCH Β HpOHeCCe KOMMyHHKaUHH.

ΠΟ-

CPEflCTBA AKTYAJlLHOrO HJIEHEHHÍI Β EOJirAPCKOM Λ3ΗΚΕ

181

3TOMy aKTyajibHoe HJieHeHHe npe^cTaBJiaeT COÖOH caMocTOjrrejibHyio eOTHHiiy Ha ypoBHe penn. IIopjwoK CJIOB HJIH rpaMMaranecKoe HJieHeHHe «BjiaeTca caMOCTOHTejibHOH eflHHHueñ Ha ypoBHe H3HKa. MMÜHHO Β npoiiecce KOMMyHHKaiJHH Π Ρ Η yCJIOBHH o6iI3aTejibHoro HajiHHHH cyôbeKma cooôufemn H oôbeKma cooônfeHun C03aaëTCH B03M0XCH0CTL· peajIH3aiJHH HJ1H Τ Ρ 3 Η Θ Φ Θ Ρ Μ 3 Ι ΐ ; Η Η aicryajibHOH Η Η Φ Ο Ρ Μ 3 ΐ Ι , Η Η nepe^aBaeMoro cooömeHH». Π ρ Η OTCyTCTBHH 3ΤΟΓΟ yCJIOBHH KaXCflOe COOÔmeHHe ο φ ο ρ Μ jiaeTCH no 3aK0HaM (npaBHjiaM) rpaMMaranecKoro HjieHeHHH Ha H3BIKOBOM ypoBHe. T a K H e C H T y a i ï H H BCTpenaioTCs npeHMymecTBeHHo Β K a n e c T B e npaiwepoB HJIH Β H30JiHp0BaHH0M T e K C T e . ü o s T O M y MHe KaaceTca, HTO iienecoo6pa3Hee 6yaeT roBopHTb o rpaMMaTHnecKOM h aKTyajibHOM HjieHeHHH TojibKO npH coôJiiofleHHH ynoM H H y T b l X yCJIOBHH. T a K O H Β3ΓΛΗΑ Ha Bonpoc, MOHCeT 6 b I T b , n p H HyacaaeT roBopHTb He o rpaMMaranecKOM Η O aKTyajibHOM HjieHeHHH, a o rpaMMaTHHecKOM HjieHeHHH h o aKTyajibHOM rpaMMaTHMecKOM HjieHeHHH, nocKOjibKy aKTyajibHoe HJieHeHHe B c e r a a orpaHHneHO paMxaMH, .zionycTHMbiMH rpaMMaTHnecKHM HJieHCHHCM. 5.1 TaKHM 0 6 p a 3 0 M

caMa no

ce6e

OTna^aeT

HeoôxoAHMOCTb

OÖbeKmueHblÜ (HOpMajIbHblH, OÖblHHblH, r p a M M a T H i e c K H H ) ηορηόοκ cnoe h cyôbeKtnueHbiù ( s M o i j H O H a j i b HbiH, 3KcnpeccHBHbiH, aKTyajibHbiñ) nopndoK cAoe. üepBbiH OTpaa c a e T OCOGCHHOCTH r p a M M a r a n e c K o r o HjieHeHH«, Β τ ο ρ ο κ — aicTyajibHoro HJieHeHH«, κ ο τ ο ρ ο ε HX npe/iycMaTpHBaeT H flonycKaeT. OnepHpOBETL· T e p M H H a M H

6. AKTyajibHoe KOTopaa C TCM

HJieHCHHe

HacjiaHBaeTcn

HeOÖXOflHMO

Ha

—

STO o n e p a n H H

Η3ΗΚΟΒΟΗ

Ha

ypoBem».

ypoBHe

penn,

OflHOBpeMeHHO

O T M e T H T b , HTO Β JIHHTBHCTHHeCKOH J I H T e p a T y p e

h y p O B e H b p e n n n o K a e m ë H e n o j i y n a j i Ο/ΙΗΟ3ΙΉΗΗΟΓΟ o n p e f l e j i e -

HHfl.

O C H O B H b l M H H C X O ß H b l M nOJIOHCeHHeM ΒΓΟ

OnpeflejieHHfl

HB-

JIAETCH o 6 H 3 a T e j i b H o e H a j i H H H e n p o u e c c a K O M M y H H K a q H H . Ü 3 M K O B b i e Φ 3 Κ Τ Η T a K O B W , HTO H a y p o B H e p e n n T H n a KOMMyHHKamiH.

Hx

xapaKTepHbie

OCYMECTBJIAIOTCH

φορΜΜ

npoaBjieHHH

M O H O J I O r H A H a j I O r . Y K a 3 a H H a X ΟΠΕΊ^ΗΦΗΚΑ K O M M y H H K a i J H H

φορΜ oimcaHbi.

6yeT onpeflejiëHHoro H a 6 o p a jobncoBbix Aojdkhu

6biTb onpeaeJieHH

h

ppa. —

Tpe-

H cpeflCTB, K O T o p w e

182

Ε. ΓΕΟΡΓΉΕΒΑ

6.1. IIpoHjiJiiocTpHpyeM 6ojirapcKoro β

H3bnca.

BbicKa3aHHyio

06meH3BecTHO,

M b i c j i b Ha

ητο

β

MaTepHajie

G o j i r a p c K O M A3 t u c e

o n p e f l e j i e H H H X CHTyaijHHX c o 3 f l a i o T c a y c j i o B H » , π ρ κ

npaMoe

flonojiHeHHe

KOTOpbix

( ο β ι ε κ τ ) y^BaHBaeTca (pe^ynjiHiinpyeTca)

—

τ . e . o h o B w p a a c a e T C H o/iHOBpeMeHHO η 3 η κ π η τ η κ ο - μ 6 ο τ ο η μ 6 η η ο η φ ο ρ Μ Ο ί ί Η n0JIH03HaHHI.IM CJIOBOM. CHHTaKCHHeCKHe yCJIOBIM T a KOBbl, ΗΤΟ AJIfl

flyÔJIHpOBaHHOrO

aOnOJIHeHHH He 0 6 H 3 a T e j I b H 0 c o -

xpaHeHHe onpeitejieHHbix π ο 3 η η ; η ο η η μ χ

ycjioBHH

—

oho

MoaceT

ynoTpeGjMTbcsi η β npeno3HiiHH, η β π ο ο τ π ο 3 η η ; η η n o o t h o i h c h h i o κ rjiarojiy, β r p y n n y KOTOporo b x o o t t . Η β τ ο μ , η β a p y r o M cjiynae, oflHaKO, ^BoiiHoe aonojiHeHHe n o j i y n a e T o n p e a e j i s H H y i o c h h T a K C H H e c K y i o 3HaHHMOCTb. C t o h k h 3peHHH a K T y a j i b H o r o u i e H e r o i a ABoiÍHoe TjonojiHeHHe n a m e B e e r ò h b j i h c t c h c u r H a j i i o a T o p o M b h χ ο λ η ο η n a c r a (ocHOBbi, H 3 B e c T H o r o ) p e n e B o r o c o o ö m e H H H . flHTejibHbiM

YGe-

ÄOKa3aTejibCTBOM τ ο γ ο , η τ ο CKa3aHHoe h b j i h c t c h n p a -

BHjibHbiM, c j i y a c a T

flpynie

rpaMMaTHnecKHe noKa3aTejin

Gojirap-

C K o r o H3biKa, a h m c h h o — y n o T p e G j i e m i e h j i c h h o h φ ο ρ Μ Η

cyme-

C T B H T e j i b H o r o . n p H M e p b i : Kmzama

h

Λ β3β 6u6nuomepKama

BU-

ÔAUomepKama Λ β3β KHueama. Β 3ΤΗΧ n p H M e p a x x a p a i c r e p H o : β n e p B O M c j i y n a e — ocHOBHoro cyöbeKT

rpaMMaTHHecKoro

6bin

β

nepBoñ,

HJieHeHHH

ββιχοληοη

Hapyrnerae

(Tpe6yiomero,

πο3ηη;ηη),

ho

htoGh

coxpaHeHHe

y c j i o B H Ì i a K T y a j i b H o r o HJieHeHHH ( T p e ô y i o m e r o , h t o G m o c H O B a c o oômeHHH 6bijia β nepBoñ, β μ χ ο α η ο η π ο 3 η η η η ) ; b o Β τ ο ρ ο Μ cjiynae, H a o ö o p o T , HaGjiiOAaeTca coxpaHeHHe ο ο η ο β η ο γ ο

rpaMMaranecKo-

r o HJieHeHHH ( c y ô t e K T H a x o f l H T c a β n e p B O H , B b i x o f l H o ñ π ο 3 η η η η ) η H a p y m e H H e y c j i o B n ñ (npaBHJi) a K T y a j i b H o r o HJieHemra (OcHOBa HaXOflHTCH Β KOHeHHOH Π03ΗΙΧΗΗ). C

tohkh

3peHHH rpaMMaranecKoro

η

a K T y a j i b H o r o HJieHeHHH

ö o j i b i i i H H HHTepec n p e f l C T a B j i a e T n e p B b r i i npHMep, n 0 3 T 0 M y o c T a HOBHMCH Ha HeM. PeaynjIHKaaHH flOHOJIHeHH« Β flaHHOM npHMepe noMoraeT η

n p e o ^ o j i e T b npOTHBopenne

aKTyajibHbiM

HJieHemieM

Meac^y

npeAJioaceHHa.

TaK

rpaMMaranecKHM OHa

HcnojiHaeT

p o j i b naflsacHoro yKa3aTejia cHHTaKCHHecKOH φyHK^HH HJieHa npeflJioxceHHH. E c j i h pe^ymiHKaiiHH HeT, τ ο

λβοηηογο

MoaceT

HHKHyTb c e M a H T H H e c K a a η j i o r H H e c K a a H e c o B M e c T H M O C T b .

bo3-

TaKHM

CPEßCTBA AKTyAJIBHOrO HJIEHEHHÄ Β EOJirAPCKOM JI3HKE

183

06pa30M, flBOÖHoe ÄonojiHeHHe öojirapcKoro H3biKa yaoBJieTBOpaeT TpeöoBaHHio aKTyajibHoro HJICHCHHH πρπ ΜΗΗΗΜΟΠΒΗΟΗ noTepe TOHHOS rpaMMaTHKO-CHHTaKcmecKOH onpeaejiëHHOCTH komΠΟΗβΗΤΟΒ npeflJlOJKeHHH, TaK KaK ΙφΗ 3TOM H3ÔeraeTCH CHHTaKCHKo^yHKUHOHajibHaH ABycMbicjieHHOCTb npefljioaceHiw. HyacHO noOTepKHyTB, Ο,Π,ΙΉΚΟ, HT O pe^ynjiHKaiïHH AonojiHeHHa Β öojirapcKOM a3bnce — STO npoiiiecc Β nprnmane xapaKTepHHH ΛΛΗ y c r a o ö , pa3r0B0pH0H pera (fluajiora). HeoôxoflHMO flonojiHHTb, HTO Cneip^HKa ΛΒΟΗΗΟΓΟ flOnOJIHCHH» Β H3BeCTHOH Μβρβ coxpaHaeTca H Β xyaoacecTBeHHbix TeiccTax, Koraa OHH BocnpomBOflHT 0C06eHH0CTH pa3r0B0pH0H pe«ffl. ^BOHHOeflOnOJIHeHHeTaKHM 06pa30M íiBjiaeTCH cpeACTBOM HHAHKaijHH aKTyajibHoro HjieHeHna Β ycTHOH, PA3ROBOPHOÑ (^AAJIORHHECKOH) pein. 6.2. Β imcbMeHHOH p e r a (KOTopaa 6ojibmeñ nacTbio 6bœaeT MOHOjiorHHecKoä) pe,nynjiHKaiiHH flonojiHeHHa ΠΟΗΤΗ He BCTpenaeTca, H3ßeraeTca, noTOMy HTO oiieHHBaeTca xaK npoTHBopenamaa CTporo HopMaTHBHoii CTpyKType BbicKa3biBaHna. H T O G H COXpaHHTb

OAHOBpeMeHHO HOpMaTHBH JIHTepaTypHOH a3bIKOBOH

npaKTHKH Η aKTyajibHoro HJieHeHHa, nHCbMeHHaa pent ορκβΗΤΗροBajiacb κ flpyrHM cpe^cTBaM a3bnca. TaKHMH HBJIHIOTCH naccHBHbie 3ajioroBbie

φορΜΗ

Η ΚΟΗΟΤΡΓΚΠ;™ jiHTepaTypHoro a3biKa.

TaK Hanp. Β nacbMeHHOM τεκοτβ yKa3aHHbie Bame ABa npHMepa ΤPAHCΦOPMHPYΚ>ΤCA Β O£HO:

Knuzama 6e e3ema (ce e3e) om 6u6AuomepKama, KOTOpoe BnojiHe cnocoÔHO npHMHpHTb rpaMMaTHHecKoe H aKTyajibHoe HJicneHne npefljioacemia. Β noAOÔHbix c j i y i a a x , ecjin aonojiHHTejibHO B03HHKaeT ποτρεδHOCTb Β

aKTyajiH3auHH npeanoxceHna,

Hcnojib3yioTca

apyrae

cpeACTBa a3biKa — nopaaoK CJIOB, HHTOHai^HH, jioraiecKoe yuapeHHe H T. G. 6.3. n p H 3TOM a3biKOBaa cncreMa, ocHOBHaa a3biKOBaa CTpyKTypa, CHHTaKCHHecKaa Mo^ejib, cxeMa He npeflycMaipneaeT H He TpeôyeT 06a3aTejn>H0Ö peaymiHKaimH AonojiHemia h Hcnojn>30BaHHH naCCHBHblX φθρΜ H KOHCTpyKHHÎÎ Β ÔOJITapCKOM a3bIKe. 3ΤΗ cpeACTBa noaenjiHCb Β a3biKe no3AHee Β CHjiy HeoôxoflHMOCTH Bbi-

184

Ε. ΓΕΟΡΓΗΕΒΑ

p a x a T b a K T y a j i b H o e HJieHeHne 6 e 3 H a p y u i e H H H o c h o b h o h c h h t e k c h HecKoñ MOAejiH npewioaceHHH. O n e p a n p n , c o B e p n i a e M t i e

Hcnojib-

3 0 B a H H C M y K a 3 a H H L I X φ θ ρ Μ Η C p e f l C T B » 3 H K a , OCymeCTBJUUOTCH H a ypoBHe p e n a

c uejn>io c o x p a H e m a

rpaMMaranecKOH

οτργκτγρω

H3HKa η e r o r p a M M a r a n e c K o r o h j i c h c h h s . 0 6 a c j i y n a « J1HMT

coßoii

rpaMMaTHKO-CHHTaKCHHeCKyiO

npeflcraB-

COOTHOCHTeJIBHOCTB

aKTyajiBHoro η r p a M M a r a n e c K o r o uieHeHHfl npejyioaceHHa, coGjiioa a a Hx ocHOBHBie T p e ß o B a H H H .

flBOHHoe

A o n o j i H e H H e η e r o 3KBH-

B a n e H T — n a C C H B H M e φ ο ρ Μ Β Ι H KOHCTpyKIXHH, — 3 Ί Ό Ä B a T p a H C φθρΜ3

OCHOBHOrO r p a M M a T H H e C K O r O HJieHeHHH,

Hbie κ

pa3HBiM

cφepaM

peneBoñ

OpneHTHpOBaH-

KOMMyHHKaqHH

— ycTHoñ

η

üHCLMeHHOH p e H H . IIOSTOMy HeOÖXOflHMO Β KaSCflOM KOHKpeTHOM c j i y i a e h m c t l β BHfly κ K a K o ä p e n e e o n ς φ ε ρ ε o t h o c h t c h pHBaeMoe

H3BiK0B0e cpe^CTBo,

Hcnojn>3yeMoe

fljiH

paccMa-r-

qejieö

aKTy-

a j i b H o r o HJieHeHHH. I l p H M e i a H H e : 3kbhbaji6hthocti>

λβοΒηογο

T e ^ b H M X KOHCTpywiHñ β öojirapcKOM κ3bixe — 3HaHeiIHÖ Η npOflBJieHHH naCCHBa.

aonojiHeHHfl h

CTpafla-

TOJibKO lacTHbifi cnyiafi

CONCERNING THE PECULIARITIES OF THE BULGARIAN GRAMMATICAL SYSTEM FROM THE POINT OF VIEW OF THE FUNCTIONAL SENTENCE PERSPECTIVE Janko Bacvarov (Sofia)

In her contribution for this volume, my compatriot E. Georgieva mentions the most characteristic features of the Bulgarian grammatical system from the point of view of the functional sentence perspective (FSP). I should like to add a few more thoughts in order to give a more complete picture of the means of expressing the category of definiteness in Bulgarian by means of comparison with other languages. This question has already been discussed in Bulgarian lingustic literature (particularly in the works of Stoyan Stoyanov, Svetomir Ivanchev and Jana Molhova).1 Of these authors only S. Ivanchev investigates this problem in the light of the theory of FSP. He emphasizes the fact that "the definite article aided by the word order and intonation is able to replace the lacking noun inflection 1

CroHHOB, CT., V A e u y e a u e ua uMenama β öbAzapatun e3UK. [The use of nouns with articles in Bulgarian.] Sofia 1965. HeameB, Ca., Edna Heonucaua ynompeöa πα uAenyeauama φορΜα [An unrecorded use of the article.] C6. Β L E C T H A AJI. TeoflopoB-EajiaH, Sofia 1955, 271—278. HeameB, CB., HaSjiioAeHHH BT>pxy ynoTpe6aTa Ha u i e r a Β ôwirapcKiM e3HK [Observations on the use of the article in Bulgarian], BbAzapcKu e3UK 7,1957, 499—524. HeaHHeB, CB., K T > M BMipoca 3a uieHyBaHeTO Ha reHepmmo ynoTpeôeinM noftnor [Concerning the question of using the article of the generically employed subject], E3HK H jiHTepaTypa 22, 1967, 61—64. HeaHHeB, CB., ΠροβπβΜΗ Ha aKryajiHOTo ιπβΗβΗΗβ Ha H3peieHHeTO [Problems of the functional sentence perspective], ΟΛαβηηοκα Φ Μ Ο Λ Ο Ζ Ι Μ Ι 10,1968, 39 — 53. Monxoea, }K., Χαρακmep u ynompeöa na HACHO β ÔBAEAPCKUN u ÜHZAUUCKUH e3W. KoumpacmeH amAU3. [The nature and use of the article in Bulgarian and English. A contrastive analysis.] Sofia 1970.

186

J. BACVAROV

in languages such as Bulgarian, and to develop from a means primarily of FSP into a grammatical means." 2 I should like to draw your attention to the fact that Bulgarian, because of its specific character of an analytical language, differs sharply in this respect from the other Slavonic languages and is extremely similar to English and French, which are also analytical. Apart from this it must be emphasized that an analogical distribution of the defined, concretely characterized sentence components mainly in the theme can be observed in all three languages. It can be said that it is namely the FSP that dictates definiteness and concreteness of those components. It is because of this, in fact, that they have an article — the definite in English and French, and the same in Bulgarian. The components with a definiteness which is weaker or even lacking are most frequently to be found in the rheme, and in English and French have an indefinite article; in Bulgarian they have a zero article, i.e. are without article; or else indefinite pronouns (HHKou "somebody", Ηηκακτ,β "some") as well as the numeral eduH "one", are used to fulfil the function of an indefinite article. In order to illustrate this, I should like to quote several examples from Antoine de Saint-Exupéry's Le Petit Prince and their equivalents in the Bulgarian, Russian and Polish versions.3

2

IJpoÔACMU πα aKmyaAHomo

3

The quotations are taken from the following publications:

H.WHenue..., p. 42.

ANTOINE DE SAINT-EXUPÉRY, Œuvres,

M o s c o u 1969; AHToaH /o>o CCHT EK3K>-

nepH, MaAKunm npunif, Οοφκβ 1966, A m y a H Ae CeHT-3K3ionepH, Majiem>KHii ΠΡΗΊΊ, MocKBa 1967;

ANTOINE DE SAINT-EXUPÉRY Ma/y

Ksiqzç,

Warsza-

wa 1970. Examples from translations in the three Slavonic languages from three different groups (Southern, Eastern and Western) have deliberately been selected in order to emphasize more clearly the separate place occupied by Bulgarian in the Slavonic linquistic family. The figures in brackets indicate the page in the respective publication.

187

PECULIARTIES OF FSP IN BULGARIAN

Here are several examples with a component bearing an article in the theme: Les e n f a n t s seuls savent ce qu'ils cherchent. (388) CaMO atixama 3HasT KaKBo T t p c H T . (72) O f l H H TOJibKO aeTH 3 H a i o T , nero HmyT. (71) Jedynie d zi e ci wiedz^, czego szukaj^. (66) Les h o m m e s ont oublié cette vérité. (387) X o p a m a ca 3a6paBHJiH TA3H HCTHH3. (71) JIK>AH 3a6wJiH 3Ty HCTHHy. (70) L u d z i e zapomnieli o tej prawdzie. (65) Le l a n g a g e e s t s o u r c e d e m a l e n t e n d u s . ( 3 8 6 ) E 3 H K 6 W e H3Bop Ha H e a o p a 3 y M e H H a . ( 6 8 )

CjiOBa

TOJibKO M e u i a i o T n o H H M a T b a p y r a p y r a .

(68)

M o w a jest z r ó d l e m nieporozumieñ. (62) Les

f l e u r s s o n t si c o n t r a d i c t o i r e s ! ( 3 6 7 )

ljBeTHma ca H3ntJiHeHH c TOjiKOBa npoTHBopeiHfl! (36) I ^ B e T b i TaK H e n o c j i e , z i o B a T e j i b H b i ! ( 3 1 )

K w i a t y maj^ w sobie tyle sprzecznoáci. (29) In the following examples the component bearing an article is in the rheme: Dessine-moi an m o u t o n . (356) HapHcyBañ μ η edna OBija. (14) HapHcyñ öapamica. (12) Narysuj mi b a r a n k a . (12) J'ai besoin d'«« m o u t o n . (356) Tpaöea ΜΗ O B i j a . (14) MHe HyaceH 6apauieK. (12) Potrzebny mi jest b a r a n e k . (12) C'est trieste d'oublier «η ami. (360) TbXHO e fla 3a6paBHm edun npnaTeji. (22) 3 τ ο oieHb nenajibHO, xor^a 3a6biaaK>T apy3eñ. (18) To bardzo przykre zapomniec przyjaciela. (18)

188

j. baCvarov

And finally, several intermediate cases where there are components bearing articles both in the theme and in the rheme: Et les r o u t e s vont toutes chez les h o m m e s . (383) A niTHiiiama Bimani b o a h t πρπ xopawzß. (64) A Bee floporH Be,ítyT κ j i i o a h m . (62) A d r o g i prowadzq. zawsze do l u d z i . (56) Les g r a n d e s p e r s o n n e s aiment les c h i f f r e s . (359) Bï>3pacTHHAne oÔHiaT ιικφρΗ/ηβ. (21) B 3 p o c j i b i e oneHL· jiioGht υ;Ηφρι>ι. D o r o s l i sq. zakochani w c y f r a c h . (16) Les e n f a n t s doivent être très indulgents envers les g r a n d e s p e r s o n n e s . (359) ¿ J e n a m a Tpaöea p.a. 6i»aaT μ η ο γ ο CHH3xoflHTejiHH ki>m Bl>3paCTHH/7ie. (22) fleTH AOJHKHM 6bITL· ΟΗβΗΒ CHH3XOflHTejTbHbI Κ B3pOCJIbIM. (18) D z i e c i muszq byc bardzo poblazliwe w stosunku do d o r o s l y c h . (17) I consider that o n the basis of the given examples it is obvious that the Bulgarian language combining the means of FSP with morphological expression of the category of definitess differs from the other Slavonic languages and displays peculiarities typical of languages of analytical type.

A K T Y A J l b H O E ÎÎJIEHEHME, r J i y B H H H W E CTPYKTYPLI H Π Ε Ρ Η Φ Ρ Α 3 Η IIpaceMbicJi AflaMeu, (Ilpara)

1.

Oahoh

H3

KapAHHantHbix

npoÔJieM

yneHHH

06

aKTyajibHOM

HJieHeHHH npoflOJiacaeT ocTaBaTBCH Bonpoc o B3anMooTHomeHHH Meacay aKTya^tHbiM HjieHeimeM h ciiHTaKCHHecKOH CTpyKTypoñ npeanoxceHHH. IIocKOJibKy Β nocjieflHHe Γ Ο Λ Η ΠΡΗΗΗΤΟ pa3JiHiaTb noBepxHocTHyio h rjiyÔHHHyio CHHTaxcHHecKHe cTpyiarypbi, το aaHHaa npoöjieMa pacna^aeTca Ha flBe n a c T H b i x : a) B3anM00TH0iiieHHe A 1 ! h noBepxHOCTHOÌi CTpyKTypbi, 6) B3aHMooTHoiiieHne A 1 ! h rjiyöüHHOH CTpyKTypbi. flo HeAaBHero BpeMeHH BHHMamie HCCJIEAOBATEJIEÑ cocpeAOTOHHBajiocb Ha nepBOM H3 STHX B3aHMooTHomeHHfi; β CBoeM AOKJiafle a xony ocTaHOBHTbca Ha ΒτοροΜ. 2. Β noHHMaHHH rjiyÖHHHOH CTpyKTypbi ( r C ) noKa HeT eflHHCTBa. OflHaKO, ecjiH He oumôaiocb, 6ojn>nraHCTBO aBTopoB Tenepb r C noHHMaeT KaK HeKOTopyio Κ0ΗφΗΐ7ρ3ΐΐΗΚ) ceMaHTHHecKHX KaTeropHH, KOTopaa BnocjieacTBHH nyTeM onpeaejieHHbix onepauHH MoaceT 6biTb τρ3ΗθφορΜΗροΒ3Η3 Β noBepxHocTHyio CTpyKTypy, τ. e. KOHKpeTHoe npeajioaceHHe. 0 6 h h h o TaKace CHHTaeTCH, MTO oimy r C MOKHO peaJiH30BaTb HecKOJibKHMH cnocoôaMH, t . e. MTO oflHa H Ta ace rjiyÖHHHaa cTpyirrypa MoaceT jieacaTB Β ocHOBe HeCKOJIbKHX nOBepXHOCTHblX CTpyKTyp, 3KBHBajieHTHbIX C TOHKH 3peHHH co^epacameHcn Β ηηχ ΗΗφορΜ3ΐ^ΗΗ. ECJIH BCTâTb Ha Taicyio no3HHHK>, το, MHe KaaceTCH, aKTyajibHoe HjieHeHHe npHAeTCH TaK HjiH HHane BKJiioHHTb Β TC, HOCKOJibKy OHO OTHacTH onpeaejiaeT B03M03CH0CTH 06pa30BaHHH OTfleJIbHblX nOBepXHOCTHblX CTpyKTyp H3 aaHHoií rjiyÔHHHOH CTpyKTypbi. IlocTapaiocb STO noKa3aTb Ha HecKOJibKHX npocTbix npHMepax. 3. ECJIH B3HTL· TpH sjieMenra — areHC, fleñcTBHe Η ο6τ>εκτ flea-

Π. Α Λ Α Μ Ε Ι ί

190 CTBHH, η

ΤΟ

ΗΧ Μ 0 5 Κ Η 0 , Β 3 a B H C H M O C T H

AaHHoro

30BaTfc

ΗΗφορΜαηιιοΗΗΟΓο

Β npe^JioaceHHa c

niecTbio

ffupexmop / U3ynaem ηροεκιη denaem dupeumopl)

(1)

ΟΤ

flaHHOH

KOHCHTyaUHH

HaMepeHHH roBopamero, rjiaBHMMH

opraHH-

BapnaHTaMH

A H :

Hmo

(¿JnarHOCTiHecKHH Bonpoc:

\

ffupexmop

(2)

my wem

/ ηροεκιη

{Hmo

myuaem

dupeKmopT)

\

Προεκηι

(3)

/ U3ynaem

dupexmop

( Hmo npoucxodum c ηροεκmoM

?)

\

(4)

Προεκηι

mynaem

/ dupeicmop

(Kmo

U3ynaem

(Hmo

όεΑαειη

ηροεκιη!)

\

(5)

ffupeKmop

ηροεκιη

/ toynaem1

dupeKmop

c

ηρο-

εκιηοΜΐ) (6)

ffιιρεκηιορ

ηροεκιη

/ U3ynaem

(M3yuaem

AU όιιρεκιηορ

ηρο-

εκηιΊ) 4.

ΠocMOTpHM

T e n e p i . , Kaie MOHCHO n e p H < j ) p a 3 H p o B a T i >

(rpaM-

M a T H H e c K H ) p a r a b l e npeflJioaceHHa. I T p e A J i o a c e H H e ( 1 ) MOJKHO

ne-

ρπφρ33ΗροΒ3Τΐ> npefljioxeHHeM \

(3a-

nam)

u3ynehueM

ηροεκιηα.

B a T i > K a K ( 8 ) M3ynemeM fljia

( 7 ) ¡^ηρεκιηορ

j 3aHUMaemcn

A H a j i o r n H H O (4) m o ä h o ηροεκιηα

3anuMaemcn

nepi^pa3npo-

(ßauHm)

ocTajibHbix n e r a p e x BapnaHTOB Taicoñ c n o c o 6

j

όιιρεκιηορ.

περκφρ33ΗροΒ-

K H , τ . e. c a H a J i H T H 3 H p o B a H H b i M B b i p a a c e H H e M r j i y ö m r a o r o K a T a , M a j i o Π Ο Λ Χ Ο ^ Η Τ : OHCBHAHO, n o B e p x H O C T H a a

npeflH-

penpeseHTauHH

r j i y Ô H H H o r o o ô b e K T a Β BHfle n o B e p x H O C T H o r o a T p a ö y T a B 0 3 M 0 » C H a TOJibKo

Β TOM

cjiynae, ecjiH

fleñcTBHe

H o6i>eKT BMecre

bxoaht

Β O f l H H H T O T ace Κ Ο Μ Π Ο Η Β Η Τ A H , H T O H M e e T M e C T O Β ( 1 ) Η ( 4 ) , H O He Β

OCTajIbHbIX.

B a p n a H T (3) moncho, npeacße Beerò, Jienco

1

τpaHcφopMHpoBaτb

Β noBepxHOCTHOü CTpyKType TaKoK BapiiaHT A H r o p a 3 ^ o

jKaeTCH φ ο ρ Μ ο ϋ ffupeKmop / u3yuàem

rame

Bbipa-

/ itpoeKm (c (J>pa30BbiM y ^ a p e r a e M β ce-

peAHHe). B o o 6 m e , KaHCflilÖ HS ΠρΗΒβΛβΗΗΜΧ BapHaHTOB Β nOBepXHOCTHOH CTpyKType ΜΟΗΜΤ 6MTI> BtipaaceH HecKOJiwcHMH CHHOHHMHHHMMH φ ο ρ Μ & Μ Η C flpyrUM MeCTOM ix TpaHCfJjopMauHH, onpeflejiHiomHX κοΗκρετHbie cHHTaKCHHecKHe οφορΜϋεΗΐω. Γραφκ^εοκΗ STO MoaceT 6BITB npeflCTaBJieHO cjie^yiomHM 06pa30M:

Ag ι (.IHJK'KTOp)

Act I {myran.)

Obj (ιψοι'κτ) —»-

(1)

Obj (npoeicT)

e I Ag I (jlipt'KTOJ)) .

Act (inyiaTi.)

l

(7)

Ai I (.IHpt'KTitp)

Act I (ii:iyimTi>)

Ob¡ (ιιροι·κτ) -

(2) (11)

(12)

« Obj I (ΠΡΟΘΗΤ)

Act I (HdyiaTb)

Ag I (AHpeKTOp) -

(3) (9)

(10)

f' 4 '

Ag I (,ιιιροκτορ)

Obj l (ηροοκτ)

Act (uayia-ri,)-·-

(8 ) (13)

(14>

'(5) (15) (16)

I Obj Affirm.mod. U ) (;ihpcKTop) (H3y>uiti>) (ιιροοκτ) ( +

(17) (18) (19)

6. rioBepxHOCTHoe οφορΜπεΗκε npeAJioaceHHH H noBepxHOCTHbie CHHTaKCHiecKHe KaTeropHH, TaKHe icaic ncwieacamee, CKa3yeK nofloßHOMy 3aMKweHHK) πρκχοΛΗτ H flpyrne HCCJiefloeaTejiH, Hanp. Π. Crajiji, K. liana, 3 . flaut, JI. flsace h flp.

AKTYAJIbHOE HJIEHEHHE Η ΓΛΥΒΗΗΗΗΕ CTPyKTYPW

193

Moe, flonojiHeHHe, onpeaejieHHe η τ. a., npeacraBjiaiOT co6oñ ρε3yjibTaT CJIOHCHOÍÍ HHrepaKijHH Meacay KaTeropHHMH ypoBHH A H h KaTeropHHMH o6o6meHHO-ceMaHTHHecKoro ypoBHH, KaK areHc, aeñcTBHe, ο6ιεκτ AeñcTBHfl, npH3HaK cyócraHimn Η Τ. Π. Te Η apyrne KaTeropHH πpeφepHpyIoτ OAHH cnocoöbi noBepxHOCTHoro οφορΜπεΗΗΗ H OTpHitaioT zipyrae (HanpHMep, φyHK^Ila caMOCToaTEJIBHOH TeMbi H, C Apyroñ CTopoHu, φyHκ^HA areHca cnocoôCTByioT οφορΜπεΗΗΐο aaHHoro 3JieMeHTa KaK no^Jieacamero, φyHKuaa caMOCTOHTenbHOH peMM, a TaK»ce 3HaHeHHe aeiiCTBHa cnoco6cTByioT οφορΜϋεΗΗΐο AaHHoro 3JieMeHTa KaK cKa3yeMoro Η t. ρ,.). ECJIH BJIHHHHE 060HX ypoBHeö HAXOFLHTCA Β corjiacHH (KaK, Hanp., Β (1) Η (5), το HeT flpyrHx B03MoacHocTeñ rpaMMaTHHecKoro ΠερΗφρΕ3ΗρθΒΒΗΜ3, ΚρΟΜβ aHajIH3aUHH rjiyÔHHHOrO npeflHKaTa. HaoöopoT, TaM, r^e BJIHHHHC A H h O6O6MEHHO-ceMaHTHieCKORO ypOBHH HaXOflHTCH Β npOTHBOpeHHH, TaM B03M05KHM pa3Hbie πepHφpa3L·I, nocKOJibKy „noôeflHTb" Μοκετ Η tot, Η apyroö ypoBeHb. Cp., Hanp., Σ 2 , rfle 3anpocaM ceMaHTHHecKiix KaTeropiiñ COOTBETCTBYET NOBEPXHOCTHOE ΟΦΟΡΜΫΟΗΐΙΕ ( 2 ) ffupenmop U3y\\

wem I ηροβκτη, Tor^a KaK TpeßoBaHnaM A H öojitme yAOBJieTBopaeT

ΟΦΟΡΜΠΕΗΗΕ

(11) 3mo ηροεκιη / nmo dupeKmop mywem

Η

(12) H3ynaeMoe dupenmopoM / — ηροεκτη η τ. r. 7. Το, Ητο μ η noKa3ajiH Ha npHMepe npeAJioxceHHH c ceMaHTHHeCKHMH KOMnOHeHTaMH Ag, Act H Obj, M03KH0 npOAeMOHCTpHpoBaTb — Η noacajiyM, em8 HarjisiAHee — Η Ha npe^JioaceHrax, cocToamnx H3 sjieMeHTOB Äpyroro THna. TaK, Hanp., ecjiH B3HTL· npeAJioaceHHH (20) flemu ¡ uzpawm eo deope

(^HarHocTHnecKHÖ Bonpoc: *frno dejiawm demuT)

w (21) fíemu uzpawm / eo deope (Tde uzpawm

demul)

\

(22) Bo deope / uzpawm demu w (23) Bo deope uzpawm / demu

( Hmo npoucxodum eo deopel) (Kmo uzpaem eo deopel), Hzpawufue demu /

ΤΟ ( 2 1 ) M03KH0 ΠEPHΦPA3HPOBAΤB KAK ( 2 4 ) •\

\

I Haxodnmcn eo deope HOT (25) Hzpa demeü / npoucxodum eo deope,

194

Π. ΑΛΑΜΕ1Ι Λ

ama ace (26) MecmoM, zde demu uzpaiom, / neAnemcn deop. AHajioγηηηο

CMMCJI (22) M03CH0 nepeflaTb nepHι β HanucaHoe

UM

\

COHUMHUe I ΗβΑΗεΤηεη

KOpOmKUM,

HTO HeB03M0)KH0

flJI«

(35).

9. J^aHHoe B b i c i y n j i e H H e npejceraBJiHeT c o 6 o ñ JIHLUL· n e p e y i o , B e c b M a n o B e p x H o c T H y i o H H e c o B e p m e H H y i o n o n t r r i c y npHMeHeHHH yKa3aHHOH MCTOAHKH. OAHaico, e c T b ocHOBamra cHHTaTb, HTo T a KOH ÜOflXOfl ΜΟΓ 6bl nOMOHb ΠΟΑ HOBbIM y r j I O M 3peHHH OCBßTHTb cjiOHCHeiíiiiyio n p o ô j i e M y a i c T y a j i b H o r o HneHeHna.

SEMANTIC CLASSES OF VERBS AND FSP Karel Pala (Brno)

1. The purpose of our paper is the following: (a) to describe briefly the apparatus of the so-called predicate expressions and to make use of it for some basic types of FSP; (b) to show certain relations between semantic classes of Czech verbs as are described by means of the apparatus of predicate expressions and FSP. 1.1. The data on the distribution of the degrees of communicative dynamism (CD), i.e. data on the positioning of thematic and rhematic component represents an indispensable part of the semantic information about text and about what meaning is to be assigned to a sentence and how it is to be interpreted. It follows, however, that the data about the distribution of the degrees of CD is necessary for any sort of synthesis, e.g. for generative grammars as well as for machine-translation synthesis. It means that for any particular transformational grammar its base component should be able to assign automatically the marker "rheme" (R) to certain elements of the deep structure and marker "theme" (T) to the other. It means further that this fact must be reflected in the arrangement of elements in surface structures and that it is necessary to know how to link up all the data with the corresponding consituations at the same time observing all the rules of cross-reference. These facts will undoubtedly change the present types of transformational descriptions of natural languages in a more radical way than it has as yet been supposed. Apparently it will be necessary to employ systematically the terms '.'communication" (in connection with the present concepts "sentence" and "set of

SEMANTIC CLASSES OF VERBS AND FSP

197

sentences"), "text", "theme", "rheme", "consituation", "communicative intention (of the speaker)", etc. Attempts of this kind have already appeared. Following the ideas of McCawley, Dahl has recently ( D A H L 1969) made an attempt to describe the meanings of sentences by means of splitting them into atomic sentences that are always formed by a predicate and a set of N P descriptions. This enables us to describe the atomic sentences by means of the apparatus of a modified predicate calculus and in this way to look for the relations to FSP. N o doubt these ideas are fruitful, but a detailed analysis shows that better results may be achieved in case we start working directly with verbs as predicates or predicate expressions (see e.g. A P R E S J A N 1967 vs. FILLMORE 1968). Then it is possible to avoid Dahl's problem how to define precisely the relations between the atomic sentences and how to cope with a "rather improbable hypothesis that all predications are perhaps in reality implications" even when "this hypothesis with some modifications works quite well" ( D A H L 1969, p. 19). 1.2. A note on terminology. In this paper the following terminology will be used: (a) on the syntactic level — sentence elements — subject, object, predicate, attribute, adverbial, NP, VP, PP. (b) on the semantic level — subject of the action (Sub) as e.g. agent, initiator, bearer of the state, bearer of the attitude, and object (Obj) of the action as e.g. patient, result of the action, goal of the action, goal of the movement, and further action, activity, state, movement, existence in space, attitude, as a broader term actant or the participant in the situation, place, time, manner. This semantic metalanguage is intentionally sketchy and incomplete and is in need of further elaboration. (c) o n the level of FSP — rheme (R), theme (T), the degrees of communicative dynamism, contextual dependence (CDP), contextual independence (CIDP), the basic distribution of the degrees of C D or some other terms all of which are taken from the works of J. Firbas (survey FIRBAS 1970). Symbol // denotes the boundary between Τ and R.

198

Κ. PALA

2. Verbs and predicate expressions (ΡE). For the description of verbs and their semantics it is convenient to employ the concept "predicate expression" (APRESJAN 1967) or "predicate and its arguments" (e.g. FILLMORE 1968). PE is represented by the predicate whose places are occupied by the symbols of predicate variables — arguments. If the terms of the above semantic level are used for predicates and their arguments then PE may describe the semantic structure of sentences; to a oneplace PE of the type P(a) may correspond semantic pattern SubState, where the Sub is the only argument and the State — the predicate, similarly to two-place PE may correspond the pattern Sub-Action-Obj or Sub-Movement-Obj effected by the movement, where Sub and Obj are arguments, Action and Movement — predicates (PALA 1967). In most cases one-place predicate correlates with intransitive verbs, two-, three- or four-place predicates with transitive verbs. The question is how to treat, in this framework, the expressions denoting place, time, manner and circumstantial modifiers and whether to regard them as arguments (or actants) or otherwise. There are more ways of solving this problem but I shall not pursue it at least for the time being. It will suffice that the so-called "inherent" adverbial modifiers (KOPEÖNY 1958) and some other obligatory adverbials may appropriately be treated as arguments whereas the so-called "free" or "relational" adverbials (which are optional) may in some way or other be regarded as constants. Further it should be pointed out that arguments syntactically realized as expanded NP are in fact predicates (PE) as well but of a lower rank. The same holds for arguments syntactically realized as PP or as dependent clauses (especially so-called content clauses in Czech grammars). E.g. in the sentence Vim, ze to η erti pravda (/ know it is not true), where the dependent clause represents the argument Obj, because there exists the semantic pattern SubInformation activity - Information content. 3. Predicate and FSP. We regard the apparatus of PE as a suitable means of describing even some phenomena on the FSP level.

SEMANTIC CLASSES OF VERBS AND FSP

199

From the point of view of a speaker or a generative device, the FSP of a given sentence depends in fact on how the arguments and their corresponding predicates are distributed and what their semantic contents are (on condition that each member of the PE is realized in the surface structure as at least one syntactic sentence element). Or in other words and more precisely (APRESJAN 1 9 6 7 , pp. 58 — 59), "Razumejetsja, ν zavisimosti ot togo, kakoj element situacii my sëitajem boleje vaznym, my mozem upomjanut' ego ran'se ili pozze. V rezul'tate peremeSòajetsja logiceskij akcent, a na jesce boleje poverchnostnom urovne menjajutsja sintaksiceskije pozicii leksiceskich morfem, oboznacaju§öich elementy situacii. Odnako eti izmenenija zatragivajut tol'ko sposob predstavlenija vesöej, no ne sami vesci; opisyvajemyj s toj ili s drugoj toòki zrenija klass situacij ostajetsja neizmennym." (In accordance with which element of the situation is considered to be most important, we may mention it sooner or later, which leads to a varied distribution of logical accent and, more on the surface, to a change of syntactic positions of lexical morphemes denoting the elements of a situation. These changes, however, have only to do with the way in which the things are presented, they have no bearing upon the things themselves, i.e., the situations are being described from different points of view, but the very class of situations remains unchanged.) The conception of semantic structure of sentence described as an η-place PE coincides with the degrees of CD and their distribution within a sentence. The predicate and its individual arguments may be regarded as the carriers of the degrees of CD. The following situations may arise: (1) One-place PE enables bipartition in Τ and R, where the predicate is Τ and the arguments are R and vice versa. The possible combinations are as follows: arg-T ¡I Pred-R arg-R ¡I Pred-T Pred-T // arg-R Pred-R // arg-T

200

Κ. PALA

(2) Two place PE offers more possible combinations. I should like to mention at least the basic one: arg-T Pred-T // arg-R arg-T ¡I Pred-R arg-R arg-R ¡I Pred-T arg-T arg-T II Pred-R 11 arg-T arg-T ¡I arg-R // Pred-T arg-T arg-T 11 Pred-R arg-T II arg-R Pred-R Pred-T arg-T // arg-R Pred-T // arg-R arg-R Pred-T // arg-R // arg-T

etc.

In this place we feel we ought to stress that what is usually termed as "normal" or "neutral" word order, is actually an intuitive expression of the fact that the character of human thinking and the necessity of linearizing the communication most often tend to constitute the order arg-T Pred-T // arg-R, where the first argument has semantic content "Subject of an action", "agent", "bearer of an action", etc., the predicate in the second place has the semantic content "Action", "Activity", "Movement", etc., the second argument in the third place denotes one of the possible "Objects of an action". This positioning is symmetrical in regard to the predicate (a fact being unjustly neglected) and is most closely connected with the basic distribution of the degrees of CD. It seems to be highly probable that the controversy "bipartition of the type Τ // R vs. tripartition T-Transition // R" has its origin in the existence of one-place predicates in regard to two-place and three-place ones. It is evident that from a certain viewpoint oneplace predicates only allow of Τ // R partition, whereas with morethan-one-place predicates the situation is much more complex. In addition to that, it is impossible to achieve a symmetric distribution of arguments in regard to their predicate with three- or more-than-

SEMANTIC CLASSES O F VERBS A N D FSP

201

three-place predicates. As has been shown by Firbas (FIRBAS 1968), the situation is more complicated, but it is clear that the distribution of the degrees of CD represents a gamut, whether one-place or more-place predicates are dealt with. (3) Three-place PE offers still more combinations than the preceding group and is more flexible in regard to potential context dependence. It would be interesting to ascertain which theoretically possible combinations do actually occur. The following may be regarded as the main ones: arg-T Pred-T arg-T // arg-R ' arg-T Pred-T / / arg-R arg-R arg-T /1 Pred-R arg-R arg-R arg-T arg-T Pred-T // arg-R arg-T arg-T // Pred-R arg-R arg-T ¡I arg-R Pred-R arg-R Pred-T // arg-R arg-R arg-R Pred-T arg-T // arg-R arg-R Pred-T arg-T arg-T // arg-R

etc.

We touched on predicate of different rank but in regard to the distribution of CD it is possible to state even more: N P or PP is from viewpoint of the degrees of CD segmentable only if they are at least one-place PE of lower rank. In our opinion this statement is in accordance with Svoboda's findings concerning the internal analysis of attributive constructions. Of course, the situation may be more complex than it might appear to be since some action nouns copy to some extent the predicate structure of the corresponding verbs. 4. Semantic classes of verbs and FSP. For some semantic classes of Czech verbs described by means of PE it is possible to attempt to formulate approximate rules of the distribution of the degrees of CD, i.e. rules of the arrangement of Τ and R provided the re-

202

Κ. PALA

spective sentences are contextually independent and have basic distribution of the degrees of CD. Our starting point will be meaning of verbs and their groups (classes) which to a great extent predicts what distribution of the degrees of CD should be assigned to the (semantic) PE (and to corresponding surface sentences). (1) Semantic copulas, i.e. semantically weak verbs cannot function as R and that is why their Sub (in fact denoting processes — as war, symposium on FSP, etc.) has the role of the R. Examples of this class: vznikat (to originate), zacít se (to begin), prestai (to stop), skonëit (to end, to finish), byt (to be), objevit se (to appear), konat se (to take place), probihat (to be in progress), existovat (to exist). Semantic (PE) and FSP pattern (rule) appearance-T 11 Sub (process)-R "inchoative"-T // Sub (process)-R "being in progression"-T // Sub (process)-R (2) The class of verbs denoting "existence in a place". Their main feature is that they require local specification. The marker " R " may be assigned either to the argument Sub or to the Local specification. E.g. byt nëkde (to be somewhere), nachâzet se (to occur somewhere), zit nëkde (to live somewhere), rozlozit se (to spread), ocitnout se, zdrzet se (to be detained), pracovat nëkde (to work somewhere), etc. Semantic and FSP pattern: Sub-T Existence (in a place)-T // Place-R Place-T Existence (in a place)-T // Sub-R (3) The verbs denoting "position in space" (sometimes termed as stative verbs, e.g. viset (to hang,) sedët (to sit), lezet (to lie), stât (to stand), valet se (to lie about) and possibly verbs such as spät (to sleep), etc. In regard to the preceding group they differ in that their local specification is not obligatory. Semantic and FSP pattern: (a) Sub-T Positioning-T // Place-R Place-T Positioning-T // Sub-R (b) Sub-T ¡I Positioning-R State-R

SEMANTIC CLASSES O F VERBS A N D FSP

203

(4) The verbs with the meaning "oriented movement requiring some goal or cause", e.g. jit se koupat (to go to bathe), bëzet pro pivo (to run and fetch), vyjit (to come out of something), dojit (to arrive at something) and as a subgroup the verbs of the type uzivat nëceho k nëcemu (to use something for some purpose), uvëznit nëkoho proc (to imprison somebody because of), odsoudit nëkoho proc (to condemn somebody for some reason) and others. Rhematic components are the goal and the cause. Semantic and FSP pattern: Sub-T Oriented movement-T // Goal-R Cause-R (5) The verbs with the meaning of "implied negation" are themselves rhematic components even if their arguments are present. This is due to the fact that expression of negation is communicatively more dynamic than the other participants (arguments) in the situation, e.g. zmizet (to disappear), propadnout (to fail), propadnout se (to drop), uniknout (to escape), vyhnout se (to avoid), chybët (to be missing), odmitnout (to refuse), etc. All of them are formally non negative. Semantic and FSP pattern: Sub-T //Impliednegation-R // Obj-T Sub-T 11 Implied negation-R Place-R (6) The verbs denoting "attitude". It is typical of them that their grammatical subjects are semantic Obj (DANES 1968). From this point of view it means that rhematic component is usually semantic Obj (but grammatical subject). This conception is based on the fact that, with action verbs, the semantic Obj is more dynamic than the verb itself. E.g. libit se (to please), tësit (to enjoy — like German "gefallen"), zdât se (to seem), jevit se (to appear), pfekvapit (to surprise), etc. Semantic and FSP pattern: Obj-R // Sub-T Attitude-T or Attitude-T Sub-T // Obj-R (with the Obj brought into prominence)

204

Κ. PALA

(7) The verbs with the meaning "behaviour " require specification as to the manner of behaviour which constitutes the rhematic element because — in regard to the meaning of PE as the whole — it is communicatively most dynamic. E.g. chovat se jak (to behave), drzet se (to do well), jednat (to act), zit nëjak (to live), postupovat (to go about), projevit se (to proved to be), oblékat se jak (to dress) and others. Semantic and FSP pattern: Sub-T Behaviour-T // Manner-R (8) The class of verbs whose objects denote "a person or a thing affected or effected by the action" and other verbs of the type "faciendi", e.g. bit nëkoho (to beat), irezat (to cut), odmënovat (to reward), stavët (to build) and many others. As a matter of fact this class of verbs probably comprises a number of more refined subclasses (maybe 10—15 possible semantic types). It has been proved that in the case of verbs of this type in particular ( F I R B A S 1970 — in fact 1959) the semantic Obj is communicatively the most dynamic (and rhematic) element. Semantic and FSP pattern: Sub-T Action, Activity-T // Obj of action-R activity-R (9) The verbs denoting "losing, seeking, finding". Their meaning suggests the conclusion that the location of the loss of the Obj and that of finding it is to be regarded as the most dynamic element even if in certain circumstances the Obj lost or found becomes rhematic. Examples of this group are: ztratit nëco nëkde (to lose st. somewhere), nafit (to find), nalézt (to find), objevit (to discover), hledat (to seek), nechat (to leave), etc. Semantic and FSP pattern: (a) Sub-T Losing-T Obj-T // Place-R Finding-T (b) Sub-T Losing-T // Obj-R Finding-T

SEMANTIC CLASSES O F VERBS A N D FSP

205

(10) The verbs with the meaning "perception", "evaluation" •whose characteristic feature is the presence of the manner of perception or evaluation as a rhematic element but here, too, the perceived object may become rhematic. E.g. brât nëco nëjak (to take), pfijimat (to accept), vyklâdat (to interpret), vysvëtlovat (to explain), vyjádñt se (to express), vnimat (to perceive), hodnotit (to appreciate), châpat, rozumët (to understand), etc. Semantic and FSP pattern: (a) Sub-T Evaluation, Perception-T // Manner-R (b) Sub-T Perception-T // Obj-R Some multiambiguous verbs may belong to several classes at the same time. On the one hand it is evident that our list of verb classes is incomplete, but on the other hand it clearly shows how the semantics of verb has a bearing on FSP and how to proceed in this direction. The final results of such a procedure should be formal rules applicable to any kind of synthesis. To prevent the search for seeming counter-examples, it is necessary to emphasize that the following rule holds: any argument or its predicate may become rhematic if it is given special intonation, if it is preceded by a special lexical means (we can speak of a kind of operator), or if it is put into a contrast position (e.g. FIRBAS 1970).

5. Semantic classification of adverbial modifiers and FSP. It might have been gathered from what has so far been put forth that semantic classification of adverbial modifiers from the viewpoint of FSP is equally necessary as the analogical classification of verbs. We don't intend to deal with this problem in detail — as it would be a topic for a separate paper — but we should only like to stress two basic things: (1) We have already mentioned the so-called inherent adverbial modifiers, which are, in fact, semantic supplements of verbs, or according to Kopecny (KOPECNY 1958), "ad-verbal attributes", expressing the degree (intensity) or quality of a verbal action. As

206

Κ. PALA

for FSP, it means that they are — in the same way as attributive expressions in N P groups — communicatively more dynamic than their headwords (verbs) and together with them (or under special c i r c u m s t a n c e s o n l y b y themselves) r e p r e s e n t a n R (SVOBODA 1968).

This fact is also reflected by the word order (UHLÍROVÁ 1969) and is the cause of the striking difference betwen their word order behaviour and that of the so-called free, relational adverbial modifiers. Inherent adverbial modifiers are most frequently posited immediately before or after the verb. (2) The so-called free or relational adverbial modifiers are connected with the verb only facultatively, they are attached to the verb from outside and do not constitute an inherent component of its content. In this case the verb functions semantically as a linking, transitive element. From the viewpoint of FSP, it means that "free" adverbial modifiers may be either thematic or rhematic element and probably do not markedly prefer one possibility to the other. As for their word-order behaviour, they represent the most mobile elements of a sentence and are posited especially at the beginning o r a t t h e e n d of a sentence (UHLÍROVÁ 1969, MATVEJENKO 1969).

The mobility is not only their characteristic feature in Slavonic languages, but it seems very probable that similar behaviour is to be expected in English or German. Interesting investigations and new procedures in this field have been dealt with by MATVEJENKO (1969b). We hope it may have been gathered, even from our sketchy account, in what way the semantics of adverbial modifiers is connected with the semantics of verbs, how the two phenomena influence each other, and what conclusions may be drawn from it within the sphere of FSP. REFERENCES APRESJAN, JU. D., Eksperimental'noje opisanije semantiki russkogo glagola. Moskva 1967. DAHL, Ö., Topic and comment: A study in Russian and general trasnformational grammar. Göteborg 1969. DANES, F., Some thoughts on the semantic structure of the sentence, Lingua 21, 1969, 5 5 - 6 9 .

SEMANTIC CLASSES OF VERBS A N D FSP

207

J., Lexical entries for verbs, FL 4, 1968, 373—393. FIRBAS, J., On the prosodie features in the modern English finite verb as means of FSP(More thoughts on transition proper),BSE 7,1968,11—48. FIRBAS, J., Some aspects of the Czechoslovak approach to problems of FSP, pp. 11—37 in this volume. KOPECNY, F., Zàklady éeské skladby [Foundations of the Czech syntax], 2 9 - 3 4 , 181. Praha 1958. MATVEJENKO, V. Α., Nëkotoryje faktory, vlijajuSdije na razmeäienije obstojatel'stv ν russkoj fraze. In: Masinnyj perevod iprikladnaja lingvistika 12, 1969a, 1-j MGPIIJA, 8 1 - 9 5 . Moskva. MATVEJENKO, V. Α., Materialy dlja avtomatiöeskogo raspoznavanija logiöeskich akcentov. In: Masinnyj perevod i prikladnaja lingvistika 11, 1969b, 1-j MGPIIJA, 153-176. Moskva. PALA, K., Otnoäenije mezdu porjadkom slov i aktual'nym ölenenijem ν Ceäskom jazyke, PSML 2, 1967, 5 1 - 6 4 . SVOBODA, Α., The hierarchy of communicative units and fields as illustrated by English attributive constructions, BSE 7, 1968, 49—101. UHLiftovÀ, L., Vztah syntaktické funkee vëtného ölenu a jeho mista ve vëtë [Relationship between syntactic function and linear position of sentence elements], SaS 30, 1969, 360— 370. FILLMORE, C H .

ON THE ROLE O F STATISTICS I N THE INVESTIGATION OF FSP Ludmila Uhlírová (Prague)

1. In accordance with the view generally accepted in the Prague School linguistic theory, as was presented in the paper by J. Firbas, 1 FSP is a phenomenon of a systemic character, i.e., a phenomenon belonging by its nature into the sphere of linguistic competence, but at the same time a phenomenon with a very strong impact on linguistic performance, too. It has been clearly pointed out by recent developments in linguistics and many times stressed during this symposium that it is both important and fruitful to investigate not only problems of linguistic competence, but also problems of linguistic performance from different aspects of human communication. One of the methods, or procedures, often applied to the study of linguistic p e r f o r m a n c e is the statistical one. Perhaps all kinds of investigation of higher levels of language so far performed with the help of statistics are concerned predominantly with its performance. They usually start with a frequency count of some elements in a representative corpus of material and aim at the establishment of some significant correlations the interpretation of which can have a certain linguistic relevance. The role of statistics lies in the fact that it helps in making a differentiation between and evaluation of what is really implemented during the language performance (i.e., the empirical linguistic reality) on the one hand, and on the other hand what is implementable potentially, i.e., what is perFIRBAS, J . , Some aspects of the Czechoslovak approach to problems of functional sentence perspective, pp. 11—37 in this volume. 1

STATISTICS I N THE INVESTIGATION OF FSP

209

mitted by the system of language), but actually does not occur in texts. One of the relationships to be investigated also statistically is the relationship between two levels of syntax, namely — according to Danes's model 2 — between the level of the grammatical syntactic structure of sentence and the level of the organization of utterance; more concretely, it can be presented as the relationship between the grammatical s y n t a c t i c functions of sentence elements and their c o m m u n i c a t i v e functions in utterance. It is this relationship that forms the topic of the present contribution. 2. It was already pointed out by V. Mathesius 3 that languages with rich morphology, such as Slavonic languages, usually have free word order and that word order is — together with intonation — the main means of expressing FSP. In contradistinction to the languages with fixed (grammaticalized) word order, such as English, it holds that the grammatical sentence elements in Slavonic languages are arranged one after another in utterances just according to their degrees of communicative dynamism, the importance of grammatical syntactic functions for the linear arrangement being very limited. Thus, it is not essential from the communicative point of view, whether, e.g., a Czech clause begins with grammatical subject, or grammatical object, or with some other syntactic element; the essential point is, whether it begins with a thematic, or rhematic element, or perhaps with a communicative element of another kind. Potentially, each syntactic element can be a carrier of different communicative functions, both thematic and rhematic; and conversely, each communicative function has various syntactic means of implementation as well. It would be possible to give plenty of examples of clauses with thematic subjects, as well of 2

DANES, F., A three-level approach to syntax, TLP 1, 1964, 225—240. MATHESIUS, V., Obsahovy rozbor souiasné angìiètiny na zâkladë obecni lingvistickém [A functional analysis of present-day English on a general linguistic basis], Prague 1961. — MATHESIUS, V., O tak zvaném aktuálním ¿lenëni vëtném [On the so-called functional sentence perspective]. In: Cestina a obecny jazykozpyt [The Czech language and general linguistics], Prague 1947, 2 3 4 - 2 4 2 . 3

210

L. UHLÍÜOVA

clauses with thematic objects, examples of clauses with transitional predicates as well of clauses with thematic predicates etc. In other words, the relationship between the grammatical syntactic element and its communicative function is never b i - u n i q u e . As has been proved by empirical research, nor the opposite extreme occurs: No grammatical syntactic element implements different communicative functions with the same p r o b a b i l i t y , e.g. no grammatical syntactic element can be the theme of utterance as often as it is the transition and as often as it is the rheme. Each grammatical syntactic element prefers one, sometimes two communicative functions and the degrees of preference can be objectively measured by means of statistics. This holds for the surface structure both of utterances in context and of isolated utterances. 3.1. The concrete statistical research may proceed in two directions, i.e. a) from the grammatical syntactic elements to their communicative functions, or briefy, from the syntactic f o r m to the communicative f u n c t i o n . It is investigated which communicative function (thematic, transitional, rhematic) is carried by a given syntactic element most often, less often etc. b) Inversely, from the communicative function to the syntactic form. It is ascertained which syntactic element implements a given communicative function most often, which one less often etc. 3.2. The former of the two just mentioned approaches leads to the conclusions which will be now illustrated by several examples from written Czech. Ex. 1. A grammatical syntactic element is a carrier of almost exclusively one communicative function, only exceptionally of other functions. E.g. more than 90% of all occurrences of the verbal predicate in the examined texts of written Czech function as transition between theme and rheme; the same holds for the copula.4 — More than 90% of the predicative attributes ("doplnëk" in the Czech terminology) function as rheme of utterance. 4 Firbas, J., On the prosodie features of the modern English finite verb as means of functional sentence perspective. BSE 7, 1968, 11—48, Brno.

STATISTICS IN THE INVESTIGATION OF FSP

211

Ex. 2. A syntactic element is a carrier of two functions, one of them being more frequent than the other. E.g. two thirds of all occurrences of subjects are thematic, and only one third of them are rhematic. — Two thirds of occurrences of the objects are rhematic, and only one third are thematic objects. Ex. 3. A syntactic element is a carrier of two functions, the m e a n frequencies of which are approximately the same, but the frequency distribution shows a strong dependence on the length of clause. E.g., the so-called inherent adverbials, i.e. adverbs qualifying the verbal predicate of clause (semantically a kind of "attribute" to verb) 5 function as transitional elements roughly in 50% and as rhematic also in 50%. They can either function as separate rhematic communicative constituents, this function statistically prevailing in short clauses, or as o n e communicative constituent t o g e t h e r w i t h the transitional predicate, this function prevailing in longer clauses. 3.3. The latter of the two approaches mentioned in 3.1. leads to such conclusions as follows: For each communicative function there is one syntactic element (for the rhematic function two syntactic elements) which is most typical of it and which renders it much more often than other syntactic elements. Thus, the theme is most frequently rendered by the subject: 6 53% of the occurrences of thematic constituents are the subjects of clauses. The transition is most often rendered by the predicate (finite verb), in 54%. The rheme is most frequently rendered either by an object o r by an adverbial (55%). These data are, undoubtedly, of a comparative value with regard to the data of the former approach. At least they give evidence of 5 KOPECNY, F., Základy âeské skladby [Foundations of the Czech syntax]. Prague 1962. — ADAMEC, P., Κ otázce uplatnëni vëtnych ölenü ν strukturních popisech jàzyka [On the role of sentence elements in structural lingustic descriptions]. In Otázky slovanské syntaxe 2 [Problems of the Slavonic syntax 2] Brno 1968, 111—114. 6 MATHESIUS, V., O funkci podmëtu [On the functioning of subject]. In: Cestina a obecny jazykozpyt [The Czech language and general linguistics], Prague 1947, 277-285.

212

L. UHLÍS.OVÁ

the fact that the approach "from the syntactic form to the communicative function" is not identical in its results with the reverse approach and that the two approaches can analyze given phenomena from different aspects. 4. The numerical data mentioned here for illustration only could be supported by more detailed statistical analysis, which was performed for Czech.7 The relationship between syntactic and communicative functions of sentence elements proves to be quite close. This closeness, however, is not due to the grammatical syntactic nature of sentence elements, but due to the semantic nature of the communicative structure. More concretely, it is the manifestation of the fact that Danes's three levels of syntax are not isolated from each other, the manifestation of a strong interference of semantics with the grammatical syntactic structure of clause. The degree of closeness, as quantitatively illustrated above, can serve — and this is the main point — as a certain objective m e a s u r e for establishing to*what degree the syntactic and communicative functions of an element of clause are determined by their general meaning as a part of speech, or by their appurtenance to semantic categories, such as actor, goal, action, circumstances etc. 5.1. A suitable syntactic category for a more detailed discussion on links among the three levels of syntax is the adverbial, because this category, as usually defined in grammars, is rather heterogeneous both from the purely syntactic and from the semantic points of view (its heterogeneity was subjected to criticism many times; recently Adamec (op. cit. in fn. 5) has elaborated a new classification according to which is split the adverbial into three diiferent syntactic categories). As for the communicative point of view, we shall try to show that some subclasses of adverbials display different communicative properties in comparison with other ones. 5.2. There is a class of the so-called inherent adverbials (already mentioned in sec. 3.2.), usually defined as adverbs qualifying the 7 UmiftovÁ, L., On the quantitative analysis of clause and utterance in Czech, PSML 4, 1972, 107-128.

213

STATISTICS I N THE INVESTIGATION O F FSP

verbal predicates of clauses. As a rule, inherent adverbials either belong to the transition (provided that the transitional verbal predicate is present), or they are the rheme proper; other communicative functions are rare and inherent adverbials carry them only under special circumstances. Besides, there is an apparent dependence on the length of clause — the proportion of the rhematic inherent adverbials decreases with the clause length (cf. changes in the ratio "length/% of rhematic adverbials": 2/75%, 3/47%, 4/42%, 5/33%, 0/49%) and at the same time the proportion of the transitional adverbials increases accordingly (cf. again the ratio "length/% of trans, adverbials": 2/25%, 3/36%, 4/47%, 5/58%, 0 /42%). Inherent adverbials are the only subclass of adverbials functioning in the communicative structure as a part of transition. In contradistinction to them, the rest of adverbials (most usually nouns in prepositional cases) are either thematic or rhematic. A more detailed quantitative examination leads to the following numbers: length thematic adverbials rhematic adverbials

2 14% 86%

3 53% 47%

4 54% 45%

5 55% 45%

It holds only of clauses of the length of 2 that the adverbial prefers just one, rhematic communicative function; frequencies of occurrence (given in %) in clauses longer than 2 should be interpreted rather as a kind of balance between the thematic and rhematic functions than a clearly marked prevalence of the rhematic function. (Note: Under the "length" we understand the number of grammatical syntactic elements in clause, excluding all attributes to nouns.) The differences in usage of inherent and non-inherent adverbials in the communicative structure are demonstrated by means of a roughly complementary word-order distribution. Whereas the inherent adverbial prefers the positions in the middle of clause (this preference increasing with length), the most frequent positions of the other adverbials are the beginning and the end of clause,

214

L. UHLÍftOVÁ

both being equally typical (the average frequency at the beginning amounts to 35%, at the end to 37%). Significant differences between the beginning and end of clause are found only in the shortest ones, where the frequency at the end prevails (57%). Positions in the middle are not typical of the adverbial, the frequencies in these positions being almost equal to one another. Thus, the different communicative functions and the different word-order distribution as well reflect the different semantic nature of the two subclasses of adverbials under consideration. The deep-structure relation between "study" and "diligently" in the phrase "to study diligently" ("diligently" being an inherent adverbial to verb) is close to that between "student" and "diligent" in the phrase "a diligent student" ("diligent" being an attribute to noun 8 ). The non-inherent adverbials are often optional complements to predicate and render time, place, manner, circumstances etc. Another word-order property characterizing the affinity of inherent adverbials and attributes is worth mentioning. Being a part of transition, the inherent adverbial most often occurs in the contact preposition before verb (contact preposition: 93%, distant preposition 7%), whereas being the rheme proper, the distant postposition after verb is quite usual. 5.3. In sec. 3.2. we claimed that each syntactic element prefers one or two communicative functions. 9 There is only one exception to this general rule, namely the adverbial (but not the subclass of the inherent adverbial). As has been already shown in sec. 5.2., both thematic and rhematic functions are equally typical of it. It is clear evidence of the semantic complexity of this category. Be8

KURYLOWICZ, J., Dérivation lexicale et dérivation syntaxique, Bulletin de la Société linguistique de Paris 37, 1936, 79—92. — SGALL, P., Generatami popii jazyka a ieskà deklinace [Generative description of language and the Czech declention]. Prague 1967. 9 PALA, Κ., Otnosenije mezdu porjadkom siov i aktual'nym ölenenijem ν öeäskom jazy ke, PSML 2,1967, 57—64. — BUTTKE, K., Gesetzmässigkeiten der Wortfolge im Russischen, Halle 1969. — ELEKFI, L., Kriterien der aktuellen Satzgliederung in ungarischen "Kernsätzen", 2t für Phonetik, Sprachwissenschaft und Kommunikationsforschung 22, 1969, 226—351.

STATISTICS IN THE INVESTIGATION OF FSP

215

sides, in this connection it is not without importance, that thematic elements can be further classified. Utterances often include elements which express various circumstances of the action, condition, time, place, manner etc. and which are thematized only very weakly, sometimes even indistinctly. E.g. in the utterance Za tohoto predpokladu ( = theme proper) mùzeme najít ( = transition) ν prenosové technice (a weakly thematized element) zdroje jiného druhu ( = rheme) (literally: Under this assumption, we can find in broadcast technique sources of a different kind). The loose contextual connection of the optional element "v pïenosové technice" (in broadcast technique) is in accordance with its relatively loose position in the linear arrangement of utterance. It usually stands either at the beginning of the utterance, or in the middle, following after thematic and/or transitional constituents. The change in its position (such as: V prenosové technice mùzeme za tohoto predpokladu najít zdroje jiného druhu) in many cases does not significantly influence the communicative structure of the whole utterance — both variants (with a changed word order) can be used in the same context types. We propose to consider this kind of constituent as a separate category different from the theme, transition and rheme and call them optional constituents. While the former three constituents are the basic constituents of the communicative structure of the utterance, or discourse (according to the ways of the consecutive thematization/rhematization, different types of "thematic progression" 10 in texts are distinguished), the role of the "optional" constituents is here only of secondary importance. Their main function consists in connecting the T-R nexus of a given utterance with a wider thematic frame of the whole paragraph, chapter, or perhaps thematic sphere. What is most important, however, is the fact that the majority of "optional" constituents (65% in our material) are expressed by one and the same syntactic category, namely by (non-inherent) 10 DANES, F., Zur linguistischen Analyse der Textstruktur, Folia linguistica 4, 1970, 72—78. — DANES, F., FSP and the organization of the text, pp. 106—128 in this volume.

216

L. UHLÍ&OVÁ

adverbiale. This leads us to the hypothesis — let us present it by the way of conclusion — that a more detailed semantic classification of adverbials should reveal some classes, called, e.g., situativa (with J. Firbas) or otherwise, of typical "optional" constituents in the communicative structure of clause.

ZUR TERMINOLOGIE DER FSP Frantisek Dane§ et al. (Prag)

Die Entstehungs- und Entwicklungsgeschichte der von Mathesius geprägten Terminologie für „aktuální clenëni vëtné" wurde bis zum Jahre 1962 von F. Danes dargelegt.1 Im Anschluß an diese Ausführungen sollen hier auch terminologische Ansätze erwähnt werden, in denen schon vor Mathesius und später unabhängig von ihm die Erfassung desselben sprachlichen Phänomens versucht wurde. H. Weil hat seine Einsichten terminologisch noch nicht fixiert. Erst G. von der Gabelentz hat die Termini „psychologisches Subjekt" und „psychologisches Prädikat" eingeführt. Diese Termini wurden zwar von W. Wundt abgelehnt, aber von H. Paul verteidigt und treffend erläutert; auch später wurden sie von vielen Gelehrten verwendet. Die Termini Subjekt und Prädikat wurden also eigentlich in doppeltem Sinne gebraucht. Eine ähnliche Doppeldeutigkeit scheint freilich schon den Termini des Aristoteles "hypokoimenon" („das Zugrundeliegende") und „kategoroumenon" („das dazu Gesagte") innegewohnt zu haben. Aristoteles hat zwar diese Termini geprägt, um logische Begriffe zu beschreiben; da er sie aber der Sprache abgelesen hat, spiegeln sie eher die Gegliedertheit des Satzes in einen thematischen (expositionellen) und einen explizierenden Teil, 1 DANEÍ, F., Prispëvek k novëjSi syntaktické terminologii. Slavjanska lingvistiina terminologija I, Sofia 1962, 46—52. — Für den weiteren Text sei im allgemeinen auf Z. TYL (ed.), A tentative bibliography of studies in functional sentence perspective verwiesen, Praha 1970 (vervielfertigt), wenn keine näheren Hinweise gegeben sind.

218

F. DANES et al.

in „Thema" und „Aussage" wider.2 (Durch die lateinischen Übersetzungen „subiectum" und „prädicatum" wurde später ihre Verwendbarkeit auf das Urteil eingeengt.) Im 19. Jahrhundert war allerdings die Wahl der Termini „psychologisches Subjekt" und „psychologisches Prädikat" für völlig vom grammatischen Subjekt und Prädikat unterschiedliche Begriffe nicht glücklich; der ziemlich klar erkannte sprachpsychologische Sachverhalt wurde dadurch nur verdunkelt. Schließlich wurde die terminologische Verwirrung auch noch dadurch vergrößert daß man darüber hinaus auch noch vom „logischen Subjekt" sprach, wobei man darunter z. T. „das handelnde Subjekt", z. T. aber auch die „Exposition" meinte.3 Außerdem tauchten auch andere Namen auf zur Bezeichnung der einzelnen Aspekte, die dann später in der Theorie von Mathesius vereinigt werden sollten. So hat z. B. M. Kuttner von „Ausgangs- und Zielvorstellung" gesprochen und später auch die Termini „Start" und „Ziel" erwogen. L. Sütterlin hat die Namen „vorbereitete" und „neue Vorstellung" verwendet. E. Richter unterschied sehr genau zwei Typen der Wortstellung: 1. die rücksichtsvolle oder sachliche und 2. die rücksichtslose oder persönliche. Einer ähnlichen Unterscheidung begegnet man auch bei E. Lerch: 1. die auf den Hörer eingestellte (soziale oder pädagogische), steigende und 2. die impulsive, fallende Wortstellung. Auch die kontextbedingten Modifikationen der usuellen Wortfolge wurden immer wieder mit verschiedenen Namen belegt, wie „Anschlußzwang", „logische Verknüpfung", „sachliche Wortstellung", „psychologische Gliederung" u. a. Schon diese terminologische Mannigfaltigkeit — obwohl wir uns hier nur auf den Bereich der deutschen Sprachforschung beschränken — ist ein Zeugnis dafür, daß es dabei an einer einheitli2 N a c h JUNKER, H . F. J., Grundfragen des koreanischen Satzbaues. fViss. Zt. der Humboldt-Universität Berlin. Gesell, u. sprachw. Reihe 7, 1957-58, 329-350. 3 WEGENER, PH., Untersuchungen Halle 1885.

über die Grundfragen

des

Sprachlebens.

Z U R T E R M I N O L O G I E D E R FSP

219

chen und allgemein anerkannten Deutung der Einzelerscheinungen mangelte. Vorherrschend war die sprachpsychologische Einstellung: die Zerlegung der Gesamtvorstellung und ihre Zusammensetzung aus zwei Hauptkomponenten in der Seele des Sprechers und des Hörers. Zu einer vertieften Auffassung der ganzen Frage hat der Sprachphilosoph H. Ammann wesentlich beigetragen. Er hat für die beiden Hauptkomponenten neue Termini „Thema" und „Rhema" gewählt und ihren Inhalt scharfsinning umrissen, indem er diese Begriffe aus der Differenz der Bewußtseinslage des Sprechenden und des Hörers entwickelt. Bei Ammann stehen die Wechselbeziehungen zwischen Denken, seelischem Leben und Mitteilungsprozeß im Vordergrund. Erst bei Mathesius gewinnt die prinzipiell linguistische Betrachtungsweise die Oberhand. Er unterscheidet scharf zwischen der formalgrammatischen und der kommunikativen, auf eine aktuelle Redeabsicht eingestellten Satzgliederung. Die wörtlichen Lehnübersetzungen des tschechischen Originalterminus „aktuální clenëni vëtné" ins Englische und ins Französische 4 haben sich nicht eingebürgert; die Lehnübersetzungen ins Deutsche und Russische („die aktuelle Satzgliederung", „aktual'noje clenenije predlozenija") sind aber z. T. schon üblich geworden (bei Elekfi, Lapteva u. a.). Mit Mathesius' Originalterminus wird der Unterschied zwischen dem abstrakten Satzschema (auf der Langue-Ebene) und seiner Realisierung in einer konkreten Äußerung in einem aktuellen (Situations)-Kontext (auf der Parole-Ebene) hervorgehoben. Andere Modifikationen des Terminus betonen entweder, daß diese Gliederung im kommunikativen (partner- und situationsbezogenen) Redeakt zutage tritt (vgl. Ivancev: „komunikativno delenie", „kommunikativnoje ölenenije"), oder daß erst dadurch der eigentliche Sinn, bzw. : die Hauptinformation der Äußerung (die Redeabsieht des Sprechers) realisiert wird (Krusel'nickaja: „smyslovoje clenenije", * Vgl. VACHEK, J., Dictionaire de linguistique de l'École de Prague. Utrecht 1960.

220

F. D A N E S et al.

Boost: „Gliederung auf der Sinn-Ebene"). Auch Garvin übersetzt den Originalterminus von Mathesius durch „information-bearing structure of the sentence". Das Wort „ilenëni" ( = „Gliederung") deutet an, daß es sich um ein bestimmtes Organisations- oder Aufbauprinzip des Satzes handelt. In seinem deutsch geschriebenen Aufsatz hat Mathesius einen anderen Terminus verwendet, u. zw. „Satzperspektive". Damit wird hervorgehoben, daß es sich um die Organisation des Satzes unter einem bestimmten Blickwinkel, aus einer bestimmten Sicht handelt. Die Art der Sicht wird aber im Terminus nicht erwähnt. Firbas hat diesen Terminus wieder aufgegriffen und durch das Attribut „funktional" erweitert: „functional sentence perspective". Dieser Terminus hat einige Vorteile; aber das Wort „funktional" ist nicht prägnant genug und eigentlich redundant. Benes verwendet deshalb lieber die Bezeichnung „Mitteilungsperspektive", womit wieder das kommunikative Moment betont wird. Doch scheint wieder dieser Terminus zu weit zu sein. Das ganze Phänomen kann auch nach den Hauptkomponenten benannt werden: So hat auch K. Boost (1955), auf Ammann fußend, den Terminus „Thema-Rhema-Gliederung" eingeführt. Die amerikanischen Linguisten gebrauchen wieder gern „topiccomment" als Grundlage für verschiedene Beziechnungen dieses Phänomens. Die Bezeichnungen für beide Hauptkomponenten, die man bei der Betrachtung des Phänomens gewöhnlich unterscheidet, sind noch mannigfaltiger als die Namen für die ganze Erscheinung. So findet man schon bei Mathesius verschiedene Termini: „základ", „téma", „vychodisko" — „jádro" ( = „Kern"), bei Boost (nach Ammann): „Thema" — „Rhema", bei Bally und de Groot: „thème" — „propos", bei Elekfi „Thema" — „Propositum", bei Krusel'nickaja: „dannoje" — „novoje", bei Ivanöev: „ekspozicija" — „kulminacija", bei Y. R. Chao (wohl zum erstenmal) und Ch. Hockett: „topic" — „comment" usw. Auch die älteren Termini „psychologisches Subjekt" und „psychologisches Prädikat" halten sich noch bis heute; von einigen sowjetischen Gelehrten (A. S. Mel'niöuk, P. Z. Panfilov) wurden

Z U R T E R M I N O L O G I E D E R FSP

221

die Termini „logiceskij Subjekt" und „logiceskij predikat" deshalb wieder eingeführt, weil es sich nach ihrer Meinung um eine Widerspiegelung der logischen Kategorien in der Sprache handelt. Tatsächlich ist eine terminologische Einigung auf diesem Gebiet deshalb schwierig, weil der begriffliche Inhalt dieser Termini von verschiedenen Forschern recht unterschiedlich interpretiert wird. Schon Mathesius hat zwar den Unterschied angedeutet, der zwischen dem kontextbedingten „Ausgangspunkt" ( = „vychodisko") und dem „Thema" als Mitteilungsgrundlage besteht; aber er hat diesen Unterschied nicht folgerichtig ausgearbeitet. Boost identifiziert wieder das „Thema" mit dem Drachschen „Vorfeld" ( = Raum vor dem Verbum fini tum im deutschen Aussagesatz). Benes hat deshalb vorgeschlagen, die Termini „Thema" ( = Thema im Sinne von Mathesius) und „Basis" ( = Thema im Sinne von Boost, Ausgangspunkt im Sinne von Mathesius) genau zu differenzieren. Eine ähnliche Unterscheidung scheinen auch Μ. Α. K. Halliday und W. Dressler zu empfehlen, die „theme" und „topic" als differenzierte Termini verwenden. Die „Basis" wäre also ein Teil des „Themas". Auch im „Rhema" wird ein Zentralteil ausgesondert, das „Intonations- und Mitteilungszentrum" (Danes), auch kurz „Sinnwort" genannt. Infolgedessen unterscheidet man weiter zwischen „einfachen" und „komplexen" Themen und Rhemen (Adamec). Außer diesen zwei Komponenten wird manchmal noch eine dritte ausgegliedert und mit dem Terminus „prechodová slozka", „transition", „Bindeglied" (so schon bei H. Paul) bezeichnet. Ein wichtiger Terminus ist „communicative dynamism" (Firbas), „kommunikativnaja nagruzka" (Krusel'nickaja), „Mitteilungswert" (Bost), der in der objektiven Abfolge der Mitteilungsperspektive allmählich ansteigt. Auch für die gegenläufigen Anordnungen der Mitteilungsperspektive: Thema — Rhema χ Rhema — Thema werden verschiedene Namen gebraucht: „objektive und subjektive Abfolge" (Mathesius), „Eindrucks- und Ausdrucksstellung" (Drach) u. a. Die Vielfalt der Termini in diesem Forschungsbereich, die mit

222

F . D A N E S et al.

unseren Bemerkungen nur angedeutet wurde, ist durchaus begreiflich. Weil es sich um einen verhältnismäßig jungen Forschungsbereich handelt, will jeder Forscher seine eigene Auffassung (auch wenn sie sich nur wenig von den anderen unterscheidet) auch durch seine eigene Terminologie erhärten. Es ist aber dann auch möglich und nicht selten, daß derselbe Begriff oder Sachverhalt mit unterschiedlichen Termini und umgekehrt auch daß unterschiedliche Begriffe oder Sachverhalte durch denselben Terminus bezeichnet werden. Eine terminologische Einigung wäre dringend nötig. Voraussetzung dafür ist allerdings, daß man sich zunächst über die Grundbegriffe einigt, die terminologisch festgehalten werden sollen. Auf diese Weise soll einerseits die Forschungsgeschichte angedeutet, anderseits die heute noch herrschende Vielfalt von verschiedenen Auffassungen übersichtlich dargestellt und miteinander konfrontiert werden.