Roman Jakobson: Echoes of his Scholarship [Reprint 2020 ed.] 9783112329788, 9783112329771

Citation preview

ROMAN JAKOBSON

ROMAN JAKOB SON ECHOES OF HIS SCHOLARSHIP edited by

DANIEL

ARMSTRONG and

C. H. VAN

SCHOONEVELb

Indiana University

LISSE

T H E PETER DE R I D D E R PRESS 1977

This edition © 1977 in The Netherlands by the Peter de Ridder Press. No part of this book may be translated or reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the Publisher. Reprint rights to individual chapters are reserved by the respective authors.

ISBN 90 316 0147 0

Printed in Great Britain by William Clowes & Sons Limited London, Beccles and Colchester

FOREWORD

In the original conception of this volume, first proposed by Peter de Ridder, the intention was that a scholar in each branch of science where the influence of Roman Jakobson's work and thinking has been felt should be invited to write a brief chapter describing and analyzing this influence. Other scholars would be asked to write more general evaluatory essays on various aspects of Jakobson's professional activities. The resulting volume, it was felt, would be a fitting tribute to the man—already honored twice previously by more conventional anniversary volumes, once more by a student Festschrift, and again by the publication of his bibliography— whose keen insight and tireless efforts have for more than sixty years contributed so greatly to so many fields. We soon found that this plan was very difficult if not impossible to realize. It was, first of all, extremely difficult to divide the vast areas of science upon which Roman Jakobson's scholarly activities have had direct or indirect impact into topics which could reasonably be handled in brief essays. Some areas, no doubt important ones, were inevitably overlooked. It was, moreover, not feasible to expect a large collective of authors, writing independently, each to analyze one of the often closely-related fields in which Roman Jakobson—one of the most far-ranging scholars of the twentieth century—has worked without touching at the same time on neighboring areas. Thus, some overlapping among the chapters was unavoidable, but, in order to preserve the integrity of each essay, we decided it best not to edit out this repetitious material. And then there are some topics of those planned which are unfortunately not represented here due to cancellations by the authors originally invited. In some instances, we were able to invite other scholars to cover these topics, but it was decided not to delay the appearance of the volume any longer by seeking authors for topics abandoned at the last minute. Represented in the volume in its present state, then, are only some—but by far not all—the fields which have been enriched by the multi-faceted scholarship of Roman Jakobson. We would like to express our appreciation to the contributors to this volume, many of whom have had to wait patiently for its publication, and especially to Krystyna Pomorska-Jakobson, who provided us with a number of excellent suggestions and saw to it that our many questions received prompt replies.

vi

FOREWORD

The book itself is presented as a token of friendship and admiration, and as a symbolization of warmest wishes and congratulations to Roman Jakobson from his friends and admirers. D.A., C.H.v.S.

TABLE OF CONTENTS

Foreword

v

By Way of Introduction: Roman Jakobson's Tenets and Their Potential C. H. van Schooneveld

1

The Study of Paleosiberian Languages Robert Austerlitz

13

Roman Jakobson's Cooperation with Scandinavian Linguists Astrid Baecklund-Ehler\

21

Roman Jakobson's Contribution to Slavic Accentology Henrik Birnbaum

29

The Influence of Roman Jakobson on the Development of Semiotics Umberto Eco

39

Roman Jakobson and the Comparative Study of Parallelism James J. Fox

59

JlHHrBHCTHHeCKafl TeOpHH P . O . ^KOÔCOHa B paÔOTaX COBeTCKHX JIHHrBHCTOB

;

91

T. B. raMKpejiud3e, T. H. Enmapemoea, B. B. Heanoe Roman Jakobson's Contribution to the Modern Study of Speech Sounds Morris Halle Jakobson's Contribution to Phenomenology Elmar Holenstein Bmiafl P. O.

flieoôcoHa

123

145

B cjiaBHHCKHe H HHAoeBponeiicKHe (J)ojibK-

j i o p H b i e H MH(J>ojiorHHecKHe HccjieflOBaHHH

B. B. Heanoe u B. H. Tonopoe

163

viii

TABLE OF CONTENTS

The Linguistic Model in Biology François Jacob

185

Jakobson and the Postwar Slavists Lawrence Gaylord Jones

193

PyccKaa

flnajieKTHaa

h

ncTOpHHecKaa

4)OHeTHKa

B

paôoTax

P. O. JlKo6coHa.

201

JI. JI. KacantKUH La Place de Roman Jakobson dans la vie littéraire et artistique tschécoslovaque Vëra Linhartovâ

219

The Contribution of Linguistics to the Theory of Aphasia A. R. Luria

237

Roman Jakobson and (Old) Church Slavonic Studies F. V. Mares

253

Comparative Slavic Poetics in the Work of Roman Jakobson Maria R. Mayenowa

259

Jakobsonian Ideas in Generative Grammar James D. McCawley

269

3 OeoôeHHOCTH, 7 npHHijHnoB H 11 pe3yjibTaTOB rpaMMaTmecKHX HccjieflOBaHHii PoMaHa ^Ko6coHa H. H. MeAbuyK

285

Phonologie théorique et phonétique ¿arko Muljacié

309

Roman Jakobson on Russian Epics and Old Russian Literature.... Riccardo Picchio

321

Jakobson's Contributions to American Linguistics Harvey Pitkin

357

Roman Jakobson and the New Poetics Krystyna Pomorska

363

Roman Jakobson and the Prague School Miroslav Rensky

379

Distinctive Feature Theory R. H. Robins

391

TABLE OF CONTENTS

ix

Child Language Studies Velta Rùke-Draviria

403

Roman Jakobson's Teaching in America Thomas A. Sebeok

411

Roman Jakobson's Contribution to the Study of Slavic Historical Phonology and Phonetics Sebastian Shaumyan

421

Roman Jakobson's Work on the History of Linguistics Edward Stankiewicz

435

Roman Jakobson and Old Czech Literature Frantisek Svejkovsky

453

Poétique générale Tzvetan Todorov

473

Roman Jakobson and Dutch Linguistics E. M. Uhienbeck

485

Roman Jakobson and Avantgarde Art Thomas G. Winner

503

Roman Jakobson and the Study of Rhyme Dean S. Worth

515

BY WAY OF INTRODUCTION: ROMAN JAKOBSON'S TENETS A N D THEIR POTENTIAL C. H. VAN SCHOONEVELD A volume describing and evaluating the work of the polyhistor Jakobson hardly needs a summarizing introduction, nor does it require a recapitulation of the eulogies which are evident from the testimonies about his work. Our gratitude for what Roman Jakobson means to so many people as a human being and as a scholar, ex aequo, and the warmth of the wishes we bear him could hardly find an appropriate signans anyway. What I would rather do is single out a few of the theoretical principles which lie at the base of his inquiries and try to argue how they, in my opinion, provide us the key for a further exploration into the predictability of linguistic phenomena. I use the word "linguistic" intentionally. Even though Jakobson's initial impact was in the field of literature and even though that impact has, during his life, tended to grow, literary phenomena have been to him basically phenomena resulting from the artistic manipulation of the materials of language. Let us concentrate our attention, therefore, upon what transpires in Jakobson's work as essential properties of language. To be sure, Jakobson will say that not all these principles, such as the semiotic nature of language and linguistic universals, were first discerned by him and will acknowledge priority for his predecessors. Yet it is with Jakobson that these principles combined into a consistent and coherent structure of which the investigative power has only partly emerged. Since I am interested in using these tenets as a synchronic point of departure for a look into the future, I will not try to trace their provenience. Rather I will attempt to show the potency of Jakobson's conceptualizations, and especially what far-reaching consequences they have if used as an integrated whole. I should like to lift one crucial concept out of Jakobson's methodological arsenal, discuss briefly his skillful handling of this concept, and pursue the consequences toward which he leads us. I have in mind the concept of the invariant in language, especially as it bears on some points of interrelatedness between linguistic invariants and linguistic universals.1

2

C. H. VAN SCHOONEVELD

The invariants in phonology are well known. Within a given language, there will be in the given speech act stylistic or positional variation. Nevertheless, from parole to parole and from environment to environment there will be a core of invariant properties which are perceptionally distinctive enough to mark off the given materialization of a phoneme as a variable of that phoneme and not of another phoneme. Another phoneme will again have its own invariants. Each phoneme provides an abstract parameter of invariants against which each individual materialization of it in each semelfactive parole is identified in each (syntagmatic) environment in which it occurs. Thus, the first layer of conditioning influences on the materialization of a phoneme in a given parole, those due to stylistic and emotional influences and those due to the (syntagmatic) environment, is discounted in order to achieve the first level of generalization, the phoneme. This is the level of invariance as it serves to demarcate and identify units within a specific language. In this manner, intralinguistic invariants, such as distinctive features, phonemes, or intralinguistic semantic features, are established. Conversely, this generalization gives us a rule which may or may not have a wider validity, but in any case provides us with a certain predictability, at least within the given language, of the actual materialization in a given environment. After we discount predictabilities due to environment, that is, relations between phonemes on the axis of contiguity, there remain the relations between phonemes on the axis of replacement. These relations, in turn, influence the latitude of the parameter. Variations in the realization of a vowel phoneme in a three-vowel system will be larger than that of a vowel phoneme belonging to a five-vowel system. Arabic pharyngealization is but an interlinguistic variant of rounding, the invariant being flatness; Hottentot clicks belong with glottal stops in other languages, to one single interlinguistic invariant: checkedness. We must, then, make a clear distinction between two levels of abstraction and generalization. The first level of abstractionis the level of paradigmatic relations between the units constituting a specific linguistic system. Their symbiosis in one language makes the entities resulting from the first abstraction and paradigmatically opposed to one another likewise condition each other. The relations on the paradigmatic axis existing in the given linguistic system are a second conditioning influence. Discounting this second category of intralinguistic conditioning influences leads us in turn to a higher, second level of abstraction, that of interlinguistic invariants, and conversely, to a higher level of predictability. Extracting the invariant between similar phonemes from language to

JAKOBSON'S TENETS AND THEIR POTENTIAL

3

language gives what de Saussure has called "panchronic" invariants, and, as Jakobson uses the term, linguistic universals and near-universals. In order to reach such universals, then, we have to go through two steps. By eliminating syntagmatic and stylistic conditioning we establish the invariant within the given language, that is, the intralinguistic invariant. At the same time, by describing the syntagmatic and stylistic conditions, we have formulated rules which make predictable the range of realization of the phoneme in the given environment. Only the next abstraction gives us the generalization which, in Saussure's words, is represented by the "lois dans le sens où l'entendent les sciences physiques et naturelles, c'est-à-dire des rapports qui se vérifient partout et toujours. ... En un mot, la langue ... étudiée au point de vue panchronique. ..." 2 This step is taken by formulating the rules which make predictable the invariant as conditioned by the systems of various languages. What remains after discounting that which is conditioned by the individual systems will constitute the universal, that is, interlinguistic invariant. Whether the rules that formulate the predictabilities of the manifestation of a given phoneme as conditioned paradigmatically and syntagmatically within a given language are wholly universal themselves or partly arbitrary from language to language is another matter. In semantics, the former option seems largely to hold. In any case, the principal difference between the phenomena of language, or, for that matter, other social phenomena, and the subject matter of the sciences seems to be the fact that in the former universal predictability is varied by conditioning by the given system, which fact thus creates invariants that have only validity within that system. It is one of the great merits of the Prague School and of Jakobson in particular to have conceptualized this two-layeredness as a matter of principle. By showing the methodological path through the two strata, the intralinguistic stratum and the stratum of the interlinguistic invariant, Jakobson has shown the way toward a linguistics that as a science has a degree of exactness comparable to that of the traditional sciences. In morphology, in the semantic interrelations between the signifieds of grammatical, word-formative, and lexical morphemes, the situation is essentially no different, although we have to do with structural relations of considerably greater complexity. Jakobson analyzed the Russian verbal categories and the Russian case system and, by discounting syntagmatic conditioning, established within the Russian language the paradigmatic semantic (intralinguistic) invariants of the grammatical categories of the verb and of the case system. The latter, consisting of eight cases, has three semantic invariants, to wit, directionality, marginality, and quantification.

4

C. H. VAN SCHOONEVELD

If we again follow Jakobson's lead, and this time apply his search for invariants intercategorically, we find his methodology vindicated again. Let us assume that such categories as substantive and preposition act as molds into which are cast such invariants as those that constitute the case system. We will ascertain that prepositions can be analyzed semantically with the help of semantic invariants again, and, moreover, that - mutatis mutandis - the three invariants that Jakobson established for the case system reoccur as three of the at least five semantic invariants of the prepositional system. In hierarchical order of complexity, the first two invariant semantic components of Russian prepositional structure are dimensionality, related to the concept of boundary in topology, and duplication, comparable to the relation between two disjunctive sets. The next three semantic features, however, extension, restrictedness, and objectiveness, are, mutatis mutandis, identical with the directionality, marginality, and quantification of Jakobson's case system. To use the terms of topology again, they are comparable, respectively, to the conjunction of two sets in a neighborhood; to the difference between such a conjunction and the first set; and to the conjunction between the complement of two sets figuring in the preceding formulations and these sets (at this point there is no longer a neighborhood). Accordingly, semantic structure, however primitively I have formulated it, appears to consist of topological relations sui generis. Thus both directionality (as represented, for instance, by the accusative) and extension (as represented by such prepositions as po, o, and iz) have as their common invariant a link of identity between the initial narrated situation and an ulterior situation, while marginality (cf. the Russian instrumental) and such prepositions or preverbs as vy~, ot, and iz have in common the cancellation of the initial narrated situation and the break between this situation and the ensuing situation. Quantification (as represented by the Russian genitives and locatives) and objectiveness (the latter represented by prepositions like pri and k) have in common the fact that their referents are related to the narrated situation independently of any neighborhood; hence, in quantification the question is posed, as Jakobson formulates it, of the "predely ucastija oznacennogo predmeta v soderzanii vyskazyvanija [the limits of the participation of the referent in the narrated situation]." A similar identity of invariants, mutatis mutandis, can be established for the grammatical categories of the verb,3 for the lexicon and for word formation. 4 Thus, in the semantic structure of the verbal lexicon, in the oppositions byt' 'to be' vs. minovaf 'to pass' and derzaf 'to hold' vs. puskat' 'to let' the second members present the cancellation ("restrictedness") type, whereas an extensional type occurs in the so-called "determi-

JAKOBSON'S TENETS AND THEIR POTENTIAL

5

nate" verbs, which are goal-directed just like the accusative. One of the varieties of lexical objectiveness occurs in the signatum of the verbal lexical morpheme budu [praesens tantum, usually translated as 'I will' or 'I will be'] (which is different from the lexical morpheme byV [verbum infinitum tantum] 'to be'). The signatum of budu is a situation which has no direct ties to the transmission of the utterance (the "speech situation"), and indirect ties to it not as a process, but only in that it has a subject. The signatum of the present is a narrated situation which is not marked by perceptional independence from the speech situation. The preterite marks a prohibition against synchronizing the perception of the narrated event with the perception of the speech event. It evidently represents a (grammatical) cancellation type, and is marked for restrictedness as a grammatical verbal category. Thus it is linked to the speech situation by a relationship of indirectness. The process represented by the signatum of budu belongs, on the other hand, to the semantic code rather than to the hie et nunc or ibi et tunc : its referent is merely potential. The only tie budu has with the speech situation is through its subject. A similar type occurs in the (transmissionally) deictic adverb ne 'not', or, for that matter, in negation in general. The referent of the unit modified by it is excluded from the parole. Most of the signata, lexical as well as grammatical, discussed so far are constituted by invariants relating the referent to the speech act. The perceptibility of the referent is expressed in terms of the only invariant common to all speech, the perceptibility of the speech act itself. This "absolute" invariant, the fact that the transmission of the utterance is perceived, is at the basis of all types of meaning traditionally called deictic. Russian verbal lexical structure seems to have six semantic invariant features of this type. Apparently, however, it has six more, based upon another "absolute" invariant, this one, however, not deictic in the traditional sense, but in another sense. These six lexical verbal features relate the perceptibility of the referent to the very act of the referent's being perceived. The only invariant common to all categories of meaning - meaning being essentially a system of cues for perceiving the referent - is the act of perception itself. Thus, one has to distinguish relatedness of the referent to the speech act (transmissional deixis) from relatedness of the referent to the act of perception of the referent (perceptional deixis). In the oppositions byf 'to be' vs. minovaf 'to pass' and derzat' 'to hold' vs. puskaf 'to let', the second members are marked by perceptional restrictedness. We find perceptional extension in stupaV 'to step' as opposed to xodW 'to go' (unmarked). Extension means a link between the initial narrated situation and the ensuing situation. The referent changes minimally. Obviously, this is

6

C. H. VAN SCHOONEVELD

the case in stupat'. The difference between the ensuing situation of stupat' (marked for perceptional extension vs. xodit') and idti 'to go' (marked for transmissional extension vs. xodit') is that in idti the ensuing situation reaches into the speech situation. Objectiveness means that the perceptibility of the referent is potentially maximally distant from the act of perception. Whereas budu (vs. byf ) represents the transmissional type of lexical objectiveness, begat' vs. xodit' is an example of lexical perceptional objectiveness. In begat', which is marked for objectiveness, the result of the cancellation of the initial situation is the potential maximal distance of the subject when one compares the ensuing situation with the initial situation. The same opposition reoccurs in kricaf 'to shout' vs. govorif 'to speak' and tvorif 'to create' vs. delaf 'to make': the objects of the first members of these oppositions are potentially maximally distant from the initial situation. (Incidentally, it follows from the preceding presentation that all the verbs of locational motion which I discussed are marked for lexical perceptional restrictedness.) There is one type of verbal lexical semantic feature, like the others occurring both as a perceptional and a transmissional (in)variant, which I have not yet mentioned. It seems, in fact, to be the first one in the hierarchy of the semantic features. It is transitivity. Transitivity can be formulated in purely semantic terms. It indicates that the agens, by performing the process, leaves independent evidence of her or his having done so. The process has results ascertainable outside the performance of the process itself. The lexical perceptional type seems to be represented largely by what are traditionally called transitive verbs vs. the (unmarked) intransitive verbs (cf. Russian byt' 'to be' vs. bavit' 'to cause to be'); the lexical transmissional type is represented, among others, by verbs of creation such as delaf 'to make', emit* 'to perform' (also marked by lexical perceptional extension), tvorW 'to create' (additionally marked by lexical perceptional objectiveness). Thus, by consistently applying Jakobson's methodology of discounting syntagmatic conditioning and extracting semantic invariants within various categorical molds in a given language, such as parts of speech and lexicon vs. grammar, we can proceed to identify paradigmatic intercategorial intralinguistic invariants. These invariants are grouped on various levels of deixis, such as perceptional and transmissional deixis. The result is a paradigmatic structure of great simplicity, yet made complex by its repetitiveness. I shall pass by here the questions of how the hierarchy of Russian semantic features is evidently a true hierarchy in the sense that these features

JAKOBSON'S TENETS AND THEIR POTENTIAL

7

appear to be ordered, and of the resulting problematics of extracting interlinguistic invariants, that is, semantic universals.5 But there is one more area toward which Jakobson's paradigmatic approach to invariants leads us. Discounting the syntagmatic context in order to achieve ever higher levels of abstraction along the axis of similarity brings us ultimately to the interrelated domains of syntax and of the predictability of grammatical categories. Syntagmatic correlation, which in turn apparently conditions the existence of the semantic categories (which are in paradigmatic oppositions to each other) engaged in syntactic relations, is achieved by deixis at various levels. Let us take as an example the trivial fact that, as in so many languages, in Russian transitive verbs have as their primary constituent an accusative object. Transitivity in the traditional sense is in Russian a lexical feature of the perceptional type. It says that the subject, by performing the process, leaves a piece of independently perceivable evidence of the performance of the given process. Thus a transitive verb refers qualitate qua to a piece of evidence, that is, an (inherent) object. But as soon as a transitive verb is used in a given parole, its signatum has semelfactively a subset in the actual denotatum of the verb. In other words, this denotatum is (transmissionally) deictically given. Any element of parole implies identification. However, the language provides the possibility of verbally identifying the object even further by means of another word in the sentence which stands in a syntactic relation to the transitive verb. This word is the accusative object. The referent of the accusative object must therefore be identical with the denotatum of the object referred to by means of the lexical category of transitivity marking the given verb. The accusative object, being the primary constituent modifying the verb, must therefore refer back to that verbal lexical object which has already been identified. Perceptionally deictic transitivity has yielded an object identified, naturally, in the parole. Through application in the parole, the inherent object has become semelfactively transmissionally deictic. Extension gives a situation connected with and extending beyond an initial situation which it presupposes. Thus, extension applied on a grammatical, that is, a transmissionally deictic level, as in the accusative, elaborates on a resulting situation which is of relevance in the speech situation and which emanates in turn from a preceding situation that is similarly identified in and of relevance to the speech situation. Thus, the accusative, which is marked by the one single feature of grammatically transmissionally deictic extension, preempts the referent of the verbal lexical object as having been transmissionally deictically given. The accusative, in other words, repeats the inherent verbal lexi-

8

C. H. VAN SCHOONEVELD

cal object, and informs us further, by means of the lexical meaning of the direct object, about the inherent verbal lexical object that is identified but not further specified. To summarize the way in which the lexical and grammatical semantic features interlock syntactically: since the extension feature, which marks the accusative, is a grammatical, that is, a transmissionally deictic feature, it indicates that the first situation of its referent is already being identified in the utterance, and that consequently the second situation can only refer to something already mentioned in the utterance. Thus, the accusative by means of transmissionally deictic extension refers back to the inherent object of the transitive verb, identifies it again and specifies it (by means of the lexical meaning of the substantive involved) with regard to the speech situation. The first semantic feature of the hierarchy of verbal lexical semantic features, transitivity, poses the problem of the perceptibility of the agens as a separate property. It signalizes a percipiend which is separate from the agens but serves as evidence of the agens. When the verbal lexical morpheme is used in the parole, the lexical transitivity feature makes a onceoccurring reference to this piece of evidence for the subject but does not identify it. The identification is tacitly assumed by virtue of the working of the parole. The nature of the piece of evidence thus referred to is circumscribed by the lexical meaning of the direct object. The essential characteristic of directionality, the only case feature marking the accusative, is, in Jakobson's apt phrasing, its focusing power. Moreover, the extension feature, of which the directionality feature is but the case-type variant, refers to a framework of properties which it assumes deictically to be identified. All the accusative does is to bring into focus the (lexical) specification of the evidence for the agens of the verbal process. Evidently, it is the most fundamental modification possible of a verb by a substantive. Accordingly, extension is the first semantic feature in the Russian case system. Although the presentation just given is very impressionistic and although the explanatory argumentation especially is in dire need of greater precision, a pattern of interplay between semantic categories and syntax nevertheless begins to emerge. The underlying principle is that syntax by its very nature always involves parole. Syntax implies semantic interlacing, in other words, the establishing of, at least, ephemeral identities. To be sure, depending on the situational context, these identities may have a longer lasting validity, but they are always established and originally validated by a semelfactive sentence, that is, by a (individual, ephemeral) parole. Syntax cannot exist without a given transmission of an utterance. Rules of syntax must involve transmissional

JAKOBSON'S TENETS AND THEIR POTENTIAL

9

deixis, since the syntagmatic axis, as de Saussure says, evolves in praesentia: "il repose sur deux ou plusieurs termes également présents dans une série effective. Au contraire le rapport associatif [Saussure's term for paradigmatic] unit des termes in absentia ... il faut reconnaître que dans le domaine du syntagme il n'y a pas de limite tranchée entre le fait de langue, marque de l'usage collectif, et le fait de parole, qui dépend de la liberté individuelle ... il est difficile de classer une combinaison d'unités, parce que l'un et l'autre facteurs ont concourru à la produire...." 6 Saussure subsumes the problem in the question: "ne s'ensuit-il pas que le syntagme relève de la parole?," 7 and he answers this question negatively. It should be answered positively at the same time. Syntax is codified parole. It operates by means of patterned transmissional deixis, whereas the semantic features themselves may be based on perceptional deixis alone. One can characterize the Russian preposition as a transmissionally deictic adverb. It modifies a referent which corresponds semantically to the categorical meaning of the substantive as a part of speech, while this referent is involved in a verbal process. In other words, on'» could say that the preposition is an adverb with a built-in substantivic reference. Like the lexical feature of transitivity it is involved in the existence of an independent substantivic unit with the difference that transitivity establishes such a unit while the preposition anticipates the existence of such a unit. The preposition assumes beforehand the existence of a substantivic unit. Accordingly, the hierarchy of Russian prepositional features does not utilize the transitivity feature but starts with the next feature, to wit, dimensionality. Its most fundamental opposition is na 'on' vs. v 'in', the latter member being marked for dimensionality. Similarly, the category of conjunction can be characterized as a type of transmissionally deictic adverb, but this type, unlike the preposition, modifies a predication instead of a substantival signified. A predication, however, is essentially an intersection between a substantival signified and a property that is assigned to it in the given utterance. Thus, the substantive is, in the given predication, uniquely characterized, that is, semelfactively separated from all its peers. Separatedness of the unit from its peers is precisely the semantic content of the dimensionality feature. This semantic content is, incidentally, none other than the semantic marking, mutatis mutandis, of the category of phoneme as such: "mere otherness." Since the category of conjunction as a whole anticipates that this characterization has taken place in the first sentence, which the conjunction is to modify, the first marking within the structure of Russian conjunctions starts with the next feature of the hierarchy, to wit, duplication. While the conjunction i 'and' is unmarked, the conjunc-

10

C. H. VAN SCHOONEVELD

tion a (a weak 'but') is marked by duplication, the conjunction da (a strong 'and') is marked by extension, and the conjunction no (a strong 'but') is marked by both. On the other hand, the category of part of speech creates potential reference. Without being a part of speech, that is, without being a substantive, verb, adjective, or adverb, a word cannot refer to reality. Evidently, only on the level of word is a semantic unit created that can operate independently. The system of parts of speech of Russian makes use exactly of the semantic features in the hierarchy that for practical purposes carry no anticipation of another semantic element co-occurring in the same syntagmatic context: dimensionality and objectiveness. While the adverb is unmarked, the verb is marked for dimensionality and the adjective for objectiveness; the substantive, on the other hand, is marked for both. Jakobson's intuition has directed his quest for invariants primarily along the paradigmatic axis. As I have tried to argue with the help of a few examples, the semantic structure of the Russian language, and, presumably, of other languages as well, mutatis mutandis, apparently consists of a small number of paradigmatic semantic features which repeat themselves a number of times on various levels of deixis. Syntactic coordinations, while based on several varieties of deixis to begin with, are made through coordination of deictic levels. Among these transmissional deixis necessarily plays a rôle. The syntagmatic axis is a more complex phenomenon than the paradigmatic axis ; it is the marked member of the opposition. When one reads Saussure's historical reconstructions, especially the Mémoire, one is struck by the radical instinctiveness of Saussure's pursuit of diachronic invariants. Later, however, when, in the Cours, Saussure shifts his attention to language as a synchronic system, he thinks more in terms of the system as it is built up of oppositions, rather than of the individual invariant units which constitute it. He posits language as a synchronic system, but the task of analyzing languages in these terms overwhelms him. 8 With the Prague School the systematic investigation of the individual synchronic structures of languages begins. Jakobson's entire work, from his literary beginnings on, is dominated by the search for what is invariable, and, as a corollary, variable, in semiotic systems. His vision has opened the way toward the invariance of invariants. Indiana University

JAKOBSON'S TENETS AND THEIR POTENTIAL

11

NOTES 1 Earlier versions of this paper were presented in lectures at the University of Texas and at Cornell University. 2 F. de Saussure, Cours de linguistique générale (im° édition, Paris, 1955), p. 134. 3 Cf. C. H. van Schooneveld, "Contribution a l'étude comparative des systèmes des cas, des prépositions et des catégories grammaticales du verbe en russe moderne," Slavica hierosolymitana I, eds. V. Raskin and D. Segal, Linguistic series I (Jerusalem 1977). 4 For the latter, see James Daniel Armstrong, A Semantic Approach to Russian Word Formation (Indiana University Ph.D. dissertation, 1973; publication forthcoming). 5 Cf. C. H. van Schooneveld, Semantic Transmutations : Prolegomena to a Calculus of Meaning, Vol. I: The Cardinal Semantic Structure of Prepositions, Cases, and Paratactic Conjunctions in Contemporary Standard Russian (Bloomington, Ind., 1977), especially Chapters VII and I. • F. de Saussure, Cours de linguistique générale, p. 173. ' Ibid., p. 172. 8 E. Benveniste, "Saussure après un demisiècle," Cahiers Ferdinand de Saussure 20 (1963), reprinted in Problèmes de linguistique générale (Paris, 1966), pp. 37ff.

THE STUDY OF PALEOSIBERIAN LANGUAGES ROBERT AUSTERLITZ

It is only a mild distortion of the truth to say that the role of the Siberian peoples and languages in Imperial Russia was, among those intellectuals who were at all interested in them, analogous to the role of the American Indians among the corresponding group in the United States during the first few decades of this century. Sadly enough, interest among linguists in these preliterate peoples and cultures was minimal. But just as this country produced a Boas and a Sapir, so in pre-revolutionary Russia some of the best minds were channeled into work on Paleosiberian1 and on other languages of the isolated remotes of Siberia. The political situation before 1917 also contributed to this in that some of the workers in this field were exiles (Bogoras-Tan, Sternberg) who had not been ethnographers or linguists to begin with. Among the very young linguists (before 1917), Trubetzkoy and Jakobson, particularly, were attracted to Paleosiberian, presumably - among other reasons - because of the structure of these languages which diverged dramatically from that of the languages studied traditionally and because of the enigmas which the genetic affiliations of these languages posed. Here, then, was a young Roman Jakobson, maturing into the Slavist and general linguist whom we now know, who must have been intrigued by the culture and languages of these aborigines and who had the benefit, already as a young man, of exposure to data and of contact with stimulating personalities. The revolution of 1917 changed the picture. Particularly during the early phases of the existence of the Soviet Union, there was a strong, sincere, and idealistic surge of interest in the autochthonous peoples of all parts of the U.S.S.R. and a parallel surge in concentrated work on Siberian languages which culminated in the publication under the editorship of Al'kor (see KrejnoviS, ed., Jazyki i pis'mennosf paleoaziatskix narodov). By this time Jakobson was in Brno and Trubetzkoy was compiling his vast repository of data on phonological systems. Jakobson's early interest (which I take for granted) must have been rekindled in the 1930's as a

14

ROBERT AUSTERLITZ

result of Krejnoviô's publication. This applies especially to Gilyak; see below. But another lasting concern of Jakobson's, the nature and distribution of areal features over large portions of continents, must have channeled his attention in the direction of Paleosiberian before 1934, as [1] K xarakteristike evrazijskogo jazykovogo sojuza (Paris, 1931) [reprinted in Selected Writings I, 146-201] indicates. The specific point of interest was the palatalizing belt in Northern Eurasia; among the examples cited there are some from Chukchi and Yukagir. These languages reappear in [2] "Uber die phonologischen Sprachbunde," Travaux du Cercle Linguistique de Prague 4 (1931), 234-40 [reprinted in Selected Writings I, 137-43]. Three other works on the question of areal affinities, all from 1931 ("Les Unions phonologiques," "O fonologiéeskix jazykovyx sojuzax," and "Sur la théorie des affinités phonologiques" - see the appended bibliography) contain no examples from Paleosiberian. This is significant because in Jakobson's next major work [3] "Kindersprache, Aphasie und allgemeine Lautgesetze" (Uppsala), 83 pp. Sprâkvetenskapliga Sâllskapets i Uppsala Fôrhandlingar, 19401942 ( = Uppsala Universitets Ârsskrift, 1942, 9) [Also republished in Frankfurt, 1969. Reprinted in Selected Writings I, 328-401. English translation, 1968] data from Ket, Kamchadal, and Gilyak are adduced in connection with [r], [j], and the nasals. Kindersprache must have been a number of years in the making, years which coincided with the trajectory Brno-CopenhagenOslo-Stockholm, years which encompass 1934, when the work by Krejnoviô was published, and years during which "continental" Prague theory (as distinct from the subsequent "Cambridge, Mass." Prague theory) matured. Jakobson's concern with system, a basic component of Prague theory, had already led him to adduce Gilyak examples (from Krejnovic) in connection with Czech in: [4] "Beitrag zur allgemeinen Kasuslehre. Gesamtbedeutungen der russischen Kasus," Travaux du Cercle Linguistique de Prague 6 (1936), 240-88 [reprinted in Selected Writings II, 23-71],

THE STUDY OF PALEOSIBERIAN LANGUAGES

15

(Another Czech/Paleosiberian parallel, in connection with ethnonyms, appears in "Die Reimworter Cech-Lech.") During his Stockholm period, Jakobson, John Lotz, and Wolfgang Steinitz held weekly three-man seminars during which Jakobson often discussed Paleosiberian, presumably in connection with the preparation of [9] below. This and Jakobson's entire European period ends in 1941, when he arrives in the United States. He must have met Franz Boas soon after landing, and I surmise that it is thanks to Boas' stimulation that [5]

"The Paleosiberian Languages," American Anthropologist 44 (1942), 602-22,

appeared so soon after his arrival. This was Jakobson's first paper entirely devoted to Paleosiberian. To appreciate it fully, one should be aware of the sparse and theoretically unsophisticated sources on which it is based: the paper is a pithy exploitation of these sources, with highly suggestive ideas about six divergent linguistic structures and about areal facts from the Yenisei River to Sakhalin. I have never understood why it did not evoke more of a response from the North American corps of specialists on Amerindian languages who are concerned with an even wider typological gamut and a much larger area. Eskimo-Aleut is only one step away from Luorawetlan. Jakobson's early New York period thus saw the appearance of: [6]

"A Note on Aleut Speech Sounds" and "A List of Works Relating to the Aleut Language," Bulletin of the New York Public Library 48 (1944), 677 and 678-80. Appended to A. Yarmolinsky, "Aleutian Manuscript Collection," ibid.

The first of these, a modest note, is an imaginative reconstruction of the vocalism and consonantism of Aleut from fairly unreliable sources and must be considered the precursor (and perhaps the original stimulus) to Bergsland's 1959 work, Aleut Dialects of Atka and Attu. The other product of this period, [7]

"A List of Works Relating to the Kamchadal Language and to the Language of Russianized Kamchadals," Bulletin of the New York Public Library 51 (1947), 667-69. Appended to A. Yarmolinsky, "Kamchadal and Asiatic Eskimo Manuscript Collections; A Recent Accession," ibid., is undoubtedly the catalyst which launched D. S. Worth's interest in Luo-

16

ROBERT AUSTERLITZ

rawetlan and his extensive work on Kamchadal (1961 and 1969). I was present when [8] "The Structure of Gilyak and Marr-ism," mimeographed outline of a paper delivered on September 8, 1949, at the 29th International Congress of Americanists, New York, 5 + 2 pages, was given at the Museum of Natural History and feel that my own interest in Gilyak dates from that moment. (By this time the original enthusiastic period of Soviet interest and work on Paleosiberian had yielded to Marrism, which, in turn, had to wait for the linguist I. V. Stalin to be liquidated. - For the connection between Marr's theories and a hypothetical form of proto-Gilyak, see Jakobson's "Les Lois phoniques du langage enfantin..." of 1949 [an expansion of "Le Développement phonologique..." of 1939] and "Typological Studies . . . " , 1958.) The sketch cited under [5] was expanded into [9] "Langues paléosibériennes" in A. Meillet et Marcel Cohen, eds., Les Langues du monde, par un groupe de linguistes... Nouvelle édition (Paris: Centre National de la Recherche Scientifique, 1952), pp. 27678 and 403-31. The discussion of phonology, grammar, and genetic affiliation is almost the same in [9] as in [5]. What gives [9] particular significance is an appended Gilyak text of some 30 words which is fully analyzed in 36 comments (pp. 424-28) in which each morpheme is fully acounted for. Many of these comments are daring sorties into internal reconstruction and as such extraordinarily suggestive for the Gilyakologist and highly educational to the general linguist and to the student of Jakobson's way of thinking. Let us remember that in this case a Slavist who has a confoundingly large body of data to draw from in his own field confronts an isolate with no data for comparison at all. Furthermore, insights from diachrony are here exploited for the purpose of a synchronic description: this still goes counter to good tradition in linguistics and is permissible only when the two sets of data are kept continuously and meticulously compartmentalized. Jakobson succeeds in doing this. (A quick check reveals that no other sample text in this, second, edition of Les Langues du monde is so carefully and rewardingly processed.) This analysis of Gilyak is also the first in a language other than Russian or Japanese. Similarly, the best synoptic introduction to the Paleosiberian field in a language other than Russian or Japanese can be found on pages 218-22 in:

THE STUDY OF PALEOS1BERIAN LANGUAGES

17

[10] Paleosiberian Peoples and Languages. A Bibliographical Guide with Gerta Huttl-Worth and John Fred Beebe (New Haven: Human Relations Area Files Press, 1957), viii + 222 pages. ("A Short Sketch of the Paleosiberian Peoples and Languages," by R. Jakobson, pp. 218-22).

This is a truly annotated bibliography of immense value and has been accorded the encomium which it deserves; see my 1959 review. (It covers archival and other unpublished material as well as works and maps by native Paleosiberians.) Already in 1949 (see [8]) Jakobson had promised a larger work on Gilyak. It appeared as [11] "Notes on Gilyak," Studies Presented to Yuen Ren Chao on His Sixty-Fifth Birthday ( = Academia Sinica, Bulletin of the Institute of History and Philology 29, Part 1) (Taipei, 1957), pp. 255-81 [reprinted in Selected Writings II, 72-79] and deserves a detailed report because it reveals something of Jakobson's manner of thinking along holistic-grammatical lines and his ability to distil important consequences from seemingly insignificant detail. A brief introductory passage covers sources, deals with the history of Gilyak studies in China, Japan, Imperial Russia, and the Soviet Union, and contains an incidental comment (pages 259-60) on the effect of poor fieldwork on grammatical analysis and interpretation. The "Grammatical Essentials," pages 261-63, are a microtheory of general grammar, snugly tailored to the particularities of Gilyak: the sentence is an intonational unit followed by an obligatory pause; the section is a stress unit followed by an optional pause. There follows a hierarchical taxonomy of morpheme-types and morpheme-combinations by function. "Phonemic Essentials," pages 262-68, is built around distinctive features, based on a very close reading of recent Soviet literature on Gilyak. (This is an early application of distinctive features, considering that Preliminaries had only appeared in 1952 and the papers on the French and Serbo-Croatian distinctive features in 1949. Note how the opposition strident/mellow is exploited here: the postvelar stop and fricative q and x are strident while the velars k and x are mellow. This permits a quadripartite deployment of the places of articulation, as one would expect in a Jakobsonian analysis: p, t, c [¿], and k\q.) The latter half of the paper is, in my opinion, the best thing ever written in this field: by concentrating on the most sensitive and hence symptomatic zone in the hierarchy of language, morphophonology, and by systematically

18

ROBERT AUSTERLITZ

exhausting the significance of the synchronic facts, Jakobson manages to reveal the essentials of Proto-Gilyak. The main process in question is lenition ( k : x , g : y - where k is voiceless-aspirated and g unaspirated) and the circumstances which condition lenition. One of the less transparent conditions of this sort is the opposition transitive/intransitive which, in the modern language, correlates with fricative/stop: fa- 'he roasts something' vs. ta- 'it is roasting'. Jakobson traces the fact that all transitives begin with a fricative to the earlier presence of the deictic element (demonstrative or third-person singular pronoun) **'-. (I would go further and say raOHCM" (J6 -

¡d - m¡, TO ECRB (JIOHCM, OTJIHHAIOMHXCH

NO 0FLH0MY

fl.n.),

n¡,

6M:IO CAEJIAHO

EME B KOHIXE 30-X ROFLOB.38 IIO3FLHEE AHAJIORHHHWH BMBOA 6MJI PA3BHT HA KJIHHHIECKOM MATEPNAJIE, ORAOCAMEMCH K TPABMATHIECKHM NOPAACEHHHM BEPXHE-3AFLHHX NOJIYUIAPHH.

39

OTACJIOB BHCOHHOH 3HAHHM0CTB

njxtñ P .

OÓIIACTH A O M H H a H T H o r o ÜKO6COHA

(jieBoro)

asín HCCJIEFLOBAHH« AA3HH

6BIJIA NOFLIEPKHYTA B NOCJIEFLHEM CNEIMAJIBHOM TPY^E E . H . BHHAPCKOIÍ NO A < | ) a 3 H 0 j i 0 r H H . 4 0 BjiHAHHe H f l e ñ P o M a H a O c n n o B H H a ^ Í K o 6 c o H a H a c o B e T C K y i o 4>OHOJiornio

yflHBHTeJIbHO HIHpOKO

H njIOflOTBOpHO. ü e H a T H b i e

Tpyflbl

He

B n o j i H e o T p a a c a i o T i n n p H H y H r j i y Ó H H y e r o BJIHSHHH. IIOMHMO " G o j i b u i o ñ " Hayraoií

neHara-KHHr

MHOrOHHCJieHHblX

H cTaTeñ - MHoroe

KOH(J)epeHnHH,

B

MONCHO H a i r r a

flHCKyCCHHX

p a Ó O T e KOMHCCHH H CeMHHapOB ( H a n p H M e p , B HecKoii

KOMHCCHH,

paGoTaBineií

B

Ha

fleflTeJIbHOCTH

HHCTHTyre

B

Te3Hcax

KOH(J)epeHUHHX,

B

(J)0H0J10rH-

BOCTOICOBEAEHH»

AH

C C C P B T e n e H H e 1 9 6 7 H 1 9 6 8 RR., H a 3 a c e a a H H H X KOTOPOÍÍ PA36NPAJIHCB B o n p o c w (JPOHOJIORHHECKOIÍ T e o p H H H M e T O f l a H a M a T e p n a j i e BOCTOHHMX H3WKOB). CTOpOHHHKH MOCKOBCKOH (J)OHOJIOrHHeCKOH IHKOJIbl, HbH npaKTHHeCKHe flOCTHMCeHHH B o 6 j i a C T K 0H0JT0rHiecKHX

fl.n.,

flHXOTOMHHeCKOH

o 6 c y » c a a j i H 3Ty T e o p H i o n p e a c ^ e B c e r o

no3HQHH Mop4>ojiorHHecKoro onpeflejieHHH ^OHeMbj, x a p a K T e p H o r o

c

ana.

HX niKOJibi H CB$rcaHHoro C c a M O H c r p y K T y p o H H3HKOB, e í í H 3 y i a B f f l H x c H . 4 1 TaK, II.

C.

Ky3HeuoB,

cHHTaa noHHTHe a . n .

fljra

«J)0H0Ji0rHH

BecbMa

BaacHbiM, cTaBHT B o n p o c o TOM, HTO flOJiacHO n p e f l i n e c T B O B a T b B c n c r e M e o n p e f l e j i e H H H : o n p e f l e j i e H H e fl.n. o n p e , a e ; i e H H i o HH, 'TjiacHbie (FCOHEMTI XHHAH", Hapodbi A3uu u Açfipwcu, 1964/4. 13 B. H. TonopoB, "IlpeflBapHTeJibHwe MaTepnaJiM K oimcaHHio 0H0Ji0rmecKHX CHCTeM KOHCOHaHTH3MaflapflCKHXJBMKOB", JIumeucmmecKue uccAedoeauuH no oôufeû u CAQ8HHCK0Ü munoAozuu (MocKBa, 1966); E. A. 3axapbHH, iìpoÓAeMbi (ßoHOAOZuu H3biKa KauiMupu (MocKBa, 1974). 14 M . H . JleKOMueBa, "®0H0JiorHHecKaa CHcreMa TaMiuibcicoro H3WKa c TOUCH 3peHHH CHHTe3a", Bonpocbi cmpyKmypbi H3biKa (MocKBa, 1964). 15 A. II. EapaHueB, (MocKBa, 1968), CTp. 57-74; CM. Taiotce "Factors in the Phonetic Evolution of Language as a Work Process", Selected Works: Articles on General Linguistics (=Janua Linguarum, series maior 72, The Hague: Mouton, 1974), crp. 65-80];

cp.

BUM.

Be.

HBAHOB,

"JlHHrBHCTHiecKHe

B3RNNABI

E.

H.

IIOJIHBAHOBA",

Bonpocbi H3MK03HÜHUH, 1957/3, CTp. 55-76; B . A. BHHorpaflOB, "Teopiw (J)OHeTHMecKHX KOHBEPREHUHÜ E .

fl.

IIOJIHBAHOBA H n p H H q n n CHCTCMHOCTH B ^ O H O J K H - H H " ,

Mame-

puajibt Koiifßepenifuu "AKmyaAbHbie eonpoevi coepeMemozo H3biK03HaHun u AUHzeucmutecKoe nacAedue E. ff. IIoAueaHoea", I (CaMapxaHA, 1964), CTp. 13. Cp. T . B . raMKpejiHfl3e, CuöuAHHmuue coomeemcmeun u neKomopvie eonpocu dpesHeüiueü cmpyxmypbi Kapmee/ihCKUx R3biKoe (T6HJIHCH, 1959); T . B . raMKpejinn3e, r . H. ManaBapHaHH, CucmeMa couanmo« u aÖAaym e KapmeeAbcKux x3bimx. TunoAOZUH oöufeKapmeeAbCKoä cmpytcmypbi (TÖHJIHCH, 1965); T. H. MaiaBapnaHH, OöufeKapmeeAbCKan KOHCOHanmHax cucmeMa (T6HJIHCH, 1965); T . B . TaMxpejiHiue, B . B . HBaHOB, OßufeuHdoeeponeücKuü H3biK. PeKOHcmpyKifun u cmpyKmypuo-munoAozutecKuH QHQAU3 UHÖoeeponeücKoao H3biKa u ¡mdoeeponeiicKOu KyAbmypbt (TÖHJIHCH, B neiaTH); H Ap. 51

52 C. K. IIIayMHH, Hcmopun cucmeMbi du(ß(fiepeHifuaAbHbix j/ieMeumoe e noAbcxoM HlbtKe. 63 T. X. EmmpeHKOBa, HccAedosamtn no duaxpomnecKoU (pOHOAOzuu undoapuiicKUX H3hiKoe (MocKBa, 1974). 54

T . B . raMKpeJiHfl3e, ",E[e3apHKaTH3aiiHH B C B 3 H C K O M " [CM. n p H M e i a m i e

19].

O HeutcKOM cmuxe npeuMyufecmeeuHO e conocmae/ieiiuu c pyccKUM [ = CöopHUKU no meopuu noimmecKozo H3bwa, V ] (Berlin-Moskau, 1923). 56 R. Jakobson, "Typological Studies and Their Contribution to Historical Comparative Linguistics", Proceedings of the VIII International Congress of Linguists, Oslo 1957 (Oslo, 1958), CTp. 17-25 [Selected Writings, I, crp. 523-32], Cp. Bhh. Be. HBaHOB, "TmiOJIOrHH H CpaBHHTejIbHO-HCTOpHieCKOe H3bIK03HaHHe", Bonpocbi H3blK03HaHUR, 55

1958/5, CTp. 3 6 ; e r o ace, " O

npepbiBHCToö HHTOHauHH B jiaTbimcKOM jnbiice" [CM.

npHMenaHne 45], 57 R. Jakobson, Kindersprache, Aphasie und allgemeine Lautgesetze (Uppsala, 1941) [Selected Writings, I, CTp. 328-401], 58 Cp. yace R. Jakobson, "Zur Struktur des russischen Verbums", Charisteria Gvilelmo Mathesio ... oblata (Prague, 1932), CTp. 78-84 [Selected Writings, II, erp. 3-15], 59 M. B. IlaHOB, " C n 0 B 0 0 6 p a 3 0 B a H H e " , PyccKuü Ü3UK U coeemcKoe oöufecmeo (AjJMa-ATa, 1962); KD. R. AnpecsH, Hdeu u Memodbi coepeMemoU cmpyKmypnoü AUHzeucmuKU (MocKBa, 1966), CTp. 73; I. I. Revzin, "Logic and Language", Contemporary Philosophy: A Survey, ed. R. Klibansky (Firenze, 1969), CTp. 329. 60

T . B . R A M K P E J H « 3 e , " C o o T H o m e H H e CMMHHMX H pmcaTHBHbix B (JWHOJIORHNECKOIT

CHCTeMe",

K npoÖAeMe MapKupoeanuocmu e

@OHOAOZUU

(MocKBa, 1974); ero

ace,

116

T. B. rAMKPEJIHfl3E, T. H. EJIH3APEHKOBA, B. B. HBAHOB

"Interrelationship of Stops and Fricatives in a Phonological System", Lingua, 35/3-4 (1975). 61 H. T. MejiHKHinBHjm, Omnomenue MapKupoeamtocmu e çfioHOAOzuu (ycAoeux MapKupoeaHHocmu e KAacce uiyMnbtx (poneM) (T6HJIHCH, 1 9 7 2 ) ; ero *e, " K myiemiio HepapxHHecKHX OTHOiiieHHfi eflHHHU OHOJiorHiecKoro ypoBHH", Bonpocbi HÌÒIKOmanu»,

1974/3, c r p . 94-105.

82

T. B. raMKpejiHfl3e, "CooTHOineHHe CMbiiHbix H (jjpmcaTHBHbix B (f>0H0Ji0rH4ecK0ii CHCTeMe" [CM. npHMeiaHHe 60]; ero «e, "Order of 'Rewrite Rules' in Diachronic Phonology", Linguistics, 126 (1974), CTp. 25-31. 63 Travaux du Cercle Linguistique de Prague, VI (1936), CTp. 240-88 [Selected Writings, II, crp. 23-71]. 64 American Contributions to the IVth International Congress of Slavisti (The Hague, 1958), crp. 127-56 [Selected Writings, II, CTp. 154-83], 65 B. B. HBaHOB, B. H. TonopoB, CancKpum [CM. npHMeiairae 4]. 66 B. B. HBAHOB, XemmcKuU H3UK [CM. n p H M e n a H H e 7 ] . 87 H. M. TpoHCKHit, OòufeuHÒoeeponeùcKoe muKosoe cocmonnue (Bonpocbi pemncmpyKifuu) (JleHHHrpaa, 1967). 68 H. M. PeB3HH, "O CHJibHbix h cjiaôwx npoTHBonocTaBJieHHflx B CHCTeMe nafleJKeft coBpeMeHHoro HeMeuKoro «buca", Bonpocbi H3biK03Hauun, 1960/3, CTp. 85. 89 E. B. HeuiKO, "CncTeMa nafleaceft flpeBHeôonrapcicoro H3biica", Bonpocbi x3biKO3HOHUH, 1967/2, CTp. 49-63; cp. ero see, "K Bonpocy o nafleacHbix KoppeJwuHHx", Bonpocvi H3viK03HaHun, 1960/2, CTp. 50-56. 70 E. B. IlanyieBa, " 0 6 onncaHHH nafle)KHOii CHCTCMH pyccKoro cymecTBHTejitHoro", Bonpocbi H3I>IKO3HQHUH, 1960/5, C T p . 104-11 ; 3. M. Bojiouicaji, T. H. MojioiHaa, T. M. HHKOJiaeBa, Onwn onucanun pyeexoeo n3biKa e ezo nucbMennoU (fiopMe (MocKBa, 1914), CTp. 33-38. 71 E. B. Hernico, "K Bonpocy o na^excHbix KoppejinuHHx" [CM. npHMenaHHe 69], 72 H. A. MeJibiyK, "Moflenb cnpsjKeHHH B H c n a H C K O M H3biKe", Maiuunnbiu nepeeod u npuKjiadnaa AumeucmuKa, B b i n . 10 (1967), C T p . 1341 ; ero ace, "OnbiT Teopira j i H H r B H C T H necKHX M O f l e J i e f i " , CMMCA - TeKcm (MocKBa, 1974), C T p . 58. 73

O.

T . PeB3HHa,

npH3HaKOB

flJIH

H.

H.

PeB3HH,

CJI0B006pa30BaHHH

"K

nocTpoemiio

CHCTCMH ,QHepeHUHajibHbix

CymeCTBHTejIbHblX

CJiaBHHCKHX

Honor Roman Jakobson, II (The Hague, 1967), CTp. 1657-66. 74 H. H. P e B 3 H H , "O C H J i b H b i x H c j i a ô b i x n p o T H B o n o c T a B J i e H H H x

H3bIKOB",

B CHCTeMe

To

naaexceii

c o B p e M e H H o r o H e M e u K o r o « b u c a " [CM. n p H M e i a H H e 6 8 ] , C T p . 8 5 . 75

C. A- KauHejibcoH, TunoAozun H3biKa u peteeoe MbiuiAenue (MocKBa, 1972), 42. 78 H. M. TpoHCKHfi, OôufeuHdoeeponeitcKoe n3biKoeoe cocmosme [CM. npHMeiaHHe 67], CTp. 73-82. 77 KpamKuù onepK cpaenumeAbnoii zpauuamuKU apuoeeponeùcKux a3UKoe (Ka3am>, 1916). 78 H. M. TpoHCKHfi, OôufeuHdoeeponeitcKoe H3biKoeoe cocmoHHue [CM. n p H M e i a H H e 67], CTP. 74. 78 C. X. EbixoBCKan, "'IlaccHBHaa' KOHcrpyiuiHji B seTHHecKnx mbucax", 23HK u MbiuiAenue, II (JleHHHrpaa, 1934); C. R. KauHejibcoH, "K reHe3Hcy HOMHHâTHBHoro npeflJioîKeHHa", Tpydti Hncmumyma n3biKa u MbiuiAenun UM. H. H. Mappa, IV: CepuH Romano-Germanica (1936); ero »ce, "SpraraBHaa KOHCTpyKnHH H apraTHBHOe npeflJioaceHHe", H3eecmux AH CCCP. OmdeAeme Aumepamypbt u a3biKa, VI/1 (1947); ero »e, TunoAozun H3WKa u peneeoe MbiuiAenue (MocKBa, 1972), CTp. 68-71; H. H. MemaHHHOB, JÎSWK BOHCKOU KAUHOHUCU {Die Van-Sprache), II: Cmpyxmypa petu (JleHHHrpa/i, 1935); ero ace, Hoeoe yneuue o H3biKe. CmaòuaAbnan munoAoeux (JleHHHrpaa, 1936); ero «e, Oôufee H3biK03namie. K npoÓAeMe cmaduaAbHOcmu e pa3eumuu CTp.

JIHHrBHCTHHECKAÄ

117

TEOPHFL

CAoaa u npedAOMcenun (JleHHHrpaa, 1940); ero » e , "SpraTHBHbiii CTpoü npewiOKeHHH", IJpoô/ieMbi cpasHumeAbHoü fßuAOAOzuu. CôopHUK cmameü K 70-jtemuto HAenaKoppecnoudeuma AH CCCP B. M. XupMyncKoeo (MocKBa-JleHHHrpaa, 1964); ero » e , Opzamueuan KOHcmpyKifun e Ribixax pamwmux munoe (JleHHHrpafl, 1967); A. C. HHKo6aBa, ITpoôjteMa spaamuenoii KOHcmpyxifuu e ußepuüCKO-KaeKajCKUx Jt3bixax, II: Teopun cyufHocmu 3peamu6Hoü KOHcmpyKifuu [Ha rpy3HHCKOM »3biKe] (Töhjihch, 1961). 80

R. Jakobson, "Beitrag zur allgemeinen Kasuslehre", Travaux du Cercle Linguistique de Prague, VI (1936), CTp. 254 [Selected Writings, II, crp. 37]. 81 Cp. T . A . Kjthmob, OtepK oôufeù meopuu spaamueiiocmu (MocKBa, 1 9 7 3 ) ; T . B . r a M K p e j m A 3 e , B . B . H s a H O B , OöufeuHdoeeponeücKuü n3biK [cm. n p H M e i a H H e 5 1 ] . 82 T. A. Kjihmob, OnepK oöufeü meopuu 3peamueuocmu, CTp. 115; cp. C. ff. KauHe.ni>coh, TWIOAOZUH Hiuxa u peneeoe MbWAeme [cm. n p H M e q a m i e 75], CTp. 70. 83

Ban. Be. ÜBaHOB, "THnojiorow h cpaBHHTejibHO-HCTopHMecKoe H3biK03HaHne",

Bonpocbi H3biK03Hamn, 1958/5, CTp. 39-40. 84 T. A. Kjihmob, OtepK ooufeü meopuu 3paamuenocmu, CTp. 191. 85 R. Jakobson, "The Relationship between Genitive and Plural in the Declension of Russian Nouns", Scando-Slavica, III (1957), CTp. 181-86 [Selected Writings, II, crp. 148-53], 86

C . FF. K a i j H e J i b c o H , TUHOAOZUH H3viKa u peieeoe

87

HanpHMep, b inyMepo-aKKaflCKo-xeTTCKHx cjiOBapHbix ciracxax: B. B. MßaHOB,

XemmcKuü

MbiuiAewte,

CTp. 4 6 .

X3WK.

88

Cp., HanpHMep, CBOflKy yHHBepcajuift: B. A. YcneHCKHii, CmpyKmypnan munoAOBUH H3biKoe (MocKBa, 1965), CTp. 186, 189, 191, 193-94, 196-97, 212. 89 R. Jakobson, "Typological Studies and Their Contribution to Historical Comparative Linguistics" [cm. npHMeiamie 56]; cp. Ban. Be. HBaHOB, "THnoJioraa h

cpaBHHTejibHO-HCTopHHecKoe H3biK03HaHHe" [cm. npHMenaHHe 83]; H. M. TpoHCKHft, OöufeuudoeeponeücKoe H3biKoeoe cocmoHHue [cm. npHMenaHHe 67], CTp. 39. T. B. raMKpejnw3e, B. B. HBaHOB, "PeKOHcrpyKijHa CHcreMbi mtaoeBponeftCKHx CMbiHHbix", KoHfßepeHifua no cpaeHumeAbHo-ucmopuuecKou epaMJuamuice UHdoeeponeitCKUX H3biKoe ( 1 2 - 1 4 deKaôpn). ÎTpedeapumejibHbie MamepuaAbi (MocKBa, 1972), CTp. 15-18; T. B. raMKpeJiHfl3e, B. B. MßaHOB, OöuteuHdoeeponeücKuü n3biK [cm. npHMeiaHHe 51]; cp. Th. V. Gamkrelidze, V. V. Ivanov, "Rekonstruktion der indogermanischen Verschlüsse. Vorläufiges Bericht", Phonetica, 27 (1973), crp. 150-56. 91 B. H. Tonopoß, " O BBeaeHHH BepoaraocTH b a3HK03HaHHe", Bonpocbi H3biK03HaHUH, 1959/6, CTp. 35. 92 Cp., HanpHMep, M. M. TyxMaH, "HHfloeBponeficKoe cpaBHHTeJibHo-HCTopHiecKoe 90

H3bIK03HaHHe

H THnOJIOTHHeCKHe

HCCJieAOBaHHa",

Bonpocbi

H3 blK03HaHUH,

1957/5,

CTp. 5 5 . 93

C. ,£{. KauHeJTbcoH, "K 0H0Ji0rHHecK0ii HHTepnpeTauHH npoTOHHfloeBponeflcKOit Bonpocbi H3UK03HQHUH, 1 9 5 8 / 3 , CTp. 4 7 ; cp. 06 3 T 0 M B o n p o c e B njiaHe o6uieü nraoJiorHH B. A. YcneHCKHÜ, CmpyKmypuan munoAoaun HJbiKoe, CTp.

3ByKOBOÖ CHCTeMbl", 27, 187.

94 P. O. äkoöcoh, K xapaKmepucmuKe eepaiuücKoao H3biKoeoao coto3a (IlaproK, 1931) [Selected Writings, I, CTp. 144-201]. 95 T. M. CyflHHK, ffuaAeKtnbi AumoecKo-cAaeimcKOzo nozpanuHbH. OiepKU (ßonoAoeunecKux cucmeM (MocKBa, 1975). Cp. Taioice pa6oTbi 06 apaScKHx anaJieKTax Cpe^Heft A3HH b hx CBH3H c flpyrHMH «3biKaMH Toro »ce apeajia: T. B. IJepeTejm, ApaöcKue duaAeKmu Cpedneü Asuu, I. Eyxapcmü duaaeicm (Töhjihch, 1957); h flp. 99 T. B. raMKpejiHÄ3e, T. M. ManaBapHaHH, CucmeMa couaumoe u aÖAaym e KapmeeAbCKUx H3UMX. TunoAoeuH oôufeKapmeeAbCKOù cmpyxmypbi (Töhjihch, 1965); T. V. Gamkrelidze, " A Typology of Common Kartvelian", Language, 42 (1966), CTp.

118

T. B. rAMKPEJIH/PE, T. X. EJIH3APEHKOBA, B. B. HBAHOB

69-83; T. B. raMiepejiHfl3e, B. B. MBaHOB, OdufeundoeeponeucKuu nibiK [CM. npHMeiaHHe 51]. 97 r . B. LJepeTeJiH, "O « M K O B O M POFLCTBE H «BIKOBBIX coio3ax", Bonpocu A3biK03HaHUH, 1968/3, CTp. 3-18. 88 B. H. TonopoB, "HecKOJibKO 3aMeiaHHfi K (JioHonorHiecKoii xapaKTepHCTHKe ueHTpanbH0-a3HaTCK0r0 H3biKOBoro coK>3a (UAilC)", Symbolae Linguisticae in honorem Georgii Kurylowicz (Wroclaw-Warszawa-Krak6w, 1965). 99 B. H. TonopoB, "OoHOJiorniecKaH HHTepnpeTauHii KOHCOHaHTH3Ma KaiiiMHpH B CBH3H c THnoJiorneiiflapflCKHXH3biKOB", CeMuomuKa u eocmomtbie H3bmu (MocKBa, 1967). 100 x % EjiH3apeHKOBa, "O BJIHHHHH TaMHJibCKoii (JwHOjiorHHecKoft CHCTeMbi Ha CHHraJibCKyio", Bonpocu 0OHOAOZUU eocmouubtx nibiKoe (MocKBa, 1975). 101 T. M. CyflHHK, ffuaAeKmu AumoecKO-cAaenHCKOZo nozpauuibH [CM. npHMenaHHC 95]; M. H. JleKoivmeBa, "O B3anMojieftCTBHH (JwHOJiorHHecKHX CHCTCM B paftoHe 6anT0-cnaBHHCK0r0 norpaHHib«", Ea.imo-c/taenHCKUu cdopnuK (MocKBa, 1972). 102 M. H. JTexoMueBa, "K ranoJiorHHecKoit xapaKTcpncTHKc (fcoHOJiorHHecKHx CHCTCM flaaneKTOB jiaTbiuiCKoro H3biKa", EaAmo-cAaeHHCKue uccjiedoeanun (MocKBa, 1974). 103 R. Jakobson, M. Halle, Fundamentals of Language; E. A. YcneHCKHft, "ITpo6jieMM jiHHrBHCTHHecKoii THnoJiorHH B acneKTe pa3JiH4eHHH 'roBopnmero' (anpecaHTa) H 'cnyiiiaiomero' (aapecaTa)", To Honor Roman Jakobson, III (The Hague: Mouton, 1967), CTp. 2087-2108. 104 H. H. PeB3HH, B. K). Po3eHUBeflr, OCHOSU odufezo u MOUIUHHOZO nepeeoda (MocKBa, 1964), CTp. 83. 105 C. A- KauHejibcoH, Tunojtozun n3biKa u peteeoe Mbiutjienue (JleHHHrpa^, 1972), CTp. 122. 106 TaM CTp. 110. 107 R. Jakobson, M. Halle, Fundamentals of Language. 108 E. H. BHHapcKaa, KAununecKue npo6/ieMbi a(fia3uu (MocKBa, 1971), CTp. 23, 156-57, 165. 109 A. P. JlypHH, "npo6jieMW H (JjaKTH HeflpojimirBHCTHKH", Teopu.t peueeou dexme.ibHocmu (JIpodaeMbi ncuxoAuuzeucmuKu) (MocKBa, 1968), CTp. 214. 110 K). K. JlexoMueB, "Crpoemie xofla H KJiaccHctrnKaijHsi npocTbix npefljioaceHHfi", MautuHHbiu nepeeod u npuKAadnan AunzeucmuKa, 1959/3 (10), CTp. 38. 111 R. Jakobson, "Shifters, Verbal Categories, and the Russian Verb" (Russian Language Project, Department of Slavic Languages and Literatures, Harvard University, 1957) [Selected Writings, II, CTp. 130-47]; B nepeBOfle A. K. JKoJiicoBCKoro: "IIlH(J)Tepbi, rnaroJibHbie KaTeropHH H pyccKHft rjiaron", ITpuHifunbi munoAozuiecKozo aHaAU3a H3UK08 pa3AuHHOZO cmpo/t (MocKBa, 1972); cp. Ban. Be. MBaHOB, " n o m n n e HefiTpaJiH3ai]HH B Mop^onoraH H jiexcHKe", EioAAemem OdbedunenuH no MamuHHOMy nepeeody, 1957/5. 112 Hanprnnep, B. B. HBaHOB, B. H. TonopoB, CancKpum (MocKBa, 1960). 113 O. r . PeB3HHa, "npeflHcnoBHe", IlpuHifunbi munoAozmecKOZo auaAU3a H3UKOB paiAUHHOZO cmpoH (MocKBa, 1972), CTp. 9. 114 R. Jakobson, "Zur Struktur des russischen Verbums", Selected Writings, II, CTp. 9. 115 B. B. BimorpaaoB, PyccKuU H3UK. rpaMMamuuecKoe yneme o CAoee (MocKBa, 1947), CTp. 465, 653. 116 Cp. B. A. YcneHCKHji, CmpyKmypuan munoAozun mviKoe (MocKBa, 1965), CTp. 55; ero ace, "npo6jieMbi jiHHrBHCTHiecKoit THnojionw B acneKTe pa3JiHieHHji 'roBopamero' (aapecairra) H 'cjiyuiawmero' (aapecaTa)" [CM. npHMeiaHHe 103], CTp. 2089. 117 H. A. MejibiyK, "O cooTHomeHHH cuHxpoHHiecKoro H AHaxpoHHiecKoro

JIHHrBHCTHHECKAfl TEOPHJI

119

onHcaHHit (K nocraHOBKe Bonpoca)", Mamepuajibt Bcecotomou KOH0epenifuu no oSufeMy H3btK03HQHUK}. OcHoeHbte npodAeMbi H3bma, M. I (CaMapxaHfl, 1966), CTp. 10; M. B. MepenaHOB, Odufee H3btK03Hame (CapaTOB, 1969), CTp. 58-59. 118 R. Jakobson, "Russian Conjugation", Word, IV (1948), CTp. 155-67 [Selected Writings, II, crp. 119-29]; cp, R. Jakobson, "Zur Struktur des russischens Verbums" (1932). 119 B. r . HypraHOBa, OtepK pyccKou MOP$OHOAOZUU (MocKBa, 1973). 120 Cp. H. A. MejibiyK, "Moflejib cnpa*eHHH B HcnaHCKOM H3biKe" [CM. npHMeiaHHe 72]; A. A. 3ajiH3HHK, "CHHxpoHUpe onncaHHe H BHyTpeHHHH peKOHCTpyxuHH", iJpo6AeMbi cpaeHumeAbHoii zpaMMamuKu undoeeponeucKux H3bmoe. Haywaft ceccwt. Te3ucbt doKAadoe (MocKBa, 1964); ero xce, PyccKoe UMemoe cnoeoMMewHue (MocKBa, 1967), CTp. 191; H ap. 121 H . A . M e j i b i y x , " M o ^ e j i H cnpraceHHH B HenaHeKOM 83MKe" [CM. npHMeiaHHe 72], CTp. 43.

122 R. Jakobson, "Linguistics and Poetics", Style in Language, pea. Thomas A. Sebeok (Cambridge, Mass., 1960), H B nepeBOfle H. A. Mejibnyica: "JlHHraHCTHKa H no3THKa", CmpyKmypaAU3M: "da" u "npomue" (MocKBa, 1975). 123 A . A . 3ajiH3H)IK, BUM. BE. HBAHOB, B. H . TonopoB, " O BO3MO*HOCTH crpyKTypHo-THnoJiorHHecKoro royHemm HexoTopwx MoaenHpyraiimx CCMHOTHMCCKHX CHCTeM", CmpyKmypHO-munoAozmecKue uccAedoeaHun (MocKBa, 1962), CTp. 134-43; M. H . PeB3HH, B. K ) . PoieHiiBefir, OcnoeW o6ufeeo u MOUIUHHOZO nepeeoda (MocKBa, 1964), CTp. 46-47.

124 H. H. PeB3HH, CoepeMeman cmpyKmypnan AumsucmuKa (MocKBa, 1976); I. I. Revzin, review of Structure of Language and Its Mathematical Aspects, Word, 19/3 (1963), CTp. 399. 125 'TIoHHTHe HeftTpajunauHH...", EwAAemeub 06bedmeHux no MammmoMy nepeeody, 1957/5; "3HaneHHe Hfleit M. M. EaxTHHa...", Tpydbi no 3HaKoebiM cucmeMQM,

VI (TapTy, 1973). 126

O teutcKOM cmuxe (1923) [CM. npHMeiamie 55].

187

C p . B. B. HBaHOB, B. H . TonopoB, " K peKOHCTpyKUHH npacjiaBHHCKoro TeKcra",

CAaenncKoe H3biK03HaHue (MocKBa, 1963). r . B. IlepeTeJiH, "MeTp H PHTM B noaMe PycTaBejiH H Bonpocu cpaBHHTenbHoft BepcH(J)HKauHH", KoHtneKcm 1973. JIumepamypHO-meopemunecKue uccAedoeauun (MocKBa, 1974), CTp. 114-37. 129 R. Jakobson, Kindersprache, Aphasie und allgemeine Lautgesetze. 130 R. Jakobson, C. G. M. Fant, M. Halle, Preliminaries to Speech Analysis (Cam128

bridge, Mass., 1952); H flp. CM. 0 6 STOK SBOJIKDUHH HfleiiP. O . ilKo6coHacneuHaJibHyio

CTATBIO H. H. PeB3HHa: "HeK0T0pwe 3aMenaHHH B CBH3H CflHxoTOMHHecKoftTeopHeft B (JIOHOJIORHH", Bonpocbi H3biK03Hamn, 1970/3. 131

E. B. naflyieBa, "BO3MOMCHOCTH myneHHH mbiica MeToaaMH TeopHH nml>opMaUHH", JJoKAadbi ua KoHtfiepeMfuu no o6pa6omKe unfpopMayuu, MauiumoMy nepeeody u aemoMamuiecKOMy Hmenuio mexema, 5 (MocKBa, 1961); O. C. AxMaHOBa, H. A. MeJibnyK, E. B. IlaayHeBa, P. M. OpyMKHHa, O motHbix Memodax uccAedoeauun H3bixa (MocKBa, 1961). 132 R. Jakobson, "Boas' View of Grammatical Meaning", The Anthropology of Franz Boas: Essays on the Centennial of His Birth (=American Anthropologist, LXI, 5, Part 2; Memoir 89), CTp. 139-45 [Selected Writings, II, CTp. 477-96], 133 H. A. MeJibiyK, "O crannapTHoft opMe H KO/iHHecTBeHHbix xapaKTepHCTHKax HeKOTOpblX JIHHrBHCTHHeCKHX OnHCaHHii", Bonpocu H3HK03H0HUH, 1963/1, CTp. 118. 134 Cp. B. B. HBaHOB, B. H. TonopoB, CAaeswcKue nsbtKoebie ModeAupyiouiue ceMuomuiecKue cucmeMbi (MocKBa, 1965); HX ace, HccAedoeanua e odAacmu CAOBHHCKUX dpeeHocmeu (MocKBa, 1974).

120

T. B. rAMKPEJIHfl3E, T. SL EJIH3APEHKOBA, B. B. HBAHOB

135 B. B. HBaHOB, "flyara»Haa opraHH3aunn nepBo6biTHbix HapoaoB h npoHCxo»fleHHe ayaJiHCTHiecKHx KOCMoroHHli", CoeemcKaH apxeoAozua, 1968/4. 138 P. O. JIKO5COH, "Mop(J)OJiorH4ecKHe Ha6nK>,neHHH Haa cnaBHHCKHM CKJIOHCHHCM** [CM. npHMenaHHe 64]. 137 Cp. 06 3tom E. B. HeuiKO, "K Bonpocy o nanexHbix Koppennumtx", Bonpocbi H3biK03Hamift, 1960/2; ero xe, "CncTeMa nanexceit apeBHe6ojirapcKoro H3WKa", Bonpocbi H3VK03HMUH, 1967/2; H. H. Pgb3hh, CoepeMewuui cmpyKmypuan AumeucmuKa (MocKBa, 1976); H flp. 138 R. Jakobson, "Boas' View of Grammatical Meaning" [cm. ripHMeiaHHe 132], 139 A. A. 3ajiH3HHK, PyccKoe UMeuuoe CAoeou3Meneme (MocKBa, 1967), CTp. 25; cp. O. C. AxMaHOBa, H. A. MeJibnyK, E. B. naflyneBa, P. M. OpyMKHHa, O mowux Memodax mynemH H3ViKa (MocKBa, 1961). 140 1.1. Revzin, "Logic and Language" [cm. npHMeiaime 59], CTp. 328; H. H. PeB3HH, CoepeMeHHan cmpyKmypuan AumeucmuKa (MocKBa, 1976). 141 H. A. Mejibiyx, "K Bonpocy o 'BHeuiHHx' pa3HHiHTejibHbix sjieMeHTax: ceMaHTHHecKne napaMeTpbi h orwcaHHe jieKCHnecKoft coHeTaeMocra", To Honor Roman Jakobson, II (The Hague: Mouton, 1967), CTp. 1341; cp. B. B. HBaHOB, "CTpyKTypHaa THnojiorHH h eflHHCTBO JiHHrBHCTHKH", JIumeucmunecKaH munoAoeuH u eocmoiHbie H3UKU (MocKBa, 1965), CTp. 33-38; 1.1. Revzin, "Logic and Language" [cm. npHMenaHHe 59], CTp. 329. 142 K). A- AnpecHH, Hdeu u Memodbt coepeMemou cmpyKmypaAbHou AumeucmuKa (MocKBa, 1966), crp. 60, 72-73; H. A. Mejibiyx, "K Bonpocy o 'BHeniHHx' pa3JiHHHTejibHbix 3JieMeHTax" [cm. npHMeiamie 141], crp. 1341; ero ace, "Oiimt TeopnH jiHHrBHCTHHecKHX MOAeJieft", CMNCA - Tencm (MocKBa, 1974), CTp. 58 h ap. 143 R. Jakobson, "On Linguistic Aspects of Translation", On Translation (Cambridge, Mass.: Harvard University Press, 1959), crp. 232 [Selected Writings, II, CTp.

261]. 144

H. M. PeB3HH, B. K). PosemiBeiir, Ocnoebi odufeeo u Matuumozo nepeeoda (MocKBa,

1964), CTp. 2 8 . 145

E. B. naayneBa, "flBa noflxoaa k ceMaHTHHecKOMy aHajnuy KaTeropHH inc.ria", To Honor Roman Jakobson, II (The Hague: Mouton, 1967), crp. 1475 [CM. "Two Approaches to the Semantic Analysis of the Category of Number", Linguistics, 103 (1973)]; ee ace, "Tojtce h Taose: BsaHMooTHomeHHe aKTyanbHoro uieHeHHH H accoUHaTHBHbix CBH3e0", TIpedeapumeAhHue nydAunatfuu npo6AeMhou zpynnn no SKcnepuMeHmaAbHou u npuKAadnou AumeucmuKe, Bbin. 55 (1974), CTp. 11; K). fl. AnpecaH, JleKcmecKan ceManmuxa (MocKBa, 1974); H. A. MeJibiyK, "OnbiT Teoprni JIHHI"BHCTHMecKHX MOflejreft", CMUCA - Tencm (MocKBa, 1974), CTp. 11. 149 E. B. naayneBa, "Toace h TaiOKe..." [cm. npHMeiairae 145]. 147 TaM ace. 148 R. Jakobson, "Quest for the Essence of Language", Selected Writings, II, crrp. 345-59, h b nepeBoae A. H. }KypHHcicoro h B. JI. BHHorpaflOBa: "B noHCKax cyiUHOCTH »3biKa", CdopnuK nepeeodoe no eonpocaM UH^opjuatfuoHHoii meopuu u npaxmuKU, 16 (MocKBa, 1970); ero ace, "Boas' View of Grammatical Meaning" [cm. npHMeiaHHe 132], 148 A. A. 3ajiH3HHK, PyccKoe uMemoe cAoeou3MeHeuue (MocKBa, 1967). 150 E. A. KpeiiHOBHi, "BbipaaceHHe npocTpaHCTBeHHoK opneHTauHH b hhbxckom «3biKe (K hctophh opneHTauHH b npocTpaHCTBe)", Bonpocbi H3UK03naHun, 1960/1, crp. 84. 151 E. B. naayneBa, O ceMaumuRe cunmaKcuca (Mamepua.iu K mpanctfiopMaifuoHHOu epaMMamuKe pyccKozo h3uko) (MocKBa, 1974), crp. 8. 152 R. Jakobson, "Linguistic Glosses to Goldstein's 'Wortbegriff'", Journal of Individual Psychology, XV (1959), CTp. 62-65 [Selected Writings, II, CTp. 227-71].

JIHHrBHCTHHECKAJI 153

Bin.

Be.

munoAoemecKue

HBaHOB,

"JlHHrBHCTmca

uccAedoaamn

h

121

TEOPHJI

HCCJieAOBamie

aa3HH",

Cmpyxmypuo-

(MoCKBa, 1962).

134

E . B . r i a f l y n e B a , O ceMaHmwe

155

T . B . T a M K p e J i i w e , " K n p o ô n e r n e npon3BOflHOCTH H3iiKOBoro 3 H a x a " ,

cuumaxcuca

[cm. n p H M e i a i m e 151],

H3*iK03HaHU!i, 1 9 7 2 / 6 , CTp. 3 3 - 3 9 ; e r o ace, " T h e P r o b l e m o f ' l ' a r b i t r a i r e d u Language, 156

5 0 / 1 ( 1 9 7 4 ) , CTp. 1 0 2 - 1 0 .

C p . H . H . PeB3HH, " 0 6 0flH0M n o f l x o f l e

l e c i e o r o a H a J i H 3 a " , ITpoôjieMbi CoepeMeHHdH

cmpytanypHan

cmpyKmypwibHoû AumeucmuKa

k moacjihm AumeucmuKu

(MocKBa, 1976).

flHcrpHÔyTHBHoro

Bonpocu signe'",

OHOJiorn-

(MoCKBa, 1 9 6 2 ) ; e r o »ce,

ROMAN JAKOBSON'S CONTRIBUTION TO THE MODERN STUDY OF SPEECH SOUNDS* MORRIS HALLE

Jakobson's early views on the phonic aspect of language were in their essence a reaction against the atomism that is endemic in much of the humanities, in general, and in linguistics, in particular. When Jakobson began his linguistic studies in the second decade of this century this atomism was especially striking in the then prevalent approach to the phonic side of language. Phonetics, as exemplified, for instance, by Jespersen's Lehrbuch der Phonetik (Leipzig, 1904) appeared to be concerned with developing notations for capturing the vocal tract configurations of ever finer nuances of sounds, whereas historical phonology, completely dominated by the Neogrammarian view of sound laws, consisted primarily of a hunt for correspondences among sounds that appeared to have little rhyme or reason. A possible way out of this morass had been indicated by F. de Saussure, with whose doctrines Jakobson first became acquainted in 1917 through reports of S. Karcevskij, who had just returned to Russia after studying for some years in Geneva.1 Jakobson was much impressed by Saussure's distinction between langue and parole, which provided a means for distinguishing between a restricted set of abstract phonemes (on the level of langue) and the much larger (perhaps unbounded) set of concrete sounds produced by speakers when actualizing phonemes in concrete utterances {parole). Since many accidental factors play a significant role in the actual production of sounds - speakers differ in the speed of speaking, they are tired, they acquire peculiar mannerisms, etc. - it is the first task of the linguist to filter out the effects of these accidental factors. To achieve this, Saussure proposed that primary attention be directed to the communicative function of language. In particular, he pointed out that phonemes serve to signal differences among words and asserted that this differentiating function constitutes the essence of the phoneme: "Every language forms its words on the basis of a system of sonorous elements, each element being a clearly delimited unit and one of a fixed number of units. Phonemes are characterized not, as one might think, by

124

MORRIS HALLE

their own positive quality but simply by the fact that they are distinct." 2 ". . . language requires only that the sound be different and not, as one might imagine, that it have an invariable quality. I can even pronounce the French r like the German ch in Bach, doch, etc., but in German I could not use r instead of ch, for German gives recognition to both elements and must keep them apart." 3 To have elaborated these functionalist notions in a highly original fashion and to have applied them in the solution of a host of linguistic problems both synchronic and diachronic is one of the great achievements with which historians of linguistics will have to credit Roman Jakobson. Like Saussure, Jakobson draws a sharp distinction between distinctive features of speech sounds ("faits . . . capables de différencier les significations dans la langue intellectuel") 4 and their other phonetic properties. For a time the latter were relegated to an almost invisible position in the scheme of things, so that Jakobson felt justified in likening the relationship between phonology, the study of distinctive features of speech sounds, and phonetics, the study of all properties of speech sounds without regard to their distinctive function, to the relationship holding between political economy and the statistics of the gross national product, or between the science of finance and numismatics. 5 He soon became aware that this was too radical a move, but in this early stage the fact that all sorts of phonetic minutiae could be relegated to the background so that maximum attention could be concentrated on a clearly limited domain of relevant data had a tremendously liberating effect: it was possible for the first time to discern functional patterns in language that appeared to be of great generality and to possess obvious significance. These were primarily the patterns constituted by the phonemes of a given language, to which the name phonological system was then applied. The study of phonological systems has remained throughout Jakobson's long preoccupation with phonology at the center of his concerns. It behooves us, therefore, to examine in detail what he understands by this term. From the very beginning, Jakobson viewed the phonological system as something much more structured than a mere list of phonemes, than an alphabet. 6 The phonological system always implied for him a specific organization of the phonemes into a multi-dimensional array, where each phoneme is distinguished from every other phoneme by one or more "significant differences." Thus, in his "Propositions," submitted in 1927 to the First International Congress of Linguists and co-signed by S. Karcevskij and N. Trubetzkoy, Jakobson wrote: "Every scientific description of the phonology of a language must above all contain a characteriza-

THE MODERN STUDY OF SPEECH SOUNDS

125

tion of its phonological system, that is to say, a characterization of the repertory, proper to this language, of the distinctive differences among its acoustico-motor units (des différences significatives entre les images acoustico-motrices)." 7 And again in the introductory section of the early masterpiece, Remarques sur l'évolution phonologique du russe, written at the same time, Jakobson observes: "We call phonological system of a language,. . . the repertory of oppositions which in a given language can be associated with a differentiation of meaning (repertory of phonological oppositions). Terms of a phonological opposition that are not susceptible to being dissociated into smaller sub-oppositions are called phonemes." 8 He is at particular pains to make clear that the phonemes are not the ultimate elements of the analysis but that they themselves are to be viewed as entities in a pattern of differential signals. Thus, he rejects the view of the phonological system as a collection of sounds, on the grounds that such a view "would involuntarily focus attention on the ideas of the acousticomotor units themselves. The types of their mutual inter-relations would then not be subject to the desired analysis; but, it is just in these (interrelations)i that the essence of a phonological system inheres. The sign in itself is fortuitous and arbitrary." 9 Perhaps the clearest illustration of Jakobson's conception of the structure inherent in a phonological system can be found in the following passage from his 1930 monograph K xarakteristike evrazijskogo jazykovogo sojuza:10 One must not imagine the phonological inventory of a language as a mechanical sum of phonemes. Phonemes stand in specific relations one to another; they constitute a system. Not infrequently a phoneme of one language is in itself (physically) identical with some phoneme of another language but the place that the two phonemes occupy in their respective phonological systems is different. As an illustration let us juxtapose the system of vowel phonemes belonging to three different languages: I a o e u i

II

III

a o e

a ui

In both the first and the second system we find the phonemes o and e, but their position within the system differs. In language II they are the corner phonemes of a triangle: o is acoustically the lowest pitched, and e the highest pitched vowel in the system; in language I, the vowels o and e are phonemes of "the middle layer," whereas the role of the acoustically lowest pitched corner phoneme is played by u and the role of the highest pitched by ». As regards their function in the system the vowels u and i in language I are closer to the vowels o and e in language II

126

MORRIS HALLE

than are the externally identical phonemes o and e of language I. The closest parallel to system II is system III. These two systems are phonologically identical; except that in system II the lowest pitched phoneme is realized as o, whereas in system III as u, and correspondingly the highest pitched phoneme is realized in one case as e and in the other as i.11 These statements call for two comments. On the one hand, it Should be noted that the basis of the distinction is phonetic (auditory). In the case under discussion, the vowels are distinguished with regard to their proper pitches and in all three languages the extreme pitch values - i.e., the lowest and highest pitched vowel - occupy a special position. What differentiates the languages are the absolute values that characterize the extrema; these may differ from language to language - as Jakobson's example shows without affecting the structure of the system in which maximally high- and low-pitched phonemes (those forming the two lower corners of the sound triangle) are distinguished from the rest of the vowels.12 The second remark appropriate here is that at least in the earlier stages of his phonological investigations Jakobson believed that oppositions differed not only with respect to the phonetic properties employed in their physical actualization but also with respect to their abstract structure. Thus in Remarques he observes that "the phonological system presents two fundamental types of oppositions: correlative and disjoint." 13 The former, exemplified by such pairs as voiced-voiceless, palatalized-unpalatalized, stressed-stressless, long-short, etc., are the precursors of the binary features that in the later development of Jakobson's thought became the only type of distinctive feature. The disjoint oppositions constitute all the rest. Jakobson gives14 the following table of the phonemic system of modern Russian, where "parentheses enclose the correlative phonemes:" (i, i) (ü, u) é ó (ä, a)

THE MODERN STUDY OF SPEECH SOUNDS

127

As can be seen from the above table the disjoint oppositions were not necessarily nonbinary, but rather they did not characterize relationships between pairs of phonemes where one member possessed a particular feature (e.g., was voiced) and the other did not (e.g., was voiceless). Instead, disjoint oppositions characterized phonemes that differed in other ways than possession vs. nonpossession of a specific property, e.g., the contrast between the stops and fricatives is binary and disjoint.15 In Jakobson's conception, an essential component of a speaker's linguistic competence, of his knowledge of his language, is his mastery of its phonological system, of its repertory of phonemes. Sound change, therefore, is for Jakobson primarily change in the repertory of phonemes. In this Jakobson dissents sharply from Saussure, for whom change was a phenomenon of parole and hence essentially outside the purview of systematic study. In the earliest paper printed in Volume I of Jakobson's Selected Writings, a translation of a "brief extract" from "a paper delivered in the Prague Linguistic Circle, January 13, 1927," we read: "F. de Saussure and his school broke a new trail in static linguistics, but as regards the field of language history they remained in the neo-grammarian rut. Saussure's teaching that sound changes are destructive factors, fortuituous and blind, limits the active role of the customary linguistic pattern as an orderly system. This antinomy between synchronic and diachronic linguistic studies should be overcome by a. transformation of historical phonetics into the history of the phonemic system. In other words, phonetic changes must be analyzed in relation to the phonemic system which undergoes these mutations." 16 Jakobson's great monograph Remarques sur l'évolution phonologique du russe comparée à celle des autres langues slaves17 is above all an attempt to defend and illustrate these new theses by following the evolution of the successive phonological systems which ultimately gave us the system of present-day literary Russian. I shall exhibit here the methodology of the book and illustrate it by reviewing a few examples cited by Jakobson. The book begins with a discussion of the so-called First and Second (Baudouin de Courtenay) Palatalizations of the Slavic velars. According to Jakobson, the First Palatalization resulted in the replacement of the velars [ k g x ] by the palatals [ c z s ] in position before front vowels, whereas the Second Palatalization replaced velars by phonetically indeterminate consonants (Jakobson represents these as ki gi xi) in the environment after long or short [i] when followed by long or short [o] or short [u]. These newly created sounds are, at this stage, not new phonemes, but positional variants of the velar phonemes in the specified environments. Subsequent to

128

MORRIS HALLE

the introduction of these new sounds, which are all marked by a palatal (i.e., front) articulation, the language is subject to a strong tendency to establish uniformity in its syllables, so that these will consist exclusively of back (velar) or of front (palatal) phonemes. This "syllabic synharmonism" leads to the fronting of vowels after the newly created [ki gi xi], for these were phonetically front. The result of this change is that the two newly created types of consonants, the [c 1 s] of the' First Palatalization and the [ki gi xi] of the Second Palatalization are in contrast before front vowels. Neither of the two classes of consonants contrasts with velars, however, since velars never appear before front vowels. The question then is whether [c 1 S] or [ki gi xi] are to be viewed as the positional variants of the velars. Jakobson decides this question in the following manner: "In view of the fact that k-ki, g-gi, x-xi are found in grammatical alternations more frequently than k-£, g-z, x-§, it was 5 z s that as a result of the revision of phonological values assumed the role of independent phonemes, while ki gi xi continued to be valued as extra-grammatical variants of k g x."18 The basic mechanism of sound change is thus the addition of rules such as the First and Second Palatalization of Slavic in the example just discussed. These rules affect classes of phonemes and hence their addition brings about change in the pronunciation not of single words but rather of the entire stock of words known to a speaker. As a result of the addition of a rule, new patterns of contrasts may arise in the language. In the case under discussion, [51S] contrasted distinctively with [ki gi xi] in position before front vowels as soon as the Second Palatalization (rule) came into the language. Once such a new contrast appears, the language responds by restructuring the phonological system.19 This picture emerges with especial clarity in Jakobson's account of the effects that the dropping of the so-called "weak jers" had on the phonological system of Slavic languages. As is well known, the "jers" - i.e., the short high vowels [i] and [u] - were dropped in certain positions, e.g., word-finally. According to Jakobson this phenomenon originated in rapid speech but ultimately was generalized to all styles of speech. In those Slavic languages where syllabic synharmonism - the tendency to limit the composition of syllables to phonemes of a single type (front or back) - had been carried out to the point where before front vowels there were only front (or "soft") consonants and where back (or "hard") consonants were to be found only elsewhere, the dropping of "weak" jers produced contrasts of "hard" and "soft" consonants. Originally kladu-klad,i were distinguished by the fact that the second syllable in the former was "hard," whereas in the latter it was "soft." The contrast between d and d, is nondistinctive since

THE MODERN STUDY OF SPEECH SOUNDS

129

the language is subject to a rule of syllabic synharmonism. The situation changes radically as soon as word-final jers are deleted because the language now no longer has syllabic synharmonism as shown by such newly monosyllabic words as klad-klad,. Once syllabic synharmonism is destroyed the distinction between "soft" and "hard" consonants takes its place. "The emergence of the opposition 'soft consonants - hard consonants' renders valueless the opposition of soft and hard sequences taken as units." 20 In his letter of 19 September 1928 Trubetzkoy wrote to Jakobson: "I have been putting together and comparing all vowel systems that I know by heart (34 in toto)... The results are most interesting. All systems reduce to a small number of types and can always be represented by symmetric schemata (triangles, parallel rows, etc.)." What Trubetzkoy had noticed here (these results were later published in Trubetzkoy's Zur allgemeinen Theorie der phonologischen Vokalsysteme,21 and also incorporated in his Grundziige der Phonologie) was that vowel systems of the most varied sorts not only exhibit the same basic phonetic contrasts (essentially, front-back and tongue height), but also that these contrasts in a very obvious fashion underlie the internal organization of each system. This was an important insight since at this point both Trubetzkoy and Jakobson saw the substantive content of the oppositions as being language particular, and only the structure of the opposition (correlative, disjoint, etc.) was regarded as a linguistic universal. In his "Retrospect" to Selected Writings I 2 2 Jakobson comments that Trubetzkoy's paper on vowel systems "came close to reducing the vocalism to a few binary features. It was gradually shown that each of these oppositions was utilized in some of the extant species of the so-called vowel harmony which lays bare the dichotomous structure of all vocalic attributes and displays their operational autonomy with particular clarity." By the middle of the 1930's Jakobson had come to the conclusion that all features are essentially binary. He first expressed these ideas in the important paper that he presented to the Third International Congress of Phonetic Sciences in 1938, "Observations sur le classement phonologique des consonnes." Jakobson begins this paper by declaring that in order to identify ( = specify) the phonemes of a language it is necessary to decompose them into their constitutive phonetic features. This was a great step forward, for in effect Jakobson declared here the primacy of the feature over the phoneme. 23 Whereas previously phonemes had been thought of as further undecomposable entities which could be characterized with the help of features, much as chemical elements were once thought to be characterized with the help of such properties as valence, atomic weight,

130

MORRIS HALLE

etc., Jakobson proposes here instead that phonemes are nothing but complexes of features, much as chemical atoms are now seen as specific configurations of protons, electrons, etc. Having thus placed the features at the center of phonological discussions, Jakobson proceeds one step further. He proposes that all features reflect binary oppositions between entities. He remarks first that this is almost self-evident with regard to the vowel system of a language such as Turkish which consists of eight phonemes exemplifying three binary features. These have traditionally been thought to be open-close, palatal-velar, rounded-unrounded. Jakobson believes that this is too narrowly phonetic a view of what is at issue here. Instead he invokes acoustic evidence which, in his opinion, shows that the open-close dimension is a special implementation of the property of perceptibility.

"In contrast to the closed vowels,

open vowels have, from an acoustic point of view, greater perceptibility and a clearer sound." 2 4 The differences between palatal and velar vowels, and between rounded and unrounded vowels, on the other hand, are related to a difference in pitch. Jakobson proposes that velar vs. palatal vowels should be distinguished as "grave" vs. "acute" sounds. A different type of pitch difference distinguishes rounded from unrounded vowels. It should be noted at once that in claiming that all features reflect binary oppositions Jakobson does not insist that all oppositions are of a single abstract structure. Instead he believes that some oppositions are contradictories, i.e., they contrast entities which possess the given property to those that do not, e.g., long vowels contrasting with vowels without length; while others are contraries, i.e., they contrast entities which, while possessing the given property, exhibit it to a maximal, respectively, to a minimal degree, e.g., acute and grave vowels (the former being maximally and the latter being minimally high pitched). Having shown that vowel systems can be characterized by a set of binary features, Jakobson turns next to the consonants. He notes that most consonantal features such as voiced-voiceless, nasal-oral, palatalized-plain, are self-evidently binary. The major non-binary feature in the consonantal system is the point of articulation, which happens also to be central to every framework that has ever been proposed. Jakobson, therefore, attempts to show that this deeply entrenched analysis can be replaced by a more adequate one employing binary features. Jakobson first proposes to distinguish velar and palatal consonants from labials and dentals by the fact that the former are more perceptible, louder than the latter. On the other hand, he observes that velars are lower pitched than palatals, and that the same pitch relations hold for

THE MODERN STUDY OF SPEECH SOUNDS

131

labials vs. dentals. This was a tremendously important move because it established not only that the major points of articulation can be characterized by means of two binary features but also that the two features involved - perceptibility and pitch - are the same as those found in the vowels. Jakobson was thus able to overcome one of the most unintuitive aspects of phonetic frameworks that were then in wide use (and continue in wide use to this day). These frameworks - cf., e.g., Bell's system - utilize separate sets of features for vowels and consonants. For some reason, with the exception of some early Hindu grammarians, no one before Jakobson appears to have been struck by the implausibility of such an arrangement, for it implies that different mechanisms are involved in the production of vowels than in the production of consonants, and yet one set of articulators is involved in the production of both types of sound, and an analogous case for a single mechanism can also be made with regard to perception. The four classes of consonants established by means of the two features pitch and perceptibility (later renamed gravity and compactness) do not suffice to characterize all consonantal points of articulation. Jakobson notes that additional distinctions divide each of the four major classes into two subclasses: "Thus, we distinguish the linguo-dentals from the hissing sounds, the palatals, properly speaking, from the hushing sounds, the bilabials from the labio-dentals, the velars, properly speaking, from the uvulars. Ordinarily these consonants have been ordered linearly according to the region of their articulation, although phonetic descriptions have allowed us to observe on numerous occasions that from this point of view the delimitation of the series in question is hardly possible. What is then the specific difference that determined these subdivisions?"25 Jakobson's answer is that the difference is due to the feature of "stridency": "Marked friction of the exhaled air eliciting an edge tone (the Schneidenton of Stumpf) contrasts the hissing, hushing, labio-dental and uvular sounds, in one word the strident consonants, as against their 'partners' listed above, which can be termed mellow consonants. A supplementary obstacle participating in this friction distinguishes the articulation of the strident Constrictives from that of the mellow... The same intense friction distinguishes the strident and mellow occlusives. The former are affricates...while the latter are proper occlusives..."26 Jakobson (or perhaps more correctly Stumpf) appears to be in error in referring to edge tones here. There are no edge tones in the strident consonants, except inadvertently produced whistles in articulating [s]. But

132

MORRIS HALLE

he is entirely right in viewing the difference between strident and nonstrident (mellow) sounds as being produced by the presence in the former of a special obstacle against which is directed the air stream. It is this extra obstacle that allows the speaker to create what is technically known as "turbulence at a boundary" which is a plausible acoustic correlate of strident sounds. The system of features just outlined, with some obvious extensions as well as modifications and improvements, has remained the basis of all of Jakobson's subsequent explorations in phonological theory. Although the Jakobsonian feature system has received extensive discussion in the literature and has affected almost all subsequent research, it has been widely misunderstood in one respect: because of the prominence given by Jakobson to acoustic considerations in setting up the distinctive features and because attention to acoustic properties was rather unusual in phonological discussions, Jakobsonian features are frequently referred to in the literature as acoustic features. Jakobson never intended to limit the features in this manner, rather he consistently held to the view that all features have manifestations on all relevant levels - articulatory, acoustic, perceptual: "The specification of distinctive oppositions may be made with respect to any stage of the speech event from articulation to perception and decoding, on the sole condition that the invariants of any antecedent stage be selected and correlated in terms of the subsequent stages, given the self-evident fact that we speak in order to be heard and need to be heard in order to be understood." 27 Jakobson regards the distinctive features as the fundamental building blocks of which all speech sounds in all languages are constructed. He is, therefore, constantly at pains to show that the role of the features is not limited to the characterization of the phonetic attributes of speech sounds but goes well beyond it. He shows that features serve to characterize the acquisition of language by children, the loss of language in aphasia, the distribution of speech sounds in the languages of the world, the restrictions of sound sequences imposed by metrical and other conventions of poetry, and, last but not least, the effects of diachronic sound laws as well as of synchronic phonological rules. He views the above not only as direct confirmation of the correctness of the postulated features, but also as a means for gaining insight into the functioning and nature of the features themselves. Perhaps the most extensive of Jakobson's efforts in this direction is Kindersprache, Aphasie und allgemeine Lautgesetze,2S which, judging by its many translations, must surely be the most popular of Jakobson's

THE MODERN STUDY OF SPEECH SOUNDS

133

works. In this book Jakobson attempts to show that the acquisition of speech sounds by children proceeds in the reverse order of the loss of control over speech sounds by aphasics. Facts of this sort had until that time rarely been taken into account in discussions of linguistic issues. Moreover, Jakobson attempted to show that these processes are governed by certain broad principles that are best captured in terms of distinctive features, and that the same principles provide an explanation for the distribution of sounds in phonological systems of different languages. Jakobson first establishes the fact that phonic contrasts are acquired in the same order by children of the most varied linguistic background: "Regardless of whether it deals with French or Scandinavian children... every careful description confirms for us the remarkable fact that for a series of sound acquisitions the relative time sequence remains everywhere and always the same." 29 In aphasics the loss of speech sounds shows the reverse order of the process of acquisition in children: "The dismantlement of the linguistic sound repertory in the case of aphasics provides an exact mirror image of the process of construction of child language. Thus, for example, the acquisition of the liquids r and / is a fairly late acquisition of child language and... one of the earliest and most frequent losses in aphasic disturbances." 30 This, however, is not all. The same principles appear to govern the distribution of particular phonetic contrasts in the languages of the world: "Contrasts which appear rarely in the languages of the world belong among the latest phonic acquisitions of the child." 31 "The number of languages with a single liquid (be it I or r) is inordinately large, and in this connection Benveniste points out quite rightly that the child is satisfied with a single liquid for quite a long time and acquires the other liquid only as one of the last speech sounds" (in its repertory - M.H.). 32 Maximally contrasting entities are first in the order of acquisition of speech sounds, which in Jakobson's view proceeds from simple unarticulated structures to complexly differentiated ones. Jakobson believes that the maximal contrast among speech sounds is represented by that of the open vowel a and the labial occlusive p: "The earliest stage of child language begins with a clear differentiation and delimitation of vowel and consonant... From the motor point of view these two classes of speech sound contrast as blockage vs. opening. The optimal opening is attained in the wide [a] vowel. The extreme contrast to the [a] vowel is provided by the occlusives, and among the occlusives it is the labials that block off the entire oral cavity." 33

134

MORRIS HALLE

Once this maximal contrast is acquired by the child, the process of articulation and differentiation can begin. The first development appears in the consonant system; the new variety that appears is the nasal consonant, a structure which combines the characteristic oral cavity closure of the consonants with a simultaneous (vowel-like) opening of the vocal tract via the velum. Thus, from a -contrast between open and shut vocal tract there develops a more refined contrast between a totally unobstructed passage from the glottis to the ambient air vs. one that is partially or fully obstructed ; i.e., a simple gross contrast evolves into a more complicated and refined one. To explain the next step in the universal order of sound acquisition Jakobson reviews certain findings of such German psychologists as Koehler and especially Stumpf, whose psychoacoustic investigations of the speech sounds, Die Sprachlaute (Berlin, 1926), exercised a profound influence on Jakobson's thought. These researchers believed that there were essential parallels between the'perception of colors and that of speech sounds. They proposed that certain attributes of speech sounds corresponded to the color attributes of saturation (Farbigkeit) and lightness (Helligkeit). In particular, they identified saturation with sonority (Schallfuelle) and lightness with subjective pitch (Tonhoehe). They noted that these two dimensions of the color space are not totally orthogonal. Highly saturated colors admit of no (or only very limited) distinctions in lightness, and the same limitation appears to hold for speech sounds: maximally open sounds such as the low vowels allow only for limited distinction in lightness (pitch), whereas minimally open sounds such as the occlusives readily allow for pitch distinction, e.g., distinctions between labial (dark) and dental (light) consonants are found in all consonantal systems: "Much like visual sensations speech sounds are, on the one hand, light or dark, and, on the other hand, saturated and unsaturated. With decreasing saturation (sonority) the contrast of lightness and darkness gains in importance. The wider the vowels, the more saturated they are and the greater is their distance from the lightness contrast. Of all vowels, a possesses the greatest saturation and is least affected by the contrast light/dark, while the narrowest vowels, which are particularly subject to this contrast, exhibit minimal saturation." 34 Since at this stage of the child's acquisition of speech the system is composed of only one vowel whose primary distinctive property is that of being maximally saturated, we do not expect lightness distinctions to develop among the vowels; we rather expect to see these developing in the consonant system, since consonants are minimally saturated and hence

THE MODERN STUDY OF SPEECH SOUNDS

135

maximally susceptible to distinctions in lightness. The light counterpart to the labial occlusive is the dental. Thus, at this early stage of speech development we find a system that contains a single vowel a and four types of consonants distinguished by the features nasal-nonnasal and light (dental)-dark (labial). Since this path of development is blocked for the vowels, these evolve first by attenuating the openness dimension. In contrast to the maximally open and saturated a there evolves a relatively closed and somewhat less saturated variety of vowel, the so-called high vowel. But since the vowels of the latter kind are less saturated than a, they are also less resistant to the development of lightness contrasts. As a result we find quite commonly three-vowel systems of the a-i-u variety. Following Stumpf, Jakobson identifies sonority (Schallfuelle) as the psychoacoustic correlate of saturation. Since a is normally louder, more audible than either i or u, sonority is ultimately equated with sound intensity and loudness. Given this definition the consonantal counterpart of the saturated a is the velar k because according to the best available psychoacoustic data (cf. Fletcher, Speech and Hearing [1928]), k is more intense, louder than p or t. The development of this (relatively) saturated consonant is the next stage in the process of language acquisition. Once this is completed the vowel system consisting of i-u-a and the consonant system consisting of dentals, labials and velars exhibit identical structure: k, a p,u dark

t,i light

saturated unsaturated

Although one cannot fail to admire the ingenuity and daring of Jakobson's constructions and of his theoretical imagination, it is not obvious - at least to this reviewer - that this account of speech sound acquisition is wellfounded. It is to be noted that the order of acquisition of phonemes concerns exclusively their production and specifically does not affect their perception. Jakobson recounts a whole series of anecdotes whose central point is that children who are unable to produce phonetic contrast are perfectly capable of distinguishing the contrasting sounds in speech of others. 35 If the acquisition of phonology concerns primarily the production of sounds and not their perception, it is puzzling what relevance should be attributed to the cited parallels with color perception. One might expect a parallelism between the perception of colors and the perception

136

MORRIS HALLE

of sounds, but it is hardly likely that the perception of color will shed any light on the order in which a child learns to manipulate various muscle groups in order to produce the different phonetic contrasts. It might be worth emphasizing that Jakobson adduces the parallelism between speech sounds and colors in order to buttress his suggestion that both vowel and consonant systems are organized in the fashion illustrated by the graph on the preceding page. Hence, by questioning the relevance of this parallelism between colors and speech sounds we are questioning this particular organization of the speech sounds, and we are leaving unaffected the more fundamental proposal that the same set of features underlies both vowels and consonants. Jakobson has continued to develop and refine his ideas on the distinctive features down to the present. Since I had the good fortune of collaborating closely with Jakobson on much of this research, I feel that it would be inappropriate for me to review it here. I omit, therefore, this aspect of Jakobson's work from consideration and turn to other matters. In the preceding, attention has been focussed almost exclusively on the phonological system: its structure, the changes that it undergoes, the manner in which command over it is acquired by a child and lost by an aphasic. These have been the main topics of discussion. A fluent speaker's command over the phonology of his language, however, crucially involves yet another component: the rules that govern the distribution and actualization of a given phoneme in different contexts. The fact that English voiceless stops are aspirated in pretonic position unless preceded by [s] or that in Slovak long vowels are shortened after syllables with a long vowel in conformity with the so-called "rhythmic law" is information that a fluent speaker of these languages must have and that therefore must be explicitly included in a phonological description of these languages. Up to this point in our discussion we have encountered rules primarily as agents of sound change (see pp. 127-29 above). It hardly needs pointing out that this is not their only role in language. In fact, they can affect change only because they are ah integral component of the synchronic functioning of the language. In view of this, phonological descriptions always include extensive discussions of various rules, and Jakobson's studies are no exception: they devote considerable space to various phonological rules among which one can find some of the most significant and lasting results obtained by Jakobson. To mention but a few, these include the "dissimilative jakan'e" rule characteristic of an important group of Southern Russian dialects,36 the treatment of vowel length and its relationship to diphthongization in standard Slovak,37 the principles of

THE MODERN STUDY OF SPEECH SOUNDS

137

word accentuation in classical Greek, 38 consonantal alternations in Gilyak,39 and the rules of the Russian conjugation. 40 It should be remarked at once that Jakobson's conception of the rules differs in many respects from that pievalent in present day generative phonology. For instance, Jakobson draws a distinction between phonology (or phonemics) proper and morphophonemics, and views most of the rules cited immediately above as belonging to morphophonemics rather than to phonology, on the grounds that they result in the replacement of one phoneme by another, whereas phonology proper should be restricted to a consideration of the sub-phonemic (positional) variants of the phonemes.41 However, there is good reason to suppose that for Jakobson this distinction is considerably less basic than it was for many other phonologists, especially on this side of the Atlantic. In an article published in 1932 Jakobson writes that "Wordphonology inventories the phonemes peculiar to a given language, their interrelations and possible combination . . . A special branch of word phonology is morphological phonology or mor(pho)phonology, which analyzes the phonological structure of the various morphological components of the word." 42 The impression that the distinction between morphonemics and phonology is quite peripheral in Jakobson's scheme of things is further supported by his observation of 1948 that "any . . . comprehensive study of a phonemic pattern inevitably runs into the problem of partial patterns mutually distinguishing and specifying the diverse grammatical categories of a given language. The limit between phonemics proper and the so-called mor(pho)phonemics is more labile. We glide from one to the other imperceptibly."43 Moreover, it should be noted that Jakobson did not take an active part in the debates of the 1960's that raged around the separateness of morphophonemics from the rest of phonology. Although we knew that Jakobson did not fully share our views, those of us who argued against the separation of levels found in Jakobson's practice some of the most striking examples to support our position. I conclude from the preceding that the distinction that Jakobson makes between morphophonemics and phonology proper should be regarded as an expository device and does not reflect the view that morphophonemics is a special level of representation, distinct and separate from phonology, having its own primitive entities and its own principles of organization. Instead, in Jakobson's practice, morphophonemics and phonology both make use of the same set of primitives (the distinctive features) and the same type of rules. For my discussion of Jakobson's phonological rules I have tried to choose instances where rules serve not only to characterize an interesting

138

MORRIS HALLE

body of linguistic data but where they also support theoretical proposals made by Jakobson. The abstract structure of these arguments is usually of the following form: given a particular theoretical proposal the configuration of facts under discussion is highly probable, whereas in the absence of the theoretical proposal that observed configuration of facts would have to be regarded as a mere accident. Since it is prima facie implausible that the configuration is accidental, its existence provides evidence in support of the theoretical proposal. This form of argumentation, though standard in most science, was long regarded as questionable by linguists, many of whom felt that theoretical proposals and constructs could be justified only if it were shown that they were discoverable in the data by following certain procedures laid down in advance. In using the type of circumstantial argument just outlined, Jakobson was thus going against a very powerful current in the field. He had the good fortune to see the current reverse direction and his own practice fully vindicated. One instance where Jakobson made use of the type of argument just sketched has already been mentioned. It was noted above that to support his proposal that gravity is a binary feature in vowels Jakobson adduced the vowel harmony rule of Turkish. A s is well known, in this language the vowels in a word must all be either grave or acute; a mixture of grave and acute vowels in a word is forbidden. The language thus treats all vowels as if they belonged to one of two mutually exclusive classes. This is precisely what one would expect to find if features are binary. If gravity were a multi-valued feature the binary treatment of the vowels by the language would have to be regarded as purely accidental. Since vowel harmony of the Turkish type is widespread among the Uralic-Altaic languages, it could not plausibly be regarded as an accident. Hence the facts just discussed must be viewed as evidence supporting the proposition that the feature grave-acute is binary. The same type of circumstantial argument is used by Jakobson in support of his suggestion that the features gravity and compactness characterize not only the vowels, but also the consonants. He cites the fact that in Rumanian [k] became [p] before [s,t] and observes that with the help of the two features the process can be shown to be an instance of partial assimilation; i.e., a type of process that is well attested in many languages. In feature terms the process would be described as one where compact stop sounds become diffuse (noncompact) "sans perdre sa gravité" if they precede diffuse acute obstruents. 44 A s an additional bit of evidence in favor of the features gravity and compactness, Jakobson cites the case of certain Czech dialects, where

THE MODERN STUDY OF SPEECH SOUNDS

139

[p b m] changed to [t d n] before [i], or, in the new feature terminology, diffuse consonants become acute if followed by an acute diffuse vowel. What is of special interest here is that acuteness is, as it were, transmitted from vowel to consonant, a fact that would tend to support Jakobson's suggestion that grave-acute is a feature of both vowels and consonants. The last example I have chosen to discuss are the rules of the "Russian Conjugation." This paper is among Jakobson's most widely quoted works because it deals with a very crucial phenomenon in a widely studied language. For our purposes the important point is that the paper attempts to establish a clear distinction between what in more recent studies would be described as surface and underlying representations and to relate the two by means of a set of explicit rules of considerable complexity. This point is brought out clearly, though somewhat obliquely, in the epigraph which appears in the version printed in the second volume of Jakobson's Selected Writings (it does not appear in the original version of the paper in the 1948 volume of the journal Word, but is referred to in a footnote). 4 5 The epigraph reproduces the following passage from Bloomfield's Language: We have seen that when forms are partially similar, there may be a question as to which we had better take as the underlying form, and that the structure of the language may decide this question for us, since, taking it one way, we get an unduly complicated description, and taking it the other way, a relatively simple one. The major problem of the Russian conjugation, and indeed of the conjugation in all Slavic languages, stems from the fact that for many verbs there are two distinct stems, one for the infinitive and past tense forms, the other for the present tense and imperative forms. The difficulty is to decide which of the two stems, if any, is basic. What all approaches prior to Jakobson have in common is the supposition that if a single stem is to be postulated as basic it must be the same s t e m - i . e . , infinitive or present stem - for all verbs. As a rereading of the passage quoted above shows, this is not the procedure recommended by Bloomfield, nor is it followed by Jakobson. Instead, in line with Bloomfield's recommendation, Jakobson chooses as basic the stem that ensures the simplest description in each case without regard to the fact that for one verb - e.g., stojat' 'to stand' - this is the infinitive stem, whereas for another verb - e.g., citaf 'to read' - this is the present tense stem. It turns out that when the two stems differ in length, the longer must be chosen as basic, or as full-stem, to use Jakobson's terminology. Moreover, "if certain phonemic constituents of the given full-stem as compared with cognate forms appear in different alternants, we take as basic the alternant which appears in a position where the other

140

MORRIS HALLE

alternant too would be admissible." 46 Once the proper choice of the basic full-stem has been made with the help of these principles, a number of simple rules permit us to derive the correct actualization of each verbal form. The rules that constitute the heart of Jakobson's description are the following two truncation rules: "Open full-stems (i.e., stems ending with a vowel - M.H.) remain intact before a consonantal desinence and lose their final phoneme before a vocalic desinence." 47 "Full-stems in j v n m drop their terminal phoneme before a consonantal desinence." 48 The effects of these rules are illustrated below: Open stems before consonantal desinence: stoj + a +1' 'to stand' Open stem before vocalic desinence: stoj + a + u - > s t o j + u 'I stand' Stem in j before consonantal desinence: c i t + a j + t ' - H i i t + a + t ' 'to read' Stem in j before vocalic desinence: c i t + a j + u 'I read' The two truncation rules clearly are part of a fluent Russian speaker's knowledge of his language; he uses them to produce the appropriate conjugational forms of newly coined verbs. 49 It is necessary, therefore, to include these rules in a complete scientific description of the language. 50 For some reason this discovery has yet to be incorporated in all standard text books of Russian; e.g., the most recent large-scale grammar of the Russian language, Grammatika sovremennogo russkogo literaturnogo jazyka,51 does not utilize Jakobson's epoch-making discovery in its treatment of the verbal conjugation. It is conceivable that one of the causes for this neglect is the feeling of disbelief on the part of many linguists that any psychological reality hides behind a solution such as Jakobson's which implicitly assumes that the production or analysis of a given form involves an abstract underlying string as well as a process of calculation utilizing the rules given above. I can testify that when I first read "Russian Conjugation" in the summer of 1949 this aspect of its description was a major stumbling block for me. I overcame it by suspending disbelief long enough to work out part of the conjugation of other Slavic languages and discovered that with some straightforward modifications Jakobson's approach held there, too. This discovery convinced me, and I strongly recommend the same procedure to anyone who has doubts about the approach on theoretical

THE MODERN STUDY OF SPEECH SOUNDS

141

grounds. In an empirical science, results like those just mentioned cannot be rejected on theoretical grounds; rather they provide the strongest argument possible for a re-examination of the theoretical grounds that appeared to exclude them. In my own case this re-examination ultimately led to what has since become known as generative phonology. But this, as Jakobson would say, is an interesting autobiographical remark which has only indirect bearing on the subject under discussion. The Massachusetts

Institute of

Technology

NOTES * This chapter was originally written for Sound, Sign and Meaning: Quinquagenary of the Prague Linguistic Circle, ed. Ladislav Matejka (Ann Arbor: Michigan Slavic Publications, 1976). It is reprinted here with the editor's permission. This work was supported in part by Grant 2P01MH13390-09 from the National Institutes of Mental Health. Except where specifically indicated otherwise, translations of texts in languages other than English are my own. 1 Roman Jakobson, Selected Writings I (2nd ed., The Hague: Mouton, 1971), p. 631. 2 F. de Saussure, Course in General Linguistics, translated with an introduction by Wade Baskin (New York: McGraw Hill, 1966), p. 119. 3 Ibid. 4 Selected Writings I, p. 21n. 5 Cf. N. S. Trubetzkoy, Grundzuge der Phonologie (Gottingen: Vandenhoek und Ruprecht, 1958), p. 14. 6 It is worth mentioning that Jakobson's concept of phonological system shows many resemblances to the concept of sound pattern developed somewhat earlier by Edward Sapir (see Sapir's celebrated "Sound Patterns in Language," Language 1 [1925], 37-51). 7 Selected Writings I, p. 3. 8 Selected Writings I, p. 8. 9 Selected Writings I, p. 9; emphasis supplied. 10 Selected Writings I, pp. 144-201. 11 Selected Writings I, pp. 151-52. 12 This relativistic character of the differentiating signals plays a central role in Jakobson's conception of the phoneme and permits him to avoid certain, though not all, problems posed by what in the American linguistic literature was to become known as "total overlap." Thus, in a rejoinder to one of his American critics, Jakobson notes that Danish pretonic d and posttonic d "which both are phonetically similar.. .nevertheless represent two different phonemes" (Selected Writings I, p. 436) because in pretonic position d contrasts with the more energetically articulated t, while in posttonic position d contrasts with the less energetically articulated opHHHa H Jiioôas Meïa(j)opa HMeeT MeTOHHMHiecKHÎî OTTCHOK"; 2 8 3Ta MbicJib HJIÜK)CTpnpyeTCH HM Ha MaTepHajie pyccKHX CBaAeÔHbix neceH. P . O . -flKoôcoHy NPHHAWIEACHT pa3Öop C STOH TOHKH 3PEHHA pyccKoro (J)OJibKjTopHoro TeKCTa, npHHaAJiexcamero K jiyôoHHoiî TpaAHUHH noBecTH H necHe o OMe H EpeMe, rae "o6a 3JiononyHHbix öpaTa cJiy«aT K)M0pHCTHHecK0H MOTHBHPOBKOH AJia ueJiH napHbix (J)pa3, napoAHpyiom H X napajiJiejiH3M, THIIHHHBIH ,zuia pyccKOH HapoAHoft no33HH, oÔHaacaioMHX ero njieoHa3Mbi H AAIOMNX MHHMopa3JiHHHTejibHyio, a B ACHCTBHTejibHOCTH TaBTOJiorHHecKyio xapaKTepHCTHKy Aßyx rope-öoraTbipeö nyTeM conocTaBJieHHa CHHOHHMHHCCKHX BbipaxceHHH HJIH »ce napamiejibHbix ccbiJioK Ha TecHO CMexcHbie H 6JIH3KO cxoacHe HBjreHH«".29 K a K AAA coôcTBeHHO (J)OJIBKJIOPHCTHHECKORO aHajiH3a 3TOTO TeKCTa H ero BapnaHTOB (npoAOJixceHHoro, co CCHJIKOH Ha paôoTy P . O . ÜKoöcoHa, n. T . B CBETE BBEAEHHORO HM PA3RPAHHHEHHFL

BKJIAfl B c i o x c e T a

OTfleJIbHblX MOTHBOB, TO Ha n p H M e p e

npH

CeBepHO-

p y c c K o ü 6 a j u i a a b i " B a c H j i H i í H Co 3ByKOBbIX H K0Mn03HIIH0HHbIX npHeMOB, THIIOB napajUiejIH3Ma H flpyrHX o6pa3Hbix cpe^cTB. 0 6 m n e cTpyKTypHbie oco6chhocth pyccKoro h K»KHOC.JiaBHHCKoro snoca, KOTOpbie aaBHo nopa3HJiH npoHHuaTeJibHoro MnKjiomHHa, kbho noKa3biBaK)T ,HBe BemH: hto RO pacna^a cjiaBHHCKoñ oóiiíhocth y cnaBAH yace 6biJia BbicoKopa3BHTaa snHHecKas TexHHKa h hto 3a Heñ cjieflOB&na HenpepbiBHaa TbicaneJieTHasi TpaziHUHH.39 Hepe3 HecKOJibKO jieT nocjie 0ny6jiHK0BaHHH uHTHpoBaHHoñ paóoTbi P. O. ÜKo6coHa Tpa/jnuHx, HaMeneHHaH b pa6oTe MHKjioniHia 06 H36HpaTejibHbix cpe^cTBax cjiaBaHCKoro snoca h aKTyajiH3HpoBaHHaa P. O. ííko6cohom, 6biJia noAXBaneHa h npoflOjraceHa b Tpy^e I I . I \ BoraTbipeBa "HeKOTopbie 3aaaHH cpaBHHTejibHoro H3yneHHH snoca cjiaBjmcKHX HapoflOB".40 C flpyroíí CTOpOHbl, BbIBO.Il o HajlHHHH pa3BHTOH SnHHeCKOH TeXHHKH yace b o6mecjiaB»HCKHií nepnofl corjiacyeTc« c pe3yjibTaTaMH HCCJieflOBaHHH B o6jiaCTH peKOHCTpyKUHH OÓlUeHHflOeBponeHCKHX nOSTHHeCKHX (J)opMyji, nacTb kotopmx HenocpeACTBeHHo OTpaxcaercJi b cnaBHHCKOM. HeKOTopbie H3 Taxnx cjjopMyji coaepxcaT mothbm, flaiomne 0CH0BaHHe JITISl peKOHCTpyKUHH UejlblX CKWKeTOB. KaK h b o6jiacTH mctphkh, cjiaBAHCKaa nepH(J)epHH coxpaHaeT apxaH3Mbl H no OTHOUieHHK) K neJIblM MOTHBaM H HX KOM6HHaUHJIM, KOTOpbie CTajiH b ueHTpe BHHMaHHx P. O. ÜKoócoHa b copoKOBbie roAbi. Pe3yjibTaTOM 3Toro HHTepeca 6 h j i uhkji pa6oT o BOCTOHHOcjiaBHHCKOM snoce (KaK 4>OJIbKJIOpHOM, TaK H nHCbMeHHOM) H ero COOTHOUieHHH C K32KHOcjiaBHHCKHM. 3 t h pa6oTbi rpynnHpyioTCH BOKpyr flByx 6ojibuiHX TeM snoc o BcecjiaBe h ero cbjbh c cepócKHM anocoM o ByKe (3Mee OrHeHHOM BojiKe) h " C j i o b o o nojiKy HropeBe" h npHMbiKaioiuHe k HeMy np0H3BeAeHHA ("3a,noHmnHa", noBecTb o 6htbc Ha KyjiHKOBOM nojie). CaMOH KpynHoií Kae Bcero Hpe3BbiHaiÍHO fleTajibHbiií aHajiH3 pyccKoñ 6biJiHHbi o BcecjiaBe c Hcn0Jib30BaHHeM KaK pa3Hbix ee BapnaHTOB, TaK

178

B. B. HBAHOB, B. H. TOnOPOB

h TeKCTOB, KOTopwe He npHHafljieHcaT k 6wjihhc, ho o6i>e,HHHaiOTca c 6bIJIHHOH CBOHM COflepJKaHHeM H OTHaeTH HMeHeM repoa. BblJIHHa aHajiH3HpyeTca no ee ceMaHTHMecKHM 6jioKaM, ripHneM asín Kaxcaoñ pa36npaeMOH nacTH 6mjihhbi bwhbjihiotch cooTBeTCTByiomHe peajiHH, HaxoaHTCH aHajioran b jictoiihch, "Cjiobc o nojiicy HropeBe" h

flpyrnx

TeKCTaX. Ha OCHOBaHHH cpaBHeHHa Bcex 3THX hctohhhkob peKOHCTpyHpyeTca npoo6pa3 repoa H HCxoAHaa cioaceTHaa cxeMa. Ilocjie BbiHJieHeHHa Tex cioxceTHbix y3JiOB, ana kotopwx Haxo/jHTCfl peajibHbiñ npoTOTHn b HCTOpHHeCKOM COÓblTHH, BOCCTaHaBJIHBaeTCa MH^JOJIOrHHeCKa» cxeMa, b ochobhwx nepTax coBnaaaiomaH co cxeMoñ cep6cKoro snoca o 3Mee OrHeHHOM BojiKe (ByKe). AKTyajibHOCTb o6meñ cxeMbi pjix Kaxmoñ H3 TpaflHUHÍI BHflHa H3 Toro, hto B Kaxcfloñ H3 hhx b yKa3aHHyio cxeMy BOBjieKaeTca HCTopHHecKHH MaTepaajr: c repoeM cxeMbi OTOHCAecTBjiajica KHH3b BeeCJiaB riOJlOUKHH ( X I B.) B pyCCKOM 3ÜOCe H ByK rprypOBHH (idespotus Regni Rasciae, XV b.) b cep6cKOM snoce. 4 2 OcHOBOH AJI a OTOXCfleCTBJieHHH 3THX 06pa30B HBJiaeTCH MOTHB OÓOpOTHHHecTBa, cnoco6HocTH npeBpameHHJi b BOJiKa hjih b flpyroro «HBOTHoro, CB5I3aHHOH C flpyrHMH HyfleCHbIMH 0C06eHH0CTHMH H cnOCOÓHOCTflMH r e p o a : oh poAHTca b copoHKe, Ha HeM ecTb oco6bie 3HaKH, oh HafleneH BceBe^eHHeM. KaxcAbiH H3 sthx aTpHÓyTOB P. O. üko6coh paccMaTpHBaeT Ha IUHpOKOM 4>OHe (J)0JlbKJ10pHbIX H 3THOrpa(J)HHeCKHX aaHHblX, CBH3aHHbix c cooTBeTCTByiomHMH noBepbHMH y cjiaBaH, npHBjieKaa b CBa3H c sthm uejiwíí p a a a p y r n x tckctob, b nacTHocTH, Teiccra pa3Hbix cjiaBaHCKHX TpaflHUHH, OTHOCaiUHeCa K pOXCfleHHK) B copoHKe H K OÓOpOTHHHecTBy-BypaajiaHecTBy. 3anaAH0CJiaBaHCKHe aaHHbie, Ka3ajiocb 6bi He OTHOcamHeca Henocpe/iCTBeHHO K AaHHOMy snHHecKOMy cioaceTy, TeM He MeHee Taicace OKa3biBaK)Tca BaxHbiMH fljia AOKa3aTejibCTBa HajiHMHa CBH3H flByX MOTHBOB - poXCfleHHa B COpOHKe H OÓOpOTHHHeCTBa (cp. KamyScKHH mothb: poflHBUiHHca b pyóamice eTaHOBHTea BaMnHpoM, a Taioice cBa3b aMyjieTa-Hay3a c MeKaHTponHeií y sanaAHbix h boctohhwx CJiaBaH). riosTOMy BoccTaHOBjieHHe o6mecjiaBaHCKoro Teiccra o KHa3eo6opoTHe 0CH0BaH0 Ha mnpoKOM xpyre (j)OJibKjiopHbix h STHorpacJjHHeCKHX TeKCTOB H3 BCeX OCHOBHblX CJiaBaHCKHX Tpa/JHUHH. BMecTe c TeM P. O. üko6coh yKa3biBaeT h cymecTBeHHbie HHAoeBponeñCKHe aHajiorHH

ana

peKOHCTpyHpoBaHHoñ

hm cjiaBaHCKoií

cxeMbi.

OcHOBbiBaacb Ha pe3yjibTaTax aHajiH3a rpenecKoro MH(J)a o /JojiOHeBOjiKe b Tpyae >KepHe, P. O. üko6coh ycTaHaBJiHBaeT, hto b HHAoeBponeííckhx HapoflHbix TpaflHHHax "rocnoACTByeT flpaMaTHiecKoe

nocToaHHaa aHTHHOMHH,

HanpaxceHHe MOKAy AByMa npoTHBonojio»CHbiMH, ho

n0CT0»HH0 nepexoAamHMH flpyr b a p y r a , pojiaMH BOJiKa" - npecjieAye-

BKJIAFL B OOJIBKJIOPHBLE MCCJLEFLOBAHH^

179

M o r o H NPECJIE/JOBATEJIA.43 3 T H BHBOAM P . O . ^KOSCOHA NPEACTABJUIIOT H H T e p e C B HeCKOJIbKHX OTHOIIieHHflX. Bo-nepBbix,

3/iecb

BnepBbie

6tiJia

cmie, CIX-CX (1916), 102-07. 5 ' Selected Writings, IV, 613. 6 "Linguistics and Poetics", Style in Language, pefl. T. A. Sebeok (New York, 1960), 374. 7 "The Kernel of Comparative Slavic Literature", Harvard Slavic Studies, I (1953), 8-9. 8 "The Kernel of Comparative Slavic Literature", 19-21.

BKJIAfl B «DOJIbKJIOPHEIE H C d l E f l O B A H H Ä 9

183

"Slavic Epic Verse" (1952), Selected Writings, IV, 456-58. n . r . EoraTbipeB, "BWKPHKH pasHoenmeoB H 6POAi nbiuiopHCTHKH H jiHTepaTypoBeaeHH«" (1931), Selected Writings, IV, 18. 22 P. £KO6COH, "Grammatical Parallelism and Its Russian Facet", Language, 42, No. 2 (1966), 399-429. 23 Linguistics and Poetics, 351. 24 TaM ace, 369. 25 TaM ace, 355. 26 P. JIKO6COH, "Two Aspects of Language and Two Types of Aphasie Disturbances" (1956), Selected Writings, II, 258. 27 P. .HKOÖCOH, "Aktuelle Aufgaben der Bylinenforschung" (1936), Selected Writings, IV, 62-63. 28 Linguistics and Poetics, 370. 29 "II033HH r p a M M a T H K H H r p a M M a T H K a no33HH", Poetics. Poetyka. IJoamuKa (Warszawa, 1961), 401. 30 II. r . EoraTbipeB, "MMnpoBH3ai(HH h HopMbi xyaoacecTBeHHbix npneMOB Ha MaTepnajie noBecreft XVIII B., Haannceft Ha ny6oiHbix KapTHHKax, CKasoK, H necHH o EpeMe H ®OMe", To Honor Roman Jakobson, I (The Hague: Mouton, 1967), 318-34. 31 P. ÄKO6COH, "Subliminal Verbal Patterning in Poetry", 306. 32 "Grammatical Parallelism and Its Russian Facet", 423. 33 "II033HH rpaMMaTHKH H rpaMMaTHKa n033HH", 401. 34 "Balladic Byliny Recorded in the South Ladoga Basin" (1956), Selected Writings, IV, 494-98. 35 "HoBwtt Tpya ororocJiaBHHCKOMsnoce" (1932), Selected Writings, IV, 41. 36 "Studies in Comparative Slavic Metrics", Oxford Slavonic Papers, III (1952), 21-66; nepeneiaTaHa noa HasBaHHeM "Slavic Epic Verse: Studies in Comparative Metrics", Selected Writings, IV, 414-63. 37 "The Kernel of Comparative Slavic Literature", 20. 38 TaM ace, 21. 39 TaM ace, 31. 10

184

B. B. HBAHOB, B. H. TOIIOPOB

40

IT. r . EoraTbipeB, "HeKOTopwe 3aflaiH cpaBHHTenbHoro H3yHCHim snoca cjiaBHHticcAedoeaHUH no CAaenHCKOMy Mimepamypoeedeniuo u 0o/ibKAopucmuKe. JJoKAadbi coeemcKux yvenbtx Ha IV MexcdynapodHOM cte3de c/iaeucmoe (MocKBa, 1960), 211-51. 41 "The Vseslav Epos", Russian Epic Studies, pea. R. Jakobson and Ernest J. Simmons [= Memoirs of the American Folklore Society, 42 (1947)] (Philadelphia, 1949), 13-86; Taxxce B Selected Writings, IV, 301-68. 42 CM. R. Jakobson with Gojko Ruziôié, "The Serbian Zmaj Ognjeni Vuk and the Russian Vseslav Epos" (1951), Selected Writings, IV, 369-79. 43 "The Vseslav Epos", Selected Writings, IV, 351. 44 TaM ace, npHMenaHHe 166 - no MaTepwajiaM paôoTM Haimopa o,aopa o ITCHXHMecKOM MexaBH3Me jiHKaHTpomiH. 46 "Pojib jiHHrBHCTmecKHx NOKAÎAHHII B cpaBHHTeJibHOit MNaKTbi. MOXCHO CnopHTb C HeKOTOpbIMH erO BbIBOAaMH, yCTapeBUIHMH OÓbflCHeHHHMH Tex HJIH HHblX HBJieHHH. Ho AeTajIbHO OnHCaHHbie e,ady. IIPHHHHOH STOTO H3MEHEHHH 6biJio "oôoômeHHe MOCKOBCKoro (JjoHeTHMecKoro 3aKOHa o noHHaceHHH apTHKyjiauHH H aena6HajiH-

3auHH npeayaapHbix jiaÔHajimoBaHHbix rjiacHbix cpe^Hero no^eMa, nepeHeceHHe 3Toro 3aKOHa TaKxce Ha rjiacHbie B nojioxceHHH nocjie

MHrKHX".28 BaacHO Taicxce 3aMenaHHe P. O. ^Ko6coHa o TOM, HTO "HCTopna CTaporo a nocjie M a r c o c r a " (rCamâK, rCanCù) "aôcojiiOTHO He CBjœaHa c HCTopHeîi yMepeHHoro aKaHba. B ôojibiiiHHCTBe roBopoB c yMepeHHbiM HKaHbCM a M e x a y AByMH MHTKHMH nepeuuio B e, HO B HeKOTopbix roBopax a Meacfly flByMH MarKHMH coxpaHseTca 6e3 H3MeHeHHii, Hanp., B KacHMOBCKOM".29 B Remarques sur l'évolution phonologique du russe P. O. >IKO6COH BHOBB B03BpamaeTca K Bonpocy 0 6 HCTOPHH yMepeHHoro aKaHbH. O H oôpamaeT BHHMaHHe Ha TO, HTO B 0CH0Be yMepeHHoro aKaHba jie»CHT TaKOH THII ceBepHopyccKoro npe/iyAapHoro B0KajiH3Ma, B KOTOPOM o ne HBJIXIOTCH BapwaHTaMH OAHOH (J)OHeMbi: o BbicTynaeT nepea TBep^biM, e nepea MarKHM corjiacHMM. HcHe3HOBeHHe STOÎÎ a, (HJIH

e>i

e f ): c'aAÔ, e'uA'y. CTapoe a Meac^y MSTKHMH corjiacHbiMH Morjio,

no MHeHHio P. O. ilKoôcoHa, eme flo 3Toro npouecca H3MeHHTbca B E H OHO, ecTecTBeHHO, pas^ejiHJio cyzjbôy CTaporo e : n'uniû. roBopax TaKoe a coxpaHajiocb : n'am'û

H o B HeKOTopwx

H T.n. M T o r ^ a OHO npensTCTBo-

Bajio ynpomeHHio npe/iyflapHoro B0KajiH3Ma. B CBJBH C STHM Meacay MHTKHMH MOrjlO np0H30HTH oSoÔLUeHHe HCTOpHHeCKHX (J)OHeM a, e, o B

i

(HJIH

P.

e'). 3 0

O . .HKOÔCOH nepBbift HAMETHJI nyTH 0Ôpa30BaHHa

AKAHBH. A B T O P H

Eepone,

yMepeHHoro

Onuma duaAeKmoAoemecKoù xapmm pyccxoeo H3bma e

NO-BH^HMOMY, T O » E c H H T a j m , KAK

nncaji B. H. CNFLOPOB, HTO

HMeHHO "yMepeHHoe aicaHbe cpeflHepyccKHx roBopoB noKa3biBaeT, HTO HX ocHOBa ceBepHopyccKaa, a HacJioeHHe io>KHopyccKoe". 31 H o "aBTopw

Onuma duaAeKmo/ioemecxoû Kapmu... 0 6

STOM

HHHero He roBopaT, He

oôtacHaa H TOTO, KaK OHH CMOTpaT Ha 06pa30BaHHe yMepeHHoro aKaHba B cpeflHepyccKHx r o B o p a x " . 3 2

JI. JI. KACATKHH

208

B . H. CHflOpOB, nOCBSTHBUIHH K

TeM ace,

P. O.

3TOMy

BOnpOCy

PHfl C T a T e ñ , 3 3 n p H X O f l H T

.HKOÖCOH, O C H O B H H M B b i B O f l a M .

B. H. CnflopoB

ace

H H a 3 B a H H e r o B o p a M c y M e p e H H b i M HKaHbCM H n p o H 3 H o i i i e H H e M u H a

flaji

MecTe P.

HTO H

MancHMH comacHbiMH,

a nepefl

O.

pa3flejiaa

ÜKO6COH,

Ha

HX

Ha KOTopbie

ToponoBCKHH

oöpaTHJi BHHMaHHe H

HOBOPOCCHHCKHH

THnbi.34

P. O.

oTMenaeT

.SKOÔCOH

B

roBope PoraieBCKoä

BOJI.

H

RJIABHBIM

06pa30M B roBopax, npHJieraiomax K Heft, nepeflBHxeeHHe 3aflHenajiaTajibHbix {rri k',

k\

Bnepefl

¿

of the formal logician are not really differences in meaning but merely reflections of the fact that ordinary speakers obey rules of fair play when speaking. Thus, while it would be misleading for me to say Either Caligula was Napoleon's valet or Coolidge was Harding's vice-president, that fact does not make the proposition which the sentence expresses false. The proposition is true (indeed, if you are asked Was Caligula Napoleon's valet or Coolidge Harding's vice-president!, an answer of Yes is appropriate), but it would be misleading to utter that sentence, since the speaker (who we may suppose to know that Coolidge was Hard-

280

JAMES MCCAWLEY

ing's vice-president but Caligula was not Napoleon's valet) could be more informative at a cost of fewer words by simply asserting that Coolidge was Harding's vice-president. Grice's approach opens the door to more semantic constancy than might otherwise be attainable; for example, it makes it possible to drop the distinction between "inclusive or" and "exclusive or" Gordon and Lakoff 46 have proposed an analysis of "indirect speech acts" (such as the use of an interrogative as a request: Would you mind passing the ketchup ?) in which a sentence is taken as having the same meaning whether used in an indirect or a direct speech act and the "conveyed meaning" is treated as resulting from the interaction between the literal meaning and principles of fairness, politeness, etc. in language use.

V. UNIVERSALITY

Finally, the widespread concern among generative grammarians for the formulation of language universals and the analysis of particular phenomena in the light of universals can also be traced back to the influence of Jakobson. In particular, Chomsky's appreciation of the importance of language universals for an explanatory theory of language did not blossom until after his contact with Jakobson and his collaboration with Halle: his early works47 display far less concern with language universals than with developing devices which allow linguistic description to proceed without reference to universals (e.g., his early policy of allowing a "cover symbol" to be assigned to any particular set of phonemes and formulating phonological rules in terms of these "cover symbols" rather than in terms of universal phonological categories). Jakobson's emphasis on the centrality of SUBSTANTIVE universals and IMPLICATIVE universals to theoretical linguistics is clearly visible in much recent generative work on syntactic universals, for example, the research stimulated by Ross48 on implicational relationships between word order and the syntactic process of "gapping" (i.e., deletion of repeated verbs in a coordinate structure), and the work of Keenan and Comrie49 on a universal "accessibility" hierarchy, which systematizes a host of implicational universals about applicability of transformations (e.g., if a language allows relativization of indirect objects, it will also allow relativization of direct objects). University of Chicago

GENERATIVE GRAMMAR

281

NOTES 1

Roman Jakobson, "Russian Conjugation," Word4 (1948), 155-67; Selected Writings II (The Hague: Mouton, 1971), 119-29. 2 M. Halle, The Sound Pattern of Russian (The Hague: Mouton, 1959). 3 For example, Roman Jakobson, "Observations sur le classement phonologique des consonnes" (1938), published in Proceedings of the Third International Congress of Phonetic Sciences (Ghent, 1939), pp. 34-41; also in Selected Writings I (2nd ed., The Hague: Mouton, 1971), 272-79. 4 J. R. Applegate, "Phonological Rules of a Subdialect of English," Word 17 (1961), 186-93. 6 Roman Jakobson, Kindersprache, Aphasie undallgemeineLautgesetze (Uppsala, 1941); Selected Writings I (2nd ed., The Hague: Mouton, 1971), 328-401. 8 D. L. Stampe, "The Acquisition of Phonetic Representation," Papers from the Fifth Regional Meeting, Chicago Linguistic Society (1969), pp. 443-54; idem, "An Essay on Natural Phonology," Ph.D. dissertation (Chicago, 1972); idem, "The Natural History of English Vowels," Papers from the Eighth Regional Meeting, Chicago Linguistic Society (1972), pp. 578-90; P. D. Miller, "Vowel Neutralization and Vowel Reduction," Papers from the Eighth Regional Meeting, Chicago Linguistic Society (1972), pp. 482-89; idem, "Bleaching and Coloring," Papers from the Ninth Regional Meeting, Chicago Linguistic Society (1973), pp. 386-97. 7 See C. Smith, "A Class of Complex Modifiers in English," Language 37 (1961), 342-65. 8 In my own work, I have explicitly accepted a generalized form of premise (2) [see J. D. McCawley, "Acquisition Models as Models of Acquisition" in Proceedings of the 1974 NWA VE Conference (Washington, D.C.: Georgetown University Press, 1977)] but have rejected premise (1): I take transformations to be rules associating surface structures to semantic structures and take "relative clause" to be an element of the child's semantic structures whose surface realizations are learned only gradually, with the simpler realizations (such as prenominal adjectives) learned before the more complex realizations (such as full relative clauses), and with the child at any stage able to form in his mind ideas far more complicated than what he is able to express. 9 Jakobson, "Russian Conjugation," Selected Writings II, 120. 10 Halle, The Sound Pattern of Russian, pp. 22-23. 11 M. Halle, "O russkom sprjazenii," American Contributions to the Fifth International Congress of Slavists (The Hague: Mouton, 1963), pp. 363-82. 12 While Chomsky and Halle [The Sound Pattern of English (New York: Harper, 1968)] adopt an analysis in which the tense vowels change height, about 1961 they were operating with an analysis in which lax vowels changed height, with derivations such as sen + ity —* sen + ity —»• seen + ity. 13 T. M. Lightner, "Segmental Phonology of Modern Standard Russian," Ph.D. dissertation, Massachusetts Institute of Technology (1965). 14 J. D. McCawley, The Phonological Component of a Grammar of Japanese (The Hague: Mouton, 1968). 15 P. Kiparsky, "How Abstract Is Phonology?" mimeograph (1968). 16 T. M. Lightner, "On the Description of Vowel and Consonant Harmony," Word 21 (1965), 244-50. 17 L. Hyman, "How Concrete is Phonology?" Language 46 (1970), 58-76; C. Kisseberth, "On the Abstractness of Phonology: The Evidence from Yawelmani," Papers in Linguistics 1 (1969), 248-82. 18 B. L. Denying, Transformational Grammar as a Theory of Language Acquisition (London and New York: Cambridge University Press, 1973).

282

JAMES MCCAWLEY

19 J. D. McCawley, Review of Derwing, Transformational Grammar as a Theory of Language Acquisition, Canadian Journal of Linguistics 19 (1974), 177-88. 20 Kisseberth, "On the Abstractness of Phonology." 21 R. Skousen, "On Capturing Regularities," Papers from the Eighth Regional Meeting, Chicago Linguistic Society (1972), pp. 567-77. 22 V. Fromkin, "The Non-anomalous Nature of Anomalous Utterances," Language 47 (1971), 27-52. 23 See also Bierwisch, "Fehler-Linguistik," Linguistic Inquiry 1 (1970), 397-414, for application of speech-error data in testing the psychological reality of proposed syntactic analyses. 24 Chomsky and Halle, The Sound Pattern of English. 25 Ibid. 28 J. D. McCawley, "The Role of a System of Phonological Features in a Theory of Language" in Phonological Theory, ed. V. Makkai (New York: Holt, Rinehart, 1972), pp. 522-28. 27 M. Halle, "Phonology in Generative Grammar," Word 18 (1962), 54-72. 28 D. L. Stampe, "On Chapter Nine" in Issues in Phonological Theory, eds. M. Kenstowicz and C. Kisseberth (The Hague: Mouton, 1973), pp. 43-52; J. D. McCawley, Review of Chomsky and Halle, The Sound Pattern of English, International Journal of American Linguistics 40 (1974), 50-88. 29 See S. R. Anderson and W. Browne, "On Keeping Exchange Rules in Czech," Papers in Linguistics 6 (1973), 445-82, for arguments that "exchange rules," one of the more controversial elements of Chomsky and Halle's analysis of English, are correct only when their environment is morphological rather than purely phonological. 30 M. Silverstein, "Hierarchy of Features and Ergativity," mimeograph (1973). 31 N. A. Chomsky, Aspects of the Theory of Syntax (Cambridge, Mass.: M.I.T. Press, 1965). 32 N. A. Chomsky, Syntactic Structures (The Hague: Mouton, 1957). 33 C. J. Fillmore, "The Case for Case" in Universals in Linguistic Theory, eds. E. Bach and R. Harms (New York: Holt, Rinehart and Winston, 1968), pp. 1-88. 34 Roman Jakobson, "Boas' View of Grammatical Meaning" in The Anthropology of Franz Boas: Essays on the Centennial of His Birth (= American Anthropologist LXI, 5, part 2 [1959], Memoir 89), pp. 139-45; also in Selected Writings II, 489-96, especially pp. 490-91. 35 Chomsky, Aspects of the Theory of Syntax. 36 J. J. Katz and P. M. Postal, An Integrated Theory of Linguistic Description (Cambridge, Mass.: M.I.T. Press, 1964). 37 For example, R. Jacobs and P. S. Rosenbaum (eds.), English Transformational Grammar (Boston: Ginn, 1968). 38 Chomsky, Aspects of the Theory of Syntax. Chomsky has since abandoned this policy (cf. "Remarks on Nominalization" in Readings in English Transformational Grammar, eds. R. Jacobs and P. S. Rosenbaum [Boston: Ginn, 1970]) and now treats all "node labels" as sets of feature specifications. 39 J. R. Ross, "On Declarative Sentences" in Readings in English Transformational Grammar, eds. R. Jacobs and P. S. Rosenbaum (Boston: Ginn, 1970). 40 J. R. Ross, "Auxiliaries as Main Verbs," Studies in Philosophical Linguistics 1 (1969), 77-102. 41 Roman Jakobson, "Shifters, Verbal Categories and the Russian Verb," mimeograph (Cambridge, Mass.: Department of Slavic Languages and Literatures, Harvard University, 1956); reprinted in Selected Writings II (The Hague: Mouton, 1971), pp.130-47. "Boas' View of Grammatical Meaning," The Anthropology of Franz Boas. Essays on the

GENERATIVE GRAMMAR

283

Centennial of His Birth (= American Anthropologist, LXI, 5, part 2, Memoir 89), 139-45; also in Selected Writings II (The Hague: Mouton, 1971), 489-96. 42 Jakobson, "Shifters, Verbal Categories and the Russian Verb," Selected Writings II, 135. 43 N. A. Chomsky, Studies on Semantics and Generative Grammar (The Hague: Mouton, 1972); and R. S. Jackendoff, Semantic Interpretation in Generative Grammar (Cambridge, Mass.: M.I.T. Press, 1972). 44 Here, as elsewhere, I use the word "surface" as a purely descriptive term, referring to the " u p p e r " ( = closest to phonetic) end of syntactic derivations, with no hint that the "surface" is a shroud that blinds us to wonders that lie beneath it. The wonders of language are to be found at all depths. 45 H. P. Grice, "Logic and Conversation" in The Logic of Grammar, eds. D. Davidson and G. Harman (Encino, Calif.: Dickenson, 1975). 46 D . Gordon and G . Lakoff, "Conversational Postulates," Papers from the Seventh Regional Meeting, Chicago Linguistic Society (1971), pp. 63-84. 47 N . A. Chomsky, "Morphophonemics of Modern Hebrew," M. A. paper (University of Pennsylvania, 1951); idem, The Logical Structure of Linguistic Theory (New Y o r k : Plenum, 1975; originally written 1955); idem, Syntactic Structures. 48 J. R. Ross, " G a p p i n g and the Order of Constituents" in Progress in Linguistics, eds. M. Bierwisch and K. Heidolph (The Hague: Mouton, 1970). 49 E. Keenan, " N o u n Phrase Accessibility and Universal G r a m m a r , " paper read at Linguistic Society of America meeting (winter, 1972).

3

O C O E E H H O C T H ,

7

P E 3 Y J I L T A T O B

I I P H H I J H n O B

H

1 1

r P A M M A T H H E C K H X

H C C J I E f l O B A H H ñ

P O M A H A

H. A.

Ü K O E C O H A

MEJIbHYK

T p o S x a , ceMepxa... Ty3Î A . C . IlymKHH, "ÜHKOBaa B

flaMa".

OflHOH HeÔOJILLUOH CTaTbe HCB03M03KH0 flaTL p a 3 B e p H y T t I H , OÔOCHO-

BaHHblH H OÔbeKTHBHblH aHajIH3 TOTO O r p O M H O r O BKJiafla, KOTOpblH BHeC P . O . - H k o ô c o h KaK b o 6 m y i o T e o p H i o r p a M M a T H K H , T a K h b

flejio

MaTHHeCKOrO

B

OIIHCaHHa

OHepeat - pyccKoro; pa6oTa.

KOHKpeTHblX

ana

3Toro

CJiaBHHCKHX

noTpeôoBajiacb

Il03T0My mm noñfleM apyrHM nyTeM. A

«3MKOB, 6bi

6ojiee

rpaM-

nepByio

cojiHflHaa

hmchho, mm

npocTO

n e p e H H C j i H M b M a i c c H M a j i b H O cacaTOH aKTopbi 4 > o p M y j i H p y i o T C H b BH^e

nPHHijHnoB,

K O T o p b i e JieacaT, n o H a m e M y m h c h h i o , b o c H O B e h k o ö c o h o b c k h x MaTHHeCKHX HCCJieAOBaHHH, H B

BHfle

PE3yjIbTATOB,

Hccjie/iOBaHHH n p H B e j i H ; p a 3 r p a H H H e H H e AocTaTOHHOH

cTeneHH

ycjiOBHO.

Ilpn

Tex

h

3TOM

K KOTOpbIM

flpymx,

öyaeT

rpaM3TH

pa3yMeeTca,

yaoÔHo

HanaTb

b c

yica3aHHH H a n ô o j i e e BaacHbix O C O B E H H O C T E H r p a M M a T H n e c K H x

imyziHH

P . Ä.;

JumreH-

o f l H a K O n o c K O J i b i c y s t h o c o ô c h h o c t h n p n c y m H Bceìi e r o

cTHHecKOH

fleaTejibHocTH,

a B O B c e He TOJibKO r p a M M a T H H e c K O H , 3 f l e c b m m

CKaaceM o h h x JiHuib HecKOJibKO c j i o b . ( I l o c y m e c T B y , s t o T e M a O T A e j i b H o ü C T a T b H : HTO-TO B p O f l e " P . O . ÜKOÖCOH KaK JIHHrBHCT".)

I. 3 O C O B E H H O C T H T P A M M A T H H E C K H X

HCCJlE/IOBAHHft

P. O . - 3 K O B C O H A A . MaKCHMajibHo

rnnpoKHÜ

H3MK

P.

(KOTopbie

oömeHayHHoii

y

R.

B 3 r j i a A Ha jiHHrBHCTHKy h Ha e c T e c T B e H H b i H

paccMaTpHBaioTca

KapTHHOÜ M H p a

h

c

Bcer^a

apyrHMH

b

tcchoh

c(J>epaMH

cbh3h

c

HejioBenecKoro

286

H. A.

n03HaHHfl) - n p H

MEJILHYK

MaKCHMajIbHO

HeTKOM

BblfleJieHHH

H

C06jIK)AeHHH

C o S c T B e H H M X TpaHHU JIHHrBHCTHKH H H3bIKa. C

o a h o h C T o p o H b i , n o H c a j i y ì i , T p y n H O H a 3 B a T b K a K o r o - H H 6 y f l b JiHHr-

BHCTa, K O T o p b i H b H a u i H flHH T a n »ce HHTeHCHBHO h P . SI.,

yrjiy6jieHHo,

KaK

3 a H H M a j i c a 6 b i CBH3HMH M e a m y JiHHrBHCTHKoft H a p y r H M H H a y K a M H .

JlHHrBHCTHKa H n O S T H K a (Ha3BaHHe H3BCCTH0H C T a T b H P . i l . ) ,

JIHHrBH-

CTHKa h My3biKOBeflCHHe (Ha3BaHHe e r o a p y r o i i C T a T b H ) , JlHHrBHCTHKa h aHTponojiorHS,

JlHHrBHCTHKa

h

Teopna

neperiana

coo6meHHH,

jiHHr-

BHCTHKa H TeOpHH n e p e B O f l a , JIHHTBHCTHKa H M a T e M a T H K a , JlHHrBHCTHKa h

ncHxojiorHH,

JlHHrBHCTHKa

h

ceMHOTHKa - h

MHoroe

CTOHHHO HaXOflHTCH B aKTHBHOH 3 0 H e BHHMEHHfl P . Si. ero

3HaMeHHTbie

MHoroM

pa6oTbi

onpeaejiHJiH

no

jiHHrBHCTHiecKOMy

pasBHTHe

apyroe - no-

H e C J i y H a ì f f l O , HTO

aHajiH3y

a4>a3Hii

aa3HOJiorHH; H e c j i y n a i m o ,

OflHHM H3 n e p B b i x n p o B e j i r n y G o K y i o a H a J i o r n i o

Me»Ay

hto

P.

bo SI.

ecTecTBeHHbiM

H3bIKOM H reHeTHHeCKHM K O f l O M ; 1 HeCJiyHaHHO, HTO HMeHHO OH OTKpblBaJI h j i h 3 a K p b i B a j i o 6 o 6 m a i o m H M AOKJiaflOM M H o r n e

Me:»mHciinnjiHHapHbie

KOH(J)epeHiiHH c y n a c T H e M jiHHrBHCTOB. ^ 3 w k Ha a K T y - H n p n f l H p H H B o THJieH

B

oTHomeHHH

cyBepeHHbix

npaB

Ero

BenHnecTBa

" L i n g u i s t a s u m ; linguistici nihil a m e alienum p u t o ! " - c nojiHbiM ocHOBaHHeM n n r n e T

OH. 3

.SfobiK, n o P . SI.,

HeCKHX

H

COUHajIbHblX

06pa30BaHHÌÌ.

B

CTaTbH,4 r ^ e y6eflHTejibHO OTCTaHBaeTca KpHTepHH

BbwejieHHH

TpaHcno3HUHH

HMnepaTHBa

HMnepaTHBHoro

3TOÌÌ

ropAocTbio H

flojraeH

B a T b c a KaK TaKOBOH, Ha ojiorHHecKHe Ha6jiiofleHHH Hafl CJiaBflHCKHM CKJIOHeHHeM" nOKa3aHO, KaK 3HaneHHH BOCbMH pyccKHX naaeaceñ (BRjHonaa napTHTHB - [HCMHOZO] cnéey - H jioKaTHB - e cHezy) onHCbiBaiOTCH nocpe^cTBOM Tpex erporo ÖHHapHbix npH3HaKOB : ( 1 ) HanpaBjieHHOCTb, ( 2 ) oôbeMHocTb H ( 3 ) nepHC^epHHHOCTb ; npn S T O M c a M H naaeacHbie 3HaneHHH BOBce He no/maiOTca nocjieflOBaTejibHOH flHxoTOMHHecKOH KJiaccH(j)HKauHH, a o6pa3yioT TpexMepHyio CHcreMy"Ky6":

PII

7» P I nepH(t>epHflHocTi>

O6Í^MHOCTT> HanpasJieHHOcT-b

B H H a p H 3 M , OTCTaHBaeMblH P .

KaK p a 3 B J i e i e T OTKa3 OT ÄHXOTOMHHC-

CKHX (H B O O Ô m e OT HepapXHHeCKHx) KJiaCCH(J)HKaUHH - B n O J l b 3 y n p H 3 H a K O -

Boro Hbie

n o f l x o f l a , CM. B b i u i e - npHHUHn 2 . ( H a n o M H H M , HTO HMCHHO ÔHHapflH^^epeHUHajibHbie

n03B0JiHJiH npeoAOJieTb

npH3HaKH

4>OHeM>

npeflJiOHceHHbie

TpaflHUHOHHyK) a n x o T O M H H e c K y i o

P.

51.,

"nponacTb"

M e a c f l y r j i a c H b i M H H c o r j i a c H b i M H H o n H C b i B a T b H T e , H flpyrne B T e p M H H a x OflHHaKOBblX aÔCTpaKTHblX n p H 3 H a K O B . )

(ii) /ÍJiH P. ft. ÖHHapH3M - STO oTHioflb He npocTO yfloÖHbiH cnoco6 onHcaHHH, nocTyjiHpyeMbiH Hccjie^oBaTejieM B nopjiflKe pa3pa6oTKH

H. A. MEJltHYK

294

MeTaH3MKa. C caMoro H a n a j i a npHHUHn 3 BbiflBHrajicH HM KaK oTpaaceHHe ^eHCTBHTejibHoro nojioxceHHH aeji B ecTecTBeHHOM «3biKe - KaK orpaaceHHC peajibHoro cBoäcTBa npoTHBonocTaBJieHHH, o6pa3yiomHx Jiioôyio A3bIKOByK) CHCTeMy. ÜKOÖCOHOBCKHe npH3HaKH ÖHHapHbl nOTOMy, HTO OHH ÔHHapHbl B CaMOM H3bIKe. H 3TOT H3bIKOBOH ÔHHapH3M yBH3bIBaeTCH P . il. C ÔHHapH3MOM HeJlOBeieCKOH nCHXHKH, npOaBJlHKJLUHMCH He TOJIbKO B H3biKe, HO H BO Bcex apyrHX ciiihm hhcjiom ouihôok), neM cHHOHHMHHHbie hm $pa3bi b aKTHBe. 3 t o no-BHflHMOMy, CBHfleTejibCTByeT o tom, hto jsjin noHHMaHHH naccHBa TpeGyioTca jihiuhhc onepauHH (jiHuiHHe - n o cpaBHeHHK) C aKTHBOM). OTMeTHM eme, h t o HeMapKHpoBaHHbiñ HJieH rpaMMaTHHecKoro npoTHBonocTaBJieHHH o6whho xapaKTepH3yeTca Gojibmeft cTeneHbio HeoflHo3HaHHocTH, Heacejm MapKHpoBaHHbiii : HecoBepmeHHbiií b h a rjiarojia h eflHHCTBeHHoe

hhcjio

cymecTBHTejibHoro

hmciot

b

pyccKOM

asbiKe

rPAMMATHHECKHE HCCJIEflOBAHH-H

297

6ojibme pa3Hbix 3HaneHHH, neM coBepuieHHwii BHA H MHOxcecTBeHHoe C 3THM fljia HeMapKHpoBaHHoro HjieHa xapaKTepHa 6ojxee uinpoKa» cepa ynoTpeôJieHHii (cp. H3BCCTHI>IH Te3Hc E. KypwjioBHHa) H ôôJibuiaH nacTOTa B TeKCTax: nacTOTa HecoBepnieHHoro BHFLA H EFLHHCTBEHHORO iHCJia B pyccKHx TeKCTax - X 5 3 % H X 7 2 % HHCJIO COOTBeTCTBeHHO. B CBH3H

( 3 . A . ULTEHH^EJIBFLT,

Hacmomnuu cjioeapb).

(iii) ^3biK H36eraeT HarpoMoacaeHHH MapKHpoBaHHwx KaTeropHH. nosTOMy B HeMapKHpoBaHHOM HacTOflineM BpeMeHH pyccKoro rjiarojia pa3^HHaiOTCH JiHua, a B MapKHpOBaHHOM npomeAiueM BpeMeHH - HCT; B HeMapKHpoBaHHOM eflHHCTBeiTHOM iHCJie pyccKoro npHjiaraTejibHoro pa3JIHHaK)TCJI poflbl, a B MapKHpOBaHHOM MHOXCeCTBeHHOM HHCJie - HeT; HaKOHeU, epeAHHH H MyXCCKOH pOA pa3JIHiaiOTCH TOJIbKO B HeMapKHpoBaHHOM HMeHHTejibHOM naAeace.16 3flecb, KaK KaaceTca, P. ft. oÔHapyxcHJi BaacHyio H3biKOByio yHHBepcajiHio : HCBO3MO»CHO, HTOÔM B MapKHpoBaHHOH KaTeropHH H3biK pasjiHiaji 6ojibine, neM B cooTBeTCTByiomeH HeMapKHpoBaHHoii.

IIpuHu,un uKommocmu.

IIpoTecTyji npoTHB CJIHHIKOM 6yKBajibHoro H njiocKoro noHHMaHHa cocciopoBCKoro nocTyjiaTa o np0H3B0JibH0CTH H3biKOBbix 3HaKOB, P . ft. B pane CBOHX pa6oT HacTaHBaeT Ha cymecTBeHHO HKOHHHHOM XapaKTepe MHOrOHHCJieHHblX 33WK0BWX C p e a C T B , T.e. Ha HajiHHHH npHMoro ("cj)aKTHHecKoro") cxoACTBa Mexyiy o3HanaiomHM H 03HanaeMbiM y 3THX cpeACTB. TaK, B CTaTbe "Quest for the Essence of Language" ( 1 9 6 5 ) 1 7 P . ft. BCKpbiBaeT HKOHHHHocrb nop»AKa CJIOB (nopaAOK oflHopoAHbix rjiarojioB OTpaacaeT nopaAOK COÔMTHH: Veni, vidi, vici; nOpflflOK OAHOpOAHMX HMeH M05KÊT OTBeiaTb KaKOH-TO HepapXHH: IJpe3udeHm u ezo cynpyza, HO He HaoôopoT; oôbiHHbiii nopaaoK CJIOB "cy6i>eKT - o6i»eKT" [AeiicTBHfl] cBasaH c ncHxojiorHiecKHM nopaAKOM BocnpnaTiw KOMHOHCHTOB CHTyauHH). ^ajiee, B cTeneHHx npHnaraTejibHoro yqjiHHeHHe 03Haiai0mHx cooTBeTCTByeT HapacTaHHK) CTeneHH KaiecTBa : high - higher /haia/ - highest /haiast/, alt-(us) - alt-ior - altissim-(us); opMbi MHoacecTBeHHoro incjia, KaK npaBHuo, AJiHHHee (J)opM eAHHCTBeHHoro HHCJia: 4>p. (1) je finis-nous finissons, (2) tu finis-vous finissez, ( 3 ) il finit - ils finissent HJIH nojibCK. ( 1 ) znam - znamy, (2) 5.

znasz - znacie,

(3)

zna - znajq ;

H e H 3 B e c T H b i H3biKH, B KOTOPBIX MH. HHCJIO

Bbipaxcanocb 6bi HyjieBoii M0p dal\ b p e 3 y j i b T a T e b p y c c K O M CTaJio o 6 a 3 a T e j i b H b i M ynoTpeôJieHHe j i h h h o t o MecTOHMeHHH : H daji, mu jiHua: dan,

OH da A] 3 T a K O H c r p y K i t H a o 6 o 6 m H J i a c b ; h b p e 3 y j i b T a T e

"pyccKHÎi

H3bIK JIHIUHJICH OflHOHJieHHblX IIOBeCTBOBaTejIbHblX npeflJIOXCeHHH; . . . O f l HOHJieHHbie npeflJioaceHHa 6 e 3 M e c T O H M e H H H - n o f l J i e a c a m e r o - B e e r ò j i H i n b SJIJIHIITHHHbie B a p H a H T b l

flByHJieHHHX.

T a K H M 0 6 p a 3 0 M , B COBpeMeHHOM

p y c c K O M H3biKe HeB03M02KHbi npe,zyio)KeHHH c jiHMHbiM r j i a r o j i o M noflJiexcamero.

B

6e3JiHMHbix

npefljioxceHHax

H y j i e B o e n o f l j i e a c a m e e " . 2 4 TaK n p o c j i e x c e H a

cjie^yeT

6e3

ycMaTpHBaTb

CB«3b a i c u e H T y a u H H

nepe3

MOp(J)OJIOrHIO C CHHTaKCHCOM (3aOflHO B 3TOM HeÔOJIblIIOM a Ô 3 a U e 4 0 JieT H a 3 a f l ô b i j i a H a M e n e H a a o c h x n o p e m e n o j i H o c T b i o He p e a j i H 3 0 B a H H a a

nporpaMMa n o c r p o e H H a T e o p H H C H H T a K C H i e c K o r o H y j i a ) . B t o h caMOH

CTaTbe

pacKpbiTa

h

3aBHCHMOCTb

Meacfly

"oTMeHoii"

ace

3axoHa

B a K e p H a r e j i H h B03HHKH0BeHHeM H y j i e B o ñ c b h 3 k h b p y c c K O M a3biice. B apyroM

M e c T e P . R.

napajuiejibHoro H3bIKa: cpa3y

"H

othctjihbo

paccMOTpeHHa

B CHHXpOHHHeCKHX,

»ce o Ö H a p y a c H B a e T c «

np0B03rjiamaeT

rpaMMaTHKH H B

h

HeoöxoflHMOCTb

(J)0H0Ji0rHH

flHaxpOHHieCKHX

cymecTBeHHaa

aaHHoro

HCCJieflOBaHHHX

oôenx

B3aHM03aBHCHM0CTb

aBTOHOMHblX C T p y K T y p - (J)OHOJIOrHHeCKOH H r p a M M a T H H e C K O i i " . 2 5 3 T y B3aHM03aBHCHM0CTb P . Ä. a i c u e H T H p y e T b flByx a c n e K T a x : (i) O o H O j i o r H i e c K H e

eflHHHUbi n o - p a 3 H O M y

OTÔHpaioTCH

h

Hcnojib-

3 y i o T c a b p a 3 H b i x r p a M M a T H i e c K H X KOHTeKCTax, e f l H H H u a x h K a T e r o p n a x . Bot

pafl

flocraTOHHo

o6biHHo

flonycKaioTca

"npe4)HKC - K o p e H b " h

cjiynaHHbix

npHMepoB:

pa3JiHHHbie

coieTaHHa

b

aH-,

hh-

(Jjohcm,

h a

aycjiayTe Ha

" K o p e m . - cy4«t)HKc" - p a 3 H b i e caHflXH; b

CTbiice hhbx-

c k o m a3biKe b coôcTBeHHbix HMeHax BCTpenaioTca 4>HKCaX;

B

H3bIKaX,

HMeiOmHX

r a p M O H H K ) r j i a c H b i x , H e K O T o p w e 4>OHOJiorHiecKHe o n n o 3 H U H H b o 3 m o 5 k h m jiHuib b K o p H a x ; b pyccKOM a3biKe cpeflH r p a M M a T H n e c K H x

Moptf) TOJibKo

CJIOBOH3MeHHTejIbHbie MOp(J)bI M O r y T COCTOaTb H3 OflHOH TJiaCHOH, H TOJIbKO MeCTOHMeHHbie KOpHH — H3 OflHOH COTJiaCHOH; H T.fl. P a Ô O T b l P . 51. n o j i H b i n o f l o Ô H b i M H H a ô J i i o f l e H H a M H . H a o c H O B e H3JioxceHHUx c o o ô p a x c e H H H b p a ô o T a x P . Ä . n p e f l J i a r a e T c a c j i e a y i o i u H H Ba^CHbiH n p n e M : T e h j i h HHbie M o p < J ) o j i o r a H e c K H e e_HHHHiibi ( K O p e H b VS. a4>4>HKC, CJ10B0H3MeHHTeJIbHbIH VS. C J I 0 B 0 0 6 p a 3 0 B a T e j I b H b I H a4«J)HKc

cocTaBa.

h

T.n.)

onpeflejiaioTca nepe3 o c o ö c h h o c t h

hx

4>OHeMHOr°

rPAMMATHHECKHE HCCJlEflOBAHHÄ

301

( i i ) H3MEHEHHA B 4>0H0JI0RHHECK0H CHCTEME ONPEAE^EHHHM

06PA30M

CBH3aHbI C H3MCHCHHHMH B TpaMMaTHHeCKOH CHCTeMe. C a M P . 51. yKa3bIBaeT neTbipe acneirra noaoÖHbix CBsneii B HanpaBjieHHH OT OHo.iiorHH K rpaMMaTHKe: 2 6 nojiHaa n e p e c r p o i i K a n a p a f l n r M (B nacTHOcTH, nepepa3JioÜCEHHE OCHOB) BCJIEFLCTBHE HOBOH rpaMMaTHHecKoii onpeaejieHHoro

KaTeropHH (B HHBXCKOM H3biKe n o c j i e npe(J)HKca HeoöbeKTa

i-

Haia^bHaa

B3pbiBHaa

Bbicrynajia

BapnaHTa: tau- 'yiHTb', i-rau 'ynHTb Koro-To',

mejießoro

B BHfle

r - no3H-

rae

UHOHHbiii BapHaHT t; n o c j i e naaeHHH HanajibHoro i- npoTHBonocTaßjieHHe t~r

eKTHbix

KaTeropHH

HTO

4>°P M

O T

nepexoflHbix

"6e306i>eKTH0CTH").

(^OHOJIORHIECKHE

H3MEHEHHA

P.

51.

BJIEKYT

rjiaronoB; npn

STOM

nepecrpoiiKy

rpaMMaTHiecKofi CHCTCMH TO;U>KO B Tex c j i y n a a x , Kor^a B Heii c a M o f i ysce HAMETHJIACB COOTBETCTBYIOMAA TEIMEHUHA.

Oco6o

PACCMATPHBAETCA

y

P . 51. H oöpaTHoe HanpaBJieHHe B03aeHCTBiw rpaMMaraHecKOH CHCTCMW HA

(J)0H0Ji0rHHecK0e

pa3BHrae;

STOH

npoöJieMbi,

O^HaKO,

MU

3flecb

KacaTbca He 6 y a e M .

7. TIpuHUfm

uHeapucmmHocmu,

POAHBIIIHHCH B He^pax IlpaaccKOH niKOJibi,

Hamen B Jinue P . 51. o f l H o r o H3 CBOHX H a n ö o j i e e CTpacrabix H BMecre c TeM 3(J)(J)eKTHBHbix rpaMMaTHiecKHx CHCTeMe

Ji«)6oro

nponaraHflHcroB.

opM JIOKHT »3biKa". 2 7

B

3Ta

"ITpoGjieMa

OCHOBC

yneHH«

KOHcraTauHH

Gesamtbedeutungen o

He

rpaMMaTHHecKoii ocTajiacb

npocro

jio3yHroM: o^Hy H3 CBOHX Han6ojiee H3BecTHbix H B M e c r e c TeM H a n ö o j i e e cnopHbix

cTaTeö P .

51.

ÜOCBHTHJI ycraHOBJieHHio

"O6IUHX

3HaHeHHH"

pyccKHx naAeaceö ( " B e i t r a g z u r allgemeinen K a s u s l e h r e " ) ; no3*ce OH e m e p a 3 BepHyjicn K 3TOH TeMe B cTaTbe " M o p ^ o j i o r H H e c K H e Ha6jiiofleHHa Hafl CJiaBflHCKHM CKJIOHeHHeM". 3aHXBUIHCb OÖUIHMH 3HaneHHflMH n a A e xceft (B »3biKax THna pyccKoro), P . 51. caM nocTaBHji c e ö a B MaKCHMajibHo TpyflHbie H, TaK CKa3aTb, HeBbiroflHbie ycjiOBHH. flejio B T O M , HTO pyccKHH (H,

rnnpe,

CJiaBflHCKHM) naaexc - STO KaTeropHH B 3HaHHTejibHofi Mepe

CHHTaKCHHecKaa, HHane r o B o p a , BO MHorax y n o T p e 6 j i e H H H x (J)opMbi

BbicrynaioT

cKopee

npocTO

KaK

MapKepw

naflexcHbie

cHHTaKCHiecKHx

3aBHCHM0CTefi, He HMeiomne co6cTBeHHofi ceMaHTHKH. (3TO cooGpaace-

H. A. MEJIbMYK

302

HHC JieacHT B ocHOBe TpaKTOBKH naaeaca y E. KypnjioBHHa, A . AE TpooTa H A . B. McaneHKo, a Taicace B pa6oTax no ajire6paHHecKoñ jiHHrBHCTHKe: A . H. KojiMoropoB, H. H. PeB3HH, A . B. DiaflKHH, C . MapKyc; cp. eme no,oxofl A . A . 3ajiH3HHKa.) IIosTOMy BbiaBjiemie o6mero 3HaieHHa Toro cjiaBHHCKoro naaexca OKa3WBaeTca HCKjnoHHTejibHo TpyflHbiM

HJIH HHoro

ítejioM: 3TO o6mee 3HaieHHe jierao pacTBopaeTca B CHHTaKCHnecKOM KOHTeKCTe. TeM He MeHee, STO o6mee 3HaeMHbiM B

OTJlHHHe

epazaMu),

OT P

H N,

CB«3AHA

B03M05KH0CTb areHTHBHoro T

HHCTpyMeHTajibHoro T ( n a p u c o e a n nepoM),

T ( i o n dyiuoü),

cpaBHHTejibHoro T (mhüach

(y6um

orpaHHHHTejibHoro

cmpenoü)

H

T.n.; K STOH »ce

nepHepHHHOCTH T CBOFLHTCH TOHKHE CMbicjiOBbie pa3JiHHHH B c j i y n a a x rana

Oh

hukom.28 CKoro

6ua

mumyAxpHbiü

coeemnuK

vs.

Oh

6ua

mumyAnpHbiM

R&yKe ecjiH HE C I H T A T B uejiecoo6pa3HbiM BBWEJIEHHE

coeem-

ceMaHrane-

HHBapnaHTHoro « A P A ¿UIA KAACFLORO naaeaca, Bce paBHO Heo6xoflH-

MOCTb KaKOH-TO HHBapHaHTHoñ xapaKTepHCTHKH naflexceñ oieBH^Ha; B HaeTHoeTH,

3.

BEHBEHHCT

aun

xopouiHH

npHMep

YCTAHOBJIEHH»,

HHBapHaHTHOH CHHTaKCHHeCKOH (jjyHKUHH JiaTHHCKOrO reHHTHBa HMEHHAA TPAHCNO3HUHX HOMHHATHBA H AKKY3ATHBA). /I,YMAETCA,

(npH-

oflHaico,

HTO HAMEHEHHAA y P . ÍL. ceMaHTHKa nafleaceñ, C T o a m a a 3A HX CHHTAKCHC0M, - BEMB

oneHb HyxcHaa H n o j i e 3 H a a ; npocTo noica B jiHHrBHcraKe e m e

njioxo yMeioT HaxoflHTb MCCTO ¿vía noAo6Hbix CBEAEHHH - aHajiorniHO TOMy, KaK noKa He Haxo^HT (B epHHHbie naaeacH HMCIOT BO MH.

HHCJie oflHy (J)opMy ( y d a p u M a ) , a Bce HenepH(f)epHHHbie coxpaHaioT

CBOH pa3JIHHHa.29 (iii) " O T K p b i B a e T c a CBa3b Meacay cocTaBHbiMH 3JieMeHTaMH naflexcHoro 3HaneHHa H oHeMaMH HJIH cocTaBHWMH 3jieMeHTaMH (JJOHCM: -rrí-

(B

303

rPAMMATMHECKHE MCCJIEflOBAHHÄ

aBTOMaTHiecKOH anbTepHauHH c - m - ) BbicrynaeT Kaie npHMeTa naaexcHoro

oöiuhh aTpn6yT -v- h -X-, oöicmhocth. OoHOjiorHH h

npH3HaKa nepHtjjepHimocTH, a mejiHHHOCTb, cjiyacaT npHMeTofi naaexcHoro

npH3HaKa

rpaMMaTHKa 0Ka3biBai0Tca cBH3aHbi u e j i o ñ r a M M o ü nepexoflHbix, MeacpañoHHbix n p o 6 j i e M , a rjiaBHoe - Hepa3flejibHOCTbK) p e i e ß o r o 3ByKa h

jihhfbhcthkh yneT jho6mx peryjiapHbix cooTHomeHHH Meac^y opMoií h cmwcjiom (CTp. 298-99) H CBfl3b MeXCfly pa3HbIMH ypOBHHMH H3bIKa (cTp. 299-301). ÜHBapHaHTHbie 3JieMeHTbi b pyccKHX naaexcHbix 3HaneHHax - 3to oahh H3 HaHÖojiee hpkhx, ho bobcc He eflHHCTBeHHbiH npHMep BbifleJieHHa 3HaHeHHa". 30 Bbiuie yace OTMenaJiocb, a c o j i b BaaceH ana

HHBapnaHTOB

P.

y

Jl.

Ot

HHBapnaHTHbix

sjieMeHTOB

b

0H0Ji0rHH

(aH(J«i)epeHUHaJIbHbie npH3HaKH) AO HHBapHaHTHblX (j)aKTOpOB pa3JIHHHbIX at}>a3HH, He r o ß o p a yace 0 6 ycTaHOBjieHHH HHBapHaHTOB a p y r n x

rpaM-

MaTHHecKHx 3HaneHHH (HanpHMep, b cTaTbe " Z u r Struktur des russischen V e r b u m s " ) - Bce jiHHrBHCTHiecKoe

TBopnecTBo P .

ikothhcho

M.

no-

CTOHHHOMy nOHCKy HHBapHaHTOB. Uraic,

bo3mo5khoctch b xanecTBe npeflBapHTejibnccjieaoBaHHH bcakoh rpaMMaTHiecKoft o6jiacTH,

- j i o r H n e c K o e HCiHCJieHHe H o r o rnara n p n

-KOMÖHHHpoBaHHe BbiHBJieHHbix skaia," Selected Writings, IV, 603-12. The same philological approach is applied by Jakobson to the study of the textual relationship between the Slovo and the Zadonscina on the basis of the expression Za lolomjanem/za solomonom (Selected Writings, IV, 534-39). 87 Cf. D. S. Lixaëev, Russkie letopisi i ix kul'turno-istoriíeskoe znacenie (Leningrad, 1947), 195, and Selected Writings, IV, 609. 88 Selected Writings, IV, 607. 89 "Kompozicija i kosmologija placa Jaroslavny," Trudy Otdela drevnerusskoj literatury AN SSSR, XXIV (1969), 32-34. 90 "Sokol v mytex," Juznoslovenski filolog, knj. XXX/1-2 (Belgrade, 1973), 125-34. 91 Jakobson's concern for those written texts which may provide a philological documentation of the language actually spoken in Medieval Russia is well expressed in his penetrating analysis of several Novgorod writings on birch bark ("Vestiges of the Earliest Russian Vernacular," Slavic Word, 1,2-6 [= Word, VIII (1952), 350-55] ; "A. V. Arcixovskij and M. N. Tixomirov, Novgorodskie gramoty na bereste," Slavic Word, II, 83-85 [= Word, IX], 407-09). 92 "O sootnolenii mezdu pesennoj i razgovornoj narodnoj reC'ju," Selected Writings, IV, 528-33. 93 "An Old Russian Treatise on the Divine and Human Word," St. Vladimir's Seminary Quarterly, IV, 1-2 (New York, 1956), 45-50. 94 Ibid., 45-46. 95 N. S. Trubetzkoy, Vorlesungen über die altrussische Literatur, mit einem Nachwort von R. O. Jakobson [=Studia Histórica et Philologica, Sectio Slavica, 1] (Florence: Licosa-Sansoni, 1973), 157 (cf. N. S. Trubetzkoy's Letters and Notes, ed. R. Jakobson [The Hague-Paris: Mouton, 1975], 86). 96 N. S. Trubetzkoy, Vorlesungen.., "Nachwort," 159.

JAKOBSON'S CONTRIBUTIONS TO AMERICAN LINGUISTICS HARVEY PITKIN

To write of one of the most prodigious and renowned scholars and scientists of our century, a scholar who is among the foremost linguists of any period in the breadth, depth, and duration of his contributions, and to attempt to indicate those contributions to American linguistics is already paradoxical: one must denigrate the contributions so as to achieve a delineable scope, or indicate a linguistics in America so impoverished previously, so limited in its development, that it must by now be either quite contingent on many of his ideas or parochially impermeable to them. None of these is possible, while they are all true to some extent. I will then restrict this homage to an indication of the influence of Roman Jakobson, even where it is indirect and mediated, rather than include the actual contribution, and enumerate those aspects of linguistics that are both prominent currently and uniquely associated with Jakobson. First, then, it is in the area of phonology and its reverberations that his greatest influence must be sought. American linguistics is not only now massively in his debt for distinctive feature analysis, which holds such a preeminent position, but the theoretical implications have informed many branches of the study of language as well as adjacent phenomena. The tradition of structuralism emanating from Kazan and Moscow sources, carried from Prague after its elaboration there to New York and Cambridge, has at each stage of its journey infused American linguistics with rich notions concerning the underlying system of organization of sound; but those notions originally European and applied at first only to phonological data have been dispersed as well to neighboring terrains through linguists as remarkable and diverse as Sapir, Austerlitz, Garvin, Halle, Chomsky and by now many others, including students of Bloomfieldian and other entirely native traditions. It is from this source, albeit variously filtered, that the crucial idea of functional contrast, when it is seen as the essential basis for characterizing relationships of entities comprising phonological systems, has been extended from those considerations originally

358

HARVEY PITKIN

embracing a spectrum from phonetics, phonemics, and morpho(pho)nemics, to morphology, syntax, and semantics, and thence to componential analyses of cognitive and ethnographic materials. The emphasis on binarism, while not universally accepted, has been at least a disciplining goad toward clarifying structural relationships. The associated concepts of functional contrast, distinctive features, markedness, dimensionality of phonological systems, and non-discrete hierarchical levels of language organization, while they may have had previous and even contemporaneous American manifestations with some degree of resemblance (compare, for example Hockett's Manual of Phonology with Trubetzkoy's Grundziige der Phonologie), were at great variance with the predominant mode of American linguistics. That they are now so obviously a part of our tradition of structuralism, in the broad sense of the term, represents a revolution if one but remembers the decades preceding Jakobson's impact. This is not to minimize the linguistic contributions of Whitney, Boas, Sapir, Bloomfield, and other earlier American scholars, nor to claim for Jakobson an entirely isolated position independent of predecessors like Baudouin or Trubetzkoy, but to note that through Jakobson great advances were facilitated at a period too long dominated by mechanistic, behaviorist, distributionist, item-and-arrangement dogma, which had obscured temporarily the ideas of relationship, process, and the attention to meaning (function) in Sapir's vision, despite an American climate very diffident toward European approaches in general. The receptivity of American linguists to Jakobsonian phonological ideas was no doubt partly due to an aridity in phonemic studies, but perhaps also to the attractive linkage between distinctive feature analysis and empirical studies of physical manifestations of sound, so appealing to American predilections for empiricism and even for a rejuvenated mechanism. There may, as well, be a debt to the not entirely theoretically motivated association with the successful Chomskyan syntactic model, felicitously linked through Halle and the intellectual geography of Cambridge. A t any rate, a balance was restored by Jakobson that had been lost after Sapir's partial eclipse during the period of the Bloomfield-Harris-Bloch orientation, and an opening to European notions was achieved. This revolution, however wrought, was to amplify the relations between surface manifestations of observed phenomena and underlying, implied, non-observable features, verifiable in more than a single analytic mode. The parsimonious representations thus made possible enhanced the opportunities for abstraction and the investigation of universals. As typology had become extinct once phonetics was replaced by phonemics, distinctive

AMERICAN LINGUISTICS

359

features introduced again a chance for comparativism and for universals to be investigated, where only linguistic relativism had been previously feasible. That also the Jakobsonian phonology was a blow to the taxonomic phonemics of neo-Bloomfieldians was most keenly felt in dealing with unwritten languages, but as a specification of the significant values of phonemics was proffered by P. Newman, S. Schane, and students of American languages who reconsidered what kind of phonemics Sapir proposed as having psychological reality, a synthesis was arrived at which offered new insights into phonological and other structures, including psychological ones, both synchronically and diachronically. On balance the phonological approach of Jakobson has left American linguistics permanently altered and has enriched and spurred entirely new lines of research that had either not been previously undertaken or had been conserved on the margins of American linguistic intellectual life: e.g., in the journal Word (various European orientations), and among Sapireans investigating native American languages (i.e., in the "underground"). The reverberations of these advances in phonology have been most manifest in generative phonology, the currently "received" approach, but even in other styles of phonology and in non-phonological investigations: matrices, features, distinctive contrasts, markedness, and neutralization are now all extensively used concepts. Perhaps the next large area to exploit very advantageously adoption of this model lies furthest from phonology and seems indebted to Jakobson via two independent routes. Semantic, ethno-semantic, folk-taxonomic, mythological, and structuralistethnological studies, contingent either on the linguistic model directly (Conklin, Frake, Goodenough, Mathiot, et al.), or indirectly, as transmitted from Lévi-Strauss after his contacts in New York with Jakobson, constitute a large and theoretically prominent area of study. Moreover, the use of this phonological model has helped reintegrate these traditional philological and linguistic concerns, traditional within the anthropological view, with the too narrowly defined fashion of linguistics that resulted in the proliferation of hyphenated subdisciplines. All of these semantically orientated kinds of studies not only are indebted to distinctive feature analysis as the model, but besides owing their structural components to Jakobsonian phonological method and semiological interests, curiously recapitulate in their substantial data traditional concerns of Jakobson for texts, style, folkloric, and literary subjects. Even studies of kinship terminologies, disease, botanical, and other lexical domains, as well as such favorites of Jakobson as deictic systems and shifters (switch-reference, as it is called in Hokan studies) typically partake

360

HARVEY PITKIN

of componential analytic methods, or often grow directly out of an interest in oral literary texts collected during linguistic field work on native American languages in the anthropological linguistic tradition. This is quite parallel to early interests of Jakobson and Trubetzkoy, who originally in their careers worked on the traditional texts of various languages of the Russian Empire: Russian as well as a number of others. In many ways the expansion into and settlement of the North American continent, the large variety of indigenous languages encountered by the European colonists and settlers, provokes interest in the parallel Russian experience. Similar research undertakings, stemming originally from ethnographic exploration in the Russian colonized Eurasian lands, were not only brought to the attention of Americans, but also knowledge of the insights and theoretical orientation that grew out of the Russian linguistic experience has been enriching. But Jakobson also provides bibliography for Americanists of archival materials originally in Tsarist collections as the result of the Russian presence on the western coast of North America, and even information about eighteenth-century American-Russian contact resulting in lexical collections of east coast languages done at the behest of Catherine the Great, intended for her projected encyclopedia. Jakobson has always confronted us with the non-American world of ideas and even sometimes with our own history. In the most general manner the spirit of American linguistics has been influenced by certain Jakobsonian "oppositions" presented in numerous university lectures as well as in publications. The paradoxical formulamodel: "the grammar of poetry and the poetry of grammar" has been employed to discuss the dynamics of synchrony and the statics of diachrony, paradigmatic and syntagmatic dimensionality, message and code, metaphor and metonym, and in this way to relate what had been artificially separated, to undo exaggerated compartmentalization that grew out of a more naive understanding of structure. To mention a single example, the work on English phonology done by Chomsky and Halle demonstrates a reaction from a strict Saussureanism toward a synchronics infused with internal reconstruction and historical dynamics in their treatment of the English sound system. These Jakobsonian theoretical juxtapositions mirror, as it were, phonological oppositions in the sense of embracing polarities. In my own direct experience of Jakobson in the milieu of American Indian linguistic studies, we have benefited on several scores from these confrontations of apparently opposed perspectives considered as non-overlapping in their integrities. The previously accepted dichotomies that he linked together raised ques-

AMERICAN LINGUISTICS

361

tions of methodology whose solutions augmented in one instance synchronic descriptivism with reinforcement by dynamic process alternatives, and likewise strengthened the understood relations between languages seen diachronically and the models of static structures which set limits on the character of their reconstructions. In the case of viewing synchrony dynamically is contained a program for understanding microlinguistic changes as they are underway. In the case of understanding the static aspect of diachrony lies a reiteration of the values demonstrated in work like Bloomfield's successful Proto-Algonquian. In the three other above-mentioned juxtapositions are not only injunctions for considering fully the aspects of two-sided relationships which continually recur, but in the case of message and code an implicit foreshadowing of sociolinguistic concerns which were to become so prominent as a reaction to a temporary overemphasis on studies of linguistic competence to the neglect of performance. Jakobson's explication of the relations of metaphor and metonym have surprisingly not only been a contribution to literary and textual work, but are a prominent rubric recently in more strictly anthropological research (e.g., E. Leach) where the nature of symbolic systems employed culturally for communication is investigated. Stemming from his important work on child language, aphasia, and the study of general sound laws, developments in psycholinguistics show a powerful Jakobsonian influence. This is not only an area of research significant in its own right, but also one, like the study of language universals, that is important for the contribution it makes to the development of general linguistic theory and the ontogeny as well as origins of language. Psycholinguistics has in fact concentrated on child language acquisition and pathology as a direct outgrowth of Jakobsonian innovations. But here as elsewhere the effects of his scholarship may in the long run turn out to be even more valuable for the climate of inquiry he has stimulated and for broadening the horizons of American linguists by introducing to their work topics and ideas which, immediately preceding his arrival on the American scene, were not considered legitimate except in the "underground," and which were essentially independent of much of the European, and especially the insights of the Prague, school. In some cases (mythological studies) it is Propp, in other cases Piaget (cognitive development), where European ideas now so much at hand have been drawn to our attention, and in yet other cases the stimulation has resulted in the recognition of American scholarship previously neglected (Peirce, the philosopher). One must conclude that for his having introduced specific scholars and subjects of linguistic import (psycholinguistics), and for his having

362

HARVEY PITKIN

linked topics not before considered as an ensemble (child language, pathologies, phonological universals), and by introducing us to other traditions of linguistics (Mathesius, Havrdnek) we are greatly indebted. While there is no question that Jakobson has dominated Slavic studies in America, and that his students and colleagues have extended his influence to topics of general linguistic importance (e.g., M. Shapiro), and that his contribution to metrics, stylistics, and semiology have been as copious as his collected writings fully testify, there has been from his elaboration of structuralism in whatever subfield of linguistics a general vitalizing effect. Not only did we profit from an assortment of the European sources he disseminated, as from his own research, but there appeared in American linguistics a change parallel to that in other disciplines which depends in no small measure on that sense of structuralism his various work embodies. Although he was not alone in accomplishing this change, he was in the forefront. And his sense of structuralism has not undergone the vicissitudes of fashionable revision, but seems a more permanent part of our scientific perspective. It has been a contribution speeding linguistics on from a model like that of the periodic table of the elements with an atomistic view of substance (matter) to a model much like that of contemporary advanced science where relationships and process are understood as crucial aspects of systems. The current admixture of idealism, rationalism, and mentalism to the previous neo-Bloomfieldian structuralism, with its exclusively empirical, inductive behaviorism, would never have come about with all the current richness of competing ideas without this foundation. The exclusive attention to distributionist surface phenomena, where directly observable, was impoverishing to the study of more abstract underlying relationships indicative of the complexity of systems "ou tout se tient." We are now, grace a Jakobson, able to exploit methodologies more sophisticated than we could before we knew him. Columbia University

ROMAN JAKOBSON A N D THE NEW POETICS NOTES ON THE JAKOBSONIAN METHOD IN THE ANALYSIS OF SLAVIC POETRY AND PROSE KRYSTYNA POMORSKA Poetics is a relatively new domain in literary studies, although the term itself has been known since Aristotle and has been used at various times to cover the most general, often vague and heterogeneous problems in the analysis of a literary text. At the beginning of the century Oscar Walzel's studies on the morphology of the novel, Kazimierz Woycicki's investigations in metrics, as well as later studies, such as Zygmunt Lempicki's philosophically-oriented analyses of literature,1 have all been called equally "studies in poetics." Roman Jakobson not only introduced a homogeneous method into this previously undifferentiated realm of investigation but also, in a sense, made the field a consistent feed-back system. This is best illustrated by his studies on Slavic metrics, where the relation between the scholar's phonological theory and the material under investigation is particularly palpable. In his first major monograph on comparative metrics, On Czech Verse, Primarily in Comparison with Russian,2 Jakobson analyzed Czech and Russian versification in terms of the dominating difference in their linguistic material: stress and vowel quantity. This method of comparative analysis was based on his first steps in the phonological theory of distinctive oppositions. It is illuminating to compare Jakobson's approach with the pioneering studies of Andrej Belyj,3 who, despite his merits as an insightful connoisseur of Russian verse and as the initiator of the statistical method in Russian metrics, nevertheless mechanically projected the principles of classical versification onto the Russian-language material. The relevance of the phonological theory for the investigation of verse structure is convincingly shown in Jakobson's later studies, such as "O lingwistycznej analizie rymu," 4 "Ob odnosloznyx slovax v russkom stixe,"5 and, above all, in his fundamental study "Linguistics and Poetics." 6 The coherence of Jakobsonian poetics has its roots in his general linguistic theory, and this framework should be constantly borne in mind. The first to write about Jakobson from this point of view was the Polish

364

KRYSTYNA POMORSKA

metrician, the late Franciszek Siedlecki, who in an article characteristically entitled "Roman Jakobson i nowa lingwistyka" 7 examined his works on verse and other areas of poetics as an outcome of his general theory of language. The basic postulates of the linguistic method as applied to the analysis of poetic texts are: (1) the principles of opposition and hierarchy that characterize every form of verbal activity; and (2) a functional analysis of all the hierarchical elements in a poetic work and, above all, of the inter-relations between parts and wholes. One could name a number of fields within Slavic poetics where Jakobson's contributions have led to significant advances: comparative Slavic metrics, with particular regard to oral tradition, Old Church Slavonic and medieval Czech versification, and the analysis of various modern Slavic forms of verse. In view of the highly specialized character of these areas, we shall leave them aside and concentrate instead on several concepts of general theoretical import which have dominated Jakobson's research and which he has applied in the analysis of Russian and Polish poetry and prose of the nineteeth and twentieth centuries: the role of phonology in poetic texts; the significance of grammatical categories for poetry; 8 the nature and role of metaphor and metonymy; and, finally, the search for invariants. There are three basic periods in Jakobson's scholarly life, each represented by certain fundamental works which clearly indicate the development of his ideas on poetics. The Newest Russian Poetry (written in 1919 and published in 1921)9 deals exclusively with Russian material; although limited in principle to the poetry and prose of the Futurist poet Velimir Xlebnikov, it attacks a much larger sphere of questions connected with the poetics of the Russian avant-garde in particular and with the concept of poetic language in general. Jakobson's late period is epitomized by the summarizing papers of the early sixties, "Linguistics and Poetics" (I960) 10 and "Poetry of Grammar and Grammar of Poetry" (1961; 1968),11 which, although not restricted to Slavic poetry and prose alone, nevertheless shed light on theoretical problems of cardinal significance for Slavic poetics. In between these two periods, in the thirties, there appeared two essays of monograph character in which the search for invariants in poetic texts was for the first time related to the biography of the poet: "On a Generation that Squandered Its Poets" 12 and "The Statue in Puskin's Poetic Mythology." 13 PHONOLOGY AND GRAMMAR

The stress on the structural role of phonological and grammatical categories in poetry is an immediately striking feature of Jakobson's theory

THE NEW POETICS

365

and analytic method. Why do phonology and grammar merit the investigator's particular attention ? If these elements simply fulfilled in poetic language the obligatory task they have in any verbal activity, it would be trivial to take them into special consideration. What gives them a special relevance in poetry can be understood in the light of Jakobson's concept of poetic language as a particular type of verbal activity. Its particularity, which Jakobson labels the "poetic function," is condensely defined as the "projection of the principle of equivalence from the axis of selection into the axis of combination." 14 Due to this operation, those elements which in non-poetic speech remain "latent" and "prerequisite" are pushed by the poetic combinatorial pattern "into the level of a patent, 'palpable' and 'perceptible' form," as was recently clarified by Elmar Holenstein.15 A consequence of this fact is that poetry consists of such indispensable phenomena as paronomasia and so-called "poetic etymology." These phenomena, in turn, give a new semantic dimension to the poetic text, since they themselves become carriers of meaning; their accumulation and careful distribution forces us to re-read the text, so to speak, in its paradigmatic dimension. The cardinal role of both the sound and grammatical level in poetry was touched upon already in Jakobson's earliest study, The Newest Russian Poetry, but the treatment of these levels remained typical of the times in that they were isolated from the entire context of language, a deficit which was finally overcome only in "Linguistics and Poetics." In his essay of 1921 the analysis of the sound pattern in poetry was not yet carried out phonologically; it was labeled "euphony," 16 and traditional phonetic terms were used for the analysis of both vocalic and consonantal repetitions in Xlebnikov's works. At the time he offered this paper in a meeting of the Moscow Linguistic Circle, Jakobson was twenty-two years old; his systematic achievements in phonology were to come some eight years later.17 The phonological approach was, however, used in the analysis of poetry earlier than in Jakobson's writings on phonology proper, in the study On Czech Verse, where it was applied to the strictly prosodic investigation of verse. Despite its limitations, The Newest Russian Poetry nevertheless offers in kernel form most of the major points in Jakobson's poetic theory, which only later was to develop into a coherent whole. In relation to the sound level and its role in carrying the new dimension of meaning, one finds, for example, the notion of the "blending" within a single word of the sound pattern which runs through an entire poem. 18 Later Jakobson analyzes this same phenomenon, under the label paronomasia, using the

366

KRYSTYNA POMORSKA

wider poetic material of the most diverse languages and epochs, paying particular attention to the special case of paronomasia which Saussure, in his unpublished manuscript, called the anagram. In fact, many of the examples cited in The Newest Russian Poetry are specifically anagrammatic in nature, e.g., "S - T - N - L - utonnyx, snega, solnce - kontaminacija stonala."19 The essence of an anagram is that the particular distribution of sounds (here consonants) prepares the pertinent word, sometimes directly the key word of the poem. Another "sound figure" presented in abundance in the Xlebnikov study is that of "poetic etymology." Classified as the case of the "sblizenie dvux edinic,"20 i.e., of parallelism, this phenomenon is explained, following Xlebnikov's own pronouncements, 21 as follows: if two words with a similar sound pattern are juxtaposed in poetry, they are perceived as being connected semantically as well. Typical examples include: "Kesar' moj svjatoj kosdr' (Batjuskov)" or "Vojna i mec\ vy casto tol'ko mjac (Xlebnikov)." 22 This figure also covers a slightly different but cognate phenomenon; when two words are in fact etymologically linked, but this connection is no longer perceived in normal speech, it may become reactivated in poetry. In Jakobson's recent works, such as the Polish version of "Linguistics and Poetics," the latter type is presented and explained in more detail; a good example is the line from Kazimierz Wierzynski's poetry, "Jeden malenki oblok bialym wldknem siq pali.. ." 23 in which the etymological connection between oblok and wldkno, automatized in normal speech, becomes a vivid poetic figure. In "Linguistics and Poetics" and a series of studies of Jakobson's recent period, i.e. from the early sixties to the present, the role of paronomasia is recognized as much more extensive than was the case in The Newest Russian Poetry. It may even embrace such categories as the entire theme of a poem. One of the reasons for this extension of the concept was the growing elaboration of phonological analysis, for which the minimal units were no longer phonemes but their autonomous constituents - distinctive features. 24 It became possible for the investigator to observe a meaningful accumulation of both vowels and consonants connected by at least one common feature, where this accumulation discloses an appropriate motif or theme in the poem. 25 Paronomasia in the narrower sense, i.e. a repetition of vowel/consonant groups in juxtaposed words, serves as an indicator of the latent semantic connection of those words. The method may be best illustrated briefly by an example from Jakobson's recent study "Stixi Puskina o deve-statue, vakxanke i smirennice."26 Puskin's twostanza poem, "Net, ja ne dorozu mjateznym naslazden'em . . . " develops

THE NEW POETICS

367

respectively two different images of an erotic ideal, one of which is declared preferable to the other. A t the same time, however, paronomasia uncovers their factual similarity: "sxodstvo v razvitii dvux kontrastirujus£ix tem skrepleno zvukovymi povtorami: WS/MPLEN'^WJ vakxanki predskazan Mig POSLEI/NIX sodroganij, a uzlovym slovam vtoroj frazy - MOLEN'JVZ - NE VNEMLCS' - v e p i l o g e o t v e c a e t PLAMEN' P o N e v o L e ! " 2 7

A phenomenon similar to that of paronomasia may be observed on the grammatical level, too, where linguistics also serves as a valuable tool in elucidating the interconnections of sound and sense. The period of intensive investigation in this area was initiated by Jakobson's fundamental study "Poezija grammatiki i grammatika poezii," first presented at the International Congress for Poetics in Warsaw in I960,28 but the first observations on this question date from the Prague period, when Jakobson edited, jointly with A. Bern, Puskin's selected works in Czech translation.29 The material of the Slavic languages, so rich in morphological and inflectional means, made Jakobson conscious of two things: first, the obligatory character of grammatical meaning, and second, the fact that these obligatory meanings turn into tropes in poetry, and furthermore, that poems may indeed be made up exclusively of such grammatical tropes. Thus, a play on genders may become the pivotal point of the lyrical drama, as in Puskin's short poem "Zoloto i bulat." 30 Gender is also the central device of poetic semantics in the expressive title of Paternak's famous collection Sestra moja - zizn' ( ' M y Sister Life'): it cannot be translated from Russian with its feminine gender for "life" - even into other Slavic languages, where " l i f e " has different genders, without losing its entire poetic value. Thus, a discovery made on the basis of Slavic poetry casts light as well on the problem of translatability and on the theory of translation in general.31 The power of grammatical genders goes beyond the borders of poetry and plays an active part in our verbal consciousness. This is reflected, for instance, in everyday mythology, for example in peoples' ideas of Death as a personified phenomenon which, due to its genders, assumes a feminine guise in the Slavic world, while appearing masculine in other cultures. Jakobson's discoveries made on the basis of Slavic poetry coincide with and confirm the theories of anthropologists and ethnologists, for instance those of Franz Boas and B. L. Whorf, 32 and of Edward Sapir, whose views on the compulsory character of grammatical meanings are quoted in "Poetry of Grammar . . ," 3 3 Another important and telling coincidence of views which should be noted is that of the theoretician Jakobson and the poet Vladimir Majakovskij. As the author of "Linguistics and Poetics" relates, during one of the

368

KRYSTYNA POMORSKA

discussions in the Moscow Linguistic Circle about epitheta ornantia, Majakovskij characteristically declared that "for him any adjective while in poetry was thereby a poetic epithet." 34 Even such obviously "nonornans" epithets as "Bol'saja Medvedica" ('The Great Bear', i.e., the Big Dipper) or "big" and "little" in the names of such Moscow streets as "Bol'saja Presnja" and "Malaja Presnja" act as full-fledged poetic epithets in Majakovskij's poetry. 35 Accordingly, in his recitations of Majakovskij's poems Jakobson, in agreement with the poet, puts a logical stress on the attribute as well as on the noun, following the rhythmical impulse of Majakovskij's verse.36 Not only does this practice show once again that the epithet, like any other linguistic means used in a poem, becomes a part of the poetic totality and is not a mere rhetorical "adornment"; it helps us to realize the declamatory type of verse, exemplified by Majakovskij's poetry. And vice versa, the style of recitation, faithful to the structure of the verse, emphasizes and helps to build "language fictions," according to the term advanced by Bentham. In accord with this, when reciting a poem by the contemporary Russian poet Bulat Okudzava, in such lines as "Ja v sinij trollejbus sazus' na xodu, / Poslednij, sluCajnyj...," not only do we qualify the last two adjectives as genuine lyrical epithets, but even sinij plays for us the same role, although it refers to the actual color of Moscow trolleybuses. Jakobson's investigation of the Slavic case system opened two new aspects in the analysis of poetic texts: it made it possible to study the patterning and interplay of different cases in a poem and also to observe the accumulation and distribution of a particular case in a given text. The system of cases acquires a distinctive importance if the text is devoid of specifically poetic tropes and figures, if it is an "imageless" poem. Puskin's poetry again provided the first model for this type of analysis. In poems like "Ja vas l j u b i l . . . " the role of grammar is quite manifest, since grammatical categories are the only tropes in the poem's structure. This is even more significant since Puskin's work is not "experimental" poetry, and it is thus not a matter of the "device laid bare" as in the case of Xlebnikov's work, but of rudimentary poetic phenomena. The same can be said of Puskin's "Zoloto i bulat" or of Radiscev's "Tobol'skie stixi." If one fails to recognize the special selection and accumulation, the symmetrical distribution and semantic functionality of grammar in such poems, they are reduced to the level of ordinary discourse. In Jakobson's method, the role of grammar within the text is treated similarly to the role of rhyme and of the metrical pattern. Like phonological constituents, grammatical categories, with their obligatory meanings and their power of abstraction

THE NEW POETICS

369

comparable to geometrical spatial relations, are transformed in verbal art into truly poetic, self-directed "fictions." POETIC SEMANTICS AND GRAMMAR

Grammatical tropes, like any trope, necessarily carry their own meaning. In the case of grammatical tropes, however, one must first distinguish the general grammatical meaning a given category has in the language system; only then can its semantic function in the poetic structure be discovered. For example, in Puskin's poem "Ja vas ljubil...," analyzed in the abovementioned study,37 the personal pronouns ja and vy occur as the sole subject and object, respectively, of the poem. The very fact that these categories alone were selected to refer to the heroes of the poem endows this poem with a specific capacity. Moreover, the relation between ja and vy is clearly indicated by the grammatical case in which each of them appears: the nominative is used for the protagonist of the drama, whereas the object of his love appears exclusively in two oblique, "directional" cases (accusative and dative). This underlying relation specifies the active and the merely passive participant in the love affair referred to. This demonstrates how the "power of abstraction" - the relational meaning of the Russian case system38 - acquires a specific semantic role in a particular poem. A different aspect of the same phenomenon is found in Pasternak's title poem from the collection Sestra moja - zizri. This example is all the more interesting because, while the poem is not limited to grammatical tropes alone, one single case does become a conspicuous trope whose accumulation, distribution, and function must be taken into account to understand the poem's semantics. Here it is the instrumental case, with its general meaning of marginality,39 which plays a significant part in the make-up of each stanza. After a careful analysis of the whole poem, and of the entire period in Pasternak's creative work to which the poem belongs, it becomes clear that the instrumental, in its various usages, carries the vital theme of marginalia and their significance for life, a theme which remains otherwise merely latent. 40 A grammatical category need not be limited in its semantic function to the theme of a single poem; it may also indicate larger tendencies in the philosophy and style of an entire poetic trend. Such is the case, for example, with the poetry of Aleksandr Blok and even that of the Russian Symbolists as a group, 41 where the conspicuous role of pronominal categories can be observed.

370

KRYSTYNA POMORSKA

METAPHOR AND METONYMY

Jakobson was the first to show in depth 42 that metaphor and metonymy are not just tropes confined to particular lexical units but are wider and more fundamental phenomena connected with the two basic linguistic operations: selection and combination. When he states that poetry and prose are, respectively, metaphoric and metonymic in nature, it does not necessarily mean that one finds a predominance of one or the other figure in either type of discourse, but that the former is primarily the result of selection, a process based on the principle of similarity, while the latter emerges out of combinatory activity, based on the principle of contiguity. At the same time metaphor and metonymy also relate to these two principles. It was again on the basis of material from Slavic poetics that this discovery was made. Long before the appearance of the theoretical study devoted to metaphor and metonymy as polar aspects of aphasic disturbances 43 Jakobson wrote, in 1935, two essays on Pasternak's prose: "Kontury Glejtu," 44 an introduction to the Czech translation of Pasternak's autobiography Safe Conduct, and an expanded version of the same, "Randbemerkungen zur Prosa des Dichters Pasternak." 45 Here he disclosed that the basic device responsible for the puzzling originality of the poet's prose is its strenuously metonymic structure. The world is presented as existing in a state of universal interrelation: anything may be interchanged with anything else which is or can be considered as contiguous to it. Objects shift, interpenetrate, and the categories of space and time change their roles dramatically. Most importantly, the central protagonist himself (or herself) is .. disintegrated into a chain of states," and his silhouette thus becomes " . . . difficult to grasp." 46 This occurs because the protagonist exchanges his own role with that of the world outside him, usually a cityscape or landscape. Although initially the outside world is presented as a fact of the character's perception, it becomes so completely autonomous and self-sufficient that it begins to act in place of the person perceiving it. Jakobson's observations on Pasternak's early prose may be confirmed and further explicated with reference to the final period of the poet's work. In Doctor ¿ivago, the metonymic description of the heroine - here, her own voice - reveals at the same time the very mechanism of making the metonymy: "Marina govorila negromko, no golosom, kotoryj byl sil'nee razgovornyx nadobnostej, i ne slivalsja s Marinoju, a myslilsja otdeVno ot

THE NEW POETICS

371

nee... etot golos byl ee zascitoj, ee angelom xranitelem."47 The very process of the emancipation of the element related to the character, its usurpation of the character's role, is shown here in progress. The contiguous relation between the heroine and her voice is obvious, although it is not a common synecdoche: her voice assumes her role (though it is not as fully developed as would have been the case in Pasternak's early prose) and acts instead of her. At the same time Pasternak stresses that it is not similar to her, but on the contrary different, one might even say alien to Marina's being ("ne slivalsja s neju"). Thus, no matter how much resemblance can be seen between the object and its metonymic representation, the very basis for the choice of a metonymy remains contiguity and not similarity. Pasternak's prose fascinated the investigator and stimulated his findings because of its expressive and consistently relational character; the principle of mere mechanical causality plays no role in Pasternak's world and is even denounced by him as wholly false. This is part of a larger phenomenon the similar tendency of the whole avant-garde art of the time. Jakobson underscores the fact that his theories, especially in phonology, were strongly stimulated by the art and literature of the Russian and international avantgarde. 48 Metaphor and metonymy, though viewed as binary poles and thus corresponding in principle to the two polar genres of poetry and prose, are by no means confined to these two genres. Metaphor and metonymy, which are "born" within the axes of similarity and contiguity, have an independent existence and can function as pure tropes in both poetry and prose. Pasternak, for one, is a poet of metonymy.49 The fact that metonymic structure was originally elaborated by him in his poetry and only then "transferred" to his prose makes the metonymic character of the latter genre all the more palpable: metonymy now not only serves as the natural principle of concatenation, but becomes an emancipated device whose function is to emphasize the relational character of the presented world, in which every element of imaginary reality can exchange its place and interrelate with neighboring elements. On the other hand, when metonymy becomes the dominating figure in poetry, the basic structure of which is metaphoric in essence, its palpability as a device is reinforced by the contrast between the basis and the superstructure. Jakobson's further analyses of other great Slavic poets, such as the Slovak Janko Krai or the Pole C. K. Norwid, show how strongly metonymic the whole work of a poet or at least his individual poems can be.50

372

KRYSTYNA POMORSKA

"POETIC MYTHOLOGY" AND BIOGRAPHY

The correspondence between a writer's "life" and "work" has been treated time and again in literary studies. The obvious and rather sorry result of such attempts was a type of "vulgar biographism" which, as Jakobson notes, " . . . takes a literary work for a reproduction of the situation from which it originated and infers an unknown situation from a work." 51 Another tendency pointed out by the same author was a reactive and equally unconsidered "anti-biographism" which unequivocably denied any interdependence between the two series or at least refused to treat it seriously. One can add still another variant of "vulgar biographism" in literary studies - vulgarized Freudianism. This variant, which insists on the causal relation between biography and literary creation, reduces the biographical background to a standard set of motifs constantly generating an identical set of symbols in the creative work, thus abolishing the work itself as an individual entity. In his studies of the 1930's Jakobson worked out a rational medium between these two opposite (biographical and anti-biographical) approaches to the relation of a writer's "life" and "work." He came to the conclusion that one cannot ignore the situation in scrutinizing a creative work, since "the situation is a component of speech" and "the work is never indifferent to it," although at the same time the work may include the situation either negatively or positively; and the situation itself is always to an extent "transformed by the poetic function." 52 On the other hand, Jakobson insists, one must pay careful attention to the actual symbolic system of a given work, which always establishes a regular pattern comparable to the regularity of the poet's personal unique rhythm or his characteristic "intonation." With this in mind Jakobson succeeded in elucidating a strikingly regular joint pattern in three of Puskin's major works - "The Stone Guest" (1830), "The Bronze Horseman" (1833), and "The Fairy Tale of the Golden Cockerel" (1834)-all of which, significantly, were written during the three autumn sojourns of the poet at his estate of Boldino. Although these three works belong to different genres, all of them exhibit a basically identical plot kernel composed of three motifs: (1) a weary man, longing for rest, is seized with desire for a certain woman; (2) a statue has supernatural power over the woman; (3) the works end when the statue, miraculously set into motion, kills the man, while the woman vanishes. Such are the basic constituents of Puskin's "sculptural myth," a myth which evolved gradually throughout his entire work. While in

THE NEW POETICS

373

Puskin's shorter poems the myth had been resolved in the modality of "the lasting sculpture and a vanishing man," in the long poems "the sculpture kills the man." An analysis of the second component of the relation, the life situation, reveals that it does not consist only of so-called "hard facts," such as, e.g., the ties and tensions between the poet and other actual persons (his father, his fiancée and later wife, Nathalie Goncarova, her family, the Tsar and his retinue, etc.), but also includes a number of intermediary factors and objects which are already symbolic in nature. One of the intriguing roles is played here by a real statue of Catherine the Great belonging to Puskin, which he and the Goncarovs tried to sell in order to raise money for the poet's wedding and subsequent living expenses. Puskin's constant references to this statue in the letters to his fiancée and others are whimsically intertwined with the images of his poems, epigrams, and drawings. The repetitive phrase used by Puskin to characterize the monument of Catherine, "the bronze Grandmother" ("bronzovaja babuska"), mixes the sculpture with the real person and thus obliterates the sharp borderline between object and symbol, or life and art. On the other hand, the poetic myth of the destructive statue, while corresponding to the peripeteias of the poet's life, should not be considered as being generated by them. In many instances it would be more fitting to say that the poet anticipated his life story with his own myth about other people's lives. The poet lives in a myth which he himself creates and consequently directs his life - more or less subliminally - towards the pattern created. In short, according to the results of Jakobson's analysis, not only is the life situation active in the process of literary creation, but the product created is likewise active and often decisive in the poet's actual biography. Thus, one cannot draw a sharp borderline between the two, nor can one generate one from the other unequivocably. Both sequences are in a relation of complementarity and form a kind of feed-back system. Mutatis mutandis the same concept can be found in Jakobson's essay on Majakovskij, "O pokolenii, rastrativsem svoix poètov," 53 written seven years earlier, which discusses Majakovskij's suicide and its poetic anticipation. It is characteristic that such works appeared in the 1930's, when general historical events and their own experience compelled scholars to take an interest in the character of poets' lives.5* Moreover, the scholarly observations actually find support in the autobiographies of the modern poets themselves. The two greatest contemporary Russian poets, Majakovskij and Pasternak, wrote their own life stories at the same time that such

374

KRYSTYNA POMORSKA

scholars as Jakobson and Mukarovsky were discussing the problem of the poet's biography from a scholarly point of view. Pasternak's Safe Conduct (1930) and Majakovskij's "I - Myself" (1922-28), each in its own way, offer a similar concept of the poet's life, in which "hard facts" mean nothing in and of themselves and are immediately transformed by the protagonist who experiences them into symbolic and strictly subjective phenomena. Thus a poet constructs his own image of life, which is at the same time his creation in a strictly symbolic sense. As Majakovskij writes: "I am a poet. I am interesting only by this fact. I write [in my autobiography] only those things which have been fermented in the Word." Or, as Pasternak puts it: "The poet gives to his life so steep an incline that it can not be found in the biographical vertical... I think that only heroes deserve a real biography, b u t . . . the history of a poet is not to be presented in such a form. One would have to collect such a biography from unessentials."

PARALLELISM

Throughout his entire career Jakobson has insisted that parallelism is the essential principle to which all the basic devices in poetry can be reduced. In the period of "Linguistics and Poetics" he suddenly found full support for his ideas in the remarkable critical works of the poet Gerard Manley Hopkins, which had only recently been published for the first time. 55 The young Hopkins observed over one hundred years ago that "The structure of poetry is that of continuous parallelism, ranging from the technical socalled parallelism of Hebrew poetry and the antiphones of Church music to the intricacy of Greek or Italian or English verse." 56 In a similar way the Russian scholar traced the role of parallelism in Slavic poetry from the Middle Ages to the present day. 57 The summit of Jakobson's investigation of this basic phenomenon is his study of 1966, devoted for the most part to the Russian material, "Grammatical Parallelism and Its Russian Facet," 58 which culminates the work he had begun in this area as early as 1917, in Baku, when he turned to an analysis of the famous folklore text "Gore-Zlocastie" after reading A. Veselovskij's work on historical poetics.59 "Grammatical Parallelism..." presents a meticulous analysis of all the various correspondences manifest in poetic structure - from the phonological to the thematic level - showing at the same time how each subsequent level grows out of the preceding one. It becomes clear from the analysis that the juxtaposition of two elements immediately sets the prin-

THE NEW POETICS

375

ciple of similarity to work: either the elements are juxtaposed because they are similar, or they become similar because they are juxtaposed. The essential point here is the switch from the formal (phonological and grammatical) to the semantic level. A basic example is the phenomenon of rhyme. Since a particular pair of words, e.g. rozy - grjozy, are similar in sound, a semantic affinity is automatically established: both rozy and grjozy are assigned to the same poetic and psychological field. Various puns, built on the same principle, carry the semantic similarity even further, as is the case with Puskin's humorous epigraph to the second chapter of Evgenij Onegin: "O, rus! / O, Rus'!" Here "Rus"'clearly equals the Latin "rus," as is confirmed by the first lines of the following stanza and, further, by the theme of the chapter in question. These problems are discussed by Jakobson in "Linguistics and Poetics" on the basis of a large number of examples ranging from folklore to the refinements of Pasternak's poetry. It is logical and legitimate that parallelism has been Jakobson's life-long interest, since it represents the basic relational principle of poetic language. As the scholar himself testifies: "Perhaps the strongest impulse toward a shift in the approach to language and linguistics... was - for me, at least the turbulent artistic movement of the early twentieth century.... Stravinsky with his 'search for the One out of the Many reveals the core of his work when he reminds us 'that one precedes the many', and that 'the coexistence of the two is constantly necessary'."60 Jakobson also speaks for his own generation of scholars: "Those of us who were concerned with language learned to apply . . . the pictorial theory and practice of cubism, where everything 'is based on relationship' and interaction between parts and wholes, between color and shape, between the representation and the represented." 61 And he quotes the penetrating declaration of the painter Braque: "I do not believe in things, I believe only in their relationship." Braque's words might well serve as a statement of Jakobson's own creed in his theory of poetics. Massachusetts Institute of Technology

NOTES 1

Cf., for example: Z. Lempicki, "Teoria ewolucji w historii literatury," Pami^tnik Warszawski, 1930, II, 4-5; also "Forma i norma," Z zagadnien poetyki, nr. 6 (Wilno, 1937). 2 O ¿esskom stixe preimuscestvenno v sopostavlenii s russkim (Berlin-Moscow, 1923;

376

KRYSTYNA POMORSKA

Brown University Slavic Reprint, Providence, 1969). There is a German translation, Über den tschechischen Vers [=Post ilia Bohemica 8-10] (Bremen, 1974), and a partial French translation, "Principes de versification," Questions de poétique (Paris: Seuil, 1973), 40-55. 3 Cf. Andrej Belyj, Simvolizm (Moscow, 1910) ; Ritm kak dialektika i "Mednyj vsadnik" (Moscow, 1929). 4 Prace Filologiczne, XVIII: 1 (1963), 47-52; cf. the Russian version, " K lingvisticeskomu analizu russkoj rifmy," Studies in Russian Philology (Ann Arbor, 1962), 1-13. 5 Slavic Poetics: Essays in Honor of Kiril Taranovsky (The Hague: Mouton, 1973), pp. 239-252. 6 Style in Language, ed. T. A. Sebeok (Cambridge, Mass.: M.I.T. Press, 1960), pp. 350-377. 7 Wiadomoici literackie (1934). 8 These two domains can be and are in fact viewed by the author himself as one area ("poetry of grammar and grammar of poetry"), but for reasons of classification it is convenient to distinguish them. 9 Novejsaja russkaja poèzija. Nabrosok pervyj: Viktor Xlebnikov (Prague, 1921). Cited according to the republication in Texte der russischen Formalisten, II : Texte zur Theorie des Verses und der poetischen Sprache, ed. W.-D. Stempel (Munich: Fink, 1972), 18-135 (with facing German translation). Partial English translation in Major Soviet Writers: Essays in Criticism, ed. E. J. Brown (London and New York: Oxford University Press, 1973), pp. 58-88. 10 See note 6 above. 11 "Poèzija grammatiki i grammatika poèzii," Poetics. Poetyka. Poètika (Proceedings of the First International Conference of Work-in-Progress devoted to Problems of Poetics) (Warsaw: PWN, 1961), pp. 397-417; "Poetry of Grammar and Grammar of Poetry," Lingua XXI (1968), 597-609. Cf. also the German version in Mathematik und Dichtung, ed. H. Kreuzer (Munich, 1965). 12 " O pokolenii, rastrativäem svoix poètov," Smert' Vladimira Majakovskogo (Berlin, 1931), 7-45 (second edition, The Hague: Mouton, 1975). English translation in Brown, ed., Major Soviet Writers, pp. 7-32 [cf. note 9 above]. 13 "Socha v symbolice PuSkinovë," Slovo a slovesnost, III (1937), 2-24 ; English translation in Puskin and His Sculptural Myth, translated by J. Burbank (The Hague: Mouton, 1975), pp. 1-44. 14 "Linguistics and Poetics" [see note 6 above], p. 358. 15 E. Holenstein, "A New Essay Concerning the Basic Relations of Language," Semiotica, XII, 2 (1974), 123. 16 Cf. for example, Novejsaja russkaja poèzija, p. 104. It is interesting to note that, despite the lack of a phonological approach in his actual analyses, Jakobson apparently had in mind the notion of the phoneme: "Evfonika operiruet ne so zvukami, a s fonem a m i . . . " (p. 94). 17 See: "Quelles sont les méthodes les mieux appropriées à un exposé complet et practique de la grammaire d'une langue quelconque ?" Premier Congrès Internationale de Linguistes (Nijmegen, 1928), pp. 36-39. 18 NovejSaja russkaja poèzija, p. 104. 19 Ibid. 20 Ibid., p. 94. 21 Quoted ibid., p. 96. 22 Ibid., p. 98. 23 "Poetyka w iwietle jçzykoznawstwa," Pamiçtnik literacki, LI (1960), 431-473. Cf. Jakobson's observations on this and other figures common to all Slavic poetries: "The Kernel of Comparative Slavic Literature," Harvard Slavic Studies, I (1953), pp. 9 ff.

THE NEW POETICS 24

377

Cf. "Observations sur le classement phonologique des consonnes," Proceedings of the Third International Congress of Phonetic Sciences (Ghent, 1939), pp. 34-41. 25 Cf. for example the analysis "Razbor tobol'skix stixov RadiSCeva," in XVIII vek, VII: Roi' i znacenie literatury XVIII veka v istorii russkoj kul'tury (Leningrad, 1968), pp. 228-236. 28 In Puskin Symposium (New York University Slavic Studies, Vol. 1), ed. A. I. Kodjak and K. F. Taranovsky (New York, 1976), pp. 3-26. 27 Ibid. 28 See note 11 above. 29 Cf. "Na okraj lyrickych bàsni PuSkinovych," Vybrané spisy A. S. Puskina, ed. A. Bém and R. Jakobson, I (Prague, 1936) [English translation in R. Jakobson, Puskin and His Sculptural Myth (The Hague: Mouton, 1975), pp. 45-50] and "PuSkinovy bâsnë v prekladu Ilji Bârta," Slovo a slovesnost, III (1937), 122-124. 30 Cf. "La facture d'un quatrain de PuSkin," La fête du texte-. Mélange Zumthor, ed. E. Vance (forthcoming). 31 "On Linguistic Aspects of Translation," Selected Writings, II : Word and Language (The Hague: Mouton, 1971), 260-266. 32 "Boas' View of Grammatical Meaning," Selected Writings, II, 489-96, and B. L. Whorf, Language, Thought, and Reality (Cambridge, Mass., 1956). 33 Cf. especially Edward Sapir, Language (New York, 1921). 34 "Linguistics and Poetics," p. 377. 35 Cf. O iesskom stixe..., p. 105. 36 Cf. Jakobson's recording of the following poems: "NaS marS," "150 000 000" (a fragment), "Celovek," "Pro èto" (fragments), deposited in the Lamont Library Poetry Room, Harvard University. 37 "Poèzija grammatiki..." [cf. fn. 11 above], 38 Cf. "Morfologiëeskie nabljudenija nad slavjanskim skloneniem" (1958), Selected Writings, II, 154-83. 39 Ibid., pp. 157 f. 40 Cf. K. Pomorska, Themes and Variations in Pasternak's Poetics (Lisse: Peter de Ridder Press, 1975), pp. 16 ff. 41 See K. Pomorska, Russian Formalism and Its Poetic Ambiance (The Hague : Mouton, 1968). 42 See the pioneering work by Mikolaj Kruszewski, "Zagovory kak vid russkoj narodnoj poèzii," Izvestija Imperatorskogo Varsavskogo Universiteta, 1876. 43 R. Jakobson and M. Halle, "Two Aspects of Language and Two Types of Aphasie Disturbances," Fundamentals of Language (The Hague: Mouton, 1956). 44 B. Pasternak, Glejt (Prague, 1935), pp. 149-162. 45 Slavische Rundschau, VII (1935), 357-374. English translation: "The Prose of the Poet Pasternak," in Pasternak: Modern Judgments, ed. D. Davie and A. Livingstone (Nashville, Tenn.-London: Aurora, 1970), pp. 135-151. 46 "Randbemerkungen..."; my translation. 47 B. Pasternak, Poktor ¿ivago (Ann Arbor: University of Michigan Press, 1958), p. 490. . v 48 "Retrospect to Selected Writings, I, pp. 632 f.; "Acknowledgements and Dedication," Selected Writings, II, pp. vi ff. 49 "Randbemerkungen..." 50 "The Grammatical Structure of Janko Krâl's Verses," Sbornik filozofickej fakulty University Komenského, XVI (1964), 29-40; "Przesztosé' Cypriana Norwida," Pamiçtnik literacki, LIV (1963), 449-450. 51 "The Statue in PuSkin's Poetic Mythology," Puskin and His Sculptural Myth (The Hague: Mouton, 1975), p. 3.

378 62

KRYSTYNA POMORSKA

Ibid. See note 12 above. 54 See T. Winner, "The Creative Personality as Viewed by the Prague Linguistic Circle: Theories and Implications," American Contributions to the Seventh International Congress of Slavists, I (The Hague: Mouton, 1973), 361-376. 55 The Journals and Papers of Gerard Manley Hopkins, ed. H. House and G. Storey (London: Oxford University Press, 1959). 58 Quoted in R. Jakobson, "Grammatical Parallelism and Its Russian Facet," Language, 42, 2 (1966). 57 Cf., e.g., "The Kernel of Comparative Slavic Literature," Harvard Slavic Studies, I (Cambridge, Mass.: Harvard University Press, 1953). 88 Cf. note 56 above. 69 See: A. Veselovskij, Istoriieskaja poitika (Leningrad, 1940), reproducing Veselovskij's works, edited and prefaced by V. 2irmunskij. 60 "Retrospect," Selected Writings, I, pp. 632 if. 61 Ibid. 63

R O M A N JAKOBSON A N D THE P R A G U E SCHOOL MIROSLAV R E N S K ?

school of linguistic thought and analysis established in Prague during the early 20th century by the Russian linguist Nikolay Trubetskoy (1890-1938) and the Czech-born U.S. linguist Roman Jakobson (born 1896)... The New Encyclopaedia Britannica, Micropaedia V (Chicago, 1974) PRAGUE SCHOOL,

The above quotation from what its publishers like to refer to as "the embodiment of authority" provides a convenient starting point for a retrospect of the close, complex, and fruitful relationship between Roman Jakobson and the Prague school; and the charming slip about Jakobson's provenience is not the only reason. With due fairness to the New Britannica, it must be added that the entry on Jakobson in the Micropaedia, while far from flawless, does more justice to his biographical data, and that John Lyons' article on linguistics in the Macropaedia presents a reasonably balanced picture of the school's multinational character and its contribution to the progress of linguistic thought and analysis. However, it is the short entries that are statistically more influential and get perpetuated in textbooks and other derived sources, and it is unfortunate that the old myth about the exclusive (and joint) leadership of the two great Russians in the Prague school is reinforced in 1974. The tenacity of such statements lends additional justification to the tendency of some Prague school historiographers to emphasize the local tradition. Vachek's Reader1 has been criticized for this, 2 and his introduction to the Prague school, 3 the most comprehensive source of information on the school, has a similar compensatory slant. Since the conscious and necessary emphasis of the Prague Circle was on phonology, where the two Russians did assume an unquestioned hegemony, further supported by their productivity, a legend about Russian theorists and their Czech epigones naturally followed in the wake of the school's fame. Vilem Mathesius, the real founder of the Prague Circle, saw

380

MIROSLAV RENSKY

it necessary to dispel this myth as early as 1936. While demonstrating its injustice to the Czech roots of the school and to the mutual give-and-take symbiosis of its truly international membership (with due credit to the uniqueness of the Russian contribution), he exposed the questionable value of such parochial evaluations: "Even if [this assertion] were true, it would hardly be objectionable because everywhere in the world progress in scientific research consists, for the greatest part, in the development and new application of ideas taken from elsewhere." 4 When free of provincial nationalism, even a superficial search for the specifically Czech and Russian contribution to the Prague school does indeed reveal many early and significant areas of common interest, and of common indebtedness to Western philosophy and culture. Thus, Mathesius' early paper, 5 which was to become a universally recognized proof of the Czech roots of the Prague Circle's theory and of its concept of dynamic synchrony (later opposed to Saussure's dichotomy), contains - long before Saussure's Cours was introduced to Prague - a lucid appreciation of Kruszewski, who was at least equal in importance to his teacher (and Saussure's precursor) Baudouin de Courtenay. And, although Russian linguists were not then familiar with Mathesius' versatile quest for alternatives to the Neogrammarian approach, A. Saxmatov's enlightened criticism of the work of J. Gebauer, the master of Czech historical grammar, attests to the critical attention of the Moscow school to Czech linguistics. 6 The atomistic historicism of the Neogrammarian doctrine, the opposition to which was to become the strongest and most universal bond of the emerging structuralist movement, inevitably found its critics, of varying caliber and theoretical depth, both in Prague and in Russia (as well as elsewhere). The Russian structuralists, however, had several advantages over their Czech colleagues. Trubetzkoy and Jakobson received their linguistic training in Fortunatov's Moscow school, which represented - in Meillet's judgment - the ultimate methodological and philosophical refinement of the Neogrammarian trend. As its dialectic opposite, they had Baudouin's established school of thought to draw on and, in Jakobson's case at least, an early input of the Cours through S. Karcevskij. However, it was probably the continuous Russian tradition of Hegelian and post-Hegelian dialectics that gave them the necessary theoretical and methodological power to grasp the significance of these opposites, to transform them into an organized system, and to proceed from programmatic statements to immediate and extensive application. Hegelian phenomenology also provided the propaedeutic basis for the enthusiastic discussions of Husserl's Logische Untersuchungen in the Moscow Circle, founded by Jakobson and

THE PRAGUE SCHOOL

381

other first-year students in 1915. The Czech and Slovak Hegelians left a less manifest imprint on Czech linguistic thought, where the influence of Husserl (though physically much closer) was less immediate, his acceptance more guarded, and where the ultimate break with the positivist tradition of the second half of the nineteenth century and the search for a teleological model were more pragmatic and eclectic. Paradoxically, but not surprisingly, this "internalized" Hegelian tradition may thus well be the most important specifically Russian contribution to the Prague workshop. One can also speculate, in this connection, on another "Russian" feature of the Russian scientific tradition, the "Turanian spirit," which apprehends the crucial nodes of its subject matter and their interrelationship in its totality through a daringly simplified but sophisticated schematism. 7 It is difficult to judge to what extent this tradition is responsible, along with Jakobson's personality, for the more specifically Jakobsonian donations to the Prague fund, i.e., for his linguistic and cross-disciplinary versatility, and for his uncanny ability to CONNECT. The means-ends relation between language and poetry, the atmosphere of the period, and the physical and academic youth of its members made the Moscow Circle concentrate on poetics, and this early interest of Jakobson's flourished in his collaboration with the Petersburg OPOJAZ (Society for the Study of Poetic Language), where he became the leading theoretician of the formalist school. In Prague (where he came as a cultural attaché of the first Soviet mission in 1920), he found a congenial atmosphere and fertile soil for this activity and immersed himself in this new medium with his characteristic vigor. His first extensive studies, devoted to Czech poetry, laid the foundations of a modern theory of verse, especially of the comparative study of verse systems. Many of his subsequent studies deal with the history of Czech verse and with prominent Czech and Slovak poets. Jakobson's evaluation of Old Czech poetry, especially his reconstruction of its relationship with the Old Church Slavonic tradition and his outstanding studies of the poetry of the Hussite period, will probably remain his greatest contribution to the cultural history of Czechoslovakia. As a passionate admirer of the greatness of the Czech cultural tradition, Jakobson became its militant defender against the attacks of the pseudoscientific Nazi propaganda of the 1930's. His intimate relationship to Czechoslovak culture was to continue after his flight from Nazi persecution in 1939. In his book about the wisdom of ancient Czechs, 8 he mobilized the inspirational resources of Czech literature to support anti-Nazi resistance. His interest in Czechoslovak culture has continued to the present day, as

382

MIROSLAV RENSKY

attested, e.g., by his analysis of the Hussite song Ktoz jsti bozi bojovnici,9 a rallying call of the Czech people against oppression of all kinds ever since the fifteenth century. Jakobson's position in the Russian formalist school and in the literary wing of Prague structuralism was not only the result of his scholarly exploits. His close ties to the Russian literary and artistic avant-garde, which he maintained for as long as it was possible, were paralleled by his intimate relationship with the Czech avant-garde. His role as its inspirer, consultant, and passionate defender in endless debates must have taken a heavy toll on the time and energy he could have used for his rapidly ramifying linguistic work, especially during the latter years of his stay in Czechoslovakia, when his drawing a line between the real function of avant-garde poetry and cheap utilitarianism led to bitter polemics within the avant-garde and when, in the complex situation of the late 1930's, what he originally came to interpret was increasingly difficult to defend and yet had to be defended in the looming shadow of fascism. This cursory reminder of Jakobson's extralinguistic activities during his Czechoslovak years cannot hope to compete with their evaluation by experts in the several disciplines involved. However, it is only against this backdrop that one can fully appreciate the scope, depth, and intensity of his more specifically linguistic work of this period, and his equally spirited involvement in the pragmatic linguistic pursuits of the Prague school. Jakobson's role in the formation of the Prague-based linguistic movement cannot be reduced to a comparison with the role of Mathesius, and yet the juxtaposition of their concepts, goals, and personalities is revealing in many ways. When - barely three months after his arrival in Prague - the prodigious novice, burning with the ambitious projects of his own background and eager to embrace the new ambience and expand into its area, called on Mathesius, he met not only a mature professor of English language and literature, but a great scholar and thinker whose impact must have been striking. Although this was two years before the eye disease, fatal to a literary scholar, which was to make him concentrate fully on linguistics, Mathesius had already formed a clear and well-rounded concept of his "linguistic characterology," a system-oriented functional analysis of languages based on their analytic comparison and on a consistent procedure from speaker to hearer, from function to form, which he viewed as complementary but superordinated in importance to the formal approach. His speech-act approach made him replace the languejparole dichotomy with that of functional onomatology and functional syntax. 10 In the latter, an early application of his principles to the comparison of Czech and English

THE PRAGUE SCHOOL

383

led to his recognition of an additional communicative level, the functional sentence perspective, for which he is best known in today's linguistics. True to his era and tradition, Mathesius was a theoretical and methodological pluralist, reluctant to accept unexceptionable laws and reductionist theories. Where Jakobson tended to reduce Mathesius' "interplay of tendencies" to a set of hierarchically ordered rules (in modern parlance) and saw the necessity of a rigorous and schematic analysis of one subsystem before moving to integration, Mathesius' first concern was to maintain the awareness of the structural and functional complexity of language. While Jakobson (and Trubetzkoy) pulled the Prague doctrine toward scientism and incipient formalism, Mathesius steered closer to the humanist approach in his quest for crucial problems rather than for exclusive solutions, guided by his belief in linguistic consciousness. And while there was enough room in Mathesius' flexible model to accept such "scientistic" exploits as long as they fit the facts and the presentation was not dogmatic, much of what he really wanted to accomplish is reflected in Jakobson's recent integrationist pursuits, enriched by the wisdom of the achievement and the trial-anderror lessons of the intervening decades. An excellent educator, Mathesius tried to reconcile theoretical and pragmatic linguistics, and the new logical orientation with local and international tradition. A cultural activist, he emphasized the responsibility of linguists for the cultivation of stylistically differentiated usage free of any prescriptivist bias, and the need for active participation of all scholars in the rebirth of Czechoslovak cultural life through international contacts. An enlightened patriot, he did not hesitate -unpopular as it was in a nation most exposed and sensitive to racist theories about Slavic inferiority - to acknowledge the geographically and historically dictated limits of the role of Czech culture and scholarship as integrators of cross-cultural currents and raise them to a new level of recognition as a positive principle (most eloquently defended by Jakobson). Independent testimonies exist of the profound respect Mathesius and Jakobson had for each other. They must have instantly recognized their common goals, their differences, the potential for cooperation and division of labor. Jakobson urged a forum analogous to the Moscow Circle; Mathesius advised a gradual approach. When several small and informal gatherings broke the ground for the formal initiation of the Prague Circle (1926), Jakobson, as one of its first speakers (January 1927), presented a masterpiece, his Concept of the Sound Law and the Teleological

Criterion,n

the birth certificate of the Prague school and a milestone of linguistic thought. Developed and formalized in Jakobson's proposition - endorsed by Trubetzkoy and Karcevskij - to the Hague congress of 192812 and per-

384

MIROSLAV RENSKY

suasively documented in his 1929 monograph Remarques sur révolution phonologique du russe,13 it vindicated the basic Neogrammarian tenet of the operation of sound laws against its negative criticism by restricting it to functionally equivalent subsystems. Substituting a teleological approach for the mechanistic view, it simultaneously bridged two Saussurian dichotomies (synchrony/diachrony, languejparole) and spelled out the foundations of Prague structuralism and functionalism by postulating the concept of language as an organic and goal-directed system, constantly responding to its speakers' needs by developing or adopting new means, while preserving the structural and functional balance within and between its subsystems and subcodes. More pragmatically, these early studies of Jakobson's (who by then was already cooperating with Trubetzkoy) were to stake out the claim to phonology for the two Russians and establish their undisputed authority in this discipline for years. A familiar and fascinating discovery route was to follow, from correlations and archiphoneme (both credited to Jakobson) to emerging binarism and markedness (credited to Trubetzkoy), and to the ultimate reduction by Jakobson of the phoneme to distinctive features, with additional support from his investigation of aphasia and language acquisition. Jakobson's "Retrospect" in his Selected Writings I 1 4 and T. Sebeok's extensive and insightful review of this volume 15 are unsurpassed guides to this area of his Prague activities and their later development. In this connection, a remark seems to be in order. While even the resident core of the Prague school, i.e., the Prague Circle, was far from monolithic, and while Jakobson's and Trubetzkoy's rapid though circuitous progress did throw the rest of the group out of step (a significant critical response emerges only in the 1930's, while the theory was still taking shape), there was no resistance to this progress within the group, as one might conclude from Sebeok's remark: [Jakobson's] dissolution of correlative phonemes into their common core and differential property, announced in 1928, was already light-years of foresight ahead of the Prague School definition, offered as a standard in 1931 and accepted without any debate for at least five years thereafter : "Phonème.. .Unité phonologique non susceptible d'être dissociée en unités phonologiques plus petites et plus simples." Exasperatingly, this definition of the phoneme, which was antiquated thirty-five years ago, has not only survived to the present but continues to be inflicted on numerous students , of descriptive linguistics through at least one of our most popular textbooks.16 The definition itself makes more sense in context with the superordinated definitions of oppositions and units, and a case could be made - though

THE PRAGUE SCHOOL

385

Jakobson chooses not to make it - for a distinction between units and elements. More importantly, the definition is Jakobson's own, as is the whole 1931 "Projet" 1 7 with its impressive fourteen-page battery of his (and Trubetzkoy's) concepts (including markedness), most of which now obviously appear at variance with the definition. Jakobson would have been the first to record a minority statement (as did other participants of the 1930 Phonological Conference), and his remark, 18 which Sebeok draws on, can only be interpreted as a lament of someone still smarting from his own failure to see the configuration of evidence through the barrier of successiveness (which was to motivate another of his lifelong crusades against Saussurean oversimplifications, that against the linearity of the signans). Moreover, this infelicitous definition - if not its deplorable fame - survived barely one year, i.e., only till Jakobson redefined (for the 1932 supplement to the Czech encyclopaedia) the phoneme as "a set of those concurrent properties which are used in a given language to distinguish words of unlike meanings." 19 If the Hague congress of 1928 was the school's international baptism, the 1929 theses for the First International Congress of Slavicists 20 represent its constitution. Thanks to some investigative reporting by J. P. Faye, we now have Jakobson's uncontested retrospect 21 on the division of work on the draft of this document. Apart from some special areas where six other members participated, the responsibility for the crucial areas of theory and method was divided largely between Mathesius and Jakobson. This statement of fields of interest, approaches, and goals is a persuasive document of the diversified and yet integrated interests of the school well beyond phonology, and of its specific notes in the polyphony of the worldwide drift of the 1930's toward a means-ends model (as later interpreted by Jakobson 22 ). It also reminds us of the active role Jakobson assumed in the development of all these areas down to the ninth (and last) item on the checklist of the 1929 theses (standard language and the culture of language). The need for a pragmatic application of the school's theoretical interest in this area was especially pressing because of the specific Czech situation with its lack of continuity between common and literary Czech, and a revival of purism after the liberation of 1918. Jakobson, though busy at the time with several major projects, selflessly lent all his erudition and eloquence to the successful campaign of his colleagues against the purists, even though - as a non-native speaker who had just learnt Czech - he had to anticipate a denial of credentials. As for the less Czech-specific areas, here again the best review of Jakobson's Prague period is provided by the selections in his Selected Writings II,

386

MIROSLAV RENSKY

and by this volume's engrossing "Retrospect" 2 3 of the philosophical roots of structuralism and its progress, viewed through the prism of dialectic negation of Saussure's tenets. This volume also contains three articles on the Prague school, interestingly complementary as their focus moves from the search for its roots and identity 24 to a mature insider's evaluation, 25 and to a retrospective integration with the general current of the 1930's. 26 Jakobson's very first sortie into linguistic areas beyond phonology and phonology-related poetics, his 1932 study "Zur Struktur des russischen Verbums," 2 7 turned out to be another trailblazer. Drawing on the scattered impulses of the Russian linguistic tradition from Vostokov to Peskovskij, and on the general principle of asymmetry of the linguistic system as exemplified best by Karcevskij, 28 Jakobson discovered an important general principle of morphological (and, potentially, syntactic) structure by extending the notion of phonological markedness to the general meanings of morphological categories. The two members of a morphological correlation are characterized by the presence or absence of a distinctive feature, formal and semantic. The presence of feature A in one member does not imply B (nor necessarily non-A) in the other member but rather its indifference to both. Formal markedness may, but need not be, paralleled by semantic markedness, the relationship between the two being arbitrary and often chiastic; the same applies to the relationship between grammatical function and lexical meaning. Under favorable contextual conditions, the unmarked form can function as a representative of both members of the correlation. This principle was further developed in Jakobson's "Kasuslehre," 2 9 further formalized in "Signe zéro," 3 0 and later reinforced in Shifters,31 where it was extended to periphrastic verbal forms. While this more abstract notion of markedness became most popular in lexis (where it is more obviously limited), it was most extensively and fruitfully applied in subsequent Czechoslovak (and other Slavic) work on the morphology of individual Slavic languages (including derivation). As a result of such extensive verification, some aspects of Jakobson's theory were challenged on several grounds, best summarized as unwarranted or excessive reductionism. Secondary form-function relationships were found to be superimposed upon this basic (but not exclusive) asymmetry in some categories (e.g., aspectual), further reducing the arbitrary relation between form and function. This is quite in agreement with Jakobson's general campaign against arbitrariness as developed most persuasively in his "Quest." 3 2 By 1939, the Prague school consolidated its theoretical basis, produced a second generation of talented linguists, and enjoyed international recognition. Mathesius' vision of a home-grown and tradition-based and yet

THE PRAGUE SCHOOL

387

internationally oriented contribution to the newly liberated discipline of linguistics had come true. No one could have envisaged then what the impact of World War II was to mean for the development of Czechoslovak linguistics beyond the loss of Jakobson to American linguistics. In an atmosphere where the prestige and support of linguistics benefited - ironically - from the same war, Jakobson successfully transplanted the substantial core of his (and the Prague school's) theory to a new school of structuralist thought, and gradually turned to increasingly epistemological concerns. His Czechoslovak colleagues had to concentrate, for several reasons, on more pragmatic matters. Having absorbed the premature loss of several leaders (including Trubetzkoy and Mathesius) and the task of post-War reconstruction of universities devastated by Nazi occupation, they turned to existing and new large-scale projects (largely lexicographical and dialectological), sponsored by the state-funded research institutes of the newly organized Academy of Sciences. Though overdue, these projects were time-consuming and only rarely conducive to theoretical innovations. At the same time, Czechoslovak linguistics had to cope with the rapidly hardening plaster cast of the Stalinist era. The equally crippling effect of its two immediate vehicles, the anachronistic imposition of Marrism and its apparent antidote endorsed by Stalin himself, was perpetuated by the fact that Prague happened to be the cradle of the structuralist heresy and Jakobson a convenient target. The zeal of a powerful and tenacious dogmatist who devoted his life largely to eradicating structuralism and Jakobson from Czechoslovak history fanned the flames even during periods of relaxation. As late as 1963, by which time Jakobson had been elected to several East European academies, his name was still missing in a new encyclopedia, and not even the award of an honorary doctorate in 1968 was to end a situation where the very name of Jakobson (and structuralism) became one of the best barometers of the political climate. With due allowance for their legitimate differences and for the humiliating but necessary tactical needs of paying lip service to one doctrine while salvaging the other (and its material base), those of Jakobson's former collaborators and students who wielded some power (however restricted) in linguistics during these repercussions will have to live with the verdict of historiography for not defending his name with more courage and dignity. While this anachronistic witch hunt meant an incalculable loss of time and effort, and considerable damage to the leadership potential of the second generation, its net effect was to reinforce, in the minds of the postw a r generation of Czechoslovak linguists, the accomplishment of the Prague school of the 1930's, and to underscore their indebtedness to Jakob-

388

MIROSLAV RENSKY

son's share in it. It also helped to disseminate the Prague school tenets elsewhere in Eastern Europe and establish there, through bonds of common interest and mutual support, an emerging modern linguistic movement, strong enough to withstand attempts at Gleichschaltung and diversified enough to absorb the advent of the strongly model-oriented methods of the 1950's. Even though some Czechoslovak linguists have adopted these models, and many more have learned from mathematics and logic without adopting any formal model, the emphasis of present Czechoslovak linguistics is, in continuity with the pre-War tradition, on a delicate formand-function analysis of individual subsystems rather than on the construction of models. Of particular interest (beyond that of the Slavicists) is its recent work in semantically oriented syntax, discourse analysis, and stylistics. This can also be its most immediate contribution to the incipient integrationist trend in international linguistics, as increasingly more proponents or developers of various models become aware of their structuralist lineage and of the need to supplement their particular emphasis with input from other levels, functions, and/or approaches. And if some linguists still doubt whether their contribution can be meaningful in the worldwide context unless restated in terms of a formalized model, the best they can do is turn to Jakobson's recent writings. In Jakobson today, they will find an inexhaustible source of strength, inspiration, and wisdom, just as their precursors did in the golden age of the Prague school. City University of New York

NOTES 1

Josef Vachek, A Prague School Reader in Linguistics (Bloomington, Indiana: Indiana University Press, 1964). 2 For example, Robert Austerlitz, "Review of A Prague School Reader in Linguistics by Josef Vachek," Word 20 (1964), 458-65. 3 Josef Vachek, The Linguistic School of Prague (Bloomington, Indiana: Indiana University Press, 1966). 4 Vilem Mathesius, "Deset let Prazskeho linguistickeho krouzku," Slovo a slovesnost 2 (1936), 137-45. 5 Vilem Mathesius, "O potencialnosti jevu jazykovych," Vestnik Krai, ceske spolelnosti nauk, trida filosoficko-historicka (1911). 6 A. Saxmatov, "Otcet IV premii A. A. Kotljarevskogo," Izvestija Imperatorskoj Akademii Nauk, V serija, Vol. IX, No. 4 (1898), 17-105. 7 See Jakobson's obituary of Trubetzkoy, "Nikolaj SergejeviC Trubetzkoy," Acta Linguistica I (1939), 64-76; cf. Roman Jakobson, Selected Writings II (The Hague: Mouton, 1971), p. 502. 8 Roman Jakobson, Moudrost starych Cechu (New York, 1943).

THE PRAGUE SCHOOL

389

9 Roman Jakobson, "Ktoz jsu bozi bojovnici," International Journal of Slavic Linguistics and Poetics VII (1963), 108-17. 10 See Whorf's lexation and patternment. 11 Roman Jakobson, "O hlâskoslovném zàkonu a teleologickém hlâskoslovi," Casopis pro moderni filologii XIV (1928), 183-84; in English in Jakobson, Selected Writings I (2nd ed., The Hague: Mouton, 1971), 1-2. 12 Roman Jakobson, "Quelles sont les méthodes les mieux appropriées à un exposé complet de la grammaire d'une langue quelconque?" Premier Congrès International des Linguistes, Propositions, Nijmegen (1928), 36-39; also in Selected Writings I, 3-6. 13 Roman Jakobson, Remarques sur l'évolution phonologique du russe comparée à celle des autres langues slaves [= Travaux du Cercle Linguistique de Prague II] (1929); also in Selected Writings I, 7-116. 14 Roman Jakobson, "Retrospect" in Selected Writings I, 631. 15 Thomas A. Sebeok, "Review of Roman Jakobson's Selected Writings I," Language 41 (1965), 77-88. 16 Sebeok, "Review of Roman Jakobson's Selected Writings I," 78-79. 17 Roman Jakobson, "Projet de terminologie phonologique standardisée," Travaux du Cercle Linguistique de Prague IV (1931), 309-23. 18 Jakobson, "Retrospect," p. 635. 19 Roman Jakobson, "Fonéma," Ottùv slovnik naucny, Dodatky II (1932), 608. 20 "Thèses présentées au Premier Congrès des philologues slaves," Travaux du Cercle Linguistique de Prague 1(1929), pp. 5-29 ; reprinted in Vachek, A Prague School Reader in Linguistics, pp. 81-87. 21 See Le Cercle de Prague [= Change, III (1969)] and Change, IV, 224-26. 22 "Efforts toward a Means-End Model of Language in Interwar Continental Linguistics," Trends in Modern Linguistics II (Utrecht-Antwerp, 1963), pp. 104-08; also in Vachek, A Prague School Reader in Linguistics, pp. 481-85, and Selected Writings II, 522-26. 23 Selected Writings II, 711-22. 24 "La Scuola linguistica di Praga," La Cultura XII (1933), 633-641, in Selected Writings II on pp. 539-46; "O predpokladech prazské linguistické skoly," Index VI (Brno, 1934), 6-9. 25 "Die Arbeit der sogenannten 'Prager Schule'," Bulletin du Cercle Linguistique de Copenhague III (1938), 6-8; also in Selected Writings II, 547-50. 26 "Efforts toward a Means-End Model of Language in Interwar Continental Linguistics." 27 In Charisteria Gvilelmo Mathesio . . . oblata (Prague, 1932), pp. 74-84; also in Selected Writings II, 3-15. 28 "Du dualisme asymétrique de signe linguistique," Travaux du Cercle Linguistique de Prague I (1929), 33-38; reprinted in Vachek, A Prague School Reader in Linguistics, pp. 81-87. 29 "Beitrag zur allgemeinen Kasuslehre," Travaux du Cercle Linguistique de Prague VI (1936), 240-88; reprinted in Selected Writings II, 23-71. 30 In Mélanges de linguistique offerts à Charles Bally (Genève, 1939), pp. 143-52, also in Selected Writings II, 211-19; cf. "Das Nullzeichen," Bulletin du Cercle Linguistique de Copenhague IV (1940), 12-14, in Selected Writings II, 220-22. 31 Shifters, Verbal Categories, and the Russian Verb, mimeographed (Cambridge, Mass., 1957); reprinted in Selected Writings II, 130-47. 32 "Quest for the Essence of Language," Diogenes, No. 51 (Montreal, 1966), 21-37, reprinted in Selected Writings II, 345-59; cf. "À la recherche de l'essence du langage," Diogène, No 51 (1965) and Problèmes du langage (Paris, 1966), pp. 23-38; and "En busca de la essencia del lenguaje," Diogenes (Buenos Aires, 1965), pp. 21-35.

DISTINCTIVE FEATURE THEORY R. H. ROBINS

In his presidential address to the Linguistic Society of America in 1964, Charles Hockett selected as one of his four major "breakthroughs" in the history of linguistics what he called the "quantizing hypothesis," 1 when the phoneme unit was established as the basic phonological entity, and the consonant and vowel sounds of every language were made referable to finite sets of mutually contrastive phonemes. The early phoneme was conceived as a linear, sequential unit of vocal communication. It was the great achievement of the Prague School to break down this exclusive linearity by the recognition and systematic formulation of the minimal and much more basic phonological entity, the distinctive feature. Trubetzkoy wrote at the end of his life, in his Grundzüge der Phonologie: "Unter phonologischem Gehalt verstehen wir den Inbegriff aller phonologisch relevanten Eigenschaften eines Phonems, d.i. jener Eigenschaften, die allen Varianten dieses Phonems gemeinsam sind und es von allen anderen, vor allem von den nächstverwandten Phonemen derselben Sprache unterscheiden."2 It would, of course, be absurd to suggest that the concept of a finite set of phonological units was unknown before the second half of the nineteenth century, or that the idea of distinctive phonological features was revealed for the first time to Trubetzkoy and the Prague School. The phoneme concept is implicit in alphabetic writing, and was foreshadowed in comments on the grämmata of the ancient Greek alphabet ("charactSres men eisin eikosi tessares, ekphön^seis de pollöi pleious" 3 ); and phonemic principles were enunciated much more scientifically in the ancient Sanskrit phonetic treatises.4 It is, however, clear from nineteenth-century phonetic literature, especially the writings of Henry Sweet, that a number of distinctions associated with the classical phoneme were being clarified at about the same time: between contrastive difference and non-contrastive difference, between narrow (phonetic) transcription and broad (phonemic) transcription, be-

392

R. H. ROBINS

tween the virtually limitless number of phonetic differentiations observable and recordable and the strictly limited sets of contrastive sound units in any language, between general phonetics (phonetic universals) and the separate phonological systems of each language.5 The intimate association of phonemic analysis and broad transcription has been apparent in the successive writings of different schools of linguists. Sweet established the technical term "broad" in relation to transcription,6 though he did not use the term "phoneme." Daniel Jones made the phoneme the justification and the definiens of a broad transcription.7 The title of Pike's classic methodological and theoretical book is self-explanatory: Phonemics: A Technique for Reducing Languages to Writing.8 Equally revealing is the reason suggested by Harris for not pursuing componential analysis to its logical end: "Phonemics is undoubtedly the more convenient stopping point in this development, because it fits alphabetic writing, but we must recognize the fact that it is possible to go beyond it." 9 Transcription stresses the unitary character of the phoneme; but, again, the analysis of phonological segments into components and the establishment of proportional relations between them on this basis were achieved by Greek grammarians: p: ph: b:: k: kh: g:: t: th: d ; 10 and they are clearly implied in the organization of the Sanskrit syllabic alphabet. 11 It was the central concern of the Prague School to develop the phoneme into a basic scientific concept for linguistics, perhaps the first such concept to be so developed in the modern era of the subject. (The nearest parallel to this outside language has been identified by Jakobson in the structuring of the genetic code in human beings as revealed by F. H. C. Crick and others : 12 the transmission of inherited characteristics is controlled by messages in a kind of "language" that is made up of different meaningful sequences of four basic and themselves meaningless phoneme-like uhits, which in turn are analyzable into combinations of more primitive features standing in binary oppositions.) In this development one sees the first thorough application of de Saussure's structural theory of language, leading, in the next decades, to the general recognition of phonology as the pace-maker in synchronic linguistic theory (cf. Harris' application of the rigorous methods of phonemics to morphological analysis;13 Longacre's earlier systematization of phonology than of grammar; 14 Pike's modelling of general linguistic terms etic and emic on phonetic and phonemic15}. Trubetzkoy's debt to Saussurian thinking in his conception of phonology vis-á-vis phonetics is recognized in the first section of the introduction to his Grundziige,16 The passage from the Grundziige quoted at the beginning of this article was a culminating statement of the analysis of the phoneme into its more

DISTINCTIVE FEATURE THEORY

393

basic distinctive features. With this statement the implied analyses of earlier centuries were explicitly formalized, and every phoneme was exhaustively analyzed into a bundle of features, these features constituting the phonological foundation of the language. The theory behind this analysis had been worked out in earlier papers of the Prague School (e.g. Jakobson's Remarques sur l'évolution phonologique de russe17). It now became possible to deal with the description of different contrast systems operative at different places in structures, through the concept of neutralization, 18 and with the Grenzsignal function of features. This foreshadowed the Firthian delimitative prosodies and the American juncture phonemes, matters that the older transcriptional unit treatment of the phoneme largely left alone, as Daniel Jones explicitly said it should do. 19 Trubetzkoy's formal introduction of the concept of markedness20 made possible the subsequent wholly binary notation of features constituting phonemes, a characteristic of distinctive feature theory which, though not espoused by Trubetzkoy, became an essential part of it at the hands of most of his successors.21 In various ways the distinctive feature, as first developed in Prague School phonology, has been one of the most pervasive and influential theoretical concepts in linguistics throughout its subsequent history. One can trace its progress by considering it from two points of view: (1) in its broadening extension from phonology over other levels of linguistic analysis, and (2) in its successive vicissitudes within theoretical phonology.

I Essentially, distinctive feature analysis in phonology exploits the closed system of the phonological level, the most narrowly closed level of all, in that the empirical reality on which it rests, or from which it is abstracted, is itself restricted. Probably no definite limit can be placed on the actual number of articulatory and auditory differences that can be produced by the human vocal tract and recognized by the human ear; but the range within which these differences can appear is circumscribed by the physiological structures and capacities of the vocal tract and of the outer and inner ear. Grammar, more specifically in its morphological aspect, necessarily involves closed systems; a standard definition of the division between grammar and lexicon turns precisely on the division between closed and open classes and systems.22 Appropriately, therefore, in the extension of

394

R. H. ROBINS

its field of application the distinctive feature concept was first used in reference to parts of the morphological structures of languages. In this regard, Jakobson's feature analyses of the Russian verb and of the Russian noun case system are taken as pioneering works. 2 3 In them the Trubetzkoyan concepts of neutralization and markedness were applied in a systematic presentation of the meanings expressed in closed morphological paradigms. Bazell used markedness in relation to cases and in relation to syntactic functions. 24 The present author briefly examined the application of neutralization to the grammatical analysis of parts of Latin, Greek, German, and English morphology. 2 5 In his classic study of grammatical case, Hjelmslev undertook a feature analysis of the semantic systems carried by the case forms of a number of languages, setting up for various categories of meaning the oppositions of positive, negative, and neutral (thus for Greek the three oblique cases are, in respect of direction, accusative positive [motion to], genitive negative [motion from], and dative neutral [absence of motion]). 26 There are obvious similarities here to distinctive feature analysis, but it would seem that the Hjelmslevian system derives more from some aspects of logical and mathematical symbolization than from analysis at the phonological level (Prague phonologists are not cited in his bibliography). In Aspects of the Theory of Syntax Chomsky established the now familiar sets of distinctive features in syntax 27 (and pointed to the possibility of "purely semantic lexical features"), 28 expressly linking them with the Jakobsonian features in phonology and in particular drawing attention to the fact that syntactically complex categories can be analyzed more readily on the pattern of feature matrices than by simple subclassifications, which are often laden with redundancies. 29 It must be pointed out that some of these syntactic features are analogous to phonological features in constituting an inherent characteristic (e.g. + N ± common, ± count, ± animate, + V), but some others are context-sensitive or distributional features like ± N P and ± Det

A recent example of the further exploitation of this

sort of use is seen in Saunders' and Davis' work on Bella Coola, where the binary distinctions ± referential and ± salient are set up in connection with certain suffixes. 30 A s well as applying feature analysis to grammar and to the closed systems of grammatical semantics, several attempts have been made to develop its insights in the far more complex task of lexical semantic analysis. Bendix, seeking to extract various semantic components from items of vocabulary, cited Jakobson on Russian cases, and referred to the economical value of setting up marked and unmarked terms in relation to

DISTINCTIVE FEATURE THEORY

395

a specific component or feature. 31 Typical components are inherence, negation, relation, cause, activity, state and change. 32 Earlier, Brendal acknowledged, in his semantic study of prepositions, the connection between his treatment of meanings and what had been done in phonology by Trubetzkoy. 33 Hammel, introducing the volume of studies Formal Semantic Analysis, acknowledged Jakobson, among others, as having provided within linguistics a "self-conscious methodology" for feature or component analysis of terminological systems. 34 Bennett, like Brandal, investigates the meanings of prepositions, in his case in terms of stratificational grammar, which as a general theory of linguistic analysis can trace lines of derivation from Prague as well as from Hjelmslev, 35 and he draws attention to the distinctive feature analysis of the Russian case system by Jakobson. 36 The big problem in applying distinctive feature theory to semantic analysis is that in semantics we are passing into a level that comprises openended systems and open classes, a level wherein the language must be able to respond to anything and everything met with or conceived in the life and experience of its speakers. In one way the search for such an analysis is one version of the finite semantic inventory that was assumed by semantic universalists in the seventeenth century (e.g. J. Wilkins 37 ) and has more recently appeared in studies like that of Katz and Postal, who wrote 38 of "atomic conceptual elements," presumably finite in number, constituting the "markers" and "distinguishers" of their semantic system. In another way, of course, it is only tidying up what we all know from common sense, for example that bull is to cow as stallion is to mare. But the value of this sort of analysis for semantics in general remains problematical, once it is extended, as it must be on any assumption of semantic totality, beyond the semantics of specific syntactic categories and structures and the more or less closed lexical systems like prepositions and pronouns and such muchworked semantic fields as kinship systems. Indeed kin terms form the bulk of the material treated in Formal Semantic Studies, and Lamb points out in his contribution to this volume that there is no general difference between distinctive and non-distinctive semantic features as there is in the case of phonetic features. 39 And Bendix refers to his book as an attempt to extend feature or component analysis "into less obviously structured domains" than kinship, etc. 40 In this connection one may cite most recently O. Duchâcek's "Quelques observations sur les structures sémantiques." 41 In view of the limitless scope of the semantic functioning of language, it may well be that no one model of analysis and description will prove adequate to the task of explicating the semantic competence of a native or

396

R. H. ROBINS

a fluent speaker. But in any case, whether or not ultimately a system of analysis based on distinctive feature theory can be devised to cover all or virtually all aspects of meaning, there can be no doubt that distinctive feature theory has thrown light onto significant areas of linguistic meaning and has pointed the way to a rigorous ordering of at least part of the lexical resources of all natural languages.

II

Trubetzkoy's distinctive features were broadly those of traditional and contemporary articulatory phonetics. His innovation was to analyze phonemes exhaustively into combinations of limited numbers of such features, and to follow up the implications of such an analysis below the level of the phoneme as a simple unit, e.g. in reference to such concepts as neutralization, markedness, etc. As early as 1929 Jakobson was arguing that acoustic categories might provide a more revealing set of distinctive features for phonological investigations.42 He followed this in 1938 by proposing that the change of Romanian /k/ to /p/ before /t/ and /s/ could best be explained as a partial assimilation of the acoustic features grave and posterior (later compact) to grave anterior (later diffuse) in the environment of a following acute anterior,43 In this Jakobson looked back to the classic work of Stumpf. 44 Insofar as a prosopographical account is appropriate for the recent history of linguistics, Jakobson's personal fortunes in relation to distinctive features is surely one such case. His movements, under the flail of World War II, took him from Central Europe first to Norway and then via Sweden to the United States, where he arrived in 1941. To his Scandinavian sojourn belongs his study of aphasia dedicated to Alf Sommerfelt, in which study he put forward the thesis that the order in which children acquire control over the distinctive features of their language is fixed and is the mirror image of the order in which under aphasia these features are lost from control. 45 This view has been discussed and challenged, but never ignored, in subsequent aphasiological work. 46 At the time of his arrival in America, research was especially active in the field of "visible speech," the visual representation of the significant sound wave patterns of spoken language. 47 Jakobson's collaboration with practitioners in the field of acoustic phonetics led to the very wide acceptance of the concept of a set of twelve binary oppositions of distinctive features, a selection of them underlying the phonemic system of every

397

DISTINCTIVE FEATURE THEORY

language and constituting a class of phonetic universals. 48 This set, together with the now standardized binary ± representation of every relevant feature, made possible the display of a phonemic system in a matrix diagram. This represents the direct line from Prague in American linguistics. The relatively few and sporadic displays of "componential analysis" in phonology and grammar, bearing only a tenuous, if any, link with the Trubetzkoyan distinctive feature theory, fall into a separate and much less influential tradition (exemplified by Hockett 49 and Harris, who rejects neutralization in favor of different distributions 50 ). The phonetics of these "componential analyses" was predominantly articulatory; just how revolutionary Jakobson's priority for acoustic data was at first is to be seen, for example, in Garvin's review of Preliminaries to Speech Analysis: "To most American linguists phonetic means articulatory." 51 There are some differences of arrangement and definitions in the presentation of the distinctive features between the two books first published in 1952 and 1956 by Jakobson and his colleagues, but what they have in common is much more important in this period in the history of linguistics. The features are identified and defined in acoustic terms, and they are shared between consonants and vowels, so that the basic vocalic triangle or quadrilateral in a language is maintained by the same contrasts as a consonantal triangle and quadrilateral: 52 a

a

ae i, p

k

k 1, p

1

Of course, every acoustic difference is produced by some articulatory difference, and the articulatory ("productive" or "genetic") characteristics were given by the authors, but the priority of acoustic characteristics is shown by the recognition in this phonological model that phonologically identical features may be produced by obviously different articulatory movements. Thus the feature of flattening (lowering or weakening the upper formants) was treated as a single feature (contrasting with plain or non-flat), whether this was produced by lip rounding (front contracting) at one end of the vocal tract or by velarization or pharyngealization ("emphasis" in traditional Arabic phonetic terminology) at the other end. Jakobson justified this identification by reference to the facts that lip rounding and pharyngealization or velarization do not occur together in one language, and that pharyngealized consonants in Arabic were taken over in loan words by speakers of Uzbek and Bantu languages as labialized consonants. 53 It is fair to say that not all linguists working in the Arabic field have been happy with this identification. 54

398

R. H. ROBINS

Jakobson's acoustic feature analysis had achieved general recognition during the structuralist period of American (and world) linguistics, alongside and as part of the classical autonomous phoneme theory, with the 1952 and 1956 books as standard texts. In the first years after the publication of Syntactic Structures,55 phonological theory, like semantic theory, was somewhat ignored by Chomsky and others working in the then new field of transformational grammar. In 1957 Chomsky accepted the existing phoneme units as units, though he redefined the place of phonology within the total descriptive system.56 In 1959 Halle set the scene, as far as phonology was concerned, for the next ten years in asserting explicitly the place of phonology, no longer autonomous, as a set of rules whose input was the surface syntactic structure and whose output was the phonetic representation of the sentence,57 and in recognizing the Jakobsonian features as the model both for the phonological representation of the output of the morpheme structure rules and for the final phonetic specifications, output of the P rules.58 This decision in favor of features, with the associated matrices, for phonological description, in preference to the earlier use of unitary phonemes (as in Syntactic Structures), was disputed by Householder in an exchange of articles,59 but the position was essentially stable until 1968. The publication in 1968 of The Sound Pattern of English, the result of

further collaboration between Chomsky and Halle, preserved the place of phonology in linguistic descriptions and the conception of a universal set of features cutting across the traditional division of consonants and vowels. But in the extensive treatment of the phonological representation (Chapter 7) Chomsky and Halle reverted from the Jakobsonian acoustically defined features to an articulatory set.60 This is not Trubetzkoy's set, though naturally it bears several points of correspondence with it. Unlike Trubetzkoy, but following Jakobson, Chomsky and Halle make their features wholly binary. As an example, their opposition of coronal and non-coronal correlates with the distinction between the Trubetzkoyan, and the traditional, categories of dental, alveolar, palato-alveolar, and retroflex on the one hand and the rest of the positional articulatory features on the other; 61 it also corresponds to Jakobson's distinction between non-grave and grave, except that palatals, which are non-grave, are also non-coronal.62 Once again the choice of a definitional basis is justified on its making possible an economical and revealing account of the data. According to Chomsky and Halle63 the contrast between palatalization and non-palatalization (between "soft" and "hard" consonants in the Slavic languages), which closely belongs both with the distinction between front and back

DISTINCTIVE FEATURE THEORY

399

vowels and that between palatal and velar consonants themselves, is arbitrarily separated from them in Jakobson's treatment by being assigned to ± sharp, while the associated distinctions are treated as ± grave. 64 The articulatory contrast of ± back brings them all within the scope of a single feature. From a general linguistic point of view one might opine that, since articulation is primarily speech from the speaker's position and acoustics is primarily speech from the hearer's position (this is not entirely clearcut in view of the speaker's constant acoustic feedback), it is reasonable to assume that sets of features established by either criterion will be descriptively relevant and that both types of features will prove factors in historical changes of sounds. The phonological chapters in The Sound Pattern of English inaugurated a fresh approach to various phonological questions that had first been raised by Trubetzkoy and the early Prague School. The notion of markedness has already been mentioned in connection with the phonological descriptions of individual languages. Trubetzkoy, for example, designated voiced and voiceless pairs of consonants in languages like English as marked and unmarked, respectively. 65 But this did not necessarily correspond to unambiguous phonetic fact: English voiced (marked) plosives can be reinterpreted as lenis (unmarked) as against the voiceless fortes (marked), or (in whispering) as - aspiration (unmarked) as against the distinctively aspirated counterparts of voiceless plosives in normal speech. The markedness of voicing in English consonants is part of the phonology of English, and it takes in factors such as the wider distribution of unmarked members in initial and final clusters (/st-/, /sk-/, /sp-/, /-sp/, /-sk/, but not /zd-/, /zg-/, /zb-/, /-zb/, /-zg/, etc.). Chomsky and Halle have suggested a reinterpretation of markedness, by relating it not, as before, to the distinctive apparatus of each separate language, but to the collective, universal apparatus of phonological distinctiveness from which every language makes its selection, and by including a set of "marking conventions." 66 The conventions, for example, whereby + voice is unmarked for vowel segments and + rounded is unmarked for high back vowels 67 are a contribution to the economy of description, not of a specific language as such, but of any language in which these categories are found; thereby the intuitive notion of "naturalness" is formalized. 68 This latest interpretation of markedness rests on the general acceptance of a set of binary features serving as the universal phonetic base which constitutes the limits within which the phonological component of any language must be organized. The concept of markedness is still a phonological one,

400

R. H. ROBINS

but it has been universalized. Work in phonology has been undertaken subsequently on these lines by others. 69 Ill

The distinctive feature has proved to be one of the most important theoretical concepts of our science during the past half-century of its history. This can be attributed to a number of factors: firstly the acknowledged status of phonology as the pace-maker in general theory during much of the interwar period, and secondly the exploitation of structural similarities associating the level of phonology with the level of grammar (morphology and syntax) and, though to a much lesser extent, with some areas of semantics. But as significant as these is the personal factor, the work of Roman Jakobson, who has involved himself in the developments and applications of distinctive feature theory throughout his professional life, to which those of us who have learned so much from him now pay our respects on the occasion of his eightieth birthday. University of London NOTES 1

C. F. Hockett, "Sound Change," Language 41 (1965), 192-96. N. S. Trubetzkoy, Grundzuge der Phonologie=Travaux du Cercle Linguistique de Prague 1 (1939), 59. 3 I. Bekker, Anecdota graeca, Vol. II (Berlin, 1816), 774-75. 4 W. S. Allen, Phonetics in Ancient India (London, 1953), pp. 50, 89. 5 H. Sweet, Handbook ofPhonetics (Oxford, 1877), pp. 100-08 [these pages are reprinted in E. J. A. Henderson, The Indispensable Foundation-. A Selection from the Writings of Henry Sweet (London, 1971), pp. 228-33]; R. Jakobson, "Henry Sweet's Paths toward Phonemics" in In Memory of J. R. Firth, ed. C. E. Bazell etal. (London, 1966), pp. 242-54. 6 Sweet, Handbook of Phonetics, p. 106. 7 Daniel Jones, Outline of English Phonetics (6th ed., London, 1947), Chap. 6; The Phoneme: Its Nature and Use (Cambridge, 1950), Chap. 2-3. 8 Ann Arbor, 1947. 8 Zellig S. Harris, "Simultaneous Components in Phonology," Language 20 (1944), 203. 10 Bekker, Anecdota graeca, II, 631, 810-11. 11 M. B. Emeneau, "The Nasal Phonemes of Sanskrit," Language 22 (1946), 86-93; Allen, Phonetics in Ancient India, pp. 7-8. 12 R. Jakobson, "Linguistics in Relation to Other Sciences" (1967), Selected Writings II (The Hague, 1971), 678-79. 13 Zellig S. Harris, "Morpheme Alternants in Linguistic Analysis," Language 18 (1942), 169. 14 R. E. Longacre, review of W. M. Urban, Language and Reality and B. L. Whorf, Four Articles on Metalinguistics, Language 32 (1956), 301. 2

DISTINCTIVE FEATURE THEORY

401

15 Kenneth E. Pike, Language in Relation to a Unified Theory of the Structure of Human Behavior (The Hague, 1967) [preliminary edition: Glendale, Calif., 1954-60], Chap. 2. 16 Trubetzkoy, Travaux du Cercle Linguistique de Prague 7 (1939), 5-17. 17 Travaux du Cercle Linguistique de Prague 2 (1929). See N. S. Trubetzkoy, "Die phonologischen Systeme," Travaux du Cercle Linguistique de Prague 4 (1931), 96-97, for further references. 18 N. S. Trubetzkoy, "Die Aufhebung der phonologischen Gegensätze," Travaux du Cercle Linguistique de Prague 6 (1936), 29-45. 19 Trubetzkoy, Grundzüge der Phonologie, 241-61 ; R. H. Robins, "Aspects of Prosodie Analysis," Proceedings of the University of Durham Philosophical Society, Series B (Arts), 1 (1957), 1-12 [reprinted in R. H. Robins, Diversions of Bloomsbury: Selected Writings on Linguistics (Amsterdam, 1970), pp. 207-26] ; Daniel Jones, "Some Thoughts on the Phoneme," Transactions of the Philological Society (1944), 119-35; idem. The Phoneme : Its Nature and Use (Cambridge, 1950), § 34, 463-69, 688-89. 20 Trubetzkoy, "Die phonologischen Systeme"; Grundzüge der Phonologie, pp. 59-80. 21 M. Halle, "In Defense of the Number Two" in Studies Presented to Joshua Whatmough on His Sixtieth Birthday, ed. E. Pulgram (The Hague, 1957), pp. 65-72; R. Jakobson, C. G. M. Fant, and M. Halle, Preliminaries to Speech Analysis (Cambridge, Mass., 1952; 7th printing, 1967), pp. 8-10. 22 Cf. M. A. K. Halliday, "Categories of the Theory of Grammar," Word 17 (1961), 247. 23 R. Jakobson, "Zur Struktur des russischen Verbums" (1932), Selected Writings II (The Hague, 1971), 3-15; "Beitrag zur allgemeinen Kasuslehre" (1936), Selected Writings II, 23-71. 24 C. E. Bazell, "Fundamental Syntactic Relations," Casopis pro moderni filologii 33 (1949), 9-15. 25 R. H. Robins, "Neutralization in Grammar and Lexicon," Travaux de L'Institut de Linguistique 2 (1957), 107-13 [reprinted in Robins, Diversions of Bloomsbury : Selected Writings on Linguistics (Amsterdam, 1970), pp. 37-45]. 26 L. Hjelmslev, La catégorie des cas (Aarhus, 1935), pp. 11-12. 27 Cambridge, Mass., 1965, Chap. 2, § 2. 28 Aspects of the Theory of Syntax, p. 88. 29 Ibid., pp. 80-83. 30 R. Saunders and P. W. Davis, "Bella Coola Referential Suffixes," International Journal of American Linguistics 41 (1975), 355-68. 31 E. H. Bendix, "Compenential Analysis of General Vocabulary," International Journal of American Linguistics 32, 2 (1966), Part 2, 2-4, and elsewhere. 32 Ibid., p. 119. 33 V. Brandal, Théorie des prépositions, trans. P. Naert (Copenhagen, 1950) [originally published in Danish in 1940), p. 39. 34 E. A. Hammel (ed.), Formal Semantic Studies [=American Anthropologist 67, 5 (1965), Part 2], p. 3. 35 J. Vachek, The Linguistic School of Prague (Bloomington, Ind., 1966), pp. 81-82. 36 D. C. Bennett, Spatial and Temporal Uses of English Prepositions (London, 1975), pp. 4-10, 54. 37 J. Wilkins, Essay toward a Real Character and a Philosophical Language (London,

1668).

38

J. J. Katz and P. M. Postal, An Integrated Theory of Linguistic Descriptions (Cambridge, Mass., 1964), p. 14. 39 Hammel (ed.), Formal Semantic Studies, p. 53. 40 Bendix, "Componential Analysis of General Vocabulary," p. 1. 41 Folia Linguistica 7 (1975), 245-52.

402

R. H. ROBINS

42 Remarques sur ¡'évolution phonologique du russe = Travaux du Cercle Linguistique de Prague 2, p. 18 ; cf. "Phonemic Notes on Standard Slovak" (1931), Selected Writings I (The Hague, 1962; 2nd edition, 1971), p. 224. 43 R. Jakobson, "Observations sur le classement phonologique des consonnes," Proceedings of the Third International Congress of Phonetic Sciences (Ghent, 1939), pp. 36-37 [reprinted in Selected Writings I, pp. 272-79]. 44 C. Stumpf, Die Sprachlaute (Berlin, 1926). 45 R. Jakobson, Kindersprache, Aphasie urtd allgemeine Lautgesetze (Uppsala, 1941) [reprinted in Selected Writings I, pp. 328-401], 46 L. F. Sies, "An Introduction to the Study of Aphasia," in Aphasia Theory and Therapy, ed. H. Schuell (Baltimore, 1974), pp. 1-68. 47 Cf. R. K. Potter, G. A. Kopp, and H. C. Green, Visible Speech (New York, 1947); M. Joos, Acoustic Phonetics (Baltimore, 1948). 48 Jakobson, Fant, Halle, Preliminaries to Speech Analysis (1967), p. 40; R. Jakobson and M. Halle, Fundamentals of Language (The Hague, 1956), pp. 28-29. 49 C. F. Hockett, "Componential Analysis of Sierra Popoluca," International Journal of American Linguistics 13 (1947), 258-67. 50 Z. S. Harris, "Simultaneous Concepts in Phonology," p. 187; see also idem, "Componential Analysis of a Hebrew Paradigm," Language 24 (1948), 87-91. 51 P. L. Garvin, Language 29 (1953), 477. 52 Preliminaries to Speech Analysis (1967), pp. 33-34; Fundamentals of Language, pp. 39-40. 53 Preliminaries to Speech Analysis (1967), p. 31; Halle, "In Defense of the Number Two," pp. 67-68; Jakobson, "Mufaxxama: The 'Emphatic' Phonemes in Arabic," in Studies Presented to Joshua Whatmough on His Sixtieth Birthday (The Hague, 1957) and in R. Jakobson, Selected Writings I (The Hague, 1971), pp. 510-22. 54 Y. A. E. El-Haleese, "A Phonetic and Phonological Study of the Verbal Piece in a Palestinian Dialect of Arabic," Ph.D. dissertation (London, 1971), Chap. 7. 55 Noam Chomsky (The Hague, 1957). 56 Syntactic Structures, p. 46; cf. also Chomsky, Current Issues in Linguistic Theory (The Hague, 1964), pp. 65-75. 57 M. Halle, The Sound Pattern of Russian (The Hague, 1959), p. 26. 58 Ibid., p. 62. 69 M. Halle, "Phonology in a Generative Grammar," Word 18 (1962), 54-72; F. W. Householder, "On Some Recent Claims in Phonology," Journal of Linguistics 1 (1965), 13-34; N. Chomsky and M. Halle, "Some Controversial Questions in Phonological Theory," Journal of Linguistics 1 (1965), 97-138; F. W. Householder, "Phonological Theory: A Brief Comment," Journal of Linguistics 2 (1966), 99-100. 80 N. Chomsky and M. Halle, The Sound Pattern of English (New York, 1968), pp. 299-300. 81 Ibid., p. 304. 62 Ibid., p. 306. 63 The Sound Pattern of English, p. 308. 64 Preliminaries to Speech Analysis (1952), pp. 29-32; Fundamentals of Language, pp. 31-32. 85 Grundzùge der Phonologie, pp. 59-80. 68 The Sound Pattern of English, Chap. 9. 67 Ibid., p. 405. 88 Ibid., p. 401. 89 E.g., C. E. Cairns, "Markedness, Neutralization, and Universal Redundancy Rules," Language 45 (1969), 863-85; S. A. Schane, Generative Phonology (Englewood Cliffs, N.J., 1973), Chap. 11.

CHILD LANGUAGE STUDIES VELTA RUKE-DRAVIl^A

"Child language and aphasia can and must be incorporated into comparative (comparative in the widest sense of the word) linguistics."1 Roman Jakobson wrote these words in 1941, nearly 25 years ago, and they hold true even today. As early as five years after that, Werner F. Leopold could state that Jakobson "has done more than others to secure the position of child linguistics as a worthwhile linguistic discipline."2 Although Leopold felt in 1956 that the contributions by linguists to the field of child language study were "not so numerous as might be wished,"3 his prophecy that "the influence of Jakobson's pioneer study is yet to come" 4 had already been fulfilled, and child linguistics today holds a prominent position within the field of general linguistics. Many contributors to the literature on various aspects of language development in children of various nationalities have appeared since then, and the number is increasing. Today, given the current stage of research in the field, we might go a step further and demand that universities establish professorial chairs in child language. As a matter of fact, Marcel Cohen has already expressed such a desire, but this plan seemed too extreme for Jakobson when he began his study on child language development. There are three complexes of general problems in Jakobson's child language studies: (1) the processes of building up individual linguistic competence, that is, defining the general rules and establishing the order in which these rules are acquired by a child; (2) the degeneration of language in aphasia and similar disorders, including the egocentric talk of children (which is an intermediate link between overt and inner speech) in presleep situations; and (3) the connection between language input and language acquisition, especially with regard to "nursery language," for example, the use of parental terms and their acquisition by children. Jakobson's work in this field of linguistics is, quantitatively, not very extensive. His publications total about 100 pages, and the most signifi-

404

VELTA RU£E-DRAVINA

cant of the articles on child language cover the period from 1939 to 1962. In connection with the Fifth International Congress of Linguists in Brussels in 1939, Jakobson presented his general theses for the first time. They were then clarified, enlarged, and supplemented in his classical monograph Kindersprache, Aphasie und allgemeine Lautgesetze.5 This valuable study was republished in Selected Writings, I, 6 and somewhat later7 it crossed the language barrier into English-speaking countries, where it has exerted a fruitful influence on the work of psychologists. In 1960 Jakobson published an article on the theme of parental terms and other kinds of modification which take place in the speech of adults when talking to young children ("Why 'Mama' and 'Papa'?"), 8 and in 1962 he continued this discussion and made suggestions for further investigations concerning deviations from normal speech rules in dreams and in presleep situations (in the introduction to Ruth H. Weir's Language in the Crib).9 Then in 1967, at the Tenth International Congress of Linguists in Bucharest, Jakobson called for an attentive inquiry into the biological prerequisites of human language and the correlation between innateness and acculturation in the process of learning the first language. His point of view on the role of biologically innate factors and what must be learned was formulated there as follows: "Children's gift to acquire any tongue whatever as their first language and, more generally, the human aptitude to command new linguistic patterns, must arise primarily from the instructions coded in the germ cell, but this genetic assumption does not authorize us to conclude that for the little apprentice the language of adults is nothing more than 'raw' material." 10 The great value of Jakobson's research lies in the pioneering character of his work and in its stimulating effect on further investigations in this field. What makes his publications so important and successful is not his use of hitherto unknown material, but rather his interpretation of the data and his proposals and hypotheses about some general rules on the gradual acquisition of phonemes in child speech. Two aspects are important in this connection: (1) interest was concentrated on the development of the PHONEMIC system and not on other aspects, as, for example, syntax or semantics in child speech; (2) the concrete material used as a basis for the theoretical conclusions came in the first place from SLAVIC languages. These two points may have some connection with the fact that other investigators after Jakobson also focused their interest on the phonemic inventory of the young child even in cases when VOCABULARY was their primary research subject (e.g. K. Ohnesorg, and J. Pacesova) and with the fact that the child language researchers who most closely followed his suggestions were in

CHILD LANGUAGE STUDIES

405

some way connected with the Prague School. It is also interesting to observe that the results of the research based on Slavic (e.g. Czechoslovakian) material mostly confirmed Jakobson's theses. In contrast to many other American linguists, Jakobson in his child language studies never bases his conclusions of "universal character" on English alone but makes a fruitful use of his enormously wide theoretical and practical knowledge of many languages. This is one of the reasons his suggestions have been received with great sympathy outside the English-speaking countries. The majority of the further investigations made by other authors after 1942 and up to the present day show that the rules formulated by Jakobson in 1942 on the whole still hold and that only a few corrections are necessary with respect to some presumptions. In order to gain a better insight into the points which seem to be indisputable, on the one hand, and the contributions which are still questionable, on the other, we will state some of the most important of Jakobson's presumptions and claims, and then we will take a look at those which illustrate the present situation and future perspectives. In such a way the evolution and expansion of child language studies will become clearer. In his works Jakobson has : (1) noted his appreciation for the importance of child language studies as a field of linguistics; (2) called for further exact linguistic records from many languages in order to examine the validity of his "working hypotheses"; (3) focused interest not only on normal and gifted children but also on speech processes which deviate from normal ones, as in dreams, in presleep situations, and aphasia, when misbuildings and destructions of the articulation and morphematic and/or syntactic models appear; (4) recommended the study of child language as a source for research on the creative activity involved in finding and building new words (thus as an important source of studies in the field of psycholinguistics); (5) awakened interest in the biological background of the speaker, the question of the critical age as well as the role of innate elements in the rapidity and perfection of the speech development process; (6) tried to find some universals, i.e. some common features characteristic of all children independent of their native language, concerning gradual developmental degrees and the use of parental terms; (7) called for cooperation between linguists, anthropologists, and experts in psychology.

406

VELTA RU£E-DRAVI1^A

During recent decades nearly all of the above suggestions made by Jakobson have been followed. The field of child language research has now branched out still more and includes the process of linguistic acquisition in normal children as well as in deaf and mentally or linguistically retarded children. Speculations on the biological foundations of language and the critical age for human language acquisition have intensified and have been enriched by new important empirical material, e.g. by the recent reports on the acquisition of language by Genie. 11 The initial interest in parental terms ("Mama" and "Papa") has now grown into a special field of sociolinguistics. It has now been about ten years since A. Avram, discussing the investigations of nursery language in Romanian, wrote that "le domaine connexe des recherches sur la langue qu'on parle aux enfants (angl. baby talk) est dans une situation moins favorable." 12 But the situation has changed radically nowadays, since rich field-work materials from diverse languages and cultural groups have been gathered and many observations on nursery or baby-talk utterances and their eventual influence on the child's speech development during the preschool period have been collected and partially published. Several international conferences on language input and output have dealt with the modifications of speech in young children of several cultures and languages. A new branch which was not present at the time of Jakobson's child language studies has developed, namely investigations of the general LANGUAGE SOCIALIZATION process in preschool children in different social groups; the research work in this field includes monologues and spontaneous and directed dialogues among children as well as between children and adults. As a new branch of applied linguistics, studies of reading and writing in early childhood (between the ages of two and six) have awakened great interest among both child language researchers and practical pedagogues. Many recent publications on aphasia underline the close correlation between this phenomenon and the language-learning process, since language disorders occurring in aphasia have numerous parallels in child language. It should also be mentioned that the creative activity of the child in finding (place and personal) names is of interest in the field of onomastics as well. 13 Most new investigations supply further support of Jakobson's theory of speech development. In order to give some illustrations which can also be of interest to broader circles outside the field of child language, we will mention here first some statements which still seem to be valid and then some samples of presumptions where corrections are necessary as a consequence of new data. According to Jakobson:

CHILD LANGUAGE STUDIES

407

(1) The language-learning process is an interaction between the influence of the surrounding input forms and the child's own creativity. Nevertheless, the loans are not a mere copy. " D a s Kind schafft, indem es entlehnt." 1 4 (2) In the babbling period the child can easily produce a number of consonants, vowels, and diphthongs (even those absent from the surrounding language), but he is unable to articulate them as phonemes in words in the later stages when the phonemes are learned as distinguishable and identifiable units. This is also true of echo repetitions. (3) Reduplication of syllables appears as a favorite device in nursery forms and in other early word units of infant language. (4) The realization of a simple vowel represents a more natural stage as compared to the realization of the vowel chain (a diphthong or a hiatus). (5) The child's speech at its early stages uses no consonantal clusters but only combinations of consonants with vowels; consonantal clusters consisting of three consonants appear as a rule later than those with only two consonants. (6) The affricates (c, dz, c, dz) are late in phonological contrasts (c:s, dz:z, etc.), even if occasional occurrences of the affricates can be noticed in early stages. During some early periods the affricates are replaced by /—>s—>c. (7) The vibrant r (when articulated as an apical sound) is a late phoneme; this phenomenon can have some correlation to the fact that this distinctive feature is relatively rare in the languages of the world. (8) The acoustic picture is accessible to the child even before he is able to coordinate it on the basis of his articulatory capabilities. Recent observations of Czech, Swedish, and Latvian-speaking children show that the child can perceive and distinguish the phoneme already at the stage when he cannot yet realize them phonetically. Some of Jakobson's theories are in full agreement with the results of the investigations in some language groups. Thus, according to J. Pacesova and her Czech material, Jakobson's interpretation by means of the "maximum contrast" theory (i.e. the principle that the consonant /p/ and the vowel /a/ prove to be the optimal pair) explains the early appearance of /p/ more satisfactorily than other theories, e.g. those denoting the relation between labial articulation and sucking activity, or those which point to the presence of strong visual support in the case of the labials, as the difference between /mI and /p/ is not visible and hardly has any importance in sucking activity. 15 In Czech children there was also an early stabilization of the phonetic realization of /u/ and its placing among those phonemes that constitute the

408

VELTA RU^E-DRAVINA

early vocalic contrasts, all in agreement with Jakobson's theory of "minimal vocalism." On the other hand, there are some statements and hypotheses in Jakobson's pioneering works which need correction and further observations. Some examples: (1) The statement that " . . . the sequence consonant + vowel proves to be the optimal sequence and therefore it is the only universal variety of the syllable pattern" 1 6 does not always find confirmation in the new material on early child utterances. The model VC appears often in the early vocabulary: as pointed out by K. Ohnesorg, for example, the dropping of the initial consonant (which results in the high frequency of the VC model) is common in the speech of Czech, French, Russian, Polish, and Latvianspeaking children. 17 (2) In my opinion, some corrections must be made in the "universalities" of the "nursery forms." In his article "Why 'Mama' and ' P a p a ' ? " Jakobson assumed that in such nursery words nearly no consonantal clusters appear, that stops and nasals predominate, that labials and dentals prevail over velars and palatals, and that wide vowels, especially /a/, are obviously preponderant. This may be true with respect to the "parental terms," but it cannot hold for the whole group of nursery words: in some cultures the number of such forms is considerable and the phonemes, their combinations, and their frequency often depend also on the phonemic structure of the respective native language. 18 Even if further research work should change still more assumptions in detail, it will not change the essence of Jakobson's theories, and the stimulating influence which has come not only through his printed works but by his personal encouragement as well will enrich this branch of linguistics with new investigations. For future child-language researchers Roman Jakobson's contribution in this field of linguistics remains as a model of sound balance between empirical data and linguistic theories. University of Stockholm

NOTES 1

R. Jakobson, Child Language, Aphasia and Phonological Universals (The Hague: Mouton, 1968), p. 92. 2 W. F. Leopold, "Roman Jakobson and the Study of Child Language," For Roman Jakobson (The Hague: Mouton, 1956), p.288. 3 Ibid., p. 285. 4 Ibid., p. 287.

CHILD LANGUAGE STUDIES

409

5 Uppsala, 1941; reprinted from Sprâkvetenskapliga Sällskapets i Uppsala Förhandlingar, 1940-42 [= Uppsala Universitets Arsskrift, 1942, No. 9], 6 The Hague: Mouton, 1962; 2nd edition, 1971, pp. 328-401. 7 Child Language, Aphasia and Phonological Universals (1968). 8 Perspectives in Psychological Theory, Essays in Honor of Heinz Werner (New York, 1960), pp. 124-34; also in Selected Writings, I (2nd edition, 1971), pp. 538-45. 9 "Anthony's Contribution to Linguistic Theory," in Ruth H. Weir, Language in the Crib (The Hague: Mouton, 1962), pp. 18-20. 10 "Linguistics in Its Relation to Other Sciences," Actes du Xe Congrès International des Linguistes 28 Août - 2 Septembre 1967,1 (Bucharest, 1969), p. 94. 11 S. Curtiss, V. Fromkin, S. Krashen, D. Rigler, and M. Rigler, "The Linguistic Development of Genie," Language 50, No. 3 (1974), 528-54. 12 A. Avram, "De la langue qu'on parle aux infants roumains," To Honor Roman Jakobson, I (Mouton: The Hague, 1967), p. 133. 13 L. Huldén, "Die Ortsnamen der Kinder," XHth International Congress of Onomastic Sciences. Summaries of the Short Papers (Bern, 1975). 14 R. Jakobson, "Kindersprache, Aphasie, und allgemeine Lautgesetze," Selected Writings, I (2nd edition, 1971), 329. 15 J. Pacesovâ, The Development of Vocabulary in the Child (Brno, 1968), p. 34. 16 R. Jakobson, "Why 'Mama' and 'Papa' ?" Selected Writings, I (2nd edition, 1971), p. 541. 17 K. Ohnesorg, "Une contribution à la pédophonétique comparée," Colloquium Paedolinguisticum: Proceedings of the First International Symposium of Paedolinguistics, held in Brno, Fall 1970, ed. Karel Ohnesorg (Mouton: The Hague, 1972), pp. 188-89. 18 V. RQlfe-Draviija, "Gibt es Universalien in der Ammensprache?" Salzburger Beiträge zur Linguistik, II (Salzburg, 1976), 3-16.

ROMAN JAKOBSON'S TEACHING IN AMERICA THOMAS A. SEBEOK

"Socrates is an evildoer, and a curious person, who searches into things under the earth and in heaven, and he makes the worse appear the better cause; and he teaches the aforesaid doctrines to others."... Socrates . . . corrupts the youth . . . does not believe in the gods of the State, but has other new divinities of his own. — Plato's Apology (tr. Jowett).

In a foreword to a recent collection of some of my essays in verbal art, I briefly recounted my first meeting with Jakobson, on the sultry 27th day of August, 1942, when we spent a long afternoon absorbed in animated conversation in the garden of Franz Boas's house in Grantwood. I recorded that, among other topics, "he spoke to me at generous length about the highly ingenious accomplishments of the Russian Formalist school and its productive reformulations by the Prague Circle, stressing the close ties of both with structural linguistics." I noted that our friendship dates from that occasion, "as does my abiding absorption with the study of the verbal arts." 1 The editors of this volume have honored me with their invitation to comment on Jakobson's "teaching" during his American period, that is, following his arrival and settlement in this country after his dramatic Hegira from Czechoslovakia, and then, successively, from Denmark, Norway, and Sweden. Since teaching is a far more intimate expression of scholarship than any formal publication can possibly be, this narrative must necessarily be laced with autobiographical observations; and since Jakobson's instructional activities are so vast in scope, I can generalize with confidence only from such glimpses of it that good fortune has bestowed upon me. These date, in the main, from the War years and immediately thereafter, while Jakobson was Professor of General Linguistics and of Czechoslovak Studies at the École Libre des Hautes Études (19421946), then Visiting Professor of General Linguistics (1943-1949) and Professor of Slavic Studies (1946-1949) at Columbia University. I never attended his courses at Harvard University, where he served, from 1949

412

THOMAS A. SEBEOK

until his retirement in 1967, as Professor of Slavic Languages and Literatures and General Linguistics, nor at M.I.T., where he became Institute Professor in 1957. However, throughout this entire period, I heard him lecture on various other campuses on occasions too numerous to recall in detail. We have attended a good many conferences together, both in the United States and abroad. In particular, he has frequently been our guest at Indiana University: I recollect four of his sojourns with especial pleasure: his first visit to Bloomington during the War; his concluding report at the fortnight's Conference of Anthropologists and Linguists, in July, 1952;2 his enthusiastic participation in the Conference on Style, in April, 1958, in the course of which he delivered what may well be his most often cited paper, "Linguistics and Poetics" ; 3 and a sequence of lectures delivered here during the 1964 Linguistic Institute. When Jakobson and I first met, I was a graduate student in transition on several levels: literally, from the University of Chicago (and utter penury) to Princeton (and the promise of relative affluence afforded by a splendid fellowship). At the same time, I was also intellectually at sea, thrashing about somewhere in the middle of the common Atlantic pool that Firth so eloquently delineated a few years later. 4 My early linguistic attitudes had essentially been molded by two men, neither of whom was at Chicago any longer: Manuel Andrade, who died prematurely, and Leonard Bloomfield, who reluctantly accepted a call to Yale in 1940. After their departure deprived me of their comfort and steady linguistic counsel, I read voraciously according to my own appetites, and thus came to discovery of the Linguistic School of Prague and a glimmer of understanding of the distinctively Russian flavor the late Trubetzkoy and the very lively Jakobson had imparted to it, transforming its classical doctrines as these had sprung from native soil.5 When Andrade suddenly died, he had already "gone a long way toward developing a semiotically grounded linguistics, much farther than any studies yet made in this field," and his ambitious program "involved the building of the whole of linguistics upon semiotical foundations; he believed that in this way linguistics would obtain a metalanguage appropriate to the description and comparison of all languages." 6 I was strongly influenced by Andrade's highly original views and their applications, but far from sufficiently equipped at the time to carry his project further, especially since his remaining notes and manuscript fragments were scarcely utilizable. In any event, no one else was much interested: he was ahead of his time in our prevailing linguistic milieu, and this pupil of Boas is remembered today, if at all, only for his technical work on Quileute and several Middle American Indian languages. As for Bloomfield, who be-

TEACHING IN AMERICA

413

came my next advisor, he had just published a masterful essay proclaiming that "Linguistics is the chief contributor to semiotic. Among the special branches of science, it intervenes between biology, on the one hand, and ethnology, sociology, and psychology, on the other: it stands between physical and cultural anthropology." 7 In his classes and private sessions, however, Bloomfield refrained from discussing broad issues; his concern, no doubt rightly, was with imparting the formal skills required of any practicing linguist. It was at his pounding insistence that I was set on the path of specialization in Finno-Ugric languages and linguistics, despite the very nearly total absence of instruction in them within the Western Hemisphere, a circumstance which forced me into an autodidactic stance, yet one which ultimately led to the institutionalization of this field in America. Bloomfield's procedures, essential for training though they may have been, left me unsatisfied and restless, for they gave a disjointed, choppy, incomplete picture of linguistics. I could, therefore, understand, in some measure, why, some twenty years later, a new generation of linguists deemed it fit to set up a straw man in his name, selecting him as an emblem for a brand on inadequate behaviorism; Jakobson was, of course, perfectly correct in maintaining that "sur bien des points Bloomfield reste supérieur au mouvement qui se réclame de l u i . . ." 8 Such, then, in capsule form, was the initial state of my affairs when, at the age of twenty-one, I journeyed across the Hudson, from New York City, to seek out the cosmopolitan linguist, already acclaimed throughout Europe, but not yet acclaimed in America. 1 was sufficiently well acquainted with some of his writings to have aroused my curiosity to enlarge my knowledge in a face-to-face meeting. The afternoon turned into an intense tutorial, extending late into that summer evening, in the course of which two subjects were discussed in the manner of a Greek symposium, as it were, liberally interspersed, that is, with drink and food : Jakobson's apperception of phonological theory, and his notions about poetic language. Although the phonemic principle was well known in America, and accepted even then by all modern schools of linguistics, Jakobson convinced me then and there that any further development of linguistic sound analysis must proceed by dissolving the phoneme into distinctive features, and that binary opposition can consistently be applied as a patterning device for the entire phonemic material. And while Bloomfield had instilled in me his conviction that "the artistic use of language by specially gifted individuals" enjoyed general favor as a substitute for the observation of language, 9 Jakobson opened my eyes and ears to the true, exciting potential of a poetics when practiced by a master of linguistics. This was, indeed, a pro-

414

THOMAS A. SEBEOK

paedeutic experience leading toward the kind of holistic vista of the language sciences that I had been vainly groping for at Chicago. I therefore resolved to stay in close touch to learn more. The first formal opportunity for doing so developed within the framework of the École Libre des Hautes Études, assembled under the auspices and on the premises of the New School for Social Research. Jakobson described it as "a university founded by French scholars who were refugees from the Nazi occupation," where we "were teachers and students of one another," 1 0 and where "dès le début les différences entre étudiants et professeurs se trouvaient abolies par le fait que les professeurs eux-mêmes allaient écouter les conférences de leurs collègues." 11 Thus Claude LéviStrauss came to introduce Jakobson (and the rest of us) to structural anthropology, while Jakobson opened the door for Lévi-Strauss (as he did for many others) to linguistics. I remember that both had the courtesy to come to my raw course on the history of the Hungarian language - my very first teaching assignment - whereas I attended as many of their lectures as my commitments to Princeton would allow. I tried never to miss Jakobson's packed seminars, after which we usually went to a nearby bistro to continue animated conversations about the topic of the evening. It was at the École Libre that I heard Jakobson lecture for the first time, and I would like to refine here my impressions of his platform style, which seldom varies whether he speaks in French (and he did in those times, in New York) or in any other language. I had once written of "the manyvalued, unmistakably Jakobsonian, rhetorical stratagems that are sprinkled among the expository statements . . . there by tactical intent, at once to persuade and to seduce." 12 With conspicous exceptions - his very carefully worded summation of the results of the Ninth International Congress of Linguists, which he read verbatim on August 31, 1962, was o n e 1 3 - t h e lectures of Jakobson give off an air of uncontrived happenings (in the semiotic sense of Nôth 14 ). He appears to rely on miniature cue cards, consulting them mostly for melodramatic effect rather than content. 15 The feeling he conveys is that he creates, shapes, and edits his topic of the moment to express it in a rhythm best suited to his auditors' pulse; after Jakobson's Madrid lectures in 1974, a Spanish newspaper report characterized them as harmonious musical performances directed by the lecturer and played by the audience. The over-all effect is that his students - all of his audiences become his students - are moved unusually close to him, to the extent of even becoming protective. One cannot help recollecting that Jakobson once professed in the Moscow School for Drama; he had kept in touch with a former actor of the Moscow Art Theater, and later made

TEACHING IN AMERICA

415

effective use of him. 16 I am strongly reminded of a passage in which Stanislavsky comments on his own methods : "To achieve a h a r m o n y . . . one needs more than outer, physical tempo and rhythm; one needs inner, spiritual tempo and rhythm. One must feel them in the sound, in the speech, in the action, in the gesture, in the movement, in fact, in the entire production." 1 7 Sometimes a genuine improvisation replaced a "happening": when, shortly before Christmas of 1942, he received a cable from Copenhagen announcing the death of Viggo Brandal, he substituted for his lecture an unforgettable, yet impromptu eulogy of his friend which deeply affected all of us present, although no one in the audience had ever met this important but remote personage of the Cercle Linguistique de Copenhague. The ambience pervading and surrounding our group was international indeed: "Il y avait des gens qui passaient par là, qui venaient nous écouter ou qui venaient parler eux-mêmes. Toutes les langues possibles s'y mêlaient," Jakobson related in an interview. 18 Americans - both established scholars from Columbia and neighboring institutions, and much younger ones, partially drawn from the Language Section of the War Department (then located at a New York address, 165 Broadway, which became the eponym for a heroic era in American linguistics) - were gradually attracted into Jakobson's orbit. (Somewhat disconcertingly, he was followed around, as well, by an indeterminate cloud made up of East European and Russian groupies, to whom he was unfailingly gracious and kind, although they did erode his time.) Many of our crowd moved on with him to Columbia, where, in 1943, he offered an evening seminar on the topic of case systems, stemming from a trail-blazing monograph he wrote in 1935.19 Each student in this seminar was assigned to analyze exhaustively the case system of a language of his choice, present his findings orally, then revise the presentation in the light of the ensuing discussion. I selected Finnish. The resulting paper became a chapter of my eventual dissertation, dealing with the form and function of several Finno-Ugric case systems. Although my Ph.D. degree was awarded by Princeton (in 1945), Jakobson served, to all intents and purposes - and with the enthusiastic concurrence of my chairman, Harold H. Bender - as my thesis supervisor; it was thus, and in this sense, that I chanced to become his "first American student." Incidentally, at Jakobson's behest, this same monograph came to forge an initial link between John Lotz and me and, in due course, the lives of the two Hungarian linguists continued variously to commingle with his, as mentioned elsewhere. 20 Jakobson's strictures of the work-in-progress of his students could be very telling: about one of my early papers, he gently hinted that he thought

416

THOMAS A. SEBEOK

I had written it with my left hand; of another, he remarked that it seemed to him especially interesting for what I had left out. On the other hand, when a finished piece of work gained his coveted approval, he would stand behind it with his full authority. While he generally tends to be reasonably equable and tolerant of criticisms leveled directly at himself - responding usually, in due course, to a coherent set of them, without inviduous identification (the "Postscriptum" to Questions de Poétique21 being a good example) - he would not countenance indirect attacks disguised as censure of his students. His loyalty to them - and, by his own count, " a b o u t a hundred of [his] former students are professors in this country" 2 2 - remains abiding and fierce; in cases such as Soffietti's Columbia thesis on Turinese phonology, his defense can result in deplorably acrimonious clashes with colleagues like Hall, who, of course, perceived the polemic in quite different terms. 2 3 When, as sometimes happens, two of his former students collide, he will not, however, hesitate to take sides strictly on the merits of the case at issue, 24 painful as that may be. Little wonder, then, that generations of his disciples, down to the youngest, many of whom declared that they went to Cambridge for the opportunity of working under him, are steadfast in their allegiance, proclaiming "his ability to illuminate a question from various points of view." 25 It would be seriously misleading to pretend that Jakobson's teachings were an instantaneous and resounding success in America. Far from it: these were roundly condemned by an influential cabal of autochthonous and lately-naturalized linguists - mostly a generation or two older than mine - clustering around "165 Broadway." In a lecture delivered on December 27, 1974, as part of the Linguistic Society's Golden Anniversary Symposium on "The European Background of American Linguistics," he characterized his foes of this era by an abusively intended Aesopian epithet, administrators, which, however, was so veiled that it was widely misinterpreted. In truth, these men were mostly misguided chauvinists, afflicted with a hubris doubtless induced by the pressures and fears of an uncertain military conflict in the backdrop. Regrettably, the behavior of this small but powerful clique - which caused Jakobson and his friends untold anguish, to say nothing of economic loss - left a sinister stain on the otherwise magnificent tapestry of achievements of American linguistics of the 1940's. Fortunately, this dark episode was transpierced by brilliant shafts of light emanating from giants like Boas and Bloomfield; their instant appreciation for Jakobson's decisive presence must be allowed to compensate for all the rest, which had better stay buried along with other, similarly motivated, wartime debris.

TEACHING IN AMERICA

417

The affidavit against the teachings of Jakobson was much the same as the charges preferred by Meletus, summed up in the epigraph at the outset of this record. "This inquisition," Jakobson might have continued in the words of Socrates, "has led to my having many enemies of the worst and most dangerous kind, and has given occasion to many calumnies There is another thing: - young men . . . come about me of their own accord; they like to hear the pretenders examined, and they often imitate me, and proceed to examine others; there are plenty of persons, as they quickly discover, who think that they know something, but they really know little or nothing; and then those who are examined by them instead of being angry with themselves are angry with m e . . . . " In the fall of 1943,1 took up permanent residence at Indiana University, where Velten, Voegelin, and I soon invited Jakobson to come for some lectures. Voegelin asked him to send some feasible topics, and was startled to receive a list of nearly one hundred titles. Jakobson arrived in Bloomington by bus, greeting me with the question, "Where are the Indians?" He spoke on the cultural and social history of Slavic languages, 26 several of which were then taught here intensively to Army personnel, and was then also asked to give an ad hoc talk in J. R. Kantor's seminar. Kantor was an extreme behavioristic psycholinguist, who relished controverting with linguists. 27 For some reason, Jakobson chose as this seminar theme "The theory of signs," which, as far as I know, was his first presentation of semiotics in this country. He had hardly finished when Kantor bounded forward, shouting, "Why, that was nothing but medieval philosophy!" " N o t at all," I remember Jakobson retorting, "it goes back at least to Plato!" So Jakobson continued to flourish, as he related to Philip Rahv, "now and then in hostile, and often in amicable contexts." 2 8 For the summer of 1946, the late Stith Thompson organized the first Folklore Institute, assembling at Indiana University a highly interesting mélange of scholars of various ages. Among the welcome participants were Roman and Svatava Pirkovà Jakobson, who, as I recall, were driven out to Bloomington by Alan Lomax, who came to join his father, John A. During approximately the same weeks, the much more venerable Linguistic Institute was in session at Ann Arbor, under the inspired direction of Charles C. Fries. I happened to be among the members of the visiting faculty of the University of Michigan that summer, and so was a well-known American specialist in Slavic languages, who, in 1944, had been the unfortunate recipient of a particularly savage review by Jakobson of his Introduction to Russian. Although the facts were beyond dispute, the tone of this piece generated

418

THOMAS A. SEBEOK

much resentment. It was the custom then to conduct during the Institute weekly luncheon conferences, led chiefly by distinguished visitors. When Fries canvassed the faculty for nominations, I proposed Jakobson, but was first hooted down. However, I kept nagging away, and, with Voegelin's sympathetic support, finally prevailed upon Fries. I then fetched Jakobson from Indiana by car, and, on July 24th, he addressed the assembled faculty and student body on "Comparative metrics as a problem of modern linguistics." His once and always antagonist 2 9 sat not far in the audience. Jakobson kept disarmingly referring - and, seemingly, deferring - to him as "my great and good friend," this unexpected warmth causing the victim (at least momentarily) to melt, and me irrepressibly to giggle. At any rate, and in spite of the novelty of the topic in linguistic circles of those times, his debut at the Institute was, by all accounts, a succès fou. Fries thereafter turned into one of his ardent admirers, and Jakobson has, since then, become one of the most sought-after guests at Linguistic Institutes. His stellar role in the 1952 session, along with a sensational performance by Lévi-Strauss, 30 was an especially memorable tour de force. Among a particularly distinguished group, leavened by such famous scientists from abroad as the late Bar-Hillel, Hjelmslev, and Sommerfelt, we listened to his pivotal presentation that introduced fundamental semiotic concepts to an essentially native audience, most of whom heard the name of their turnof-the-century compatriot, Charles Sanders Peirce, for the first time; many were, I think, incredulous to have the Russian colleague characterize this American as "one of the greatest pioneers of structural linguistic analysis," as well as " a genuine and bold forerunner of structural linguistics," 31 but, as usual, time was to bear out amply his farsighted assessment. 32 The hallmark that stamps all but the most solemn of his public utterances is Jakobson's wit. He sprinkles his lectures with humorous asides, often calculated to point up the discrepancies between reality, with its shortcomings, and a state considered desirable by the speaker in temporary collusion with his listeners. Anecdotes about him could fill a modest-sized monograph; although I have always suspected that he secretly engendered most of them himself, such stories tend to take on a life of their own, becoming collective property much in the manner of his and Bogatyrev's "Die Folklore als eine besondere Form des Schaffens." 33 Possibly the best known has to do with Jakobson's arresting accent. While his mastery of the grammatical and lexical resources of spoken English is elegant, and of its rhetorical effects superb, his pronunciation has remained shockingly alien, giving rise to a remark most often ascribed to Kurylowicz, 34 but, in fact, circulating in numerous variants: "Jakobson can lecture perfectly in six

TEACHING IN AMERICA

419

languages - unfortunately, all of them Russian." His proficiency in handling discussion is histrionic and, partly as a consequence, a lot of fun to watch. I was once chairing a lecture where he spoke for a scheduled hour or so to a large assembly of students. When the time came for questions, his mostly young auditors were shy, and too overawed to speak up. After a few moments of awkward silence, Jakobson turned to me, holding his hand high: could he address a query to himself, he wondered ? I nodded, he put his question, then went on to answer himself, thus expanding his lecture for another rapt hour. Some years ago, in the early 1960's the Director of the Newberry Library convened a meeting of a dozen or so linguists in Chicago, culminating in a convivial banquet, where we were called upon to relate "Jakobson stories" in turn, most of which I have now forgotten. What does linger in my memory is the spontaneous outpouring of affection with which the many hilarious incidents - true or alleged - were suffused. Not a trace of malice disfigured that glow. I remember remarking on that pleasant atmosphere to friends who walked me back to my hotel: I felt that, after nearly twenty years of searching, Jakobson, who was born on October 11th and hence jocosely fancies his affinity with Columbus, had found the symbolic Indians he vainly looked for upon first alighting in Indiana, and they had finally made him their honored chief. After this mutual discovery, the benefice of his teaching continued to radiate serenely out of Cambridge, prompting even his callowest followers to proclaim that he had, indeed, "played a key role in the development of linguistics in America," and acknowledge that "he has a lasting, and often decisive, influence on our scholarly development." 3 5 The youth of America thus turned out to be luckier than the youth of Athens, whose elders succeeded in killing the man of whom Crito said "that of all men of his time whom [he has] known, he was the wisest and justest and best" (Plato's Phaedo, tr. Jowett). Indiana University NOTES 1

Thomas A. Sebeok, Structure and Texture (The Hague: Mouton, 1974), pp. vii ff. R. Jakobson, "Results of a Joint Conference of Anthropologists and Linguists" (1952), Selected Writings, II (The Hague: Mouton, 1971), 554-67. 3 Published in Style and Language, ed. T. A. Sebeok (New York: Wiley, 1960), pp. 350-77. 4 J. R. Firth, "Atlantic Linguistics," Archivum Linguisticum, I (1949), 95-116. 5 Joseph Vachek, The Linguistic School of Prague: An Introduction to Its Theory and Practice (Bloomington: Indiana University Press, 1966). 6 C. Morris, Signs, Language and Behavior (New York: Prentice-Hall, 1946), p. 223. 7 Leonard Bloomfield, "Linguistic Aspects of Science," International Encyclopedia of Unified Science 1/4 (Chicago: University of Chicago Press, 1939), p. 55. 2

420 8

THOMAS A. SEBEOK

J. P. Faye, J. Paris, and J. Roubaud, "Entretien de Roman Jakobson," Hypothèses: Trois entretiens et trois études sur la linguistique et la poétique (Paris: Seghers/Laffont, 1972), p. 47. 9 Bloomfield, "Linguistic Aspects of Science," pp. 5 f. 10 V. Mehta, John Is Easy to Please: Encounters with the Written and the Spoken Word (New York: Farrar, Straus & Giroux, 1971), p. 232. 11 Faye, Paris, and Roubaud, "Entretien de Roman Jakobson," p. 34. 12 T. A. Sebeok, "Review of Roman Jakobson, Selected Writings, I: Phonological Studies," Language, 41 (1965), 86. 13 R. Jakobson, "Results of the Ninth International Congress of Linguists," Selected Writings, II, 593-602. 14 W. Nöth, Strukturen des Happenings (Hildesheim: Olms, 1972), pp. 130 ff. 15 Once I had to miss a class where he was to have dealt with glossematics, and, the next time, I asked him to lend me his notes. He handed me a small stack of cards. The top card read: "The dog." The rest were equally uninformative. Plainly, the secrets of glossematics were not concealed in that cache. Years afterward, the meaning of the legend on that card dawned on me - but that is another story, involving - to mention it only briefly - the various contextual meanings of the vocable "dog" in their relation to its general meaning. 16 R. Jakobson, "Linguistics and Poetics," Style in Language, p. 354. 17 K. Stanislavsky, My Life in Art (Moscow: Foreign Languages Publishing House, 1959), p. 443. 18 Faye, Paris, and Roubaud, "Entretien de Roman Jakobson," p. 34. 19 "Beitrag zur allgemeinen Kasuslehre: Gesamtbedeutung der russischen Kasus,'' Selected Writings, II, 23-71. 20 T. A. Sebeok, "John Lötz: A Personal Memoir," Ural-Altaische Jahrbücher, 46 (1974), 3 ff., 5 ff. 21 Paris: Seuil, 1973. 22 Mehta, John Is Easy to Please, p. 232. 23 Cf. R. A. Hall, Jr., Stormy Petrel in Linguistics (Ithaca, N.Y.: Spoken Language Services, Inc., 1975), pp. 141 ff. 24 E.g., "Concluding Note," International Journal of Slavic Linguistics and Poetics, 14 (1971), 209. 25 Charles E. Gribble (ed.), Studies Presented to Professor Roman Jakobson by His Students (Cambridge, Mass.: Slavica, 1968), p. 7. 28 R. Jakobson, "The Beginnings of National Self-Determination in Europe," The Review of Politics, VII (1945), 29-42, and Readings in the Sociology of Language, ed. J. A. Fishman (The Hague: Mouton, 1968), pp. 585-96. 27 Cf. J. R. Kantor, An Objective Psychology of Grammar (Bloomington : Indiana University Publications, 1936). 28 "The Editor Interviews Roman Jakobson," Modern Occasions, Winter 1972, p. 18. 29 Jakobson subscribed to a journal edited by this man, and told me that one of his checks was cashed with the erudite endorsement, Pecunia non olet. 30 C. Lévi-Strauss, Structural Anthropology (New York : Basic Books, 1963), pp. 67-80. 31 Jakobson, "Results of a Joint Conference of Anthropologists and Linguists," Selected Writings, II, 555, 565. 32 T. A. Sebeok, "The Pertinence of Peirce to Linguistics, Presidential Address to the Linguistic Society of America, December 30, 1975," Language, 53 (1977). 33 R. Jakobson, Selected Writings, IV (The Hague: Mouton, 1966), 1-15. 34 Mehta, John Is Easy to Please, p. 229. 35 Gribble, Studies Presented to Professor Roman Jakobson by His Students, p. 7.

ROMAN JAKOBSON'S CONTRIBUTION TO THE STUDY OF SLAVIC HISTORICAL PHONOLOGY A N D PHONETICS SEBASTIAN SHAUMYAN

The appearance of Roman Jakobson's famous Remarques sur révolution phonologique du russe comparée à celle des autres langues slaves (1929) and "Prinzipien der historischen Phonologie" (1931) gave birth to the theory of diachronic phonology and, more specifically, of Slavic diachronic phonology. More than 40 years have passed since Roman Jakobson published these revolutionary works, but the basic ideas of his theory still dominate contemporary research in all branches of diachronic phonology. We can properly appreciate Roman Jakobson as a founder of diachronic phonology in the context of his pioneering contributions to the theory of phonology and, more generally, to the theory of structural linguistics. The cornerstone of the theory of diachronic phonology is the conception of the phoneme as a bundle of distinctive features. In his article on phoneme and phonology published in the Second Supplementary Volume to the Czech Encyclopedia (Ottùv slovnik naucny [Prague, 1932]) Jakobson wrote: "[The] phoneme is the basic concept of phonology. By this term we designate a set of those concurrent sound properties which are used in a given language to distinguish words of unlike meaning. In speech diverse sounds can implement one and the same phoneme." 1 The concept of the phoneme as a bundle of distinctive features was advanced by J. Baudouin de Courtenay, but it was Roman Jakobson who drew revolutionary consequences from this concept. "Only when brought up to the level of distinctive features," wrote Jakobson in his paper "On the Identification of Phonemic Entities," "does the linguistic analysis enable us to verify Saussure's cardinal statement on phonemic units as first and foremost 'entités oppositives'. The phoneme by itself is not a term of opposition." 2 Being aware of the revolutionary significance of his concept of the phoneme, Jakobson stated: "Linguistic analysis, with its concept of ultimate phonemic entities, signally converges with modern physics, which has revealed

422

S. K. SHAUMYAN

the granular structure of matter as composed of elementary particles." 3 It would be no exaggeration to say that the rediscovery of the phoneme as a bundle of distinctive features made by Jakobson became a turning point in the development of phonology and had a decisive influence on linguistic methodology. Starting from the concept of the phoneme as a bundle of distinctive features, Jakobson conceived of phonological evolution as a specific interplay of distinctive and redundant features of language sounds. Let us summarize the main ideas of Jakobson's theory of diachronic phonology. A sound change has taken place. What changes within the phonological system have resulted from this sound change? Have certain phonological distinctions disappeared, and if so, which ones? Have new phonological distinctions arisen, and what are they? Or has the inventory of phonological distinctions remained unchanged, while the structure of certain phonological distinctions has been transformed ? Any sound change should be considered only in relation to the phonological system within which it has taken place. A sound change may be devoid of phonological significance and may only increase the number of combinatory variants of a phoneme. If a sound change is phonologically significant, it is called a phonological mutation. The formula of the phonological mutation is: A : B > A x :Bi One should distinguish two categories of phonological mutations: either only one of the two relations (A: B or Ai: Bi) is phonological, or both are phonological: A: B as well as A i : Bi are different varieties of phonological relations. The first category divides into two types: the suppression of a phonological distinction may be called a dephonologization, and the formation of a phonological distinction may be called a phonologization. Dephonologization is defined thus: A and B are opposed to each other phonologically, while there is no phonological distinction between Ai and A2. The investigation of a dephonologization involves the following questions: what is the phonological nature of the phonological opposition A : B ? Is it a disjunction or a correlational pair? If it is a correlational pair, is its suppression a particular case of a more general process (that is, of the suppression of the whole correlation), or has the correlation remained ? What is the nature of the extraphonological relation A: B ? Is it a relation between variants, and if so, between which ones: combinatory ones? stylistic ones? Or it is a case of phonetic identity (two

SLAVIC HISTORICAL PHONOLOGY AND PHONETICS

423

identical realizations of one and the same phoneme) ? If the relation A : B is extraphonological, then A j is phonetically similar to Ai and Bi is phonetically similar to B, and only the conditions of the occurrence of A and B have changed. But if Ai is phonetically similar to Bi, then either A i ^ A and Bi # B, i.e. A and B have merged into a sound C which differs from both A and B, or A i # A but Bi = B, i.e. A > B . Phonologization is defined thus: there is no phonological distinction between A and B while there is one between Ai and Bi. The investigation of a phonologization involves the following questions: do Ai and Bi represent a disjunction or a correlational pair? If it is a correlational pair, does the mutation enrich an existing correlation, or is it part of a more general process, that is, of the formation of a new correlation ? The second category of mutations is called rephonologization. Rephonologization is the transformation of a phonological distinction into a heterogeneous phonological distinction which has a different place in the phonological system than the first phonological distinction. There is a phonological distinction between both A and B and Ai Bi, but the phonological structure of the two oppositions is different. There are three types of rephonologization: (1) the transformation of a correlational pair of phonemes into a disjunction, (2) the transformation of a disjunction into a pair of correlative phonemes, (3) the transformation of a pair of phonemes that belonged to one correlation into a pair of phonemes that belong to another correlation. It should be pointed out that there are sound changes that modify not the phoneme inventory, but the inventory of the phoneme groups of a language. These sound changes do not modify the phonological system of a language, but only the frequency of the occurrence of different phonemes. The mutations we have considered so far have the following feature in common: all terms of these mutations are equal as to their extension. If the phonemes A and B are equal in their extension, then the phonemes Ai and Bi are also equal in their extension; if A and B are phoneme combinations, then Ai and Bi are phoneme combinations with the same extension. But from the point of view of diachronic phonology we should attach equal importance to mutations in which the resultant term Ai is not similar in its extension to its prototype A. These mutations divide into the following types: 1. A phoneme splits into a group of phonemes. As a result of this mutation a distinction between two phonemes transforms into a distinction between a group of phonemes and a phoneme (rephonologization).

424

S. K. SHAUMYAN

2. A group of phonemes transforms into a phoneme. There are two possibilities: (a) the result of the transformation is a phoneme that already existed in the system, (b) the result of the transformation is a new phoneme that did not exist in the system. When we discover several mutations that took place simultaneously, we should analyze the set of these mutations as a whole. There are inherent relations between these mutations and we have to bring to light the laws that control these relations. One of these laws is: a phonologization is always accompanied and controlled by a dephonologization. The description of mutations does not exhaust diachronic phonology. We have to face the task of interpreting mutations as well. A description has a bearing on two linguistic situations: on one that precedes and the other that follows the mutations, and makes it possible to raise the problem of the direction and sense of the mutations. This problem carries our investigation from the domain of diachrony to that of synchrony. Mutations may be the object of synchronic research no less than linguistic elements that remain unchanged. It is a serious mistake to consider statics and synchrony as synonymous terms. A static slice is a fiction: it is only an auxiliary scientific construct rather than a mode of existence. We can consider the perception of a film not only from a diachronic, but also from a synchronic, point of view, and the synchronic aspect of a film is not identical with a still of the film. The perception of movement has its synchronic aspect as well. The same can be said about language. When we consider a linguistic mutation in the context of synchronic linguistics, we bring it into the domain of teleological problems. The problem of purposefulness is meaningful for a chain of successive mutations, that is, for diachronic linguistics. If a given mutation was preceded by a lack of equlibrium in a phonological system and if this mutation restored equilibrium in the system, we can pinpoint the function of this mutation: its purpose is to restore the equilibrium. Although a mutation may restore the equilibrium in one part of the system, it can, at the same time, upset the equilibrium in other parts of the system, providing thereby the impetus for a new mutation. In this way there arises a chain of stabilizing mutations. But it would be a mistake to think that the purpose of every phonological mutation is the restoration of an equilibrium. Basically, the phonological system of an intellectual language tends toward equilibrium, whereas the infringement of the equilibrium is one of the constitutive features of emotional and poetic language. That is why a static phonological description goes against reality when the object of this description is affective language.

SLAVIC HISTORICAL PHONOLOGY AND PHONETICS

425

Affective discourse acquires its expressive power by exploiting extraphonological phonic differences that exist in a given language since, in order to attain a higher degree of affectivity, it does not stop short even at deformation of the phonological structure. An increase of emphasis and the acceleration of tempo may lead to the violation of prosodie correlations, to the suppression of certain phonemes, or to the convergence of different phonemes into one phoneme. All this is due to the fact that in affective language information cedes its place to the expression of emotions with the ensuing attenuation of certain phonological distinctions. In the same way, the poetic function enables language to overcome the automatism and imperceptibility of the word leading to displacements in the phonological structure. The interplay of statics and dynamics is one of the fundamental antinomies that determine the essence of language. One cannot understand the dialectic of linguistic development without taking into account the existence of this antinomy. Attempts to identify synchrony, statics, and the sphere of teleology, on the one hand, and diachrony, dynamics, and the sphere of mechanical causality, on the other hand, unnecessarily confine the field of synchrony and transform historical linguistics into an agglomerate of disparate facts creating a superficial and harmful illusion of a gulf that separates synchrony and diachrony. These are the main ideas of Jakobson's theory of diachronic phonology. Roman Jakobson's Remarques sur l'évolution phonologique du russe comparée à celle des autres langues slaves was a breakthrough in the study of the history of Slavic languages : the application of the theory of diachronic phonology to Slavic linguistics marked the birth of Slavic diachronic phonology. In applying this theory to Slavic linguistics, Jakobson saw his central task in the exploration of the interplay of phonological correlations in the course of the evolution of the Slavic languages in terms of the general laws of the structure of phonological systems. These laws define the various dependences between phonological correlations. Given two correlations A and B, there are three types of dependences between them: 1. If A exists, B exists. An example of this is the relation between the correlation "one ~ another structure of the syllable intonation" and the correlation "length ~ shortness of vowels." If the first of these correlations exists in a given phonological system, the second must exist, too. But the reverse is not true. 2. If A exists, B is absent.

426

S. K. SHAUMYAN

An example of this is the relation between the correlation "dynamic accent ~ absence of accent" and the correlation "length~ shortness of vowels." 3. If A is absent, B is absent. An example of this is the relation between the correlations "one ~ another structure of the syllable intonation" and "musical accent ~ absence of accent." The main content of the Remarques is a penetrating analysis of the history of phonological correlations in the Slavic languages. Jakobson shows that the evolution of phonological correlations was in the Slavic languages teleologically defined by the tendency of the phonological system to avoid the co-existence of incompatible correlations. For instance, no Slavic language admitted the co-existence of the musical correlations ("musical accent ~ absence of accent," "one ~ another structure of intonation") with the correlation "softness ~ hardness of consonants." This tendency had a profound influence on the evolution of Slavic phonological correlations. Thus, the Slavic languages which had lost the correlation "softness ~ hardness of consonants" after the disappearance of the weak jers were able to preserve the correlations "musical accent ~ absence of accent" and "one ~ another structure of intonation" (Serbo-Croatian and Slovene). In the Slavic dialects where the correlation "softness ~ hardness of consonants" and musical correlations clashed, the musical correlations were excluded from the phonological system and replaced by the correlation "dynamic accent ~ absence of accent." The result of this was the emergence of a new clash: the correlation "dynamic accent ~absence of accent" conflicted with the correlation "length~ shortness of vowels." This clash was resolved in two ways: (1) an eastern solution (Russian and Bulgarian), where the correlation "length~ shortness of vowels" was eliminated, and (2) a western solution (Czech and Slovak; Old Polish, Lusatian) where the correlation "dynamic accent ~ absence of accent" was eliminated, and where the stress became bound to a definite syllable of the word and quantity was preserved. As regards the correlation "softness ~ hardness of consonants," it either prevailed over the correlation "musical accent ~ ~ absence of accent" (Russian, East Bulgarian, Polish, Lusatian), or it disappeared together with the musical accent. In order to exhibit the power of the method adopted in the Remarques, I shall consider in some detail Jakobson's hypothesis on the origin of the so-called "dissimilative akarie" characteristic of an important group of Southern Russian dialects. Jakobson advanced his hypothesis as an alternative to the hypothesis of

SLAVIC HISTORICAL PHONOLOGY AND PHONETICS

427

A. Saxmatov. According to Saxmatov's theory, dissimilative akarCe arose at the time when the distinction between short and long vowels still existed. The process that brought about dissimilative akarCe involved the lengthening of the unaccented reduced vowels e, o, a. But, as Jakobson showed, the assumption of the lengthening of the unaccented reduced vowels e, o, a clashed with the following fact: the elimination of the weak jers preceded the reduction of the unaccented e, o, a, but the dialects that have dissimilative akarie lost the weak jers only after the correlation "long vowels ~ short vowels" was eliminated. Jakobson's hypothesis regards the akarCe as a process resulting from the transformation of the correlation "musical accent ~ absence of accent" into the correlation "dynamic accent ~ absence of accent." The last one involves the process of the reduction of unaccented vowels which brings accented vowels into prominence. Every vowel has a definite degree of intensity. A comparative analysis of vowels, and of Russian vowels in particular, shows that the most intense vowel is a (degree 1), and the less intense vowels are o, e (degree 2), w, e (degree 3) and i, u (degree 4: weak vowels), in that order. Just as a singer who is inconvenienced by a divergence between the course of the melody and the distribution of the individual heights of the vowels is often forced to modify the text in order to avoid the coincidence of high tones with inherently low vowels, and vice versa, so in a language with the correlation "dynamic accent ~ absence of accent" the presence of strong vowels in weak position creates an internal contradiction which the language users try to overcome. The analysis of the dialects with dissimilative akaiie shows that unaccented syllables that do not precede the accented syllable show a tendency to eliminate strong vowels (1-3). While the weak vowels remain unchanged in these syllables, vowels 1-3 are converted into a reduced vowel that is more backward and more open after hard consonants and more forward and more closed after soft consonants. This vowel tends to converge with the unstressed i. Both these variants belong, as regards their intensity, to weak vowels. In the syllable which precedes the stress, the intensity relations are more complicated. In the dialects with dissimilative akan'e this syllable, as well, is weak so that strong vowels cannot retain their individuality. When the stressed syllable has a strong vowel, the strong vowel of the preceding syllable undergoes a reduction that is similar to the reduction of the strong vowels in other unstressed syllables. On the contrary, when the stressed syllable has a weak vowel, the preceding syllable retains the strong vowel, but the range of the intensity of this syllable is reduced to the vowel

428

S. K. SHAUMYAN

a because it is not strong enough to preserve the individuality of every strong vowel. Thus, the strong vowels of the pretonic syllable depend on the vocalism of the stressed syllable and are subject to the action of the balance of forces. The pretonic vocalism undergoes simplification: the intermediary types 3 and 2 disappear and type 1 is generalized. This theory makes it possible to regard the different types of dissimilative jakan'e as different realizations of the possibilities inherent in this process. Consider the Don type of dissimilative akarte. In this type the intensity categories 1,2, 3 became 4 in the pretonic syllable when the stressed syllable is of degree 1, 2, 3, and they became 1 when the stressed syllable is of degree 4. This type is most clear because the intensity categories 1-3 are treated on a par both in the stressed syllable and in the pretonic one. While in the Don type a new weak vowel appears in the pretonic syllable in all classes when there is no weak vowel in the stressed syllable, in the Zizdra type, on the contrary, the pretonic syllable generalizes the strongest vowel in all cases when the stressed syllable has no vowel of this category, 1.e. only the strongest vowel in the stressed syllable, rather than intermediary categories, brings on the reduction of the strong vowels in the preceding syllable: 1, 2, 3 > 4 in the pretonic syllable when the stressed one is 1 and 1, 2, 3 > 1 when the stressed one is 2, 3, 4. The Don type lumps the intermediary categories, in the stressed syllable, together with the strongest category, but the 2izdra type lumps them together with the weak category. As regards the Obojan' type, here either of the intermediary categories is treated differently: 2 in the same way as 1, and 3 in the same way as 4. In this type we have: 1, 2, 3 > 4 before the stressed 1, 2; and 1, 2, 3 > 1 before the stressed 3, 4. Dissimilative akarie is a special case of a more general process that is called akan'e in Russian dialectology. All dialects with akatie have in common the following feature: the vocalism of the unstressed syllables is reduced to three vowels: /, u, a. What distinguishes one dialect from the other are the conditions of the alternation of the a and i in the pretonic syllable. Akan'e, in its turn, should be regarded as a special case of a still more general process, the tendency to simplify the vocalism of unstressed syllables in comparison with the vocalism of the stressed syllables. Thus, the Russian dialects with the so-called "okatie" have reduced the vocalism of the unstressed syllables to four vowels in the majority of cases: i, u, a, o. Scrutinizing the process of simplification of the vocalism of unstressed syllables in the Russian dialects, Jakobson advanced a hypothesis concern-

SLAVIC HISTORICAL PHONOLOGY AND PHONETICS

429

ing the expansion of akan'e at the expense of okan'e based on purely internal linguistic criteria. According to Jakobson, the victory of akan'e over okan'e in the Russian literary language cannot be explained (as it was by Saxmatov) in terms of the numerical superiority of people speaking with akan'e, but should be interpreted with reference to the structural difference between the akan'e and okan'e dialects: the speakers of akan'e were unable to reproduce the vocalism of the okan'e, while the speakers of okan'e could easily assimilate the vocalism of akan'e through lexical borrowing without any disturbance of their own phonological system. The works of Jakobson in Slavic historical phonology overcame the antinomy between synchrony and diachrony that was introduced by F. de Saussure. As is well known, de Saussure taught that sound changes are blind destructive factors external to the linguistic system. Being a revolutionary in his synchronic approach to language, Saussure treated the history of language in the spirit of the Neogrammarians. Through his analyses of the evolution of the Slavic phonological systems, Jakobson showed that a description of sound changes makes sense only when diachronic processes are analyzed with relation to synchronic states. In overcoming the antinomy between synchrony and diachrony, Jakobson has at the same time overcome another antinomy, that between teleology (associated with synchrony) and mechanical causality (associated with diachrony). By showing that sound changes are subject to phonological laws, he was able to show that teleology dominates both synchrony and diachrony. The resolution of these antinomies led Jakobson to deeper interpretation of the interrelation between dialects. The crucial problem he posited was this: should a structural approach apply only to independent phonological innovations, or should it apply also to the propagation of such innovations ? "Since Saussure excluded the concept of system from diachrony," Jakobson wrote, "there inevitably arose a profound antinomy between the synchronic and diachronic treatment of borrowings. For Saussure, the synchronic state of language {la langue) is 'un système qui ne connaît que son ordre propre' ; he consequently concluded that 'l'emprunt ne compte plus comme tel, dès qu'il est étudié au sein du système ; il n'existe que par sa relation et son opposition' to other facts of the system 'au même titre que n'importe quel signe autochtone' (42-43). But when Saussure turned from synchrony to diachrony, he rigorously separated the 'emprunt de phonème' from 'changements phonétiques', insisting that one should carefully distinguish 'les foyers d'innovation, où un phénomène évolue unique-

430

S. K. SHAUMYAN

ment sur l'axe de temps, et les aires de contagion' where 'il ne s'agit pas de la modification d'un prototype traditionnel, mais de l'imitation d'un parler voisin, sans égard à ce prototype' (283-88). "Since we cannot accept Saussure's starting point, i.e. the postulate of the absence of any connection between diachrony and synchrony, we cannot accept the consequences of this postulate. "The relation of a diachronic law to a synchronic one (cf. Saussure 131) may be defined as a relation of a means to a result achieved. We speak about the result rather than about the goal not because we deny that a diachronic law is by its nature a tendency and has a teleological meaning, but because the realization often need not coincide with the initial aim; for as in other realms of human activities, especially collective activities, the goal is not always achieved. " N o innovation in the system of a language can be interpreted apart from the system in which the innovation occurred, no matter whether the innovation is an independent one or a reproduced one assimilated from outside. The reference to borrowing cannot serve as the complete explanation of a fact that takes place in the life of a linguistic system." 4 According to Jakobson contagion is possible without predispositions towards convergent evolution, and convergence is possible without contagion, but this does not mean that we should oppose contagion and convergence. In most cases the propagation of a phonological innovation cannot be traced to simple imitation. Jakobson wrote: "This unity of borrowing and convergence reminds us very much of mimetism in modern biology: 'The factors of resemblance belonged from the very beginning both to the imitator and to the model, and a certain impluse might be enough to make them manifest (Berg, 224). The conclusive theory of biologists which claims that mimetism is a particular instance of convergence to which it would not be reasonable to attribute a special origin or significance (o.c., 229) finds its counterpart in linguistics." 5 Jakobson gives quite a few examples from the evolution of the Russian dialects which support his thesis of the unity of contagion and convergence. This thesis may be regarded as a cornerstone of a new conception of structural dialectology. Among others, it throws light on the problem of the dislocation of the protolanguage. Referring to the unity of contagion and convergence, Jakobson wrote: "Under these conditions, the debates concerning the dislocation of a 'common language', e.g. of Proto-Slavic or Proto-Russian,

SLAVIC HISTORICAL PHONOLOGY AND PHONETICS

431

lose their acuteness. The notion of dislocation refers to the incapacity of certain dialects to undergo common changes. We can hardly ever decide what constitutes the basis of this concept of 'common' and whether we are dealing with pure convergence or with convergence supported by contagion. If convergence dominates over divergence, it is reasonable to use the concept 'common language'. "There is no need to explain that this conventional historical conception of language unity cannot be transposed without reservation into synchrony. The latter has no criteria that would enable us to answer the question of whether we are dealing with one or with several languages. The answer is determined by the attitude of the collective of language users. The critic of this attitude should take into account the notion of conformity to a goal rather than conformity to historical phonology. Only a parochial science can still solve the question of the legitimacy of such or such linguistic separatism in the light of purely genetic problems. On the social plane the notion of class has for a long time replaced the notion of caste; on the national plane self-determination forced the question of origin into the background; by the same token, the topical linguistic questions in which we are interested relate not to genesis, but to function." 6 Let us return to the phonological laws presented by Jakobson. It should be noted that these phonological laws are empirical generalizations rather than statements deduced from some fundamental principles obtained in an analytic way. These generalizations have a probabilistic character and should have the form "If A exists, probably B exists," "If A is absent, probably B is absent," rather than the cited forms "If A exists, B exists," etc. Jakobson is aware of the probabilistic character of his laws. He writes: "Typology discloses laws of implication which underlie the phonological ahd apparently the morphological structure of languages: the presence of A implies the presence (or on the contrary the absence) of B. In this way we detect in the languages of the world uniformities or near-uniformities, as the anthropologists used to say. " N o doubt a more exact and exhaustive description of the languages of the world will complete, correct, and perfect the code of general laws. But it would be unsound to postpone the search for these laws until a further broadening of our factual knowledge. The question of linguistic, particularly phonemic, universals must be broached. Even if in some remote, newly recorded language we should find a peculiarity challenging one of these laws, this would not invalidate the generalization drawn from the imposing number of languages previously studied. The uniformity observed

432

S. K. SHAUMYAN

becomes a 'near-uniformity', a rule of high statistical probability. Before the discovery of the duckbilled platypus in Tasmania and Southern Australia, zoologists in their general definitions of mammals did not foresee the egg-laying ones; nonetheless these obsolete definitions retain their validity for the overwhelming majority of the world's mammals and remain important statistical laws. "But even at present, the rich experience of the science of languages permits us to uncover constants which will hardly be degraded to nearconstants. There are languages lacking syllables with initial vowels and/or syllables with final consonants, but there are no languages devoid of syllables with initial consonants or of syllables with final vowels. There are languages devoid of fricatives but none deprived of stops. There are no languages with an opposition of stops proper and affricates (e.g. /i/-/tt/) but without fricatives (e.g. /s[). There are no languages with rounded front vowels but without rounded back vowels." 7 It should be emphasized that, in cybernetic terms, there is a feedback between phonological laws and linguistic observations. Linguistic observations lead us to the discovery of phonological laws. But once a phonological law is discovered it serves as a guide to new observations. If new observations disclose new facts that contradict the established phonological law, we should refine it in order to accommodate it to the new facts. Having reformulated it, we should use the reformulated version of the phonological law as a guide to new observations that might lead us to new difficulties that would make it liable for further refinement. The following remarks of Jakobson are instructive in this respect: "Furthermore, partial exceptions in the case of some near-universals call merely for a more supple formulation of the given law. Thus, in 1922, I noted that free dynamic stress and the independent opposition of long and short vowels are incompatible within the same phonemic pattern. This law, which satisfactorily explains the prosodic evolution in Slavic and some other Indo-European groups, is valid for an overwhelming majority of languages. The few cases of allegedly free stress and free quantity have proven to be illusory: thus, Wichita (in Oklahoma) was said to have both phonemic stress and quantity, but according to Paul Garvin's re-examination, Wichita is actually a pitch language with an opposition of rising and falling accent hitherto overlooked. Nonetheless, this general law requires a more cautious formulation. If in a language phonemic stress co-exists with phonemic quantity, one of the two elements is subordinate to the other, and three, almost never four, distinct entities are admitted: either long and short vowels are

SLAVIC HISTORICAL PHONOLOGY AND PHONETICS

433

distinguished only in the stressed syllable or only one of the two quantitative categories, length or brevity, may carry a free, distinctive stress. And apparently in such languages the marked category is not the long vowel opposed to the short, but the reduced vowel opposed to the non-reduced. I believe with Grammont that a rule requiring amendment is more useful than the absence of any rule." 8 The concept of phonological law advanced by Jakobson is one of the cornerstones of diachronic phonology, and this concept should guide any researcher who investigates the evolution of a phonological system. In one of his papers Jakobson wrote: "A proverb says that it is no good to discover America too late, after Columbus, but the premature discovery of America may turn out to be no less painful, when there are not yet the means for adopting and exploiting new territories." 9 Jakobson discovered diachronic phonology in time. That is why after his works we have seen a rapid development of the new discipline in both the West and East. But Jakobson not only discovered a new territory, he also provided the proper means for exploiting this territory, and one of these means is the concept of phonological law. These means have not as yet been used to a full extent. And this is why Jakobson's works will serve for a long time as a source of inspiration for future researchers. Yale

University

NOTES 1 Roman Jakobson, Selected Writings, I: Phonological Studies, second expanded edition (Mouton: The Hague-Paris, 1971), p. 231. 2 Ibid., p. 421. 3 Ibid., p. 425. 4 "Remarques...," Selected Writings, I, p. 106. 5 Ibid., p. 107. 6 Ibid., pp. 108-09. 7 "Typological Studies and Their Contribution to Historical Comparative Linguistics" (1957), ibid., p. 526. 8 Ibid., pp. 526-27. 9 "The Kazan' School of Polish Linguistics and Its Place in the International Development of Phonology" (1960), Selected Writings, II: Word and Language (Mouton: The Hague-Paris, 1971), p. 414.

R O M A N JAKOBSON'S WORK ON T H E HISTORY OF L I N G U I S T I C S EDWARD STANKIEWICZ

Roman Jakobson's work on the history of linguistics can be characterized, like his work in general linguistics, by one outstanding feature: its universality. Thus there is hardly an aspect or phase in the history of linguistics that he has not at one time or another submitted to a penetrating scrutiny, and to which he has not brought some fresh insights. References to the history of linguistics are scattered in many of Jakobson's writings, but the main bulk of his historiographical research is now conveniently gathered in Volume II of his Selected Writings, under the heading Toward a Nomothetic Science of Language.1 Jakobson's historical interests are far from antiquarian: he probes the past for its relevance to the present and traces the growth of those ideas which have made linguistics into a pivotal science of man, a science which now bears the decisive imprint of his thought. The section in question covers some of the most seminal moments in nineteenth-century linguistics and deals with such outstanding linguists of the twentieth century as N. S. Trubetzkoy, Franz Boas, and Antoine Meillet. Several articles were originally written as necrologies (N. Trubetzkoy, A. Meillet, and S. Karcevskij); their brevity, however, does not constrict their content: with a few strokes Jakobson makes us see the man and scholar in question and appreciate his place in the development of modern linguistics. The bulk of Jakobson's historiographical research is, however, reserved for those figures of the late nineteenth century who through their efforts and genius brought about that "grand renouvellement de la linguistique" spoken of by Meillet, a renovation that turned linguistics away from purely historical pursuits towards synchronic analysis and a general, "nomothetic" science of language. The initiators of the new movement emerged almost simultaneously in various parts of the world to challenge the monopoly of the German comparative-historical school, i.e. in Russia, England, America, France, and Switzerland. And it is of such men as J. N. Baudouin de Courtenay, M. Kruszewski, H. Sweet, W. D. Whitney, F. de Saussure,

436

EDWARD STANKIEWICZ

and J. Winteler that Roman Jakobson writes with unique knowledge, perspicacity, and empathy. Of their heroic and at times frustrating and tragic struggles he writes as follows: "The proverb says that it is no good to discover America too late, after Columbus, but the premature discovery of America may turn out to be no less painful, where there are not yet the means for adopting and exploiting the new territories. Such was the fate of the Vikings, who supposedly perished in such a brave but premature expedition. Although the nineteenth-century precursors of modern linguistics were not swallowed up by the waves of the North Atlantic, the history of their lives also bears the mark of tragedy." 2 To understand the extent of the change in methods and scientific outlook, it is enough to compare the programs of some of the leading Neogrammarians with those of their opponents and forerunners of modern structuralism. As late as 1891 Brugmann and Streitberg declared that "the true task of linguistics is to describe the total development of the Indo-European languages since their darkest origins and to discover the laws that determined their direction," whereas H. Paul saw around the same time an unbridgeable gulf between linguistics as a science of "real objects and facts" and the "sciences of law" or of general principles (Gesetzwissenschaften). The former, according to Paul, could be studied only by "isolating individual factors" {Die Wirksamkeit der einzelnen Faktoren isoliert zu behandeln), and by comparing facts of a different chronological order. 3 Paul's ideological distaste for "abstractions" and "inference" was the dominant attitude of the late nineteenth century, when the positivist fascination with the "bare fact" pervaded almost all humanistic, historical disciplines. In a letter to Fresenius, J. Burckhardt wrote, not without pride: "By nature I cling to the concrete, to visible nature and to h i s t o r y . . . . I am unsuited to speculative theorizing and never even for a minute in a whole year, feel I disposed towards abstract thought." 4 A new undercurrent of thought was, however, in the making and, according to the testimony of C. S. Peirce, it made itself felt first in the abstract sciences, such as mathematics, logic, physics, astronomy, and in music. "Their ideal," wrote Peirce, was "the universal and a b s t r a c t . . . in contradiction to the ideals formed upon outward experience." 5 The grossly empirical and individual thus came to be supplemented or replaced by the abstract and universal, the factual and inductive by the formal and deductive, and the contingent and variable by what is obligatory and invariant. "The comparison of languages is but a single phase in the history of our science" and is "only a means and not an end," wrote Baudouin de Courtenay in 1871. He later elaborated on this idea: "The lin-

THE HISTORY OF LINGUISTICS

437

guist must not be content with the registration of haphazard and sporadic facts" but must attempt to discover the "genuinely distinctive properties of language" and "the laws hidden in the depth, in the intricate combination of various elements" (1910).6 Similar ideas were voiced almost at one and the same time in the writings of all those linguists whom Jakobson has brought close to us through his writings. These ideas were given forceful expression, perhaps for the first time, by one of the great and unjustly forgotten pioneers of modern linguistics, M. Bréal, who in his opening lecture at the Collège de France (1868) reminded his listeners that, besides comparative grammar, there is another kind of study, "c'est cet assemblage de principes et d'observations ... qui est connu sous le nom de grammaire générale et philosophique," 7 and who consciously returned to the eighteenth-century tradition of French linguistics, when language was treated not as a biological or physical phenomenon, but as a system of signs. The history of linguistics, like the history of a language, affords a dual perspective, and hence alternative possibilities of emphasis and viewpoint: it can be treated retrospectively ("genetically"), with a view to the precedents and conditions which brought about a change in the intellectual climate, or it can be studied prospectively ("teleologically") with an emphasis on the goals and consequences of the new attainments. A many-sided interpretation of history requires, of course, a combination of both of these viewpoints. "The method of science," Peirce said, "is itself a scientific result. It did not spring out of the brain of a beginner; it was a historical attainment." 8 Jakobson's analyses can be said to be "teleological," as they tend to emphasize those aspects of the past that remain vital for the present and to underestimate the complexities of the background from which they have sprung. Thus they also tend to play down the features of continuity in periods of transition, and the tensions between the old and the new that persist in the theories of the innovators. Closer attention to those aspects might, for example, reveal to what extent the new "paradigm" is a response to an existing crisis, or highlight the internal contradictions which pervade the works of such outstanding figures as Baudouin, Kruszewski or Meillet. One may also wonder whether the quest for laws, or for "a nomothetic science of linguistics," was the foremost feature of the modern trend in linguistics, as Jakobson seems to suggest. He emphasizes this point (apart from the title of the historiographical section) by citing the youthful letter of Kruszewski to Baudouin de Courtenay (of 1871) in which the young Polish scholar inquires of his future mentor whether it would be possible to subsume all linguistic phenomena under one all-embracing law "without which a science cannot be a science." 9 This insatiable appetite for laws,

438

EDWARD STANKIEWICZ

that was treated with irony by Baudouin, was, in effect, the hallmark of all late nineteenth-century linguistics (beginning with the "blind" phonetic laws of the Neogrammarians and continuing through the laws of uninterrupted progress) and led to a veritable hypertrophy of laws that soon gave place to total skepticism and to a belief in mere statistical regularities (as, for example, in works of Delbrück or Paul). As Baudouin pointed out in his review of Kruszewski's work (1889), all science aims at the formulation of general rules and regularities,10 and the achievement of the men of the late nineteenth century lay rather in the fact that they broadened the framework of linguistics and posited a new set of scientific questions and laws. The longest and most important article in the historiographical section is devoted to the development of phonology in modern linguistics. Jakobson's interest in the problem is quite understandable, not only because he is the acknowledged master in this branch of linguistics, but because phonology, as he puts it, provided "the methodological model for all other areas of linguistic analysis." 11 In his customary manner, the author ranges far beyond the subject announced in the title of the article ("The Kazan' School of Polish Linguistics . . ."). 12 Jakobson gives us, in fact, an outline of the development of the concept of the phoneme from antiquity (the Sanskrit grammarians, Plato, and Aristotle) to modern times. At the center of the article are, of course, Baudouin de Courtenay and Kruszewski. Jakobson traces the development of Baudouin's thought on phonology and morphophonemics from his youthful paper on the alternation of the Polish sibilants with /x/ (in 1869) until the end of his long scholarly career (Baudouin died in 1929 at the age of 84). As Jakobson indicates, Baudouin de Courtenay was the first modern linguist to posit anew "the question of the relationship of sound and meaning" and the first to launch a functional approach to language. In an article published about ten years later (in 1969), Jakobson had the opportunity to show that a similar functional approach to the sounds of language germinated in the early writings of F. de Saussure. In an unpublished "treatise" on general phonetics Saussure wrote: "Toutefois tout ce qui est dans le langage est un fait de conscience, c'est à dire le rapport entre le son et l'idée. La valeur sémiologique du phonème, peut et doit s'étudier en dehors de toute préoccupation historique; l'étude sur le même plan d'un état de langue est parfaitement justifiée (et même nécessaire quoique néglegée et méconnue) quand il s'agit de faits sémiologiques."13 This two-pronged orientation towards a synchronic analysis and a functional interpretation of the speech-sounds was thus to become the basis for a new general linguistics. The parallelism between the ideas of the

THE HISTORY OF LINGUISTICS

439

Geneva linguist and Baudouin de Courtenay were soon recognized by some of Baudouin's students (Polivanov and Sôerba), but Jakobson is perhaps the first to point out the concrete resemblances and areas of mutual influence in the works of the two founders of structuralism. A central problem in Baudouin's elaboration of phonological theory and in the development of phonology up to this day is the interpretation of the relation between phonology and morphophonemics. While Baudouin and his co-worker Kruszewski were able quite early to separate the study of the physical and physiological aspects of sounds (a study Baudouin defined as "anthropophonetics" in his first university lecture in 1870) from the study of "the role of sounds in the mechanism of the language," they failed for a long time to see the distinction between phonology and morphophonemics, and were consequently vacillating and inconsistent in defining the phoneme. It is known that at the beginning the Kazan' linguists, and especially Kruszewski, identified phonology with the study of sounds in the morphological system of a language and hitched the problem of phonetic variants to that of the alternation of sounds in different morphological environments. Since the morphological alternants of a language need not bear any phonetic similarity (their link being given by the morphological units in which they occur), Baudouin spoke from the beginning of the relativity of sounds, or of "the discrepancy between the physical nature of sounds and their role in the mechanism of language and in the perception of language." 14 The preoccupation with the morphological role of sounds was not a sudden discovery of the two Polish linguists. As Kruszewski has made clear in his Vber die Lautabwechslung (1881), it was their response to a methodological crisis in late nineteenth-century linguistics concerning the relation between presumably "regular" sound-change and "irregular" analogical leveling. The question of sound-alternations has, furthermore, been on the agenda of comparative linguistics since the early nineteenth century (with its controversies about the origin and "virtues" of the Indo-European Ablaut) until its culmination in the Mémoire of F. de Saussure. Baudouin's familiarity with the early-nineteenth-century work of Mrozinski (see below) could likewise have influenced the morphological orientation of his early phonological theory, an orientation reinforced by Kruszewski's comparativist outlook (the alternants within one language were treated by Kruszewski on a par with the "comparents" or correspondences between historically related languages). As Jakobson points out, the subsequent crystallization of phonological theory involved the elaboration of two pivotal problems: the interpretation of the role of phonetic variants and their relation tô the phoneme as a

440

EDWARD STANKIEWICZ

distinctive unit, and the emancipation of the phoneme from historicalcomparative and purely morphological considerations. The clarification of these two problems gave direction to the entire development of Baudouin's phonological theory and constitutes his chief accomplishment. In the first place, he rejected the purely naturalistic interpretation of the "divergencies" that was proposed by Kruszewski, and before him by the Neogrammarians. Morphemes, he wrote in his review of Kruszewski's work, are not divisible into physical-physiological sounds, but are made up of distinctive semiotic units. 15 Some years earlier (in 1881, in a postscript to an article entitled "Suum cuique") he had already insisted on the separation of the two types of variants, i.e. "divergencies" and "correlatives," and on the need to distinguish them by a different terminology and notation. The firm grasp of this distinction enabled Baudouin to elaborate a functional theory of speech-sounds that would include both the concepts of distinctive and concomitant features (i.e. those features Baudouin labeled "phonetic cement") and a refined theory of morphophonemic alternations. Jakobson's account of Baudouin's post-Kazan' development lacks, however, nuance, and his evaluation of the relationship between Baudouin and Kruszewski is clearly tilted in favor of the latter. Thus he assumes that Kruszewski's death (in 1887) dealt a deep blow to Baudouitr who, he claims, "wearily abandoned the common achievements . . . of the whole Kazan' period," and that it was only thanks to the influence of his younger disciples Scerba and Polivanov that Baudouin was able (around 1910) to "extract the core" of his earlier sound insights from the accumulated "superfluous chaff." 16 A close look at Baudouin's publications after his departure from Kazan' shows, on the contrary, a progressive broadening and deepening of his thought. It is in such works as the Lectures on Latin Phonology (1893), the Versuch einer Theorie phonetischen Alternationen (1895), the Linguistic Remarks and Aphorisms (1903), and last, but not least, the "necrology" of Kruszewski (1888-89), in which Baudouin comes very close to the modern notion of the phoneme as an invariant ("the sum of generalized anthropophonemic properties"), to a theory of distinctive features, and to a many-sided analysis of morphophonemic alternations. Jakobson's conclusions that the "liberation of phonemic analysis from morphological 'command' has become possible only today," 17 or that "the problem of resolving the phoneme into its ultimate discrete components had been raised concretely in contemporary research" 18 can be accepted only with some qualifications. The idea that not the phonemes, but their "discrete auditory-articulatory components" ("acousmemes" and "kinamemes"), constitute the ultimate phonetic units of language was forcefully

THE HISTORY OF LINGUISTICS

441

stated in Baudouin's study of 1910 ("On Phonetic Laws") and repeated in many of his subsequent papers (inklings of this idea are scattered in several of his earlier writings as well). The fact remains that neither Scerba nor Polivanov saw the full import of Baudouin's approach, and it was left to Jakobson to put the theory of distinctive features on a solid scientific foundation. Jakobson ascribes Baudouin's decline in his Dorpat and Cracow periods to his psychologizing tendency, whereas this tendency was, in fact, neither as pervasive nor as pernicious as Jakobson suggests. Its most harmful consequence lies, perhaps, in the fact that it opened up a gulf between phonology and phonetics, and reintroduced that dualism for which Baudouin had himself taken Kruszewski to task. However, such a dualistic approach to phonology dominated linguistics for a long time and is still rampant in some quarters. Thus, for example, Sapir claimed that the "underlying," "abstract," "ideal sound-system" can be totally at variance with the "objective system of sounds peculiar to a language" and that the latter may change without affecting the former, 19 while Trubetzkoy saw a "profound difference between phonology and phonetics" in that phonemes, in contradistinction to sounds, are totally immaterial ("incorporels") just as the elements of meaning they serve to distinguish.20 In the light of these pronouncements Baudouin was, indeed, and in spite of his reputed "psychologism," far ahead of his time. Baudouin's contribution to historical phonology is mentioned by Jakobson in a perfunctory reference to the "facultative speech-sounds" (i.e. optional, stylistic variants) which Baudouin discussed in one of his last papers (1929), though this question was raised by him in a number of earlier papers as well. Jakobson discusses also, though very briefly, Baudouin's notion of the stability and variability of sounds both in the synchronic state and in the historical development of a language, but ignores Baudouin's most important study in the field of historical phonology, his Outline of the History of the Polish Language, written in 1922, which is still a model of precision and depth, and which was the first treatise on the phonological history of any language. "Linguistics," wrote Baudouin in one of his earliest lectures, "must be recognized as an autonomous discipline, not to be confused with physiology or with psychology," 21 and it was this program that guided the pioneers in modern linguistics, especially Saussure, who shared with the Polish linguists their basic methodological premises and concepts. At the end of his study Jakobson draws some striking parallelisms between the Kazan' and Geneva schools, and points out similar tendencies in the contemporary works of Winteler, Sweet, and Passy. The second article on the Kazan' school is devoted to Kruszewski,22 and

442

EDWARD STANKIEWICZ

represents not only a critical analysis of his works, but a tribute to a brilliant explorer. Kruszewski's contributions to modern linguistics and the teachings of the Kazan' school are still in need of a thorough, objective evaluation. In contradistinction to Baudouin, K r u s z e w s k i ' s work has received little critical attention and his writings are still to be collected and edited. Jakobson's study is an important step in this direction. The first analysis of Kruszewski's work was, as was mentioned, provided by Baudouin himself, though this analysis is surely not free from ambiguities that reflect the uneasy relation between teacher and student. On the one hand, Baudouin praised Kruszewski's philosophical "bent," his talent for precision and terminological inventiveness, while on the other hand he scoffed at Kruszewski's premature generalizations, his "greed" ("lapczywosc") for laws and his neglect of hard empirical evidence. However, the harshest judgment of Kruszewski concerned his originality: in the opinion of Baudouin there was none, for "Kruszewski repeated merely the things he has learned from others." During his short but extremely intense scholarly life, Kruszewski produced two important books: Über die Lautabwechslung (1881) and Ocerk nauki o jazyke [An Outline of the Science of Language] (1883). The first of these books incorporates Baudouin's teachings about morphophonemic alternations, while the second expands and develops his theory of morphological absorption, or what Kruszewski called "perintegration." There can be little doubt that Kruszewski did not merely repeat the ideas that were elaborated in the Kazan' workshop, but gave them a clearer, "more philosophical" interpretation and put them in a broader linguistic framework. In the second book especially he goes far beyond the questions of morphological absorption and in fact strives for a synthesis of the most central phonological and morphological problems of the day, that had farreaching consequences for the future. As if to compensate for Baudouin's severe verdict on Kruszewski, Jakobson seems to go to the other extreme: he sees in Baudouin's review a sign of intellectual apathy and decline, and hails Kruszewski as one of the boldest innovators in linguistic theory. Consequently, he tends to ignore those elements in the Outline which reflect some of the commonplace ideas and blind spots of nineteenth-century linguistics, such as belief in the incessant striving of language towards an ideal "harmony," in the inexorable conflict (couched in Darwinian terms) between the destructive forces of phonetic change and the constructive forces of analogy, the creation of meaning by analogy to the referent, and others. Like most nineteenth-century linguists, Kruszewski also stays strictly with-

THE HISTORY OF LINGUISTICS

443

in the confines of the word, and Baudouin was undoubtedly correct when he pointed out that the proper title of Kruszewski's book should have been An Outline of the Science of the Word, rather than of Language. Kruszewski, furthermore, oversimplified the concept of morphological "absorption." Thus he saw in it only a process of a cyclic, leftward expansion of the ending at the expense of the suffix, and of the suffix at the expense of the stem, but failed to notice the opposite type of "absorption" that changes a part of the prefix into an element of the stem (a phenomenon that Baudouin had observed in the Slavic languages in connection with such forms as Russian k nemu, s nim), or a full word into a prefix (a phenomenon that was discussed in European linguistics at least since Adam Smith's Dissertation of 1759). But otherwise the Outline is full of extremely penetrating observations and insights, the most interesting being Kruszewski's idea about the two types of "associations" that integrate the words of any language into a system. The importance of this idea was immediately recognized by Baudouin, and it proved extremely fruitful in subsequent theories of language, including Jakobson's studies in linguistics and poetics. According to this idea, every word is connected with other words by an association of contiguity, and by an association of similarity which yields "families" or "nests" of formally or semantically similar words. It is the latter (as we say now, paradigmatic) type of "association" which serves, according to Kruszewski, as a source of innovation and creativity and which eventually makes language into a "harmonious whole," while the first, syntagmatic type of relation acts as a conservative, retrograde force. In his treatment of contiguity Kruszewski had obviously considered only fixed phraseological expressions, failing to notice that the axis of succession, too, allows for creativity, i.e. for new collocations and shifts in the use of meanings and forms. But, like Saussure, Kruszewski did not touch upon syntactic problems, nor on the relation of the word to higher structures. The distinction between the syntagmatic and paradigmatic axes found its way into modern linguistics mainly through the work of Saussure, who spoke of the former as "la suite linéaire" and of the latter as "la solidarité associative" or "groupment par familles." Jakobson hypothesizes that Saussure, whose admiration for Kruszewski is well-known ("Baudouin and Kruszewski," he wrote in 1908, "came closer than anyone else to a theoretical view of language") took the distinction from Kruszewski. Although the possibility of such influence cannot be excluded, the recognition of the two axes of language has an older history and points up the general complexity regarding the origin of scientific ideas.

444

EDWARD STANKIEWICZ

The concept of the linear character of speech, which figures so prominently in Saussure's work, was vigorously discussed in seventeenth and eighteenth-century linguistics, culminating in Condillac's "méthode analytique," according to which language is to be viewed as an instrument which converts ("decomposes") ideas that are simultaneous in the mind into signs that must follow each other in a linear seqùence. It is thus likely that Saussure owes the idea about the linearity of speech not so much to Kruszewski as to an older tradition of French linguistics, to which he is also otherwise indebted for a number of concepts and terms. Kruszewski's treatment of the paradigmatic axes is, on the other hand, closely tied with the work of his contemporaries. The division of the lexical stock of a language into primary words and derived words that are created by "association" with certain "groups" or "families" of words is discussed by a number of linguists of that time (e.g. Bréal and Whitney). Kruszewski's immediate inspiration might have been, in fact, H. Paul's study (of the late seventies) which states that "speakers do not learn ready-made words" but create new words according to patterns, and that the existence of such patterns or "groups" of words makes possible the formation of words into new combinations. "Reproduction by memory and new formation by means of association" are for Paul, as they are for Kruszewski, the "two indispensable factors" in the development of language. 23 This convergence or diffusion of ideas among linguists of various persuasions should remind us that the new theories of language did not mark in each instance a break with the past, but were as much an attempt to adapt the tradition to a new mold, and that the history of ideas proceeds not only by dialectical jumps, but also by series of transitions and adoptions. In addition to the articles dealing with the founders of the Kazan' school, two more studies by Jakobson discuss Polish linguistics. One of them is devoted to Jôzef Mrozinski, a general in the Napoleonic army, an educator and scholar, and the other to the question of "Polish-Russian Cooperation in the Science of Language." 24 The work of General Mrozinski might have been forgotten in linguistics had not Baudouin de Courtenay written that he was "one of the most outstanding linguists of the first half of the nineteenth century." This judgment was subsequently endorsed by Baudouin's disciple K. Appel and by Baudouin's follower T. Benni, who in 1913 wrote a special study on Mrozinski the "psycho-phonetician." 25 A series of studies on Mrozinski was presented in 1957 at a meeting of the Poznan Society of the Friends of Science. 26 Jakobson's study is both a continuation of the high tribute that has been paid to Mrozinski and a correction of some misinterpretations of Mrozin-

THE HISTORY OF LINGUISTICS

445

ski's work that crept into Benni's study. Mrozinski, who authored two descriptive studies on the Polish language (in 1822 and 1824), was, according to Benni, interested in "anthropophonetics" and "psychophonetics" (both terms of Kazan' vintage), but was guilty of confusion of these two aspects. It is Jakobson's contention that the confusion of the two aspects is not so much due to the Polish general Mrozinski as to the Polish phonetician Benni. Mrozinski was probably the first Polish linguist (and according to some Polish scholars the first linguist) to classify the speech-sounds according to a two-fold criterion, i.e. according to their inherent phonetic properties and according to their function within the morphological system. The similarity of this conception to Baudouin's and the appeal (and possibly influence) which Mrozinski held for Baudouin are thus apparent. Jakobson gives a neat outline of Mrozinski's career and works, and points out that Mrozinski, who until he began his career as a grammarian had a better knowledge of French than of his mother tongue, set out to apply the high standards of French grammatical tradition to the study of Polish, which he felt was perverted in the hands of native grammarians. The focal points of that tradition are reflected, according to Jakobson, in the following excerpt from the program for the study of the Polish language formulated by the general-turned-linguist: " W e cannot know Polish grammar unless we learn the philosophical part of our language, or its internal structure . . . The rules of the language can be deduced only from the mechanism of that language [and] the exposition of the sounds and of the grammatical variants serves in my work only to explain and prove my theory of the general principles of the structure of language." 2 7 This mingling of principles pertaining to general grammar and to a specific language is, indeed, strongly reminiscent of the style of various French grammars of the eighteenth century that went under the generic title of grammaire générale, but to see in Mrozinski simply a pupil of French "grammaire générale" would be an injustice and an oversimplification, and would explain neither the true significance of Mrozinski in the history of Polish linguistics nor his appeal to Baudouin, whose critical attitude towards "universal" and "logical" French Grammar is well known. Since Jakobson does not dwell on this question (though he states that Mrozinski was able "to find and creatively use the highest achievements" of French philosophy of grammar), some explanatory remarks are in order to indicate Mrozinski's position in early nineteenth-century linguistics in Europe and in Poland. Although Mrozinski was intimately acquainted with the Western tradi-

446

EDWARD STANKIEWICZ

tion of general grammar, he probably came closest to those late eighteenthcentury theoreticians of language (such as Thiebault and Girard) who rejected the older tradition of grammaire générale, for which the study of language consisted of two qualitatively and hierarchically different components, i.e. of a "science" that dealt with universal principles, valid for all grammars, and an " a r t " that dealt with the external and formal properties of individual languages. The later grammarians and Mroziriski saw the relation between "science" and " a r t " in a new light: the former represented for them the system of a given language, and the latter its usage. The principles of grammar were thus to be drawn from the study of the "genius of a given language and from its practice" (Girard), or as Mrozinski would have it, "science" (umiejqtnosc) referred to descriptive grammar, whereas " a r t " (sztuka) was concerned with normative grammar. The works of these grammarians 28 thus mark an end to speculations about the "nature of thought" and its reflections in grammar, and the beginning of a healthy empirical tradition (continued by J. Madvig and W. von Humboldt) that explores the distinctive properties of diverse languages. When Mrozinski appeared on the Polish linguistic scene, Polish grammatical thought and teaching was totally dominated by Kopczynski's National Grammar, which incorporated all the rationalist and logicizing principles of the tradition of "general grammar." It is Mrozinski's sharp critique of Kopczynski's grammar, of its confusion of letters and sounds, its haphazard separation of meanings and forms, its blind imitation of "foreign models" that have " n o basis in the structure of Polish," that leads him to a new comprehension of the two-fold function of Polish sounds and to a new analysis of the Polish grammatical system. Like Baudouin a hundred years later, Mrozinski distinguishes two aspects of language, an internal ("philosophical") and an external one, but he insists that a correct interpretation of the former is predicated on a precise description of the latter, or of what he called "the mechanism of the language." A m o n g the basic discoveries of Mrozinski is his distinction of " h a r d " and soft consonants, and of the complementary distribution of the Polish vowels i and y (in pairs such as byljbil). A s Jakobson points out, the discrimination of distinctive sounds and positionally conditioned variants is still incompletely grasped by some linguists, including Benni, who tried to revindicate the phonemic status of y and who objected to Mrozinski's treatment of the sonorants as being outside the voiced/voiceless ("strong/weak") opposition on the grounds that these sounds may appear in some positions as voiceless (e.g. krwi, plwac). Mrozinski's other important contribution concerned the morphopho-

THE HISTORY OF LINGUISTICS

447

nemic relation of the Polish consonants. "In Polish," Mrozinski wrote, as if anticipating the formula of Baudouin, "we decline words not only by changing endings, as in all languages, but also by softening consonants." Unlike Kopczynski, who spoke in this case (like some modern transformationalists) of the phonetic change of sounds, Mrozinski introduced (again like Baudouin) the term "alternation" (zamiana), and he subdivided the consonants into "primary," "non-alternating" (niezamienne) and "secondary," "alternating" (zamienne) elements of an alternation. The thrust of Benni's critique concerns Mrozinski's identification of the former with hard consonants and the latter with soft, or palatalized, consonants. For Mrozinski himself such a supplementary phonetic criterion of classification was a matter of convenience, a notational device. "My physiological observations," he admitted; "may be wrong, but my classification of the consonants is correct." Jakobson expresses, on the other hand, the view that such a common denominator between the morphological and phonetic ("organogenetic") properties of the alternants does indeed exist, and he sees it in the "narrowing of the channel" that characterizes all the "secondary," "derived" alternants, i.e. the palatalized, palatals, and strident acute (c, j ) consonants. But this insistence on a common phonetic denominator between morphological alternants seems to me (as it seemed to Benni) forced, both on theoretical grounds and with respect to the Polish consonants. In the first place, phonetic change may completely alter the original phonetic affinity between alternants (e.g. the Common Slavic alternation t~t', d~d' which became in Bulgarian t~st, d~st); in the second place, Polish has such alternations as c ~ c , s,z~s,z (e.g. nasz~ nasi) or w ~ / (bialy ~ biali) which cannot be fitted into Jakobson's formula. Jakobson's article on Mrozinski remains, nevertheless, an excellent and fitting tribute to a linguist who, at the very beginning of the nineteenth century, recognized that progress in linguistics hinges on the inseparable conjunction of abstract theorizing with meticulous observation. The popular article on Polish-Russian cooperation in linguistics draws attention to the contributions of Polish linguists to the study of Russian and vice versa, and to the joint Polish-Russian efforts in the field of linguistics and poetics. Among the pathfinders in phonological theory, Jakobson reserves a special place for Henry Sweet, the founder of the English school of modern linguistics who was never offered a chair at an English university. 29 The practical tradition of the English and their colonial and missionary activities combined to impart to English linguistics a pragmatic bent whose most important result was the refinement of phonetic studies and the invention

448

EDWARD STANKIEWICZ

of an International Phonetic Alphabet. Sweet's immediate predecessors Ellis and Bell had already hit upon the idea of a two-fold phonetic notation which Sweet labeled Narrow and Broad Romic\ the first type of notation would render all the minutiae of sound (including "the superfluous sounddistinctions"), while the second, "less accurate" transcription would symbolize only those distinctions which are "independently significant in any language." By emphasizing that "each language uses only a few of the [relevant] distinctions," Sweet not only discovered the importance of phonological entities, but also foreshadowed the modern exploration of phonological universals. Jakobson underscores this aspect of Sweet's theory with special vigor: "If two vowels, even 'formed in a totally different way', 'are never employed together in the same language to distinguish the meanings of words, they must be considered as variations of the same vowel'." Thus, Jakobson concludes, "the extraction of invariants from intralingual variations is quite logically complemented by a daring and novel search for interlingual universal variations and for corresponding invariants." 3 0 Jakobson mentions briefly Sweet's broader activities in comparative grammar, dialectology, and general linguistics, finding striking parallelisms between his thought and that of his continental colleagues (especially Baudouin and Passy). It is not without interest that this parallel and as if spontaneous international movement in linguistics was bolstered by a process of cross-fertilization in the form of a lively correspondence (between Sweet and Baudouin) and of personal contact (between their students Daniel Jones and Lev Scerba). To H. Sweet, Jakobson reminds us, the English-speaking world also owes the notion of "distinctive features" which Bloomfield has, on his own admission, taken from Sweet. In turning to American linguistics, Jakobson focused his attention on the three leading linguists of the New World: Whitney, Boas, and Sapir. He wrote a special survey, "The World Response to Whitney's Principles of Linguistics," 31 which reveals the great esteem in which Whitney was held by his European colleagues. Of interest is Saussure's appraisal (unpublished in his lifetime) of Whitney's work, which provides a clue to his appeal to the Geneva scholar. On Sapir, Jakobson has never written a special study, but Sapir's presence is felt in many of his works, and he is probably the American linguist Jakobson quotes and admires most. Fairly long reviews are devoted to Boas: one (1943) gives a thorough analysis of Boas' ideas and work, and the other (of 1959) uses his work as a springboard for some of Jakobson's own ideas. 32 In the first of these reviews (the tone of which conveys Jakobson's admiration and friendship for Boas), we are presented with the portrait of the founder of American anthropological linguistics, a

THE HISTORY OF LINGUISTICS

449

man who stood at the intersection of two traditions: the older tradition of European historical linguistics and the new tradition of American descriptivism. The lasting result of the latter is, no doubt, the Handbook of American Indian Languages which was initiated by Boas in 1911 and continued by Sapir and by Boas' and Sapir's disciples. This great synthesis not only put on record a large variety of the languages of the New World, but aimed at a description of these languages that was still free of the taboos and idiosyncrasies of later American descriptivism. Jakobson singles out Boas' firm grasp of the questions of grammatical meaning: "Above all, he showed convincingly that each language has a peculiar tendency to select this or that aspect of the mental image which is conveyed by the expression of the thought. Different languages differently select those aspects of experience that must be expressed. Such 'obligatory aspects are expressed by means of grammatical devices', whereas some other aspects are taken as non-obligatory and are expressed by lexical means." 3 3 How this selection operates in concrete languages is illustrated by Jakobson in the second article (of 1959) on the basis of the English sentence "the man killed the bull" (taken from Boas' "Language"), which is shown to consist of a set of binary, marked and unmarked grammatical categories. The awareness that language is a classificatory system which combines (on all its levels) universal and language-specific categories places Boas in the tradition of W. von Humboldt, in opposition both to the particularistic tendencies of the Indo-European grammarians and the reductionist tendencies of older and recent "universal" grammar. Jakobson emphasizes Boas' relativism in the study of language and languages, which is not surprising for a linguist and ethnographer who had to confront a great variety of structurally diverse languages; on the other hand, Boas never came to embrace Sapir's or Whorf's belief in the dominance of language over thought, but believed instead that language is a "sufficiently pliable" tool to adapt to the needs of culture and society. Like most linguists of his generation (and like many younger American linguists) Boas never found a bridge between descriptive linguistics and historical grammar, being convinced to the end that only the latter holds a key to the explanation of linguistic facts. In observing the phenomena of structural similarity between lexically remote Amerindian languages, he was, nevertheless, ready to abandon the nineteenth-century dogma about the direct lineage of languages from a single "Ursprache" and to embrace the notion of linguistic diffusion that was at the same time (i.e. in the thirties) making headway in European linguistics. Boas, it seems, was quite unaware of this coincidence of ideas until Jakobson brought it to his

450

EDWARD STANKIEWICZ

attention and put into his hands Trubetzkoy's "Gedanken über das Indogermanenproblem" (1939). It is now, unfortunately, clear that this "brilliant" article ascribed exaggerated importance to the role of diffusion and that the theory of language mixing which reached such perverse proportions in the works of a Schuchardt or Marr, cannot replace, but may only supplement, the theory of genetic affinity. The remaining articles on the history of linguistics go back to Jakobson's Prague period and deal primarily with the program and figures of that school. A sketch on Masaryk 34 shows that highly cultured, democratic president of the Czechoslovak Republic as a man who was interested not only in his chosen fields of philosophy and history but also in problems of linguistics and in the activities of the Prague Circle. The necrologies of Trubetzkoy (of 1939) and Karcevskij (of 1955)35 underscore the merits of these prominent co-founders of the Circle in the areas of general and Slavic linguistics, and in the case of Trubetzkoy also in Caucasian linguistics. Noteworthy is Jakobson's reminiscence of Trubetzkoy's efforts in the field of Slavic diachronic phonology, which was meant to be crowned by A Prehistory of the Slavic Languages. Neither the conditions of his life nor the state of theoretical and Slavic historical phonology enabled Trubetzkoy to complete this work. A lasting tribute has now been paid by Jakobson to his former collaborator and friend by his publication of Trubetzkoy's letters and notes (to Jakobson), 36 a book which is a precious document on Trubetzkoy's development as linguist, literary historian, and social philosopher, and a monument to his times. Also of considerable historical interest is the necrology of A. Meillet (1937),37 whom Jakobson honors as a great Slavist and comparativist. Meillet's sympathies with the Prague School and with structural linguistics are well known. Jakobson's estimate of Meillet as the French champion of structuralism appears, however, from the perspective of time, overstated. Even though Meillet repeatedly emphasized that a language is a system (popularizing the phrase that language is "une système où tout se tient") and proclaimed loudly his support of Saussure and his structuralist ideas, he remained in his orientation and works a fairly orthodox comparativist. Thus he questioned the possibility of a general linguistics without a preliminary historical description of the languages of the world, reduced the question of the social aspects of language to that of the influence of culture upon language (mostly on its lexicon), and failed to produce a French school of structural linguistics. The judgment pronounced on Saussure by Trubetzkoy in a letter to Jakobson (in 1932) can be applied almost verbatim to Meillet: "What is valuable in his ideas about language as a system is

THE HISTORY OF LINGUISTICS

451

general and abstract, and explains the attitude of his students who speak a great deal about "system," but who have not written a single descriptive work of a living language."38 Two other articles by Jakobson provide an outline of the main tenets of the Prague School. Written in the heyday of the Circle (in 1932,1933)39 they now sound programmatic and very general. What they make exceedingly clear, however, is that the program and subsequent achievements of that school have splendidly withstood the test of time, and have expanded and vitalized the study of language more than any other school of modern linguistics. The merits of Jakobson in deepening and spreading the teachings of that school are manifest: pars magna fuit. Yale University

NOTES 1

Roman Jakobson, Selected Writings, II: Word and Language (The Hague: Mouton, 1971), 369-593. 2 Selected Writings, II, 414. 3 Indogermanische Forschungen, 7 (1891), V-X; H. Paul, Prinzipien der Sprachegeschichte (6th ed., Tübingen, 1960), p. 24. 4 The Letters of Jacob Burckhardt, ed. and trans. Alexander Dru (New York, 1955), p. 73. 5 C. S. Peirce, Selected Writings, ed. Ph. P. Wiener (New York: Dover Edition, 1966), p. 263. 8 A Baudouin de Courtenay Anthology, trans., ed., with introduction by Edward Stankiewicz (Bloomington: Indiana University Press, 1972), p. 276. 7 George Mounin, Histoire de la linguistique des origines au XX* siècle (Paris, 1967), p. 218. 8 Selected Writings, II, 350. 9 Ibid., 431. 10 "Nikolaj Kruäevskij, ego zizn' i nauinye trudy," I. A. Boduèn de Kurtenè, Izbrannye trudy po obslemu jazykoznaniju, I, eds. V. P. Grigoriev, A. A. Leont'ev (Moscow, 1963), 197. 11 Selected Writings, II, 428. 12 "The Kazan' School of Polish Linguistics and Its Place in the International Development of Phonology" (1960), Selected Writings, II, 394-428. 13 Cahiers Ferdinand de Saussure, 26 (1969), 12. 14 A Baudouin de Courtenay Anthology, p. 93. 15 Izbrannye trudy, I, 125. 16 Selected Writings, II, 420. 17 Ibid., 410. 18 Ibid., 413. 19 Edward Sapir, Language (1921), pp. 55-56. 20 N. S. Trubetzkoy, "La phonologie actuelle," Journal de psychologie, 30 (1933), 331-32. 21 A Baudouin de Courtenay Anthology, p. 60.

452 22

EDWARD STANKIEWICZ

R. Jakobson, "ZnaCenie Kruäevskogo v razvitii nauki o jazyke" (1966), Selected Writings, II, 429-50. 23 The citation comes from Henry Sweet, who in his Presidential Address of 1878 spoke in admiration of H. Paul's general linguistic ideas advanced in Beiträge IV; see H. Sweet, Collected Papers, ed. H. C. Wyld (Oxford, 1913), pp. 111-12. 24 "Jazykovedieskie boi Generala Mrozin'skogo" (1967), Selected Writings, II, 375-88; "Polish-Russian Cooperation in the Science of Language" (1943), ibid., 451-55. 25 Tytus Benni, "Jeneral Mrozinski jako psychofonetyk," Sprawozdania z posiedzen Towarzystwa Naukowego Warszawskiego, Wydzial ftzykoznawstwa i literütury, 6, 9 (1913), 77-93. 26 Sprawozdania Poznanskiego Towarzystwa Przyjaciöl Nauk, 3, 6 (1957). 27 Selected Writings, II, 378. 29 A decisive influence on Mrozinski was, according to S. Rospond, exercised by the book of H. L. Jakob, Grmdriss der allgemeinen Grammatik (1814). See Stanislaw Rospond, "Onufry Kopczynski," Rozprawy Komisji Jgzykowej, Warszawskie Towarzystwo Naukowe, (1971), 1-39. The book by Jakob has, unfortunately, not been available to me. 20 "Henry Sweet's Paths toward Phonemics" (1961), Selected Writings, II, 456-68. 30 Selected Writings, II, 462. 31 Whitney on Language - Selected Writings of William Dwight Whitney, ed. M. Silverstein (Cambridge, Mass.: MIT Press, 1971), XXV-XLV. 32 "Franz Boas' Approach to Language" (1943), Selected Writings, II, 477-88; "Boas' View of Grammatical Meaning" (1959), ibid., 489-96. 33 Selected Writings, II, 481. 34 "Jazykovye problemy v trudax T. G. Masarika" (1930), Selected Writings, II, 46876. 35 "Nikolaj Sergeeviö Trubetzkoy" (1939), Selected Writings, II, 501-46, and "Sergej Karcevskij" (1956), ibid., 517-21. 36 N. S. Trubetzkoy's tetters and Notes, prepared for publication by R. Jakobson [=Janua Linguarum series maior, 47] (The Hague: Mouton, 1975). 37 "Antoine Meillet zum Gedächtnis," Selected Writings, II, 497-500. 38 Letters and Notes, p. 241. 39 "La scuola linguistica di Praga" (1932), Selected Writings, II, 539-46; "Die Arbeit der sogenannten 'Prager Schule"' (1936), ibid., 547-50.

ROMAN

JAKOBSON

AND

OLD

CZECH

LITERATURE

FRANTISEK SVEJKOVSKY

If I were to summarize the basic relationship and contribution of Professor Roman Jakobson to the study of Old Czech literature, I would try to formulate it in two parts. The first would be that Jakobson entered the field of Old Czech literature in an untraditional manner and continued to follow untrodden and untraditional paths. His approach to Old Czech literature was truly "from another direction." This was primarily due to the fact that he was a FOREIGNER who had joined the ranks of native scholars. Furthermore, he chose a method DIFFERENT from those acceptable in this field of scholarship. It was indeed unusual and practically unique for a foreign scholar to become interested in Old Czech literature and specifically to such an extent that he began to study it systematically. Foreign scholars would treat only individual aspects of phenomena in this period, and they were primarily motivated by comparative interests. Jakobson, however, especially in the earliest periods of his work, up to the end of the 1930's, found Old Czech literature the object of his concentration both from the point of view of historical problems as well as the development of his theories and his method of literary analysis. It can therefore be said without exaggeration that this study created an organic and significant part of his scholarly activities as a whole. This brings us to the second part of our aphoristic characterization: If Jakobson himself influenced significantly the study of Old Czech literature by the novelty of his conclusions and motivations, Old Czech literature also repaid him in that its problems served as a further motivation, since it was the foremost factor that led to the expansion of his literary study. This it did primarily by means of its historical dimension. The attention of the scholar, whom life brought to Czechoslovakia at the beginning of the 1920's, was attracted to works which led him to the roots of the Slavic literary tradition, to the beginnings of Old Church Slavic literature, which, in further stages of its development, became the oldest and most extensive medieval literature in a native Slavic tongue,

454

FRANTISEK SVEJKOVSKY

Old Czech literature. These cultural and literary traditions were attractive also because, from the geographical point of view, they represented an area where history and creativity had for a long time reflected the confrontation of the two basic European spheres of political, social, cultural, and artistic concepts. This is the region which the modern era recognizes as the EastWest terminus. But in reality much more was involved than, for example, the impact of the meeting of the Greco-Byzantine and Romano-Latin tradition in the Middle Ages, a topic which especially interested Jakobson. Thus an area opened up which was undoubtedly broad enough and sufficiently motivating for the scholar, for whom there was always a significant breadth of interests oscillating between the Slavic sphere and the rest of the world, between the present and the past, between theory and practice, between concentrating on details and taking a broader approach to a topic. In connection with this, a significant share of Jakobson's work concerns primarily Czech topics: (1) his study of Old Czech literature, concentrated from the thirteenth to the fifteenth centuries, and (2) his study of Old Church Slavic literature. Both form an interesting antithesis to his interest in contemporary literature, e.g., literature of the nineteenth century. Jakobson's entrance to the study of Old Czech literature was not, as we have already indicated, a mere continuation of the scholastic direction current at the time, but it followed its own course and more than once carried the tone of a polemic with tradition. Undoubtedly there is always something refreshing and productive when a tradition is disrupted by a stimulus outside of convention and stability, if this stimulus shakes up the calm and simultaneously brings something positive. And this is exactly what happened here. Jakobson began to change the view of and approach to the study of the past and along with it the basic concept of "old" or "older" literature. His contribution included the transfer to the Czech milieu of the new principles of theory and method which had been born just shortly prior to this time and were destined to become one of the most progressive literary scientific trends of the twentieth century under the étiquette of Russian formalism. This was the first application of these new principles outside the Russian sphere, and at the same time their first systematic application to older phases of literary work, the literature of the Middle Ages. 1 There is no doubt that from the point of view of the traditional study of Old Czech literature, or literature in general, methodological principles began to be questioned and at the same time opinions concerning the achievements of the past began to change. The scholarly situation with which Jakobson came in contact upon his arrival in Czechoslovakia was dominated by the heritage of the nineteenth

OLD CZECH LITERATURE

455

century. On the one hand, the philological orientation prevailed; on the other hand, there was the sociological interpretation which presented a broad historical classification of a work and also tried to describe that which was common to a majority of authors and periods. Around the turn of the century, Jan Gebauer (1838-1907) brought the philological method, especially the study of medieval literature, to a high and productive level. He also turned attention primarily to the period of the initial development of literature written in Czech, i.e., the thirteenth to the fifteenth centuries; the older Old Church Slavic or Latin literature did not represent a special interest for him. His concentration on and high estimation of specific problems (paying special attention to the history of the language) predetermined, in fact, the main direction of further studies of Old Czech literature. In working on his main project, Historicka mluvnice jazyka ceskeho [Historical Czech Grammar],2 and on his extensive studies of old literature, he stabilized the methods of philological analysis and interpretation which were then adopted as a basis for the study of our literary past. Here also lay the foundation for a further feature of this study: concentration on the individual work and an analytical rather than a synthetic approach. The concept of a historical approach from a philological basis stressed another significant aspect: attention to the language and to its development produced between the "old" and the "new" eras a strong dividing line, which, according to concepts of that time, fell into the last decades of the eighteenth century. This stabilized the differentiation between "old" and "new" literature. Especially at the end of the nineteenth century this dividing line began to be felt not only by the scholars but in cultural life in general, when other phenomena (an orientation toward current life, "modernism") came into being. Under the influences of contemporary interpretation, creativity of the past became something removed from the present and its interests—especially in the sense of the work's individual qualities. A work's significance was sought primarily in its historical documentary value, indispensable, for example, to the study of history, religious ideas, and philosophy. This approach was strengthened by a broader conception of historical study that stemmed from nineteenth-century positivism. With a causal and sociological interpretation of the works, emphasis on ideological aspects came to the fore. In the Czech milieu the theory of H. Hettner was of special interest in this regard. The literary work and its evaluation throughout almost the entire nineteenth century was accompanied by a regard for the social function of the work as an important manifestation of the national existence and historical tradition within the political and cultural pressure of the Austrian

456

FRANTISEK SVEJKOVSKY

monarchy. Expressions of these tendencies were the two most important synthetic works : Jaroslav Vlcek's Dëjiny ceské literatury [History of Czech Literature] (published gradually beginning with 1893),3 and Jan Jakubec's Dëjiny literatury ceské [History of Czech Literature] (first appearing in 1911 and in a revised edition in 1929-1934).4 These works already included the conclusions of study conducted primarily by Gebauer's students. Their devotion to their teacher was expressed by continuing the work outlined by him. They primarily dealt with a critical analysis of individual works, the designation (or, when necessary, the reconstruction) of the archetype in harmony with the classical concepts of textual criticism ("Textkritik"), and finally with the classification of the work into the historical and thematic context of its period. The analytical approach dominated these synthetic works, a textual study peaked in its thorough description (possibly designating sources) and in editing. Any attempts at synthesis appeared as the sum of the conclusions obtained in the characterizations of individual works. This is what can be said about the basic direction of the study of Old Czech literature and consequently later literature linking it to the Middle Ages. Neither the study of literature in Latin nor interest in the sixteenth century influenced these approaches to literature; the study of literature of the seventeenth and eighteenth centuries still awaited systematic development. However, what was lost in all this was the actual LITERARY substance of these works. Interest in the literary substance was exhausted in the limited scope of predominantly descriptive data. It was mainly determined by the rhetorical tones of the work, by the character of the style, and, in the case of poetry, by verse description. All of this found support in traditional descriptive poetics. The individual characteristics of such works remained isolated, without any characterization of their mutual relationships or their relationships to the concept of the work as a whole, which would have permitted their actual value and significance to stand out. Evaluation of these aspects was usually based not on the reconstruction of the actual principles of the work itself, but was done under the influence of criteria obtained by contact with modern literature, or, in some cases, under the influence of subjective conclusions about old poetry. Besides methodological problems, severe shortcomings were felt in the area of historical poetics (and with respect to the main current of the study of medieval poetics), which was actually not a phenomenon typical only of the Czech lands but a general problem proposed by the Russian scholar A. N. Veselovskij (1838-1906) which initiated a thorough study of theoretical texts as demonstrated by E. Faral's Les Arts Poétiques du XIIe et du XIIIe siècle.5

OLD CZECH LITERATURE

457

If we ponder these important motivations which in the beginning of Jakobson's scholarly work reached into the study of older phases of European literatures, then of course we must not overlook other impulses which often without any direct connection with the specific problematics of the older works would nevertheless influence a change in the approach to them. What we mean by this is the fermentation of the artistic method and its echo in contemporary literary theory, which was so typical for the period at the turn of the century. Interest in a work, in its concrete linguistic form, in its uniqueness, in a work as an expression of a continuously repetitious act to conquer in a unique way by artistic expression the problem of life—all of this revealed the literature of older times in a new light. Older works ceased to appear only as the result of a stage of development leading from the past to the present, but rather began to invigorate genuine creativity at its foundations. They began to be felt as a living value close to the present day because of their very creative foundations. The roots of this, however, originate far before the turn of the century, beginning with philosophical concepts and ending perhaps with the new linguistic associations. And at the same time there is also that link with other spheres of art which similarly share the "uncovering" of the relations with the past, as can be detected, for example, in the relationship of symbolism to medieval art and expressionism to mannerism and baroque. Similar rejuvenation had affected folklore, or, in a broader connection, "primitive" folk art. This is the background against which one must view Jakobson's entrance and input into scholarship in Czechoslovakia, for at that time an analogous situation existed there. Study of Old Czech literature represented a calm which was not disturbed in a significant way by the flow of the modern current of theoretical views and criticism. Therefore the new activity in these studies was not primarily the result of scholarship in the field of old literature itself, but grew out of interest in the problems of artistic work or in new concepts of literature, as we can see from the works of F. X. Salda or A. Novak, or in the expressions of the writers themselves, who were turning to the literary heritage of the past. Nevertheless, or possibly for this very reason, the response coming from other fields represented a significant pattern which indicated that it was necessary and possible to have a new approach to old literature and to introduce it into direct contact with new literature, even the newest literature. There is no doubt that these tendencies prepared the way for the acceptance of concepts which Jakobson brought from his own Russian background. In addition, it was possible to link these tendencies with

458

FRANTISEK SVEJKOVSKY

another basic phenomenon of his scholarly orientation, with linguistics. It is worthy of note here that Jakobson's orientation toward linguistics was of double value to him in the Czech milieu. It brought him into contact with Czech linguists who were interested in becoming acquainted with his new ideas, linguists who had grown up in an atmosphere of "postNeogrammarian scientism" characterized by the views of V. Mathesius and J. Zubaty and by a development toward synchronic language study.6 And at the same time linguists oriented Jakobson's scholarship toward the literary text in its uniqueness or as understood primarily linguistically. Here the Czech milieu provided the foundations for a close relationship between the study of language and literature. There is no need to mention that in a short while the meetings of these various tendencies gave birth to an important center for the development of twentieth-century scholarship, the Prague Linguistic Circle. The fiftieth anniversary of its founding (1926) comes just this year when we celebrate Jakobson's eightieth birthday. From the point of view of literary history and theoretical problems, Jakobson's approach to literature embraced all the basic principles which brought him closer to linguistic concepts and specific problems of literary study in the broadest sense of the word. It gave him the opportunity to respond to actual questions from these non-traditional points of view. However, at the same time, his starting point in Russian formalism imprinted the aspects of specificity and novelty on his concrete work. His concept of a literary work managed to embrace the creativity of all historical periods while at the same time respecting their uniqueness. He significantly expanded the importance of folklore and expressions of oral literature which prior to this time had a secondary place in the study of medieval literature. The work itself became the focal point of Jakobson's study, and his approach to a work was carried out with particular stress on its specific literary qualities given in its form of expression. Let us not forget that in the period when Jakobson began to make use of these approaches to Czech literature, he already had a statement behind him which even by today's standards is classic: "The object of a literary study is not literature, but literalness."7 From the point of view of the current situation in the study of Old Czech literature, this concept included much that was basic and novel and capable of rejuvenating and productively developing further work in this field. It is important to note that in suggesting a synchronic approach Jakobson supplanted the approach of traditional criticism based on criteria which were derived from works of later stages and transferred back into the past. Together with this there

OLD CZECH LITERATURE

459

was also the question of historical poetics and its theoretical nucleus. The normative concept of medieval poetics, in particular, contained several suggestions close to Jakobson's theoretical interest and to his study of the connection between poetics and linguistics. In connection with this Jakobson had a great share in building an important bridge between the entrenched differentiation of old and new literature. He supported it by both historical explanations and theoretical concepts as well as the intertwining of interest in the works of both the past and the present. 8 Thus, shortly after his arrival in Czechoslovakia in June 1920, Jakobson's interest began to move through the entire field of Czech literature with his major concentration in medieval literature or the nineteenth and twentieth centuries. Thus, in his linguistic works, alongside the documentation from Russian literature there frequently appear also Czech themes. Orientation in the period of Old Czech literature therefore cannot be viewed in the context of Jakobson's entire work as an incidental phenomenon but, on the contrary, as we have indicated, as an important and motivating part of his work. From the relationship to Old Czech literature, along with other literary and linguistic studies, there arose the first expressions of what we today find to be his major achievements in literary study. It is characteristic of his work that the solution of historical problems goes hand in hand with the solution of theoretical problems. That is how we should view and evaluate the importance of his works devoted to the topics we shall examine. B. Ejxenbaum, in his article "The Theory of the 'Formal Method'," is correct when he says about this type of approaches to literature: "In our studies we value a theory only as a working hypothesis for helping us discover and interpret facts." 9 Jakobson marked his entrance into the study of Czech literature of the Middle Ages by the selection of a subject that at the beginning of the 1920's was characteristic from the point of view of both the Russian origin of his method and the Leitmotiv of his entire future work: he began to work with Old Czech verse, and his main interest settled on poetry. Not only could he use his linguistic interest in connection with these themes, but he could also link them to his other work which had introduced him in the international forum, that is, the studies Novejsaja russkaja poezija (1921) and O cesskom stixe preimuscestvenno v sopostavlenii s russkim (1923).10 In the study of Old Czech verse he, of course, had precursors among native scholars. He continued their tradition, but at the same time he re-evaluated their approach and the overall development of the study of prosodic problems. This is evident primarily in his first study published in Russian

460

FRANTISEK SVEJKOVSKY

"Starocesskie stixotvorenija slozennye odnorifmennymi cetverostisijami" (1924).11 The general problem of this work is closely tied to the study O cesskom stixe. Here he tried to resolve, for example, the problem of the syllable and stress in Old Czech poetry; in his conclusions, he critically settles current views of Czech scholars compiled at that time in J. Krai's Czech Prosody.12 We should point out that even in this work he does not concentrate primarily on the problems of prosody, but rather indicates the possibility of exploiting conclusions from this level and applying them to other aspects of literary work, primarily the problems of style and genre. In connection with this work it is necessary to add a number of other studies, for the time being only in manuscript form, that stem from the same period. 13 In them Jakobson studies further significant verse compositions such as, for example, the Alexandreida and the collection of poems in the Hradecky Manuscript. It was in the analysis of these verse techniques that he showed the possible function of verse study on a broad level. He utilized these methods in connection with the problem of their authorship, dating, and other textological problems. The peak of this series of studies was the synthetic work "Vers starocesky" [Old Czech Verse] (1934), published in the representative encyclopedia entititled Ceskoslovenskd vlastiveda. Here Jakobson not only worked out the typology of Czech medieval verse but also classified it briefly according to its relationship with Slavic verse, especially with Old Church Slavic verse composition. 14 This part of the study may be considered as one of the first pieces of evidence of Jakobson's interest in Slavic versification. From the point of view of method, it is characteristic for his study of Old Czech verse that Jakobson views verse in connection with a number of other aspects of the poem: verse-lyric, epic, drama; verse-song, recitation ; verse - genres; verse - style. This is the way in which he wishes once again in harmony with the effort to develop the ideas of the Russian formalists - to embrace more broadly than before the structure of a literary work from a single chosen basis, in this case verse. Besides this, in the study "Old Czech Verse" there was another new conceptual unit: the utilization of diachrony. It appeared, however, not to lead to the examination of the history of Old Czech poetry, but to focus attention on developmental changes on the level of prosodic composition. This brought about the realization of new advances from the original positions of Russian formalism, signaled in 1928 by the theses in the article "Problemy izucenija literatury i jazyka," which Jakobson published together with Ju. Tynjanov. 15 Among other things they emphasized the need to "re-examine principles of diachrony" (Thesis 4). To show how organically this

OLD CZECH LITERATURE

461

orientation in Old Czech verse falls into the whole of Jakobson's work of that period, we need to add merely that this was the time when he introduced the principle of diachrony into his phonological studies and thus made a fundamental contribution to their development. We can continue in this historical confrontation and follow the tone of coordination of various spheres of Jakobson's language and literary study. Among his themes from 1928 on there appeared a determination to study phenomena in literature in connection "with other historical series" (Thesis 2) and at the same time further determination to replace the "aggregate of p h e n o m e n a . . . by the concept of a system" (Thesis 4). Jakobson applied all this productively in a further study which in its origin and subject matter was linked to Old Czech verse but which differed significantly in the concept of historical method. We are speaking of the study "Ovahy o basnictvi doby husitke" [Treatise on the Poetry of the Hussite Era] (1936), presented at a lecture in the Prague Linguistic Circle a year earlier. 16 This expanded Jakobson's view of Old Czech poetry into the fifteenth century. 17 In it the author did not abandon the basic fundamentals of the text but broadened their interpretation first of all by the introduction of linguistic elements which had an extra-textual relevance to the historical social situation and Hussite ideology. It is understandable that to him it was not a matter of a traditional formulation of "period background for literature," but rather the pinpointing of forms, how the extra-texual situation comes through in the formulation of the text, what is its importance to the content and the structuring of the poetic composition - that is, the "motivation" of means and approaches which were utilized in the poetry of the Hussite era. Thus he remained on the textual level, but in his approach to it, aspects of the work as a whole came to the foreground as a system and became the foundations for the study of "a correlation between the literary series and the other historical series." His attention turned to the concept of the work and its semiological problems. In order to identify the principles for the study of the interconnection of these series, he first approached the questions linked with the concept of the work and its semiological aspects from the plan of the epistemological and theological trends that determined the broad social and cultural context of the Hussite era; from there he turned to the specific problem of literature. The concrete situation of the fifteenth century brought him first of all to one of the important questions of that period, to the change in the hierarchy of the function of liturgical and cultural Latin and the native language, Czech. Then he switched to an analysis of the concept of sign and significance during the Hussite period in a chapter entitled significantly "Signum et

462

FRANTISEK SVEJKOVSKY

Signatum." This formed the basis from which he then came to the linguistically oriented analysis of the means, with special attention to the prosodic plan. Here, finally, he met the complex of propositions observed in "Old Czech Verse." There is no doubt that we can witness here significant developmental changes in the approach to a literary work: from the study of the individual levels of a work or the relationship of two extracted levels, as was frequent in the work of the Russian formalists, we come here to the complex view of a literary work. This approach, of course, also carries theoretical importance that increases when regarded from today's retrospective points of view. This study introduced the THEORY OF A LITERARY TEXT as a specific sign system and its relation to the extra-textual situation. In connection with this we see here the SEMIOLOGICAL aspect in his interest not only for studies of the literary creation but also for other levels of social and cultural life. If we consider along with this the fact that this study arose so near the studies frequently recalled today about Pasternak and Puskin ("Randbemerkungen zur Prosa des Dichters Pasternak" in 1935 and "Socha v symbolice Puskinove" [The Statue in Puskin's Symbolism] in 1937), 18 which are considered to be symptomatic articles in the development of Jakobson's theory of literature, then we have to add to them also this study, "Treatise on the Poetry of the Hussite Era." It belongs to them not just by its chronology but also because of the importance of its new conceptual characteristics. So much for the context of Jakobson's work as a whole; but it cannot be pinpointed in totality if we at the same time ignore a broader look at the important development of ideas within the Prague Linguistic Circle, into which the overall orientation in the study of language and literature is classified organically and which actually represents in persona one of the principal links between linguistic and literary study. 19 It was around the mid-1930's when interest in developing the problems of structure and sign came into the forefront. In 1934, J. Mukarovsky announced at the International Philosophical Congress in Prague: ". . . The problems of sign and meaning are becoming more and more pressing, for every spiritual content which extends over the limits of individual consciousness acquires, by the very fact of its communicativeness, the force of a sign. The science of the sign (semiology, according to de Saussure; sematology, according to Biihler) must be worked out in all its breadth the same way as contemporary linguistics (compare the study of the Prague school, i.e., the Prague Linguistic Circle) expands the field of semantics, showing from this point of view that all the elements of linguistic structure, possibly even sounds.

OLD CZECH LITERATURE

463

must be the results of linguistic semantics applied to all the other sequences of signs and differentiated by their special characteristics." 20 In connection with this let us recall older studies by S.I. Karcevskij (for example, "Du dualisme asymétrique du signe linguistique"), 21 J. Mukarovsky's study from the area of literary scholarship and aesthetics (Estetickâ funkce, norma a hodnota jako sociâlnifakty [Aesthetic Function, Norm and Value as Social Facts] in 1936),22 or the ethnographic and folkloristic works of P. Bogatyrev. 23 This evoked a number of other problems which had so far remained outside the main direction of scholarship and at the same time focused attention on language and the literary text. I do not wish to influence the view of Jakobson's work as a whole by concentrating attention on his studies of old literature. However, I would like to repeat that this is in truth a significant part of his work in the 1920's and 1930's. In view of the scope of his studies and their continuity, it was an important sounding board for his later studies of literary works and general questions on poetics. And, on the other hand, by which I mean the situation in the study of Czech literature of the older period, this work played an important role in its development toward enriching and changing the current study methods. At the same time, nevertheless, Jakobson more than once surpassed the scope of medieval literary study by his activities and embraced the entire area of literary history and theory. Possibilities in these directions were increased even more when in 1933 he had the opportunity to develop his lecturing activities at the university. Jakobson's relationship to older literature is not at all exhausted. In connection with the works mentioned above, he was more than once brought to other themes where the dual polarity of his literary and linguistic interests could be utilized. He became interested in TEXTOLOGICAL problems, as was reflected in his publication of several medieval texts. First of all he was interested in the problem of the original version of the song Hospodine, pomiluj ny; this text was known from much later manuscripts from the beginnings of the fourteenth century. In his publication Nejstarsi ceské pisnë duchovni [The Oldest Czech Spiritual Songs] (1929),24 he attempted to reconstruct and at the same time analyze texts closely linked to his other research on medieval poetry in the Czech lands. In this work he also treated the theory of the various levels of the surviving versions of the Song of St. Wenceslas. To them he added an analysis of other Old Czech religious works, Hail Mary and The Logos at the World's Creation,25 He applied his conclusions on prosodie problems especially constructively in the difficult reconstruction of the poem The Dispute Between Body and

464

FRANTI§EK SVEJKOVSKY

Soul (1927). 26 In the introduction to the transcribed edition of this text together with another Old Czech poem, On the Dangerous Time of Death, we also find the first extensive attempt by Jakobson to review the situation of Czech medieval poetry with special attention to opening up access to these verbal works to a broader cultural consciousness among contemporary Czech readers and to present them as actually living works. 27 Finally Jakobson tried at that time to resolve another difficult problem: to reconstruct the original version of a secular song from the fifteenth century from a defective manuscript, a task made especially difficult since the manuscript was marked by the influence of a Polish language milieu. This was in the article "Slezsko-polska cantilena inhonesta ze zacatku XV stoleti" [A Silesian-Polish Cantilena Inhonesta from the Beginning of the Fifteenth Century] (1935). 28 On the one hand, the poem produced interest in analyzing some aspects of worldliness in medieval literature and at the same time oriented attention to the relationship between literature and folklore. Besides this, it is the first signal of Jakobson's interest in the Czecho-Polish literary relations during the older historical periods, i.e., themes to which Jakobson later returns several times. If it would be possible to say at this time that the above-mentioned works include everything significant that Jakobson achieved during the two decades, one other area would remain, and this we have excluded on purpose. Briefly this is a group of articles not primarily concerned with the study of specific problems of literary works but rather with problems of the broader social function of literature. In Jakobson's work as a whole, there is definitely a single approach, but it is interesting for that very reason. In these works we see one important characteristic: at a certain point during the second half of the 1930's the scholar felt the need to speak in a voice different than the one so characteristic for him. Here he brought to the foreground the ethical and social responsibility of his work in relationship to actual problems of contemporary life. This was also his approach to literature. What were the roots of this motivation ? It is concerned with the history of the period shortly before Hitler's incursion into Czechoslovakia and into Europe as a whole. It is contained in the history of the face-to-face confrontation of Europe and the whole world with the threat of war. From time immemorial Czechoslovakia had had written into its history a tendency to become from time to time the first sounding board for an approaching catastrophe that would reach far beyond the borders of the country. And at the end of the 1930's the feeling of an existential tension, the threat of a loss of freedom, began more and more to penetrate its entire life. It transformed the paths and aims of artists and scientists, each of whom

OLD CZECH LITERATURE

465

began to examine his relationship to the basic values of life and the world. Roman Jakobson was one of them. At this point he showed how much he had merged with the Czechoslovak mileu. He emerged on the battleground that held the meaning of his life's effort, in the field of scholarship. And it is significant that once again he spoke out by means of his relationship to Old Czech literature. His first important expression of this sort was an extensive polemic with the conclusions of a recent work by a German cultural historian, K. Bittner, which Jakobson called "Usmérnéné názory na staroceskou kulturu" [Revised Views of Old Czech Culture] (1936).29 Bittner's ideas were in harmony with the tendencies of the German scholars to support the Nazi policies toward Czechoslovakia with historical data. Jakobson came out with a piercing, documented criticism which, through analysis of the problems of medieval culture and first of all literature, reached all the way to the present and unmasked the principles of Bittner's actual intent. The truth lay on Jakobson's side, but against it, against him, and against the whole country stood the argument supported not so much by historical data but by the will to subjugate and conquer. Later Jakobson added to this position a sequel "Neni pravda, ze . . . Odpovéd' na brozuru K. Bittnera 'Deutsche und Tschechen. Eine Erwiderung'" [It is not true that . . . Answer to the Brochure by K. Bittner: "Deutsche und Tschechen. Eine Erwiderung"] (193 8). 30 But then he himself was compelled to leave Czechoslovakia to escape the threat of force. It is clearly symptomatic that this happened the very instant Hitler's troops were moving onto Czech soil. The book published after Jakobson's arrival in the United States was the high point and at the same time the new form of expression of the scholar's involvement in the existential and cultural struggle of a country under attack by an enemy not satisfied by sheer physical force alone but reaching out to attack even the highest values, the spiritual values. Its title was adopted from the work of J. A. Komensky: Moudrost starych Cechü [ Wisdom of the Old Czechs].31 Its polemic tone was blended with the tendency to observe through the medium of literature the development of the Czech political and cultural consciousness. On this basis Jakobson re-evaluated more than one of the previous conclusions from his study of Czech literature. The same tone was carried by the small publication Cesství Komenského [The Czech National Feeling of J. A. Komensky].32 However, upon departing from Czechoslovakia, he left there another variant of genuine scientific legacy, speaking to the present by recalling significant characteristics of the past. He contributed to the work that was actually born, in the atmosphere of the Prague Linguistic Circle, of the

466

FRANTISEK SVEJKOVSKY

scholars' effort to say something of utmost importance for the times. This was a collection of articles entitled: Co daly nase zeme Evrope a lidstvu [What Our Countries Have Given Europe and Mankind].33 As his part Jakobson prepared two themes: "Cesky podil na cirkevneslovanske kultufe" [Czech Contributions to the Church Slavic Culture] and "Cesky vliv na stredovekou literaturu polskou" [Czech Influence on Old Polish Literature]. These studies could no longer be published under his name; he hid behind the pseudonym Olaf Jansen. Since the 1940's Old Czech works began to recede from the main field of Jakobson's literary research. This does not mean, however, that they disappeared altogether. He would return to this topic - and returns to it even today - as though to an area familiar and dear to him. The distant land leaves traces in his entire later work. Were we to review the bibliography of these years, we would find in it "Old Czech" themes, which suggest that the studies begun before the War still influence him. New echoes of Old Czech poetry have been heard since the 1960's when his focus on the semantic analysis of the various poetic texts began to expand to the basis of the "grammar of poetry." When he turned to diverse material in order to demonstrate and probe his theory, texts from the history of Czech poetry emerged among the others. Attention to the older works and to metric and textual reconstruction gave rise to a further study "Staroceska basen o prvotnim hrichu" [The Old Czech Poem on Original Sin] (1957).34 Another echo of past interests can be heard from his return to the so-called Kunhuta's Song in the article "Nepovsimnute filiace" [Neglected Filiation] (I960). 35 This time his motif concerns the echoes of one text within other texts (here in the prayers of a fourteenth-century priest, Jan Milic from Kromeriz). In the same publication he presented another part of a study on "Slovanske duchovni dejiny v pojeti Jana Amose Komenskeho" [Slavic Spiritual History in the Conception of J. A. Komensky]. 36 Two other studies are important for their methodological and historical conclusions: The work "Medieval Mock Mystery (The Old Czech Unguentarius)" dealt with the medieval theater. 37 On the basis of the Old Czech dramatic text from the fourteenth century, he resurrected general questions of comic and religious symbolism in medieval theater. At the same time he touched upon the connection of this work with folklore and through it upon the traditions of pre-Christian ritual. It is worthy of note that Jakobson had at one time acquainted his students in Czechoslovakia with some of the ideas of this study, so that they became a part of Czech scholarship even before he

OLD CZECH LITERATURE

467

put them into this definitive form in 1958. Something new in Czech contemporary medieval studies was represented by the attention devoted to medieval Jewish literary culture, which Jakobson characterized in his article "Rec a pisemnictvi ceskych zidu v dobe premyslovske" [Language and Writings of Czech Jews in the Premyslid Times] (1957).38 Here he pointed out the European and domestic interrelations between Jewish education and verbal culture in Czech lands. Moreover, this contribution called attention to theoretical works relating to medieval literature and language. And now our chronological survey brings us to the recent decades, into the period of Jakobson's study of problems of "linguistics and poetics." Among the studies which he has analyzed on this basis, there are selected texts from Dante and Shakespeare all the way to Brecht. We also find an analysis of the Hussite Choral work Who Are God's Soldiers ? (1963) and the oldest known song Hospodine, pomiluj ny, included in the study "Staroslovenska kantilena v dSdictvi ceskem" [Old Slavic Canticles in the Czech Heritage] (1972).39 Although it was our task to concentrate in this article on the scope of studies devoted to Old Czech literature, the fact that Jakobson's work keeps returning to the above-mentioned composition Hospodine, pomiluj ny forces us to glance at another sphere of study: studies devoted to the relationships between Old Czech and Church Slavic creativity. We have mentioned before how Jakobson's principal interest was concentrated on poetry from the thirteenth to the fifteenth centuries and from here it spread deeper into the past rather than in opposite directions. It was this period, the beginning of writing on Czechoslovak territory, that produced an important complex of Church Slavic literature which Jakobson began to study systematically from the end of the 1930's and especially in the following decades. Special studies in the present volume treat these subjects, but we would nevertheless like to mention at least the concept so vehemently pursued and studied by Jakobson: I have in mind the idea of continuity and the meaning of the relationship between Old Church Slavic and Old Czech literature and at the same time the important role of Cyril and Methodius's bequest to the further development of the Czech lands. Jakobson formulated these interrelations clearly at that time in his "Treatise on the Poetry of the Hussite Era" (1936), in the "Glosy k legende o cv. Prokopu" [Glosses to the Legend of St. Prokopius] (1937),40 and primarily in the studies devoted to the song Hospodine, pomiluj ny, that became the direct expression of the ties between these two spheres in his work. Several times he would return to the questions of its archetype, while at the same

468

FRANTI§EK SVEJKOVSKY

time he tried to obtain from his conclusions an answer to a most difficult question, the question of dating. Together with this he returned over and over again to the broader question: does this work belong to the period of Slavic ritual in Bohemia or to the oldest Cyrillo-Methodian period and Old Church Slavic literature ? An important link in the development of his hypothesis of the song Hospodine, pomiluj ny was the study "O stixotvornyx reliktax rannego srednevekov'ja v cesskoj literaturnoj tradicii" (1950),41 where he supported his theory: it is the first Czech monument. At the same time he returned there to other compositions that first interested him in the early work The Oldest Czech Spiritual Songs, i.e., the compositions St. Wenceslas and Hail Mary.42 The last word on the question of the song Hospodine, pomiluj ny is in the work "Old Slavic Canticles in the Czech Heritage" (1972). There the analytical approach to the text and to its literary problems is expanded. Jakobson himself deserves credit in this expansion also for the fact that he inspired new musicological studies of the Slavic song tradition in relationship to the Byzantine tradition, especially to the Hirmoligion. He reached a conclusion which supported another hypothesis (mentioned above) that moved the origin of the composition to the ninth century - the period of Constantine and Methodius. Nevertheless the tradition during later centuries gave it a lasting place among CZECH creativity. These studies Jakobson linked also with the analysis of the prosodic and grammatical plan of the work and included an important confrontation with the musical plan. This last study bears the date 1972; the first song study falls into the 1920's. Thus the intensive relationship and the return to the song Hospodine, pomiluj ny is clear proof that the circle of Jakobson's interest in the literature of the Czech past is remarkably uniform and is clearly delineated from the very beginning. I have tried to show that within his works there has taken place a lively evaluation of old verbal values, which at the same time have become a living component of Jakobson's theoretical efforts - a process which is still in progress. Thus from both points of view he has offered and still offers many valuable conclusions and suggestions concerning the study of Old Czech literature. But it is also true that the spiritual energy and creative force stored up in the legacy of the distant Czech past has given him and still gives him, JAKOBSON THE SCHOLAR, in return more motivations so that he returns to it today more than fifty years later. And it brings enjoyment to JAKOBSON THE READER, who not only is capable of grasping artistic values precisely but can sample them, whether they belong to modern authors or to the authors of the past, removed perhaps by a thousand years. It goes without saying that a long

OLD CZECH LITERATURE

469

time has passed since Jakobson was a "foreigner," as I labeled him at the beginning of this study, to Czech literature and its traditions. University of Chicago

NOTES 1

In the studies of the Russian formalists there was no special interest in problems of old literature, especially medieval. One author who was inclined in this direction, Aleksandr Skaftymov [Poétika i genezis bylin (Saratov, 1924)], developed his work outside of the main trend. As proof of an older interest of Jakobson in Old Slavic literature, we must mention one of his first published works: "Zamétka o drevnebolgarskom stixoslozenii," Izvestija Otdelenija russkogo jazyka i slovesnosti Rossijskoj Akademii Nauk XXIV, No. 2 (1922), 351-58. 2 Prague: I. Hláskoslovi, 1894; III. Tvarslovi, 1. Skloñováni, 1896; III. 2. Casování, 1898; IV. Skladba, 1929. 3 Prague: 1893-1920. 4 Prague: J. Laichter, 1911; Prague: J. Laichter, 1929-1934. 5 Paris: É. Champion, 1924. 6 To throw light on the situation in that period we find interesting the description in Vilém Mathesius' article "New Currents and Tendencies in Linguistic Research" [MNHMA, Sbornik na poces/ Josefa Zubatého (Prague, 1927), pp. 188-203] and in the introductory chapter to Josef Vachek's The Linguistic School of Prague [Bloomington, Indiana: Indiana University Press, 1966]. See also in Vachek's book the translations of articles by V. Mathesius and B. Trnka. 7 Printed in the study Novejsaja russkaja poézija (Prague, 1921), originally a lecture given at the Moscow Linguistic Circle in 1919. For additional data on all of Jakobson's works cited in this article, I refer the reader to Roman Jakobson: A Bibliography of his Writings (The Hague-Paris: Mouton, 1971). 8 A similar tone can be seen in some of the comments and articles Jakobson suggested to the reading community beyond the scope of their special interest. 9 Boris M. Éjxenbaum, "Teorija 'formal'nogo metoda'" in Literatura, Teorija, Kritika, Polemika (Leningrad, 1927). 10 Novejsaja russkaja poézija (Prague, 1921); O cesskom stixe .. . (= Sbornikipo teorii poéticeskogo jazyka, V) (Berlin-Moscow, 1923; reprint Providence, R. I.: Brown University Press, 1969). 11 Slavia III (1924), pp. 272-315. 12 Josef Král, Ceskáprosodie (Prague: J. Otto, 1909). Jakobson paid special attention to Král's views in his article "O Králové Ceské prosodii," Kritika II (1925), 110-14. 13 An edition of this manuscript has been prepared for Jakobson's Selected Writings, Vol. V: Early Slavic Ways and Crossroads (The Hague, in press). His own studies of Old Czech verse were later used by Jakobson in his article "Starocesky vers a Rukopisy" ( = Part I, "K casovym otázkám nauky o ceském versi"), Slovo a slovesnost, I (1935), 46-53. Here he undertook an analysis of the "fruits of Czech romanticism," i.e., poems put forth as original medieval texts. By comparing them, he pointed out the difference in the metric base of these texts and the traditions of medieval compositions. 14 Further thoughts and conclusions on this theme were offered in his article "Cesky verá pred tisíci lety" (=Part II, "K casovym otázkám nauky o éeském versi"), Slovo a

470

FRANTISEK SVEJKOVSKY

slovesnost, I (1935), 46-53. Besides the song Hospodine, pomiluj ny, he focuses his attention especially on the so-called Kievan Missal and the Polish monument Bogurodzica. Here Jakobson reacts simultaneously to the conclusions of modern scholarship on this topic. 15 Jakobson and Ju. Tynjanov, "Problemy izucenija literatury i jazyka," Novyj Lef, No. 12 (1928), 36-37; English translation in The Structuralists from Marx to LéviStrauss, ed. Richard and Fernande De George (Garden City, N.Y., 1972), pp. 81-83. 16 Slovo a slovesnost, II (1936), 1-21. 17 In the area of fifteenth-century poetry, he devoted special attention to the isolated historical song about the Battle of Varna in "Staroceskâ pisen o bitvë u Varny 1444," Slovo a slovesnost, III (1937), 189-90. Here he continues the work of the Czech historian R. Urbânek and adds several versological comments. 18 "Randbemerkungen . . .," Slavische Rundschau, VII (1935), 357-74; "Socha . . . , " Slovo a slovesnost, III (1937), 2-24. 19 This role of Jakobson was emphasized by V. Mathesius in his retrospective "Deset let Prazského lingvistického krouzku," Slovo a slovesnost, II (1936), 142. Published in slightly abridged form as "Ten Years of the Prague Linguistic Circle" in The Linguistic School of Prague, ed. Josef Vachek (Bloomington, Ind., 1966), pp. 137-51. 20 Jan Mukarovsky, "L'Art comme fait sémiologique," Actes du Huitième Congrès International de Philosophie à Prague (Prague, 1936). 21 Travaux du Cercle Linguistique de Prague, I (1929), 33-38. 22 Praha: Fr. Borovy, 1936; in English translation by Mark E. Suino, Aesthetic Function, Norm and Value as Social Facts ( = Michigan Slavic Contributions, 3) (Ann Arbor, 1970). 23 A bibliography of P. Bogatyrev's work from this period is in Materiâly k retrospektivni bibliografii ceské etnografii a folkloristiky, 3 (Brno: Università J. E. Purkynë, Fakulta filosofickâ, Katedra etnografii a folkloristiky, 1968). 24 — Närodni Knihovna, VI (Prague, 1929). 25 He returned once again to the problems included in his book Nejstarsi ceské pisnë duchovni in the "Odpovëd' dr. Frantiäku Tichému," Slavia, X (1931), 396-400 (i.e., a reply to an editor) and then primarily after the publication of the work by the musicologist D. Orel. In his study of it, Jakobson had the opportunity both to compare his conclusions with this work and also to add further details from his own observations. This is what we find in the article "Z dëjin staroceského zpëvného bâsnictvi" [From the History of Old Czech Choral Poetry], Slovo a slovesnost, IV (1938), 41-44. 26 Spor duse s tëlem; O nebezpecném casu smrti (=Närodni Knihovna, IV) (Prague, 1927). 27 Another example of popularization is his edition of several Old Czech poems with a brief introduction, "O cestâch k ceské poesii gotické" [On the Roads to Czech Gothic Poetry], ¿ivot, XIV (1936), 57-63. 28 Närodopisny vëstnik ceskoslovansky, XXVII-XXVIII, 56-84. 29 Slovo a slovesnost, II, 207-22. 30 Slovo a slovesnost, IV, 117-23. 31 New York, 1943. 32 New York, 1942. 33 Prague, 1939. 34 Listy filologické, V (LXXX), 204-10. 36 "Kunhutina skladba a modlitby Milicovy," Scando-Slavica, VI, 26-29. 36 Scando-Slavica, VI, 30-34. 37 Studia philologica et litteraria in honorem L. Spitzer (Bern, 1958), pp. 245-65. 38 Rok 1957 (New York: Moravian Library), 35-46. 39 Ricerche slavistiche, XVII-XIX (1970-1972), 259-72.

OLD CZECH LITERATURE 40

471

XL1X. rocenka Chudym detem: iivy Vrchlicky (Brno), pp. 65-77. Slavisticna Revija, III (Ljubljana), 267-73. 42 In this study Jakobson takes a stand on several new views presented in the editions and analyses of texts by A. Skarka in his book Nejstarsi ceska duchovnl lyrika (Prague, 1949). 41

POÉTIQUE G É N É R A L E TZVETAN TODOROV

Lorsqu'on tente de saisir, dans une vue d'ensemble, l'œuvre d'un poéticien, la question qui surgit la première est: qu'est-ce que la littérature? Il se trouve que, tout au long des écrits de Jakobson, cette question - et sa réponse - ont toujours été présentes; au point que l'une des études porte comme titre: "Qu'est-ce que la poésie?" La réponse, à travers de légères variations terminologiques, reste étonnamment stable. En 1919 Jakobson écrit : "Je qualifie de moment unique et essentiel de la poésie cette visée de l'expression, de la masse verbale... La poésie n'est rien d'autre qu'un énoncé visant à Vexpression." En 1933: "Le contenu de la notion de poésie est instable et varie dans le temps, mais la fonction poétique, la poéticité, comme l'ont souligné les Formalistes, est un élément sui generis... Mais comment la poéticité se manifeste-t-elle? En ceci que le mot est ressenti comme mot et non comme simple substitut de l'objet nommé ni comme explosion d'émotion." Et en 1960: "La visée (Einstellung) du message en tant que tel, l'accent mis sur le message pour son propre compte, est ce qui caractérise la fonction poétique du langage." L'emploi poétique du langage se distingue des autres emplois par ceci que le langage y est perçu en lui-même et non comme un médiateur transparent et transitif d'"autre chose". Le terme ainsi défini est, en 1919, la poésie; il devient plus tard le poétique (la fonction poétique), c'est-à-dire la catégorie abstraite que l'on saisit à travers le phénomène perceptible. Mais la définition même n'a pas changé. Le langage poétique est un langage autotélique. Cette idée n'est certes pas une invention des Formalistes; et ils ne sont pas les seuls à la défendre à notre époque. Elle leur vient des futuristes russes (le "mot autonome" de Xlebnikov) et, moins directement, de Mallarmé, qui lui-même résume une tradition romantique dont le point culminant se trouve chez Novalis. (Mallarmé et Novalis, ces deux- noms apparaissent fréquemment associés dans les premiers écrits de Jakobson.) Novalis écrit, par exemple: "Le langage à la deuxième puissance, par exemple la fable, est l'expression d'une pensée totale - et appartient à la hiéroglyphique de la deuxième puissance - au langage sonore et écrit de la

474

TZVETAN TODOROV

deuxième puissance. Elle a des mérites poétiques et n'est pas rhétorique subalterne - lorsqu'elle est une expression achevée - lorsqu'elle est correctement et précisément euphonique à la deuxième puissance - lorsqu'elle est, pour ainsi dire, en même temps une expression pour l'expression lorsque du moins elle n'apparaît pas comme moyen - mais qu'elle est en elle-même une production achevée du pouvoir linguistique supérieur." "L'écrivain n'est en vérité qu'un passionné du langage." L'expression pour l'expression, le langage pris comme but et non comme instrument: ce sont des formulations qui rappellent de près "la visée du message en tant que tel" ou "l'accent mis sur le message pour son propre compte". C'est de la même manière que Du Marsais définit la figure et Goethe, le symbole. A notre époque, cette même thèse romantique est soutenue par Sartre dans Qu'est-ce qu'écrire?: "Les poètes sont des hommes qui refusent & utiliser le langage. (...) Le poète s'est retiré d'un seul coup du langage-instrument; il a choisi une fois pour toutes l'attitude poétique qui considère les mots comme des choses et non comme des signes. Car l'ambiguïté du signe implique qu'on puisse à son gré le traverser comme une vitre et poursuivre à travers lui la chose signifiée ou tourner son regard vers sa réalité et le considérer comme objet." Cette définition de l'"art" par l'autotélisme de son matériau marque chez Novalis2 le renversement d'une tradition vieille de deux mille ans, selon laquelle le "moment unique et essentiel" de l'art et de la poésie est Y imitation. Ce n'est pas un hasard si la musique, dont le statut artistique ne pouvait être que marginal chez Aristote, devient chez Novalis, ainsi que dans toute l'esthétique romantique et symboliste postérieure, le modèle de tout art, et ceci précisément à cause de son caractère non mimétique: "Le musicien prend en lui-même l'essence de son art; et le moindre soupçon d'imitation ne peut l'effleurer. La nature visible semble préparer le travail du peintre et être le modèle qu'il ne pourra jamais atteindre; mais en soi, l'art du peintre est aussi indépendant, aussi apriorique que l'art du musicien. Le peintre ne se sert que d'une langue hiéroglyphique plus difficile que celle du musicien." Ce n'est pas un hasard non plus si l'exemple privilégié d'Aristote (et de toute la tradition classique) est la fiction - que ce soit épopée ou tragédie; alors qu'un Mallarmé, tout comme les Formalistes, érigera le poème (non représentatif) en quintessence de la littérature - de ce qu'on appellera à partir de ce moment le langage poétique. L'art n'est pas la représentation d'autre chose mais le jeu de ses propres éléments. L'autotélisme du matériau en art n'est peut-être pas sans rapport avec la "finalité sans fin" kantienne; il est cependant irréductible à la doctrine de l'art pour l'art, même si les deux ont une origine commune: dans le

POÉTIQUE GÉNÉRALE

475

premier cas, il est question de la fonction du l a n g a g e s littérature (ou du son en musique, etc.) ; dans le second, de la fonction de la littérature, ou de l'art, dans la vie sociale. Jakobson aura donc raison de protester contre les accusations injustes: "Ni Tynjanov, ni Mukarovsky, ni Sklovskij, ni moi, nous ne prêchons que l'art se suffit à lui-même; nous montrons au contraire que l'art est une partie de l'édifice social, une composante en corrélation avec les autres..." La fonction sociale de la poésie dont il se préoccupera plus particulièrement (dans "Qu'est-ce que la poésie?" par exemple) est celle-là même que résume le précepte mallarméen: "donner un sens plus pur aux mots de la tribu..." Jakobson dira: "La poésie nous protège contre la rouille qui menace notre formule de l'amour et de la haine, de la révolte et de la réconciliation, de la foi et de la négation. Le nombre des citoyens de la République tchécoslovaque qui ont lu, par exemple, les vers de Nezval, n'est pas très élevé. Dans la mesure où ils les ont lus et acceptés, sans le vouloir, ils vont plaisanter avec un ami, injurier un adversaire, exprimer leur émotion, déclarer et vivre leur amour, parler politique, d'une manière un peu différente..." 3 Le refus radical de la théorie de l'imitation a été libérateur chez Novalis ; près de deux cents ans plus tard le besoin de synthèse se fait sentir, entre une conception de la littérature comme "imitation de la nature" et une autre, qui la définit comme le jeu intransitif de ses éléments. Non l'un ou l'autre, mais l'interaction et l'équilibre des deux. Il est instructif d'observer, dans ce sens, l'évolution de la pensée jakobsonienne. En 1919, le refus total de la représentation, de la relation entre les mots et ce qu'ils désignent, est, sinon la norme de toute poésie, tout au moins son idéal. "La poésie est indifférente à l'égard de l'objet de l'énoncé." "Ce que Husserl appelle dinglicher Bezug est absent." En 1921 il consacre une étude intégrale au "réalisme en art", en dénonçant la polysémie du terme, mais sans trancher sur l'existence ou non d'un rapport de représentation. Dix ans plus tard, dissociant le poétique de la poésie, il voit celle-ci comme une "structure complexe" dont l'autotélisme poétique n'est qu'une des composantes. Dans l'étude consacrée à Pasternak, il considère que la "tendance à la suppression des objets" est propre à certaines écoles poétiques seulement, telle le Futurisme russe ("Nous avons constaté dans la poésie de Pasternak et des poètes de sa génération une tendance à porter à un degré extrême l'émancipation du signe par rapport à son objet.") Enfin en 1960 il écrit: "La suprématie de la fonction poétique sur la fonction référentielle n'oblitère pas la référence (la dénotation) mais la rend ambiguë." Sera ainsi parcouru, en quelque quarante ans, le trajet intégral de l'esthétique romantique.

476

TZVETAN TODOROV

Pour dépasser la doctrine de Pautotélisme pur, Jakobson (solidaire en cela des autres Formalistes) indique deux directions principales. La première est l'étude de la motivation : "parfois on appelle réalisme la motivation conséquente, la justification des constructions poétiques"; on étudiera donc, non la "réalité" que désigne la littérature, mais les moyens par lesquels le texte nous donne l'impression de le faire. Sa vraisemblance plutôt que sa vérité. Deuxièmement, on étendra l'analyse des éléments non significatifs (sons, prosodie, formes grammaticales) au sémantique, à la "structure thématique"; ce n'est pas, encore une fois, la "réalité" comme telle qui devient objet de l'analyse, mais son mode de présentation dans le texte. Le modèle inégalé de ce type de travail reste les "Notes marginales sur la prose de poète Pasternak", 4 où, par un véritable tour de force, Jakobson englobe dans la même "figure" non seulement le jeu rhétorique et les configurations sémantiques ou narratives ("Prenant comme point de départ les particularités structurales fondamentales de leur poétique, nous avons tenté d'en déduire la thématique de Pasternak et de Majakovskij"), mais aussi la biographie poétique (par opposition à anecdotique) de l'écrivain: "Prenant ainsi comme point de départ la structure sémantique de la poésie de Majakovskij, nous avons pu en déduire son véritable livret et découvrir le noyau central de la biographie de ce poète." La seule différence entre Novalis (ou Sartre) et Jakobson n'est pas que les premiers définissent la poésie comme pur autotélisme du langage, alors que le second permet d'entrevoir l'interaction de ces deux composantes, imitation et jeu. Il y a plus : le discours poétique ou prophétique de Novalis, le discours pamphlétaire de Sartre sont qualitativement distincts du discours scientifique de Jakobson. Il y a peut-être une grande ressemblance entre les formules des uns et de l'autre, lorsqu'on les extrait de leur contexte ; mais la dissemblance apparaît, tout aussi importante, dès qu'on envisage l'utilisation qui en est faite. Leur sens est proche, leur fonction ne l'est pas. Ce qui intéresse Jakobson n'est pas d'énoncer des révélations ou de dénoncer ses adversaires, mais de poser une base à partir de laquelle sera possible la description, la connaissance des faits littéraires particuliers. Et en cela il n'a pas de précurseur. Dès son premier texte sur la littérature Jakobson écrit: "L'objet de la science de la littérature n'est pas la littérature mais la littérarité... Si les études littéraires veulent devenir science, elles doivent reconnaître le procédé comme leur 'personnage' unique." 5 Et, cinquante ans plus tard, dans le "Postscriptum" à ses Questions de poétique. "La 'littérarité' Qiteraturnost'), autrement dit, la transformation de la parole en une œuvre

POÉTIQUE GÉNÉRALE

477

poétique, et le système de procédés qui effectuent cette transformation, voilà le thème que le linguiste développe dans son analyse des poèmes." 6 L'objet de la science n'est pas, n'a jamais été, un objet réel, pris tel qu'en lui-même; ce ne sont donc pas, dans le cas des études littéraires, les œuvres littéraires elles-mêmes (tout comme les "corps" ne le sont pas, pour la physique - ou pour la chimie, ou pour la géométrie). Cet objet ne peut être que construit: il est fait des catégories abstraites que tel ou tel point de vue permet d'identifier au sein de l'objet réel, et des lois de leur interaction. Le discours scientifique doit rendre compte des faits que l'on observe mais il n'a pas pour but la description des faits en eux-mêmes. L'étude de la littérature, que Jakobson appellera plus tard poétique, aura pour objet non les œuvres mais les "procédés" littéraires. Ce choix fondamental place le discours de Jakobson dans la perspective de la science. Il faut écarter ici deux malentendus fréquents et complémentaires. Le premier est celui que commettent les "techniciens": ils croient que la science commence avec les symboles mathématiques, les vérifications quantitatives et l'austérité du style. Ils ne comprennent pas que ce sont là, dans le meilleur des cas, des instruments de la science; que le discours scientifique n'en a pas besoin pour se constituer: il consiste dans l'adoption d'une certaine attitude à l'égard des faits, non dans l'emploi de telle ou telle technique. Le second est celui des "esthètes": ils crient au sacrilège dès qu'on parle d'abstraction, car on risque ainsi d'oblitérer la singularité précieuse de l'œuvre d'art. Ils oublient que l'individuel est ineffable: on entre dans l'abstraction dès l'instant où l'on accepte de parler. On n'a pas le choix de se servir ou non de catégories abstraites mais de le faire consciemment ou non. L'évocation simultanée, d'une part, de la science, dans la perspective de laquelle se range la poétique, et d'autre part, de la sémantique ("ce sont les problèmes sémantiques qui à tous les niveaux du langage préoccupent à présent le linguiste, et s'il cherche à décrire ce dont le poème est fait, la signification du poème ne présente qu'une partie intégrante de ce tout", écrit Jakobson en 1973), ne pose pas moins un problème, qui mérite qu'on s'y attarde. Contrairement aux autres parties de la linguistique, la sémantique ne possède pas de doctrine universellement admise; on continue, de nos jours, à débattre de sa possibilité même. Les critiques littéraires, dont le témoignage importe dans le présent contexte, se rangeraient plutôt, dans ce débat, parmi les sceptiques. A les entendre, dès qu'on a affaire au sens, il n'y a plus de limite infranchissable entre description et interprétation (donc ici entre science et critique), toute nomination du sens est subjective, ce qui expliquerait l'extraordinaire abondance d'interprétations différentes d'un

478

TZVETAN TODOROV

seul texte, au cours des siècles ou même simplement selon les individus. Les lectures poétiques d'un linguiste permettent-elles de lever ces objections, nous amènent-elles à introduire la certitude scientifique jusque dans les problèmes du sens? Pour mieux situer la position nuancée de Jakobson face à ce problème essentiel, observons de plus près sa pratique d'analyse littéraire. Une section de son étude du sonnet de Dante "Se vedi li occhi miei" 7 est consacrée au niveau sémantique. Quel genre de faits sera évoqué? Il relève quatre termes, pietà, giustizia, paura, vertu, et remarque: "L'angoisse et Yeffroi sont les réponses respectives du poète et de chacun, réponses indissociables des souffrances infligées à la justice-vertu" ; il note aussi que "les références directes et les références déplacées se succèdent selon une alternance régulière", ou qu' "il s'établit un lien étroit entre pietà et vertu". Dans l'analyse d'un sonnet de Du Bellay, Jakobson affirme que "l'actualité de Vadore vient remplacer le cachet potentiel du verbe pouras", ou que "les deux circonstants désignent l'un - au plus hault ciel - la distance maximum et l'autre - en ce monde - la proximité la plus intime dans l'espace". Parlant du Spleen de Baudelaire, il relève que "le sujet, Angoisse, nom abstrait personnifié, par contraste avec les corbillards de la première proposition, appartient à la sphère du spirituel. Or l'action de ce sujet abstrait, ainsi que l'objet direct qu'elle régit, sont en revanche tout à fait concrets", etc. Qu'est-ce qui unit ces divers exemples d'analyse sémantique ? On pourrait, dans un premier temps, distinguer deux séries: appartiennent à la première tous les faits de sémantique syntagmatique, tous les cas où Jakobson indentifie la valeur positionelle, relative, de tel segment linguistique par rapport à tel autre (rapport de parallélisme, de contraste, de gradation, de subordination, etc.). Un second groupe est formé par des faits qui s'établissent non plus in praesentia mais in absentia, dans le cadre d'un paradigme, dont un seul des termes figure dans le poème analysé: on remarquera ainsi que tel nom est abstrait, tel autre concret; que telle strophe participe de la substance, telle autre de l'accident; que l'actuel ici s'oppose au virtuel là, et ainsi de suite. En fait les deux séries de faits, syntagmatiques et paradigmatiques, trouvent leur unité en ce que ce sont toujours des faits relationnels. On ne nommera pas autrement V angoisse, mais on précisera qu'elle s'apparente à Yeffroi et s'articule avec la vertu. On ne dira pas ce que "veut dire" le ciel pour Du Bellay, mais on relèvera qu'il appartient à la classe des objets distants - et par là contraste avec les termes appartenant à la classe opposée des objets proches. On ne parlera pas du sens mais toujours et seulement des sens.

POÉTIQUE GÉNÉRALE

479

Dans le débat, donc, qui réunit et oppose sémanticiens et critiques, l'attitude de Jakobson reviendrait à attribuer à chacun un domaine qui lui est propre, et pour lequel il est le seul qualifié. Au critique affirmant la subjectivité du sens, Jakobson oppose implicitement le fait que le rapport des sens est, lui, identifiable dans et par la langue: les mots se parlent entre eux. Au sémanticien avide de s'emparer de la totalité du domaine du sens, Jakobson oppose cependant, tout aussi implicitement, un autre argument: on ne peut décrire, dans un langage cohérent et incontestable, que des rapports formels (y compris les rapports formels des sens); les contenus sémantiques individuels ne se prêtent pas au métalangage, mais seulement à la paraphrase - laquelle reste affaire du critique. Plus proche de Saussure qu'il ne paraîtrait à première vue, Jakobson réserverait à la linguistique la seule sémantique relationnelle, faites des différences et des identités des termes au sein des syntagmes et des paradigmes, en laissant à l'interprétation (à la critique) le soin de nommer le sens d'une œuvre - pour une époque, pour un milieu, pour une sensibilité donnés. Mais revenons à l'ensemble des "procédés" dont Jakobson a fait l'objet de la poétique. Quels sont-ils ? Leur identification découle de la définition que Jakobson a donnée de la poésie: un langage qui tend à devenir opaque. Ce seront donc tous les moyens mis en œuvre par les poètes qui nous amènent à percevoir le langage en lui-même, et non comme le simple substitut des choses ou des idées. Les figures, les jeux avec le temps et l'espace, le lexique singulier, la construction de la phrase, les épithètes, la dérivation et l'étymologie poétiques, l'euphonie, la synonymie et l'homonymie, la rime, la décomposition du mot... Il est une tendance du langage poétique qui retient particulièrement l'attention de Jakobson : la tendance à la répétition. Car "on ne perçoit la forme d'un mot à moins qu'elle ne se répète dans le système linguistique", écrit Jakobson en 1919. Et en 1960, se demandant "selon quel critère linguistique reconnaît-on empiriquement la fonction poétique", il formule cette réponse: " L a fonction poétique projette le principe d'équivalence de l'axe de la sélection sur l'axe de la combinaison." C'est ce qui explique l'attention particulière qu'il accorde, tout au long de son travail, aux différentes formes de répétition, et plus spécifiquement encore, au parallélisme (lequel inclut aussi bien la ressemblance que la différence). Il aime citer cette phrase de G. M. Hopkins: " L a partie artificielle de la poésie, peut-être serait-il juste de dire toute forme d'artifice, se réduit au principe de parallélisme", et il écrit lui-même: "Sur tous les niveaux du langage l'essence de l'artifice poétique consiste en retours périodiques." La majorité de ses études concrètes porte sur trois de ces niveaux: celui des sons

480

TZVETAN TODOROV

ou des lettres, celui de la prosodie, celui des catégories grammaticales; mais ce choix est dicté par des raisons pratiques plutôt que théoriques. Depuis 1960, Jakobson s'est surtout consacré à l'illustration de ce principe, à l'aide d'analyses concrètes de poèmes, qu'il choisit volontairement dans des langues différentes et à des époques très éloignées l'une de l'autre. Cet échantillonnage universel inclut des textes de Dante et Shakespeare, Puskin et Baudelaire, Mâcha et Norwid, Pessoa et Brecht... Le but de ces analyses est double: théorique, en ce qu'elles tendent à prouver son hypothèse sur le fonctionnement de la poésie (bien que le théorème initial disparaisse presque derrière l'abondance des preuves) ; et historique, en cela qu'elles rendent possible une meilleure compréhension de certains textesclés de la tradition littéraire européenne (ce qui est une contribution marquante à la connaissance de la littérature universelle). Nommer le sens d'une œuvre individuelle, on l'a vu, ne fait pas partie des tâches de la poétique, pas plus que l'explication de son effet esthétique ; mais la description exacte des procédés poétiques mis en œuvre permet d'invalider les interprétations abusives. Il suffit, pour s'en convaincre, de lire la conclusion de sa récente plaquette Shakespeare's Verbal Art in "Th' Expence of Spirit" (écrit en collaboration avec Lawrence G. Jones): 8 une bonne dizaine de lectures antérieures du même sonnet se révèlent inconsistantes, une fois confrontées avec la description rigoureuse des structures verbales propres à ce texte. Valéry disait: "La littérature est, et ne peut pas être autre chose qu'une sorte d'extension et d'application de certaines propriétés du langage", et Paulhan : "Que toute œuvre littéraire soit essentiellement une machine si l'on aime mieux, un monument - de langage, voilà qui saute aux yeux au premier coup." Quelques vingt ou trente ans plus tôt, Jakobson avait déjà voué sa vie à la passion du langage - et donc nécessairement à la littérature. Ceux qui l'accuseront de "formalisme", ou s'empresseront de nous assurer que le formalisme est dépassé, ne se rendent pas compte que leurs accusations reposent sur une dichotomie préalable, qui oppose la "forme" au "fond", ou aux "idées". Le choix de Jakobson - de ne jamais cesser de percevoir le langage, de ne pas le laisser sombrer dans la transparence et dans le naturel, quelles que soient les excuses données, - ce choix a une signification idéologique et philosophique bien plus grave que telle ou telle "idée" qui a cours sur le marché. Cependant le refus de reconnaître l'autonomie du langage, de chercher à connaître les lois qui lui sont propres, participe d'un geste séculaire, constitutif de toute une partie de notre culture - et il faut bien plus d'un Jakobson pour le combattre.

POÉTIQUE GÉNÉRALE

481

Ce n'est pas seulement au niveau de la phrase que l'observation des formes linguistiques sera pertinente pour la connaissance de la littérature; mais aussi à celui du discours. Les types de discours, appelés traditionnellement genres, se forment, selon Jakobson, autour de l'expansion de certaines catégories verbales. Les deux genres littéraires les plus étendus, poésie lyrique et poésie épique (ou à un autre niveau mais de manière parallèle: poésie et prose) ont attiré le plus souvent son attention. " R a menant le problème à une simple formulation grammaticale, on peut dire que la première personne du présent est à la fois le point de départ et le thème conducteur de la poésie lyrique, alors que ce rôle est tenu dans l'épopée par la troisième personne d'un temps du passé", écrit-il en 1934; il précise plus tard : " L a poésie épique, centrée sur la troisième personne, met fortement à contribution la fonction référentielle; la poésie lyrique, orientée vers la première personne, est intimement liée à la fonction émotive; la poésie de la seconde personne est marquée par la fonction conative, et se caractérise comme supplicatoire ou exhortative, selon que la première personne y est subordonnée à la seconde, ou la seconde à la première." Mais c'est le rapprochement entre ces mêmes deux types de discours et deux figures de rhétorique, la métaphore et la métonymie, qui est la plus célèbre tentative de Jakobson d'observer la projection des catégories verbales dans les unités transphrastiques. En 1923 un autre Formaliste, Boris Èjxenbaum, identifiait déjà ainsi les deux grandes écoles poétiques de l'époque, symbolistes et acméistes, dans le livre qu'il consacrait à Anna Axmatova, un des principaux représentants de l'acméisme: "Les symbolistes mettent l'accent précisément sur la métaphore ('en la marquant parmi tous les moyens représentatifs du langage'": Andrej Belyj), comme une manière de rapprocher des séries sémantiques éloignées. Axmatova rejette le principe de l'extension, qui repose sur la puissance associative du mot. Les mots ne se fondent pas les uns dans les autres, mais se touchent, pareils aux morceaux d'une mosaïque. (...) A la place des métaphores apparaissent, dans toute leur variété, les nuances latérales des mots, fondées sur des périphrases et des métonymies." Jakobson généralise cette observation dans son étude sur Pasternak, en l'appliquant aux deux genres fondamentaux, et conclut vingt ans plus tard: " L a métaphore pour la poésie et la métonymie pour la prose constituent la ligne de moindre résistance." Il n'y a pas de frontière nette entre les écrits de Jakobson qui relèvent de la linguistique et ceux qui traitent des problèmes de la poétique; et il ne peut pas y en avoir. Son travail de grammairien peut nous intéresser tout

482

TZVETAN TODOROV

autant que celui qu'il consacre à la prosodie: puisque, précisément, les catégories verbales se projettent dans l'organisation des discours. D'autres ont déjà tenté de poursuivre cette recherche, à partir de sa théorie des types doubles (citation, nom propre, autonymie, embrayeurs) ou des fonctions du langage; d'autres encore, sans doute, sauront trouver un jour dans ses écrits de "linguiste" une source d'inspiration pour la connaissance des discours, poétique et autres. Toutes les catégories discursives viennent de la langue; mais pour les identifier, il faut reconnaître auparavant la pluralité de systèmes qui fonctionnent à l'intérieur de celle-ci. Jakobson n'a jamais cessé de combattre les réductionnistes de tous bords, tous ceux qui veulent ramener le langage à un seul des systèmes se manifestant à travers lui. De même qu'il a fallu un jour reconnaître que l'Europe n'est pas le centre de la Terre, ni la Terre, le centre de l'univers, il faut, dans le même mouvement de différenciation entre soi et les autres, le même combat avec l'égocentrisme infantile, cesser d'identifier le langage avec la partie que nous en connaissons le mieux. Réciproquement, les mêmes figures, les mêmes procédés se retrouvent en dehors du langage: au cinéma, en peinture. Car le langage en lui-même, pas plus que les œuvres littéraires, ne saurait être l'objet immédiat d'une science. "De nombreux traits poétiques relèvent non seulement de la science du langage mais de l'ensemble de la théorie des signes, autrement dit, de la sémiologie (ou sémiotique) générale." Ce sont donc les différents types de procès sémiotique qui constitueront l'objet de chaque discipline, et non les différentes substances. La métaphore et la métonymie se définissent par le rapport (différemment) motivé entre deux sens d'un mot; mais toute image comporte un rapport motivé entre elle-même et ce qu'elle représente; il y a donc lieu d'étudier simultanément tous les rapports motivés de signification; et, en un autre lieu, tous ceux qui sont immotivés. Ainsi le même mouvement qui naguère avait fait passer des études littéraires à la poétique, nous poussera un jour de la poétique à la sémiotique ou à la symbolique. S'il me fallait choisir un fait de la biographie de Jakobson pour en faire le symbole, ce serait celui-ci : un adolescent âgé de dix-huit ans est brûlé par les vers de trois poètes contemporains, de peu ses ainés: Xlebnikov, Majakovskij, Pasternak; il se promet de ne jamais oublier cette expérience. Même si ce fait n'a pas eu lieu, il est nécessaire pour comprendre les lignes maîtresses de la carrière de Jakobson. Il y a tout d'abord comme un pari, d'être linguiste et de vivre en même temps intensément une poésie aussi audacieuse. La solution de tran-

POÉTIQUE GÉNÉRALE

483

quillité aurait été de pratiquer la linguistique mais de ne lire que des énoncés "moyens" ; ou inversement, de se passionner pour la poésie mais de laisser de côté la science du langage. Jakobson n'a voulu renoncer à rien, et il a gagné: sa théorie du langage a ceci d'exceptionnel, qu'elle n'admet pas l'opposition entre norme et exception. Si une théorie linguistique est bonne, elle doit pouvoir rendre compte non seulement, disons, de la prose utilitaire neutre, mais des créations verbales les plus sauvages d'un Xlebnikov, par exemple. C'est en cela que Jakobson linguiste nous est particulièrement précieux. Cette même expérience a été décisive pour sa théorie poétique. N o n seulement il a consacré trois études fondamentales à ces poètes, mais toute sa conception de la poésie repose sur une généralisation de son expérience première. Aurait-il pu avoir la même chance avec Puskin? Non, à moins d'être né cent ans plus tôt: le langage contemporain fait partie de la structure du texte, la poésie se consomme à chaud. Son expérience n'aurait jamais été aussi intense avec des poètes d'un autre âge, et n'aurait donc pas pu déterminer de la même façon sa vision de la poésie en général. Il a su lire Puskin par l'intermédiaire de Majakovskij ; l'inverse aurait donné le résultat médiocre qui nous est familier à tous, depuis nos années d'études universitaires. Morale pour le jeune poéticien : il faut vivre la poésie de son temps. Il y a plus. C'est faire preuve, à la fois, d'une extrême ambition et d'une grande humilité, que de lier sa vie à la connaissance des faits, comme c'est le vœu de tout savant: ici, humilité, parce qu'on ne se propose que de décrire et d'expliquer ce qu'ont fait les autres; ambition, parce que ces autres s'appellent Pasternak, Majakovskij, Xlebnikov. C'est renoncer à la fois à la facilité du discours non référentiel et à l'ennui de la description inutile. En cela encore, Jakobson est gagnant; il nous donne aujourd'hui tout autant à savoir qu'à penser. Ainsi les trois fées-poètes auxquelles il a confié ses destinées l'ont bien recompensé; elles ont exaucé tous les vœux qu'il avait formulés à leur place. Aujourd'hui Jakobson n'est plus seulement le symbole d'un grand savant, linguiste et poéticien, mais d'un certain bonheur, qui est celui des créateurs; "amoureux fervent" plus encore que "savant austère"... Paris NOTES 1

"Co je poesie?" Volné smëry, 30 (Prague, 1934), 229-39; "Qu'est-ce que la poésie?" Poétique, II, No. 7 (1971), 247-307, et dans Roman Jakobson, Questions de poétique (Paris: Éditions du Seuil, 1973), 113-26.

484 2

TZVETAN TODOROV

Dans ses premiers écrits, Jakobson étend également cette définition à tous les arts: "Si les arts plastiques sont une mise en forme du matériau visuel à valeur autonome, si la musique est la mise en forme du matériau sonore à valeur autonome, et la chorégraphie, du matériau gestuel à valeur autonome, alors la poésie est la mise en forme du mot à valeur autonome..." Novejsaja russkaja poèzija (Prague, 1921); "La nouvelle poésie russe", Questions de poétique (Paris, 1973), 15. 3 Développement parallèle chez Eliot, dix ans plus tard: "Le devoir direct du poète, en tant que poète, est envers sa langue, qu'il doit d'abord sauvegarder, puis enrichir et améliorer... A la longue, l'influence de la poésie joue sur la langue, la sensibilité, la vie de tous les membres d'une société, de tous les membres de la communauté, et du peuple tout entier, que les gens lisent et goûtent la poésie ou non", etc. Cf. T. S. Eliot, "The Social Function of Poetry", The Adelphi, XXI (1945), 152-61, et dans Critiques and Essays in Criticism, ed. Robert W. Stallman (New York: Ronald Press, 1949), 105-16. 4 "Randbemerkungen zur Prosa des Dichters Pasternak", Slavische Rundschau, VII (1935), 357-74; "Marginal Notes on the Prose of Pasternak", Pasternak, ed. D. Davie and A. Livingstone (Glasgow, 1969), 135-51 ; "Notes marginales sur la prose du poète Pasternak", Poétique, II, No. 7 (1971), 308-21, et dans Questions de poétique (Paris, 1973), 127-44. 5 Novejsaja russkaja poèzija (Prague, 1921); "La nouvelle poésie russe", Questions de poétique (Paris, 1973), p. 15. 6 Questions de poétique (Paris, 1973), 486. 7 (avec P. Valesio) " Vocabulorum constructio in Dante's Sonnet 'Se vedi li occhi miei' ", Studi Danteschi, XLIII (Florence, 1966), 7-33; Questions de poétique (Paris, 1973), 299-318. 8 (Mouton: The Hague, 1970); "L'art verbal dans 'Th' expence of spirit' de Shakespeare", Questions de poétique (Paris, 1973), 356-77.

ROMAN JAKOBSON A N D DUTCH LINGUISTICS E. M. UHLENBECK

I To describe the influence of a prominent scholar upon his environment is far from a simple matter. This is true in particular in the case of Roman Jakobson, who personally and by his many writings has reached a multitude of colleagues and students in different parts of the world during the sixty years of his scholarly activity. Like linguistic borrowing, influence is a process with historical and psychological aspects, and in order to understand it one needs to study not only the influence itself, that is, in the first place the ideas which emanate from the scholar in question, but also those who are affected by that influence. This requires extensive knowledge of the discipline involved in its historical development and an understanding of the leading scholars who determined its course. The present article has a limited theme: Jakobson's impact upon the development of Dutch linguistics. It does not take long to discover that we do not possess an overall picture of the state of affairs in that field at the time when Jakobson's influence began to make itself felt in 1928. For the period after that date we are somewhat better informed. There are personal memories, and there is still the possibility of collecting oral information from contemporaries of Jakobson. Moreover, a large part of the more recent period falls within one's own period of scholarly life, so that one has become familiar with many of the relevant writings. However, even for that period, one is confronted, to one's surprise, with uncertainties and doubts about seemingly simple matters. To gauge the influence of one scholar upon another turns out to be a subtle task which only rarely can be fulfilled to one's own satisfaction. And at every step one becomes painfully aware that a linguist is not automatically also a historian or a psychologist. II

Rereading the literature of the period one gets the impression - apparently shared by those who lived in it - that in the first quarter of the present cen-

486

E. M. UHLENBECK

tury linguistics in Holland was prospering. As in other smaller European countries, there were attempts to go beyond the still reigning Neogrammarian framework. Dialect geography, which stressed the importance of the so far neglected factor of space, developed in this period in Holland into a mature and respected branch of linguistics. J. van Ginneken, J. Schrijnen and G. G. Kloeke in the northern part, L. Grootaers and E. Blancquaert in the southern part of the Dutch language community were the most prominent scholars in this new field. Kloeke's 1927 monograph on the cultural expansion of the province of Holland in the sixteenth and seventeenth century and its reflection on the Dutch dialects became an internationally known classic.1 At the turn of the century, G. C. Uhlenbeck, steeped in the Neogrammarian doctrine like no other Dutch linguist of his generation, and the most influential and respected linguist in the country during this period, started to move away from the classical type of comparative work and tried to get a deeper, more realistic view of prehistorical linguistic realities by drawing upon non-Indo-European languages: first Basque and Eskimo, later the Algonkin languages. Gradually he completely immersed himself in the new world of American Indian languages, b u t - i n spite of his descriptive studies based on fieldwork - never lost his predominantly historical orientation. Van Ginneken's famous Principes de linguistique psychologique2 may be viewed as another attempt to break away from the Neogrammarian position and to establish general laws of language change with the help of the psychology of Pierre Janet. Experimental phonetics, introduced by P. J. Rousselot, E. W. Scripture, and others, began to attract the attention also of Dutch scholars. A series of descriptive phonetic studies published in the first decade of the present century marked the beginning of this new interest, which culminated in the solid Leerboek der Phonetiek [Handbook of Phonetics] by H. Zwaardemaker and L. Eykman, 3 a book which served later generations of Dutch linguists as a reliable guide to Dutch phonetics. Certainly no less important were the descriptive efforts. It is in this period that Holland developed into a prominent center for the description of modern English. The great names are H. Poutsma and E. Kruisinga. Both worked outside the Dutch universities, and it is for that reason, and also because descriptive work still lacked the prestige of historical studies, that the high quality of what they accomplished remained insufficiently appreciated, nationally and internationally. Between 1904 and 1926 Poutsma, a teacher at a secondary school in Amsterdam and self-taught as

DUTCH LINGUISTICS

487

a linguist, produced the 3,000 pages of his Grammar of Late Modern English,4 a mine of factual information which, in spite of the intermingling of data from different periods of the history of English, has retained much of its value to the present day. Kruisinga was many years ahead of his time. He applied basic Saussurian insights in his descriptive work, such as the distinction between the synchronic and the diachronic point of view, years before the Cours de linguistique générale appeared in print. 5 Between 1909 and 1911 Kruisinga's grammar of present-day English came out. 6 Subsequent editions of the three volumes, appearing at regular intervals, incorporated the newest results of Kruisinga's impressive descriptive efforts and showed the unceasing struggle of the author to describe the structure of modern English in its own terms, free from the distortions due to the adoption of a traditional grammatical framework. There are many signs that at least the leading Dutch linguists of this period were aware that the supremacy of the Neogrammarians was coming to an end, and that linguistics was changing. Van Ginneken had made it clear in a paper delivered at the Groningen congress of philologists in 1913 that linguistics was not just a historical discipline and that psychology, sociology, and aesthetics also had a role to play. 7 There was not only a cry for a broader view of linguistics but also a strong emphasis on the study of the "living" language, that is, on the language as spoken at present in all its psychological and sociological complexities. However, a unifying theory had not yet developed. In his inaugural address of 1919, B. Faddegon probably expressed a common opinion when he stated that four general language sciences had to be distinguished : the physiology of language, the psychology of language, ethnolinguistics, and genetic linguistics, which had the task of studying language, respectively, in relation to the human body, in its true essence, in its relation to peoples and races, and in its historical development. General linguistics did not yet exist. The term is mentioned in Schrijnen's manual, 8 but left without any definition or discussion. In his lecture to the Royal Academy on "The Causes of Language Change," van Ginneken's conception of general linguistics was still largely historically colored, and in the animated discussion which followed his paper, this conception was apparently not challenged. 9 The one who came closest to our present conception of general linguistics was H. J. Pos. Appointed in 1924 to the chair of general linguistics and classical philology expressly created for him at the Free University of Amsterdam, he delivered an inaugural lecture which even today is still worth reading. This is especially true of the passage in which he touches upon the simultaneous presence of a universal and an idiosyncratic aspect in each language. 10

488

E. M. UHLENBECK

In view of the development of linguistics after 1928 the question may be asked whether the work of the two main forerunners of phonology and more generally of structural linguistics, namely that of Ferdinand de Saussure and Edward Sapir, met with much response in Holland. It is not easy to give a clear answer to this question without a detailed historical investigation. De Saussure's Cours de linguistique générale and Sapir's Language11 are titles found in the bibliographies appended to several publications of that period. Pos, in his admirable and original introduction to linguistics written in the form of a dialogue, has both titles in his short, two-page bibliography. 12 Schrijnen, author of a more pedestrian introductory book on comparative linguistics, dutifully records the "division" of linguistics into a "synchronistic" part and a "diachronistic" part "to which" - as he remarks in a non-committal way - "de Saussure attached much value," without intimating that here a distinction is made that is of more than passing interest for linguistic theory. 13 Even earlier van Ginneken had occasion to make a comparison between the course on general linguistics given by C. C. Uhlenbeck around 1906 with the one given by de Saussure "which had only recently become available to me." 14 His opinion was that Uhlenbeck's course had to be strongly preferred, because of its factual richness and its world-wide horizons, even though de Saussure surpassed him "in the acuity of his systematic exclusions." 15 However, at least one Dutch scholar reacted quite differently. The Romanist C. de Boer, being fully conversant with the publications of the Geneva school, was convinced of their importance, and, as has been pointed out in his obituary, 16 they were a kind of revelation for him. This is all the more remarkable since between de Saussure's Cours and the work of his two illustrious students quite important differences of opinion existed. De Boer, however, was a rather isolated scholar, with few outside contacts, and although he repeatedly mentions de Saussure in his inaugural lecture of 1917, he made no attempt to propagate his newly acquired theoretical insights among his colleagues. As for Sapir, it is certain that his Language of 1921 did not remain unnoticed. Kruisinga praised his book, "probably unknown to most students of English," in a review in English Studies, concluding that "serious students of English cannot afford to ignore Mr. Sapir's book." 1 7 Kruisinga confessed, however, that he experienced difficulties in grasping Sapir's notion of "an inner sound system that characterizes each language." 18 Therefore, in his announcement of the newly founded journal Language, in the next volume of English Studies, he especially welcomed Sapir's famous paper on sound patterns as "the very article we had asked

DUTCH LINGUISTICS

489

for: a fuller explanation of what he means by sound-patterns." 19 Also A. W. de Groot was probably well acquainted with Sapir's article in Language, as seems to be proven by the stenographic notes made by him in the margins of this issue of the journal, to which he was one of the very few Dutch subscribers. In the absence of further evidence, the conclusion seems justified that in the late 1920's de Saussure and Sapir were fairly widely known in Holland, but that only very few - even of the leading linguists - suspected that one day these two scholars would belong to that small and illustrious group which is recognized as having put their stamp on the development of their science in the present century. Such, then, seems to have been in rough outline the Dutch linguistic soil on which Jakobson's phonological seeds fell, when in April 1928 he arrived in Holland for the first time to attend the first postwar meeting of linguists convened by a committee of Dutch linguists, headed by C. C. Uhlenbeck and J. Schrijnen, who felt that the wounds of World War I were sufficiently healed to permit an international exchange of ideas. It is this congress which marks the beginning of Jakobson's influence on Dutch linguistic thinking. In the history of Dutch linguistics it was a true watershed.

Ill The First Congress of Linguists in The Hague was largely a European affair: with the exception of eight members from the United States (among them, however, F. Boas, L. Bloomfield, and E. Sapir) all came from European countries. Of the 300 scholars present, more than one third came from the host country. Nearly all occupants of Dutch university chairs in the study of language and literature attended the congress and also many of those who were to be their successors. For the congress Jakobson had drawn up the famous five theses, upon his request 20 also signed by N. S. Trubetzkoy and S. Karcevskij, which were put on the program for discussion as Proposition 22. 21 Contrary to Jakobson's own expectations, 22 they were received by many with warm approval and by all with interest and sympathy. At the second congress in 1931 the same attitude prevailed. 23 The main notions of the new phonology seem to have been understood and assimilated by Dutch linguists with remarkable ease. 24 Seven linguists are in particular responsible for the development of Dutch phonological studies in the interwar period. Four of them belonged to the older generation; three were contemporaries of Jakobson.

490

E. M. UHLENBECK

N. van Wijk, van Ginneken, Kruisinga, and Faddegon were all born in or before 1880. From the point of view of phonology van Wijk was by far the most influential. From 1928 until his death in 1941, he published a series of articles on various aspects of phonology, showing his by no means uncritical but very positive attitude towards the new approach. 25 Several of them appeared in periodicals such as De Nieuwe Taalgids and Levende Talen, which were read by language teachers, and in this way the phonemic principle and the notion of sound system became gradually known among a fairly extensive group of people with a university education in language and literature, although the fact that the Prague school concentrated on languages little known in Holland was felt by some to be a barrier. 26 Moreover, van Wijk stressed in several of his publications that the notions underlying phonology could be applied fruitfully to the study of morphology and syntax.27 His Phonology, with the apt subtitle: A Chapter of Structural Linguistics, was certainly his crowning achievement within this field of his scientific endeavor.28 It was favorably received by C. C. Uhlenbeck29 and by Kruisinga, 30 and it became the handbook through which most Dutch linguists of later generations acquired their first knowledge of phonemics. It was even circulated in mimeographed form among linguists in the Dutch East Indies in 1940 when Dutch scholars working in that part of the world were cut off from their occupied home country and could no longer obtain Dutch books. It is in this manner, through a hardly readable copy, that the writer of this article learned about the activity of the Prague school and the role played by Jakobson and his associates at the Hague Congress. Like van Wijk, van Ginneken immediately realized the importance of what Jakobson, Trubetzkoy, and V. Mathesius had to say31 and tried in his usual exuberant way to incorporate their findings in his grand syntheses, which, however, often failed to impress the more sober-minded. In 1930 van Ginneken took part with de Groot in the "Réunion phonologique internationale" held in Prague. With van Wijk he founded the Dutch Phonological Society and tried his h a n d - w i t h questionable results32 - at a phonological description of Dutch. His articles on Dutch phonology which appeared within a period of ten years (1931-1941), mostly in the periodical Onze Taaltuin, of which he was the founder, proved both his keen interest in the further development of the new approach and his inability to understand that his very personal, at times even fantastic, biological notions of language could hardly be reconciled with the principles of Prague phonology. Within the framework of this article there is no need to say more than a

DUTCH LINGUISTICS

491

few words about Faddegon and Kruisinga. The former scholar wrote on phonology only once, in a remarkable article.33 It was a sequel to a paper published more than thirty years earlier,34 in which he set out - partly with the help of certain psychological experiments, partly on the basis of assonances from nursery rhymes - to prove that there existed "subconscious associations between certain speech sounds, some by kinesthetic, most others by auditive resemblances."35 This view led him to challenge Trubetzkoy's opinion that phonetics is a physical science, but phonology "a mental science teleological in character." 36 As may be concluded from his positive review of van Wijk's Phonology, Kruisinga was certainly not unsympathetic towards phonology, but his main interest lay in English morphology and syntax, and he rarely discussed phonological issues in his journal Taal en Leven or elsewhere, although there is sufficient reason to assume that he kept himself informed about further developments. Next to van Wijk there were three Dutch linguists, namely de Groot, Pos and A. Reichling, all born in the 1890's and thus belonging to the same generation as Jakobson himself, who were not only deeply influenced by the work of the Prague phonologists but also contributed materially to the further development of phonology, be it in quite different but compatible ways. They occupied chairs at the Municipal University of Amsterdam. 37 In many respects their linguistic views were similar, while being distinct from those of other centers, so that people sometimes referred to them collectively as "the Amsterdam school." The course of intellectual development of Jakobson and de Groot shows a striking similarity in spite of the fact that they were emotionally quite different personalities. Both Jakobson and de Groot were at first interested in poetics and metrics and both began their scholarly career in these fields, before turning their main attention to linguistics proper, although also in their later years they kept working in the field of initial preference. In view of this parallel and their common view that the study of language included the study of its poetic function, 38 it is not too surprising that in many publications of de Groot one finds the author agreeing or disagreeing (but in any case reacting) to opinions expressed by Jakobson. De Groot did not write any reviews after 1935, but made an exception for the Preliminaries to Speech Analysis.39 In his Algemene Versleer of 194640 de Groot paid extensive attention to Jakobson's paper given at the Congress of Phonetic Sciences in Amsterdam on the verse structure of Serbo-Croatian folkepics41 and several more recent publications by him on related topics. The Hague Congress seems to have had a great impact upon de Groot. One finds this clearly reflected in the topics of his publications after 1928. From

492

E. M. UHLENBECK

that date onward his linguistic attention switched from Latin to presentday Dutch. It is true that he kept an interest in the case system of Latin the study of case systems was another point of contact with Jakobson and in Latin metrics, two areas in which he continued publishing. His main field, however, was general linguistics and the descriptive study of Dutch. The two were for him intimately connected. The introduction to general linguistics which he wrote towards the end of his life bears the supplementary title Also Dutch.42

an Introduction

to the Grammar

of

Present-day

Pos, from the 1930's until his premature death in 1955, generally recognized as one of the most - or, in the eyes of many, the most - prominent philosopher of the country, had a quite different influence on Dutch phonology than de Groot. Pos never did any descriptive work, but he was deeply interested in the wider philosophical and general linguistic implications of the new theory. Two of his linguistic articles are particularly influential. Among Dutch linguists of the younger generation (including the writer of the present article), his academy lecture of 1938 was found highly illuminating.43 His perspicacious discussion of the relation between the background of the older experimental phonetics and the new phonology contributed more to a clear understanding of the shifting of viewpoints brought about by Jakobson cum suis than any other article published in the interwar period. Moreover, it made abundantly clear that phonology set an example, albeit for a limited part of language, as to the direction which the study of grammar, and also the study of semantics, had to take. In this respect his article "Perspectives du structuralisme"44 was also a source of inspiration for linguists outside Holland, including Jakobson himself, as we know from a personal communication. As Pos wrote in this stimulating and far-ranging article: "En effet, le mouvement phonologique paraît appelé à fructifier les autres domaines de la linguistique où les présuppositions du 19me siècle sont encore en vigueur." 45 Reichling's contribution to Dutch linguistics and to linguistics in general was again quite different from those made by de Groot and Pos. Like Pos, Reichling was never personally involved in descriptive linguistics. Unlike Pos he was a general linguist and not at all a philosopher of language. Reichling became the main theoretician of modern Dutch linguistics. His not easily accessible monograph of 1935, which is still insufficiently known internationally, was one of the earliest attempts to develop a full theory of language of which word and sentence were the pivotal theoretical notions. 46 It is within this framework that Reichling, with full recognition of the importance of Sapir's Language article and the work of the Prague school,

DUTCH LINGUISTICS

493

was able to assign to the phoneme its essentially correct position as a potentially distinctive, subordinate element of the word-form. 47

IV

In the preceding section the influence of Jakobson on Dutch linguistics was largely equated with the birth and development of Dutch phonology under the impetus of the Hague Congress. The first - and important influence was easy to trace, since there is ample internal evidence that those scholars who in the 1930's became most active in this new field had studied the Prague Travaux and Jakobson's contributions to them. Because of Jakobson's prominent role at the Hague Congress and through the writings of Dutch phonologists in which his papers were quoted and discussed, his name became well known in Holland. Moreover, Jakobson visited Holland again in 1932 on the occasion of the First Congress of Phonetic Sciences, held in Amsterdam, and after World War II he lectured in Holland at least twice. By correspondence and exchange of reprints, Jakobson kept in regular contact with Dutch linguists and closely followed their development. It goes without saying that Jakobson's writings cover more than just phonology and that in phonology his views kept developing. Certainly for a country like Holland, where the study of language is normally combined with the study of literature and philology, it was a most attractive aspect of Jakobson's personality that he was not a narrow-minded specialist but conceived the study of language in the broadest terms. It is perhaps for this reason that Jakobson's activities at Harvard University and the Massachusetts Institute of Technology were closely followed by those Dutch linguists who were directly or indirectly influenced by "the Amsterdam school." In this fourth and final section I will try to show which of the quite different aspects of Jakobson's activities had a demonstrable effect on Dutch scholarship. For various reasons this account cannot be exhaustive, but it is hoped that the main lines of influence are correctly indicated. Jakobson's early publications on poetics became known in Holland through his above-mentioned paper on Serbo-Croatian folk epics.48 It drew the attention of de Groot - understandably enough, given his interest in Latin and Greek verse. Apart from earlier papers, de Groot's indebtedness to Jakobson in this field comes out most clearly in his Algemene versleer. It is via this book that M. Rutten 49 became acquainted with the

494

E. M. UHLENBECK

work of Jakobson, J. Mukarovsky, and those other members of the Prague school with an interest in poetics. Via de Groot and Rutten, Jakobson's congress paper of 1932 also may have influenced the inaugural lecture of A. Teeuw.50 Knowledge of Jakobson's early literary work remained for a long time limited to those few scholars who could read Russian and Czech. It is true that C. L. Ebeling's Dutch inaugural lecture devoted to the Formalists had some influence,51 but the book that did much to make this aspect of Jakobson's activity better known was Victor Erlich's excellent monograph. 52 Two other early articles by Jakobson, both in the field of historical linguistics, have to be mentioned at this point, as they called forth immediate reactions from van Wijk, whose main field of interest was still historical and comparative linguistics. Van Wijk's paper called "'Taalbond' en 'Taalfamilie'" 53 was prompted by Jakobson's monograph of the same year,54 in which he had tried to give a characterization of the Eurasian "linguistic alliance" ( = Trubetzkoy's Sprachbund). As far as we know, this topic has not been taken up again by Dutch linguists. Jakobson's challenging "Principes de phonologie historique," 55 however, had a more durable influence. It evoked a critical appraisal by van Wijk, 56 and also in more recent publications the impact of this study is demonstrable.57 As could be expected, Jakobson's gradual development of his ideas concerning distinctive feature analysis was closely followed by several Dutch linguists. It was again van Wijk 58 who was the first to realize the importance of Jakobson's congress paper entitled "Observations sur le classement phonologique des consonnes," 59 which contains - in nuce several important principles developed later in the Preliminaries of 1952.60 The influence of this paper is also clearly visible in de Groot's article on "Voyelle, consonne et syllabe,"61 in my analysis of the Javanese phonemic system,62 and in A. H. Kuipers' description of the phonology of Squamish.63 »

After phonology we must turn our attention to morphology, because this is the area-jn which Jakobson's influence on Dutch linguistics perhaps has come out even stronger than in phonology. Two of his older articles on morphology, namely his article on the structure of the Russian verb64 and his extensive study of case systems, have had a strong impact on several Dutch scholars engaged in descriptive work in quite different fields. First of all these two articles - which because of their general importance have been reprinted more than thirty years after their initial appearance -have been studied by de Groot, who has had a lifelong interest in case systems. In particular de Groot's contribution to the Mélanges Bally has to be

DUTCH LINGUISTICS

495

mentioned, 6 5 followed many years later by his contribution to the first Festschrift for Jakobson 6 6 and his detailed description of the Latin genitive. 67 De Groot's appreciation for Jakobson's study of 1936 is expressed by him in the following words : "Without agreeing with Jakobson and Kurylowicz 68 on all points, I still believe that - with a few modifications - the distinctions made by them are of fundamental importance, and among the most valuable tools of structural linguistics." 69 Next to de Groot four Dutch Slavists, all of whom were for some time students of Jakobson, have shown their indebtedness to Jakobson's morphological studies, namely C. H. van Schooneveld, 70 A. G. F. van Hoik, 7 1 and C. L. Ebeling 72 in their articles on Russian morphology, Ebeling in his study of the Dutch tenses, 73 and A. H. Kuipers in his description of the nominal system of Circassian. 74 In my descriptive studies on the morphology of Javanese, 75 the same two articles have been instrumental in shaping my theoretical views. Especially the distinction between marked and unmarked has proved to be indispensable for gaining an insight into the structure of the Javanese verb system. In addition, Jakobson's short but fundamental paper on deixis and the verbal categories in Russian 7 6 has exerted a beneficial influence on my study of the Javanese pronominal system. 77 Finally a few words have to be said about two other important publications of Jakobson's, namely his "Kindersprache, Aphasie und allgemeine Lautgesetze" 78 and "Linguistics and Poetics," his brilliant contribution to the volume Style in Language.79 As to the first, one can be certain that it has been read and admired by Dutch linguists, including the writer of the present article. In his study of the use of visual language by the deaf, B. Tervoort recognized its fundamental importance, 8 0 while W. Kaper showed his indebtedness in many pages of his careful report on certain stages in the process of language acquisition of his two sons. 81 As to "Linguistics and Poetics," it is difficult to point out specific Dutch publications which have been influenced by it, but it is certain that Jakobson's exposé of the six main functions of language found in it has become widely known. In many places where linguistics is being taught it has been integrated into the course on general linguistics, as have distinctive feature analysis and other products of Jakobson's creative personality. Leiden University BIBLIOGRAPHY Actes du Premier congrès de linguistes tenu à La Haye, du 10-15 Avril 1928 (Leiden: Sijthoff, 1928).

496

E. M. UHLENBECK

Actes du Deuxième congrès international de linguistes, Genève, 25-29 Août 1931 (Paris: Maisonneuve, 1931). Blancquaert, E., and H. Vangassen, Dialect-atlas van Zuid-Oost-Vlaanderen (Antwerp: De Sikkel, 1931). Blancquaert, E., Dialect-atlas van Noord-Oost-Vlaanderen en Zeeuwsch-Vtaanderen (Antwerp: De Sikkel, 1935^11). de Boer, C., Autour d'un mot (Amsterdam: Portielje, 1917). Bühler, K., "Phonetik und Phonologie," Travaux du Cercle Linguistique de Prague 4 (1931), 22-53. Dresden, S., "Herdenking van Cornelis de Boer (26 mei 1880-7 augustus 1957)," Jaarboek der Koninklijke Nederlandse Akademie van Wetenschappen 1957-58 (Amsterdam, 1958), 239^16. Ebeling, C. L., "On the Semantic Structure of the Russian Sentence," Lingua 4 (1954), 270-22. —, Taal- en Letterkunde: Aspecten van het Russische Formalisme (The Hague: Mouton, 1955). —, Linguistic Units ( = Janua Linguarum 12) (The Hague: Mouton, 1960). —, "A Semantic Analysis of the Dutch Tenses," Lingua 11 (1962), 86-99. Erades, P. A., "In Memoriam Etsko Kruisinga (December 8th 1875-February 15th 1944)," Lingua 1 (1948), 140-48. Erlich, V., Russian Formalism: History — Doctrine (The Hague: Mouton, 1955). —, "Roman Jakobson," Orbis 7 (1958), 287-90. Faddegon, B., "Geleidelijke en Springende Klankverandering," Tijdschrift voor Nederlandse Taal- en Letterkunde 26 (1907), 204-23. •—, Het sprookje van den verliefden grammaticus (Amsterdam: Muller, 1919). —, "Phonetics and Phonology," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, Nieuwe Reeks, Deel 1, No. 10 (1938). van Ginneken, J., Principes de linguistique psychologique (Paris: M. Rivière, 1907). —•, Als ons moedertaalonderwijs nog ooit gezond wil worden (Nijmegen : Malmberg, 1917). —, De nieuwe richting in de taalwetenschap (Utrècht-Nijmegen : Dekker/van Leeuwen, 1923). —, "De oorzaken der taalveranderingen," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde 59, Serie A2 (1925). —, Grondbeginselen van de schrijfwijze der Nederlandsche taal (Hilversum : Brand, 1931). —, "Presidential Address," Proceedings of the International Congress of Phonetic Sciences, Amsterdam, 1932 (Amsterdam, 1933), p. 15. de Groot, A. W., "Het eerste wereldcongres van linguïsten," Vragen des Tijds 55, Deel 1 (1929), 423-41. —, "De wetten der phonologie en hun betekenis voor de studie van het Nederlands," De Nieuwe Taalgids 25 (1930), 2 2 5 ^ 3 . •—, "De Phonologie van het Nederlands," De Nieuwe Taalgids 26 (1932), 10-19. —, "Les oppositions dans les systèmes de la syntaxe et des cas," Mélanges Bally (Geneva, 1939), 107-27. —, "Voyelle, consonne et syllabe," Archives Néerlandaises de Phonétique expérimentale 17 (1941), 21-41. —, "Woord of Woordgroep," De Nieuwe Taalgids 38 (1944), 1-6. —, Algemene Versleer (The Hague: Servire, 1946). —, Structurele Syntaxis (The Hague: Servire, 1949). —, Review of Roman Jakobson, M. Halle, and C. G. M. Fant, Preliminaries to Speech Analysis, Word 9 (1953), 58-64. —, "Classification of Cases and Uses of Cases," For Roman Jakobson (The Hague: Mouton, 1956), 187-94.

DUTCH LINGUISTICS

497

—, "Classification of the Uses of a Case Illustrated on the Genetive in Latin," Lingua 6 (1956), 8-66. —, lnleiding tot de Algemene Taalwetenschap, tevens inleiding tot de grammatica van het hedendaagse Nederlartds (Groningen: Wolters, 1962). Grootaers, L., and G. G. Kloeke, Handleiding bij het Noord- en Zuid-Nederlandsch dialectenonderzoek ( = Noord- en Zuid-Nederlandsche Dialectbibliotheek, 1) (The Hague: Nijhoff, 1926). van Haeringen, C. B., "Herdenking van J. J. A. van Ginneken," Jaarboek der Koninklijke Nederlandsche Akademie van Wetenschappen 1945-46 (1946), 221-30. van Hamel, A. G., Geschiedenis der Taalwetenschap (The Hague: Servire, 1945). van Hoik, A. G. F., The Semantic Spectrum of the Russian Infinitive (Leiden: Sijthoff, 1953). Jakobson, R., "Principes de phonologie historique," Travaux du Cercle Linguistique de Prague 4 (1931), 247-67. —, K xarakteristike evrazijskogo jazykovogo sojuza (Paris: Izdanie Evrazijcev, 1931). —, " Z u r Struktur der russischen Verbums," Charisteria Gvilelmo Mathesio Quinquagenario...oblata (Prague, 1932), 74-84. —, " Ü b e r den Versbau der serbokroatischen Volksepen," Proceedings of the International Congress of Phonetic Sciences (1933), 44-53. —, "Beitrag zur allgemeinen Kasuslehre: Gesamtbedeutungen der russischen Kasus," Travaux du Cercle Linguistique de Prague 6 (1936), 240-88. —, "Observations sur le classement phonologique des consonnes," Proceedings of the Third International Congress of Phonetic Sciences, Ghent, 1938 (Ghent, 1939), 34-41. —, "Nikolaj Sergejevic Trubetzkoy," Acta Linguistica 1 (1939), 64-76. —, "Kindersprache, Aphasie und allgemeine Lautgesetze," SpräkvetenskapUga Sällskapets i Uppsala Förhandlingar, 1940-42 (Uppsala, 1942). —, M. Halle, and C. G. M. Fant, Preliminaries to Speech Analysis ( = Acoustics Laboratory, Massachusetts Institute of Technology, Technical Report 13) (Cambridge: M.I.T., 1952). —, et al., "Results of the Conference of Anthropologists and Linguists," International Journal of American Linguistics, Memoir 8, XIX, No. 2 (April, 1953), 11-21. —, "Shifters, Verbal Categories and the Russian Verb," Russian Language Project, Harvard University (1957). —, "Linguistics and Poetics" in Style in Language, ed. Thomas A. Sebeok (Cambridge, Mass.: Technology Press of M.I.T., 1960), pp. 350-77. de Josselin de Jong, J. P. B., "In Memoriam Christianus Cornelius Uhlenbeck (18th October 1866 - 12th August 1951)," Lingua 3 (1953), 243-68. Kaper, W., Einige Erscheinungen der kindlichen Spracherwerbung erläutert im Lichte des vom Kinde gezeigten Interesses für Sprachliches (Groningen: Wolters, 1959). Kloeke, G. G., De Hollandsche Expansie in de 16de en I7de eeuw en haar weerspiegeling in de hedendaagsche Nederlandsche Dialecten ( = Noord- en Zuid-Nederlandsche Dialectbibliotheek, 2) (The Hague: Nijhoff, 1927). Kruisinga, E., A Handbook of Present-Day English, 4 vols. (Utrecht: Kemink en Zoon, 1909-11). —, Review of Language. An Introduction to the Study of Speech, English Studies 7 (1925), 177-79. —, "American Linguistics," English Studies 8 (1926), 18. —, Review of N. van Wijk, Phonologie, een Hoofdstuk uit de structured Taalwetenschap, Taal en Leven 3 (1939), 77-79. Kuipers, A. H., "The Circassian Nominal Paradigm: A Contribution to Case Theory," Lingua 11 (1962), 231-48.

498

E. M. UHLENBECK

—, The Squamish Language, 2 vols. ( = Janua Linguarum, series practica, 73/1-2) (The Hague: Mouton, 1967-69). Kurylowicz, J., "Le problème du classement des cas," Bulletin de la Société Polonaise de Linguistique 9 (1949), 20-43. Mukarovsky, J., "La phonologie et la poétique," Travaux du Cercle Linguistique de Prague 4 (1931), 278-88. Nuytens, E. T. H., De tweetalige mens (Assen: Van Gorcum, 1962). Polak, M., "Phonetiek en Phonologie," Levende Talen 1940, 111-17. Pos, H. J., Algemene Taalwetenschap en Subjectiviteit (Amsterdam: H. J. Paris, 1924). —, Review of Karl Vossler, Gesammelte Aufsätze zur Sprachphilosophie (München, 1923), Neophilologus 10 (1925), 310-12. —, Inleiding tot de Taalwetenschap (Haarlem: Bohn, 1926). —, "Quelques perspectives philosophiques de la phonologie," Archives Néerlandaises de Phonétique expérimentale 8-9 (1933), 1-4. —, Review of A. H. Gardiner, The Theory of Speech and Language, English Studies 15 (1933), 220-21. —, "Phonologie en betekenisleer," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, Nieuwe Reeks, Deel 1, No. 13 (1938). —, Perspectives du structuralisme," Travaux du Cercle Linguistique de Prague 8 (1939), 71-78. •—-, Keur uit de Versprèide Geschriften van Dr. H. J. Pos, I: Taal, Mens en Cultuur; II: Beginselen en Gestalten (Assen: Van Gorcum, 1957-58). Poutsma, H., A Grammar of Late Modem English (Groningen: P. NoordhofT, 1914-29). Reichling, A. J. B. N., Het Woord (Nijmegen : Berkhout, 1935). Rutten, M., "Dichtkunst en Phonologie," Revue Belge de Philologie et d'Histoire 28 (1950), 871-920. Sapir, E., Language: An Introduction to the Study of Speech (New York: Harcourt, Brace & World, 1921). de Saussure, F., Cours de linguistique générale (Lausanne-Paris: Payot, 1916). van Schooneveld, C. H., "The Aspect System of the Old Church Slavonic and Old Russian Verbum Finitum byti" Word 7 (1951), 96-103. —, Over de woordsoorten in het Moderne Russisch (Leiden: Brill, 1953). —, A Semantic Analysis of the Old Russian Finite Preterite System ( = Slavistic Printings and Reprintings, 7) (The Hague: Mouton, 1959). Schrijnen, J., Handleiding bij de Studie der Vergelijkende Indogermaansche Taalwetenschap, vooral met betrekking tot de Klassieke en Germaansche Talen (Leiden : Sijthoff, 1918, 2nd ed., 1924). Teeuw, A., Taal en Versbouw. (Amsterdam : Djambatan, 1952). Tervoort, B. Th. M., Structurele analyse van visueel taalgebruik binnen een groep dove kinderen (Amsterdam: North Holland Co., 1953). Uhlenbeck, C. C., Review of N. van Wijk, Phonologie, De Nieuwe Taalgids 33 (1939), 274-77. Uhlenbeck, E. M., "De structuur van het Javaanse morpheem," Verhandelingen, Koninklijke Bataviaasch Genootschap van Künsten en Wetenschappen 78 (1949). —, "Verb Structure in Javanese," For Roman Jakobson (The Hague: Mouton, 1956), pp. 567-73. —-, "De systematiek der 'Javaanse pronomina," Verhandelingen van het Koninklijk Instituut voor Taal-; Land- en Volkenkunde 30 (1960). Vachek, J., Dutch Linguists and the Prague Linguistic School (Leiden : Universitaire pers, 1968). de Vooys, C. G. N., "Het onderzoek van de Nederlandse Dialekten," De Nieuwe Taalgids 15 (1921), 1.13-22.

DUTCH LINGUISTICS

499

—, "Herdenking van A. G. van Hamel (5 Juli 1886 - 23 November 1945)," Jaarboek der Koninklijke Nederlandsche Akademie van Wetenschappen 1945-46 (Amsterdam, 1946), 231-37. van Wijk, N., " 'Taalbond' en 'Taalfamilie'," De Nieuwe Taalgids 25 (1931), 284-90. —, "De moderne phonologie en de omlijning van taalkategorieën," De Nieuwe Taalgids 26 (1931), 65-75. —, "Grammatika en woordvorming," De Nieuwe Taalgids 28 (1934), 362-74. —, "Umfang und Aufgabe der diachronischen Phonologie," Mélanges van Ginneken (Rotterdam, 1937), 93-99. —, "Klankhistorie en Phonologie," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, Nieuwe Reeks, Deel 1, No. 3 (1938). —, "Parallélisme tussen 'Phonologie' en 'Grammatika'," De Nieuwe Taalgids 33 (1939), 109-22. —, Phonologie, een Hoofdstuk uit de Structurele Taalwetenschap (The Hague: Nijhoff, 1939; repr., 1965). —, Review of N. S. Trubetzkoy, Grundzüge der Phonologie, Slavia 17 (1939-40), 230-36. —, "Quantiteit en Intonatie," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, Nieuwe Reeks, Deel 3, No. 1 (1940). —, "De betekenis der phonologie," Levende Talen 108 (1940), 13-23. Wils, J., "In Memoriam Jacques van Ginneken (21 April 1877 - 20 October 1945)," Lingua 1 (1948), 133-39. Zandvoort, R. W., De taak der Nederlandse Anglistiek (Amsterdam: Swets en Zeitlinger, 1936). —, Eindrapport (Groningen: Wolters, 1964). Zwaardemaker, H., and L. P. H. Eykman, Leerboek der Phonetiek (Haarlem: Bohn 1928).

NOTES 1 G. G. Kloeke, De Hollandsche Expansie in de 16de en 17de eeuw en haar weerspiegeling in de hedendaagse Nederlandsche Dialecten ( = Noord- en Zuid-Nederlandsche Dialectbibliotheek 2) (The Hague: Nijhoff, 1927). 2 J. van Ginneken, Principes de linguistique psychologique (Paris: M. Rivière, 1907). 3 H. Zwaardemaker and L. P. H. Eykman, Leerboek der Phonetiek (Haarlem: Bohn, 1928). 4 H. Poutsma, A Grammar of Late Modem English (Groningen: P. Noordhoff, 1914-1929). 5 F. de Saussure, Cours de linguistique générale (Lausanne-Paris: Payot, 1916). 6 E. Kruisinga, A Handbook of Present-Day English, 4 vols. (Utrecht: Kemink en Zoon, 1909-1911). 7 J. van Ginneken, De nieuwe richting in de taalwetenschap (Utrecht-Nijmegen: Dekker/van Leeuwen, 1923), pp. 22-24. 8 J. Schrijnen, Handleiding bij de Studie der Vergelijkende Indogermaansche Taalwetenschap, vooral met betrekking tot de Klassieke en Germaansche Talen (Leiden: Sijthoff, 1918; 2nd ed., 1924). 9 J. van Ginneken, "De oorzaken der taalveranderingen," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde 59, serie A2 (1925).

500 10

E. M. UHLENBECK

H. J. Pos, Algemene Taalwetenschap en Subjectiviteit (Amsterdam: Paris, 1924). E. Sapir, Language: An Introduction to the Study of Speech (New York: Harcourt, Brace & World, Inc., 1921). 12 H. J. Pos, Inleiding tot de Taalwetenschap (Haarlem: Bohn, 1926). 13 Schrijnen, Handleiding bij de Studie der Vergelijkende Indogermaansche Taalwetenschap (1924), pp. 42-43. 14 J. van Ginneken, Als ons moedertaalonderwijs nog ooit gezond wil worden (Nijmegen : Malmberg, 1917), p. 12. 15 Ibid., p. 12. 18 S. Dresden, "Herdenking van Cornelis de Boer (26 mei 1880-7 augustus 1957)," Jaarboek der Koninklijke Nederlandse Akademie van Wetenschappen 1957-1958 (Amsterdam, 1958), 239-46. 17 E. Kruisinga, Review of Language: An Introduction to the Study of Speech, English Studies 7 (1925), 177-79. 18 Ibid., p. 178. 19 E. Kruisinga, "American Linguistics," English Studies 8 (1926), 18. 20 See R. Jakobson's obituary of Trubetzkoy in Acta Linguistica 1 (1939), 73 ; also in Selected Writings II (The Hague: Mouton, 1971), 501-16. 21 The 42 propositions were distributed in printed form among all participants before the beginning of the Congress. 22 Personal communication from Jakobson. 23 Cf. the reactions of A. Meillet and O. Jespersen at the Geneva congress. 24 J. Vachek, Dutch Linguists and the Prague Linguistic School (Leiden: Universitaire pers, 1968). 25 N. van Wijk, "De moderne phonologie en de omlijning van taalkategorieën," De Nieuwe Taalgids 26 (1931), 65-75. 29 M. Polak, "Phonetiek en Phonologie," Levende Talen 1940, p. 116. 27 N. van Wijk, "Parallélisme tussen 'Phonologie' en 'Grammatika'," De Nieuwe Taalgids 33 (1939), 109-22. 28 N. van Wijk, Phonologie, een Hoofdstuk uit de Structurele Taalwetenschap (The Hague: Nijhoff, 1939; 2nd ed., 1965). 29 C. C. Uhlenbeck, Review of N. van Wijk, Phonologie, De Nieuwe Taalgids 33(1939), 274-77. 30 E. Kruisinga, Review of N. van Wijk, Phonologie, een Hoofdstuk uit de Structurele Taalwetenschap, Taal en Leven 3 (1939), 77-79. 31 J. van Ginneken, "Presidential Address," Proceedings of the International Congress of Phonetic Sciences (Amsterdam, 1933), p. 15; also in Archives néerlandaises de Phonétique expérimentale VIII-IX (1933), p. 106. 32 A. W. de Groot, "De Phonologie van het Nederlands," De Nieuwe Taalgids 26 (1932), 10-19. 33 B. Faddegon, "Phonetics and Phonology," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, nieuwe reeks, deel 1, No. 10 (1938). 34 B. Faddegon, "Geleidelijke en Springende Klankverandering," Tijdschrift voor Nederlandse Taal- en Letterkunde 26 (1907), 204-23. 35 Faddegon, "Phonetics and Phonology," p. 485. 36 Ibid., p. 486. 37 Pos moved from the Free University of Amsterdam to the Municipal University in 1932. 38 See, for instance, A. W. de Groot, Inleiding tot de Algemene Taalwetenschap, tevens inleiding tot de grammatica van het hedendaagse Nederlands (Groningen : Wolters, 1962), p. 306, and R. Jakobson et al., "Results of the Conference of Anthropologists and 11

DUTCH LINGUISTICS

501

Linguists," International Journal of American Linguistics, Memoir 8, XIX, No. 2 (April, 1953), 11-21 ; also in Selected Writings II (The Hague: Mouton, 1971), 554-67. 39 A. W. de Groot, Review of Roman Jakobson, M. Halle, and C. G. M. Fant, Preliminaries to Speech Analysis, Word 9 (1953), 58-64. 40 A. W. de Groot, Algemene Versleer (The Hague: Servire, 1946). 41 R. Jakobson, "Über den Versbau der serbokroatischen Volksepen," Proceedings of the International Congress of Phonetic Sciences (1933), 44-53; also in Archives néerlandaises de phonétique expérimentale VIII-IX (1933), pp. 135-44, and Selected Writings IV (The Hague: Mouton, 1966), 51-60. 42 De Groot, Inleiding tot de Algemene Taalwetenschap. 43 H. J. Pos, "Phonologie en betekenisleer," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, nieuwe reeks, deel 1, No. 13 (1938). 44 H. J. Pos, "Perspectives du structuralisme," Travaux du Cercle Linguistique de Prague 8 (1939), 71-78. 45 Ibid., p. 74. 49 A. J. B. N. Reichling, Het Woord (Nijmegen : Berkhout, 1935). 47 Reichling, Het Woord, Chapter 5. 48 Jakobson, "Über den Versbau der serbokroatischen Volksepen." 49 M. Rutten, "Dichtkunst en Phonologie," Revue Belge de Philologie et d'Histoire 28 (1950), 871-920. 50 A. Teeuw, Taal en Versbouw (Amsterdam: Djambatan, 1952). 51 C. L. Ebeling, Taal- en Letterkunde; Aspecten van het Russische Formalisme (The Hague: Mouton, 1955). 52 V. Erlich, Russian Formalism: History - Doctrine (The Hague: Mouton, 1955). 53 N. van Wijk, "'Taalbond' en 'Taalfamilie'," De Nieuwe Taalgids 25 (1931), 284-90. 54 R. Jakobson, K xarakteristike evrazijskogo jazykovogo sojuza (Paris: Izdanie Evrazijcev, 1931); reprinted in Selected Writings I (2nd ed., The Hague: Mouton, 1971), 144-201. 55 R. Jakobson, "Principes de phonologie historique," Travaux du Cercle Linguistique de Prague 4 (1931), 247-67; also in Selected Writings I (2nd ed., The Hague: Mouton, 1971), 202-20. 56 N. van Wijk, "Umfang und Aufgabe der diachronischen Phonologie," Mélanges Van Ginneken (1937), 93-99. 57 See, for example, C. L. Ebeling, Linguistic Units (=Janua Linguarum 12) (The Hague: Mouton, 1960), pp. 136-43; and E. T. H. Nuytens, De tweetalige mens (Assen: Van Gorcum, 1962). 58 N. van Wijk, "Quantiteit en Intonatie," Mededeelingen der Koninklijke Nederlandsche Akademie van Wetenschappen, Afd. Letterkunde, nieuwe reeks, deel 3, No. 1 (1940). 59 Roman Jakobson, "Observations sur le classement phonologique des consonnes," Proceedings of the Third International Congress of Phonetic Sciences (1938), 34-41 ; also in Selected Writings I (2nd ed., The Hague: Mouton, 1971), 272-79. 60 R. Jakobson, M. Halle and C. G. M. Fant, Preliminaries to Speech Analysis, Acoustics Laboratory Technical Report 13 (Massachusetts Institute of Technology, 1952). 61 A. W. de Groot, "Voyelle, consonne et syllabe," Archives néerlandaises de phonétique expérimentale 17 (1941), 21-41. 62 E. M. Uhlenbeck, "De structuur van het Javaanse morpheem," Verhandelingen Koninklijke Bataviaasch Genootschap van Künsten en Wetenschappen 78 (1949). 63 A. H. Kuipers, The Squamish Language (The Hague: Mouton, 1967). 64 Roman Jakobson, "Zur Struktur der russischen Verbums," Charisteria Gvilelmo Mathesio Quinquagenario . .. oblata (Prague, 1932), 74-84; also in Selected Writings II (The Hague: Mouton, 1971), 3-15.

502

E. M. UHLENBECK

65 A. W. de Groot, "Les oppositions dans les systèmes de la syntaxe et des cas," Mélanges Bally (1939), 107-27. 66 A. W. de Groot, "Classification of Cases and Uses of Cases" in For Roman Jakobson (The Hague: Mouton, 1956), 187-94. 67 A. W. de Groot, "Classification of the Uses of a Case Illustrated on the Genetive in Latin," Lingua 6 (1956), 8-66. 68 De Groot is referring here to J. Kurytowicz, "Le problème du classement des cas," Bulletin de la Société Polonaise de Linguistique 9 (1949), 20-43. 69 De Groot, "Classification of Cases," p. 188. 70 C. H. van Schooneveld, "The Aspect System of the Old Church Slavonic and Old Russian Verbum Finitum byti," Word 7 (1951), 96-103; idem, Over de woordsoorten in het Moderne Russisch (Leiden : Brill, 1953) ; idem, A Semantic Analysis of the Old Russian Finite Preterite System (The Hague: Mouton, 1959). 71 A. G. F. van Hoik, The Semantic Spectrum of the Russian Infinitive (Leiden: Sijthoff, 1953). 72 C. L. Ebeling, "On the Semantic Structure of the Russian Sentence," Lingua 4 (1954), 207-22. 73 C. L. Ebeling, "A Semantic Analysis of the Dutch Tenses," Lingua 11 (1962), 86-99. 74 A. H. Kuipers, "The Circassian Nominal Paradigm: A Contribution to Case Theory," Lingua 11 (1962), 231-48. 75 . E. M. Uhlenbeck, "De structuur van het Javaanse morpheem"; idem, "Verb Structure in Javaanse" in For Roman Jakobson (The Hague: "Mouton, 1956), 567-73; idem, "De systematiek der Javaanse pronomina," Verhandelingen van het Koninklijke Instituut voor Taal-, Land- en Volkenkunde 30 (1960). 76 Roman Jakobson, "Shifters, Verbal Categories and the Russian Verb," Russian Language Project, Department of Slavic Languages and Literatures, Harvard University (1957); also in Selected Writings II (The Hague: Mouton, 1971), 130-41. 77 Uhlenbeck, "De systematiek der Javaanse pronomina." 78 Roman Jakobson, "Kindersprache, Aphasie und allgemeine Lautgesetze," Sprdkvetenskapliga Sällskapets Förhandlingar, 1940-1942 (Uppsala, 1942); reprinted in Selected Writings I (2nd ed., The Hague: Mouton, 1971), pp. 328-401. 79 Roman Jakobson, "Linguistics and Poetics" in Style in Language, ed. Thomas A. Sebeok (Cambridge, Mass. : Technology Press of M.I.T., 1960), pp. 350-77. 80 B. Th. M. Tervoort, Structurele analyse van visueel taalgebruik binnen eengroep dove kinderen (Amsterdam: North Holland, 1953). 81 W. Kaper, Einige Erscheinungen der kindlichen Spracherwerbung erläutert im Lichte der vom Kinde gezeigten Interesses für Sprachliches (Groningen: Wolters, 1959).

ROMAN JAKOBSON A N D A V A N T G A R D E ART THOMAS G. WINNER"1

Tjia3 KOCH

B neiaTH cypryia Ha nponeT 6ojrraJi o PoMKe XKo6coHe H CMeniHo noTen, CTHXH yna...

(MaHKOBCKHit, "ToBapnm Herre")

Majakovskij's verse, invoking his friend, conveys to us something of that stimulating period, so familiar to the young Roman Jakobson, when, during the early decades of this century, Russian avantgarde painters, poets, and artists of all kinds were freely experimenting with revolutionary new forms. Jakobson himself recalls, in several of his writings, the fundamental impact of this movement upon his developing thinking.1 In fact, in 1914 Jakobson contributed his own futurist poetry, and at that time he also translated a poem by Majakovskij into French and, with a typical futurist twist, into Old Church Slavic, as we shall discuss later. At the same time, as one of the founders and leaders of the Moscow Linguistic Circle and the OPOJAZ, Jakobson was beginning to construct the aesthetic theories which later, under his leadership, were to become the foundation of the poetics of the Prague Linguistic Circle. Unfortunately, we possess all too scant documentary materials bearing upon the personal relations between Jakobson and this group of artists. We do know that Jakobson was particularly influenced by his friend, the Suprematist painter Kazimir Malevic. The interesting question of the relation of Malevic's minimal units of spatial forms and Jakobson's minimal units of speech sound has been brought to the fore in a provocative essay by the French art historian Dora Vallier.2 Another close friend was the painter and composer Mixail MatjuSin, the husband of the poet Elena Guro, who composed the music for Krucenyx's opera, Pobeda nad solncem. Jakobson's most intimate associates among the futurist poets, many of whom of course were also painters, were Majakovskij, Xlebnikov, and

504

THOMAS G. WINNER

KrucSenyx. He was also in close contact with Elena Guro and D. Burljuk. A large correspondence between Jakobson and these practitioners of avantgarde art, unfortunately lost in Prague during the German occupation, included a wealth of theoretical discussions, especially between Jakobson and Krucenyx, 3 as well as-many friendly communications. Among the materials which have been preserved are some dedications to Jakobson inscribed by Majakovskij on his works, a letter to Xlebnikov containing some theoretical material, as well as some of Jakobson's own zaum poetry, some of which are the following: 1. Dedications by Majakovskij to Jakobson: At the opening of the book Vse socinnenoe Majakovskim (Petrograd, 1919), presented to Jakobson in 1919: Te6e

PoMKa

XBajIH rpOMKO

On the book Majakovskij dlja golosa (Berlin, 1922): MHJIMH POMHK ! flJIH HOBblX XBaJieHHH HOBblH TOMHK. 20 V I I 23

B MOHK

On a copy of the poem "Xoroso": MHJIBIM — PoMHKy

5KeHe ero h ero AOMHKy BOA MOHK

(Cited from Jakobson, "Novye stroki Majakovskogo." 4 ) 2. A letter by Jakobson to Xlebnikov, dated February 1914, with a discussion of some theoretical issues of futurism. 3. Some of Jakobson's own zaum poetry. These examples bear witness to the intimate relations between Jakobson and the Russian avantgarde, which also continued with the members of the Czech avantgarde in the twenties and thirties. These early influences were, as we shall show, of fundamental importance in Jakobson's early career and have, moreover, continued to affect very significantly the evolution of his later thinking.

AVANTGARDE ART

505

In addition to the direct influences of these friends, Jakobson's early, and unceasing, attempts to understand the deeper relations between natural language, the verbal and the visual arts, and the nature of the aesthetic function in communication in general, can be related to various general causes or themes. For one thing, the milieu of the times encouraged such questions. As we know, during the early decades of the twentieth century in Europe the various arts were turning their attention toward the characteristics of their specific media. Thus the verbal arts exploited to the utmost the artistic potential of natural language, and the visual arts those of forms and color; and furthermore, in Russia, the verbal and visual arts were in unusually close contact. As Jakobson recalls, this was the subject of many of his discussions with Malevic and Krucenyx : Les visites de Malévitch, chez moi, et nos conversations de 1913-1914 étaient consacrées à ce que nous appelions "la libération de l'énergie" dans la peinture et la poésie: peinture non-représentationelle et poésie non-référentielle, slogans avec lesquelles nous projettions de partir pour Paris. Au cours de l'été 1914 avec mon grand ami le poète futuriste Aleksei Kroutchenykh, j'ai rendu visite à Malévitch, ...et nous avons eu des discussions vivifiantes, tous ensemble, autour des mêmes sujets... Parmi les sujets de discussion se trouvaient mes recherches en matière de poésie zaoumnaja (supraconsciente).5 In fact, cubo-futurist painting and the minimalist experiments of Malevic served as principal inspirations to Russian futurist poetry. 6 Furthermore, many futurist poets were also active as painters (e.g. David Burljuk, Elena Guro, V. Majakovskij, and A. Krucenyx, to name only a few). Futurist poetry itself was highly syncretic, frequently combining the auditory and lingual with the visual aspects of the poem, through particular arrangement of the lines and/or letters and, at times, the use of special paper on which the verse was printed. (We know, for instance, that some poems were printed on wallpaper.) The hegemony of the visual arts in the 1910s, noted by Grygar, 7 is expressed aptly by Xlebnikov, when he says : "We want the verbal arts (slovo) to follow boldly in the footsteps of painting." 8 This same close interconnection between the visual and the verbal arts also characterized the Czech avantgarde during the twenties and thirties, when Jakobson worked in Prague and in Brno. Suffice it to remember that Karel Teige, the foremost theoretical spokesman of the Czech avantgarde, also first entered the sphere of art as a painter, strongly influenced, as Effenberger has pointed out, 9 by analytical cubism. Secondly, the philosophical developments of the early twentieth century, representing a fundamental break with the past values and traditions,

506

THOMAS G. WINNER

underlay many of the specific movements we have discussed. To a perceptive and analytical thinker like Jakobson, these early currents were not lost. In the arts, traditional realities based on the mimetic principle no longer seemed interesting, as new ways of perceiving space and time became increasingly important, and as understandings of the inner dynamics of structures became the subjects of the new scientific and artistic thinking. Finally, of course, there is that elusive efficient cause, the individual, who catalyzes and creates new syntheses of newly-phrased universal questions. Here one can only hazard a few guesses. As Jakobson has said of himself : ... il ne s'est jamais agi d'un seul thème. La pluralité des thèmes, c'est mon habitude depuis ma jeunesse. Ce qui m'intéressait déjà, c'était de confronter des thèmes divers, d'obtenir ainsi une certaine complémentarité entre diverses disciplines.10

In the history of Western thought, indeed of human civilization, there have always been individuals whose work and life served to bridge, mediate, and synthesize, diverse areas of culture. Jakobson's work throughout his lifetime has been an outstanding example of the innovative bringing into focus of various interrelationships not previously detected. As we know, his conceptions span numerous cultural spheres; particularly those of the arts and the sciences, most specifically linguistics and poetics understood in the broadest of contexts. We turn now to Jakobson's specific writings on the avantgarde, both theoretical and artistic. Interestingly, Jakobson's earliest writings in 1914, signed with the pseudonym of Aljagrov, were a series of futurist zaum poems, a part of a collection of futurist poetry published by Krucenyx, entitled Zaumnaja gniga.11 The title itself is a typical futurist word play : the word gniga, built on the Russian root gm- (associated with the meaning of rotting) replaces kniga (book). The following brief zaum verse, written by Jakobson at the age of eighteen and appearing in Zaumnaja gniga, well illustrates Jakobson's early interest in radical experimentation with abstract poetry: AJIJirPOB M3rjIb/6)KByO HHXbHHbflpbW

HTJI3IIÎK XH (j)H CMI CKbinOJ13û

a BTfl6-flJiKHH T b a n p a KaKaii3HflH eBpeeu; nepHHjibHHua

PA3C£flHOCTb.

yflyiiia

HHKK

apKaH

KaHKOH apM^HK Ayiua/ZHKH KHTajZHKH kut bi Tax h Hi/ieafl

AVANTGARDE ART

507

apM/jK 3THK3TKa TKXafl TKflHb TMK TKOHI« KFLHTHK a O O p i l i a T K ^ H T H T/OK TOKH MSIK

TM/lHTbl XHjiKy HIK^M aHM/l Kb/Kb aTpa3HKciio HaMe'K yMe'H M^HK - yiliaTH He a B a o n o c T H e

T&MH

nepeaoBHua

nepeAHHK r y 6 j i n u i o c T o n TJIAK B

Baro nepeflaBACb

In another edition of zaum poetry, also edited by Krucenyx and entitled Zaumniki,12 we find a single poem by Jakobson, also written in 1914. Zaumnaja gniga, which has not been available in its totality to this author, appears to be an excellent example of the syncretic character of the art of this period, and especially of the strong visual elements of the poetry of those days. The volume is described elsewhere as follows: The cover is a collage: a heart of vermilion color with a real button in the center. The epigraph of the book prohibits its being read in a state of clear mind. Inside are ten etchings ... representing various playing cards [with a strongly cubist character, TW] by Ol'ga Rozanova (1886-1918), one of the first Russian futurist painters.13 In the same year in which Jakobson wrote these poems, he discusses, in a letter to V. Xlebnikov, experimental poetry. Anticipating his later thought about paradigmatic and syntagmatic axes, he argues with Xlebnikov that the alphabet is not too poverty stricken for futurist poetry, if new approaches are used, such as the superimposition of letters (Jakobson uses the term splety bukv), analogous to musical chords. Thus, said Jakobson, we could achieve simultaneity, as well as variety of graphic combinations. 1 4 I am not aware of any poetry ever written with this kind of graphic arrangement; but it would of course have been pure collage. In 1918, Jakobson attempted a tour de force, the translation of Majakovskij's poem "Nicego ne ponimajut" into Old Church Slavic, a typical futurist joke, given the great interest of futurist poets such as Xlebnikov in etymological word plays, Slavic roots, and Old Church Slavic. The final poetic attempt of the young Jakobson was a translation into French of a stanza of Majakovskij's "Oblako v stanax," which was never completed

508

THOMAS G. WINNER

nor published. It is, however, supplied here, following the recollections of Roman Jakobson. BouiJia Tbi, p e 3 K a a , Kaic ' H a T e ! ' ,

Myna nepnaTKw 3aMui.

CKa3ajia : '3HaeTe — .H Bbixoacy 3aMy»c'.

Apre, comme un mot de dédain tu entras, Marie, et tu m'as dit, en tourmentant tes gants de daim : 'Savez-vous je me marie'.

In January 1917, Jakobson remembers, in Brik's apartment, Majakovskij listened to Jakobson's recital of this translation and asked him to translate this stanza for him orally word for word from French into Russian. When it was done, Majakovskij said, "It's all right. I was afraid that one and the same word was repeated to make a quasi-rhyme, but now I see that 'Marie' - 'marie' are mere homonyms." 15 Jakobson's first theoretical article, entitled simply "Futurizm," 1 6 is in a sense prophetic of his entire life work, for it deals with general, broad questions of aesthetics, and specifically with a searching examination of cubist painting. The essay is probably the first structural discussion of these new directions in painting, and it takes up a number of important concepts which are later to become significant aspects of structural thought in general, and of Jakobson's and Mukarovsky's aesthetics in particular. The close relation of cubist painting to the cultural-scientific revolution of the early twentieth century is first elaborated in this essay of the twenty-threeyear-old poet-scholar who discusses the "fundamental transformation of the ancient scientific edifice" that has taken place, " a transformation which the history of science has not known before [and which] destroys verities that...have never been affirmed, because they appeared self-evident." 17 Jakobson comments penetratingly upon the departure from static space and linear progressive time that dominates the new cubist aesthetics, where traditional perspective is abandoned and objects are presented in a multiple temporal-spatial fashion. The new view of time and space forces the cubist painters to overcome the discreteness of temporal and spatial units which now may overlap, that is, for instance, the front and the back of a face may be shown simultaneously. This brings cubism to a radically new conceptualization of the relation between art and reality. For the cubists, the painting is no longer a static object that catches a moment in time and space, but it has a distinct temporal dimension. Thus several time slices may coexist in it; hence the cubist painters are no longer preoccupied with the mimetic relation of the painting to an external reality, but with the exploration of

AVANTGARDE ART

509

pictorial possibilities. 18 Jakobson put it this way in his essay: "Painting emancipates itself from the elementary illusion." 19 As Picasso said: "I am not painting a woman, I am painting a picture." During the same period when Jakobson was investigating the relation of the new art to new conceptualizations of time and space, the Russian philosopher P. A. Florenskij was also commenting upon new philosophies of time and space, as has recently been demonstrated by V. V. Ivanov. 20 Such views by Florenskij and Jakobson anticipated by several decades many contemporary discussions of the relation of the arts to the new view of time and space, as for example Sartre's discussion of the use of time in Proust and Faulkner. 21 Concerning the resulting revolutionary approach to internal relations in cubist art, Jakobson wrote in his early essay that the new painting intensely exploits different fields of pictorial expression, which brings a fundamentally new stress on internal relationships. "Correlations of volume, constructive assymetry, dissonance of colors, density, emerge in the painter's consciousness." 22 The problem of the new relation of cubist painting to external referents, and of the intense preoccupation with the internal relationships of the elements within the painting to the total structure of the work of art, continue to preoccupy Jakobson over four decades later in his discussion of the semiotics of cubism: The mode in which the signatum stands relative to the signans on the one hand, and to the denotatum on the other, has never been laid bare so plainly before, nor the semantic problems of art been brought forward so provocatively as in cubist pictures, which delay recognition of the transformed and obscured object, or even reduce it to zero.23 Introduced into this early discussion are three focal ideas which took much of their inspiration from the data of cubist art itself: Firstly, the concept of the "laying bare of the device" (obnazenie priema)2i which became a key concept in the Russian "Formal" school (Jakobson was soon to call the device "the sole hero of literary investigations"); 25 secondly, the autonomy of the aesthetic structure which, with certain modifications, became the theoretical cornerstone of Prague school aesthetics a decade later (as Jakobson wrote in his early essay, the artistic work itself, its inner structure, no longer needs to be justified by an outside reality because "it becomes autonomous"; 2 6 and thirdly, artistic deformation which, according to Jakobson's early discussion, while always realized "in small doses" in art, "becomes canonized in cubism." 27 This idea, later to be called ostranenie by the "Formalists," became transformed into the structural concept of aktualizace (best translated by Paul Garvin as "foregrounding") in the aesthetic system developed by the Prague Linguistic Circle, which led to a

510

THOMAS G. WINNER

radically new theory of artistic evolution based on the notion of the aesthetic norm and its violation. Even in those early years, however, Jakobson suggests the concept of norm violation and, most importantly, its relativity to its period, which he is later to elaborate along with Mukarovsky and others in Prague. In a recent interview with Emmanuel Jacquart, Jakobson stated: Dès ma jeunesse, j'ai découvert une perspective qui m'a frappé moi-même et qui a frappé mes amis. A l'époque où j'étais encore jeune étudiant, on ne manifestait, en Russia et dans d'autres pays slaves, aucun intérêt pour l'art des icônes et des fresques médiévales. On croyait que cela relevait de l'archéologie. Tout à coup, on a été très étonné quand au début du siècle, lors de sa visite à Moscou, Matisse a déclaré : ce que vous avez devant vous, ce sont des maîtres de l'art moderne et vous ne le remarquez pas ! C'est à cette époque qu'on s'est mis à étudier les grands valeurs artistiques du Moyen Age russe....28 In this recent interview, Jakobson compares medieval Russian art to the modern collage, since different biblical motifs and texts are frequently combined with the original text. 29 But even in his first essay Jakobson detected elements of cubist perspective in primitive art. Thus, he noted that some medieval Russian icons may represent the same saint two or three times in contiguous moments of action, suggesting a plurality of spatial-temporal points of view. 30 Jakobson never abandoned his interest in medieval art, which is well represented in many interesting essays concerning Russian and Czech medieval poetry, as well as in his latest essay on a fourteenth-century Serbian poet, Siluan. 31 Two years after the appearance of "Futurizm," two more crucial essays by Jakobson on avantgarde trends in the arts appeared: There was first the lengthy study of futurist poetry, Novejsaja russakaja poèzija, one of Jakobson's most important early works and his first solid study of the aesthetics of verbal art. The essay was actually written in 1919 (the same year in which "Futurizm" had appeared) as a preface to an unrealized edition of Xlebnikov's works, and it had been presented orally at the Moscow Linguistic Circle and the OPOJAZ. The second essay was inspired by Jakobson's observation in the West of dadaism, another radical art movement. The resulting brief essay was sent to the Moscow journal Vestnik teatra, where it appeared in the rubric "Letters from the West," under the title "Dada." 3 2 The article discusses the writings and manifestos of Hiilsenbeck and Tzara and establishes points of contiguity and of contrast between dada and futurism. The important essay Novejsaja russkaja poèzija is a discussion almost entirely of poetry, and appears to be a companion piece to the earlier

AVANTGARDE ART

511

essay "Futurizm," which had dealt primarily with the visual arts of the epoch, and cubo-futurism. Again, we see the germs of several of Jakobson's later fundamental concepts in this work of his youth, including the following: (1) The relation of aesthetic norm and aesthetic norm violation, more fully expressed in the Jakobson-Tynjanov manifesto of 1928,33 is developed here in greater detail and complexity than it was in "Futurizm." Jakobson argues that, over time, the aesthetic norms of a given epoch (he gives the example of those of Puskin's time) lose their freshness, become cliches, and call for replacement by unfaded norms. "Form," Jakobson argues, "exists for us only as long as it is difficult for us to perceive (vosprinjat') it, as long as we feel the resistance of the material." 34 (2) Secondly, the concept of the aesthetic function, based on self-orientation of the aesthetic object, is first clearly introduced in this essay. As Jakobson wrote, "the Russian futurists are the initiators of 'the autonomous, self-valuable' [samovitogo, samocennogo] word." 35 "Poetry...is nothing else than an utterance with an orientation on the expression,.. .Poetry is language in its aesthetic function." 36 The second part of the essay illustrates the theoretical statements by demonstrating in great detail how the device is "laid bare" in the poetry of Xlebnikov, in his syntax, lexicon, rhyming, etc. This early statement of the aesthetic function based on self-orientation culminates in Jakobson's later well-known model of communication, in which the aesthetic function is seen as one of the six functions of the communicative act.37 In this early version, Jakobson speaks only of three functions: emotive, referential, and poetic. Where the poetic function is dominant, "the referential function is minimal." 38 The view of the strict independence of the artistic work, conveyed in this article, will later be modified, beginning with the 1928 manifesto written with Tynjanov, which qualified the concept of the total independence of the artistic work. Increasingly, the concept becomes one of autonomy, which is still further modified when the semiotic view is adopted. While we cannot find explicit reference, in these early essays, to the semiotic point of view, the germs of this later approach were nevertheless also present. The semiotic approach, which transformed the aesthetics of the Prague school, no longer saw the aesthetic system as strictly self-contained, since it is always qualified by its complex interrelations with other systems, the analysis of which was brought to the fore as equal in importance to that of the internal structure of the system itself. All forms of reduction, as well as reification and isolation of systems, were finally eschewed in the search for interrelationships between the verbal, visual, tonal, dramatic, and cinematic arts, the natural language itself, and larger contexts. While, as Jakobson has

512

THOMAS G. WINNER

recently commented, the term semiotic was not employed in his youth, his early discussions with Russian artists suggested the theme to him: Par hazard, mon adolescence s'est passée parmi de jeunes peintres et j'ai eu avec eux des longues discussions sur les rapports entre la peinture et la poésie, entre les signes visuels d'une part, et les signes verbaux d'autre part. C'est ainsi que la sémiotique s'est posée à moi pour la première fois, et, comme vous le savez, je ne l'ai jamais quittée.39 It may sound like an oxymoron for Jakobson to speak of his scientific biography as conservative but, just as in the case of Majakovskij, his work is of one piece. The fundamental ideas which so interested Jakobson, the eighteen-year-old zaum poet and friend and associate of Majakovskij, Xlebnikov, Krucenyx, and Malevic, never lost their fascination for him. They were developed and transformed during the sixty-two years that have elapsed since Jakobson published his first poem, and they are still continuing to agitate his mind. His opposition to mechanistic and reductive interpretations of art, his rejection of simplistic answers, his dialectical view of oppositions and complementarities which have been manifest in all his work, in linguistics, and aesthetics, and semiotics, his openness to new forms of art, all mark him not only as a friend, of the avantgarde, who was stimulated by the revolutionary artistic practice of the futurists and cubists, but also himself as a true avantgarde personality who has led the way in all the fields in which his seminal mind has been active. Clearly, his early active participation in, and association with, the cubo-futurist painters and the futurist poets of Russia had a fundamental impact on his entire personality and on his evolving scientific biography. As he says himself: Ceux qui ont exercé sur moi le plus d'influence, ce sont des artistes, non les savants...Picasso, Bracqiie, Khlebnikov, Joyce, Stravinsky.40 Brown University

NOTES * I wish to express my appreciation to Roman Jakobson for making accessible to me some invaluable materials, as well as for having shared with me some of his personal memories. 1 E.g., Jean Pierre Faye, "Questionner Jakobson," Nouvelle Revue Française (1966), cited from Le récit hunique (Paris, 1967), pp. 273-85 ; "Entretien avec Emmanuel Jacquart : autour de la poétique," Critique 348 (Paris, 1976), p. 462. 2 "Malévitch et le modèle linguistique en peinture," Critique 334 (Paris, 1975), pp. 284-96.

AVANTGARDE ART 3

513

As related to the author by Roman Jakobson in a personal conversation, July 28. 1976. 4 R. Jakobson, "Novye stroki Majakovskogo. I: Tekst i primecanija; II: Kommentarij k pozdnej lirike Majakovskogo," Russkij literaturnyj arxiv, published under the auspices of the Department of Slavic Languages and Literatures, Harvard University (New York. 1956), p. 180. 5 R. Jakobson, "Message sur Malévitch," La peinture. Change 26-27 (1976), pp. 293-94. 6 Cf. N. Xardziev, "Majakovskij i zivopis'," Majakovskij: Materialy i issledovanija (Moscow, 1940). 7 Mojmir Grygar, "Kubizm i poèzija russkogo i ceäskogo avangarda," Structure of Texts and Semiotics of Culture, eds. Jan van der Eng and Mojmir Grygar (The Hague: Mouton, 1973), pp. 59-101. 8 V. Xlebnikov, Neizdannye proizvedenija, eds. N. Xardziev and T. Gric (Moscow, 1940), p. 334. 9 Vratislav Effenberger, "Od 'stavby bâsnë' k surrealistické semiologii" unpublished manuscript (1974), pp. 6, 10. 10 "Entretien avec Emmanuel Jacquart: autour de la poétique," Critique 348 (1976). p. 461. 11 A. Krucenyx, ed. (Moscow, 1914). 12 Moscow, 1921. 13 L'Arc 60 (Paris, 1975), p. 19. 14 R. Jakobson, letter to V. Xlebnikov, cited from Poèticeskaja kul'tura Majakovskogo, eds. N. Xardziev and V. Trenin (Moscow, 1970), p. 37. 15 Personal communication from Roman Jakobson, who supplied the translation of the stanza from memory. 16 R. Jakobson, "Futurizm," Iskusstvo, August 2, 1919 (Moscow), cited from R. Jakobson, Questions de la poétique, ed. Tzvetan Todorov (Paris: Seuil, 1973), pp. 25-30. 17 Ibid., p. 27. 18 Ibid., pp. 25, 27-28. 19 Ibid., p. 26. 20 V. V. Ivanov, "Kategorija vremeni v iskusstve i kul'ture XX. veka," Structure of Texts and Semiotics of Culture, eds. Jan van der Eng and Mojmir Grygar (The HagueParis: Mouton, 1973), pp. 103-150. 21 Jean Paul Sartre, "A propos du bruit et la fureur: la temporalité chez Faulkner," Nouvelle Revue Française, 52 (1939), pp. 1057-61. 22 R. Jakobson, "Futurizm," op. cit., p. 26. 23 R. Jakobson, Selected Writings, II, "Retrospect" (The Hague : Mouton, 1962), p. 632. 24 Ibid. 25 R. Jakobson, Novejsaja russkajapoèzija (Prague, 1921); cited from Texte der russischen Formalisten, ed. Jurij Striedter, Vol. II (Munich: W. D. Stempel, 1972), p. 31. 26 Ibid. 27 Ibid., p. 25. 28 "Entretien avec Emmanuel Jacquart: autour de la poétique," Critique 348 (1976), p. 468. 29 Ibid. 30 R. Jakobson, "Futurizm," op. cit., p. 25. 31 R. Jakobson, "Slavoslovie Siluana Simeonu," Xenia Slavica. Papers Presented to Gojko Ruzicic on the Occasion of His Seventy-Fifth Birthday, 2 February 1969, eds. R. L. Lencek and B. O. Unbegaun (The Hague-Paris: Mouton, 1975), pp. 75-83. 32 R. Jakobson, "Pis'ma s zapada. Dada," Vestnik teatra 82 (Moscow, 1921), p. 3. The author is indebted to Stephen Rudy for a handwritten copy of this essay.

514 33

THOMAS G. WINNER

R. Jakobson, and Ju. Tynjanov, "Problemy izucenija Iiteratury i jazyka," NovyjLEF (Moscow, 1928). 34 R. Jakobson, Novejsaja russkaja poèzija, op. cit., pp. 20-21. 35 Ibid., p. 28. 36 Ibid., p. 30. 37 R. Jakobson, "Linguistics and Poetics," Style in Language, ed. Thomas A. Sebeok (New York, 1960), pp. 350-77. 38 R. Jakobson, Novejsaja russkaja poèzija, op. cit., p. 30. 39 "Entretien avec Emmanuel Jacquart: autour de la poétique," Critique 348 (1976), p. 462. 40 Jean Pierre Faye, "Questionner Jakobson," Nouvelle Revue Française (1966), cited from Le récit hunique (Paris, 1967), p. 281.

ROMAN JAKOBSON A N D THE STUDY OF RHYME DEAN S. WORTH

Roman Jakobson's work on rhyme covers over a dozen languages and a thousand years of verse history. It is rich in both theoretical insights and factual knowledge. The following essay does not pretend to catalogue all of Jakobson's accomplishments - that would be foolish, since they continue to accumulate daily - but attempts only to give some idea of the range of problems he has addressed and, here and there, to convey something of the flavor of his scholarship. I have avoided polemics and extended digressions (my few questions and doubts are relegated to the footnotes), nor have I repeated - as far as I know - what is already available in the literature.1 The essay has been divided - somewhat artificially - into two parts. The first discusses Jakobson's views of the nature and functions of rhyme and the second sketches his contribution to the rhyme histories of Czech and Russian, and to allied historical topics.

I. THE NATURE AND FUNCTIONS OF RHYME

Jakobson has always insisted that rhyme is only a special case of parallelism and, as such, is intimately connected to other types of parallelism (metrical, grammatical, semantic, etc.).2 Rather than the "isolationist" Zirmunskij, who opposed rhyme as an "organizing" factor to all other sound correspondences ("orchestration"), 3 Jakobson has always been an "integrationist," viewing rhyme itself in its constant interplay with other sound patterning, even where the latter is clearly not rhyme in the narrow sense, e.g. in Brecht's Einzelne - zwei - Partei.4 It is this view of rhyme as one integral part of a structured poetic whole that has led Jakobson's work on rhyme into so many contiguous areas of (phonological, morphological, etc.) structure.

516

DEAN S. WORTH

The Phonology of Rhyme. Throughout his career Jakobson has emphasized that rhyme is a phonological, not a phonetic phenomenon. 5 His insistence that the sound texture of verse operates "with phonemes, not with sounds" 6 has been misinterpreted in some quarters to mean that phonemic identity is a precondition of rhyme, i.e. that separate phonemes cannot rhyme with each othei' and that subphonemic entities can play no role in rhyme - two statements which Jakobson has not only not made but which are clearly contradicted both by such oft-cited rhymes of phonemically distinct entities as classical German /u/ - /i/ and by many of the phenomena discussed in Jakobson's own work, e.g. the frequent rhyme of /e/ - /e/, etc. in Old Czech7 or of /k/ - /x/, /j / - /»/, etc. in eighteenth and nineteenth-century Russian verse,8 not to mention more remote correspondences such as Old Czech syllabic /j/ rhyming with /i/, /i/, /y/, /y/, /e/, /ie/.9 As far as subphonemic elements are concerned, Jakobson is himself explicit: "an autonomous rhyme role can be played not only by entire phonemes, but also by their individual differential elements: either both members of a binary opposition, e.g. soft and hard consonants, are identified with each other in a rhyme set, or, on the other hand, the rhyme is restricted to the marked members of such a binary opposition, e.g. the identification of various soft consonants in open masculine rhymes." 10 The Jakobsonian view of rhyme as phonemic must be interpreted to mean that assonance, alliteration, etc. - in brief, all poetically organized sound - consists not of extrasystematic phonetic correspondences (i.e. of linguistically irrelevant physical data), but rather of structured, phonologically relevant entities, i.e. of entities of the phonemic system; these may be phonemes themselves ([i] rhyming with [y] because both=/i/), one or more distinctive features (/j/ rhyming with any [ +sharp] consonant, e.g. tvoja - ditja = /tvaja/ - /d,it,a/ 11 ), or a series of distinct phonemes joined by some common feature or features, as in the Old Czech instance just adduced, in which the [+vocalic] /¡•/ (and often /}/) rhymes with any [ +vocalic], [ + high tonality] phoneme, or the Nitsch series of voiced or of voiceless consonants. To sum up the misunderstanding: whereas some critics have glossed Jakobson's term "phonemic" as "consisting of (and only of) phonemes," Jakobson himself has clearly intended it to mean "phonemically ( = phonologically) relevant." It is against the background of this latter interpretation that some of Jakobson's parallels become particularly convincing, e.g. the comparison of the [y] - [i] rhyme with the admissibility and rarity, respectively, of choriambic onsets like s nej ubezai' and *s neju bezat': the Russian iambic line can begin with a stressed, but not with a distinctively stressed syllable.12 The phonological rather than physical nature of rhyme is what permits

THE STUDY OF RHYME

517

rhyme canons to vary in space and time, as when the former French distinction between masculine and feminine rhyme is transcoded, after the loss of final e muet, into the non-pronunciation vs. pronunciation of the final consonant, 13 or when one and the same rhyme type is felt as rich rhyme in one genre but merely adequate in another (disyllabic rhyme in sung and spoken verse in fourteenth-century Czech 14 ). "Deep" Rhyme. The complex interrelations of end-rhyme to other sound correspondences have occupied Jakobson since he first discussed Trediakovskij's internal rhyme over sixty years ago. 15 He has studied initial rhyme in Mâcha's verse 16 and has established the basic laws governing "Reimwôrter" of the figli - migli type, 17 but more of his attention has been devoted to deep sound correspondences other than those in absolute anlaut. For example, in discussing the phonological structure of rhymed words in a Hussite battle song, Jakobson notes the song's emphasis on vowels preceding as well as within the rhyme (e.g. i-u-e in slibuje - mien Aui/e);18 such deep correspondences can involve consonants as well as vowels, and go back well beyond the Hussite period, as in Czech Gothic bylo bosti - biele kosti - I/bos//. 19 Internal rhyme is investigated by Jakobson in several languages and in several connections, e.g. in the Czech Gothic lines On Radvana, sen Mladotu a sen /ana, on Radotu, where internal rhyme underscores the dieresis (i.e. the caesura falling on the border between two feet). 20 Internal rhyme is typical of Dante, 2 1 and is used by Baudelaire both to reinforce the end-rhyme (serrants austères sédentaires) and to emphasize the syntactic structure of the line ( Qui comme eux:sont frileux).22 Similar strengthening of end-rhyme by deeper correspondences can be found in Croatian (Drzic's /i/àdnê rodicë - m/âdë deklice; my emphasis, DSW), 23 Serbian (slawz s/ova - «asledova), 24 Russian (Radiscev : kuda. ja edu - ne byvalo s/edu), 25 and Polish (Norwid's interior monosyllables in é reinforcing his e'-rhymes).26 Nowhere, however, was the interrelation of internal and end-rhyme more elaborate than in Czech Gothic verse, where one has end-rhyme, internal-rhyme, and hybrids where line-end words rhyme not with each other but with the first autosemantic word of the following line; the three types, though interleaved, are mutually exclusive, i.e. end-rhymes never correspond to internal words nor internal rhymes to end words. For example:

518

DEAN S. WORTH

(1) (2) (3) (4)

Ach, fot' jsem 1 smutny i pracny2 i nevzacny2 v ciziem /craji3 dalekot' jsem 1 v neznamosti4, v ne/Mi'/osti4 túhú //caji 327

The role of vowel length and tone in establishing or preventing rhyme has been examined by Jakobson in several languages, most notably in various periods of Czech (see "Rhyme in the history of Czech verse," p. 522 below) but also in Norwegian, where rhymed words with different tones provide contrast within similarity 28 in much the same way as in Czech Gothic verse, where the confrontation of vowels differing only in length was a deliberate poetic device (e.g. viere - vérie).29 In one type of Chinese verse, on the other hand, length is a concomitant of rhyme itself, the final syllable of rhymed lines carrying a long beat, in contrast to the obligatory short close of the unrhymed line; 30 Chinese, using both quantity and rhyme in this demarcative function, thus provides a polar contrast to Serbian folk poetry, in which rhymed line-ends and the quantitative close are nearly in complementary distribution. 31 Rhyme in Relation to Morphological Entities. A recurrent theme in Jakobson's rhyme studies has been the relation of rhyme to units of the morphological system (word-boundaries, grammatical affixes and the meanings they express, etc.). The connection between rhyme and word-boundaries is clearest in his work on Czech verse, perhaps only because he has done more work on this language than on any other. It was he who first called attention to the fact that in Czech Gothic verse, in those (rare) cases where the line ends in a stressed monosyllable, this monosyllable is connected to the preceding word by the rhyme (c/ovek - tento vek), thus helping to suppress the word-boundary and to avoid stressed even syllables at lineend; 32 half a millennium later K. H. Mácha and other Romantic poets had difficulty introducing masculine rhymes because this could be accomplished only with the final monosyllables; 33 in general, Czech tends to use only disyllabic rhymes, unless a word-boundary intervenes before a final monosyllable.34 The importance of morpheme classes (derivational vs. flexional affixes, lexical morphs) in rhyme has been emphasized both in general 35 and specifically for Slavic material, in which - unlike Germanic and Romance grammatical rhyme is said to play a decisive role, 36 since rhyme types often correspond to grammatical categories like singular - plural, masculine feminine, etc. 37 The correlations of grammatical classes and rhyme types have been demonstrated for Baudelaire, who rhymes not grammatical

THE STUDY OF RHYME

519

endings alone but entire suffixes (¿//«celles - /?arcelles - prunelles),39 for Old Czech, where rhymes "connect divergent grammatical forms of proper and common nouns," 39 in a Dante sonnet, 40 and for eighteenth-century Russian, 41 while in a Brecht poem it is the suppression of rhyme and metric norms that makes grammatical patterning stand out all the more sharply. 42 The interaction of rhyme and grammar sometimes involves a third factor as well: the syllabic structure of verse. The monosyllabic rhyme of Czech Gothic verse is usually grammatical or derivational if the final word is nonmonosyllabic (cinem - prikladtm, radost - mj'/ost),43 while the development of 3- and 4-syllable rhymes in late Gothic (e.g. in the Stockholm Catherine Legend: tociti - osociti, Katefina - materina), with their low semantic correlation (since differing lexical morphs are included in the rhyme) tends to emphasize their primarily phonic function. 44 The development of such multisyllabic rhyme words obviously affects the metrical structure of the verse, reducing, for example, the number of monosyllabic clitics at line-end. 45 Homonymic rhymes have been investigated by Jakobson on Old Czech material. He establishes three types: (1) phonologically identical but grammatically distinct rhyme-words, e.g. nedelii (substantive) - nedelti (verb); (2) pairs where the phonological shape of one member is included entirely in the other, e.g. piti - oslepiti (length is usually irrelevant in Old Cz. rhyme); (3) rhymes where one or both members of the set are prefixed, e.g. zvierti - yen/.46 Jakobson has been concerned with the grammatical or anti-grammatical orientation of poets at least since his early study of Xlebnikov, in which he stated that modern Russian poetry "develops to its limit the tendency to rhyme different parts of speech"; this "form of form," with identical sound but completely different content, is contrasted by Jakobson with paregmenon (etymological figures), in which the basic meaning is identical but the form differs. 47 Jakobson has insisted again and again that rhyme must be either grammatical or anti-grammatical: "The poet's rhyme principle is either grammatical or anti-grammatical, but it is never agrammatical" ; 48 "Thus we observe in Slavic rimed verse either a capitulation to the grammatical rime or resistance to it: indifference toward the grammatical aspect of the rime is excluded"; 49 poetic schools are "oriented toward or against grammatical rhyme; rhymes must be either grammatical or anti-grammatical; an agrammatical rhyme, indifferent to the relation between sound and grammatical structure, would, like any agrammatism, belong to verbal pathology"; 50 rejecting both 2irmunskij's orthographic and Tomasevskij's orthoepic principles, Jakobson repeats that the "innovations

520

DEAN S. WORTH

of rhyme [in] Lermontov and Tjutcev [are based in] the transition from a grammatical to an anti-grammatical orientation." 51 This programmatically binary approach to a complex problem should not be interpreted simplistically to mean that the rhyming elements themselves must either consist only of grammatical morphemes or contain no grammatical elements whatsoever, but merely that the poet or poetic school either concentrates on grammatical rhyme (for semantic, metrical, or other reasons which vary in time and space) or deliberately avoids such rhyme to the extent possible (also for various reasons, including the desire for change itself); "indifference toward the grammatical aspect of the rhyme is excluded" for the simple reason that the grammar itself cannot be excluded, i.e. the grammatical morphemes of the given language are present and cannot simply be ignored (any more than the phonological structure of the language can be ignored; cf. pp. 516-17 above), but must be either used or avoided in rhymes. 52 Rhyme and Semantics. It is not only the grammatical morphemes and their meanings which have occupied Jakobson's attention in his rhyme studies, but also more general semantic problems. Just as poetic sound is phonological, i.e. systematic, and just as no poet can be indifferent to the everpresent grammatical structures and their patterning, so too, "Any attempt to confine such poetic conventions as meter, alliteration or rhyme to the sound level are speculative reasonings without any empirical justification," because "Rhyme necessarily involves the semantic relationships between rhyming units." 53 The semantics of rhymed words can differ from context to context, as Jakobson demonstrates with the pair ráj - máj in Mácha, 54 and one must be wary of subjective, psychological interpretations imposed on the verse by the reader, 55 but rhyme invariably implies either equivalence or contrast between sound and (lexical and grammatical) sense.56 Jakobson's question, "Is there a semantic propinquity, a sort of simile between rhyming lexical units ?" 57 is best classed as a rhetorical question, since he had already answered it positively, as in his discussion of punning rhymes in Old Czech quatrains 58 and in publicistic verse of the Hussite period; 59 such rhymes can be used for parodistic effect, e.g. in the Old Cz. Mam psáno v starém záchodé ('in an old outhouse'), echoing and parodying v starém zákoné 'in the Old Testament', 60 or to provide a comic framing in the prologues and epilogues to Russian prose fairy tales. 61 Jakobson's views of the semantic functions of rhyme are illustrated by his studies of, among other poets, Shakespeare, 62 Codax, 63 and of Dante's "positive" vs. "negative" images and male vs. female agents in the semantics of love, 64 and he even ventures the assertion that Baudelaire's combination of fem-

THE STUDY OF RHYME

521

inine substantives with masculine rhymes underscores the androgyny of his feline subjects. 65 Rhyme and Stanzaic Form. In addition to its multiple connections with other phonological features and with grammatical and lexical semantics, rhyme of course serves as a stanzaic organizer, i.e. rhyme plays an important role in the poetic macrostructure. This has been shown time and again by Jakobson, using material from many different verse traditions. In the Czech Gothic epic, for example, the rhyme pattern creates quatrains, in which internal rhyme regularly appears only in the third of the four lines; the same is true at a somewhat later date of Czech songs, 66 while in other types of verse the onset of larger sections ("periods") is signaled by a rhymed couplet (Vaclave - ceske, Boha- Ducha, etc.) 67 and in a later, Hussite song the rhyme structure varies from one stanza to the next, thus emphasizing their individuality. 68 Both rhyme and syllabic structure distinguish the first quatrain from the remainder of some Janko Krai' poems, 69 while in Chinese regulated verse it is the final syllables of even-numbered lines throughout the poem which are linked by rhyme. 70 Substantival rhyme distinguishes Dante's tercets from his verb-rhyming quatrains. 71 While Blake uses grammatical rhyme to join the odd-numbered couplets of his verse, 72 a Shakespearian sonnet associates non-contiguous quatrains via the similarity of their rhyme sets, 73 and Puskin used his rhyme scheme to divide Cto v imeni tebe moem ? into two symmetrical eight-line groups. 74 Even in a line of Gogol' 's prose, Jakobson finds the author using rhyme to link the distichs of a sentence. 75 This verse-level organizing function of rhyme has been noted and illustrated throughout Jakobson's work on poetics, and, like many other recurrent themes of his rhyme studies, lends an astonishing consistency to investigations spanning over half a century and touching on more than a dozen poetic traditions. 76

II. COMPARATIVE A N D HISTORICAL STUDIES

Common and Comparative Slavic Rhyme. Jakobson's study of Common and comparative Slavic rhyme has been limited to passing remarks ancillary to his major work on comparative Slavic metrics and comparative Slavic literary studies. Noting Nitsch's discovery that Slavic approximate rhymes must remain within either the voiced or the voiceless series {boty can rhyme with boky but not with body), Jakobson attributes this restriction (which is most evident in West Slavic, e.g. Slovak) to "a common

522

DEAN S. WORTH

Slavic patrimony" rooted in the phonemic voicing distinction. 77 The interrelations of rhyme and meter, one of Jakobson's major theoretical interests, appear in their diachronic dimensions in his remark that West Slavic (esp. Czech) folk poetry substitutes final rhyme for the "quantitative close" preserved in the Serbo-Croatian deseterac, a substitution repeated in eighteenth-century SCr. literary imitations of the deseterac, which contain both rhymes and a concomitant weakening of the quantitative close. 78 In another context, Jakobson notes that Slavic (esp. West Slavic) spoken verse (as opposed to song) has rhyme or homoeoteleuton at line-ends, 79 while in the discussion of his new etymology of cex ( < *ceNd-), he compares rhyme pairs with initial labials opposed to various stridents (Cz. souka - louka, Pol. lap - cap, Russ. cuxo - luxo).so Rhyme in the History of Czech Verse. One of Jakobson's major contributions to Slavic scholarship is to be found in his books and articles on Czech verse, especially in its medieval and Romantic periods. Although here, too, as throughout his work on poetics, rhyme is often subordinated to his paramount concerns with metrical and grammatical patterning, it is in Jakobson's study of Czech verse that one finds his most detailed treatment of rhyme, especially its phonological aspect, and finds also the first steps toward his later concern with the relation of phonology to grammar. 8 1 Jakobson's interest in rhyme for its own sake is most apparent in his investigation of Old Czech verse structure. Old Czech rhyme was unrelated to stress and ictus; the extent ("depth") of rhyme was a function of genre, songs having monosyllabic and spoken verse disyllabic rhyme. Rhyme pairs consisted of final vowels, plus the following consonants in the case of closed syllables, e.g. ukvapil - vzbudil, and often but not obligatorily the supporting consonants in the case of open syllables, e.g. mnohe - drahe but also divna - tela; open and closed syllables could rhyme with each other, e.g. hodiny - cervenym,82 There are only two phonological constants in Old Cz. rhyme: i rhymes with y (as in Polish and Russian) and long vowels rhyme with their short counterparts (male - krale, viere - verie [note the complementary distribution of length - DSW]); as Jakobson notes, the role of length in Old Cz. rhyme has yet to be investigated. Other phonological features of Old Cz. rhyme tend to vary in time or by genre; u and iu ( = [U]) rhyme with u, u in the older verse but, losing their own rounding, with i, i in the latter fourteenth century. In some manuscripts syllabic /•, / rhyme with no other consonants, while in others they form pairs with sequences of r, I plus vowel (clovek - crvek, tvrdy - bltidi) and in many mss. with i Iy y (cini - crrti). In early verse, hard and soft consonants rhyme but

THE STUDY OF RHYME

523

rarely (doby - pobi), but with the loss of the sharping correlation in the fourteenth century, formerly opposed consonants fall together and can then of course form rhyme pairs. The oldest manuscripts do not permit pairs matching t d n with V d' ri, but in later texts such "approximate" rhymes occur occasionally. The dental and palatal series sometimes rhyme (s-s,zz), but, in accordance with Nitsch's law, the voiced and voiceless series are carefully distinguished.83 A more complete typology of Old Cz. rhymes was planned for Part II of Jakobson's book O cesskom stixe, but never got beyond the few remarks in a footnote to "StaroceSskie stixotvorenija"; the typology was based on three factors, whether or not (1) ti rhymed with ci, (2) /• / with i, (3) iu with u and/or i.M Rhyme types can be correlated with the evolution of Old Czech verse and often provide a key to textual doors that might otherwise remain locked. Old Czech progresses in a full circle from unrhymed Romanesque through elaborately rhymed Gothic to again unrhymed Hussite verse. Lack of rhyme in the Hospodine, pomiluj ny is considered a sign of this song's great age (by analogy with Latin texts, in which rhyme became obligatory only in the eleventh century),85 but almost all Czech Gothic verse is rhymed (Jakobson distinguishes two types, with and without couplet structure),86 and by the fourteenth century rhyme is extensive and deliberate.87 It develops from purely phonetic rhyme, e.g. in the Alexandreida, to partially grammatical rhyme in parts of the Hradecky rukopis, but returns to phonetic rhyme in the Stockholm Legend (=the expanded version of the Life of St. Catherine);88 here as elsewhere Jakobson calls for a statistical study of the percentage of grammatical rhyme throughout the history of Czech verse, a call which has apparently yet to be answered.89 Late Gothic verse is characterized by "broken rhyme" (lomeni rymu), in which there is no syntactic pause after rhyming line-ends; the presence or absence of broken rhyme helps Jakobson to establish the relative chronology of Old Czech texts.90 Unrhymed verse appears again in songs of the Hussite period, when sung verse had dissociated itself from spoken verse, although the anti-Hussite epic verse still preserved rhyming line-ends.91 A comparison of rhyme types enables Jakobson to solve many textological problems of Old Czech verse, for example, in attributing the three poems O sedmimezcietma blazniech, O peti studeniciech, Anzelmus to the pen of a single author 92 or in repudiating FlajShans' attempt to do the same thing with the Kunhutina pisen and Vitaj, mily Jezu Kriste (Jakobson shows that the latter song has only 50% of exact rhyme, while all but 3 of the rhymes in Kunhutina pisen are exact).93 The presence of assonance (rytierze - witieze) distinguishes O bazni Boii from O smrtedlnosti,94 and details of

524

DEAN S. WORTH

rhyme structure not only contribute to the relative chronology of Old Czech texts but also help to identify borrowings from one text to another. 95 Rhyme as a correlate of literary genre is studied most successfully in Jakobson's work on Old Czech, where he shows monosyllabic and disyllabic rhyme to be characteristic of sung and spoken verse, respectively, as homonymic rhymed couplets (aabb) characterize the epic and cross-rhyme (abab) the song, while rich rhyme is typical of satirical and macaronic verse96 and develops to calambour (mnicha - mecha) and grotesqueries in the democratic verse of the Hussite period (during which alliteration, incidentally, is found primarily in the aristocratic epic of Hus' opponents, e.g. Malo mudre rady maje - Potom to prijal pykaje). 97 Needless to say, all of Jakobson's longer studies of Old Czech verse contain a wealth of minor remarks on the rhyme schemes of individual works, which need not be iterated here. Jakobson's work on nineteenth-century Czech verse is somewhat less extensive than his investigation of Old Czech, but here too he succeeded in establishing regularities that had escaped the attention of earlier scholars. The view of e.g. Gebauer and of Frinta that vowel length was irrelevant to Czech rhyme could not stand up against Jakobson's statistical studies, which proved conclusively that the connection - or lack of it - between rhyme and length was bound to the axes of time and genre, Romantic poets tending to avoid rhyming long vowels with short while late nineteenthearly twentieth-century verse paid little attention to vowel quantity. 98 Equally new was Jakobson's establishment of the relation of rhyme and word-boundary: Czech verse requires a two-syllable rhyme, unless one member of the rhyme pair is a monosyllabic word; i.e., nality can rhyme with temno ty but not with temnoty," the word-boundary functioning as a kind of deuces-wild universal vowel (my terms - DSW). Most of Jakobson's other comments on nineteenth-century Czech rhyme are tangential to his main interest in metrics, e.g. throughout his major study of M&cha,100 which contains passing remarks on the rhyme of other poets as well.101 Rhyme in the History of Russian Verse. Jakobson's studies in Russian verse history deal both with universals and with the rhyme techniques of individual poets. It was Jakobson who discovered the basic structural limitation on the choriambic onset, namely that the choriambus is possible only with initial monosyllabic words; i.e., dam uvedet, net u tebja, noz unesi are possible, but not *damu vedet, *netu tebja, *nosu nesi, in spite of the phonetic near-identity of the two series. For Jakobson this limitation, based on

THE STUDY OF RHYME

525

the role of the phonetically zero word-boundary, illustrates the decisive significance of the phonological system in verse, as does, for example, the consistent rhyming of phonetically disparate [y] and [i], both phonemically /i/. 102 Jakobson's monograph on Czech verse compared to Russian contains remarks on the vocalism of nineteenth and twentieth-century Russian rhyme, connecting the emergence of "new" rhyme (with its disregard of unstressed vowels, e.g. Majakovskij's oxalo - zagloxla) with the reduction of unstressed vowels that set in after the mid-nineteenth century. 103 Another early study outlines Trediakovskij's views of rhyme: the young Trediakovskij considered rhyme essential, except in the epic, while in his mature years he thought rhyme artificial; in both cases, he adduced evidence from folk poetry in support of his views.104 It is fair to assume that Jakobson's lifelong interest in the structure of folk verse can be traced back to this early (1915) study of folk influence on Trediakovskij, who noticed the use of rhyme in such genres as the "lubocnyj stix" (together with assonance: Bogdan - bog dal105) and in marketplace cries like "to zdes' pirogi gorjaci edjat golodny pod"jaci" and in "balagannyx buffonadax" (dva brata - s Arbata);106 fifty-five years later, Jakobson is still interested in the rhyme of distich riddles such as Zverek s versok a xvost' seni verst.107 Jakobson's most important contribution to the history of Russian rhyme is found in his Novejsaja russkaja poezija of 1921. Although devoted primarily to Xlebnikov, this monograph makes a number of important generalizations about both the sound and the semantics of Russian rhyme from Puskin to the present. Noting that Xlebnikov (and modern poets generally) treat consonants as more important than vowels but at the same time, and at first glance paradoxically, tend to disregard the distinction between palatalized and non-palatalized consonants, Jakobson resolves the paradox by observing that vowels are characterized by a high basic tone and that the + / - sharp distinction in consonants is also one of basic tone height, which leads to his conclusion that modern rhyme has developed away from tonality correspondences towards rhyme sets based on what would now be called sonority features. 108 Whereas the Puskin school cultivated posttonic consonantal correspondences, modern poetry cultivates the support consonants instead; corresponding consonants can be similar but not necessarily identical, and "metathetic rhyme" permits them to occur in different sequences. Rhyming words can have differently-placed stress.109 Jakobson closes Novejsaja russkaja poezija with an outline of the relation of sound and meaning in the history of Russian verse, in which the general development is from an original symmetry of both euphony and semantics

526

DEAN S. WORTH

toward a poetry in which rhymed words are almost unconnected semantically; this "denuded" (obnazennaja) rhyme acquires special importance when the verse is only weakly organized by rhythmic constants (meter).110 Rhyme in Other Verse Histories. In addition to his extensive studies of Czech and Russian verse, Jakobson has contributed analyses of the metrical, grammatical and rhyme structure of poems in over a dozen other languages. Rhyme is seldom at the center of attention in these articles, however, being mentioned only cursorily or in connection with the distribution of metric structures or the grammatical categories that have occupied his attention in the recent, grammar-of-poetry-and-poetry-of-grammar period. In his studies of English verse, for example, Jakobson notes in passing that a sonnet from Sydney's Arcadia is rhymed ababcdcdefefgg, which is hardly surprising,111 records the rhymes of Poe's Raven,112 emphasizes the interrelation of rhyme and grammatical patterns in Blake113 and of rhyme and couplet or quatrain structure in both Blake and Shakespeare,114 and treats "echo-rhyme" in the political slogan I like Ike (/ay layk ayk/), 115 but the rhyme material itself, consisting at most of a few dozen lines, does not lend itself to statistical treatment or the type of original interpretation that characterizes Jakobson's best work on Russian and Czech. The same is true of most of Jakobson's comments on the rhymes of poets in other languages: Bulgarian,116 Chinese,117 Croatian, 118 French, 119 German, 120 Italian, 121 Kashubian, 122 Norwegian, 123 Polish,124 Portuguese,125 Rumanian, 126 Serbian,127 Slovak.128 Although one naturally marvels at the breadth • of the material which Jakobson adduces to confirm his views of the interrelations of grammatical and verse patterning, these studies were not intended as contributions to the rhyme histories of the many languages in which he has worked. To the extent that they deal with non-languagespecific aspects of rhyme, they have been discussed in Part One above. CONCLUSION

I have presented Roman Jakobson's rhyme studies from the dual perspectives of general verse theory and individual verse histories. Some things have been omitted, while in other cases more detail has been included than will perhaps appeal to the reader, and it is possible that parts of the poetic forest have been obscured by a too-close examination of its trees. I shall not attempt to provide an overall evaluation of Jakobson's work on rhyme, since such an evaluation would be both presumptuous and premature; his

THE STUDY OF RHYME

527

work is ongoing, and will stand as its own best monument. But, even without evaluating, one may venture to characterize. Jakobson's work on rhyme, which spans sixty-one years as these lines are written, has been permeated since the beginning by a profound sense of structure, by the conviction - passionately held, forcefully expressed, generously illustrated - that rhyme and meter, sound and meaning, form and substance are inseparably linked, that in poetry as in all language, things do not fall apart, the center does hold. Of the many traits that characterize Jakobson's scholarship - vast memory, incomparable powers of synthesis, unending curiosity, and unflagging energy - none is more impressive than the passion which informs his que§t for understanding. Jakobson's work springs from the heart as well as from the mind. I hope he will forgive the the subtext if I close this essay by saying that in the best and broadest sense of the term J. is a sensuous scholar. University of California, Los

Angeles

BIBLIOGRAPHY The following selected bibliography contains those of Jakobson's works mentioned in the text or footnotes: "An Unknown Album Page by Nikolaj Gogol'," Harvard Library Bulletin, 20 (1976). With B. Aroutunova. "Analyse du poème Revedere de Mihail Eminescu," Cahiers de linguistique théorique et appliquée, 1 (1962), 47-54. With B. Casacu. "Arbeiten über die öechische dichterische Form," Slavische Rundschau, 4 (1932), 275-79. "[Closing statement:] Linguistics and Poetics," Style in Language (New York, 1960), 350-77. "Cesky vliv na staropolské pisemnictvi," Co daly nase zemë Evropë a lidstvu, 1 (Prague, 1939), 48-51. (In 2nd ed. [Prague, 1940], title is "Cesky vliv na stiedovëkou literaturu polskou.") "Der grammatische Bau des Gedichts von B. Brecht 'Wir sind sie'," Beiträge zur Sprachwissenschaft, Volkskunde und Literaturforschung, W. Steinitz dargebracht (Berlin, 1965), 175-89. "Die Reimwörter Cech - Lech," Slavische Rundschau, 10, no. 6 (1938), 10-15. "K lingvisticeskomu analizu russkoj rifmy," Studies in Russian Philology, 1 [=Michigan Slavic Materials, 1] (Ann Arbor, 1962), 1-13. "K popisu Mâchova verëe," Torso a tajemstvl Mdchova dila (Prague, 1938), 207-78. "Ktoz jsu bozi bojovnici: slovni stavba husitského chorâlu," International Journal of Slavic Linguistics and Poetics, 7 (1962), 108-17. '"Les chats' de Charles Baudelaire," VHomme, 2 (1962), 5-21. With C. Lévi-Strauss. "Les oxymores dialectiques de Fernando Pessoa," Langages, 12 (1968), 9-27. With L. S. Picchio. "Lettre à Haroldo de Campos sur la texture poétique de Martin Codax," Change, 6 (1970), 53-59.

528

DEAN S. WORTH

"Medieval Mock Mystery (The Old Czech Unguentarius)," Studia philologica et litteraria in honorem L. Spitzer (Bern, 1958), 245-65. Nejstarsi íeské pisné duchovnl [=Národni knihovna, 6] (Prague, 1929), 48 pp. Novejsaja russkaja poézija (Prague, 1921), 68 pp. O cesskom stixe preimuscestvenno v sopostavlenii s russkim [=Sbornik po teorii poétiéeskogo jazyka, 5] (Berlin-Moscow, 1923), 170 pp. "O sootnoSenii mezdu pesennoj i razgovornoj narodnoj reS'ju," Voprosy jazykoznanija, 11,3 (1962), 87-90. "Ob odnosloznyx slovax v russkom stixe," Slavic Poetics. Essays in Honor of Kiril Taranovsky (The Hague-Paris, 1973), 239-52. "On Russian Fairy Tales," Russian Fairy Tales (New York, 1945), 631-56. "On the Role of Word Pitch in Norwegian Verse," Lingua, 11 [=Studia gratulatoria dedicated to A. W. de Groot] (1962), 205-16. With A. Sommerfelt. "On the Verbal Art of William Blake and Other Poet-Painters," Linguistic Inquiry, 1 (1970), 3-23. "Poézija grammatiki i grammatika poezii," Poetics, Poetyka, Poétika (Warsaw, 1961), 397-417. "Postscriptum" [to his] Questions de poétique (Paris, 1973), 485-504. "'Przeszlosc' Cypriana Norwida," Pamigtnik literacki, 54 (1962), 449-56. "Razbor tobol'skix stixov Radiséeva," XVIII vek, 7 [=Rol' i znacenie literatury XVIII veka v istorii russkoj kul'tury ( = Berkov-Festschrift)] (Leningrad, 1966), 228-36. Shakespeare's Verbal Art in "Th'Expence of Spirit" (The Hague-Paris, 1970), 33 pp. With L. G. Jones. "Slavoslovie Siluana Simeonu," Xenia Slavica. Papers Presented to Gojko Ruzicic on the Occasion of His Seventy-fifth Birthday, 2 February 1969 (The Hague-Paris, 1975), 75-83. "StaroéeSskie stixotvorenija, slozennye odnorifmennymi ietverostiSijami (8a-4)," Slavia, 3 (1924), 272-315. "Stroka Maxi o zove gorlicy," International Journal of Slavic Linguistics and Poetics, 3 (1960), 89-108. "Struktura dveju srpskohrvatskih pesama," Zbornik za filologiju i lingvistiku, 4-5 (1962), 131-39. "Struktura na poslednoto Botevo stixotvorenie," Ezik i literatura, 16 (1961), 1-14. "Studies in Comparative Slavic Metrics," Oxford Slavonic Papers, 3 (1952), 21-66. "Subliminal Verbal Patterning in Poetry," Studies in General and Oriental Linguistics (Hattori-Festschrift) (Tokyo, 1970), 302-08. "The Grammatical Structure of Janko Král"s Verses," Sbornlk filozofickej fakulty University Komenského, 16 (1964), 29-40. "The Grammatical Texture of a Sonnet from Sir Phillip Sidney's 'Arcadia'," Studies in Language and Literature in Honor of M. Schlauch (Warsaw, 1966), 165-74. "The Kernel of Comparative Slavic Literature," Harvard Slavic Studies, 1 (1953), 1-71. "The Modular Design of Chinese Regulated Verse," Échanges et communications. Mélanges offerts á Claude Lévi-Strauss (The Hague-Paris, 1970), 597-605. "Úvahy o básnictví doby husitské," Slovo a slovesnost, 2 (1936), 1-21. "Vers starocesky," Ceskoslovenská vlastivéda, 3 (Prague, 1934), 429-59. "Vlijanie narodnoj slovesnosti na Trediakovskogo," Selected Writings, IV: Slavic Epic Studies (The Hague-Paris, 1966), 613-33. Written in 1915. " Vocabulorum constructio in Dante's Sonnet 'Se vedi li occhi miei'," Studi Danteschi, 43 (1966), 7-33. With P. Valesio. "Z déjin staroceského zpévniho básnictví," Slovo a slovesnost, 4 (1938), 41-44. Základy ceského verse (Prague, 1926), 140 pp.

THE STUDY OF RHYME

529

NOTES 1 E.g. the comparison of Jakobson's views of Russian rhyme with those of ¿irmunskij and TomaSevskij as outlined in D. S. Worth, "On Eighteenth-century Russian Rhyme," Russian Literature, 3 (1973), 47-74. Publications cited in the footnotes without an author are by Jakobson. In a few cases, the titles have been slightly abbreviated; the full references are given in the appended bibliography. 2 Cf. for example "Linguistics and Poetics," 368; "Struktura na poslednoto Botevo stixotvorenie," 12-13. 3 V. M. ¿irmunskij, Rifma, ee istorija i teorija (Petersburg, 1923). 4 "Der grammatische Bau des Gedichts 'Wir sind sie'," 400. 5 Novejsaja russkajapotzija(1921), 48,62; "Klingvisticeskomu analizu russkoj rifmy" (1962), 12 and passim. 6 " K lingvisticeskomu analizu russkoj rifmy," 12. 7 Zaklady ceskeho verse, 65-66. 8 "K lingvistideskomu analizu russkoj rifmy," 9, 11. 9 "Vers starocesky," 453-57. 10 "K lingvisticeskomu analizu russkoj rifmy," 13. As to the second type of correspondence, it is not necessarily only the marked members of the opposition which can form rhyme sets, as proved by Jakobson's own illustration of Nitsch's Law (boty - boky, with two [-voiced] consonants; redefining the opposition as [+tense] will not save the situation when faced e.g. with the fact that Slovak rhymes can still be formed by members of either series /p 11' k c c s 5 x f/, /b d d' g 3 3 z z h v/; cf. A. V. Isaienko, "Iz nabljudenij nad 'novoj rifmoj'," Slavic Poetics: Essays in Honor of Kiril Taranovsky, ed. R. Jakobson et al. [The Hague-Paris, 1973], 219). Jakobson's statement should be modified to read, "the rhyme is restricted to either the marked or the unmarked member " 11 "K lingvisticeskomu analizu russkoj rifmy," 11-12. 12 "Ob odnosloznyx slovax v russkom stixe," 243. 13 " 'Les chats' de Charles Baudelaire," 6. Another way of putting this would be to say that the masc.-fem. distinction is preserved in the underlying structure but transcoded by a phonetic rule. 14 "VerS starocesky," 453. 15 "Vlijanie narodnoj slovesnosti na Trediakovskogo," 615. 18 "Stroka Maxi o zove gorlicy," 92. 17 Novejsaja russkaja poezija, 55-56. The general shape of such pairs is A + A ' , where A ' = A except for /m/ substituted for the first phoneme of A (gogoF - mogoV, gusli musli, etc.), unless A itself begins with a labial, in which case A' begins with a dental or s and the order of elements is reversed (turja - burja, salovat' - balovat'); the secondary labial /, unknown to Old Russian, is treated as non-labial (figli - migli). Cf. similar pairs (Cz. souka - louka, Pol. lap - cap, Russ. cuxo - luxo) in "Die Reimworter tech-Lech." 18 "Ktoz jsu bozi bojovnici," 110. 19 "Vers starocesky," 454. Cf. also I ty take se nedvofe / otevfi mi spiese dvefe adduced in "tJvahy o basnictvi doby husitske," 14. 20 "VerS starocesky," 434 and passim. 21 " Vocabulorum constructs," 20-21, 30-31. 22 '"Les chats' de Charles Baudelaire," 11,9. 23 "Struktura dveju srpskohrvatskih pesama," 136. 24 "Slavoslovie Siluana Simeonu," 80. 25 "Razbor tobol'skix stixov Radis5eva," 229 ff.; NB. also the corresponding diffuse vowels / u / - / y / in Kvtda-byvalo and /j/ corresponding to sharped /I,/ in /jedu//sl,edu/.

530 26

DEAN S. WORTH

"'Przeszloác' Cypriana Norwida," 455. "VerS staroéesky," 448; I have added the superscript numerals. The emphasis is Jakobson's, but note also the syntactic and grammatical parallelism of the end-plusinternal rhymes (i pracny inevzacny; v nezamosti, v nemilosli)'a.nd the enriching sound parallels of ne - ne in 11. 2,4 (nevzacny - nemilosti), of stressed a in the beginning of 11. 1, 3 (ach - dalekot'), of z . . . m - z... m in 11. 2, 3 (ciziem kraji - v neznamosti), the metathetic liquid-velar clusters of 11. 2, 4 (kraji - lka/0, etc. Although these additional correspondences were not pointed to specifically by Jakobson in this particular instance, it seems fair to claim that, were it not for his pathbreaking example, the rest of us would never think to look for such things. I am reminded of a 1953 Harvard seminar in which Jakobson remarked that the hardest part of philology is to see what is right there on the page in front of one. 28 "On the Role of Word-Pitch in Norwegian Verse," 208. 29 "Ver5 staroCesky," 456. 30 "The Modular Design of Chinese Regulated Verse," 599, 602-04. 31 "Studies in Comparative Slavic Metrics," 31. 32 "VerS staroCesky," 433. 33 "K popisu Máchova verSe," 236. 34 Základy íeského verse, 69; cf. 524 below. 35 "Linguistics and Poetics," 367. 36 "The Kernel of Comparative Slavic Literature," 9. 37 "Postscriptum," 499-500. Jakobson's work on non-Slavic languages has shown how generally pervasive grammatical patterning is in all poetry, which leads to the conclusion that the "decisive role" of grammatical rhyme in Slavic poetry is merely a reflection of the highly flexional structure of the languages themselves and not a characteristic of Slavic poetics as such. 38 '"Les chats' de Charles Baudelaire," 7, 11, 16-17. 39 "Medieval Mock Mystery," 253. 40 " Vocabidorum constructio," 14. 41 "Razbor tobol'skix stixov RadiSCeva," 232-33. 42 "Der grammatische Bau des Gedichts 'Wir sind sie\" 397. Here a question could be raised: if the poetic architectonics of rhyme and of grammar are independent of each other, is their interrelation there where both occur anything more than coincidental? And, if the absence of rhyme makes grammatical architecture stand out more clearly, why is it the presence of grammatical patterning which strengthens the rhyme structure ? 43 "VerS staroCesky," 454; this is not surprising, since the presence of a grammatical ending automatically adds at least one syllable. 44 Ibid., 455. 45 "StaroCeSskie stixotvorenija," 279, 281. 46 "VerS staroCesky," 455. The classification could be refined by specifying whether the "included" portions of the rhymes in types (2) and (3) are identical morphologically as well as phonologically, e.g. by distinguishing the type piti - oslepiti from, e.g., plti vypiti. 47 Novejsaja russkajapoézija, 43-44. The actual situation is somewhat more complex: in paregmenon one portion of both form and meaning is identical in both members of the rhyme set (the root, the stem where the latter contains a root plus a derivational suffix or other thematic element, etc.), whereas in completely anti-grammatical rhyme only the sound shapes are identical (e.g. Russ. sox past masc. of soxnut' 'to dry up' sox gen. pi. of soxá 'plough'). Jakobson's remarks were of course not intended to exhaust the subject, which has still to receive adequate attention. 48 "Poézija grammatiki i grammatika poezii," 408. 49 "The Kernel of Comparative Slavic Literature," 9. 27

THE STUDY OF RHYME 50

531

"Linguistics and Poetics," 368. " K lingvistiöeskomu analizu russkoj rifmy," 5. In fact, the term "agrammatical" appears in some of Jakobson's early work, e.g. "Verä staroiesky," 455 (where it is applied to non-homonymic rhyme). 52 Grammatical elements are seldom totally absent from a poet's rhyme, and must be viewed from the point of view of whether they rhyme primarily with each other (pure grammatical rhyme, e.g. duSoj - rukoj, letajut - sverkajut) or with derivational or lexical morphemes or segments thereof (lexicogrammatical rhyme) (prax - oiax, sprosjas'knjaz'). Pure grammatical rhyme itself varies widely from author to author and school to school, depending on whether the sound envelop and the grammatical categories are identical (Kurganov - bajanov) or differ partially (na poitovyx [loc. pi.] - svoix rodnyx [gen pi.]) or completely (udar [acc. sing] - tatar [gen. pi.]); furthermore, the rhyme can involve the above (and other!) types of grammatical relations coupled with a variety of lexical correspondences of both form and meaning (strast'ju - vlast'ju, supruga [nom. sing, fem.] - supruga (gen. sing, masc.], etc.). We shall return to these matters elsewhere. 53 "Linguistics and Poetics," 367. The latter statement is true only if "involves" is glossed to include both "emphasizes" and "deliberately downplays," since, as Jakobson himself has shown, rhymes can have a primarily phonic, i.e. asemantic function (see the discussion of tri- and quadrisyllabic rhymes in the Catherine Legend, p. 519 above). 54 "Stroka Maxi o zove gorlicy," 104. 65 "Arbeiten über die Cechische dichterische Form," 507-508 (contra Jirdt, who also confused phonemes with letters). 86 "Postscriptum," 497-98. 67 "Linguistics and Poetics," 367. 58 "Medieval Mock Mystery," 259. 59 "Üvahy o bäsnictvi doby husitskd," 14. 60 "Medieval Mock Mystery," 259. 81 "On Russian Fairy Tales," 647. 82 Shakespeare's Verbal Art, 20. 83 "Lettre ä Haroldo de Campos," 55. 84 "Vocabulorum constructio," 13, 20. 85 '"Les chats' de Charles Baudelaire," 21. Less inspired students may doubt whether the grammatical agreement categories labeled "masculine" and "feminine" have any rhyme structures for which the same labels are used connection with the J - and — - except, perhaps, in the mind of the investigator long familiar with both uses of the terms. 88 "VerS staroiesky," 437-38, 447-48. 67 " Z dSjin starocesk£ho spevniho bäsnictvi," 42. 88 "Ktoz jsu bozi bojovnici ?" 115. 89 "The Grammatical Structure of Janko Kxäl' 's Verses," 33-34. 70 "The Modular Design of Chinese Regulated Verse," 598-99. 71 " Vocabulorum constructio," 28. 72 "On the Verbal Art of William Blake," 3 if. 73 Shakespeare's Verbal Art, 21, 25. 74 "Poezija grammatiki i grammatika poezii," 412. 75 "An Unknown Album Page by Nikolaj Gogol'," 242. 78 Cf. for example Novejsaja russkaja poizija (1921), 64 or "Postscriptum" (1973), 498-500. 77 "The Kernel of Comparative Slavic Literature," 8-9; cf. also "Linguistics and Poetics," 374. 78 "Studies in Comparative Slavic Metrics," 31. 51

532 79

DEAN S. WORTH

Ibid., 59. "Die Reimworter Cech - Lech," 13 f.; cf. his analysis of gogol' - mogoV type repetitions in Novejsaja russkaja poizija, 55-56 fn. 81 This is true not only as far as rhyme itself is concerned. Jakobson's intellectual biography, when it is written, will probably show that his career-long concern with the interconnections of phonological and grammatical patterns had its origins in his early studies of Old Czech verse. 82 "VerS staroCesky," 453-54. 83 Ibid., 453-57; Nejstarsi cesfce pisrte duchovni, 30 ff. 84 "StaroCeSskie stixotvorenija," 288 fn. 85 Nejstarsi ieske pisni duchovni, 20. 86 "VerS staroiesk?," 430. 87 Nejstarsi ieske pisni duchovni, 10; on rhyme in the Svaty Vdclave see ibid., 30. 88 "Verg staroCesky," 454-55. 89 "StaroCeSskie stixotvorenija," 279. 90 "Ver5 staroCesky," 443 fF. 91 "Uvahy o bdsnictvi doby husitske," 19-20; Nejstarsi ceske pisni duchovni, 20. 92 "VerS starodesky," 430. 93 "StaroCeSskie stixotvorenija," 284 fn. 94 "StaroieSskie stixotvorenija," 298. 95 Ibid., 292-93. Rhyme is often a key to interlingual borrowings as well. For example, the oldest example of disyllabic rhyme is in the Ostrovskd pisen (latter thirteenth century) and imitates Latin verse of the time, while certain Old Polish texts contain rhymes which indicate borrowings from Old Cz., e.g. Old Pol. biskupi - skqpi, cf. Old Cz. skupi("Ceskjf vliv na stfedovekou literaturu polskou," 50); borrowed rhymes appear more frequently in songs than in other verse forms ("O sootnoSenii mezdu pesennoj i razgovornoj reC'ju," 88). 96 "VerS staroCesky," 453-55; Nejstarsi ieske pisni duchovni, 38; "StaroCeSskie stixotvorenija," 302-03. 97 "Ovahy o bdsnictvi doby husitskd," 14, 20. 98 O iesskom stixe, 57 ff.; Zaklady ceskeho verse, 65-66. In the latter study, Jakobson shows that quantitative non-correspondence is more frequent in auslaut than elsewhere and more usual with e-e than with other vowels [why?-DSW]. Another feature distinguishing Romantic from later poets is the latter's use of rhyming support consonants in open masculine rhymes {Zaklady ceskeho verse, 71). 99 Zaklady ieskeho verse, 69; O iesskom stixe, 60-61. This line-end law in Czech mirrors the restrictions established by Jakobson on the choriambic onset in nineteenthcentury Russian verse, cf. pp. 516, 524; one notes also the parallel with the zero - j equivalence of eighteenth and nineteenth-century Russian rhyme (volny - polnyj, etc.). 100 "k. popisu M&chova verSe," passim (e.g. 238-44). 101 E.g. on Vrchlicky and RabeS, "K popisu Machova verSe," 222, 263-64. 102 " o b odnosloznyx slovax," 242-43. Cf. the similar role of word-boundaries at lineend in Czech rhyme, p. 524 above. 103 o iesskom stixe, 101-02. These early comments, echoing those of Belyj on Blok, were perhaps somewhat oversimplified, as eighteenth-century comments on the "soft Moscow pronunciation" (i.e. akanje) make clear. 104 "viijanie narodnoj slovesnosti na Trediakovskogo," 624-31; O iesskom stixe, 112. ins Uncharacteristically, Jakobson fails to catch the grammatical pun Bog(om) dan(nyj) - Bog dal. log "viijanie narodnoj slovesnosti na Trediakovskogo," 624-31; NB. the metathetic supporting consonants (bra/a - rbata) which precede by 150 years the metathetic rhymes Jakobson notes in Novejsaja russkaja poizija, 62. 80

THE STUDY OF RHYME

533

107 "Subliminal Verbal Patterning in Poetry," 304. The sound patterning of this riddle is even more complex than Jakobson noted; cf. for example the sharping oppositions of zverè'k - sveisôk and xvdst' - vé'rst and the consonant-zero pairs in zverëk - îoerSok and xuost' - oërst. 108 Novejsaja russkaja poèzija, 61. Such generalizations have lost none of their force in more than half a century. 109 NovejSaja russkaja poèzija, 62. Recent research indicates that some of these "modern" features are widespread in eighteenth-century verse and may even originate in syllabic virsi. The rhyming of words with different stresses is, like the weakening of the sharping opposition noticed by Jakobson, indicative of the anti-tonality orientation of modern rhyme. 110 Novejsaja russkaja poèzija, 62-64. One can extend Jakobson's remarks to conclude that the semantic and the organizing functions of rhyme are in complementary distribution, as are rhyme and meter as organizing constants. i n "xhe Grammatical Texture of a Sonnet," 166. 112 "Linguistics and Poetics," 372. u s " T h e verbal Art of William Blake," 3 ff. 114 Ibid.; Shakespeare's Verbal Art, 21, 25. 115 "Linguistics and Poetics," 357. 116 "Struktura na poslenoto Botevo stixotvorenie," 12-13. 117 "The Modular Design of Chinese Regulated Verse," 598-99. 118 "Struktura dveju srpskohrvatskih pesama," 130, 136. 119 "'Les chats' de Charles Baudelaire," 6, 11 etc.; "On the Verbal Art of William Blake," 11, 13 (on Rousseau-Douanier). 120 "Der grammatische Bau des Gedichts 'Wir sind sie'," 400. 121 " Vocabulorum constructs," 14, 20-21, 28, 30-31. 122 "O sootnoSenii mezdu pesennoj i razgovornoj narodnoj rec'ju," 529. 123 "On the Role of Word-Pitch in Norwegian Verse," 208. 124 "Subliminal Verbal Patterning in Poetry," 304; "'Przeszlosc' Cypriana Norwida," 452-55. 125 "Lettre à Haroldo de Campos," 54-55, 59. 126 "Analyse du poème Revedere de Mihail Eminescu," 49. 127 "Slavoslovie Siluana Simeonu," 80. 128 "xhe Grammatical Structure of Janko Krâl"s Verses," 31, 33-34.