·

·

Permanent Dynamic Synchrony (1)

Excerpts on PDS (Permanent Dynamic Synchrony) selected from the writings of Roman Jakobson.


Jakobson, Roman 1962[1928a]. The concept of the sound law and the teleological criterion. In: Rudy, Stephen (ed.), Selected Writings I: Phonological Studies. s-Gravenhage: Mouton, 1-2. Online access: https://archive.org/details/selectedwritings01jako/page/n15/mode/2up

The basic assumption of the neo-grammarian linguistic methodology, that of the sound law operating without exceptions in a given language at a given time, has, up till recently, repeatedly met with negative criticism, since the neo-grammarians have not been able to give a theoretical foundation for this working hypothesis. The revision of the traditional tenet leads to the recognition of the fact that language (and in particular its sound system) cannot be analyzed without taking into account the purpose which that system serves. Once this amendment is made, the objections to the doctrine of the sound law lose their validity.

  1. The idea of a sound law operating without exceptions in a given language must be limited to a linguistic system characterized by one and the same function, i.e., to linguistic entities which are functionally equivalent.
  2. The neo-grammarians did not succeed in explaining the social character of sound change (why a speech community accepts and sanctions individual slips), but this problem too finds its solution once it is posed teleologically. The same requirement applies if one attributes the decisive role in sound changes to the succession of generations.
  3. The overlapping between territorial, socially or functionally distinct linguistic patterns can be fully comprehended only from a teleological point of view, since every transition from one system to another necessarily bears a linguistic function.

The first attempts at a goal-directed interpretation of sound changes, in particular their explanation with reference to the law of the economy of energy or to fashion and esthetic factors, are one-sided and greatly oversimplify the problem. It is impossible to deal with the sound of a given language without regard to its phonological system, i.e., to the repertory of meaningful distinctions among the acoustico-motor images proper to the given language. (Jakobson 1962[1928a]: 1)

["The concept of the sound law and the teleological criterion"] [Online access] ["Published in Czech in Časopis pro moderni filologii, XIV (Prague, March 1928), as a "brief extract" from a paper delivered in the Prague Linguistic Circle, January 13, 1927."]

Jakobson, Roman; Tynjanov, Jurij 1981[1928d]. Problems in the study of language and literature. In: Rudy, Stephen (ed.), Selected Writings III: Poetry of Grammar and Grammar of Poetry. The Hague; Paris; New York: Mouton, 3-6. Online access: https://archive.org/details/selectedwritings0003jako/page/n23/mode/2up

1. The immediate problems facing Russian literary and linguistic science demand a precise theoretical platform. They require a firm dissociation from the increasing mechanistic tendency to past together mechanically the new methodology and old obsolete methods; they necessitate a determined refusal of the contraband offer of naive psychologism and other methodological hand-me-downs in the guise of new terminology.

Furthermore, academic eclecticism and pedantic "formalism" - which replaces analysis by terminology and the classification of phenomena - and the repeated attempts to shift literary and linguistic studies from a systematic science to episodic and anecdotal genres should be rejected.

2. The history of literature (art), being simultaneous with other historical series, is characterized, as is each of these series, by an involved complex of specific structural laws. Without an elucidation of these laws, it is impossible to establish a scientific manner the correlation between the literary series and other historical series.

3. The evolution of literature cannot be understood until the evolutionary problem ceases to be obscured by questions about episodic, nonsystemic genesis, whether literary (for example, so-called "literary influences") or extraliterary. The literary and extraliterary material used in literature may be introduced into the [|4|] orbit of scientific investigation only when it is considered from a functional point of view.

4. The sharp opposition of synchronic (static) and diachronic cross sections has recently become a fruitful working hypothesis, both for linguistics and for history of literature, inasmuch as it has demonstrated that language, as well as literature, has a systemic character at each individual moment of its existence. At the present time, the achievements of the synchronic concept force us to reconsider the principle of diachrony as well. The idea of a mechanical agglomeration of material, having been replaced by the concept of a system or structure in the realm of synchronic study, underwent a corresponding replacement in the realm of diachronic study as well. The history of a system is in turn a system. Pure synchronism now proves to be an illusion: every synchronic system has its past and its future as inseparable structural elements of the system: (a) archaism as a fact of style; the linguistic and literary background recognized as the rejected old-fashioned style; (b) the tendency toward innovation in language and literature recognized as a renewal of the system.

The opposition between synchrony and diachrony was an opposition between the concept of system and the concept of evolution; thus it loses its importance in principle as soon as we recognize that every system necessarily exists as an evolution, whereas, on the other hand, evolution is inescapably of a systemic nature.

5. The concept of a synchronic literary system does not coincide with the naively envisaged concept of a chronological epoch, since the former embraces not only works of art which are close to each other in time but also works which are drawn into the orbit of the system from foreign literatures or previous epochs. An indifferent cataloguing of coexisting phenomena is not sufficient; what is important is their hierarchical significance for the given epoch. [|5|]

6. The assertion of two differing concepts - la langue and la parole - and the analysis of the relationship between them (the Geneva school) has been exceedingly fruitful for linguistic science. The principles involved in relating these two categories (i.e., the existing norm and the individual utterances) as applied to literature must be elaborated. In this latter case, the individual utterance cannot be considered without reference to the existinc complex of norms. (The investigator, in isolating the former from the latter, inescapably deforms the system of artistic values under consideration, thus losing the possibility of establishing its immanent laws.)

7. An analysis of the structural laws of language and literature and their evolution inevitably leads to the establishment of a limited series of actually existing structural types (and, correspondingly, of types of structural evolution).

8. A disclosure of the immanent laws of the history of literature (and language) allows us to determine the character of each specific change in literary (and linguistic) systems. However, these laws do not allow us to explain the tempo of evolution or the chosen path of evolution when several, theoretically possible, evolutionary paths are given. This is owing to the fact that the immanent laws of literary (and, corresponding, linguistic) evolution form an indeterminate equation; although they admit only a limited number of possible solutions, they do not necessarily specify a unique solution. The question of a specific choice of path, or at least of the dominant, can be solved only through an analysis of the correlation between the literary series and other historical series. This correlation (a system of systems) has its own structural laws, which must be submitted to investigation. It would me methodologically fatal to consider the correlation of systems without taking into account the immanent laws of each system. (Jakobson & Tynjanov 1981[1928d]: 29-30)

["Problems in the study of language and literature"] [Online access] ["Written in Russian during Jurij Tynjanov's visit to Prague in the winter of 1928 and first publisted, under the title "Problemy izučenija literatury i jazyka", in Novyj Lef, 12 (1928), pp. 36-37."]

Jakobson, Roman 1962[1931d]. Phonemic notes on standard Slovak. In: Rudy, Stephen (ed.), Selected Writings I: Phonological Studies. s-Gravenhage: Mouton, 221-230. Online access: https://archive.org/details/selectedwritings01jako/page/220/mode/2up

The close linguistic, cultural, political and ethnographic communality and interconnection of the Czechs and Slovaks is beyond all doubt. If, however, we wish to arrive at a thoroughgoing characterization of a whole, we must pay attention not only to its unifying features, but also to the peculiarities of each of its individual parts.

It is the purpose of these pages to outline a number of properties which differentiate the phonemic pattern of standard Slovak from that of standard Czech. Just as there are differential dictionaries of Czech and Slovak, this will be a contribution to a differential phonology of the two languages. I am here concerned solely with the standard language, that is with those elements which have been incorporated into the normative textbooks and codified as a set of valid orthoepic prescriptions. I leave aside such questions as how large is the number of educated individuals who consistently observe the literary norm and how widespread are the deviations therefrom, i.e., the dialectal variants of the standard language. (Jakobson 1962[1931d]: 221)

["Phonemic notes on standard Slovak"] [Online access] ["Written in Prague and published in Czech in the Studies presented to Albert Pražák - Slovenská Miscellanea (Bratislava, 1931)."]

Jakobson, Roman 1981[1971f]. The dominant. In: Rudy, Stephen (ed.), Selected Writings III: Poetry of Grammar and Grammar of Poetry. The Hague (etc.): Mouton de Gruyter, 751-756. Open access: https://archive.org/details/selectedwritings0003jako/page/750/mode/2up

The first three stages of Formalist research have been briefly characterized as follows: (1) analysis of the sound aspect of a literary work; (2) problems of meaning within the framework of poetics; (3) integration of sound and meaning into an inseparable whole. During this latter stage, the concept of the dominant was particularly fruitful; it was one of the most crucial, elaborated, and productive concepts in Russian Formalist theory. The dominant may be defined as the focusing component of a work of art; it rules, determines, and transforms the remaining components. It is the dominant which guarantees the integrity of the structure.

The dominant specifies the work. The specific trait of bound language is obviously its prosodic pattern, its verse form. It might seem that this is simply a tautology: verse is verse. However, we must constantly bear in mind that the element which specifies a given variety of language dominates the entire structure and thus acts as its mandatory and inalienable constituent dominating all the remaining elements and exerting direct influence upon them. However, verse in turn is not a simple concept and not an indivisible unit. Verse itself is a system of values; as with any value system, it possesses its own hierarchy of superior and inferior values and one leading value, the dominant, without which (within the framework of a given literary period and a given artistic trend) verse cannot be conceived and evaluated as verse. For example, in Czech poetry of the fourteenth century the inalienable mark of verse was not the syllabic scheme but rhyme, since there existed poems with unequal numbers of syllables per line (termed "measureless" verses), which nevertheless were conceived as verses, whereas unrhymed verses were not tolerated during that period. On the other hand, in Czech Realist poetry of the second half of the nineteenth century, rhyme was a dispensable device, whereas the syllabic scheme was a mandatory, inalienable component, without which verse was not verse; from the point of view of that school, free verse was judged as unacceptable arrhythmia. For the present-day Czech brought [|752|] up on modern free verse, neither rhyme nor a syllabic pattern is mandatory for verse; instead, the mandatory component consists of intonational integrity - intonation becomes the dominant of verse. If we were to compare the measured regular verse of the Old Czech Alexandriade, the rhymed verses of the Realist period, and the rhymed measured verse of the present epoch, we would observe in all three cases the same elements - rhyme, a syllabic scheme, and intonational unity - but a different hierarchy of values - different specific mandatory, indispensable elements; it is precisely these specific elements which determine the role and the structure of the other components.

We may seek a dominant not only in the poetic work of an individual artist and not only in the poetic canon, the set of norms of a given poetic school, but also in the art of a given epoch, viewed as a particular whole. For example, it is evident that in Renaissance art such a dominant, such an acme of the aesthetic criteria of the time, was represented by the visual arts. Other arts oriented themselves toward the visual arts and were valued according to the degree of their closeness to the latter. On the other hand, in Roman art the supreme value was assigned to music. Thus, for example, Romantic poetry oriented itself toward music: its verse is musically focused; its verse intonation imitates musical melody. This focusing on a dominant which is in fact external to the poetic work substantially changes the poem's structure with regard to sound texture, syntactic structure, and imagery; it alters the poem's metrical and strophical criteria and its composition. In Realist aesthetics the dominant was verbal art, and the hierarchy of poetic values was modified accordingly.

Moreover, the definition of the artistic work as compared to other sets of cultural values substantially changes, as soon as the concept of the dominant becomes our point of departure. For example, the relationship between a poetic work and other verbal messages acquires a more exact determination. Equating a poetic work with an aesthetic, or more precisely with a poetic, function, as far as we deal with verbal material, is characteristic of those epochs which proclaim self-sufficient, pure part, l'art pour l'art. In the early steps of the Formalist school, it was still possible to observe distinct traces of such an equation. However, this equation is unquestionably erroneous: a poetic work is not confined to aesthetic function alone, but has in addition many other functions. Actually, the intentions of a poetic work are often closely related to to philosophy, social didactics, etc.. Just as a poetic work is not exhausted by its aesthetic function, similarly aesthetic function is not limited to the poetic work; an orator's address, everyday conversation, newspaper articles, advertisements, [|753|] a scientific treatise - all may employ aesthetic considerations, give expression to aesthetic function, and often use words in and for themselves, not merely as a referential device.

In direct opposition to the straight monistic point of view is the mechanicistic standpoint, which recognizes the multiplicity of functions of a poetic work and judges that work, either knowingly or unintentionally, as a mechanical agglomeration of functions. Because a poetic work also has a referential function, it is sometimes considered by adherents of the latter point of view as a straightforward document of cultural history, social relations, or biography. In contrast to one-sided monism and one-sided pluralism, there exists a point of view which combines an awareness of the multiple functions of a poetic work with a comprehension of its integrity, that is to say, that function which unites and determines the poetic work. From this point of view, a poetic work cannot be defined as a work fulfilling neither an exclusively aesthetic function nor an aesthetic function along with other functions; rather, a poetic work is defined as a verbal message whose aesthetic function is its dominant. Of course, the marks disclosing the implementation of the aesthetic function are not unchangeable or always uniform. Each concrete poetic canon, every set of temporal poetic norms, however, comprises indispensable, distinctive elements without which the work cannot be identified as poetic.

The definition of the aesthetic function as the dominant of a poetic work permits us to determine the hierarchy of diverse linguistic functions within the poetic work. In the referential function, the sign has a minimal internal connection with the designated object, and therefore the sign in itself carries only a minimal importance; on the other hand, the expressive function demands a more direct, intimate relationship between the sign and the object, and therefore a greater attention to the internal structure of the sign. In comparison with referential language, emotive language, which primarily fulfills an expressive function, is as a rule closer to poetic language (which is directed precisely toward the sign as such). Poetic language and emotional language often overlap each other, and therefore these two varieties of language are often quite erroneously identified. If the aesthetic function is the dominant in a verbal message, then this message may certainly use many devices of expressive language; but these components are then subject to the decisive function of the work, i.e., they are transformed by its dominant.

Inquiry into the dominant had important consequences for Formalist views of literary evolution. In the evolution of poetic form it is not so much a question of the disappearance of certain elements and the emergence [|754|] of others as it is the question of shifts in the mutual relationship among the diverse components of the system, in other words, a question of the shifting dominant. Within a given complex of poetic norms in general, or especially within the set of poetic norms valid for a given poetic genre, elements which were originally secondary become essential and primary. On the other hand, the elements which were originally the dominant ones become subsidiary and optional. In the earlier works of Šklovskij, a poetic work was defined as a mere sum of its artistic devices, while poetic evolution appears nothing more than a substitution of certain devices. With the further development of Formalism, there arose the accurate conception of a poetic work as a structured system, a regularly ordered hierarchical set of artistic devices. Poetic evolution is a shift in this hierarchy. The hierarchy of artistic devices changes within the framework of a given poetic genre; the change, moreover, affecs the hierarchy of poetic genres, and, simultaneously, the distribution of artistic devices among the individual genres. Genres which were originally secondary paths, subsidiary variants, now come to the fore, whereas the canonical genres are pushed toward the rear. Various Formalist works deal with the individual periods of Russian literary history from this point of view. Gukovskij analyzes the evolution of poetry in the eighteenth century; Tynjanov and Èjxenbaum, followed by a number of their disciples, investigated the evolution of Russian poetry and prose during the first half of the nineteenth century; Viktor Vinogradov studies the evolution of Russian prose beginning with Gogol'; Èjxenbaum treats the development of Tolstoj's prose against the background of contemporaneous Russian and European prose. The image of Russian literary history substantially changes; it becomes incomparably richer and at the same time more monolithic, more synthetic and ordered, than were the membra disjecta of previous literary scholarship.

However, the problems of evolution are not limited to literary history. Questions concerning changes in the mutual relationship between the individual arts also arise, and here the scrutiny of transitional regions is particularly fruitful; for example, an analysis of a transitional region between painting and poetry, such as illustration, or an analysis of a border region between music and poetry, such as the romance.

Finally, the problem of changes in the mutual relationship between the arts and other closely related cultural domains arises, especially with respect to the mutual relationship between literature and other kinds of verbal messages. Here instability of boundaries, the change in the content and extent of the individual domains, is particularly illuminating. Of [|755|] special interest for investigators are the transitional genres. In certain periods such genres are evaluated as extraliterary and extrapoetical, while in other periods they may fulfill an important literary function because they comprise those elements which are about to be emphasized by belles lettres, whereas the canonical literary forms are deprived of these elements. Such transitional genres are, for example, the various forms of littérature intime - letters, diaries, notebooks, travelogues, etc. - which in certain periods (for example, in the Russian literature of the first half of the nineteenth century) serve an important function within the total complex of literary values.

In other words, continual shifts in the system of artistic values imply continual shifts in the evaluation of different phenomena of art. That which, from the point of view of the old system, was slighted or judged to be imperfect, dilettantish, aberrant, or simply wrong or that which was considered heretical, decadent, and worthless may appear and, from the perspective of the new system, be adopted as a positive value. The verses of the Russian late-Romantic lyricists Tjutčev and Fet were criticized by the Realist critics for their errors, their alleged carelessness, etc. Turgenev, who published these poems, thoroughly corrected their rhythm and style in order to improve them and adjust them to the extant norm. Turgenev's editing of these poems became the canonical version, and not until modern times have the original texts been reinstated, rehabilitated, and recognized as an initial step toward a new concept of poetic form. The Czech philologist J. [Král rejected the verse of erben and Čelakovský as erroneous and shabby from the viewpoint of the Realistic school of poetry, whereas the modern era praises these verses precisely for those features which had been condemned in the name of the Realistic canon. The works of the great Russian composer Musorgskij did not correspond to the requirements of musical instrumentation current in the late nineteenth century, and the contemporaneous master of compositional technique, Rimskij-Korsakov, refashioned them in accordance with the prevalent taste of his epoch; however, the new generation has promoted the path-breaking values saved by Musorgskij's "unsophisticatedness" but temporarily suppressed Rimskij-Korsakov's corrections and has naturally removed those retouchings from such compositions as "Boris Godunov".

The shifting, the transformation, of the relationship between individual artistic components became the central issue in Formalist investigations. This aspect of Formalist analysis in the field of poetic language had a pioneering significance for linguistic research in general, since it provided important impulses toward overcoming and bridging the gap between [|756|] the diachronic historical method and the synchronic method of chronological cross section. It was the Formalist research which clearly demonstrated that shifts and change are not only historical statements (first there was A, and then A1 arose in place of A) but that shift is also a directly experienced synchronic phenomenon, a relevant artistic value. The reader of a poem or teh viewer of a painting has a vivid awareness of two orders: the traditional canon and the artistic novelty as a deviation from that canon. It is precisely against the background of that tradition that innovation is conceived. The Formalist studies brought to light that this simultaneous preservation of tradition and breaking away from tradition form the essence of every new work of art. (Jakobson 1981[1971f]: 751-756)

["The dominant"] [Open access] ["From the unpublished Czech text of lectures on the Russian Formalist school delivered at Masaryk University in Brno in the spring of 1935."]

Jakobson, Roman 1971[1953d]. Results of a joint conference of anthropologists and linguists. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 554-567. Online access: https://archive.org/details/selectedwritings02jako/page/556/mode/2up

Returning now to the linguistic functions - I mentioned the emphasis on the topic, on the sender, on the receiver; and we see how many new things we are able to do when analysing this paramount problem of [|558|] sender and receiver. Moreover, there is the possibility of an emphasis either on the code or on the message. This emphasis of the message on its own self is called the poetic function. I am very happy that, if not at this Conference at least at the next, as it was said, this function will form part of the discussion. A Hill's and H. Whitehall's successful seminar on poetic language at this Linguistic Institute is one of the eloquent proofs that the problems of poetic language enter into the foreground of American linguistics. I am happy that, as Whitehall intimates in his excellent pamphlet recently published by the Foreign Service Institute, a bridge is finally being built between linguistics and literary criticism in this country. The proper subject of inquiry into poetry is precisely language, seen from its preponderant function: the emphasis on the message. This poetic function, however, is not confined to poetry. There is only a difference in hierarchy: this function can either be subordinated to other functions or appear as the organizing function. The conception of poetic language as language with a predominant poetic function will help us in understanding the everyday prosaic language, where the hierarchy of functions is different, but where this poetic (or aesthetic) function necessarily exists and plays a palpable role both in the synchronic and diachronic aspect of language. There are instructive border cases: the largest linguistic code unit functions at the same time as the smallest poetic whole, and in this marginal area the research of our friend D. B. Shimkin on proverbs is a fascinating theme, since the proverb is both a phraseological unit and a poetic work.

We mentioned the factors involved in the speech event, but we did not touch upon their interchangeability, the roles of sender and receiver merging or alternating, the sender or the receiver becoming the topic of the message, and the other interactions of all these factors. The most essential problem for speech analysis is that the code common to both sender and receiver and underlying the exchange of messages. No communication is feasible without a certain stock of what the engineers, and especially D. M. MacKay, one of the nearest to linguists among them, call preconceived possibilities and prefabricated representations. When I read all that was written by the communication engineers, especially American and English (in particular E. C. Cherry, D. Gabor, and MacKay), on message and code, I realized of course that both these conjoined aspects have been for a long time familiar to the linguistic and logical theories of language here and abroad under various dichotomous names such as langue/parole, Language/Speech, Linguistic Pattern/Utterance, Legisign/Sinsign, Type/Token, Sign-design/Sign-event, etc. [|559|] But at the same time I must confess that the Code-Message concepts of communication theory are much clearer, must less ambiguous, and much more operational than the traditional presentation of this dichotomy in the theory of language. I believe that it's preferable to work at present with these well-defined, measurable and analysable concepts without replacing them by new, once again somewhat vague terms, such as the "common core".

Communication theory seems to me a good school for present-day linguists, just as structural linguistics is a useful school for communication engineering. I think that the basic reality facing a linguist is the interlocution - the exchange of messages between sender and receiver, between addresser and addressee, between encoder and decoder. There occur attempts to revert ot a very, very old, I should say pre-Whitneyan, stage of our science in considering individual speech as the only reality. As I already mentioned, individual speech doesn't exist without an exchange. There is no sender without a receiver - oh, yes, there is, if the sender is drunk or pathological. As to non-exteriorized, non-uttered, so-called inner speech, it is only an elliptic and allusive substitute for the more explicit, enunciated speech. Furthermore, dialogue underlies even inner speech, as demonstrated from Peirce to L. S. Vygotskij.

With the customary great interest I read the paper on Idiolect, distributed by my old friend C. F. Hockett. This paper confines the idiolect to a single individual's habits of speaking at a given time, not including his habits of understanding the speech of others. If my Cambridge utterances over a long period were observed and tape-recorded, one would never hear me use the word "idiolect". Nevertheless now, when speaking with you, I use it because I am adapting myself to my potential opponents, for instance, to Hockett. I use many other terms in the same way. Everyone, when speaking to a new person, tries, deliberately or involuntarily, to hit upon a common vocabulary: either to please or simply to be understood or, finally, to bring him out, he uses the terms of his addressee. There is no such thing as private property in language: everything is socialized. Verbal exchange, like any form of intercourse, requires at least two communicators, and idiolect proves to be a somewhat perverse fiction.

There are indeed many stimuli to be gained for linguists from the theory of communication. A normal communication process operates with an encoder and decoder. The decoder receives a message. He knows the code. The message is new to him, but, by virtue of this code, he interprets the message. To comprehend this operation we now have the great [|560|] help of psychology. One of the most pleasant experiences we had during this conference was Osgood's brilliant report on the psycholinguistic analysis of decoding and encoding processes.

The receiver understands the message thanks to his knowledge of the code. The position of the linguist who deciphers a language he doesn't know is different. He tries to deduce the code from the message: thus he is not a decoder; he is what is called a cryptanalyst. The decoder is a virtual addressee of the message. The American cryptanalysts who, during the war, read the Japanese secret messages were not the addressees of these messages. Obviously, the linguist must develop the technique of cryptanalysts; and, naturally, when one deals too long with a technique, one begins to believe that it is the normal procedure. But, as a matter of fact, such a procedure is quite marginal and exceptional in usual communication, and even the task of a linguist is to start with the job of the cryptanalyst but to end up as a normal decoder of this language. His ideas is to become like a member of the speech community studied. The cryptanalyst observes allpohones and looks for the phonemes. But the phonemes, the invariants, are much more intimately known to the decoder, the member of the speech community, than are the variations. He doesn't care what the allophones are. He wants to pick out the phonemic contrasts in order to understand the text. (By the way, the terms "allophone" and "contrast" are in my parlance further examples of the verbal adaptation of the speaker to his listeners; otherwies, I would say "variant" and "opposition".)

In the field of interaction between message and code, this Conference has shown great progress. We have discussed here, on various levels, the relation between two participants in speech communication. As we well know, one of the essential duties of language is to bridge space - to span distance - to create a spatial continuity - to find and establish common language through the air. Of course, where distance is involved, there emerge still greater and more numerous dialectal differences. If there eare two neighboring speech-communities, the code is not the same, but still there is no hermetic isolation of either speech community. This may occur only as an abnormal, rather pathological case. As a rule, there is a tendency to understand the members of the other speech community, and we heard the illuminating paper of my tried friend Twaddell, which shos us how such a mechanism works. This is the "code switching" of the communication engineers. W. F. Twaddell always senses not only the problem of present-day linguistics but also the problem of tomorrow. Just as his monograph on defining the phoneme [|561|] was a spur to search for a strictly scientific phonemic analysis, his new paper calls for thorough attention to the focal linguistic problem of code-switching.

We proceed now to the puzzles of bilingualism, graphically discussed by Mary R. Haas and J. B. Casagrande. We are still on the same problem of bridging space. Here almost nothing stems from a common core. The codes become still more and more different. But there is always a certain correspondence, a certain relation between the two codes. There is the possibility of a search for at least a partial understanding, and there are in such relations intelingual mediators, interpreters - bilingual people. Here we reach a very relevant, decisive point. Bilingualism is for me the fundamental problem of linguistics, because the division into departments is artificial - the department of French, the department of Italian, etc. Are the contiguous languages in complete segregation? If there is an iron curtain, we know how easily such a certain is penetrated by various forms of verbal communication. We know that there exist bilingual areas or bilingual groups of speakers, and the sociology of language presents us with interesting accounts of them. Since bilingual people can obviously speak to and influence a higher number of listeners, they consequently have a higher power, a higher prestige. What is then the result? There is an adaptation on the part of the bilingual person from one language to another and a subsequent diffusion of certain phenomena stimulated by bilingual people among non-bilingual people. As was pointed out in that most important paper of A. Sommerfelt's, we face the question of the diffusion of patterns - of phonemic patterns, of grammatical categories, of what Sapir called the grammatical processes. We shall see how enormous this diffusion is when we obtain the Atlas begun in Oslo before the last war, the atlas of such phenomena, cartographed regardless of the boundaries and relationships of the languages carrying these phenomena. I spoke with one of the most sober among linguists, Haas, and with one of the most sober among anthropologysts, Ray. The extent of such a phonemic and grammatical diffusion among neighboring languages of clearly different origin appeared to us so surprising, so difficult to explain that we were unanimous in stressing the urgent need for a systematic, international study of these phenomena. This task does not at all eliminate the problem of genetic kinship, but the problem of affinity is no less important - and, without knowing exactly what affinity is, we will never detect the genetic features.

So much for space. Now we must confront the factor of tme. It was not discussed at this Conference, but it was examined in the luminous [|562|] mimeographed paper by Hill, distributed here. We were accustomed to textbooks advocating a complete split between synchronic and diachronic linguistics. They were presented as two entirely diverse methodologies, two basically different problems. This is, in my opinion, obsolete and I am in complete agreement with the views of Hill: the history of a language can only be the history of a linguistic pattern, a linguistic system, which undergoes different mutations. Each mutation must be analysed from the point of view of the pattern as it was before the mutation and after it. Here we come to an important point. I formulate it in other terms than Hill, but I hope that we will be no less in accord. It seems to me that the great mistake and confusion, the sharp separation between synchrony and diachrony, was to a high degree due to a confusion between two dichotomies. One is the dichotomy of synchrony and diachrony, and the other is the dichotomy between static and dynamic. Synchronic is not equal to static. When at a movie I ask you what you see at a given moment on the screen, you won't see statics - you'll see horses running, people walking, and other movements. Where do you see statics? Only on the billboards. The billboard is static but not necessarily synchronic. Suppose a billboard remains unchanged for a year - that is static. And it is completely legitimate to ask what is static in linguistic diachrony. I'm sure it would interest Hahn if I try to define what has been static, unchangeable, immutable in Slavic from the early Middle Ages, or from Common Indo-European, until the present. This is a static problem but at the same time a diachronic one.

Let us tackle the dynamic problems. I'll use as an example a change I have observed from my childhood: there has occurred a certain salient change in the vowel pattern of contemporary Standard Russian. In the unstressed, especially pretonic position, the two phonemes /e/ and /i/ were distinguished by our grandparents' generation, these two phonemes merged in one /i/. For the intermediate generation, that of our parents, this distinction is optional. What does this mean? It means, the intermediate generation has a code that contains this distinction. When discrimination is needed to avoid ambiguities or to make speech particularly clear, both phonemes are distinguished in pronunciation. But in a slurred, slovenly, so to speak elliptical style, this distinction, along with certain others, may be omitted: speech becomes less explicit. Thus, for a certain time, both the starting point and the finish of a mutation appear to coexist as two stylistic layers and, moreover, when the time factor enters into such a system of symbolic values as language, it becomes a symbol itself and [|563|] may be used as a stylistic means. For instance, when we speak in a more conservative way, we use the more archaic forms. In Moscow Russian, the generation of our parents did not use the distinction between unstressed /e/ and /i/ in familiar talk: rather the newer fashion of fusing both phonemes was followed to produce the impression of being younger than one really was. Suppose that one generation always makes this distinction, and the succeeding one doesn't make it at all. But it can never happen that only one generation exists, and that the whole preceding generation dies on one and the same day. Thus both patterns must for a time coexist, and the receiver of one of them is accustomed to recode the message from the sender of the other generation. Thus a change is, at its beginning, a synchronic fact, and, insofar as we don't wish to oversimplify, the synchronic analysis must encompass linguistic changes, and, vice versa, the linguistic changes can be comprehended only in the light of synchronic analysis. (Jakobson 1971[1953d]: 557-562)

["Results of a joint conference of anthropologists and linguists"] [Online access] ["Concluding report at the Conference of Anthropologists and Linguists, Indiana University, July 21-30, 1952, published in the supplement to Int. Journal of American Linguistics, XIX, No. 2, April, 1953."]

Jakobson, Roman 1971[1953e]. Pattern in Linguistics (Contribution to Debates with Anthropologists). In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 223-228. Online access: https://archive.org/details/selectedwritings02jako/page/224/mode/2up

There are no universal laws of linguistic evolution. The development [|227|] may proceed in different ways, with a single limitation: it cannot lead to a state contradictory to general laws. The general laws, often sought in evolution, as a matter of fact govern any synchronic state. Nor is this tenet confined to linguistics. It corresponds strikingly to the view of quantum mechanics formulated by L. Tisza, its eminent representative in this country: "Quantum mechanics is morphically deterministic, whereas the temporal processes, the transitions between stationary states, are governed by statistical laws." In other words, structural linguistics as well as "quantum mechanics gains in morphic determinism what it loses in temporal determinism."

Here we see an essential prerequisite for further studies in the evolution of linguistic patterns. It must help us to overcome the methodological heritage of the neogrammarian doctrine, the tradition of the late nineteenth century in historical linguistics, even though we accept from this tradition a high number of technical devices. Its bias permitted only a history of scattered facts, piecemeal dealings with separate phenomena, without reference to the pattern to which these components belonged, and therefore the real evolution of the phonemic or grammatical structure was overlooked.

In analyzing a change of A into B, we must consistently render account of what the whole pattern was before the change and what it has become after the change; and here, again, we must see what was the sense of the change, its role for the whole pattern. Then we come to very new results, and many events which were considered completely separate appear to be simply fractional and predictable details of a single comprehensive mutation which can be described in much simpler terms.

The gulf between descriptive and historical linguistics so vehemently emphasized as inevitable by Saussure was a temporary gap due primarily to the circumstance that he used a new methodology for synchronic linguistics, while in diachronic linguistics he remained in the neogrammarian rut. This gulf was, moreover, considered necessary, because of a fallacious identification which endangered, I think, not only linguistics but anthropological research in general, the identification of two dichotomies: (1) synchronic - diachronic and (2) static - dynamic.

A synchronic system is never static. Let me resort to a simple comparison. If you are watching a movie and I ask you, "What do you see at this moment?" you do not see static situations. You see gangsters at world and horses running and various other motions. Only in front of the box office do you see merely statics pasted on the billboards. Neither is synchrony confined to statics, nor statics to synchrony. we can take a [|] static approach to history and ask in particular what has remained immovable from Latin or even Indo-European to contemporary Italian. On the other hand, there is room for a dynamic approach in synchronic linguistics. A linguistic change is not like a change to Daylight Saving Time, so that on May 1 all members of the speech community would no longer use element A, but only element B. There are periods when both elements A and B coexist with different stylistic connotation, and speakers may be aware that element A is an archaism and B an innovation and will use them accordingly just as in na audience with a high dignitary you will not wear the tie you wear on a date. Thus, the start and finish of a change originally belong to the same synchronic state. The time factor itself, upon entering into such a symbolic system of language, assumes a symbolic value. The artificial barrier between synchronic and historical linguistics vanishes, since one can analyze changes in terms of a synchronic pattern just as one does with its static constituents. (Jakobson 1971[1953e]: 227-228)

["Pattern in linguistics (Contribution to debates with anthropologists)"] [Online access] ["Statement made at the International Symposium on Anthropology (New York, June 1952) and published in An Appraisal of Anthropology Today (Chicago, 1953)."]

Jakobson, Roman; Halle, Morris 1962[1956a]. Phonology and phonetics. In: Rudy, Stephen (ed.), Selected Writings I: Phonological Studies. s-Gravenhage: Mouton, 464-504. Online access: https://archive.org/details/selectedwritings01jako/page/464/mode/2up

4.3 The spatio-temporal pattern of phonemic operations. If there is a difference between the linguistic pattern of two speech communities, interlocution between members of the two communities demands an adjustment of the listener to the speaker and/or of the speaker to the listener. This adjustment may involve all the aspects of language or only a few of them. Sometimes the phonemic code is the only one affected. Both on the listener's and on the speaker's side there are different degrees of this adjustment process, neatly called code switching by the communication engineers. The receiver, trying to understand the sender, and/or the sender, in trying to make himeself understood, concentrate their attention on the common core of their codes. A higher degree of adjustment appears in the effort to overcome the phonemic differences by switching rules, which increase the intelligibility of the message for its addressee. Having found these clues, the interlocutor may try to use them not only as a listener, but also in a more active manner, by adapting his own utterances to the pattern of his addressee.

The phonemic adjustment may cover the whole lexical stock, or the imitation of the neighbor's phonemic code may be confined to a certain set of words by his use of them. Whatever the adjustments are, they help the speaker to increase the radius of communication, and if often practiced, they are likely to enter his everyday language. Under favorable circumstances they may subsequently infiltrate into the general use of the speech community, either as a particular speech fashion or as a new pattern fully substituted for the former norm. Interdialectal communication and its influence on intradialectal communication must be analyzed from a linguistic and, particularly, from a phonemic point of view.

The problem of bridging space stops neither at the borders of distant and highly differentiated dialects, nor at the boindaries of cognatie or unrelated languages. Mediators, more or less bilingual, adapt themselves to the foreign phonemic code. Their prestige grows with the widening radius of their audience and may further a diffusion of their innovations among their monolingual tribesmen. [|502|]

Not only the interdialectal, but also the interlingual adjustment may affect the phonemic code without limitation to borrowed words or even without any lexical borrowing. In all parts of the world, linguists have been surprised, as Sapir confesses, to observe "the remarkable fact that distinctive phonetic features tend to be distributed over wide areas regardless of the vocabularies and structures of the languages involved." This far-reaching phonemonon still awaits systematic mapping and study in connection with the equally urgent inquiry into the typology of phonemic patterns.

The other possibility of phonemic adjustment ot a different dialect or foreign language is a partial or total preservation of its phonemic structure in borrowed words. As noted repeatedly in the phonemic literature and closely examined by Fries and Pike, "the speech of monolingual natives of some languages is comprised of more than one phonemic system." Such a coexistence of two systems within one lnaguage is due either to a phonemic difference between the original vocabulary and unassimilated loanwords, or to the use of two patterns, one native and the other imitative, as different styles of speech. Thus, spatial phenomena, namely interdialectal or interlingual isoglosses, especially isophones, may be projected into the framework of a single dialect, individual or social.

The same statement, mutatis mutandis, can be made about the time factor in language, particularly in the phonemic field. Any sound change, at its proceeding, is a synchronic fact. Both the start and the finish of a change coexist for a certain length of time. If the change differentiates the younger generation from the older, there is always some intercourse between the two generations, and the receiver belonging to one is accustomed to recode messages from a sender of the other. Furthermore, the initial and the final stage may co-occur in the use of one and the same generation as two stylistic levels: on the one hand, a more conservative and solemn, on the other, a more fashionable way of talking. Thus, synchronic analysis must encompass linguistic changes, and, vice versa, linguistic changes may be comprehended only in the light of synchronic analysis.

The decisive factor in phonemic change and in the diffusion of phonemic phenomena is the shift in the code. The interpretation of events in time and space is primarily concerned with the question: in what respect [|503|] is the structure of the code affected by such shifts. The motor and physical aspects of these innovations cannot be treated as self-sufficient agents, but must be subordinated to the strictly linguistic analysis of their role in the coding system. (Jakobson & Halle 1962[1956a]: 501-503)

["Phonology and phonetics"] [Online access] ["Written in Orleans and Cambridge, Mass., 1955, and published first in Fundamentals of Language (The Hague, 1956), then in a slightly revised and shortened version in Manual of Phonetics, ed. by L. Kaiser (Amsterdam, 1957), and in German translation by G. F. Meier (Berlin, 1960)."]

Jakobson, Roman 1981[1960d]. Linguistics and poetics. In: Rudy, Stephen (ed.), Selected Writings III: Poetry of Grammar and Grammar of Poetry. The Hague (etc.): Mouton de Gruyter, 18-51. Open access: https://archive.org/details/selectedwritings0003jako/page/18/mode/2up

Sometimes we hear that poetics, in contradistinction to linguistics, is concerned with evaluation. This separation of the two fields from each other based on a current and erroneous interpretation of the contrast between the structure of poetry and other types of verbal structure: the latter are said to be opposed by their "casual" designless nature to the "concasual", purposeful character of poetic language. In point of fact, any verbal behavior is goal-directed, but the aims are different and the conformity of the means used to the effects aimed at is a problem that evermore preoccupies inquirers into the diverse kinds of verbal communication. There is a close correspondence, much closer than critics believe, between the question of linguistic phenomena expanding in space and time and temporal spread of literary models. Even such discontinuous expansion as the resurrection of neglected or forgotten poets - for instance, the posthumous discovery and subsequent canonizaton of Emily Dickinson (d. 1886) and Gerard Manley Hopkins [|29|] (d. 1889), the tardy fame of Lautrémont (d. 1870) among surrealist poets, and the salient influence of the hithero ignored Cyprian Norwid (d. 1883) on Polish modern poetry - finds a parallel in the history of stardard languages that tend to revive outdated models, sometimes long forgotten, as was the case in literary Czech, which toward the beginning of the nineteenth century leaned toward sixteenth-century models.

Unfortunately, the terminological confusion of "literary studies" with "criticism" tempts the students of literature to replace the description of the intrinsic values of a literary work with a subjective, censorious verdict. The label "literary critic" applied to an investigator of literature is as erroneous as "grammatical (or lexical) critic" would be applied to a linguist. Syntactic and morphologic research cannot be supplanted by a normative grammar, and likewise no manifesto, foisting a critic's own tastes and opinions on creative literature, can serve as a substitute for an objective scholarly analysis of verbal art. This statement should not be mistaken for the quietist principle of laissez faire; any verbal culture involves programmatic, planning, normative endeavors. Yet why is a clear-cut discrimination made between pure and applied linguistics or between phonetics and orthoëpy, but not between literary studies and criticism?

Literary studies, with poetics as their focal point, consist like linguistics of two sets of problems: synchrony and diachrony. The synchronic description envisages not only the literary production of any given stage but also that part of the literary tradition which for the stage in question has remained vital or has been revived. Thus, for instance, Shakespeare, on the one hand, and Donne, Marvell, Keats, and Emily Dickinson, on the other, are experienced by the present English poetic world, whereas the works of James Thomson and Longfellow, for the time being, do not belong to viable artistic values. The selection of classics and their reinterpretation by a novel trend is a substantial problem of synchronic literary studies. Synchronic poetics, like synchronic linguistics, is not to be confused with statics; any stage discriminates between more conservative and more innovative forms. Any contemporary stage is experienced in its temporal dynamics, and, on the other hand, the historical approach both in poetics and in linguistics is concerned not only with changes, but also with continuous, enduring, static factors. A thoroughly comprehensive historical poetics or history of language is a superstructure to be built on a series of successive synchronic descriptions.

Insistence on keeping poetics apart from linguistics is warranted only when the field of linguistics appears to be illicitly restricted, for example, [|21|] when the sentence is viewed by some linguists as the highest analyzable construction, or when the scope of linguistics is confined to grammar alone or uniquely to nonsemantic questions of external form or to the inventory of denotative devices with no reference to free variation. Voegelin has clearly pointed out the two most important and related problems that face structural linguistics, namely, a revision of "the monolithic hypothesis about language" and a concern with "the interdependence of diverse structures within one language". No doubt, for any speech community, for any speaker, there exists a unity of language, but this over-all code represents a ssytem of interconnected subcodes; every language encompasses several concurrent patterns, each characterized by different functions.

Obviously we must agree with Sapir that, on the whole, "ideation reigns supreme in language [...]", but this supremacy does not authorize linguistics to disregard the "secondary factors". The emotive element of speech, which, as Joos is prone to believe, cannot be described "with a finite number of absolute categories", are classified by him "as nonlinguistic elements of the real world", he concludes, "which we refuse to tolerate in our science". Joos is indeed a brilliant expert in reduction experiments, and his emphatic demand for the "expulsion" of emotive elements "from linguistic science" is a radical experiment in reduction - reductio ad absurdum. (Jakobson 1981[1960d]: 19-21)

["Linguistics and poetics"] [Open access] ["First presented as the Presidential Address at the Annual Meeting of the Linguistic Society of America, December 27, 1956 [...]" (cf. Jakobson 1976e: 121)]

Jakobson, Roman 1962[1958a]. Typological studies and their contribution to historical comparative linguistics. In: Rudy, Stephen (ed.), Selected Writings I: Phonological Studies. s-Gravenhage: Mouton, 523-532. Online access: https://archive.org/details/selectedwritings01jako/page/522/mode/2up

5. Morphic Determinism. Since the "invariant points of reference for description and comparison" are (one must agree with Kluckhohn) the focal point of typology, I venture to illustrate these relatively new problems in linguistics by a salient analogue from another science.

The development of the science of language and particularly the transition from a primarily genetic standpoint to a predominantly descriptive approach strikingly corresponds to the contemporary shifts in other sciences, particularly to the difference between classical and quantum mechanics. This parallelism seems to me highly stimulating for the discussion of linguistic typology. I quote a paper on Quantum Mechanisms and Determinism delivered by the eminent specialist, L. Tisza, at the American Academy of Arts and Sciences: quantum mechanics [and let us add, modern structural linguistics] is morphically deterministic, whereas the temporal processes, the transitions between stationary states, are giverned by statistical probability laws. Both structural linguistics and quantum mechanics gain in morphic determinism what they lose in temporal determinism. "States are characterized by integers rather than by continuous variables", while "according to classical rules these systems would be characterized by continuous parameters", and "since two empirically given real numbers can never be rigorously identical, it is not surprising that the classical physicist objected to the idea of definite objects of perfect identity".

The structural laws of language are an ever nearer and clearer goal of typology and of the whole of descriptive linguistics in its newest phase, as I attempted to sum up the development in my linguistic obituary to Boas. And while one can only approve Greenberg's and Kroeber's illuminating remarks about the statistical character of "diachronic typologies" with their direction indices, the stationary typology must operate rather with integers than with continuous variables. [|528|]

We have avoided the current label "synchronic typology". If, for the modern physicist, the "peculiar interplay of quasi-permanent identity and random temporal change appears to be a most fundamental feature of nature", likewise in language "statics" and "synchrony" do not coincide. Any change originally belongs to linguistic synchrony: both the old and new variety co-occur at the same time in the same speech community as more archaic and more fashionable respectively, one pertaining to the more explicit and the other to the more elliptic style, i.e., two subcodes of the same convertible code. Each subcode in itself is for the given moment a stationary system governed by rigid structural laws, while the interplay of these partial systems exhibits the flexible dynamic laws of transition from one such system to another. (Jakobson 1962[1958a]: 527-528)

["Typological studies and their contribution to historical comparative linguistics"] [Online access] ["Report in the First Plenary Session of the Eighth International Congress of Linguists, Oslo, 5 August 1957. Published in the Proceedings of that Congress (Oslo, 1958)."]

Jakobson, Roman 1971[1960c]. The Kazan' School of Polish linguistics and its place in the international development of phonology. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 394-428. Online access: https://archive.org/details/selectedwritings02jako/page/396/mode/2up

On the other hand the basic contents of the theory elaborated by Kruszewski and the young Baudouin were assimilated by Ferdinand de Saussure, who, in the last decade of the nineteenth century and especially at the beginning of the twentieth, resolutely approached the basic problems of general linguistics. His first notes on the subject included both Polish scholars in the list of those few names "that should be cited" when discussing cardinal contributions to the theory of language. In 1908, while working on his course in general linguistics, Saussure sketched a review of the first systematic outline of the Geneva linguistic doctrine, which his student A. Sechehaye had just published. Saussure began with the observation that past attempts in the area of theoretical linguistics - from Humboldt to H. Paul and Wundt - had contributed nothing but raw material, while "Baudoin de Courtenay and Kruszewski were closer than anyone else to a theoretical view of language, without [|421|] digressing from purely linguistic considerations; yet they are unknown to most Western scholars". As a matter of fact, at the time when the Geneva linguist was attributing international significance to Kruszewski, the name of the researcher who had died twenty years before was quite forgotten by the Slavic as well as the Western academic world, and Baudouin's linguistic ideas were still unfamiliar to the international scholarship. Saussure carefully studied the theory of language elaborated by the two linguists, whom history has linked forever, and in his lectures, which were refashioned by Bally and Sechehaye into a posthumous book, he took from the teachings of Baudouin and Kruszewski and eloquently discussed such fundamental dichotomies as linguistic statics and dynamics (or in Baudouin's and likewise Saussure's favorite formulation, kinematics); constancy and changeability (immutabilité and mutabilité), and correspondingly "the eternal antagonism between a conservative force, based on associations by contiguity, and a progressive force based on associations by similarity" (solidarité avec le passé and infidélité au passé); language and speech (langue and parole; centrifugal and centripetal forces in language (force particulatrice and force unifiante); the coherent whole of the system and its parts; association by similarity, viz. "the bonds of kinship" (solidarité associative or groupement par familles), as opposed to "bonds of contiguity with concomitants" (solidarité syntagmatique); finally the "inseparable pair" of signans and signatum - oboznačajuščee and oboznačaemoe (signifiant and signifié). The general statements about shortening of stems in favor of ending (or "the process of morphological absorption", to use Kruszewski's term), which Baudouin had launched and his disciple widely developed since his student years, entered in full into Saussure's 1906-1907 lectures on general linguistics.

Saussure also obviously joined Baudouin's and Kruszewski's trend in their approach to the phonetic side of language. When at the end of the 1890's Saussure states that in the forms srutos, sreumen, sreuo "le phonème u noun apparaît sous deux formes acoustiques", he is trying to "perceive unity in diversity" (entrevoir l'unité dans la diversité), just like Baudouin [|422|] when discovering i mutabile. It is interesting that Baudouin's later formulations find a close analogue in Saussure's course on general linguistics, in which the phoneme is defined as a complex psychic unit that joins images of the articulatory act and of the acoustic effect. At the same time, however, Saussure was apparently deterred by the ambiguity of the term "phoneme", which was still used by French linguists in the meaning given to it by its inventor, Dufriche-Desgenettes. In his third and last course he suggests that it would be safer to omit from the analysis of the linguistic sign such terms as phonème, "qui contient l'idée d'action vocale, de parole". However in rejecting the term, which had earlier been adopted from him by Kruszewski and Baudouin and used by them in a new way, Saussure was by no means burying the idea of elementary linguistic invariants, which he had originally put into that term and for which broad outlooks were opening in the joint research of the two Polish linguists. Saussure emphasized in his lectures that first of all "il faut dresser le système phonologique de l'idiome qu'on aborde". The components of that system, "the irreducible phonological units, or phonic elements of the language", constitute a finite total within any given system. (Jakobson 1971[1960c]: 421-422)

["The Kazan' School of Polish linguistics and its place in the international development of phonology"] [Online access] ["The Polish version of this study appeared in the Biuletyn Polskiego Towarzystwa Językoznawczego, XIX (1960), and was based on the paper given at a meeting of the Linguistic Commuttee of the Polish Academy of Sciences on January 2, 1958, in Warsaw, under the title "The Origin of the Concept of Phoneme in Polish and World Linguistics"."]

Jakobson, Roman 1980[1962e]. Sign and system of language: A reassessment of Saussure's doctrine. Translated by B. Hrushovski. Poetics Today 2(1): 33-38. DOI: https://doi.org/10.2307/1772350

We deal here, apparently, as in all modern sciences, with the significant idea of invariance. We speak about combinatorial, context-dependent variants on the level of sound as well as on the level of grammar. But it would be impossible to speak about variants as long as we have not clarified the nature of the basic invariant, the unit to which all these variants are related. The search for the invariants is now the most substantial problem not only in phonology, but in grammar as well. When dealing with the sign, the bilateral signum as a link [|35|] between the signans and signatum, how do we discover such invariants on one hand in the domain of the signans and on the other hand in the field of the signatum? The basic difference between the two, from a linguistic point of view, is that the signans must necessarily be perceptible whereas the signatum is translatable. In both cases the principle of equivalence obtains. In the domain of the signans the relative equivalence must be extremely perceivable; it can be ascertained, however, only in respect of the function of these sound relations in a given language. We recognize such distinctive features, by means of a spectograph, we are able to translate them from the acoustic field into the visual level. And like the signans, the signatum too must be studied in a purely linguistic and objective manner. A purely linguistic semantics can and must be constructed, if we agree with Peirce that the basic property of any verbal sign lies in its capability of being translated into another verbal sign, either a more deployed, explicit sign, or, on the contrary, a more elliptical sign, of the same language system or of a different one. This translatability lays bare that semantic invariant for which we are searching in the signatum. In such a way it becomes possible to submit semantic problems of language to distributional analysis. Metalinguistic identifying sentences, such as "A rooster is a male of a hen" belong to the inventory of the English language community; the reversibility of both expressions - "A male of a hen is a roostel" - demonstrates how the meaning of words becomes a real linguistic problem through a distributive analysis of such common metalingual utterances.

Among the basic features of the Cours de linguistique générale is the split nature of linguistics: the separation of synchrony from diachrony. The thorough work done over several decades in both partial areas, as well as the refined methodology developed in this research, brought about a serious danger of a flagrant gap between these two descriptions, and also the necessity of overcoming this gap. Saussure's identification of the contrast between synchrony and diachrony with the contrast between statics and dynamics turned out to be misleading. In actual reality synchrony is not at all static; changes are always emerging and are a part of synchrony. Actual synchrony is dynamic. Static synchrony is an abstraction, which may be useful to the investigation of language for specific purposes; however, an exhaustive true-to-the-facts synchronic description of language must consistently consider the dynamics of language. Both elements, the point of origin and the final phase of any change, exist for some time simultaneously within one language community. They coexist as stylistic variants. When taking this important fact into consideration, we realize that the image of language as a uniform and monolithic system is oversimplified. Language is a system of systems, an overall code which includes various subcodes. These variegated language styles do not make an accidental, mechanical aggregation, but rather a rule-governed hierarchy of subcodes. Though we can tell which of the subcodes is the basic code, it is nevertheless a dangerous simplification to exclude the discussion of the other subcodes. If we consider langue as a totality of teh conventions of language, then we must be very careful not to be researching fictions. [|36|]

I believe that today our chief task should be to become realists, to build a realistic study of language and combat any fictionalism in linguistics. We must ask ourselves: what is the real linguistic convention that enables exchange of speech in a given language community and serves effectively the various tasks of communication? Some linguists ask, why should linguistics differ from physics in its methodology? Why could not the scholar of language impose his own system of symbols, his creative model, upon the investigated material, as is common in natural sciences? Indeed, one observes, in many respects, an ever more meaningful and fruitful contact between the natural science and linguistics; nevertheless one must keep in mind the specific differences as well. In the London school of mathematical information theory the cardinal difference was clearly recognized and the problem of communication was separated from other aspects of information. First of all, one must distinguish between two classes of signs - indices and symbols, as Peirce called them. Indices, which the physicist extracts from the external world, are not reversible. He transforms these indices given in nature into his own system of scientific symbols. In the science of language the situation is cardinally different. The symbols exist immediately in language. Instead of the scientist, who extracts certain indices from the external world and reshapes them into symbols, here an exchange of symbols occurs between the participants of a communication. Here the roles of addresser and addressee are interchangeable. Hence the task of a science of language is quite different. We are simply trying to translate into metalanguage this code, which is objectively given in the language community. For the natural scientist symbols are a scientific tool, whereas for the linguist they are beyond that, and first af all, the true object of his research. The physicist Niels Bohr understood perspicaciously this natural realism of the linguist's position.

Having mentioned Niels Bohr, I would like to recall his methodological dictum essential both for physics and linguistics. Namely, that, when an observation is made, it is imperative to determine exactly the relation between the observer and the observed thing. A description that does not comply with this requirement is imprecise from the point of view of today's physicist, as it is from today's linguistics. It is our task to clarify the various positions of scholars vis-à-vis language. The so-called crypto-analytical position is the point of view of an observer who does not know the language code, and who could be compared to a military crypto-analyst, attempting to decipher an enemy's encoded message. He tries to break the foreign code through a careful analysis of the text. In the study of unknown languages such devices may obviously bring fruitful results. This, however, is merely the first stage of research, and it is by no means the only, but rather one of many methodologial, a first approximation. Then the observer attempts to reach the second, more advanced stage, the stage of a pseudo-participant in the given language community. He does not any more move from the text to the code, but rather absorbs the code and tries to use the code for a better understanding of the message. (Jakobson 1980[1962e]: 34-36)

["Sign and system of language"] [DOI: 10.2307/1772350] ["Vorgetragen in Erfurt, 2. Oktober 1959, am 1. Internationalen Symposium "Zeichen und System der Sprache", veröffentlicht in den Schiften zur Phonetik, Sprachwissenschaft und Kommunikationsforschung, IV (Berlin, 1962)." (cf. Jakobson 1962e: 279)]

Jakobson, Roman 1971[1961b]. Linguistics and communication theory. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 570-579. Online access: https://archive.org/details/selectedwritings02jako/page/570/mode/2up

Preconceived possibilities, according to MacKay, "is the key phrase in communication theory", and a similar claim comes from linguistics. In neither discipline has there been any doubt about the fundamental role of selective operations in verbal activities. The engineer assumes a "filing system" of prefabricated possibilities more or less common to the sender and receiver of a verbal message, and Saussurian linguistics speaks correspondingly about langue, which makes possible an exchange of parole between interlocutors. Such an "ensemble of possibilities already foreseen and provided for" implies a code, conceived by communication theory as "an agreed transformation - usually one-to-one and reversible" - by which one set of informational units is converted into another set, for [|573|] instance, a grammatical unit into a phonemic sequence and vice versa. The code matches the signans with its signatum and the signatum with its signans. Today, with respect to the treatment of coding problems in communication theory, the Saussurian dichotomy langue/parole can be restated much more precisely and acquires a new operational value. Conversely, in modern linguistics communication theory may find illuminating information about the stratified structure of the intricate linguistic code in its various aspects.

Although the framework of the linguistic code has been adequately outlined in linguistics, it is still frequently overlooked that the finite ensemble of "standard representations" is limited to lexical symbols, their grammatical and phonological constituents, and the grammatical and phonological rules of combination. Only this portion of communication may be defined as mere "activity of replicating representations". On the other hand, it is still opportune to recall that the code is not confined to what communication engineers call "the bare intelligence content" of speech, but that likewise the stylistic stratification of the lexical symbols and the allegedly "free" variation, both in their constitution and in their combination rules, are "foreseen and provided for" by the code.

In his program for the future science of signs (semiotics) Charles Peirce stated: "A Legisign is a law that is a Sign. This law is usually established by men. Every conventional sign is a legisign." Verbal symbols are cited as a salient example of legisigns. Interlocutors belonging to one given speech community may be defined as actual users of one and the same linguistic code encompassing the same legisigns. A common code is their communication tool, which actually underlies and makes possible the exchange of messages. Here is the essential difference between linguistics and the physical sciences, and this difference has been distinctly and repeatedly singled out in the theory of communication, especially by its English school, which insists on a clear-cut line of demarcation between the theory of communication and of information. Nevertheless, this delimitation, strange as it seems, is sometimes disregarded by linguists. "The stimuli received from Nature," as Colin Cherry wisely stresses, "are not pictures of reality but are the evidence from which we build our personal models." While the physicist creates his theoretical construct, imposing his own hypothetical system of new symbols upon [|474|] the extracted indices, the linguist only recodes, translates into symbols of a metalanguage those extant symbols which are used in the language of the given speech community.

The constituents of the code, for instance, the distinctive features, literally occur and really function in speech communication. Both for the receiver and the transmitter, as R. M. Fano points out, the operation of selection forms the basis of "information-conveying processes". The set of yes-or-no choices underlying any bundle of these discrete features is not an arbitrary concoction of the linguist, but is actually made by the addressee of the message, insofar as the need for their recognition is not cancelled by the prompting of the verbal or non-verbalized context.

Both on the grammatical and on the phonological level, not only the addressee in decoding the message, but also the encoder may practice ellipsis; in particular, the encoder omits some of the features, or even some of their bundles and sequences. But ellipsis, too, is governed by codified rules. Language is never monolithic; its overall code includes a set of subcodes, and such questions as that of the rules of transformation of the optimal, explicit kernel code into the various degrees of elliptic subcodes and their comparison as to the amount of information requires both a linguistic and an engineering examination. The convertible code of language, with all its fluctuations from subcode to subcode and with all the current progressing changes which this code is undergoing, is to be jointly and comprehensively described by the means of linguistics and communication theory. An insight into the dynamic synchrony of language, involving the space-time coordinates, must replace the traditional pattern of arbitrarily restricted static descriptions.

The linguistic observer who possesses or acquires a command of the language he is observing is or gradually becomes a potential or actual partner in the exchange of verbal messages among the members of the speech community, a passive or even active fellow member of that community. The communication engineer is right when defending against "some philologists" the absolutely dominant "need to bring the Observer ont the scene" and when holding with Cherry that "the participant-observer's description will be the more complete". The antipode to the participant, the most detached and external onlooker, acts as a cryptanalyst, who is a recipient of messages without being the addressee [|575|] and without knowledge of their code. He attempts to break the code through a scrutiny of the messages. As far as possible, this level of linguistic investigation must be merely a preliminary stage toward an internal approach to the language studied, when the observer becomes adjusted to the native speakers and decodes messages in their mother-tongue through the medium of its code.

As long as the investigator knows no signatum of a given language and has access to nothing but signans he willy-nilly has to strain his detective capacities and obtain any possible information about the structure of this language from the external evidence. The present state of Etruscology is a good example of such a technique. But if the linguist is familiar with the code and has mastered the conventional transformations byw hich a set of signantia is converted into a set of signata, then it becomes superfluous for him to play Sherlock Holmes, unless he aspires to finding out how wide and reliable data could be obtained by such a mock scrutiny. It is difficult, however, to stimulate ignorance of a familiar code: smuggled-in meanings distort one's allegedly cryptanalytic approach.

Obviously "the inseparability of objective content and observing subject", singled out by Niels Bohr as a premise of all well-defined knowledge, must be definitely taken into account also in linguistics, and the position of the observer in relation to the language observed and described must be exactly identified. First, as formulated by Jurgen Ruesch, the information an observer can collect depends upon his location within or outside the system. Furthermore, if the observer is located within the communication system, language presents two considerably different aspects when seen from the two ends of the communication channel. Roughly, the encoding process goes from meaning to sound and from the lexicogrammatical to the phonological level, whereas the decoding process displays the opposite direction - from sound to meaning and from features to symbols. While a set (Einstellung) toward immediate constituents takes precedence in speech production, for speech perception the message is first a stochastic process. The probabalistic aspect of speech finds conspicuous expression in the approach of the listener to homonyms, whereas for the speaker homonymy does not exist. When saying /sʌn/, he knows beforehand whether "sun" or "son" is meant, while the listener depends on the conditional probabilities of the context. For the receiver, [|576|] the message presents many ambiguities which were unequivocal for the sender. The ambiguities of pun and poetry utilize this input property for the output. (Jakobson 1971[1961b]: 572-576)

["Linguistics and communication theory"] [Online access] ["Presented in the Symposium on Structure of Language and Its Mathematical Aspects, New York, 15 of April, 1960, and published in Proceedings of Symposia in Applied Mathematics, XII (1961)."]

Jakobson, Roman 1971[1963e]. Parts and wholes in language. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 280-284. Online access: https://archive.org/details/selectedwritings02jako/page/282/mode/2up

Nagel's statement (1-e) - that "the word 'whole' may refer to a pattern of relations between certain specified kinds of objects or events, the pattern being capable of embodiment on various occasions and with various modifications" - finds a wide application in language with its relational invariants and manifold contextual and stylistic variations. This kind of whole-part relation, which was for a long time underrated by linguists, has finally attracted their attention, especially in regard to the contextual variants in phonemics and grammar. In the lexical field, which so far remains the underdeveloped province of linguistics, the study of invariance and of variables is still inferior to the medieval doctrine of the mody significandi.

Stylistic variations, particularly in phonology, gradually have begun to disturb students of language who until recently had been possessed by the isolationist idea of a monolithic verbal code. The variety of functional, mutually convertible subcodes requires a careful and consistent structural analysis. Such an analysis makes possible a synchronic study of the phonemic and grammatical changes in progress, which initially present a necessary coexistence of the older and newer form in two related subcodes, and thus there emerges a bridge between descriptive and historical linguistics. On the other hand, the inquirer into the system of subcodes encompasses the various forms of interdialectal and even interlingual code switching and thus establishes an intimate bond between the description of an individual or local dialect and the vast horizons of linguistic geography. [|284|]

If the whole is "a pattern of relations", then the part, as Nagel notes, may also refer to "any one of the elements which are related in that pattern on some occasion of its embodiment". Thus he touches upon the findamental difference between design and token, a whole-part relation which linguists have recognized, but without drawing all the obvious and far-reaching inferences.

Finally, with the progress of typological studies, the science of language will be able to answer Nagel's question about systems "whose parts stand to each other in various relations of dynamical dependence" (1-h). The universal and near-universal laws of implication which underlie this taxonomy reveal a rigorous phonemic and grammatical stratification, which likewise determines the gradual acquisition of language by children and its decay in aphasia.

A systematic consideration of multiform whole-part relations broadly extends the scope of our science; it allows a systematic analysis of verbal messages with respect both to the code and to the context; it uncovers the complex interaction of the various levels of language, from the largest to the smallest units, and the constant interplay of diverse verbal functions. It introduces time and space factors into descriptive linguistics and, in the search for general, universal laws, is near to proving the scientific verity of Henry Delacroix's keen anticipation: "Une langue est une variation historique sur le grand thème humain du langage." (Jakobson 1971[1963e]: 283)

["Parts and wholes in language"] [Online access] ["Lecture at the Hayden Colloquium of the Massachussetts Institute of Technology, 1960, published in Parts and Wholes, under the editorship of D. Lerner (New York; London, 1963)."]

Jakobson, Roman 1962c. Retrospect. In: Rudy, Stephen (ed.), Selected Writings I: Phonological Studies. s-Gravenhage: Mouton, 629-659. Online access: https://archive.org/details/selectedwritings01jako/page/630/mode/2up

My reiterated thesis on the distinctive oppositions inherent in the structure of language was meant as a literal intrinsic description of actual phenomena and not at all as a picturesque and metaphorical way of expression. All the distinctions functioning in a language are acquired, performed, perceived, and interpreted by the participants of verbal communication, and the linguist recodes them as [|650|] he does all other superposed constituents of the symbolic possessed by the language users. The linguist translates this system of symbols into a correlated system termed "metalanguage". In this respect there is an essential difference between a physical science which imposes its own code of symbols upon the "indexes" observed (in C. S. Peirce's meaning of this term) and the phenomenologoy of language, whose task is to break up the inner code actually underlying all verbal symbols and, as Sapir used to say, all "symbolic atoms". The verbal code is a real property of any given speech community, and therefore the notorious linguistic controversy between the "hocus-pocus" position and the "God-given truth" is aimless. Any phonemic or grammatical opposition is neither fictional nor metaphysical, but simply and solely a code-given truth.

In positions of "neutralization" the phonemes reduce the number of their distinctive components, whereas on the level of features every distinctive opposition is endowed with a perceptual constancy; and as far as the features are properly defined in purely relational terms, no overlapping can arise. The relational invariant of each oppositional pair is per definitionem actualized in every context where the given feature occurs, unless this feature is omitted in an elliptic variety of speech. Any such variety, however, may be translated in case of need by the speaker or listener into a more explicit subcode of the same language. The shipshod forms are judged precisely as reduced, slurring, slovenly, and each request for repetition and every danger of misunderstanding prompt the restoration of the distinctions omitted. The existence of optimum explicitness both on a phonemic and on a grammatical level is a sine qua non of all ellipsis; otherwise an historically elliptic sequence is no longer elliptic from a synchronic viewpoint: the optional omission of a feature has changed into its mandatory absence. The explicit phonemic subcode or "full style" of pronunciation, in Ščerba's terms, is an inward resource of spoken language, quite different from those extrinsic auxiliaries used by speakers to decipher homonyms, such means as an ad hoc contrived spelling-pronunciation or a recourse to the spelling names of the letters, or simply to their writing.

Any suggestion to dismiss the problem of translation from one subcode into another (C. L. Ebeling) is to be rejected, like all endeavors to rob linguistics of some of the vital properties pertaining to language. The elliptic subcode has its own structural laws, and its coexistence with the explicit subcode is the indispensable synchronic phase of every phonemic merger, since in general the start and finish of a phonemic change are first conceived as belonging to two coexistent subcodes. This [|651|] synchronic approach to linguistic changes abolishes the customary identification of synchrony with statics and of dynamics with diachrony. The concept of dynamic synchrony calls for a strictly relational treatment of changes "en fonction du système phonologique qui les subit" (p. 3). Convinced from the start that "the pending task was to overcome statics and to discard the absolute" (Iskusstvo, August 2, 1919), I concentrated my research work of the late twenties on mutability as a constant essential component of any phonemic system, and on the systematic character of phonemic mutations. The subsequent decades of international discussion over the principles of historical phonology and over their application to diverse languages, in particular to the Slavic material, call for a new, deepened, and extended view of the same theoretical and concrete problems. The transition from the listing of phonemes to a consistent analysis into features affords a much more synthetic scope of phonological processes. Traditionally, on the level of phonemes, only those alterations which depend on the preceding or following segments of the chain were seen as conditioned, combinatory, contextual changes, whereas research on the level of features radically cuts down the number of ostensibly "spontaneous" changes, because most of the featural changes are confined to combinations with specific concurrent features. For example, the loss of vocalic nasality does not affect nasal consonants and is thus a typical example of contextual changes.

Phonemic change is a recoding: like any question of the linguistic code and of coding economy, it is first and foremost a semiotic question; yet despite Sapir's forceful warning (Language, Ch. VIII), some students of linguistics still maake "the fatal error of thinking of sound change as a quasi-physiological" phenomenon and bandy about such easy catchwords as the "ease of articulation".

The question of invariants and variables in time is paralleled by the problem of invariants and variables in space. The "increase in the radius of communication" and the process of switching the code to accommodate the interlocutor help to explain the broad expansion of phonemic features and the widespread phonemic affinities between neighboring languages, whether cognate or unrelated. My first attempt at outlining certain instances of this phenomenon, in particular the "Eurasian" area of the consonantal opposition sharp/non-sharp, can now be revised and improved, since a much more exhaustive and precise body of phonemic material concerning the various languages and dialects involved has become available. We face the impending need for collective international work on a phonemic atlas of the world. [|652|] Unquestionably such a coherent cross-language mapping of isophones will provide far deeper insight into the paths of phonemic expansion and change, since expansion is an integral part of any change, and since the distinction between "sources" (foyers d'innovation) and "affected zones" (aires de contagion) proves to be rather illusory.

In the over-all code of any individual speaker and of any speech community, the observer, insofar as he refrains from factitious filtration, unfailingly detects the permanent coexistence of phonemic variants relating to different subcodes of one and the same convertible code. Thus from fieldwork of 1916 in a village north of Moscow I first learned that we cannot properly speak of a uniform dialect, but only of "a multitude of individual and short-term parlances, and instead of sound-laws one deals here for the most part with mere bents and tendencies". Like modern thermodynamics, linguistics too treats both the reversible and the irreversible aspects of time. The former aspect is exemplified by the fluctuation of Paris speech between the original distinction of /ã/ - /õ/ and the optional merger of the two nasal vowels: there still exists a virtual reversal, from the latter, an elliptic innovation, back to the conservative phonemic discrimination of blanc and blond. On the other hand, the regression from the loss to the maintenance of a distinction between /ẽ/ and /ǿ/ is obsolete in some dialectal varieties of Frech, and the reversible fluctuation has given place to the product of a completed mutation.

Since in the process of a change its two terms, the start and the finish, necessarily co-occur and can be compared as to their place and function in the system, we are enabled and even compelled to seek the purpose of the change. If mutations are a constituent part of the purposive linguistic system, then the application of a "teleological criterion" to the analysis of phonemic changes must be accepted as a corollary following from these premises. I cannot share that antiquated superstitious fear of teleology which is still professed by some students of linguistics. As has been made clear by the productive and inspiring discussion of "behavior, purpose, and teleology" carried on during the last twenty years in the philosophy of science (from A. Rosenbleuth, N. Wiener, and J. Bigelow to R. Taylor, I. Scheffler, and others), "the adoption of a teleological approach simplifies the analysis of goal-directed behavior and enlarges the scope of this analysis." The theoretical elucidation of such notions as "goal-attainment", "goal-failure", and "negative feedback" opens new possibilities for their use in linguistic operations. (Jakobson 1962c: 649-652)

["Retrospect" to SW I] [Online access] ["Parts of this Retrospect are taken over from my communication to the Fourth International Congress of Phonetic Sciences (Helniski, September 7, 1961), "The Phonemic Concept of Distinctive Features", which has appeared in the Proceedings of that Congress ('s-Gravenhage, 1962)."]

Jakobson, Roman 1971[1966e]. Henry Sweet's path toward phonemics. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 456-467. Open access: https://archive.org/details/selectedwritings02jako/page/458/mode/2up

In his famous Presidential Address of 1877 Sweet condemned the exclusively genealogical orientation of comparative philology which values "the forms of later languages solely according to the amount of light they throw on older forms". Both in his address and in a later study devoted to Linguistic Affinity, Sweet outlined a new set of comparative problems. On the one hand "divergence between cognate languages [...] raises the question, how far does the possibility of change of structure extend?" On the other hand, "nothing can be more important than the comparisoon of the 'parallel developments' in such [|465|] distinct languages as the Romance and the Neo-Sanskrit, English and Persian, etc.". The diffusion of linguistic phenomena asks for a systematic study, since the possibilities of mixture has proved to be "greater than was suspected by the founders of comparative philology". "There is no necessary limit to mixture of vocabulary. [...] The possibility of syntactic influence is clearly proved" and "there is clear evidence that different languages may influence one another morphologically".

Again, it was Sweet who had broached an unwonton problem which has become a crucial topic in present-day linguistics: "In the first place, there can be no doubt that contiguous languages often show striking phonetic resemblances even when they are not cognate or only remotely so." This statement is supported by references to "marked phonetic peculiarities" spread out "without regard to linguistic relationship" in the Caucasus, in Eastern Asia or in Southern Africa.

Beside similaritise due to kinship or contiguity of languages Sweet observed genetically independent "agreements in general structure" and for instance proposed a comparison of some features of Modern English "to those of Chinese, the Turanian, and even of some savage languages". From such a typological comparison independent of genetic relations, Sweet infers "the all-important principle that every language and every period of a language has an individuality of its own, which must be respected. The corollary of this principle is "the recognition of a science of living, as opposed to dead, or antiquarian philology". This general conclusion of Sweet's, exactly like his Broad Romic, he links with the practical study of language; he lucidly foresees, however, that this science "is the indispensable foundation" of the various linguistic branches, and even "of historical and comparative philology".

On the practical level Sweet displays a consistently functional approach, and discussing "the delicate distinctions of the English verb" he asks ironically: "What can historical philology contribute to the analysis of will love, shall love, is loving, etc.?" The 'injurious' misuse of the historical viewpoint is drastically attacked by Sweet and appears to him "as reasonable as it would be to insist on every one having Macaulay's History of England permanently chained round his neck, because history [|466|] is an improving study. Meanwhile, Sweet's view of synchrony is far from deadly statics, and 'changes in progress' are discussed by him with gripping ingenuity.

There appears one more concept, conjugate with the functional method of approach and quite alien to the predominant linguistic doctrine of the late nineteenth century. This was the idea of totality, emphasized by Sweet against the disintegrating spirit of the current dogma: "I, for one, am strongly of the opinion that our present exaggeratedly analytical methods [...] are a failure comared with the synthetic methods of the Middle Ages, by which sentences were grasped as wholes", whereas at present they are "put together like pieces of mosaic work". The critic concludes that "any real reform will involve, partially at least, a return to these older methods". It is remarkable that similarly to Sweet, his older American contemporary Charles Sanders Peirce also, while anticipating the development of a semiotic science in the future, deplored and assailed its present status and invoked the Schoolmen's superior legacy. (Jakobson 1971[1966e]: 464-466)

["Henry Sweet's path toward phonemics"] [Open access] ["Written in the Center for Advanced Study in the Behavioral Sciences (Stanford, California), 1961, for the volume In Memory of J. R. Firth (London, 1966)."]

Jakobson, Roman 1971[1963d]. Efforts toward a means-ends model of language in interwar continental linguistics. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 522-526. Online access: https://archive.org/details/selectedwritings02jako/page/524/mode/2up

The emphasis on the duality of any verbal sign, taken over by F. de Saussure from Stoic and Scholastic tradition, necessarily brought new results when the relation between the two aspects of the sign, its signans and signatum, was consistently revised from the means-ends angle, and the two Saussurian "basic principles" - the arbitrariness of the sign and the linearity of the signans - proved to be illusory.

In the study of the two basic linguistic operations selection and combination, or, in other terms, the paradigmatic and syntagmatic aspects of language - it is the paradigmatic aspect that was particularly elucidated in the work on the means-ends model. The selection of units and their combinations is a purposive operation, in contradistinction to those purely redundant combinations which admit no selection. The problem of a careful discrimination between autonomous and combinatory variants was successfully attacked on both the phonological and grammatical levels. One of the most intricate networks, the strikingly hierarchic [|525|] make-up of the paradigmatic pattern, was subjected to penetrating scrutiny, particularly in the research of Kuryłowicz. The consistent concern with meaning, a true yield of the entire trend, and the systematic analysis of grammatical meanings with a rigorous distinction between general and contextual meanings demanded a similar exploration of lexical meanings, and the imperative need to treat vocabulary as "a complex system of words mutually coordinated and opposed to each other" was comprehensively advocated by Trubetzkoy at the First Congress of Slavists.

In the "Thèses" inaugurating the first volume of the Travaux and in later deliberations the Prague Circle, insisting on purposiveness in language, outlined an inquiry into languages of diverse functions and paid due attention to their different patterns. In this study of the various linguistic aims, the poetic function obtained the most fruitful treatment. The sense for the multifarious character of language saved the Prague group from an oversimplified, bluntly unitarian view; language was seen as a system of systems and especially Mathesius' papers on intralingual coexistence of distinct phonemic patterns opened new outlooks.

The regard for the various "functional dialects", or, in other words, the different styles of language radically altered the view of linguistic change. The two stages of a change in progress were reinterpreted as two simultaneous styles of language; the change was conceived as a fact of linguistic synchrony, and as any fact of synchrony it demanded a means-ends test with respect to the whole system of language. Thus historical linguistics experienced a complete metamorphosis. If in the previous stage of Indo-European studies, as Benveniste stated in 1935, "l'effort, considérable et méritoire, qui a été employé à la description des formes n'a été suivi d'aucune tentative sérieuse pour les interpréter", henceforth, he pointed out, it would be necessary to consider the reconstructed language no longer as a repertory of immutable symbols but "comme une langue en devenir", and, furthermore, to envisage the functions of the elements involved. (Jakobson 1971[1963d]: 525)

["Efforts toward a means-ends model of language in interwar continental linguistics"] [Online access] ["Written in 1962 for Trends in Modern Linguistics (Utrecht-Antwerp, 1963)."]

Jakobson, Roman 1985[1970j]. Linguistic evidence in comparative mythology. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 12-32. DOI: https://doi.org/10.1515/9783110855463.12

The synchronic approach to language, which replaced the one-sided genetic conception, cast new light on the fundamental problems of comparative-historical linguistics as well. In turn the structural analysis of myths and mythological systems, which has creatively mastered the methods of contemporary linguistics and which, following the courageous initiative of V. Ja. Propp, advances newer and newer tasks in the latest works of C. Lévi-Strauss, A. J. Greimas and the recent Moscow Symposium on the Structural Study of Sign Systems, will undoubtedly make possible a more integral and systematic approach to the religions of the distant past. Typological comparison, which is being practiced ever more widely in contemporary linguistics, shows mythology the way to cenostruct the whole on the basis of its fragmentary remains. (Jakobson 1985[1970j]: 32)

["Linguistic evidence in comparative mythology"] [DOI: 10.1515/9783110855463.12] ["First presented as a report to the Seventh International Congress of Anthropological and Ethnological Sciences, Moscow, 6 August 1964, and published in the proceedings of the Congress [...]"]

Jakobson, Roman 1966n. Retrospect. In: Rudy, Stephen (ed.), Selected Writings IV: Slavic Epic Studies. The Hague; Paris: Mouton & Co, 637-704. DOI: https://doi.org/10.1515/9783110889581.645

The oral tradition of the Slavic peoples shows such a powerful continuity that recently a bold and stimulating attempt was made by Ivanov and Toporov to reconstruct entire Common Slavic texts, especially compositions of ritual shamnistic character, namely, spells and exorcisms, as well as survivals of mythological motigs in lamentations, epics and faily tales.

Comparative linguistics brings out 1) intrinsic typological parallels, 2) analogous developmental traits due to similar social and cultural preconditions, 3) correspondences based on common ancestly, and 4) points of likeness caused by diffusion. Finally, etymologists trace the connection of the vocabulary and phraseology of a given language with historical events, morals and manners. However, no linguist today would declare a single one of these problems the only legitimate task while discarding or discrediting the rest, as has quite often been the case in the study of Russian folklore, especially epics. Both aspects of linguistics - diachrony and synchrony - complement each other; nevertheless students of Slavic epic traditions frequently perisst in a single-tracked historical approach, notwithstanding the exhortation of the young Veselovskij to interpret any vital fact of folklore first and foremost within the framework of its time and environment. On the other hand, linguists could hardly share the opposite creed of some folklorists to whom the history of oral epos seems an unfeasible task, since, as Skaftymov puts it, the bylina "is not something definitely created but undergoes a process of continuous creation and perpetual change". This definition is equally applicable to [|649|] language; nonetheless, even investigators of inwritten languages would never feel handicaped by the alleged "lack of pivots".

Linguists are aware that in the history of language certain innovations may be confined only to the upper or only to the lower class, and their downward or upward expansion may change some of them into a common property of the whole people, while in folklore we still encounter vulgar, pseudo-sociological biases ascribing the creative role only to the top or, inversely, only to the bottom of the social pyramid. Ancient lexical and phraseological units are subject to gradual shifts in meaning, and possibly in form, under the pressure of posterior historical events and constellations, and on the other hand, neither borrowing nor loan translation precludes the immediate or later use of these foreignisms for the designation of obviously indigenous novelties. (Jakobson 1966n: 648-649)

["Retrospect"] [DOI: 10.1515/9783110889581.645]

Jakobson, Roman 1971[1969c]. Linguistics in relation to other sciences. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 655-696. Online access: https://archive.org/details/selectedwritings02jako/page/656/mode/2up

The linguistic framework of courtship, marriage, and kinship rules and taboos is their indispensable implement. The careful and exhaustive observations of Geneviève Calame-Griaule on the pragmatics of language in the erotic, societal, and religious life of a community is a telling illustration of the decisive role of verbal behavior in the entire domain of social anthropology.

In the century-old history of economics and linguistics, questions uniting both disciplines have repeatedly arisen. One may recall that economists of the Enlightenment Period used to attack linguistic problems, as, for example, Anne-Robert-Jacquues Turgot, who did a study on etymology for the Encyclopédie, or Adam Smith, who wrote on the origin of language. G. Tarde's influence upon Saussure's doctrine in such matters as circuit, exchange, value, output/input, producer-consumer is well-known. Many common topics, as, for instance, "dynamic synchrony", contradictions within the system, and its continual motion, undergo similar developments in both fields. Fundamental economic concepts were repeatedly subjected to tentative semiotic interpretations, coined the catch-phrase "a ruble is not silver, a ruble is the ruler's word", and John Law taught that money has only the wealth of a sign based on the prince's signature. At present, Talcott Parsons systematically treats money as "a very highly specialized language", economic transactions as "certain types of conversations", the circulation of money as "the sending of messages", and the monetary system as "a code in the grammatical-syntactical sense". He avowedly applies to the economic interchange the theory of code and message developed in linguistics. Or, according to the formulation of Ferruccio Rossi-Landi, "l'economia in sense proprio è studio di quel settore del segnico non-verbale, che consiste nella circolazione di un particulo tipo di messaggi solitamente chiamati 'merci'. Piú in breve, e con una formula: l'economia è studio dei messaggi-merci". In order to avoid a metaphorical extension of the term "language", it is, perhaps preferable to interpret money as a semiotic system with a particular destination. A semiotic interpretation of the processes and concepts involved is necessary for the extract scrutiny of this [|666|] medium of communication. Since, however, "the most general matrix" of symbolic systems, as Parsons points out, "is language", linguistics actually appears to offer the most helpful model for such an analysis. Yet there are further reasons for connecting economics with linguistic studies: the exchange of utilities "converted" into words, the direct concomitant role of language in all monetary transactions and the translatability of money into purely verbal messages, such as checks or other obligations.

Thus, communication of mates and goods or services proves to be to a high degree an interchange of auxiliary messages, and the integrated science of communication incorporates semiotic proper, i.e. the study of sheer messages and their underlying codes, plus those disciplines wherein messages play a relevant yet solely accessory role. In any event, semiotic occupies a central position within the total science of communication and underlies all other provinces of this science, while semiotic, in turn, comprises linguistics as its central section which influences all other semiotic provinces. Three integrated sciences encompass each other and present three gradually increasing degrees of generality: 1) Study in communication of verbal messages = linguistics; 2) study in communication of any messages = semiotic (communication of verbal messages implied); 3) study in communication = social anthropology jointly with economics (communication of messages implied).

Studies developing at present under such overlapping labels as sociolinguistics, ethnolinguistics or folk linguistics, and anthropological linguistics represent a sound reaction against some still frequent survivals of the Saussurian tendency to curtail the tasks and aims of linguistic research. Yet such constraints of aims and purposes superimposed by individual linguists or linguistic teams on their own investigatory program should not be labeled "pernicious"; any particular emphasis upon some limited sections of linguistic science or any degree of self-restriction and rigorous specialization is perfectly legitimate. What would be, however, erroneous and pernicious is any degradation of all the other facets of language as supposedly residual, second-rate linguistic questions, and, especially, any attempt to expel these topics from linguistics proper. Linguistic experimentation may deliberately cut off certain inherent properties of language. Such were, for example, experiments in American linguistics with the exclusion of meaning, first from linguistic analysis in general, and later at least from grammatical analysis. Such have been, also, the recently revived Saussurian propensities to confine analysis merely to the code (langue, competence) in spite of the [|667|] indissoluble dialectic unity langue/parole (code/message, competence/performance).

None of such eliminative experiments, however useful and instructive they are, can be viewed as a compulsory narrowing of the total scope of linguistic science. The various tasks and questions lately advanced and discussed under such labels as sociolinguistics all deserve a thorough study, and, one must add, many of these topics have behind them a long history of international research, and their local oblivion is of short duration. All these items, however, form an integral part of linguistics and require the same structural analysis as any other intrinsic constituent of language. By the way, I deliberately refer to "structural analysis", because whatever the factional dissensions among contemporary linguists and their divergences in terminology, technique and chief interests are, the analysis of linguistic structures is the common denominator of all the contemporary scientific currents, one which sharply distinguishes the quest of the last four or five decades from the main ways and targets of linguistics in the early twentieth and late nineteenth centuries.

The domain of ethnolinguistics and sociolinguistics - we can but agree with a farsighted promoter of their program, Dell Hymes - must and, finally, will be simply incorporated into linguistics, because the latter cannot be separated and isolated from "questions of the actual functioning and role of language in human life".

Any verbal code is convertible and necessarily comprises a set of distinct subcodes or, in other words, functional varieties of language. Any speech community has at its disposal 1) more explicit and more elliptic patterns, with an orderly scale of transitions from a maximal explicitness to an extreme ellipsis, 2) a purposive alternation of more archaic and newfangled dictions, 3) a patent difference between rules and ceremonial, formal and informal, slovenly speech. The areally distinct and manifold sets of rules permitting, prescribing, or prohibiting talk and science are destined to serve as a natural preface to any veritably generative grammar. Our linguistic performance is, furthermore, governed by a competence in dialogic and monologic rules. In particular, the varied verbal relations between the addresser and the addressee build a substantial part of our linguistic code and border directly upon the grammatical categories of person and gender. The grammatical and lexical rules relating to the present and absent differences in the hierarchical standing, sex, and age of the interlocutors cannot be bypassed in a thorough and accurate scientific description of a given language, and the place of these rules in the total verbal pattern raises a challenging linguistic question. [|668|]

The diversity of interlocutors and their mutual adaptability are a factor of decisive importance for the multiplication and differentiation of subcodes within a speech community and within the verbal competence of its individual members. The variable "radius of communication", according to Sapir's felicitous term involves an inteldialectal and interlingual exchange of messages and usually creates multidialectal and sometimes multilingual aggregates and interactions within the verbal pattern of individuals and even of entire communities. An exact comparison of the usually wider competence of the individual as hearer and his narrow competence as speaker are pertinent but frequently overlooked linguistic task.

Centrifugal and centripetal forces displayed by territorial and social dialects have been already for many decades a favorite subject in world linguistics. The recent application of structural analysis to the field work in social dialectology once more disproves the myth of homogeneous speech communities, discloses the speaker's awareness of variations, distinctions, and changes in the verbal pattern, and, thus, brings new illustrations to our view of metalanguage as a crucial intralinguistic factor.

The necessity to cope with the problems of standardization and planning, and herewith to put an end to the last survivals of the neogrammarian noninterference in the life of language belongs to the urgent linguistic tasks vitally connected with the progressively increasing radius of communication.

Our cursory survey of topics itemized in the recent programs of sociolinguistics shows that all of these questions require a strictly and intrinsically linguistic analysis for they are pertinent and inalienable part of linguistics proper. William Bright shrewdly points out the common denominator of these programs: "linguistic diversity is precisely the subject matter of sociolinguistics". Yet this same diversity may be characterized as the chief target of international linguistic thought in its endeavors to overcome the Saussurian model of langue as a static, uniform system of mandatory rules and to supplant this oversimplified and artificial construct by the dynamic view of a diversified, convertible code with regard to the different functions of language and to the time and space factors, both of which were excluded from the Saussurian conception of the linguistic system. As far as this narrow conception finds ts adepts again and again, we must repeat that any experimental reduction of linguistic reality can lead to valuable scientific conclusions so long as we do not take the deliberately narrow frame of the experiment for the unrestricted linguistic reality. [|669|]

Since verbal messages analyzed by linguists are linked with communication of nonverbal messages or with exchange of utilities and mates, the linguistic research is to be supplemented by wider semiotic and anthropological investigation. As foreseen in Trubetzkoy's letter of 1926, the integrated science of communication is intended to show, according to Bright's formulation, "the systematic covariance of linguistic structure and social structure". Or, in Benveniste's terms: "le problème sera bien plutôt de découvrir la base commune à la langue et à la société, les principes qui commandent ces deux structures, en définissant d'abord les unités qui dans l'une et dans l'autre, se prêteraient à être compareées, et d'en faire ressortir l'interdépendance".

Lévi-Strauss contemplates the path of such future interdisciplinary research: "Nous sommes conduits, en effet, à nous demander si divers aspects de la vie sociale (y compris l'art et la religion) - dont nous savons déjà que l'étude peut s'aider de méthodes et de notions empruntées à la linguistique - ne consistent pas en phénomènes dont la nature rejoint celle même du langage [...] il faudra pousser l'analyse des différents aspects de la vie sociale assez profondément pour atteindre un niveau où le passage deviendra possible d'exprimer les proriétés communes aux structures spécifiques relevant de chaque aspect. L'emploi de ce code devra être légitime pour chaque système pris isolément, et pour tous quand il s'agira de les comparer. On se mettra ains en position de savoir si l'on a atteint leur nature la plus profonde et s'ils consistent ou non en réalistés du même type". He envisages a "dialogue" with linguists on relations between language and society. One may recall Durkheim's complehension of the ever increasing superiority of linguistics among social sciences and his paternal admonition to build up a "linguistic sociology". Until now, however, the initial steps in this direction were taken precisely by linguists, as, for instance, in the stimulating attempts toward a correlation of language and sociocultural problems made in Russian linguistic literature on the threshold of the 1920's and '30's. Sociologists acknowledge "the cruel truth" that awareness of language can do more for sociology than sociology can do for linguistic studies, and that the lack of training "in formal linguistics" hinders workers in the social sciences from achieving a productive concern with language.

The variable radius of communication, the problem of contact between the communicants - "communication and transportation" - aptly advanced by Parsons as the ecological aspect of the systems, prompts [|670|] certain correspondences between language and society. Thus, the striking dialectal homogeneity of nomads' languages bears an obvious relation to the wide radius of nomadic roaming. In hunting tribes, for long periods hunters remained out of communication with their women but in close contact with their prey. Hence, their language undergoes a noticeable sexual dimorphism reinforced by the multiform taboo changes which hunters introduce in order not to be understood by animals. (Jakobson 1971[1969c]: 665-670)

["Linguistics in relation to other sciences"] [Online access] ["Based on the report in the Plenary Session of the Tenth International Congress of Linguists, Bucharest, 30 August 1967, and written as a part of the author's survey "Linguistics" for the Unesco publication Main Trends in Social Research [...]"]

Jakobson, Roman 1971d. Retrospect. In: Rudy, Stephen (ed.), Selected Writings II: Word and Language. The Hague; Paris: Mouton, 711-724. Online access: https://archive.org/details/selectedwritings02jako/page/712/mode/2up

The late nineteenth and very early twentieth centuries were marked by a continuous upsurge of comparative historical studies. At the same time, however, tentative writings of lone seekers in different countries reveal the first, precursory inklings of a prospective, structural approach to language. These anticipation and efforts culminate in Ferdinand de Saussure's Cours de linguistique générale, a posthumous edition of 1916 arranged by Ch. Bally and A. Sechehaye on the basis of students' records. The five subsequent decades have witnessed an unprecedented, strenuous rise and capital revision of the linguistic science, and the clearest way to pount out the essential innovations will be to confront them with the [|717|] Saussurian doctrine, which has been viewed as the start of an ew era in the science of language.

Most of the cardinal theoretical concepts and principles introduced by Saussure go back to his older contemporaries, Badouin de Courtenay and Kruszewski; but in the Cours some of these notions were presented in a more perspicuous and expanded manner, and an affective emphasis was placed on the mutual solidarity of the system and its constituents, on their purely relative and oppositive character, and on the basic antinomies which we face when we deal with language. It must be added, however, that the factual analysis of linguistic systems was a task passed on to the future researchers, and the elaboration of the more appropriate methods for such an analysis has become a vital qusetion of linguistic theory and praxis for several decades.

The consistent attention focused upon the antinomies "qu'on recontre dès qu'on cherche à faire la théorie du langage" is one of the greatest assets of the Cours. It was important to realize these dichotomies, but as long as they remained unresolved, the wholeness and unity of linguistics was imperiled. In Husserl's terms, "Halbheiten oder unzulässige Verabsolutierungen von nur relative und abstraktiv berechtigten Einseitigkeiten" had to be overcome, and gradual efforts to bridge and synthesize these "inner dualities" actually mark the post-Saussurian stage of linguistics.

At the very end of his scientific activities, Saussure adopted the Stoic conception of the twofold verbal sign composed of the perceptible signans and the intelligible signatum. He realized that these two elements are intimately united "et s'appellent l'un l'autre', but taught that the bond between the signans and signatum is arbitrary and that "the whole system of language is based on the irrational principle of the arbitrariness of the sign". This assumption has been submitted to a gradual revision whereby the role of relative, grammatical motivation as invoked by Saussure to restrict the arbitrariness of the connection between the two aspects of the verbal sign proved to be quite insuffirient. Inner, iconic ties of the signans to its signatum and, in particular, intimate connections between grammatical concepts and their phonological expression cast doubt on the traditional belief in "the arbitrary nature of the linguistic sign" maintained in the Cours. In post-Saussurian linguistics the question of relationship between the signans and signatum has been extended also to the phonological aspect of language, and the entangled questions of interplay between the phonological and grammatical levels as well as of their mutual demarcation have come to the foreground of linguistic [|718|] attention. The essential difference between the phonological oppositions which are rooted in the signans and the grammatical oppositions founded in the signatum has been apprehended.

"The linearity of the signans", which was decreed by Saussure to be a self-evident fundamental principle fraught with incalculaable consequences for the science of language, has been shaken by the dissociation of phonemes into the ir concurrent components ("distinctive features"); and, on the other hand, thequestion of the successive order in the structure of the signatum regains the importance it had in the classical age, and the increasing attention to the hierarchy of immediate constituents has removed the shortcomings of the former, straightforward approaches to the sequence. Saussure's remarks on the irrelevance of the "substance" in which the linguistic form is expressed and on the arbitrariness of the relation between form and substance were put to the test, and have finally yielded to a hierarchical view of primordial speech and its graphic substitutes and to a tenacious request for an exhaustive, comparative inquiry into the distinct autonomous properties of the oral and written varieties of language; sound patterns utilized for the building up of meaningful distinctions proved to be based on a semiotic selection and adaptation of natural phonic means; a typology of the extant phonological systems based on a strictly relational standpoint was attempted, and implicational laws of universal validity were inferred from it. A grammatical (morphological and syntactic) typology proves to be the next urgent task of such an inquiry, with a watchful attention to the multiple structural interrelations between these two dissimilar levels.

The Saussurian inner duality of langue and parole (which mirrors the synonymous distinction of jazyk and reč' launched by Baudouin de Courtenay in 1870) or, to use a modern, less ambiguous terminology, 'code' (Saussure's code de la langue) and 'message' - alias 'competence' and 'performance' - gives rise to two divergent approaches within the same section of the Cours: "Sans doute, ces deux objets sont étroitement liés et se supposent l'un l'autre", and, on the other hand, the author claims the impossibility of grasping "le tout global du langage", insists on a strict bifurcation of the inquiry into langue and parole, and even declares the former as the sole object of linguistics proper. Although this restrictive program still finds its theoretical adherents, in fact the absolute separation of the two aspects turns into a recognition of two different hierarchical relations: an analysis of the code with due regard for the message, and vice versa. Without a confrontation of the code with the message, no insight into creative power of language can be achieved. [|719|] Saussure's definition of langue as "la partie sociale du langage, extérieure à l'individu" in opposition to parole as a mere individual act does not consider the existence of a personal code which removes the temporal discontinuity of the single speech events and which confirms the preservation of the individual, the permanence and identity of his ego; nor does he take into account the interpersonal, social, mutually adaptive nature of the "circuit de la parole" which implies the participation of at least two individuals.

The uniformity of the code, "sensibly the same" for all the members of a speech community, posited by the Cours and still recalled from time to time, is but a delusive fiction; as a rule, everyone belongs simultaneously to several speech communities of different radius and capacity; any overall code is multiform and comprises a hierarchy of diverse subcodes freely chosen by the speaker with regard to the variable functions of the message, to its addressee, and to the relation between the interlocutor. In particular, the subcodes offer a scale of transforms ranging from explicitness to the gradual degrees of phonological, grammatical, and narrational ellipsis. When one-sided concentration on the cognitive, referential function of language gave way to an examination of its other, likewise primordial, underivable functions, the problems of the code-message relationship showed much greater subtlety and multivalence.

La langue, according to the Cours, "must be studied in itself", and it "never requires premeditation" on the part of speakers. The new rapid progress of applied linguistics with such items as language planning and policy, language teaching, communication engineering, etc., is a natural and predictable offshoot of the modern goal-oriented linguistic thought, but it remains alien to Saussure's view of linguistic science and to the predominant scholarly ideology of his time.

Saussure obviously followed Kruszewski in teaching that the 'generative' operations of language involve two kinds of - one, relying upon selection, was characterized by him as 'associative', 'intuitive', or 'paradigmatic', while the other, based on combination, was named 'syntagmatic' or 'discursive'. The terms 'paradigmatic' and 'syntagmatic' have entered into general use, but the interpretation of these two notions and of their interdependence has undergone substantial changes. The Cours affirmed that the members of a paradigmatic series have no fixed order "et c'est par un acte purement arbitraire que le grammairien les groupe d'une façon plutôt que d'une autre"; at present, however, this agonistic habit is being supplanted by an insight into the objective stratification within any series which displays a [|720|] set of correlations between the lack and presence of 'markedness' or, in a different formulation, between relatively nuclear ('deep') and accessory, secondary structures. (Jakobson 1971d: 716-720)

["Retrospect" to SW II] [Online access]

Jakobson, Roman 1985[1972b]. Verbal communication. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 81-92. DOI: https://doi.org/10.1515/9783110855463.81

In the Neogrammarian tradition the notions and labels "comparative" and "general" linguistics nearly merged, and the comparative method was confined to a merely historical or, strictly speaking, genealogical study of cognate dialects and languages. Today virtually any linguistic problem whatever has received a thoroughly comparative treatment. Any question of language and languages is conceived of as being a comparative operation in search of the equivalent relations that underlie the structure of a given language, and that furthermore allows us to interpret the structural affinities and divergences between languages, however distant they may be in origin and location. The decisive procedure for scientific inquiry into the different levels of linguistic structure is a consistent elicitation and identification of relational invariants amid the multitude of variations. The variables are investigated with reference to the set of diverse transformations that they undergo and that can and must be specified.

Whatever level of language we deal with, two integral properties of linguistic structure force us to use strictly relational, topological definitions. First, every single constituent of any linguistic system is built on an opposition of two logical contradictories: the presence of an attribute ("markedness") in contraposition to its absence ("unmarkedness"). The entire network of language displays a hierarchical arrangement that within each level of the system follows the same dichotomous principle of marked terms superposed on the corresponding unmarked terms. Any second, the continual, all-embracing, purposeful interplay of invariants and variants proves to be an essential, innermost property of language at each of its levels.

These two dyads - markedness/unmarkedness and variation/invariance - are indissolubly tied to the be-all and end-all of language, to the fact, as Edward Sapir (1884-1939) put it, that "language is the communicative process par excellence in every known society". Everything language can and does communicate stands first and foremost in a necessary, intimate connection with meaning and always carries semantic information. The promotion of meaning to a pivotal point of structural analysis has been an ever stronger claim of international linguistic endeavors during the past five decades. Thus, for instance, 20 years ago the French linguist Émile Benveniste, one of the leading figures of the structural trend, declared in a programmatic study that in the final account careful reflection on the makeup of any language points [|86|] to the "central question of meaning", and that a deepening insight into this problem will open the way to the future discovery of "transformational laws in linguistic structures".

True, various reductionist experiments were conducted in America. At first repeated efforts were made "to analyze linguistic structure without reference to meaning". Some later tests confined the removal of meaning to the study of grammatical structures under such slogans as "Linguistic description minus grammar equals semantics". All these tentative operations were undoubtedly of considerable interest, particularly since they succeed in providing us with a graphic demonstration of the omnipresent semantic criterion, no matter what level and constituent of language is examined. One can no longer continue to play hide-and-seek with meaning and to evaluate linguistic structures independently of semantic problems. Whatever end of the linguistic spectrum we deal with, from the phonic components of verbal signs to the discourse as a whole, we are compelled to bear in mind that everything in language is endowed with a certain significative and transmissible value.

Thus in approaching speech sounds we must take into account the fact that they are cardinally different from all other audible phenomena. An astounding discovery of the recent past is that when two sounds are presented simultaneously to both ears, any verbal signals such as words, nonsense syllables and even separate speech sounds are better discerned and identified by the right ear and all other acoustical stimuli such as music and environmental noises are better recognized by the left ear. The phonic components of language owe their particular position in the cortical area, and correspondingly in the aural area, solely to their verbal functions, and henceforth a constant regard for these functions must guide any fruitful study of speech sounds.

In its sound pattern any language contains a certain limited number of "distinctive features", discrete and ultimate relational invariants that can, under a set of transformations, endure even drastic alterations in every respect save their defining attributes. "The categorial nature of perceptual identificatio", pointed out by the psychologist Jerome S. Bruner in his memorable study "Neural Mechanisms in Perception" (1956), maintains the constancy and validity of these features in verbal communication, where they exercise the fundamental faculty of semantic discrimination.

The pattern of distinctive features is a powerful and economic code: each feature is a binary opposition of a present mark and a missing [|87|] mark. The selection and interconnection of distinctive features within any given language reveal a remarkable congruity. A comparison of the existing phonological structures with the laws underlying the development of children's language enables us to outline the typology of feature systems and the rules of their internal hierarchical arrangement. The communicative relevance of distinctive features, which is based on their semantic value, brings to naught any chance occurrence and contingency in their patterning. The list of distinctive features that exist in the languages of the world is supremely restricted, and the coexistence of features within one language is restrained by implicational laws.

The most plausible explanation of these either totally or nearly universal principles in regard to the admissibility and interconnection of features apparently lies in the internal logic of communication systems that are endowed with a self-regulating and self-steering capacity. The quest for a universal table of distinctive features must certainly apply the same method of extracting invariants that has been used with respect to single languages: in the context of different languages the same feature with unaltered categories attributes may vary in its physical implementation.

Transformations that provide the invariants with diverse concomitant variations can be roughly divided into two kinds of alteration: contextual and stylistic. Contextual variants point to the concurrent or consecutive neighborhood of the given feature, whereas stylistic variants add a marked - emotive or poetic - annex to the neutral, purely cognitive information of the distinctive feature. Both of these invariants and variations belong to the common verbal code that endows interlocutors with the competence to understand one another.

For the study of verbal communication it is necessary to face the fact that any speech community and any existing verbal code lack uniformity; everyone belongs simultaneously to several speech communities of different extent; he diversifies his code and blends distinct codes. At each level of the verbal code we observe a scale of transitions that range from maximum explicitness to the briefest elliptic structure, and this scale is subject to a set of rigorous transformational rules. The cardinal property of language noted by the initiator of semiotics, Charles Sanders Peirce (1839-1914), namely the translatability of any verbal sign into another, more explicit one, renders an effective service to communication in that it counteracts ambiguities caused by lexical and grammatical homonymy or by the overlapping of elliptic forms.

People usually display a narrow competence as senders of verbal [|88|] messages and a wider competence as receivers. The differences in patterning and extent between the codes of the addresser and the addressee attract ever closer attention from students and teachers of language. The core of this divergence was grasped by St. Augustine: "In me it is the word which takes precedence over the sound [In me prius est verbum, posterior vox], but for thee who looks to understand me, it is first the sound that comes to thine ear in order to insinuate the word into thy mind." The two-way transformations that make it possible to determine the state of the outputs from that of the inputs and vice versa are an essential prerequisite for all genuine intercommunication.

Both spatial and temporal factors play a significant role in the structure of our verbal code. Various forms of interdialectal code-switching are among the daily devices in our verbal intercourse. Bilingualism or multilingualism, which allows total or partial shifts from one language to another, cannot be rigidly separated from interdialectal fluctuations. The interaction and interpenetration of single languages in a polyglot's use follow the same rules that apply in the case of translations from one language into another.

As for the time factor, I refer to my earlier objections to the tenacious belief in the static character of the verbal code: Any change first appears in linguistic synchrony as a coexistence and purposive alternation of more archaic and new-fashioned dictions. Thus linguistic synchrony proves to be dynamic; any verbal code at all its levels is convertible, and in any conversion one of the competing alternants is endowed with a supplementary informational value and hence displays a marked status, in contradistinction to the neutral, unmarked character of the other. A historical phonology of grammar, for example, develop into as tudy of extractable constants and temporal transformations that both demand an adequate explanation.

The unparalleled expediency of language is rooted in a consistent superposition of several interconnected levels, each of which is differently structured. The system of a few distinctive features serves to build a more differentiated morphological code of entities endowed with inherent meaning, namely words and, in those languages where words are decomposable, their minimal meaningful constituents (roots and affixes), termed morphemes. The analysis of the morphological units once again reveals a system of relational invariants - but there is a difference of basic importance between a phonological and a grammatical [|89|] opposition: in the former case the coupled contradictories reside in the perceptible side of language (signans, or "signifier"), whereas in the latter they lie in its intelligible side (signatum, or "signified").

To illustrate this difference let us first cite an opposition of a phonological mark and its absence: nasalized/non-nasalized implemented by such pairs of consonants as m/b and n/d or the French nasal vowel in bon as opposed to beau. On the other hand, in a grammatical opposition such as preterit/present the first, marked tense signals the precedence of the narrated event over the speech act, whereas the general meaning of the unmarked present tense carries no information about the relation between the narrated event and the speech act. This relation varies and itst specification depends on the context. Compare the diverse contextual meanings of the same present-tense form in the four sentences "Spring begins today"; "A year from today he begins a new trip"; "With the death of Caesar a new era begins for Rome"; "Life begins at 50."

Here, again, as when treating the sound pattern, we come across the momentous property of natural languages, namely their context-sensitivity. Precisely this property sets them apart from their formalized, artificial superstructures, which tend to a context-freedom. The significant difference between context-free and context-sensitive sign systems had been perspicuously noted by Noam Chomsky, but as Daniel A. Walters complains in Information and Control (1970), the specific properties of context-sensitive grammar still receive much less attention than context-free grammars. It is the context-sensitivity of a natural language at all levels that provides it with a unique abundance of free variations. The dialectical tension between invariants and variables, which in their own way also appear to be pertinent, ensures the creativity of language.

Morphology answers the phonological pattern of distinctive features with an equally coherent and step-like organization of equally binary conceptual features; they remain invariant while undergoing a set of transformations that convert the general meanings of grammatical categories into varied contextual (including situational) meanings. In this way we proceed from one grammatical area to a superior one, namely from morphology as a study of coded units to the analysis of syntactic structures that combine coded matrices with a free or, as is always the case in verbal communication, relatively free selection of words that fill them up.

Words display two patently distinct kinds of semantic value. Their compulsory grammatical meaning, a categorial relational concept or [|90|] group of concepts that words constantly carry, is supplemented in all autonomous words by a lexical meaning. Like grammatical meanings, any general lexical meaning is in turn an invariant that under diverse contextual and situational transformations generates what Leonard Bloomfield (1887-1949) precisely defined as "marginal, transferred" meanings. They are sensed as derivative of the unmarked general meaning, and these tropes either stand in agreement with the verbal code or they are an ad hoc digression from it. (Jakobson 1985[1972b]: 85-90)

["Verbal communication"] [DOI: 10.1515/9783110855463.81] ["Written in Cambridge, Mass., June 1972, for a special issue of Scientific American on communication (vol. 227, no. 3, Sept. 1972)."]

Jakobson, Roman 1985[1977a]. The grammatical buildup of child language. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 141-147. DOI: https://doi.org/10.1515/9783110855463.141

Works such as the books of Grégoire or of the zealous and keensighted investigator of Russian child language acquisition Aleksandr Nikolaevič Gvozdev might do justice to their precise results, and to their wealth of linguistic and psychological data, their abundance and thoroughness. In sharp contrast are today's ever-increasing snapshots of linguistic behavior and current linguistic ability, taken of the child a few times weekly or even monthly by the learned visitor, which chop up and destroy the total picture. The latter conceal the individual mainsprings of the changes and their order of succession, and thus make it impossible for us to draw conclusions regarding the dynamic laws of language acquisition.

To be sure, important new observations have been made concerning children's beginning attempts and efforts to learn to communicate at the pre-verbal age, and at the threshold of verbal exchange with their teacher, in particular their mother. Much new material has also been observed and clarified regarding the sequence of children's phonological acquisition, but the grammatical buildup of child language still remains unexplained in many, or perhaps even in most, respects. [|142|]

The question which from the very beginning interested me in particular was that of the relative role of imitation and of the child's creative gift in the acquisition of his first language. While certain investigators stress the importance of replication, others emphasize that of creative ability. It appears that a synthesis would be preferable. What does take place of the void. Imitation offers broad possibilities for the beginner's creative powers. The existing model permits a selection from the completed acquisitions and their proper sequence, thanks to which the child can acquire first one thing and only then the next. Universal laws of one-sided implication (or foundation) are at work here: no "B" could arise as long as "A" had not developed, whereas in the linguistic system "A" can exist independently of "B".

The investigation and explanation of the strict lawfulness observable in the development of child language progress only slowly. The Neogrammarian, unilaterally diachronic tradition, which was prominent in linguistics only a short time ago, showed no appreciation whatever either for the pure description of language structure or for the search for universal laws of structuring and restructuring. On the other hand, the Geneva school and its numerous epigones preached a strict and narrowly static version of synchrony and, as a matter of fact, of linguistic analysis as a whole, and concealed the means of comprehending child language, the nature and structure of which are subject to the laws of synchrony applied dynamically.

If, in spite of the residue of protracted skepticism, one now finally observes the wealth and meaning of universals in the structure of the world's languages and in the buildup of child language, many strikingly dynamic laws manifest themselves, especially in the development of the phonological inventory. These laws show either a general validity or at least a nearly universal probability. Two kinds of monopolistic attempts to explain this wide-ranging uniformity in the fundamentals of human language are apparent. On the one hand, slogans for a fanatical nativism issue forth and multiply; on the other hand, a persistent sociologism tries to convince us that all linguistic laws and tendencies toward uniformity are determined by the unswerving social usage and character of language.

As proof of the "innateness" of the fundamenal laws of language, nativists adduce the relative ease and speed of language acquisition in small children the world over. As a matter of fact, however, children also learn, with the same naturalness, accuracy and ease, all of the external local particulars of the linguistic milieu to which they owe their first [|143|] knowledge of language. The alleged rapidity of complete language acquisition likewise proves to be an exaggerated over-generalization. But the universally human and uniquely human desire and gift of mastering a language actively and passively from early childhood on is a capability embodied in the biological fact of being human, for, as Goethe says: "Each one learns only what he can."

Still, one must not forget - and it is forgotten, strangely enough, all too often - that, strictly speaking, what is acquired is a dialogue. For the child's speech and language two speech partners are necessary: on the one hand, the underage novice, and on the other, an older, more experienced companion, in particular the child's mother. Therefore, the development of language cannot be understood without bearing in mind, from the very first, two parties, two participant, of whom the one learns and the other in fact teaches. Today we often hear frivolous statements to the effect that the child requires no language instruction and quite independent production. Critics who deny or belittle the role of learning or teaching in the child's acquisition of language are neither laboring under the bureaucratic conception of an official educational establishment, or else are hypnotized by the once fashionable idea of a gap between parents and the younger generation. Close reciprocal accommodation remains an essential concomitant feature of language learning.

We are already in a position to establish the existence of numerous structural principles which languages hold in common, laws operating universally or nearly universally, among the latter - widely distributed in space and time - tend toward exceptionlessness, without, to be sure, attaining it. Naturally, the biological basis plays a certain role, although one that remains indeterminate. On the other hand, one must never forget that language presents an essentially social totality, and that this dynamic whole, in the terms of the German philosophers, shows a constant self-motion (Selbstbewegung) - self-stirring, according to the terminology of cybernetics. The dynamic laws of language are furthered by the fact that an all-embracing collective system is involved, which its user comes to experience at the boundary between the conscious and the unconscious. To the subliminal domain belongs the striving after appropriate structures, which are followed by the linguistic community, and partly also consciously, under more metalinguistic direction.

Language is first and foremost characterized by its essentially universal stratificational structure and is subject to a basically uniform [|144|] principle of superimposed arrangement. Each level consists of internal relations exclusively its own, and of relations which bind each level with the others. The analysis of these relations, both the internal as well as the interlevel ones, is indispensable for understanding the system of language and its buildup. (Jakobson 1985[1977a]: 141-144)

["The grammatical buildup of child language"] [DOI: 10.1515/9783110855463.141] ["First presented at the 204th Session of the Rheinisch-Westfälische Akademie der Wissenschaften on May 28, 1975 in Düsseldorf and published in German, with the discussions following the report, in the Vorträge of the Academy (Geistwissenschaften, G 218, 1977)."]

Jakobson, Roman; Waugh, Linda R. 1988[1979d]. Speech sounds and their tasks. The Sound Shape of Language. Rudy, Stephen (ed.), Selected Writings VIII: Major Works 1976-1980. Completion Volume 1. Berlin; New York: Mouton, 7-82. DOI: https://doi.org/10.1515/9783110862744.7

XXIII. MULTIFORMITY AND CONFORMISM
Linguists, even when interested chiefly in oral speech, often unwittingly give way to the hypnosis of written language. It is peculiar that in discussing the order of some verbal units in a sequence they use the terms "left" and "right" instead of "before" and "after" and speak about the "left-hand" and "right-hand" environment of a speech sound. In the examples of linguistic ambiguities they quote sentences ambiguous merely in writing and perfectly distinguishable in their explicit oral form. It is perhaps under the influence of the higher uniformity proper to the code of written lnaguage that sometimes the idea of a rigorously monolithic code of language in general captures theoreticians and tempts them to believe in the puerile myth of a perfectly invariable speect community with equally competent speaker-hearers and to apply the delusive idea to concrete operations. However, "real individuals command a variety of related linguistic systems", a variety of styles of speech used in a range of social situations (as was succinctly noted by Chomsky & Walker 1976): "Individuals within a speech community may differ in these respects and speech communities sometimes may vary quite widely in the systems represented within what is popularly called a single language" (p. 21). Hence, any actual linguistic status contrasts strongly with the assumption of a grammar "uniformly represented as a single invariant system in the mind of each English speaker". [|78|]

Witold Doroszewski (1899-1979), who was hostile to the idea of relational invariance in the sound structure of language, paid particular attention to the abundant diversity of variants in the everyday speech of Polish peasants which he recorded in his field work. These minute observations are particularly valuable because, contrary to the observer's anti-unitarian tenet, they bring to light the orderliness within the obvious variety. The several exponents of the Polish nasal ę used by all members of a rural Polish speech community near Plock were recorded and described in his French paper of 1935 and, with more detail, in a previous Polish report of 1934. The basic variants stand out against the marginal ones and the preponderant cases display competition and compromises between opposite tendencies: nearer vs. more distant in space or time, and either disappearing or developing; rural traits compete with urban influence; articulatory memory clashes with lexical borrowings preserving their sound shape. A closer interpretation of this eloquent "jeu des formes flottantes" could to a large degree find the key to this selection in the changeability of contexts and topics, momentary universe of discourse, difference of speech functions, and switching relations between speakers and hearers. According to Doroszewski's belief, the individual himself is completely passive and makes no deliberate choice between diverse possible solutions; he is influenced by opposite tendencies, whatever the different between them: "les deux le travaillent, coexistent en lui". Doroszewski concluded that the system is a "Procrustean bed", but his dialectal material is in fact an excellent plaidoyer for the pluralism of the verbal code both against the narrowly empirical denial of system and again the reductionist denial of multiform systems.

The expert observer William Labov (1964, 1970) exemplified the motivated and structured variability of sound forms in their interpersonal and personal usage and uncovered "a pattern of continuous and regular variation through different styles and contexts". He noted specific properties of speech "used in informal situations where no attention is directed to language" and in particular occurrences of spontaneous, excited speech "when the constraints of a formal situation are discharged". The science of language cannot but agree with Jacob's reference to the two apparently opposite properties of living beings - stability and variability - as "an inherent quality of the very nature of living systems". Any unbounded generalization of the stability and uniformity principle proves to be an impoverishing, stultifying "idealization" of the heterogeneous linguistic reality. "The association [|79|] between structure and homogeneity is an illusion", as Weinreich et alii state, following in the footsteps of the Praguians; "the concept of a variable as a structural element makes it unnecessary to view luctuations in use as external to the system, for control of such variation is a part of the linguistic competence of members of the speech community" (1968: 185-187).

Like any other social system, language is in continual motion and self-generating development (cf. Lange 1962: 73 ff). The verbal code and in particular the sound pattern of any language constantly undergoes changes. In contradistinction to daylight savings time or to spelling reforms, which can be decreed and enter into common practice on a definite date, the strat and finish of a sound change in spoken language go through a period of coexistence; they belong to two styles, two subcodes of the same language, and are actively used either by different speakers or by one speaker who oscillates between "archaïsm" and the "modernism". Speakers and hearers may be aware of the time axis to which both items belong, and time itself thereby enters into the verbal system as a semiotic value. The belief earlier voiced among linguists that the process of linguistic change is never directly observed does not take into account the vital phenomenon of speaker's preoccupation with speech itself and their habitual metalinguistic talk about talking. There are frequent cases of a generational difference between interlocutors, the youngest of whom make use of the nascent innovations which the older ones understand but have not included in their speaker repertory. Similarly, the younger speakers comprehend the older ones altohugh the younger no longer actively use the elements they deem "outdated". Besides such cases of manifest division between speakers and listeners, there obviously also occur frequent instances of mutual adaptation in intercommunication between people of different generations. Members of a speech community are competent to use both th start and the finish of the change, and the overall code of the given language must correspondingly be conceived of as convertible. Thus the two stages of a change in progress should be interpreted in terms of a dynamic synchrony. Concurrence and successivity are, therefore, interrelated both in single utterances and in the "aoverall code" of language (cf. Hockett 1958).

The tendency of Saussure's Cours to reduce the structure of any language system to concurrence (simultaneity) with disregard for successivity (temporal succession) - "l'axe des simultanéités, concernant les rapports entre choses coexistantes, d'où toute intervention du temps [|80|] est exclude" (I: 177) - is an equally arbitrary and impoverishing a tendency as is the reverse attempt of the same Cours to discard concurrence from the structure of linguistic units and to confine them to a mere linearity - "caractère linéaire du signifiant" (Saussure I: 157).

With respect to the variable radius of communication from the nuclear family to interdialectal or even crosslingual verbal intercourse, the multiplicity of tasks again involves different subcodes adapted to changing interlocutors; herewith space enters into the overall code of the speech community and its members as a semiotic value. In both their temporal and spatial aspects the code and the circuit of messages exhibit a perpetual interplay between conformism and nonconformism. A spatial conformism, adaption to neighbors, usually implies a temporal nonconformism, in other words, a temporal discontinuity. On the other hand, there is normally a linkage between temporal conformism and discontinuity with the neighbor's pattern, alienated from it. The repeated assumption of an essential difference between the 'source' (Saussurian foyer) of a linguistic innovation and the area of its "contagion" and propagation clashes with the fact that any change is a phenomenon of propagation, from a slip of the tongue to its repetition and acceptance first by a narrow and then by a wide collective body; a change and its diffusion appear to be but two facets of one and the same ongoing "contagion". The plurality of subcodes and the transition from one subcode to another is an essential constituent of linguistic competence possessed by an individual and mutatis mutandis by his milieu. Doroszewski's records, discussed above, belong to a series of papers compiled by various field workers in order to prove that individual speakers do not use any integral system, since the texts recorded and publish exhibit an extremely mixed character. If, however, one subjects these variables to an attentive analysis, then, may we repeat, all the supposedly conflicting textual peculiarities easily find a natural explanation in the alternation of thematic and stylistic factors and of addressees, and these variations display a complex integral system. The belief of the recorders in variability without integration is no less illusory than the belief of a theoretician in integral competence without inner variation. Also, in conditions of close contact between two dialects or two languages, people with bidialectal or bilingual competence can enjoy a prestigious position in their linguistic environment, and the partial merger of two neighboring codes which is typical of bilingual individuals or groups favors a wider expansion of certain [|81|] particularities of one language, especially of its sound shape, to the adjacent tongue. (Jakobson & Waugh 1988[1979d]: 77-81)

["Speech sounds and their tasks" in The Sound Shape of Language] [DOI: 10.1515/9783110862744.7]

Jakobson, Roman; Waugh, Linda R. 1988[1979d]. The network of distinctive features. The Sound Shape of Language. Rudy, Stephen (ed.), Selected Writings VIII: Major Works 1976-1980. Completion Volume 1. Berlin; New York: Mouton, 125-179. DOI: https://doi.org/10.1515/9783110862744.125

X. DYNAMIC SYNCHRONY
Linguistic facts and linguistic theory imply each other, and are equally indispensable for a clear insight into the buildup of child language and of its sound shape. A lack of theory is the most hazardous and most speculative theory. The idea of language as a structured, coherent system of devices from the smallest to the highest units has for ages been enrooted in sciences striving against the superstitious and lifeless image of a fortuitous aggregate of scattered particulars. Wilhelm von Humboldt, great heir to a centuries-old tradition of philosophical grammar and precursor of present-day linguistic views, insistently claimed that "nothing in language stands by itself but each of its elements acts as a part of a whole"; he contended "dass in der Sprache Alles durch Jedes und Jedes durch Alles bestimmt wird". The designation of language as a system "où toit se tient" has entered inte the primers of French and international linguistics.

Language remains a coherent system both in its being and in its becoming. Children's nascent language is but a particular case of language in becoming, and must be viewed as a developing system. Such a consideration is fully valid also with respect to the sound shape. The universal rules and tendencies disclose the intimate conjunction between the constituents of the pattern and the implicational order of these relations in the structuration of language as well as in the successive manifestations of this order in the child's linguistic [|169|] maturation. It is necessary to see the forest and not just the trees, and in the given case to see the whole network of distinctive features and their simultaneous and sequential interconnections and not just an apparent mosaic of unrelated acquisitions. In present-day scientific thought, according to the terse formulation of Oskar Lange (1904-1965), "system is taken to mean a set of elements together with the set of relations between the elements; the set of such relations (and of all their isomorphic transformations) is called the structure of the system".

The notion of structure is inseparable from that of 'transformation' (transmutation) and, as Lange points out, "wholes can never remain in a changeless state, they must change constantly". In this respect, language is not an exception among systems, and the frequent question "why do all languages keep changing all the time?" loses its edge. In linguistics, the incessant attempts to disconnect the concept of a system and of its structure from that of change lead to an unrealistic view of language and an oversimplified conception of its network. These attempts to inhibit the explanation of changes through a reluctance to see theme as necessary components of the system.

Between the changes in children's language during its acquisition and the changes in extant languages of entire speech communities there are appreciable differences. Thet ask of the former development is to construct language, while the latter process is aimed at the partial restructuration of language. However, in children's process of constructing their language, the restructuration of the adult model is vitally involved. In any case, there is a set of essential correspondences between both kinds of changes. Any change involving distinctive features exhibits a temporal distance between its initiation and conclusion. The interval between these two boundary markers is frequently characterized by students as a temporary disorganization - chaos, heterogeneity, mixture, an irregular state of 'unstructuredness' - whereas in fact the coexistence of entities which were compulsory before a change began with those which presomably will have monopoly after its completion does not mean a disruption of orderliness. If some of the members of the same speech community, in particular the senior ones, use solely the older forms, while the others, especially the junior ones, have constant recourse to the innovations, a new rule is added to the linguistic code of the whole community, concerning a difference in the forms employed by speakers and listeners of diverse ages. Another [|170|] limiting variation of basically one and the same linguistic status would be a preference for the older form by the younger generation when addressing older people and a corresponding concession by the older speakers to the younger listeners.

If, however, free choice between kinds of forms characterizes the same speakers independently of the addressees' age (and this seems to be a quite frequent case), the freedom of this variation is subject to significant restraints: the selection of older and newer forms depends on the style of speech - more or less hurried, slipshod, elliptic, informal, expressive, or the opposite of each of these modes. The elliptic origin of many sound changes is indisputable, for instance the developing loss of the vocalic lax ~ tense feature in contemporary French. Often the sound shapes of diverse words undergo the same change with different speed, a discrepancy which is closely tied to their different stylistic coloring. Thus we must for a given moment in the course of a certain sound change distinguish between the different subcodes which belong to the multilayered overall code of the whole community or of certain of its subgroups. An observer's heightened attention to such subgroups and their subcodes is necessary in order to avoid the myth of temporary disintegration.

As noted above, any change is a gradual diffusion of one of the variables among diverse speech uses and diverse speech users, with a gradual widening of the radius of intercommunication. As long as they co-occur for diverse communicational purposes, the older and newer shapes must be viewed as synchronically coexistent. They are part of the code of the same collective body even if theyare felt to pertain to temporally diversified layers of the overall code. Since the competing phases of the change prove to be copresent, the change in progress constitutes a part of one and the same synchronous cut (Saussure's section transversale). It is encompassed in a dynamic synchrony (cf. the stimulating contributions by Fónagy 1956 and Wang 1969 to the still insuffiriently examined questions of "competition between coexisting rivals").

The neogrammarian division of sound changes into 'spontaneous' vs. 'conditioned' has undergone a double modification, and most of the alleged 'spontaneous' changes in fact appear to be 'conditioned' ones. Firstly, a change constrained by the surrounding sequential context was formerly the only type to be labeled 'conditioned', whereas the change of a feature constrained by the concurrent, simultaneous features was not conceived of as 'conditioned' but was treated as if it [|171|] were 'spontaneous'. Secondly, it was not seen that the confinement of a given change to a single verbal style means a specific context through which the change is 'conditioned'. At present ever greater attention is being paid to the stylistic context of changes, regarded as the "embedding problem" (Weinreich et al. 1968: 185) and defined as the "location of the change within the lniguistic and social matrix which governs its development" (Labov 1972: 114). One may add that those factors of changes which had been envisaged as external now demand a revised interpretation. Take, for instance, the prestige pattern, which furthers the diffusion of features from dialect to dialect or from language to language. Whatever the reason for the privileged position of the sound pattern or vocabulary or grammatical structure of a given dialect or language with respect to others, this position becomes as intrinsic attribute assigned to one linguistic system in relation to some other one and hence requires an inner, strictly linguistic interpretation.

The neogrammarian belief, tenacious in historical linguistics, that grammatical constraints on a sound change are to be viewed as posterior to the completion of the change itself proves to be in disagreement with the observed facts and gives away to the recognition of (at least) an initial confinement of a sound change to some certain class of morphemes. The wider the scope of 'conditioned' changes becomes, the more obselete the ribric of 'sporadic exceptions' to a sound law appears to be. They can be "better accounted for in terms of internal factors of the implementation of sound change" (Chen 1972: 494). The state of a change in progress appears to Hans Vogt "as a more or less free variation between forms of expression equally admissible within the system" (1954: 367). This observer is right when he accents the limiting "more or less": the image of "free variants" equally admissible within a system pushes us to investigate the supposed "freedom" and "equality", which actually depend on an interplay of internal factors.

Scholars have been accustomed to assigning the accomplishment of a sound change to the movement when the fluctuations of the earlier stages have disappeared and no residual forms remain, so that the action of the sound law appears to be achieved. If these conditions were not fulfilled, one was supposed to state that the change "has come to a premature end" (cf. Chen: p. 493) and consequently remained unsuccessful (cf. Wang 1969: 16). However, if the grammatical and/or stylistic constraints are an inseparable part of the sound law in question, the [|172|] "unsuccessfulness" becmes a subjective characterization imputed by the detached onlooker who has been accustomed to the ausnahmslose Gesetze of the Leipzig tradition. As it was densely formulated in Osthoff & Brugmann's declaration of 1878: "Any sound change in its mechanical progression is affected according to exceptionless laws, viz. the direction of the sound alteration is for all the partakers of a speech community except the case that dialectal scission enters, alwys the same, and all words in which the sound undergoing alteration appears under the same conditions are subject without exception to the change" (p. XIII).

Several supposedly controversial questions thus seem to lose their edge. One of these concerns the duration of the change and in particular the notion of "changes that continue in the same direction over several generations" and perhaps even span centuries. If the phenomena discussed overstep the limits of synchrony and the generations involved actually do not coexist in time, then we are dealing with a train of successive changes, each of which is witnessed by a speech community at one of its consecutive historical stages. Each of these synchronously experienced changes has its own particular traits with regard to the other changes of the same historical chain, although the tendencies of the whole set of changes may show a homogeneous direction, especially in view of the continuoun overlap of generations. The historical comparison of these successive changes and of the temporal variables as well as static invariants in the evolution of the given linguistic system is the task of diachronic investigation.

Notwithstanding protracted discussions, there is no real contradiction between the view of gradual change and change per saltum (cf. Wang 1969: 14). The mutative, abrupt character of a change does not mean a sudden disappearance of old forms for the sake of the new one in the overall code of the speaker or of the speech community. It means on the one hand the newborn possibility of abandoning the old form and of accepting the new one in one of the individual or communal subcodes, and on the other hand the irreversible loss of the old form and the emergent monopoly of the new one.

One of the facets attributed to the idea of "gradual changes" is the "route by which one state of the language passes into another" (Labov 1972: 114): "intermediate" articulations between the original sound form and the concluding one have been claimed to be improbable. According to Labov's report, instrumental studies on the location of the [|173|] first formant have allowed him to delineate the transitional stages of a Martha's Vineyard dialectal change, namely new centralization of [aw] (1972: 126ff.); cf. observations on the optional weakening of rounding in the pretonic /o/ on its way to a merger with /a/ in the dialects in the northern proximity of Moscow. The distinction oft he two pretonic vowels becomes a feature of lesser importance for those speakers under the influence of neighboring, "prestige" dialects and according to the style of speech is either totally abandoned or at least readily lessened. Transitional implementations are possible but not compulsory in certain sound changes while they are improbable in other instances. The most relevant and generas aspect of gradualness is the interplay of styles, which is sometimes imprecisely labeled 'overlapping' and which furthers the expansion of a sound change from one speech style to another.

Attempts to explain linguistic shifts, and sound changes in particular, by the discontinuity between the parent's and the child's models of language keep cropping up and have been discussed since the threshold of this century. These surmises leave several questions unanswered and unanswerable: Why would the restructuration of the elders' model show a sameness in all the children of the speech community? Why should we assume that the spatial continuity between the coevals of the speech community is stronger than the temporal continuity between parent and child? And finally, if there should exist in the speech community an inner demand for just such a change, and if we take into account the inner dynamics of certain linguistic styles and the collision between the different styles within an individual linguistic framework and within a dialogue of any given interlocutors, would these conditions not suffice for the urge toward innovation and restructuration and for its accomplishment and its diffusion? Briefly, a parent-child linguistic tension may occur as a factor furthering some given changes but can scarcely be viewed as a necessary or adequate reason for change. The chief thing is that any linguistic system, both individual and collective at any moment of its life, necessarily implies the coaction of two forces, stability and mutability, so that the need for change is constantly inherent in any verbal code, with its permanent switching between subcodes. Thus the ongoing sound changes invariably belong to the proper essence of any living system: permanent variability in space and time and in speech praxis is the main universal of language. [|174|]

Featural changes of different kinds - merggers, splits, transfers (cf. Moulton 1967) - display a goal-directed character in regard to the sound system which undergoes them. Such a character, one could add, implies not only goal attainment but also goal failure.

The once-popular concept of "blind changes" has, not only in linguistics but also in all studies of human activities and of living systems in general, yielded to a patent or at least latent recognition of a means-ends relation and to a search for the inner motivation of changes. Besides changes with an equilibratory, prophylactic, and stabilizing role for the language system, those with a reverse, disequilibratory aim attract the ever-closer attention of the explorer; he has to take into account the interaction of verbal styles, in particular those styles which imply a disautomatization of linguistic constituents and a collision with a static order of things and which in this way favor structural changes. The purposefulness of the change merges with that of the system which undergoes the change (cf. Shaumyan 1977).

The question of purpose is sometimes obscured by its submission to the involved discussion of the speaker's awareness, although the inner logic of language is evident independently of the oscillation of verbal activities between the unconscious and consciousness. Moreover, one should not forget that metalinguistic operations, which prove to be one of the cardinal functions of language, ensure a high-level awareness of both the stability and the mutability of the linguistic system. The constituents of the linguistic network and of its ongoing changes are, whether consciously or subliminally, perceived and recognized as functional both by the adult and by the child (cf. Waterson 1971b).

Even during the earliest stages of language acquisition the child evidences the presence of subcodes in his linguistic stock. The difference between the addresser's and the addressee's competence is manifest in children. Various kinds of conformism and nonconformism play a great part in the gradual changes of children's language. Natalie Waterson's son at 1 year 6 months was "capable of producing a clearer version of his form, i.e. a form closer to the adult's, when his first effort was not understood. This suggests that he had some idea of certain features of the adult form that he did not normally use and that were redundant for him at this stage" (1970: 3). As this careful observer of children's speech noticed, they frequently recognize their recent linguistic increments as innovations, but on the other hand they preserve their memory of and feeling for the "archaïsms" of [|175|] their earlier verbal experience and use them on occasion in utterances of a retrospective tinge. A telling example is given by Waterson: at 1 year 8 months her child's form for pudding is [pʊpəŋ], but the old form, [pʊpʊ], reappears "under emotional stress, an urgent request: [bɪʔɪ dzæm pʊpʊ] 'a bit of jam for my pudding'. This indicates that the form which was last used by the child two months yreviously was still in his competence" (1970: 14).

In discussing the diffusion of a sound change within and between speech communities, we can extend this notion also to cases of the convergence of similar but independence changes in different local sources (Saussure's foyers d'innovation), because both processes complement each other: a convergent anticipation of the change-to-occur is a furthering condition for a "successful" diffusion. In an analogous way, the submission of the little tutee to his language tutor is furthered by the former's anticipation of the structural rules to be mastered, an anticipation due to the antecedent stages of his linguistic training and to his inherent abilities and predispositions for the adoption of language design and especially for the featural network. (Jakobson 1988[1979d]: 168-175)

["The network of distinctive features"] [DOI: 10.1515/9783110862744.125]

Jakobson, Roman; Waugh, Linda R. 1987[1979b]. Afterword. In: Rudy, Stephen (ed.), Selected Writings VIII: Major Works 1976-1980. Completion Volume 1. Berlin; New York: Mouton, 235-239. DOI: https://doi.org/10.1515/9783110862744.235

Neither the divergences nor the convergences in the makeup of the distinctive features can be minimized. A rigorously relational approach to the features shows, for instance, that neither the hidebound concentration on convergence nor the biased negation of convergence does justice to such basic phenomena as the mutual affinity and diversity of vowels and consonants. Let us repeat that there is no autonomy without integration and no integration without autonomy.

The two essential focuses of linguistic investigation, one upon invariance and the other upon variation, are sterile separately from each other, and any one-sided exaggeration - one might even say monopolization - of one of the two facets with a disregard for the opposite one distorts the very nature of language. Any system is by definition always mutable, hence the notion of an individual or collective linguistic system, without variation, proves to be a contradiction in terms. The notion of context which furthers variation becomes ever wider and encompasses not only sequential and concurrent neighborhoods in the sound flow but also the diversity of speech styles. And on the other hand, the concept of invariance can no longer be confined to a single language, but the further logical step would be to view the sound patterns of single languages as varying implementations of universal invariants. Spreading efforts to expand the world inventory of distinctive features are enrooted in the unmotivated abandonment of the relational approach with respect to the comparative analysis of different sound patterns.

The growing sense for the interplay of variants and invariance introduces and strengthens the notion of dynamic synchrony and removes the traditional antinomy synchrony/diachrony. It becomes really outdated from a relativistic standpoint to confine one's treatment of linguistic matter to a mere description or a mere history. Also, such notional dualism as competence/performance or innate/acquired [|237|] prove to be indivisible, as for instance may be exemplified by explanatory truisms such as "language cannot be learned without the capacity to learn it" and "language cannot arise without acquisition from the environment"; likewise, "a perceiver's and/or an emitter's performance is the implementation of his compentence" and "competence means the competence of the perceiver and/or emitter to perform". The separation of the concepts, sound change and its diffusion, is an artificial one also, because change in order to be achieved implies a personal repetition and an interpersonal diffusion.

The role of the addresser and the addressee in verbal communication are two inseparable topics of investigation; and inner speech, an important variety of this double theme, means the assignment of the two roles to one and the same person. Both forms of speech (inner and outer) are two closely interconnected processes of communication, and cognition plays a substantial role in both of them. One must take persistently into account the activities of communication and cognition in both their interpersonal and their intrapersonal aspects.

The distinctive features and their concurrent and sequential bundles (phonemes and syllables) differ from all other constituents of language through the lack of their proper, immediate signification. Their only signatum is that of 'mere otherness', or in Sapir's terms, they carre "no singleness of reference." Without having their own meaning they serve to differentiate the meanings of the grammatical units to which they pertain, morphemes and words. Their inner organization is built on the principle of the most effective perception and recollection. And this merely mediate characteristic appears as their only load so long as language is taken in its narrowly rational application.

However, any distinctive feature is built on an opposition which, taken apart from its basic conventional linguistic usage, carries a latent synesthetic association and thus an immediate, semantic nuance. This immediacy in signification of the distinctive features acquires an autonomous role in the more or less onomatopoeic strata of ordinary language. The habitual relation of contiguity between sound and meaning yields to a bond of similarity . This phenomenon goes beyond the limits of onomatopoeias proper and succeeds in creating submorphemic links between words of diverse origin. It is this similarity in sound and meaning which even assumes an active role in reviving and condemning lexical archaisms and in furthering viable neologisms. [|238|]

The significance of the play on words (jeu de mots) in the life of language should not be underestimated: the vocalic association of the designations for day and night, with the light ~ dark contrast for a Slav, and with slower ~ quicker for a Frenchman, brings about, as Lévi-Strauss put it, une petite mythologie.

The tension between two structural principles - contiguity and similarity - permeates the whole of language. If, as mediate building blocks of meaningful entities, the distinctive features serve to connect sound and meaning by virtue solely of contiguity, the inner sound symbolism peculiar to these features strives to burst forth and to sustain an immediate similarity relation, a kind of equivalence between the signans and the signatum. Besides the conventional thései relations, such a direct semantization of the sound shape comes into play.

And it is precisely 'play' and the mythopoeic transforms of language which help to dynamize the autonomous semantic potential of the distinctive features and their complexes. Poetry, as a purposeful, mythopoeic play, is the fullest, universal accomplishment of the synthesis between contiguity and similarity.

The analysis of the two closely interconnected synthetic powers of poetry - that of similarity and contiguity and that of selection and combination - is a burning task faced by our science. Any fear of or reluctance about the analysis of the poetic transformations of language impairs the scientific program of those linguists who pull back from the pivotal problem of this vital transformation; and likewise it curtails the research of those literary scholars who, in treating poetry, pull back from the innermost problems of language.

Among the varied and intricate questions proffered by the sound shape of language, that of the spell of the speech sounds appears to acquire a particular attention. In her recent survey, Eli Fischer-Jørgensen rightly states that not only the "potential symbolic values" of the distinctive features but also their universal character and has been widely documented. Certain instances, and in particular her comparison of West African linguistic data with her own experiments on Danish subjects, "show clearly that these values are not dependent on specific languages and cultures" (1978). The perceptual universality of these values shows certain limitations and an unevenness of distribution, due to the differences in the repertory of features in given languages. Therefore, the author wisely supports "the hypothesis of almost universal values". The frequent predilection of linguistic inquirers for 'absolute', exceptionless universals, as preferred to 'near' universals, clashes once [|289|] more with Sapir's warning against the "dogged acceptance of absolutes". Probability near to certainty but still less than 1.0 is as signal a phenomenon as probability 1.0.

The inquiry into the acquisition of language, into the changes of language, and into the aphasic losses - or in the words the inquiry into the structuration, restructuration, and destructuration of language - must deal primarily with lawful tendencies without any superstitious fear of exceptions and without any prejudice which would assume that there are only two possibilities: either an absolute rule or an absolutely blind chance. The Saussurian vision of linguistic dust (poussière linguistique) far from disintegrating linguistics, widens its vistas int he search for general laws. And just as Saussure's notion of linguistic dust does not undermine his idea of system and its general laws, in the same way, the "linguistic antinomies" discovered throughout centuries do not authorize us to undermine the unity (or more exactly, the 'bi-unity') of approach to language and to its sound shape in all its real multiformity, nor to close our eyes to the contradictions to be removed. (Jakobson & Waugh 1987[1979dx]: 236-239)

["Afterword" to The Sound Shape of Language] [DOI: 10.1515/9783110862744.235]

Jakobson, Roman 1985g. The immediate quests and accomplishments of comparative linguistics. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 314-317. DOI: https://doi.org/10.1515/9783110855463.314

For the first time the indissolubility of the notions of a law-governed system and its changes, in turn of a regular nature, is realized. The boundaries of linguistic comparison widen in an essential way; new tasks are added to the study of the common legacy of language families. Forst, the common features acquired by the phonological and grammatical structure of languages contiguous in space are subjected to clarification, and thus enter, as it were, into relations of alliance. Second, the possibility and even necessity of juxtaposing diverse language (above all, phonological) systems, irrespective of the question of the presence of absence of genetic relationship or geographical propinquity, becomes an established fact.

As a result of the comparative analysis of such systems there arises in turn the possibility of their consistent typological classification and its theoretical grounding. With such prerequisites the data of living languages, supported by the documentation of historical languages, make it possible to verify the probability of the proto-language systems reconstructed by the comparative method and convincingly prompt the most plausible solution of difficult problems in reconstruction. In a word, typological comparison proves to be a benificial tool in comparative-historical procedures.

All these principles of general linguistics, newly discovered or at least conceived anew, now present every concrete investigative work with inescapable and pressing demands. The collective work of Tomaz Valer'janovič Gamkrelidze and Vjačeslav Vsevolodovič Ivanov, Indoevropejskij jazyk i indoevropejcy ("Indo-European and the Indo-Europeans"), completely fulfills the mission formulated by the authors in the subtitle: "A Reconstruction and Historical-Typological Analysis of the Proto-Language and Proto-Culture". Those deep shifts and transformations which characterize the stage now attained in the development of linguistic science, and in which a considerable creative role fell to the lot of both mentioned investigators, lay at the methodological base of their searches. Here the early approaches of international seekers to all the particular questions of Proto-Indo-European linguistic antiquity are taken into account, and a fascinating reply is given to the theses which gained scientific currency on the border of the two centuries. This stage in the scientific work of Ivanov and Gamkrelidze is marked not only by the unusual anwers they offer to the given questions, but also by their [|316|]very formulation of such questions and by their unprecedentedly wide thematic horizon.

In correspondence with the dialectical reduction of the dichotomy of the temporal progression and the cross-section (diachrony and synchrony) and with the parallel inclusion of temporal diffusion into the number of internal linguistic factors, the book naturally transforms the scheme of Common Indo-European, viewed by the scholarly tradition as statis and uniform in time and space, and creates a model of dynamic synchrony with an integral grasp of the foundations of the proto-language, its evolutionary shifts, its internal, regional differentiation, and its sequential crossing with neighboring language areas. Precisely in light of questions about the mutual relations of dialects of the Indo-European proto-language and about its interrelations with adjacent proto-languages arises the promising work of the two linguists on the geographical determination of the primordial Indo-European homeland (evidently Asia Minor) and on the proposed pats of the initial migration undergone by the different branches of the common Indo-European territory.

The broad interpretation of two concepts - comparison and system - is contemporary linguistics is attended by the steadfastly progressing relativization of the entire linguistic structure and ever more consistent transformation of linguistics into a science of intralinguistic relations. Moreover, the attention of linguists, especially Ivanov and Gamkrelidze, is held in the first place by the indissoluble interconnection between the parts and the whole. The main key point of this complex problematics is the interrelation of the invariant and variation, that vital theme of all contemporary scientific thought. The dependence of variation on the diversity of contexts becomes ever more clear in harmony with the development of the thesis of linguistics of our day which opposes natural language, adjusting itself to the context (context-sensitive), to languages irrespective of the changing context (context-free), i.e., artificial, formalized systems. Here, of course, variations of form and meaning play an essential role: both on the level of sound, as well as the various levels of grammatical meanings, the systematic extraction of invariants becomes the central linguistic task.

This entire methodological program is widely developed on the example of the reconstruction of Proto-Indo-European. With the decomposition of the phoneme into minimal sense-discriminative components, the context of context, which earlier had been limited to he temporal sequence of combinations of phonemes, has been expanded to include combinations of simultaneous components, and the twofold [|317|] dimension of phonological combinations reveals step by step new, uninvestigated typological regularities in the interrelation between both classes of combinations and within each class. See, for example, the remarkable typological works of Ivanov and Gamkrelidze on preferred or, on the contrary, avoided combinations of differential components "on the axis of simultaneity" and on the principal varieties of symmetrical relations, which, as V. V. Ivanov has shown, form the basis of linguistic structure.

With the development of the problematics of context, the simplistic interpretation of stylistic variants as "free" variations gives way to an understanding of style as a distinctive context, and the conditions offered to language by the different speech functions enter clearl into the circle of the general conceptions of contextuality. We are indepted to the correct initiative of Ivanov and Gamkrelidze in including Indo-European poetics, in particular metrics and the question of the anarammatic tradition, first raised by Saussure, with the number of pressing tasks of linguistic reconstruction. (Jakobson 1985g: 315-317)

["The immediate quests and accomplishments of comparative linguistics"] [DOI: 10.1515/9783110855463.314] ["Written in February 1979 as a preface to Indroevropejskij jazyk i indoevropejcy by V. V. Ivanov and T. V. Gamkrelidze."]

Jakobson, Roman 1985[1980f]. Einstein and the science of language. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 254-264. DOI: https://doi.org/10.1515/9783110855463.254

Perhaps the most telling concordances between the innovation in physics and that in contemporary linguistics are those coincidences that seem to be due to purely convergent, independent development. Such latent correspondences reveal a substantially parallel course in these different sciences. Both Einstein's demand of the theoretical physicist to strive for the highest possible standard of rigorous precision in the description of pure relations, as in his address at Max Planck's sixtieth birthday (1918), quote above, and the close counterpart of this demand - namely, the even stricter inquiry into the physical world as a network of interrelated components - stand in eloquent correspondence with the tasks of advanced linguistics. Careful comparisons between the fundamental concepts of relativistic physics and the constituents of language as analyzed and defined by contemporary linguists disclose a salient isomorphism that could be easily exemplified on the different levels of verbal structure.

A few widespread phonological cases may suffice to reveal the generality of the problem. The distinctive feature that fulfill the main task of speech sounds are, as Einstein would term them, rigorously relational ideas, intuited as binary oppositions. Thus, for instance, in those consonantal systems that make sense-discriminative use of the so-called "flatness" feature, flat consonants are phenomenologically equivalent; in our perception they are mainly distinguished by a particular [|263|] lowering of their inferior formant. In diverse languages we observe certain differences in the sensorimotor modalities of this process. For example, a fairly similar auditory effect is obtained by labialization and by pharyngealization, or, in other terms, by the narrowing of the frontal or of the buccal ends of the mouth cavity. But since the difference between these two special cases is never used for sense-discriminative purposes, the common denominator outweighs the difference (as well as some other, likewise superficial modalities). The typology of languages asserts the structural invariance of the feature in question, and the universal laws of language prove to admit no more than one single opposition of present and absent flatness. In linguistics the principle of equivalence (instead of mechanical sameness) puts limits on the significance one can expect from a search for separate, uncoordinated particulars of experience and yields instead the gradual discovery of law-governed paucity in the fundamental relationships that underlie the verbal (as well as the physical) universe.

"Nur fiel mir ein" (Now it came to me) - that is the item from Einstein's "Autobiographical Notes" that sounds like a first glimpse of the general theory of relativity and like a joint slogan of the contemporary sciences, all tending to transform an abundance of raw stuff into a parsimony of general laws. The problemss of equivalence proved to be as pertinent for the principle of relativity as for the discovery of linguistic universals. A substantial revision of the time-space model, notwithstanding the differences in the statement of these questions with respect to diverse sciences, leads us far from the previous mechanical rutine. Among such new linguistic vistas calling for a vital interdisciplinary discussion, one may bring to the foreground the notion of dynamic synchrony, the reversible course of current events, and the conception of any change in progress as an intrinsic simultaneity of sensible oscillations. Niels Bohr repeatedly insisted on the deep links that at present tie together physics and linguistics, to whose interrelation both of us devoted a joint MIT seminar at the end of the 1950's. The "exigencies of relativistic invariance", in Bohr's favored term, were intently discussed with respect to the search for and structure of the ultimate constituents of both physical and the linguistic universe, the "elementary quanta", as they were termed in physics and were picked up from physics by linguists. The endeavor of our linguistic generation to [|264|]conceive the verbal mass as a "discontinuous" matter, which is composed of elementary quanta and hence reveals a "granular" structure, partly continues an older set of efforts. At the same time, however, it exhibits an evident dependence on the development of the exact sciences, which was, I testify, a genuine source of inspiration for the linguistic avant-garde of the first third of our century, both in the West and the Eastern scientific centers.

Let us finally mention that two polar and inseparable problems - namely, symmetry (with its various transformations) and asymmetry, on the one hand, and the breaking of symmetry, on the other - permeate the diverse sciences. In his Thematic Origins of Scientific Thought Gerald Holton pointed out the initiative, vital role that symmetry arguments have acquired in Einsteinian physics. Substantially analogous concepts find an ever wider application in the analysis of any linguistic structure. Yet the entire symmetry-asymmetry complex in linguistic research, both in its ontological commitment and in the role of a pure formal device, must be seen to belong more to the victories of tomorrow rather than to the solutions of yesterday and today. However, we may perhaps console ourselves on that point with a thought that Einstein wrote down only four weeks before his death: "For us [...] the distinction between past, present, and future is only an illusion, albeit a stubborn one."

An epochal scientific theory may regenerate in contemporaneous poetry into an elementary myth. Thus, for instance, Vladimir Majakovskij, the Russian avant-garde poet, from his first, anxious glimpses of 1920 into the theory of relativity and until the eve of his suicide in 1930, praised "the futurist brain of Einstein" and devoted his ultimate drama of 1929, Banja (Bathhouse), to the crushing victory of such an unusual brain over the alleged absolute of time. (Jakobson 1985[1980f]: 262-264)

["Einstein and the science of language"] [DOI: 10.1515/9783110855463.254] ["Presented at the Einstein Centennial Symposium in Jerusalem, March 16, 1979, and published in Albert Einstein: Historical and Cultural Perspectives, ed. G. Holton and Y. Elkana (Princeton, 1982)."]

Jakobson, Roman 1985[1981m]. My favorite topics. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 371-376. DOI: https://doi.org/10.1515/9783110855463.371

Returning to the inferences I have made on the basis of my phonological and grammatical research, I should like merely to list several further areas that belong among my favorite themes of investigation.

Time and space, usually regarded as extrinsic factors in relation to the verbal code, prove to be veritable constituents of the latter. In the speakers' and listeners' code any change in progress is simultaneously present in its initial and final forms as stilistic variants, one more archaic and the other more advanced, both being mutually interchangeable in the speech community and even in the use of its individual members [...]. Since my earliest report of 1927 to the newborn Prague Linguistic Circle I have pleaded for the removal of the alleged antinomy synchrony/diachrony and have propounded instead the idea of permanent dynamic synchrony, at the same time underscoring the presence of static invariants in the diachronic cut of language [...].

The verbal code is convertible also with respect to the factor of space. It contains a set of variants serving for different degrees of adaptation to interlocutors of diverse dialectal and social distance. Diffusion of linguistic characters results from such variations, and during the 1930's I devoted several essays [...] to one of the extreme manifestations of the space factor in the life of languages, in interlingual rapproachment termed Sprachbund by N. Trubetzkoy; and later I repeatedly, though so far in vain, appealed for phonological atlases on vast territories, a task with undoubtedly surprising vistas. (Jakobson 1985[1981m]: 374)

["My favorite topics"] [DOI: 10.1515/9783110855463.371] ["Written in Peacham, Vermont, August 1980, and first published in Italian in Premi "Antonia Feltrinelli" 1980 (Rome: Accademia Nazionale dei Lincei) on the occasion of the author having been awarded the Premio INternazionale per la Filologia e Linguistica."]

Jakobson, Roman 1985d. Retrospect. In: Rudy, Stephen (ed.), Selected Writings Volume VI: Early Slavic Paths and Crossroads. Part 2, Medieval Slavic Studies. Berlin; New York: Mouton Publishers, 889-897. DOI: https://doi.org/10.1515/9783110863901.889

There is, no doubt, a particularly deep divergence between Hermann Paul (1846-1921), the main theoretician of the Neogrammarian school as well as an outstanding researcher, on the one hand, and Saussure's Cours de linguitique générale on the other. While for Paul the science of language was confined to the historical study of its alterations over the course of time, the Genevan Cours expressed the view that the scientific approach to the system of language must strictly follow "the horizontal axis", i.e., concentrate solely on a given state of language, with no account of any linguistic changes. The changes are, in turn, the only topic in diachronic linguistics, while system as such is beyond the competence of this discipline, since allegedly changes go on without any regard to the system.

Notwithstanding the discrepancy between the Neogrammarians' absorption in the history of changes and Saussure's predilection for a longitudinal section of linguistic systems, there still remains one striking feature which unifies both these standpoints and opposes them to the impetuously spreading conviction that "synchrony" does not refer merely to the statis aspect of our science, but rather that any change in progress necessarily pertins to the synchronic aspect of language.

There do not and cannot exist linguistic changes decreed and utterly accomplished at once, with a wave of the wand. The starting and finishing forms of a change which is in progress both occur at the same time within the same speech community; they may specify the community's different generations and be used intermittently in the verbal exchange of the elder and younger members of the community. At this stage of a change, earlier and later forms may be employed alternately by the same individual as specimens of differing verbal styles: the more archaic of the two may be felt as more prestigious, the other as fashionable and somewhat frivolous. The position of the two competitive forms, one more conservative and the other more altered, may be realized with a greater or lesser level of consciousness by the speaker and [|890|] listener, thus revealing their awareness of the vertical axis at a given synchronic stage of the language. Any current change is experienced in its entire course simultaneously by all members of the speech community: it represents a part of the linguistic system, and as such cannot be discussed out of the context of the system in question. Moreover, there are no immutable systems, and linguistic synchrony demands the constant attention of the analyst to changes in progress.

The "absolute prohibition" in Saussure's Cours against dealing "simultaneously with relations in time and relations in the system" is at present being forced to yield to the ascertainment of the mutual indissolubility of these two seemingly separate topics. A dynamic synchrony cancels the artificial break between system and mutation.

Any verbal code at any given moment embraces a number of temporal subcodes. Both time and space function as inner factors of any language system, notwithstanding the efforts of a few linguists to justify the view of language as irrespective of either time or space.

Time in language displays its two axes: simultaneity as well as successivity. This was illustrated on the grammatical level by Charles Bally's reference to the concurrence of grammatical meanings ("cumul des significations"), and on the phonological level by the decomposition of phonemes into concurrent distinctive features, thereby rejecting Saussure's "linéarité du signifiant". On the other hand, language in time displays its two measurements: the concurrence of all constituents, whether stable (static) or mutable (dynamic), within any synchronic stage of language, and the successivity of stable and mutable stages in linguistic diachrony.

The manifold character of any verbal code, namely its divisibility into temporal subcodes, is paralleled by the variety of coexisting subcodes based on spatial proximity: we have different forms of address for people depending on their closeness to or distance from us. The degree of our verbal adaptation to the spatial location of our interlocutors is also distinctly expressed here. The interpenetration of dialects or even of neighboring languages contributes significantly to the inquiry into the problems of multispaced codes and their composition.

The distribution of social functions assigned to the temporal and spatial subcodes is one of the most powerful factors in the history of nations and cultures, one might even say in the destiny of our multifaceted world. As Aleksandr Blok, the astute Russian poet of the beginning of our century, wrote in 1919 in the preface to his historiosophic poem Vengeance: "I'm accustomed to confronting facts from all spheres of life which are at present accessible to my sight, and I'm [|891|] convinced that taken together they always create a single musical sense." The mission of language to impart a new, inspiring sense to diverse, seemingly unrelated phenomena has always lured me. (Jakobson 1985d: 889-891)

["Retrospect" to SW VI, Part 2] [DOI: 10.1515/9783110863901.889]

Jakobson, Roman 1985[1984e]. On the Dialectics of Language. In: Rudy, Stephen (ed.), Selected Writings VII: Contributions to Comparative Mythology. Studies in Linguistics and Philology, 1972-1982. Preface by Linda R. Waugh. Berlin; New York; Amsterdam: Mouton, 377-378. DOI: https://doi.org/10.1515/9783110855463.377

The hic et nunc of linguistic reality brings every human being face to face with a multitude of spatial fields and temporal moments, and any production and perception of language reaches its steady manifestation through the selection and combination of suitable entities from within this double multitude. Our selective and combinatory verbal activities are generally restrained and directed by a system of acting rules. It has frequently occurred to linguistic interpreters that the use of language was being conceived without respect to these rules. Speech production and perception in their temporal changes remained the only focus of scholarly observation. The opposite trend wsa the view that rules confine the production and perception of language at any given stage, and these rules, promoted as the chief subject of linguistic study, wer termed langue versus parole, or "code" versus "message", or "competence" versus "performance". A singleness was attributed to the rules of competence, and this was resolutely superposed on the plurality of performances. I have objected to this strict mechanistic rupture between invariant and variants: no speaker appears to be limited to one single code. In essence he holds to the same language with the closest and most distant members of his environment, yet constantly modifies his manifold code and thus adapts his competence to diverse interlocutors, [|378|] different topics, and his ceaselessly varying verbal styles. There is, as in any system, an incessant linkage of variants and invariants, a permanent unity and diversity of phonological, morphological, syntactic, lexical, and variational means. The universal phenomenon of dynamic synchrony points to a constant interchange of the code.

Both in various self-adaptations to the interlocutors (verbal conformism) and in different degrees of mutual repulsions (verbal nonconformisms), we submit our code to a maximal variability, an inconstancy both in space and in time. Such has been my recognition of the inseparability between invariance and variability. This thesis appeals to me as the conditio sine qua non of scientific analysis from the early steps of Hegel's dialectics to the present-day sciences, especially linguistics, and our indebtedness to the Master's inspirations is far from exhausted. In particular, time and space are two mutually inseparable, inner factors of language. Every verbal activity implies incessant selections and decisions between locomotor opportunities which suggest themselves, regardless of whether it concerns an intimately merged idiom or a distant coincidence, as well as which stage of the mutation in progress - an imminent archaism or the final stage of innovation. (Jakobson 1985[1984a]: 377-378)

["On the dialectics of language"] [DOI: 10.1515/9783110855463.377] ["Written in June, 1982, on the occasion of the author's having been awarded the Hegel Prize of the city of Stuttgart."]

0 comments:

Post a Comment