Roman Jakobson's Conception of «Sprachbund»

Helmut SCHALLER, Univ. de Marburg : «La conception du Sprachbund chez R. Jakobson» 
Pour la première fois en 1929 Troubetzkoy, lors du Premier Congrès international des linguistes, développe la conception du Sprachbund, dans laquelle il prend en considération non seulement le point de vue phonique, mais aussi le point de vue morphologique, syntaxique et lexical de proximité de langues relevant de familles différentes ; avant tout, il prend pour objet de réflexion la proximité de langues balkaniques slaves et non slaves. 
Pour Jakobson, cependant, c'est le point de vue phonologique qui est au cœur du concept de Sprachbund, ce qui a eu pour conséquence qu'il a pu, sur la base de corrélations phonologiques, rassembler en Sprachbund de vastes zones de langues voisines. C'est ainsi qu'il a développé le concept de Sprachbund eurasien, avec pour marque commune la corrélation de mouillure. 
Ce à quoi on a affaire entre Troubetzkoy et Jakobson, ce sont des particularités dans les différences, mais certainement pas des contradictions quant à la conception du Sprachbund.

To illustrate the vagueness of the notion «Sprachbund» since Tru betzkoy and Jakobson, 1 should like to make a survey of its usage and then attempt to come to sorne definition with special reference to the «Bal kansprachbund». The notion «Sprachbund» was first mooted by N. Tru betzkoy first of ail known as the fo under of the phonological method in 1923 in «Vavilonskaja baSnja i smesenie jazykov», then at the First Inter national Congress of Linguists in The Hague in 1928, in order to add to language families and groups another term, which takes into account the linguistic peculiarities which have arisen from mutual influences between languages. Trubetzkoy writes : Viele Missverstii ndnisse und Pehler entstehen dadurch, dass die Sprachforscher die Ausdrücke Sprachgruppe und Sprachfamilie ohne genügende Vorsicht und in zu wenig bestimmter Bedeutung gebrauchen.
Trubetzkoy therefore made the fo llowing suggestions : Unter den Sprachgruppen sind zwei Ty pen zu unterscheiden: Gruppen, beste hend aus Sprachen, die eine groBe Ahnlichkeit in syntaktischer Hinsicht, eine ahnlichkeit in den Grundsatzen des morphologischen. Baus aufweisen, und eine groBe Zah) gemeinsamer Kulturworter bieten, manchmal auch auBere Ahnlich keit im Bestande der Lautsysteme, -dabei auch auBere Ahnlichkeit im Be stande der Lautsysteme, -dabei aber keine systematischen Lautentspre chungen, keine Übereinstimmungen in der lautlichen Gestalt der morphologi schen Elemente und keine gemeinsamen Elementarworter besitzen, -so1che Sprachgruppen nennen wir Sprachbünde. Gruppe, bestehend aus Sprachen, die eine betrachtliche Anzahl von Elementarwortem besitzen, Übereinstimmungen im lautlichen Ausdruck morphologischer Kategorien aufweisen, und vor allem konstante Lautentsprechungen bieten, -solche Sprachgruppen nennen wir Sprachfamilien. (Trubetzkoy, 1928, p. 17-18)  thuanian, Latvian, Estonian and North Kashubian and finally also sorne of the North German dialects. The German Baltist Viktor Falkenhahn (1963) had attempted to establish a Lithuanian-Polish Sprachbund on the grounds of the similarities between two languages in the verbal rection. The two languages belong to different families, yet there is only one pattern which they share. If one compares the various Sprachbünde within and outside Europe' contrastively, above a11 the question arises which the common characteris tics of the constituent languages of a Sprachbund are. Thus there are postu lated Sprachbünde which have only one linguistic characteristic, e. g. the polytony in the Baltic Sprachbund as a phonological conformity or as a syntactical conformity the verbal rection in the Polish-Lithuanian Sprach bund. In contrast to these, there are quite a nurober of linguistic similarities in the Balkansprachbund. But strictly speaking, aIl the Sprachbünde mentio ned above share only one characteristic, namely that they consist of lan guages of various families, as Trubetzkoy laid down as early as 1923 and 1928. The question which arises again and again as to how many similari ties are required to constitute a Sprachbund has led to subdivision of two kinds of Sprachbünde, namely the intensive and extensive. As an example of the extensive Sprachbund, we may take the Baltic Sprachbund, with its only one characteristic of polytony, whereas for the intensive Sprachbund the Balkansprachbund may serve as the prime example, which stands out on account of its various correspondences, in the phonetical field as weIl in other linguistic fields. Unlike «language» which is a fixed concept, the word Sprachbund can be replaced by more or less synonymous terms like «Sprachverband» or «Sprachenbund», at least with regard to German ter minology. If we try to find the genus proximum, 1 would suggest the notion of the family of languages in which the languages are doser to each other in their genetic similarities than in a Sprachbund the similarities of which are of a typological order : differentia specifica. The characteristics of a Sprachbund, as mentioned above, arise from mutual influences. Therefore a definition of the term Sprachbund could be made as follows : ln contrast to the genetically defined family of languages (genus proximum), the Sprachbund comprises a typologically defined group of geographically neighbouring language whose common features are derived from mutual influences (differentia specifica).
Neither an extensive nor an intensive Sprachbund can consist of two languages, with the exception of the Lithuanian-Polish Sprachbund. It is questionable whether the extensive Sprachbund with only one common feature is in Hne with the definition of a Sprachbund. Therefore one rnight arrive at the following extended definition : .
In contrast to genetically definedfamilies of languages, the Sp rach b' und comprises a typ ologically defined group of at least three geographi cally neighbouring languages, whose common fe atures are derived from mutual influences.
Not only Trubetzk.oy's, but also Jakobson's influence on the deve lopment of linguistics has been a very great one. The latter was one of the founders and movers of the Prague Linguistic Cercle. On the basis of the new structuralist concepts, he set fo rth theories like that of an extensive Sprachbund and illustrated it with demonstrations based on Slavic and 0ther languages. So he examined prosodie problems of languages as diverse as Ancient Greek, Norwegian and Chinese. AIso, Siavic accentological evi dence plays a small or secondary role in his works on phonological «con vergence 'areas», Sprachbünde, particularly the «Eurasian linguistic allian ce», said to be characterized by the combination of accentuai monotony and distinctive palatalization in consonants.
Yet, although the phenomenon is fa miliar, the term «Sprachbund» introduced by Trubetzkoy and Jakobson, is admittedly unsatisfactory . Its fu ndamental fauit seems to be that it implies a unit, as if a language either were or were not a member of a given Sprachbund. U. Weinreich (1948, p. 378) proposes that it would be preferable to abandon these terms and speak simply of cases of convergent development and, if necessary, of conver gence areas. He would then say that in the Caribbean area, as for example in the Balkans, a number of Indo-European languages have undergone in tensive convergent developments 3 .
So we can sum up, in the sense of N. Trubetzkoy, that a lot of mis leadings and mistakes were originated by the fact that linguists used the no tions «language group» and «language fam ily» without sufficient examina tion and in not sufficiently defined meaning. Within language groups we have to see two different groups, consisting of languages which show a great similarity in syntax, similarity in the principles of morphologie al structure, and also a great number of common cultural words, sometimes an external sim ilari t y in the stock of their phonetic systems, but no systematic phonetic correspondences, no identity in the phonetic shape of morpholo gical elements and no common words.
These groups of languages are named 'Sprachbund', but groups consisting of languages which show a great number of common words, identity of morpholo gical categories, and last not least fixed ph onetic correspondences, -these groups of languages are named language fa milies.
So we have two categories of Sprachbund : the intensive one, consti tuted by N. Trubetzkoy, the extensive one, constituted by Roman Jakobson, based on phonological marks in contrast to phonological, morphological, syntactic and even lexical marks of Balkansprachbund. Both concepts of Sprachbund, the intensive and the extensive one are discussed up to today and so we remember in 1996 the great ideas of Roman Jakobson and the Prague School of linguistics.