OUTLINE OF THE COMPARATIVE GRAMMAR OF THE SLAVIC LANGUAGES

S. B. Bernstein, An Outline of the Comparative Grammar of the Slavic Languages. USSR Academy of Sciences Publishing House, Moscow, 1961, 352 pp.

S. Bernstein’s “Outline of the Comparative Grammar of the Slavic Languages” is part of a comparative grammar of the Slavic languages, covering introductory questions and a chapter on phonetics. It represents the very latest word in linguistic scholarship on the comparative-historical phonetics of the Slavic languages and general problems of Slavic studies. In our view, this work far exceeds its modest aim as a “textbook for students” and takes its place alongside the classic works on the historical linguistics of the Slavic languages by A. Meillet, N. van Wijk, A. Selishchev, and A. Vaillant. Both in methodological terms and through its extensive material concerning Baltic studies, and, finally, its treatment of the problem of direct Baltic-Slavic linguistic relations, S. Bernstein’s work is important for researchers into the history of the Baltic languages. It is gratifying to note that this work devotes considerably more space to questions of Baltic studies than earlier works of this kind.

The chapter of S. Bernstein’s book entitled “Linguistic Relations between the Baltic and Slavic Languages” (pp. 27–37) will probably be of greatest interest to our readers. Unlike earlier works, for example those by V. Vondrak and J. Mikola, whose authors did not specifically investigate Baltic-Slavic linguistic relations but were content to state the existing situation, the author of the new work not only sets out the earlier theories but also presents clearly formulated views of his own. S. Bernstein unhesitatingly argues against the theory of a Baltic-Slavic proto-language, which, in his words, has done considerable harm to the comparative grammar of the Slavic languages: this theory encourages researchers to mix phenomena from different periods and devise artificial constructions and hypotheses that hinder the study of the history of the Proto-Slavic language (cf. p. 36). This is why, in S. Bernstein’s view, the theory of a Baltic-Slavic proto-language cannot be accepted:

a) Many of the oldest processes in the Baltic and Slavic languages developed differently. The history of the vowels ā, ō, ā̄, ō̄ is particularly characteristic in this respect.

b) Many of the most important processes in the Proto-Slavic language, which took place in the earliest period of its existence, were entirely absent in Proto-Baltic or were only weakly reflected in individual dialects. This is especially true of the fate of s in the Baltic and Slavic languages.

c) In the past, the entire system of evidence offered by proponents of a Baltic-Slavic proto-language theory was based on comparing facts from Proto-Slavic and Lithuanian. Very often, facts from Lithuanian were presented as Proto-Baltic. Yet Proto-Baltic

313

had split in remote antiquity into two principal proto-languages, one of which comprises Lithuanian and Latvian, while the other comprises Prussian. Comparison primarily with Lithuanian is also problematic because Proto-Slavic tribes had most contact with the Prussians and Yotvingians. There is no doubt that Baltic-Slavic community (сообщность) initially encompassed the Proto-Slavic, Prussian, and Yotvingian languages. The history of the Proto-Slavic language shows that, during the period of close contact between the Baltic and Slavic tribes, there was in fact no longer a unified Proto-Baltic language. It is therefore no coincidence that the most important innovations of this period link Proto-Slavic differently with individual Baltic languages.

d) Proponents of the Baltic-Slavic proto-language theory, despite various attempts, cannot solve the problem of Baltic-Slavic reconstructions. The problem is further complicated by the fact that reconstructing Proto-Baltic is largely impossible: many Baltic languages disappeared without a trace, while those that survived do not make it possible to reconstruct many Proto-Baltic phenomena.

e) The data of archaeology do not support the theory of a Baltic-Slavic proto-language.

S. Bernstein then concludes that the similarity between the Baltic and Slavic languages is explained by the prolonged contact between Baltic and Slavic tribes, which gave rise to the formation of a Baltic-Slavic community. This community is corroborated by shared isoglosses. The task facing researchers is to determine the nature, geographical extent, and date of formation of these Baltic-Slavic isoglosses (cf. p. 36). While rejecting the theory of a Baltic-Slavic proto-language, S. Bernstein also emphasizes that data from the Baltic languages are exceptionally important for studying the history of Proto-Slavic.

In examining individual questions of Slavic studies, the author of the work offers a number of interesting ideas that in one way or another concern matters in the history of the Baltic languages. The periodization of the history of Proto-Slavic is certainly of interest to us. S. Bernstein divides the history of Proto-Slavic into two major epochs: the epoch before the loss of closed syllables and the epoch of the loss of closed syllables. He further divides the first epoch into three periods: the period immediately following the breakup of the Indo-European proto-language, the period of the Baltic-Slavic community, and the period from the dissolution of this community to the epoch of the loss of closed syllables. The second epoch is divided into two periods: the first is the period of the loss of closed syllables, and the second is the period of the breakup of Proto-Slavic and the formation of the languages of the Slavic peoples. S. Bernstein dates the latter period to the second half of the first millennium CE, noting that it is impossible to establish a more precise date, even within a span of two centuries, as most earlier researchers had done (cf. pp. 51–52). In S. Bernstein’s view, contacts between the Baltic and Slavic tribes began in the second half of the second millennium BCE and may have continued as late as the middle of the first millennium BCE. Thus, by the middle of the first millennium BCE, the Baltic-Slavic community had dissolved (cf. pp. 60–61).

S. Bernstein agrees that these claims are still highly hypothetical. However, there is material that makes it possible to refine them. First and foremost, these are the elements of Baltic and Slavic vocabulary that formed within the common Baltic-Slavic gen-

314

during their existence. The most important point here is to distinguish the common Indo-European lexical stock from innovations shared by the Balts and Slavs.

The author of the work is justified in emphasizing that earlier researchers of Baltic and Slavic linguistic relations paid too little attention to the South Slavic languages. It turns out that analogues for Baltic linguistic phenomena are often found in the South Slavic languages, or only in Bulgarian. Here are a few examples.

The tree and its fruit are called gruša in the East and West Slavic languages (Russian груша, Ukrainian груша, Polish grusza, Polabian grūsoi, Czech hruška). However, a variant of this name with k- is already found in Polish dialects, as well as in Kashubian and the Sorbian languages (Polish krusza, Kashubian kreša, Lower Sorbian krušva). The variant with k- is also used in Bulgarian and Serbo-Croatian (Bulgarian круша, Serbo-Croatian крушка). Thus, this is a phonetic isogloss common to the South Slavic, some West Slavic, and Baltic languages (cf. Lithuanian kriaušė, Old Prussian crausy, crausios).

Shared word-formation variants also occur. Lithuanian gerklė, gurklys and Old Prussian gurcle correspond to Bulgarian гръклян and Serbo-Croatian гркљан, which are unique among the Slavic languages. Lithuanian āvinas, Latvian auns, and Old Prussian awins form a shared isogloss with Old Church Slavonic овьнь, Bulgarian овен, Serbo-Croatian ован, and Slovenian óvan. Lithuanian milžti, melžiu corresponds to Old Church Slavonic млъзѫ, Bulgarian мълзя, Serbo-Croatian мўзем, and Slovenian móžem.

Lexical variants are particularly numerous: Lithuanian burnà — Bulgarian бърна ‘lip’, Serbo-Croatian брњица ‘muzzle’; Lithuanian žiaunos — Bulgarian джуна ‘lip’; Lithuanian tárpas — Bulgarian трап ‘pit’, Serbo-Croatian тра̀п; Lithuanian draskýti — Bulgarian драскам; Lithuanian dẽšinas — Old Church Slavonic деснь, Bulgarian десен, Serbo-Croatian дѐсан; Lithuanian šalnà, Latvian salna — Bulgarian слана ‘hoarfrost’, Serbo-Croatian сла́на, Slovenian slána.

Some semantic variants are also found. For example, Lithuanian drebulė (: drebeti) corresponds to Bulgarian треперлика, Serbo-Croatian трепељљка, and Slovenian trepetljika.

On the basis of these facts, it is concluded that in the distant past close contacts existed for a long time between the Balts and the languages of the present-day South Slavs, especially Bulgarian. Since the territory inhabited by the Baltic tribes did not extend farther south than the Pripyat, it may be argued that tribes speaking South Slavic languages also lived in the north of the Proto-Slavic territory (cf. p. 75).

It should be noted, however, that the recent studies by V. Toporov and O. Trubachev on the hydronymy of the upper Dnieper¹ show that the traditional Baltic-Slavic boundary along the Pripyat River probably needs to be revised, since traces of Baltic hydronymy can also be found south of the Pripyat. S. Bernstein’s statement should therefore be qualified somewhat. Finally, elsewhere (cf. p. 60), when discussing the Baltic-Slavic boundaries, he himself notes that small Baltic groups may have penetrated south of the Pripyat.

¹ Cf. B. H. Топоров, О. Н. Трубачев, Балтийская гидронимия Верхнего Поднепровья, „Lietuvių kalbotyros klausimai, t. 4, Vilnius, 1961, p. 198; Лингвистический анализ гидронимов Верхнего Поднепровья, Москва, 1962, p. 232—233.

315

Speaking about lexical borrowings, S. Bernšteinas notes that the oldest borrowings to enter Proto-Slavic from related languages are Baltic. Many elements shared by the Baltic and Slavic languages were not inherited from Proto-Indo-European but borrowed from one another during the period of their common development. However, there are no firm criteria for distinguishing these borrowings from inherited words. The author of the work believes that the Slavic word for ‘hand’ (Russian рука, etc.) could be classified among such Baltic borrowings: it cannot be etymologized within the Slavic languages, whereas it has a reliable etymology in the Baltic languages (rankà: riñkti).

The work also examines some interesting questions of phonetic substitution. These, however, concern mainly the chronology of Germanicisms. But phonetic substitution also affected the old Slavic borrowings from Baltic, of course. The phenomena of phonetic substitution show that, in establishing the chronology of borrowings, one cannot always rely on data concerning phonetic changes. For example, until recently, the signs of the first palatalization in Germanic borrowings had led scholars into a dead end (cf. šelm<xelm, kbneže<kuning-). This is because words borrowed from Germanic otherwise show traces of the second and third palatalizations in Proto-Slavic (cf. cbrky<kirk, cęta<kintus), while by the time of contact between the Slavic and Germanic tribes, the process k’>č’, g’>ž’, x’>š’ must have ended long before. Therefore, the emergence of Slavic šelm-, kbneže is now explained as follows. The result of the first palatalization here is not a living phonetic process, but a phenomenon of substitution. Since the syllables ke, ge, xe had already disappeared from Proto-Slavic by the time of borrowing as a result of the first palatalization, the ke, ge, xe of borrowed words were regularly rendered as č’e, ž’e, š’e. These, and not other, equivalents were chosen because the consonants k, g, x in Proto-Slavic could alternate with č’, ž’, š’ in the same morphemes. Thus, in this case, the rendering of Germanic ge as ž’e in Slavic borrowed words is not reliable evidence for the chronology of the borrowings. The question of the chronology of the earliest Germanicisms in Proto-Slavic also bears indirectly on the age of Germanicisms that entered the Baltic languages through Slavic. As we know, K. Būga maintained that “the very oldest words of Germanic origin among the Slavs do not go back further than the 3rd century B.C.”². S. Bernšteinas dates the beginning of Slavic-Germanic contacts to the 2nd century A.D. (cf. pp. 97, 169). The bulk of Germanic borrowings entered Proto-Slavic in the 2nd–5th centuries A.D. At that time, the Slavs also borrowed the words kotъ, šelmъ, osьlъ, from which our katilas, šalmas, asilas later developed.

S. Bernšteinas’s claims concerning the relations between the Slavic and Germanic languages contradict the hypothesis, recently revived, of a Slavic homeland near the Vistula-Oder³. However, the arguments of the proponents of this hypothesis are, as yet, in fact highly speculative.

S. Bernšteinas points out that we have no evidence that the Slavs lived by the Baltic Sea in the first millennium B.C. Earlier scholars who advanced such ideas based their argument on the presence in Slavic languages of ap-

² Cf. K. Būga, Rinktiniai raštai, vol. 2, Vilnius, 1959, p. 84.

³ Cf. В. В. Мартынов, К лингвистическому обоснованию гипотезы о висло-одерской прародине славян, «Вопросы языкознания», Москва, 1961, № 3, р. 51—59.

316

propriate word mor'e. This word is known in many Indo-European languages, cf. Latin mare, Old Irish muir, Gothic marei, Lithuanian mãrės, mãrios, Latvian mare, Old Prussian mary, etc. However, S. Bernšteinas, relying among other things on the claims of K. Būga and A. Isačenko, points out that the meaning ‘sea’ of the Slavic word mor'e is more recent. Earlier the word meant ‘a pool, standing water’. Thus, it arose in Proto-Slavic before the Slavs became acquainted with the open sea (cf. pp. 63–64). In this way, it is no argument at all that Proto-Slavic speakers lived by the sea.

The author of the work discusses individual questions of Baltic studies at considerable length, providing a characterization of Proto-Slavic changes associated with the period of the Baltic-Slavic linguistic union. He points out that the Baltic protolanguage had already split into many dialects during the period of Baltic-Slavic common development. Consequently, many Baltic-Slavic isoglosses link Proto-Slavic with individual Baltic languages in different ways. Among the very oldest Baltic-Slavic isoglosses one should include the loss of syllabicity by vocalic sonants and their transformation into the diphthongs ir, ur, il, ul, im, um, in, un. Admittedly, many other Indo-European languages also lost syllabic sonants, but this process unfolded differently in them, whereas in Proto-Slavic and Proto-Baltic it coincided remarkably closely.

During the period of Baltic-Slavic common development, intonation began to play an important role in the phonetic and grammatical structure of Proto-Slavic. It became an autonomous factor in the linguistic system. Many processes in this domain are common to Proto-Baltic and Proto-Slavic. Individual differences are likewise explained by the dialectal fragmentation of Proto-Baltic. It is thought that, in terms of intonation, the Old Prussian and Yotvingian dialects were closest to Proto-Slavic (cf. pp. 159–60).

The change of the consonantal sonant m at the end of the Proto-Slavic syllable into n should probably also be assigned to the period of Baltic-Slavic common development (cf. p. 165).

The phenomena of changes in consonants before j are also assigned to this period. Alongside features common to the Baltic and Slavic languages, we also find a number of essential differences here. There is therefore reason to believe that this process, which continued for a very long time in Proto-Slavic, began as the Baltic-Slavic common development was already breaking up. As regards the fate of consonants before j, Latvian and, in part, Old Prussian dialects are closest to Proto-Slavic (cf. pp. 166–167).

In conclusion, it should be noted that the examples from the Baltic languages in S. Bernšteinas’s work are very well chosen and carefully verified. Perhaps, instead of the forms žvaigždė (cf. p. 153) and dėdė (cf. p. 148) given by F. Kurschat, the forms of literary Lithuanian žvaigždė, dėdė should have been used; and instead of the infinitive rinkt, which is not attested in Old Prussian monuments (cf. p. 92), the form senrinka of this verb should have been cited.

Having read with pleasure one part of S. Bernšteinas’s “Outline of Comparative Grammar of the Slavic Languages,” we shall look forward to the appearance of the next part of this work, in which we also hope to find a good many facts significant for the study of the Baltic languages.

Vilnius

A. Sabaliauskas

317