to etymologize and on the basis of Baltic-language material; cf. Lithuanian talė ‘a path through an overgrown meadow’. For the semantics, cf. Latvian taluot ‘to turn white, grow pale’, tals: audeklus liek uz tala (pļavas, nuoras) talāt, balināt¹¹.

The inaccuracies or oversights mentioned here are, of course, minor and do not affect the substance of the work at all.

In conclusion, we would add that the chosen topic (the area under study is geographical, not ethnolinguistic) entailed a very real danger that the conclusions of the work, too, would remain within the bounds of a regional study. Apart from one or two cases in which a definitive resolution of the question was impossible because of the regional profile of the topic (e.g., the eastern boundary of the distribution of Baltic hydronymy, the problem of the origin of the suffixes -анка, -янка, -ейка, -ня, -на, etc.), the authors generally avoided this danger, supporting their conclusions not only with an analysis of hydronyms in the Dnieper basin but also with that of neighboring territories, or with studies by other authors.

The monograph “Linguistic Analysis of the Hydronyms of the Upper Dnieper Region” is truly of great scholarly importance. It is an interesting and significant linguistic work that can serve as an example of how the analysis of linguistic facts can yield conclusions important to many other branches of scholarship. It contains much interesting information about the Baltic homeland, the long period of Baltic-Slavic “symbiosis,” former neighbors of the Balts, and so on—not only for linguists, but also for historians, archaeologists, ethnographers, folklorists, and scholars in other fields. This work will shed much light on the problem of Baltic-Slavic linguistic relations.

A. Vanagas

QUESTIONS OF BALTIC AND SLAVIC ACCENTUATION

V. M. Illich-Svitych, Nominal Accentuation in Baltic and Slavic (The Fate of Accentual Paradigms), Moscow, 1963, 178 pp.

The origin of the intonation and accentuation systems of the Baltic and Slavic languages is an important problem in Indo-European studies. The interpretation of the earliest relations between these languages depends in part on its resolution. It is therefore understandable that, in studying Baltic-Slavic linguistic relations, considerable attention is paid to the accentuation of these languages. In recent years, research into the accentuation of the Baltic and Slavic languages has intensified considerably. Several studies and other articles have appeared¹, in which considerable space is devoted to Baltic accentuation itself

¹¹ K. Mūlenbacha Latviešu valodas vārdnīca, Redigējis, papildinājis, nobeidzis J. Endzelīns, IV sējums, Rīgā, 1929—1932, p. 128.

¹ A. Vaillant, Grammaire comparée des langues slaves, I, 1950, Paris; Le problème des intonations balto-slaves, BSL, 37, 1936, p. 114 tt.; J. Kuryłowicz, L’accentuation des langues indo-européennes, Wrocław—Kraków, 1952, II leid. 1958; Chr. S. Stang, Slavonic accentuation, Oslo, 1957; L. Sadnik, Slavische Akzentuation. I. Die vorhistorische Zeit, Wiesbaden, 1959; E. Nonnenmacher-Pribić, Die baltoslavischen Akzent und Intonationsverhältnisse und ihr quantitativer Reflex im slovakischen, Wiesbaden, 1961; J. Kazlauskas, Dėl lietuvių kalbos vardžodžio kirčiavimo sistemos raidos, „Kalbotyra“, VII, Vilnius, 1963, p. 171—181.

231

system, as well as its relations to the corresponding systems of Slavic and other Indo-European languages.

At present, two opposing theories have been proposed concerning the origin of the accentual systems of the Baltic and Slavic languages.

The founder of the first theory is F. de Saussure. Having analyzed the Lithuanian system of intonations and stress and established its oldest state by means of internal reconstruction, in 1896 he concluded that the Lithuanian opposition “barytone accentual paradigm — mobile-stress accentual paradigm” had developed from the Indo-European opposition “barytone accentual paradigm — oxytone accentual paradigm”^². Thus, F. de Saussure assumed that the Indo-European oxytone paradigm with a fixed accent had been transformed into a mobile-stress paradigm in Lithuanian. This theory therefore posits a direct connection between the intonation systems of the Baltic-Slavic languages, on the one hand, and Greek, on the other, as well as between the paradigmatic systems of the Baltic-Slavic languages, on the one hand, and Greek and Old Indo-Aryan, on the other. In A. Vaillant’s view, the Balts and Slavs have an acute where a laryngeal has been lost. This theory is held by A. Vaillant, L. Sadnik, Ch. Stang, and others.

J. Kurylowicz took a completely different view of the relationship between the accentual systems of Indo-European and those of the Baltic and Slavic languages. He concluded that the accentual systems of the Baltic and Slavic languages are innovations in those languages and therefore have nothing in common with the corresponding systems of Greek, Old Indian, and, in part, Germanic. In J. Kurylowicz’s view, the Baltic and Slavic acute arose when stress was retracted from a short syllable to a long one. Moreover, the opposition characteristic of the Indo-European linguistic unity, “barytone accentual paradigm — mobile-stress oxytone accentual paradigm,” had become mixed up in those languages^³.

It is understandable that, given this situation, researchers in accentology faced the task of using new methodological principles and a larger body of critically assessed material to test the validity of both theories through comparison.

V. Illich-Svitych undertook to address this task in his book “Accentuation of Nouns in the Baltic and Slavic Languages”^⁴. In its preface, the author writes: “It is unlikely that there is another field of Baltic and Slavic linguistics in which disagreements among individual researchers are as great as they are in accentology. The reason is clear: although the Proto-Baltic and Proto-Slavic accentual systems have already been reconstructed in their general outlines, and some details of their structures have been clarified, the main question—the origin of these systems—remains unresolved” (p. 3). And further: “This book is an attempt to find the nominal accentua-

^² F. de Saussure, Lithuanian Accentuation, IF, 6 (Anzeiger) 1896, p. 157—166.

^³ A fuller discussion of this appears in R. Aitzetmüller, Das baltoslavische Akzent- und Intonationssystem. Grundzüge und Grundprobleme, dargestellt nach dem heutigen Stand der Forschung, „Die Sprache“, 8, 1962, p. 46—58.

^⁴ В. М. Иллич-Свитыч, Именная акцентуация в балтийском и славянском (Судьба акцентуационных парадигм), Москва, 1963, 178. р.

232

the common Indo-European basis of the system of accentuation paradigms” (p. 3).

In our opinion, the author of the book has carried out the task he set himself excellently. His research findings allow us to conclude that the accentuation systems of the Baltic and Slavic languages are continuations of the Indo-European accentuation system. The material, meticulously collected, analyzed, and checked, clearly shows that all Lithuanian nominal forms of the barytone paradigm (except for the group with a long non-apophonic vowel in the root) correspond to nominal forms of the same paradigm in Greek, Old Indian, and Germanic. Likewise, Lithuanian nominal forms of the mobile-stress paradigm correspond to nominal forms of the mobile-stress oxytone paradigm in the aforementioned languages (p. 83). The Indo-European opposition “barytone accentuation paradigm — mobile-stress oxytone paradigm” has essentially been preserved in the Slavic languages as well.

V. Illich-Svitych is one of those who regard the accentuation systems of the Baltic and Slavic languages, on the one hand, and of Old Indian and Greek, on the other, as equivalent. By contrast, proponents of the traditional view considered the accentuation systems of the Baltic and Slavic languages to be derived from those of Indian and Greek. V. Illich-Svitych’s view also entails another fundamental proposition: stress mobility in the Baltic languages is ancient, inherited from the period of Indo-European linguistic unity. Consequently, the paradigm reconstructed for Indo-European linguistic unity is not the oxytone paradigm (as representatives of classical linguistics did, relying primarily on evidence from Old Indian and Greek), but the mobile-stress paradigm (p. 12). Moreover, it has been observed that Old Indian and Greek showed a clear tendency to fix paradigmatic stress on a single syllable (in the root or the ending), and that stress mobility in these languages is therefore of a relic nature. For example, Old Indian vṛkas ‘wolf’ and Greek λύκος ‘wolf’ are barytones, whereas Lithuanian vilkas (pp. 40–41, 85, 115, but not 11, 44) belongs to the mobile-stress paradigm (accentuation class 4). According to the apophony theory, a vowel could be reduced only when unstressed; therefore, it must be assumed that these Old Indian and Greek words, too, once belonged to the mobile-stress paradigm. In this respect, Lithuanian is more archaic than Old Indian and Greek. With all this in mind, the author of the work under review writes: “Thus, stress mobility in the Baltic and Slavic languages may reflect not a Balto-Slavic innovation, but an archaism that disappeared in Old Indian and Greek” (p. 162).

Proponents of Baltic-Slavic linguistic unity consider the retraction of stress to an acute root in some ancient oxytones to be an innovation in those languages⁵.

This phenomenon can now be interpreted anew. The considerable number of Baltic and Slavic barytones corresponding to Old Indian and Greek oxytones shows that stress retraction in Baltic and Slavic is a regular phenomenon: ancient oxytones with a long non-apophonic vowel in the root became Bal-

⁵ Cf. Я. Отрембский, Славяно-балтийское языковое единство, ВЯ, № 5, 1954, p. 42.

233

tic and Slavic barytones (Hirt’s law). V. Dybo noted a similar phenomenon in Italic and Celtic, where long stressed root vowels remained intact, while unstressed ones were shortened; cf. Latin fūmus, Lithuanian dūmai, Latvian dūmi, Serbo-Croatian дим and Greek θυμός ‘courage’, Old Indian dhūmaḥ. Thus, in Italic and Celtic (as in Baltic and Slavic), stress must have been retracted from the ending to the root. Even in the oldest parts of the Rigveda, the stress of the aforementioned oxytones, occurring as the second components of compounds, is retracted to the root; that is, they become barytones, e.g.: śakad(h)āmas⁷. In this connection, the author of the work under review writes: “Thus, stress retraction to a syllable with non-apophonic length was apparently not a specifically Baltic-Slavic phenomenon, but an Indo-European dialectal process, possibly directly connected with the weakening (and loss) of laryngeals and with the corresponding tonal strengthening of newly arising (non-apophonic) long vowels” (p. 164).

The evidence presented in the work under review contradicts J. Kurylowicz’s claim that Indo-European o-stem nouns belonging to the barytone and oxytone mobile-stress paradigms merged into a single mobile-stress paradigm in the Baltic and Slavic languages. V. Illich-Svitych gives nineteen examples of Lithuanian nouns with fixed stress corresponding to barytones in other Indo-European languages, thereby clearly showing that this merger was an exclusively Slavic process (p. 119). Thus, this merger cannot be regarded as a common Baltic and Slavic innovation.

The author was able to confirm the connection between the accentuation systems of the Baltic and Slavic languages and that of Indo-European linguistic unity, and to offer new interpretations of certain accentuation phenomena, thanks to the abundant and authentic material collected from old Lithuanian texts, grammars, and dictionaries, as well as from Lithuanian dialects⁸.

We will next mention a few inaccuracies or claims that, in our view, are debatable.

The author’s claim that some northwestern Samogitian dialects contrast two intonations in original acute syllables (one arising from stress retraction and the other belonging to originally stressed syllables), but have only one intonation in original circumflex syllables (p. 66), is inaccurate. He relies on J. Aleksandravičius’s article “Kirtis ir priegaidė Kretingos tarmėje” (LKK, I, 1957, p. 97 ff.). However, the article states that original circumflex syllables also have two intonations: a medial one, arising from the retraction of stress to the root (cf. láikã~laikai and žẽimã~žiemà), and a circumflex one in originally stressed syllables (cf. láiks~laikas, žẽimà~žiemà). Moreover, the article characterizes the medial intonation as follows: “The medial intonation is considerably shorter than the circumflex intonation and is pronounced without prolongation. Its

⁶ V. A. Dybo, Shortening of Long Vowels in the Celtic-Italic Languages and Its Significance for Balto-Slavic and Indo-European Accentology, ВСЯ, No. 5, 1961, p. 18.

⁷ H. Sköld, Zur Akzentzurückziehung auf Akutsilben, ZslPh, IV, 1927, p. 143.

⁸ In the bibliography, we note the absence of K. Būga, Kirčio ir priegaidės mokslas, „Lietuvių kalbos žodynas“, I sąs., Kaunas, 1924, p. XIX—LIII (=K. Būga, Rinktiniai raštai, III, Vilnius, 1961, p. 19—84) and A. Laigonaitė, Literatūrinės lietuvių kalbos kirčiavimas, Vilnius, 1959.

234

The accented phoneme is less tense than that of the circumflex intonation. It also differs from the circumflex intonation in the rise and variation of its pitch” (LKK, I, 1957, p. 105). Thus, the medial intonation is a distinct intonation, different from the circumflex. Consequently, circumflex syllables, like acute syllables, distinguish two intonations.

One cannot agree with the author’s view concerning the accentuation of Lithuanian pūrai, Latvian pûri⁹ (pp. 77–78). The author was unaware of the fact that in Samogitian dialects, alongside pūrai, 2nd and 4th accentuation paradigms, pūrai also occurs (1st accentuation paradigm; cf. Serbo-Croatian pȕr, pȕra). This is the accentuation in Babrungėnai, Darbėnai, Klaipėda (alongside pūrai), Kretinga, Palanga, Pikeliai, Seda, and Skuodas (alongside pūrai, 2nd and 4th accentuation paradigms). If we were to agree with the author that the word pūrai is supposedly an old barytone, Latvian should have *pūri (= Lithuanian pūrai) with a rising intonation. In that case, it would be difficult to explain the Latvian broken intonation, which indicates that the word belonged to the mobile-accent paradigm.

Most likely, we have here a neuter o-stem noun (cf. Old Russian pyро “spelt,” perhaps Old Prussian pure “brome,” alongside old masculine o-stem nouns; cf. Greek πυρός, Sanskrit pūraḥ). In that case, the singular of the Latvian mobile-accent paradigm (pūri, from Baltic singular *pūrōn; cf. Greek πυρός) could be contrasted in accent placement with the plural of the Lithuanian fixed-accent paradigm (pūrai, from Baltic plural *pūrā; further cf. linai, 4th accentuation paradigm, from Baltic linā, alongside singular *linon; cf. Greek λίνον). The Lithuanian circumflex may have had either of two origins: it could have arisen through the later generalization of the singular forms of the mobile-accent paradigm to the plural (in place of older barytonic forms), or through the borrowing of this regionally restricted word from other Baltic dialects (most likely Curonian). Given that both the Lithuanian and Latvian forms of this word are used in western regions, mostly in areas inhabited by Curonians, this cannot be ruled out.

The comparison of Lithuanian rankà with Germanic *wərnhō “corner” (p. 24) is doubtful both phonetically and semantically; moreover, this comparison provides no clear support for the author’s argument (cf. the neighboring oxytone Germanic form). The parallel between Germanic *armaz “arm, upper part of the arm” (cf. Gothic arms) and Lithuanian armai (alongside ārmos, feminine) “part of a cart into which the drawbar is inserted,” illustrating the possibility of a semantic development from “that which is curved” > “arm,” must be rejected, since Lithuanian armai/ārmos¹⁰ is considered a Germanism. The etymology rankà: riñkti is much more reliable.

The Baltic material can hardly help establish the original intonation of the Serbo-Croatian words tȕk, tȕka (p. 151). Both Serbo-Croatian tȕk, tȕka “fat” and Latvian tàuki “fat,” tàuks “fat, stout” may show no metatony; cf. Lithuanian

⁹ An older form pūri (with -r-); pūri may have arisen on the model of kuieš = Lithuanian kviečiai “wheat”; see E. Fraenkel, Litauisches etymologisches Wörterbuch, Heidelberg-Göttingen, 1955 ff., p. 671.

¹⁰ K. Būga, Rinktiniai raštai, II, Vilnius, 1959, p. 531; K. Alminauskis, Die Germanismen des Litauischen, Teil I: Die deutschen Lehnwörter im Litauischen, Leipzig, 1934, p. 27; Lietuvių kalbos žodynas, Vilnius, 1941, p. 251; E. Fraenkel, LEW, p. 16.

235

taũkus (acc. sing. taũkų) “fat(ty), luxuriant”: tūkti “to become fat” (like daũbą: dúbti, sausą: sùsti, raũkti: rùkti, etc.); also belonging here is taũkti, -ia, -ė “to thrive (of grain, livestock).” On the other hand, the Baltic languages also have an acute root; cf. Lithuanian taukai (acc. pl. táukus) “fat”: Latvian tūkstu “I become fat” (like láužti: lùžti: traukti: trūkti, etc.); cf. also tū-las, Old Church Slavonic ty-ti.

Instead of the singular instrumental galvà, the form should be gálva (p. 9), since nouns with an acute root do not shift the accent to the ending in this case (except instrumentals with the ending -mi); cf. lángų, vēlniu, žálčiu, giesme, mėnesiu, vándeniu, juosmeniu).

The Baltic cognate of Slavic *męso (plural męsą) is Samogitian mensà (Kaltinėnai, Vainutas¹¹, Laukuva, Saugos). This word, like its Slavic counterpart, has a circumflex and belongs to the mobile-accent paradigm. The origin of the neighboring Samogitian form meisa is unclear (cf. Old Prussian mensā|menso), although in some Samogitian dialects en can become ei; cf. Samogitian lẽšis, skēisk¹² (from *lenšis “lentil, lens,” *skeńsk “I submerge, drown”). The author incorrectly regards the Latvian form mièsa as borrowed from Slavic. Latvian mièsa also derives from Baltic *mensā (cf. Latvian pièci= Lithuanian penki).

These few critical remarks do not diminish the great value of V. Ilič-Svitič’s work. Such a work was needed for the further development of Baltic and Slavic accentology, and the author of the book, having substantiated the connection between the Baltic and Slavic accentuation systems and the Indo-European system, made a significant contribution to Baltic and Slavic accentology.

S. Karaliūnas

ON THE ORIGIN AND FORMATION OF CERTAIN HYDRONYMS

Alfred Senn, Zur Bildung litauischer Gewässernamen, „Annali dell’ Instituto Universitario Orientale, Sezione Slava“, II, Napoli, 1959, p. 41—47.

Recently, articles and studies on Lithuanian place names have been appearing ever more frequently in foreign linguistic and specialized toponymic literature. Only occasionally, when incomplete or unreliable toponymic data are used, are unfounded conclusions drawn. This can also be said of A. Zen’s article discussed here.

Drawing on the proper names presented in the “Dictionary of Lithuanian Spelling” (1948) and the introductory remarks on Lithuanian place names, A. Zen attempts to describe the more prominent cases of formation among Lithuanian hydronyms. Unfortunately, many of his claims are unsupported, and he even explains the formation and etymology of some Lithuanian hydronyms incorrectly.

The author claims that the proper names of Lithuanian and Latvian have not yet been subjected to genuinely scientific study, and that people are still content with what is now-

¹¹ K. Būga, RR, I, p. 113.

¹² P. Skardžius, ZslPh, XXVI, 1958, 375 tt.; E. Grinaveckienė, LKK, III, 1960, p. 196.

236