LIETUVIŲ KALBOTYROS KLAUSIMAI, IV

LIETUVOS TSR MOKSLŲ AKADEMIJA

BALTIC HYDRONYMY OF THE UPPER DNIEPER REGION

V. N. TOPOROV, O. N. TRUBACHEV

This article briefly presents the main conclusions we reached concerning Baltic hydronymy in the Upper Dnieper basin in our recently completed major monograph, “Linguistic Analysis of the Hydronyms of the Upper Dnieper Region.” In addition to a detailed analysis of the Slavic, Iranian, and Finnish elements of Upper Dnieper hydronymy, this study gives particular attention to traces of the Baltic ethnic element, as reflected in names of bodies of water. The main objectives of this part of the work were to collect, as exhaustively as possible, Baltic hydronyms in the region; to establish the boundaries of the Balts’ distribution in the Dnieper region and their main centers; to describe the word-formation types of Baltic hydronymy in the Upper Dnieper; and, finally, to attempt a preliminary dialectal stratigraphy of the Baltic element in these territories.

As for the first of the tasks mentioned, its solution is presented in an etymological dictionary of Baltic hydronymy in the Dnieper region, which cannot be reproduced here because of its length. We must therefore limit ourselves to a few excerpts.

From our point of view, one of the most significant results is the establishment of the southeastern boundary of Baltic names in the Dnieper region. There is no doubt that it must be placed much farther to the southeast than has hitherto been assumed, in the Seim basin. This is evidenced, first, by instances of direct Balto-Iranian linguistic contact in the Seim basin and, second, by the presence of at least two dozen Baltic river names along the Seim. Balto-Iranian contacts, supported more generally by a number of remarkable Balto-Iranian lexical parallels noted very recently (cf. the names of a whole range of agricultural products, etc.), can also be inferred from some highly specific hydronymic examples. One of the best is the name of a river in the Seim basin Ропши (from an Iranian word meaning ‘fox’), calqued not only by means of the Slavic hydronym Лисичка, but also by means of the name Лопанка (a tributary of the Ропши), formed-

195

from a Baltic root meaning ‘fox’. In this connection, it is interesting to mention a unique case of calquing within a single word: the Irano-Slavic hybrid formation Хартислова (where the first part contains Iranian har- ‘to flow’, while the second reflects Slavic slov- with the same meaning) has a Balto-Slavic doublet, Серцысловка (whose first part derives from *Serti- and has certain Baltic features; cf. Lithuanian Sartai, etc., and Сертея in the Berezina basin, among others). Finally, even the Iranian-origin river name Свапа (from *xvār- ‘good water’; compare nearby Доброводка), as regards its initial s, presupposes the participation of a Baltic linguistic element (otherwise it would be *Хвана). On the other hand, the certain presence of the Balts on the Seim is confirmed by a whole cluster of Baltic hydronyms on the right bank of this river: Обеста, вар. Абеста, Обста (cf. Lithuanian Abista), Клевень (cf. Lithuanian Klevinas), Кубрь, вар. Кубарь, Куберь (cf. Lithuanian Kaubri, Kavbre, Sprogė), Вопка (cf. Вопь, from Baltic ирёlare ‘river’), Лекта (cf. Old Prussian Likutte), Вабля (cf. Old Prussian Wobel, Lithuanian Vabalė; Вабаля, Воболя, Sprogė), Молч (cf. Молча in the Berezina basin, Милча in the Sozh basin; Old Prussian Miluke, Lithuanian Mिलkutis, Milčius), Ратень, вар. Ратень (cf. Lithuanian Rotenis, Old Prussian Roteniten), and others. It is curious that only two Baltic hydronyms have been found to the left of the Seim: Мороча, вар. Мароча (cf. Lithuanian Merkys) and Рехта (see below). This distribution of Baltic names makes it possible to regard the Seim as a kind of boundary of the Balts’ distribution in the southeast, which they did not cross southward. If in this area the Balts were neighbors of the Iranians, who perhaps represent a more ancient ethnic stratum here, then in the other parts of the Upper Dnieper basin the Balts were undoubtedly the principal ethnic component, beginning in the most remote antiquity accessible to linguistic investigation.

The grouping of Baltic hydronyms along the Seim noted below represents an extension of a rather large cluster of corresponding names in the Desna basin. As for the peripheral territories where Baltic-type hydronymy is found, the first to be mentioned is the lower course of the Desna, from its confluence with the Seim to its mouth. For this area, only the name Верепето/Верепата has been cited as a Baltic hydronym (cf. Lithuanian ver- pelas ‘whirlpool’). Nevertheless, there are many more here, although not all the examples are equally convincing: Мордосы (at the very mouth of the Desna; cf. Lithuanian Mardosai, a village name), Тутва (cf. Lithuanian Tytuva), Жевень, вар. Жеведь (cf. Lithuanian Žiaunė, from žiauna ‘jaw’), Шмородовка (cf. Lithuanian Smurgainė), Рехта, Мена, Сперия, Вербца, Силенка, Салова, Кокес, and others.

Already in earlier works on the study of Baltic traces in the hydronymy of territories that are now Slavic, the Pripyat had come to be viewed as the southern boundary of the distribution of Baltic names. In any case, Kochubinsky, Pogodin, and Buga noted a number of examples of this kind, and Buga associated some of them with the Yotvingians (names in -da along the Pripyat and the Western Bug). Nevertheless, the basin

196

The Pripyat had not previously been considered a place with a very large concentration of Baltic-type hydronymy. At present, at least fifty names of Baltic origin can be cited from the left bank of the Pripyat. This seems sufficient to assert that the hydronymy north of the Pripyat bears a clear imprint of Baltic origin. Unlike the areas along the Seim and lower Desna, where Baltic examples are in the minority and, in any case, sporadic, on the left bank of the Pripyat Baltic hydronymy forms the principal stratum. These include Moroč (see above), Maža (cf. Lith. Maž-upē), Nača, variant Nač (cf. Old Prussian Notis, Lith. Notija, Nočia, etc.), Nenac (see Nača), Beržica (from Baltic berḡ- ‘birch’, cf. Beržūona), Mytvica (cf. Lith. Mituva, Latv. Mitava), Orazhnia, variant Orizhnia (cf. Old Prussian Aryngine), Osvica (cf. Lith. Ašva), Nertka (cf. Lith. Neriata; Sprogis), Shalbenka, variant Sholbenka (from Baltic *Skalb-/*Skelb-), Plisa (cf. Lith. Plisa), Moroža (cf. Lith. Margis), Uča (cf. Lith. Ančia), Cna (cf. Old Prussian tusnan ‘quiet’, Tuseine), Yaselda, Pina, Pipin, Temra, Lan, Pershnitsa, Vessia, variant Eseika, Sivolka, variant Sivelga, Dokolka, Dupolka, Shatska, Shanka, Yashna, Nemeglia, Kaklanka, variant Orazhnia, Neratovka, Oressa, Snerolka, Yazelitsa, Nesla, Nesetnia, Tremlya, Kovna, Narovlya, Sverizh, and others. Many of these names have fairly clear correspondences in the hydronymy of present-day Baltic territories. Taken together, all these facts suggest that the Baltic ethnic element in this area continued to persist for a considerable time. Of great interest—especially since this has not been discussed before—is the presence of several river names, apparently of Baltic etymology, south of the Pripyat. These include in particular the names Mytva, right-bank tributary of the Pripyat; Verpa, right-bank tributary of the Slavechna; Cheressa, right-bank tributary of the Devoshin; Zhelon, right-bank tributary of the Slavechna, with its tributaries Prizholonok, Zhelonka; Zhelon, a right-bank tributary of the Sitovka. Of course, considered in isolation, these names may be etymologized differently; however, in the general context of the hydronyms adjoining this area from the north, the conclusion that the names in question are of Baltic origin seems very likely, especially since a number of names recur in territories where Baltic elements are beyond doubt (cf. Zhelon, Vereneto, Mytvitsa north of the Pripyat). Incidentally, the presence of Baltic elements on both sides of the Pripyat seems doubtful only if the Pripyat constituted an impassable boundary, as is commonly held in the literature. This view, it must be supposed, now requires serious revision. In addition to the examples cited, we have other evidence that casts doubt on the isolating role of the Pripyat. First and foremost, this concerns the recurrence of river names north and south of the Pripyat. In particular, this applies to the name Sluch, found on both sides of the Pripyat, not to mention some less significant hydronyms, sometimes occurring with slight modifications. It is therefore unsurprising that south of the Pripyat there are a number of hydronyms apparently of Baltic origin for which correspondences can be found not only on the left bank of the Pripyat but also in other parts of the Upper Dnieper region.

197

For example, cf. Sazhelka, a right-bank tributary of the Rodcha, a right-bank tributary of the Ubort, a right-bank tributary of the Pripyat, and Sazhelka (Yelninsky district, Smolensk province); Bastova, left-bank tributary of the Usha, a right-bank tributary of the Pripyat; likewise Bastovka, left-bank tributary of the Usha, cf. Bostya; Vishka, a right-bank tributary of the Ubort, a right-bank tributary of the Pripyat, cf. Visha, a right-bank tributary of the Ipa, a left-bank tributary of the Pripyat (Lith. Viesia); Svidovka, a right-bank tributary of the Ubort, a right-bank tributary of the Pripyat; likewise Svidovets, a lake in the Svidovka basin, cf. Svida, in the Berezina basin, and Svidnia, in the Desna basin; Losha, a left-bank tributary of the Cherten, a left-bank tributary of the Slavechna, a right-bank tributary of the Pripyat, cf. Losha, in the Sozh and Desna basins, and some others. In addition to these cases, we should probably mention Baltic hydronyms south of the Pripyat that usually have no correspondence in other parts of the Upper Dnieper region: Skolodinka, a right-bank tributary of the Pripyat, cf. Old Prussian Skalden, Scolden, Skoldainiai; Rodcha, a right-bank tributary of the Ubort, a right-bank tributary of the Pripyat, variant Radcha, cf. Old Prussian Rudicus; likewise Racha, a left-bank tributary of the Grezli, a left-bank tributary of the Usha, a right-bank tributary of the Pripyat, variant Radcha); Vigovka, a left-bank tributary of the Usha, a right-bank tributary of the Pripyat, cf. Lith. Vygala, a river, Prussian Wigal; Narovlya, a right-bank tributary of the Pripyat, variant Narovlyanka, cf. Polish Narew, Prussian Narussa; Sitovka, a right-bank tributary of the Pripyat, cf. Prussian Seyte, a river, Lith. Sietuva; Cholnitsa, a right-bank tributary of the Styr, a right-bank tributary of the Pripyat, cf. also Cholna, Cholkna in the Iput basin; Perega, a right-bank tributary of the Ubort, a right-bank tributary of the Pripyat; Zolnia, a left-bank tributary of the Ubort, a right-bank tributary of the Pripyat; likewise Zolonka, and others. It is noteworthy that traces of Baltic hydronymy are concentrated in the eastern part of the right bank of the Pripyat (along the Ubort, Slavechna, and Usha), whereas we have found none in the western part (along the Horyn). Thus, the considerations presented compel us to revise the traditional view of the Pripyat as a Baltic-Slavic watershed (cf. Buga) and, at the same time, to clarify the character of this southern periphery of the Baltic hydronymic area along the Dnieper.

One of the more substantial concentrations of Baltic hydronymy in the Upper Dnieper region is in the Sozh basin. It is significant that for a long time researchers considered the presence of Baltic names in this area impossible. Only in the 1920s did Buga, contrary to his earlier view, consider it permissible to identify several Baltic river names along the Sozh. Vasmer likewise acknowledged the possibility that several names of Baltic origin were present in this area. However, as a result of the work of these scholars, fewer than ten names had been regarded as Baltic to date: Volchesa, Chetovka, Liamna, Krunka, Uzny, Luchesa, Lyza, Chechera/Chechora. A comprehensive survey of the hydronymy of this area has shown that the number of hydronyms cited can be increased at least fifteenfold (not counting variants). It should be noted at once that, in terms of reliability and the presence of correspondences in the hydronymy of present-day Baltic territories, the large number of names presented below is no less convincing than those identified earlier. Also noteworthy is the even distribution of Baltic hydronymy throughout the Sozh basin, which, among other things, indicates that Baltic settlements did not end within this basin but extended farther east. Without claiming to provide an exhaustive list here, and without repeating what

198

as has already been noted by previous researchers, we may cite such names as Vipenka, Upynka (from Baltic upė ‘river’), Natopa (from Baltic *Nat-up-/*Nat-ap-, cf. Lithuanian noterė, Prussian noatis ‘nettle’), Krotka (cf. Lithuanian Krauta, Prussian Kreutho, Kreuto), Remistovo (cf. Prussian Ramestow), Gizhinka, Geishinka (cf. Prussian Gezyelauken), Vorminka (Prussian Wormen), Rudeya (from Baltic *Rudēja, cf. Lithuanian Rudė, Rudya. Sprogis, Prussian Rudenik, Ruditen, etc.), Zhubr (cf. Lithuanian Zhauberiškiai. Sprogis, Prussian žiauburys ‘elevated place’), Molinka (cf. Prussian Moleyn, Lithuanian Molupis, -ys, Moline. Sprogis, etc.), Ochesa (from Baltic *Akesa, cf. Lithuanian Akele), Nevezha (cf. Lithuanian Nevėžis), Kolpita (a modification of a Baltic name, cf. Prussian Gulbiten), Vorzhanka (cf. Verzha, Lithuanian Viržuva), Gabya (cf. also Gabel, Gobenka, Goben, Gobik, Gobesk, Gobya; Lithuanian Gabinės, Prussian Gobayn, Gabelin, or Latvian Gauja?), Nadva (Prussian Nedwe), Resta (cf. the Baltic name for ‘nut’; comparison with Lithuanian raistas ‘swamp’, Prussian Raystopelk, Reisten is also possible), Rekta/Rekhta (possibly from Baltic dialectal *riekst->Slavic *rėkt->*rėxt-), Ovchesa, Sapelka, Perenka, Piryanka, Lobzha, Udoga, Loptsa, Vopek, Shelbitsy, Cholkna, Otkilnya, Chelena, Voprat, Oporot, Boldryzha, Domsha, Shcholbnya, Selnya, Selinka, Skvereta, Dymalki, Senzha, Sitinka, Nedna, Navlya, Lyaduzha, Kalita, Drogonka, Selenka, Bovsha, Zhdovnitsa, Shuma, Zhadunka, Zhadyn, Paluets, Polonka, Lobno, Besed, Basya, Pronya, Chavenka, Koshanka, Koshka, Rybnitsa, Dubrech, Kasna, Salozha, Stolet, Stolyatka, Kartel, Garista, Bestevenka, Veprenka, Polna, Turosna (Trosna), Sudinka, Gzhun, Slun, Slun-Milcha, Ivol, etc. Thus, the Sozh basin is one of the most significant in terms of the number of hydronyms and the density of the Baltic cluster extending from the headwaters of the Ugra and Oka, where Baltic hydronymy is encountered far less frequently, to the lower Desna, where the number of Baltic names declines noticeably.

East of the Sozh basin, Baltic hydronymy is concentrated in the middle and upper Desna basin. Until now, Baltic hydronymy in this area had practically not been identified. Nevertheless, it undoubtedly exists here (about a hundred names), although quantitatively it falls short of the Baltic names of the Sozh region by at least one third, and is often less transparent etymologically. Among the Baltic names of the middle and upper Desna region, the following should be noted: Strena, variant Stryana (cf. Lithuanian Stirna, Stirnė), Stryannitsa (see Stryana), Bolva (cf. Latvian Bolva, Lithuanian Balvis, Prussian Balowe), Vedga (cf. Vedoga, Veduga, Uduga; Lithuanian Vadaga, Vedega), Vodocha (cf. Lithuanian Vadak(s)tis), Leven (cf. Lithuanian Lėvuo), Chechera (cf. Lithuanian Čičirys, Latvian Ciecere), Kitseta (cf. Prussian Keikot), Verzheika (see Verzha), Lokch (Prussian Lukte, Lucten), Lopanka (cf. Prussian Lopino, Lapraupen, Lithuanian Lapinas, Lapainė, Latvian Lapaina; see also above), Zalyatka (cf. Lithuanian žolėtas ‘covered with grass’), Svidnya (cf. Prussian Swyden, Sweiden), Gabya, Gobya, Gobesk (see above), Lokna (cf. Latvian Lukna, Lithuanian Luknė, Prussian Löckeneyn), Lemnitsa (cf. Lithuanian Liminac), Orlya (cf. Prussian Arle, Lithuanian Orlya. Sprogis), Nerusa (cf. Prussian Narussa), Degna

199

(cf. Latvian Deguns, Deguni), Seriazha (cf. Prussian Seringe), Steneva (cf. Prussian Stenge, Lithuanian Stangė-is), Krupela (cf. Prussian Craupolythen, Kraupol-), Venera, Inochka, Ulepka, Zhevanka, Ledmen, Navlya, Ramosukha, Vablya, Vara, Varya, Svesa, Nakot, Selnya, Polna, variant Palnya, Ustar, variant Tara, Ratcha, Nemda, Litiya, Dumcha, Ponemel, Sheverka. Lomna, Nedna, Prismara, Prismarka, Bestan, Vosavka, Tureyka, Damynka, Salozha, Nepolob, Peretesna, Lokaten, Smorzhinovka, Osova, Solya, Sudost, Lomnya, Rudomlya, Lopuzna, Labezna, Berezna, Tuitsa, Stogorevka, Golbrosovka, Trosna, Ulitsa, Ivot/Ivot, Vetma, Vitemka, Bereshchanka, Nessa, etc. The decrease in the number of Baltic names in the Desna basin compared with the Sozh basin, as well as their decline within the Desna basin itself from west to east, may apparently be interpreted as evidence that the easternmost boundaries of the Baltic tribes’ distribution lay nearby. However, further clarification is possible only through a survey of hydronymy along the upper Oka. West of the Sozh basin (in the upper Dnieper), there are territories where many traces of Baltic hydronymy had already been identified earlier. Nevertheless, the gaps here, too, were considerable. At present, the number of known Baltic hydronyms in the upper Dnieper (as far as the mouth of the Sozh, excluding the Berezina basin) can be substantially increased. The following river names in this area may be classified as Baltic: Abolna (cf. Latvian Abula, Abuls, Obole, Obol, Obolyanka in the Western Dvina basin), Stabna/Stabenka, Stabnya (cf. Prussian Stabingen, Stabynotilte), Korsa (cf. Prussian Corseen, Corslauken), Styabura/Stebula (cf. Lithuanian stimbras, Latvian stiebps), Vop (cf. Baltic upė/ape), Vopets (see Vop, etc.), Osma (cf. Lithuanian Ašmena), Kolpitka (see Kolpita), Matolka (cf. Prussian Matulen, etc.), Skorodka (cf. Lithuanian Skardupis, Prussian Scharde, Zkarde), Vedosa/Vedosa (cf. Lithuanian Bedoca. Sporis), Vorlinka/Vorlyanka (cf. Lithuanian Valinis), Morzhovka (cf. Lithuanian Murgenis), Losmina/Losmena, Losmenka (cf. Lithuanian Lašmuo), Vodva (cf. Lithuanian Vaduva), Orleya/Arleya (see Orlin), Dryagnovka (cf. Lithuanian drėgnas, Latvian dregns), Kidel (cf. Lithuanian Kidulis. Sporis, Latvian Cidulupe), Neroplya (from Baltic *Ner-upė/*Ner-apė), Serteya (cf. Lithuanian Sartė, Sartai, etc.), Votrya, Votrica, Apelka, Klitsa, Chetovo, Vedosa/Veduga, Iskozha, Bostya, Obisha, Opisana, Lomna, Lemnya, Lyamna/Lemna, Lyamnitsa, Levna, Kolonka, Degovka, Sherevenka, Zhalobyka, Mereya/Mireya, Mereyka, Labarovka, Ageya, Zapakulka/Sapakulka, Lomsha, Bolonovka, Svidera, Peresna, Dobosna/Dobysna, Babich, Rekta, Rekhta, Rekuta, Rekotka, Rekotna, Yaseyka, Politnya, Svesovka, Plinki, Kolyaya, Bestenka, Be stvenka, Ocheska, Elcha, Varka, Gitsenka, Balechuka, Varsoha/Varsoha, Polna/Polnya, Polna, Zhagulina, Vyazovenka/Dubrovinka, Birzheevka, Polnya, Dorogotnya, Selnya, Negra, Rasna/Ryasna, Kletka, Metelka, Zalazenka, Zelezenka, Zhilo zhka, Solya/Sol, Podsolok, Zalazenka/Zalazinka, Zherty/Zherti/Zherd, Azhovka, Molyanka/Molenka/Molinka, Ulitsa, Shedalovka, Gzholka/Gzhelka, Verzha, Shakolovsky, Smogorovka/Smogoryanka, Morchanka, Nadva, Nadva/Nagovny, Zalazna, Smolka, Verchanka, Drut, Adrov/Odrov, Rudeya/Radeya, Boreyshovka, Ruz-

200

ka, Kureyka/Koreyka, Bebrya, Bobreyka, Volcheyka, Neresno, Zhupur, Yernevka, Berezinka, Berezenets, Berezina, Berezinka/Berezina, Ivatal/Ivota, Lupa/Lupogolova, Sertsislavka, Trosna, Madora/Maderka, etc. (a total of at least one and a half hundred). It is noteworthy that in the upper reaches of the Dnieper, the number of Baltic hydronyms in relative terms is somewhat lower than along the Sozh. On the other hand, within the uppermost part of the Dnieper itself, the distribution of Baltic names is somewhat uneven: there are relatively few along the Dnieper up to its turn southward, and considerably more on the right bank, adjacent to the Berezina basin, and farther south, as one approaches the mouth of the Sozh. The reliability of most of the Baltic traces in the hydronymy of this region noted above is beyond doubt. It should be noted that the Baltic settlements apparently were not among the peripheral ones. To their east, Baltic names continue, though in considerably smaller numbers, in the Oka and Moskva basins (in their upper reaches). To the northeast, the upper reaches of the Dnieper adjoin a small number of entirely reliable Baltic names on the right-bank side of the Upper Volga. Northward, the Baltic hydronymy of the Upper Dnieper region extends into the basin of the Western Dvina. It is interesting to note that Baltic names analogous to those identified in the upper reaches of the Dnieper occur in all the territories listed, for example,

Obisha — Obsha, Kidel — Tsidel, Vop — Ula, Sertega — Serteyka, Abolna — Obolyanka, Obol, Negra — Negryanka, etc. The Baltic hydronymy of the Berezina basin has the greatest relative share. The relative number of such names here is so large, and their Baltic character so evident, that the idea of a relatively recent presence of a Baltic population in this territory leaves no room for doubt. This is also confirmed by some direct historical evidence. The East Baltic affiliation of the Berezina hydronymy is equally reliable; it reveals numerous connections with the hydronymy of modern Lithuania and Latvia. Overall, the following river and lake names in the Berezina basin may be considered Baltic (since their etymology is most often transparent, we shall simply list the hydronyms concerned): Berezina, Berezanka, Berezhnitsa, Beresnyanka, Myadziol, Myadelitsa, Ponya, Varlynka, Mozhayka, Nidalka, Serech, Mantso, Manets, Mancha, Osovina, Osova, Osovka, Kleva, Klevitsa, Ola, Nacha, Zhortayka, Pelik, Plaviya, Stav nitsa, Vyacha, Vencha, Grivka, Bolocha, Molcha, Rinelka, Sved/Svida, Olsa, Neseta, Tsitovka/Tsitva, Myana, Mozha, Vozha, Vazha, Diy/Diya, Volma, Brozha, Dobrelina, Klupcha, Bruch, Svisloch, Mray, Pyadynka, Plis(s)a, Mesreda, Tsna, Isa, Skha/Tskha, Usyazh/Usyazha, Gayna/Goyna, Volodut, Rova, Rava, Pustometushka, Sinyaya/Sin, Ochicha, Pentritsa, Nerchay/Nerchanka, Bobruyka, Yanchanka, Solenka, etc. Thus, the density of Baltic hydronymy in this basin exceeds all cases noted so far (about 80 Baltic names out of a total of about 300).

201

The identification of a very substantial number of Baltic hydronyms in the Upper Dnieper region (700–800 names) and the study of their distribution in this territory—which revealed a continuous and, on the whole, fairly even distribution of Baltic names, with no major gaps—compel us to draw a conclusion that could not previously have been reached on the basis of a small number of geographically restricted examples. The conclusion is as follows: in the first millennium CE and the first centuries of the second millennium, the territory of the Upper Dnieper region was inhabited by tribes of Baltic origin; it is impossible to explain the names, preserved in such numbers over many centuries, by assuming the existence of isolated colonies of prisoners of war and later settlers, as some scholars have tended to think. Special emphasis should be placed on the idea that the Baltic population of the Upper Dnieper region, in its main body, did not retreat to the northwest as the Slavs spread across this territory, but was gradually assimilated by them. Apparently, only this assumption makes it possible to understand both the preservation of such a significant number of names of Baltic origin and their Slavicized phonetic form, which misled scholars who approached the identification of Baltic traces using the yardstick of traditional phonetic Balto-Slavic correspondences, derived chiefly from originally related vocabulary. In light of what has been said, the question of the contribution, in the broadest sense, of the ancient Baltic population to the history of the Upper Dnieper region assumes particular importance. Of course, this question cannot be resolved in this article.

Until now, the question of Baltic word-formation types in Upper Dnieper hydronymy has practically not been raised. Researchers have confined themselves to searching for new Baltic lexemes, apparently considering the material too scant for conclusions about word formation. For our part, we believe it possible to assert that word-formation analysis promises the most significant results, including in addressing questions that are difficult to answer on the basis of lexical material alone.

Analyzing the word-formation types of Baltic hydronymy in the Dnieper region, we conclude that, in their main features, they reflect the most authoritative hydronymic formatives known in Lithuania, Latvia, and former East Prussia. Naturally, it is most difficult to identify those hydronymic types that, in their Baltic form, lacked distinctive word-formation features and must ultimately have lost their distinctiveness in the process by which the Slavic population adopted the corresponding Baltic hydronyms. The widespread Baltic word-formation types in -(a)s, -us, -a, fully current in the hydronymy of the Baltic territories as well, were undoubtedly represented in the Baltic hydronymy of the Dnieper region. However, the types in -as, us, characte-

202

…characterizing Baltic words of masculine gender were accordingly rendered by masculine formations in the Slavic languages (-ъ, and after its disappearance, -о or the absence of an inflection). The clarity of this kind of phonetic-morphological substitution was apparently explained by the existence of Baltic-Slavic bilingualism in the Upper Dnieper region during a certain period. Admittedly, the condition just mentioned alone would not have sufficed for this. Apparently, the decisive role was played by speakers’ perception of the closeness of the languages in contact, in particular, by the approximately identical principle of contrasting masculine and feminine gender; cf. Baltic -as (masc.): -a (fem.) = vs. Slavic -ъ (masc.): -а (fem.). This correspondence created conditions for the almost automatic conversion of Baltic hydronyms ending in -as, -us into the Slavic form ending in -ъ. All grammatical characteristics were preserved in this conversion, which was very important for both languages. For these reasons, the absence of -as, -us is understandable both among hydronymic traces in the Upper Dnieper region and among Baltic borrowings (common nouns) in Russian and Belarusian dialects. It is all the more understandable that Baltic names of bodies of water in the Dnieper region lack an independent Baltic hydronymic type ending in -а. It would have coincided completely with the corresponding Slavic type, with the same feminine-gender characteristic -а. The complete absorption of the Baltic hydronymic type ending in -а into a similar Slavic type was facilitated by the fact that formations of this kind are used with roughly equal frequency in hydronymy in both Slavic and Baltic languages, since they are associated, respectively, with feminine-gender words meaning ‘river’: Baltic upė/ape. Therefore, the absence of Baltic hydronymic types ending in -as, -us, -а in the Upper Dnieper region does not mean that they were not present here earlier. On the contrary, their former presence here is beyond doubt, as can be established through etymological analysis of non-derived hydronymic stems of Baltic origin. Everything we know about the Baltic-Slavic symbiosis in the Upper Dnieper region makes the absorption of the Baltic types mentioned above seem natural; the opposite outcome would seem more surprising. The adoption of Baltic feminine hydronyms with a soft-stem variant ending in -а proceeded somewhat differently; cf. Lithuanian Almė, Kurmė, Labė, Nevežė, Taurė. Since the corresponding soft-stem variant ending in -а in East Slavic dialects differed somewhat from the Baltic one (cf. Baltic -ė — Slavic -’а) and, above all, was not stable in all contexts, direct adoption of Baltic hydronyms ending in -ė as Slavic forms ending in -’а is relatively uncommon. It should be noted, however, that even these few cases often appear in a somewhat obscured form; cf. Невежа alongside Lithuanian Nevežė, Nevė-ža; Толща alongside Lithuanian Tilžė; Ольша — Lithuanian Alšia, Вопля — Lithuanian upė ‘river’.

The presence of Baltic hydronymic traces with the soft variant of the masculine stem -as-, as well as with masculine and feminine stems ending in -i-, can be traced much more clearly, since these hydronyms correspond in the Upper Dnieper region to forms with a specific feature—the softness of the final consonant, for-

203

formally, by the presence of -ь. In addition to direct reflexes in the hydronymy of the Upper Dnieper, the inventory of names in -i- may be augmented by forms with a soft-stem variant ending in -а, which, as is known, was closely related to the stem in -i-. In addition to examples such as Морочь — Мороца, Мелечь — Мильча, Мильча, Молча, Молча, Nevėžis — Neveža, it is worth drawing attention once again to the pairs Наць — вар. Наца. The usual correspondences here allow us to propose as a source the Baltic hydronymic stem in -i- Noti-; cf. Old Prussian Notis, and also Lithuanian Notė. Traces of Baltic hydronymic types of this kind in the Upper Dnieper region were apparently fairly numerous, as evidenced, in particular, by the somewhat unusual quantitative ratio of hydronyms ending in -ь to the west and east of the Dnieper.

The presence in Baltic hydronymy of a number of suffixes containing -к- makes it conceivable in principle that some of these formations were retained in river names in the Dnieper region. However, the substantial differences between Baltic and Slavic languages in the ways -к- is attached to a stem, together with chronological considerations, probably make it inadvisable to seek traces of Baltic hydronymic types among Slavic water names ending in -ка. Nevertheless, a very considerable number of Baltic hydronymic roots in our region have been extended directly by secondary Slavic -ка, or as part of more complex suffixes; cf. Лучесня, Вытрейка, Мерейка, Варлинка, Колпитка, and others. This fact is important for characterizing the details of the process by which Baltic hydronyms of the Upper Dnieper region became Slavic word-formation types. Of particular note is a small, compact group of hydronyms with the element -ол-, characteristic of Baltic word formation, concealed by the Slavic suffix -ка: Матолка, Старолка, Вихолка, Снеролка, Дуполка, Доколка, and others. The Baltic character of the names in this group is indicated by the hydronym Мядзиол in the Berezina basin, which retains greater similarity to the corresponding Baltic type in -ol-, not yet extended by Slavic -ка; cf. Lithuanian Medžiolai, a village near Vievis. In principle, Baltic hydronymic formations in -al- may also be concealed behind the formation in -ол- mentioned above; cf. Lithuanian Vygala, Snie-tala, Latvian Jumala, Makale, and others. The element -i- in names of this kind is a sufficiently clear indication of Baltic origin, especially since Slavic hydronyms do not have this -i- in the characteristic position next to a vowel forming part of the stem.

Hydronymic types of the Upper Dnieper region containing -н- in various suffixal combinations are for the most part part of the Slavic stratum, and are often of fairly recent formation. Accordingly, the proportion of possible Baltic word-formation reflexes in this section is also relatively low. Despite this, it is possible—even on the basis of isolated examples—to identify some very important Baltic hydronymic formatives. First of all, this may concern the name of the Berezina and a number of other ‘birch’ rivers, in which

204

The presence of the Baltic word-forming element -n- is beyond doubt. The difficulty of identifying a specifically Baltic type in this case lay in the popularity, also in Slavic hydronymy, of the word-forming type with -н- and even -ин(-а). Only comparison of certain differences in word-formation patterns and in the meaning of the element -in-, seemingly insignificant at first glance, makes it possible to identify a specifically Baltic type in an outwardly Slavic hydronym. From the large group of Slavic hydronyms of adjectival origin in -на(-вна), an entirely Baltic name such as Стабна/Стабня (historically Стабья; cf. Стабенка) can be distinguished, in which one may justifiably recognize the Baltic hydronymic formant -inas (cf. Lithuanian dialectal Šabinas, Old Prussian Stabinotilte), rather than the Slavic suffix mentioned above. It is possible that in the hydronyms Вупенка, Упинька we are dealing with a Slavic reinterpretation, as diminutive suffixal formations in -енька-, -инька, of the Baltic formant -in- (cf. Lithuanian upinis ‘riverine’), which does not have this meaning at all. Many examples of this kind of reinterpretation can presumably be found among Slavic hydronyms in -енька, -инька; cf. Чавенка, Чапенка, Доренка, Седеленка, Шуешенка, Шервенка, Вороженка, Вупенка, Варлинка, Варльинка, Ворминка. Among the Baltic hydronyms of the Upper Dnieper region, several end in -ень, concealing two independent word-forming types of Baltic hydronymy. On the one hand, in examples such as Клевень, Верепень, it is reasonable to see continuations of the Baltic suffix -inis (cf. Klevinė, a village near Kapsukas; Verpetinė, a meadow near Pabiržė), which was adapted on Slavic soil as -ень, hence -ень, which coincided with the corresponding Slavic suffix. As a possibility, one may mention a permissible reflex of the Baltic suffix -en-, especially since this would be consistent with the evident absence of a mobile -е- in our example. On the other hand, examples in -енье, such as Левень, provide grounds for positing an old consonant stem in -n-. The likelihood that other Baltic consonant stems are reflected in our territory is confirmed by the presence of a name such as Лосьмена, which points to an old stem in -men-, also known in Lithuanian hydronymy; cf. Lašmio. Interestingly, this name is known in the Upper Dnieper region in a whole range of variants: Лосьмина, Лосьменка, Лосменка, Лосменная, Лосьмонька (with the characteristic vocalism of this suffix), Лосвина, Лосвянка. Although stems in -men- also occur in Slavic hydronymy, cf. Вязьмень, Сухмяна and others, they cannot be confused with the corresponding Baltic stems. A Baltic hydronym with the corresponding stem may also be suspected as the source of the river name Ледмен. Apparently, an old stem in -men- survives in residual form in the hydronymy of Осьма in the very upper reaches of the Dnieper; cf. Lithuanian Ašmena. Notably, the name Каменка occurs frequently in the Osma basin and in the areas immediately adjacent to it, which together closely resembles the pattern observed in the eastern part of Lithuanian territory.

205

A distinctive word-forming type in -ēja is characteristic of the hydronymy of all Baltic territories. The distribution of Baltic hydronyms in -ija is more limited (they are absent in Latvia). Hydronyms in -auja are even rarer; cf. Miltauja, Pienauja, Vidauja, etc. These and similar formations constitute a distinct group of hydronymic formants closely related to one another in the Baltic languages. The most prominent of the Baltic hydronymic types listed are also found in the Upper Dnieper region. This applies first and foremost to hydronyms in -ея (Мерея, Сертей, Локнея, etc.), some of which are reflected in names in -ейка (Мерейка, Веселейка, Верхнейка, and others). For the latter case, it is reasonable to suppose that at least some examples may be directly connected with the Baltic formant -eika. The Baltic hydronymic type in -ija is also attested by several examples in the Upper Dnieper region: Орлия, Весия, Луния, and others. Special mention is due to the sole hydronym in -уй-ка—Бобруйка, an expanded form of the name Бобруя, for which Lithuanian *Bebr auja, *Babraja would be the corresponding forms.

A fairly widespread Baltic class of hydronyms with the element -t-, which includes a number of variants depending on the nature of the preceding vowel, is represented in the Upper Dnieper region primarily by two of its most common types. On the one hand, there are names in -ита, for which a corresponding Baltic stem can be identified only in isolated cases (cf. Колпута—Prussian Gulbiten). It is also useful to bear in mind names expanded with the Slavic -ка: Колпутка, Хохлитка, and others. On the other hand, the Baltic word-forming type with the suffix -eta is reflected in Upper Dnieper hydronymy; cf. Скверета, Несета, Верепета, Кичета. Interestingly, beyond the Dnieper, Гордота is known (Upper Volga), in which one may see a reflex of another variant of the Baltic suffix containing the element -t-. Some further Baltic word-forming types may be discerned in the hydronymy of the Upper Dnieper region, but only upon closer analysis; cf. Водоча, Lithuanian Vadaktis, Vadakstis. It is possible that among hydronyms in -ча there are some names concealing old Baltic formants with the element -t-, but at present it would be imprudent to make a more specific claim.

Finally, we can discern traces of another Baltic word-forming type in Upper Dnieper hydronymy. This is the formant -ista, which within Baltic word formation is related to the suffixes -asta, -esta, -osta, -uosta, -usta. In our territory, the hydronym Обеста/Обиста/Обста/Абеста should be noted in this connection. Beyond the Upper Dnieper region, besides a number of Baltic hydronyms such as Lithuanian Aknysta, Lokysta, Laukysta, etc., there is an exact parallel in the Lithuanian water name Abista.

The regional Baltic hydronymic type in -(u)va is represented in the Upper Dnieper region by the hydronyms Водва, Титва, Cit-

206

sha, Mitva, Nadva, possibly, Bolva, and others. Baltic hydronymy is characterized by the presence of two distinct formations with the suffixal element -s-. On the one hand, this concerns the marker -sa, which can be clearly distinguished only in certain cases; on the other hand, it concerns several suffixal variants depending on the nature of the vowel element preceding this -s-: -esa, -asa-, -osa-, -usa-. Apparently, all these types are reflected to some extent in the hydronymy of the Upper Dnieper region; cf. those in -sa: Vodsa, Korsa, Usa, Plisa; in -asa: Volchasa, Ochasa; in -esa: Luchesa, Volchesa, Ovchesa, Ochesa, Oressa; in -osa: Vodosa, Orossa, Mordosy; in -usa: Nerusa, Voronusa. The certainty of these cases makes it unnecessary to discuss them in greater detail.

In the formant -sha of a number of Upper Dnieper hydronyms, at least in individual cases, the reflection of specific Baltic word-formation types is highly probable. However, the small number of examples and, above all, the absence of exact Baltic correspondences encompassing the root as well do not allow us to determine with confidence whether we are dealing with the old suffix -sa, -usa, or some other suffix. The only possible correspondence may be found in Prussian Dargow (cf. Dorogovisha), which, however, lacks suffixal marking. Thus, Upper Dnieper hydronyms in -sha have less definite Baltic correspondences, which are further obscured by cases such as Ropsha, whose Iranian origin is evident. Although Baltic names are undoubtedly present among Upper Dnieper hydronyms in -zha, it is quite difficult to speak here of the reflection of a specific Baltic word-formation formant. The point is that some hydronyms of Baltic origin in -zha have sufficiently convincing correspondences in the hydronymy of present-day Baltic territories, clearly indicating that the element -zh- belongs not to the suffix but to the root; cf. Nevezha — Lithuanian Nevėža, Morocha — Prussian Margis. Only in individual cases can -zh- be considered suffixal; this is also confirmed by correspondences from Baltic hydronymy, for example, Lobzha alongside Latvian Lubūzis. The suffix is somewhat harder to interpret when the element -zh- is preceded by a vowel (-yzha, -izha). If these suffixes are taken to reflect Baltic word-formation types, one might be tempted to compare them with the suffix -ng- (-ing-, -ung-, -eng-), common in Baltic hydronymy and toponymy. However, the regional restriction of this suffix in present-day Baltic territories, as well as the absence of fully reliable parallels, makes the assumption that the suffix -ng- occurs in Upper Dnieper hydronymy highly conjectural for now.

In the hydronymy of the Upper Dnieper region, certain types of Baltic word formation with the formant -g- are undoubtedly reflected in names containing the element -ga. Precisely determining the type of suffix in such cases is often difficult, although the most likely suffixal variants are -aga/-ega/ūga; cf. Vedoga/Veduga, Vedzga — Lithuanian Vedega, Vadaga; Rydga, Rydoga — Baltic *Rūdūga/*Rūdaga, *Raudaga, and others.

207

River names ending in -da are characteristic of the southwestern outskirts of the ancient Baltic area. Their narrowly regional character, as well as their belonging to the Yatvingian hydronymic type, is apparently beyond doubt. The phonetic form of the element -da itself, presupposing an older Slavic -ъda from Baltic *uda ‘water’, is a useful criterion for establishing a name’s dialect affiliation, since this formant is entirely unknown in hydronymy in East Baltic and Prussian dialects, and the appellative meaning ‘water’ itself has a different form. Water names with the element -da adjoin the Upper Dnieper basin directly from the west, occurring mostly in the upper reaches of the Neman and along the Western Bug; cf. Golda, Grivda, Nevda, Sichda, Sokol’da (Sukogda), and others. Names in -da extend into the Dnieper region in only one area, directly continuing the aforementioned range of hydronyms in -da; cf. Yaselda (variants Yaselda, Yasel’da, Yasolda, Yasolna, Yasnolda). Apart from this hydronym, there are apparently no reliable examples of the river formant -da continuing into the Upper Dnieper region. The main types of Baltic hydronymy are not exhausted by models formed through the use of various formants. Alongside this word-formation method, Baltic hydronymy also employs a highly characteristic way of forming water names through compounding—a type extremely rare in ancient Indo-European hydronymy and one that points instead toward the Finno-Ugric sphere, where this method is fundamental. This is the favored type of compounding in Baltic hydronymy, with a second element denoting a river or lake. Several hundred formations of this kind with the second member -upė/-upis/-upys have been noted in the East Baltic area, where this type is highly productive; cf. Lithuanian Alguvys, Alkupis, Bartupė, Daugupys, Ėglupys, Gelžupis, Juodupė, Laukupė, Šešupė, Šilupė, Vandupė, Vilkupis, and others. Examples of this type are no less numerous in Latvian hydronymy: Aunupe, Āžupe, Baluožupe, Caujupe, Ežupe, Gaiļupe, Guovupe, Jerupīte; Kazupe, Lāčupe, Raudupis, Taurupīte, Ūdrupe, Vēžupe, Zivupe, and others. This type is also represented by a whole series of examples in Prussian hydronymy; cf. Berukapy, Caymenape, Caypeaps, Kirsappen, Kirsnappe, Laukappe, Nycape, Perapien, Rinappe, Sarape, Serappin, Serenappe, Swereppe, Wangaria, Wogenapp, Vogoeaps, and others (this type is reflected in German onomastics of regional Prussian origin; cf. Glasenapp). As for the Upper Dnieper territory, traces of this type may include Neroplya, whose second member has previously been taken to reflect the Baltic river name *Ner-up-/*Ner-ap-; cf. Lithuanian Nar-upė. The river name Natoma, in the Sozh basin, which has been disregarded in this connection until now, apparently belongs here as well. It is worth noting that traces of Baltic formations of this type in Russian territories occur primarily in immediate proximity to ancient Finno-Ugric territories, i.e., on the eastern and northern periphery of the distribution of Baltic tribes; cf. beyond the Upper Dnieper region, but in its immediate vicinity—

208

names related to it include Жукона (Upper Volga), Ужена (Oka basin), Каспля (Western Dvina), Кудень (a tributary of the Velikaya), cf. also the village of Горспле, etc. The situation is more complex with traces of names whose second element denotes a lake; this is apparently related in part to the formal similarity between the Baltic and Slavic names for this word. There is similar uncertainty regarding the less common type of predominantly toponymic names with a second element denoting ‘end’, ‘edge’; cf. formations in -gala, galas in the East Baltic area: Lithuanian Ariogala, Maišiogala, Latvian Latgale, Zemgale, etc. It is rather difficult to speak of reliable examples of this kind in the Upper Dnieper region. However, one may note the use of the second element of these compounds as an independent name for marshes in Belarus, cf. Галы, as well as the enigmatic river name in the upper reaches of the Dnieper Лупоголова — variant Луп(n)а, recorded in the Book of the Great Drawing in the form Лупогалова, in which the issue may not simply be akanye. This is all the more so because at the far eastern (as, indeed, at the western) periphery of the Baltic area, names in -gala are quite clear; cf., for example, the Galindians mentioned by Ptolemy in Prussian territory and the Goljad’ of Russian chronicles on the Protva.

In addition to the aforementioned traces of Baltic hydronymy, reflected in the derivation and compounding of names of bodies of water in the Upper Dnieper region, it is appropriate to discuss various hybrid formations involving Baltic and Slavic elements in different proportions. If we consider hydronyms that are entirely of Baltic origin, without any traces of Slavic adaptation, on the one hand, and hydronyms that are entirely Slavic, devoid of any signs of Baltic influence, on the other, we find a whole scale of gradual transitions—sometimes difficult to discern—from Baltic to Slavic hydronymy. The gradual nature of these transitions and the variety of intermediate stages make the problem of hybrid formations, their classification, and, finally, the geographical distribution of particular hybrid types especially important. The study of such formations may reveal the main aspects of the process by which the Upper Dnieper territories were освоены. Naturally, we cannot claim here to account for all the facts; at the present stage, it is more important to draw attention to the question itself and to the need to resolve it. Nevertheless, some specific observations on the classification of hybrid types can already be summarized in general terms.

Thus, Baltic names showing no signs of Slavic influence—or even any possibility of reinterpretation according to Slavic patterns—can be seen in hydronyms such as Бедоса, Невежа, Водва, Болва, Волма, and the like, which in fact have exact counterparts in the hydronymy of Baltic territories and contain no features in their phonetic form that would conflict with the Slavic phonetic type and require substitution. Close to these examples are cases such as Лучеса, Водога /Ведъга, Титва, Обеста, etc., which also have an exact counterpart

14. Kalbotyros klausimai. IV

209

in Baltic hydronymy, but have already undergone Slavic phonetic adaptation (sometimes regular, sometimes optional). The next category consists of Baltic names in which Slavic reinterpretation of the derivational formant is possible, though with some difficulty; cf. Ремист-ов by analogy with Дед-ов. Alongside the possibility of a Slavic interpretation of the suffixal part, there are hydronymic examples in which a Baltic root is replaced by a Slavic root similar in form, as we have already encountered in the name Колпита. An example of the Slavic reinterpretation of both the root and suffix of a formerly Baltic name is the hydronym Волчица — variants Волчеса, Волчаса; cf. Lithuanian Vilkesa. The most numerous type of Balto-Slavic hybrid formations in hydronymy is represented by examples in which a Baltic root, usually already subject to Slavic phonetic adaptation, takes a clearly Slavic suffix, most often one without a counterpart in Baltic hydronymy: Лопанка, Вор-минка, Варлинка, Какланка, Лучесянка, Мяделица, Можайка, Мерейка, Стабенка, Пелик, Аболъна, Вупенка, Вержица, etc. Cases of the opposite type, in which a Slavic root is extended by a Baltic suffix, are, by contrast, rare; cf. Хмелита, Калита, etc. The last two categories of cases essentially provide compelling evidence that the Baltic layer in the Upper Dnieper region is older than the Slavic one; at the same time, contacts between Baltic and Slavic dialects were so close that certain Baltic formants may have become common or even productive with Slavic roots. Such facts should caution researchers against depicting the process of retreat as too straightforward. In this regard, names such as Хохлитка, Деснокитка are particularly revealing: a Slavic root is accompanied by a Baltic suffix, which in turn is extended by a Slavic suffix. As a result, such examples appear Slavic. The Slavic element plays an even greater part in examples where both root and suffix are interpreted as Slavic and no obviously Baltic elements are present at all; cf. the name Березина, in which traces of a Baltic element are revealed only through special analysis. It is even more difficult to identify traces of an old Baltic name in cases of calquing. This is possible only when certain specific indications are available, particularly when differences in the popularity of a given name type are taken into account. Such examples practically merge with hydronymy that is Slavic proper.

An overview of Baltic derivational types in the hydronymy of the Upper Dnieper region, together with an analysis of the diverse types of Balto-Slavic hybrid derivation, combined with the fact that Baltic names are borne even by small rivers and a number of populated places (i.e., objects whose names change comparatively easily), provides grounds for a conclusion that is highly important from our point of view. Its essence is that the Baltic ethnic ele-

210

ment in the Upper Dnieper region did not retreat to the northwest on any significant scale as the East Slavs advanced northward. Apparently, Slavic expansion proceeded as a natural, gradual penetration, with assimilation of the Baltic element at different times in different parts of this territory. We must assume that even when the Slavs had advanced far to the north, beyond the limits of the Upper Dnieper region, numerous enclaves of Baltic population still remained in our territory; through a stage of bilingualism, they adopted Russian speech, while in turn leaving traces in the linguistic and ethnographic features of the local population. These traces make it possible to conclude that certain independent Baltic dialects, different from those known to us, ceased to exist in the territories that had been settled. The latter, it seems, persisted for quite a long time, at least in remnants, down to the beginning of the 2nd millennium AD.

The question of dialectal differentiation in the Baltic hydronymy of the Upper Dnieper region has, until now, scarcely been raised. This had to be accepted, since the total number of Baltic names in this territory was too small and, moreover, they were distributed over a relatively limited area. Now that several hundred such names are known, spread across a very extensive territory, there can be no doubt that dialectal differences existed in the speech of the Baltic population of these areas. In view of these circumstances, it seems timely to draw attention now to the importance of this problem and to attempt to present an overall picture, as far as the current state of research on Upper Dnieper hydronymy permits. Unfortunately, a number of circumstances impede the solution of this problem, some of them apparently insurmountable. In particular, in a number of cases we are virtually unable to draw definite conclusions even with respect to the isolated examples that would seem to invite conclusions about dialect affiliation. Several river names may serve as an example: Vasmer saw in them different dialectal variants of the Baltic name for ‘river’: прё, ape. Close analysis convinces us that such an interpretation is difficult, and even partly impossible. First of all, the material on which Vasmer based his argument in this connection was incomplete. It may now be considered that this root occurred in a much larger number of names of very different types: Упа, Упинька, Уперта, Описа, Апелка, Вупенка, Вопь, Вопец, Вопка, Вопля, Вопрат, Непопля, Каспия, Горсиле, Жалобыка, Янчанка and others. Analysis of these names compels us, for a whole range of reasons, to reject any overly straightforward interpretation of the phonetic form of hydronyms. Without denying the possibility that these names do in fact contain two dialectal variants, we must take into account factors such as the clearly different dates at which the Slavs adopted some of the names listed above; compare, on the one hand, Упа, which implies the latest acquaintance of speakers of East Slavic dialects with this Baltic for-

211

m, on the other hand, Вопь or Кастля, which were borrowed so early that they had time to undergo a long course of specifically Slavic phonetic evolution. It is highly interesting that geographical and general historical considerations confirm what has been said above. For example, east of the Upper Dnieper region, in the Oka basin, where the Slavs are known to have appeared later than in the Dnieper region, names occur only with the root Уп-, whereas examples indicating an early date of borrowing are common in the Upper Dnieper region (the isolated Упинка may also be explained otherwise). Particularly interesting in this connection are cases in which variants of the same name reflect different dates of adoption by the later population of these areas; for example, Лучеса, borrowed at a time when Baltic -ke- must have been substituted by Slavic -че-, and the nearby Лукесы, which preserved the more ancient Baltic state intact (with regard to the change under discussion). Naturally, such borrowing of Baltic names at different times was carried out by East Slavic dialects at different stages of phonetic evolution. Hence the different principles of Slavic substitution of Baltic names, which also preclude unambiguous conclusions. Finally, we must take into account a very large number of examples in which discussion of regular Balto-Slavic phonetic correspondences is altogether impossible. We mean cases in which Baltic names underwent secondary reinterpretation according to Slavic folk etymology.

Another reason for the difficulty of identifying dialectal differentiation among Baltic names in the Upper Dnieper region is that we still lack materials on the Baltic hydronymy of Lithuania, Latvia, and former East Prussia that are sufficient in both quantity and reliability. As for the materials that are available, they generally remain unanalyzed; consequently, our knowledge of hydronymic dialect isoglosses in modern Baltic territories is so limited that it scarcely allows the corresponding comparison to be made with the Upper Dnieper material. While this obstacle could in principle be overcome through appropriate efforts by specialists in Baltic hydronymy, there are other obstacles that will scarcely be avoidable. The point is that, although our knowledge of the hydronymy and toponymy of the Old Prussian territory is relatively satisfactory, we know very little about the Old Prussian appellative lexicon. In this respect, it is unlikely that any substantial improvement can be expected. Our far from complete knowledge of Old Prussian vocabulary seriously hinders the resolution of the question of dialectal differentiation, since whenever we fail to find in Old Prussian a word corresponding to a given East Baltic word, we are generally unable to determine whether this is due to dialectal divergence or simply to the scarcity of data on Old Prussian vocabulary. Needless to say, what has been said applies to an even greater extent to other extinct Old Baltic

212

dialects from which nothing has survived, or in which only isolated words have been preserved. This gap is particularly regrettable, since it may be assumed that some of these dialects (Yotvingian, Goladian) existed in the territory under study and may have left traces in certain hydronyms. Naturally, under these circumstances we can speak only of the Baltic type of a given hydronym (and even that not always), but not of its specific dialect affiliation.

Finally, even if we move from Baltic dialectal distinctions in hydronymy to the dialect features that distinguish Baltic dialects as a whole, it will become clear that these features will not always allow us to differentiate the Upper Dnieper hydronymy dialectally. This is explained, first, by the fact that Baltic names in the Upper Dnieper region were adopted by the East Slavs at a time when some of the dialectal distinctions historically known in the Baltic languages had not yet developed; second, in a number of cases, different Baltic dialectal variants could be reflected in the same way in Slavic; and, finally, third, some East Slavic dialect phenomena are formally analogous to corresponding Baltic ones, making it difficult to distinguish between them. As for the first category of cases, it should be borne in mind that distinctions relevant to modern dialectology—between sibilants and hushing sounds (ž, š, — z, s), between k and s before a front vowel, between the diphthong ei and the vowel ie, and between retention of a nasal consonant before another consonant and its loss in the same position—could hardly have been reflected (except in isolated cases) in the hydronymy of the Upper Dnieper region, since they arose later. In particular, the sound -и- in the name Ситовка may equally reflect either West Baltic ei (cf. Old Prussian Seyte) or East Baltic ie (cf. Lithuanian Sietuvà, ‘river’, Latvian siets ‘a deep place in a river’). Likewise, it would be unjustified to regard the hydronym Залятка as Latvian in type on the basis of its initial З- (rather than Ж-) (cf. Latvian zāle ‘grass’ alongside Lithuanian žolė). Similarly, we would not venture to insist on a specifically Lithuanian origin for the name Биржевка, since -ж- could reflect the corresponding sound not only in the Lithuanian word beržas, but also the same sound in other Baltic dialects, where it later developed into z; cf. Latvian berzs ‘birch’. The distinction between k and s before a front vowel, significant for modern Baltic dialects, is not a reliable criterion in analyzing Upper Dnieper hydronymy, since, for example, К- in Кидель may reflect either a Lithuanian dialect variant or Old Latvian (before the shift to s); on the other hand, Цидель in the Oka basin does not necessarily reflect a Latvian-type form (cf. Latvian Cidulupe), but may instead be the result of a later change from К- in the local language. Likewise, a name of the type Венча (from *Ventia) does not provide sufficient grounds to assign it to either the Lithuanian or the Prussian type; it may equally reflect an earlier state of those dialects,

213

in which simplification of a tautosyllabic cluster later took place, leading to the loss of the nasal consonant. Despite the numerous difficulties involved in establishing the dialectal differentiation of Baltic hydronymy in the Dnieper basin, it may be useful to draw attention to certain local features of these names, as well as to analyze some hydronyms represented by two or more phonetically similar variants. In the field of word formation, it seems advisable to us to identify not only those formatives whose dialectal affiliation can be determined without much difficulty through corresponding examples from living Baltic languages, but also individual formations of a local character, even if we lack definite information about their dialectal affiliation. A number of hydronyms ending in -да are characterized by a relatively narrow distribution in the far southwestern part of the area under consideration. Formations of this type are not found elsewhere in the Upper Dnieper region, but they do occur in the Western Bug basin. This circumstance, together with etymological considerations, makes the traditional view that names with this suffix are Yotvingian quite plausible. Perhaps there is no other Baltic name area in our territory so clearly delimited whose dialectal affiliation is beyond doubt. The distribution of hydronyms with the suffixes -ия and -ея in the Upper Dnieper region is interesting. The former tend toward the southern part of the Upper Dnieper region, concentrated mainly along the Pripyat, near the area of names ending in -да, and generally do not extend north of 53° north latitude. By contrast, river names with the suffix -ея are generally concentrated precisely north of the 53rd parallel. It is noteworthy that this mutually exclusive (though not entirely consistent) distribution of the Baltic formatives -ija и -ēja contrasts with the situation in Lithuania and former East Prussia, where they coexist; a different situation is observed in Latvia, where names ending in -ija apparently do not occur. One of the clearest dialectal hydronymic isoglosses in the modern Baltic area, -uva/-ava, unfortunately cannot be traced with certainty in the Upper Dnieper region (the same, incidentally, applies to the possibility of distinguishing Baltic hydronyms in -uva/-va because their phonetic development was identical in Slavic). However, we know, first, the approximate distribution of names ending in -ва in the Upper Dnieper region (a belt extending southwestward from the upper reaches of the Dnieper as far as the right bank of the Pripyat) and, second, that names ending in -ва (from -uva) are more likely related to corresponding Old Prussian and Lithuanian formations than to Latvian ones, where the suffix -uva in hydronymy is usually replaced by -ava. Hydronyms ending in -асал-есал-оса occupy roughly the same territory as names ending in -ва and are found in the modern Baltic area predominantly in Lithuania. As for other hydronymic formatives of Baltic origin, their material makes it difficult to judge the dialectal

214

of the Upper Dnieper region: in some cases, the number of available examples is too small to draw any definite conclusions; in other cases, the question cannot be resolved because hydronymic isoglosses have not been identified in the present-day Baltic territory. The same difficulties arise in attempting a dialectal differentiation of individual hydronyms of prepositional type. The names in question are, first of all, names such as Устарь, cf. Тара, Усперт and perhaps Пипин. Although reliable results in assigning these names to dialects can hardly be expected, it is nevertheless noteworthy that Ус- in the first two examples more likely reflects the East Baltic type of prepositional formations known in toponymy; by contrast, names with the corresponding preposition are absent from Old Prussian toponymy. As for the name Пипин, if our proposed etymology (from *Apei-pin-) is correct, it too may be thought to be connected more likely with East Baltic examples, particularly perhaps Lithuanian ones, than with Old Prussian examples.

The difficulties of establishing the dialectal differentiation of Upper Dnieper hydronyms by means of phonetic criteria have been discussed above, so it is difficult to say anything definite in this regard. Nevertheless, when some other data are taken into account, the rather varied examples of phonetic variation in the name of a single river, or in the names of different rivers derived from the same root, may prove useful in certain cases. Naturally, one should bear in mind the possibility of change already on Slavic soil, as well as the possibility that a given Baltic hydronym was adopted at different times. Recently, Arumaa suggested that the name of a river in the Sozh basin, Лыза, was of Golyad origin, on the basis of -з- instead of -ж- (cf. Latgalian Lyūža). It is possible that in this case we really are dealing with a particular Baltic dialect form, although not necessarily a Golyad one. In connection with the name Лыза, it is worth considering a whole series of examples in which -з- alternates with -ж-, cf. Азарза — вар. Заржа, Жарка; Жубр — вар. Зубир, Зуберь, Зубер, тут же поблизости Зубар, Зуберь; Жавуница — вар. Завуница; it turns out that all the names listed are concentrated along the Sozh, which makes these variants perhaps more indicative. Characteristically, outside this area we are dealing not so much with different variants of the name of one and the same river as with different hydronyms reflecting the alternation ж: з, cf. Залазенка, Залазинка, Залазка, — Жалижа, Жалож, Жиложка; Жагулинка — Загуленка. К такому же типу вариантов относится Шморговка — Смороговка. Венца — Вяча, Морчанка — Морочь, Морожа — Моржовка, Лучеса — Лучесы, Гейшинка — Гижинка, Мильча — Молц, Аткильня — Чолна, Чолка, Сеньжа — Усажа, and a whole range of others. The variation in these hydronyms is, of course, primarily connected with differences in the chronology of their adoption. The significance of phonetic criteria in determining the dialectal affiliation of Upper Dnieper hydronyms may perhaps increase somewhat in future research in connection with the establishment of lexical correspondences between these names and the appellatives associated with them in the Baltic languages.

215

Having established lexical correspondences between these names and their associated appellatives in the Baltic languages. Bearing in mind the limitations noted above in determining the dialectal affiliation of river names on the basis of vocabulary, we nevertheless have to acknowledge that, at present, lexical material may be the most convenient for this purpose. Naturally, the dialectal affiliation of hydronyms can be established most reliably where they have parallels in Old Prussian and none in Lithuanian or Latvian. One such case is represented, in particular, by the names Стабна, Стабня, Стабенка, Стабница, which can reasonably be regarded as hydronyms of the Prussian type; cf. Old Prussian Stabingen, Stabynotilte, Stabuniten, Stabayen, Stabelauken, all from stabis ‘stone’, whereas Lithuanian and Latvian use a different root for ‘stone’. The name Поня apparently reflects the Old Prussian type; cf. Old Prussian Panyen from Old Prussian pannean ‘swamp’, whereas the East Baltic languages have other words for ‘swamp’. Comparing these reliable lexical facts about the dialectal affiliation of the Baltic hydronymy of the Dnieper region with the corresponding facts concerning word formation and phonetics presented above leads us to a very interesting conclusion. It turns out that the combination of the identified phonetic, word-formational, and lexical features, in cases where all of them can be assigned to approximately the same location, corresponds to none of the known Baltic dialects. In other words, we encounter cases in which features known from different Baltic languages occur together. For example, if we take the geographically restricted area around the upper reaches of the Dnieper and along the Sozh, it is easy to see that alongside names clearly of the Prussian type, such as Стабна, there are word-formational formatives of the East Baltic type, such as -esal-asa, and, on the other hand, phonetic features attributed to Golyad dialects (the alternation ж: з). This is further evidence that it would be naive to identify the vanished Baltic dialects of the Dnieper region with any particular Baltic language; therefore, the existence of a distinct group of Baltic dialects in the Upper Dnieper region must be taken into account. Unfortunately, the greatest difficulty lies precisely in identifying the lexical composition of these unknown dialects, since we have nothing concretely attested to compare them with, as we do in the case of Lithuanian, Latvian, and Old Prussian. Only in exceptional cases can individual lexemes without exact equivalents in the known Baltic languages be tentatively reconstructed. The name Откильня/Аткильня is probably such a case (cf. also Чолна, Чолкна). In this word we see an old designation for a spring or source, expressed by means of the prefix at- and the weak grade of the root *kel-/*kil- ‘to rise, originate’ (with regard to the semantics, cf. Russian род-

216

to give birth). In the known Baltic languages, however, a spring or source is designated in quite different ways. Among other cases that make it possible to speak of dialectal differentiation on the basis of vocabulary, one can single out a group of examples with separate appellative correspondences in the known Baltic languages; cf. Geishinka alongside Old Prussian geeuse ‘heron’, Dokolka, cf. Old Prussian Dakow alongside d lacke ‘starling’, and others. Examples with separate hydronymic connections also belong here; cf. Abesta alongside Lithuanian Abista, Remistovo alongside Old Prussian Ramestow, Orizhnya alongside Old Prussian Aryngine, Nyuchka alongside Old Prussian Inacus, Kubar alongside Lithuanian Kaubri, and others. There are quite a number of examples of this kind with separate correspondences in the vocabulary, providing grounds for identifying a whole series of particular lexical isoglosses linking the Upper Dnieper region with individual areas of the Baltic region. In addition to what has been said, it would be of interest to determine the dialect affiliation of those hydronyms that recur repeatedly in our territory and are often linked by a chain of complex transitions; cf., for example, Rekta, Rekhta, Resta, Roshty and other variants. In this etymologically difficult case, the question of dialect affiliation is decided depending on which etymological interpretation is adopted. If it is connected with the name for ‘nut’, its dialect affiliation would have to be explained as East Baltic (cf. Lithuanian riešutas, Latvian rieksts alongside Old Prussian reisis). If, however, it is thought to be connected with the Baltic word for ‘swamp’, then only the variant Resta can definitely be said to be of East Baltic origin (cf. Lithuanian raistas ‘swamp’ and the numerous hydronyms derived from this root in Lithuania).

Although the hydronymic material presented alone is quite sufficient to leave no doubt about the presence of a Baltic population in the Upper Dnieper region, there is also another source attesting to the same fact: numerous traces of the Balts’ presence here, preserved in Russian dialect vocabulary. Taken together, all this compels us to take a fresh look at the ethnic and linguistic history of the Upper Dnieper region.