Ономастика Куршей

LIETUVIŲ KALBOS SANDAROS TYRINĖJIMAI LIETUVIŲ KALBOTYROS KLAUSIMAI, XXVI (1987) ОЯР БУШ THE ONOMASTICS OF THE CURONIANS One of the most pressing questions in Baltic studies today concerns the place of the Curonian language among the other Baltic languages. In seeking an answer, it is necessary not only to identify and bring into scholarly circulation new Curonian-language material, but also to review and verify material that has already been repeatedly subjected to linguistic analysis, and, where necessary, to reinterpret it. In this connection, particular attention should naturally be paid to V. Kiparsky’s monograph „Die Kurenfrage“ (KF). As is well known, this book remains to this day the only monographic study of the Curonian language. However, not all of V. Kiparsky’s conclusions can be accepted without reservation: for this renowned linguist, the principal fields of activity were Finno-Ugric studies (and, of course, Slavic studies), rather than Baltic studies. Moreover, nearly half a century has passed since this study was published, and during that time much new material (in particular, toponymic material) has been recorded, systematized, and studied2. V. Kiparsky’s monograph analyzes chiefly onomastic material recorded in ancient documents. This article examines four toponyms from the oldest legal document associated with Courland, namely, the 1230 treaty between the Curonian king Lamekin and papal vice-legate Baldwin of Alna. The treaty lists 17 toponyms—the names of lands (terrae) and so-called kiligunds, relatively small territorial units. Among these names are some that are undoubtedly Baltic (or, at any rate, Indo-European), for example, in Latinized spelling, Ardus, Durbis, Riwa, Saccze; some are clearly Finno-Ugric (Normis, Kiemala); and for others, their origin cannot be stated categorically (Aliswangis, now Alsunga, Edualia – Ēduole, Sarnitus – Sārnate, etc.). Perhaps the greatest number of hypotheses to date have concerned the origin of the name Alsunga, which in Lamekin’s treaty appears as Aliswanges or Alis- 1 See the list of abbreviations. 2 For an analysis of some specific shortcomings of V. Kiparsky’s monograph (while recognizing its overall positive assessment), see the reviews by J. Endzelīns [and E. Blese: Endzelins J. [Rev.] Die Kurenfrage von V. Kiparsky. – In: Darbu izlase. III sēj. 2 d. Rīga, 1980, 535.—565. lpp.; Blese E. [Rev.] V. Kiparsky. Die Kurenfrage. – Latvijas Vēstures institūta žurnāls, 1940, Nr. II, 295—308. lpp. 96 wangis, in other documents Alswanghen 1253, 1341, 1355—62, Alswangen 1338, Aliswangen 1434, Alszwangen 1503, etc., in the German interpretation Alschwangen /KF 204—205/. Finno-Ugric linguists (V. Kiparsky, L. Kettunen, I. Sjögren) tend to consider a Finno-Ugric origin for this toponym the most likely (KF 205; Kett. VII). V. Kiparsky’s proposed comparison with Livonian al’l’i ‘greenish’, ōl’az ‘green’ + Livonian (in Vidzeme) vanga ‘flood meadow’ is not without a certain degree of plausibility. Nevertheless, all Balticists—from A. Bielenstein onward—are convinced of the Baltic origin of the toponym Alsunga. In his monograph of 1892 (now perhaps undeservedly forgotten), A. Bielenstein had already compared Alsunga with the Old Prussian toponym Alexwdnga, with other Old Prussian toponyms containing the element vang-, and also with the Latvian appellative alksnis ‘alder’³. In fact, J. Endzelīns, who first pointed out this connection (with Lithuanian dialectal aliksnis ‘alder’ and with vanga) in the footnotes to a 1912 article⁴, also regarded essentially this same etymology as the most likely; 40 years later, he substantiated it in considerable detail in the toponym dictionary of the Latvian SSR /Lvv I 23/. All Balticists apparently agree that the second part of the toponym under consideration should be traced back to Baltic vanga; only the origin of the initial element remains disputed. If it is compared with Latvian alksnis (and its equivalents), the loss of the sound k- must be explained. J. Endzelīns initially explained this loss by dissimilation⁵, and later, in the toponym dictionary, reconstructed two alternative Curonian names for alder—*aliskā or *alisā, i.e., without -k-, cf. Russian ольха. However, alongside this possibility, another supposition also seems quite likely: the initial element of this toponym was originally als- (as K. Būga believed⁶); the spelling with -i- (Alis-) in some documents may be inauthentic, especially since other documents of no less antiquity more often have Als-. The toponym dictionary of the Latvian SSR lists 7 toponyms with the root als- (alsava, alsi, alsupite, etc.) and 5 with the root alš- (alši, alšu-valks, etc.) /Lvv I 22—23/. It is noteworthy that almost all of them (with 1–2 exceptions) are located either in Courland (more often) or in Selonia. The root als- has no direct equivalent in Latvian appellative vocabulary, whereas such equivalents occur in Lithuanian and Prussian onomastics; in Prussia, there is the toponym (name of a hill) Alsitten and several anthroponyms /see Lvv I 23/; in Lithuania, a whole series of hydronyms and other toponyms beginning with Als-/Lhež 40/. It should be noted that of the Lithuanian hydronyms mentioned— ³ Bielenstein A. The Boundaries of the Latvian Ethnic Group and the Latvian Language in the Present Day and in the 13th Century. — St. Petersburg, 1892, p. 284. ⁴ Endzelins J. On the Nationality and Language of the Curonians. — In: Darbu izlase. II. R., 1974, p. 447. ⁵ Endzelins J. /Review of/ Die Kurenfrage.., p. 564. ⁶ Būga K. Lietuvių kalbos žodynas. I. — In: Būga K. Rinktiniai raštai. V., vol. 3, 1961, p. 420. 4. 1835 97 Of the Lithuanian hydronyms mentioned, 5 are located in western Lithuania, 5 in eastern Lithuania and Belarus, and only 1 in central Lithuania. Broadly speaking, this range seems to continue the range of Latvian toponyms in Als-, allowing the cautious supposition that Baltic toponyms in Als- may be genetically related to the peripheral Baltic languages. V. N. Toporov compares such Baltic toponyms primarily with Lithuanian alsà, alsúoti, but also points to Belarusian альса, алёс, альсе, альсавина ‘swamp, boggy place’ and to the hydronym Ольса7 in the Dnieper basin. These Slavic parallels are indeed highly “suspicious,” but their own etymological status is not entirely clear. It is quite likely that they themselves are Balticisms. Yu. Laučiūtė derives the Belarusian words mentioned (as well as their Russian, Polish, and Ukrainian equivalents) from Latvian (dialectal) aluots ‘spring’8. This comparison accords well with G. Krahe’s view that Lithuanian hydronyms in Als- derive from the Indo-European hydronymic root al-, expanded by the formant -s- (as does K. Kuzavinis) /see Lhež 40/. However, deriving the Slavic words mentioned directly from Latvian aluots is unlikely (although a more distant genetic relationship between these words is possible). As for G. Krahe’s hypothesis, A. Vanagas in turn doubts the reality of the identified formant and considers comparison with Lithuanian alsà, alsùs the most likely, explaining this relationship semantically in terms of slow water flow (Lhež 40). This digression was necessary here not in order to settle definitively the question of the origin of these toponyms themselves, but only to show that the initial element of the Courland toponym Alsunga (most likely from *Als-vanga) should, perhaps, nevertheless be considered alongside the other toponyms mentioned that begin with Als-. As for the further etymology of this element and its semantic motivation, might the verb alsuoti have provided the basis for a name meaning a quaking, seemingly “breathing” bog? (Alsunga could then be interpreted etymologically as ‘boggy vanga’.) V. Kiparsky also considers Ēduole, Edualia in the Lamekin treaty, German Edwahlen, a Finno-Ugricism, deriving it from Livonian ed’d’i ‘place in front (of something)’ + val’ ‘embankment, enclosure’ (KF 209). J. Endzelīns notes that this etymology is unlikely because of the long ē in the Latvian (Curonian) toponym⁹; nevertheless, he too considers the toponym to be of Finno-Ugric origin, comparing it with the Estonian name of the village Heedli and the appellative vald, genitive valla ‘parish’. On the basis of the German Edwahlen, J. Endzelīns supposes that -io- (in Latvian Ēduole) < corresponds to -va- /Lvv I 281/. However, the letter sequence -ia- in the first recorded form Edualia could already have denoted the diphthong -io-, in which case later attestations such as Edevalen (1253, 1338) reflect a secondary pronunciation, altered in the speech of foreigners (and perhaps also reinterpreted— 7 V. N. Toporov. The Prussian Language. — Moscow, 1975, vol. 1. A–D, pp. 78–79. 8 Лаучюте Ю. А. Словарь балтизмов в славянских языках. – Л., 1982, с. 9. 9 Endzelīns J. /Review of/ Die Kurenfrage..., p. 563. 98 thought?). If -io- is original (which is quite possible, although, of course, not indisputable), then the toponym Ēduole contains the ancient Baltic suffix -uol- and the root ēd- (cf. Latvian ēst, Lithuanian ēsti). The semantics of this root fully accord with the patterns of hydronym formation; Baltic hydronyms with the root grauz- /grauž- may be cited as a parallel [see Lvv I 315; Lhež 121/. Furthermore, Courland also has a river (a tributary of the Venta) called Ēda; V. Kiparsky likewise derives its name from Baltic ēd- (although it should be noted that a Finnic origin for this hydronym has also been hypothesized10). A. Vanagas connects Lithuanian ēsti ’to gnaw, graužti’ with the name of a part of Lake Ēdėlė in the Ignalina district (though not without some reservations): („apparently, Ēdėlė – tai išgraužta, vandens išplauta vieta or something similar“ [Lhež 99/). What is striking is the phonetic coincidence of this hydronym with the Livonian variant of the name Edole, ādel /see KF 209/. Of course, this coincidence may be entirely accidental; nevertheless, these parallels between Ēduole (and Ēda) in Courland and Ēdėlė in Lithuania are most likely also etymological; in that case, Ēduole may be considered a toponym of Baltic origin. The toponym Sārnate (in the Lamechin treaty, Sarnitus) is still generally considered a Finnic loan, which can be compared either with Finnish saarni ‘ash tree’ (although, as V. Kiparsky already noted, the corresponding Livonian form sārna has a different vocalism), or with reconstructed Finnic *sārin̄aiset ‘islanders’ [KF 234]. In the manuscript section of his dictionary of toponyms of the Latvian SSR (held at the Institute of Language and Literature of the Latvian SSR Academy of Sciences), J. Endzelins writes only: „word of Livonian origin“ (though he marks this observation with a question mark). The Finnic interpretation of this toponym is certainly quite plausible; nevertheless, another interpretation of the word is also possible, based on Baltic and Indo-European linguistic material. The toponym Sārnate has a remarkably exact Illyrian counterpart—the ethnonym Σ α ρ ν ο ᾶ τ ε ς (already noted by K. Kasparsons in 193811; this and some other Balkan-Illyrian parallels can be found in the works of H. Krahe12). However, regardless of whether the coincidences mentioned are accidental or not, the toponym under consideration may be compared with Latvian sārti, sārni ‘mud, filth’. The meaning, as one can see, is not the most appealing, but it is quite typical of ancient hydronyms (and Sārnate is not only an oikonym but also a hydronym). In Lithuania, the closest counterpart, both etymologically and semantically, is the name of the river Sárva, as well as the name of the lake Sėrvas, which A. Vanagas compares— 10 См. Rudzīte M. Somugriskie hidronīmi Latvijas PSR teritorijā. — In: Latviešu leksikas attīstība. R., 1968, 180. lpp. 11 Kasparsons K. Parallēles. — In.: Filologu Biedrības Raksti. XVIII. Rīga, 1938, 55. lpp. 12 См., напр., Krahe H. Die alten balkanillyrischen Namen. — Heidelberg, 1925, S. 34, 35, 97. 4° 99 connects with Lithuanian sarvalai, which is etymologically related to Latvian sārņi /Lhež 291/. Latvia also has a number of toponyms whose etymological connection with the word sārņi is beyond doubt, e.g. Sārnes, a farmstead in Birzgale; Sārņi, a village in Zvirgzdene; and others. A Baltic origin for the toponym Sārnate is also supported by the formant -at- (if it is not -<ast-, cf. Sārnastvalks, a stream in Dundaga). Overall, it seems that the hypotheses of Finnic and Baltic origins for the toponym in question are approximately equally probable. One of the most enigmatic toponyms mentioned in the 1230 treaty is Pygawas. The exact location of this kiligund is currently unknown, and no equivalent in modern toponymy is known either. It has been suggested that this is a mistaken record in place of the correct Tygwe (=Tīgve), which is mentioned in a document of 125313. However, a toponym Pigow is also mentioned in one of the documents of 1461, and in 1858 a farmstead Piegenēek (i.e. Pigeniek) was recorded in Dzīre (see KF 137). V. Kiparsky compares Pygawas with the Lithuanian toponyms Pigėnai and Pigainė (from I. Sprogis’s publication), and notes that the spellings in -awas, -ow reliably („sicher“) attest to the presence of the Baltic suffix -ava (ibid.). Among modern Lithuanian toponyms, one may cite the name of the village Pigailiai in the Raseiniai district (Lasž 241). We have no direct evidence that the name of the kiligund *Pigava or *Pigava also functioned as a hydronym, but this is unlikely, especially given the presence of the typically hydronymic formant -ava. A possible hydronym *Pigava could perhaps be compared with Lithuanian pigùs, pỹgis. Semantically, this comparison is supported by one of the meanings of its Latin counterpart piger-, ‘slow’, and even more so by the possible analogy with the semantic relationship between Lithuanian lėtas ‘slow’ and Latvian lēts ‘cheap’ (E. Fraenkel had already noted the semantic parallels between pigùs and lėtas /LEW 588/). Thus, the original meaning of the toponym *Pigava (apparently Baltic) may have been ‘slow river’. To conclude, a few more general observations concerning V. Kiparsky’s monograph „Die Kurenfrage“. The specific analysis of linguistic material in this study is meticulous and well considered; there is not very often occasion to disagree with the author. Nevertheless, the monograph’s general conclusions (about the character and location of the Curonian language) seem unconvincing. Why? The reasons apparently relate primarily to the criteria used to select material for the monograph. First, this material is drawn mainly from documents, and these documents date from the 13th–16th centuries, that is, from a period when the Old Curonian language had probably already ceased to be used, while by the 16th century its successor—the New Curonian language—had also disappeared. Therefore, linguistic facts from this period cannot show precisely what the language of the Curonians was like at the time 13 См. Bielenstein A. Die Grenzen..., S. 209. 100 of its flourishing. Secondly, in selecting the linguistic material, the administrative boundaries of Courland were taken into account, and these often did not coincide with ethnic boundaries. Therefore, some of the Baltic linguistic material analyzed in the monograph is genetically related not to the Curonian language, but to other Baltic languages. The two circumstances just mentioned (along with several other, less significant reasons) meant that the “Curonian question” posed in the monograph did not receive a sufficiently convincing answer. The search for this answer remains a task for the present generation of Balticists. ABBREVIATIONS Belarusian — Belarusian language Old Prussian — Old Prussian language Liv. — Livonian language Lith. — Lithuanian language Latv. — Latvian language Ger. — German language. Kett. — Kettunen L. Livonian Dictionary with Grammatical Introduction. — Helsinki, 1938, LXXI + 648 pp. KF — Kiparsky V. The Curonian Question. Suomalaisen Tiedeakatemian toimituksia. B XLII. — Helsinki, 1939. — 474 pp. Lašž — Directory of the Administrative-Territorial Division of the Lithuanian SSR. Part II. — V., 1976. — 398 pp. LEW — Fraenkel E. Lithuanian Etymological Dictionary. — Heidelberg—Göttingen, 1955—1965. Lhež — Vanagas A. Lietuvių hidronimų etimologinis žodynas. — V., 1981. — 408 p. Lvv I — Endzelīns J. Place Names of the Latvian SSR. Part I, Volume 1. A—J. — Rīga, 1956. — 426 pp.