Acta Linguistica Lithuanica

LV (2006), 89 – 106

The Mediating Role of High Latvian in the Old Russian and Lithuanian Loanwords in Latvian¹

ILJA A. SERŽANT

Lietuvių kalbos institutas

In this article it is argued that most Latvian borrowings from Old Russian must have entered the language through Proto-High-Latvian, whence they were borrowed into Proto-Middle-Latvian. This would explain their phonetic shape, particularly Middle Latvian uo as the reflex of Old Russ. ѡ [u:] in borrowings like suoma (OR. съма), muoka (OR. мъка), presumably through Proto-High-Latvian *sūma, *mūka with subsequent substitution of [uo] for [u:] according to the pattern of sound correspondences between the dialects. This assumption would also account for many other cases where Middle Latvian forms diverge from those of Old Russian (and Lithuanian), cf. Middle Latvian baznīca with broken tone on a suffix usually bearing sustained tone (-nīca). This broken tone cannot be due to the Old Russian pronunciation and must reflect the High Latvian glottalization of internal syllables containing original ī or ū.

1. LOANWORDS FROM SLAVIC

1.0.1 It is well known that the modern standard Latvian language is based on the Central Latvian dialect. The distribution area of this dialect can be described very roughly as “in the middle of Latvia,” with the middle

¹ I would like to express my sincere thanks here to Professor Sergej Temčūn (Vilnius, LKI) for discussing my work and for his many valuable comments. I would also like to thank Everita Andronova (Riga, St. Petersburg), who provided me with an electronic list of the loanwords from (Old) Russian recorded in ME, thereby greatly facilitating the search for such items.

89

Ilja Seržant

lying along the west-to-east axis. Modern Latvian has a number of loanwords of various ages from Russian and Belarusian dialects. Since, then, the Central Latvian dialect had no points of contact with East Slavic dialects in historical times, viewed purely geographically (and has none to this day), loanwords from East Slavic dialects must have entered Central Latvian through the mediation of the High Latvian / Latgalian dialect. This obvious consideration allows us to explain the sound form of many loanwords from Old Russian and Lithuanian, which has so far gone unnoticed.

1.0.2 It is important in this context to emphasize that the correspondence East Slavic [u:] versus Latvian uo, discussed in § 1.1, has hitherto been interpreted as a reflex of the Proto-Slavic pronunciation supposedly preserved by Latvian loanwords (e.g. Kiparsky 1952: 74; idem 1963: 80–81; Koškin 2006: 36). In what follows, however, it will be shown that this assumption is unfounded. The explanation of such an irregular correspondence is to be sought not on East Slavic, but—as with a number of other loanwords—on Latvian soil, specifically in the mediating role of the Proto-High Latvian dialect.

1.0.3 This paper considers exclusively Old Russian loanwords that probably entered the Latvian-speaking area around the period from the 10th to the 12th century. The lower limit (10th century) is set on the assumption that loanwords associated with the Christian mission could not have been borrowed before Christianity was officially proclaimed the state religion of Rus’. Note that there is a large number of terms associated with the mission, which can serve as a guide for dating the loanwords. The upper time limit (12th century), too, is only approximate. From the 13th century onward, the strengthening of the German crusading order in western Latvia changed the geopolitical situation, causing Old Russian influence to lose significance.

Finally, the linguistic characteristics of a loanword classified here as Old Russian are listed:

90

Die Vermittlungsrolle des Hochlettischen …

– the retention of reduced vowels in weak position and the rendering of strong reduced vowels with their original sound value, i.e. Old Russian ъ > Latvian u, ь > Latvian i; – the retention of vowel quantity, i.e. Proto-Slavic long vowels are also rendered as long in Latvian.

East Slavic borrowings that do not meet these criteria are regarded as later borrowings.

1.0.4 The oldest East Slavic borrowings into Proto-High Latvian

probably originate from the dialect of the Kriviči, spoken with some variation in old Novgorod, Pskov, and Smolensk. In addition to this Old Russian dialect’s immediate geographical proximity to Latgale, a number of other arguments support this. For some time after the arrival of the German crusading order, the eastern Latvian area was also within the sphere of influence of the Novgorod city-state. Thus, it is known that in the 12th century the city of Novgorod had ceded its tribute rights in Latgale to its daughter city Pskov, which was also confirmed by the order in a treaty around 1224 (Nazarova 2002: 592–593). The later designation Middle Latvian krievs, subsequently applied to all Russians, should also be mentioned; initially, it surely referred only to the Kriviči.

Finally, linguistic arguments should also be cited: some old borrowings from East Slavic show (where possible) features typical of the Novgorod-Pskov dialect, primarily cokanje (Semenova 1959: 599): cf. buca ‘barrel’ from Old Russian *ъbča (cf. modern Pskov dialect боча (POS₂ 136 s. v.)) (ME₁ 344)²; c’iešk’i adv. ‘frequently, often’ (Zvirgzdine, Pilda, Bērzgale, Baltinava (EH₁ 269)) from *cēst-ji ← adj. *cēsts (from Old Russian часть / часть, Srezn₃ 1475); rūcis (gen. sg.

² Here we have the Proto-Slavic cluster *-tj-. This cluster also exhibits cokanje; cf. хочѫ and хочю (for хотѣти in the Old Novgorod dialect, see Zaliznjak 1995: 677). Moreover, Latvian dialectal buca cannot possibly have been influenced by the synonymous muca, since the latter itself must be explained as resulting from the influence of the former (MEⅡ 657–658). Furthermore, S. Temčin (Vilnius) kindly draws my attention to Croatian boca ‘bottle’. I cannot explain the Croatian form, but the modern Pskov form boča indicates that the presumed Old Russian *ъbъца contains an instance of cokanje.

91

ILJA SERŽANT

rùča) from Old Russian порчами, порчии (SreznIII 199), tyuc’s from Old Russian торча, ceista from Old Russian чисто, svece from Old Russian свѣча, plecs from Old Russian плечѣ/ плече (Semenova 1959: 600; Reķēna 1962: 371, 399); High Latvian cinêt ‘to peel potatoes’ from Old Russian чинити (Laumane 1977: 74)³.

1.1. East Slavic -оγ- > Latvian -uo-

1.1.1 The root syllable

The usual correspondence for Old Russian оγ is Latvian uo, cf. bluōda < Old Russian бладо; šuòģis < Old Russian сѫдьи, (Old Church Slavonic бладо, сѫдии), šuòds (Old Russian сѫдъ, Old Church Slavonic сѫдъ). Since Latvian has a ū sound, one might ask why these loanwords are not pronounced **blūda, **sūdis / **sūģis. This can be explained by the mediation of the High Latvian dialect, as follows: the Old Russian word бладо must first have been borrowed into the High Latvian dialects as *blūda, as expected, and the Old Russian word сѫдъ as *sūds (also attested in Kalupe, cf. Reķēna 1962: 473), since the stressed long vowels in Old Russian (at least originally) are likewise rendered by a long vowel in the High Latvian dialects; cf. ruocs < *rāds < Old Russian радъ, slùops < *slāps < Old Russian слаѣть (Reķēna 1962: 26). As the now Proto-High Latvian lexeme subsequently migrated westward, for example *sūds, speakers of the Proto-Middle Latvian dialect substituted the corresponding sounds of their own dialect for the Proto-High Latvian sounds, just as they also do when borrowing inherited words. Cf. Hermann Paul (1995: Chap. XXII §284): “Thus a word taken from a related dialect quite commonly appears in the same phonological form it would have acquired had it been preserved from the period of the former linguistic unity.” For the speaker of the Proto-Middle Latvian dialect, the relation of

³ Some more recent borrowings also exhibit the tsokanye preserved to this day in the Pskov dialects: càrnica < Russian dialectal черница ‘blueberry’ (Laumane 1977: 73); casnàks from Russian чеснок (cf. Lithuanian česnākas), kànčs from Russian кончик in Kalupe (Reķēna 1962: 371, 399). These entered High Latvian later, brought by settlers from the Pskov region.

92

Die Vermittlungsrolle des Hochlettischen...

approximately Proto-Middle Latvian ruoka to Proto-High Latvian rūka as a model. It therefore applies the sound substitution Proto-High Latvian *ū → Proto-Middle Latvian *uo⁴.

This fact also explains: ḷuoti and Old Russian пюто, pūōsts and Old Russian поуеть. Regarding the word suōgīs, it should also be noted that at the time this word was borrowed from Old Russian, the phoneme ǥ was presumably pronounced *d’ in High Latvian, cf. Seržant 2006, as it still is today in northeastern Latvia (Ābele 1940: 213).

The vowel systems of these two forms seem explicable only through the mediation of High Latvian. This hypothesis was already put forward in Reķēna (1962: 31). The Old Russian word соудии/соуд- ани was also borrowed into Lithuanian, where it has the regular vowel system ū; cf. Lithuanian sūdžia vs. Latvian sùoǵis and Lithuanian sūdas vs. Latvian sùods, Lithuanian bliūdas vs. Latvian bļùoda.

Further examples are: sùolīt (Lithuanian siūlyti, Old Russian соулити), duōma (Lithuanian dūmà, Old Russian доума), uōma (Lithuanian ūmas, Old Russian оумъ), rùobs / rùobīt from Old Russian ровити, skūops (Old Russian скопъ), as well as mūoka from Old Russian моука, puōga from Old Russian поугъвица/ поугъва / поугъы (Srezn₍II₎ 1723), skūops from Old Russian скопъ, rùobeža from Old Russian ровежь (Srezn₍III₎ 179), suōma Old Russian соума, Latvian (alongside the regular kūms), also dialectal kuoms (Lithuanian kūmas, Old Russian коумъ) (ME₍II₎ 336).

1.1.2 The suffix syllable

Furthermore, in some cases Middle Latvian suffixal -uo- also seems to belong here; cf. Middle Latvian karuogs from Old Russian хороунъ/ хороунъ-вы, kaltuons (dialectal) corresponding to Russian колтун < Old Russian *колтоунъ, kažuoks Old Russian кожохъ (examples from Endzelin 1899: 294).

1.1.3 Later borrowings usually have a short -u- in this position,

cf., e.g., dial. suma from Russian сума, dial. budka from Russian будка (Endzelin

⁴ It is conceivable that in some cases the loanwords really travelled from east to west, from one place to the next. Along the west → east line, various stages of the transition uo > ū are attested (cf. DA: 54a., 55e.), which was probably also the case in earlier times. Thus, while *bļūda travelled farther west, the vocalism changed from ū (in the far east) to uo (in the west), depending on the pronunciation of the phoneme in question in each locality. The phonetic form of words of the type bļūoda can also be explained in this way.

93

ILJA SERŽANT

1899: 306), which is attributable to the shortening of the vowel in question in Russian.

1.1.4. Kiparsky’s interpretation

Some of the examples mentioned above (§ 1.1) are cited from Kiparsky (1952: 74), who, on the basis of Latvian (and also, in some cases, the uo found in the Finnic languages of the Baltic Sea region), concludes that borrowing into Latvian occurred earlier here than into Lithuanian, which has ū in this position in almost all borrowings. As evidence that Latvian uo does indeed go back to uo/ō—in other words, that this correspondence cannot be explained as a peculiarity of Latvian—he cites borrowings from Swedish and Middle Low German, in which Swedish or Middle Low German ô [ō] and û [ū] are indeed rendered by Latvian uo and ū, respectively; cf. Middle Low German bôm ‘turnpike’ – Latvian buomis. For several reasons, however, this interpretation by Kiparsky seems untenable.

(a) The Middle Low German and Swedish loanwords must have entered Middle Latvian or Livonian Latvian directly, and the Livonian dialect (“lībiskais dialekts”) differed little from Middle Latvian in vowel quality, both formerly and today. Consequently, these borrowings were not altered by another Latvian dialect, so that there was phonetic identity here. The situation with regard to East Slavic borrowings is fundamentally different: they were first taken up by High Latvian, which had undergone a series of sound changes in earlier times (cf. the relative chronology in Seržant 2005: 84).

(b) The pronunciation of Proto-Russian *ъ as *uo or *ō seems unlikely for the following reasons:

– to my knowledge, there is no evidence in present-day dialects and their subdialects for an earlier pronunciation of Old Russian ѫ as [uo];

94

Die Vermittlungsrolle des Hochlettischen...

– the East Lithuanian dialects provide indirect evidence for Proto-Slavic *ǭ from Proto-Slavic *on and *un. The nasal vowels disappeared in East Slavic before the beginning of written records (cf. the table in Kiparsky 1963: 150). In High Lithuanian, the nasal vowels have generally been preserved; cf. Lithuanian rankà. It has long been known that, although nasal vowels were preserved in East Lithuanian under the influence of the East Slavic languages, their vowel quality changed (Zinkevičius 1987: 69). Thus, here, around the 9th–10th centuries, the sounds ą, an, am became ų, un, um, and ę, en, em became į, in, im (Zinkevičius 1966: 96; idem 1981: 94 and 1987: 69). This indicates that the relative chronology of the change from Proto-Slavic *on > to East Slavic u proceeded as follows: Proto-Slavic *on > *un > ų: (Old Bulgarian ж) > Old Russian ѫ [u:]. The intermediate stage *uo is therefore unlikely. Kiparsky’s assumption also encounters relative-chronological difficulties more generally: on the one hand, it postulates a very ancient pronunciation of ѫ/ ѧ, while on the other hand, it presupposes the merger of the two Proto-Slavic vowels (ѫ > ѧ)—a process that, as already mentioned, took place shortly before the beginning of East Slavic written records. Given the merger of the two phonemes in the loanwords, their great antiquity is therefore hardly plausible; – in the parallel case of East Slavic -и-: Latvian -ie- (see below, § 3), such an explanation—that the East Slavs pronounced ie here—is altogether inconceivable, since some East Slavic -и- goes back to Proto-Slavic *ī. An East Slavic development *i > *ie cannot be demonstrated; – furthermore, the High Latvian borrowings (as well as the Lithuanian ones) that were not adopted into Middle Latvian also support the original sound value of Old Russian ѫ. In kìukūl’i (Kalupe, Reķēna 1962: 405, cf. Old Russian коукокль (Srezn₁ 1362)), its archaic character is confirmed by the change Proto-High Latvian *ū > High Latvian iu/ yu, one of the relatively oldest sound changes in High Latvian (Seržant 2006: 44–47 and 84).⁵

Kiparsky’s view (loc. cit.) must therefore be revised.

⁵ kìukāli in Baltinava and Viļāni (ME₁₁ 342) is an artificially reconstructed Middle Latvian form for *kiukāli / *kyukāli, in accordance with the dictionary’s design, in which all dialect forms are recorded in their Middle Latvian pronunciation.

95

Ilja Seržant

1.2. East Slavic -ъ- > Latvian -i-

Another borrowing, Old Russian ѿрѡвь corresponding to Middle Latvian siruobs_III (ME 848), can be explained phonetically only through the mediation of the High Latvian dialect. That the borrowing must be old is attested by the vowel of the first syllable, which corresponds to the Old Russian reduced vowel ъ in a weak position; in later Russian this reduced vowel was syncopated, cf. Modern Russian сруб. The Old Russian reduced vowel had the sound value [ŭ] (ultrashort [u]). In accordance with its phonetic form, the Old Russian word was borrowed into High Latvian as *syrūbs (where y represents a sound similar to the ы-sound of modern Russian). This lexeme is also attested in almost the same form in Kalupe; cf. syrups, gen. sg. syruba (Reķēna 1962: 459). Here the medial vowel was shortened, thus *ū > u⁶. The Middle Latvian -i- in siruobs is thus a substitution for the High Latvian sound -y-. The assumption of direct borrowing by Middle Latvian would not explain the substitution of -i- for Old Russian -ъ-.

1.3. East Slavic -и- > Latvian -ie-

Analogous to the correspondence presented in § 1 (East Slavic -u- > Latvian -uo-) is the correspondence for its palatal counterpart. The explanation of this correspondence likewise rests on the early monophthongization of Middle Latvian ie > High Latvian ī. We therefore assume that East Slavic loanwords were also borrowed into High Latvian with ī and only then underwent the sound substitution ie > ī in the reverse direction (as in § 1, cf. Paul 1995: ch. XXII §284). This yielded Middle Latvian ie as the correspondence of East Slavic и and Proto-High Latvian *ī. The following examples attest to this development: the name of the Old Pskov tribe, Old Russian кривичь, hence High Latvian *krīv- > Middle Latvian krievs, which remains in use in Latvian to this day as a designation for Russians; Latvian miers < Old Russian миръ, among others. The widespread Latvian suffix -(i)nieks), which generally denotes a person belonging to an occupation, should probably be included here. The-

⁶ In this dialect, unstressed long vowels are often shortened; cf. ūzuls (< High Latvian elsewhere ūzūls), uobuls (< High Latvian uobūls), among others (Reķēna 1962: 39).

96

Die Vermittlungsrolle des Hochlettischen...

This suffix, which has no old parallels in Lithuanian, must have replaced the corresponding Baltic suffix Latvian *-iniks, Lithuanian (regularly) -ininkas under East Slavic influence (here -никъ). The penetration of the East Slavic form of the suffix was made possible by numerous borrowings containing this suffix; cf. Latvian burtnieks ‘beekeeper’ and Old Russian бъртникъ.

1.4. East Slavic -ъі- > Latvian -ie-

The same applies to Middle Latvian siērs, which was borrowed early from Russian (for the intonation, see § 3 below): Old Russian сыръ could only be rendered in High Latvian as *sīrs, since the Russian vowel ы is stressed in this case and therefore retains its length or is perceived as long, so it could not be substituted with Latvian y—what was needed was a long vowel. Moreover, a “misspelling” [i.e., a deviation from the orthographic norm of the time] is occasionally also found in the Pskov monuments of the 15th century, which may attest to confusion between Old Russian ы and и (Karinskij 1909: 176). An assimilation of [ъі] to [и] that occurred on East Slavic soil is therefore also conceivable. That the form *sīrs existed in High Latvian is proved by the form seirs in Jaunroze₃₈₄, Jaunlaicene₃₈₅, Veclaicene₃₈₆ (Ābele & Lepika 1928: 39), where ei regularly develops from an old ī (Endzelīns 1951: 146–147; Seržant 2005: 44–47). Other examples also attest to the substitution of Old Russian ы by High Latvian ī; cf. High Latvian skrein’e (Kalupe, Reķēna 1962: 460) < *skrin’a < Old Russian скрина / скрина, as well as the borrowing seits < sits (= standard Latvian sāts) ‘sated’, from Russian сыть (МЕ₃ 855), which is found only in the Latgalian area. The form *sīrs thus moved westward and, by retracing the attested intermediate stages of the transition Middle Latvian ie -> High Latvian ī in reverse, became siers; cf. § 1.3 above. Similarly, High Latvian ķēiļš Kalupe (Reķēna 1962: 404), which goes back to Proto-High Latvian *ķilis, from East Slavic къіль⁷.

⁷ Strictly speaking, these borrowings are probably somewhat later in origin than, for example, Old Russian сърѹбъ Latvian siruobs, since here the vowels ъ and u in ъi are not rendered as *ūi / *ūi (as in the oldest borrowings, such as muita and from Proto-East Slavic мыто).

97

ILJA SERŽANT

1.5. East Slavic -ъ- > Latvian -ē-, Lithuanian -ie-

The correspondence ē in Latvian versus ie in Lithuanian for Old Russian ъ is explained by Kiparsky (1952: 76) through Proto-Latvian umlaut (Endzelīns 1951: 96–97); cf. Latvian grē̄ks and Lithuanian griēkas versus Old Russian гъхъ. That is, the Latvians first substituted Proto-Latvian (closed) *ē here and then umlauted it according to Proto-Latvian umlaut, yielding ē when followed by a back vowel, e.g. grē̄ks, and ē when followed by a front vowel, e.g. vēsts, plural vēstis versus Old Russian вѣсть. Unlike Lithuanian, High Latvian could not substitute the inherited Proto-Latvian closed *ē (from Proto-Baltic *ei / *ai), since by the time of our borrowings the change Proto-Baltic *ē > Proto-Latvian ie > High Latvian ī had already been completed in High Latvian (see § 1.3 above, and also § 1.4).⁸

The metatony discussed here had already been completed in the Proto-Latvian period. Therefore, no specifically High Latvian influence can be identified in these borrowings.

1.6. East Slavic -а- > Latvian -ē- / Lithuanian -ē-

The Old Russian word свѧть was borrowed into Latvian as svēts. In the very oldest Old Russian texts⁹, ѧ merges with а, which indicates that ѧ was pronounced as /’a/; cf. also Modern Russian свят /sv’at/. A. Holvoet pointed out to me that the two East Baltic languages in

⁸ It is unclear whether the Latvian borrowing žēl belongs here, from Old Russian *жълъ > жалъ, since this word was probably borrowed at a precursor stage (see Krys’ko 1994: 31, with reference to G. A. Il’jinskij, n.d.). Furthermore, Krys’ko (1994b: 27; idem 1998: 80) assumes, on the basis of the Finnish rendering of Proto-Slavic ъ as ää (in mäẗa – määrä), the East Slavic rendering of Finnish ja as ъ (in Karjala – Коръла), and some dialectal relics, that the North Russian or “ильмено-словенские” dialects had preserved the Proto-Slavic pronunciation of ъ as [ä:] (open e). In that case, conversely, Latvian vēsts (with closed ē) would have to be regarded as the result of Proto-Latvian metatony and Latvian grē̄ks as borrowed without change. This, however, does not seem likely for the dialect of the Krivichi, since even in the oldest written Novgorodian monuments the merger of ъ and е, or ъ and и, occurs, clearly indicating here the closed character of ъ; cf. Zaliznjak (1995: 57).

⁹ The Ostromir Gospel (“Остромирово Евангелие”) of 1056 already shows this confusion.

98

Die Vermittlungsrolle des Hochlettischen...

In earlier times, most consonants did not have a palatality correlation. The palatality of the consonant in the source language was therefore expressed in the borrowing by the vowel itself, that is, C’a > Ce: cf. Lith. česas < Old Russian часъ, Lith. čėdyti < Old Russian цадити, Lith. rėdyti < Old Russian радити, Lith. mėsa < Old Russian мясо. In this case, then, the divergent pronunciation of the Latvian form is to be explained by phonologically conditioned substitution, unlike in the other cases discussed here.

Cases such as prèst and prèšlica from Old Russian прасти and праслица (Srezn. II 1719; examples from Laumane 1977: 85, ME III 390) also belong here; likewise Old Russian блядь ‘deceiver’ (Srezn. I 123; Zaliznjak 1995: 595 ‘meretrix’) corresponding to Latvian blēdis ‘the same’. Also noteworthy is the root vowel’s falling tone, which can be explained only through the mediation of High Latvian (cf. § 3 below). A further example of the adoption of a High Latvian form is Middle Latvian lẽča ‘lens’ (ME II 455), from Old High Latvian *l’āca < Old Pskov лаца (Old Russian лача; cf. § 1.0.4). Also belonging here, though attested only in High Latvian, is ràdâ ‘nearby’ (loc. sg.), from Old Russian радъ (according to Laumane, ibid.).

1.7. East Slavic -ь- > Latvian -u-

The phonetic value of Old Russian / Proto-Slavic ъ was [ĭ] (ultrashort [ĭ]). The -u- in the Middle Latvian word krusts could be explained through the mediation of High Latvian; the word was certainly borrowed from Old Russian кръсть [kristŭ]. One may assume that this word was borrowed from Old Russian in precisely this form, that is, as *krists. It would then have undergone umlaut in High Latvian in accordance with the second High Latvian umlaut law (Seržant 2005: 68–72), yielding krists > krysc (for example, in Kalupe₄₄₅, Reķēna 1962: 411). It is conceivable that speakers of Middle Latvian adopted this very form from speakers of High Latvian and substituted the closest equivalent, namely -u-, for -y-, which was absent in Middle Latvian¹⁰. Evidence for an underlying root form *krist- is provided by

¹⁰ Bügas Erklärung (1925: 44), nach der das Lettische dieses Wort aus dem weißrussischen **кръсть entlehnt haben soll, scheitert daran, daß diese Form im Weißrussischen nicht belegt ist, vgl. das neuweißrussische хрест, das allerdings wohl aus dem polnischen chrzest entlehnt ist (Temčīn, briefl.). Die von Büga angenommene Depa-

99

Ilja Seržant

the fact that those derivatives of this word which contain a palatal vowel after the root vowel—and thus could not undergo the second High Latvian umlaut here—preserve the root’s old -i-; cf. Middle Latvian kristīt ‘to baptize’ (Old Russian крьстити Srezn₁ 1342)¹¹, High Latvian krístęniski ‘krusteniski’ in Krāslava (Alksnis 1932: 43).

1.8. East Slavic -o# > Latvian -e#

Another case of East Latvian mediation is the e-stem ending in pekle ‘deep abyss, bog, hell’ (ME₃ 193), which is a borrowing from East Slavic пекло. The following development is involved: пекло > Proto-High Latvian *pekla (cf. Lithuanian peklà) > pęklę (cf. High Latvian mùotia < mâte, zam’a < zeme Seržant 2005: 74–5), Middle Latvian pekle. The open -e in this word’s ending cannot be old, as indicated by the open -e- in the root. An open -ę- is impossible in the root of an old Latvian e-stem, since the old word-final -e would have umlauted the root’s -e- to a close **-ę-, hence **pękle (cf. Endzelīns’s description of umlaut, 1951: 94). The same probably applies to svece from Old Russian свѣча. Overall, the confusion or merger of Baltic ā- and ē-stems is characteristic of the High Latvian dialect. It is probably to be explained here by the phonetic merger of -e and -a, which occurs in some cases.

2. THE CORRESPONDENCE OF PROTO-SLAVIC AND LATVIAN TONES IN BORROWING

Endzelin (1913: 115–118; 1922: 206–207) stellt die These auf, daß (i) die altruss. Entlehnungen ins Lettische die urslavischen Töne widerspiegeln; daraus ergibt sich notwendigerweise die folgende These, daß (ii) das Urrus-

Palatalization of -pb- in Proto- or Old Belarusian to -pb- occurs only with reduced vowels in weak position and does not yield **-pb-, but rather -pbl-. (Durnovo 1969: 171); крьстъ would therefore have yielded nom. sg. крьсть and gen. sg. *крысма.

¹¹ Here, moreover, the semantics makes clear that the verb kristīt ‘to baptize’ was also borrowed, rather than originating on Latvian soil, since the lexical association between ‘cross’ and ‘baptize’ is exclusively Slavic.

100

Die Vermittlungsrolle des Hochlettischen...

Slavic or Old Russian had preserved the Proto-Slavic suprasegmental features. I must, however, agree with Lehr-Spławiński (1918), who, despite the length of the list of correspondences given by Endzelin (1913: 116–118), points out that the Latvian tone depends on the position of stress in the Russian counterpart: if the root syllable is stressed, it receives the lengthened tone in Latvian; if it is unstressed, it receives the falling tone. This rule also applies to other languages: thus, all borrowings from Middle Low German are pronounced with the lengthened tone, since the first syllable is always stressed there. This is probably connected with the fact that a Latvian perceives a syllable stressed in another language as slightly overlong (provided, of course, that it is long in the source language) and renders it accordingly in Latvian, since the lengthened tone—as its name indicates—denotes an overlong syllable. The following examples contradict the conjecture put forward by Endzelin (loc. cit.):

puōsts versus Serbo-Croatian pūst, pūsta,

ļeca versus Serbo-Croatian léca

uōma versus Serbo-Croatian gen. sg. ūma

Wie Endzelin (1922) in seiner Antwort an Lehr-Spławiński ausführt, gibt es im Lettischen auch russische Entlehnungen, die, obwohl im Russischen durchgehend betont, im Lettischen den Fallton aufweisen. Es handelt sich um die folgenden Beispiele: lett. vèsts versus aruss. вѣсть, lett. ļuoti versus aruss. лѹто, lett. blèdis ‚Betrüger‘ versus aruss. бладь (nach Srezn₁ 123,ds.‘). Diese Beispiele lassen sich sehr gut erklären, wenn man annimmt, daß sie nach dem Zusammenfall des Dehn- und Falltons im Fallton ins Hochlettische entlehnt wurden und deshalb das Mittellettische (das diesen Zusammenfall nicht durchgemacht hat) bereits falltonig erreichen. Es handelt sich also auch hier um dasselbe Prinzip der Vermittlung des Hochlettischen bei den ostslavischen Entlehnungen, das mittels der Annahme einer hochlettischen Zwischenstufe die phonetischen und (in diesem Fall) auch die tonalen Besonderheiten erklärt (vgl. die in § 1.1–1.9 besprochenen Fälle).

101

Ilja Seržant

3. SUMMARY

One further, not insignificant result of these considerations must be highlighted. As the East Slavic borrowings in § 1–2 show, it was the speakers of Proto-High Latvian who adopted the loanwords and then—and this is especially worth emphasizing—themselves passed them on to the speakers of Proto-Middle Latvian. Since most East Slavic borrowings are cultural terms, we must assume that the Proto-High Latvian area was a certain social center of the Latvians of that time. This conclusion applies in particular to the East Slavic mission to Latvia. To explain the linguistic mediation by the ancient High Latvians (Latgalians), we must assume that they themselves, and not the East Slavs, had already carried out missionary work in the ancient Middle Latvian-speaking area. Alongside the many cases presented above of forms adopted from Proto-High Latvian, such as krusts (§ 1.7), muoka (§ 1.2), pekle, svece (§ 1.8), which might still have been attributed to East Slavic missionaries, there are also purely suprasegmental traces of Proto-High Latvian in Middle Latvian, such as the broken tone in the suffixal syllable of baznīca, which is secondary and can be explained ONLY with the help of High Latvian (Seržants 2003: 102; for the suprasegmental transition, see 92–94). The explanation presented in § 2 for the occurrence of the falling and lengthened tones in East Slavic borrowings should also be noted. It may now be considered proven that it was the ancient Eastern Latvians (Latgalians) who carried out missionary work among the western Latvians and passed certain cultural terms on to them.

Andere von Endzelin (1899) genannte Beispiele sind anders zu erklären. Das Wort lett. cieņa ist kaum eine Entlehnung aus dem altruss. цѣна, insbesondere wegen der *-i̯ā-Deklination im Gegensatz zur *ā-Deklination im Slav., wobei der lettische Vokalismus regelmäßig mit -ie- den zu erwartenden urbalt. Diphthong *-ei- / *-ai- wiedergibt.

Overall, the tendency presented here must of course not be transferred a priori to other borrowings as a “sound law.” One must bear in mind that individual borrowings can also pass from one language to another by roundabout routes¹².

¹² Thus, the following borrowings cannot be explained within the correspondences and developments presented: Latv. pūkas: Slav. пухъ, Latv. pirāgs: Slav. пирогъ.

102

Die Vermittlungsrolle des Hochlettischen...

4. LOANWORDS FROM LITHUANIAN

The mediating role of the High Latvian dialect can be further illustrated by an example of a loanword from Lithuanian. Although this loanword cannot be dated with any certainty and may therefore also be recent, in this case too the Middle Latvian sound form becomes explicable only if one assumes mediation by the High Latvian dialect.

While the word ģidrs (dialectal in Dundaga, Sauka, ME₁ 700, standard-language dzidrs) can likewise be explained as a loan from Lithuanian giẽdras, mediated by the northern Samogitian dialects (where Aukštaitian Lithuanian ie is pronounced as ẹi, sometimes also as iį; see Zinkevičius 1966: 84), or as a shared Curonianism, the Middle Latvian word ģibt is clearly a loan from Lithuanian (here geibti) via the High Latvian dialect. The word was borrowed into the High Latvian dialects in precisely this form, namely *ģeibt (g is impossible before a palatal vowel throughout the entire Latvian-speaking area). The word then passed through the stages of the transition ī > ei from east to west (that is, ei > ī) and arrived in Middle Latvian as ģibt.

ABBREVIATIONS

DA Atlas of Latvian dialects. Phonetics. In preparation. LU Latvian Language Institute.

ME Mülenbachs, Latviešu valodas vārdnīca. Edited, supplemented, and continued by J. Endzelīns. 2nd unchanged edition. In five volumes. Chicago Baltic Philologists’ Group. Chicago, 1954.

POS Pskov Regional Dictionary with Historical Data. Leningrad State University. Leningrad University Press.

SREZN I. I. Sreznevsky, Materials for a Dictionary of the Old Russian Language from Written Monuments. Publication of the Department of Russian Language and Literature of the Imperial Academy of Sciences. Volumes I–III. St. Petersburg 1893. Printing House of the Imperial Academy of Sciences.

Standard Latvian kūms, kūma ‘godfather, godmother’ could be interpreted as a direct loan from Lithuanian into the western Latvian dialects.

103

Ilja Seržant

LITERATURVERZEICHNIS

Ābele, A. & M. Lepika 1928: Par Apukalna izloksnēm. Filologu Biedrības Raksti 8, 19–49.

Alksnis, M. 1932: Krāslavas izloksne. Filologu Biedrības Raksti 12, 26–48.

Būga, K. 1925: Die litauisch-weißrussischen Beziehungen und ihr Alter. Zeitschrift für Slavische Philologie 1, 26–55.

Endzelin, J. [I. Эндзелинъ] 1899: Латышскія заимствованія изъ славянскихъ языковъ. Живая Старина. Журнал о русском фольклоре и традиционной культуре 3, 265–312.

Endzelin, J. 1913: Weiteres zu den lettischen Intonationen. Indogermanische Forschungen 33, 104–118.

Endzelin, J. 1922: Zur Intonation der lettischen Lehnwörter aus dem Russischen. Slavia 1, 206–207.

Karinskij, N. 1909: Языкъ Пскова и его области в XV вѣкѣ. (Записки Историко-Филологическаго Факультета Императорскаго С.-Петербургскаго Университета. Часть XCIII.)

Kiparsky, V. 1952: The earliest contacts of the Russians with the Finns and Balts. Oxford Slavonic Papers 3, 67–79.

Kiparsky, V. 1963: Russische historische Grammatik. Bd. I. Die Entwicklung des Lautsystems. Heidelberg: Carl Winter.

Koškin, I. 2006: Антонс Брейдакс – выдающийся латышский языковед. In: Antons Breidaks, Darbu izlase I. Rīga: LU Latviešu valodas institūts, 33–37.

Krys’ko, V. B. 1994: Заметки о древненовгородском диалекте. Вопросы Языкознания 1994 [5], 28–45.

Laumane, B. 1977: Лексический материал диалектологического атласа латышского языка, отражающий латышско-русско-белорусско-польские контакты. In: Контакты латышского языка. Riga: Zinātne, 48–95.

Lehr-Splawiński, T. 1918: Ślady dawnych różnic intonacyjnych w językach ruskich. (Z powodu prac Szachmatowa, Rozwadowskiego i Endzelina.) Rocznik Slawistyczny 8, 250–263.

104

Die Vermittlungsrolle des Hochlettischen …

Lubben, A. 1980. Mittelniederdeutsches Handwörterbuch. Nach dem Tode des Verfassers vollendet von Christoph Walthner. Darmstadt: Wissenschaftliche Buchgesellschaft.

Nazarova, E. L. 2002: Псков и Ливония в 40–90 гг. XIII в. In: Civitas et Castrum ad Mare Balticum. Baltijas arheoloģijas un vēstures problēmas dzelzs laikmetā un viduslaikos. Rakstu krājums – veltījums LZA īstenajam loceklim Prof. Dr. habil. hist. Andrim Caunem 65 gadu dzīves jubilejā. Rīga: LU Latvijas vēstures institūts, 591–609.

Paul, H. 1995: Prinzipien der Sprachgeschichte. 10., unveränderte Auflage. Studienausgabe. Tübingen: Niemeyer.

Reķēna, A. 1962: Kalupes izlokšne. Fonētika. Morfoloģija. Leksika. Kandidāta disertācija. (Unveröffentlicht.)

Rūķe-Draviņa, V. 1964: Die Expansion der russischen Buchweizenbenennungen nach Westen. Scando-Slavica 10, 107–131.

Semenova, M. F. 1959: По поводу двух фонетических явлений русских и латышских говоров Латгале. In: Rakstu krājums. Veltījums Akadēmiķim Profesoram Dr. Jānim Endzelīnam viņa 85 dzīves un 65 darba gadu atcerei. Rīga: Latvijas PSR Zinātņu Akadēmijas Valodas un Literatūras Institūts, 595–606.

Seržants, I. 2003: Intonationen in den suffixalen und Endsilben im Lettischen. Synchronie und Diachronie. Baltu filoloģija 12 (1), 83–123.

Seržant, I. A. 2005: Относительная хронология основных фонетических изменений в истории верхнелатышского диалекта. Acta Linguistica Lithuanica 53, 39–89.

Seržant, I. (im Druck): Относительная хронология процесса аффрикатизации в верхнелатышском.

Zaliznjak, A. A. 1995: Древне-новгородский диалект. Москва: Российская Академия Наук. Институт Славяноведения и Балканистики.

Zinkevičius, Z. 1966: Lietuvių dialektologija. Lyginamoji tarmių fonetika ir morfologija. Vilnius: Mintis.

105

Ilja Seržant

Zinkevičius, Z. [З. П. Зинкявичус] 1981: К вопросу о генезисе литовских диалектов. In: Проблема этногенеза и этнической истории балтов. Тезисы докладов. Март, 1981. Вильнюс: Академия Наук Литовской ССР, 93–95

Zinkevičius, Z. 1987: Lietuvių kalbos istorija, Bd. II. Iki pirmųjų raštų. Vilnius: Mokslas.

Ilja A. Seržant

Lietuvių kalbos Institutas

P. Vileišio

[email protected]

106