Liet. „barzdà“ ‘barba’: samplaikos „-zd-“ kilmės klausimu

ACTA LINGUISTICA LITHUANICA LI (2004), 15–27 Lith. barzdà ‘beard’: on the origin of the consonant cluster -zd- ROLANDAS KREGŽDYS Vilniaus universitetas The article deals with the origin of the consonant cluster in Lith. barzdà ‘barba, mala’. Whereas Būga assumed the cluster to be inherited, Endzelins ascribed it to Slavic influence. Otrębski described its derivation as obscure. The forms borrowed into Finnic point to -d- rather than -zd-, but they may go back to Germanic barda ‘barba; securis’ as well as to Baltic words. In this article the view is advanced that Lith. barzdà is inherited and reflects an IE dialectal formation. The analysis confirms Būga’s assumptions concerning the original morphological structure of the word. It follows that Lith. barzdà, Latv. dial. bàrzda are older than Latv. bārda. 1. INTRODUCTION The origin of the consonant cluster -zd- in the structure of Lithuanian nominal forms was discussed long ago by Kazimieras Būga. He provided several examples with -zd-: -z- alternants and assumed the historical status of consonantal -zd-, i.e., he had no doubt that this cluster was ancient: Lith. liždas ‘a place set up for laying eggs, brooding and raising young (of birds, insects); nest’ (DLKŽ⁵ s.v.) – dial. lizàs ‘the same’¹, Lith. blauszdà ‘the part of the leg from the ankle to the knee’ – dial. blauzà ‘the same’, Lith. lazda ‘a polished stick used for support when walking; to strike someone, etc.’ – dial. lazà ‘the same’, Lith. barzdà ‘hair on the lower part of the face’ – dial. baržà ‘the same’ (Būga 1958: 273). Zigmas Zinkevičius states that “in the past, the Lithuanian cluster zd sporadically changed into z” (1980: 129 ff.), i.e., he considers this cluster to be very ancient. The functionality of the consonant cluster -zd- (-d-) in the Lithuanian verb system has been briefly discussed by Erdvilas Jakulis. It has been observed that the fricative /z/ is usually retained, although it is not always etymological (Jakulis 2002: 35, 62): bėzda – bėza ‘to fart; to stink, to emit a bad smell’; Latv. bezdēt, -u, -ēju ‘the same’; Lith. kūza, -ėti, -ėjo – kūzda, -ėti, -ėjo ‘to whisper; to tremble, quiver’; Lith. splàzda – splàza ‘to flutter’, and others. A similar tendency to retain the fricative consonant can also be traced in Latvian dialects: Latv. dial. brāzēt, -u, -ēju – brāzdēt ‘to crackle’ (Jakulis 2002: 92, 108); cf. also an example with the cluster -z(ž)g-: Latv. dial. blīzgēt ‘to glimmer, shimmer’ = Latv. dial. blizēt (Jakulis 2002: 87). On the basis of Latvian and Prussian data, Endzelin attempted to demonstrate the late origin of the Lithuanian name for ‘beard’ with -zd- (Endzelin 1902a). 1 Cf. also PEŽ III 54 ff. 16 | ROLANDAS KREGŽDYS Aim of the study: 1. to determine the areal boundaries of Lithuanian and Latvian forms with -zd-, -z-, -d-; 2. to reinterpret the significance of Finnish loanwords in determining the Proto-Baltic form of the name for beard; 3. to identify the reasons that may have led to the emergence of different morphophonological variants with -zd-, -z-, -d-. 2. ANALYSIS OF THE MATERIAL Viena iki šiol neišaiškinta baltų kalbų barzdos pavadinimų problema – šaknies struktūrinio modelio neatitikimas lietuvių ir atitinkamai latvių bei prūsų kalbose, mat rytų baltų dialektuose vartojamos formos su šakniniu junginiu -zd- (latvių kalboje ir -d-), o prūsų kalboje – tik forma su -d-: liet. barzdà resp. lat. dial. bàrzda – lat. bārda resp. pr. bordus E 101. Yra dvi tokio neatitikimo aiškinimo hipotezės: 1. Formos su samplaika -zd- yra senos, paveldėtos iš ide. prokalbės, todėl traktuotinos kaip reliktinės, lygintinos su s. i. bhr̥stís ‘ašmenys, kraštas’, s. isl. ir s. v. a. burst ‘šeria’, s. ang. byrst ‘šeria’ ir kt. (Būga 1958: 273; ME 273; LEW 36; Kurylowicz 1935: 118; Vasmer I 196) – rekonstruotina prablt. *barzdā ‘t. p.’ (LEV 108). Tokios formos rekonstrukcija suponuoja ir ide. prolyties korekciją: vietoj iki šiol visuotinai pripažįstamos ide. *bhar-dhā² rekonstruotina *bhars-dhā (ME ibd.), iš kurios iki šiol buvo vedami slavų „vagą“ reiškiantys žodžiai (ESSJ II 220); 2. Lytys su -zd- yra vėlyvos, paveiktos panašių žodžių į sl. *borzda³ ‘vaga’ > rus. borozda ‘t. p.’ ir kt. (Endzelin 1902; ESSJ II 198; Varbot 1984: 216; PEŽ I 151), arba kildintinos iš bl. *bard- + *-da > *barzda (Skardžius 1943: 99; Endzelynas 1957: 100) – rekonstruotina prablt. *bardā ‘barzda’ (fem.) (Trautmann 1923: 27). Tokiai rekonstrukcijai pritaria Toporovas (PJA 241), Mažiulis (PEŽ I 151; dar plg. ESSJ II 198, 220), nors kai kurie lingvistai -zd-/-z- kaitą tiesiog įvardija kaip neaiškią (Otrębski 1958: 334 t.). 2 Pagal Marki hipotezę, ide. *bhardhā yra inovacinis barzdos reikšmę, pakeitęs ide. *smonkru ‘barzdą’. Ši lytis savo morfologine struktūra artima s. i. bhr̥stís ‘ašmenys’ (fem.), lot. far ‘duona, miltai’, got. barizeis ‘miežis; Hordeum sativum’, s. isl. bar ‘grūdai; miežiai’ ir kt. Spėjama, kad šis žemdirbystės terminas buvo pasiskolintas vakarų ide. dialektinės zonos gyventojų iš semitų dar iki kraustymosi į Europą. Tokios hipotezės šalinin kai ide. *bhardhā aiškina kaip dirbinį: pirmasis dėmuo *bhar- interpretuojamas kaip senas indoeuropiečių miežio pavadinimas (Hordeum sativum – Marki hipotezė), kilęs iš s. hebr. bar ‘beicuotas grūdas’, arab. bur ‘javai’, pietų arab. bor, barr ‘grūdai, javai’ (Otkupščikov 1963: 89). Taip pat manoma, kad šis grūdų pavadinimas galėjo būti pasiskolintas vėlyvuoju ide. prokalbės laikotarpiu (skilimo į dialektus pradžia), o ne suskilus jai į dialektus (Ivanov 1985: 162 t.). Skolinimosi iš arabų hipotezę anksčiau yra pasiūlęs Meillet (1914: 143). Borealinės lingvistikos atstovai teigia, kad ide. *bhardhā sememos barzda vėlyvumą vietoj senesnės „augalinės“ patvirtina rekonstruojama borealinė lytis *bard- „krūmynas“ ~ gruz. bard- ‘t. p.’ ir kt. (Klimov 1988: 153). Pasak Porcigo, ide. *bhardhā iš pat pradžių reiškė ‘šeriais apaugusi veido dalis (kūno dalis [gyvulių]); šeriai’, prokalbėje funkcionavęs kaip nomen collectivum, vėliau pakeitęs ide. *smek- (Porcig 1964: 302). Jis remiasi hipoteze, kad ide. *bhar- rekonstruotina semema ‘aštrus’ (plg. Otkupščikov 1963: 89). Tradiciškai rekonstruojama proly tė *bhardhā traktuojama kaip indoeuropiečių veldinys < ide. *bhār- ‘išsikės į priekį; šerys; smailuma; akuotas’ (Pokorny 110; Fraenkel 36) < verb. ide. *bher- ‘trinti; pjauti; rėžti; mušti, smogti; skelti, skaldyti’ (LEV 108) > verb. ide. *bhar- ‘kyšoti, būti išsikišusiam’ (Pokorny 110; ESSJ II 197) → ‘šeriai’ (Porcig ibd.; Gamkrelidze, Ivanov 1984: 69). 3 An attempt has also been made to relate it to Russian dialectal borozdavka ‘wart’, Kshubian barzovka ‘ibid.’ (Nepokupnas 1983: 20, 47; Endzelin 1902b/1971: 221). LIET. barzdà SAMPLAIKOS -zd- KILMĖ 17 The second hypothesis is difficult to prove, since there is no consensus on the CVRC reconstruction of the structural model of the root of the Prussian word for beard. On the basis of Prussian onomastic material, the root *barsd- is reconstructed, that is, a Prussian CVRCC- model for beard is posited, as mentioned by V. Toporov⁴ (PJA₁ 197 ff.) and A. Kaukienė⁵ (1999: 21); cf. Prussian Barsde (name of a deity), Prussian Claus Barsde (proper name—Trautmann 1925: 17), Curonian Barsdemin (proper name) (Toporov PJA₁ 199), and perhaps the Prussian proper name Borssythe – Herman Borssythe 5a, 51 = 5b, 62 Schalwe – Trautmann 1925: 20 (for more on the origin of this personal name, see Kregždys 2004: 35). Before examining the reasons for the emergence of the -zd-/-d- alternants in the Baltic languages and the possibilities for reconstructing their proto-forms, the areal boundaries within which one form or another is used should first be defined. In old Lithuanian texts, the word barzdà ‘hair on the lower part of the face; tuft of hair on a horse’s fetlock; small jaw’ (LKŽ₁ 673 ff.) is attested only with the cluster -zd-: barłdą (acc.sg.) BB Ez 5,1; Barzdà, -õs f. ‘beard’ R₁ 13, R₁₁ 52, R₂₁₉; Gaidžio Barzdà ‘comb’ MŽ₂ 281 = R₁₁ 211; Barždà, -õs f. MŽ₂ 27; Barzda SD⁵ 19; barzdà ‘scythe blade, turned over before beginning to cut the grass’ MŽ² 292; Strėlà su Barzdà ‘arrowhead’ R₁₁ 276 = Strėlõs barzdà K₁₁ 107; barzdà BzF 99; barzdà K₁ 167, 675; bárzdà M 28; barzdà Leskien B 214; Barzda, bardzėlė ISPL 84; barzda ‘beard’ ZCh 183₉, 184₇...; barzd’õs (gen. sg.) P 230, bard’õms (instr. pl.) V 444 (Buch 1961: 41); or with -z-: barze MPDP 149; barza SG II 59; barzā, -ās Jrk 62; barza BsP I 99. On the basis of data from LKŽŽK, LKŽŽKP and TA, it can be stated that the form with the cluster -zd- is used over a considerably wider area: barzdà—in Western Samogitian (SG), Northern Samogitian Kretinga subdialects (bazdà PLT, bárzda PLN, barzdà ŽEM, MS), Telšiai subdialects (ŠARK [bárzda EIG, LK; barzda VGR, SKD, TRŠ]), Southern Samogitian Varniai and Raseiniai subdialects (bárzda ŠLL, RDN, KRKL), Western Aukštaitian Šiauliai subdialects (bárzda Sk, Škt; barzda Sk, Kdn), Kaunas subdialects (ZP, JRB [bazdà BLB]), Eastern Aukštaitian Panevėžys subdialects (RD, MLN [bárzda SLC; barzda KRIAUŠ, GĖL, ANG, PS, NMR]), Kupiškis subdialects (KP [barzda PD, DRŠ]), Anykščiai subdialects (SV [barzda AND, LDK]), Utena subdialects (LLN [barzda UŽB, ŠPL]), Vilnius subdialects (DKŠT), Southern Aukštaitian (VS, AUK, DRSK, DG, VRN [bazdà VLKN, MCL]), and peripheral dialects (LZ, PLS). Forms with root -z- are used in a considerably smaller area: ERŽ, VDŽ, ŠAUK, KRŽ—barzà, VLKŠ, VDK, J-E; PGR⁶) (TA), GRV (Kardelytė 1975: 114); see also LKT 102, 442. LKŽK records that such forms are used in the area around Jurbarkas (GRDŽ— 4 Toporov explains the absence of -z- in Prussian bordus ‘beard’ E 101 by the use of other PIE words meaning beard without a fricative pre-consonant: Old Church Slavonic brada, Old High German bart, etc. (PJI 241). On the basis of Prussian Barsde (name of a deity), Prussian Claus Barsde (proper name—Trautmann 1925: 17; PJI 199), Curonian Barsdemin (proper name), and the Upper Dnieper place-name Borzna < *Borz(d)na (?) [origin and formation unclear], Toporov cautiously reconstructs Prussian *barsd-/barzd- (PJI 197). 5 Kaukienė derives Prussian bordus from *bārdzā and cites PEŽ₁ 150, but Mažiulis does not make this claim. On the contrary, he posits a secondary -z- resulting from the influence of Slavic *borzda. 6 Jonikas notes that in the Pagramantis dialect the consonant cluster zd becomes z in all words, not only in the form under discussion, baržà (Jonikas 1939: 28). 18 ROLANDAS KREGŽDYS Zinkevičius 1966: 179) ir Skirsnemunėje. Zinkevičius taip pat užsimena, kad barzdą vartojama ir rytiniame lietuvių kalbos ploto pakraštyje – Lazūnuose (Zinkevičius 1966: 180). Vadinasi, forma be -d- vartojama ne tik pakraštinėse pietryčių žemaičių šnektose, bet ir pietų žemaičių raseiniškių šnektoje bei su ja besiribojančiame vakarų aukštaičių pakraštyje, taip pat vakarų aukštaičių kauniškių, rytų aukštaičių vilniškių paribio Lazūnų šnektoje. Fonetškai toks reiškinys, matyt, apibūdinamas kaip progresyvinės asimiliacijos nulemtas kitimas, plg. liet. dial. lazã = liet. b. k. lazdà ir kt. (plg. Zinkevičius 1966: 180). MAP 1: BLACK TRIANGLES MARK THE USE OF FORMS WITH THE CLUSTER -zd-; WHITE ONES MARK FORMS WITH -z-. Some Latvian linguists treat morphological mismatches between Lithuanian and Latvian forms of barzdos as the influence of Lithuanian dialects on Latvian border dialects. This presupposes that forms with root -z- are later (Staltmane 1981: 65; Steinbergs 1996–1997: 24), and the origin of Latvian personal names containing the first component bārd- is associated with the process of appellativization, since they lack the pejorative connotation characteristic of the corresponding Lithuanian examples: Latv. bārda ‘beard’ → bārdiņa ‘little beard’ → Bārdiņš (surname), Bārdainis ← bārdainis ‘bearded’ (-zd- is explained as a dialectal feature). It should be noted that Staltmane’s claim that surname variants with -zd- could have determined the emergence of Latvian dialectal bārzda is difficult to prove (cf. Staltmane, ibd.). This lexical variant may have been used alongside Latv. bārda or may be treated as an archaism (cf. Nilsson 1998: 124). Of course, this way of solving the problem would be more convincing if these forms were used in Latvian only in border dialects (e.g., Bsk (235)); however, such forms are used far from the Lithuanian border: bārzda² – KND (84), VNP, ARLV (65), LIET. barzdà SAMPLAIKOS -zd- KILMĖ 19 b'ʒrda² – PRE (439), KRTP (357), barzdə² DND (63), VND (75) (also cf. Latvian dialectal būorzda ‘beard’ (the Livonian subdialect) – an example from www.ailab.lv – see below), or are separated by 2–3 dialects: b'ʒrda² – SNK (353) (ME, 273); also cf. Latvian dialectal būorzda ‘beard; a long tuft of horsehair near the legs; a long tuft of bristles on the lower part of an animal’s body’ (KV, 167 ff.), and examples from Latvian folklore: Borzda izauga, a gudreibys nav 527 33449; Boarzda koa krīvam 1268 793 (an example from the Selonia subdialect area, where ā (-'r- < -ār – LEV₁ 108) > oa (Rudzīte 1993: 237). MAP 2: FORMS WITH THE CLUSTER -zd- ARE USED IN THE INDICATED AREAS; TRIANGLES MARK LOCATIONS WHERE SUCH FORMS MAY HAVE BEEN USED (BASED ON LATVIAN FOLKLORE COLLECTIONS). A form with the root cluster -zd- is used in Latvian and Lithuanian toponymy: Latv. Bārdzaini, Bārdzēni < nom. prop. Latv. Bārzda ~ Lith. Barzdai, Barzdinė (Endzelīns 1956: 93). Toporov’s theory on the emergence of doublet word forms also contradicts this solution to the problem: according to it, traces of forms with root -d- (rather than -zd-) should also be found in Lithuanian dialects (in the Latvian border area)—yet no form *barda (without -z-) has been recorded on the Lithuanian-Latvian border. Of all Lithuanian proper names with roots related to barzda, only surnames Bardišius, Bardyšius (used in various regions of Lithuania—Maciejauskienė 1991: 98) might reflect the root *bard-; however, these words are treated as borrowings from Slavic languages, i.e., as deriving from Pol. Bardyszewski ~ Pol. bardysz → Lith. bardišius ‘long-handled axe for chopping enemies; a blunt tool; a large man; a poor-quality animal’ (LPŽ₁ 194; LKŽ₁ 657). Also cf. Lithuanian toponyms Bardiškėliai (UKM), Bardiškiai (PKR, RS), Bardžiai (ŠLL), and others (LAŽ 27), without root -z-, which may be 20 | ROLANDAS KREGŽDYS of anthroponymic origin and related to the Polish substratum mentioned above, such as the toponym Barzdūnai (village, ŠLU), with the Lithuanian personal name Barzdūnas (Razmukaitė 1993: 154). Similarly, the origin of the Lithuanian surname Bardžius (ALVT), used alongside Barzdžius (cf. Bardzis Ukм) (LPŽ 195), could probably be explained in this way, although it may be treated as the only case of -z- loss in the Lithuanian-speaking area. Būga explains the emergence of morphonological variants of Barzdos as a linguistic feature of the Southeastern Samogitians: simplifying the word structure according to whether the sound sequence was morphologically significant: “zd sokratilos’ v z vo vsech tech formach, v kotorych d ne imelo morfologičeskoj cennosti” (ibid.). According to Būga, examples of similar change in other dialects, where fricative -z- is used instead of the expected -zd-, should be explained as sporadic occurrences. The -zd- cluster in all these words derives from Proto-Lithuanian *-zd- (Būga 1958: 273). Mažiulis (PEŽIII 55) and Derksen (2003: 12) agree with this solution. Endzelīns (1974: 267–268) also discusses the old use of the cluster. Zinkevičius simply states that “some Southern Samogitians and their closer Aukštaitian neighbors do not pronounce d between z and a vowel,” and that “the loss of the consonant d between z and a vowel has also been observed at the eastern edge of the Lithuanian-speaking area” (1966: 179 ff.). One argument that could support the Proto-Baltic origin of the cluster *-sd- might be toponymic examples showing that forms with the cluster -zd- are specifically the older ones: lazda – Lazdėnai, Latv. lazda ‘hazel’ – Lazduone (see also Gerullis 1922: 254). The Lithuanian form *Lazėnai does not exist. Zinkevičius reasons similarly—in illustrating this phonetic change, he cites the Lithuanian dialectal adjective barzyla ‘one with a beard’ (JRB) and the Lithuanian dialectal verb barzuoti < Lith. barzdūoti ‘to rub something with a beard’ (Zinkevičius, ibid.). A convenient example to demonstrate the complexity of this problem may be the verbal pair Lith. bėžda – bėža, where -zd- is etymological (cf. Jakulis 2002: 87, 107, 109); therefore, it is unreliable to suppose that some innovation caused by phonological or semantic development is involved here. Jakulis attempts to explain the emergence of these sound variants by attaching the extension -d- to the Indo-European root *pes- or *bes- (Jakulis op. cit., 109), although it is doubtful that such a claim can be reliably formulated on the basis of Lithuanian pėža, -ėti, -ėjo alone, especially when reconstructing a Proto-Indo-European form: examples from other languages imply that this infix was already present in Proto-Indo-European (LEW 42; Vasmer I 163). Examples retaining the stop consonant, where the fricative /z/ is clearly a parathetic consonant, are less common: spùrzda – spùrda ‘spūrda’. Incidentally, Jakulis does not examine the causes of this phenomenon in greater detail, but notes that the use of consonant clusters such as -zd- alongside -z- may be determined by a semantic factor—a function of iconicity is presupposed (Jakulis 2002: 42). Thus, it can be concluded that in Lithuanian dialects the cluster -zd- can become -z-. In the nominal system, this change may be associated with the phonetic principle of “easier pronunciation,” provided, of course, that it does not violate the rules of Lithuanian morphology; cf. barskutis ‘razor’ DP 563₆ < *barzd-skutis (Otrębski 1958: 367). In Latvian, the emergence of a form without -z- may have been caused by the same factors invoked to explain the emergence of -z- in Lithuanian barzdà. One may suppose that the forms without the cluster -zd- used by neighboring Slavic and Finnic peoples may have brought about the emergence and establishment of Latv. bārda, and the displacement of the old forms with the cluster -zd- from usage LIET. barždà SAMPLAIKOS -zd- KILMĖ 21 nos. This assumption is supported not only by the use of forms with -zd- in the Livonian dialects, but also by its particularly frequent occurrence in old Latvian folklore collections published on the website www.ailab.lv. The ratio between forms with the clusters -zd- and -d- could be estimated at 2.3 to 1—the form with -zd- is used much more frequently7. Thus, the changes Lithuanian -zd- > -z-, Latvian -zd- > -d- can tentatively be linked to the principles of linguistic “convenience” and “economy,” as well as to the influence of other languages. All these factors may have determined the emergence and spread of the variants in question in different dialectal areas of the same language. Therefore, such phonetic variants cannot be ancient or directly linked to similar words in other Indo-European languages meaning beard, whose formation may have been influenced by entirely different factors; compare Fraenkel’s attempt to archaize the problem under consideration as an Indo-European one: Old High German bart ‘beard’: burst ‘bristles’, Old English beard ‘beard’: byrst ‘bristles’, which may have different derivations (of verbal origin, with -st-?) (LEW 36). The influence of neighboring Slavic and Finnic languages on the variation in the Baltic name for beard should be discussed separately and in depth, since data from these languages are used in attempts to seek evidence for the Slavic origin of the cluster -zd-. First, it should be noted that proponents of the hypothesis that the cluster -zd- in Lithuanian barždà is of Slavic origin emphasize only the importance of the morphological factor, which may be an example of accidental coincidence—the lexical-semantic paradigms of these nouns do not match. Compare Old Church Slavonic brazda ‘furrow’, Bulgarian бразда ‘same’, Serbo-Croatian brázda ‘same’, dialectally ‘ditch full of water’, Slovak brázda ‘furrow, wrinkle’, Polish brózda ‘furrow’, 7 [proverbs]: Bārzda aug, bet pats neaug 1225 12801; 699 191; Strīdās kā gar ķeizara bārdu 382 39; [songs]: <… Piecī gadi bārzdu dzina …> (14999); <… Ka žīdam liela bārzda…> (34300); [tales]: <…such a bearded man, such a fool; <… an old man with a gray beard …> (The animals go to confess their sins and fall into a pit. 1. A. 1. 47. 20. 31. R. K. Porietis Šelpis Aizupēs, Zin. Kom. kr. LP, VI, 43, 1. AŠ, II, 3c.); <… You have no beard, so I cannot let you in …> (The Goat and the Wolf. 123. L. Ivanovskis in Kapiņi, Daugavpils district, in N. Rancāns’s collection); <… an old man with a gray beard …> (Wolves tear apart the unfortunate man. 1. LP, V, 83, 1. AŠ, II, 41c.); <… entered Spičbārdis …> (The hero rescues three abducted princesses. 9. A. 301 A. A. Māturs in Grāvendāle. I.P, VI, 115, 15 d. note AŠ, II, 86 f.); <… a little man with such a long copper-colored beard …> (The hero rescues three abducted princesses. 10. A. 301 A. P. Lielupis in Vec-Piebalga, Brīvzemnieks’s collection LP, VI, 115, 16. AŠ, II, 86 h.); <… On the third day, the strongest man remained, and he … the old man’s beard> (The Beautiful Daughter of the King of Zemgale. 5. A. 301A. J. Ozols in Sauka, Brīvzemnieks’s collection LP, VI, 115, 24. Continued AŠ, II 86 b); <… a little old man with a very, very long beard …> (The Strongman and His Companions. 3. A. 301B. J. Ozols in Sēce, Brīvzemnieks’s Sborņik, 99 note. LP, VII, II, 18, 6.); <… a man with such a long beard …> (The Strongman and His Companions. 4. A. 301 B. (650 A.) J. Bergmanis in Bārta, Magazin der Lett.-Lit. Gesellschaft, XVIII, 1887, 17.); <… a man with a very long beard …> (The Strongman and His Companions. 12. A. 301B. T. Štaūeris in Iecava. LP, IV, Iv, 5. AŠ II, 87e.); <… a little man comes crawling along with a very, very long beard …> (The Strongman and His Companions. 13. A. 650 A. 30t B. J. Gobziņš in Durbe. LP, IV, 1, 4 v. AŠ, II, 8i f.); <… the lad seized a span and will nail his beard to a tree with a peg …> (The Strongman and His Companions. 16. A. 301B. L. Makevica in Nīca. LP, VI, 115, 24 continued. AŠ, II, 87 d.); <…a little man with a long white beard …> (The Blacksmith’s Son. 6. A. 301B. Nameitu Juris, Franka-Sesava. Jkr. II, 2. LP, VII, II, 18, 4, 3. AŠ, II, 92 a.); <… a little man like a beet, with a beard longer than a horse’s tail …> (The Iron Son. 2. B. 301A. In Dzirciems, LP, VI, 115, 19 note.); <… a tall little man with a cubit-long beard …> (The Iron Son. 5. A. 30113. J. Saržants in Garlene. LP, VII, II, 18, III, l.); <… a little man the size of a thumb and with a very long beard …> (The Bear’s Son. 4. A. 301B. M. Šimiņš (Pērkons) in Brukna, Kurzeme. Jkr. II, 1. LP, VII, II, 18, 5. AŠ, II, 90d.); <… with a cubit-long beard and a loaf of bread …> (The Bear’s Son. 8. A. 301B. Pogu Jānis, Bechmanis, I879. In Bramberg and Ezere, Brīvzemnieks’s collection LP, VI, 115, 10a. AŠ. II, 90h.); <… a little man with a long beard …> (The Mare’s Son. 3. A. 303. 300. 30IA. 513A. Lauris in Biksti. LP, VI, 115, 2b. AŠ, II, 92c.) there are several hundred other examples as well, too numerous to present here. 22 | ROLANDAS KREGŽDYS rus. borozda ‘arklio išvagota eilė’ < prasl. *borzda (Brückner 1957: 42). Toks spėjimas būtų kur kas patikimesnis, jei šis slavų vardažodis kuriame nors dialekte reikštų „barzdą“ ar kokį nors kūno organą. Deja, tokių duomenų nėra (plg. ESSJ₂ 220). Suponuoti, kad semema „vaga“ kažkaip susijusi su nagrinėjamos kūno dalies išvaizda taip pat negalima, kadangi slavų kalbose iš šios šaknies pasidaryti kūno dalių pavadinimai išlaiko savo denotatinę vertę: s.-ch. (čak.) brazgotina ‘randas’, slov. brazgotina ‘t. p.’ < s. sl. brazda < prasl. borzda (fonetinė kaita -zd- > -zg-) (Kurkina 1982: 182). Vadinasi, slavų kalbose išlaikomas denotatinis dėmuo „tai, kad įrėžta; išvagota“, kuris visiškai nesusijęs su liet. barzdà denotatu „apatinės veido dalies plaukai“. Proponents of Slavic influence often rely on the protosememe “plant” reflected in these languages; compare Russian boroda ‘beard’, which in Russian dialects can mean: 1. “a swath while mowing or harvesting rye (and other cereal crops)” (Pskov region); 2. “the main prop in the East Slavic ritual devoted to the ‘beard’ (associated with harvesting grain)—an artificial beard made of straw”; 3. “the last sheaf of grain or rye” (Ternovskaja 1977: 80ff.). This is thought possibly to be connected with a mythological beard made from the last sheaf of grain or rye during the harvest festival. This ritual was particularly characteristic of Slavic peoples; compare the Russian expressions zavit’ Velesu borodu and zavit’ borodu ‘to finish the grain (rye) harvest’ (Dal’ I 116). It is conjectured that during sacred rites a beard was attached to a person (apparently a priest); the ritual symbolized an anthropomorphized deity (for a fuller discussion, see Ivanov 1985: 163ff.). It is claimed that a similar ritual was also practiced in Baltic rites (Greimas 1990: 479, 484f.), but its influence on changes in the word’s morphological structure has not been demonstrated. It is difficult to believe that a similar naming process associated with the mythological beard also occurred in Germanic dialects—Old High German burst ‘bristles’ is semantically much closer to the East Baltic examples meaning beard than to the Slavic words belonging to other semantic paradigms. The morphonological hypothesis of a Slavic substrate is also not very convincing, since it does not take into account the structure of words in other Indo-European languages: Sanskrit bhrṣṭís ‘blade, edge’, Old Icelandic and Old High German burst ‘bristles’, Old English byrst ‘same’. It is therefore difficult to deny the possibility that Indo-European *bhars-dhā existed. Admittedly, Germanic root -s- has so far been ignored in reconstructing Proto-Germanic *barda- (Kluge 1975: 54). Karulis, following Endzelin’s (DI₁ 251) and Vasmer’s conjecture about the antiquity of the cluster -zd-, indicates that its origin may be linked to the development of -s- into the Indo-European proto-form (LEV ibid.). This view is much more plausible, since forms without root -z- are not used in Lithuanian; by contrast, in some dialects root -d- and -r- are lost, but never -z- (see the Lithuanian dialect forms baržà, bazdà discussed above). Thus, there is no basis for reconstructing Proto-Lithuanian *bardā (Trautmann 1923: 27; Mažiulis I 151). Nor can the possibility be ruled out that a form with -z- existed in Prussian as well (see footnote 4). On the basis of the linguistic facts described here, one might suppose that it was precisely the Lithuanian nouns meaning beard that preserved the older morphological structure, and therefore the Slavic- LIET. barzdà SAMPLAIKOS -zd- KILMĖ 23 The factor of analogy of the former type is applicable to Latvian examples, while dialectal forms of the latter language with -z- should be treated as archaisms. Another important issue is how Baltic loans in Finnic languages should be treated, since it is assumed that the Finns borrowed the name for barzdos from the Balts: Suo. parta ‘beard’, Est. pard ‘the same’, Veps. bard ‘the same’ (Sabaliauskas 1966: 6; 1990, 9; PJĄ₁ 242 with references). To support the authenticity of forms with -d- and, correspondingly, the secondary origin of -zd-, the Finnic form parta ‘beard’ is invoked; its borrowing from Baltic is apparently beyond doubt. Nevertheless, this is not the only name for barzdos in Finnic languages. The Livvi dialect of Karelian uses the form barde ‘beard’ (Barancev 1967: 8). Liudikų tarmė lokalizuojama Onegos ežero baseino ribose (Bubrich 1971: 4), t. y. ribojasi su rusų kalbos šiaurinėmis šnektomis, kurių įtaka ypač ryški šios tarmės leksikos darybos kaitai. Labai svarbi aplinkybė fonetiniam skolinio interpretavimui yra tai, kad šioje tarmėje nėra priešbalsių duslusis ↔ skardusis kaitos (konsonantinis sinharmonizmas) (Bubrich op. cit., 5). Todėl kaip tik šioje srityje galima lokalizuoti senesnės formos vartojimo pavyzdžius nei Pabaltijo finų, kuriems ši kaita būdinga. Thus, one would have to conclude that two different loanwords exist in the Finnic group for naming the same object. This would be difficult to believe. Since the adaptation of this loanword is not entirely clear, one might suppose that Livvi barde ‘beard’ came not from the Baltic languages but from other neighboring peoples. There are several possible sources of borrowing: northern Slavic dialects and northern and western Germanic dialects. These loanwords may have entered the Livvi dialect via the Veps, from whom they are also derived (Salo 1971: 57; Serebrennikov 1971: 25, 30). Analysis of the borrowings in Barzdos has not been carried out—A. Sabaliauskas very cautiously considers suo. parta, Est. pard, and Veps. bard to be Baltisms (1963: 124), perhaps for this reason not even mentioning them later (1966: 65). Here are the reasons why they should be discussed as thoroughly as possible, since whether they can be treated as Baltisms depends on this. They may be Germanisms: German Bart ‘beard’ (< Proto-Germanic *barda- ‘ibid.’, Kluge 1975: 54)8. The Livonian example cited here casts doubt on the Baltism hypothesis, since Est. pard, (gen.) parra ‘beard’ is certainly not a form used by Livonians. How, then, can the emergence of the word used in the Livonian dialect be explained? It is difficult to believe that it could have given rise to Russian borodá ‘beard’, i.e. barde differs morphologically greatly from this word, while the morphonological alternation -oro- > *-or-9, later becoming -ar-, is not characteristic of Finnish loans; compare the example of the CVRC structure in Baltic Finnic laudus ‘wooden fishing-net float’ < Russian dial. lovdus, lovdas, lovda, lovtus ‘ibid.’ (Matveev 1959: 14). If we assume that Veps. bard ‘beard’, phonologically closest to Livvi barde, could have determined the emergence of the name for barzdos in the Livvi dialect, the first thing to explain would be— 8 Compare also Norwegian-Danish Bart, New High German Bart (Middle High German bart) = Old High German bart, Anglo-Saxon beard, Old Scandinavian barð (neuter) ‘beard, hem, edging, edge, rim’ (NDW 52). 9 Vowel alternation in the roots of loanwords in Finnic languages is very common—/o/ may change not only to /a/ but also to /e/ (cf. Kregždys 2003: 34). Whatever the cause of this alternation, it could be substantiated by knowing the regularities of this dialect’s morphonological structure. Endzelin established that in Finnic loanwords of Slavic origin, /a/ occurs in place of the Slavic base /o/ (Endzelin 1911: 106f.). 24 | ROLANDAS KREGŽDYS to correct the inflectional mismatch10 -V: -Ø11. Thus, it is possible to cautiously conjecture that the attested bardė is indeed the original borrowing, having best preserved the phonetic structure of the borrowed word. It is also possible to conjecture that it was borrowed a very long time ago and reflects the old Germanic *bardą. Of course, this is not easy to prove, since A. Sabaliauskas also points out that “it is most difficult to establish the precise authenticity of one borrowing or another because here we are dealing with very long time spans” and that “the Finnic languages of the Baltic region are very different in structure, and a word that passes from one language to another is subject to the laws of the new language over such a long period, changing so much that it becomes impossible to recognize” (1963: 130; 1966: 67). Nevertheless, the principle of borrowing adaptation in the Livonian dialects presented here—preserving a word’s morphological (and phonetic) structure as authentically as possible—strengthens the evidential value of the hypothesis. It is therefore a much more reliable assumption that these borrowings are of Germanic rather than Baltic origin. This would explain the vocalic form of the Livonian bardė, discussed in greater detail here, although the word-initial consonantism (cf. Vepsian bard) still requires a separate explanation. It is most likely necessary to posit that Finnish parta underwent consonant harmony twice: first, Germanic bart(a)12 (< Germanic *barda, Pokorny 110) became *para (the Estonians borrowed this form from them), and later, due to the aforementioned consonant harmony, it developed into part. Thus, it is possible to cautiously conjecture that Karelian dialect bardė is certainly not a borrowing from the Latvian or Lithuanian linguistic area, where examples reflecting the cluster -zd- predominate. The problem of adapting Finnic borrowings should also be examined from a temporal perspective; however, this requires a detailed internal phonological analysis of Finnic dialects, which is lacking in the aforementioned descriptions of the dialect (Livonian). K. Būga gives some phonetic differences among Finnic borrowings, characteristic of the same borrowing that came into use at different times (1959: 233ff.), but these are not informative for the problem under consideration. This problem may be approached in much the same way as examples from East Baltic that reflect the doublet formations -z-:-zd- used in different dialect zones. The reasons for the emergence of these variants may indeed be sought in the lexicon of other languages, although the factors that could have brought them about should first be assessed—the aforementioned semantic factor, the principle of linguistic economy, and the possibility of simplifying the difficult-to-pronounce cluster -zd- (this was said with reference to the diversity of the phonetic structure of words used in Lithuanian dialects). In conclusion, it can be stated that the causes and usage of denominal alternants with root -zd-/-z-/-d- are still poorly understood. There is considerable uncertainty about the morphological structure of some words; cf. Lithuanian skruzdìs, -ìės ‘ant’ – Lithuanian dialect (Eastern Highlanders) skrudė ‘same’, skrud-ėlė (Ds, LNKM), where -zd- is interpreted differently by various authors: Būga unequivocally claims that Proto-Lithuanian *skrus, gen. 10 The Livonian inflectional -ę in bardė is a front mid unrounded vowel, more open than /e/, sometimes also represented graphically as ä (Barancev op. cit., 16). 11 This Livonian inflectional -ę may reflect a considerable archaism, i.e. the word may have preserved a much more archaic form than the other Finnic examples; cf. Germanic *barda, ‘beard’. Thus, there is no need to connect the borrowings with Baltic inflectional examples (for more on the adaptation of endings in Baltic borrowings, see Dolja, Suchanova 1971: 53). 12 Cf. the examples: Old Icelandic barða ‘axe’ and Old Saxon barda ‘beard’ (Pokorny ibid.). LIET. barzdà SAMPLAIKOS -zd- KILMĖ 25 *skruzdės (1959: 368), therefore posits a proto-form with -zd- (1958: 289), although forms without -z- are abundant in the old written sources (cf. LEW 819; Ambrazas 2000: 54). Fraenkel (“Im Ostlit. kommen z-lose Formen vor...”) follows this line of reasoning as well, but Karulis disagrees (LEV₂ 208f.). It is possible that the interpretation of this consonant cluster also affects the explanation of the word’s origin. 3. CONCLUSIONS 1. The morphophonological variants of the name for beard, with -zd-: -z-, may be explained not only by the tendency in some dialects to simplify word structure depending on whether the consonant cluster is morphologically significant, but also by the principles of economy of speech and ease of pronunciation. 2. The spread and use of Latvian words for beard in old written sources and folklore allow us to posit that an older form had -zd-, rather than, as has been thought until now, -d-. 3. The argument that the -zd- cluster in Lith. barzdà is secondary in origin, linking this noun with Finnic loanwords, is highly questionable, since the words meaning “beard” in the Baltic Finnic languages may be of Germanic origin. ABBREVIATIONS ARLV – Ārlava (65) BSK – Bauska (235) DND – Dundaga (63) KND – Kandava (84) KRTP – Krustpils (357) PRE – Preiļi (439) SNK – Sunākste (353) VND – Vandzene (75) VNP – Ventspils ABBREVIATIONS OF SOURCES DLKŽ⁵ – S. KEINYS, ed., Dictionary of Contemporary Lithuanian (computerized version). ESSJ – O. N. TRUBAČEV, ed. Etimologičeskij slovar’ slavjanskich jazykov. Moscow: Nauka, 1974–. KV – A. REKĒNA. Kalupes izloksnes vārdnīca 1–2, Riga: Latviešu valodas institūts, 1998. LAŽ – Directory of the Administrative-Territorial Division of the Lithuanian SSR 2, Vilnius, 1976. LEV – K. KARULIS. Latviešu etimoloģijas vārdnīca 1–2, Riga: Avots, 1992. LEW – E. FRAENKEL. Lithuanian Etymological Dictionary 1–2, Heidelberg, 1962–1965. LKŽ – Dictionary of the Lithuanian Language, Vilnius, 1941–2002. LKŽK – The Lithuanian Language Institute’s Dictionary of the Lithuanian Language card index. LKŽKP – Supplements to the Lithuanian Language Institute’s Dictionary of the Lithuanian Language card index. LPŽ – Dictionary of Lithuanian Surnames 1–2, Vilnius, 1985–1989. ME – K. MŪLENBACHS. Latviešu valodas vārdnīca 1–4. Edited, supplemented, and continued by J. Endzelīns. Riga, 1923–1932. 26 | ROLANDAS KREGŽDYS PEŽ – V. MAŽIULIS. Prūsų kalbos etimologijos žodynas 1–4, Vilnius, 1988–1997 NDW – H. S. FALK, A. TORP. Norwegisch-Dänisches etymologisches Wörterbuch 1, Oslo–Bergen: Universitetsforlaget, 1960. PJa – V. N. TOPOROV. Prusskij jazyk: Slovar’ 1–5. Moskva: Nauka, 1975–1998. TA – Index-card file of the Dialect Archive of the Institute of the Lithuanian Language. LITERATŪRA AMBRAZAS, S. 2000: Daiktavardžių darybos raida 2, Vilnius: Mokslo ir enciklopedijų leidybos institutas. BARANCEV, A. P. 1967: Fonematika ljudikovskogo dialekta karel’skogo jazyka. Pribaltijsko-finskoe jazykoznanie, 3–19. BREJDAK, A. B. 1970: Vlijanie pribaltijsko-finskich jazykov na latgal’skije govory Ludženskogo rajona Latvijskoj SSR. Vzaimosviazi baltov i pribaltijskich finnov, Riga: Zinātne, 157–164. BRÜCKNER, A. 1957: Słownik etymologiczny języka polskiego, Warszawa: Wiedza Powszechna. BUBRICH, D. V. 1971: Russkoe gosudarstvo i sformirovanie karel’skogo naroda. Pribaltijsko-finskoe jazykoznanie, 3–22. BUCH, T. 1961: Die Akzentuierung des Christian Donelaitis, Wrocław–Warszawa–Kraków: Ossolineum. BŪGA, K. 1958, 1959, 1961: Rinktiniai raštai 1–3. Sudarė Z. Zinkevičius, Vilnius: Valstybinė politinės ir mokslinės literatūros leidykla. DAL’, V. 1989: Tolkovyj slovar’ živogo velikorusskogo jazyka 1, Moskva: Russkij jazyk. DERKSEN, R. 2003: On the reception of Winter’s law. Baltistica 37 (1), 5–13. DOLJA, T. G., SUCHANOVA, V. S. 1971: Struktura pribaltijsko-finskich zaimstvovannych slov v russkich dialektach Karelii (imena suščestvitel’nye). Pribaltijsko-finskoe jazykoznanie, 52–56. ENDZELIN, J. 1902a/1971: Lettische Etymologien. Bezzenbergers Beiträge 27, 329–331. Cituojama pagal: id., Darbu izlase 1, Rīga: Zinātne, 251–252. ENDZELIN, J. 1902b/1971: Issledovanija v oblasti kašubskogo jazyka. Sbornik statej, posvjaščennych.... Fortunatovu, Varšava, 535–572. Cituojama pagal: id., Darbu izlase 1, Rīga: Zinātne, 214–242. ENDZELIN, J. 1911/1974: Slavjano-baltijskie etjudy, Char’kov. Cituojama pagal: id., Darbu izlase 2, Rīga: Zinātne, 167–354. ENDZELINS, J. 1956: Latvijas PSR vietvārdi 1, Rīga. ENDZELY̌NAS, J. 1957: Baltų kalbų garsai ir formos, Vilnius: Valstybinė politinės ir mokslinės literatūros leidykla. GAMKRELIDZE, T. V., IVANOV V. V. 1984: Indoevropeiskij jazyk i indoevropejcy 1–2, Tbilisi: Universiteta. GERULLIS, G. 1922: Die altpreußischen Ortsnamen, Berlin–Leipzig: Walter de Gruyter. GREIMAS, A. J. 1990: Tautos praeities beieškant. Apie dievus ir žmones, Vilnius-Chicago: Mokslas. IVANOV, V. V. 1985: Indoevropejskie etimologiji. Etimologija, 160–166. JAKULIS, E. 2002: Lietuvių kalbos tekėti, teka tipo veiksmažodžiai ir jų istorija. Dakтaro disertacija, Vilnius. JONIKAS, P. 1939: Pagramančio tarmė, Kaunas: Švietimo ministerijos leidybos komisijos leidykla. KARDELYTĖ, J. 1975: Gervėčių tarmė, Vilnius: Mintis. KAUKIENĖ, A. 1999: Prūsų kalbos u-kamienas. Vakarų baltų kalbos ir kultūros reliktai 3, Klaipėda: Universiteto leidykla, 20–30. KLIMOV, G. A. 1988: Dopolnenija k etimologičeskomu slovarju kartvel’skich jazykov 3, Etimologija, 151–165. KLUGE, Fr. 1975: Etymologisches Wörterbuch der deutschen Sprache, Berlin–New York: De Gruyter. KREGŽDYS, R. 2003: Dėl liet. sparnas ir Pabaltijo finų perna ‘bluźnis’. Baltistica 37 (1), 34. KREGŽDYS, R. 2004: Prūsų nom. prop. Borssythe. Tarptautinė Aleksandro Vanago konferencija: Baltų onomastikos tyrimai. Pranešimų tezės, Vilnius, 35. LIET. barzdà SAMPLAIKOS -zd- KILMĖ | 27 KURYŁOWICZ, J. 1935: Études indoeuropéennes 1, Kraków: Drukarnia Uniwersytetu Jagiellońskiego. KURKINA, L. V. 1982: [rec.] Blaž Jurišić, Rječnik govora otoka Vrgade. Etimologija, 1980, 16–30. MACIEJAUSKIENĖ, V. 1991: Lietuvių pavardžių susidarymas, Vilnius: Mokslas. MATVEEV, A. K. 1959: Finno-ugorskie zaimstvovanija v russkikh govorach severnogo Urala, Sverdlovsk: Tipolaboratorija UrGu. MEILLET [MEJE], A. G. 1914: Grammatika indo-evropejskich jazykov, Jur’ev: Tipografija K. Mattisena. NEPOKUPNAS, A. 1983: Baltai slavų giminaičiai, Vilnius: Mokslas. NILSSON, T. K. 1998: Baltische Sprache mit deutscher Feder. Smoczyński, W., ed., Colloquium Pruthenicum Secundum, Kraków: Universitas, 120–131. OTKUPŠČIKOV, Ju. V. 1963: Ob etimologii russkich slov borona i boroda. Etimologičeskie issledovanija po russkomu jazyku 4, Moskva: Izdatel’stvo MGU, 87–95. OTRĘBSKI, J. 1958: Gramatyka języka litewskiego 1, Warszawa: Państwowe Wydawnictwo Naukowe. POKORNY, J. 1959–1969: Indogermanisches etymologisches Wörterbuch 1–2, Bern–München: Franke. PORZIG [PORCIG], V. 1964: Členenie indoevropejskoj jazykovoj oblasti, Moskva: Progress. RAZUKAITĖ, M. 1993: Oikonimų, atsiradusių iš asmenvardžių su patroniminėmis priesagomis, paplitimas Lietuvoje. Lietuvių kalbotyros klausimai 32, 133–165. RUDZĪTE, M. 1993: Latviešu valodas vēsturiskā fonētika, Rīga: Zvaigzne. SABALIAUSKAS, A. 1963: Baltų ir Pabaltijo suomių kalbų santykiai. Lietuvių kalbotyros klausimai 6, 109–136. SABALIAUSKAS, A. 1966: Lietuvių kalbos leksikos raida. Lietuvių kalbotyros klausimai 8, 7–142. SABALIAUSKAS, A. 1990: Lietuvių kalbos leksika, Vilnius: Mokslas. SALO, I. V. 1971: Iz istorii vzaimootnošenij vepsov, karel i russkich Karel’skogo Belomor’ja. Pribal’tijsko-finskoe jazykoznanie, 57–62. SEREBRENNIKOV, B. K. 1971: K probleme smešenija dialektov. Pribal’tijsko-finskoe jazykoznanie, 23–32. SKARDŽIUS, P. 1943: Lietuvių kalbos žodžių daryba, Vilnius: Lietuvių kalbos institutas. (= id., Rinktiniai raštai 1, parengė A. Rosinas, Vilnius: Mokslo ir enciklopedijų leidykla, 1996.) STALTMANE, V. Ē. 1981: Latyšskaja antroponimija. Familii, Moskva: Nauka. STEINBERGS, A. 1996–1997: The reconstruction of Proto–Baltic body parts. Linguistica Baltica 5–6, 21–31. TERNOVSKAJA, O. S. 1977: Leksika, svjazannaja s obrjadami žatvennogo cikla. T. M. Sudnik, T. V. Civ’jan, red., Slavjanskoe i balkanskoe jazykoznanie, Moskva: Nauka, 77–130. TRAUTMANN, R. 1923: Baltisch-slavisches Wörterbuch, Göttingen: Vandenhoeck & Ruprecht. TRAUTMANN, R. 1925: Die altpreußischen Personennamen, Göttingen: Vandenhoeck & Ruprecht. VARBOT, Ž. Ž. 1984: Praslavjanskaja morfonologija. Slovoobrazovanie i ètimologija, Moskva: Nauka. VASMER [FASMER], M. 1986–1987: Ètimologičeskij slovar’ russkogo jazyka 1–4, Moskva: Progress. ZINKEVIČIUS, Z. 1966: Lietuvių dialektologija, Vilnius: Mintis. ZINKEVIČIUS, Z. 1980: Lietuvių kalbos istorinė gramatika 1, Vilnius: Mokslas. Rolandas Kregždys Gauta 2004 12 21 Vilniaus universitetas Kauno humanitarinis fakultetas Lietuvių filologijos katedra Muitinės g. 8, 44280 Kaunas, Lietuva