Baltic Loanwords in Saami

51straipsniai / articles Lucie hofírkoVá Masaryk university, brno fields of research: fenno-ugric studies, especially finnish, estonian, saami. VácLaV bLaŽek Masaryk university, brno fields of research: indo-european studies, especially slavic, baltic, celtic, anatolian, tocharian, indo-iranian languages; fennougric; afro-asiatic; etymology; genetic classification; mathematic models in historical linguistics. baLtic Loanwords in saaMi baltų skoliniai samių kalboje annotation the purpose of the present study is to summarize and interpret the lexemes of baltic origin in saami, in the context of fenno-saamic and fenno-Volgaic relations. our conclusion is that about 20% of lexical parallels between baltic and saamic from about 40 balto-fennosaamic comparisons are without the balto-fennic counterparts. it means that it is probable that the ancestors of the balts and the saami were in direct contact. In this study we evaluate the role of words of baltic origin in the saami vocabulary in a wider fenno-saamic and fenno-Volgaic perspective. recent works of finnish authors were the main sources we used, such as aikio (2006, 2009), häkkinen (2010), kallio (2009), koivulehto (2006), sammallahti (1998, 1999), but we also took account of earlier studies of thomsen (1890), itkonen (1961) and korhonen (1981). according to Mikko korhonen (1981), the first contacts between the ancestors of the saami and finnic people began at the time when fennosaamic languages were already clearly separated from fenno-Volgaic languages and were in the beginning of separation into balto-fennic and saami languages at the same time. thus, there was just a slight, dialectal, difference between the baltofennic and saami languages. Vilhelm thomsen (1890) dated these contacts much later as he assumed that the contacts had not begun until the turn of the era or shortly before. erkki itkonen (1961) provided as the latest possible period the keywords: contact, loanword, borrowing, neighbourhood, influence. 52 Lucie hofírkoVá, VácLaV bLaŽek acta Linguistica Lithuanica LXIV–LXV year 500 bc. rather, he was inclined to the opinion that the contacts with the speakers of baltic languages had begun much earlier. according to korhonen, archaeologists are inclined to the opinion that the arrival of the balts to the area of the baltic sea corresponds with the arrival of the corded ware in this area at the end of the third millennium bc. however, this dating does not agree with the above-mentioned statement that the first contacts with the baltic people began after the separation of Proto-fenno-Volgaic (the 19th–18th century bc) (blažek 2010; see apendix). according to korhonen, we may agree with this hypothesis if we take into account several factors that somehow influence this dating. one of these factors is the time elapsed from the arrival of the baltic people to the area of the baltic sea before the beginning of loanwords. it certainly took some time before the two culturally and linguistically very different nations began to influence each other. according to korhonen, the first baltic loanwords were moving into fenno-Volgaic between 1800–1500 bc (korhonen 1981: 32–34). however, it is not necessary to assume that fenno-Volgaic continuum was divided at the time of contacts. Vanagas (1980: 119) defined the upper boundaries of the baltic hydronomic area as follows: the north – the northern boundary of Latvia, Pskov, toropec, Zubcov and kalinin towns; the east – the cities of Moscow, kaluga, orel and kursk; the south – the seim, the Pripjať and the western buh rivers; the west – the wisla river. thus, the baltic people were in the immediate neighbourhood with the ancestors of the Mordva and Mari people in the northeast and the borrowings could have happened later, after the separation of fennoVolgaic, independently on balto-fenno-saamic contacts. this is also supported by classification scheme 2 (see appendix) that reflects a closer relation of the Mari language to Permic languages than to the Mordva language. thus, the presence of baltic loanwords in Mari can be explained as a result of borrowings independent of the genetic relations but resulting from the territorial relationships (five of six baltic loanwords discussed in Mordva occur in Mari too). Proto-baLtic Loanwords 1. saan biebmat ‘to feed’ < Ps *pēmmō (yss: 936) < fs *pämmi < Pb *penima- (sammallahti 1998: 127; 231; aikio 2006: 40). the root *penappears in Lith. pẽnas ‘food’, penjti ‘to feed, to fatten’, the derivate of -mis attested in pẽ nimas ‘fattened’, penmis ‘fattener, fattened pig’ (Lew 569; smoczyński 2007: 449–50). 2. saan buorgg’s ‘forbidden, prohibited’ < Ps *puoɢē < Pres *pō/arkos < Pb *bār gas (cf. Latv. bags ‘strict, unkind, uncharitable’) (sammallahti 1999: 77). baltic Loanwords in saami 53straipsniai / articles however, this comparison is semantically ambiguous; the word could also be of scandinavian origin, cf. sw. dial. bark ‘a mulish unkind man’, barkun ‘rude, gruff’ (Pokorny 1959: 163). 3. saan čiehkat ‘to hide’ (~ finn. säetä id.) < Ps *ćēke (yss 146) < fs *śäke- < Pb *seg()e/a- (cf. Latv. segt ‘to cover; to turn on, to tuck’) (sammallahti 1998: 127; 234). according to fraenkel (Lew ii 770), however, the original meaning of Latv. segt was the same as Lith. sègti ‘to fix, to attach’ and the verb has the meaning ‘to hide’ by prefixation: apsegt = Lith . apsègti ‘to fix’. a likely cognate old Pruss. seggīt ‘to do, to make’ is semantically further. then the whole comparison is unsure. 4. saas daktere ‘daughter’ (~ finn. tytär id.) < Ps *tektēr (yss 1211) < fs *tüktär < Pb *duktē : gen. *dukteres (cf. Lith. duktė, old Pruss. duckti id. (Lew i 110)) (sammallahti 1998: 121; korhonen 1981: 30; thomsen 1890: 167). 5. saan duovli ‘tinder’ (~ finn. taula id.) < Ps *tōvlē (yss 1311) < Pres *takla < Pb *dagla- (cf. Latv. dagla ‘id., birch bark tinder’; Lith. dãglas, dẽglas ‘black spotting’ (Lew i 86)) (aikio 2006: 31; korhonen 1981: 30; sammallahti 1998: 127; thomsen 1890: 165). 6 . saan gahpir ‘cap’ (~ finn. kypärä ‘cap; helmet’) < Ps *kepērē (yss 317) < fs *küpärä < Pb *kepuriā (cf. Lith. kepùrė ‘cap, coif’, Latv. cȩpure id. (Lew i 241)) (sammallahti 1998: 127; aikio 2009: 250; thomsen 1890: 185). 7. saan geardi ‘times’ (~ finn. kerta id.) < Ps *krtē (yss 391) < fs *kerta < Pb *kerdā (cf. old Pruss. kērdan ‘time’ (Lew i 242; toporov iii 315–23), whereas Lith. kartas ‘times’ (Lew i 224))1 stands aside (sammallahti 1998: 243; korhonen 1981: 30; thomsen 1890: 185–86). cf. Mordva Mokša kyrda, erzja kirda ‘times’ (thomsen 1890: 186); Mari kb kerδə in a phrase pülä-γerδə ‘very long time ago’ (uew 659). 8 . saan giehpa ‘soot’ < Ps *kēpw (yss 425) < Pb (cf. Latv. kvêpi ‘soot, smoke, steam ’ , kvêpêt ‘to smoke ’ , lit. kvepti ‘to smell ’ (Lew i 325) ) (sammallahti 1998: 127; aikio 2006: 40). 9 . saan giekka ‘cuckoo’ (~ finn. käki) < Ps *kēke (yss 418) < fs *käke < Pb *gegiā (cf. Lith. gėgė id., that it is explained as a regressive shortening of a longer form geguž  , Latv. dzȩguze, old Pruss. geguse (Lew 142–43)) (?korhonen 1981: 88; thomsen 1890: 172). 10. saan guksi ‘ladle’ (> finn. kuksa ‘small wooden bowl, wooden ladle’, beside the original continuant kauha ‘ladle’ – see skes 172) < Ps *kuksē < fs *kavša < Pb *kaśa (cf. Latv. kaûss ‘large dish, drinking dish; skull; peel’, Lith. 1 semantically more distant is gmc. *χerðō ‘herd’ < Pre-gmc. *kerdhā. 54 Lucie hofírkoVá, VácLaV bLaŽek acta Linguistica Lithuanica LXIV–LXV káušas ‘ladle; skull; shell; snail shell’ (Lew i 231–232)) (sammallahti 1998: 123; thomsen 1890: 184). 11 . saan guoibmi ‘companion’ (~ finn. kaima ‘namesake’) < Ps *kōjmē (YSS 504) < fs *kajma < Pb *kamā (cf. Lith. kaimýnas ‘neighbour’, káima(s) ‘village (of farmers)’, kiẽmas ‘farmstead’; Latv. cems ‘village’, kaĩmiņš ‘villager (resident of the same village), neighbour’; old Pruss. caymis ‘village’ (Lew i 251–52)) (sammallahti 1998: 127; korhonen 1981: 30; thomsen 1890: 177). 12. saan guovllas ‘tag on a dog collar’ (~ finn. kaula ‘neck ’ ) < Ps *kōvle s < fs *kakla < Pb *kaklas and *kaklā (cf. Lith. kãklas ‘neck ’ , Latv. kakls id. (Lew i 205) ) (sammallahti 1998: 127; 1999: 83). 13. saan jávri ‘lake ’ (~ finn. järvi id.) < Ps *jāvrē (yss 258) < fs *jävrä < Pb *jarā- (cf. Lith. jáura ‘marsh ’ , j ū ra ‘sea ’ , Latv. jūra , jūŗa, old Pruss. iūrin id. (Lew i 198)) (sammallahti 1998: 249; aikio 2009: 246). cf. Mordva erzja eŕke, Mokša (j)äR’kä and Mari kb jär, u b jer ‘lake ’ (uew 633). 14. saan leaibi ‘alder’ (~ finn. leppä id.) < Ps *leajpē < fs *lejpä < Pb *lepā (cf. Lith . lepa, -ė, lepas ‘linden’, Latv. liẽpa, -e, lieps, old Pruss. lipe id. ( Lew i 366) ) (sammallahti 1998: 127; aikio 2006: 40). cf. Mordva erzja ľepe, Mokša ľepä ‘alder’ (uew 689). 15 . saan loapmi ‘hole’ (~ finn. loma id.) < Ps *lmē (yss 609) < Pb *lōmā (cf. Lith . lomà ‘hole, depression’, Latv. lãma ‘low-lying place on the field or meadow’ < *lāmā, beside Latv. luôms ‘hole in a fence’ < *lōma- ( Lew i 385; smoczyński 2007: 344–45) ) (sammallahti 1998: 127). cf. Mordva Mokša luv in the phrase käď-luv ‘space between two fingers’; Mari kb lo in phrase parńalo ‘space between fingers’ (uew 692). 16 . saan luossa ‘salmon’ (~ finn. lohi id.) < Ps *lōse (yss 627) < fs *loše < Pb *lašsa- (cf. Lith. lašišà, lãšis, Latv. lasis id., old Pruss. lalasso, i.e. *lasaso ( Lew i 341) ) (sammallahti 1998: 127; korhonen 1981: 30; thomsen 1890: 194). 17 . saan luovdi ‘wooden vessel’ (~ finn. lauta ‘board’) < Ps *lōvtē (yss 634) < fs *lavta < Pb *platā (cf. Lith. plaũtas ‘board lying in the steam bath; side board in the boat’, Latv. plauts ‘shelf’ < *platas (Lew i 608)) (sammallahti 1998: 127). comparison of fs isoglosses with the samoyedic material (uew 239) is less convincing. 18. saan luovvi ‘structure for storing supplies’ (~ finn. lava ‘stand, pillar; bench in a sauna; timbering furnace’) < Ps *lōvē (yss 632) < fs *lava < Pb *lāvā (cf. Lith. lóva ‘bed’, Latv. lâva ‘sweat bench’ (Lew i 387)) (sammallahti 1998: 127; aikio 2009: 262). 19. saai lyepi ‘bladebone’ (~ finn. lapa id) < Ps *lōpē (yss 625) < fs *lapa < Pb *lapā (cf. Lith. lãpas ‘leaf, sheet’, Latv. lapa id., lãpa ‘pad’ (Lew i 339–40)) baltic Loanwords in saami 55straipsniai / articles (sammallahti 1998: 127). other parallels from Mansi, samoyedic languages and yukaghir language (uew 236) question the baltic origin of saami and finnish ‘blade bone’, as well as the semantic difference between fs and b. 20. saaL niehpē ‘husband´s nephew or niece’ (~ finn. dial. nepaa, arch. nepas ‘cousin´s child’) < Ps *npē < fs *nepa- < Pb *nepō(ts), acc. *nepatin (cf. Lith . nepuots ‘grandson, nephew’ < *nepōta- (Lew i 494; smoczyński 2007: 420)) (skes 373; sammallahti 1998: 127; 1999: 84). 21. saan ráigi ‘hole’ (~ finn. reikä id.) < Ps *rājkē (yss: 1019) < fs *rVjkA < Pb *rakā (cf. Lith. riẽkti ‘to cut’, riek ‘slice of bread’, Latv. rekt ‘to cut’, rika ‘slice of bread’, ràika ‘track, furrow’ (Lew ii 729; smoczyński 2007: 514–15)) (sammallahti 1998: 127). 22. saan ruoida ‘leg (of an animal)’ (~ finn. reisi ‘thigh’) < Ps *rōjte ‘thigh’ (yss 1063) < fs *rajte < Pb *retiā (cf. Lith. retas ‘thigh’, Latv. curonian riẽta ‘leg’ < *ret-a-/-ā- (Lew ii 731)) (sammallahti 1998: 127; korhonen 1981: 133; thomsen 1890: 212). Protoform *retiā is supported by Lith. parieti ‘lower part of the perineum’ (smoczyński 2007: 516). 23. saas saertie ‘heart (as food)’ < Ps *sāᴅē (sammallahti 1999: 82) (*sārtē (YSS 1112)) < Pb *śird- (cf. Lith. širds ‘heart’; Latv. sids; old Pruss. seyr id., further eastern Lith. šerds, Latv. sede ‘marrow, pith’ (Lew ii 986–987)) (sammallahti 1998: 127; aikio 2006: 40). 24. saan sápmi ‘saami’ (~ finn. Häme ‘historical region in south-central finland’) < Ps *sāmē (yss 1106) < fs *šämä ‘? earth, land’ < Pb *źemē (cf. Lith . žẽmė, southern dial. žãmė, Latv. zeme; old Pruss. semmē, same ‘earth, land’ (Lew ii 1299)) (?sammallahti 1998: 127, 262; korhonen 1981: 130). 25. saan sarvva ‘elk’ (~ finn. hirvi id.) < Ps *serve (yss: 1091) < fs *širve < Pb *śira- (cf. old Pruss. sirwis ‘roe-buck’, Lith. šivis ‘white horse; hare’ : šivas ‘grey, griseous’ (Lew ii 989)) (sammallahti 1998: 127; korhonen 1981: 30, 130; aikio 2009: 276; thomsen 1890: 225). 26. saan sarvvis ‘reindeer bull’ (~ finn. hirvas id.) < fs *širvas < Pb *śiras – etymologically the same as sarvva (sammallahti 1998: 127; korhonen 1981: 30, 130; aikio 2009: 276). 27. saan searvi ‘company’ (~ finn. seura id.) < Ps *sprē- (yss 1127) < fs *sepra < balt. *sebra- (cf. Lith. sibras ‘friend, relative’, Latv. sē brs, sebris ‘neighbour, companion’ (Lew ii 768)) (sammallahti 1998: 127, 262; thomsen 1890: 215). 28. saas sesnie ‘tanned reindeer leather’ (~ finn. hihna ‘belt’) < Ps *sesnē < fs *šišna < balt. *šikšnā- (cf. Lith. šikšnà ‘gently tanned leather’, Latv. siksna ‘belt’ (Lew ii 981)) (korhonen 1981: 30; aikio 2009: 150–151). cf. Mordva erzja (k)šna, Mokša šna ‘belt’; Mari u b šüštö id. (uew 786). 56 Lucie hofírkoVá, VácLaV bLaŽek acta Linguistica Lithuanica LXIV–LXV 29. saan siepman ‘seed’ (~ finn. siemen) < Ps *sēmen < fs *sēmen < Pb *sēmen (cf. Lith. sėmenys, old Pruss. semen ‘seed’ (Lew ii 774)) (sammallahti 1999: 85; thomsen 1890: 216). 30. saan suoidni ‘grass, hay’ (~ finn. heinä) < Ps *sōjnē (yss: 1171) < fs *šajna < Pb *šena- (cf. Lith. šiẽnas, Latv. sens ‘hay’ (Lew ii 980)) (sammallahti 1998: 124, 127; korhonen 1981: 30; thomsen 1890: 223). 31. saan suoldni ‘dew’ (~ finn. halla ‘(night ground) frost’ < Ps *sōlnē (YSS 1176) < fs *šalna < Pb *šalnā (cf. Lith. šalnà ‘rime, light frost’, Latv. salna ‘(night ground) frost’ (Lew ii 960)) (sammallahti 1998: 124, 127; korhonen 1981: 30). 32. saan suolu ‘island’ (~ finn. salo ‘wasteland, deep forest’) < Ps *sōlåj (YSS 1177) < fs *saloj < Pb *salā (cf. Lith. salà, Latv. sala ‘island’ (Lew ii 758)) (sammallahti 1998: 124, 127; korhonen 1981: 30; saarikivi 2004: 204; thomsen 1890: 214). 33. saan vietka ‘adze’ < Ps *vētke (yss: 1415) < Pb *edegā (cf. Lith. vedegà ‘kind of adze’, eastern dial. vėdegà, Latv. vȩdga ‘crowbar; chisel’, old Pruss. wedigo ‘adze’ (Lew ii 1211)) (aikio 2006: 40; 2009: 288). Proto-baltic reconstruction *ōdegā endzelina refers to the etymon ‘tail’: Lith. uodegà (see Lew 1164). the coherence of the meanings ‘tail’ and ‘axe’ is doubtful. 34. saan vuoras ‘old (of people)’ < Ps *vōres (yss 1458) < Pb. *āras (cf. Lith. võras, old Pruss. urs ‘old’ (Lew ii 1274; smoczyński 2007: 767)) (sammallahti 1998: 127; aikio 2006: 40). 35. saau vyöy’jee ‘wedge’ (~ finn. vaaja id.) < Ps *vōvjē (yss: 1477) < fs *vakja < Pb *aga- (cf. Lith. vágis, Latv. vadzis ‘id.’ (Lew ii 1179)) (korhonen 1981: 30–31). controVersies 36. saan gal’le ‘how much’ < Ps *kellē < fs *küllä < Pbs ?*kel- (cf. Lith. kel ‘how much, a few’, keliñtas ‘der wievielte, mancher, einige’ (Lew i 236)). however, the borrowing could be in the opposite direction, cf. finn. kyllä ‘yes’; kylläinen ‘full’; kyllin ‘enough’ (sammallahti 1998: 242). 37. saan galmmas ‘cold’ (~ finn. kylmä id.) < Ps *kelme- < PfP *külmä (uew 663) < Pbs *gel(u)mā- (cf. Lith. gelumà, gelmuõ ‘severe frost’ (Lew i 145: gélti ‘to stab’)) (sammallahti 1998: 127, 242). cf. Mordva erzja keľme, Mokša keľmä ‘cold; frost’; Mari kb kəlmə ‘frozen’ (uew 663: *külmä). if the fennoVolgaic isogloss is of the baltic origin, the first syllable vowel ü can be explained baltic Loanwords in saami 57straipsniai / articles by compensatory labialization caused by the elimination of the vowel -uin the assumed baltic source. 38. saan luokta ‘gulf’ (~ finn. lahti id.) < Ps *lōkte ( yss 619) < fs *lakte < balt. (cf. Lith. lañkstas ‘bend ’ , lañktis ‘handle ’ , Latv. jūŗas līcis ‘sea bay ’ (Lew 356 – 57)) (sammallahti 1998: 127; ssa ii 36). however, the fennosaamic isogloss is compared with the ob-ugric parallels of the same meaning in uew 234. 39. saan muohta ‘snow’ < Ps *mōttē ‘to snow’ (yss 696) (~ est. matma ‘to bury’) < fs *matta- < Pbs *mat- (sammallahti 1998: 255; koivulehto 2006: 185). the expected baltic source should be a derivate of the verb attested in Lith. mèsti ‘to throw’: ãt-mata ‘waste’, š-mata, -mota ‘scrap; waste’, pã-matas ‘base’, pre-mota ‘plaster’; Latv. mest ‘to throw’, at-mats ‘support’, pa-mats, -mata ‘base’; old Pruss. pomests ‘submitted, subject’, metis ‘throw’ (Lew 442–43); however, from the examples it is clear that the meaning ‘snow’ does not occur in the baltic languages. slavic words such as russ. metel’ ‘snow storm’, Pol. zamieć ‘snow drift’ < old slav. *metati are closer. romanian omát, omete ‘snow’ is of slavic origin too (cf. russ. dial. omét ‘lump’. however, the closest parallel we can found in osset. iron. mīt, digor. met ‘snow’ (abaev ii 124). the correspondence of iron. ī: digor. e reflects the diphthong *a, that may be original (mīzyn : mezun ‘to flow’ ~ avest. maēz- < iran. *maz-) or secondary (iron. mīd-: digor. med- ‘inside’ ~ avest. maiδiia- ‘middle’ < iran. *mada-). if the diphthong is secondary the ossetian iron. mīt, digor. met ‘snow’ can be derived from the protoform *mata-, that is compatible with Ps *mōttē . 40. saan šearrat ‘clear’ < Ps * e a e ᴅ ē < *śerä- < Pbs *źer- (cf. Lith. žėrėti ‘shine, blaze ’ (Lew ii 1301) ) (sammallahti 1999: 79). however, the saami word with šcan be relatively late as in the case of saan šūvon ‘welltrained dog’. 41. saan šūvon ‘well-trained dog’ < Ps * uovunj e < fs *śōvonji ~ *śavonji < Pbs śu ō n(i)- (cf. Lith. šuõ ‘dog ’ , gen. šuñs, old Lith. gen. šunès, and especially Lith. dial. (Šiaulia, grodno) nom. šovà, (samogitia) šovà; next Latv. suns, gen. suņa, old Latv. gen. suns; old Pruss. sunis id. (Lew ii 1033) ). sammallahti (1999: 79) notes that the saami šrepresents another reflex than the adaptation of baltic *šthat transforms into saa sand finn. h- ( saan sarvvis ‘reindeer bull’ ~ finn. hirvas id. < fs *širvas ). sammallahti assumes that the reason is an older chronological level of borrowing. however, jaakko häkkinen (häkkinen 2010) and Petri kallio (kallio 2009: 35) are inclined to the view that the two above-mentioned words (saan šearrat and saan šūvon) are later borrowings from such a baltic lan- 58 Lucie hofírkoVá, VácLaV bLaŽek acta Linguistica Lithuanica LXIV–LXV guage in which the alternation of the phoneme is š- (e.g. Lithuanian). they base this claim both on low the distribution of the words (saan šearrat occurs only in Lule, north and inari saami; saan šūvon occurs in south, ume, Lule and north saami) and on the phonological structure of the words. it seems to them that these words were borrowed into saami after secondary *š had appeared in Proto-saami. this secondary *š is preserved in present saami, whereas the original *š (that appears in the latest baltic borrowings) was changed into *s in Proto-saami (as well as in present saami languages). this solution seems to be easier and therefore more likely than the idea of sammallahti. on the other hand, kallio is wrong if he manipulates with the projection of Lith. šuõ < balt. *šō(n). if we take into account the similar nominatives such as Ved. śv, śuv and gr. κύων the Lithuanian form also represents the continuant of ie *uō(n) > balt. *śuō. Lith. dial. šovà ‘dog’ could also originate by internal development from this protoform. häkkinen adds that Ps * uovunj e was borrowed into saami before the vowel rotation, but Ps * e a eᴅē (häkkinen: *šear e tē) could be borrowed even after. according to him, this claim is supported by the non-etymological vowel combination *eae (regularly: early Ps *e-a > Late Ps *ea-ē and early Ps *e-i > Late Ps *iee , but never *eae ) ( häkkinen 2010; kallio 2009) . 42. saan vuos’si ‘handle’ (~ finn. ansa ‘snare’) < Ps *vōssē (yss: 1470) < fs *ansa < Pb *anšā, however, it is evidenced just the only one form *anšas (cf. Lith . vąšas ‘handle’, samogitian ąnšas id. (Lew ii 1207)) (sammallahti 1998: 127). alternatively, finn. ansa ‘snare’ (> saan hānsâ- ‘board’) was derived from the baltic source represented by Lith. ąsà ‘handgriff, handhabe, henkel’, Latv. ùosa, ùoss id., old Pruss. ansis ‘kesselhaken’ (Lew i 18), see thomsen 1890: 159; skes 19. in ssa i 77, both saan vuos’si ‘handle’ and finn. ansa ‘snare’ are derived from Lith. ąsà etc . concLusion we can find the words of baltic origin in all fenno-Volgaic languages, about seven in Mari (five in our minicorpus) and thirteen in Mordva (six in our minicorpus); three of them do not have an equivalent in balto-fennic languages (sammallahti 1984: 140). it is very difficult to differentiate between some words if they got directly into saami or through balto-fennic because we can expect only slight differences between the Proto-balto-fennic and Pre-saami language baltic Loanwords in saami 59straipsniai / articles at the time when borrowings began. thus, there were no major sound changes yet. some of the listed baltic loanwords (eight of forty if we eliminate ##39, 40, maybe 19) do not have an equivalent in balto-fennic languages. although it is only 20 %, we can deduce that the balts did not live in immediate contact only with the balto-fennic people but also with the ancestors of the saami people, who had to be their neighbours for some (but not long) time. we can assume that the ancestors of saami used to live near the gulf of finland and near the river neva at the time of contacts with the balts. this also corresponds to tacitus’ description of fenns2 that can be likely applied to the saami people than to the ancestors of the finnic people. we can read from poor geographical data that in the first century ad tacitus’ fenns lived somewhere in eastern baltic, north from the aesties, who are considered to be of baltic origin. the words that have some equivalents in balto-fennic were likely borrowed into saami during the Proto-balto-fennic period or even earlier, during the fenno-saamic period. these words are four times more than saami-baltic parallels, so we can reasonably assume that baltic populations had closer and more long-term relationships with their north neighbours than with the ancestors of the saami people in the balto-fennic period or the fenno-saamic period. several tens of words of baltic origin are known only in northern balto-fennic languages (finnish, karelian, ingrian and Veps) and in saami; on the other hand, about ten of baltic borrowings occur only in the southern balto-fennic (estonian, Votic and Livonian). it is likely that the ancestors of the balto-fennic people used to live on both sides of the gulf of finland at the time of baltic contacts, and that the words of baltic origin were borrowed independently (sammallahti 1984: 140). 2 Fennis mira feritas, foeda paupertas: non arma, non equi, non penates; victui herba, vestitui pelles, cubile humus: solae in sagittis spes, quas inopia ferri ossibus asperant. Idemque venatus viros pariter ac feminas alit; passim enim comitantur partemque praedae petunt. Nec aliud infantibus ferarum imbriumque suffugium quam ut in aliquo ramorum nexu contegantur: huc redeunt iuvenes, hoc senum receptaculum. Sed beatius arbitrantur quam ingemere agris, inlaborare domibus, suas alienasque fortunas spe metuque versare: securi adversus homines, securi adversus deos rem difficillimam adsecuti sunt, ut illis ne voto quidem opus esset. „the fenni are strangely beast-like and squalidly poor; neither arms nor homes have they; their food is herbs, their clothing skins, their bed the earth. they trust wholly to their arrows, which, for want of iron, are pointed with bone. the men and the women are alike supplied by the chase; for the latter are always present, and demand a share of the prey. the little children have no shelter from wild beasts and storms but a covering of interlaced boughs. such are the homes of the young, such the resting place of the old. yet they count this greater happiness than groaning over field-labour, toiling at building, and poising the fortunes of themselves and others between hope and fear. heedless of men, heedless of gods, they have attained that hardest of results, the not needing so much as a wish.“ (Germania §46; translated by a.j. church et alii).