Daugiskaitiniai asmenvardiniai oikonimai ir jų reikšmė linksniavimo bei kirčiavimo sistemų rekonstrukcijai

LIETUVIŲ ONOMASTIKOS TYRINĖJIMAI LIETUVIŲ KALBOTYROS KLAUSIMAI, XXI (1981) BONIFACAS STUNDŽIA PLURAL PERSONAL-NAME OIKONYMS AND THEIR SIGNIFICANCE FOR THE RECONSTRUCTION OF DECLENSION AND ACCENTUATION SYSTEMS 0. A considerable number of Lithuanian toponyms are place names of personal-name origin. Examined here are oikonyms formed from the plural of ā- and ē-stem personal names. 0.1. The oikonyms were collected from the “Directory of the Administrative-Territorial Division of the Lithuanian SSR” (V., 1976, part 2), and the personal names1 from the surname card file of the Institute of the Lithuanian Language and Literature of the Lithuanian SSR Academy of Sciences and the forthcoming dictionary of Lithuanian surnames2. Analysis of the collected data encounters several difficulties: 1) surnames in the card file are often recorded by non-specialists and contain errors, sometimes even lacking accent marks, so determining their accentuation is out of the question; 2) o-stem surnames frequently occur alongside ā-stem surnames (Survilà: Survilas), so the geographical distribution of both variants had to be established more precisely; 3) we do not have an etymological dictionary of Lithuanian surnames (it is still being prepared). Consequently, some oikonyms could not be used: a) those for which the stem of the underlying personal name could not be determined, e.g.: Ašmonai: Ašmona(s), Buzai: Buzà || Būzas, Jovaišai: Jóva iša(s); Meškėliai: Meškėlė || Meškėlis (about 20 oikonyms in all); b) those whose underlying personal name’s etymology is doubtful or unclear (some of them were evidently recorded in an inauthentic form): Čiulada, Dubravà (cf. Lithuanian dubravà, -āvas LKŽ II 78, Polish Dubrawa, Belarusian Дубрава, etc.), Kiena, Medekšà (cf. Belarusian Мадэкша, Polish Medeksza)3, Pažerà [: žėrėti (?)], Šilva [< Šiluvà (?), cf. šilmà < šilumà or šilimà], Tūita, Valečkà ← Valentinas (?); Vinča ← Vinca(s)? etc. (about 25 oikonyms in all)4. 0.2. A total of 226 aforementioned plural oikonyms were found; 183 are examined (slightly more are used for geographical analysis; see footnote 4). Most of the selected oikonyms (156 of 183) are derived from ā-stem personal names (27 oikonyms from ē-stem personal names). 1Personal names (avd) were identified on the basis of present-day surnames. The term personal name (anthroponym) is used here because most oikonyms are old, having arisen considerably earlier than our present-day surnames took shape. 2Taking this opportunity, I wish to thank A. Vanagas, Candidate of Philological Sciences and head of the Contemporary Lithuanian Literary Language Section at LKLI, for his help and advice, and for permission to use the aforementioned card file and the typescript of the first volume of the surname dictionary. 3 Nei M. Biryla (Бирула М. В., 1969, 265), nei J. Bystronis (Bystroń J., 1936, 281) šio avd kilmės nepaaiškina. 4Oikonyms formed from some of the latter personal names are included in the distribution map (see footnote 11). 185 1.1. When examining the plural oikonyms of interest to us, it is useful to distinguish two models (according to their relationship to the singular stem or paradigm of the underlying personal name): 1) (i)o-stem oikonyms, i.e. a model involving alternation (neutralization) of stems or paradigms; 2) ā- or ē-stem oikonyms (the simple model). 1.1.1. The (i)o-stem model is considerably more productive, comprising 118 oikonyms (i.e. ≈ 64% of all oikonyms)5: Barzdai: Barzdā, Bulkai: Bùlka, Juškai: Jùška, Linkai: Linka, Noreikai: Norēikà, Blõžiai: Blõžė, Dirsiai: Dirsė etc. Of the underlying personal names in this model, 104 have an ā stem, while 14 have an ē stem. 1.1.2. Sixty-five oikonyms have a paradigm without stem alternation6 (16 of them also have a parallel model with stem alternation; see footnote 5), e.g.: Alseikos: Alsēika, Andreikos: Andreikà, Dindos: Dindà, Karéivos: Karēiva; Dailidės: Dailidė, Lāpės: Lāpė. 1.2. Plural oikonyms of the first model [(i)o stem] exhibit not only a change in declension, but also a certain alternation in accentual paradigms. More than two thirds (39 of 55) of these oikonyms have final-syllable stress (accentual paradigm 4), in contrast to the baritone (or suffixal) accentuation of the singular form of the underlying personal name7 (accentual paradigms 1, 2): Dìlbai (4): Dìlba (1), Guobai (4): Gúoba (1), Kirkai (4): Kìrka (1), Minelgai (4): Minélga (1), Paurai (4): Páura (1), Šarkai (4): Šárka (1), Varžai (4): Várža (1); Blõžiai (4): Blõžė (2), Dirsiai (4): Dirsė (1), Lapiai (4): Lāpė (2) etc. Most such ā- and ē-stem personal names derive from common nouns (33 of 39). Thus, it should be assumed that in this case the patterns discernible in onomastics can help solve problems in historical morphology and accentology. Analogous distribution of accentual paradigms is characteristic in some dialects of other plural (i)o-stem oikonyms as well (without stem alternation). The “Directory of the Administrative-Territorial Division of the Lithuanian SSR” (V., 1976) yielded over 200 non-disyllabic plural oikonyms of personal-name origin, which are stressed according to accentual paradigm 4, in contrast to the baritone or suffixal accentuation of the singular forms of the underlying personal names. Such oikonyms are concentrated in the area of the western Aukštaitian Kaunas subdialect (≈ 44%) and in the northern and western parts of Samogitia (≈ 27%). Smaller numbers are found elsewhere, including peripheral dialects, e.g.: Banaičiai (4) Šk*: Banáitis; Grigaliūnai (4) Kps, Prn: Grigaliūnas, cf. Grigaliūnai (1) Šr, Švnč; Tūmeliai (4) Šlu: Tūmėlis; Perkūnai (4) Klp: Perkūnas; Jurgeliai (4) Brž: Jurgėlis; Petreliai (4) Viln: Pet- rėlis; Šeštokai (4) Lzd: Šeštõkas etc. 1.3. The underlying personal names of oikonyms in both models have diverse origins8. Of the 183 personal names, 105 (≈ 57%) are associated with Lithuanian common nouns, e.g.: Bilda: bilda (:bildėti), Bùrba: 5 Sixteen such oikonyms also have parallel ā-stem (13 oikonyms) or ē-stem (3 oikonyms) forms: Bimbai || Bimbos, Linkai || Linkos, Petrõškai || Petrõškos etc.; Lapiaí || Lāpės, Laumiai || Laũmės, Pėmpiai || Pėmpės. The geographical distribution of these variant forms differs. 6Of these, 52 oikonyms have an ā-stem, while 13 have an ē-stem. 7The singular form, therefore, because the plural of a personal name, as we shall see later, may have root or suffixal stress as well as final-syllable stress. *Abbreviations for settlement names are used as in LKŽ. 8The underlying personal names in the individual oikonym models do not differ substantially in quantitative terms with respect to origin. The etymologies of the personal names are based on the following works: Lietuvių pavardžių žodynas, 186 bùrba (:burbėti), Linka (:liñkti), Kimšà (:kimšti), Rimšà (:rimti); Bimba: bimba, Diršė: diršė, Pėmpė: pėmpė, Vaštakà: vaštaka ‘a kind of bird = coot’. Among the others, 10 two-stem anthroponyms stand out (almost all old): Dalgėda, Gerdvilà, Norėikà, Nórgėla, Sangáila, Survilà⁹ etc. The remaining 78 anthroponyms are either formed from borrowed appellatives or simply borrowed, usually from the Slavic name stock (38): Bùlka (cf. Polish Bulka, Belarusian Булка), Dùdà (cf. Polish Duda, Belarusian Дуда: Polish duda, Belarusian dūda > Lithuanian dūdà), Kìrka: kirka (Slavic)¹⁰, Šaškà: šaškà (Slavic) ‘a kind of sword’; the others are formed from Christian names (40). The base words of most Christian-name anthroponyms did not originate in the Lithuanian linguistic environment. These are most often folk forms of East Slavic Christian names, e.g.: Belarusian Іваука ← Іван, Polish Iwaszko ← Iwan: hypocoristic Ivoškà; Belarusian Юхно ← Евфимий: hypocoristic Juknà; Belarusian Сау́ка ← Савва, Polish Sawka ← Sawa: hypocoristic Sáukà etc. Only an occasional Christian name may also have developed in Lithuanian: Dòmas ← Dominykas: hypocoristic Domėikà, Razmà ← Erázmas: hypocoristic Razmà, Valentà ← Valentinas: hypocoristic Valentà. In general, more than two thirds (123) of the personal names with ā- and ē-stems from which plural oikonyms are formed are related to appellatives. Most such anthroponyms (85%) are formed from Lithuanian or Baltic appellatives, while the rest are from borrowed ones (usually Slavic). 1.4. Having presented a concise quantitative, accentual, and etymological characterization of the data under investigation, their geographical distribution should also be discussed. Let us designate the models of the oikonyms examined here as follows: 1. a) Burbai: Bùrba; b) L apiai: Làpė. 2. a) Bùrbos: Bùrba; b) Làpės: Làpė¹¹. 1.4.1. The map shows that oikonyms of the first model (with stem alternation) are concentrated in the Western Aukštaitian Kauniškiai area¹² (38 oikonyms) and in Samogitia (31 oikonyms). (i)o-stem oikonyms also occur elsewhere, e.g.: 13 oikonyms each in the Western Aukštaitian Šiauliškiai and Eastern Aukštaitian Panevėžiškiai areas, 20 in other Eastern Aukštaitian dialects, and 5 oikonyms in the southwestern corner of the Southern Aukštaitian area. Thus, the Western Aukštaitians and Samogitians stand apart from the others (not vol. 1. A—K (manuscript); Zinkevičius Z. Lithuanian Anthroponymy. — V., 1977; Vanagas A. Our Surnames. I. — In: Words and People. V., 1974, pp. 87—152; Vanagas A. Our Surnames. II. — In: People and Language. V., 1977, pp. 49—92; Kuzavinis K., Savukynas B. Etymological Dictionary of Lithuanian Names. — In: Names and Words. V., 1971, pp. 43—157; Бірыла М. В. Беларускія антрапонімы. — Мінск, 1966, т. 2, 1969; Bystroń J. St. Polish Surnames. Lwów, 1936; Dictionary of Old Polish Personal Names. — Wrocław—Warszawa—Kraków, 1965—1979, vol. 1—5; Gottschald M. German Onomastics. — Berlin, 1954. 9 Most oikonyms formed from these anthroponyms have o-stems (8): Gerdvilai, Noreikai, Survilai etc. 10 The anthroponym Kirka is also related to Polish Kirka and Belarusian Kipko. 11 In determining the distribution of oikonyms, care was taken to ensure that the personal name came from the same or a very nearby locality as the oikonym derived from it. Therefore, some oikonyms (mostly o-stem ones) were not included in the map. The numbers of oikonyms used for the individual models are as follows: 1) Burbai — 113 oikonyms, Lapiai — 7 oikonyms; 2) Bùrbos — 76 oikonyms, Làpės — 15 oikonyms. Note. The counts also include geographical variants of the same oikonym; for example, at least 5 villages named Dailidės are known. In general, identical oikonyms are relatively few. 12 This article follows the classification of Lithuanian dialects proposed by A. Girdenis and Z. Zinkevičius (see Kalbotyra, vol. 14, pp. 139—147). 187 all, see the map), northern Panevėžys dialect speakers, etc. In these areas (especially in the Kauniškiai region), o-stem oikonyms are quite numerous; ā-stem oikonyms are rare here (see 1.4.2), except among Samogitians, where they occur more densely in some places (for example, in Kelmė district). 1.4.2. Oikonyms of the second model are concentrated in the eastern Aukštaitian Vilnius dialect (34 oikonyms). Fewer such oikonyms occur elsewhere: they are scattered in small groups across roughly the whole of Lithuania, e.g., 12 oikonyms in the southern Aukštaitian area, 21 among the Samogitians (8 of them in Kelmė district), 4 in the south of the western Aukštaitian Kauniškiai area (Prienai district), and 16 oikonyms scattered across the eastern Aukštaitian territory (excluding the Vilnius dialect). 1.4.3. To summarize the data on the distribution of oikonyms, it should be noted that the geographical distribution of both models is fairly well-defined: oikonyms of the first model (Burbai, Lapiai) are concentrated in the western Aukštaitian territory (especially among the Kauniškiai), across most of the Samogitian area, and in some eastern Aukštaitian areas (the northern Panevėžys and Širvintos dialects); oikonyms of the second model (Būrbos, Lāpės) are concentrated in the eastern Aukštaitian Vilnius area (see the map). 2.1. The oikonym models examined are undoubtedly associated with the plural forms of ā- and ē-stem personal names, which also have two variants in the dialects: Burbai, Lapiai and Būrbos, Lāpės, respectively¹³. The geographical distribution of these variants essentially coincides with the distribution of the oikonym models discussed. 2.1.1. Forms with stem or ending alternation [(i)o stem] are used by most western Aukštaitians and the western corner of the southern Aukštaitian area (Lzd, Vs, Kpč, Lp), e.g.: Kabelkai: Kabelka, Kalėdai: Kalėda, Švitrai: Švitra, Gerviai: Gėrvė¹⁴. Such usage is also characteristic of many Samogitians (Mžk, Trk, Tl, Slnt, Krtn, Skdv, Šauk, Krš, Varn, Rs), and of some eastern Aukštaitians: the southern Panevėžys and Širvintos dialects, and part of the Utena and Anykščiai dialects. In almost all Lithuanian dialects, the (i)o-stem model is also characteristic of the plural of (i)u-stem personal names, e.g.: Butkai: Būt­kus (cf. oikonym Butkai in Kš, Klp, Tl, Trg), Kubiliai: Kubilius (cf. oikonym Kubiliai in Kš or Kubiliai in Brž, Klm, Šk, Ut). There are dialects in which plural forms of ā- and ē-stem personal names with stem alternation are characteristic of only part of the paradigm. Around Tverečius (eastern Aukštaitian Vilnius dialect), the following forms are (or at least were) used: Garlai, Garlų, Garlāms, Garlāis, but Gārlas: Gārlą (Otrębski J., 1934, 209 ff.). Thus, only the plural accusative retains the singular stem and accentuation of the personal name. 2.1.2. Almost everyone (the southern Aukštaitians (except the western corner) and most of the eastern Aukštaitian Vilnius, Kupiškis, and Utena dialect speakers (Rk, Km, Žb, Ob, Ds, Zr, Krns, Ktk), as well as the northern Panevėžys dialect speakers (Lnkv, Ps, Jnšk, Pš)) use plural forms of ā- or ē-stem personal names without stem alternation, e.g.: Būrbos, Stričkos, Dailīdės. This is most often the case 13 Burbai, Lapiai and Būrbos, Lāpės ‘the Burbas’ or Lapės family (several families); people with these surnames’ → ‘a place where those families or people live’. This is a conventional way of naming: a property of an object, agent, and so on is transferred to a place associated with that object, etc. 14 Perhaps the first to describe an almost analogous stem alternation in the Tverečius dialect was J. Otrembskis (1934, 209 ff.). E. Grinaveckienė (1959, 179; 1960, 161) drew attention to such usage in the Mituvos basin dialect (around Jurbarkas). A. Girdenis wrote about it in greater detail (1962, 4 ff.). 188 Fig. 7. Distribution map of oikonyms in the plural formed from ā- and ē-stem personal names 189 This is also how the Lithuanian dialects along the Latvian border behave (apparently under the influence of Latvian, which seems to have only such plural forms of the aforementioned anthroponyms). 2.2. Personal-name plurals with a change of stems or paradigms [as in the (i)o-stem oikonyms discussed above, see 1.2] are characterized by final stress (accentual paradigm 4), in contrast to the otherwise baritone, or initially stressed, singular paradigm, e.g.: Baužai (4): Bauža (1), Mažeikiai (4): Mažeikė (1,2); Gerviai (4): Gervė (1), Kateliai (4): Katėlė (2). Where such ā- or ē-stem personal-name plural forms are used, plurals of personal names of other stems, too, tend to have final stress (accentual paradigms 4, 3): Gailiai (4): Gailius, Sprindžiai (4,3): Sprindis, Kazlauskiai (4,3): Kazlãuskis, Šimkūnai (4,3): Šimkūnas etc. (The third accentual paradigm is more characteristic of Samogitians and eastern Aukštaitians; the fourth accentual paradigm is best represented by the western Aukštaitian Kaunas subdialects15). This stress pattern is apparently followed by all Lithuanian dialects that have ā- or ē-stem personal-name plurals in -(i)ai16. 2.2.1. If the plural of ā- or ē-stem personal names is formed on the basis of the singular stem (Dūrbos, Lāpės), the plural paradigm of other stems usually has root or initial stress (accentual paradigms 1, 2): Sprindžiai, Kazlauskiai etc. Apparently, only some Samogitians follow this rule with greater regularity. Aukštaitians (including peripheral dialects outside Lithuania) who have a variant of ā- or ē-stem personal-name plurals without a change of stems are characterized by final stress (accentual paradigms 3, 4; paradigm 3 is more frequent): Trinkūnai, Mackoniai, Ūdrai: Ūdras, Bulbieniai (4) Str (cf. bulbieniai (2) ‘potato haulms’) etc. Given the relationship between ā- or ē-stem personal-name plurals with stem alternation and final stress (see 2.2), it may be assumed that these Aukštaitians, too, must have had a plural variant of ā- or ē-stem personal names in -(i)ai (nominative). 2.2.2. Since the plural forms of (i)o- and (i)u-stem personal names (usually not disyllabic) have two variants with stress on the ending—the fourth and third accentual paradigms—the question arises as to which is primary, that is, older. It may cautiously be conjectured that in this case the older is the fourth accentuation, because: 1) only it is manifested by the plural forms of ā- and ē-stem personal names with stem alternation (see 2.2); 2) among Samogitians, who have mobile stress in the plural of (i)o- and (i)u-stem anthroponyms (accentual paradigm 3), plural oikonyms of personal-name origin that are not disyllabic and are stressed according to the fourth accentuation occur fairly often, especially in the western and northern parts; for example: Skurvydai (4) Als: plural of Skūr vydas, Pakauskiai (4) Tl: plural of Pakāuskis (see 1.2 for more detail); 3) in Eastern Aukštaitian dialects (for example, some Utena and Vilnius subdialects), (i)o- and (i)u-stem personal names of accentual paradigms 1 and 2, and the corresponding oikonyms, have a very prominent secondary stress on the plural accusative ending (Šim'kū,nius, Tu'mė,nus, Ru'giė,nius), which most likely presupposes original stress on the ending here, in some cases also manifested (cf. Bulbienius, Šimkūnius etc.)17. 2.3.1. It is thought that the plural of (i)ā- and ē-stem personal names with -(i)ai (Dūdai, Gauliai: Gául ia, Lapiai) arose by analogy with the more productive (i)o- 15Noun variants with oxytone stress are also concentrated in the Kaunas subdialect area; see Petraitienė Z. Stress variants of disyllabic nouns in Lithuanian. Diploma thesis. — VVU, 1981. 16Naturally, this is after eliminating the effects of stress retraction characteristic of some dialects. 17The scope of the article does not allow a more detailed analysis of this interesting problem. 190 kamienių avd daugiskaitą (Grinaveckienė E., 1960, 161). Kitokių nuomonių, rodos, nėra. Thinking this way, it is first of all difficult to explain why ā- and ē-stem personal names, which otherwise have root or suffix stress, tend in the plural, with a change of stem, to shift to ending stress, which is no more productive than barytone or suffix stress. 2.3.2. It seems most reasonable to assume that the two plural variants of personal names discussed above had different meanings18. The distribution of these variants can still be found in dialects (according to the data available, in places around Jurbarkas and elsewhere along the edges of the western Kaunas highlanders, here and there in the southern Panevėžys dialect, and in the southwestern corner of the western highlanders)19. In such cases, a plural of personal names with a change of stems and/or accentual paradigms20 denotes the name of a family or kin group, derived from a personal name—thus it has a collective meaning, as is also evidenced, among other things, by its usage: western Kaunas highlanders and some eastern Vilnius highlander dialects (the latter only optionally) use the plural of personal names in -(i)ai with collective numerals21, e.g.: dvėji, treji … Dūdai, Gudeliai ‘the two, three … Dūdė and Gudėlio families’. The second plural variant, formed according to the singular stem and accentuation, is used to refer to individual members of a family or kin group, e.g., brothers, cousins … Dūda, Dirsė. It would be appropriate to call the first plural, which has a specific oxytone form in (i)ai that neutralizes the opposition among stems, genders, and accentual paradigms, and specific semantics (it denotes an indivisible whole), a collective plural; the second, a regular plural. Personal names of different stem classes could manifest two plural variants, for example, as follows: 1) a model with neutralization of the opposition among stems, genders, and accentual paradigms in the collective plural: 18 Tokia prielaida atitinka J. Šmidto (1889) išplėtotą tezę apie dvi ide. daugiskaitos funkcijas; be to, žr. Kuryłowicz J., 1964, 204. 19 For personal information, the article’s author is grateful to Candidates of Philological Sciences J. Paulauskas and A. Vidugiris, Associate Professor A. Piročkinas, B. Savukynas, and others. 20 A change of stem then also entails a change of accentual paradigm if the personal name has barytone or suffix stress (see 1.2, 2.2). A model of plurals differing only in their accentual paradigms (Šimkūnai: Šimkūnai) is known in places around Jurbarkas, but not everywhere. E. Grinaveckienė (1960, 162) only suggests that it may have existed. 21 In this case, collective and group numerals differ only in that the collective form is the strong (marked) member of the opposition, usually expressing an approximate number (treji metai: trejetas metų ‘three years’). Associate Professor V. Labutis brought this fact to the author’s attention. 22 Such forms of personal names with an (i)u-stem are not used even where the plural of personal names is formed according to the singular stem and accentuation. Burbai L apiei Barkai Bùrba Lãpė Barkus Bùrbos Lãpės *Bárkūs 191 2) a model with neutralization of the opposition between accentual paradigms in the collective plural: 2.3.3. In practically all Lithuanian dialects, the distribution of two plural variants of personal names (avd) was generalized in favor of one or the other member²³. The prevailing plural variant is the one in which the opposition between stems and accentual paradigms is neutralized, i.e., the variant with -(i)ai24. Place names in this model formed from personal names with ā- and ē-stems also predominate (see 1.1.1 and the diagram). Such place names also occur in areas (along the Latvian border, in the northern Panevėžys and Kupiškis dialects, and here and there in other eastern Aukštaitian areas, including the Vilnius dialect) where, according to the available data, the plural of personal names with (i)ā, ē stems ending in -(i)ai is unknown (see the diagram). Since plural anthroponymic place names are characteristically collective in meaning (see note 13), they are derived from the collective plural of personal names, whose forms in the dialects mentioned above were displaced by the ordinary plural paradigm²⁵. 2.4. The hypothesis may be put forward that the opposition between ordinary and collective plurals, traceable in place names, was also characteristic of appellatives. This hypothesis is partly supported by the fact that most personal names (≈ ⅔) from which plural place names are derived are of appellative origin (see the relevant parts of this article). Most importantly, however, remnants of the opposition between the two plurals can still be found in appellatives themselves. As in toponymy, these remnants are of two kinds: 1. The ordinary plural contrasts with the collective plural, which neutralizes the opposition between declensional and accentual paradigms²⁶, e.g.: liepos, sg. liepa: liepai (4) ‘linden grove’ Št (information from V. Mažiulis), cf. also liepūkai ‘linden blossoms’ Dkš; pėdos, sg. pėda ‘lower part of the leg, paw; an impression left by a foot, a hollow’ Lp, Dp 178: pėdai (3) ‘marks imprinted by someone’s feet, footprints’ Lp, Dp 183, 192, 210 (LKŽ IX 736); plaušos ‘fibres, strands of some material’ S. Dauk: plaušai ‘the lower fibrous layer of the bark’ S. Dauk²⁷. 2. The collective plural is manifested by neutralization of the opposition between accentual paradigms, i.e., by stress on the ending, as compared with the ordinary plural (or 23 The mixed plural paradigms of personal names in some dialects (see 2.1.1) should be regarded as the result of the decline of the aforementioned distribution (contamination of two paradigms). 24 The personal names with (i)u stems generalized only the collective plural variant in probably all Lithuanian dialects (see 2.1.1). 25 The displacement of the ordinary plural model in some areas and the collective plural model in others depended on various factors, which there is insufficient space to discuss here. 26 The number of examples (illustrations) here and below is fairly limited. 27 Cf. the following sentences by S. Daukantas (LKŽ X 109): [the Samogitians] worked maukny palagus, punes <…>, and tied and bound them with hop vines and fibres, as well as with willow shoots || Plaušai [both instances italicized by me. — B. S.] cover the whole tree with that upper bark. Note. LKŽ X 109 does not define quite accurately the meaning of the word plaušos, given in the first sentence. What is meant here is not the fibrous layer of the bark, but the individual fibres of that layer, “threads.” 192 (singular) a barytone or pre-accent paradigm, e.g.: úošviai (1), sg. úoš-vis: uošviai (4) ‘father-in-law and mother-in-law; their home = uošvija’; dial. tėvai (1), sg. tėvas (the barytone paradigm is indicated by Latvian tēvs): tėvai (4) ‘father and mother, родители’; váistai (1), sg. váistas ‘usually some kind of medicinal herb’: váistai (4) ‘лекарство’ Dkš (information from V. Labutis); šimtai (2), šímtas (denotes an exact number): šimtai (4) – denotes an approximate number, ‘many’ Glv, Nm; úoksai (1), sg. úoksas: oakn Uoksāi (4); vartai (2): oakn Vartāi (4)²⁸; vókyčiai (1), sg. vóky tis: vokyčiai (3ª) ‘the German people, etc.’ Trk (information from A. Girdenis); lietùviai (2), sg. lietùvis: lietuviai (4) ‘the Lithuanian people, etc.’ Grv; siuvėjai (1), sg. siuvėjas: siuvėjai (4) ‘the tailors’ family’ Ldvn, Jrb (information from J. Kabelka; Grinaveckienė E., 1960, 162 ff.)²⁹. 2.4.1. To clarify the thesis posited here, let us compare the following sentences: These herbs are good medicines and They gave Father a portion of medicine Dkš; Two hundred (rubles) and Hundreds of them are flying there Glv (Morkūnas K., LKK XI 127); Three Germans walked across the yard and During the war those Germans did a lot of mischief Trk; Across the river, the Belarusian woman married a Lithuanian and In the fourteenth year <…>the Poles and Lithuanians fought Grv (Kardelytė J., 1975, 101, 131); Tailors, hairdressers … work at the factory and Here live the tailors of our village… ‘the tailors’ family’ Ldvn. 2.4.2. The distribution of stress among passive participles in some Samogitian dialects is probably also related to the distribution of noun accentual paradigms discussed above; cf. pristatýta, ištaisýta [pristatìˑtą, ištaisìˑtą] – feminine singular with pre-accent stress: pristatytà, ištaisytà [pristatìˑtą, ištaisìˑtą] – the so-called neuter form with end-stress Lc, Lž, Ktk, Akm, Mžk, Vkš, Kal, Pln. Cf. such sentences: That meat will be delivered every morning and Everything will be delivered in the morning; Your load will be corrected at once and I think that the road has already been repaired, they are driving on it Pln³⁰. 2.5.1. The idea that a collective plural in -(i)ãi existed in Lithuanian (and at least in East Baltic) is not entirely new. J. Otrebski had already advanced a similar hypothesis, albeit very fragmentarily, in explaining the reasons for the emergence of variants among o-/a-stem nouns. He argued that, as East Baltic collective nouns [nomina collectiva with *-(i)ã] became pluralized, a new (secondary) plural form in -(i)ãĩ [≤ *-(i)ã] arose, retaining a collective meaning for some time (Otrebski J., 1956, 3, 28). 28 Attention has long been drawn to variation in accentual paradigms, e.g.: kálnai (1): kalnai (3), krántai (2): krantai (4), etc. (Grinaveckienė E., 1960, 162; Ильич—Свитыч В. М., 1963, 22 ff.; Kazlauskas J., 1968, 48; Kardelytė J., 1975, 46). It seems that no one has noted that the meanings of these variant forms may differ. 29 A. Girdenis has written more extensively about the fact that derivatives with the same suffixes often have, in individual dialects, either pre-accent or root stress, or end-stress (mobile stress) (šáutuvai). He believes that “a considerable number of the aforementioned examples of ‘root’ or ‘end’ stress are relics, detached from other derivatives not only in their form (stress), but also in their content” [this word is italicized by me. – B. S.]; see Girdenis A., 1972, 69. Conversations with A. Girdenis also crystallized the thesis advanced here: end-stressed variants of nouns that are otherwise barytone (cf. vókiečiai: vokiečiai) may be relics of a former collective plural. 30 Associate Professor A. Girdenis also noticed this accentual fact and brought it to my attention. 13. 91 193 2.5.2. J. Otrembskis’s hypothesis is not entirely original either. It rests on J. Schmidt’s view that the Balts had a collective (= neutrum pl.)31 with the ending -ai (< *-āi), which, when the masculine and neuter genders of nouns merged (coincided), was generalized as the nominative of masculine o-stem nouns. Thus, according to J. Schmidt, the usual association of Lithuanian vilkai with Slavic vьlci, Greek λύκοι, etc., is impossible32 (Schmidt J., 1889, 231). The ending (formant) of Lithuanian -ai (ežer-ai, ger-ai, t-ai) was derived by J. Schmidt from *-āi (i.e. *-ā + i), which he considered to be of Indo-European antiquity and one of the variants of the neuter plural (= collective) nominative (alongside *-ā)33 (Schmidt J., 1889, 227–233). J. Schmidt also argued for the origin of -ai from *ā + i on the basis of accentuation34. He observed that the singular of many Lithuanian o-stem nouns has barytone stress, while the plural is stressed on the ending, e.g.: namas (cf. Old Indian dámas, Greek δόμος): namai, cf. Russian дома́ (Schmidt J., 1889, 41). As is known, such a distribution of accentual paradigms is characteristic of the neuter singular and plural (= collective). 2.5.3. Some linguists agreed with J. Schmidt’s view on the origin of Lithuanian -ai, while others disagreed35. The aforementioned hypothesis was criticized in a special study by E. Nieminen (1922). In recent times, apparently no one believed that the Balts could have had a collective (more precisely, a collective plural, i.e. a pluralized collective) in -(i)ai (Stundžia B., 1978, 113, footnote 8). 2.5.4. To clarify the relationship between the conceptions of J. Schmidt, J. Otrembskis, and the one advanced here, as well as the use of terminology, it is necessary to consider, at least briefly, the relation between the collective plural of nouns and the collective. There is no doubt that the relation of the collective plural to the singular (or ordinary plural) is derivational rather than inflectional. This is indicated by the specific (neutralizing) form of this plural [in -(i)ai] and its semantics (it denotes a collective, i.e. an indivisible whole). Consequently, with the subsequent reconstruction in mind, the collective plural is to be derived from the derivational category of the collective (collectivum, *ā/a-collectivum) and treated as a pluralized collective, i.e. as a transitional stage in the grammaticalization of the collective (collectivum > pluralis)36. Relative chronology 31J. Schmidt was the originator of the theory that Indo-European neuter plural was an ancient collective (*ā-collectivum), which he regarded as a derivational category of the feminine singular (Schmidt J., 1889, 6 et passim). A. Meillet treated the collective as a neuter form (Meillet A., 1938, 30, 330 ff.). Chr. S. Stang (1970, 16 ff.) advanced an interesting hypothesis, assigning the *ā-collective the status of a nominal class category. On the use of the term collective, see Valeckienė A., 1978. 32Even before J. Schmidt, another founder of the study of the Indo-European collective, G. H. Mahlow, held a similar view (Mahlow G. H., 1879, 81). 33Regarding J. Schmidt’s hypothesis, we will note for now that the author of this article holds a somewhat different view on the origin and antiquity of Lithuanian -ai and its relationship to the neuter gender. 34This was one of the more important arguments for this hypothesis, advanced by J. Schmidt’s supporter A. Meillet and given considerable attention by its critics, J. Endzelīns (Endzelīns J., 1974, 291 ff., 591 ff.). 35 Plačiau ir bibliografiją žr. A. Valeckienės straipsnyje (1978, 69, 2 išnaša). Dėl liet. -ai kilmės žr. Mažiulis V., 1970, 170 t. t. 36 The grammaticalization of the collective (a derivational category), i.e. its transformation into a plural (an inflectional category), is a diachronic universal (Kurylowicz J., 1964, 5 et passim). For a fuller discussion of inflectional and derivational categories and their relationship, see ibid., pp. 35–37. 194 Questions concerning the grammaticalization of the collective and the origin of the collective plural ending (formant) -(i)ai are beyond the scope of this article (see also footnote 33). 2.6. Instead of conclusions. An analysis of personal-name-derived plural oikonyms, plural personal names, and certain relic appellatives shows that Lithuanian (and at least the eastern Baltic languages) had two contrasting variants of the noun plural (probably of the entire nominal inflection): the ordinary plural and the collective plural. The following features of the collective plural can be reconstructed: 1. A specific form with final stress and the ending -(i)ai (nom.) and specific semantics (denoting an indivisible whole), e.g.: liepai ‘a lime grove’: liepa, Burbai ‘the Burbas family’: Būrba; uošviai ‘father-in-law and mother-in-law; their home =, in-law family’: úošvis, lietuviai ‘the Lithuanian nation, etc.’: lietùvis, lietùviai (denoting a countable group of people). 2. Neutralization of oppositions in stems, genders, and accentual paradigms (in the i)o-stem paradigm. Consequently, in opposition to other plural paradigms, this plural paradigm should be regarded as the weak (unmarked) member. 3. A derivational, rather than inflectional, relationship with the singular (or, correspondingly, the ordinary plural). Thus, the collective plural should be treated as an intermediate stage in the grammaticalization of the derivational category of the collective (*āla-collectivum). 2.6.1. To summarize, the following models of the opposition between the two plurals can be reconstructed: 1) a model in which the oppositions in declension, accentual paradigms, and gender are neutralized in the collective plural, e.g.: 2) a model in which the opposition between accentual paradigms is neutralized in the collective plural, e.g.: 2.6.2. Even these models, presented merely for illustrative purposes, show that the concept of two plural variants has considerable explanatory power. Assuming a declension model in which oppositions in stems, genders, and accentual paradigms are neutralized in the collective plural [(i)o stem], the mixing of not only noun plural but also singular paradigms becomes more comprehensible; the spread of final stress and changes in accentual paradigms in general become clearer (especially in cases where Baltic final stress conflicts with the root stress of Ancient Greek and Indo-Aryan), or vice versa; cf. vilkai: Gr. λύκοι, sg. λύκος; OInd. vṛ́kas, sg. vṛ́kas or dūmai, Latv. dūmi: Gr. θυμός, OInd. dhūmās (for more examples, see Иллич—Свитич В. М., 1963, 13* 195 p. 85). The declension pattern mentioned above makes it possible to take a fresh look at forms such as šūnai, žvėriai, tai etc.; it also helps explain some of the alternations between o/ā- and io/ē-stem nouns. The main reason for their emergence was probably the neutralization of the opposition between stems and genders in the collective plural system, which must have encouraged the frequent introduction of the o or io stem into the singular as well—especially in words whose collective use was dominant. This “invasion” of the o or io stem did not affect the same words everywhere and proceeded unevenly; consequently, in individual dialects, competing singular forms with o/ā or io/ē stems arose. Such forms (words) may also occur in the same dialect, but they usually have different meanings, or one of them functions only in idioms and other fixed expressions, e.g. liepa: liepas Trk [this form (word) is used only in comparisons, e.g. aukštas kaip liepas ‘tall as a linden’]. Such competing forms must have become especially numerous as the plural opposition involving neutralization of stem and gender distinctions for collective meaning was disappearing37. Once this meaning had been “forgotten,” the opposition between the two plurals became unnecessary and disappeared in favor of one form or the other (this opposition persisted longest in peripheral and relatively specialized microsystems, such as family names derived from personal names, kinship terms, etc.). Those ā- or ē-stem words whose plural generalized the variant with stem neutralization shifted to the o or io stem in the singular, e.g. plaušai: plauša → plaušas. Words with ā or ē stems survived when the ordinary plural variant was generalized for them. As we can see, this explains not all, but only those alternations between o/ā- and io/ē stems in which the ā- or ē-stem form is older. Thus, some of the alternations mentioned may have arisen for other reasons as well (see Stundžia B., 1978, 112 ff., 118)*. ADDITIONAL ABBREVIATIONS a.p. — accentual paradigm avd — personal name (anthroponym) dpv. — current (Lithuanian) surname inf. — information okn — oikonym p. — south r. — east š. — north v. — west 37 Similarly, o/ā-stem noun alternations were explained by J. Otrembskis. He argued that, since the endings of the secondary plural “-ai paradigms” (see 2.5.1) had for a time been attached not only to o-stem nouns but also to ā-stem nouns, alternations such as laža: plural lažai could have existed. “Where the collective meaning of forms in -ai was forgotten and these forms became merely plural forms of o-stem nouns, corresponding singular forms were also formed on their basis, namely with -as (nominative). Thus, alongside the feminine forms (vyža, laža, spūrga), masculine forms arose—vyžas, lažas, spūrga[s]” (Otręb ski J., 1956, 28). It would follow from this that ā-stem forms are older than o-stem forms. But this cannot be the case in every instance (see Stundžia B., 1978, 118). *The author is grateful to Associate Professor V. Čekman, Associate Professor A. Girdenis, Professor V. Mažiulis, and Candidate of Philology A. Vanagas, who read the manuscript of this article and improved it with valuable comments. 196 LITERATŪRA Bystron J. Nazwiska polskie. – Lwów–Warszawa, 1936. Endzelīns J. Darbu izlase. – Rīga, 1974, t. 2. Girdenis A. Tirkšių tarmės morfologija. – V., 1962 (rankraštis). Girdenis A. Lietuvių kalbos vardažodžio priesagų kirčio susiformavimas. – Kn.: Baltistica. V. 1972, 1 priedas. Grinaveckienė E. Būdingesnieji giminės, skaičiaus ir linksnio vartojimo atvejai Mituvos upyno tarmėje. – Kn.: Lietuvių kalbotyros klausimai. V., 1959, t. 2. Grinaveckienė E. Mituvos upyno tarmės daiktavardis. – Kn.: Lietuvių kalbotyros klausimai. V., 1960, t. 3. Kardelytė J. Gervėčių tarmė. – V., 1975. Kazlauskas J. Lietuvių kalbos istorinė gramatika. – V., 1968. Kuryłowicz J. The Inflectional Categories of Indo-European. – Heidelberg, 1964. Mahlow G. H. Die langen Vocale AEO in den europäischen Sprachen. – Berlin, 1879. Mažiulis V. Baltų ir kitų indoeuropiečių kalbų santykiai. Deklinacija. – V., 1970. Nieminen E. Der urindogermanische Ausgang -āi des Nominativ-Akkusativ Pluralis des Neutrums im Baltischen. – Helsinki, 1922. Otrębski J. Wschodniolitewskie narzecze twereckie. – Kraków, 1934. Cz. 1. Gramatyka. Otrębski J. Gramatyka języka litewskiego. – Warszawa, 1956, t. 3. Schmidt J. Die Pluralbildungen der indogermanischen Neutra. – Weimar, 1889. Stang Chr. S. Zum indoeuropäischen Kollektivum. – In: Stang Chr. S. Opuscula linguistica. Oslo, Bergen, Tromsö, 1970. Stundžia B. Dėl baltų ó-/ā-kamienių daiktavardžių gretybių. – Baltistica. V., 1978, t. 14(2). Valeckienė A. Lietuvių kalbos bevardės giminės įvardžių kilmė. – Kn.: Gramatinės kategorijos ir jų raida (Lietuvių kalbotyros klausimai, t. 18). V., 1978. Бырыла М. В. Беларуская антрапонімія. – Мінск, 1966, т. 1; 1969, т. 2. Ильич-Свитыч В. М. Именная акцентуация в балтийском и славянском. – М., 1963. Мейе А. Введение в сравнительное изучение индоевропейских языков. – М.–Л., 1938. ПЛЮРАЛЬНЫЕ АНТРОПОЙКОНИМЫ И ИХ ЗНАЧЕНИЕ ДЛЯ РЕКОНСТРУКЦИИ СИСТЕМЫ СКЛОНЕНИЯ И АКЦЕНТУАЦИИ Резюме На основании анализа плюральных антропойконимов и некоторых реликтовых апеллятивных форм делается попытка обосновать возможность существования в восточнобалтийских языках противопоставления двух вариантов множественного числа: простого и собирательного. Реконструируются следующие особенности собирательного множественного числа: 1. Специфическая окситонная (4 а. п.) форма на -(i)ai в им. п. и специфическая семантика (значение неразложимого целого), напр.: liepai ‘липняк’: liepa ‘липа’, Burbai ‘семья Бурбы’: Burba ‘Бурба (антропоним); tėvai ‘родители’: tėvas, tėvai ‘отцы’. 2. Нейтрализация в парадигме (i)o-основы: а) оппозиции основ, рода и акцентуационных парадигм, напр.: liepai: liepa, мн. ч. liepos; sūnai ‘сыновья’: sūnūs, мн. ч. sūnūs ‘сыны’; б) оппозиции акцентуационных парадигм: tėvai (4): tėvas (1), lietuviai (4): lietuvis (2) и т. д. 3. Не флективное, а деривационное отношение к единственному (resp. простому множественному) числу; при этом собирательное множественное число относится к типу плюрализованного коллектива, т. е. представляет собой промежуточный этап грамматикализации деривационной категории коллектива (*āja-collectivum). 197 Существование модели деклинации с нейтрализацией оппозиций основ, рода и акцентуационных парадигм в системе собирательного множественного числа делает более понятным процесс смешения парадигм существительного не только во множественном, но и в единственном числе; проясняется также распространение окситонной акцентуации и вообще изменения в акцентуационных парадигмах (особенно в тех случаях, когда „конечное“ ударение балтийских языков противоречит „корневому“ ударению древнегреческого и индийского языков или наоборот, ср. vilkai: др.-гр. lýkoi, др.-инд. vṛkás или dā́tai, латыш. dāti: др.-гр. ὀψός, др.-инд. dhūmās). Эта модель позволяет заново взглянуть на такие формы как debesiai ‘облака’, žvėriai ‘звери’, tai ‘это’ и подобные, и дает ключ к объяснению части дублетных форм (слов) с о/ā- и i̯o/e̯-основами, напр.: plaušas || plaušà, dalgis || dalgė.