Dėl rytų aukštaičių tarmės skirstymo

LIETUVIŲ KALBOTYROS KLAUSIMAI XXXVIII (1997) Apžvalgos. Recenzijos. Pastabos Opinions. Reviews. Observations Überblicke. Rezensionen. Anmerkungen ON THE CLASSIFICATION OF THE EASTERN HIGHLAND DIALECT A dialect can be classified well only when the linguistic area has been thoroughly studied and all the more important isoglosses are already known. However, the practical work of dialect research necessitates classifying (dividing) dialects, even imperfectly, from the very beginning of the work. Before the war, the Eastern Highland dialect had been studied incomparably less than the Western (and Southern) Highland or Samogitian dialects, and it was therefore difficult to divide it into subdialects and even smaller units. Prof. Antanas Salys, following Kazimieras Jaunius in part, proposed in the interwar period that the Eastern Highlanders (whom he called eastern Highlanders) be divided into three subdialects: 1) the pántininkai, who preserve the acute an, am, en, em intact, but change the circumflex forms to ɔn, ɔm, ɛn, ɛm, and likewise change circumflex ą, ę to ɔˀ; ɛˀ; 2) the pón tininkai, who change these sounds to ɔn, ɔm, ɛn, ɛm and ɔˀ; ɛˀ, respectively (from ą, ę), not only in circumflex but also in acute syllables; and 3) the pùntininkai, who have un, um, in, im, ū, ī instead of an, am, en, em, ą, ę. However, closer study showed that things are in fact much more complicated in the areas of his pantininkai, pontininkai, and peripheral puntininkai. In some places there are other sound changes unknown to A. Salys. For instance, around Krekenava and Ramygala, acute an, am, en, em are preserved intact, but circumflex forms change to un, um, in, im; thus, this is essentially a puntininkai-type variant of pantininkai, which A. Salys had not envisaged. Around Panevėžys, these sounds in acute syllables are changed as by the pontininkai, and in circumflex syllables as by the puntininkai, so here again there is a distinctive puntininkai-type hybrid of the pontininkai. Around Pasvalys and Saločiai, conversely, acute forms are pronounced as by the puntininkai, and circumflex forms as by the pontininkai, yielding a pontininkai-type hybrid of the puntininkai. Moreover, the change of en, em is not everywhere analogous to that of an, am: there are dialects in which án, ám become ɔn, ɔm, but én, ém remain intact (Joniškėlis, Vaškai...), making them half pontininkai, half pantininkai. What is more, the change of circumflex ẽn, ẽm often depends on the following syllable: before a soft consonant one finds ẽn, ẽm, while before a hard consonant one finds ˀɔn, ˀɔm. The diversity is further increased by numerous differences in unstressed syllables. The change of vowels ą, ę also often does not conform to A. Salys’s account. Many pantininkai and pontininkai change stressed ą to o regardless of accent, while they pronounce the unstressed vowel as ɔ; only on the periphery around Krekenava is ā found under stress and a in an unstressed syllable. Around Pasvalys, ū or ɔ is pronounced depending on 237 stress (not accent!). In place of ę in the middle of a word, many speakers pronounce an open e (long when stressed, short when unstressed, regardless of accent), while in the ending they pronounce a close ę (around Krekenava, e). In the east, around Pasvalys, ī or ę is pronounced depending on stress (not accent). In these circumstances, it is impossible to define precisely who the pantininkai and pontininkai are. They could be distinguished from one another only by the pronunciation of stressed acute ān, ām (these two only!), but this would have theoretical, and by no means practical, significance. For there are very few words with stressed acute án, ám that the pontininkai actually change to ón, óm—only a dozen or so (the more common ones are ántis, lángas, pántis, sámtis...), excluding their suffixed derivatives. Much more often, the pontininkai area uses unchanged (unshifted) acute ān, ām, specifically those of secondary origin (e.g., támsta, dative singular tám, gerám, dative plural vaikám, first person plural dírbam, etc.), in loanwords and words borrowed from the standard language (ángelas, bánkas, táŋkas, etc.), in words of onomatopoeic origin (ámtelėti, bámbtelėti...), and so on. A person can speak all day without using a single one of those dozen or so words with shifted án, ām. What is more, for consistency we would also have to distinguish p ú o n t i n i n k a i; cf. the pronunciation púontis ‘pantis’ in Jotainiai. Finally, the meaninglessness of such a classification is shown by the fact that it allows us to classify only the northwestern corner of the Eastern Highland area, while the rest—several times larger and belonging entirely to the puntininkai—would thus remain unclassified.1 In 1964, while refining the classification of dialects with Aleksas Girdenis, we proposed classifying the Eastern Highlanders on a different basis: in the Eastern Highland area west of roughly the Papilys–Vabalninkas–Kavarskas–Giedraičiai–Maišiagala line, where long and short vowels are not distinguished in unstressed syllables (they are all short), we proposed distinguishing the Panevėžys subdialect, in which all short final vowels merge into a single reduced vowel (back after a hard consonant, front after a soft one), and the Širvintos subdialect, where there is no such merger. East of that line, where the distinction in vowel length is preserved in unstressed syllables, we proposed distinguishing the Utena, Anykščiai, Kupiškis, and Vilnius subdialects according to the same criteria as in A. Salys’s classification, but giving the subdialects different names—not nicknames, but names based on the most important town in each area. This classification of the Eastern Highland dialect, like the entire revised classification of dialects, gradually took hold and is now, one might say, in universal use. Later, as the antiquity and origin of the distinguishing features of dialects were studied in greater depth and the history of Lithuanian dialects was investigated more generally, the view emerged that the established Anykščiai and Kupiškis subdialects had, over time, split in two the older Utena area (the most typical Eastern Highland region). After that, an innovation that had advanced from the west 1A detailed critique of A. Salys’s classification is presented in: Lietuvos TSR Mokslų akademijos darbai, A serija 3(31), 1969, 142–146. 238 The Panevėžys and Širvintos dialect wave gradually flooded the western side of the so-called “Utena” area; only in the north, next to the Kupiškis dialect, did the so-called Vabalninkas speech remain, classified as Northern Panevėžys but still relatively close to the Utena dialect. Moreover, it also became clear that, besides the Vabalninkas speech, another remnant of the “Utena” dialect exists on the right bank of the Šventoji between Kavarskas and Vidiškės (Gelučiai and other villages).2 Recently, Žaneta Urbanavičiūtė-Markevičienė published a very interesting article3, arguing that the latter “Utena” remnant around Gelučiai once covered a larger area—namely, all the villages of the Kavarskas and Vidiškės parishes (a very important observation!)—but that, in her view, this area should be assigned not to Southern Panevėžys but to Širvintos. She raised the question of whether it was worthwhile to separate the Širvintos dialect from Southern Panevėžys at all, since these two East Aukštaitian subdialects are very close. There is supposedly more justification for separating Northern Panevėžys from Southern Panevėžys and treating it as a distinct East Aukštaitian subdialect. Yet, she argues, there is no longer any point in dismantling the established classification. The following may be said in response to the doubts and arguments raised. First, as to assigning the Kavarskas and Vidiškės parishes to the Širvintos dialect. Like the Vabalninkas area, these parishes form peripheral Panevėžys dialects (the Vabalninkas area is Northern Panevėžys, and these parishes Southern Panevėžys), interspersed with features of the Utena dialect preserved as relics (in the Vabalninkas area) or with features of its present-day continuation, the Širvintos dialect (in the Kavarskas and Vidiškės parishes). The transitional nature of the Kavarskas–Vidiškės speech is obvious. For example, in Vidiškės one says dárbus ‘work’, acc. pl. šakùs ‘branches’, as throughout the Southern Panevėžys area (cf. Širvintos dárbas, šakàs), but šakà (cf. Širvintos šaką, Southern Panevėžys šakų), dirba (Širvintos dírba, Southern Panevėžys dírbu), kója (Širvintos kóje, Southern Panevėžys kóji), with the preserved “Utena” reflex -a (also in the Širvintos form šakà). In these circumstances, this transitional speech could be assigned either to Southern Panevėžys or to Širvintos. However, since the development of the ending a proceeded along the path a → α → u (ų), this speech, which has reached the final outcome u (though not in all forms), is more properly assigned to Southern Panevėžys than to Širvintos, which has no such u (there the stage is only α). Of course, dialectologists need to reach agreement on this, as in other cases involving transitional dialects. As for the question of whether it was worthwhile to separate the Širvintos dialect from Southern Panevėžys, the doubt is justified in this case too. In fact, Southern Panevėžys, especially its transitional dialects closer to Širvintos, is quite close to the Širvintos dialect. However, the fate of the short vowels in word endings clearly distinguishes the two: Southern Panevėžys merged those vowels into a single sound (after hard consonants 2 Z i n k e v i č i u s Z. Dėl rytų aukštaičių patarmių istorijos. – Baltistica 16(2), 1980, 138–139. 3 U r b a n a v i č i ū t ė - M a r k e v i č i e n ė Ž. Dar kartą apie rytų aukštaičių tarmių ribas. – Kalbotyra 43 (1), 1994, 85–87. 239 the back vowel υ, after palatalized consonants—the front vowel ę), whereas the Širvintos dialect has no such merger. This resulted from the different stages reached in the development of a → α → υ in word endings: Panevėžys has υ, while Širvintos has reached only α. In the transitional Kavarskas–Vidiškės speech, as we have seen, the merger process is not yet complete: some forms have retained a and u, and after palatalized consonants the distinction between e (ia) and i remains. If Širvintos and Southern Panevėžys were combined into a single subdialect, it would be difficult to formulate its defining criterion. All Panevėžys dialects—Southern and Northern—are united into one subdialect by a phonologically very important feature: the merger of short vowels in word endings. Separating Northern Panevėžys as a distinct East Aukštaitian subdialect, despite its considerable differences from Southern Panevėžys, would create substantial practical difficulties. Consider the most striking difference between Northern and Southern Panevėžys: the reduction of short vowels in word endings to murmured sounds and their subsequent loss. It is very difficult—indeed impossible—to determine where a sound has become murmured and where it is merely strongly reduced. There is even disagreement about where it has disappeared altogether and where some remnants remain. Clearly, a feature that does not form a distinct isogloss cannot serve as a criterion for distinguishing a subdialect; it can only be regarded as a difference between parts of that subdialect. The same applies to the other important differences, many of which likewise do not form clearly discernible isoglosses (perhaps with the exception of the universal retraction of stress, though this is not characteristic of all Northern Panevėžys—the Biržai corner is an exception), such as the more pronounced vowel pronunciation in stressed syllables across the Northern Panevėžys area (compared with Southern Panevėžys), the more pronounced pitch accents, and so on, as noted by the article’s author. Moreover, all the distinguishing isoglosses are highly dispersed and do not form a clear boundary. On the other hand, diachronically speaking, Southern Panevėžys is still, we might say, not entirely typical Panevėžys, if Northern Panevėžys is taken to be the typical variety. Cf. the development of short vowels in word endings across the Panevėžys area: a, u, e, i → υ, ę →. Therefore, the article’s author’s doubts are entirely understandable, especially if we view these dialects synchronically. But dialect classification cannot disregard diachrony (the history of the dialect); it cannot ignore it. All this makes it clear that changing the established classification of the East Aukštaitian dialect would achieve nothing and would only create confusion. Moreover, the article’s author also writes that there is no point in dismantling the established classification. Let us continue to treat the Kavarskas–Vidiškės and Vabalninkas speeches as transitional varieties of the Panevėžys subdialect. At the boundaries between dialect units, isoglosses do not always form distinct bundles; they are often highly dispersed (even reaching the middle of a dialect). Consequently, selecting the more important isoglosses that determine dialect classification is not an easy task, and researchers often have to rely on agreement. In the article mentioned above (and in other recent works), I generally discuss the development of dialect features, not the classification of dialects, as readers sometimes seem to understand. This gave rise to Ž. Urbanavičiūtė-Markevičienė’s inaccurate claim that I had assigned the Gelučiai speech to the Utena dialect (p. 85). I did not do so. I consider it merely close to the Utena dialect (as are the Vabalninkas area and its surroundings). 240 a dialectal variety reminiscent of the Širvintiškiai (p. 139). By the way, I put the word Uteniškiai in quotation marks because I had in mind the former area of that dialect before it was split in two. Žaneta Urbanavičiūtė-Markevicienė’s article also expresses other ideas of interest to dialectologists, with which one cannot but agree. Her reflections on the dialectal affiliation of the town of Ukmergė are interesting: she regards it as the object of contact between the Širvintiškiai and southern Panevėžiškiai dialects. I have always held the same view4; I have never unconditionally assigned Ukmergė to either the Panevėžiškiai or the Širvintiškiai (the author seems inclined to assign it to the Širvintiškiai). The town lies at the point where three subdialects of the Eastern Aukštaitian dialect meet (Anykštėnai approach from the northeast), but it belongs to none of them. The speech of the town’s residents is mixed, and elements of all three subdialects can be observed in it. As in all towns, it is strongly influenced by the standard language. Because it lies at the intersection of dialects, I could not use the name Ukmergė to name a subdialect (for which the people of Ukmergė sometimes reproach me), although this town is my birthplace. The author’s reflections on the so-called “ruotininkai,” who pronounce rōtas, even ruōtas ‘wheel’, are interesting. Regarding this pronunciation, she writes: “The reason is probably the system of two vowel lengths. For here the vocalism is essentially like that of the Širvintiškiai, except for ‘rotininkavimas’. If r̄tas, ǵēras had been retained, as among the other Anykštėnai, ɔ, ę would have to lengthen, and a, e would merge with o, ė. And in place of lapas, and in place of lopas, we would have lōpas in stressed position, and lapėlis in unstressed position. The merging of vowels o and ė with a and e would be very inconvenient; the vowel system would become very undifferentiated. It must be borne in mind that in these dialectal varieties, a, o and uo, as well as e, ė and ie, already merge in unstressed position. The merging of two more vowel pairs would certainly be too much. Therefore ɔ, ę were diphthongized” (p. 86). One could agree with this conclusion. It should be noted, however, that the southernmost corner of the Anykštėnai area itself—the southwestern part of Lyduokiai parish toward Ukmergė (the villages of Nuotekai, Juodausiai, Jogvilai, etc.)—does not “ruotininkavimas” (the pronunciation here is rōtas), although its vocalism is closer to that of the Širvintiškiai than that of those who “ruotininkavimas” (where there is a tendency toward “žadininkavimas”). “Puontininkavimas” (Jotainiai) should also be recalled. The vocalism of all these dialectal varieties ought to be thoroughly investigated from the standpoint of phonology. The author’s observations on the unusual monophthongization of au, ai, ei in falling-accent and unstressed syllables in the Anykštėnai and Širvintiškiai dialectal varieties of this region are noteworthy, especially at the end of an open word (pronounced matū ‘I see’, miškī ‘forests’, gáudi ‘you heard’), which she relates to the complete transfer of stress to the second component of the diphthong. One cannot but agree with the author, who highlights the importance of parish boundaries in the establishment of isophones. This has long been observed. Parish boundaries are old and stable. District and local-area boundaries are another matter: they are relatively new and are also continually changing. Dialectologists may disregard the latter, but parish boundaries must not be ignored. Zigmas Zinkevičius 4 See Z i n k e v i č i u s Z. The dialect of the Ukmergė district. – Mūsų kalba 4, 1978, 28–31. 241