Kirčio atitraukimas XX a. 6–7 dešimtmečių rytų aukštaičių uteniškių patarmės rankraštiniuose šaltiniuose

Acta Linguistica Lithuanica 2025, vol. 92, pp. 1–15 Contents link / Turinio nuoroda ISSN 2669-218X (online / elektroninis) STRESS RETRACTION IN MANUSCRIPT SOURCES OF THE EASTERN AUKŠTAITIAN UTENA SUBDIALECT FROM THE 6TH–7TH DECADES OF THE 20TH CENTURY Stress Retraction in the Manuscript Sources of the Eastern Aukštaitian Subdialect of Utena from the 1950s and 1960s VIL IJA RAGAIŠIENĖ Institute of the Lithuanian Language E-mail: [email protected] ORCID ID: https://orcid.org/0000-0002-5026-4667 Research areas: accentology, dialectology, ethnolinguistics, lexicography. https://doi.org/10.35321/all92-07 ABSTRACT The article examines, from the perspective of accentuation, material from manuscript sources of the Eastern Aukštaitian Utena subdialect from the 6th–7th decades of the 20th century, preserved in the Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language. The study seeks to identify which stress retraction rules operated in the Utena subdialect in the mid-20th century and to discuss the nature and distribution of the phenomenon across the subdialect area. The analysis reveals an uneven and rather late formation of stress retraction as a rule in the Utena subdialect. Based on data from texts of the Eastern Eastern Aukštaitian subdialects—Kupiškis, Anykščiai, and Utena—published in Leipzig by the German linguist Franz Specht in Litauische Mundarten gesammelt von A. Baranowski in the 7th–9th decades of the 19th century, and on descriptions of subdialects from the 6th–7th decades of the 20th century, new insights are provided into the nature of stress retraction in the Utena subdialect. It is hypothesized that in the northern subdialects and western peripheries of the Utena area, assigned by the compilers of the Atlas of the Lithuanian Language (1982) to the area of weak conditional stress retraction, a more intensive rather than weak conditional stress retraction rule was forming in the 6th–7th decades of the 20th century. KEYWORDS: Eastern Aukštaitian Utena subdialect, stress retraction rule, “two-peaked” stress, syllable, manuscript source. ABSTRACT The article examines manuscript sources of the Eastern Aukštaitian subdialect of Utena from the 1950s and 1960s, stored in the Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language, from the perspective of accentuation. The study aims to identify the stress retraction phenomena that were in effect in the Utena subdialect in the mid-20th century and to discuss the nature and distribution of this Copyright © 2025 Vilija Ragaišienė. Published by the Institute of the Lithuanian Language. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. // Išleido Lietuvių kalbos institutas. Šis straipsnis yra atviros prieigos, platinamas pagal „Creative Commons“ priskyrimo licencijos sąlygas, leidžiančias neribotai naudoti, platinti ir atkurti turinį bet kokioje laikmenoje, nurodant autorių ir šaltinį. Received: / Gauta: 2025-04-18. Accepted: / Priimta: 2025-05-14. 1 phenomenon within the subdialect area. The analysis reveals an uneven and rather late formation of stress retraction as a rule in the Utena subdialect. Drawing on 19th-century texts (from the 1860s to the 1880s) covering the subdialects of Kupiškis, Anykščiai, and Utena (which belong to the eastern part of the Eastern Aukštaitian dialect), published by the German linguist Franz Specht in Litauische Mundarten gesammelt von A. Baranowski in Leipzig, as well as 20th-century descriptions of subdialects (from the 1950s and 1960s), the article provides new insights into the nature of stress retraction in the Utena subdialect. It is assumed that between the 1950s and mid-1960s, a more intensive rather than weak conditional stress retraction rule was forming in the northern part and western periphery of the Utena subdialect, an area previously assigned to the weak conditional stress retraction by the compilers of the Atlas of the Lithuanian Language (1982). KEYWORDS: Eastern Aukštaitian subdialect of Utena, stress retraction rule, double-peaked stress, syllable, manuscript source. INTRODUCTION The stress retraction phenomena operating in the Eastern Aukštaitian Utena subdialect have so far received little attention from researchers. In the scholarly literature, the nature of stress retraction in this subdialect area is discussed rather fragmentarily. It is most often indicated that stress is retracted from a short ending only in part of the Utena area—in the northwestern subdialects and western peripheries adjoining the Anykščiai and Širvinti subdialects (Zinkevičius 1966: 447, map no. 2; LKT 1970: 34–35). Earlier dialectological studies stated that two rules operated here—weak conditional and more intensive conditional stress retraction: around Molėtai, Suginčiai, Debeikiai, Panemunėlis, and Šuvieka, stress is retracted from a short ending onto a syllable consisting of a long vowel or ie, uo, while in the vicinity of Rokiškis and Obeliai it is retracted onto any long penultimate syllable (LKA II 123–124, map no. 105; Urbanavičiūtė-Mark evičienė, Grinaveckis 1992: 12; Zinkevičius 1994: 74; Markevičienė 1999: 16). In recent years, the view has been adopted that only more intensive conditional stress retraction occurs in the area of the subdialects under discussion (LKTCH 2004: 126–127; RAU 2010: 35). At the very edge of the Utena area—in the northern part of the Vabalninkas subdialect—stress is retracted from a short ending onto either a long or a short penultimate syllable (Zinkevičius 1966: 447, map no. 2; LKA II 124, map no. 105; LKTCH 2004: 126). The data presented in the scholarly literature on changes in stress retraction as a rule in the northwestern corner and certain western peripheries may indicate ongoing changes in this phenomenon throughout the subdialect area. Sources, aim, and methods. The research material was collected from manuscript sources of the Utena subdialect from the 6th–7th decades of the 20th century preserved in the Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language. In collecting material for the Atlas of the Lithuanian Language¹, over two decades dialectologists investigated the populated localities (points) belonging to the Eastern Aukštaitian Utena area² and prepared dialectal descriptions of them (see Fig. 1)³. Data were collected from nearly all ¹ LKA I; LKA II; LKA III. ² The localities to be investigated (the names and numbers of the points) were established by the compilers of the Atlas of the Lithuanian Language. ³ A manuscript source for a point usually consists of a presentation of the locality investigated, its linguistic features, and its informants; answers to questions on the lexicon, phonetics, morphology, and syntax in the program for collecting material for the Atlas of the Lithuanian Language; and transcribed texts (see Ragaišienė 2024: 181–183 for more detail). 2 all Utena localities—89 sources (approximately 2,100 pages) from 66 points were found in the archive4. The analysis of the manuscript-source material also draws on texts of the Utena subdialect published in the 1970 anthology Lithuanian Dialects. Summarizing the data found in the sources under study, it may be stated that they are sufficient for discussing the stress-retraction tendencies prevailing in the subdialect area in the mid-20th century. The aim of the article is to describe the stress retraction phenomena operating in the Eastern Aukštaitian Utena subdialect in the mid-20th century and to reveal their areal distribution across the subdialect area. The main objectives of the study are: 1) to compile an inventory of words found in manuscript sources in which stress may be retracted; 2) to assess the ratio between forms retaining stress on the word ending and those retracting it onto the penultimate syllable; 3) to establish the nature of stress retraction across the subdialect area. The study applies a descriptive-analytical method, as well as comparative and internal-reconstruction methods. The research should be relevant not only to scholars investigating changes in the accentuation phenomena of Lithuanian dialects but also to researchers of the development of accentuation in the Baltic languages. This is primarily because the conclusions presented in the article supplement the published results of studies of 6th–7th-decade-20th-century manuscript sources of the Kupiškis and Anykščiai subdialects and provide new data on stress retraction phenomena in the area of the Eastern Eastern Aukštaitian subdialects (Ragaišienė 2018: 176–201; 2020a: 85–94; 2020b: 211–226). STRESS RETRACTION PHENOMENA: CHANGE AND TENDENCIES The manuscript-source material shows that the accentuation system of the Eastern Aukštaitian Utena subdialect in the 6th–7th decades of the 20th century was rather complex. During the period under study, several interrelated accentuation phenomena operated: leveling of accentual paradigms, the formation of two accentuation types—stem and ending accentuation, accentual analogy, retention of relics of ancient root accentuation (for example, u-stem adjectives), and the formation of the stress retraction rule (Ragaišienė 2024: 184–185). It is not always clear which stem-stressed forms arose through paradigm leveling or the formation of the stem-accentuation type, and which are associated with stress retraction. This is particularly true of subdialects in which stress retraction is irregular. On the other hand, the accentuation phenomena under discussion, except for stress retraction, are associated with the accentuation of inflected words. Thus, in assessing the nature of stress retraction in specific subdialects, it is necessary to rely on the accentuation of verb forms and indeclinable words. The research material shows that in the 6th–7th decades of the 20th century stress was retracted with varying intensity in part of the Utena subdialect area—regularly in some places, inconsistently in others, or only in isolated forms (see Fig. 1). According to the nature (intensity) of the phenomenon, the area under discussion can be divided into three areals. In the northern Utena subdialects, that is, in the Užūbaliai (87)5 – Ūnuškis 4 No data are available from three points—Vyžuonos (275), Aknystėlės (307), and Deguliai (345). The anthology Lithuanian Dialects contains only a text recorded in the Vyžuonos subdialect, in which no instances of stress retraction were found (LKT 1970: 321–322). 5 The number of the point in the Atlas of the Lithuanian Language is given in parentheses. 3 (88) – Rokiškis (147) – Obeliai (148) – Svėbiškis (150), stress was regularly retracted from a short ending onto any long penultimate syllable, e.g.: aĩni⁶ ~ eini⁷, dur̃nų⁸, gal̃va, ruñka ~ ranka, bėda, kãrtus, luñkus ~ lankus, nuveĩni ~ nueini, nuveĩsi ~ nueisi, su mė́sų ~ mėsa, septintì, svìrnì ~ svirne, uñksti ~ anksti, pìnki ~ penki, par vaŕtus, vìena, óška ~ ožka, mér̃ga, žmõna Užb; skū́ra, aĩni ~ eini, aĩnų ~ einu, gal̃va, liẽpsna, liẽtus, jaũni, kãlbi Bč; aĩni ~ eini, aĩšų / aĩšiu ~ eisiu, aļga, ãrti, keŕti ~ kerti, šìrdis, bárždu ~ barzda, gìrdžių ~ girdžiu, su karū́nų ~ karūna, laĩku, mãrti, meŕgų ~ merga, pìñkus ~ penkius, sõdi ~ sode, šáltą žiemą, šmaĩkštus, šlaũnis ~ šlaunis, vaĩkus, jaũni, vìeni, nuõga Rk; aĩnų ~ einu, neužeĩnų, balsù, diẽna, kãlbu maldẽlas ~ maldeles, laĩki ~ laike, prakeĩkta meŕga, šeĩma, žmõna Ob; aĩsi ~ eisi, diẽna, diẽnas, dur̃na, gal̃va, laĩki ~ laike, piẽla⁹, pìñkis ~ penkis, pìñkus ~ penkius, ruñka ~ ranka, vìena Svb¹⁰. Stress was also regularly retracted in the northwestern corner, in the Utena subdialects beyond the Anykščiai and Kupiškis areas, e.g.: aĩ.ši ~ eisi, bùrobùs, su dãžiaugsmu ~ džiaugsmu, gal̃va ~ galva, gãmta, gì.vi ~ gyvi, juõsva, kapù.stus ~ kopūstus, luñ.kus ~ lankus, nebėra, pìl.ka ~ pilka, pìj.ki ~ penki, su ruñ.ku ~ ranka, šál.ta ruñ.ka ~ šalta ranka, vìena, vìeni Gtč; aĩ.ni ~ eini, baĩ.gu ~ baigiu, bal̃.ti ~ balti, blaũ.zda ~ blauzda, gė́l.svi ~ gelsvi, diẽna, dùñ.gus ~ dangus, ì.ra ~ yra, kaŕ.tu ~ kartu, kir̃.vu ~ kirviu, laũ.ki ~ lauke, lìñ.ksma ~ linksma, lìñ.ksmi ~ linksmi, meŕ.ga ~ merga, miẽgu, pavaŕ.ksti ~ pavargsti, pìj.ki ~ penki, põña ~ ponia, ruñ.kas ~ rankas, sãl.di ~ saldi, sukal̃ti ~ sukalti (kuolai), trùm.pi ~ trumpi, tvõra, val̃.džia ~ valdžia, vaŕ.tus ~ vartus, vìena, žù.si.s ~ žąsis Drs¹¹. In the manuscript sources of the Vabalninkas and Darsiškiai subdialects, isolated instances of intensive conditional stress retraction were found, e.g.: su kìtu ~ kita, vėdì Vb; ãkis, šãku ~ šaka, vìsi Drs¹². In the manuscript sources of the subdialects assigned to the area discussed, no forms were found in which stress was not retracted from a short ending onto a long syllable, nor were any instances of secondary or “two-peaked” stress¹³ found. ⁶ The transcription of words varies in the sources under study, and inaccuracies or accentuation errors occur. The transcription signs in the examples presented in the article have not been corrected, and errors have not been amended. ⁷ After the tilde sign (~), transposed word forms, which are not stressed, are given. If the transposed form differs from the transcribed form only in stress, it is not provided. ⁸ The loanword durnas, -à (4) is used in the meaning ‘stupid, unintelligent’. ⁹ The Slavicism pielà (4) is used in the meaning ‘saw’. ¹⁰ Not all examples found in the manuscript sources are presented in the article, and their frequency of recording is not indicated. Quite a few words are attested several or a dozen times in the subdialect descriptions. For example, in the description of the dialectal features of the Darsiškiai subdialect, the form vìena is given 11 times, ì.ra ~ yra—10 times, su põnų ~ ponia—4 times, and so on. ¹¹ Stress is also regularly retracted in the text of the Darsiškiai subdialect published in the anthology Lithuanian Dialects (LKT 1970: 273). ¹² These forms should not be considered errors. The description of the dialectal features of the Darsiškiai subdialect states that “the dialect retracts stress conditionally. <...> universal stress retraction is sometimes observed, but such cases are very rare” (Gargasáitė, Misiūnaitė 1958: 6–7). ¹³ For more on the term “two-peaked” stress, see Girdenis 1978: 75–76. “Two-peaked” stress functions in the Western Aukštaitian Kaunas, Eastern Aukštaitian Anykščiai, Utena, Vilnius, and other subdialects (Girdenis 1978: 75–76; Mikulėnienė, Morkūnas 1997: 16; Markevičienė 2001: 15–16; Urbanavičienė 2006: 224; Ragaišienė 2020b: 214, 219 et seq.). This stress has also been recorded in the Southern Aukštaitian area (Leskauskaitė 2004: 164). 4 Figure 1. The nature of stress retraction in Utena manuscript sources from the 6th–7th decades of the 20th century From the perspective of stress-retraction intensity, the same areal includes the northern Panemunėlis (177) – Aleksandravėlė (180) – Kriaunos (179) – Sūvieko (181) subdialects and the western peripheries of the subdialect area—the environs of Debeikiai (274) – Suginčiai (379) – Bijutiškis (479) – Giedraičiai (504) – Alanta (377) (see Fig. 1). In some places stress was retracted rarely (only isolated cases were found), elsewhere irregularly (forms with stress retraction usually account for approximately 26–32 percent). The area under discussion was likewise not uniform in terms of the nature of stress retraction during the period under study. The Atlas of the Lithuanian Language states that a weak conditional stress-retraction rule operated here in the mid-20th century (LKA II 123–124, map no. 105). The manuscript-source material shows that more intensive conditional stress retraction was forming in the northern Utena subdialects—stress tended to be retracted here from a short ending onto any long penultimate syllable, e.g.: ai̇̃nu ~ einu, pīlna karieta, liėtus, par dirvas, diėna, dirva, pī̃kas ~ penkias, prigim̃tis, daikteĩ ~ daikte, viėna, gir̃džu ~ girdžiu, pōna ~ ponia, rūnka ~ ranka, ūntri ~ antri, užmiẽgu, žmōna Alk. The formation of the rule is also evidenced by the use of forms with “two-peaked” stress (cf. Urbanavičiūtė-Markevicienė, Grinaveckis 1992: 12; Stundžia 1995: 70; Ragaišienė 2020b: 214–215; 2024: 185), cf. forms with “two-peaked” and retracted stress recorded in the same subdialect, e.g.: cviekùs ~ cviekus¹⁴, kalavār.tū ~ kalavartų¹⁵, ¹⁴ Cviėkas (4) ‘nail’. ¹⁵ Kalavartas (2) ‘spinning wheel’. 5 laũ.kùs ~ laukus, piñ.kùs ~ penkius, piẽtus ~ pietus, piñ.kì ~ penkti, sergù ~ sergu, skūrã ~ skūra, šìltã ~ šilta, vaĩ.kùs ~ vaikus, viẽnà / kàrtì, laũ.kus ~ laukus, piñ.kus ~ penkius, viẽnà Pnm¹⁶; aĩnì ~ eini, viẽnà ~ viena / aĩšù ~ eisiu, viẽnà, šìmtus, vaŕpus Krn. The establishment of intensive rather than weak conditional stress in the northern subdialects is shown by two facts. First, forms in which stress should be retracted onto a long syllable are pronounced with “two-peaked” stress, e.g.: šùnùs ~ šunus, taŕ.nùs ~ tarnus, saldõtùs ~ saldotus¹⁷, karūnãs ~ karūnas, ĩrã ~ yra, eĩ.nì ~ eini, maĩ.šù ~ maišu, piñ.kùs ~ penkius, šõkùs ~ šokius, mẽr.gù ~ merga, žmõgus ~ žmogus, vaŕ.dù ~ vardu, ke.kvienà dienà ~ kiekviena diena, devỹnì ~ devyni, uñ.kštì ~ anksti, kaŕ.tùs ~ kartus, kẽr.tù ~ kertu, viẽnà ~ viena, nebepra keĩ.kta ~ nebeprakeikta, õrì ~ ore Rg. Second, the alternating use of forms retaining stress and retracting it onto the long penultimate syllable, e.g.: biẽdnà / biẽdna ~ biedna, i.rã / ĩrã ~ yra, piñkì / piñkì ~ penki, šìrdis / šìrdis ~ širdis, etc. Rg. The situation during the period under study was different in the western peripheries of the subdialect area adjoining the Anykščiai and Širvintiškiai subdialects. In the texts of more than half of the subdialects, irregular stress retraction onto any long penultimate syllable is attested, e.g.: aĩnù ~ einu, aĩšù ~ eisiu, bẽda, juõdi, mẽrga, piŕštu, šìmtus, švìesa, ùntra ~ antra, viẽnà, žmõną / kìrviù ~ kirviu, miegù, vaĩkùs, etc. Čkn; daĩktì ~ daikte, dūdù ~ dūda (groja), juõda, kaũlas ~ kiaules, pasṍeikši ~ pasveiksi, piñkis ~ penkis, prõtu ~ protu, šìmtus, viẽnà / aĩsmì ~ eisme, baisùs, barzdù ~ barzda, burnã, neĩslaidùs ~ neišlaidus, neĩšù ~ neisiu, nusperkì, piñkãs ~ penkias, sunkù, šeimù ~ šeima, vaikù, vaikùs, etc. Skd; aĩšù ~ eisiu, kelìntà, laũkì ~ lauke, lìndì ~ lendi, lìėtus, miẽstì ~ mieste, nuõga põna ~ ponia, piñkus ~ penkius, piñkì ~ penki, šìrdis, švìesa, vaŕkstu ~ vargstu, viẽnà, viẽtą / aĩnù ~ einu, aĩšì ~ eisi, ausìs, baigì, dainã, kalbì ~ kalbi, neb eĩnù, neĩšù ~ neisiu, su šeimù ~ šeima, piñkì ~ penki, piñkus ~ penkius, suprantù ~ suprantu, urkstì ~ anksti, verkšì, etc. Sgn; daržùs, nekũntrì ~ nekantri, tvõra, šìmtus, viẽnà / lietùs, mergã, miršù ~ mirsiu, piñkì ~ penki, viršùs, etc. Grs; aĩšù ~ eisiu, liẽtus, piñ.kùs ~ penkius, riẽstą, õšką ~ ožką, skìlas ~ skyles, vaĩ.kùs ~ vaikus / aĩšù ~ eisiu, barzdã, ilgì, jaunã, i.rã ~ yra, sunkù ~ sunku, uždėta, etc. Vdn¹⁸. In other subdialects, isolated cases were recorded in which stress is retracted from a short final syllable onto a syllable consisting of a long vowel, ie, uo, or the diphthongs ei, ui, e.g.: kariẽta¹⁹, kuĩnu, kuĩnus, rū.bus ~ rūbus (cf. ru.bùs ~ rūbus) Lln. Few such cases were found in the area under discussion²⁰. ¹⁶ Linguists who investigated the dialect of the village of Robliai, belonging to the Panemunėlis point, noted that the oldest-generation informants did not retract stress from the short ending onto the long root, representatives of the middle generation “pronounced two stresses” in such words, while the young retracted stress consistently (Grinaveckienė, Gargšaitė 1956: 4). In the narratives of the oldest-generation representatives presented, isolated cases of stress retraction were found, while somewhat more instances of “two-peaked” stress were recorded. ¹⁷ Saldõtas (2) ‘soldier’. ¹⁸ In the texts of the Suginčiai subdialect, forms in which stress is retracted from a short ending onto any long penultimate syllable account for approximately 38 percent (51 cases out of 134); in the narratives from the Gireitiškiai point, approximately 26 percent (8 cases out of 31); and in the material collected in the vicinity of Videniškiai, approximately 22 percent (7 cases out of 33). ¹⁹ The form was recorded three times. ²⁰ Some of these forms may also be associated with morphophonological paradigm leveling or the formation of the stem-accentuation type (Ragaišienė 2020b: 222; 2024: 185). 6 Manifestations of stress retraction in the 6th–7th decades of the 20th century were observed in the eastern and southeastern Utena subdialects, which the compilers of the Atlas of the Lithuanian Language did not assign to the area of weak conditional stress retraction. The manuscript-source material shows that this area and the western peripheries do not essentially differ in terms of stress retraction. Forms with “two-peaked” stress were also recorded in the eastern and southeastern subdialects, e.g.: aĩ.nù ~ eĩnu, audė́ja, sùnkù ~ sunku, telõkùs ~ teliokus, ùnkstì ~ anksti, vaĩ.kù ~ vaiku, viẽną Spt, etc.²¹ As in the sources from the western subdialects, few instances of this stress type were marked in the manuscripts of the eastern and southeastern subdialects. This indicates that in the 6th–7th decades of the 20th century the phenomenon in question may have been of varying strength in the Utena subdialect, and therefore secondary stress was probably not marked in all cases. The penultimate long syllable—the first component of “two-peaked” stress—may have been pronounced even longer than the stressed syllable and may not have differed greatly from retracted stress (cf. Stundžia 1989: 435; Girdenis 2003: 95; Ragaišienė 2018: 184). It cannot be ruled out that in some cases researchers of the subdialect features may have indicated stress only on the penultimate syllable. The spread of the phenomenon in the eastern and southeastern part of the subdialect area is also evidenced by occasional isolated instances of stress retraction (cf. Ragaišienė 2024: 185). In some subdialects stress is more often retracted from a short ending onto a syllable consisting of ie, uo, or a long vowel, e.g.: liẽtus (cf. liẽtus ~ lietus, lẽ.taũs ~ lietaus), raudõ.ni ~ raudoni (cf. raudonõ.s ~ raudonos), sū̃nus ~ sūnus (cf. su sūnum̃), viẽdrus ~ viedrus²², viẽna ~ viena (cf. vienà ~ viena), viẽni ~ vieni (cf. viẽni ~ vieni) Kkt; diẽna, geltõna (cf. gelta.nũ ~ geltonų), negì.va ~ negyva, nẽra, sū̃nus, viẽna (cf. viẽną) Brč²³. In other subdialects, as in the western peripheries of the subdialect area, stress is retracted onto any long syllable, e.g.: aĩme ~ eime, su bar̃zdu ~ barzda, diẽnas, grìbus ~ grybus, su plūgù ~ plūgu, laĩškì ~ laiške, liẽtus, ruñkas ~ rankas, suprantũ ~ suprantu, šìntus, viẽna kapeĩka, žmõna Jnš; balti, jermẽkus²⁴, kalbi, kàrtu aĩnu ~ einu, kir̃vu ~ kirviu, krū́va ~ krūva, miẽgu, piñkas ~ penkias, diẽnas, širdis nesveĩka, blogà žmõna ~ bloga žmona, supilti Arn²⁵. In both areas, forms with retracted stress may have been used as variants, e.g.: dienà / diẽna, skurà / skūrà, sū̃nus / sū̀nus, vienà / viẽna, tvorà / tvõra, etc. Lb²⁶; baĩsi / baĩsì, baĩsi ~ baisi, gýva / gì.va ~ gyva, jì.rà / jì.ra ~ yra, kapeĩkas / kapeĩkas, põne / po.nė ~ ponia, su.nùs / sū̀.nus ~ sūnus, šìntus / šìntus, viẽną / viẽna, etc. Bjt; i.rą / ì.ra ~ yra, pinkì / piñki, piñkì ~ penki, vienà / viẽna Jnš. It should be noted that stress is retracted from a short ending onto the preceding long syllable, both circumflex and acute (analogous ²¹ The formation of retracted stress has been associated with “two-peaked” stress. The description of the linguistic features of the Spitrėnai subdialect states that “the process of stress retraction has already affected this dialect as well. In some words two stresses can be clearly heard” (Urbanavičiūtė 1966: 5). ²² Viẽdras (2) ‘bucket’. ²³ Forms with retracted stress account for less than 10 percent. ²⁴ The loanword jermė́kas is used in the meaning ‘long overcoat, military greatcoat’. ²⁵ All forms retracting stress found in the manuscripts of both subdialects are indicated (accounting for 11 percent and 9 percent, respectively). ²⁶ In the description of the Labanõras dialect, Jonas Šukys (1958: 5) stated that “stress retraction has been observed only in disyllabic words when the root is long and the ending short”. 7 accentuation tendencies are also characteristic of the Eastern Aukštaitian Anykščiai subdialect, see Ragaišienė 2020b: 215–216, 222). The analysis of the manuscript-source material reveals an uneven and rather late formation of stress retraction as a rule in the Utena subdialect: in the mid-20th century, the phenomenon had already become established in the very north of the area and in the northwestern corner; in the northern Panemunėlis (177) – Aleksandravėlė (180) – Kriaunos (179) – Sūvieko (181) areal and in the western peripheries adjoining the Anykščiai and Širvintiškiai subdialects, it was still developing. The fact that the phenomenon began to form relatively late in the Utena area is confirmed by the texts of Antanas Baranauskas from the second half of the 19th century (Specht 1920: 52–83)²⁷. In the 1920 publication Litauische Mundarten gesammelt von A. Baranowski, in the narratives from the Vabalninkas (142) subdialect, stress was regularly retracted from a short ending onto any long syllable; in the texts from the Žiobiškis (147)²⁸ and Vyžuonos (275) parishes it was not retracted, and only isolated instances of “two-peaked” stress were found, which are not statistically significant (Ragaišienė 2015: 180–181). In A. Baranauskas’s time, the intensive conditional stress retraction established in the northern part of the Kupiškis area, the northwestern corner of the Utena area, and the western periphery of the Anykščiai area was moving southward (see Fig. 2; for more detail, see Ragaišienė 2015: 180–183; 2018: 181–184; 2020b: 215). The manuscript-source material from the 6th–7th decades of the 20th century shows that stress was already regularly retracted in the Kupiškis subdialect, though with varying intensity, and throughout the Anykščiai subdialect area (Ragaišienė 2018: 185; 2020a: 90–93; 2020b: 215–216)²⁹. Thus, the retracted stress established in these subdialects in the mid-20th century was forming in the western periphery of the Utena subdialect and moving into the eastern and southeastern subdialects. ²⁷ While teaching at the Kaũnas seminary for priests, Antanas Baranauskas organized his students to collect dialect texts around 1867–1884, which he later sent to the German linguist Hugo Weber. ²⁸ Žiobiškis was assigned by Antanas Salys to point 147 (Geržotaitė et al. 2014: 412; cf. also Geržotaitė 2015: 147). Žaneta Markevičienė (2011: 306) has likewise indicated this locality as a point of the Utena subdialect. ²⁹ The Selian substrate may have influenced the formation of the stress-retraction rule in these subdialects (Zinkevičius 2006: 30, 73). 8 Figure 2. The nature of stress retraction in the Eastern Eastern Aukštaitian texts presented by A. Baranauskas Based on the material from the texts of A. Baranauskas and the manuscript sources of the Eastern Aukštaičiai subdialects, it may be assumed that in the mid-20th century a more intensive, rather than weaker, conditional stress retraction was forming in the western periphery of the Utena subdialect. Lithuanian Language Atlas compilers may have found it difficult to determine the type of stress retraction for several reasons. First, because during the period under study the phenomenon was of varying intensity in neighboring subdialects (see Fig. 1). In some manuscripts, forms with retracted stress account for approximately 21–38 percent, whereas in others only isolated cases of stress retraction and/or forms with “double-peaked” stress were found. It cannot be ruled out that the researchers may have associated cases of morphophonological leveling of the accentual paradigm with stress retraction (Ragaišienė 2024: 185). In the manuscripts of the western-peripheral, eastern, and southeastern subdialects that retract stress infrequently, more than half of the cases under discussion consist of the singular instrumental and plural accusative forms of circumflex-stem words (68 cases out of 115)30. It should be noted that this accentuation is more common in second-accentuation-class words whose penultimate syllable has a long vowel and ie, uo, e.g.: 30 In the stem, singular nominative forms also tend to be stressed, e.g.: ma.kkla ~ mokykla, sta.vkla ~ stovykla ‘statue’ Int et al. (Ragaišienė 2024: 185). 9 gaidžōkus ~ gaidžiokus, su pōnu, sagōnus³¹, viščiōkus ~ viščiokus Šd; ma.kìklas ~ mokyklas, su plūgu, su pōnu, rūbus, rūtas ~ rūtas, samagōnù³² Int; su diẽdu, baravì.kus ~ baravykus, ragutẽlãs ~ raguteles, rūbus, sijōnus, viščiōkus ~ viščiokus Mdk; burōkus, niẽkus (cf. niẽkai), miẽstus (cf. miestẽ), miežius, su vỹnu, žỹdus (cf. žỹdų) Gdr³³. Analogous accentuation tendencies are also characteristic of the Eastern Aukštaičiai Vilnius and Southern Aukštaičiai subdialects, which do not belong to the area of stress retraction (Grumadienė 1994: 101–104; Mikulėnienė 1996: 149; Urbanavičienė 2006: 220; Ragaišienė 2010: 53; 2016: 162–163 et al.). In summary, it can be stated that the material from manuscript sources of the 1950s–1960s attests to two important points. First, it demonstrates the formation of stress retraction as a phenomenon in the western peripheries of the Utena area and in the eastern and southeastern subdialects. Second, it raises doubts about the operation of weak conditional stress retraction as a separate rule in this area during the period under study. SUMMARY The material from manuscript sources of the Utena subdialect of the Eastern Aukštaičiai from the 1950s–1960s shows that stress retraction as a rule began to form in the subdialect relatively late and operated unevenly. In the mid-20th century, the phenomenon had already become established in the very north and in the northwestern corner of the area; in the northern Panemunėlis (177) – Aleksandravėlė (180) – Kriaunos (179) – Sūviekas (181) area and in the western peripheries bordering the Anykštėnai and Širvintiškiai subdialects, it was only beginning to form, while in the eastern and southeastern subdialects it was rare. The analysis conducted substantiates the current theory of the origin of Lithuanian dialects and supports the conclusions of studies of A. Baranauskas’s texts from the second half of the 19th century concerning the Eastern Aukštaičiai Kupiškėnai and Anykštėnai, as well as of descriptions of the dialectal features of subdialects from the 1950s–1960s, namely, that stress retraction proceeded from north to south. It is likely that the innovative wave of stress retraction that moved from the northern part of Kupiškėnai toward the eastern periphery in the second half of the 19th century could have reached the western Utena subdialects no earlier than the mid-20th century. Lithuanian Language Atlas compilers’ area of weak conditional stress retraction, the northern Panemunėlis (177) – Aleksandravėlė (180) – Kriaunos (179) – Sūviekas (181) area and the western peripheries, appears to have had only intensive conditional stress retraction in the 1950s–1960s. The manuscript-source material does not show a continuous area in which weak conditional stress retraction operated. Cases of morphophonological leveling of the accentual paradigm were probably regarded as manifestations of this rule (in the penultimate syllable of second-accentuation-class words, whose nucleus is a long vowel or ie, uo, the singular instrumental and plural accusative forms are stressed). Assumptions concerning the operation (and existence) of the rule of weak conditional stress retraction in the Utena area should be based on material from the stress-retaining subdialects of this subdialect from the second half of the 20th century to the beginning of the 21st century. ³¹ Sagōnas (2) ‘a large cast-iron pot with or without handles’. ³² Samagōnas (2) ‘homemade vodka’. ³³ Few forms of these cases retaining stress on the ending were found, e.g.: gri.būs ~ grybus, kisielùs ~ kisielius, miežius ~ miežius, po.nū ~ ponu Gdr. 10 REFERENCES AND SOURCES Gargasa itė Dalia, Misiūnaitė Aldona 1958: Description of the Daršiškiai Village Dialect: manuscript source [Description of the Daršiškiai Village Dialect: Manuscript Source], Vilnius: Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language. Geržotaitė Laura, Kaikarytė Aušra, Šinkūnas Mindaugas 2014: Comparative Analysis of the Lithuanian Language Atlas and the Antanas Salys Survey 1 Residential Areas (Compilation of Points) [Comparative Analysis of the Lithuanian Language Atlas and the Antanas Salys Survey 1 Residential Areas (Compilation of Points)]. – Lithuanian Dialects at the Beginning of the 21st Century: Geolinguistic and Sociolinguistic Research (Maps and Their Commentaries), ed. D. Mikulėnienė, V. Meiliūnaitė, Vilnius: Briedis. Geržotaitė Laura 2015: Geolinguistic Analysis of Antanas Baranauskas’ Classification of Lithuanian Dialects [Geolinguistic Analysis of Antanas Baranauskas’ Classification of Lithuanian Dialects]. – Problems of the History of the Language and Dialectology 4, ed. V. Ragaišienė, V. Meiliūnaitė, Vilnius: Institute of the Lithuanian Language, 133–165. Girdenis Aleksas 1978: Accentological Trivial [Accentological Trivial]. – Baltistica 14(1), 75–76. DOI: http://dx.doi.org/10.15388/baltistica.14.1.1254. Girdenis Aleksas 2003: Theoretical Foundations of Lithuanian Phonology [Theoretical Foundations of Lithuanian Phonology], Vilnius: Institute of Science and Encyclopaedia Publishing. Grinaveckienė Elena, Gargasa itė Dalia 1956: Description of the Robliai Village Dialect: manuscript source [Description of the Robliai Village Dialect: Manuscript Source], Vilnius: Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language. Grumadienė Laima 1994: Some Newer Features of Inflection and Stress in the Punsk Dialect [Some Newer Features of Inflection and Stress in Punsk Dialect]. – Questions of Lithuanian Linguistics 34, 97–106. Leskauskaitė Asta 2004: Features of Vocalism and Prosody of the Southern Aukštaičiai [Features of Vocalism and Prosody of the Southern Aukštaičiai], Vilnius: Institute of the Lithuanian Language. LKA I: Lithuanian Language Atlas 1: Lexis [Lithuanian Language Atlas 1: Lexis], ed. K. Morkūnas, Vilnius: Mokslas, 1977. LKA II: Lithuanian Language Atlas 2: Phonetics [Lithuanian Language Atlas 2: Phonetics], ed. K. Morkūnas, Vilnius: Mokslas, 1982. LKA III: Lithuanian Language Atlas 3: Morphology [Lithuanian Language Atlas 3: Morphology], ed. K. Morkūnas, Vilnius: Mokslas, 1991. LKT 1970: Lithuanian Dialects: reader [Lithuanian Dialects: Reader], ed. E. Grinaveckienė, A. Jonaitytė, K. Morkūnas, B. Vanagienė, A. Vidugiris, Vilnius: Mintis. 11 LKTCH 2004: Lithuanian Dialects Reader [Lithuanian Dialects Reader], ed. R. Bacevičiūtė, A. Ivanauskienė, A. Leskauskaitė, E. Trumpa, Vilnius: Institute of the Lithuanian Language. Markevičienė Žaneta 1999: Texts in Aukštaitė Dialects [Texts in Aukštaičiai Dialects] 1, Vilnius: Vilnius University Press. Markevičienė Žaneta 2001: Texts in Aukštaitė Dialects [Texts in Aukštaičiai Dialects] 2, Vilnius: Vilnius University Press. Markevičienė Žaneta 2011: Differentiation of the Anykštėnai Subdialect [Differentiation of Anykštėnai Subdialect]. – Lituanistica 57(1), 305–311. Mikulėnienė Danguolė 1996: On the Stress Variations of Inflected Words in the Southern Aukštaičiai Dialects [On the Stress Variations of Inflected Words in the Southern Aukštaičiai Dialects]. – Questions of Lithuanian Linguistics 36, 148–155. Mikulėnienė Danguolė, Morkūnas Kazys 1997: Texts of the Dieveniškės Dialect [Texts of the Dieveniškės Dialect], Vilnius: Institute of Science and Encyclopaedia Publishing. Ragaišienė Vilija 2010: Accentuation Variants of Inflected Words in the Southern Aukštaičiai and Eastern Aukštaičiai Vilnius Dialects: doctoral dissertation [Accentuation Variants of Inflected Words in the Southern Aukštaičiai and Eastern Aukštaičiai Vilnius Dialects: Doctoral Dissertation], Vilnius: Institute of the Lithuanian Language. Ragaišienė Vilija 2015: Features of Aukštaičiai Accent in the Texts of Antanas Baranauskas [Features of Aukštaičiai Accent in the Texts of Antanas Baranauskas]. – Problems of the History of the Language and Dialectology 4, ed. V. Ragaišienė, V. Meiliūnaitė, Vilnius: Institute of the Lithuanian Language, 166–210. Ragaišienė Vilija 2016: The Accentuation System and Its Changes [The Accentuation System and its Changes]. – Change in Lithuanian Dialects at the Beginning of the 21st Century: Lithuanian Subdialects in Poland, D. Mikulėnienė, A. Leskauskaitė, V. Ragaišienė, N. Birgelienė, Vilnius: Institute of the Lithuanian Language, 120–204. Ragaišienė Vilija 2018: Changes in the Accentuation System of the Eastern Aukštaičiai Kupiškėnai from the 7th–9th Decades of the 19thth Century to the 6th–7th Decades of the 19thth Century. – Eastern Aukštaičiai Subdialect: Change and Changes, ed. J. Urbavičienė, R. Petrokienė, Vilnius: Institute of the Lithuanian Language, 176–203. Ragaišienė Vilija 2020a: New Systematic Approach to Stress Retraction in the East–Eastern Aukštaitian Subdialects. – Lithuanian Dialectology Profiles: Problems and Findings, Editorial Board: D. Mikulėnienė (Editor-in-Chief), A. Stafecka, M. Jankowiak, E. Trumpa, I. Lemeškin, Special issue editor V. Meiliūnaitė, 85–94. DOI: https://doi.org/10.35321/e–pub.8.problems–and–findings. 12 Ragaišienė Vilija 2020b: Stress Retraction in the Texts of the Eastern Aukštaičiai Anykštėnai Subdialect [Accent Withdrawal in the Anykštėnai Dialect Texts of the Eastern Aukštaičiai]. – Acta Linguistica Lithuanica 83, 211–226. DOI: https://doi.org/10.35321/all83–10. Ragaišienė Vilija 2024: Accentuation Tendencies of Declinable Words in the Manuscript Sources of the Eastern Aukštaičiai Utena Subdialects from the 1950s–1960s [Accentuation Tendencies of Declinable Words in the Manuscript Sources of the Eastern Aukštaičiai Subdialects of Utena from the 1950s–1960s]. – General Language 97, 179–205. DOI: https://doi.org/10.35321/bkalba.2024.97.09. RAU 2010: Eastern Aukštaičiai from Utena [Eastern Aukštaičiai from Utena], ed. R. Rinkauskienė, R. Bacevičiūtė, V. Salienė, Vilnius: Institute of the Lithuanian Language. Specht Franz 1920: Litauische Mundarten gesammelt von A. Baranowski 1: Texts [Lithuanian Dialects Collected by A. Baranowski 1: Texts], Leipzig: Verlag von K. F. Koehler. Stundžia Bonifacas 1989: Antanas Baranauskas the Accentologist [Accentologist Antanas Baranauskas]. – Baltistica 3(2) supplement, 434–442. DOI: http://dx.doi.org/10.15388/baltistica.0.3.2.870. Stundžia Bonifacas 1995: On Stress Retraction in the Eastern Aukštaičiai Subdialects of Lithuanian [On the Question of the Retraction of Stress in East Aukštaitian subdialects of Lithuanian]. – Res Balticae, 67–72. Šūkys Jonas 1958: Description of the Labanoras Dialect: manuscript source [Description of the Labanoras Dialect: Manuscript Source], Vilnius: Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language. Urbanavičienė Jolita 2006: Accentuation Peculiarities of Peripheral Subdialects of the Eastern Aukštaičiai Vilnius Dialect [Accentuation Peculiarities of Peripheral Subdialects of East Highlanders of the Vilnius Dialect]. – Baltistica 41(2), 219–231. DOI: http://dx.doi.org/10.15388/baltistica.41.2.974. Urbanavičiūtė Žaneta 1966: Description of the Spitrėnai Subdialect: manuscript source [Description of the Spitrėnai Dialect: a Manuscript Source], Vilnius: Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language. Urbanavičiūtė–Markevičienė Žaneta, Grinaveckis Vladas 1992: Lithuanian Dialects: Phonetics, Morphology [Lithuanian Dialects: Phonetics, Morphology] 2, Vilnius: Vilnius Pedagogical University. Zinkevičius Zigmas 1966: Dialectology of the Lithuanian Language [Dialectology of the Lithuanian Language], Vilnius: Mintis. Zinkevičius Zigmas 1994: Dialectology of the Lithuanian Language [Dialectology of the Lithuanian Language], Vilnius: Institute of Science and Encyclopaedia Publishing. Zinkevičius Zigmas 2006: The Origin of Lithuanian Dialects [The Origin of Lithuanian Dialects], Vilnius: Institute of the Lithuanian Language. 13 Abbreviations of the Names of Residential Areas of the Eastern Aukštaičiai from Utena (Lithuanian Language Atlas points) Alk (180) – Aleksandravėlė, Rokiškis District. Arn (526) – Arniōnys, Molėtai District. Bč (121) – Bučiūnai, Rokiškis District. Bjt (479) – Bijūtiškis, Molėtai District. Brč (245) – Berčiūnai, Anykščiai District. Čkn (306) – Čekó̇nys, Molėtai District. Drš (172) – Daršiškiai, Kupiškis District. Gdr (504) – Giedraičiai, Molėtai District. Grs (412) – Girsteitiškis, Molėtai District. Gtč (114) – Gataučiai, Biržai District. Int (480) – Įnturkė, Molėtai District. Jnš (506) – Jõniškis, Molėtai District. Kkt (380) – Kūktiškės, Utena District. Krn (179) – Kriaūnos, Rokiškis District. Lb (414) – Labanõras, Švenčionys District. Lln (344) – Leliūnai, Utena District. Mdk (481) – Medeikiai, Molėtai District. Ob (148) – Obeliai, Rokiškis District. Pnm (177) – Panemunėlis, Rokiškis District. Rg (178) – Ragėliai, Rokiškis District. Rk (147) – Rõkiškis Sgn (379) – Sugiñčiai, Molėtai District. Skd (378) – Skudutiškis, Molėtai District. Spt (309) – Spitrėnai, Utena District. Svb (150) – Svõbiškis, Rokiškis District. Šd (478) – Šiaudžiai, Molėtai District. Užb (87) – Užubal̇iai, Rokiškis District. Vb (142) – Vabalniñkas, Biržai District. Vdn (450) – Vidėniškiai, Molėtai District. Stress Retraction in the Manuscript Sources of the Eastern Aukštaitian Subdialect of Utena from the 1950s and 1960s SUMMARY The article examines manuscript sources of the Eastern Aukštaitian subdialect of Utena from the 1950s and 1960s, stored in the Dialect Archive of the Centre of Geolinguistics at the Institute of the Lithuanian Language, with a focus on accentuation. This topic has so far 14 received little attention from linguists. The article aims to identify the stress retraction phenomena that were in effect in the Eastern Aukštaitian subdialect of Utena in the mid-20th century and to reveal their areal distribution within the subdialect area. The research material provides new data on the development and nature of stress retraction in the Eastern Aukštaitian subdialect of Utena. The study shows that stress retraction, as a rule, began to form relatively late in the subdialect and functioned rather unevenly. By the mid-20th century, the phenomenon had already become established in the northernmost and northwestern corner of the area. Meanwhile, it was still in the process of formation in the northern area (around Panemunėlis, Aleksandravėlė, Kriaunos, and Suviekas) and along the western periphery bordering the Anykščiai and Širvintos subdialects, and remained rare in the eastern and southeastern parts. The analysis supports the current theory of the origins of Lithuanian dialects and reinforces the findings from studies of Antanas Baranauskas’ texts from the second half of the 19th century covering the Eastern Aukštaitian subdialects of Kupiškis and Anykščiai and the descriptions of dialectal features of local dialects from the 1950s and 1960s, indicating that stress retraction spread from the north towards the south. Likely, the innovative wave of stress retraction, which moved from the northern part of the Kupiškis and Anykščiai subdialects towards the eastern edge in the second half of the 19th century, may have reached the western part of the Utena subdialect only by the mid-20th century. The northern area and the western periphery, which were previously assigned to the area of the weak conditional stress retraction by the compilers of the Atlas of the Lithuanian Language, seem to have only experienced a more intensive form of conditional stress retraction in the 1950s and 1960s. The manuscript sources do not indicate a uniform area where the weak conditional stress retraction would have been in effect. The manifestations of this rule are likely to be associated with cases that involve the morphonological levelling of the accentual paradigm. In the descriptions of local dialects where stress retraction occurs irregularly, more than half of the instances of stress on the stem occur in the instrumental singular and accusative plural forms of the words from the second accentual paradigm (whose penultimate syllable contains a long vowel or the diphthongs ie or uo). The insights presented in the article regarding the functioning (and existence) of the weak conditional stress retraction rule in the Utena subdialect are supported by the data from the local dialects of this subdialect, collected from the second half of the 20th century to the early 21st century, which do not exhibit stress retraction. 15