Balsių kaitos aprašai nuo Augusto Schleicherio iki Kazimiero Jauniaus

ACTA LINGUISTICA LITHUANICA LIETUVIŲ KALBOTYROS KLAUSIMAI XLIII (2000) Gražina AKELAITIENĖ Vilnius Pedagogical University DESCRIPTIONS OF VOWEL GRADATION FROM AUGUST SCHLEICHER TO KAZIMIERAS JAUNIUS §1. Changes in the expression of roots have attracted the attention of many linguists. Hints of this can already be found in the first printed grammar of the Lithuanian language. In the chapter “On the various positions of letters, or accidents,” Danielius Kleinas mentions three kinds of “accidents of letters”: change, contraction, and dropping. The chapter discusses dialectal changes in greater detail and states that other changes, i.e. those concerning inflection, will become clear from declensions and conjugations (see Kleinas 1957: 427). In the chapters on conjugations, D. Kleinas focuses primarily on changes in the marked letters1, but also mentions vowels in the penultimate syllable. For example, when discussing the past tense of verbs, he notes that some verbs in –auju form the past tense “quite irregularly,” namely: palowiau – from the present tense palauju, and the future tense palausu2. In discussing exceptions to the formation of other tenses, Kleinas separately mentions verbs of the sirgau type, which change the penultimate-syllable vowel e “to i in the future tense for the sake of better euphony, and it remains so in the past tense formed from it” (Kleinas 1957: 427). The chapter “On the noun and its accidents” also discusses “changes in letters.” Besides observations on vowel changes in the case endings of simple forms, a similar phenomenon is mentioned in pronominal (in D. Kleinas’s terminology, emphatic) adjectives and participles. D. Kleinas attributes changes in the expression of the pronominal affix, like other “changes in letters,” to “more pleasant sound”—in other words, to euphony. §2. August Schleicher occupies a special place in the history of the study of Lithuanian vowel gradation. He was the first scholar not only to survey and describe the full diversity of Lithuanian roots, but also to provide a scientific account of the place of changes in vocalism within the general system of sound development in Proto-Indo-European. In 1856, the first, phonetics chapter of Lithuanian Grammar (Litauische Grammatik) gives a detailed description of vowel laws. It is significant that this scholar, as 1 D. Kleinas calls the letters before the final u marked letters. 2 D. Kleinas understands the relationship among verb tense forms differently from the way it is generally understood today. According to him, all tenses are derived “directly” or “indirectly” from the indicative present tense (see Kleinas 1957: 479). 3 pats pabrėžia (žr. Schleicher 1856: 34), aiškiai skiria tuos dėsnius, pagal kuriuos balsiai keičiasi žodžių daryboje ir kaityboje, nuo garsų sandūros dėsnių. Pirmieji kitimai sudarą sistemą, kuri lyginamojoje indoeuropiečių kalbotyroje vadinama balsių kaitos laipsniais. Anot A. Schleicherio, lietuvių kalbos balsiai kitę silpnėjimo (vocalschwächung) ir stiprėjimo (vocalsteigerung) būdu. Pastarojo skiriamos dvi pakopos, atitinkančios iš indų kalbų gramatikų žinomus guṇo ir vṛddhio reiškinius. Kaip ir lyginamosiose germanų kalbų gramatikose, jis skiria tris balsių klases, arba eiles, kurių pagrindiniai balsiai esą a, i, u. Šie balsiai ir jų atbalsiai, anot jo, sudarą balsių kaitos sistemą (žr. Schleicher 1856: 62). Pavyzdžiui, a-eilės vokalizmo žodžiai susidarę, iš vienos pusės, pakitus pagrindiniam balsiui a į o („steigerung“), iš kitos pusės, susilpnėjus („schwachung“) pagrindiniam balsiui, atsiradę variantai su e (arba pailgintu ė), i (pailgintu ī): virti, svyruoti, sverti, svėriau, svarstyti. Atskirai minima, kad kartais šioje eilėje gali atsirasti u (tampyti, tempti – įtumpas). Iš viso a-eilės žodžių pateikta daugiausia (per 80), iš kurių dar atskirti bredu: braid(y)ti tipo žodžiai, rodantys sumišimą su i-eile. i-eilės balsių kaitos nariai išdėstyti šia tvarka: i(y), e (=ie), ai (dygti, diegas, daiginti); u-eilės kaitą sudarą šie nariai: u(ū), u (=uo), au, kartais av, ov (krūva³, krauti, kroviau). A. Schleicheris relies chiefly on the writings of Ferdinand Nesselmann and Konstantinas Sirvydas. In a few places, dialectal words or words from other Indo-European languages are given. The position of the main vowel is sometimes indicated not by the whole word, but only by the root. This allows us to infer that in the past a word with such vocalism must have existed; for example, in the ėd-mi word’s series, only the root ad- is indicated in the a vocalism position; for the word slėgti, the root slag-; for the word rėkti, the root rak-; while stogas is represented by the root stag-. A. Schleicher’s rather comprehensive list of cognate words is particularly valuable to researchers today in several respects. In grouping cognates, he made only an occasional mistake; for example, with the word mirti he needlessly compares the Slavic loanword morai “bier,” and with dievas the Slavic loanword dyvas; gardus with gerti (see Sabaliauskas 1979: 68). A. Schleicher concluded that the great diversity of vowels in Lithuanian, as in other languages, was not original: more archaic features had supposedly been preserved in the u-series, which in Lithuanian was also only three-membered; the novelty of the i-series was indicated by its having distinct ie and ei, while the a-series had diverged most from the original form (see Schleicher 1856: 62). As noted, A. Schleicher clearly emphasized that sound laws are of two kinds. According to him, under the laws discussed above, sounds change in word formation and inflection. Therefore, he did not overlook variation in root vocalism, and 3 In A. Schleicher’s examples, the words krūva, šūvis, lūžti and other roots with a long vowel are written with u. He wrote ū only in present-tense forms that had the infix n (grūvu, pražūvu, kliūvu). 4 in another, word-formation, chapter. Here vowel alternations are no longer examined in detail4, but the roots with a different vocalism presented alongside the derivatives show that the scholar had observed vowel alternation participating in word formation together with suffixes. It is simply not entirely clear on what principle A. Schleicher relied in indicating the vowel of the selected roots in his list of derivatives, e.g.: žiema – Vžim⁵, Laima – Vlam, atlaida – Vlid, atviras – Vvar, senas – Vsan, kreivas – Vkriv, pilis – Vpal, kova – Vku. It seems that what is most often indicated is the basic member of the relevant vowel alternation, which for A. Schleicher, as for the Indian grammarians, was the weak grade. When presenting examples of a certain type of derivative, he mentions vowel alternations sporadically, sometimes separately, as an already self-evident matter of phonetics. For example, it is noted that in forming nouns with the suffix -a, the root vowel often strengthens rather than weakens: sargas, dagas, kvapas, takas, tvanas, maras, svaras, and so on (Schleicher 1856: 102). The detailed classification of primary Lithuanian verbs presented in A. Schleicher’s Grammar6 is also based on differences in the expression of the root (or, more generally, the stem). The basis for class division is the characteristics of the present tense, for example, a lengthened or altered root vowel (Class II), a nasal consonant in the present tense (Class I), and so on. The verb is described with attention to the system of vocalic change: the grade represented by the verbs of the respective classes is specified (see Schleicher 1856: 234-254). In essence, A. Schleicher’s classification of verbs is also followed in grammars today. §3. The typological classification of Lithuanian word-root structure devised by A. Schleicher undoubtedly occupied an important place in Indo-European linguistics. In the history of Baltic-language grammars, this question was examined scientifically for the first time, and thus later works by Balticists already drew on Schleicher’s theory. The account of vowel alternation in Friedrich Kurschat’s grammar, published in 1876 (Grammatik der litauischen Sprache), differs very little from A. Schleicher’s theory. F. Kurschat, however, discusses vowel alternations not in the phonetics section (where Sanskrit vowel-alternation grades are merely mentioned), but in the introductory chapter on word formation. This seems to express the view that vowel alternation should be regarded as part of word formation. That F. Kurschat drew on A. Schleicher’s examination of vowel alternations is shown by the same examples and, at times, by direct references to the work cited above (cf. Kurschat 1876: 80–82). §4. In 1884, August Leskien’s study Der Ablaut der Wurzelsilben im Litauischen was published (see Leskien 1884). As he himself states in the introduction, his account of vowel-alternation series was not intended to support or refute the Indo-Germanic theories current at the time— 4 A. Schleicher himself emphasized that the material in the word-formation chapter was not presented comprehensively, but rather was “whatever happened to come to hand” (Schleicher 1856: 102). 5 A. Schleicher used this mathematical symbol to denote the root in his account of derivatives. 6 Lietuvių kalbos veiksmažodžių klasifikacija jau buvo pateikta ankstesniame, 1853m. išspausdintame, A. Schleicherio darbe Lituanica. Tačiau Gramatikoje ji išsamesnė. 5 of vowel-system theories. According to A. Leskien, it arose as an auxiliary means for examining the vowel system of the Slavic languages. A. Leskien notes that the examples were collected from F. Nesselmann’s dictionary, but were later supplemented with material he himself collected not only from Lithuanian but also from Latvian. Unlike earlier researchers, A. Leskien distinguishes five ablaut series: I. i y(ī) ē ei(ej) ej ai(aj) II. u ū ū au ov III. a) i y (=ī) e ē ą o (=ā) b) e ē ą o (=ā) IV. ė a o (=ā) V. a o (=ā) (see Leskien 1884: 268–269). The traditional distinction of three ablaut series was not arbitrary. It was an abstracted schema of the law of vowel alternation in related words. The system presented by A. Leskien is not more perfect—it is simply a more substantiated realization. In essence, series III, IV, and V correspond to the a-series in A. Schleicher’s system. They are separated on the basis that the possibilities of alternation are realized differently in roots of the corresponding type. For example, the fourth and fifth series are represented by words with essentially the same vowel system, except that the alternations in the fifth series are illustrated by roots that have no variant with ė (ariu: orė, blaškai: bloškiau, privalau: prievolė). The third series, according to A. Leskien, is heterogeneous: he distinguishes series III a (bildu: beldžiu: baldas) and III b (degu: nuodegulis: dagas: ato-dogei), although the introduction itself states that these series are closely related and may be combined into one (see Leskien 1884: 268). This subdivision of the series may simply be more informative: it allows one to judge the productivity of ablaut types more accurately. In general, A. Leskien’s clear desire to find as many alternants as possible for each root is noticeable, that is, to show all the vowel grades of every root. Very often, when no Lithuanian word was found, the alternation is illustrated with Latvian words (sometimes with Latvian words alone); for example, the ie: ai, i: ī vowel pattern is illustrated by Latvian kniebju: knaipīt, miju: miju (see Leskien 1884: 138). Such a selection of examples sometimes somewhat distorts the actual picture of how ablaut functions in Lithuanian. The semantic relationship between some words regarded as having the same root might scarcely have been discernible synchronically (e.g., diržti: diržas: daržas (p. 324); galas: gelti (p. 325); tiltas: patalas: toli (p. 349)); therefore, A. Leskien’s study would more accurately be considered diachronic7. The second part of the work shows the distribution of ablaut across various subsystems of the Lithuanian language. It, too, consists essentially of indexes of finely classified examples with brief comments. 7 A. Leskien published a considerable number of new and correct etymological comparisons (see Sabaliauskas 1979: 74), but because semantics was disregarded (cf. also Latvian didels – Latvian didit (“song”), his understanding of root identification cannot be considered synchronic. 6 A. Leskien’s Der Ablaut der Wurzelsilben im Litauischen remained for many years the only work devoted in its entirety to a specific question in Lithuanian historical grammar—vowel alternations—and was highly significant for all researchers, perhaps not so much for its theoretical foundations as for its abundant illustrative material. §5. The grammars of A. Schleicher and F. Kurschat undoubtedly laid the theoretical foundations for the works written for teaching purposes by Antanas Baranauskas and Kazimieras Jaunius. A. Baranauskas’s lectures, given at the seminary for priests, were published without his knowledge under the title Kalbomokslio lietuviškos kalbos. In the second part of Kalbomokslio, entitled “Žodžugis,” there is a section on vowel “reversals.” A. Baranauskas distinguishes three “basic” vowels, a, i, u, from which, in his words, “all the others have flowed forth.” A. Baranauskas explains: “the basic a, as it weakens, becomes e, and as it weakens still further, becomes i, while as it strengthens⁸ it becomes i a, o” (the other “reversals” are explained accordingly) (see Baranauskas 1896: 9). Summarizing the regularity of the alternations, he also presents three series of vowels. “From the three basic vowels arise three series of vowels, diphthongs, and nasal vowels: 1st series a e, i, y, ė, ā, o — — — ia, io ę, į, — ą, ią 2nd series i i, y, ė, ei, ai i, — — — — — — iai 3rd series u u, ū, au, av, ov iu, iū, iau ų, ių” (see Baranauskas 1896: 10–11). A. Schleicher’s influence is strongly felt in A. Baranauskas’s explanation: he adopted not only the model of the alternation series but also the method of explanation. In providing examples illustrating the series, A. Baranauskas repeats some of A. Schleicher’s inaccurate root associations; for example, the Slavic borrowing morai is presented alongside words of the root mirti. Kitame A. Baranausko darbe Zamietki o litovskom jazyke i slovare (Baranovskij 1898) balsių kaita aiškinama daug plačiau. Jis kalba apie dvejopus, t. y. etimologinius ir fonetinius procesus, vykstančius žodžių sudedamosiose dalyse. Fonetiniais procesais vadina tarmių fonetikos skirtybes, o etimologiniais procesais – žodžių daryboje ir kaityboje atsirandančius skirtumus. Kaip A. Baranauskas pats nurodo, jis iš esmės pateikia A. Schleicherio teoriją. Terminai taip pat rodo sekimą A. Schleicheriu: vartojami vokiškų terminų rusiški 8 In quotations here, the spelling of A. Baranauskas’s Kalbomokslio (and, below, K. Jaunius’s grammars) is not followed; for example, Kalbomokslyje spells the word stėbjasi. 7 pakaitalai: oslabenie, količestvennoe uveličenie, podjom (žr. Baranovskij 1898: 10 ir kt.). A. Baranauskas also describes differences in the expression of endings on the basis of a three-grade vowel model. But when discussing the expression of endings, he does not always agree with A. Schleicher. Drawing on his own (Anykščiai) dialect, he distinguished three vowel lengths and criticized A. Schleicher for not doing so (see Baranovskij 1898: 18). §6. As is well known, K. Jaunius also attended A. Baranauskas’s Lithuanian language course, studying at the seminary from 1871 to 1875. Later, in the final decade of the century, almost at the same time as A. Baranauskas’s works, K. Jaunius’s own Lietuviškas kalbomokslis (1897) saw the light of day. This was a hectographed work which, essentially unchanged and prepared by K. Būga, was published in 1911 as Lietuvių kalbos gramatika. K. Jaunius undoubtedly drew not only on A. Baranauskas’s lectures, but also on published grammars. In the dialect section of his own work, he explicitly states that he preserves A. Schleicher’s and F. Kuršaitis’s “good things,” but does not adopt their “mistakes and follies” (see Jaunius 1911: 27). K. Jaunius identified the distinction between the “half-long” acute- and circumflex-accented vowels as one of A. Schleicher’s “follies.” By contrast, he spoke favorably of F. Kuršaitis in this respect: “Kuršaitis, a true Lithuanian, understood the Lithuanian accent and the manner of pronouncing syllables better, and therefore there are no maimed words in his writings.” But K. Jaunius criticized F. Kuršaitis for confusing uo with o, and ie with ė (see Jaunius 1911: 27, 28). In the section on syllable structure10 “Vowel gradation,” K. Jaunius describes the “alternation of vowels in word formation.” The principal members of the series are a, ie and au. The latter apparently seemed more informative to K. Jaunius than i and u, which were customarily used to mark the second and third series (i and u can also occur in roots with a vocalism). Here is one of K. Jaunius’s “laws” of vowel gradation: “If a word has the vowel a in its root, its other cognates may have e, ē, i, y, o, u, uo, ū, but will have neither ie nor au” (Jaunius 1911: 38). He argues that “at most, only two or three members of an entire series can be found most often; therefore, it is especially useful a to divide the series into four subseries” (Jaunius 1911: 39). (It is possible that A. Leskien’s account of vowel gradation influenced this division, although there is no reference to A. Leskien’s study11.) a series is divided by him into “subseries” a, ē, o, uo. Among the examples illustrating these “subseries,” especially uo, are words whose roots are hardly suitable to be considered roots with a vocalism, for example: duoti, pulti, puoliau, gulėti, guolis, as well as juosti, juosta. They are more likely to represent u, that is, the au series in Jaunius’s model, rather than the a series. 9 Apie tai savo darbe yra užsiminęs ir A. Baranauskas (žr. Baranovskij 1898: 9). 10 According to Antanas Balašaitis, a researcher of linguistic terminology, K. Jaunius was the first to give the phonetic discipline its proper name, “tarsnymokslis,” that is, the science of sounds, although at the beginning of his career K. Jaunius also called a sound by another term, gaisinis (see Balašaitis 1961: 184, 186). 11 Notes and bibliographic references in K. Jaunius’s other works show that he had studied A. Leskien’s works (see Drotvinas, Grinaveckis 1970: 99, 138, etc.). 8 One may also question some of the words cited as evidence for the subseries of the vowel system under discussion. For example, the o subseries is illustrated, among other words, by bažmas “a multitude, a crowd,” and buožė, bužulas “a bulb,” “a large knot,” which are connected with bokštas. This connection is doubtful (cf. Polish baszta). Not all the words K. Jaunius gives as examples of the vocalism of a particular series can even be connected etymologically. He regarded rimstu, rymai, remiu, rèmiau, ramus, ramentas, romus, romyti and rūmai as words with the same root (vocalism of the a series). Apparently, on the grounds that the Russian word pokoi (“peace, tranquility”) once also had the meaning “room,” he proposes a semantic comparison between rūmai12 (Latin palacium) and Russian pokoi, počit “to die, to fall asleep.” This is one of K. Jaunius’s unfounded etymological connections. In assessing such connections, historical linguists have observed that K. Jaunius, as a self-taught linguist, sometimes wandered down the byways of comparative linguistics; the basis of this wandering was his effort to derive semantically related word forms that differed in vocalism, and sometimes in consonantism as well, from a single proto-form, explaining them by certain sound-change laws of his own invention (see Zinkevičius, Kabelka 1958: 23). K. Jaunius also discusses the “transfer from one vowel series to another,” that is, the mixing of series. He compares this phenomenon to the mixing of word inflection (“declension”) and calls it “word association.” In his view, “according to the laws of human thought, two series become mixed if several members in one are the same as those in the other” (Jaunius 1911: 39). K. Jaunius argues that people can transfer word roots from one series to another through association, thereby creating a word with such a root whose form “does not fit at all” with the other words of that root (see Jaunius 1911: 40). K. Jaunius paid no particular attention to word formation, so it is difficult to say how he assessed the role of vowel gradation in derivation. However, quite a few references to alternations can still be found in the section “Verb Inflection” (i.e., conjugation). When comparing stem vowels, he labels their distinctions with terms customary in descriptions of vowel gradation: “higher-grade vowel,” “weaker vowel,” and “broader vowel.” §7. A comparison of K. Jaunius’s account of vowel gradation with the works of other researchers shows that K. Jaunius essentially follows A. Schleicher’s theory of vowel gradation, but formulates the laws in his own way. The more complex model of the system of vowel alternations presented here (dividing the series into subseries) may have resulted from his own reflections, although A. Leskien’s works may also have influenced it. On the other hand, one need not look for ways in which K. Jaunius’s account of vowel gradation differed from earlier ones, or for how it supplemented or corrected the work of his predecessors. Something quite different may be mentioned—something emphasized by Juozas Tumas-Vaižganas, who attended K. Jaunius’s lectures— 12 Etymologists argue that the word rūmas should be connected with Latin rūs, rūris “village, estate” (see Sabaliauskas 1990: 265). 9 …: when Jaunius had led them through some sort of “alternations, gradations, disappearances, or changes,” it became clear to them how individual people, having fallen in love with an idea, can move half the world with it (see Tumas 1916: 3). LITERATŪRA Balašaitis A. 1961: Lietuvių kalbos fonetikos terminai. – LTSR MA darbai, serija A. 1, 183–206. Baranauskas A. 1896: Kalbomokslis lietuviškos kalbos, išdūtas per LL, Tilžė. Baranovskij A. 1898: Барановский А. Заметки о литовском языке и словаре, Санктпетербург. Drotvinas V., Grinaveckis V. 1970: Kalbininkas Kazimieras Jaunius, Vilnius. Jaunius K. 1911: Lietuvių kalbos gramatika, Peterburgas. Kleinas D. 1957: Pirmoji lietuvių kalbos gramatika, ats red. J. Kruopas, Vilnius. Kurschat F. 1876: Grammatik der litauischen Sprache, Halle. Leskien A. 1884: Der Ablaut der Wurzelsilben im Litauischen, Leipzig. Sabaliauskas A. 1979: Lietuvių kalbos tyrinėjimo istorija, iki 1940 m., Vilnius. Sabaliauskas A. 1990: Lietuvių kalbos leksika, Vilnius. Schleicher A. 1856: Lituaische Grammatik, Prag. Tumas J. 1916: [K. Jauniaus lietuvių kalbos gramatikos rec.] – Lietuvių balsas 69(98), 3. Zinkevičius Z., Kabelka J. 1958: Kazimieras Būga. – Rinktiniai raštai 1. Vilnius, 13–100. THE DESCRIPTION OF VOWEL ALTERNATIONS FROM AUGUST SCHLEICHER TO KAZIMIERAS JAUNIUS Summary August Schleicher occupies an exceptional place in the history of the investigation of Lithuanian vowel alternations. The authors of subsequent grammars of Lithuanian who dealt with vowel alternations could be said to be his followers. Kazimieras Jaunius, for example, basically adhered to Schleicher’s theory though he formulated the rules of vowel alternation in an original way. Jaunius presented a complex model of the system of vowel alternations, with the series of alternating vowels further divided into subseries. This system could have been devised on the basis of some considerations of his own. On the other hand, it could also have been influenced by A. Leskien’s works, which were undoubtedly known to Jaunius. Gražina AKELAITIENĖ Vilniaus pedagoginis universitetas Ševčenkos g. 31 LT-2009 Vilnius, Lietuva El. paštas [email protected] Gauta 2000 02 07 10