Dvibalsių ir mišriųjų dvigarsių priegaidės Eržvilko šnektoje: kiekybė ir spektrai

LITHUANIAN LANGUAGE QUESTIONS XL (1998) Daiva ATKOČAITĖ Lithuanian Language Institute, Vilnius SUBJUNCTIONS OF DIVALENT AND MIXED DIVALENTS In the Snore of the Wolf: Quantity and Spectra 1.The distribution and acoustic characteristics of the Southern Samogitian Raseiniškis diphthongs have not yet been studied and described in detail. There is little evidence from the southern edge of the dialect, where the Eržvilkas dialect is studied. The authors of both monographs and review works usually rely on empirical observations, which are not always accurate and can only approximately describe the acoustic features of the cliff. There are no works in which the prepositions of individual speakers or of the whole dialect have been experimentally studied, and nowhere is it said how the prepositions of Raseiniškiai are close to the corresponding prepositions of other Žemaičiai, or how much more similar to the prepositions of the common language. In this article, we are trying to prove the distinction of two accents — the firm and the continuous — in diphthongs, to identify the most important distinctive features and to compare them with the corresponding studies of the closer Žemaičių dialektai and the common language." 2.Although it is common in the Raseiniškiai dialect to distinguish between a firm beginning and a continuous preposition, dialect researchers sometimes use other terms. For example, M. Slušinskaitė refers to the acute preposition in Lomių dialect as the sudden acute, which is similar to the firmly beginning preposition of common language, although there is no clear description of how this preposition differs from the preposition of common language (Slušinskaitė 1981: 20). Z. Zinkevičius also does not use the term "fixed-start waiting period". In his opinion, in the southeastern edge of the Žemaites (and therefore in Eržvilkas dialect) the Žemaites' loud-sounding accent is moving into the Vysoké step. In the case of emphasized pronunciation, the pusher turns into a real log (Zinkevičius 1966: 35). The continuous consonant of circumflex origin differs more from its equivalent in general language: it is commonly assumed that its pressure is divided between the two diphthong components. However, although this is typical of the entire area of Raseiniškiai, V. Tamošaitytė-Jokubauskienė uses the term tvirtagalės priegaidė when describing Pagramančio šnektą (Tamošaitytė-Jokubauskienė 1974: 27). It remains unclear whether this is the case here.'This work will mainly be compared with the nearest southern Samogitian Varniškis dialects. (Bagvilaitė 1975; Bukantis 1984), northern Samogitian (Girdenis 1974) and common language (Pakerys 1984). 1968; Pakerys et al. 1972; Pakerys 1974; Pakerys et al. 1974; Pakerys 1982) on the basis of double-sided premonitions. The methodology of the study of other dialects was also used (Girdenis, Pupkis 1974; Kosienė 1982; Girdenis 1996). 88 called the usual continuous waiting (this term was used by ir P. Jonikas (Ionic 1939: 12)), ar In fact, the usage of the word in this context iš coincides with the su common sense. Since there is no clear motivation to use any new terms, this article follows the traditional opinion that in the Raseiniškiai dialect there are two distinct prepositions: — ir — (plg. Grinaveckis 1973: 83; Zinkevičius 1994: 108), o other occasionally occurring incidents are just variants of these (See section 4.4.) § 3). 3. Already empirical observations have shown that in diphthongs the speech distinguishes well the opposition of acute and circumflex prepositions, e.g.: ld.uk “lauk' —lū.ū.k laūk', kd.lts ‘guilty’ — ka. Lt ‘guilty’. The preposition of acute origin in speech is only solidbeginning. The sound of the fire is rarely heard. It appears only sporadically, usually in a syllable with a secondary accent, e.g.: ta_tas_antrasis | valei ji “kit's iš juo || tas tė.i.ni: sa.u | rūsas || In expressive speaking, sometimes the fricative prefix can also be heard in a syllable with a main accent, e.g.: tai —gerai padd.ri | pliešiks da: būtu plieše(s) žmūonis || kuokde būwa Zudrni || This rather rare and irregular variant of the fricative prefix, usually depending on the intonation of the phrase, is almost not found in diphthongs, so the distribution of the fricative prefix or its acoustic characteristics will not be discussed here. The origin of the circumflex is more difficult to describe. Sometimes a continuous consonant is pronounced quite clearly in the trunk, with the pressure spread across both consonants, but sometimes it is difficult to determine whether one or both consonants are emphasized. As a positional variant in the absolute ending, the middle consonant is more often pronounced, the pressure of which tends to be concentrated in the first accent, and the second accent is greatly reduced, e.g.: pavd.saris | tetervindi šū.ū.k || pakrūoviem | ta višta i tupieie geerai || In this case, the majority of the vocal pressure goes to the first accent, and the second is pronounced much weaker." For this reason, when selecting the word pairs for the study, more attention was paid to the opposition of the consonants in the root, where both consonants are clearly pronounced, although all diphthongs were also examined in the absolute endings. The main tasks of the study are to determine: 1) how the duration of the whole diphthong and its individual elements depends on the prefix, 2) how the quality of the elements is related to the prefix, 3) whether the duration of the whole diphthong and its individual elements depends on the quality of the elements. * Further in the article, such a middle preposition is not distinguished and is called a continuation. 89 4. Since the quality-quantity relationship of doubles in ir prepositions was concerned, only su the pronounced a’, e+u, i, I, m, n, r. Not at all a questionable diphthong «i, because it is rarely used in conversation, it is difficult to find suitable ir word pairs for the experiment, and the consonants of this diphthong be are poorly distinguished. We are not talking about the mixed diphthongs u-+/, ir i, m, n, r, because the jų quality and quantity relationship of the first demens with the antecedents has already been discussed in a previous publication (see section 4.4). Atkočaitytė 1997), o the relationship ir of the length of the second dot of the whole diphthong with the consonant should not differ from the tendencies that are characteristic of other mixed diphthongs studied here. For the experiment, words were taken whose pronounced syllables have a firm beginning or a continuous prefix. The minimum word pairs were included by the speakers of this dialect: two speakers (a man, born in 1932, and a woman, born in 1939), living in the northern area of the dialect, in the village of Varlaukis, and the author of the article, who also speaks this dialect. In total, 148 words were analyzed, pronounced in affirmative intonation. The following groups of words were read: 1) diphthongs: gd.uti “to get”: ga.ū.sti ‘to get’ —buvd.u ‘I was’; k'd.ura “kiaurą': k'ū.ū.le “piaūlę'—jū.ū. ‘jal’:ajd.u ‘¢jal’; allow “Ičisti'": spread. Lsti ‘spread’ —kvi'té.i “kvietei': kvi'tė.i. ‘kviečiai’; ld.ida ‘laido (mint)': la.i.da ‘laido’ — sakd.i ‘vksm. Say' :saka.i ‘dktv. say'; 2) mixed diphthongs: ld.pks ‘window’: la.j.ks ‘lap’; kd.lts ‘guilty’: ka.lts ‘guilty’; vaikd.ms ‘for children’: ka.r.(p)s® ‘corner’; kd.rs ‘karas’: ka.F.sts ‘kafstas’; pé.rs ‘péras’: pé.i.st ‘pefsti’; ké.lti ‘kélti’: ke. Lti “keltį', lé.mpa ‘lémpa’: le.m.ts ‘lemtas’, lé.nks ‘1énkas’lė.rj.kts "benktas". The pairs of words read by the informant were recorded on a tape recorder and injected into the computer memory, and the author’s examples were read directly into the computer. The acoustic analysis was performed using P. Kasparaitis’ program “Kalbam44”. Oscillograms, cross-sectional and dynamic spectrograms of the investigated binary axes were made. Measured the total and individual duration of the double-sided in milliseconds (ms). In diphthongs, especially au, ei, the two parts merge, without a clearer break, so it was difficult to determine the boundaries and the exact duration of these diphthong parts. The data of mixed binary stars are more reliable because their components are easier to distinguish—the limit is shown by the change in the spectrum. The aim of the spectral analysis of diphthongs and mixed diphthongs is to determine the effect of whims on the quality of vowels. * In the root au is studied both after the hard and after the soft consonant, because in the position [C'—-] a vowel 'a of different quality appears and a diphthong 'au is formed. * There is no pair that differs in the ending d.u and a.i, but the pronunciation of the verb endings is not consistent in the speech, so the same informants can pronounce such ending diphthong twice — once pronounced as a strong-pronounced consonant, and the other time as a continuous one. * In the common language, the first accent of the -ia-echoing diphthong 'in the Ažemaičių dialect has long since been prefixed and turned into -e, no different from the old diphthong -ei of the first sando. 6 The labial p and b between m and s are almost always discarded. 90 frequency determined at three points: beginning, ir middle end. In segments u ir i made a po single cut, because these sounds are shorter už a, e, so it's difficult to find jų the beginning, ir the middle end. It is particularly difficult to # determine the formative frequencies. There are no to sonnets. Measurement data ir are statistically processed in the Student program. The following statistical parameters are calculated: (a) the arithmetic mean (x), (b) the standard deviation (s), (c) the coefficient of variation (v), (d) the confidence interval of 9596, (e) the Student criterion (t) and its critical value (t,)’. The compactness, bemolarity and tonality indices of the segments were calculated using A. Girdenis program,FORMANTES.PASTM. Although there are some quantitative and qualitative similarities, diphthongs and mixed diphthongs are discussed separately in this article. 5.1. Two-voiced. 5.1.1. Data on the total duration show that the total duration of diphthongs with the first elements a, e, pronounced in affirmative intonation is a constant sign of prepositionality. The 1.2:1 ratio is longer than the 1.2:1 ratio of the continuous vowels. This difference is relatively reliable, since each of the solid-started diphthongs taken individually is longer than the continuous one (see Table 1), their confidence intervals, except for ei and au (cf. ei —316+ 341 ms and 288 + 316 ms; au —297+ 348 ms and 228 + 322 ms)’, do not overlap, and the values of the f criterion (except for au) exceed the critical limit of significance. 9596 Table 1. Total duration of diphthongs (ms) * Diphthong | n | x | s | v(96) |confident (95%) | tq, di 12 366 8.6 23 355 + 376 82 > 5.0** to 12 324 7.5 23 315 = 333 of 12 328 9.9 3.0 316 + 341 3.9 >3.8* and 12 302 1.4 3.8 288 + 316 du 12 359 8.8 24 348 +370 52 >5.0** ad. 12 317 15.8 5.0 297 + 337 du 8 322 27.9 8.6 297+ 348 23> 22 id 8 275 44.7 16.2 228 + 322 ' More about the methodology of mathematical statistics zr. Girdenis 1974; Kosiené 1982. * For calculation formulas and methodology, see Piotrovskis 1960.? Since the average was calculated independently of the position, this could be due to the structure of the diphthongs. Since the two stems are in the same order, the second stems of these dichotomous stems are more assimilable and shorter than the root stems in the open ends. This means that the total number of points gained by each player will be reduced by the number of points gained by the other player, thus reducing the total number of points gained. * Table ** marked pairs differ 99.9%, and * —99% probability. Mean x and confidence interval numbers are rounded to one. 91 The total average ir length of the ir stem, end diphthongs was calculated, although in the open end the continuous strong-startir ed diphthongs are shorter than in the root, because in the speech there is a tendency to slightly reduce the accentuated syllable of the end, e.g.: tuo ir dvd.ra kūmetei| ten ši'na pjūovi|| kūr ten priič.i.si| miškai didž'a.usi|| The beginning of the second element merges with the end of the first element so much that at the end of the first element it is already clear what elevation the second su sound will have, so this ir jį element can not be pronounced completely, shortened. In the first case, the fractional reduction is less than the reduction of the second, because the first is less than the second. However, the length ratio between the continuous diphthongs of the strong ir beginners does not depend on the position: in the stem, the ir strong ir beginners are longer. In this respect, Eržvilkas dialect differs from the northern Žemaičių dialects, where circumflex diphthongs are significantly longer. (plg. Bukantis 1984a: 72; Girdenis 1974: 175), and more similar to the common language, in which the lengths of the compound diphthongs do not depend on the preposition: sometimes there are longer firm beginnings, sometimes — firm endings (Pakerys 1974: 176). 1968: 107; Pakerys ir et al. 1972: 10). 5.1.2. An even clearer ir and more important feature of the pretemporary is the duration of the individual segments (cf. Pakerys 1982: 158). The duration ratio shows that the first elements of the strong-started diphthongs are significantly longer than the second elements (in a ratio of 2.3: 1), while the duration of the two continuous elements differs less, although the first element is also longer than the second (in a ratio of 1.5: 1) (see tables 2 and 3). This is a rather unexpected result, since it is assumed that in continuous diphthongs the duration and voice pressure should be distributed equally between the two parts. The length of the two parts of both the firm and continuous diphthongs is slightly shorter in the ending, but the ratio between them remains similar. Table 2. 1996: 1st place in the 100m freestyle with a time of 9.96s. int. (95%) | tz tį d.(i) 12 255 9.6 3.8 243+ 267 13.5>5.0** ū.li.) 12 188 5,3 2,8 182 +195 €.(i) 12 237 17.4 7.3 216 + 259 5.1 >5.0** é.(i.) 12 197 2.7 1.4 194 + 200 'a.(u) 12 252 10,3 4,1 239+ 265 12,9 >5,0** 'ū.(ū.) 12 190 3,3 1,7 186 + 194 d.(u) 8 211 30,5 14.4 183 + 239 4,4 >3,6* a.(a.) 8 146 12,7 8,7 130 + 162 The first elements of both strong-started and continuous diphthongs a and 'a are longer than e because, as in common language, they have a greater proper length (see Pakerys 1982: 48). Apparently, this also leads to a longer overall duration: diphthongs with the first accent a or ‘a are significantly longer than ei (see table 1). 92 Table 3. Duration of second segments (ms) Diphthong| n x s | v(96) | patik.int.- (9596) | 1%q, (d.)i 12 100 3,8 3,8 96+ 105 10.1 >5,0** (a.)i. 12 126 43 3.4 121 + 131 (e.)i 12 101 6.8 6.7 93+110 1.7 <23 OF 12 109 6.7 6.1 101 +117 (d.)u 12 103 3.6 3.5 99 + 108 5.7 >5.0** (a.)ū 12 121 5,8 4,8 113 + 128 (d.)u 8 116 18.0 15.5 97+ 135 15 <23 (a.)ū 8 132 16.7 12.6 112 + 153 The second elements of the firm and continuous diphthongs differ much less than the first. This is best seen in the diphthongs ei and au: the confidence intervals of the second elements intersect ((€.)i 93 + 110 ms —(ė.)i 101 + 117 ms; (d.)u 97 + 135 ms — (a.)ii 112 + 153 ms), and t is less than the critical value of f, ((e)i -1.7 < 2.3, (a)u -1.5 < 2.3). The lower reliability of the duration of the second segments is confirmed by higher variation coefficients, especially in continuous diphthongs (see tables 2 and 3). Thus, at least the duration of the second elements ei and au is not an important distinguishing feature of the preposition. In other diphthongs, the difference between the first and second accents is also greater. The results of the measurement show that although the duration in the continuous diphthong is distributed similarly as in the northern Samogitian dialects, the more important indicator of precocity is the first dot. In this respect, Eržvilko- ’s language is closer to the common language (cf. Pakerys 1968: 114) and differs fundamentally from Varniškiai, where the duration of the second accent is more significant (cf. Bukantis 1984a: 73). 5.1.4. Already from hearing it can be said that the solidelements of initial and continuous diphthongs differ not only in quantity, but also in quality. More pronounced differences are found in the first elements: the solid-start elements are more similar to the corresponding monophthongs, they are less assimilated with another diphthong element, as in the common language (cf. Pakerys 1974: 149), have a lower elevation and somewhat posterior articulation, while the first elements of the continuing ones have a higher elevation, their articulation is more anterior (the first form F, lower, and the second F, higher) (see Figure 1). The lowest elevation after the spot is in the diphthong d.u: F; =700 Hz, and the highest, as expected,—c.i. : Fy = 440 Hz (see Figure 1.1.c). In the upper case, the ascending and descending vowels are separated by a point. A more diffuse and higher-tone sound is pronounced (see table 2), similar to the e of the middle rise. Similarly, i acts in a diphthong ai: both the continuous and the firm a(i) of higher rise and more forward articulation than a in a diphthong au (see figure 1 bir d), because « somewhat pulls back the articulation of the first point“" and only slightly elevates its rise. In continuous diphthongs, the influence of the second accent is even stronger than in the strong accents: '@.(i.) and 4.(u.) are more labialized, of lower tone than the corresponding strong accents (see table 4).' In common language, the first elements of the diphthongs also depend on the quality of the second: d.(i) moves forward, and d.(u) moves backward (Pakerys 1974: 150). 93 (a) 'a.(u) —'a.(ū.) (b) a.(u) — a.(ū.) Hz Hz vani aaa, 20005 aaa. there aaa — i 1400 T 1200 Fn a I S i i 1500 T 1000 + so 2034 0 De S —————————0 0 == — (c) is.(i) —is.(i) (d) a.(i) —a.(i) Hz Hz 2500 T 2000 T 2000 amas | | —| ta ta a ma 1500 1500 1900 | ios ae 500 E———————————- —500 + + Bae m 0 Figure 1. "The first two-parts of the 4th volume". Qualitative indices of diphthongs Diphthongs Formants Compactness Tonality Bemolarity F, F, yeah. 740 1860 878 360 163 u 490 980 901 -521 176 'Ain. 650 1750 867 250 165 th. 450 990 886 -510 177 d. 730 1350 915 -150 167 u 540 870 929 -628 176 a. 640 1250 906 -250 169 i. 440 920 892 -575 176 d. 690 1490 894 -10 166 i 330 2240 753 736 170 a. 620 1640 868 135 167 é 570 2120 829 620 164 i 370 2180 770 677 169 é. 440 2150 793 650 165 i. 340 2290 754 790 170 310 2250 744 755 171 '* The values of the forming elements of the solid base are marked with dotted (---), and the forming elements of the solid tip with solid (—) lines. 94 The secondary elements of continuous diphthongs are slightly more diffuse and of higher tone. Their formant values are further away from the center of the spectrum, so the sounds are more tense. The dynamics of all the diphthongs shows that the first elements pronounced with continuous consonant are more smoothly transferred to the second ones without greater fracture (see Figure 2). Continuing education The trajectories of the formative frequencies change little, thus both sounds are not only of the same order, but also of similar rise, while the F of the firm beginning €.i, much lower, only rises sharply at the end, when pronouncing i. In the diphthong d.i. the fracture is even more pronounced because the points are of different orders. J. Bukantis presents a similar picture of the ai-forming letters in the Varniai dialect (see Bukantis 1984b: 63, fig. 43), although, in his opinion, it is often not a sufficient indicator of preposition (Bukantis 1984a: 79). a) is..(~-) —is (—) (b) G.i.(—) —d.i (—) Hz Hz = = S = 2500 £ 2500 = 8 2000 = —==x 8 2000 — 's 1500 + a 1500 £ 1000 + Co E 1000 | E 500 —£ —— © ——a 500 + £ Di E 0 +t —— į LL 1 2 4 ¢)a.. (---) —d.u (—) (d) ‘aa. (---) —'d.u (—) a N on N = in ‘E2000 > 2 2500 »N | = S 1500 + + X 8 2000 N 8 1000 Za, WE = 0 = £ | = 1000 NN E500 #7 with BT 500 erm ———— £ 0. — ttt E 0 41 o o LL 1 LL 1 Figure 2. The trajectories of the u-forming diphthongs with the second accent are not so straight: about the beginning of au, but in the middle of 'au, a greater or smaller peak is visible, and then the curves of both u-forming diphthongs descend. However, the two forms of strong-started diphthongs, although the frequency values fluctuate, remain in the higher frequency range, with a sharper drop only at the very end of the segment. "A.S.O.S.- ". and a.ii. the fall of both formants is slightly smoother. * Formants measured in stem diphthongs. Nine measurement points were taken in each diphthong. 95 The results of the spectral analysis do not contradict the conclusion made in § 5.1.2 that the features of the first diphthong spot are more important. Although the spectral characteristics of the second element change due to the prefix, this does not have much importance for the perception of the prefix, because even in jų continuous diphthongs u, i are too short for small differences to be noticed. 5.2. Mixed diphthongs. 5.2.1. The ratio of the duration of continuous mixed diphthongs to the duration of continuous mixed diphthongs is similar to that of pure diphthongs: continuous diphthongs are about 1.2 times longer than continuous diphthongs. The total duration of the whole diphthong depends not only on the prefix, but also on the natural duration of the individual elements, since the longest are those strong-started diphthongs, whose second elements are nasal sonates (see Table 5). As in common language, the diphthong is most prolonged by n (Pakerys 1982: 51); m is slightly shorter, so the diphthongs of the a, e+m structure are also shorter. The shortest of the sonates is r, but with it formed strongstart diphthongs are quite long. Since the first accent of a strong-started diphthong has a greater proportion of the vocal pressure and duration, it becomes longer (a —up to 259 ms, e —up to 253 ms; see table 6) and compensates for the shortness of r. Table 5. The duration of continuous mixed diphthongs Diphthong | n | x | s | vw | patikint (95%) | 1771, al 8 334 5.9 1.8 326 + 341 8.0 >5.0** th.l. 8 272 16.3 6.0 251 + 292 and 8 342 5.6 1.6 336+ 349 13,4 >5,0** af 8 265 11,8 44 250 + 279 im 8 338 3,9 1,1 333+ 342 10,9> 5,0** am. 8 279 114 4,1 265 + 293 in 8 393 4,1 1,0 388+ 398 8.4 > 5.0** if. 8 348 11.2 3.2 335 + 362 el 8 358 11.8 3.3 343+ 373 4.5 >3.8* el 8 334 1.0 0.3 333 + 335 ér 8 342 16.7 4.9 316 + 369 5.3 >4.3* or 8 271 20.9 77 238 + 304 em 8 363 6.6 1.8 354+ 371 6.1 >5.0** n.m. 8 340 4,7 1,4 335 + 346 én 8 397 1,9 0,5 395+ 400 11,1>5,0** č.ū. 8 358 7,6 2,1 349 + 368 5.2.2. As with bivowels, the duration of the accents in mixed doublets is quite accurate. The ratio of the lengths of the first and second syllables is similar to that of the common language: the first syllables are significantly longer than the second (about 1.8 times). The natural length of the first note is not so important, the lengthening of the vowel depends more on the sonant: mainly the d.r and €.r first notes are lengthened, because 96