چکیده:
از آنجایی که گذار سازهای همخوان به واکه حاوی یکی از سرنخهای آکوستیکی برای تشخیص محل تولید همخوان است، بررسی میزان گذار سازههای یک واکه که همان شیب صعودی یا نزولی سازه در بافت cv است هدف اصلی پژوهش حاضر است. جهت نشاندادن مسیر تغییر مقادیر سازههای F1 و F2 هر یک از واکههای پیشین و پسین ترکی، شکلها با نرمافزار اکسل ترسیم شد و از آنجایی که فضای سازهها با توجه به ماهیت واکه در زنجیرة گفتار تغییر قابل توجهی دارند تاثیر بافت آوایی برروی مقادیر فرکانسی سازههای F1 و F2 در ناحیه گذار مربوط به همخوانهای لبی، تیغهای و بدنهای مورد بررسی قرار گرفت، به عبارتی با دستهبندی و تعمیم گذار سازهها میتوان پارامتری را تعیین کرد که به تشخیص محل تولید همخوانها کمک کند. در نهایت مشخص شد گذار F1همه واکهها به صورت صعودی است. افزایش مقدار گذار F1 واکه پس از پایان رهش همخوانهای لبی، تیغهای و بدنهای یک سرنخ آکوستیکی برای بازبودن مجرای دهان و عدم وجود گرفتگی در دستگاه گفتار است. تحلیلهای آماری برای مقایسه فرکانس سازة دوم هر یک از واکهها در دو نقطة آغاز و مرکز گذار همخوانهای لبی، تیغهای، بدنهای نشان داد که تنها در واکههای [o، u، a] تفاوت معناداری بین فرکانس سازه دوم در دو نقطة آغاز و مرکز گذار پس از همة همخوانهای لبی، تیغهای و بدنهای وجود دارد و در بقیه موارد اختلاف این دو متغیر معنادار نیست.
Abstract Acoustic correlates of consonantal place of articulation are numerous and varied. Burst information, onset spectra, and formant transition motion have all been implicated as the significant cues for consonant place of articulation within each labial, alveolar, and velar place category. The main purpose of this study was to investigate the transition rate of a vowel formant, which was the ascending or descending slope of the formant in CV syllables. Therefore, this analysis offered additional insights to the role of F2 transition at the onset and mid-points when serving as predictor variables. The results showed that F1 vowel transitions had an ascending slope. Increasing the value of F1 vowel transition after the releases of labial, alveolar, and velar consonants was an acoustic clue to the opening of the oral cavity and absence of obstruction/closure in the vocal tract. The statistical analyses demonstrated that only in the vowels [o, u, æ], there was a significant difference between the second formant frequencies at the two onset and the mid-points of the transition after all coronals, labials, and velars. The difference between these two variables was not significant in the other cases. Keywords: Vowel, Transition, Formant, Syllable, Ascending or Descending Slope Introduction Azerbaijani language is mainly spoken in Azerbaijan and Iran (Johanson, 2010). After Farsi as the official language, Azerbaijani has the most (approximately 15-20 million) speakers in Iran (Crystal, 2010). Most of the speakers inhabit in four provinces of the northwestern part of Iran. In the literature, two striking features of Turkic morphology has been presented: first, its agglutinative nature (low level of fusion in Sapir's terminology), indicating the fact that there is basically one-one correspondence between grammatical categories and their exponents and second, its highly synthetic nature (high level of synthesis again by using Sapir's terminology), demonstrating the fact that a given word can contain a large number of morphemes. Furthermore, the syllable structure of Azeribaijani language is consonant-vowel-consonant (consonant). This indicates that tautosyllabic consonant clusters with sonority drop are acceptable. In Turkic languages, the stress falls on the final syllables of all parts of speech (noun, adjectives, adverbs, and verbs). Most of the derivational and inflectional suffixes carry the stress. However, some of the derivational suffixes and clitics are unstressed. When roots combine with an unstressed suffix, the stress falls on the immediately preceding syllable and stress shift may not take place. According to these stress-placement rules, we used target vowels in simple words with the cvc.cvc syllables. Materials and Methods Generally, the first 3 formants of vowels are the most important indicators providing information for distinguishing vowels from each other (Hagino et al., 2008). There is a close relation between the first formant frequency (F1) and articulatory and/or perceptual dimension(s) of vowel height. The second formant frequency (F2) corresponds with the place of maximal constriction during the production of vowels (Wang and Van Heuven, 2006). Measurement of formants is one of the elements of the acoustic analysis of vowels. The first two formants are most important for identifying a vowel (Raina, Chakraborty, and Velankar, 2014). A formant is a concentration of acoustic energy around a particular frequency in the speech wave and corresponds to a resonance in the vocal tract. Formants can be seen very clearly in a wideband spectrogram where they are displayed as dark bands. They are labeled as F1, F2, F3, F4, etc. starting with the lowest frequency (Gunasekar et al., 2017). F1 is primarily related to tongue height with high vowels having a low F1 frequency and low vowels having a high F1 frequency. F2 is mostly related to tongue advancement with back vowels having a low F2 frequency and front vowels having a high F2 frequency (Ludlow, Kent, and Gray, 2019). This study included 10 native Turkish speakers aged 20-45 years. They pronounced Turkish vowels following consonants like xɯr.dɑ, nɑ.xɯr, rɯ.zɑ, and sɑ.rɯ. Acoustic analysis of the WAV files was performed in the computer program Praat, Version 6.1.08 (Boersma and Weenink, 2019). Segmentation of vowels in words was carried out in a manual approach. Formant frequency analysis was conducted using the Formant Pro (Xu, 2018) available in Praat with its default standard settings (range of 5,000 Hz for 5 formants). We used International Phonetic Alphabet (IPA) symbols for Turkish sounds. Discussion of Results and Conclusion Coarticulation is generally defined as the influence of one phonetic segment on another segment (Marchal, 2009). It is an overlap between the articulations of neighboring phonemes. Articulators constantly move as we talk so that the shape of the vocal tract for a particular phoneme is influenced by the shapes for the phonemes that both precede and follow it (Goldstein, 2010). Vocal tract movements produce transitions of formant frequencies. Formant transitions reflect the overall change in the shape of the vocal tract during speech production (Story and Bunton, 2010). The study objective was to analyze F1 and F2 transitions in Turkish vowels in a consonant context with different places of articulation (labials: m, b, p, f, and v; coronals: s, z, d, t, n, and l; alveolo-palatals: ʃ, ʒ, ʣ, r, and ʦ; palatals: ɟ, c, and j; uvulars: ɣ and x. Therefore, this analysis offered additional insights into the role of F2 transition at the onset and mid-points when serving as predictor variables. Formant transitions in consonant-vowel sequences from real words were analyzed. Consonants with different places of articulation were included in the analysis. Finally, statistical analysis of T-test was performed for two similar groups by calculating the significance level of the difference between the F2 frequency values of the onset and the mid-points of the target vowels after the consonants. The results revealed that F1 vowel transitions had an ascending slope. Increasing the value of F1 vowel transition after the releases of labial, alveolar, and velar consonants was an acoustic clue to the opening of the oral cavity and absence of obstruction/closure in the vocal tract. The statistical analyses demonstrated that only in vowels [o, u, æ], there was a significant difference between the frequencies of the second formant at the two onset and mid-points of the transition after all coronals, labials, and velars. The difference between these two variables was not significant in the other cases.
خلاصه ماشینی:
بیجن خان (١٣٧٤) وضعیت هندسی گذار سازة دوم برای همپوشی همخوانی-واکه ای در زبان فارسی را به صورت جدول (١) خلاصه میکند: جدول ١- وضعیت هندسی گذار سازة دوم برای همپوشیهای همخوانی-واکه ای در زبان فارسی (بیجن خان ،٢٨٨:١٣٧٤) Table 1- Geometric status of the second formant transition for consonant-vowel coarticulations in Persian (Bijankhan, 1995: 288) واکه لبی دندانی لثوی لثویکامی سختکامی نرمکامی ملازی [i] صعودی/شیب کم صعودی/شیب کم صعودی /شیب کم نزولی/شیب کم نزولی/شیب کم نزولی/شیب کم صعودی/شیب زیاد [e] صعودی/شیب کم تراز صعودی/شیب کم تراز نزولی/شیب کم نزولی/شیب کم صعودی/شیب کم [a] صعودی/شیب کم نزولی/شیب کم تراز نزولی/شیب کم نزولی/شیب زیاد نزولی/شیب کم نزولی/شیب کم [u] تراز نزولی/شیب زیاد نزولی/شیب زیاد نزولی/شیب زیاد نزولی/شیب زیاد تراز تراز [o] تراز نزولی/شیب زیاد نزولی/شیب زیاد نزولی /شیب زیاد نزولی/شیب کم تراز تراز [ɑ] تراز نزولی/شیب زیاد نزولی/شیب کم نزولی/شیب کم نزولی/شیب کم تراز تراز بیجن خان (١٣٧٤) معتقد است در نظام آوایی زبان فارسی، پارامتر محل گرفتگی برای الگوی [بست کامی] در جایگاه آغازة هجا با 1 Hillenbrand, J.
(به تصویر صفحه رجوع شود) نمودار١- فضای واکه ای ترکی به صورت تابعی از مقادیر فرکانس F١ و F٢ در دو موضع تکیه بر و بیتکیه (چپ ) در دو جایگاه هجای باز و بسته (راست ) (صادقی و محمودی، ٢٤٤:١٤٠٠) Fig 1- Turkish vowel space as a function of frequency values F1 and F2 in two positions of stressed and unstressed (left) in both open and closed syllable positions (right) (Sadeghi and Mahmoodi, 244: 2021) ١این مسئله دو دلیل دارد: اول آن که اگر فرض کنیم بازة فرکانسی F١ برای تولید واکه های افراشته بین ٢٠٠ تا ٣٥٠ هرتز باشد (١٩٨٨ ,Stevens)، واکه های افراشته [i]،u[ ]،y[ ] در ترکی آذری متوسطی نزدیک به بیشینۀ این بازة فرکانسی دارند؛ یعنی این واکه ها به لحاظ آکوستیکی کاملاً افراشته نیستند و از سطح ارتفاع بیشینۀ بدنۀ زبان فاصلۀ نسبی دارند.