Nasals, Approximants, Rhotics, and Laterals
Total Page：16
File Type：pdf, Size：1020Kb
Nasals, approximants, rhotics, and laterals Cơ sở âm vị học và ngữ âm học Formant ranges, F1 and F2: F1: 300 ∼ 1000 Hz Lecture 15 F2: 800 ∼ 2500 Hz Recap Recap • Last time we saw that stops consist of both a closure • Voiceless stops can be identified by concentrations of (cấu âm tắc) and a burst (tiếng nổ ). energy (khu vực năng lượng) in the release burst. 5000 p t k Frequency 0 0 0.738 Time (s) From Kent & Read (1992), The Acoustic Analysis of Speech Recap Nasals • Like stops, nasals involve a vocal tract constriction • Voiced stops, on the other hand, were better identified (cấu âm thắt trong đường dẫn âm)... using formant transitions( chuyển tiếp formant). simmer http://www.cns.nyu.edu/∼david/courses/perception/lecturenotes/speech/speech.html From Ladefoged (2003), Phonetic Data Analysis Nasals Nasals • ...but like vowels, nasals have a periodic glottal source • Energy (amplitude) is lower than for vowels – in part (nguồn thanh hầu có chu kỳ). because nasal membranes absorb the sound (màng simmer nhầy hút âm thanh). From Ladefoged (2003), Phonetic Data Analysis Nasals Nasals • • Since the oral tract is completely closed, this effect is Like stops, the crucial information is contained in the enhanced, giving rise to antiformants( formant yếu). formant transitions. • In practice, the velar nasal is usually the clearest. dimmer dinner dinger Nasals Formant transitions preceding nasals • Like stops, the crucial information is contained in the formant transitions. − [m]: falling F2 • In practice, the velar nasal is usually the clearest. − [n]: level F2 − [N]: rising F2 dimmer dinner dinger (reverse for transitions following nasals) From Ladefoged (2003), Phonetic Data Analysis Nasals Approximants • F1 tends to be low (250-300Hz) and F2 around 2500Hz. In between (where vowels have F2) there is little energy. • Vowels are periodic – complex repeating waves ama ana aNa generated by regular vibrations of the vocal folds. • Consonants may have an aperiodic component, but can still be voiced or voiceless • There are also (largely periodic) segments that lie somewhere in between: approximants. Approximants Approximants • Vowels are periodic – complex repeating waves • Vowels are periodic – complex repeating waves generated by regular vibrations of the vocal folds. generated by regular vibrations of the vocal folds. • Consonants may have an aperiodic component, • Consonants may have an aperiodic component, but can still be voiced or voiceless but can still be voiced or voiceless • There are also (largely periodic) segments that lie • There are also (largely periodic) segments that lie somewhere in between: approximants. somewhere in between: approximants. Approximants Approximants • Acoustically, [j] [w] are very similar to vowels, with • Acoustically, [j] [w] are very similar to vowels, with striations and formants, but often with less energy striations and formants, but often with less energy • For this reason they are sometimes called semivowels. • For this reason they are sometimes called semivowels. • The constriction for [j] [w] is narrower than for [i] [u], but • The constriction for [j] [w] is narrower than for [i] [u], but formant values are similar formant values are similar Approximants Approximants • Acoustically, [j] [w] are very similar to vowels, with striations and formants, but often with less energy • For this reason they are sometimes called semivowels. • The constriction for [j] [w] is narrower than for [i] [u], but formant values are similar From Borden, Harris, & Raphael (1994) Approximants Approximants • What constitutes a diphthong vs. a semivowel + a vowel • [w] and [j] tend to differ from [u] and [i] in that they tend can be controversial... not to have a steady-state portion (một phần không đổi). From Kent & Read (1992), The Acoustic Analysis of Speech From Kent & Read (1992), The Acoustic Analysis of Speech Rhotics (‘r’-tính) Rhotics (‘r’-tính) • The defining acoustic signature of [ô] is the third formant • The defining acoustic signature of [ô] is the third formant (F3) making a steep fall to below 2kHz. (F3) making a steep fall to below 2kHz. • This effect is referred to as ‘r-colouring’ (r-hoá). • This effect is referred to as ‘r-colouring’ (r-hoá). • [ô] – and other rhotics ([r], [R], [ö]) – are the only • [ô] – and other rhotics ([r], [R], [ö]) – are the only approximants for which F3 is (perceptually) important approximants for which F3 is (perceptually) important Rhotics (‘r’-tính) Tongue configuration types of American English /r/ ([ô]) • The defining acoustic signature of [ô] is the third formant (F3) making a steep fall to below 2kHz. • This effect is referred to as ‘r-colouring’ (r-hoá). • [ô] – and other rhotics ([r], [R], [ö]) – are the only approximants for which F3 is (perceptually) important Examples of corresponding American English tongue configuration types for sustained /r/ as identified from MRI by Tiede et al. (2004). Adapted from Tiede et al. (2007), ‘Variability of North American English /r/ Production in Response to Palatal Perturbation’, Haskins Internal Workshop on Speech Production and Motor Control. Rhoticization Taps (âm vỗ) and trills (âm rung) 3: Ç: From Ladefoged (2003), Phonetic Data Analysis Where F3 is useful Lateral approximants • [ô]: characterised by low F2 and especially low F3. • [l]: faint formants at a) low frequency and b) about 1000-1500 and 2500-3000 Hz, followed by abrupt change in amplitude when the tongue tip breaks away. Frequency (Hz) Time (ms) From Ladefoged (2006), A Course In Phonetics From Ladefoged (2006), A Course In Phonetics Clear vs. dark [l] Clear vs. dark [l] • Catalonian (clear, top) vs. Mallorcan (dark, bottom) /l/s • Allophones in many 5000 English dialects: dark /l/ [l] F2 in coda (‘feel’), clear /l/ in onset (‘leaf’) • In Newcastle English, initial /l/ is light; in 0 Recasens & Espinosa (2005), ‘Articulatory, positional and coarticulatory characteristics for Leeds, initial /l/ is dark 5000 clear /l/ and dark /l/: evidence from two Catalan (Carter & Local, 2007) [ɫ] dialects’, JIPA 35(1), 1-25. F2 Frequency(Hz) • In Scottish English, /l/ can be dark in both positions. 0 Time Clear vs. dark [l] Clear vs. dark [l] • Allophones in many • Allophones in many English dialects: dark /l/ English dialects: dark /l/ in coda (‘feel’), clear /l/ in coda (‘feel’), clear /l/ in onset (‘leaf’) in onset (‘leaf’) • In Newcastle English, • In Newcastle English, initial /l/ is light; in initial /l/ is light; in Leeds, initial /l/ is dark Leeds, initial /l/ is dark (Carter & Local, 2007) (Carter & Local, 2007) • In Scottish English, /l/ • In Scottish English, /l/ can be dark in both can be dark in both positions. positions. Spectrogram practice 5000 [l] F2 0 5000 [ɫ] F2 Frequency(Hz) • What distinguishes these approximants? 0 Time .