An Acoustic Study of VOT in Pahari Stops.
This study examines voice onset time (1707) for stops in Pahari. Ten Pahari native speakers participated in the study. The participants were asked to pronounce twelve Pahari stops thrice in VCV context. It is found: 1) voiceless stops have short VOT, voiceless aspirated have long VOT, while voiced stops are pronounced with prevoicing 2) VOT of dental and alveolar is less than those of bilabial and velar stops; this supports Hardcastle's (1973) claim that faster the movement of the articulator, the shorter the VOT, as both dental and alveolar are articulated with the tip of the tongue. The study suggests following hierarchy in Pahari: velars greater than bilabial greater than alveolar greater than dental.
Keywords: voice onset time (VOT), aspirated stops, unaspirated stops, lag VOT
For the past few decades, a considerable number of studies have investigated voicing contrasts in stops by the use of voice onset time (V01). VOT is considered as one of the most important methods for examining the timing of voicing in stops and has been applied in studies of many languages. By analyzing VOTs in stop consonants, linguists have concluded that for most languages, VOT values get longer as the place of articulation moves backward (Lisker and Abramson, 1964; Cho and Ladefoged, 1999). However, there are some exceptions, such as Mandarin, which does not follow the general rule (Lisker and Abramson, 1964; Cho and Ladefoged, 1999; Chao, Khattab and Chen, 2006). However, no attempt has been made to examine VOT patterns in Pahari so far. Therefore, the purpose of present study is to provide information for a general VOT pattern of Pahari stops.
Pahari has 12 oral stops articulated from four different places of articulation namely bilabial, dental, alveolar and velar. It has three way voicing contrast: voiced, voiceless unaspirated and voiceless aspirated. It is also important to mention here that aspiration is phonemic in Pahari. Following table illustrates Pahari oral stops.
Table 1: Stops of Pahari
* To investigate differences in VOT by voicing in Pahari stops
* To explore the differences in VOT by place of articulation in Pahari stops
Lisker and Abramson (1964, p. 422) defined VOT as "the time interval between the burst that marksrelease of the stop closure and the onset of quasi-periodicity that reflects laryngeal vibration". Ladefoged (2001) states that stops differ along a variety of acoustic dimensions: closure duration, voice onset time (VOT), the spectral characteristics of their release burst, and formant transitions of surrounding sonorant. Lisker and Abramson (1964) argue that VOT is a strong cue to voicing contrasts between stops. It is also well known that VOT varies with place of articulation. The principal findings regarding VOT with respect to place of articulation of stops are: (1) the further back the closure, the longer the VOT (Fischer-Jorgensen, 1954; Peterson and Lehiste, 1960); (2) the more extended the contact area, the longer the VOT (Stevens, Keyser and Kawasaki, 1986); and (3) the faster the movement of the articulator, the shorter the VOT (Hardcastle, 1973). These patterns are known for many years.
In a crosslinguistc study, Lisker and Abramson (1964) explore that velar stops always have a longer VOT. Furthermore, they show that in both aspirated and unaspirated stops, VOT is shortest for bilabial stops and intermediate for alveolar stops, with the exception of the unaspirated stops in Tamil and the aspirated stops in Cantonese and Eastern Armenian.
In recent years many studies (the most important are Cooper, 1991; Docherty 1992; Jessen, 1998) report variations in VOT. These variations depend on a number of factors, including laws of aerodynamics, articulatory movement velocity, and differences in the mass of the articulators (Cho and Ladefoged, 1999). There is an alternative analysis that
suggests that there is a temporal adjustment between stop closure duration and VOT (Weismer, 1980; Maddieson, 1997).
Maddieson (1997) suggests that one of the factors which contribute to VOT differences is the relative size of the supraglottal cavity behind the point of constriction. According to Cho and Ladefoged (1999), there are two ways of considering this: firstly, the cavity behind the velar stop has a smaller volume than that behind the alveolar or bilabial stops; secondly, the cavity in front of the velar stop has a larger volume than that in front of the alveolar or bilabial stops.
Hardcastle (1973) finds that VOT difference can be due in part to the fact that the tip of the tongue and the lips move faster than the back of the tongue. Kuehn and Moll (1976) support his claim and add that the articulatory movement is fastest for the tongue tip, intermediate for the lower lip, and slowest for the tongue body. As Maddieson (1997) notes,
If the articulatory velocity is the primary physiological factor for the voice onset difference, we would expect that the VOT would be shorter for apical alveolar stops than for either bilabials or velars, which is not the general finding. This suggests there may be some other factors accounting for the place-related voice onset difference.
Stevens (1999) states that the rate of change in intra-oral pressure following the release depends on the rate of increase in cross-sectional area at the constriction. This is significantly different for different places of articulation, primarily due to the differences in the extent of articulatory contact. When there is a long narrow constriction, the Bernoulli effect causes the articulators forming the constriction to be sucked together. Because the velar stop has extensive contact between the tongue body and the palate, there is a larger Bernoulli force so that the change in cross- sectional area is relatively slow compared with that for the bilabial or alveolar stops. Consequently, the decrease in intraoral pressure after the closure is gradual for the velar and rapid for the bilabial.
The above mentioned literature shows that phoneticians do not agree on a common ground, possibly due to the cross-linguistic variations. Following
section aims to determine the acoustic cues of Pahari stops and to discuss their idiosyncrasy in the light of previous studies.
A total of 10 Pahari speakers (five male and five female) participated in the study. The subjects were born in the Poonch dialectal area. They all are in their twenties and graduate students. All the speakers are multilingual. They speak Pahari at home and Urdu in market places. They had also formal classes of English in schools and colleges. All the participants reported no history of any language impairment. None of them had any phonetic training and knowledge for this kind of experiments.
The stimuli consist of Pahari words and non-words with a VCV sequence, in which C is one of the target consonants. To reduce variation from the vowel context, an identical vowel /a/ is used for both V1 and V2 positions for all stimuli. A list of 12 Pahari stops in VCV has been prepared for recording.
Table 2: Pahari stoDs
A total of 360 tokens (12 plosive phrases x 3 repetitionsx 10 participants) were acoustically analyzed for VOT. VOT was measured by visually examining the spectrograms in the PRAAT software. Positive VOT was measured by placing a time marker at burst onset of each stop and another marker at the onset of the second formant in the following vowels while for the negative VOT the portion of voicing during the closure duration preceding the release of stops was measured. The PRAAT software displays spectrogram, in terms of time denoted in milli seconds (ms) along the horizontal axis. This allows for direct measurement of the VOT between the
markers. All VOT measures were carried out similarly for all 12 stop consonants of Pahari. Mean VOT values were calculated for the stops produced by each participant. The analysis involves displaying spectrogram and waveform on the screen, as shown in Figure 1.
Results and Discussion
The following section reports the VOT of Pahari stops:
Voice Onset Time (VOT)
Voice Onset Time (V01) was measured for each stop for each speaker. The results are summarized in tables 3, 4, 5 and 6.
Table 3: Mean VOT and standard deviation for Pahari bilabial stops
Bilabials###No. of Tokens###Mean###Standard Deviation
Table 4: Mean VOT and standard deviation for Pahari dental stops
Dental###No. of Tokens###Mean###Standard Deviation
Table 5:Mean VOT and standard deviation for Pahari alveolar stops
Alveolar###No. of Tokens###Mean###Standard Deviation
Table 6: Mean VOT and standard deviation for Pahari velar stops
Velars###No. of Tokens###Mean###Standard Deviation
VOT by Voicing Type
As mentioned earlier, Pahari stops have three way voicing: voiced, voiceless unaspirated and voiceless aspirated. Pahari voiced stops are characteristically fully voiced throughout their closure duration, with VOTs ranging between -94 ms for [b] and -109 ms for [g]. But some occurrences are found during the analysis of voiced stops where voiced stops are not fully voiced during closure. Figure 2 shows the spectrograms of the fully- voiced closure of a [g] in the [aga] context.
The /g/ in the above [ag a] context is fully voiced and has a negative VOT of about 81 ms. This is clearly seen from the pitch contour, which is present during the whole articulation of /g/. Since pitch can only be measured on voiced parts, the pitch contour can be used to identify voicing.
The spectrogram above shows that the pitch contour is absent for /k/ during the closure, but voicing starts as soon as the closure is released and the vocalic part begins. In above spectrogram the VOT of /k/ is 22 ms. This reveals that voiceless unaspirated stops have positive VOT. Voiceless aspirated stops in Pahari also have positive VOT, and it is longer than VOT of unaspirated voiceless stops. The following spectrogram shows the VOT of aspirated voiceless velar stop /kh/.
The spectrogram of /kh/ above indicates that there is a positive long VOT of 101 ms. VOT of /kh/ is positive and is the time between the onset of the burst and the onset of the following vowel /a:/. The spectrograms above exhibit a pattern of VOT for Pahari: voiceless stops have short VOT, voiceless aspirated have long VOT and voiced stops are pronounced with prevoicing. Table 8 shows the average VOT by voicing type.
Figure 5 illustrates that the VOT of the three stop voicing types (i.e. voiceless unaspirated, voiceless aspirated and voiced) are clearly separated and their VOT shows strength hierarchy: voiceless aspirated greater than voiceless unaspirated greater than voiced. A single-factor ANOVA confirms that the p-value is smaller than 0.001 that means the difference between the three voicing types is highly significant. Moreover a pairwise t-tests indicates highly significant differences between any two voicing types as the p-value is smaller than 0.001. It can be concluded that Pahari exhibits a pattern found in many other languages: voiceless stops have short VOT, voiceless aspirated have long VOT and voiced stops are pronounced with prevoicing.
By Place of Articulation
In addition to voicing type, Cho and Ladefoged (1999) claim, in a cross- linguistic study, that VOT characteristically increases with the backness of a stop's place of articulation. Pahari stops are not consistent with this cross- linguistic generalization, as shown in Figure 6.
To see the effect of places of articulation one-way ANOVA test is performed and the p-value is found smaller than 0.001 which means that the place of articulation is statistically significant. Moreover, to determine the differences between bilabials and dentals; bilabials and alveolars; bilabials
and velars; dentals and alveolar; dental and velars; alveolar and velars, a pairwise test shows the places of articulation differ from each other significantly as the p-value is found smaller than 0.001 except for dental and alveolar where the p value is greater than 0.001.
The previous studies show that the VOT is shortest for bilabials, longer for alveolars, and longest for velars as shown in hierarchy: bilabial less than alveolar less than velars. This suggests that the more we go back in oral cavity, the greater the VOT. However, this study shows that the VOT of dental and alveolar is not consistent to previous studies and is less than bilabial and velar. Velar has longer VOT for unaspirated, aspirated and voiced stops than all the other places of articulation. Dental and alveolar show similar VOT duration (35 ms and 36 ms respectively), and this smaller difference between them is not statistically significant. On the other hand, VOT durations are statistically significant between bilabial and dental, alveolar and velar, and between bilabial and velar.
The above results show that Pahari has three way voicing: voiced, voiceless unaspirated and voiceless aspirated. Statistical test also proves that VOT of three way voicing is clearly separated from each other. It is important to make a comparison of the VOT values of Pahari stops with those measured for other languages. For four American English speakers, the VOT values of /p/, /t/, /k/, /b/, /d/, and /g/ are found to be 58, 70, 80, 1, 5, 21 ms, respectively (Lisker and Abramson, 1964). In the study of Macleod and Stoel- Gammon (2005), the VOT values for three Canadian English speakers are measured and the mean VOT values of voiced stops /b/, /d/ and voiceless stops /p/, /t/ are found 19.8 and 87.9 ms, respectively.
Lisker and Abramson (1964) report that for two Puerto Rican Spanish speakers the VOT values of voiceless stops /p/, /t/, /k/ are 4, 9, and 29 ms, respectively whereas the voiced stops /b/, /d/, /g/ are -138, -110, and -108 ms, respectively. They further report the VOT for three Thai speakers and the VOT values of /p/, /t/, /k/, /b/, /d/ are found to be 6, 9, 25, -97, -78 ms, respectively. Kessinger and Blumstein (1997), in their study on four Thai
speakers, report that the VOT values of /p/, /t/, /b/, /d/ are 15, 13, -70, -65 ms, respectively.
VOT is found to be an effective means to distinguish between voicing categories in oral stops. The above discussed studies show that the voiceless stops /p/, /t/, /k/ have long positive VOTs while voiced stops /b, d,, d, g/ have negative VOT except English where VOT is positive. For Italian, Spanish, and Thai, the voiced and voiceless stops have voicing lead and short voicing lag, respectively. Thus the value of VOT is a good indicator for voiced and voiceless stops.
In Pahari language both the voiceless unaspirated /p, t , t, k/ and voiceless aspirated /ph, r t h, th, kh/ stops have positive VOT values in the range of 16 to 27 ms and 53 to 77 ms, respectively. In the voiced stops /b, d, g/, the VOT values are negative and are in the range -94 to -109 ms This indicates that voiceless unaspirated have short voicing lag while voiceless aspirated have long VOT. On the other hand, voiced stops have voicing led. It is also found that the manner and place of articulations affect the values of VOT. Thus it can be used as a reliable acoustic parameter that indicates the contrast between voiced and voiceless stops. Also, VOT can be used for the distinction of bilabials, dentals, alveolar and velars from one another.
Regarding the effect place of articulation on VOT, the study finds that:
* Velar stops have the longest VOT.
* VOTs of dental stops and alveolar stops are overlapping. The mean of VOT value for dental stops is 35 ms and that for alveolar stops it is 36 ms. A paired t-test showed that difference between these means is not significant.
* VOTs of dental and alveolar are less than bilabial and this provides the evidence against the claim that VOT difference is the relative size of the supra-glottal cavity behind the point of constriction (Maddieson, 1977).
* A significant difference is found in the VOT of bilabial and velar stops in Pahari.
The study suggests the following hierarchy in Pahari: velars greater than bilabial greater than alveolar greater than dental.
The data have demonstrated that Pahari contrasts voiceless unaspirated, voiceless aspirated and voiced stops at labial, dental, alveolar, and velar places of articulation. It is found that voiceless stops have short VOT, voiceless aspirated have long VOT and voiced stops are pronounced with prevoicing. Regarding place of articulation, a few interesting trends emerge from this study. Cho and Ladefoged (1999) claimed that VOT characteristically increases with the backness of a stop's place of articulation. Pahari stops are not consistent with this cross-linguistic generalization, as VOT of dental and alveolar is less than those of bilabial and velar stops. The findings support Hardcastle's (1973) claim that faster the movement of the articulator, the shorter the VOT, as both dental and alveolar stops are articulated with the tip of the tongue. The study suggests following hierarchy in Pahari: velars greater than bilabial greater than alveolar greater than dental.
Abbasi, G. A. (2010). Dilemma of usage and transmission: A sociolinguistic investigation of Dhundi-Pahari in Pakistan. Language in India, 10. Retrieved from www.languageinindia.com/may2010/pakistanpaharifinal.html.
Amina, L. (2002). Phonemic inventory of Siriaki and acoustic analysis of voiced implosives. Center for Research in Urdu Language Processing (CRULP), 43-48. Retrieved from www.scribd.com/doc/19620943/Sindhi-Implosive.
Billerey-Mosier, R. (2003). An acoustic analysis of Kiowa Stops. UCLA, 1-15. Retrieved from less than www.linguistics.ucla.edu/people/grads/billerey/ papers/RBM_Kiowa_stops.pdf.
Cho, T., and Ladefoged, P. (1999). Variations and universals in VOT: Evidence from 18 languages. Journal of Phonetics, 27, 207-229.
Clark, J., and Yallop, C. (1992). An introduction to phonetics and phonology. Oxford: Blackwell.
Davenport, M., and Hannahs, S. J. (2005). Introducing phonetics and phonology. London: Hodder Arnold.
Dixon, R. M. W. (1980). The languages of Australia. Sydney: Cambridge University Press.
Docherty, G. (1992). The timing of voicing in British English obstruents. Berlin: Foris Publications.
Fischer-Fergensen, E. (1954). Acoustic analysis of stop consonants. Miscellanea Phonetica, 2, 42-59.
Gussenhoven, C., and Jacobs, H. (1998). Understanding phonology. London: Arnold.
Hardcastle, W. J. (1973). Some observations on the tense-lax distinction in initial stops in Korean. Journal of Phonetics, 1, 263-271.
Hyman, L. M., and Kenneth,V. (2002). Tone and syllable structure in Hakha-Lai. Berkeley: University of California.
Karnai, M. K. (2007). Pahari aor Urdu: Ik taqabali jaiza. Islamabad: National Language Authority.
Kent, R. D., and Charles, R. (2002). The acoustic analysis of speech. San Diego: Singular Publishing Group.
Koul, 0, N. (2008). Studies in Kashmiri linguistics. Delhi: Indian Institute of Language Studies.
Ladefoged, P. (2001). A course in phonetics. USA: Harcourt College Publishers.
Ladefoged, P. (2003). Phonetic data analysis: An introduction to fieldwork and instrumental techniques. Malden, MA: Blackwell.
Ladefoged, P., and Maddieson, I. (1996). Sounds of the worlds languages. Oxford: Blackwell.
Lindau, M. (1985). The story of /r/. In V.A. Fromkin (Ed.), Phonetic linguistics: Essays in honour of Peter Ladefoged (pp. 157-168). Orlando: Academic Press Inc.
Lisker, L., and Abramson, A. (1964). A cross-language study of voicing in initial stops: Acoustical measurements. Wora 20, 384-422.
Lisker, L., and Abramson, A. (1970). The voicing dimension: Some experiments in comparative phonetics. In Proceedings of the 6th International Congress of Phonetic Sciences, Prague, 563-567.
Lothers, M., and Lothers, L. (2003). Mirpuri immigrants in England: A sociolinguistic survey. Unpublished paper.
Lothers, M., and Lothers, L. (2010). Pahari and Pothwari: A sociolinguistics survey. Islamabad: Summer Institute of Linguistics.
Maddieson, I. (1984). Patterns of sounds. Cambridge: Cambridge University Press.
Maddieson, I. (1997). Phonetic universals. In W. J. Hardcastle and J. Laver (Eds.), The handbook of phonetic sciences (pp. 619-639). Oxford: Blackwell.
McDonald, M. (2002). A study of the phonetics and phonology of Yaraldi and associated dialects. Munich: Lincom Europa.
Peterson, G. E., and Lehiste, I. (1960). Duration of syllable nuclei in English. Journal of the Acoustical Society ofAmerica, 32, 693-703.
Rehman, A. K., and Baart, L. G. (2002). The language of the Qurashis of Kundal Shahi in Azad Kashmir. In L. G. Baart (Ed.), Pakistani languages and society (pp. 1-20). Islamabad: National Institute of Pakistan Studies and Summer Institute of Linguistics.
Roach, P. (1991). English phonetics and phonology. Cambridge: University Cambridge Press.
Saghir, M. (2000). Pahari lok Beet. Muzaffarabad: Kashmir Academy. Shackle, C. (1979). Problems of classification in Pakistan Punjab. Transactions of the Philological Society, 191-210.
Stevens, K. N. (1998). Acoustic phonetics. Cambridge: MIT Press.
Stevens, K. N., Keyser, S. J., and Kawasaki, H. (1986). Towards a phonetic and phonological theory of redundant features. In J. S. Perkell and D. H. Klatt (Eds.), Invariance and variability in speech processes (pp. 426-449). New Jersey: Lawrence Erlbaum.
Tabassum, N. (2003). A phonological analysis of Pothwari/Mirpuri language (Unpublished MA Thesis). University of Glasgow, United Kingdom. Weismer, G. (1980). Control of the voicing distinction for intervocalic stops and fricatives: Some data and theoretical considerations. Journal of Phonetics, 8, 427-38.
Yip, M. (2002). Tone. Cambridge: University Press.
|Printer friendly Cite/link Email Feedback|
|Publication:||Kashmir Journal of Language Research|
|Date:||Jun 30, 2011|
|Previous Article:||EMP Needs of Medical Undergraduates in a Saudi Context.|
|Next Article:||Error Analysis of the Writing of Pakistani College Students: From Causes to Types to Rectification.|