Abstract
Stress is the property of a language to exhibit prominence or distinction in one or more syllables in a given domain. The existence of word stress has not been suitably explored in previous acoustic studies of the Mizo language, which is a tonal language of the Kuki-Chin sub-category in Tibeto-Burman language families. In this study, we attempt to analyze word stress on disyllabic target words, specifically in three lexical categories—adjectives, nouns, and verbs. Utterances of the target words are recorded in isolated setting (out of focus) and in sentence frames (in focus). First, averages of features, namely—duration, intensity, F0, formants, and spectral tilt, are extracted and investigated for identification of stressed and unstressed syllables on a total of 2,880 samples. Next, the interaction of word stress on the four tones of Mizo is investigated. While it is found that H-tone is generally stressed, inferences are made that stressed syllables are not unique to a specific tone. Third, significance of the selected features are validated using a two-tailed paired sample t-test. Our analysis indicates that the mean differences in duration, intensity, and F0 of the stressed and unstressed syllables are significant across the lexical categories at p < 0.05. Next, validations on the significance of the mean differences are carried out using Cohen’s d effect size and Pearson’s Correlation Coefficient (r). Finally, three machine learning models—Support Vector Machines (SVM), Naive Baye’s, and Ensemble learning methods (AdaBoost and Boosted Aggregation), are used to identify stressed and unstressed syllables associated with tones in Mizo. Discriminating differences, especially in disyllabic verbs, are observed between stressed vs. unstressed syllables. Conclusions are drawn that duration is a strong and robust cue for acoustic correlates of stress, while intensity is a medium cue for stress and F0 a weak cue for stress.
- [1] . 2016. Identification of rules for recognition of named entity classes in Mizo language. In Proceedings of the 15th Mexican International Conference on Artificial Intelligence. IEEE, 8–13.
DOI: Google ScholarCross Ref
- [2] . 2014. The Use of Praat in Corpus Research. Oxford University Press, Oxford, UK, 342–360.
DOI: Google ScholarCross Ref
- [3] . 1997. Stress, prominence, and spectral tilt. In Proceedings of ESCA Workshop on Intonation: Theory, Models and Applications. ISCA, 67–70.Google Scholar
- [4] . 2017. Stress patterns and acoustic correlates of stress in Balti Tibetan. Himal. Ling. 15, 2 (2017), 1–49.
DOI: Google ScholarCross Ref
- [5] . 2008. Phonetic cues to stress in a Tonal language: Prosodic prominence in San Lucas Quiaviní Zapotec. In Proceedings of the Annual Conference of the Canadian Linguistic Association.Google Scholar
- [6] . 2000. Tone Sandhi: Patterns across Chinese Dialects. Vol. 92. Cambridge University Press, Cambridge, UK.Google Scholar
Cross Ref
- [7] . 1986. A Preliminary Grammar of the Mizo Language (Tibeto-Burman). Ph. D. Dissertation. The University of Texas at Arlington.Google Scholar
- [8] . 1978. Modern Spectrum Analysis. IEEE Computer Society Press, New York.
DOI: Google ScholarCross Ref
- [9] . 2002. It’s the effect size, stupid. In Proceedings of the British Educational Research Association Annual Conference, Vol. 12. BERA, Exeter, 1–18. Retrieved from http://www.leeds.ac.uk/educol/documents/00002182.htm.Google Scholar
- [10] . 1960. A coefficient of agreement for nominal scales. Educ. Psychol. Measur. 20, 1 (1960), 37–46.
DOI: Google ScholarCross Ref
- [11] . 1992. Statistical power analysis. Curr. Direct. Psychol. Sci. 1, 3 (1992), 98–101.
DOI: Google ScholarCross Ref
- [12] . 2018. Robust Mizo continuous speech recognition. In Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH’18). ISCA, 1036–1040.
DOI: Google ScholarCross Ref
- [13] . 2009. The phonetics of register in Takhian Thong Chong. J. Int. Phonet. Assoc. 39, 2 (2009), 162–188.
DOI: Google ScholarCross Ref
- [14] . 2020. A survey on ensemble learning. Front. Comput. Sci. 14, 2 (2020), 241–258.
DOI: Google ScholarDigital Library
- [15] . 2002. Articulation of extreme formant patterns for emphasized vowels. J. Phonetica 59, 2-3 (2002), 134–149.
DOI: Google ScholarCross Ref
- [16] . 1992. Some Aspects of the Lexical Phonology of Mizo and English: An Autosegmental Approach. Ph. D. Dissertation. CIEFL, Hyderabad. Retrieved from http://hdl.handle.net/10603/130415.Google Scholar
- [17] . 2014. Efficient and scalable multi-class classification using naïve Bayes tree. In Proceedings of the International Conference on Informatics, Electronics & Vision (ICIEV’14). IEEE Computer Society Press, 1–4.
DOI: Google ScholarCross Ref
- [18] . 2014. Hybrid decision tree and naïve Bayes classifiers for multi-class classification tasks. Exp. Syst. Applic. 41, 4 (2014), 1937–1946.
DOI: Google ScholarDigital Library
- [19] . 2017. Tone and stress at the word level. In Intonation and Prosodic Structure, Key Topics in Phonology. Cambridge University Press, Cambridge, UK, 178–224.
DOI: Google ScholarCross Ref
- [20] . 1989. The vowels of stressed and unstressed syllables in nonnative English. Lang. Learn. 39, 3 (1989), 341–373.
DOI: Google ScholarCross Ref
- [21] . 2020. Lexical tone recognition in Mizo using acoustic-prosodic features. In Proceedings of the 12th Conference on Language Resources and Evaluation (LREC’20). European Language Resources Association, 6458–6461. Retrieved from https://www.aclweb.org/anthology/2020.lrec-1.795.Google Scholar
- [22] . 2021. Learning Mizo tones from F0 contours using 1D-CNN. In Proceedings of the International Conference on Speech and Computer (LNAI, 12997). Springer, 214–225.
DOI: Google ScholarDigital Library
- [23] . 2017. Acoustic correlates of word stress: A cross-linguistic survey. Ling. Van. 3, 1 (2017), 1–11.
DOI: Google ScholarCross Ref
- [24] . 2012. Role of pitch slope and duration in synthesized Mizo tones. In Proceedings of the 6th International Conference on Speech Prosody. ISCA, 39–42.Google Scholar
- [25] . 1987. Stress and the cycle. Ling. Inq. 18, 1 (1987), 45–84.
DOI: Google ScholarCross Ref
- [26] . 2007. Signals and Systems. John Wiley & Sons, New Delhi.Google Scholar
- [27] . 2017. Evaluation of spectral tilt measures for sentence prominence under different noise conditions. In Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH’17). ISCA, 3211–3215.
DOI: Google ScholarCross Ref
- [28] . 2018. Spectral tilt as a correlate of Papuan Malay word stress. In Proceedings of the 9th International Conference on Speech Prosody. ISCA, 339–343.
DOI: Google ScholarCross Ref
- [29] . 2015. T test as a parametric statistic. Kor. J. Anesthesiol. 68, 6 (2015), 540.
DOI: Google ScholarCross Ref
- [30] . 2018. Production and perception of rising tone sandhi in Mizo. In Proceedings of the 6th International Symposium on Tonal Aspects of Languages (TAL’18). ISCA, 114–118.
DOI: Google ScholarCross Ref
- [31] . 2020. Interaction of tone and voicing in Mizo. In Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH’20). ISCA, 1903–1907.
DOI: Google ScholarCross Ref
- [32] . 1977. On stress and linguistic rhythm. Ling. Inq. 8, 2 (1977), 249–336.Google Scholar
- [33] . 2014. A preliminary study on acoustic correlates of Tone2+Tone2 disyllabic word stress in Mandarin. In Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH’14). ISCA, 179–183.Google Scholar
Cross Ref
- [34] . 2021. version 9.10.0 (R2021a). Retrieved from https://in.mathworks.com/products/matlab.html.Google Scholar
- [35] . 2012. Interrater reliability: The kappa statistic. Biochem. Medic. 22, 3 (2012), 276–282.Google Scholar
Cross Ref
- [36] . 2009. The Interaction of Tone and Stress in the Prosodic System of Iquito (Zaparoan).
Technical Report . University of California, Berkeley. 1–23.DOI: Google ScholarCross Ref
- [37] . 2001. Acoustic correlates of lexical stress in Hindi. In Linguistic Structure and Language Dynamics in South Asia-Papers from the Proceedings of SALA XVIII Roundtable. Motilal Banarsidass Publishers Pvt. Ltd., Delhi, India, 123–143.Google Scholar
- [38] . 2010. Acoustic correlates of stress in Mizo, a tonal language. Lang. India 10, 4 (2010).Google Scholar
- [39] . 2013. Influence of stress pattern of native language on non-native language: A comparative study between tone and non-tone language speakers. Lang India 13, 8(Aug.2013), 239–250.Google Scholar
- [40] . 2018. Test for significance of pearson’s correlation coefficient. Int. J. Innov. Math., Statist. Energ. Polic. 6, 1 (2018), 11–23.Google Scholar
- [41] . 1996. Acoustic correlates of stress in Thai. J. Phonet. 53, 4 (1996), 200–220.
DOI: Google ScholarCross Ref
- [42] . 1994. F0 correlates of stress in Thai. Ling. Tibeto-Burm. Area 17, 2 (1994), 1–27. Retrieved from http://sealang.net/sala/archives/pdf4/potisuk1994f0.pdf.Google Scholar
- [43] . 2019. Acoustic correlates of stress in tone language: A comparison between Indian and Chinese. Lang. India 19, 1 (2019), 32–38.Google Scholar
- [44] . 2018. Robust Mizo digit recognition using data augmentation and tonal information. In Proceedings of the 9th International Conference on Speech Prosody. ISCA, 621–625.
DOI: Google ScholarCross Ref
- [45] . 2015. Detection of Mizo tones. In Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH’15). ISCA, 934–937.Google Scholar
Cross Ref
- [46] . 2010. A preliminary acoustic study of Mizo vowels and tones. J. Acoust. Societ. India 37, 3 (2010), 121–129.Google Scholar
- [47] . 2009. New effect size rules of thumb. J. Mod. Appl. Statist. Meth. 8, 2 (2009), 26.
DOI: Google ScholarCross Ref
- [48] . 2004. Speech emotion recognition combining acoustic features and linguistic information in a hybrid support vector machine-belief network architecture. In Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP’04), Vol. 1. IEEE Computer Society Press, New York, I577–I580.
DOI: Google ScholarCross Ref
- [49] . 2012. Pearson’s correlation coefficient. Brit. Med. J. 345 (2012), 1–2.
DOI: Google ScholarCross Ref
- [50] . 2020. Decision tree classification: Ranking journals using IGIDI. J. Inf. Sci. 46, 3 (
June 2020), 325–339.DOI: Google ScholarDigital Library
- [51] . 1992. On tone sandhi and tonal coarticulation. In Acta Linguistica Hafniensia, Vol. 25. Taylor and Francis Online, 83–94.
DOI: Google ScholarCross Ref
- [52] . 1996. Spectral balance as an acoustic correlate of linguistic stress. J. Acoust. Soc. Amer. 100, 4 (1996), 2471–2485.
DOI: Google ScholarCross Ref
- [53] . 2010. Spectral-tilt features of emotional speech -research on emotional-speech synthesis based on voice-quality conversion. In Proceedings of the International Conference on Kansei Engineering and Emotion Research.Google Scholar
- [54] . 2020. The acoustic correlates of stress and tone in Chácobo (Pano): A production study. J. Acoust. Societ. Amer. 147, 4 (2020), 3028–3042.
DOI: Google ScholarCross Ref
- [55] . 2021. Audacity ® – Audio Editor and Recorder. Muse Group. Retrieved from http://audacityteam.org/download/.Google Scholar
- [56] . 1975. Componential Analysis of Lushai Phonology. Vol. 2. John Benjamins Publishing.
DOI: Google ScholarCross Ref
- [57] . 2014. Tones, tonal phonology, and tone sandhi. Handb. Chin. Ling. (2014), 443–464.
DOI: Google ScholarCross Ref
- [58] . 2009. Ensemble learning. Mach. Learn. 1 (2009), 181–210.
DOI: Google ScholarCross Ref
Index Terms
A Preliminary Analysis on the Correlates of Stress and Tones in Mizo
Recommendations
Analysis and modeling of F0 contours for cantonese text-to-speech
For the generation of highly natural synthetic speech, the control of prosody is of primary importance. The fundamental frequency (F0) is one of the most important components of speech prosody. This research investigates the variation of F0 in ...
Learning Mizo Tones from F0 Contours Using 1D-CNN
Speech and ComputerAbstractThis work attempts to build an automatic 1D-CNN based tone recognizer of Mizo, an under-studied Tibeto-Burman language of North-East India. Preliminary research findings have confirmed that along with four canonical tones of Mizo (High, Low, ...
A study on tone statistics in Chinese names
This paper describes a study on tone statistics of peoples' names in Mandarin Chinese. The problem was brought out when we tried to apply an English version of a speech recognizer to a Chinese voice tag dialing task. The questions were: (1) How serious ...






Comments