Audio-visual Analyses of Differences Between Natural and Teaching Styles for Mandarin Tone Production

Yueqiao Han, Martijn Goudbeek, Maria Mos, Marc Swerts


In order to examine the acoustic and visual information that is used by speakers to facilitate the perception of Mandarin tones, 4 Chinese native speakers were video-taped and asked to produce Mandarin tones/words in natural and teaching speaking styles as the experimental materials for a tone identification task. Acoustic and visual analyses of the produced materials were conducted to evaluate the characteristics of the tones/words produced in both speaking styles. Acoustic results showed that as compared to natural style, speakers in teaching style produce Mandarin tones in a more “exaggerated” way, represented mainly by prolonging the pronunciation of the tones. Visual analyses revealed that speakers in teaching mode signaled more visual information/facial motions than in the natural style. Furthermore, the types of facial motions (horizontal and vertical movement) displayed are associated with tone variations.


 DOI: 10.21437/SpeechProsody.2018-148

Cite as: Han, Y., Goudbeek, M., Mos, M., Swerts, M. (2018) Audio-visual Analyses of Differences Between Natural and Teaching Styles for Mandarin Tone Production. Proc. 9th International Conference on Speech Prosody 2018, 729-733, DOI: 10.21437/SpeechProsody.2018-148.


@inproceedings{Han2018,
  author={Yueqiao Han and Martijn Goudbeek and Maria Mos and Marc Swerts},
  title={Audio-visual Analyses of Differences Between Natural and Teaching Styles for Mandarin Tone Production},
  year=2018,
  booktitle={Proc. 9th International Conference on Speech Prosody 2018},
  pages={729--733},
  doi={10.21437/SpeechProsody.2018-148},
  url={http://dx.doi.org/10.21437/SpeechProsody.2018-148}
}