The Opposite Effects of Vowel and Onset Consonant Lengthening on Speech Segmentation

Shu-Chen Ou, Zhe-Chen Guo


This study examines the use of vowel and onset consonant lengthening in speech segmentation. It is well-documented that a longer vowel tends to be perceived as a finality cue, improving speech segmentation when occurring in the final position of a unit but leading to no facilitation when occurring the initial position. Research on domain-initial strengthening suggests that a longer syllable-onset consonant may also be a segmentation cue, but one that signals initiality. We investigated this possibility with an artificial language (AL) learning experiment with Taiwanese Southern Min listeners. The listeners first learned an AL by listening to long speech streams in which the “words” of the AL (e.g., /ba.nu.me/) were concatenated without pauses and then identified the words in a test. Higher identification accuracy indicated better segmentation during the learning. Results replicated previous findings on vowel lengthening and further demonstrated that the effects of onset consonant lengthening were opposite to those of vowel lengthening: a lengthened onset consonant improved segmentation in the initial position but resulted in no facilitation in the final one. It is assumed that lengthened vowels and onset consonants are analyzed by some prosody-computing mechanism as signaling the end and beginning of a prosodic unit, respectively.


 DOI: 10.21437/SpeechProsody.2020-16

Cite as: Ou, S., Guo, Z. (2020) The Opposite Effects of Vowel and Onset Consonant Lengthening on Speech Segmentation. Proc. 10th International Conference on Speech Prosody 2020, 76-80, DOI: 10.21437/SpeechProsody.2020-16.


@inproceedings{Ou2020,
  author={Shu-Chen Ou and Zhe-Chen Guo},
  title={{The Opposite Effects of Vowel and Onset Consonant Lengthening on Speech Segmentation}},
  year=2020,
  booktitle={Proc. 10th International Conference on Speech Prosody 2020},
  pages={76--80},
  doi={10.21437/SpeechProsody.2020-16},
  url={http://dx.doi.org/10.21437/SpeechProsody.2020-16}
}