Statistical modelling of phonetic and phonologised perturbation effects in tonal and non-tonal languages

Si Chen, Caicai Zhang, Adam G. McCollum, Ratree Wayland

Research output: Journal article publicationJournal articleAcademic researchpeer-review

11 Citations (Scopus)


This study statistically models perturbation effects of consonants on f0 values of the following vowel in order to quantify the differences between phonetic perturbation effects (i.e., phonetic variation) and phonologised perturbation effects (i.e., tone distinctions). We investigated perturbation effects in a non-tonal language, Japanese and a tonal language, Chongming Chinese. Traditional methods of modelling cannot distinguish phonetic and phonologised effects on surface f0 contours, as variation caused by both effects reached statistical significance. We therefore statistically modelled and tested the differences in underlying pitch targets, which successfully distinguished between phonetic and phonologised effects, and is robust to data variability. The methods used in this study can be further applied to examine perturbation effects cross-linguistically and shed light on the development of tones and stages of phonologisation more broadly.
Original languageEnglish
Pages (from-to)17-38
Number of pages22
JournalSpeech Communication
Publication statusPublished - 1 Apr 2017


  • Functional data analysis
  • Growth curve analysis
  • Perturbation
  • Phonologisation
  • Underlying pitch targets

ASJC Scopus subject areas

  • Software
  • Modelling and Simulation
  • Communication
  • Language and Linguistics
  • Linguistics and Language
  • Computer Vision and Pattern Recognition
  • Computer Science Applications


Dive into the research topics of 'Statistical modelling of phonetic and phonologised perturbation effects in tonal and non-tonal languages'. Together they form a unique fingerprint.

Cite this