The production and perception of pitch and glottalization in Yucatec Maya Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 22, 2019
  • Frazier, Melissa
    • Affiliation: College of Arts and Sciences, Department of Linguistics
  • This dissertation uses the Bidirectional Stochastic OT model of the phonetics-phonology interface to analyze the production and perception of pitch and glottalization in Yucatec Maya. The Gradual Learning Algorithm (GLA) is used to develop mean ranking values of constraints. I show that, when using this algorithm, a simulated learner must be trained on both production and perception tableaus in order to reach an accurate adult grammar (contra Boersma 2006, who proposes that perception learning alone is sufficient). This simulated learner is trained on phonetic data obtained from tokens of real speech, and these results show that bidirectional constraints can account for the symmetrical relationship between production and perception. However, because the symmetries are not exact, the production grammar does not simply fall out of perception learning. Production and perception studies were conducted with native speakers of Yucatec Maya in Yucatan, Mexico. The results of these studies are analyzed with Bidirectional Stochastic OT, but they are also presented in detail in order to document the phonetics of pitch, length, and glottalization in Yucatec Maya. One important result of the production studies is that there is previously undocumented dialectal variation in the production of pitch and length such that tone may be a dialectal feature of Yucatec Maya. Furthermore, there is variation in the perception of pitch that mirrors the variation in production; the cues that differentiate phonemic categories in production are the same cues that are attended to in perception. These results thus provide further support for the idea that production and perception grammars are defined by the same constraints. This research fills in two gaps in the literature. First, despite the robust literature on its morphosyntax, there is little research on the sound system of Yucatec Maya, especially at the phonetic level. The production study thus provides the first thorough account of the suprasegmentals of the vowel system, and the perception study is one of the first conducted with this language. Second, this work is the first to test the Bidirectional Model with actual (and not simulated and idealized) language data.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Smith, Jennifer L.
Degree granting institution
  • University of North Carolina at Chapel Hill
  • Open access

This work has no parents.