Dissertations, Theses, and Capstone Projects

Date of Degree

9-2026

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy

Program

Speech-Language-Hearing Sciences

Advisor

Douglas H. Whalen

Committee Members

Mark Tiede

Christina Hagedorn

Jason Shaw

Timothy Vance

Subject Categories

Phonetics and Phonology

Keywords

Ultrasound tongue imaging, Vowel Devoicing, Tokyo Japanese, Articulatory Phonology

Abstract

In Tokyo Japanese, high vowels /i/ and /u/ are devoiced with high frequency between two voiceless obstruents. There is an ongoing outstanding debate in the literature on whether devoiced vowels maintain their lingual articulation or are deleted. To contribute to this debate from a lingual articulatory perspective, the current study uses ultrasound imaging of the tongue and various acoustic measurements. Data from 11 native speakers of Tokyo Japanese are analyzed. This is the first lingual articulatory study to examine both devoiced /i/ and /u/ from more than three speakers. All speakers provide articulatory and acoustic evidence that devoiced /i/ and /u/ are contrastive. An articulatory contrast by vowel is observed in the preceding consonant of devoiced vowels as well. A majority of speakers consistently show an association between devoicing and tongue root retraction whether the vowel was /i/ or /u/, the effect of devoicing on lingual articulation that has not been reported elsewhere in the literature of high vowel devoicing. Inconsistent with what is known as the norm in the literature, the occurrence of devoicing is affected by whether a devoiceable mora appears in the first or second mora of a target item. In the first mora, devoicing occurs with an extremely high frequency and is independent of speech rate. In contrast, how often devoicing happens in the second mora is variable across speakers and is affected by speech rate. Based on the duration of a few acoustic intervals, we provide two potential temporal coordination among consonant and vowel gestures associated with vowel devoicing.

Share

COinS