Year of Award


Document Type


Degree Type

Master of Arts (MA)

Degree Name


Department or School/College

Linguistics Program

Committee Chair

Mizuki Miyashita

Commitee Members

Leora Bar-el, Jeanie Castillo, Heather Bliss


phonetics, phonology, connected speech, discourse, intonation unit, indigenous language


University of Montana

Subject Categories

Applied Linguistics | Discourse and Text Linguistics | Language Description and Documentation | Linguistics | Modern Languages | Morphology | Other Linguistics | Phonetics and Phonology | Typological Linguistics and Linguistic Diversity


This thesis presents a study of final vowel devoicing in Blackfoot, an indigenous language of Montana and Alberta. Previous research on final vowel devoicing in Blackfoot variously suggests word-final, phrase-final, and utterance-final vowel devoicing processes (e.g. Taylor 1965, Bliss & Gick 2009, Frantz 2017), though, the conditioning environment for this phenomenon had not been a research focus prior to this study. The present study investigates intonation units (IUs) as the conditioning domain for final vowel devoicing in Blackfoot.

Final vowel devoicing in Blackfoot is investigated here by examining the common word-final suffixes –wa (3SG.AN) and –yi (4SG) in two recordings of connected speech. Each recording features a different native speaker of Blackfoot. Speakers were asked to generate a narrative to go along with illustrations in a picture book.

These recordings are interlinearized using ELAN annotation software. Next, tokens of –wa and –yi are analyzed acoustically using Praat phonetic software. Then, –wa and –yi tokens are analyzed in terms of their position within the intonation unit (IU-medial or IU-final). Finally, the data are collated, giving the frequencies of different phonetic variants as well as the distribution of phonetic variants across IU-medial and IU-final environments.

The findings of this study are that fully-audible variants of –wa and –yi almost always occur IU-medially, while devoiced variants are most frequently found in IU-final position. Based on these findings, this thesis proposes an IU-final vowel devoicing rule to describe the phonetic variation and distribution of –wa and –yi in connected speech.

The analysis put forth in this thesis has implications for the theoretical classification of vowel devoicing phenomena, for linguistic research methodologies, and for the typology of intonation units cross-linguistically. Furthermore, the findings of this work bear on language documentation, revitalization, and pedagogy.



© Copyright 2019 Samantha Leigh Prins