91泡芙

Date & time: Tuesday, May 6th, 11:30am
Location: RCB 6152 & Zoom (email lingcomm@sfu.ca for the link)

91泡芙 the speaker

Dr. Ted Kye currently works at the , University of Washington, Seattle. His areas of research include phonetics, speech production, corpus linguistics, and phonetics of underdocumented languages. Learn . 

Recent publications include:

Title

Shedding new light on denasalization in Lushootseed: A phonetic analysis of voiced stops using legacy archival recordings

Abstract

Denasalization is a well-known sound change that occurred in Lushootseed and other languages of the Pacific Northwest, where nasal stops */m n/ became voiced oral stops /b d/. While previous accounts maintain that these languages lost nasality entirely (Hockett 1955, Thompson & Thompson 1972, Campbell 1997), there has not been an account of whether this loss could be observed at the phonetic level. In this study, I examine whether residual nasality is present in the production of voiced stops in Lushootseed, as well as address what implications the phonetic findings poses in our understanding of denasalization in Lushootseed and other neighboring languages in the Pacific Northwest. A female native elder speaker was examined from several archival recordings dating to the 1950鈥檚. I conduct a qualitative and quantitative acoustic analysis on the voiced stops /b d/ by listening for audible nasal murmur, inspecting spectrograms and waveforms, as well as measuring the amplitude of F1 (A1) and the extra peak above F1 (P1, which is near 1kHz) that corresponds to energy introduced from nasal airflow.

The findings reveal that residual nasality was present in voiced stops in the form of partial (weak) nasality. Moreover, partial nasality occurred exclusively in prosodic domain-initial position (i.e., initializing a stressed syllable and preceded by a pause) and never occurred when they initialized an unstressed syllable or occurred intervocalically. A1 for the partially nasalized variants [岬恇 鈦縟] was greater than the plain voiced stop variants [b d]. P1 was significantly lower in [岬恇 鈦縟] than full nasal stops [m n] (where [m n] was observed when the speaker was speaking Sahaptin in one of the recordings), which suggests that the amount of nasal airflow was considerably weak in the production of [岬恇 鈦縟]. This suggests that the speaker partially nasalized her voiced stops to enhance the voicing contrast of the stops, making it easier to hear and produce voicing during the constricted interval.

These findings provide several important implications on our understanding of denasalization in Lushootseed. It is possible for denasalization to have occurred diachronically due to strengthening of nasals into voiced stops in prosodically strong boundaries, and that partial nasality was preserved within those boundaries in order to enhance the voicing contrast of the stops /b d/. Moreover, languages within the linguistic area (i.e., languages neighboring Lushootseed and other languages that have undergone denasalization) are susceptible to strengthening their sonorants into obstruents, making it feasible for some of those languages to undergo denasalization.