TY - JOUR
T1 - Effects of word position and flanking vowel on the implementation of glottal stop
T2 - Evidence from Hawaiian
AU - Davidson, Lisa
N1 - Funding Information:
I would like to thank members of the NYU PEP Lab and participants at the 2020 Linguistic Society of America meeting in New Orleans and the 2020 Acoustical Society Meeting (?Acoustics Virtually Everywhere?) for their feedback on this work. I am especially grateful to Professor Larry Kimura and the invaluable voices of the ??lelo Hawai?i speakers who appeared on the Ka Leo Hawai?i radio program and are reflected in this work: Henry Hanal? Machado, Lydia L?lia Wahinemaika?i Hale, Alice Kema, Abraham Pi?ian?i?a, Alice Ku?uleialohapoina?ole N?makelua, M?lia Craver and Louis Kaulana?ula Grace. I would also like to thank Katie Drager and Andrea Berez-Kroeker at the University of Hawai?i for their guidance in how to work with the voices that have been generously made available on the recordings of the Ka Leo Hawai?i radio program. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Publisher Copyright:
© 2021 Elsevier Ltd
PY - 2021/9
Y1 - 2021/9
N2 - Much of the literature on the phonetic realization of phonemic or allophonic glottal stop has shown that it is often not produced with full glottal closure. Some studies of languages like English or German suggest that full glottal closure might be more likely in stressed syllables or positions of prosodic prominence (Garellek, 2014; Kohler, 1994), but the conditioning factors in the realization of contrastive glottal stop are less well understood. This study focuses on Hawaiian, which has phonemic glottal stop that is contrastive in both word-initial (/ʔaka/ ‘laugh’ vs. /aka/ ‘shadow’) and word-medial position (/pua/ ‘flower’ vs. /puʔa/ ‘to excrete’) (Parker Jones, 2018). Glottal stop realization is examined with respect to three factors: word position, different vs. identical flanking vowel (/puʔa/ ‘to excrete’ vs. /puʔu/ ‘hill’), and duration of the target /V(#)ʔV/ sequence. Recordings from the Ka Leo Hawaiʻi Hawaiian language radio program that aired from 1972 to 1988 were examined. Results show that the majority of phonemic glottal stops are produced as a period of creaky voice, most often in a modal voice-creaky voice-modal voice configuration, but also as modal-creaky or creaky-modal. Full glottal stops were more likely in word-initial position, and identical flanking vowels led to longer periods of creaky voice. Shorter target intervals led to longer proportions of creaky voice. These findings for phonemic glottal stop are consistent with research on the timing of contrastive voice quality in vowels, which has shown that modal-nonmodal-modal patterns are preferred to ensure that vowel quality, voice quality, and tone (for languages with all three) are recoverable (Silverman, 1995/1997). The effects of word position and flanking vowel are also related to recoverability and segmentation. The potential articulatory configurations common to glottal stops and creaky voice which may explain why they are on a continuum are also discussed.
AB - Much of the literature on the phonetic realization of phonemic or allophonic glottal stop has shown that it is often not produced with full glottal closure. Some studies of languages like English or German suggest that full glottal closure might be more likely in stressed syllables or positions of prosodic prominence (Garellek, 2014; Kohler, 1994), but the conditioning factors in the realization of contrastive glottal stop are less well understood. This study focuses on Hawaiian, which has phonemic glottal stop that is contrastive in both word-initial (/ʔaka/ ‘laugh’ vs. /aka/ ‘shadow’) and word-medial position (/pua/ ‘flower’ vs. /puʔa/ ‘to excrete’) (Parker Jones, 2018). Glottal stop realization is examined with respect to three factors: word position, different vs. identical flanking vowel (/puʔa/ ‘to excrete’ vs. /puʔu/ ‘hill’), and duration of the target /V(#)ʔV/ sequence. Recordings from the Ka Leo Hawaiʻi Hawaiian language radio program that aired from 1972 to 1988 were examined. Results show that the majority of phonemic glottal stops are produced as a period of creaky voice, most often in a modal voice-creaky voice-modal voice configuration, but also as modal-creaky or creaky-modal. Full glottal stops were more likely in word-initial position, and identical flanking vowels led to longer periods of creaky voice. Shorter target intervals led to longer proportions of creaky voice. These findings for phonemic glottal stop are consistent with research on the timing of contrastive voice quality in vowels, which has shown that modal-nonmodal-modal patterns are preferred to ensure that vowel quality, voice quality, and tone (for languages with all three) are recoverable (Silverman, 1995/1997). The effects of word position and flanking vowel are also related to recoverability and segmentation. The potential articulatory configurations common to glottal stops and creaky voice which may explain why they are on a continuum are also discussed.
KW - Creaky voice
KW - Gestural timing
KW - Glottal stops
KW - Hawaiian
KW - Word position
UR - http://www.scopus.com/inward/record.url?scp=85109166972&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85109166972&partnerID=8YFLogxK
U2 - 10.1016/j.wocn.2021.101075
DO - 10.1016/j.wocn.2021.101075
M3 - Article
AN - SCOPUS:85109166972
SN - 0095-4470
VL - 88
JO - Journal of Phonetics
JF - Journal of Phonetics
M1 - 101075
ER -