While I was pursuing a PhD in the Department of Linguistics at the University of Hawaiʻi at Mānoa, I studied ʻōlelo Hawaiʻi at Kawaihuelani Center for Hawaiian Language, part of the Hawaiʻinuiākea School of Hawaiian Knowledge. I realized that there had been very little linguistic research on Hawaiian pronunciation and decided to focus my doctoral research on the phonetic properties of the language’s vowels. I have been measuring the acoustic properties of elders whose voices were recorded in the 1970s-1980s on Ka Leo Hawaiʻi, a Hawaiian language radio show. Some outputs of this research have included:
[abstract] [poster] Kettig, Thomas, and Lisa Davidson. 2024. Acoustic correlates of stress in contrastive short and long vowels in ʻōlelo Hawaiʻi. LabPhon19. Seoul, Korea, 27–29 June.
[slides] Kettig, Thomas. 2023. Inter- and intra-speaker /ai~ei/ variation in ʻōlelo Hawaiʻi. Acoustic and historical-comparative evidence. NWAV 51, New York, NY, 14 October.
[abstract] [slides] Kettig, Thomas. 2023. Word class and frequency effects in Hawaiian stressed vowel clusters. International Congress of Phonetic Sciences 2023. Prague, Czechia, 9 August.
[slides] Kettig, Thomas. 2022. Word class and diphthong reduction in ʻōlelo Hawaiʻi. NWAV 50. San Jose, CA, USA, 14 October.
[video] Kettig, Thomas. 2021. Nānā i ke kumu: Acoustic phonetic research on archival recordings of ʻōlelo Hawaiʻi. International Conference on Language Documentation and Conservation 7. Online.
[paper] Kettig, Thomas. 2021. Haʻina ʻia mai ana ka puana: The vowels of ʻōlelo Hawaiʻi. PhD Dissertation, University of Hawaiʻi at Mānoa.
[abstract] [poster] Kettig, Thomas. 2019. Spectral coarticulation in Hawaiian /aV/ and /aCV/ sequences. International Congress of Phonetic Sciences 2019. Melbourne, Australia, 8 August. (Poster)
I conducted postdoctoral research at the University of York on Vincent Hughes’ and Carmen Llamas’ AHRC-funded project Humans and Machines: Novel Methods for Assessing Speaker Recognition Performance. We developed a bespoke computer game that elicits judgments of voice similarity in pairs of voices with various UK regional accents. We wanted to explore what factors affect sameness judgments in humans (for instance, jurors in trials involving voice-matching evidence) and we wanted to compare these judgments to the outputs of automatic speech recognition systems.
[abstract] Hughes, Vincent, Carmen Llamas, and Thomas Kettig. 2022. Eliciting and evaluating likelihood ratios for speaker recognition by human listeners under forensically realistic channel-mismatched conditions. Proceedings of Interspeech. Incheon, Korea.
[abstract] Hughes, Vincent, Carmen Llamas, and Thomas Kettig. 2022. A game-based approach to eliciting and evaluating likelihood ratios for speaker recognition. 30th Annual Conference of the International Association for Forensic Phonetics and Acoustics. Prague, Czechia, 10-13 July.
I’ve done various work on the Canadian Shift in Montreal.
[slides] Kettig, Thomas and Katie Drager. 2018. The social evaluation of TRAP-backing in Montreal. Sociolinguistics Symposium 22. Auckland, New Zealand, 30 June.
[paper] Kettig, Thomas and Bodo Winter. 2017. “Producing and Perceiving the Canadian Vowel Shift: Evidence from a Montreal Community.” Language Variation and Change 29(1): 79–100. DOI:https://doi.org/10.1017/S0954394517000023
[poster] Kettig, Thomas. 2017. “Diachronically stable, lexically specific variation: The phonological representation of secondary /æ/-lengthening”. Phonetics and Phonology in Europe. Cologne, Germany, 12-14 June. (Poster)
[poster] [extended abstract] Kettig, Thomas. 2016. “The BAD-LAD split: Secondary /æ/-lengthening in Southern Standard British English” (Poster). 90th Annual Meeting of the Linguistic Society of America. Washington, DC, 8 January. DOI: http://dx.doi.org/10.3765/plsa.v1i0.3732.
[paper] Kettig, Thomas. The BAD-LAD Split: A Phonetic Investigation. Masters Thesis, University of Cambridge.
I have a soft spot for vowel shifts.
