Japanese allows for almost no consonants in syllable codas. In loanwords, illegal codas are transformed into onsets by means of vowel epenthesis. The default epenthetic vowel in loanwords is [ɯ], and previous work has shown that this [ɯ]-epenthesis reflects Japanese listeners’ perception of illegal coda consonants. Here, we focus on one of the non-default cases: following coda [ç] and [x] the epenthetic vowel is a copy of the preceding vowel. Using an identification and a discrimination task, we provide evidence for the perceptual origin of this copy vowel phenomenon: After [ç] and [x], Japanese listeners perceive more often an epenthetic copy vowel than the default vowel [ɯ], whereas after [k] it is the reverse.
Cite as: Guevara-Rukoz, A., Yu, S., Peperkamp, S. (2021) Speech Perception and Loanword Adaptations: The Case of Copy-Vowel Epenthesis. Proc. Interspeech 2021, 4004-4008, doi: 10.21437/Interspeech.2021-1481
@inproceedings{guevararukoz21_interspeech,
author={Adriana Guevara-Rukoz and Shi Yu and Sharon Peperkamp},
title={{Speech Perception and Loanword Adaptations: The Case of Copy-Vowel Epenthesis}},
year=2021,
booktitle={Proc. Interspeech 2021},
pages={4004--4008},
doi={10.21437/Interspeech.2021-1481}
}