| Assuming the Perceptual Assimilation Model developed by Peperkamp (2005) and the Principle of Phonological Perception put forth by Iverson & Lee (2006), this dissertation investigates (i) the perceptual basis for the varying quality of epenthetic vowels in loanwords, (ii) the role of recipient language phonology in determining the quality of the inserted vowel during speech perception, and (iii) the influence of non-perceptual factors in determining vowel quality.;The point of departure here is that inserted vowels in loanwords are "perceptually intrusive" in that they fill the perceptual gap between two consonants (or gestures) clustered in a given phonetic context. Experiments with Korean speakers show that English ;This study also highlights the influence of recipient language phonology in determining the quality of an inserted vowel during perceptual decoding of the source acoustic signal. Despite the striking similarities between Korean and Japanese with respect to [s] in the native phonology (e.g., [s] is the manifestation of /s/ before /i/ or /y/ in both languages), English word-final [s] is rendered with inserted /i/ in Korean (rather than default /barred i/) but with default /[turned m]/ in Japanese, not /i/. It is argued that such asymmetry in the choice of inserted vowel between the two languages correlates directly with the phonemic status of the sound [s] in Korean and Japanese: [s] is an allophone of /s/ in Korean but represents a separate phoneme in Japanese contrasting with /s/. |
