Korean listeners’ perception of consonant clusters in English.
Gwanhi Yun
Abstract
Gwanhi Yun
Abstract
Perception experiments were conducted to see whether Korean listeners perceive the illusionary vowel in the legitimate sequence of English consonants, especially in onset positions. Korean listeners were presented three types of licit consonant clusters in English (e.g., sC in spee; Cy in kyoo; obstruent + approximant in plee) and had to choose between the original sequence and vowel-repaired sequence. First, Korean listeners did not perceive the illusionary inserted vowel between two consonants with all three types of onset clusters (sC, Cy, OA). Perception of the insertion of unmarked vowel was obtained only 20.1 %, while consonant clusters (disallowed in Korean) were not repaired in perception. Second, perceptual repair with illusive vowels was adopted the most in the sequence of obstruents and approximants (pree; 36 %), next of s+consonant (spee; 20%), and the least of Cy sequences (pyoo; 15%). Our result indicates that loanword phonology might be primarily computed on the basis of the phonological grammar of the borrowing language rather than perceptual assimilation based on phonetic-fine representation [Peperkamp et al. (2008)] . Further, it shows that perceptual repair by illusory vowels in the illicit onset clusters are gradual, not categorical unlike phonetic implementation of loanword phonology.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Perception experiments were conducted to see whether Korean listeners perceive the illusionary vowel in the legitimate sequence of English consonants, especially in onset positions. Korean listeners were presented three types of licit consonant clusters in English (e.g., sC in spee; Cy in kyoo; obstruent + approximant in plee) and had to choose between the original sequence and vowel-repaired sequence. First, Korean listeners did not perceive the illusionary inserted vowel between two consonants with all three types of onset clusters (sC, Cy, OA). Perception of the insertion of unmarked vowel was obtained only 20.1 %, while consonant clusters (disallowed in Korean) were not repaired in perception. Second, perceptual repair with illusive vowels was adopted the most in the sequence of obstruents and approximants (pree; 36 %), next of s+consonant (spee; 20%), and the least of Cy sequences (pyoo; 15%). Our result indicates that loanword phonology might be primarily computed on the basis of the phonological grammar of the borrowing language rather than perceptual assimilation based on phonetic-fine representation [Peperkamp et al. (2008)] . Further, it shows that perceptual repair by illusory vowels in the illicit onset clusters are gradual, not categorical unlike phonetic implementation of loanword phonology.
Key concepts: Obstruent, Vowel, Consonant, Linguistics, Perception, Psychology, Phonology, Categorical perception