Identification of vowels of two different varieties of English by native speakers of Japanese and Korean

Native speakers of Japanese and Korean heard and identified /i, ɪ, eɪ, ɛ, æ, ɑ, ʌ/ uttered in /bVd/, /dVd/ and /kVd/ frames by native speakers of American and New Zealand English. New Zealand English has gone through an idiosyncratic vowel shift. For instance, /æ/ and /ɛ/ are raised and /ɪ/ is centralized. Overall American English vowels are identified more accurately by the two listener groups. Both listener groups identified American English /ɛ, æ/ better than New Zealand English equivalents, but on the contrary American English /ɑ/ is less accurately identified than New Zealand English /ɑ/ (or /ɒ/). Despite these similarities, some differences are observed between the two listener groups. While Japanese listeners identified New Zealand English /i, ɪ,/ less accurately than American English /i, ɪ/, Korean listeners identified New Zealand English /ɪ/ more accurately than American English /ɪ/. Japanese listeners outperformed Korean listeners in identifying American English /eɪ/ and New Zealand English /ɑ,/, but Korean listeners identified American English /ʌ/ and New Zealand English /ɪ/, /æ/ and /ʌ/ better than did Japanese listeners. The results point to the effect of L1 phonology and the differences in what each listener group believe each English vowel sounds like. For instance, Korean non-high front vowels are acoustically more similar to New Zealand English vowels /æ/ than to American English corresponding vowel. [Work partially supported by Grant-in-Aid for Scientific Research (C)16K02650.]