The Effect Of Loanword Status In Japanese On The Length Of The Vowel Inserted Between Consonant Clusters In L2 English By L1 Japanese Speakers Of English

In L2 phonological acquisition, the difference in phonotactics between two languages might result in misperception and non-target like pronunciation (Dupoux et al., 1999; Nogita & Fan, 2012). If a language does not allow consonant clusters (CC), the language speaker may perceive an illusory vowel between the two consonants in the cluster (e.g., Japanese, Dupoux et al., 1999) and also they may insert a vowel between two consonants (i.e., vowel epenthesis). Similarly, the pronunciation of loanwords when adopted is usually modified according to the phonological system of the recipient language (Kay, 1995). The current study investigates the effect of loanwords in Japanese on the length of the vowel inserted between CC which are illegal in Japanese phonotactics by comparing the pronunciation of loanwords and non-loanwords (henceforth, loanwords refer to English words which exist in Japanese as loanwords, whereas non-loanwords are the ones which do not). For example, trumpet is a loanword because it is widely used in Japanese as a loanword but pronounced as torampetto [to.ɾa.m.pe.t.to] whereas trash is a non-loanword because it is not used in Japanese as a loanword, rather gomi, the Japanese counterpart of trash, is used. The status of loanwords and non-loanwords seems to be different in learners’ mental lexicon (Nomura & Ishikawa, 2018). The potentially resulting differences in the length of inserted vowels will, therefore, be discussed in terms of the differences in the mental representations of loanwords and non-loanwords in a Japanese learner’s mental lexicon. In order to determine L1 influence on these potential differences, the participants will be L1 Japanese/L2 English learners (JEL) and L1 Mandarin Chinese/L2 English learners (MEL). Two tasks are employed: a picture-naming task and a reading-aloud task. The results revealed that the status of loanwords in the mental lexicon of L1 Japanese speakers is different from that of non-loanwords, which caused a longer duration of vowel insertion. These results would suggest a pedagogical implication that loanwords should be treated in pronunciation teaching differently