Overview

It is well known that Japanese speakers have difficulty in differentiating the liquid consonants /l/ and /r/. This is because /l/ and /r/ are not contrastive in Japanese, and allophonic variations of both /l/ and /r/ occur in Japanese speech. The most common realisation is alveolar tap [ɾ], but [l] also occurs in natural speech. However, these variants are phonemically all recognised as /r/ in Japanese. A study of Japanese speakers’ English pronunciation errors using a large English learner corpus, J-AESOP, found that Japanese speakers produced more mistakes in /l/ than /r/. The Japanese speakers substituted /r/ for /l/ (418 examples out of 2,142 consonantal errors) much more often than they substituted /l/ for /r/ (124 examples out of 2,142 consonantal errors). Research is being conducted to examine the acquisition of English liquid consonants by Japanese speakers in relation to the concept of new phonetic categories proposed by the Speech Learning Model. Furthermore, it is important to assess how L2 learners use a new phonetic category when they face a new variant of a phoneme, produced in a different accent to that which they had already studied in the target language. Previous analysis of Japanese speakers’ mimicry speech of (a) American English and (b) English-accented Japanese suggested that Japanese speakers were aware of acoustic and articulatory features of English approximant [ɹ] (Kondo 2016). The Japanese speakers overused approximant [ɹ] and r-coloured vowels in their mimicries of both (a) and (b). Further articulatory analysis of Japanese and English consonants showed that the English approximant [ɹ] is quite distinct from Japanese consonants, all of which lack lip rounding and tongue retraction. The results of these studies suggest that Japanese speakers may not be able to recognise English /l/ and /r/ as separate phonemes, but that they can hear the approximant [ɹ] as it forms a new phonetic category, i.e. /ɹ/. In contrast, they recognise English /l/ as a sound in the same category as Japanese /r/. Based on these earlier data, the current chapter assesses the effectiveness of pronunciation training for Japanese students. We discuss the abilities of ninety trained and ninety untrained Japanese speakers to form a new phonetic category of an approximant [ɹ].