Phoneme Duration Sensitivity in L2 Adaptation in Japanese –Phonological Parsing Rules in Different Phonological Units

This study investigated Japanese speakers’ second language (L2) phonological adaptation processes by examining the production of L2 with particular reference to speech rhythm and timing. In addition it examined the effectiveness of proposed L2 phonological adaptation rules in different prosodic units, namely word and phrase. Languages use different prosodic units to maintain their speech rhythm; for example, Japanese is mora-timed (Otake et al., 1993; Cutler & Otake, 1994) whereas French is syllable-timed (Mehler et al., 1981). Loanword adaptation reflects phonological parsing of foreign sound sequences, and L2 perception and production often reveal the true nature of the first language. Shinohara (1997 & 2004) analyzed Japanese loanword adaptation processes phonologically and proposed four main rules regarding speech rhythm and timing: (i) phonemic lengthening of L2 vowels by phonetically long vowels, e.g. Pete [pi:t] > /piito/ and cool [ku:l] > /kuuru/, (ii) orthographic effects, e.g. au, ou, eau, â, ô, û are often treated long, (iii) ‘pre-final lengthening’ (i.e. adding an extra mora by either consonant gemination or lengthening of preceding vowel) of word-final consonants, e.g. pit |pɪt#| > /pit.to/ and picnic |pɪk.nɪk#| > /pi.ku.nik.ku/ but not */pikkunikku/, and (iv) syllabification of L2 stray consonants in order to preserve Japanese syllable structure (C)V(C/V), e.g. cream > /ku.rii.mu/, print > /pu.riN.to/. These rules add mora count in Japanese adapted forms, and consequently word duration is increased by the number of added morae. In this study we will first review studies of speech rhythms and phonological rules (Section 2). Then in sections 3, 4 and 5 we will report the results of our experimental study where Japanese speakers’ French utterances were examined phonetically based on the loanword adaptation rules proposed by Shinohara, and compared against utterances of the same phrases by native French speakers. In the study we measured durations of 6 pairs of French phrases, where each pair had the same syllable number but different mora counts in the Japanese adapted forms. In addition we also measured durations of the individual syllables which lead to the different moraic interpretations. Based on the proposed loanword adaptation rules, the durations of the Japanese speakers’ pairs of phrases should differ by the number of morae in the phrase, whereas the French speakers’ phrase durations would be based on the syllable count. We tested whether all the rules for the word level are also applied at the phrase level.