Acquisition of Yo-on (Japanese contracted sounds) in L1 and L2 phonology

This study observes the prosodic constraints that govern the pattern of simplification of Japanese contracted sounds in the acquisition processes of first (L1) and second language (L2) learners: Japanese children and English speaking learners of Japanese. Japanese language has a group of palatalized consonants, known to scholars of Japanese language as the 'contracted' sounds, [CjV]. In accordance with the prosodic constraints of Japanese, which is mora-timed, a CjV mora has to be the same length as a CV mora. However, in addition to typical moraic timing errors, learners of Japanese whose L1 is English appear to treat Japanese contracted sounds initially as consonant + glide clusters, where there is an equivalent [Cj] cluster in English, or otherwise tend to insert an epenthetic vowel [CijV]. The phonological status of the palatal glide [j] is still controversial in both English and Japanese. The acquisition process of the Japanese "CjV" unit by English speaking learners could offer the clue to clarify this issue. Given the similarity between palatalized consonants and consonant clusters in their phonetic status, the acquisition literature of consonant clusters is considered initially. In general, L1 learners at an early stage of acquisition employ a reduction strategy to cope with the production difficulties associated with the clusters. Sonority appears to be the causal factor referred to most frequently. If the sonority hypothesis governs the reduction pattern, syllable initial clusters reduce to whichever consonant in the cluster creates a maximal sonority rise, and syllable final clusters reduce to whichever consonant in the cluster creates a minimal sonority descent; e.g. Onsets "pl" ? "p"? Codas "lp" ? "l"(Ohala,1999). The former would possibly be the case if the "CjV" unit in Japanese were treated as a consonant cluster. The results of this study show a contrastive reduction pattern between L1 and L2 learners. Due to the moraic timing constraint, Japanese children would not epenthesize vowels in producing contracted sounds, while L2 learners violated moraic timing by lengthening a vowel or inserting gemination as well as an epenthetic vowel. The reduced consonant also differed between the two groups, and the result shed light on the clarification of the phonological status of /Cj/ sequences in Japanese and English. The findings of the study also point to Japanese children's early acquisition of palato-alveolars, which accounts for the prevalence of the recognition that children are good at producing the contracted sounds.