Syllabification of intervocalic consonants by English and Japanese speakers

This study investigated how English and Japanese speakers syllabify two-syllable English words and nonwords with single intervocalic consonants. In oral tasks, participants inserted a pause between two parts of English words and nonwords. In written tasks, they selected their preferred syllabification of English words between two alternatives. In Experiment I both English and Japanese speakers generally tended to place intervocalic consonants in the second syllables, following the maximum onset principle. Japanese speakers showed a stronger preference for this type of segmentation than English speakers. Responses by speakers of both languages were affected by the position of stress, the length of the stressed vowel, and the sonority of the intervocalic consonant, reflecting the universal effects of these factors on syllabification. Japanese speakers produced a relatively large number of CVCV/C (C=consonant, V=vowel) responses (e.g., sala/d) which were also influenced by vowel length and consonant type. Experiment 2 tested the Japanese speakers who had been trained to recognize English syllables. They showed more ambisyllabic responses and fewer CVCV/C responses than the untrained participants, suggesting their tendency toward syllabic segmentation, rather than moraic segmentation. The results are discussed in light of linguistic and psycholinguistic theories of syllabification.