Production of word-medial voicing contrasts in Japanese plosives:Comparing native Japanese speakers and Chinese learners

This study investigates the acoustic realization of word-medial voicing contrasts of Japanese plosives, comparing the production of native Japanese speakers with those of native Chinese speakers learning advanced level Japanese. Ten speakers of each group participated in a recording experiment. There were clear differences in patterns between the two groups. The Japanese speakers exhibited a much higher voicing ratio during the closure of voiced plosives than the Chinese speakers. In cases where voice plosives lacked pre-voicing, the Chinese speakers showed a larger between-category difference in aspiration than the Japanese speakers. Also, while the Japanese speakers consistently showed a shorter closure duration for voiced plosives than for voiceless plosives, more than half of the Chinese subjects showed either no difference or the opposite pattern. Consonant weakening in voiced plosives, characterized by a weak or absent release burst, was frequently observed in the Japanese speakers’ data, but much less frequently in the Chinese speakers’ data. Also, the Japanese speakers exhibited a significant between-category difference in burst energy, whereas the Chinese learners showed only minimal contrast. These results help to explain the difficulties that Chinese speakers face with Japanese word-medial plosives, and they may also have implications for improving their perception and production.