Prosodic Features of English Yes-No Questions Produced by Japanese Speakers

This paper examined the characteristics of the English rising tone in yes-no questions produced by native speakers of Japanese according to the different levels of the intelligibility of intonation. Five native English speakers evaluated the intelligibility for three kinds of yes-no questions read by 10 Japanese speakers: the questions were those with the nucleus placed in the sentenceinitial, sentence-medial, or sentence-final position. The F0 and the duration of the Japanese speakers’ (JS) productions were measured and compared with those of native English speakers’ (NS) ones to elucidate acoustic features. It was found that the intelligibility was lower in the JS productions with the sentence-initial and the sentence-medial nucleus. The acoustic analysis showed that the overall range of the major pitch change was smaller in the JS productions than in the NS ones. The intelligibility was lower in the JS productions with the sentence-initial or sentencemedial nucleus when the magnitude was smaller, and it tended to be higher in controlled JS productions with the sentence-final nucleus when the nucleus had longer duration and was correctly identified. Lower intelligibility was also attributed to the wrong placement of the nucleus and the inappropriate use of nuclear tones. The rise in pitch on the sentence-final word suggested both negative and positive L1 transfer in JS productions.