紀要論文

紀要論文

text

言語文化研究. 2013, 39, p. 221-244

日本語

text

2013-03-31

The first part of the paper describes the acoustic characteristics of neutral yes-no questions (i.e., without assumptions regarding whether the answer is affi rmative or negative) that are composed of a single verb. The verb has a straight or concave-up pitch rise that begins in the final mora. The final mora is lengthened by 40-100 ms (depending on the verb's lexical accent) when compared to the final mora in a neutral statement. If the verb has an accentual pitch fall, as in No'nda?‘ Have you drunk (taken) it?’, the final rise in da concludes at the level of the maximum pitch in No'n. The onset of the accentual fall is delayed by an average of 7 ms. The portion that precedes the interrogative rise is pronounced at a faster tempo than it is in a neutral statement (10% reduction in duration) and is pronounced with a wider pitch range. In the second part of the paper, the acoustic characteristics of four types of questions, i.e., (Type 1) questions with a high degree of necessity for answers, (Type 2) doubtful questions, (Type 3) questions with positive assumptions, and (Type 4) questions with negative assumptions, are investigated through the analysis of laboratory speech and perception tests using synthetic speech. The results indicate that Type 1 questions have the widest pitch range and the slowest tempo in the portion that precedes the fi nal rise. A large amount of final rise is required to be perceived as Type 2 and 3 questions.

10.18910/24700

application/pdf

https://doi.org/10.18910/24700

日本語

2013-03-31

39

221

244

https://hdl.handle.net/11094/24700

10.18910/24700