聞き取りにおける予測の機構 : 音声知覚の「合成による分析」理論とその意味するもの

概要

論文の詳細を見る
The ability to predict what will follow next plays an important roie in listening comprehension. When we use our native language, we are generally unaware of this, but in listening to a foreign language, this factor looms large. We believe that the mechanism of prediction can be explained by the motor theory or, preferably, by the analysis-by-synthesis theory of speech perception. After examining these, we claim that the analysis- by-synthesis model is the more explicit one. Its main inadequacy is that a number of trials and errors may have to be repeated before the correct hypothesis for matching can be synthesized. In order to avoid this, we suggest that more than one hypothesis might be synthesized at one time whenever necessary. With this modification, we believe, this model can offer a plausible explanation of the mechanism ot prediction in listening comprehension, and, accordingly, it can give us some useful suggestions about how to train students in listening comprehension of foreign languages.
外国語教育メディア学会の論文
1978-03-31