聴能訓練器の特性に関する研究(1) : 語音の物理的要素とイントネーションの知覚

抄録

This study was intended to show what acoustic componeneuts were used at percieving a intonation of a speech, and to discuss how they are used in the acquisition of language. The components treated in this study were an amplitude and a pitch of a speech, as a linguistic signal. I thought that they made a intonation of a speech and an accent position was a mathematic pole of the intonation. Identified the ability of a listening of the accent position was a mathematic pole of the intonation. I defined the ability of a listening of the accent position was that of the intonation. I examined whether ten subjects could listen an accant position of words and sentences with a frequency distortion or an amplitude distortion. A frequency distortion was made through a 425Hz low-pass filter, and an amplitude distortion by a original AGC circuit. Tts rise time of the operation was less 1 msec, and its fall time of it was approximately 0.1 sec. It was examined whether an accent position of eleven words and ten sentence were listened constantly in twice by each subject, and whether those with a frequency distortion, and an amplitude distortin, were listened identicaly them with no distortion (original sounds). This result was as follow: [table] Non identical numbers of an accent position of words and sentences between listenings of a subject in twice. This table indicated that an accent was listened by a pitch of a speech We could know that an accent position was listened in pitch more than sound pressure, and the speech trainer could have a low frequency a AVC circuit but not cut all over a gain at (100-400 Hz) which was not useful for the recognition of the phoneme. I thought that some impaired hearing children could not speak in grammaticaly because some hearing aids were made with a nonlinear distortion at low frequency, referring Lievermnn's book and Lenneberg's paper.