By use of a measuring apparatus of speech articulation utilizing in our department, the mishearing phenomenon of thirty mono-syllables in Auditorially Balanced (AB-) List of Japanese language proposed by Prof. Kawata was observed simultaneously with the frequency analysis by a visible frequency analysator (TS-6 type), the recording of this procedure being made by 8 mm cinematography.
Furthermore, the interrelation between the mishearing and the frequency analysis of monosyllables was investigated by successive tracing of the temporal change of speech spectrum.
(1) Even in normal ears, the mishearing can occur in the particular speech sounds on the same vowel elemtal line. From a fact that, frequency-analytically, the speech sounds on the same vowel elemental line, of which back vowels are common, has the similar intensity distribution of a chief frequency area and the tone which gives the property characterizing the speech is the front consonant, it is observed that the mishearing results chiefly from the similarity of distortion of a front consonant in amplitude, frequency or duration.
(2) In the examination of deafness, the pure tone hearing curve was classified into eight types and the pure tone hearing loss was observed in five stages such as 0-20, 21-40, 41-60, 61-80 and above 81 db. And the speech hearing curve was described by the classification of Tateishi.
(3) The mishearing in deafness, as in normal ears, frequently occurs in the particular speech sounds on the same vowel elemental. line, especially occurs in the speech sounds containing a Formant distorted in frequency, observing from the pure tone hearing curve. The hearing loss has no direct relationship with the mishearing; namely when the intensity of tone is increased there occurs no change in mishearing. The tendency of mis-hearing is common in various kinds of deafness.
(4) In both intact and deaf ears, the mishearing occurs most frequently in Mi, Me, Nu, Bu, Be, Pe, in which “Mi” and “Me” or “Ma” column and “Na” column have frequency-analytically very similar Formant and shows a very slight difference due to the time shift of development of frequency area. From these facts, the increase in the ratio of mis-hearing can be explained. Since the voiced consonants tend to become noise because of the vibration of vocal cords by phonation the mis-hearing of a sonant is frequently observed in “Ra” column, being a voiced consonant. This fact is presumably due to the semi-vowel character of speech sounds on “Ra” column.(author's abstract)
抄録全体を表示