Some relation is considered to exist between voice raising or lowering and the articulation of vowels. Because there exist not a few dialect in Japanese whose accent have some restrictions according to vowel hight. The common characteristics of those dialects are that low vowels seldom follow the accented high vowels.
The present paper reports the results of the investigation about the timing relation between fundamental frequency change and vowel hight, and also about physiological mechanism of /kusa/ or /kusa/ with low vowels following accented high vowels.
The subjects of the experiment were selected from native speakers of the Kinki dialect which is free from the above restrictions. The words mainly studied here were, (1)/ai/, /ei/, /ie/.(2)/aCa/, /aCi/, /iCa/, /iCi/(C: voiced consonant), with three accent types A (ao), B (of) and C (oo).(3)/kusa/, /kuse/, and /kusi/.
Narrow and wide band spectrograms were obtained from those speech materials, and onsets of Fo change and those of formant transition were measured and their timing relations were compared.
The results were as follows:
(1) 1./ai/ and /ei/: the onset times of voice raising and lowering (P_t) lag behind the onset times of formant transition (F_t) as was reported before in regard to /ai/.
2./ie/ was found to have the tendency of synchronizing P_twith F_t, while the timing of P_tin /ie/ and /ie/ precede that of F_t (Fig. 1, 2, 3, and Table 1).
(2) Similar results were observed in words containing /VCV/ sequences (Fig. 4, 5, 6, 7). The results sugest that phonatory controls have something to do with the articuratory controls.
(3) Fo contours of [a] in /kusa/ start high and show a sharp descent, while those of [i] in /kusi/ start comparatively low and level (Fig. 10, 12).
The second vowel of /kusa/ has more sharp descent.
As the results of electromyographical study, it is known that there is a positive correlation between the pattern of cricothyroid (CT) activity and the raising and lowering of F₀. It was revealed that the activity of sternohyoid (SH) was related to the sharp descent in the Fo contours of type A and type B in the Kinki dialect (Fig. 8, 9). SH is also known to participate in jaw opening, tongue lowering and tongue retraction.
Synchronizing the timing of voice lowering in F0 and toungue lowering in words containing high and low vowel sequences is considered to be related with activity of SH. Accurate tongue lowering for articulation and sharp decent in F₀may be needed in the pronounciation of words like /kusa/ or/kusa/(Fig. 13, 14). It is considered to have relationship with the fact that low vowels seldom folow the accented high morae in some dialects.

View full abstract

From the phonemic point of view, the author attempts to review the reconstructions of Archaic Chinese sound system by B. Karlgren, Tung T'ungho. He modifies their reconstructions and presents his phonemic interpretation of the initials and the finals in Archaic Chinese. To summarize the conculusion of his interpretation, the phonemic system of Archaic Chinese is as follows.
1) Initial consonants:
p-p′-b-m-
t-t′-d-n-l-ç-j-
ts-ts′-dz-s-z-
k-k′-g-η-x-γ-
2) Medials:
-φ-(zero)-r-u-
3) Vowels:
(i)(u)
e e A
a a
4) Final consonants:
-m-p-b
-n-t-d -r
-D-k-g
-uη-uk-ug

View full abstract