Laryngeal Muscle Activity in Thai Tones

Donna Erickson

The primary acoustic realization of the tones in Standard Thai is the fundamental frequency (F0), which maintains a typical, relative contour for each tone with reference to the speaker's voice range. The F0 contours for tones are determined by variation in the rate of vocal fold vibration, which is to a large part controlled by tension of the vocal folds regulated by laryngeal muscles.

This paper, based on the author's doctoral dissertation (Erickson, 1976; Erickson, 1993) discusses the laryngeal muscle activity associated with each of the tones, an articulatory description of tones in terms of laryngeal articulation features, and laryngeal control mechanisms for F0.

Acoustic and electromyograpic recordings were made and processed at the Haskins Laboratories Research Laboratories. Hooked wire electrodes were inserted by Dr. Hajime Hirose, of Tokyo University, RILP, (Hirose, 1971). The laryngeal muscles examined were the cricothyroid (CT), vocalis, and the three strap muscles, the sternohyoid (SH), the sternothyroid (ST), and the thyrohyoid (TH). EMG data were analyzed using the Haskins Laboratories computerized EMG processing system, (Kewley-Port, 1973).

The speakers were 4 native speakers of standard Thai, 3 men and 1 woman, students at the University of Connecticut.

The samples examined were nine syllable types, made up of one of the three vowels, /a, i, u/, one of the three initial consonants, /b, p, ph/, and one of the five tones, yielding a total of 45 utterance tokens. The syllables were spoken in a carrier phrase. The phrases were randomized and spoken 16 times each.

The results indicate that both the cricothyroid muscle and the strap muscles are active during production of the tones of Thai. For the high F0 of the high, rising, and falling tones, the CT shows increased activity, while the strap muscles show decreased activity. For the low F0 of the low, falling, and rising tones, the CT shows decreased activity while the strap muscles show increased activity.

An articulatory description of the tones in terms of laryngeal articulation features is proposed based on the patterns of CT and strap muscle activity associated with each of the tones. The HIGH TONE can be described as [+CT, -Strap], the LOW TONE, as [-CT, +Strap], the MID TONE, as [-CT, -Strap]; the dynamic FALLING and RISING TONES can be described as a sequence of HIGH TONE + LOW TONE [+CT, -Strap][CT, +Strap] and LOW TONE + HIGH TONE [-CT, +Strap] [+CT, -Strap], respectively.

In terms of laryngeal control of F0, the CT shows either an increase or decrease in activity 50 to 150 ms prior to the corresponding increase or decrease in F0, which suggests the CT plays a significant role in the control of F0. Since strap muscle activity shows decreases in activity roughly 150 ms prior to a drop in F0-- but only when the F0 falls from a mid-F0 range--the strap muscles may function additionally to adjust the physical conditions of the larynx to allow F0 to drop to a lower-than-mid F0 range. A modal control of F0 is proposed such that at above mid-F0 range, the CT controls the F0, whereas below mid-F0 range, the strap muscles adjust the laryngeal setting so that the folds are relaxed to vibrate at a lower frequency. Specifically, activity of the SH and ST may result in the vocal folds relaxing when the larynx is pulled down. Furthermore, when the SH contracts, it may anchor the hyoid bone, so that contraction of the TH may result in a tilting upwards of the thyroid cartilage, contributing to an expansion of the cricothyroid space, and a consequent shortening of the vocal folds.

Physiological studies of speech production such as this lay important ground-work for understanding how historical changes came about in languages, and specifically, how tonal systems developed.

References
Erickson, D. (1976). A Physiological Analysis of the Tones of Thai. Ph.D.Dissertation, University of Connecticut.
Erickson, D. (1993). "Laryngeal muscle activity in connection with Thai tones." Ann. Bull. RILP, 27, 135-149.
Hirose, H. (1971). "Electromyography of the articulatory muscles: Current instrumentation and technique." Haskins Laboratories Status Report on Speech Research, SR-25/26, 73-86.
Kewley-Port, D. (1973). "Computer processing of EMG signals at Haskins Laboratories." Haskins Laboratories Status Report on Speech Research, SR-33, 173-183.