Phonetics and Phonology of Vocative Chant Variation in Korean

Eon-Suk Ko
University of Pennsylvania

This study investigates the nature of prosodic prominence in Seoul and Chonnam Korean through the realization of vocative chants. I claim that Seoul is a pitch accent language while Chonnam is a stress one, according to the definition of Beckman (1986). Evidence is drawn from the assignment of the accentual H tone (H*) in vocative chant, and acoustic measurements that shows the correlation between the H tone and the vowel length only in Chonnam, but not in Seoul. I argue that Chonnam provides evidence that both underlying tones and metrical structure are necessary for explaining phonological prominence. The nature of the distinct prosodic systems of Seoul and Chonnam have been investigated in Jun (1993), where she argues that the tonal events in Korean do not function to mark prominent syllables. She, therefore, classifies both dialects as pitch accent languages, which differ only in the inventory of the tonal shapes, LHLH for Seoul, and LHL and HHL for Chonnam. In her analysis, Chonnam has a vowel length distinction, but Seoul does not. Thus, she defines that the TBU is syllable in Seoul, but mora in Chonnam.
However, an experiment on the realization of the vocative chant of the two dialects suggests that we need a metrical structure to explain the assignment of the H* in vocative chants of Chonnam. Four speakers stratified by dialect and sex were solicited to read and then and sing the calling contour for 60 different names, each twice in random order. All the names used are composed of two syllables, followed by a vocative suffix -(y)a. An acoustic analysis was conducted to measure the F0 and the length of each syllable of the names. The result shows that the H tone and vowel length is closely related in Chonnam, but not in Seoul. As is well known, the accentual H tone (H*) in a vocative chant is aligned with the most prominent syllable of the name (Liberman 1975, Ladd 1997). In Seoul, the location of the H* is fixed on the second syllable, as is predicted by the pitch accent anlaysis of Jun. However, the location of the H* varies on the first or second syllable in Chonnam. Jun's pitch accent analysis predicts a HHL pattern for every name that begins with an aspirated consonant in Chonnam. Thus, both 'Sunja-ya [su:njaya]' and 'Sujin-a [sujina]' should be assigned the tonal pattern of HHL(L). However, their realization in vocative chant shows an opposite tonal pattern of HLM and LHM for each name. This suggests that the location of the H* is determined by a metrical structure in Chonnam rather being a reflex of the pitch accentual pattern.
It has been argued that Korean has an iambic foot system (Lee 1987). However, I contend that the metrical structure of Chonnam cannot be weight-driven. This is because names of equal distributional pattern of weight show different tonal patterns in vocative chants. For example, 'Sangwon-a [saNw+na]' and 'Dongchul-a [doNc`+ra]' are realized as LHM while 'Suntaek-a [suntQga]' and 'Hyunchul-a [hy+nc`+ra]' are realized as HLM.
I propose that the metrical structure of Chonnam is determined by the underlying tonal specification of each syllable, which I assume to have been lexicalized from Middle Korean tones. The stress assignment rule of Chonnam I propose is head initial unbounded, with an initial L tone extrametrical. The duration of the vowel is concomitantly lengthened if the stressed syllable is underlyingly specified as H, but not if L. Thus, Chonnam shows that both underlying tones and metrical structure are necessary and interactive components of the grammar of phonological prominence as has been also proposed for languages such as Huave (Noyer 1991).
The difference in the aspect of surface vowel length distinction in Seoul and Chonnam is argued to be due to the loss of stress system in Seoul rather than different underlying vowel length distinction per se. I suggest that the loss of stress system in Seoul is an on-going historical sound change, concomitant with the gradual loss of vowel length distinction among younger generation speakers (Blumstein & Magen 1993). This could be attributed to the diffusion of lexical tonal specification due to the change in writing system in Korean that excludes Chinese characters, which happened only within a few decades. An investigation of the prosody of the native Korean vocabulary is expected for the suggested theory to be able to provide a whole picture of the prosodic prominence of Korean.


Beckman, Mary (1986) Stress and Non-stress Accent Language, Dordrecht, Holland.
Jun, Sun-Ah (1993) The Phonetics and Phonology of Korean Prosody, Ph.D. dissertation, OSU
Ladd, Robert. (1997) Intonational Phonology, Cambridge University Press
Lee, Hyun-Bok (1987 Korean Prosody: Speech Rhythm and Intonation, Korea Journal 27.2
Liberman, Mark (1975) The Intonational System of English, Ph.D. dissertation, MIT
Magen, Harriet and Sheila Blumstein (1993) Effects of Speaking Rate on the Vowel Length Distinction in Korean, JPhon 21.4
Noyer, Rolf (1991) Tone and Stress in the Saan Mateo Dialect of Huave, ESCOL 8

< Back to List of Abstracts | Back to PLC23 Home Page >

About the PLC23 Committee
Previously held Penn Linguistics Colloquium: PLC22 (1998), PLC21 (1997)

Penn Department of Linguistics
University of Pennsylvania