Korean laryngeal contrast revisited: An electroglottographic study on denasalized and oral stops

Abstract

In several Korean dialects, domain-initial nasal onsets undergo denasalization as a recent sound change. Nasal stops may be realized as prevoiced or even devoiced stops. This makes it necessary to examine the interplay of phonetic properties of the denasalized and the three oral stop series as a whole, in synchrony and diachrony. What are their concomitant and conflicting properties? Our study provides a bigger picture of the laryngeal contrast in Seoul and Gyeonggi Korean by examining the acoustic distributions related to the laryngeal properties of the four stop series, using acoustic and electroglottographic data. VOT and f0 play important roles in the distinction of the four stop series, in line with previous studies. While the contribution of voice quality is relatively minor, we show that it plays an essential role of disambiguation when the VOT–f0 space gets crowded: When lenis stops can be confused with other stops, there is an enhancement of breathy voice. Finally, we discuss stop variation according to prosodic contexts. We highlight the basis of both syntagmatic variation and paradigmatic contrast in their phonetic implementations. They illustrate a constant reorganization to reconcile contrast maintenance with constraints from articulatory and perceptual systems, as well as language-specific structures.