An acoustic study on the domain of final lengthening in Korean

Abstract

Phrase-final lengthening is universally known to be the phenomenon where the last segment/syllable is the most influenced by boundary-related lengthening; however, the specific pattern is different for each language. The present acoustic study explores the domain of final lengthening in Korean and demonstrates how the pattern is different compared to other languages, including English. The pattern of final lengthening is different in not only the phonological structure of each language, but also in the phonetic contents. Korean, which is often assumed to be a syllable-timed language, might show a different lengthening pattern from other Indo-European languages by emphasizing the entire word, not a specific lexically stressed segment. Korean would also be influenced by its specificity—Korean has strong domain-initial strengthening—but the specific final lengthening pattern in Korean has not yet been fully understood. Therefore, this paper focuses on the domain of final lengthening in Korean and on language specificity’s influence on Korean. The results of this paper indicate that Korean shows a progressive lengthening pattern, where the lengthening effect gradually weakens as the segment gets farther away from the boundary. This pattern is similar to Hebrew. The results of all syllable types (two-syllable and three-syllable words) show that the lengthening effect does not generally appear (or weaken) on the initial syllable; however, the final segments are robustly influenced by the phrase-final boundary. Differences in the lengthening pattern related to focus type and the initial strengthening effect are not found. This is the opposite result of the predictions that the lengthening might be influenced by phrase-final lengthening and initial strengthening in Korean, and that the final lengthening interacts with the prominence system as in English. In addition, the presence and absence of a coda consonant does not affect the lengthening pattern. However, the specific pattern is different, depending on the phonetic contents. In particular, a phonetically short vowel (i.e., a high vowel) that is located further from the boundary has a longer lengthening pattern than a phonetically long vowel (i.e., a low vowel). It is proposed that a vowel type (short vowel vs. long vowel) is a significant factor in deciding the final lengthening pattern in Korean, implying that the lengthening is influenced by the vowel’s phonetic content (i.e., vowel height). It is estimated that the articulatory movement of a short vowel is relatively less than a long vowel; therefore, the short vowel should have more influence on lengthening effect than the long vowel does.