최적성이론에 의한 센다이방언 일형악센트의 분석

Abstract

The objective of this study is to examine the accent and metrical structure of Sendai Japanese in the framework of Optimality Theory (Prince & Smolensky 1993/2002). Sendai Japanese is often called ‘accentless’ dialect, because there is no distinctive accent pattern in this dialect. But this does not mean there is no accent (pitch fall) in the dialect. When a lexical word is uttered in isolation, an accent appears in the dialect. This is because of the markedness constraint HaveAccent, which requires that every prosodic word has at least an accent. Sendai Japanese accent has fixed location in prosodic words.: it appears on the penultimate mora in more than 1-mora words. To account for this penultimate accent, I adopt two well-known constraints NonFinality(μ') and Aligh-R, and proposes the ranking NonFinality(μ') ≫ Aligh-R. Aligh-R demands that any accent falls as far to the right as possible. Nevertheless, because NonFinality(μ') dominates Aligh-R in the dialect, the final accent is not allowed in more than 1-mora words. Exceptionally an accent appears on the antepenultimate mora, when a coda obstruent is located on thepenultimate mora in the dialect. To account for this phenomenon, I suggest a markedness constraint *μ'/Obstruent, which bans accented moraic obstruents. *μ'/Obstruent is more highly ranked than Aligh-R inthe dialect, therefore when a coda obstruent is located on the penultimate mora, the accent does notappear on the penultimate mora, but on the antepenultimate mora.