陸中宮古方言アクセントの実相  [in Japanese] The Current State of Word Accents in the Rikuchu-Miyako Dialect  [in Japanese]

岩手県宮古市方言のアクセントは,柴田1955で「のぼりアクセント核」体系とされて以来注目されてきたが,不明な点が多かった。この方言のアクセントの全体像を詳細な調査により明らかにした。音声レベルでは,(A)1基本アクセント節内でアクセント節例によっては2箇所が高まる重起伏調・(B)語によっては声の上昇位置が定まっている・(C )上昇した音節からは少しずつ下降する調子がある(後続語によっては下降しない)・(D)有声音無声化が音調に影響がある・(E)特殊音との関係で音調を担う単位は音節,の5特徴が認められた。音韻論的解釈の結果,アクセント体系は東京式アクセントの「さがりめ」体系とは異なる「のぼり核」体系であることが明確になった。また,同一語の単独1語文と付属語接続時の各実現音調で高い部分が異なる,以前"山が動く"と表現された現象や,基本アクセント節の音調が「起伏」でも「無核」であることなどに合理的説明が得られた。

The Rikuchu-Miyako Dialect-spoken in the central area of Miyako City in Iwate Prefecture-has a very unique word accent system, one first reported by SIBATA Takesi (1955) as a system that is made up of "Ascending word accent kernels". However, after SIBATA'S report, many problems still remained unsolved. Now, this paper's detailed survey will verify the whole word accent system of this dialect : [1] It finds five phonetic features in word accents of this dialect : A) In some words, there are two tonal peaks (initial high sequence and middle or final high) in single tonal phrase-in most Japanese word accent systems, words have only one tonal peak (one high sequence or syllable) in single tonal phrase- ; B) Some words have a fixed point of pitch rising ; C) After pitch rising, the pitch lowers little by little ; D) Vowels or originally voiced consonants influence the pitch contour when they become devoiced ; E) In relations between word accent and asyllabic sounds, the pitch contour is controlled on syllables (not on morae). [2] Using a phonological interpretation based on those phonetic features in linguistic sources, this paper points out that the word accent system of this dialect is different from "Descending kernel" systems such as those in the Tokyo dialect. It rather confirms that this is an "Ascending kernel" system. [3] Other problems are settled in this analysis. One problem is "high tone shift", in which the initial high sequence and final high tone syllable in utterances of a noun with following particles is different from those tonal peaks of that noun when it is offered alone ; for example, 'uTA〜'Uta-GA, SAkaNA〜SAKAna-GA, NI'WAtoRI-NI' WATOri-GA ('uta=song, -ga=subject particle, sakana=fish, ni'watori=fowl ; capital letters signify high tone syllables) . The other problem is the matter of kernelless phrase in undulation. In previous theory on Japanese word accents, undulating phrases were seen to invariably involve kerneled words. But, in this dialect, there are undulating phrases filled with kernelless words. Both of those two matters can be considered in relation to the phrase tone theory of UWANO Zendo (1984 etc).