Focusing on the dialects in Setouchi-chō in Amami-ōshima, which constitute one of the subgroups of Ryukyuan, this paper presents a diachronic analysis of the occurrences of different types of heavy syllables. These are represented by word-final closed syllables in k'uˑp (neck) and Ɂusaˑk (rabbit), the word-internal long syllables in kutuːba (word) and waɾaːbï (child), or word-internal closed syllables in ɸuk'ɾu (paper bag) and Ɂapɾa (oil), in the Shodon dialect of Kakeroma-jima in Setouchi-chō.

To explain how and why such different types of heavy syllables were generated in this area, this paper proposes that a relative chronology between accentual changes, which were supposed to have occurred in this area, and various phonological changes that motivated the appearance of the heavy syllables, must be taken into consideration. The paper also contends that the vowel raising of mid to high vowels (^*o> u, ^*e> ï) must have taken place after the phonological changes that motivated the emergence of those heavy syllables.

In reference to the Qieyun (切韻) included in the Shinsen-jikyō (新撰字鏡), Ueda (1981) pointed out that it was consistent with Zhangsun Neyan's Qieyun (長孫訥言切韻) system. On the other hand, Sadakari (1989) stated that only the lower level tone parts were different.

In this article, I will first show that, indeed, only the lower level tone parts are different, as observed by Sadakari (1989), and will take the position that they display the special characteristics of the Tangyun (唐韻) system. Next, I will review in detail the problems with the example presented by Ueda (1981) from the revised edition of Zhangsun Neyan's Qieyun (長孫訥言切韻). In addition, I will show that there are other errors that are similar to the example of the incorrect fanqie of Qieyun (切韻) as presented by Ueda (1981,1984).