The article is composed of the following: 1. Ryukyuan language, its definition, history and the present state 2. Some comments on the methodology of the phonetic research 3. Phonological characteristics of the Ryukyuan dialects as compared to mainland Japanese 4. Differences between the phonetic characteristics of the Amami-Okinawa dialect group and those of the Miyako-Yaeyama dialect group 5. Vowel phoneme systems of Amami, Okinawa, Miyako, Yaeyama, and Yonaguni Dialects 6. Some characteristics of Ryukyuan consonant phonemes, including the phonemic contrasts between the laryngealized unaspirated consonants and the consonants without laryngealization 7. Further sub-classification of the above two dialect groups
Some sub-dialects of the Northern Ryukyuan Dialect of Japanese keep the voiceless bilabial stop /p/ of ancient Japanese, while in some of its sub-dialects, the stop has changed to /h/, as seen in present-day Japanese dialects on the mainland of Japan. Moreover, in some Northern Ryukyuan dialects, /k/ has historically changed to /h/, and /t/ has also changed to /s/ or /θ/. These cases of spirantization of voiceless stops are all mutually related. These changes are also connected to vowel raising (o > u, e > i) and laryngealization of consonants, both of which have taken place in the Northern Ryukyuan Dialect. The degree of spirantization differs from one sub-dialect to another.
The South Ryukyu dialects are divided into two categories: one of which basically has three vowels /i ・ a ・ u/, and the other, four /i ・ ï ・ a ・ u/. The former is assumed to be derived from the latter by the decline of the neutral vowel /ï/. This neutral vowel has been interpreted and transcribed differently in phonetic notation by many researchers, e.g., [ï, ɨ, 〓]. The purpose of this article is to analyse the acoustic substance of these neutral vowels, using the digitized phonetic sources of the Aragusuku dialect as reported in Kuno et al. (1993), and the Karimata dialect and Oogami dialect as reported in Oono et al. (1998). According to Imaishi (1997), the characteristics of the neutral vowel can be seen from examination of F2. Although the distribution of F2 of the neutral vowels in all of these dialects locates between [i] and [u], the values of F2/F1 differ significantly among these dialects. This is supposed to be affected by the difference of articulatory position of the vowels in question. The value of F2/F1 of Oogami dialect is especially low, and in part overlaps with that of [u]. This means that the tongue position of the neutral vowel in Oogami is very close to that of [u]. Presumably, the distinction between the two vowels depends on the configuration of the lips, spread vs. round.
Miyako dialect which belongs to Ryukyu dailect is peculiar among Japanese dialects in its consonant-based characteristics. Its so-called "mid-vowel" is more like a consonant than a vowel in substance, and it has several syllabic consonants which can compose a syllable by themselves. Various phonological interepretations of the dialect have been proposed, but they are all insufficient sometimes with their redundancy, sometimes with their internal contradiction. The best solution is not yet found.
The aim of this paper is to describe some characteristics of the accents of the Amami dialects, as compared with those of Mainland Japanese dialects. The following points are discussed with concrete examples: (1) there is great variation among the dialects, (2) many dialects have multi-peak patterns, (3) N-pattern accent systems are dominant, with multi-pattern accent systems found as well, and (4) there are various kinds of accentual alternations. The historical aspects of Amami accent are also sketched: (5) I propose proto-categories for accent, and (6) three types of origins for accent are found and discussed. Lastly, I propose that the traditional analysis of syllable structure must be revised to capture the general nature of accentual alternations.
This paper seeks to evaluate phonetic (articulatory and acoustic) and phonological explanations of vowel coalescence over intervening consonants, using data mainly drawn from Ryukyuan dialects. Coalescence over an intervening h is attested in Yaeyama, Okinawan, and Amami dialects. This coalescence receives both an acoustic phonetic explanation, and an explanation within a featuregeometrical view of phonology. Also attested in Okinawan and Amami dialects is a coalescence of vowels across an intervening r. It is argued that acoustic phonetics fails to account for this, but a straightforward phonological account is available.
Based on the regular correspondence between the Kindaichi tonal classes, originally developed for mainland Japanese dialects, and those of Ryukyuan dialects, this study proposes that the accentual system of proto-Ryukyuan had at most a three-way tonal distinction for words of up to three syllables. Then, assuming that these tonal classes in proto-Ryukyuan can be categorized into at least three large groups, each being supposed to share certain common characteristics in the proto-Ryukyuan accentual system, these are referred to as Pattern Categories (Keiretsu) A, B, and C in this paper. This study then proposes lists of words belonging to each category, based on research conducted on Okinoerabu Island in the Amami region, where the dialect system is known to retain the maximum numbers of tonal distinctions within the Ryukyuan dialect area.