@article{oai:u-ryukyu.repo.nii.ac.jp:02005436,
 author = {宮良, 信詳 and Miyara, Shinsho},
 issue = {1},
 journal = {ヒューマン サイエンス, Ryudai human science review},
 month = {Mar},
 note = {In the previous analyses of the Shuri dialect and many other Okinawa main-island dialects, palatalized consonants have been proposed as phonemes /c/ and /z/. However, this analysis of /c/ and /z/ is unsatisfactory in that: (1) in essentially the same environment, it becomes neccessary to make a phonological distinction of whether /c/ should derive [t] or [k] and of whether /z/ should derive [d] or [g] ; (2) phonemic representations containing /c/ and /z/ cannot provide any means to make distinctions in many sets of homonyms ; (3) although correspondences are expected in the phonemic representations of words of the same origin among the dialects of Japanese, the expected correspondences of / c/ or /z/ are opaque. For example, /c/ in /cici/(→[tʃitʃi] 'moon') corresponds to /ts/ or /k/ in /tsɨkɨ/(→[tsɨkɨ]) of the Ishigaki dialect, one of the southern Ryukyuan dialects, and to /t/ or /k/ in /tuki/ (→ [tsuki]) in Standard Japanese. Similarly, the correspondence of /z/ in /zii/(→ [dʒii] 'soil') or /zintoo/ (→ [dʒintoo] 'true') is also opaque when compared with [dzɨɨ] or [gintoo] in the Ishigaki dialect and [dʒi] or [gentoo] in Standard Japanese. The most serious problem of the above three is the one stated in (1). By contrast, the present analysis, which renders it unnecessary to set up such phonemes as /c/ and /z/ and whose palatalization is applicable to /t/, /d/, /k/, /g/ as well as /s/, /h/, has an automatic consequence of solving the problems above., 紀要論文},
 pages = {1--34},
 title = {沖縄本島首里方言における口蓋化について},
 year = {1995}
}