抄録
The central vowels in the Northern Ryukyu dialects are historically thought to have been formed by the following four processes: (I) centralization of the vowel [e], (II) centralization of the vowel [u] following the consonants [s・z・c], (III) the coalescence of diphthongs [ai・ae] and IV) the assimilation to the central vowel in an adjacent syllable. Among these, process II brought about the condition of the so-called "Hitotsugana-ben" throughout most of the Ryukyu dialects except the Northern Amami dialect. So the absense of process II can be considered one of the remarkable characteristics of the Northern Amami dialect. Again, the proto-condition for process I, prevalent in all dialects of the Amami regeon, and process II must have included such a contrast as ^*Ci/Ce/Cu (C=s,z,c). This contrast can be interpreted to have changed into the following situations present in current Northern Amami dialects: (1) Ci[i]/Ce[ï]/Cu [u] observed in Yoro, (2) Ci[i]/Ce = Cu[ï] observed in peripheral area of Ooshima, Ci=Ce[i]/Cu[u] observed in Kikai, Ci = Cu[ï]/Ce[i] observed in north part of Tokunoshima, and (3) Ci = Ce = Cu[ï〜i] observed in Naze and often refered to as "Hitotsugana-Ben". Historically (1) is considered to be the oldest situation. This was followed by (2), which was brought about by the merger of the various parts of (1). Lastly, stage (3) was brought about by the merger of the remaining independent part of (2).
