Linguists mostly agree on the validity of the
Amami–Okinawan languages as a family. The subdivisions of Amami–Okinawan, however, remain a matter of scholarly debate, with two major hypotheses: • In a two-branch hypothesis, posited by
Nakasone (1961),
Hirayama (1964) and
Nakamoto (1990), among others, Amami–Okinawan divides into
Amami and
Okinawan, with the northern and southern varieties of Amami Ōshima both falling within the Amami branch. • In a three-subdivision hypothesis, proposed by
Uemura (1972) as one of several possible classifications and supported by
Karimata (2000), Northern Amami Ōshima (perhaps together with
Kikai) and
Central/Southern Okinawa form two branches, while the intervening varieties – Southern Amami Ōshima (Setouchi),
Kunigami, and the dialects/languages of the islands between – form a third branch. In this proposal, Amami Ōshima does not constitute a single language, and the northern and southern varieties are not even more closely related to each other than they are to other Ryukyuan languages. The two-subdivision hypothesis is convenient for discussing the modern languages since the posited linguistic boundary corresponds to the centuries-old administrative boundary that today separates Kagoshima and Okinawa Prefectures. In addition, several
isoglosses do group Northern and Southern Amami together. In Amami, word-medial is changed to or even dropped when it is surrounded by , or . This can rarely be observed in Okinawan. Japanese becomes in Amami and in Okinawan. tentatively group Kikai dialects together. ==Dialects==