Evidence for seven vowels in proto-Japanese

On one more source of Old Japanese i 2

“…203-206). The raising of o 1 to u is well attested in EOJ and in the history of Japanese in general (Hattori 1978(Hattori -1979; (Hayata 1998); (Frellesvig and Whitman 2008), etc. ; for EOJ cf.…”

The Transeurasian Languages

“…Furthermore, he supported the view that Tungusic vowel harmony is retracted tongue root (RTR) based, as it is in Manchu and Evenki, and that the reduced vowel harmony in Contemporary Korean derives from a tongue root based system in Middle Korean. As far as the harmony-like opposition between o 2 and u, o 1 or a in Old Japanese is concerned, the recent reconstruction of a seven-vowel system in Proto-Japonic by Frellesvig and Whitman (2008) implies an underlying opposition between pJ *ɨ, *ə and *u, *o, *a, which does not exclude an original RTRbased contrast. Whereas Vovin (1993: 50-51) and Bugaeva (2015: 26-28) reconstruct palatal harmony in Ainu, Shibatani (1990: 15) speculates that the Ainu opposition between o and u, a might have its origin in tongue root harmony, but here the indications are even weaker than in the Japanese case.…”

Proto-Japonic *e and *o in Eastern Old Japanese

“…Their hypotheses are supported by very few examples and have not received general acceptance. (Whitman 1990, Martin 1987, is thus now generally accepted for PJ (Hattori 1978-1979, Frellesvig & Whitman 2008 …”