[10] For example, Bjarke Frellesvig (2010) distinguishes three dialect areas:[11] He states that these dialects form a continuum with the varieties of Nara Old Japanese, with North Eastern Old Japanese constituting the most divergent variety.

[15] Eastern Old Japanese is a SOV language[a] with a structure including a modifier at the start of the sentence, although there are exceptions.

[10] The phonotactic structure of Eastern Old Japanese is strictly (C)V, without consonant gemination nor long vowels.

Subsequently, Kantō would have been populated by Japonic speakers directly from Kyūshū, without passing through central Japan.

[20][21] However, Alexander Koji Makiyama (2015) finds the results of diachronic changes in Eastern Old Japanese such as in denasalization, fortition and vowel raising unconvincing in comparison with the Ryukyuan languages.

In fact, he finds: The hypothesis of a linguistic contact or a resemblance is therefore, in the state of current knowledge, only speculative.