Of the nine scholars, three dealt with “sound”: Ueda Kazutoshi, Hashimoto Shinkichi, and Arisaka Hideyo. The three talk about the same subject, but the evidence each held is different. We line up their work along this difference. (Overall timeline)
The three kinds of evidence
| Scholar | Object | Kind of evidence | Core confirmed |
|---|---|---|---|
| Ueda Kazutoshi | The old sound of the ha-row | Small facts of modern Japanese (the po of “okorippoi,” “shimeppoi,” etc.) | The reasoning that the ha-row was once P |
| Hashimoto Shinkichi | Sound distinctions in the Nara period | The use of man’yōgana (how characters were written apart) | The fact that man’yōgana are strictly written in two groups |
| Arisaka Hideyo | Nara-period vowels; what a sound is | A hidden regularity in the combinations of vowels within words | Arisaka’s Law. And in Onin-ron, a rethinking of how sounds are grasped |
Ueda’s evidence starts from the language being spoken now. Hashimoto’s is the use of characters thirteen hundred years ago. Arisaka’s is a regularity hidden in the combinations of vowels within words. Since the evidence differs, the kind of certainty differs too.
Ueda Kazutoshi — reasoning an old sound from modern speech
Why is the po of “okorippoi” a po? From this small fact, Ueda argued that the ha-row of Japanese was anciently a P sound (the Ueda Kazutoshi feature).
This reasoning became the skeleton of the present standard view. More precisely, the verdict article’s summary is that the course is p → ɸ → h, and whether it was p or ɸ is still debated.
Hashimoto Shinkichi — the “fact” of a writing distinction
Hashimoto confirmed that Nara-period man’yōgana were strictly written in two groups (jōdai tokushu kanazukai). The verdict of the Hashimoto feature treats this as an unshakable premise (the Hashimoto Shinkichi feature).
However, the actual nature of those two sounds is not yet settled. Multiple views coexist, including the eight-vowel theory. What Hashimoto confirmed was the fact that there is a writing distinction; what the two sounds were was a separate question.
Arisaka Hideyo — asking the “identity” behind the fact
Arisaka took up that separate question. He supported, from the regularity inside words, a path of interpreting the kō-otsu distinction Hashimoto had organized as a difference of vowels (the Arisaka Hideyo feature).
Arisaka’s Law says that within a single word root, the otsu-type o-row vowel does not co-occur with a, u, or the kō-type o-row vowel. The verdict article treats this as an unshakable foundation. The eight-vowel theory (the actual nature of the otsu-type vowels), on the other hand, is influential but unsettled.
And in Onin-ron (1940), Arisaka rebuilt the question of what a sound is. The verdict article summarizes it as “living on not as an answer but as a way of posing the question.”
What to keep in mind when reading
First, the line from Hashimoto to Arisaka can be drawn within what the Arisaka feature confirmed, but the relation between Ueda and Hashimoto/Arisaka has not been confirmed in this group of features. Ueda too dealt with sound. But how Ueda’s reasoning connects to Hashimoto’s or Arisaka’s work, we do not assert here. In the chart too, we draw no line.
Second, “unshakable,” “influential,” and “unsettled” are different states. The fact that there is a writing distinction and theories of the nature of those sounds must not be mixed.
Summary
| Point | Detail |
|---|---|
| The three kinds of evidence | Facts of modern speech (Ueda) / the writing distinction (Hashimoto) / vowel combinations (Arisaka) |
| Unshakable | The fact of jōdai tokushu kanazukai. Arisaka’s Law |
| Influential but unsettled | The eight-vowel theory (nature of otsu-type vowels). Whether P was p or ɸ |
| Living as a question | Onin-ron — what is a sound? |
| Caution | The influence between Ueda and Hashimoto/Arisaka is unchecked |
In a nutshell — In the sound thread, three ways of working stood each on its own evidence: reasoning an old sound from modern speech, confirming the fact of an old writing distinction, and then asking its identity.