In the plain-language article we saw the story of inferring the old sound of the ha-row from the “po” of okorippoi. Here we check, in the original, in what order Ueda Kazutoshi’s “P-on-kō” (in Kokugo no tame, Part 2) actually piles up its evidence. (The English renderings of the quotations below are ToL translations.)
First evidence: working back from voiced sounds to voiceless
Ueda first turns to the relation between voiceless and voiced. He shows it in the form of an equation.
D＝T G＝K B＝（？）＝P！！
The da row (D) came from the ta row (T), and the ga row (G) from the ka row (K). So from which voiceless sound did the ba row (B) come? Ueda continues:
B音の出し清音は、決してハ行（H）音にもあらず、ファ行（F）音にもあらず、卽ち純粋なる唇的清音パ行（P）音ならざるべからず。
(The voiceless sound from which B arose must be neither the H sound of the ha row nor the F sound of the fa row, but the pure labial voiceless sound of the pa row: P.)
This is the skeleton of the argument. If a voiced sound is a voiceless one with “voicing” added, then the voiceless partner of the ba row must be the same lip sound — that is, P. Yet the ha row, which today is lined up as the voiceless partner of the ba row, is not a lip sound but a throat sound (H). In this mismatch Ueda saw a clue to the old form.
Second evidence: questions to the opposing view
Against those who hold that “the ha row was a throat sound from the start; there was no P in high antiquity,” Ueda lines up questions.
何故に今日の如き喉的H音が濁る必要ありたるか。よし濁る必要ありたりとするも、喉的H音が濁るに当りて、何故に唇的濁音とはなりたるか。
(Why was it necessary for a throat H sound like today’s to be voiced? And even granting the need, why, when the throat H was voiced, did it become a labial voiced sound?)
A throat sound gets voiced, and why does it become a lip sound? Ueda notes that the old theory’s treatment of the ha row as a “labial” also overlaps with this question. This way of questioning is characteristic of Ueda’s reasoning: first accept the opponent’s theory, then press in turn on the points it can’t explain.
Third evidence: how Sanskrit sounds were written
Ueda looks to evidence outside Japan as well.
梵漢の二国語に於けるP音が、日本の波行にて写され居るにかゝはらず、梵漢の二国語に於けるH音が、総て我国にては加行に写さるゝ事これなり。
(It is this: although the P sounds of the two languages, Sanskrit and Chinese, are written in Japan with the ha row, the H sounds of those two languages are all written in Japan with the ka row.)
Looking at the table Ueda gives, for instance, the “ha” of Sanskrit Hami (哈密), Arahan (阿羅漢), and Maha (摩訶) is read in Japan in the ka row. If the ha row of Japanese at that time had been an H sound, it could have written foreign H. In fact, though, the throat sound H was written with the similar throat sound K (the ka row). Ueda reads it as: Japanese of that time had no H sound; instead the ha row was P.
Fourth evidence: Japanese inside Ainu
Ueda also turns to Japanese words that entered Ainu.
Ueda first notes that Ainu distinguishes the three sounds P, F, and H, and then shows that its words include ones that look like borrowings from Japanese.
Ueda lists words like pashui, corresponding to Japanese hashi (chopsticks); pishako, to hishaku (dipper); pone, to hone (bone); and pakari, to hakari (measure). In every one, the Japanese ha row appears in Ainu as P.
So Ueda poses this question.
もし新しく入りし者なりとせば、何故にF・Hを有するアイヌは、之を其の音にて伝へざるか。
(If these were newly entered words, why did Ainu, which has F and H, not carry them in those sounds?)
Ainu has both H and F. Yet it received Japanese ha-row words as P. Ueda’s line of argument: is it not because the ha row of Japanese in the era when these words entered was P?
What this paper shows
“P-on-kō” is not a paper that reaches its conclusion on one decisive piece of evidence. It piles up several pieces of evidence of different kinds toward the same conclusion: facts within Japanese (the correspondence with voiced sounds, the habit in compounds, onomatopoeia), facts outside Japan (how Sanskrit sounds were written), and facts of a neighboring language (Ainu).
Summary
| Point | Detail |
|---|---|
| 1. Voiced pairs | If D=T and G=K, the voiceless source of B must be a lip sound, P |
| 2. Questions to opponents | Why does throat H get voiced, and why into a lip voiced sound? |
| 3. Sanskrit transcription | Sanskrit and Chinese H was written in the ka row, P in the ha row |
| 4. Ainu | Ainu, which has both H and F, takes Japanese ha-row words as P |
| Style of argument | Pile up different kinds of evidence toward the same conclusion |
In a nutshell — starting from one point, “the voiceless source of a voiced sound must be a lip sound,” Ueda Kazutoshi piled up evidence — through how Sanskrit was written and Japanese words in Ainu — to argue that the ha row of high antiquity was P.