Introduction
This pages summarizes key aspects of the usage of kana in Japanese prior the the reforms of the 1940s. This known in Japanese as きゅうかなづかい (kyuukanadzukai - 旧仮名使い or 旧仮名遣い). For example, the きょう (kyou) reading of the kanji 京 was written きゃう, and the きょう reading of 轎 was written けう. Verbs which now have an う ending such as 叶う were written 叶ふ.
The changes were made in the 1940s (after sixty years of often heated debate) to bring the writing into line with the modern pronunciation.
Many kanji dictionaries, e.g. Morohashi, use the old readings for the On readings (音読み), and major Japanese dictionaries such as the Koujien (広辞苑) show the old readings as well as the modern.
General
The kana や, ゆ and よ when writing syllables such as きょう were often written in their large form, e.g. きよう rather than きょう. Also the つ used to indicate a sokuon (促音) was often written in its larger form: がつこう rather than がっこう.
Verbs
Verbs for which the "dictionary form" now end in う used to end in ふ. This extends to conjugations of the verb, for example for 叶ふ we see 叶はず (叶わず), 叶ひつつ (叶いつつ), 叶へば (叶えば), etc.
The "ましょう" volitional/hortative verb ending was written "ませう", and similarly "でしょう" was "でせう".
Sino-Japanese Words (漢語)
These are just a few words showing the
旧仮名使い usage with 漢語. You see them in dictionaries like the
広辞苑, where the entry for 荒城 (ruined castle) reads:
| Current Reading | Old Reading | Kanji | Example(s) |
| きょう | きゃう | 京 | 京都 (きゃうと) |
| きやう | 饗 | 饗宴 (きやうえん) | |
| 狂 | 狂歌 (きやうか) | ||
| きょう | けう | 驕 | 轎輿 (けうよ) |
| 教 | 教育 (けういく) | ||
| きゅう | きう | 旧 | 旧悪 (きうあく) |
| かん | くわん | 緩 | 緩慢 (くわんまん) |
| ちょう | てう | 鳥 | 啄木鳥 (たくぼくてう) |
| しょう | せう | 昭 | 昭然 (せうぜん) |
| そう | さう | 装 | 正装 (せいさう) |
| お | を | 悪 | 悪心 (をしん) |
Native Japanese Words (和語)
| Current Reading | Old Reading | Example(s) |
| きょう | けふ | 今日 |
| たおる | たふる | 倒る |
| つかい | つかひ | 使ひ, 遣ひ |
| おとこ | をとこ | 男 |
Notes by Bart Mathias
(These notes have been extracted and edited from Bart's postings to the sci.lang.japan Usenet newsgroup.)
A few centuries back, all cases of vowel + "u" (which "u" may have been written originally as う, being the end of a Sino-Japanese reading or a pure Japanese う resulting from the loss of "k" in く, or else written ふ of which the intervocalic "h [p/f]" had ceased to be pronounced) assimilated to a single long vowel. If the preceding vowel was "u," obviously it's just a long "u." If it was an "i" (as in the きう of き うかなづかひ or as in the case of 言ふ) the result was a long "u" with a "y" in front.
The actual kana spelling remained rather random (although there were schools of orthography based on all kinds of weird ideas) until rekishiteki kanazukai (歴史的仮名使い) was recovered and given the nod during the Meiji period, and the spellings I'm using here became correct. Then about half a century ago a spelling reform resulted in things being written (with some strange exceptions) as pronounced, and きう became きゅう, い ふ became いう (oops, there's an exception--we'd expect ゆう--due to its being a verb, I suppose).
The other changes involving う and ふ were their being pronounced "o" after an "e" or an "o." In the case of えう and えふ," the frontness of the first vowel left a trace, the initial "y" sound of the resulting よう.
In the case of what would be like "(-)あう/あふ" in 歴史的仮名使い both vowels changed to a lower "o" than the case of "(-)おう/おふ" and "(-)えう/えふ"--we know this only because the Portuguese distinguished them in romanization--but before long the two kinds of long "o" merged into one.
Another oddity is the word for "get seasick/drunk" (now 酔う/よう). Early in the history of Japanese it was pronounced "wepu" and thus spelled ゑふ. Then "w" ceased to be pronounced except when followed by the lowest vowel, "a," and the original "p" changed to "h" and people quit pronouncing it in the middle of most words (what true exceptions can you think of besides あひる and あふれる?). The remaining pronunciation "eu" then went through the assimilation-to-long-vowel bit, resulting in "yoo." And finally, in the dialects that ultimately preserved a "(-)u" at the end of verbs, we end up with the pronunciation "you" (not "yoo") for the spelling よう.
Acknowledgements
Thanks to the people in the sci.lang.japan Usenet newsgroup for discussing this topic and getting me started on this page, and most particularly to Bart Mathias, who is a goldmine of information on this and so many other topics.
Further References
The Wikipedia article on historical kana
Chris Seeley: The Japanese Script since 1900, Visible Language, XVII 3 (Summer 1984), pp 267-302.
Chris Seeley: The History of writing in Japan, University of Hawai'i Press, 2000.
Jiri Neustupny, Introduction to Japanese Writing, Japanese Studies Centre, Melbourne, 1984