上代日本語における母音組織と母音交替

概要

論文の詳細を見る
この論文は国立情報学研究所の学術雑誌公開支援事業により電子化されました。It is now well known that ancient Japanese, seen chiefly in 8th Century documents, reveals the existence of eight vowels instead of five. The eight vowels can be classified into front (palatal) and back (velar) ones, and most scholars are of the opinion that we find here clear indication of an old vocal harmony thereafter extinct - the vowel i being a neutral element as in most of the Uralic and Mongolian groups of languages. However, minute observation of the distribution of the vowels and of the " modi inothum " of their interchanges discloses the facts : 1) that these harmonic-seeming indications are but the rather scanty remmants of an older, stricter vocal harmony, stemming from long before the time of their documentary attestation; 2) that vowel i was in origin not a neutral one, but an element of the front series of vowels, exactly in opposition to i of the back series (this wasprecisely the case with i and/or c of ancient Uralic and Mongolian, too); 3) and that, accordingly, we have to assume, in opposition to the u, also the existence of an *u in the more complete vocal system of the more ancientJapanese before its documentary fixation; 4) and thus that the original system of vowels seems on the whole to show a closer resemblance to that of Uralic and Altaic (including Turkic) vowelsin origin.
京都大学の論文
1956-11-20