I was surprised to learn that 飯 has the reading いい such as in 飯{いい}田{だ}橋{ばし}, and this is also listed as an "archaicism" in my dictionary. I know this kanji has the more usual kun-yomi めし (and on-yomi はん), but what is the etymology of this alternate reading? In particular, is it originally a native Japanese word (Yamato kotoba)?

2 Answers 2


Yes, い[ひ]{い} is indeed a yamato kotoba which can be found in 万葉集. From 学研全訳古語辞典:

いひ 【飯】 名詞 飯(めし)。

出典 万葉集 一四二 「家にあれば笥(け)に盛るいひを」 [訳] ⇒いへにあれば…。

いひ is even older than めし. めし gradually replaced いひ around the Muromachi period, according to this page.

  • 1
    While I realize that you are quoting a dictionary, it is important to note that poem #142 does not actually contain any evidence that 飯 is supposed to be read as いひ. Rather, it contains the Chinese word 飯, but no 万葉仮名 to indicate how it is supposed to be read. It is certainly a reasonable guess considering that c. 11th century 類聚名義抄 contains the reading イヒモルケ (=飯盛る笥) which closely fits this context, but it is still a guess as well as being several centuries removed. Surely there are more definite textual sources that can be quoted.
    – Dono
    Aug 8, 2016 at 11:50
  • 1
    Poem 888 uses phonetic 比{ひ} to spell the -hi portion of 干{かれ}飯{ひ} "dry rice, dry grain". Poem 3609S also uses 比{ひ} to spell 飯{ひ} phonetically in the compound 笥飯{けひ} (literally "container grain"), the ancient name of Tsuruga in Fukui Prefecture. Aug 12, 2016 at 0:52

Origins of the ii reading

A native origin?

This appears as ihi in historical hiragana renderings (i.e. using kana spellings from before the Post-WWII spelling reforms), likely realized in ancient times as ipi.

Digging around, this is a bit of an odd one. There are not many ancient native-Japanese words (Yamato kotoba) that have this phonetic structure. [VOWEL] + /pi/ could suggest a verbal derivation, but the only verb /ipu/ is the root of modern 言{い}う, and that doesn't fit.

It could be a native term that's just a little odd.

An ancient nativized borrowing?

I have bumped into interesting suggestions that this might not be a Yamato kotoba after all.

The older meaning of 飯 with the /ipi/ reading appears to be the more generic sense "grain", rather than specifically "rice".

We know from archaeology that the Yayoi settlers who crossed over to the Japanese archipelago from the mainland were also grain farmers. The geographic origins are a bit murky, but be it modern Jiangsu Province or somewhere in Manchuria, these pre-Yayoi people would probably have had contact with Chinese speakers, and trade activities could lead to the borrowing of a few relevant words.

In Middle Chinese, 粒 "grain" was pronounced as /liɪp̚/, which we also see evidenced in now-obsolete Korean hanja reading [립]{rip}. It seems that Old Japanese (and probably prehistoric Japanese) did not allow native words to begin with /r/ (see quote and link in this other JSE thread). If Middle Chinese /liɪp̚/ were imported into prehistoric Japanese, it may well have nativized as /ipi/ rather than the /ritu/ form we see for the later borrowing of 立 with a similar MK reading of /liɪp̚/.

And given the nearly-millenia-long Japanese propensity for applying kanji spellings in creative ways to write words with related meanings but different readings, it is not much of a stretch to find the 飯 character with a Chinese meaning of "food, cooked rice" used for the /ipi/ "grain" sense.

Note that this is all very circumstantial.

Side note: Origins of the meshi reading

It bears noting that the kun'yomi of めし is the 連用形{れんようけい} or continuative form of verb 召{め}す. This started out as an honorific for 見{み}る, using the -e vowel stem + ancient honorific verb ending す (which interestingly has parallels with the -si- or -se- honorific verb infix in Korean). Over time, and through various evolutions of usage and meaning, this verb mesu gained a sense of "to put into or onto the body""to eat; to drink; to wear". The "eat" sense is already apparent in the Man'yōshū in poem 3853, which spells out the imperative form 召{め}せ in phonetic man'yōgana as 賣{め}世{せ}.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .