Okurigana: Kana Tails on Kanji | Japanese Grammar

Chinese characters never change shape — Chinese has no conjugation, so its characters never needed to. Japanese verbs and adjectives, by contrast, conjugate ceaselessly. Okurigana (送り仮名, おくりがな) is the beautifully simple invention that reconciles the two: the kanji holds the fixed meaning, and a hiragana tail trailing behind it carries all the changing grammar. Understanding okurigana is what turns a wall of kanji into readable, parsable Japanese.

The kanji stays; the tail conjugates

Take the verb 食べる (たべる, taberu, "to eat"). The kanji 食 supplies the meaning "eat" and reads た (ta); everything after it is okurigana, and that is what moves as the verb conjugates:

毎日、朝ごはんを食べる。

mainichi, asagohan o taberu

I eat breakfast every day. (食べる — plain present)

もう昼ごはんを食べましたか。

mō hirugohan o tabemashita ka

Have you eaten lunch yet? (食べました — polite past)

昨日は忙しくて何も食べなかった。

kinō wa isogashikute nani mo tabenakatta

I was busy yesterday and didn't eat anything. (食べなかった — plain negative past)

Line the forms up — 食べる, 食べました, 食べなかった — and the pattern is obvious: 食 never budges, while the hiragana tail flexes to mark tense, politeness, and negation. Read the tail and you have read the grammar. This is the everyday engine of the whole verb system; the base form the tail attaches to is covered on the dictionary form page.

💡

Okurigana is why you can parse a verb you have never seen. If you meet 泳いだ and know 泳 means "swim," the tail いだ tells you "plain past" before you know anything else about the word. The kanji gives the concept; the kana gives the sentence its tense.

Okurigana pins down which word you mean

Here is where okurigana earns its keep. A single kanji is often the seed of several different verbs, and the only thing distinguishing them on the page is the kana tail. Consider 上 ("up"):

エレベーターで五階に上がる。

erebētā de gokai ni agaru

I go up to the fifth floor by elevator. (上がる — agaru, 'to rise', intransitive)

その荷物を棚の上に上げてください。

sono nimotsu o tana no ue ni agete kudasai

Please put that luggage up on the shelf. (上げる — ageru, 'to raise', transitive)

Same kanji 上, two completely different verbs — agaru (something rises on its own) versus ageru (someone raises something). Written 上る with yet another tail it would be noboru ("to climb"). Strip the okurigana and the sentence becomes unreadable; the tail is not decoration, it is the disambiguator.

The same is true of 生, which you met on the overview page:

祖母は九十歳で、まだ元気に生きている。

sobo wa kyūjussai de, mada genki ni ikite iru

My grandmother is ninety and still living in good health. (生きる — ikiru, 'to live')

私は大阪で生まれ育った。

watashi wa ōsaka de umare-sodatta

I was born and raised in Osaka. (生まれる — umareru, 'to be born')

生きる (ikiru) and 生まれる (umareru) share the kanji 生 and its meaning "life," but the tails きる and まれる make them different verbs with different readings of 生 (い vs う). The kanji carries the concept; the okurigana names the exact word.

Adjectives conjugate on their tails too

It is not only verbs. The i-adjectives inflect through okurigana in the same way — the kanji root holds the meaning, the kana tail carries tense and negation. Watch 新しい (あたらしい, atarashii, "new"):

この店は先月できたばかりで、まだ新しい。

kono mise wa sengetsu dekita bakari de, mada atarashii

This shop only opened last month, so it's still new. (新しい — plain present)

うちの車はもう新しくない。

uchi no kuruma wa mō atarashikunai

Our car isn't new anymore. (新しくない — negative)

やっと新しいスマホを買った。

yatto atarashii sumaho o katta

I finally bought a new phone.

新 stays fixed and means "new"; the tail runs しい → しくない → しかった to give present, negative, and past. The mechanics live on the i-adjectives page — here the point is only that the kanji/kana division is the same as for verbs.

Where does the boundary fall?

The standard convention (codified in Japan's official okurigana guidelines, 送り仮名の付け方) is this: the kanji absorbs the fixed head of the word's reading, and the okurigana begins where the word starts to change. Practically, line up all the conjugations and look at which sounds stay put and which move — the unchanging head belongs to the kanji, and the kana takes over from there so that everything that inflects lands in hiragana.

For a godan (u-) verb like 書く (かく, kaku, "to write"), the kanji 書 reads か and the tail starts immediately: 書く, 書きます, 書かない, 書いた — every changing sound is in kana. For ichidan (ru-) verbs like 食べる, convention sends back one extra fixed syllable (べる, not just る) so the reading and the verb class stay unambiguous. That is why you write 食べる and 見る, but 食べる keeps its べ in kana.

💡

There is no need to reverse-engineer these rules from scratch — dictionaries and this guide's inline glosses give you the correct spelling. But knowing the principle (kanji = invariant meaning-head, kana = everything that conjugates) lets you predict most okurigana correctly and read unfamiliar verbs on sight.

Be honest with yourself that the official rules have a tail of exceptions, and a handful of words have accepted alternate spellings: 行う and 行なう (okonau, "to carry out") are both standard; 表す and 表わす (arawasu, "to express") coexist. When a word looks borderline, check a dictionary rather than guess — but these gray zones are a small minority.

The insight most apps skip

Step back and see what okurigana actually accomplishes. Japan borrowed a writing system built for an uninflected language (Chinese) and used it to write a heavily inflected one (Japanese). That should not have worked. Okurigana is the hinge that made it work: the borrowed character keeps the meaning frozen in place, and native hiragana rides behind it to supply the grammar the characters could never express on their own. Meaning in the kanji, inflection on the kana — that division of labor is the quiet genius of written Japanese.

Orthography reminder: okurigana is always hiragana

Okurigana is never written in katakana — not ever. This matters because a few hiragana and katakana are near-twins: hiragana べ versus katakana ベ, hiragana へ versus katakana ヘ. Writing 食ベる with a katakana ベ is a real and common slip, and it is simply wrong. The rule is absolute: the meaning-root may be kanji, but the inflecting tail is hiragana, full stop.

Common mistakes

❌ たべる (writing the whole verb in kana when the kanji is standard).

Incorrect at N4+ — the meaning-root belongs in kanji; the tail in kana.

✅ 食べる

taberu

to eat — kanji root 食 + hiragana tail べる.

❌ 食ベる (okurigana written with katakana ベ).

Incorrect — okurigana is always hiragana; katakana ベ only looks like hiragana べ.

✅ 食べる

taberu

to eat — the tail べる is hiragana.

❌ 上る for 'to raise something', expecting it to mean 上げる.

Incorrect — the tail decides the verb. 上る is noboru ('to climb'), not ageru.

✅ 荷物を上げる。

nimotsu o ageru

to raise the luggage — the tail げる selects ageru.

❌ 新らしい (boundary too far back, an extra ら in the tail).

Incorrect — the standard spelling keeps 新 = あたら and sends only しい.

✅ 新しい

atarashii

new — 新（あたら）+ しい.

Key takeaways

Okurigana is the hiragana tail on a kanji; the kanji holds the fixed meaning, the tail carries the conjugation (食べる, 食べました, 食べなかった).
The tail disambiguates which word you mean: 上がる (agaru) vs 上げる (ageru), 生きる (ikiru) vs 生まれる (umareru) — same kanji, different tails, different verbs.
Adjectives inflect on their tails too: 新しい → 新しくない → 新しかった.
The boundary convention: the kanji keeps the invariant head of the reading; the kana begins where the word changes — with a small tail of official exceptions and a few accepted alternate spellings.
Okurigana is always hiragana, never katakana.

Now practice Japanese

Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.

Start learning Japanese→