Sometimes, when learning kanji, I get tripped up because I can't remember which kana are the furigana of the kanji, and which are the okurigana.

Take たべる(食) for example. I might think that たべ are the furigana for 食, and る is the okurigana, which is obviously not true. I can tell easily enough when the word is a verb, and it ends in る or す or some such, but it's not always easy to remember whether sounds in the middle are part of the kanji or not.

Is there a system behind all of this? Or do you simply have to know the kanji readings and memorise the okurigana?

  • 1
    There is something of a system. For instance, for ichidan verbs, typically the last two syllables are spelled out in okurigana. – Zhen Lin Dec 28 '14 at 0:57


As had been pointed out, as a general rule, that part of the word - in terms of kana syllables - that changes or inflects is written with okurigana.

See also 「送り仮名の付け方『国語を書き表すための送り仮名の付け方のよりどころ」』・単独の語1・活用のある語・通則1」, which states as a general principle that the inflectional ending is added in kana.



When kanji had been introduced to Japan, and when people began to write Japanese with kanji as well, there was the problem that when writing everything in kanji, you had to indicate somehow which inflection of a word was intended. Chinese didn't have these inflections and thus the writing system had to be adapted.

What the Japanese people came up with was (a) the 弖爾乎波【てにをは】 system(s) which consisted of adding small dots and strokes to indicate a certain inflection (e.g. ~たり or ~て) or particle (e.g. に or を); and (b) - like with Egyptian hieroglyphs - a more flexible dual system, some kanji used for their meaning (semantically) and some for their sound (phonetically).

For example, you could write 恐美多利 for 恐【かしこ】みたり, the last three kanji used phonetically. The latter became man'yougana, then hentaigana and evolved into the well-known kana eventually.


Today, there are 五段 ("u", e.g. 泣く) and 一段 ("ru", e.g. 語る) verbs, but historically, there had been slightly different groups. If you take a verb like taberu, its conjugation had been:

終止形【しゅうしけい】 たぶ

連体形【れんたいけい】 たぶ

連用形【れんようけい】 たべ

未然形【みぜんけい】 たべ

已然形【いぜんけい】 たぶれ

命令形【めいれいけい】 たべよ

That's called the 下二段【しもにだん】 conjugation; 二段 because the vowel changes between e and u (cf. the ablaut in German singen-sang-gesungen-Song, and English sing-sang-sung-song or tooth-teeth), 下 because these vowel come last (=at the bottom in vertical writing) in the Japanese order a-i-u-e-o.

二段 verb became 一段, getting rid of the vowel change, and the 終止形【しゅうしけい】 fell out of use, leaving you with only a final る that changes during inflection. But as a remnant of Classical Japanese, the correct spelling with okurigana is 食べる instead of ✘食る, which would be analogous to 見【み】る. The same principle holds for many other verbs.

The same story holds true for 起きる・起こる, 掛かる・掛ける and many other pairs. They are usually derived from some earlier form of one (shorter) verb, e.g. 掛【か】く.

One reason why new verbs such as 掛かる and 掛ける were formed from an existing verb 掛く was to create transitive/intransitive pairs, or potential/causative verbs such as 見【み】る・見える or 覚【さ】める・覚ます, which gives the explanation for this pattern detailed in the answer by @fosskers.

Furthermore, okurigana help to distinguish between the two words, 起る could theoretically be either おきる or おこる.

For 五段 verbs, historically 四段 (=four vowels; a, i, u, e) verbs, only the last syllable changes, thus 食【は】む, 蝕【むしば】む, 読【よ】む 坐【ましま】す, 嗤【わら】う, 貪【むさぼ】る etc. The inflection of 読む is, for reference:

終止形 よむ

連体形 よむ

連用形 よみ

未然形 よま

已然形 よめ

命令形 よめ

To illustrate this, consider the two verbs 統【す】べる and 滑【すべ】る. 統べる derives from the 二段 verb 統ぶ (inflection: すぶ・すぶる・すべ・すべ・すぶれ・すべよ), thus its non-inflecting stem is す and it is written 統べる. On the other hand, 滑る is a perfectly normal 四段 (now 五段) verb on る (inflection: すべる・すべる・すべり・すべら・すべれ・すべれ), whose non-inflecting stem just happens to be すべ, thus it is spelled 滑る. In classical Japanese, these two verbs were referred to in their dictionary forms すぶ and すべる and were as different as 行く and 言う, which just happen to exhibit the same past form いった.

Verbs derived from other words retain their inflectional endings. 怪【あや】しむ・怪しがる and the adjective 怪【あや】し, 悲【かな】しむ and 悲し, 苦【くる】しむ and 苦し.


The story continues with adjectives. As for i-adjectives, the conjugation is as follows for 赤【あか】い:

終止形 あかし

連体形 あかき / あかい

連用形 あかく / あこう

あこう is a dialectical variation (and used before ございます, as in 有【あり】難【がと】うごさいます, from 難【かた】い) and arose from あかく→あかう→あこう via a sound shift. Thus, only the last syllable is inflected and the proper spelling is 赤い, 赤う, 赤く.

There is a variation of this conjugation for adjectives whose last syllable before the inflecting one is し, e.g. 美【うつく】しい:

終止形 うつくし

連体形 うつくしき / うつくしい

連用形 うつくしく / うつくしゅう

The final inflectional し is not added for the 終止形 (dictionary-form), thus the proper spelling is 美し or otherwise it would not get any okurigana to indicate it is an adjective, thus it is written 美しい and 美しく.

The same applies to the compound conjugation of adjectives. 赤くない, because that is 赤く + the word 無い. 赤かった, because that is 赤し+あり+つ+あり→あかかりたり→あかかった.

Some adjectives that are formed from an inflected form of another word retain that inflection, e.g. 羨【うらや】ましい, derived from the verb 羨【うらや】む, plus し + the adjectival inflectional ending い; or 頼【たの】もしい, from 頼む.

Compounds & misc.

Next are compound words whose elements are buried under etymology and may have become obscure. The suffix is rarely written with kanji and left in kana, which may look like okurigana.

The proper spelling is 速【すみ】やか, 艶【あで】やか, 健【すこ】やか, 雅【みやび】やか, 明【あき】らか, 滑【なめ】らか, 朗【ほが】らか, 安【やす】らか etc. because they are compounds containing the suffices -やか and -らか, e.g. 艶【あで】+やか and 安【やす】+らか.

静【しず】か, 愚【おろ】か, 確【たし】か, 豊【ゆた】か, 僅【わず】か consist of a combination with some suffix -か. The above -やか might be a combination of や+か.

See this question concerning the three suffixes -ら, -や, -か.

嬉しげ, 在【あ】りげ, 大人【おとな】げ, 可愛【かわい】げ etc. contains the suffix -気【げ】, now often written in kana only.

古【ふる】めく, 唐【から】めく, 春【はる】めく ひしめく, ざわめく etc. contain the auxiliary verb めく ("appear/look like").

欲【ほし】しがる, 強【つよ】がる, 度【た】がる etc. contain the auxiliary verb がる "show signs of".

昂【たか】ぶる (高【たか】+ぶる), 偉【えら】ぶる, 揺さぶる etc. contain the auxiliary verb ぶる (振【ぶ】る) "pose as, pretend to".

綺麗【きれい】の, 綺麗【きれい】な because they are compounds containing the particle の and the verb (as a copula) なり shortened to な.

全【ま(っ)た】く, 全【すべ】て, 決【けっ】して, 妄りに because they are derived from the adjective 全【また】し, the verb すべる, the Sino-Japanese (on-reading + する) verb 決【けっ】する, and the verb 乱【みだ】る, respectively.

曰【いわ】く (from 言う), 見【み】らく, 散【ち】らく etc. because they are all formed with some old element く.

No okurigana

Some nouns derived from verbs can be written without okurigana, e.g. 話【はなし】, 取調室【とりしらべしつ】, 折紙【おりがみ】, 振仮名【ふりがな】, 受付嬢【うけつけじょう】, 立入【たちいり】禁止【きんし】. Sometimes, the okurigana other than for the last verb are omitted in verbal compounds, e.g. 問合【といあわ】せ, 振舞【ふるま】う, 受付【うけつ】ける.

Exceptions / Special cases

Some words are written with additional okurigana: 後【うし】ろ, 蔑【ないがし】ろ (from 無【な】きが代【しろ】). Sometimes you can find variant okurigana spellings such as 既【す】でに, 故【ゆ】えに and 此【こ】れは (instead of 既に, 故に, and 是は).

Additionally, おこなう used to be written as 行う (行ふ), but the spelling 行なう has gained in popularity.

Thinking about おこなう, I would suggest that the main motivation behind these okurigana usages is to help the reader and to avoid ambiguity. 行って could be either おこなって or いって (or ゆって), but 行なって must be おこなって. Likewise, 後 could possibly be のち, ご or うしろ with only the context left to decide which, but 後ろ must be うしろ, no argument about the context required.

既 or 蔑 might be hard to read for some people, but giving part of its reading as 既で or 蔑ろ helps you to guess the word from context, even if you didn't know the kanji at all.


There are explanations for these pattern regarding okurigana usage, and they require some basic understanding of the earlier stages of the language. However, that doesn't change the fact that there are patterns, and even without understanding them in full detail, you will soon start to notice them and be able to tell how to spell a word. For example, you don't need to know about the history of the suffix -やか to notice that many words ending on -やか are spelled with やか as the okurigana.

  • 3
    Regarding remembering this: When something seems random to me and I don't understand why, I've got a hard time memorizing all that when it doesn't make any sense. But when I've read an explanation how all of that fits together and that there's some logic to it, I get rid of those negative emotions associated with it and memorizing becomes easy and fun, even when I can't claim I understand the explanation fully. – blutorange Dec 28 '14 at 9:47
  • Haven't yet had time to read this in detail, but looks like an incredible answer. – Lou Dec 28 '14 at 12:15

You can think of Furigana and Okurigana as the root and stem of the word respectively.

Do you know what transitive/intransitive verbs are? Unlike English, nearly all verb meanings have a pair of these. In case you don't (or some future reader doesn't), transitive verbs can (sometimes must) take a direct object (using を). They are actions you do to something. Intransitive are the opposite; actions that you yourself do. Some examples in English first:


  • I ate an apple.
  • I ate.
  • I cut my finger.
  • I cut it.


  1. to eat is transitive, but it doesn't always need an object.
  2. to cut is also transitive, but it must have an object. We can't say I cut. In cases like this we naturally feel compelled to add an it.


  • I fell.
  • I died.

Notice it makes no sense to say "I died him", etc. to die cannot take an object.


This is a nice feature of English. We reuse many verbs as both transitive and intransitive.

  • I woke up.
  • I woke him up.
  • I ran.
  • I ran the marathon.

Japanese makes extra work for us by having distinct pairs of these words. Example (transitive on left):

  • 起こす・起きる
  • 切る・切れる
  • 落とす・落ちる

Notice the furigana remain the same regardless of which form the verb takes. When studying it's best to learn these in pairs, and eventually you'll know enough that you'll be able to make decisions on-the-fly about how they need to be written. No memorization necessary.

Say you forget how to split おとす and you'll say to yourself: well, its pair is おちる, and since お is what stays the same it must be the furigana. 「落とす!」

Hope this helps.

  • Informative answer, but verbs aren't the only words that can have okurigana. い-adjectives and nouns can have them as well (in the case of い adjectives, they always have okurigana, and often there's an additional character besides the い in the okurigana). – Ataraxia Dec 28 '14 at 5:23
  • Okay. I wonder about words like 教える, though, which doesn't appear to have an intransitive equivalent. Does the え have any inflectional function here? – Lou Dec 28 '14 at 12:09
  • @LeoKing There's the incoming/outgoing pair 教わる・教える, although both verbs are transitive. – snailboat Dec 28 '14 at 15:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.