On vs. Kun readings

From LingWiki

Jump to: navigation, search

Kanji are logographic characters which have been borrowed from written Chinese. Japanese has two sets of readings that are in common use: On readings (音読み) and Kun readings (訓読み). On readings are those derived from classical Chinese pronunciations, whereas Kun readings are native to Japanese. Which reading is appropriate to use depends on the context in which it appears. There also exist Na-nori readings, which are reserved for use in names.


When to use On readings

For the most part, On readings are used in compound words formed from multiple kanji. In this way they also function as a system of roots for Japanese, much the way Latin and Greek roots are used in European languages. For example:

Compound words formed in this fashion form a special class of words in Japanese. Like with English, some words formed in this way can be used in a rather utilitarian fashion:

When used as verbs, most words formed from kanji compounds are followed by する (suru), which is the verb for to do. Likewise, if you see a verb formed with a single kanji followed by する (e.g. 関する, kan suru) it should be read by its On reading.

Choosing the right reading

While there is no simple rule for learning the right On reading in any given context for a given character, there are a couple rules and systematic changes that can be useful to remember. One key point is that each distinct On reading for a kanji character is usually associated with a specific meaning.

Only one On reading

Some characters only have one On reading. For example, is always read as デン (den).

Initial syllable changes

Some readings change their initial consonant to fit the ending of the character before it. For example, has an On reading of ハイ (hai), but when preceded by (itsu) it becomes パイ (pai) (thus, the whole word is read as いっぱい (ippai).

Final syllable changes

If the last syllable of the first kanji is つ (tsu), it usually becomes a small tsu (っ) when it's placed before another kanji, as in 一杯 (ippai) above, or in 出発 (shuppatsu):

  • 出発 (shuppatsu)
    • (出) しゅつ (shutsu) + (発) はつ (hatsu) = しゅっ (shu + small tsu) + ぱつ (patsu) = しゅっぱつ (shuppatsu)

Another case where the final syllable can change into a small tsu is when the last syllable of the first character and the first syllable of the next begin with the same consonant. Using 格好 (kakkou) as an example:

  • 格好 (kakkou)
    • (格) かく (kaku) + (好) こう (kou) = かっ (ka + small tsu) + こう (kou) = かっこう (kakkou)

Multiple Possible Readings

Some characters, like , have several possible readings:

  • じょ (jo)
  • にょ (nyo)

Looking through the dictionary though, you aren't likely to find many words that use にょ (nyo); it only appears in words referencing ladies of the courts of the old emperors of Japan. Thus, in most modern contexts じょ (jo) is the default reading.

When to use Kun readings

Kun readings are used in most cases where you see a kanji character on its own, and in all cases where it has okurigana (hiragana characters that add inflected endings). The only exception is when a single kanji is followed by する (suru), as explained above.

Kun readings are usually much more numerous and varied than On readings, however there are a few rules that can be applied to figure out when to use each.


Some kanji have several kun readings, but attach hiragana characters to the end. These characters can be used as a hint to what the reading of the kanji should be. When you look the characters up in a kanji dictionary, the reading of the character itself is written first, with any okurigana following a dot, dash, or similar marker. For example, the entry for might list its kun readings as follows:

What this means is that in the first case the character itself would be read as た (ta), and 〜べる (-beru) would be written after the kanji. Thus, た・べる would be written as 食べる. In the case of the second reading, 食 would be read as く (ku), and would be followed by 〜う (u), producing 食う.

There are three primary ways to use okurigana to choose the appropriate Kun reading for the character:

One base reading, to which okurigana are attached

  • e.g. can be read as やす・らか (yasu-raka), やす・らぐ (yasu-ragu), やす・い (yasu-i), or やす・んじる (yasu-njiru)
    • All of these readings add okurigana to the base reading of やす (yasu). Therefore, we can safely deduce that any time you see 安 and it isn't part of a kanji compound, it should be read as やす (yasu).

Several base readings, to which different okurigana are attached

  • e.g. can be read as しろ (shiro), しろ・い (shiro-i), しら・ける (shira-keru), or しら・む (shira-mu)
    • If the kanji is on its own, or if it is followed by い (i), it is read as しろ (shiro)
    • If the kanji is followed by ける (keru) or らむ (ramu), it is read as しら (shira)

Several readings, some of which have similar okurigana, but can be distinguished by context

  • e.g. can be read as い・り (i-ri), い・る (i-ru), い・れる (i-reru), or はい・る (hai-ru)
    • はいる (hairu) in the potential form is read as はいれる (haireru), written in kanji as 入れる. This could be confusable with the verb いれる (ireru), also written as 入れる
      • はいれる indicates that it's possible to enter something. For example, 私は剣道部に入れる (watashi wa kendou-bu ni haireru) means "I can enter (join) the kendo club", not "I can insert into the kendo club."
      • いれる means to put in. For example, 私はカップにコーヒーを入れる (watashi wa kappu no naka ni koohii wo ireru) means "I put coffee into the cup", not "I can enter coffee into the cup", "I can enter the cup", or any other similarly ridiculous-sounding possibilities.

Repetition Symbol

Occasionally you may see written after a kanji character. This is used to indicate repetition, and is shorthand for writing the character out a second time. When reading it aloud, the first syllable of the second character's reading is voiced if the original reading is unvoiced (thus, if the first character is か (ka), さ (sa), た (ta), or は (ha)、they become が (ga)、ざ (za)、だ (da), ば (ba), and so on). Some common examples include:

  • 時々
    • Equivalent to 時時, pronounced ときどき (toki doki)
  • 人々
    • Equivalent to 人人, pronounced ひとびと (hito bito)
  • 山々
    • Equivalent to 山山, pronounced やまやま (yama yama)
      • Note that the pronunciation did not change for the first syllable of the second character. This is because や (ya) cannot take a dakuten, and thus must remain the same.

See Also

Personal tools