Some languages, but not English, have regulators such as the Académie française (French Academy). Amongst other things, it decides whether or not English words such as email, software and ウォークマン ought to be in the French language, or should be replaced with French-derived terms like courriel, logiciel and baladeur.

Wikipedia says the Agency for Cultural Affairs (文化庁) at the Ministry of Education of Japan (文部科学省) plays a role in language regulation. Previous questions have mentioned that the government has deprecated hentaigana and certain kanji. Does any other form of language regulation occur, either by the Japanese government, or organizations theoretically independent of it?

3 Answers 3


文化庁 also publish guidelines on okurigana, the writing of gairaigo, and the use of romaji. Those can be seen here. Other documents here include guidelines on the use of punctuation, iteration marks, etc.

There is also 国立国語研究所. Although they're researchers not regulators, they have produced documents on how to reword difficult-to-understand gairaigo using kanji terms (「外来語」言い換え提案).

Finally, the Japanese Industrial Standards Committee (日本工業標準調査会), through the JIS X 208 and JIS X 0213 standards, has a big influence. If you can't (or can't easily) enter/display certain characters on your computer/phone, that's going to affect usage, even if you don't really care about government standards and guidelines.


This is a partial answer specifically about government kanji lists (prior answers did not go into detail).

MEXT (文部省) and MOJ (法務省) publish official kanji lists (漢字表) like 常用漢字表 (MEXT) and 人名用漢字票 (MOJ). The former is a list of about 2000 kanji "for every day use" (常用) that schools teach and children are expected to learn in school. The second is a supplementary list of about 1000 kanji "for use in names" (人名用), that are not judged to be particularly common outside use in names. The union of these sets represents kanji that are "legally recognized" (can be used in legal documents without issue, etc).

Note that native speakers generally learn to recognize/read many kanji beyond what is in the government lists (the lists account for about 3000 "official" kanji). (See note below on numbers beyond "government recognized" kanji)

MEXT further divides the 常用 list. A subset of 1006 kanji is called the 教育漢字表, or "educational kanji list", which is further separated by grade level and is taught in elementary school. First graders learn 80 kanji (plus hiragana and katakana). Second graders learn another 160, third and fourth graders learn 200 a piece, fifth graders learn 185, and sixth graders learn 181.

These lists are not the first versions. The government has published kanji lists in the past, but the older ones have been replaced by the newest iterations.

Aside: How many kanji are there, and how many do native speakers know? It is hard to give an estimate. To give an idea of how many kanji exist that are relevant to Japanese, Wikipedia suggests that a dictionary containing 50,000 unique kanji is considered comprehensive (but of course many of these are outdated or specialist-only), and that JIS standards call for computer encodings of 13,000 kanji, with the main range including 6,000. I queried the JMDIC data set, which contains spellings of Japanese words representing close to 6,000 kanji, with close to 3,000 of these being in words marked "common". The Kanji Kentei Pre-1 and 1 levels have low pass rates even among adult native speakers, but they require very in-depth reading and hand-writing knowledge of ~3,000 and ~6,000 respectively. To me it seems reasonable to say that a typical native speaker will have some familiarity with readings or meanings of between 3,000 to 13,000 kanji, but a more precise number would be very sensitive to a definition of "familiarity" and background of the native speaker.

  • 1
    "some familiarity with close to 10,000 kanji" do you have a source? (10,000 sounds really huge) or at least what do you define as familiarity? (seen once in a lifetime?) I am asking because 漢検1級 is barely 6,000 kanji and the success rate is really low (and I don't think it is only due to ことわさ and 四字熟語) Jun 11, 2016 at 14:39
  • "some familiarity" was used rather loosely, as in, if they saw the kanji, they couldn't necessarily read it, but they at least didn't feel like "huh? I've never seen that in my life before", and they could maybe take a reasonable guess at meaning with enough context. I'll see if I can dig up a source, it's been a while. Jun 11, 2016 at 14:59
  • On a side note: japanese.stackexchange.com/questions/11123/… Jun 11, 2016 at 15:03
  • Couldn't find the source I had in mind, unfortunately. en.wikipedia.org/wiki/Kanji says that Dai Kan-wa Jiten contains 50,000 and that intersection of various JIS sets contains 13,000. I queried JMDIC for number of unique characters in "spellings" of Japanese words, and got ~6K overall, or ~3K if limiting to "common words" (JMDIC is English-Japanese, so smaller than a Japanese-only set). Somewhere in ~3K to ~13K range but hard to say for sure, and as you said, very sensitive to the definition of "familiar". Kanken1 asks for deep knowledge though. Jun 11, 2016 at 17:00

Well the government can and does regulate the words that and characters that can be used in both official documents, government signage and legal names.

They also set the minimum education standards for Japanese language. So kids have to learn kanji and learn it in a specific form.

While they can't go and take your book of out circulation or take down your store's signage if it doesn't match their standard, they effectively ensure that future generations will tend to favor not using characters and words that the government considers obsolete or useless.

You must log in to answer this question.

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