How has the percentage of writing using each script changed since 1945? With script reform, increasing influx of loanwords, and stylistic choices, several factors affect how much certain parts of the Japanese script are used in writing. Modern Japanese is very different from how it looked in 1945, by any measure.

Taking this into the context of my question, what proportion of words used in writing are written in hiragana, katakana, and kanji? I've been unsuccessful in finding any information on this so far. Keeping in mind that this has nothing to do with whether we agree or disagree with the changes that have happened (after all, language is just communication whatever words you use), does there exist data for something like:

% Words by script in 1945:
Hiragana: .%
Katakana: .%
Kanji: .%

% Words by script in 2015:
Hiragana: .%
Katakana: .%
Kanji: .%

Any sources of written language is fine, whether newspapers (the easiest to find), journals, handwritten letters, websites and so on.

  • 3
    I think you're going to have a hard time finding something that can accurately quantify this since it will vary greatly on the subject material. You can probably find which kanji have come in and gone from the joyo kanji list and extrapolate (probably not actually), but other things will be extremely arbitrary. I could see this happening if literary works were digitized, but Japan is in the stone ages on this front. Interesting question nonetheless. Dec 1, 2015 at 21:30
  • 1
    More importantly you may want to distinguish between 旧字体 and 新字体 for the kanji parameter. What you're looking for is a corpus set that lets you search by date.
    – virmaior
    Dec 1, 2015 at 22:34
  • 1
    I kind of see your point, but my reason for suggesting including 旧字体 vs 新字体 is that I assume you're looking for changes in language use and I take the reforms to have simultaneously affected kanji vs kana % and kanji-type vs kana percentages.
    – virmaior
    Dec 2, 2015 at 1:40
  • 4
    Using words may not be as easy as you initially think. You may have to consider a few potential issues: 1. A word can consist of kanji and kana letters: 走る, 高い, 歩きスマホ, and so on. 2. Do you want to give both content words and function words equal weights? 3. Not everyone agrees on the definition of words. For example, some people treat 静かな as one word while others treat it as two words. These may not matter for your purpose in the end, but just wanted to make sure you are aware of them. Dec 2, 2015 at 18:38
  • 1

2 Answers 2


I believe one can write a program to do this, but you can also estimate usage in the 2010s by using Table 5.9 of this book and assuming that the ratio of katakana to hiragana usage hasn't changed much since 1985 (hiragana remains dominant because most okurigana and particles are written in hiragana, although this assumption may underestimate % katakana considering the influx of loan words since), and with the Kanji usage represented by Figure 1 of this article.

As for the 1945 proportion, kanji and katakana were mostly used based on the Wikipedia article, but because of my limited exposure to these documents, I cannot provide a good estimate.

  • 1
    It's very interesting the idea of figuring out these proportions via software. If we considered the proportions of kanji, kana and katakana character by character, it would be pretty straightforward to write a program that counts them, but the proportion of words, that's another story
    – jarmanso7
    Jun 12, 2020 at 16:30
  • 1
    I agree. When we realize that words are mostly a composite of kanji and kana, how do we do that? What should the algorithm be? Search each word, find its kanji-kana composition, then do a weighted average depending on relative frequency? And parsing is another story--where does a word start and where does it end? Although there are codes written for that purpose, I think.
    – rebuuilt
    Jun 12, 2020 at 23:54

I don't know how to help with hiragana and kanji, but I can help with katakana. The Katana alphabet's word count has increased massively, but there is little empirical data to prove their increase in such writings. The increase of katakana words in Japanese writing poses a problem for foreign learners of Japanese whose L1 is English.

I hoped I helped. :)

This might help with kanji http://www.japanesestudies.org.uk/ejcjs/vol12/iss3/premaratne.html

  • 5
    Use of katakana per se was reduced since then because it was the main kana before.
    – user4092
    Aug 16, 2016 at 13:05
  • 1
    Your answer itself does not seem very clear to me, but the linked article was very helpful.
    – oxr463
    Sep 25, 2019 at 16:45

You must log in to answer this question.

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