I'm learning a bit of Classical Japanese recently, and of course the spelling of words is pretty different, due to sound changes over the centuries. For example, きょう was spelled けふ. That I can understand, since no language has static sounds.

I then went to YouTube and listened to a Japanese guy explain Classical Japanese. In the first lesson what he did was explain "historical kana orthography" and gave a whole bunch of ridiculous pronunciation rules. Yes, I understand that was how Japanese pronunciation changed over the centuries, but why should we emulate the sound shifts into Modern Japanese when we are reading Classical Japanese? What bad is there in reading いろはにほへと、ちりぬるを as "i ro fa ni fo fe to, ti ri nu ru wo" as it used to be pronounced?

  • 1
    Sorry, I'm not qualified to answer, but I have a guess: because it's easier? Also, sometimes bits of 文語 show up in Modern Japanese, right? Wouldn't it be confusing to learn those with two different pronunciations and switch back and forth?
    – user1478
    Dec 3, 2012 at 19:37
  • 1
    I actually thought it would be harder since you'd need to remember a huge bunch of spelling rules and sound shifts. The 文語 used in Modern Japanese is usually written with modern orthography, so that shouldn't be an issue.
    – ithisa
    Dec 3, 2012 at 21:38
  • Also CJ could be read according to modern spelling without, say pronouncing the ha-line as f-*. Even は could go as wa when used as particle. That would still be much closer to the original pronunciation than retracing all the sound shifts.
    – ithisa
    Dec 3, 2012 at 21:42
  • 2
    One could ask the same question about classical Chinese (why read it in Mandarin!?) or Latin (which is not just a funny way of spelling Italian words)...
    – Zhen Lin
    Dec 3, 2012 at 23:39
  • 2
    The sound change rules for Japanese are almost trivial in comparison to other languages I know about. I can give a complete list without thinking too hard. Also, Chinese is largely phonetic. How do you think they managed to reconstruct Old Chinese, if not by analysing the use of phonetic components in characters? Let me play devil's advocate: from what we know of Middle Chinese phonology – and there are many sources – if there is a modern dialect that is closest to the old pronunciation, it is Cantonese, not Mandarin. So why don't we read it in Cantonese?
    – Zhen Lin
    Dec 4, 2012 at 7:58

3 Answers 3


As you pointed out, there is no single correct pronunciation of Classical Japanese. It would be more accurate to teach different pronunciations used in different periods, but it would be probably too complicated to teach at schools. The pronunciation of Classical Japanese taught at high schools is the newest one used in Meiji period and later. (I do not know if the same is true for the YouTube courses which you watched.)

I do not know why this choice was made at schools. My guess would be that it is more likely to encounter recent text in Classical Japanese than very old text in Classical Japanese.

It is not only the pronunciation that varies over time. The grammar of Classical Japanese is not static. I believe that the grammar taught at schools also follows the newest part of Classical Japanese, but I would like someone who knows better to give a more complete picture of the situation.

  • My question is as much relevant to Meiji-era schools and people too. The "logical" way, for say the Meiji period, in my point of view would be: 1. When old spelling are used in writing modern works as 歴史的仮名遣い, use the messy sound shift system so that it matches speech when read out. (i.e. けふ as "kyou") 2. When old spelling is used in pseudo-old CJ, read as per hiragana chart (i.e. けふ as "kefu") Pseudo-old CJ in laws etc are supposed to be mainly written anyway, so why invent messy pronunciation rules for CJ?
    – ithisa
    Dec 4, 2012 at 6:42
  • This would be like why don't we read "thou" as "you" in modern English and "shouldst" as "should" by inventing rules about th and st.
    – ithisa
    Dec 4, 2012 at 6:44
  • 1
    The problem with teaching the classical pronunciations instead of a system to turn them into modern Japanese is that you have to them a whole new vocabulary, because instead of teaching them a set of transformations that allows them to arrive at きょう from けふ, you have to teach them that けふ means きょう, which may be fine for just one word, but is way less efficient when you have to teach them all the words.
    – rintaun
    Dec 5, 2012 at 3:31
  • 1
    I agree with that, since such a system would also help reading purely modern Japanese works written in rekishiteki kanadzukai, such as prewar novels. However, this brings up an interesting question. Why don't Japanese publishers "update" the kana to modern phonological equivalents, while not translating the old grammar and vocab? Classical Chinese books for students in China often include pinyin (i.e. modern) pronunciation guides for characters pronounced in a different "literary" way (i.e. bo1 rather than bai1 for 白) rather than true period pronunciation like the Japanese do with old kana.
    – ithisa
    Dec 5, 2012 at 23:09
  • 1
    @Eric Dong: I do not think that there is a reason Japanese publishers cannot do that. They could do that if it were an acceptable practice. For whatever reason, publishers are not supposed to replace the historical kana orthography with the modern kana orthography when publishing the classical literature in Japanese. I guess that Chinese and Japanese publishers have different practices. Dec 5, 2012 at 23:26

I realize that this is a very old question and you probably know the answer already, but I'll put it here just for reference.

The answer is actually very simple. Because prior to 1946, historical kana orthography was used for BOTH Modern Japanese AND Classical Japanese.

People who were reading Classical Japanese texts were reading Japanese spelled in the same way they spelled their words in the present. So, they read the Classical Japanese words in the same way with the same pronunciations as they read their Modern Japanese words. The sound shifts and changes we use to pronounce Classical Japanese in the present day are directly based off of the sound shifts pre-1946 people pronounced the Modern Japanese language with. Because up until the 1946 reform, the pronunciations and phonology changed but the orthography stayed the same. As けふ changed into きょう, the people started saying きょう yet they still wrote down けふ, all the way until 1946 until the kana spelling was reformed to directly match the modern phonology (with the exception of は -> わ / へ -> ゑ -> え / を -> お, which are the only remnants of the sound shifts people regularly read Modern Japanese in prior to 1946 that still exist in the language today).

Really up until the reform, nobody found them "weird sound shifts". It was just the way everyone in Japan read and wrote text, that everyone had to learn and everyone used naturally. People didn't really question it until around the time of the reforms. When you were writing [今日]{きょう}, you were writing [今日]{けふ}. 今日 was spelled けふ, and you just pronounced it as /kyou/. You didn't spell 今日 as きょう or きよう, it was けふ. People were educated in the historical kana orthography spellings as they learned how to read and write - because there was no "modern kana orthography" at the time; this would just be the way Japanese was read and written, with this set of sound changes. They would be taught to you as you learned how to read.

It's like how "tough" in English is pronounced like "touf" but we still spell it as "tough". The final "gh" is a remnant of a sound that is no longer pronounced in the word in the modern day, but it was once there, and is the way people have been spelling the word for centuries. You don't really question why it's spelled "tough" or label it as a "weird sound change", you just know it's pronounced "touf".

Historical kana orthography has stuck around for Classical Japanese but is no longer used for Modern Japanese after the 1946 reform. There hasn't really been an incentive to change the way Classical Japanese is pronounced; after the reform, people who studied Classical Japanese just pronounced the historical kana orthography used to write it the same way they always knew - the way they learned and pronounced Modern Japanese written it it with.


First of all, I am also no expert, but I have been looking at classical Japanese orthography recently and noticed that many of the "spelling-change rules" seem to follow the same logic as some modern Japanese's collocations/"slang".

For example the simplifying of words by seemingly merging sounds: わからない → わかんない。If you take けふ and pronounce ふ as hu not fu, and try to "combine/merge" the sounds, to me you can almost hear きょう. I think the language sounds might have changed in a similar way. Just a thought. Again, I am NO expert.

You must log in to answer this question.

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