Onyomi reading of kanji

Japanese
#1

Hi fellow learners :slight_smile:

I’d like to get a better understanding of why there are often several readings for a single kanji, and especially why the onyomi reading even exists.

The japanese language borrowed the kanji from chinese rather than inventing its own symbols (ignoring hiragana and katakana).

Therefore it makes sense to me, that a single kanji is pronounced differently when used for different meanings. E.g. the kanji 相 is pronounced そう in 相談 where it has the meaning ‘mutual’, but しょう in 首相 where it has the meaning ‘minister’.

But there are a few things which I don’t understand.

  1. Why was it necessary, to also borrow the pronounciation from chinese (onyomi)?

  2. Was there no onyomi pronounciation in japanese before kanji symbols were adopted?

  3. Are there also multiple readings for one kanji in chinese?

  4. From a linguistic/historic point of view, how did jukugo arise?

Thanks in advance for your help :slight_smile:

3 Likes
#2

Really good questions, following for the answers and discussion.

#3

Quick preface: i use the word ‘chinese’ as a catch-all for all chinese languages in this post. It would be a bit complicated if I were more specific. The dominant language in china changed with the dynasty in charge. It’s a complicated history. Even today there are upwards of 10 language families that can rightfully be called ‘chinese’

  1. It was not initially used to transcribe Japanese. Rather, Chinese was used for official documents. It was not until later that people started using Chinese characters to transcribe Japanese. This was around the 9th century, iirc, so around 400 years after the introduction of chinese writing.

  2. Nope. On-yomi is just the chinese reading of the character. There was no writing system prior to kanji.

  3. There are not multiple readings in chinese, unless you count how the different languages pronounce them. Kanji have both a phonetic and logographic component. This is only sometimes useful i. japanese because there are so many readings for the same phonetic components. In chinese, you can generally blind-read hanzi if you know the phonetic component. Every kanji can have up to, i believe, three different on-yomi readings, because there were three major periods where kanji and loan words were imported from china. The dominant language in china was different each time they did this.

  4. This has roots in Kanbun (a way of annotating chinese writing to be intelligible in Japanese) and Bungo (a highly literary court version of Japanese that heavily used chinese words and readings which remained unchanged for almost a thousand years.) They stopped writing official documents in Bungo after world war two.

5 Likes
#4

The Chinese reading of a kanji is referred to as the “on” reading and the native Japanese reading as the “kun” reading. “Kun” reading would be their pronunciation before symbols were adopted.

Kanji were reintroduced to Japan several times throughout its history with 3 generally distinct periods/regions. So the “on” reading could be broken down into general groups that include:

  • Go-on (Wu dialect of southern China, 5-6th century)

  • Kan-on (Tang Dynasty, 7-9th century)

  • Tōsō-on (Hangchow dialect, 14-16th century)

image
image1528×614 122 KB

Source: Japanese by Soichi Iwasaki

2 Likes
#5

Also, you were curious about 相. I actually remember looking this one in a big huge paper kanji reference up at one point, curious which one was the original meaning. That one strangely did not have that noted. Most of the other entries marked one if the senses as the original. Wikipedia had my back, though. The original meaning was ‘Look’. Eventually the meaning shifted in chinese to ‘compare’, and the different senses in which it’s used in japanese (minister, mutual, aspect) come from this shift.

#6

https://www.kanjidamage.com/kanji_facts

Quotes from Kanji Damage

  1. Japanese traders needed to communicate with their Korean and Chinese counterparts. For every existing Japanese word (what we today call the ‘KUNyomi’ word), they tried to find the corresponding Chinese kanji, and pair them up. Furthermore, they decided to use the Chinese pronunciation of the words, too, but got it wrong … It’s kind of like Canada: everyone in Canada has to learn Quebeqois French, even though real French can’t understand Quebeqois-French

  2. yes

  3. China has hella different dialects. So one Japanese trader would come back from Shanghai, where they pronounce 青い (blue) as SEI, and he’d teach everyone in his town to say SEI. Meanwhile, another Japanese trader would come back from Hong Kong, where they pronounce 青い as SHOU, and he’d teach everyone in HIS town to say SHOU. So there’s that.

  4. see 1

#7

Seems like they’re uhhhhhhhh playing a little fast and loose with how it went down.

1 Like
#8

As I understand it, it was necessary because they were also adopting new words they didn’t have before. If kanji compounds could be used to represent a concept that already had a native word, then the imported word often became a more formal one.

I had a huge conversation with ChatGPT about this the other day. E.g. why do we have both 減法 and 引き算 for subtraction. Supposedly the first one is used more in academic contexts, the second is more common speech. And here’s where it gets wild. There’s a parallel with English and Latin. E.g. (IIRC) we had the word “ask” from the Germanic path before we got the fancier word “inquire” from Latin. There were many other examples (like “minus” vs “subtraction”).

Here is a link to the entire conversation, I think it was pretty interesting: ChatGPT - 減法 引き算の違い

#9

We didn’t get the word ‘ask’ from german. We got it from English. English is Germanic. We also didn’t get ‘inquire’ from latin (directly). It came from French, just like pretty much all of our latin words.

There are some parallels between english and japanese with the different roots, particularly in how scholars tried to re-latin-ify our french words in the 15th century, but fundamentally these are different things.

English was a synthetic (note: this does not mean fake or constructed) heavily declined (note: does not mean diminished) SOV germanic language until the Norman invasion. What happens when two peoples who speak very different languages start to live next to eachother is that the dominant language changes to become more learnable for the newcomers. It stops being synthetic and becomes more analytic (this means that grammatical case is marked by word order rather than case-marking particles or declensions, which is easier to learn than a bunch of declensions, and in general the language becomes much more reliant on word order to confer meaning) and it adopts many of the words and grammatical features of the newcomer’s language (like SVO word order).

This is all to say, the acquisition of latinate words into english was accompanied by the fundamental altering of our grammar at the most basic level.

This did not happen in Japan.

Bungo proceeded on its own track as the language of the court and samurai, while spoken japanese proceeded in a much different direction (multiple directions, really). Spoken Japanese among regular folks didn’t really use kango words at all, aside from fully lexicalized, primarily religious words. And while the spoken language developed and evolved as all languages do, and にて became で, なり became である and its contraction だ, the endings to all the verbs changed, etc etc etc, Bungo was locked in with Heian-period classical grammar. It was not until Edo and really Meiji where the wall between these two forms fell. Bungo borrowed heavily from Chinese but these words were primarily full-on loan words. Then they opened the country up, and had to come up a bunch of new words for all the different stuff they were learning about. This is when the wasei-kango word was born. That is, japanese-created on-yomi compounds. Those are pretty new. Relatively speaking.

In english, french derived words are fully lexicalized. IE, the language treats them as it does any other word. Kango words are generally not fully lexicalized. This is why so few on-yomi compounds are affected by rendaku. There are special grammar rules for them (ご instead of お in keigo, する instead of 一段 or 五段 verb endings).

So, yes there are some parallels. But that comparison is extremely limited and potentially misleading. Japanese never adopted chinese grammar because they didn’t adopt chinese words to speak with chinese people, but to speak with eachother.

(Anyway, apologies for going on a tear here, I admittedly was kind of just looking for an excuse to post a bit more in depth about this, and your post was it)

1 Like
#10

Fair enough, my wording, along with my knowledge there, were imprecise. My point was that the two words “ask” and “inquire” came from different paths, and have different nuances attached to them today.

It’s not misleading to say “hey look here is one interesting parallel” which is exactly what I was doing. It was intended to be limited. I made absolutely no claims about grammar or any of the other things you’re pointing to. I was simply talking about a few dozen vocabulary words. Apologies if it misled anybody somehow.

And in an actual different direction: something else interesting I remember reading somewhere was that before the Chinese came along, Japanese often didn’t have a way to formalize categories of some things. E.g. they had words for apple, orange, pear… but not a word for fruit. So if they wanted to synthesize the concept of “fruit” in a conversation, they’d have to say “oh I have a bucket of those edible things that are like apples or oranges or pears”. Supposedly Chinese brought a lot of new abstract conceptual words, which is probably another reason why onyomis got used in those cases. I’m sure I’m remembering this wrong, please feel free to fine tune me again if you know better. :grin:

#11

I just see the claim that ‘oh it’s just like latin and english’ a lot and people end up coming away with weird ideas about japanese. Seen people start ascribing chinese grammatical features to japanese and stuff like that. Saying it’s fundamentally monosyllabic and stuff.

And, as I said, I wasn’t really responding to your post, more just, like, using it to talk. I think for those that are curious, it can make learning the language a bit easier if you understand why it’s Like That.

Also, I can’t really say for sure about your edit, but I have a hunch that it might be because of 動物? But くだもの is a wago (ie japanese-native) word. And けもの is the native word for animal.

1 Like
#12

Interesting, it was a first for me.

The only other parallel I ever noticed is the calendar thing, i.e. China and Latin both got their days of the week from Babylon, IIUC, and passed that logic onto both the Eastern and Western worlds, which is why in these completely unrelated language families, we call all 7 days of the week essentially the same thing. That blew my mind when I first noticed it. I’ve spoken English for nearly half a century and never noticed that “moon” and “month” and “Monday” are all related.

#13

I am no expert in the origins of readings and the history of Japanese.
But when I try to explain to somebody that there are different readings,
I use the following made up analogy.

If English were written in Kanji:

The word for 水 would be “water” in English.
In compounds we often don’t use the “reading” water, but we might use
the greek “hydro” or the latin “aqua”.

hydrostatics, hydroelectric, hydrocarbon, etc.
aquaculture, aquatic, aquarium, etc.

But there are still words like “waterproof” and “watermark”.

I guess most of those words will be written with 水 in Japanese. Often read すい (like aqua, hydro) and not read みず (like water). But you would not call water “hydro” in English. But “aqua” might be correct or understandable in many languages.

#14

To give a different type of answer; it clicked for me when someone reminded me that all languages start as a spoking language FIRST and become a written language SECOND. (And this is especially true for Japanese, of course.)

This means that the kanji didn’t get multiple readings - but instead different spoken words received the same kanji.