I've searched for an answer to this question online but all I can find are articles discussing when to use one reading over the other based on how you are referring to your dad or someone else's dad. What I am looking for is some rule or guideline (if such rules exist) that I can use to figure out the reading for kanji given the surrounding prefixes and/or suffixes, not just for this kanji but for others as well. So really this is a question about figuring out readings for kanji and not about "how to refer to my dad". Is chichi an exception to some rule I don't know about. Please help!
by JesseAlexandro
21 comments
It’s actually all just memorization and context clues. Have fun!
Because Japanese just don’t give a fuck.
You will just have to memorize each reading for each vocabulary separately. There isn’t any rule you can use to figure out which pronunciation goes with which vocabulary reliably, just familiarity with the words.
Google onyomi and kunyomi. In short: kanji pronunciation is a mess because it is a writing system designed for Chinese that was forced on Japanese, a language that is very different to Chinese. Therefore most kanji have a Japanese (kunyomi) and one or more, well, let’s say “similar to Chinese” pronunciations (onyomi). There are nightmare kanji like 生 that have A LOT of pronunciations.
Similar to English with its messed up spelling, you only can guess the pronunciation of a word that is written with kanji. Basically you have to learn the pronunciation of every word individually, just like with English. The writing system is too irregular.
Fun fact: the kun (訓) in kunyomi (訓読み), is actually the onyomi reading of that kanji. 🤪
There aren’t really rules like that. It’s better to think of words as being primarily spoken, and the kanji just being a messy attempt at spelling Japanese using a foreign writing system imported from ancient China. They had two different words for father, titi and toto, and they wrote both as 父. The first one later regularly became chichi and is still spelled the same, while the other one had an honorific o- and -sama added to it, and later the pronunciation changed somewhat irregularly from ototosama to otōsan, hence the spelling お父さん, recognizing the historical structure of the word.
You’ll run into a myriad other situations like this, and it’s best not to think of characters being given different pronunciations, but of different words being spelled with the same character because kanji were adapted to represent meanings. This is why e.g. 日 has four different readings nichi, jitsu, hi and ka, plus whatever it phonetically represents in 今日: there was a native word pi (that became hi), there was a native morpheme -ka that is still used in numerals counting days, and there are borrowings from Chinese from different eras and areas, bringing the pronunciations nichi and jitsu in Chinese-formed compounds. They all mean “day” so whatever word was formed using them will use the same spelling, and the pronunciation depends on which word was used.
What I’d recommend going forward is to not learn kanji readings on their own, but rather as part of a word. In this case 父 isn’t actually read as とう, but rather 父さん is read as とうさん. You’ll never see that reading outside of (お)父さん, so remember the reading as part of that word.
This applies not just to this word, by the way. 食 isn’t read た, rather 食べる is read たべる and you’ll never see that reading without at least the べ there as well.
(I do not speak English so I use DeepL to translate… Oops, since that couldn’t be used for translation, I had no choice but to use ChatGPT. I want to clarify that this is not text automatically generated by AI.)
Your question made me curious as well. I did a quick search, but please understand that this is merely a summary from a web search and that I have not confirmed any academic basis for it.
To explain the rule of change first, when honorifics like ～さん/様 or polite forms お～ are used, it becomes とう. In most other cases, it remains ちち. The exceptions might be old honorific expressions like 父上(ちちうえ) and 父君(ちちぎみ). (I couldn’t think of any other exceptions)
Originally, 父 seems to have been pronounced as ち or ちち, or something similar. This trace can also be found in 古事記. Over time, this pronunciation gradually changed to てて, and eventually became とと. This sound was originally considered a childish term, but it seems to have become a commonly used word.
This is something I heard before, but within Japanese families, forms of address tend to align with the perspective of the youngest child. In other words, parents refer to themselves as お父さん/お母さん, and in family conversations, they refer to their own parents as お祖父ちゃん/お祖母ちゃん. This suggests that since とと was originally a childish term, it became established and came to be used in お父さん.
the general pattern is that onyomi is used in “kanji compounds” (words with more than one kanji right after another) and kunyomi is used elsewhere, especially when okurigana (hiragana that immediately follow a kanji and are part of the reading of the kanji) are present. unfortunately, this is a very loose rule, and there are many, many exceptions.
by and large, you will not be able to “figure out” the on/kun readings of a kanji you have never seen before, nor will one help you determine the other. some kanji have components that hint at their (onyomi) reading, but these are generally more “complicated” ones that you will encounter infrequently early on. for example, these kanji can all be read as “retsu”: 列 裂 烈
finally, certain words/compounds have readings that were assigned to them. these readings do not correspond to any of the readings of the component kanji themselves. this practice is called “gikun” and is relatively common.
otousan is one such case. 明日 being read as “ashita” is another example. these are truly just memorization.
good luck!
There is a rule of thumb. When you see a word with only one Kanji, or one Kanji with hiragana beside it, the reading is likely to be kunyomi. Eg: 車 (kuruma). If the word is made up of two Kanjis without any hiragana beside them, the reading is likely to be onyomi. Eg. 電車 (densha).
So many new learners want a language to make sense. It doesn’t, and it’s constantly changing, these days it changes daily with the internet.
It’s not a math problem. There is no perfect formula, every single rule has exceptions. Try not to get hung up on why. It just is, the sooner you learn that, the easier things will be.
You’re being your own road block.
My sister once had a linguistics professor who had words of wisdom on this, and many other similar things:
“Language…. is arbitray.”
It’s read as tou because that’s how someone decided to do it many years ago. Or おとう was just a word people used somewhere and since it meant father they used that kanji to write it. Or who knows? It’s fun to track down the etymology of words and I’m into it, but don’t let it be an impediment to your learning either.
Kanji will have pronunciations you don’t expect sometimes when used in some combinations. That’s just how it is.
Unlike what others are saying, there are actually hints on how to read kanji.
First of all, the vast majority of kanji commonly use only one kunyomi or one onyomi. The other readings are obscure, so you don’t really have to worry about them. Just remember the special cases as vocabulary words when they come up in your studies.
If you see a kanji by itself or with hiragana, it’s generally read with the kunyomi. If it’s with another kanji, it’s generally read with the onyomi.
Also, kanji are often made up of multiple smaller kanji put together, and something like 70-80% of kanji are read the same way as one of their smaller parts. For example, 五 is go, but so is 語 吾 悟. 古 is ko, so is 湖 個 故 胡.
When I first learned Japanese, I never studied individual kanji readings. I just studied vocabulary words. If you learn enough words, you’ll begin to see the patterns and you’ll be able to guess the readings of unfamiliar kanji words.
the reading is in the 付表 of the official 常用 list.
Interesting question. As someone has already pointed out, it all depends on the context. As a native speaker, when I come across words related to “父” (such as 父親, お父さん, 父上, 父の日, 父方, etc.,), I instantly recognize them as a single unit and don’t consciously think about whether they use on’yomi or kun’yomi readings. I suppose that for those who are not familiar with Japanese, it’s common to read kanji characters individually, considering which reading to use, on’yomi or kun’yomi. However, in practice, native speakers read in units of words, rather than individual characters, without thinking about on’yomi or kun’yomi. It’s more like “seeing” than “reading”.
Good luck learning all the ways to read 生、上、下、重、長、行、当、and others that have around 15-20 different ways to say them.
Basically it’s pure memorization. We like to think of kanji as words, but really they still act a lot like letters – you need to put letters together in a specific way to get a specific meaning. Japanese is less free form about that than english, but it still matters.
there are no guaranteed rules, one does not read by memorizing character pronunciations, one reads by memorizing words. there are patterns but also tons of exceptions. every word must be memorized, along with it’s spelling, pronunciation, meaning, and usage, just like english. and just like english there’s concepts simialr to greek and latin roots, but also just like english, knowing them won’t tell you what the word means for sure or how to pronounce it for sure, it’s just added context.
Does there need to be a reason? Sometimes languages just do things
It’s just one big chaos, there are so many exceptions to the ‘rules’ for kanji that there is almost no point. You actually gave a great example with the とうreading, since you wouldn’t always see it in a kanji explanation. But it’s a very common word so you will see this reading all the time for 父さん。
It’s better to just accept that kanji are a way to represent words. And multiple different words can use the 父 kanji because they all mean father.
Sometimes while learning more kanji you will randomly come across a word that will confuse you again, because the reading is so unexpected. But the good thjng is, this happens to Japanese people too!
There is no reason. Why do you add an S in English when there is more than one of something. Because. That’s how it’s done. Same thing for Kanji. Sorry bro
I have bad news… in this case and in general with Japanese grammar (maybe the grammar for every language), there is no 100% rule for how to use words or read them. There are rule of thumbs and maybe exceptions only occur a fraction of a time (I have no clue of that percentage), but you’ll encounter those exceptions enough that the rule itself feels unreliable.
To illustrate this, here’s an example:
A rule of thumb could be “if you see kana, you’ll read it kun-yomi”, if it’s a pair of kanji with no kana, you read it on-yomi”. So お父さん is おとうさん, there’s kana in the word, and 和牛 is わぎゅ as it’s two kanji. Fits the rule so far. But then you have 母親 which is read ははおや rather than something like ぼしん. See this [comment thread](https://www.reddit.com/r/japanese/s/XOaUWzjcSi) which is where I first noticed this. I have no clue why 母親 (and 父親) use their kun-yomi, nor do I know deeper rules that encompass these exceptions, nor do I know all of the exceptions. All I can hope for is to see/hear it used enough to *just know*.