Does anybody have Kanji-list with the frequencies of its Onyomis?
Something like that:
京　ケイ　　２５
京　キョウ　１２
I guess it woul be pretty easy to write a programm for that, but I have no idea how to extract the meaning for a single kanji from a compound.

Thank you much for your help!

Does that data even exist? You can't just look at how many words use the reading because that ignores the frequency of the words themselves. As you learn Japanese you'll get a feeling for which readings are rare and which are common.

The data could be approximated using one of the compound frequency lists, but it's a lot of work for pretty much no benefit and wouldn't give very complete coverage.

Looking at the list of readings given in the jouyou list is good enough I suppose, since they are the more common of the readings.

Learning readings in isolation is a waste of time though, so you really shouldn't bother.

that advice comes too late, because I learned them already ;P

when I learned the rtk1 kanji, I incorporated the ON-readings in the stories, and it worked very well for me. But when a Kanji had 2 or more readings, I always took one reading as the main reading and the others where just side readings. Unfortunately I sometimes took an obscure reading as the main reading.

I would like to learn the rtk3 kanji the same way, that's why I was asking for the list.
Thank you for the tip with the jouyou list, is there a similar list, that includes the rtk3 Kanji?

yudantaiteki Wrote:Does that data even exist? You can't just look at how many words use the reading because that ignores the frequency of the words themselves. As you learn Japanese you'll get a feeling for which readings are rare and which are common.

It doesn|t have to be perfect. I was thinking about something like that:
脚 can be read kyaku or kya.
Search the 脚 in a Kanji dictionary.
there are 42 compounds for 脚.
in 38 compouns 脚 reads kyaku
only in 4 compunds 脚 reads kya.

For other Kanji the readings will occur equally often.

I just would like to have a feeling for the onyomi-frequency when I'm learning, because this method worked very well for me.

[Katsuo]

This spreadsheet contains Joyo Kanji readings and also marks which of those are considered to be less common. Of course this information is too late as you have already finished RTK1.

The spreadsheet may still be of some use though, as it also contains details of the 196 proposed new Joyo kanji of which 157 are not in RTK1 (23 are not in RTK3 either). I'm afraid that for the rest of the RTK3 kanji it doesn't indicate which readings are the most useful.

Similar information from Kanji Jiten with the more useful readings marked by heavy brackets.

aovora Wrote:It doesn|t have to be perfect. I was thinking about something like that:
脚 can be read kyaku or kya.
Search the 脚 in a Kanji dictionary.
there are 42 compounds for 脚.
in 38 compouns 脚 reads kyaku
only in 4 compunds 脚 reads kya.

What if the four compounds that use kya occur 100x more frequently than the ones read as kyaku? Counting the frequency of a reading within a list of all compounds says nothing about the frequency of the reading in real life.

kanji-vocab can help to find the common readings, but only for 201/964 RTK3 kanji

https://spreadsheets.google.com/ccc?key=...RN2c&hl=en

Not exactly what you're asking for, but the above used Cangy's list to generate a joyo onyomi count (minus rare readings). In addition, I added a page that gives the count for Kanji using just the 1110 kanji from 2001KO.

Main reason I did this was I would then use that count order to learn Onyomi systematically. This is easy if you use the "RTK 1 and 3" anki deck I loaded up.

Say you want to look at all the kanji that are 2001KO and have a コウ as a joyo onyomi (since it has the most onyomi). In the search items, search for "tag:2001KO_1 tag:2001KO_2 コウ -コウ*". Now you can highlight these and convert them to something like "Test Onyomi Cards" template where you test Kanji to Onyomi and Meaning.