[Userscript] Keisei 形声 Semantic-Phonetic Composition

I don’t really understand all this semantic-phonetic thing, but I just want to let you know that your script fixed me one problem! :slight_smile: During radical lessons, now I can see in advance which ones are actually Kanji, so I just add a synonymy to the radical. Just by doing this, you simplified my Kanji learning because it was hard to forget the meanings from the radicals. I can even learn the real ones in advance ^^

Yeah, it also bothered me that later radicals don’t include kanji from lower levels that were built from parts. It should also help for “similar kanji”.

And maybe with the script you can find out about the composition, in the best case you get lots of ON readings for free, and understand that not all kanji are “pictures” of something like 日 is a sun.

If you look at for example 滝 it’s not a water dragon or something, it’s “the word read like dragon 竜 that is related to water”. If you know all readings like a native but not the kanji this would actually be helpful (water,
りゅう~ろう? => waterfall!), but at least you will wonder less why there is a dragon in the waterfall.

You should really do some reading on kanji phonetic components. I just looked at the level 30 kanji as an example. 9 of the 31 kanji in that level have a phonetic component that I recognize, meaning I “learned” those readings for free and didn’t need to spend mental effort learning them. It makes learning new kanji much easier!

New version 0.9.8 released, almost finished! Tested for all modes, should work fine.

New features:

  • Added the (61) non-jouyou kanji from WK to the DB
  • Added a new button for the “more info fold” for more consistent interface
  • Click the question mark for version and database information

1 Like

Thank you so much for this script yo!
I was using the other old one, but this one is a lot easier to grasp/understand, just works better better too… :sweat_smile:… SO nice to have this info accessible, this script and the pitch accent script = totally essential scripts imo.

Version v1.0.0 is out! Changes are under the hood. I also updated the database a bit, it has now 1930/2201, only a few more to go.

Please also check out the new poll in the top message!

1 Like

The kanji 「拠」 was created using semantic-phonetic composition, the tone mark is 「豦」

Are you quite sure about this one?
Yes, sometimes the phonetic component is simplified or modified in kanji, but at least it resembles the original in some way.
You’d really have to stretch it to say these two look alike in any way.

I get most things from here: 拠とは (キョとは) [単語記事] - ニコニコ大百科 (other sources gracefully skip this kanji).

The old form was 據 and then 㨿. The second one already has a different component, which was 處/䖏 and turned into 処 (another tone mark), but it has reading しょ. I think they swapped the old component because it looked basically the same, but didn’t update the reading, so 処 is not the tone mark.

I think it’s a sound theory, but for this one the tone mark is not super helpful. However this is the only instance I remember where they actually swapped the tone mark.

1 Like

Anyway, please keep on looking for errors, I edit the files by hand so readings/compounds may be wrong, some kanji appearing multiple times, etc.

The database is now almost finished, maybe 60 kanji to go.

@LucasDesu @GabrieleCK @antagonisten @swigder @seanblue @Arnagos @gzovak @sgtowns

Version 1.0.3 is out, all kanji in WK are now added and the code refactored to potentially interfere less with other scripts. I didn’t test it for lessons yet, the section may look strange when something went wrong.

I didn’t go through the WK radicals if there are more matches, please mention them here if you find some.

2 Likes

So far everything working swell.
Just one question, is there a way to permanenly toggle this to be on?
Cause so far I’ve had to click it every time and sometime I miss that there is more info hidden.
Capture

1 Like

That one was even too big for my tastes sometimes. I can add a toggle to the options menu.

1 Like

Is the phonetic component for 視 correct? It says the component is 示, but that’s not really helpful. Are you including historical components that are no longer included in the kanji?

For this one I’m pretty sure, the old version was 視, the reading fits, and even the meaning is related to the “real radical part” 見. It happens rarely that the tone mark is also looking like a radical, but for this one it fits. 視とは (シとは) [単語記事] - ニコニコ大百科

The pelican can come from at least two different components, for example
礻示 ~ 祈
⻂衣 ~ 初

New version 1.1.0 out now, two new features:

  • not in WK kanji now include a meaning gathered from jisho and get the locked style to be distinguishable. Clicking on them now leads to jisho.

  • @gzovak Full info mode toggle that always lets you start with the “more info fold” open.

1 Like

I’ve been using this script and really like it, but I found a minor… glitch? For certain kanji it incorrectly reports them as if they were the phonetic mark itself. On level 11 there’s two such ones, and , which have the exact same description as 方 and 皮 respectively. For instance, the page for 放 says:

The kanji 「方」 is used as a phonetic mark in other compounds!

When it should say:

The kanji 「放」 was created using semantic-phonetic composition, the tone mark is 「方」

It doesn’t really impact usability, I just found it strange.

Thanks, I will check this. These kanji are tone marks on their own and also built from other ones, it seems I pick the wrong one in the end as the main tone mark. You can see the real solution in the “more info fold” (leftmost button).

@DeadlyFugu Version 1.1.2 is out now, the problem should be solved now. It now also shows the component info.

I have another question. When a kanji includes a phonetic component, are all kanji displayed supposed to show one of the rating badges (like 天)? I’ve occasionally seen ones that are missing the badges.

This should only happen to the first kanji occasionally.

The first 1–3 items in the character grid have a special status, they are added in front of the compounds (I didn’t want to mess around with the style to make them stand out more).

  • Always green, the phonetic used
  • If phonetic element is also a WK radical, add it (blue).
  • If phonetic element is also a kanji in WK, add it (pink) << that one

You could say that the kanji of a tone mark is a perfect compound with itself, so that one will always have 天. I left the badge out so that it is easier to see that the tone mark is also a common kanji.

1 Like