I've been reading many questions here about radicals, and it seems there is some confusion over them. My question is:

Apart from the 214 radicals of Kangxi, what are the other components/elements/primitives which are contained in kanji from the Joyo kanji list but not listed as 'official radicals' in the 214? In other words, I am looking for a list of sub-elements which are not considered radicals (according to Kangxi) but do appear in some kanji on the Joyo kanji list. I'm sure there are quite a few, but how are they categorized? I'm trying to make two lists, comprising (a) 'official' radicals, and (b) all other components which appear in joyo kanji but not in list (a).

To give an example, take the sub-element ⺤ which is part of kanji like 受、妥、採, etc. As far as I know, this is not on the list of 214 kangxi radicals, but does appear in several joyo kanji. Therefore, it will be on my list (b). Does anyone know of a complete list of other such elements?

  • I have a list of those components written in paper, which might not be complete, some of them are kanjis in its own. They are over 300. Then I have another list of more complex components made with components of the list of 300, but in this second list all of them are used to make more complex kanjis. This second list has over 700. To be honest the second list I dont remember where I left it, but the first list it would be worth posting I suppose, if I knew how to turn it to digital format. Taking photos of the sheets doesnt seem like a good idea, is it?
    – Pablo
    Oct 9, 2017 at 0:27
  • Can I ask how you compiled the list? How did you cross-check the non-Kangxi components against the Joyo kanji list?
    – kandyman
    Oct 9, 2017 at 7:51
  • 1
    If you don't have a digital version, then photos will be good too. At least I will have something to work with, and I can compile a digital version. Thanks.
    – kandyman
    Oct 9, 2017 at 7:51
  • I found the list you're looking for, although they list every character ever encoded rather than just joyo Kanji. You'll be able to pull out the components you want with some (recursive) scripting. See chise.org/ids/index.html.
    – dROOOze
    Dec 20, 2017 at 14:31

2 Answers 2


Here you go:

Data mined from

  • Forgot to include: Full 2,136 Joyo decompositions into the components described in the answer. See pastebin.com/8GaYEdKs
    – dROOOze
    Jan 7, 2018 at 14:33
  • @drooze This is incredible! I'm amazed by how detailed this info is. Were you directly involved in producing this?
    – kandyman
    Jan 18, 2018 at 13:47
  • @drooze Also, by any chance do you know anything about editing kanji to generate custom glyphs? There is an editor here (en.glyphwiki.org/glyphEditor.cgi) but I'm having trouble generating preview images. Just a shot in the dark.
    – kandyman
    Jan 18, 2018 at 13:51
  • one more question. do you know of any list which contains all joyo kanji and a breakdown of each character into its components. For example, the list would have 針 and then its components 釒and 十. For every Joyo kanji character.
    – kandyman
    Jan 18, 2018 at 15:28
  • @kandyman I mined that list using scripting from the sources mentioned. For the full Joyo decompositions, I believe that I've provided that in my first comment (I forgot to place this in the answer). As for GlyphWiki, I'm able to generate custom glyphs fine, what are you having trouble with?
    – dROOOze
    Jan 18, 2018 at 21:01

Apologies for not answering this question directly, but for all intents and purposes, such a list is not useful. This is especially true with regards to the character forms in Japan after simplification (shinjitai).

Kangxi radicals themselves do not serve a purpose outside of being dictionary header sections, which is the literal translation of 部首, and only act as the Chinese character equivalent of an alphabetical order to organise dictionaries. While many modern forms of Chinese characters can be decomposed into primitives, many also cannot, as they may have undergone extensive graphical changes from the original form.

For example, the ancient forms of and were near identical, and were distinguished by a V-shaped mark near the top. They both originally referred to some kind of officer in the imperial court. In Kangxi, they're grouped under the radicals 口 and 亅, respectively, and the modern form of 史 completely obscures its primitives whereas 事 largely preserves it (a shape like , depicting a writing-brush container, and morphed into a shape like 彐, depicting a hand holding this container).

Note, ⺤ is just another graphical form of a hand:

  • depicts a hand passing a boat (originally , simplified/corrupted into ) to another hand 又, indicating the meaning to receive and 舟 doubles as the sound cue;
  • was originally just written , which depicts a hand gathering something from a tree , indicating the meaning to harvest.
  • Thanks for your answer. You say that such a list wouldn't be useful, but with respect, I disagree. I am studying kanji by a method which focuses on learning sub-components. For the purpose of being able to write all words covered by Joyo kanji, I'm trying to create a list of all the sub-components I need to learn. Such a list would be useful because it can be transformed into digital flashcards and learned with a review program like Anki. Individual radicals/primitives could be practiced in isolation, which is partly my goal.
    – kandyman
    Nov 6, 2017 at 17:09
  • 1
    I say it's not useful because many modern forms don't reveal the original meaningful components of the characters, which is why there isn't actually - as you said - a comprehensive list which contains all character primitives, because in many characters they've undergone changes over the course of history and have been obscured. This means that many real primitives aren't digitised! If you're able to handwrite and upload them, maybe you can create flashcards out of them.
    – user26355
    Nov 6, 2017 at 23:50
  • I forgot to mention - GlyphWiki (sandbox located at en.glyphwiki.org/glyphEditor.cgi) is a very useful resource if you want to delete components off characters or make arbitrary ones that don't exist. This means that you can make images of components that aren't yet digitised.
    – user26355
    Nov 7, 2017 at 0:34
  • I'm not sure I follow. Can you describe some characters where the components have not been digitized?
    – kandyman
    Nov 7, 2017 at 10:13
  • E.g. the primitives of 事 are not digitised - what was originally a hand 又 is described as 彐 (which is defined as pig snout, if you look it up in Wiktionary), but this is incorrect; the middle stroke should go beyond the vertical stroke, and the component should strictly mean 'a hand'.
    – user26355
    Nov 7, 2017 at 10:59

You must log in to answer this question.

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