Some kanji have alternative writings that are significantly different from each other (e.g. different numbers of strokes). For example, see here.

I'm looking for ways to find alternative writings for a given kanji in electronic form (that I can copy-paste as text; IOW, bitmapped images won't do).

Does anyone know of an online reference for finding alternative drawings of a character?

NB: I wrote the example below before I was aware of the issues that Shen Kuo explained in his first answer. I now realize that this example is likely to be confusing rather than illuminating to most readers. I've kept it as is, since several comments and at least one answer refer to it. If you do find it confusing, please ignore it.

For an example of the sort of search I'm trying to perform, consider the character

enter image description here

Every place I've looked online shows the second character of the compound そうち (device) like this:

enter image description here

Note that the last stroke of this character is a simple horizontal one across the bottom.

In contrast, I have a several printed sources that show a different writing for the second character of this compound, where the last stroke consists of a vertical component along the left side of the character, followed by a horizontal one across the bottom; some instances are shown below:

enter image description here

Note the last two characters in both books' titles. These two characters are supposed to be the compound 装置, but the second one of the characters is drawn differently from 置1. Incidentally, my old Nelson kanji dictionary (2nd revised edition, 1974; 17th printing, 1984) shows the same drawing of this character as that shown in the picture above (see character 3644). AFAICT, it does not mention the form 置 at all.

FWIW, none of the search tools that I've tried that accept handdrawn characters as input have retrieved the alternative form of 置.

1These books use this alternative drawing of 置 throughout, not just in their cover titles. If you right-click on the image, and open it at full resolution in a different tab, you will be able to find other instances of this alternative drawing of 置 besides those appearing in the main titles.

  • What differences do you see between 置 online and 置 on that book cover? I think I'm going crazy here cus I don't see it...
    – Robin
    Dec 13, 2015 at 14:53
  • @Ash: I posted an image of the form of this character that I find online. I hope that you can see the difference. The "online" version has 13 strokes, with the 13-th stroke being the horizontal one across the bottom. The version on the book covers has 14 strokes; its 13-th stroke is a vertical one along the left side of the character. (BTW, Nelson confirms that this version has 14 strokes.)
    – kjo
    Dec 13, 2015 at 15:08
  • Ah thanks! That makes sense along with Shen Kuo's answer. I have my browser in Japanese so I couldn't see any difference xD
    – Robin
    Dec 13, 2015 at 15:17
  • 1
    Although the Japanese version is still 13 strokes btw. The last vertical and horizontal are done with one stroke (kakijun.jp/page/13111200.html , kanji.quus.net/kakijyun/1364.htm)
    – Robin
    Dec 13, 2015 at 15:18
  • 1
    And just to confirm, the Chinese version is also 13 strokes: bihua.51240.com/e7bdae__bihuachaxun
    – user1478
    Dec 13, 2015 at 23:38

2 Answers 2


I'll leave my old answer up there, and write up what I said in the comments now that you've edited your question.

Character variants are uncountable. Officially, you already have:

  • Japanese 新字体, used in official print, according to the 常用漢字 list and 人名漢字 list
  • Japanese 旧字体 that weren't actually phased out fully, especially in names, according to the 人名漢字 list
  • 簡体字, simplified Chinese as published by the Chinese gov
  • 繁体字, traditional Chinese as used in HK and published by their gov
  • 繁体字 as used in Taiwan and published by their gov (slightly different to that in HK)
  • Hanja (漢字) taught in Korea in high school as their gov decrees

Then you get non-standard characters, like 飴, that people aren't taught in school but most can read and use, and you see often in written language anyway.

Then you get 朝日文字 such as 𦜝 (officially 臍), which aren't standard 新字体, but apply their logic to characters that weren't simplified.

On top of that, there's ゲバ字, which pretty often simply use simplified Chinese in Japanese. In calligraphy too, many characters are written like simplified Chinese (many simplified characters were based off calligraphic variants). Fonts also have variations based upon calligraphy and handwriting style.

Then do we count ryakuji, too, like 㐧 rather than 第, and 门 rather than 門?

The creation of characters is ongoing, and there are always new abbreviations coming about as part of their evolution, so you won't find any cohesive works. The Kangxi dictionary contained 47,000 characters, of which about 40% are graphical variants, and there are now new variants invented in the 200 years since.

Digitally, there isn't even one system of encryption, and not all characters that you see written are encoded. Unicode is possible the most commonly used standard, but it's not totally comprehensive. But, most people use it, and you seem to be too, hence the problems with Han unification.

You imply in your question that the chinese 直 is a valid variant of 直 in Japanese. In this case, wiktionary gets closest, because it lists out every language's simplification of any character, and has articles on most encrypted variant characters, such as 𣥂 for 步 etc.

But it's not finished, and new characters are made and encoded everyday.

You can't beat it, but you can sit along for the ride.


置 is the Japanese form of the character, used in Japan.

Zhi, as you gave it, is the Chinese form of the character used in simplified and traditional Chinese.

Because of Han unification, most browsers display characters using their Chinese form unless your browser's language tells them to do otherwise (i.e. your browser is in Japanese). Most people have their browsers in English here, so they have to see the Chinese forms. Some websites do actually allow you to display forms selectively, but to great hinderance, Stack Exchange is not one of them despite having JSE that has to put up with this a lot.

I wrote more about this in another very similar question recently, which has other useful information: The component 曷 and the kanjis 褐, 喝, 謁, 渇

  • 1
    I took it out of the answer for obvious reasons, but I encourage everyone to support this proposal on meta, which means we don't need to use images every time to deal with this type of question, and we can properly differentiate between Chinese and Japanese. Right now it's a discredit to us as a Q&A site that we can't display all Japanese writing on a site about the Japanese language. (meta.stackexchange.com/questions/251743/…)
    – sqrtbottle
    Dec 13, 2015 at 15:12
  • 1
    If you're looking for alternative character forms, you can find some at en.wikipedia.org/wiki/Han_unification. There are also some variant character forms that come from different simplifications like ゲバ字, which will replace 戦争 with 战争, for instance. The only place I know that actually lists every character variant is wiktionary, and you'd have to go through each character turn by turn. The issue being that actually looking for variants is impossible due to the stylistic variations in written writing. Some people will write Shinjitai, others will use traditional in calligraphy (cont)
    – sqrtbottle
    Dec 13, 2015 at 15:52
  • 1
    and then you get variations like ja.wikipedia.org/wiki/%E3%82%B2%E3%83%90%E5%AD%97 that are literally just Chinese characters adopted into Japanese. Throw in en.wikipedia.org/wiki/Ryakuji to that, and you've got a big soup but no complete answers. Wiktionary has them all (or 99.9% of characters and variants), but not on one page, and you'll have to look up each character individually to find its alternative forms. Other than doing a quad-lingual cross reference of all CJKV languages, it's not possible to assemble a complete list of variants. Some old books do a decent (cont)
    – sqrtbottle
    Dec 13, 2015 at 15:56
  • 1
    the en.wikipedia.org/wiki/Kangxi_Dictionary, but even then more variants have arisen since then. It's an ever ongoing process of character evolution and creation. The standard characters are fixed, though. The joyo kanji + jinmei kanji lists give all the kanji you're expected to know in Japanese, published and sometimes slightly updated by the Japanese gov, but EVEN THEN there are characters like 飴 that aren't on the list that most people use anyway...you can't really win. tl;dr, it's not possible to find a 100% comprehensive resource, but wiktionary gets damn close for digital forms
    – sqrtbottle
    Dec 13, 2015 at 15:59
  • 1
    Thanks again. Your last 3 comments answer my question. I'll be happy to accept this as an answer if you care to re-post it as such. BTW, I love kanji, I think they're beautiful, almost magical. I have to admit, though, from the point of view of sheer practicality, I find it difficult to defend a writing system based on ideographs. I can only wonder how much more productive the Chinese and Japanese economies would be if they had a more practical writing system...
    – kjo
    Dec 13, 2015 at 16:06

You must log in to answer this question.

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