The screenshot below is from Kotoba for iPhone.


It shows the character for "cool", 冷 as used in the word 冷たい. However, the character in the stroke order diagram is slightly different to the main one displayed top-left!

Why is this? Which one is correct? Or are both correct?

I suspect the stroke order diagram actually comes from a Chinese font, rather than a Japanese font. I'd like to confirm this.

Thanks to Lukman for the following screenshot:

different radicals

  • 8
    マ is not even a radical, though, strictly speaking. It looks to me like that program is using the word "radical" in a very broad sense, roughly meaning "oft-encountered visual element", to make it easier to find characters. WWWJDIC and its descendants do the same thing for their kanji search, sometimes referring to these as "parts" or "elements" to distinguish them from radicals. This is convenient because it means you can find 予 without knowing that its radical is actually 亅, but it muddies the waters when you actually want to discuss radicals as traditionally defined.
    – Matt
    Commented Sep 20, 2011 at 7:19
  • For what it's worth, the printed form is the form in the Kangxi dictionary: see the bottom right of this page, and the bottom left of this page. The Kangxi dictionary was the former gold standard in character shapes, before the promulgation of national standards in the modern era.
    – Zhen Lin
    Commented Sep 20, 2011 at 9:51
  • 1
    @Matt: Kotoba (the program above) and about 99% of all Japanese-learning software out there all use the same WWWJDIC dataset. So you can expect the same mistake/idiosyncrasies. Your point is interesting (I did not know that WWWJDIC's set of "radicals" differed from, say standard dictionary bushu), but I also think convenience is by far the priority for such a tool...
    – Dave
    Commented Sep 20, 2011 at 13:48
  • 2
    @Dave For sure! I won't deny that I make good use of the "element" search at www.jisho.org myself. Just pointing out that since マ isn't a "real" radical, it's not quite correct to say that the two forms "use different radicals."
    – Matt
    Commented Sep 20, 2011 at 14:02
  • 1
    Related, posted over at Chinese StackExchange: Characters which have several different shapes, which specifically discusses Chinese vs. Japanese printed shapes, and includes a query on「令」.
    – dROOOze
    Commented Nov 20, 2018 at 6:52

6 Answers 6


It's no big deal, just that the most common standard handwritten form of the character is different from the most common printed form of the character. This doesn't even rise to the level of "variant character" in the strictest sense (like 悪 vs 惡). The two are the same character, just like a joined-up printed さ is the same as a disjoint handwritten one, or a cursive [a] is the same as a printed one in English.

The Chinese/Japanese thing is a red herring: here are two more sources clearly aimed at Japanese people showing this handwritten form. It is true that the printed Chinese form looks more like the diagram, but this is just because the "official" printed Chinese form was revised to be more in line with the pre-existing standard written form, shared by both Japanese and Chinese.

If you won't take my word for the above, check out the jōyō kanji guidelines [PDF] from the Ministry of Education. Scroll down to the section headed "明朝体と筆写の楷書との関係について" and you will see many similar cases of difference between standard printed and written forms, along with the Ministry formally declaring that these differences do not result in "different characters", or that the written form is "wrong". ("... 筆写の楷書における書き方の習慣を改めようとするものではない。 ... 印刷文字と手書き文字におけるそれぞれの習慣の相違に基づく表現の差と見るべきものである。")

Note that the character 令 is actually one of the examples in their "筆写の楷書では,いろいろな書き方があるもの" section, and the form with a final vertical is recognized as a possible "correct" handwritten version, so if it makes you feel more comfortable, go for it! Just don't be ragging on people who write it diagonally, because that's cool too.

  • Thanks for a well-reasoned answer. I managed to find the ja.wiktionary.org link myself, and I thought it actually supported my theory that it was a Chinese/Japanese thing. Yes, the page is aimed at Japanese, but the use of the form containing the マ radical is restricted to the section on the Chinese language. The Joyo kanji PDF is more convincing though.
    – MatthewD
    Commented Sep 20, 2011 at 7:07
  • So, what do you make of this discussion?: forum.jisho.org/discussion/902/… - "The crossing is common in Japanese, touching in Chinese・Cantonese. The character is the same in Unicode, the rendering depends on the font (which should be selected to suit the text's lanugage)." (Discussing a different character, but similar concept.)
    – MatthewD
    Commented Sep 20, 2011 at 7:35
  • Hm, I guess Wikipedia isn't as unambiguous as I had hoped. My thinking: the stroke order diagram (that looks like マ) is up at the top in the 漢字 section, separate from the "Chinese" and "Japanese" sections. The "Chinese" part does show a printed version that looks like the stroke-order diagram, but if you view source you'll see that it's the same character with a different "lang" attribute. Also compare the pages for 馬/門, which show the simplified forms as 異体字 (with the note 簡体字). ... But I suppose all this could be written off as Han unification issues, so I'm glad the PDF was convincing.
    – Matt
    Commented Sep 20, 2011 at 7:36
  • 3
    Re the discussion: It is true that Chinese/Cantonese and Japanese can have different "standard" forms for the same character, and sometimes different/multiple "standard" forms within the same language for printed/written text. Whether this means that the characters in question are actually not the same can get philosophical. Han unification is a complex issue. But in the context of Japanese alone, I see no reason to deem written and printed forms "different characters" when native intuition is that they are the same and the closest thing to an authority (MEXT) agrees.
    – Matt
    Commented Sep 20, 2011 at 7:42
  • 2
    Can't rule out anything from an app developer, but at least in the case of this character, the difference could just be because the form that looks like マ is indeed the most common handwritten form in Japanese. Put another way, I would expect the difference shown in your screenshot to show up even in apps hand-crafted from scratch by Japanese people for Japanese people. BTW, I would be willing to bet that "you should use JP fonts for JP and CN fonts for CN because there can be slight differences" was a reference to Han unification in general. It really is a vexatious issue!
    – Matt
    Commented Sep 20, 2011 at 8:13

It's nothing to worry about, I would go as far to say that it's not even a different "radical". (How can it be? It's the same Kanji.) Just like in English, things get, shall we say, "corrupted" in hand writing. Nothing is ever as neat and pretty as the pixels on a finely crafted character.

Fortunately, this is an easier one to remember. Ignore my awful mouse handwriting.

Bend that baby, 60 degrees, roughly.

I've seen entire sections reduced to squiggles. Look at this for example:

enter image description here

In summation, keep it in mind and move on to the next Kanji. You're going to find a lot these types of "differences" along the way. Enjoy!

  • 5
    I find myself in awe of the intended recipient's presumed ability to read that. Commented Sep 23, 2011 at 20:13

「令」(command; order) contains two components:

  • 「亼」, an upside-down「口」(mouth), with handwritten shape sometimes as「亽」and sometimes the same as the Traditional print shape「亼」(cf.「食」・「今」・「合」・「倉」);
  • 「卩」, picture of a kneeling person. Only in the character「令」(and derivatives), this component is uniquely handwritten as「龴」and uniquely printed as「ㄗ」.

enter image description here

enter image description here

enter image description here

As mentioned elsewhere,

  • The shape「{{zh-CN:令}}」is the universal (Chinese, Japanese, and elsewhere) handwritten shape, and doubles as the Mainland Chinese printed shape, which merged the handwritten form into the printed form;
  • The shape「令」is the Traditional printed shape, which Japanese has kept unchaged.

Many differences in the printed shapes of some characters are due to the different regions importing different sets of handwriting features into the print shape. For example, the Republic of China (Taiwan) standardised the character as somewhat of a hybrid:「{{zh-TW:令}}」, with handwritten form「龴」and printed form「亼」.

Note that「龴」is not actually a meaningful component. It may be useful as a graphical element to describe characters specifically for an ideographic description sequence decomposition, but「龴」is not originally a character or common mark, and it is only coincidence that some modern shapes of characters happen to contain this. For instance:

  • In「予」,「龴」was originally the top component of「呂」.「予」(Baxter-Sagart: /*laʔ/) is composed of phonetic「呂」(Baxter-Sagart: /*[r]aʔ/) and a distinguishing mark「亅」, added to differentiate「予」from「呂」.


    enter image description here

    enter image description here

    enter image description here

    enter image description here

    enter image description here


  • In「矛」(pike),「龴」was originally the shape of the blade.


    enter image description here
    enter image description here
    enter image description here

    enter image description here




The screenshots below are from some font website.

s1 s2 s3 s4 s5 s6

They show the character "S", as used in the word "Script". However, the character in the pictures are entirely different to one another! I have seen even more variations!

Why is this? Which one is correct? Or are they all correct?

  • 5
    @sawa is correct. It's simply handwriting vs print. Different styles. 冷 vs storage.kamezo.cc/asp/katchnetwork/img_17/178038/k693197271.jpg (The last stroke is simply angled instead of perfectly vertical.)
    – phirru
    Commented Sep 20, 2011 at 5:43
  • 10
    Different as they look, it is still a font issue more than a variant issue. Prickly as sawa's answer might be, he is basically right. But I think the question was still legitimate and could have been answered in a somewhat gentler way.
    – Dave
    Commented Sep 20, 2011 at 6:32
  • 4
    @rdb I don't see the seething hatred, just a partial answer. There are indeed printing differences in English as well, just look at the printed letter "a" vs the commonly handwritten "a". In fact, depending on what font you use, the "a" character can be displayed differently. Switch to something like Avant Garde and you see the handwritten version of the "a". It's similar in Japanese too, there are certain fonts which mimic the handwritten version of kanji as opposed to the version often seen in newspapers.
    – Troyen
    Commented Sep 20, 2011 at 7:05
  • 6
    I think the trouble here is that those of us (myself included) who take kanji-recognition for granted see this as a trivial stylistic variation. sawa's answer merely shows that the same kind of stylistic variations in the forms of the Latin alphabet. I wouldn't normally upvote such an answer, but I don't think it deserves -4. So +1 from me.
    – Zhen Lin
    Commented Sep 20, 2011 at 9:37
  • 5
    With all due respect, I think that interpretation is itself the result of a cultural difference. sawa's approach here (this seems like a typical example, actually) strikes me as somewhat Zen-inspired, answering a question with another question in order to highlight an implied misconception. Commented Sep 23, 2011 at 21:21

Don't panic. They're variant forms of the same character. You will encounter others. Chinese characters are very old, and have evolved in a variety of ways, including scriveners' errors, simplification, vulgarization, invention, etc., etc., etc. Fortunately, unless you're reading pre-war texts, most variations in use now are pretty easy to remember.


It’s just differences in font/handwriting. But as a side note, I've always written the version in the top-right of your screenshot, that's just how I was taught for Chinese.

See also nciku's Chinese dictionary entry for the character.

You must log in to answer this question.

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