Learning to Write Chinese through Character Components

So I wrote this way last year, summarizing a research paper I read about learning to read/write Chinese.  Upon re-reading it, I realized how so much of the info is relevant to my current obsession of writing!  Thus making me finish this post finally.  

Learning how to write character components is the 4th step in our Learning how to handwrite Chinese study thread.  I provided all the links in that post.  The Cool Chinese website has lots of info.  But I’m using the research paper PDF I found online as a way to teach how to recognize characters.  I thought I would sum up the research paper so that I can internalize what I learned.  It is a 160 page paper!  The abstract is in English.

Hacking Chinese also has several good articles about components.


  • Some really old and dead dude (wrote a book on how to teach children and he said that knowing 1000+ characters is for speaking, and 2000 for writing.  He also said to learn characters you start from simple ones and then use them to construct and recognize complex ones.
  • In the classroom teachers spend a lot of times teaching how to write a character and students mechanically write it many times in order to remember them.  But it doesn’t teach them to recognize characters.   Now a days with the advent of computers children are having a harder time recognizing characters forms.  They confuse and use the wrong characters.
  • Taiwanese classroom teach characters as part of a line of text.  The base component in a lesson is “reading text’, and from it comes writing, reading, composition.  This helps children with critical thinking and writing.  But kids don’t have a good grasp of how characters are composed and have a hard time with writing them later on.  The emphasis when teaching characters are usually on how many strokes and stroke order rather than character components, which aids in remembering the character.
  • The author’s basically advocating analyzing a character’s components to help you recognize a character, rather than trying to force memorization of a character via stroke order.  This is because most Chinese characters have a method to the way they’re composed or formed.  So rather than learning to recognize characters through number of strokes and stroke order, it’s better to do it through character components.
  • Taiwan has 440 character components out of 4808 characters, China has 441 character components out of 3500 characters.
  • China is ahead of Taiwan in using character components to help teach character recognition.  They’re using it in school and overseas teaching.  Taiwan is only doing it for special ed.  Taiwan uses character composition (left right, inside/outside).  China uses verbal, stroke order, hand gestures simultaneously to describe a character.
  • Taiwan typically uses the ‘copy a word multiple times in order to memorize a character method’.  So that’s why we write it so much!


The Details

When learning to write Chinese, knowledge of character strokes, stroke order, character structure and shapes, and finally components are required.  The author went into details about different types of components and comparing how different researchers decide what makes a component.  I’ve included definitions of various types of components at the end of the post.

She then basically looks at each of these aspects of a Chinese character and discussed them in depth. It’s a really good resource if you want to use this info to teach.

  1.  Character Strokes – 筆畫

    There are really just 5 basic strokes:  橫 , 豎, 撇, 點, 折.  This is much less than the 24 I see in the 生字簿 online.  Or the one I see at Huayu World. It’s easier to remember and I can just use these 5 to describe strokes.

    Stroke is the smallest component of a character.  One stroke is where the brush begins and picks up.  This is the first step one should teach when teaching writing.

     Of particular note is that the stroke will change for a character if it gets relegated to the left side.  For example, the  木 in 林。

  2. Stroke Order – 筆順

    There are 17 basic stroke order rules.  For example, 自左至右 (left to right), 先上後下 (top then bottom), etc.  I believe I taught Thumper this really quickly last year, but obviously it didn’t stick.  I don’t know if this is one of those things where, when you write a lot, you kind of just get it.

  3. Components

    Pages and pages talking about how different researchers define character components.  For example, for 香, you could split it into 禾 and 日.  You see this in MDBG.  But, what happens when you see it in a character like 馨?It’s actually better when you teach it to just refer to it as 2 components, 香 and what’s on top.  Or 教.  You can divide it into 2 components:


    Or, you can divide the 孝 into two other components if you like. But, there is no reason to do so because 孝 is a phonetic component.

    Basically one researcher says that there’s a difference between character component (部件) vs the components of a character (字符) that gives it meaning, be it phonetic component or semantic component.  And it’s better to break it down to the meaningful component level instead of just character component level.  Because components are just defined as that gray space between a single stroke and the whole character, and there are myriad of ways you can split a character to find its component.

    If you’re interested, pg 28 of the paper lists the number of components as defined by different researchers.  I‘ll just highlight one from each country:

     Total Characters  Character Components  Standalone Components  Radical Components  Unique Components
    China 3500 441  311
    Taiwan  4808  440  95  225  120

    The author continues to go on and on about this.  And sums it up by talking about how she defines components and also how she will “name” these components.  Because different researchers have different names.

    In any case, for the author, she found ~380 components:

    • 214 radical components, with 172 actual radicals and 42 radicals that got divided into smaller components
    • 74 non radical standalone components
    • 91 unique components

    The important thing about this is that this is a much smaller number than the 500+ that other researchers found.  Because her aim is to get it to just small enough for children to remember to write, not necessarily to divide it for categorizing characters components.

  4. Character Structure & Shapes

    The author then goes on to discuss character shapes and structures because in oder to describe the layout of components, you kind of have to study the shapes and structures of Chinese characters.  I will refer you to the posts I wrote about the two topics.  Basically the characters structure is how components fit together.  For example, most characters are left-right like 林 or top down like 早.  Then there are the shapes of characters when we write them (so that they’re pretty).  For example, when you write 日, it’s kind of skinny and tall rather than flat and wide.


Putting it altogether

What does it all mean?  I’m sure if you’re a lazy person like I am, you just want to know how to use it.

Right now, I’m doing it half heartedly with Astroboy when we go through Sagebooks.  Because I hadn’t bothered to re-read the research paper till I rewrote the post.  (ha!)  The author detailed how she arrived at her methodology in naming the components of a character.  It requires that you remember the character structures and stroke names and character shapes and I can’t remember all that.  Instead, I just use whatever knowledge I retained from grade school.

For example, we’re on fourth book of third (orange) series:

  • 拉 – 手字旁,立字邊
  • 塊 – 地字旁,鬼字邊
  • 才 – 一橫,一直,一撇
  • 星 – 日字頭,生字底

And here’s how the research paper says it:

  • 拉 – 提手旁立字邊
  • 塊 – 土部旁鬼字邊
  • 才 – nothing
  • 星 – 日字頭生之底

There is really only 1-2 additional steps between this method and how many parents will point out components to their kids when teaching.  The author is describing, via a mnemonic, the relationship of a character’s components so you can use it to recall or to write.  

I don’t even do it for every character.  But what I do is, when Astroboy cannot remember a character, I ask him to tell me what he sees.  This at the minimum makes him really look at the character and how it’s written.  Most of the time he does not remember the components.  But sometimes, if I were to repeat the way you write it (e.g. 手字旁,立字邊), he remembers the character.  Other times, he starts trying to tell me how it’s written (e.g. There’s a 日 in there; or names whatever stroke name he remembers)  I figure at the minimum it is introducing him to other characters and stroke names.

Now, for Thumper, I didn’t even use this method because I didn’t really know about it when we were zooming through Sagebooks earlier this year.  She’d had some schooling on radicals and all I had to do was point out the two components and how one is a sound and the other is meaning and she remembered the characters super fast.  Here is where having a good command of the language helps.  Because there are many components that just “contributes” to the final sound but isn’t exact.  For example 時(shi2) has the component 寺 (si4) or 狼 (lang2) has the component 良 (liang2).  However, now that we’re focusing on writing more, I think I’m going to try this method on her so she remembers how to write.

So here are the author’s rules for naming components for writing.  Maybe one day I will post that cheat sheet I made of the character structure names (旁/邊, 頭/底, etc)

The way she names the component and how to write:

  • Standalone components: use their name
  • Unique components : use their stroke order
  • Use character’s common names (e.g. 三點水)
  • Name component’s place in the character
  • Try to be simple and use modern wording
  • Follow general stroke order in naming how components are “assembled”
  • Split character to first level and only split to additional “levels” if the child doesn’t understand (prob due to not knowing enough characters)
  • Use 字 and 之 to distinguish between stand long components vs unique components

I guess when I actually start naming things according to the paper, I will write another super detail post about it.  Otherwise you can scan through the paper yourself.  Right now, just naming components we see through semantic vs phonetic component and some description of how the components fit together is good enough for me.  I’m hoping it will have the same effect on Astroboy as on Thumper in learning to recognize new characters past Sagebooks 500*.  Page 75 has the names of the character structures.  Page 165 has the list of 4808 characters and how to describe the components.  To find your character, first look it up on the variants dictionary, find it’s character number (星 is A01792) and then look for the equivalent number in the paper.

*Note:  Coincidentally, the new books I bought from Greenfield does teach components.  Maybe I will have a chance to pull it out to teach Astroboy and Thumper next semester.

Definition of Components: 

  • Character components:  Parts of a character that stands between individual strokes and the whole character.
  • Standalone Components:  Components that can also stand alone as characters
  • Unique components:  Components that do not stand alone as characters.
  • Basic Component:  When a character can not be divided into smaller components.
  • Combined Component:  Composed of 2+ basic components
  • Transformed Radicals:  Radicals are transformed in order to “fit” into a square for writing.
  • Combined radical:  A radical with two basic component
  • Separating space:  Spaces in character which separates 2+ strokes.  It’s a way to figure out what the character components are.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s