August 21, 2014

Derived characters

While I was still at the Chinese department, during our lectures on Chinese writing, our professors taught us about 6 Chinese character types: pictograms (象形字), simple indicatives (指事字), semantic compounds (會意字), phono-semantic compounds (形聲字), phonetic loans (假借字) and derived characters (轉注字). (For further reading on Chinese character types see this post).

While they explained the first 5 quite in detail, when talking about the last sixth category, we were told that these still require further research and that no one really understands them well. Or so they said.

 wang4 'hope, expect' is a derived character. Let's look at its definition from the 說文解字 (100 CE) dictionary first:


It took me quite some time to figure out what this means. 出亡在外,望其還也 is the definition itself. After quite a bit of research on all the characters in the definition I think the translation should be: : To run away (from home) and disappear. Looking into the distance for the disappeared one. The从亡,朢省聲 part defines the elements in the character. It says:  formed by  wang2 ‚perish, disappear‘ semantic and a reduced  wang4 ‚full moon‘ phonetic.

If you look closely at the character, the  chen2 ‚subject, servant‘ element at the top left has been replaced by  to form .

Now the theory goes, that derived characters are characters where one element in an existing character that has a sound we need for a certain new character is removed and replaced by another element to form this new character. The original character acts as a phonetic element and the newly inserted character acts as the semantic element. 

Some other examples:

 hui3 ‚destroy, ruin‘ formed by  tu3 ‚earth, soil‘ semantic and a reduced  hui3 ‚beat grain‘ semantic. The  mi3 ‚rice‘ element has been extracted from  replaced  by . The residual combination of  and  would be an inexistent character.

 jin4 ‚soak‘ formed by () shui3 'water' semantic and a reduced  qin1 ‚invade’ phonetic.  qin1 ‚invade‘ originally meant ‚to proceed‘ and was formed by () ren2 'person' semantic and  zhou3 'broom' semantic (today written with  instead of  at the bottom). It was thus a semantic compound (a person sweeping the floor with a broom ‚proceeding‘ in a certain direction), not a phono-semantic compound. The () ren2 'person' semantic has been replaced by () shui3 'water' semantic in, but zhou3 'broom' has phonetically or semantically nothing to do with  and is just a residue of  after() ren2 'person' has been removed.

 ji1 ‚territory around the capital‘ formed by  tian2 ‚field‘ semantic and a reduced ji3 ‚several, few, how many?‘ phonetic. The  ren2 ‚person‘ element at the bottom left has been extracted from and replaced by . This character is even more messy, because originally  was written as a combination of two  yao1 'small‘ over shu4 'patrolling soldier' with the meaning ‚dangerous‘ (small patrolling soldier or few patrolling soldiers, danger, dangerous). In thecharacter it just so happens to be that scholars responsible for the formatting created a character where the left bottom part of  resembles  ren2 ‚person‘ which was then replaced by  to form.

Another interesting example is you2 ‘wander, walk around’. To understand its etymology we have to go back two steps. The base character for 遊 you2 is  qiu2 ‘float, hover, drift’. Later a phono-semantic  you2 ‘swim, float; walk around, travel’ was created with the original meaning ‘movement of a flag in wind’ ( yan3 ‘flag’ semantic jammed into  qiu2 ‘float’ phonetic to create 游 with 㫃 corrupted in the center of ). Finally  you2 ‘wander, walk around’ was created by removing() shui3 'water' and replacing it by  () chuo4 'go, walk' semantic. If you remove  from  what is left is a non-existent character, only a residue of  without with no meaning or sound.

The key difference between these derived characters and phono-semantic compounds, where one part of the character represents the sound another part its meaning is, that while phono-semantic compound characters can be nicely separated into two full-quality standalone characters, derived characters cannot, or if separated, standalone characters will not represent their original phonetic values as was the case with 浸 jin4 ‚soak‘. You can separate  qing1 ‚clear‘ into() shui3 'water' semantic and  qing1 ‚green, blue‘ phonetic, or  qing2 ‘emotion’ into() xin1 'heart, mind' semantic and qing1 ‚green, blue‘ phonetic, but you can’t separate  that way.  wang2 'perish' is the semantic element in  but the rest is a non-existent character ( over  doesn't exist and means nothing), which is just a leftover of the original  character after 臣 has been removed.


  1. Hello. First of all, I'd like to thank you for the interesting post. Although I am not yet actively learning spoken Chinese, I am currently interested with the study of Chinese characters. I'd like to comment on the character 朢 because it intrigues me quite a bit. By the way, some of my sources for character definition are as follows:

    Ayway, I'd like to make some points. First of all, you mentioned that 望 was a derived character from 朢. Actually, I was wondering that if that's the case, what kind of character is 朢 then ? Is it a derived character as well ? Looking at the yellowbridge dictionary, it says that 朢 can be explained as pictophonetic with 王 (the one below) representing the sound, and the minister looking at the moon as semantic (meaning to hope).

    Now, as you have said, the character may have been changed at some point in time ???
    and got changed to 望. However, the meaning is still the same (as mentioned too in the Korean dictionary, but will be differentiated later). It mentioned 望 and 朢 as 同字. Now, in the "new" character 望, yellowbridge says that we can now define this as an associative compound or as a pictophonetic character. If we want to define it as pictophonetic, 亡 would suggest the sound and the remaining part, a person looking up at the moon would be the semantic part.

    I think this is pretty interesting. Although I am not saying that I disagree that 望 is a derived character, we can certainly get a point of view that 望 can be explained as pictophonetic as well.

    By the way, I think that the defintion given and your interpretation is accurate. Although as I've said earlier 望 and 朢 have the same definitions, I think one of the original meanings of 朢 is "full moon"as you have mentioned. And then 望 was defined in this way as "to hope/to expect" with reference to the full moon. Of course, this is because as we know it, Chinese culture values the full moon as an auspicious sign. Hence, during full moon, Chinese would have a special celebration for it and to wish for good fortune upon the moon. Most notably, the 中秋節 festival occurs during a full moon. And also, the 1st full moon of the lunar year is viewed as an auspicious sign by the Chinese (and also Koreans). This tradition has been well-preserved by the Koreans actually. On the first full moon of the lunar year, Koreans will make a wish upon the full moon so that the whole year will bring them good fortune. Koreans call this celebration "Jeongwol Dae Borum Dal Machi".

    That's it. Thanks for reading my comment. Hope to read more of your research in the future. Have a nice day. :)

  2. Hello dear 김동령 ,

    Thank you for your the interesting comment.

    I understand what you're aiming at.

    朢 wang4 is not a phono-semantic compound. According to 說文解字 its original meaning was either full moon or imperial audience and it is a semantic compund character formed by 臣 chen2 'subordinate, servant' 月 yue4 'moon' and �� ting3 ‚imperial hall ‚ (not 王 wang2 ‚king‘) (說文解字: 朢:月滿與日相朢,以朝君也。从月从臣从��。��,朝廷也。you need special encoding to view ��. it is similar to 王 but the top stroke is a 丿 stroke not 一 stroke). Kings servants looking at the king at the full moon imperial audience. This meaning got later extended to ‚gaze at‘. Based on numerous other similar examples, I think that because this character suddenly had two very different meanings, to avoid confusion a new character (望) was created to represent the meaning ‚look at, gaze at’. Now this was usually done by creating a phono-semantic compound character by adding another full quality semantic element to the existing character, but in this case the new character was created as a derived character by extracting 臣 from the character and replacing it by 亡 wang2 'perish. We know this by looking at 100 AD 說文解字 definitions of 望 and 朢:

    望: 出亡在外,望其還也。
    朢 : 月滿也。

    The fact that the two characters today have the same meaning again can have many reasons. I personally think, again based on many other similar examples, that the meaning ‚full moon‘ was also represented by another completely different character which became more frequent, and the meaning was not associated with朢 so strongly anymore and subsequently 朢 was erroneously used to represent the meaning 'gaze at' again.

    I am sure however that 望 is not a phono-semantic compound (pictophonetic) but a derived character. 月 over �� does not exist as a standalone character. I’m also pretty sure that 朢 is not a phono-semantic compound either, because based on all data I have, there is no phonetic relation between 朢 and ��. 望 does have a 亡 phonetic element in it but wasn't intended to be a full scale phono-semantic compound when it was created. I think it is a coincidence that the character can be viewed as a phono-semantic compound (as you say 亡 phonetic and 月 over �� semantic) but it was not intended to be this way. Or, the fact that 月 over �� semantic is related to the meaning of 望 was not the primary intention, but a secondary coincidence. 朢 is semantically related to 望 and served as a semantic basis for 望 and not月 over ��.

  3. Just for a quick summary I’ll show you two examples of derived characters and phono-semantic compounds (pictophonetics).

    Phono-semantic compounds:

    青 qing1 'green, blue' phonetic

    清 qing1 'clear'氵(水) shui3 'water' semantic.
    情 qing2 'emotion’忄(心) xin1 'heart, mind' semantic.
    請 qing3 ‘please’言 yan2 'words' semantic.
    睛 jing1 ‘eye’ 目 mu4 'eye' semantic.

    Derived characters:

    熒 ying2 ‚shining, shine‘

    營 ying2 ‚camp, campfire‘宮 gong1 ‘palace, temple’ semantic.
    瑩ying2 ‚jade-like stone, glittering‘ 玉yu4 ‚jade ‚ semantic
    螢ying2 ‚firefly‘ 虫 chong2 ‚insect‘ semantic.
    榮 rong2 ‚glory, honor‘Original meaning 'Aleurites cordata (tree name)'. 木 mu4 'tree, wood' semantic

    I wrote this in the article, but just for a short recap, as you can see the main difference between the two character classes is, that in phono-semantic compounds you can clearly separate a character into two full quality characters with their own pronunciations and meanings (as that is how the character was created in the first place) whereas with derived characters this is impossible. The original 熒 ying2 character in the series was created as a semantic compound formed by 焱 yan4 'flame' over 冂 jiong1 'house’. Flames of torches illuminating a house (說文解字 definition: 熒:屋下鐙燭之光。从焱、冂。) So i f you want to separate the character, you have to separate it into 焱 and 冂. This is not the case with the following series as you can see, because in the series, only the bottom of the character, that is 火 huo3 'fire' has been extracted from the character and replaced by other elements. This means that if you separate 睛 into 目 and 青 all three characters will have a meaning and a pronunciation, whereas if you separate 瑩 into �� (two 火 over 冖) and 玉, �� doesn’t mean anything and has no pronunciation. It is just a residue after 火 has been separated from 熒 which wasn’t meant to be separated that way in the first place.

  4. Hello. Thank you for the explanation. I'd just like to add something from the characters you mentioned. Although you can say that 營, 瑩, 螢, 榮 can be explained as "derived characters" from 熒, I'd just like to comment that it is not the whole story and does not give a complete picture of how 營, 瑩, 螢, 榮 were formed. To complete the explanation, please also keep in mind in this case that when 營, 瑩, 螢, 榮 were formed, they were also intended to be interpreted as a picto-phonetic character in a certain sense. I'd like to explain a bit.

    Take for instance 瑩 which means "luster of gems" among others.
    This character can be explained picto-phonetically as well. I'll give the explanation of character formation from the Korean dictionary:

    瑩: This character is formed from 玉 which suggests the meaning, and from the truncated form of 熒 (which is 炏 over 冖) which suggests the sound.

    The other characters you mentioned is explained similarly and gives a very clear description of how the characters were formed.

  5. Dear 김동령,

    again I understand what you are saying, but I think we both have something different in mind.

    There are six classes of Chinese characters and I'm merely saying that 營, 瑩, 螢, 榮 belong to the Derived characters class and 清, 情, 請, 睛 belong to the Phono-semantic compound characters class.

    The fact that 營, 瑩, 螢, 榮 can be interpreted as pictophonetics is of course absolutely evident and an undeniable fact, but they are not the same character class as 清, 情, 請, 睛.