Incorrect Characters

Since Chinese is a language of tones and homonyms, speaking is one thing, and probably the easiest part of it. Even if you know a large number of characters, when you have to write anything down, there is a lot of room for confusion in determining which character goes where. This is especially true for those of us who are learning Chinese as a second language.

In Chinese, these types of “incorrect character” problems are generally called “tso bie zi,” which one might roughly translate as “wrong other characters.” In fact, this entire issue is additionally complicated by the fact that a large number of characters have quite similar meanings.

Is there any sort of computer program that can assist in dealing with this problem? (I use a desktop computer.) By way of comparison to English, in my Microsoft WORD program, any words which the program considers misspelled are lightly underlined, whereupon I can hit the right mouse button and get a list of spelling selections.

My Chinese input program offers little specific help in this regard. One input option allows you to “pick a second character” from a list of options, however that is not always the ULTIMATE SOLUTION in my opinion. Sometimes I can only remember the “second character” of the expression I am seeking, and that input option doesn’t allow me to “pick a first character”. Also, it is often true that in picking a second character, the very characters which I am trying to differentiate between are all listed there one by one.

Interestingly, my Chinese associates’ solution to dealing with this problem is to say “Ask a friend.” However, when I am typing up a Chinese document late at night or very early in the morning, there are no friends to ask.

Although I have several hardback Chinese dictionaries on my shelf, looking up words in them is time consuming . . . . moreover, if the exact combination which I am searching for is not listed, I am still in the dark.

Any suggestions?

Hi Richard,
I used to have a program that did exactly this --it was absolutely wonderful for catching just what you are talking about. It was like a “spell checker” that thought about whether you were using the right characters one with the other. It was called “chu2 cuo4 da4 shi1” [eliminate-mistake-great-teacher]. I did a search on Google but all I can find is references to serial numbers (I suppose to make illegal CD copies run).

The version I had used to run on Win 95. I got a new computer later (with Win2000) and when I did, they didn’t have support for it. I never went back to figure out whether they had a Win2000 (or now, XP) compliant version, and I stopped doing work into Chinese [that was only ONE client, and a special case, but the program was very helpful for getting rid of really bonehead errors!] but now that you mention it, it might be worthwhile to see if we can scare it up again.

Originally I bought the package at a software store just at the corner of Pate [Bade] Road and Hsinsheng [Xinsheng] S. Road.


Another possible (quicker, although still dictionary-based) solution is to use a dictionary like Lanbridge’s (wen2 qiao2 chu1ban3 she4) excellent “Pinyin Chinese>English Dictionary”. It’s listed by Pinyin, but WITHOUT TONES (the tones are marked but the arrangement of words is done without regard to tone or which character the syllable’s phonetic form represents). That means that you can quickly look up the spoken word you’re thinking of, even if you’re not sure what the first character is, and see the correct characters there for your typing pleasure.

Also great for listening to anything, because you can find things quickly without knowing the first character.


Maybe you can try to learn “how to type in Chang1 Jie2”, it’s the kind of input method which you don’t even need to know how to read the word (ex, when you want to type “ni hao”, you have to know it is “n” and “i” with 3rd tone, and chang1 jie2 doesn’t need it), you just have to know how does this word look like, (use the normal english-chinese dictionary, and you will find out the chinese characters (of course, you have to have a little idea about how the character “about” look like since one english words could translate into many different chinese characters)

Unfortunately, this is a big part of the problem. I often find myself trying to make a determination of which character should properly go where, and some of the selections only differ by a slight amount, such as the left side radical, or whatever.

For example, there are two words which roughly translate as “make”, both of which are pronounced zuo4. Over the years I have asked several local Chinese teachers (elementary school, junior high school, vocational school, etc. levels) if they could provide me with the definitive list of combinations for these two characters, so that I could refer to the list and immediately know which one to use in which expression.

None of these teachers could provide me with such a list. In fact they were surprised by the question. Why would anyone be confused by these characters? They couldn’t understand it.

P.S. Can anyone provide me with such a list? Or if I have other groups of characters with which I am totally confused, can someone come up with a definitive list of combinations for correct usage?


Perhaps a comprehensive list is too long to put together. An alternative is to learn the meaning for the words. For example, I had trouble with the three “de”'s until I learned their definitions. In Taiwan and the rest of the Chinese world this is actually very often misused.

For example, you would see


only #2 above is correct. The other 2 sentences should be


So usage in this case depends on the part of speech:

的 is used to show possession
得 is an adverb (modifying verbs)
地 is an adverb (modifying adjectives)

Another example is “那” – when do you use 那 and when do you use 哪?

So the best way I find is to learn the grammar, and the 2nd is to read a lot.

As for software, have you tried NJStar? Their lianxiang (聯&#24819 feature is pretty good. When I type in zuoshi the only choice I get is 做事, and when I type in zuoye the only choice I get is 作業。

oops, speaking of 錯字, the above “那知筆是我的” should have been “那支筆是我的”

I blame it on the word processor.

I appreciate your comments. You have misused a “zyr” character. This is another group of confusing characters. As in the expression “zyr1 zi4 bu4 ti2” – not a single character was mentioned.

Which zyr is it? I can never remember. 支 or枝 or只 or枳 or椥, to me this is a giant source of confusion. Using the Chinese input program on my computer, I find the following possibilities for characters pronounced zyr (by tone) – 1st tone: 33 selections, 2nd tone: 32 selections, 3rd tone: 27 selections, 4th tone: 73 selections.

That is 165 characters more or less, since some of them are repeats. If you were to write down all of these characters on notebook paper and ask me to make a sentence with each, I doubt that I could get a grade higher than 25%. I have been actively studying Chinese for over 27 years, and I have written and had published nearly fifteen books of my collected essays in Chinese (CROWN PUBLISHERS, Taipei). At the present time, I write Chinese legal documents. I write Chinese every day, hence I deal with these confusing character issues every day.

You suggest learning the meanings of the words (characters). I wonder how practical that is. Out of the 165 zyr characters, I know perhaps 42. That leaves 123 that I don’t know. In order to deal with my lack of knowledge in this area, you are suggesting that I learn the meanings of the remaining 123? So that I can use them correctly if the need arises? Frankly, I don’t think that such a methodology is practical or efficient. It would be much better for someone to put together definitive lists of the characters which are commonly confused, (in consideration of a commonly used core group of perhaps 6,500 characters). That would be a time consuming task. I would be willing to contribute to such an effort, or suggest possible combinations, but I wouldn’t want to be the one heading it up.

The second Chinese sentence you gave could be translated as “That pencil is mine.” Well, pencils are made of wood. Why doesn’t the “jyr” you used to modify pencil have the wood radical? I would also like to see a definitive explanation of how the common radicals relate to the characters where they are used. For example, “yan jiang” (I am not sure of the tones) means “a speech”. But “yan” has the water radical. That took me many years to remember. I learned it by rote. Is there a grammatical rule for such things? What does “a speech” have to do with water?

You can make up a funny story to somehow relate “a speech” to the water radical, but I can also make up a funny story to relate “a speech” to the metal radical, or to the dragon radical, or to the hand radical. So, making up funny stories, or having the learner try to remember STORIES on which characters are (supposedly) based is an exercise in nonsense in my opinion.

Well, you can certainly consider it “an exercise in nonsense” if it doesn’t happen to work for YOU, but there are thousands of students (and teachers) who successfully incorporate mnemonic devices into their teaching and learning, and who do not consider them “nonsense” because the devices help them to retain information.

Each learner needs to consider his or her own strengths, both the innate ones and the ones that he or she has acquired because of the learning environment he or she has been in. For example, I don’t believe that Chinese learners are innately more “able” to memorize than Westerners, they have just had a lot more practice doing so – that would be environmental.

Have you made an effort to find the program I mention above? I think its dictionary and correction function would take care of 90+% of the problems you mention. With that program, you first type your document, then run it (on the line of an English-language “spell checker”) and it will mark combinations that it doesn’t feel are right.

There is also a dictionary (the name of which escapes me just now, but I believe the author or chief editor is Shou-hsin Teng) that is specifically for foreign learners in Chinese and addresses just this problem: which characters to use in which situation. It addressses not just characters but also usage and word choice for easily-confused combinations. I think I bought mine at the Lucky Bookstore at Shita a couple of years back. The title has something to do with “usage”, I think. Accept no substitutes – this one is written specifically for us foreigners.


As an aside, I would also suggest that even if one has one’s works published in a foreign language, one shouldn’t expect to be a native speaker as a result. I don’t say this to berate you, and I’m not suggesting that you consider yourself one – but you seem to be putting that kind of expectation on yourself, and I think that might be an unrealistic amount of pressure!?!

I see the same thing in reverse with Chinese learners of English beating themselves up because they “still make mistakes after 20 years”, especially in writing. Hey…that’s what editors are for. Even native speaking writers usually try to build and maintain a good relationship with one or two outstanding editors, and many famous novelists or writers wouldn’t have achieved the success they did without these editors. Maybe it’s more important for you to ensure that your meaning is unequivocally clear and just concentrate on your message, and “throw money” at the problem by paying somebody for whom “cuo zi” will not be a problem to catch them for you.

I think it’s like the law of diminishing returns. When you start learning, every new word is like another 10% of your repertoire. After a couple of years, progress slows a lot. I usually advise Chinese professors, for example, to write their paper clearly but not necessary perfectly in English and get a competent native-speaking academic editor to vet it for them. Their time is usually worth more if it is spent on other endeavors rather than spending hours and hours to catch a few mistakes, only to miss others. (These are for the folks whose English is pretty darn good…the others I usually advise to just write in Chinese and let a translator handle it, becuase it can take longer to figure out what they meant in English than to translate it from the Chinese.)

Just my NT$0.05



I’ve never been a fan of the strange little stories myself but I find that I do need some way of remembering those darned characters and memorising stroke by stroke just doesn’t cut it for me. My preference is to look up the etymology on

Unfortunately yan3 jiang3 is a particularly bad example.

This would be one where only brute force memorisation is suitable. However many other characters have more meaningful deriviations and are easy to remember.

Also I’m not too sure what it is you want as far as words involving zuo4 go, but if all you require is a dictionary look up of every compound involving a zuo4 charcter I can easily do that.

Using Hao’s frequency list I calculated the percentage of his corpus that was contained within varying numbers of the most frequently occuring charcters.

For instance the top 500 most frequently occuring charcters comprised 79% of the corpus.

500 79.1568%
1000 91.0692%
2000 97.8480%
3000 99.4230%
4000 99.8081%
5000 99.9165%
6000 99.9482%
7000 99.9622%
8000 99.9717%
9000 99.9793%
10000 99.9859%
11000 99.9915%
12000 99.9962%
13000 99.9998%
13060 100.0000%

The corpus is the entire collection of internet newsgroups posts made in 1993 and 1994. It contains 13060 distinct characters.

So what does this mean? Well let’s say you already know the top 3000 characters and learn another 1000. That next thousand takes you from understanding 99.4230% of the corpus to 99.8081%.

Progress is very slow at the high levels.

I agree that this is a real hassle learning Chinese. Another example: There’s 3 characters ‘mi’ (same tone?) which all mean secret. I can memorise all three, but wil that help me write ‘mimi’ (meaning secret) or mima (meaning PIN number)? Not really. I actually find that a lot fo Taiwanese make mistakes and these ‘mistakes’ often become common and accepted. I believe that probably hundreds of years ago many of these homonyms were just the same word (how would they be any different to an illiterate) but the scholars just decided to complicate things to protect their arcane knowledge.


A complaint was made about mi mi mi, or de de de de. Actually, it is
all your fault for not also learning a second Chinese “dialect”. For
instance the several de’s are all still pronounced differently in
Cantonese. Having thus two eyeballs instead of one, all these
suddenly make sense, and you will never screw up even when a native
would. You can get 50 cassettes and books for Taiwanese from in Taizhong or
Taibei. Anyways, it’s like all those French endings that sound the
same spring back to life. The right side of characters, the
“phonetic” suddenly makes sense. Plus you can out-cool Hartzell, who
doesn’t speak Taiwanese, and being probably twice the age of the
normal poster here, probably will never learn it. :slight_smile: :slight_smile:

With command of Taiwanese your visit to the presidential palace will
be a big success, as they might even assign a personal tour guide to
your party, complimenting you all the way… makes you feel like or something. The tour guides are all A-bian
believers. [I might have been too, but A-bian let the Tongyong jerk go nuts with the street signs, instead of using real pinyin.] The
hours go by as you discuss Pres. CKS’s shoes or something, and before
you know it it’s 12:00, closing time. By the way, the presidential
palace is shaped the character ri4 of ri4ben3, which makes sense, as
it was the Japanese who built it. I bet that makes some folks

Yeah, Jidanni’s right. Learning another dialect (or even Japanese or Korean) helps.

秘密 (mandarin) - mi4 mi4 (or bi4 mi4 I have heard sometimes)

秘密 (cantonese) - bei3 mat6

秘密 (japanese) - himitsu

비밀/秘密 (korean) - mi mil

don’t know what it is in Taiwanese, though, ashamed to say. So besides mandarin, the 3 other examples listed above have different pronunciations for the 2 characters!

d’OH! Screwed up again.

The Korean pronunciation should have been

bi mil, not mi mil.