I need to write a program that can insert the count of objects into a text - without knowing virtually anything about the Chinese language. While for English this needs to distinguish between singular and plural, I don't know which special cases need to be handled for Chinese. Using an online translator to translate "one house, ..., thirteen houses", I obtained

一所房子,两个房屋、 三栋房子、 四个房屋、 五之家、 六套房子、 七个之家、 八个公司、 九个房屋、 十家、 十一间房屋、 12 间房、 十三个房子.

The same with cats:

一只猫,两只猫、 三只猫,四只猫、 五只猫、 六只猫、 七只猫、 八猫、 九猫、 十猫、 11 个猫、 十二个猫,十三个猫

Some of this may be caused by imperfections of the online translator used, but I wonder if some of the patterns should be observed by my program. And I need to provide it with suitably many "declination" forms to allow it to create the correct output:

  • Arabic digits for 12 in the first test and 11 in the second test are obviously an artefact of the online translator. I assume that it is never customary to use arabic numbers in Chinese?
  • I wonder why "two" seems to be translated by 两 instead of 二. Is that correct? Or does it depend on something?
  • What follows the numeral seems to differ depending on the number. For example I see 所房子 with 1; 个房屋 with 2, 4, 9; 栋房子 with 3; and all remaining numbers seem to have their very individual signs, at least when it is about "houses". For the cats I see 只猫 with 1-7, 猫 with 8-10, 个猫 with 11-13. While this looks less complicated, I wonder: Are the different forms required by grammar? Do they follow a specific pattern (e.g. for living versus dead things)? Which variants have to be used for which number? (Or is this just another translator artefact?)

Thanks for any hints on this, but please remember to keep them understandable for someone who knows no Chinese and will handle Chinese texts with a mere "pattern substitution" technique.

  • As stated in the answer, there is no such thing as a plural in Chinese. It's not an agglutinative language, which mutates depending on the context. I really think some knowledge of Chinese grammar is necessary here. Otherwise you will run into a lot of unforeseen issues. Commented Oct 10, 2012 at 21:53
  • My suggested algorithm: 1. Find nouns (assume you can do this, or that it is provided). 2. Look up the classifier for the noun (you can scrape a dictionary site for entries on all your nouns). If multiple classifiers, pick one. If no classifier available, use 个. 3. Insert <number><classifier> before the noun. 4. As to the preferable form of the number, I'd do something easy, like chinese numerals for n < 20 (and remembering 2 = 两) and arabic numerals for n >= 20. Commented Oct 11, 2012 at 20:45
  • 1
    Oh, totally forgot! Some words don't take classifiers, but there are very few of them. and are the only ones I can think of off the top of my head. Commented Oct 12, 2012 at 0:21

3 Answers 3


Regarding arabic numerals: I think modern usage sees arabic numbers to be generally acceptable. However the context of the material, or even the region, may influence whether they're commonly used or not. For example, in metropolitan areas you might commonly see phone numbers or prices expressed in digits, whereas in more traditional material you probably wouldn't. You may or may not be able to get away with this in your project.

Regarding 两 vs 二: 两 is generally used as "two" for specifying a number of objects, whereas 二 is used as the digit, such as in a phone number. See this question for more on this distinction.

Regarding the classifier: As far as I'm aware, there's no such rule as 1-6 uses this classifier, 7-9 uses this, etc. The classifier is determined by the noun. In your example, the classifier for cats (猫) is 只. For house (房子) you can use 栋.

The classifier for people is (usually) 个. Perhaps more importantly, 个 can be used as the catch-all classifier in casual language, or when an obscure noun has a classifier that you simply cannot remember. Again, whether you can use this "shortcut" will depend on the specifications of your project. In more formal situations, the correct classifier is preferred.

If you have a finite list of nouns in your project, you could spend some time looking up the relevant classifier for each noun.

http://www.mdbg.net/chindict/chindict.php specifies the classifier in the dictionary entries, denoted by the "CL:" text.

Hope that helps!

  • I'm ok with classifiers as I can simply require them as part of zhe input. However, I am concerned with the classifier changing depending on the number. As I understand you, however, this change should not happen and Chinese really has no such things as singular, dual, plural (or the funny variants used in Russain)? Commented Oct 10, 2012 at 11:42
  • 1
    @HagenvonEitzen There is no such thing like classifier changing in Chinese. In a consistent context you should always use the same classifier for the same noun (otherwise it seems weird). And Chinese nouns don't reflect, in other words, they remain the same almost all the time.
    – Nihil
    Commented Oct 10, 2012 at 16:16
  • 1
    个 suits most nouns, but not all. 一个人 is perfect, 一个房子 not perfect yet acceptable, but *一个猫 sounds weird.
    – Nihil
    Commented Oct 10, 2012 at 16:20
  • 1
    @HagenvonEitzen a large proportion of the online translation results you've cited is wrong or inconsistent; what is said in this answer is right. Just stick with the answer and ignore those translations.
    – NS.X.
    Commented Oct 10, 2012 at 18:06
  • @HagenvonEitzen - Just to confirm what was said by Nihil. This answer is mostly correct, except for using 个 as a catch all. This might be acceptable for someone just starting to learn the language where it can be forgiven, but it will just look wrong anywhere else.
    – going
    Commented Oct 11, 2012 at 3:40

For classifiers there are actually a large range and, contrary to the answer above, there doesn't exist a catch-all classifier which can be used in all situations.

The classifier 个 is the most common one and is used by new learners when they cannot remember which one to use, but it is important that the correct classifier is used in each case.

What is tricky is that sometimes different classifiers can be used with the same item (as you saw with 房子. This is usually region specific or dependent on what the object itself looks like. Some examples for this include the word for dog, you can use 一只狗 which is used for many animals or 一条狗 "条" is used for something that is elongated.

As per your question, when counting from 1 to 10 the classifier is not going to change. However if you asked different people to type in 3 houses, you may get 5 or 6 different inputs. How you hand someone typing 3 or 三 into your program will need to be handled by the program as some people may type it differently. However, I think more than likely although 3 is acceptable, you will find more people would type 三个房子 than 3个房子.


Head(n.) --- Measure Word / Quantifier

房子 --- 间(for Room);栋or幢(for House/Building);所(for the abstract)

房屋 --- 间(for Room);套(for the abstract esp. real estate property right)

家 --- 户(for Family) e.g. 一户人家(a family)

公司 --- 家(more formal);间

猫 --- 只(always use this word only but not else!)

That's natural Chinese:)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.