How many Chinese words do you know?

Here's one for English. Pretty interesting to compare.

http://testyourvocab.com/

33,600 English words, 8,685 Chinese words. I reckon that I am possibly about level with an ambitious 8 year old Chinese kid or a very lazy 11 year old, but even that is a troublesome measure because I have a different vocab set to most kids.

I tend to agree that the test over-estimates your count of Chinese words but I am not able to say why. I am reminded of that game - you get one eight or nine letter English word and you have make 35 or 40 words out of the letters, only using each letter at most once, no plurals. With any dozen chinese characters you should be able to make hundreds of words, so that would suggest that your word count should be way above your 'character' count.

None of this seems very scientific, but good fun - thanks for posting RJ.

Hmmm. When I chose the option to test using words, I also scored way above my expectations, and decided the test does overestimate one's vocabulary level. But, when I chose the option to test using characters, I scored way below what I expected, and decided I must focus more on recognizing characters out of context. I found it surprising that the test says I know over 10 times as many words as I know characters.

Rachel - how does this one work? (Silly question?) There was one short very easy reading on the home page - say I could read it all. I don't see how it quantifies the characters I know. Missing something. There are only three tabs - the others are 'research findings' and T-shirts. What did you think of the research findings by the way - I was surprised he said that he didn't know how to tell what a Chinese word is. He lacks the certainty of a native Chinese speaker? :)

Don't worry, I had a similar problem when I first encountered this. If you say "too easy, know them all" it doesn't LOOK like it, but the page refreshes super fast and a new text appears for you to read. There are a total of three texts to get through. Assuming you know all the words and mark none as "unknown" then the algorithm extrapolates that you would have the vocab of a native speaker. But if you mark words out as unknown, then the algorithm counts and evaluates, and reduces the estimate.

Regarding "what is a Chinese word" I actually think it's a somewhat valid question. The question comes down to how wide a grey band there is between compound words and short phrases (hint: in Chinese it's a very wide grey area). Think of German, which tends to compound a lot: "geschwindekeitsbegrenzung" is a single word, but why isn't it gescwindekeits (speed's) begrendzung (limit), two words? Because that's just how German is.

English has, I think, a pretty wide grey area because (at least in the American varieties) we tend to argue about spelling compound words. I used to be responsible for maintaining the style manual for my marketing department, and we'd get into real knock-down drag-outs over words like setup. Set up? Set-up? I think we ended up compromising that "set up" is a verb, "setup" is a noun, and "set-up" is an adjective.

For instance in the test referenced here, he chose to count 现代社会 (modern society) as one word, not two words 现代 (modern times) 社会 (society)。When you're taking the test, you can't say you know 社会 but don't recognize 现代。So his choices about how to parse the text mattered.

John did a post on it on Sinosplice donkey's years ago.

Rachel - thanks for the low down on the test. Maybe I need to re-fresh it, I stayed on the page quite a while; is it possible I didn't notice a change in text!? Yes ... :)

On word counts .. I suppose so .. but I expected him to be something of an expert, there are rules. I wouldn't think that a knowledge of either German or English helps much, this is Chinese.

I did a reading course and you were set reading tests and you either got the word count right or wrong. This is China. There is something of a probability question when you get ambiguous sentences but I wouldn't think this happens with great frequency, and there is always a most likely option. Fun when it does happen, as in John's example on sinosplice (thanks for that) - but his example kind of proves my point, there is a single most likely interpretation.

现代社会 is I think two words, but it is quite a while since I looked at this. I take your point that using his rules (one word) you had to understand the whole expression to 'pass' - something that knocked my score around.

Taking this test was more fun that the other tests! It says I know far fewer words than the test listed above, but I think it's probably more accurate. At any rate, it was enjoyable to see what meaning I could glean from the passages it gave me.