New York high school Chinese test

« previous post | next post »

Zhuang Pinghui, in the South China Morning Post (1/18/17) has an article that is truly baffling:  "US high school Chinese test stumps internet users in China".

A high school in New York has produced an exam paper for its pupils learning Chinese which features questions that have daunted internet users in China and even a college professor.

The final exam for pupils at the Franklin Delano Roosevelt High School’s Foreign Language Department comprises four sections, according to a photograph of the test paper shared on Chinese social media.

The first two focus on words and idioms not commonly used in conversational Chinese.

In one part pupils were asked to give synonyms for 10 words, but they are more often used in ancient Chinese writing than in everyday speech.

Many Chinese social media users admitted they struggled to read even the first word – jiu ju – which means to live in a rented apartment. Many didn’t even know what the word meant, let alone come up with a synonym for it.

“At first, I thought the question was to write down the pinyin, but after reading it I realised I didn’t even know how to read the word,” one internet user wrote on social media.

Another question in the first section requires a synonym for the word he, meaning to bite, which was also mainly used in ancient writing.

Pupils were also asked to give antonyms for 10 words and idioms in the test and were required to write a 300-word essay.

Topics ranged from “The Inspiration of Lotus”, a reference to the Song dynasty (960-1279) philosopher Zhou Dunyi’s essay Ode to the Lotus Flower, to Reflections on “Fat Rat”, referring to a piece of writing by the Qing dynasty (1644-1911) scholar Pu Songling.

“I felt uneasy while trying the test. Now I wonder if I might have learned fake Chinese,” one user wrote on social media.

Another agreed: “I might well be a fake Chinese!”

Wang Hongtu, a professor of Chinese language at Shanghai’s Fudan University, told the news website that the questions were “very difficult”.

The first question alone got him thinking for a while, he said.

Wang said most Chinese people knew 5,000 to 6,000 words, but some words used in the test were very uncommon.

Ohhhhh!  This article raises so many thorny issues that I hardly know where to begin.  So I'll just fire away.

Starting at the very end, Professor Wang Hongtu of Fudan University is wrong to say that most Chinese people know 5,000 to 6,000 words.  In the first place, he's probably confusing "words" with "characters".  Even if he has made the perennial mistake of mixing up zì 字 ("characters") and cí 詞 ("words"), there are exceedingly few people who "know" (i.e., are able to recognize, write, pronounce accurately, and define correctly) more than 4,500 characters.  The average literate person knows about 3,000 characters.  As for words, the vocabulary of most literate Chinese ranges somewhere between 15,000 and 25,000 items.

If you want to get an idea of the 10,000 most frequent Chinese words, they are listed here, with romanization and translation.

There are some useful comments on the matter of how many characters are required for literacy on Quora here:  "How many characters does the average Chinese person know?".  I especially recommend the fifth comment by Shawnxuande Li, posted on July 10, 2016, for its trenchant, informative, insightful remarks covering the rise and fall of Chinese characters from their beginning on the oracle bones more than three millennia ago to the present time.  The final comment, by Anonymous on January 7, is also pertinent:  "One of China's illiteracy standards is knowing less than 1500 characters."  Despite what the Chinese government may tell us about there being near universal literacy in China, applying the 1,500 character standard, from my own experience in the field, I'd wager that well over half the people of China are illiterate, and the late and much lamented Zhou Youguang privately admitted the same thing to me.

Now, jumping back to the very top of Ms. Zhuang's article, there is an image of the opening part of Franklin Delano Roosevelt High School’s CLA Final Exam.  I would like to know who is this teacher, Ms. Luo, who has come up with these fiendishly difficult exam questions for her students.  Somebody needs to inform the head of the Foreign Language Department, Ms. Florio-Fintz, and the Principal of FDR HS, Ms. Katz, that Ms. Luo is teaching her students wildly impractical Chinese.

The first question on the exam is to give a synonym for jiùjū 僦居.

僦 is ranked #5438 in this list of the 10,000 most frequent characters and has a frequency of less than .05% in  a large corpus of Chinese texts collected from online sources.

Jiù 僦 is not the sort of character that a student of Mandarin should be spending time to memorize, and jiùjū 僦居 ("rent a place to stay in") is not a term that a student of Mandarin needs to learn.

It should be pointed out that this is a "CLA Final Exam", where "CLA" means "Collegiate Learning Assessment".  But I wouldn't expect high school students in Mandarin classes to be able to answer these questions, and I wouldn't expect graduate students in Mandarin classes to be able to answer them either.  The ONLY way I would expect any students to perform adequately on this exam is if they had followed a syllabus of Literary Sinitic (Classical Chinese) specifically designed to prepare the students for these very questions.

See also:

"US school's Chinese-language exam leaves native speakers speechless" (, 1/17/17)

"US school's Chinese-language exam leaves native speakers speechless" (China Daily, 1/16/17)

These articles show the complete exam, the entirety of which is totally out of touch with MSM.  It was supposedly for course FMS63 Chinese 3.  The exam consists of four sections:  1. give synonyms for vocabulary items; 2. give antonyms for vocabulary items; 3. make sentences with vocabulary items; 4. write an essay of 300 or more characters on a literary theme from premodern times.

I didn't exhaustively check every single item on the test, but a quick scan of the whole gives the strong impression that it is made up of archaisms and classicisms that one would seldom, if ever, encounter in MSM conversation or even in typical reading.

If this material is being taught as Third-year Mandarin, as seems to be the case from the title ("Chinese 3"), it is a travesty.  The contents are utterly inappropriate for a third-year high school Mandarin class.  If it is being offered as a course in Literary Sinitic (Classical Chinese), that is an entirely different matter.  But I still would question how many students at Franklin Delano Roosevelt High School are in need of such a course and will benefit from it.

[h.t. Mark Metcalf, Bill Holmes]


  1. AntC said,

    January 18, 2017 @ 7:45 pm

    Thank you Victor for your unambiguous assessment.

    I always struggle to reconcile the various metrics I've seen for MSM speakers'/writers' knowledge of words and characters.

    If The average literate person knows about 3,000 characters. ; but well over half the people of China are illiterate [applying the 1,500 character standard]; are those 1,500 characters nevertheless sufficient, by using their pronunciation only, to 'sound out' every syllable for the 10,000 most frequent Chinese words?

    Is 'sounding out' what the semi-literate do? (Which leads to the nonsense readings/translations that you regale us with.) These days I guess people could use pinyin input to 'sound out', but would then be uncertain which of the matching characters carries the right meaning.

    Although over human history, for most cultures, most of the population has been illiterate, it's hard to imagine in the C21st. Everybody has a handheld device. How do they read it? (I guess this explains the popularity of voice messaging.)

  2. Jake said,

    January 18, 2017 @ 9:27 pm

    Has it been confirmed that this test was actually given to students and is not just some kind of urban legend? (Sort of like a less jerky version of that 'ghetto math test' thing)

  3. Chris Barts said,

    January 18, 2017 @ 9:41 pm

    Hm. It sounds like the course was the equivalent of teaching students Latin while claiming to teach them Italian, with your comments about the questions being more appropriate to a course focused on Literary Sinitic. A thought occurs: I wonder if they think that language is more correct than MSM, in the same way the "grammar rules" typically taught at the high school level spectacularly fail to capture how anyone outside of a schoolroom uses English, and are, of course, all the more correct because of it.

  4. krogerfoot said,

    January 18, 2017 @ 10:00 pm

    From the China Daily article: "A netizen, named Xizigebaozhale, said, 'Now I understand what it feels like when English speakers sit for English-language tests for China's College Entrance Exam.'"
    If these exams are anything like the English tests that Japanese students endure, this is a really perceptive observation. Even the better tests, such as the 英検/STEP Test, contain maddening examples of usage that are nearly nonexistent or unknown among native speakers.

  5. VV said,

    January 18, 2017 @ 10:09 pm

    Is there any chance that the words in sections 1-3 of the exam all came from a specific Literary Sinitic text/passage that the the students had recently read? This is a pretty common assignment-and-assessment method in language classes – assign reading, then give a quiz on a sampling of "new" vocabulary from that reading – and while it would still be a pretty unproductive exercise given the obscure vocabulary items chosen, the exam would at least make a lot more sense in such a context.

    (Btw, another V complained that there were two Vs commenting here, so I've switched to VV)

  6. Jonathan Smith said,

    January 18, 2017 @ 11:32 pm

    It seems VV is correct though can't be bothered to check exhaustively — Zhou Dunyi's 周惇頤 Ai lian ji 愛蓮記 contains some of this stuff so the pieces involved must be those to which the essay prompts refer and perhaps others.

  7. Jenny Chu said,

    January 19, 2017 @ 12:09 am

    I was trying to figure out whether Franklin Delano Roosevelt High School is the kind of place where you might find super-students who can pass graduate-level Chinese.

    1. It is 81% "minority" students. Are any (most? none?) of those from a Chinese background, e.g. speaking Chinese at home and/or going to Chinese school on Sundays?

    2. The AP® participation rate at Franklin Delano Roosevelt High School is 25 percent. Is that high or low?

    3. The student teacher ratio is 18:1. That's a lot of faculty, isn't it?

    4. 79% qualify for free student lunches, i.e. are "economically disadvantaged".

    5. The graduation rate is 58%. Whoa! That's low, isn't it?

    Now I feel more confused than before.

  8. Fluxor said,

    January 19, 2017 @ 2:54 am

    "Wang said most Chinese people knew 5,000 to 6,000 words, but some words used in the test were very uncommon."

    Did Professor Wang really say "words", or was this simply a mistranslation from the journalist writing the article? Given Professor Wang is a Chinese language professor, I had my doubts his interview was conducted in English.

    The English article sites "" as the source of Prof. Wang's comments. I was able to find this article:

    Here's what Professor Wang was reported as saying in, "又比如,大部分中国人可以认识五六千个汉字,少有人可以认识一万以上的汉字,所以汉字题想要考倒人家是很容易的。" (My translation: Another example, most Chinese people can recognize five to six thousand characters (hanzi), but few can recognize more than ten thousand; thus, it is quite easy to devise an extremely difficult Chinese test.)

    Clearly, he did not say "words", but mentioned "hanzi" (Chinese characters) explicitly. The English translation of his quote is different enough that one certainly wonders about the quality of reporting at SCMP.

  9. Avi Rappoport said,

    January 19, 2017 @ 2:57 am

    So I have a question: how does this compare to the HSK? While I understand that's very character- and memorization-oriented, it's a standard.

    In answer to Jenny Chu:

    1. I don't know, maybe it's immigrants? Clicking through says 41% Asian, but also high proficiency in English.

    2. 25% is very high, but if 65% pass the test out of that 25%, that's suddenly not so high.

    I was curious, so I checked my daughter's Berkeley High School, 58% take at least one AP and the participant passing rate is 72%.

    3. 18:1 faculty – about the same as Berkeley High (we pay extra taxes for art and music though).

    4. Very high poverty rate – Berkeley High as has 32% – gentrification issues

    5. That's an abysmal graduation rate, Berkeley is 92%

    (please ignore the "District" comparison, it's messy).

  10. Amy Stoller said,

    January 19, 2017 @ 7:28 am

    I can’t answer the questions regarding Chinese, but official information on FDR High can be found at, and the full list of courses offered, including Chinese 1–6, is at Information on Federal and State testing, graduation requirements, and so forth, is also available via the website.

    Wikipedia’s explanation of AP tests reads "Advanced Placement (AP) is a program in the United States and Canada created by the College Board which offers college-level curricula and examinations to high school students. American colleges and universities may grant placement and course credit to students who obtain high scores on the examinations. The AP curriculum for each of the various subjects is created for the College Board by a panel of experts and college-level educators in that field of study. For a high school course to have the designation, the course must be audited by the College Board to ascertain that it satisfies the AP curriculum. If the course is approved, the school may use the AP designation and the course will be publicly listed on the AP Course Ledger.”

    25% participation in AP strikes me as pretty decent, especially when compared to the low (and very poor) graduation rate. It looks to me like academic achievement is improving, especially considering the massive educational disadvantage of low income/poverty.

  11. Andreas Johansson said,

    January 20, 2017 @ 10:19 am

    How comparable are vocabulary scores between languages?

    Frex, acc'd a test I took my English vocabulary is about 40,000. Is it meaningful to say this is better than the average literate Chinese person's 20,000 or so, or are English and Mandarin vocabulary sizes incommensurate in some way?

    (The test I took was at – I can't vouch for the accuracy, but it doesn't really matter for the purposes of my question.)

  12. Victor Mair said,

    January 20, 2017 @ 1:35 pm

    "Dàbùfèn Zhōngguórén kěyǐ rènshi wǔliù qiān gè hànzì 大部分中国人可以认识五六千个汉字" ("Most Chinese people can recognize five or six thousand Chinese characters")."

    Not true.

  13. Peter S. said,

    January 21, 2017 @ 10:39 am

    My grandfather grew up in a rural German-speaking household in Indiana. When he took the entrance exam for Purdue University, the only section he flunked was the language section – German – because the dialect they spoke was so far from Hochdeutsch. (He got in anyway.)

  14. Kevin Li said,

    January 29, 2017 @ 7:05 am

    "How does it compare to the HSK?"

    According to the HSK vocabulary lists I used not even the first word 僦居 is in the entire HSK 6… And Pleco doesn't know this word either (though this is definitely no indicator).

    These are the lists I have consulted:

RSS feed for comments on this post