Why Literary Sinitic is so darn hard

« previous post | next post »

Two days ago, in "Difficult languages and easy languages, part 2" (5/28/19), we listed scores of languages from easiest to hardest to learn.  Spanish came out overall as the easiest widely spoken language for many people to learn, while Arabic and Turkish struck many people as quite difficult to master.

If you want to experience a language learning task that is really and truly daunting, give Literary Sinitic (LS) / Classical Chinese (CC) a whirl.  To me it is far more challenging than Sanskrit or Classical Greek or Latin.  With their moods, tenses, gender, number, conjugations, declensions, paradigms, and whatnot, learning these languages is most assuredly tremendously demanding, but they cannot begin to compare with the humiliating experience of trying to make sense of a passage in LS when its subject is not explicit, the time when it happened is not clear, and all you know is a succession of uninflected morphemes interspersed with some often highly ambiguous particles that may indicate more or less pertinent relationships among the morphemes.

I remember how I began the learning of Sanskrit.  We started with Perry's Primer to pick up the basic grammar and core vocabulary.  Having mastered the foundations and clutching Whitney's Grammar and a good dictionary (I was attached to Monier Monier-Williams), you jump right into real texts.  For me it was Lanman's Reader, starting with Mahābhārata, then Hitopadeśa and Kathāsaritsāgara, and off we went.  It was hard, but if you were conscientious it was doable.  It was all very precise and explicit, so long as you knew the rules.

With LS it is different.  To comprehend a LS text, you have to possess immense learning and uncanny intuition, or else you just memorize it and rely on your teacher's explication of the meaning.  To understand LS well, you have to read vast amounts of it.  It's similar to learning how to write Chinese characters:  practice practice practice!

I've been teaching First-year Classical Chinese for forty years, and I love to do it.  Everybody who comes knows that I spend the first few days dissuading those who are not genuinely serious about the class from taking it.  I tell them this will be the hardest class they have ever taken, and it is (except perhaps for organic chemistry).  Those who stay the course usually receive an A or A-, seldom a B+, B, or B-.  Occasionally I have to give a C, D, or alas, an F, when the poor student really wants to learn LS but just cannot do it.  Some of them take it a second time, but it still doesn't sink in.  They just don't have what it takes.

Usually the students work together to prepare for the daily recitations.

Those who stick with it have a mind-blowing / bending / altering / expanding experience.  They discover different dimensions of thinking (nooks and crannies of the brain) that they never knew existed.  Even though I've been going through the beginning stages of learning to read LS / CC with a new batch of students in the fall semester for four decades (it's a two semester course), every year it's just as exhilarating as it was the first time.  I never get tired of bringing these ancient, dead texts alive for the students.

By the way, I still use Harold Shadick's A First Course in Literary Chinese (originally from the University Press of Cornell, where he taught), even though it's been out of print for decades, uses Wade-Giles Romanization, and is built upon a linguistically daunting apparatus, because its three big volumes provide the best selection of texts with vocabulary notes, commentaries, and systematic grammar.  If Bryan Van Norden's Classical Chinese for Everyone:  A Guide for Absolute Beginners (Hackett) comes out in time later this year, I may give the students a spin through it first before diving into Shadick, or I may just assign it as supplementary reading throughout the year.

Oh, incidentally, unlike almost all other courses of LS / CC across the country, there are no prerequisites for my course, and students may pronounce the characters in Mandarin, Cantonese, Japanese, Korean, etc.


Which is harder: Western classical languages or Chinese? ” (3/6/16)




Selected readings

"Why Chinese Is So Damn Hard" (8/91)



"How to teach Literary Sinitic / Classical Chinese" (9/6/18)

"Aphantasia — absence of the mind’s eye" (3/24/17)

"The miracle of reading and writing Chinese characters" (3/26/17)


  1. Philip Taylor said,

    May 30, 2019 @ 6:45 am

    One thought occurs to me. It is no secret that you (VHM) believe that modern Chinese (MSM/Putonghua/Hanyu) would be better taught (and perhaps even better written) through the medium of pinyin rather than hanzi; have you ever given thought to developing a transcription system analogous to pinyin which could be used to transcribe, and then to teach, literary Sinitic ?

  2. Neil Kubler said,

    May 30, 2019 @ 7:04 am

    Completely agree with Professor Mair's comments above. Is Literary Sinitic/Classical Chinese (CC) a separate language or is it "Chinese"? This is a question where non-Chinese and Chinese will typically disagree. For most Chinese, CC is "of course" Chinese; how could it be otherwise? In addition to history, culture, and education (some CC is taught in China and Taiwan beginning in middle school), one important reason for this view is the nature of Chinese characters, which obscure the fact that pronunciations have changed greatly, as have words (no. of syllables and meanings) as well as grammar. Moreover, there are important differences between the CC of different historical periods. For the foreign student learning high-level Modern Chinese, learning Classical is a little like "learning a language within a language" and could be compared to studying Latin along with Italian or learning Old Norse in conjunction with Norwegian. When I was with the U.S. State Department, officers who hoped to test at the R-4 level or above (scale ranges from 0 to 5) had to possess elementary reading proficiency in Classical Chinese. This reflects the fact that high school and college grads in China or Taiwan can more or less get the gist (though often not all the details) of poems and simple texts in CC, which do appear in part or as a whole from time to time in newspapers and other modern writings.

  3. Kristian said,

    May 30, 2019 @ 8:01 am

    Is Literary Sinitic a language one could learn to compose new texts in? Harry Potter and the Philosopher's Stone was translated into Classical Greek and Latin. Could someone translate it into Literary Sinitic and what would be involved in doing that?

  4. R. said,

    May 30, 2019 @ 8:44 am


    Yes, Harry Potter (or at any rate, retellings of the tales therein) have been translated into Literary Sinitic (Wenyanwen) https://www.zhihu.com/question/48292884

    As for new texts composed in Literary Sinitic, there is the glorious Classical Chinese Wikipedia:


    Here is the article for 'mayfly' which i always found somewhat poetic:


    I recall reading a book, written and published around 2003 in Taiwan that (I think) was written entirely in Literary Sinitic. For what it's worth, the book was a divination manual.

  5. Alex T said,

    May 30, 2019 @ 9:39 am

    I definitely appreciate the smatterings of LS poetry that I can read (real basic stuff), but where LS has the most interest for me is as a written lingua franca in East Asia. I'm speaking of the type of LS that would have been written by Vietnamese scholars meeting up with Korean scholars, or any other literate people of a time before vernacular writing was popular in Asia.

    I have a couple of books from the late 19th century that seem to have instructions for this sort of LS, but I wondered if any readers here know of good examples of authentic texts that fit this category.

  6. Victor Mair said,

    May 30, 2019 @ 10:15 am

    From Ronan Maye (who took LS / CC with me a few years back):

    Having just taken organic chemistry last year, I can tell you LS is definitely harder! One of my organic chemistry teachers did a master's in EALC at Harvard (Japanese) and knew classical Japanese, and he seemed to think chemistry was easier than that too, so maybe that's the only subject that can compete with LS. After a couple years of working in organic chem, you become completely fluent in it, but LS has a much steeper and longer learning curve it seems to me. I remember spending hours on just one small page of text when writing my thesis. Once I get through my med school applications I want to start working on my LS again and try to read Zhuangzi

  7. Doctor Science said,

    May 30, 2019 @ 10:46 am

    Dr. Mair:

    Those who stick with it have a mind-blowing / bending / altering / expanding experience. They discover different dimensions of thinking (nooks and crannies of the brain) that they never knew existed.

    Can you elaborate (or have you done so already in an article or post)?

  8. Katelyn said,

    May 30, 2019 @ 11:23 am

    "some CC is taught in China and Taiwan beginning in middle school"

    古诗 (ancient poems) may be taught to Mainland Chinese students, starting in first grade. In first grade, children read relatively easy passages, look up words in the dictionary, and figure out the meaning. I speculate the teachers may give an explanation as well.

    Before primary school, or during the kindergarten years, children may recite ancient poems as part of 国学.

  9. Victor Mair said,

    May 30, 2019 @ 11:59 am

    From David Knechtges:

    I must confess that I never found classical Chinese especially difficult. I began formal study of it as an undergraduate at U.W. The instructor was Erwin Reifler, who used the Creel "inductive method" textbooks. We had to memorize virtually every text we read. Part of the final exam was to write out selected passages. Like Creel, Reifler was a strong advocate of the "pictographic" theory, and even then, I was skeptical of his explanations of the meaning of words and his analysis of the grammar. He claimed the particle 焉 was a picture of a fist pounding on a table, and thus when we translated it, we had to render it as "thump." Despite Reifler's misconceptions of the language, I managed to learn to read quickly. The first quarter of his class I began reading Han fu, and I managed to translate two long pieces for a course with Hellmut Wilhelm. By the way, I never thought fu were especially difficult. Even today, I find more problems with some short 詩 pieces, especially 四言詩, a number of which are included in the Wen xuan. Recently I have had to spend several days on a single line. The main problem stems from determining how a Wei-Jin-Nanbeichao writer reuses an earlier phrase especially from the Shi jing and Shu jing. Despite their "canonical" status, there was still no fixed interpretation of these texts, and even though Li Shan often cites the Mao commentary version, it is clear that the early medieval writer was following a different version. However, that is not a problem unique to classical Chinese texts.

  10. Victor Mair said,

    May 30, 2019 @ 12:05 pm

    @Doctor Science

    If others don't weigh in to answer your question within a day or two, I'll tell you a thing or two about what you'll discover when coming to grips with LS / CC.

  11. David Marjanović said,

    May 30, 2019 @ 12:21 pm

    I've studied enough organic chemistry to concur that it's much easier than LS. Electrochemistry, though, which requires a lot of advanced math, may get there.

    From David Knechtges:

    I must confess that I never found classical Chinese especially difficult. […] We had to memorize virtually every text we read.

    For most people, these two sentences contradict each other.

  12. Katelyn said,

    May 30, 2019 @ 1:34 pm

    On the Which is harder: Western classical languages or Chinese? page, julie lee gives a glimpse at the mind of non-native speaker. She lists examples:

    "moon" , yueliang (MM), yue (CC)
    "today", jintian (MM), jin (CC)
    "or", huozhe (MM) , huo (CC)
    "time", shijian (MM), shi (CC)
    "happy" kuaile (MM) , le (CC)
    "garden" , huayuan (MM) , yuan (CC)

    This is not how first-language speakers think. yueliang is meaningless. It must be yuè liang, because that is how first-language speakers are taught. yuè liang is the Pinyin form of 月亮. 月亮 forms a 词 ("word") in Chinese, but this word can be broken down into two words – 月 and 亮. 月 means "moon". 亮 means "brightness of the moon" in this word. 举头望明月 is a line from a famous Tang poem 《静夜思》.

    jintian is not Chinese. 今天 is Chinese. 今日 may also be used. In SWC, 今日 may be used because it is more formal. However, in some Mandarin dialects (dialects within the Mandarin branch of Chinese), 今日 may be used colloquially to reflect the spoken language. 今天 is the "word" that can be broken down into two words 今 and 天. 今 is the adjective. 天 is the noun. 今月,今年,如今,等等。

    huozhe is not Chinese. 或者 is Chinese. It may be shortened to 或. It is NOT shortened to 者, probably because 者 may refer to 人.

    I am reminded of a Wikipedia article called "Westernised Chinese language": https://en.wikipedia.org/wiki/Westernised_Chinese_language
    I personally see this style of writing a lot among second-language learners whose mother tongue is one of the European languages, and among Chinese translations from English-language works or other European-language works. The Chinese translations always sound too wordy. For example, a Chinese translation of an American English work, I recall, has 红颜色的旗子. A Chinese storyteller writing toward a Chinese audience would not have to worry about the limitations of speech, and will likely write 红旗 to convey the same meaning.

    As for Why Literary Sinitic is so darn hard, this post gives a brief explanation why it is hard for non-native students. 文言 (Chinese term; Koreans, Vietnamese, and Japanese may use 漢文) is difficult for them too, because of its heavy use of literary references and allusions as well as its extremely abbreviated style. Truthfully, the Chinese characters of 文言 can be all looked up in the 字典, and this dictionary provides literary and colloquial meanings of each character. But reading classical texts is more than just knowing the meaning of the characters; the reader must know the literary references and allusions, hence the work is extremely high-context.

  13. Alec Story said,

    May 30, 2019 @ 2:02 pm

    Perhaps my opinion on this is because I am a relative novice at Literary Sinitic, but in my reading of later, more technical, texts, such as 齊民要術 or 北山酒經, I haven't run into the problems Dr. Mair describes in this post.

    I wonder – but certainly do not have the breadth of study to be able to evaluate – whether later texts, and texts that are not philosophy or theology, are simpler and more straightforward.

    It was certainly my perception that a lot of the texts from the classical period are intentionally obtuse.

    I took two semesters of Literary Sinitic / Classical Chinese at Cornell in 2010-2011, and our class had a similar requirement to yours: you had to have *some* familiarity with Sinitic characters, but where that familiarity came from didn't matter. However, at least in my class, everyone was reading in Mandarin.

  14. Scott P. said,

    May 30, 2019 @ 2:19 pm

    But reading classical texts is more than just knowing the meaning of the characters; the reader must know the literary references and allusions, hence the work is extremely high-context.

    What if you start by reading the early texts, before there is any literature to make reference to?

  15. Christopher Coulouris said,

    May 30, 2019 @ 3:03 pm

    Professor Mair,
    In previous posts you exampled how you went about learning to speak MSM and how you learned to read Chinese. If you have time could you please recount your approach to learning Classical Chinese? Did you find supplementary materials on you own like you did for learning MSM or did your strictly use the materials your professors prepared. Thank you.

    Chris Coulouris

  16. cliff arroyo said,

    May 30, 2019 @ 3:26 pm

    "have you ever given thought to developing a transcription system analogous to pinyin which could be used to transcribe, and then to teach, literary Sinitic ?"

    I'm wondering if General Chinese might not work for that (and if not, why not?)

  17. Michael Sinclair said,

    May 30, 2019 @ 4:12 pm

    This is a marvelous description of my experience studying Classical Chinese!

    I was a graduate student at Stanford, specializing in European history when the Ford Foundation created a program — essentially a double major — to continue in European history while learning the Chinese language and Chinese history. We had the regular language courses during the nine-month academic year, but in the summer we had intensive courses.

    During the summer of my third year I studied Classical Chinese with Allyn Rickett. Two days ago I was reminiscing about that course, which I regard as the most demanding and rewarding experience that I ever had in my many years of education. Professor Rickett included a segment that was especially valuable to me; we were required to read Qing documents. We spent many hours outside class in the Hoover East Asian Library, and we soon realized that we must wait until the course was completed before we could get more than four or five hours of sleep.

    It was as close as I’ve ever come to total immersion in a subject, and I treasure the experience as I rapidly approach my eightieth birthday. I spent thirty-eight years teaching at Wake Forest University, where I taught history and spent ten years teaching the first two years of Chinese (which were the only Chinese language courses that WFU had) while trying to persuade the administration to recruit a full-time faculty. Success came when a committee that I chaired brought us Pat Moran, who was an undergrad at Stanford and completed his Ph.D at Penn.

    I still have the final exam in Classical Chinese and remember the pleasure I had when I made a 97.

    My study of foreign languages is quite limited compared to scholars like yourself. I studied, in order, Spanish, German, French, Chinese, Japanese (not enough), and Russian. The latter seemed remarkably easy after Chinese.

    Thank you, sir, for your expertise and insight into the difficulties and rewards of Classical Chinese.

  18. Victor Mair said,

    May 30, 2019 @ 5:23 pm


    Most of the students in my LS / CC classes are native speakers from Taiwan, Hong Kong, PRC, Singapore, etc. There have also been students from Korea and Japan, as well as from many other countries.

    You have a lot of catching up to do in orthography, morphology, phonology, grammar, prosody, psychology of language, historical linguistics, language reform during the past century and more, and in linguistics generally.

  19. AJ said,

    May 30, 2019 @ 5:31 pm

    On the subject of a transcription system for Literary Sinitic:

    Recently, I've taken to using Early Middle Chinese as a medium to pronouncing the characters. As not all Middle Chinese reconstructions are in agreement, and the likelihood that "Middle Chinese" as we know it from rime books is likely an artificial diasystem, I've probably only come up with a mere approximation to some dialect a literati might use to read aloud a poem in the Tang dynasty.

    The homophones are reduced enough that I could likely read aloud a text to someone else who understood this 'Middle Chinese' and have it be fairly intelligible.

  20. Bathrobe said,

    May 30, 2019 @ 6:36 pm

    I had trouble understanding Katelyn's comment. It seemed to boil down to: pinyin isn't real, only Chinese characters are real, but I may have missed the point.

    It did seem to miss the fact that Literary Sinitic is largely a written medium and a historically prior form of the language. Since characters generally (but not always) spell out morphemes, they are the bridge to LS for the literate native speaker. Would it work for illiterates? I would be curious to know how easily an illiterate speaker of MSM (if you could find one) could learn LS.

    The point about Westernised Chinese is that it insists on physically hauling every single word from the Western language over into the Chinese text, an extreme form of literalism, where Chinese would conventionally use more succinct turns of expression. This certainly ties into real linguistic differences.

    But there is a certain kind of cart-before-the-horse-ism here. Spoken MSM can be very wordy indeed. It is when learning to write that the emphasis is placed squarely on Chinese characters (look at children's textbooks if you want to understand this) and from the very start children are taught to write in a terse manner — by relying on the ability of Chinese characters to spell out meaning. The use of Chinese characters and the tendency to terseness are culturally induced characteristics and are not necessarily inherent in the language. The fact that LS is the main source of this terseness, and that Chinese characters form a bridge to LS for the MSM speaker, would appear to explain much of what you seemed to be saying — if I understood it correctly.

  21. Bathrobe said,

    May 30, 2019 @ 9:15 pm

    Let me clarify my previous comment.

    The Chinese child learns at her mother's knee that yuèliang refers to the moon.

    She goes to school and learns that this is written 月亮, made up of 月 yuè 'moon' and 亮 liàng 'bright'.

    She learns it is more 'elegant' to use 月 alone than in the combination 月亮. She might learn some poems by heart, like 举头望明月. She will learn that 月 also means month, and that colloquial 这个月 zhèige yuè is more elegantly expressed in written Chinese as 今月 jīnyuè. This is in keeping with the traditional emphasis on brevity in written Chinese.

    If she later studies Literary Sinitic, her knowledge of the character 月 and her familiarity with succinct non-colloquial expressions will help her to understand this older written language.

    My point is that your statement 'This is not how first-language speakers think' ignores the process of language acquisition that all native speakers of a language go through. Yuèliang is not meaningless at all to the native speaker; it's the Chinese word for 'moon'. Only a skewed emphasis on literacy could result in a denial of its meaningfulness.

  22. Chas Belov said,

    May 30, 2019 @ 11:37 pm

    In case anyone is interested, it is my understanding that 稀微的風中 by Wu Bai from the Legend of the Sacred Stone movie soundtrack is in Classical Chinese.

  23. liuyao said,

    May 31, 2019 @ 10:47 am

    One must conclude that David Knechtges was "born" to study Classical Chinese. (I somehow got hold of a copy of 中華書局 edition of 抱樸子 with Knechtges's name inside.)

    On a more serious note: How much would it alleviate the difficulty in learning LS as compared to Sanskrit if we had better tools for looking up characters? (I apologize if you have already talked about it in one of your earlier posts.) In some ways, what you regard makes the Classical Western languages (and Sanskrit) "easy" might be partially afforded by a familiarity with Sinitic characters, and some key phrases that have come down to modern Chinese.

    I too am interested in knowing in what ways a course in LS is a mind-bending experience, particularly for non-Asians who had had no exposure to it. If someone just took it for one or two years (and does not use it on a daily basis), how much do they retain their reading abilities, or is it mostly a sense of the general features of LS?

  24. Victor Mair said,

    May 31, 2019 @ 10:52 pm

    Difficulties similar to those I've outlined for LS apparently also abound, and there are differing opinions among the professoriate on how to approach them:

    "Is 'Reading' Latin Impossible?", by Tom Keeline, Latinitium

    What a tour de force blog post this is, all focused on a single difficult word, inuleo, in one of Horace's Odes, 1.23.

    Joe Farrell, comments: …The "nature method" approach to Latin has been gaining ground, to the point that some people have blogged that people who have been taught Latin the "classical" way are the only ones who can't actually read it. Keeline's blog (I think) is meant to be an implicit rebuttal to that point of view.

RSS feed for comments on this post