Spoken Classical Chinese

« previous post | next post »

Tom Mazanec saw an announcement of a course in Hong Kong, in which the teacher, Dr. Lai Chi Fung 黎智豐, proposes to teach Classical Chinese by focusing on vocabulary for everyday use, just as if one were learning a foreign language. So you learn greetings, introductions, and the like. The idea intrigued Tom, especially since the language of instruction is Cantonese (which would make spoken Classical a little more intelligible than if it were in Mandarin).

If the course were taught in Mandarin, I think it would be a big flop, but since it is being taught in Cantonese, there is a somewhat higher possibility of limited, meaningful spoken communication.  Cantonese preserves more features (phonological, lexical, grammatical, syntactical) of earlier stages of Sinitic than does Modern Standard Mandarin (MSM).  Just in terms of phonology, Cantonese has more than 2,200 possible different syllables, almost twice as many as MSM (source).

Classical Chinese / Literary Sinitic (CC/LS) has been a dead language (not a living, spoken language) for at least two millennia.  As I have said countless times, one cannot have a spontaneous, unrehearsed conversation in CC/LS.  This truism is reflected in Y. R. Chao's (1892-1982) concept of "Sayable Chinese" (see "Selected readings" below).  We have spoken Latin, spoken Sanskrit, etc., but we can't really have "spoken CC/LS", because it is a purely "book language" (shūmiànyǔ 書面語) or "written language".  Of course, our dictionaries of English, German, French, Russian, and so on are sprinkled with entries that are designated as "lit(erary)" or "writ(ten)", but can you imagine trying to hold an interesting, unfettered conversation while limiting yourself only to "lit" terms?


Selected readings


  1. AntC said,

    June 29, 2021 @ 5:19 pm

    I thought the CCP is trying to suppress all topolects other than Putonghua(?) And that Cantonese had all but died out on the mainland.

    So is this course a back-door way to keep Cantonese alive in HK? That is, in an Academic context.

    I wish Dr. Lai well, but I fear he'll get arrested. Just as soon as the CCP has dealt with those opposing the National Security Law.

  2. Julian said,

    June 29, 2021 @ 7:53 pm

    Just curious: By 'purely book language/ written language', do we mean-
    1. We don't know how it was spoken; and/or –
    2. The written remains are an inadequate record of how the language would have been used in daily life, and/ or –
    3. The written remains, even at the time they were first written , were a prestige code that did not correspond to the language in daily use (like writing in Latin in Dante's time)
    BTW, I resist the idea that Latin is a 'dead language'. It's the language of the people of Italy. It's just evolved quite a lot and at some point changed its name (but that's a detail). I concede that this is really a point about the definition of terms.
    How much a language has changed over time, and whether the written remains of some earlier time are a good guide to the language in daily use at that time, are separate variables.

  3. Nick Williams said,

    June 29, 2021 @ 8:38 pm

    I don't agree with the rigid distinction you are making here between a speakable language and classical Chinese. It's unfair to treat as a linguistic matter what is in fact the consequence of a cultural transformation. It's hard to chat about the US Open in classical Chinese, for sure; but what does that really signify? When I studied Latin in high school, my classmates and I would never have dreamed that anyone in the 20th century could converse in the language, and yet, mirabile dictu…

    From the point of view of actual practice today, I'm reasonably sure I've heard well-educated Chinese speakers have conversations, or, especially, make formal spoken remarks at conferences, that were largely in 文言. You can witness it being spoken in certain TV dramas:
    though perhaps other readers can offer better examples than that?

    To some extent I think your original post is like saying that nobody has spontaneous conversations that sound like the writing of Edward Gibbon or like Hamlet's monologues, which is true, but more a matter of formality, register, diction, complexity of argument, concentration of rhetorical devices, and similar factors than of language per se.

  4. Victor Mair said,

    June 29, 2021 @ 8:46 pm


    #3, but even more different than Latin was from the vernacular in Dante's time.

  5. stephen reeves said,

    June 29, 2021 @ 8:49 pm

    Latin is basically classical Italian

  6. Alex said,

    June 30, 2021 @ 1:15 am


    Cantonese has certainly not "died out" in Mainland China. Social media is full of videos in, about, and celebrating all kinds of topolects, and a recent major TV drama was entirely filmed in Fujian and Ningxia topolects, including the characters representing Party officials. Here's a relevant article in a state-owned newspaper: http://www.chinadaily.com.cn/a/202102/05/WS601c9ee2a31024ad0baa76bb.html

    The decline of topolects and rise of Modern Standard Mandarin is a significantly more complicated picture than "CCP says Mandarin good and all else bad."

  7. David Marjanović said,

    June 30, 2021 @ 4:22 am

    I thought the CCP is trying to suppress all topolects other than Putonghua(?) And that Cantonese had all but died out on the mainland.

    There are still lots of people alive who were born before the People's Republic was even founded, let alone before it became official policy (whatever exactly that means, see above) to equate Putonghua with "Chinese". It's possible that some topolects are moribund now, in the sense that there aren't any children learning them anymore, but I doubt even that.

    I resist the idea that Latin is a 'dead language'. It's the language of the people of Italy. It's just evolved quite a lot and at some point changed its name (but that's a detail). I concede that this is really a point about the definition of terms.

    Sort of. First, Italian isn't that much closer to Classical Latin than any other Romance language is (French and Portuguese have moved far away in pronunciation, but not in other respects) – except that Standard Italian, ironically enough, is based on 13th-century Florentine, while the standards of other Romance languages have much more recent bases.

    Second, no Romance language is descended from Classical Latin. They're descended from so-called Vulgar Latin. In Cicero's time those were simply the literary and the spoken register of the same dialect*, but 200 years later the difference was well beyond that.

    * The Roman dialect of Latin, as opposed to the Praenestine or the Faliscan one. The latter two are extinct by any definition, though nihil may be a Faliscan loan in Roman (with the Faliscan f > h shift from *ne filum "not even a thread").

    By 'purely book language/ written language', do we mean-
    1. We don't know how it was spoken; and/or –

    Very detailed reconstructions of the pronunciation used at the imperial court at two points in the Middle Ages do exist now, and less detailed ones for other times, but these are 1) only known to a few historical linguists and 2) not easy to learn, e.g. because none of these sound systems lines up exactly with that of any one, or the sum of two, topolects spoken today.

  8. Christopher Nugent said,

    June 30, 2021 @ 9:23 am

    In a similar manner to Nick Williams, I would respectfully disagree with the sharp distinction behind the claim that Literary Sinitic "has been a dead language (not a living, spoken language) for at least two millennia." Even if they were not holding conversations about their daily lives in LS, hundreds of thousands of people (perhaps millions) were writing, and thus thinking, in LS up through the first half of the 20th century. They contemplated their world and expressed new and original thoughts about it in LS. I would assume their thoughts were phonetic, i.e. they where thinking in the sounds of a language. To my mind (so to speak), this at least indicates a "living" language, if not a true "spoken" one.

  9. Victor Mair said,

    June 30, 2021 @ 12:15 pm

    I humbly request that Nick Williams and Christopher Nugent wait until the publication of the forthcoming massive dictionary of Middle Vernacular Sinitic that Zhu Qingzhi and I have been laboring on for more than two decades before they come to the conclusion that there is no significant difference between vernacular and LS/CC.

    And read Lu Xun's magisterial An Outsider's Chats about Written Language (item 91 in The Hawai'i Reader in Traditional Chinese Culture). And familiarize yourself with Y. R. Chao's concept of "Sayable Chinese" referred to in the o.p. and "Selected readings".

  10. Antonio L. Banderas said,

    June 30, 2021 @ 12:38 pm

    @Victor Mair

    Will that be the next publication in the ABC Chinese series ? If so, any ETA?
    Otherwise, any info. about which one might be?

    Literariness (Wen) and Character (Zhi): From Baihua to Yuluti and Dazhongyu

  11. Christopher Nugent said,

    June 30, 2021 @ 12:47 pm

    Victor– I do indeed look forward to this new publication! It sounds wonderful. I also have students read Lu Xun's piece (from that very reader) in my class on the Sinitic script.

    I'm not at all saying that "there is no significant difference between vernacular and LS/CC" and do not in the slightest believe that to be the case. What I took issue with was the claim that LS/CC has been a dead language for over 2000 years. These are completely different points. Vernacular and LS/CC can be utterly different in every respect and LS can still have been a "living" language 100 years ago, as a substantial number of people continued to write and, the key aspect of my point, THINK in it.

    We may simply have different notions of what counts as "living" here. For example, I would say that Latin was a "living" language for Newton, as he clearly formulated many of his most profound thoughts in it (at least when writing the "Principia"). If by "living" you mean "regularly used in everyday conversation," then I would agree with you. I do believe, however, that we can and should make finer distinctions than that.

  12. julie lee said,

    June 30, 2021 @ 11:34 pm

    Somehow I've always believed classically educated Chinese gentlemen of as late as the first half of the 20th century could converse among themselves in Classical Chinese.

    I'm reminded of how the philosopher Mou Tsung-san once translated Hamlet's words "to be or not to be" into Classical Chinese and then into Modern Mandarin:

    Classical Chinese: "生歟? 死歟?“ (sheng yu? si yu?)
    Modern Mandarin: "存在呢? 抑或不存在?“ (cunzai ne? yihuo bu cunzai?)

    I always marvel at the brevity of Classical Chinese.

  13. Mehmet Oguz Derin said,

    July 1, 2021 @ 1:09 am

    I always wondered about how modern native Mandarin speakers thought about the vernacular use of Classical Chinese. Old Turkic (both inked and inscribed texts) never felt remotely distant in a way that it couldn't act along with a daily conversation in Turkish, but maybe that's just my confirmation bias!

  14. Victor Mair said,

    July 2, 2021 @ 8:44 am

    Since the time he assumed office, I have observed Xi Jinping make numerous speech errors, especially when he tries to throw in something literary. Here are some examples:

    "Annals of literary vs. vernacular, part 2" (9/4/16)


    "Latin Caesar –> Tibetan Gesar –> Xi Jinpingian Sager" (3/20/18)


    "Pinyin for the Prez" (10/25/18)


    "Xi Jinping's reading errors multiply" (12/28/18)


    "More literary troubles for Xi Jinping" (1/3/19)


    A similar error made by the president of Peking University:

    "Peking University president misreads an unobscure character: monumental implications" (5/5/18)


  15. Ronan Maye said,

    July 8, 2021 @ 1:13 am

    This is an interesting idea, but I think there is a better direction they could take this. Instead of focusing on conversation (which would be stilted and artificial), it would be super interesting to see textbooks for Classial Chinese (C.C.) that use the ideas from language acquisition research about comprehensible input. There is no point in trying to speak a written language, but if there were a C.C. textbook that replicated what Lingua Latina Per Se Illustrata did for Latin, it would probably be a huge hit.

RSS feed for comments on this post