Chinese Prose Rhetoric.

You can read Christoph Harbsmeier’s “The Rhetoric of Premodern Chinese Prose Style” in a draft version (pdf) or in its final form as a chapter in Victor Mair (ed.), Columbia History of Chinese Literature (; either way, it’s a very interesting take on an ancient tradition, with suggestive comparisons to the classical West. Some excerpts:

Confucius maintained that when words get their message across, one should stop (Analects 15.41). What was admired in Confucius was his flair for wei yen (subtle speech), which, without being yin (hidden, arcane, riddle-like), achieved that peculiar subtle variety of ming (translucence, perspicuousness) which became so essential to the classical Chinese aesthetic. It was of the essence of this translucent, limpid effect that it was preferably achieved with an austere economy of stylistic means, an apparent sparseness of effort, a naturalness, the elegant light touch.

This ideal of translucence and perspicuousness, then, is not an intellectual clarity brought about by elaborate explicitness, definiteness of meaning. The text is designed to inspire in the reader the congenial but active and even creative production of artistic sense. The texts do not impose meaning, they are designed to inspire the creation of sense. […]

The Greeks were known in antiquity as loquacious and contentious people, and Greek texts are also semantically pugnacious: through definition and explicitness they push the reader around, aim to force an intended meaning on him or her, compel him or her to acknowledge an objective truth which the text sets out to make explicit, they aim to appear to give a complete picture of a certain reality. […]

If one surveys classical Greek and Latin literature as well as the Christian literary tradition in so far as it was shaped by this classical heritage, there is no doubt that in traditionalist Chinese eyes these suffer from chronic obsessive overexplicitness and thus ultimately aesthetic and rhetorical crudeness. […] For the core of traditional Chinese rhetorical aesthetics is the ethereal, intellectual light touch, the esthetic and reflective pinch of cultural salt, an acute sense of what is inevitably left unsaid. […]

Classical Chinese words and phrases crop up in texts a little bit like cobras: whenever we spot one, we do well to look for a corresponding mate not too far away. Chinese prose is pervasively patterned through echoes and resonances with the past, on the one hand, and through echoes and resonances within any passage, on the other. Traditional European literature, of course, contains plenty of parallelism, but no variety of European prose shares the cobra-like quality which so pervades nearly all the widely different varieties of mainstream classical Chinese literature from the earliest times. […]

One striking feature of classical Chinese rhetorical practice is the extraordinary absence of basic ephemeral communicative phrases […]. Phatic communion […] must have been common in ancient China, but it is not recorded in classical Chinese. Classical Chinese rhetoric eschews the ephemeral and inconsequential. The rhetoric of classical Chinese excludes such small talk, whereas the rhetoric of Homer encourages it. […]

Greek texts, as part of a literary market economy, read as if they were written for a general critical intellectual, and quite probably that is the kind of public for whom these texts were publicly read out and performed: a potentially hostile, rebellious, and even quite literally impatient audience. The great Roman writers, such as Seneca or Cicero, wrote with an explicit concern for audience reaction. Pre-Buddhist Chinese texts, however, read as if one were eavesdropping on a discourse directed at a ruler and oblivious of the natural reactions of a general reading (much less listening) public. […]

It is not part of Chinese style to comment on one’s style as one is using it. This contrasts sharply with the Western practice of stylistic self-assessment in midcourse.

Thanks, Bathrobe!


  1. David Eddyshaw says

    I wonder (without actually knowing anything about it, and being keen to hear from those who do) whether Confucius’ “allusive subtlety” might be to some extent be an ex post facto reinterpretation of the fact that he’s difficult to understand because what we’ve got is a lot of scattered remarks collected by his disciples delivered in particular contexts which we no longer have any way of retrieving. He himself might not have intended to be subtly allusive at all.

    Mencius strikes me as the opposite of subtly allusive. One feels quite sorry for poor old King Hui getting his head bitten off after attempting to greet the Sage politely and all.

  2. J.W. Brewer says

    Not all English versions seem to have same intra-chapter numbering, but I take it the opening sentence paraphrases the line rendered by Legge as ‘The Master said, “In language it is simply required that it convey the meaning”‘ and by Pound as “He said: Problem of style? Get the meaning across and then STOP.” If what is intended is in fact the sort of subtle and non-explicit allusiveness argued for in the rest of the passage by Harbsmeier, all three of these English renderings seem rather misleading, as Anglophones will take them to be recommending the sort of clarity of expression that Harbsmeier says is neither found nor desired in the Confucian texts.

    FWIW, one online source gives the key Chinese as 辭達而已矣, which Google Translate renders into English as “It’s just a word.” When you add the introductory 子曰 and play around a bit you can get “Zi said: It’s only a good word.”

  3. Pound is wonderful, but he’s not yer man for subtle and non-explicit allusiveness.

  4. J.W. Brewer says

    One can certainly imagine Pound’s wording here as a somewhat loose and zingy paraphrase of a hypothetical Sinitic text that Legge is a more literal translation of, but I am intrigued by the (possibly highly unreliable) implication from google translate that Legge stuck a lot of exegesis/paraphrase into his version because a “literal” translation would be too bafflng for the Western reader.

  5. Yes, now I’m curious about that too. I hope someone can enlighten us.

  6. David Eddyshaw says

    It’s a Gricean maxim: Say enough, then stop.

  7. I wouldn’t expect Google Translate to do anything useful with this — it’s presumably trained on modern written Chinese, which is a rather different language. You could respell the Aeneid into French orthography, but it wouldn’t mean GT would be able to render it sensibly as French > English.

    My Classical Chinese is not great, and perhaps someone can come along with a better gloss, but I’d understand this to very literally be something like:

    ‘Speech gets there and stops.’

    達 (Pinyin dá) is what I’ve rendered as ‘gets there’. It has a whole range of meanings, from ‘lead’ to ‘reach’ to ‘convey’ to ‘express’. I’d imagine most of the work in really interpreting this would lie in deciding just exactly what this should mean in context.

    My impression would be that 已矣 (yī yǐ — they probably already sounded fairly similar in Old Chinese) would have less the sense of ‘cuts off, ceases prematurely’, and more ‘finishes up, is done to the point of completion’, but someone who’s more familiar with this stuff than me should really confirm or deny. So maybe more ‘speech gets there and is done’? Where ‘speech’ can obviously include writing.

  8. David Marjanović says

    I dare say allusive subtlety was, at least originally, not so much intended as it was an inevitable side-effect of an extremely abbreviated writing style – which was used because writing was so time-consuming.

    As previously featured here: “An uninitiated westerner can no more be expected to understand such writing than Confucius himself, if transported to the present, could understand the entries in the ‘personal’ section of the classified ads that say things like: ‘Hndsm. SWGM, 24, 160, sks BGM or WGM for gentle S&M, mod. bndg., some lthr., twosm or threesm ok, have own equip., wheels, 988-8752 lv. mssg. on ans. mach., no weirdos please.’”

    Once you’re fluently writing with this kind of efficiency, it grows a poetry all on its own. (And now that I’ve hit “Post”, I’m wondering if this is actually what attracted Pound.)

  9. In my (admittedly rather limited) experience, the comparison to classifieds isn’t very on point (though it’s funny, which can excuse a lot). You can’t talk about that many things with classified-ese, and a lot of its actual obscurity is just a matter of graphic abbreviations. Surely the better comparison would be telegraphese, if people decided they should write works of philosophy and history in it?

  10. I just happened on a 2004 thread in which Zizka (now John Emerson) says:

    Many of the same things are true of translating Chinese poetry, and one very influential scholar (Edward Schafer) believes that poetry translation from the Chinese is more or less impossible […]

    I think that Schafer greatly overstates the case, but by now the translations by Pound, Waley, and Rexroth that got me into Chinese poetry seem almost as distant from the Chinese as I think Fitxgerald’s Omar Khayyam are. In both cases, something new unquestionably came into English poetry from Chinese and Persian, but Chinese and Persian poetry per se didn’t come across. This is supposedly always true of translation, but the difficulties between even classical Greek and English are much less than from Persian or Chinese.

  11. Most of what I’ve been focusing on re Classical Chinese has been poetry. I’m inclined to be sceptical about the supposed impossibility of translating it, and feel that a lot of it just comes down to the collision of two mystic auras: that of Chinese and the alien, ineffable East; and that of poetry in general.

    Of course, no perfect translation of poetry is ever possible (basically by definition, if we acknowledge that poetry has a phonological component), and when you get right down to it things like _Beowulf_ are extremely hard to translate fully satisfyingly. So I think we need to manage expectations a little. It’s fine to say Chinese poetry also can’t be translate ‘fully satisfyingly’, as long as we don’t make it out to be unreachably more so than other traditions (even just looking to slightly older stages of English and its close relatives). I don’t think anyone would dispute that most Chinese poetry is far easier to translate than most dróttkvætt. Chinese poetry is _reasonably translatable_, and that’s really about all we can ask for most poetry.

    A conversation along these lines came up not too long ago on Language Log:


    (ETA: I suppose cultural distance can be an issue too, but honestly, a lot of Classical Greek is now very alien to many modern westerners, and one can do a bit of contextual reading to appreciate, say, Persian or Chinese verse. This seems largely more about the biases and backgrounds of an aggregated or idealized audience for published poetry, and less about anything strictly linguistic, or even about how a particular reader will respond to a translated poem.)

  12. David Eddyshaw says

    I’m inclined to be sceptical about the supposed impossibility of translating it, and feel that a lot of it just comes down to the collision of two mystic auras: that of Chinese and the alien, ineffable East; and that of poetry in general.

    Yes; that seems very likely to me, too.

    I wonder if I am imagining it, but I thought I detected something of this “alien, ineffable East” vibe in Harbsmeier’s essay, too. (Said with due diffidence, as he is plainly an expert and I most definitely am not.) I think this comes out a bit in the way he over-eggs the pudding with his contrast with Western classical literature; some of his assertions about that strike me as something between cherry-picking and being frankly tin-eared. Talking of cherry-picking, there seems to be rather a lot of Classical Chinese literature which doesn’t display the supposedly typical characteristics he flags up.

    I was hoping Bathrobe would chip in at some point …

  13. J.W. Brewer says

    I did realize the danger of using google translate in that the modern Mandarin meaning of a particular string of characters, even if coherent, might be rather different than the Classical Chinese meaning of the same string of characters. But as I was cut-and-pasting the characters I’d found elsewhere into the google translate interface what did strike me was that it seemed rather likely that virtually any language would need more than five (inflection-free) morphemes to unequivocally convey what Legge said these five characters said.

  14. PlasticPaddy says

    Re contrast in philosophy, is something like this meant?
    Cicero (De amicitia, 64.4-8): “haec ut omittam, quam
    graves, quam difficiles plerisque videntur calamitatum societates! ad quas non est facile inventu qui descendant. Quamquam Ennius recte:
    Amícus certus ín re incerta cérnitur,…”

    Confucius (Analects 9.28): 歲寒、然後知松柏之後彫也。
    “Only after it turns winter are we aware of the survival of the Pine and Cypress.”

    The Ennius sententia is more explicit, as is the Cicero exposition of it. But I think it might be better to compare the Confucius with the pre-Socratics.

  15. Seong of Baekje says

    I think 辭達而已矣 is a straightforward sentence which means “language just needs to convey meaning”. There doesn’t seem to be any confusion about it on the Chinese internet.

  16. “language just needs to convey meaning”

    Presumably most of the confusion on the part of Anglophones is more about how this should be interpreted in a wider sense. Is it a pragmatic statement against rhetorical flourish? A call to concision? A more abstract philosophical observation? Something else?

    But this would seem to fit with a general pattern: it’s not ‘literal’ translatability that’s really the problem (at least not unusually so), so much as broader interpretation. There’s a reason for the popularity in China of commentaries like the Zuo Zhuan, instead of everyone just reading the Chun-Qiu.

  17. January First-of-May says

    So far my impression of the assorted translations seems to be that the most plausible meaning is something to the effect of “say as much as you need, then stop”. David Eddyshaw’s “Gricean maxim” version comes pretty close. Given Nelson Goering’s commentary, in particular, I’m tempted to translate it as “speak until done” [on the pattern of “cook until done”].
    I’m not sure how we’re getting “language just needs to convey meaning” from this, though.

    That said, I don’t know any Classical Chinese (I’m literally just going off the translations in this thread), so I might be missing something obvious.


    Presumably most of the confusion on the part of Anglophones is more about how this should be interpreted in a wider sense. Is it a pragmatic statement against rhetorical flourish? A call to concision? A more abstract philosophical observation? Something else?

    This is also important: the statement as stated (as interpreted by me) seems at first glance to be a borderline tautology. “You should say exactly as much as you should say.” Duh.
    On second glance, it’s not a tautology but in fact the maxim of quantity; but even so, is it the first part, the second part, both parts, or something else similar?

  18. Presumably most of the confusion on the part of Anglophones is more about how this should be interpreted in a wider sense.

    Perhaps we could avoid the confusion is we assume for a second that there is no wider sense, because Confucius said exactly what he wanted to say and nothing more.

  19. PlasticPaddy says

    Do you mean that Confucius is (recorded as) making a naturalist/botanical observation in Analects 9.28?????

  20. There is this saying that everything is easy when you know it. I wonder how much of the difficulties is based on the lack of familiarity with the philosophers from a foreign tradition. If you grow up in the West, you are introduced to the western classic philosophy in school, through children’s books, and so when you start to study them for real you already have a certain familirity and the concepts seem less opaque. With Confucius, Mencius and other eastern classic philosophy you have to start from zero, like a middle school child. From that starting point, everything looks vague and difficult to understand.

    If we understand subtlety to mean the opposite of hammering home your point, I guess “say as much as you need, then stop” is good enough, but that seems like a very odd kind of subtlety.

    What kind of poems are we talking about here? There are a lot of classical Chinese poems that are easy to understand with knowledge of modern Chinese and only a minimal knowledge of Classical Chinese. Many popular poems have very universal themes, such as nature, homesickness, love and friendship. For the experts, there is deeper meaning based on the history, context and so on, but you don’t need to be an expert to enjoy the poems. Are we reading different poems, or are we reading them differently?

    By the way, what’s up with the cobras? That sentence puzzled me the most of the whole article.

  21. Seong of Baekje says

    One source of confusion seems to be people analyzing 已 alone instead of as part of the collocation 而已. This is how I would break down the sentence:
    辭 (words; language)
    達 (express [meaning])
    而已 (that is all; nothing more (rough equivalent in modern Mandarin is 就可以))
    矣 (final particle similar to modern Mandarin 了)

  22. Seong, ah, thank you, I’d missed that about 而已 (though in retrospect I think even I should have caught this — at least, looking back I see it was something I technically learned once, without apparently relearning it enough to stick). Interestingly, taking it compositionally is still pretty close to the mark, which I suppose just means that this is an intuitive line of linguistic development.

    I’m still interested in just which nuance of 達 is meant. I’d learned the word as ‘to reach, to get in contact with, to be successful’, but found (while looking at this line) that it can mean things like ‘express’ (among other things). Is there an easy way to know how lexicalized a meaning like ‘express’ would have been in Confucius’s day?

  23. people analyzing 已 alone instead of as part of the collocation 而已

    This is exactly what Pound was doing with “and then STOP.” But it doesn’t change the meaning too much. Zhu Xi seems to have thought along similar lines: “Language should seek to convey meaning and then stop. Craft does not consist in richness and beauty.” (辭,取達意而止,不以富麗為工)

    From the traditional commentary by Kong [Yingda (574-678)]: “In all matters, nothing is more important than concreteness. If language conveys the meaning, that is enough. Do not trouble with colorful language.” (孔曰:凡事莫過於實,辭達則足矣,不煩文豔之辭)

    (“concreteness” is my awkward translation for shí 實, whose semantic range also covers “replete, solid, fruit, seed, facts.”)

  24. John Cowan says

    Confucius said exactly what he wanted to say and nothing more.

    Doubtless. But that doesn’t make him that latter-day philosopher 大象霍顿, who always meant what he said and said what he meant.

  25. Confucius is faithful, one hundred percent.

  26. “Whereof one cannot speak, thereof one must be silent.”

    Confucius seems to be equally gnomic, neither more nor less.

  27. Bathrobe has said nothing because 1) he’s been preoccupied this last week and 2) he’s not really very knowledgeable about Classical Chinese.

    But there is one thing that hasn’t been brought up here. Modern Mandarin is not nearly as succinct as Classical Chinese.

    For example, spoken Chinese, in the mouth of a Beijinger, at least, can be extremely prolix. Beijingers really know how to natter on.

    And the written language, which has been heavily influenced by Western languages and Japanese, is also more expansive than Classical Chinese. Certain features of Western style, such as the use of attributive clauses, especially in officialese, have made it so. There is also the 把 form (which fronts objects) and its equivalent 将 in written Chinese, as well as the common use of 将 as the equivalent of English ‘will’ (so-called future tense), which make the language less succinct than Classical Chinese. There are also other features (the use of 了, the tendency to multi-syllable words) that lead in the same direction. That is one reason that Chinese find Classical Chinese so “beautiful” — its succinctness contrasts with the pedestrian talkativeness of spoken Chinese or the greater prolixity of modern written Mandarin.

    Of course, the language does retain some of its former features, including the use of fixed classical expressions (成语) that express a lot of meaning in a few syllables, but even this can be less than concise if they are used just for the sake of using them. The language of official documents and party directives is not concise at all; it’s been heavily influenced by foreign languages. There have been books published in Hong Kong criticising the way that officialese on the Mainland has ditched the conciseness of Classical Chinese, including 成语, in favour of stolid translationese. I don’t know if they’re available on the Mainland because I get the impression they are regarded as being negative about the party and the bureaucracy.

  28. (Other things also add to loss of succinctness, such as the greater use of measure words. I’m sure there are others, too, but I can’t think of them.)

  29. David Eddyshaw says


    I was confused by Brett’s response initially, but I have now found links to the Tractate in question. I see that it is a work of moral philosophy rather than epistemology, as I had supposed from JC’s comment; that was what threw me.

  30. Indeed, as the line “It should be! It should be! It should be like that!” makes perfectly plain. Telegony and all.

  31. John Cowan says

    GT renders 大象霍顿 as ‘Elephant Holden’, presumably because Mr. Caulfield was commoner in the sources than Mr. Horton.

    The moral element of the Tractate was pointed out by one of the earliest commentators, but later ones have used it “in discussions on a wide variety of topics, including economics, Christianity, feminism, and adoption.” In regard to the latter two, one middle-period commentator believes it to be a tendentious argument in favor of fetal rights.

  32. @John Cowan: The fetal personhood argument is associated with Horton’s second, much later tractate. (Dr. Seuss did not usually reuse his protagonists, but the Horton, the Cat in the Hat, and Bartholomew Cubbins each get two volumes.)

  33. John Cowan says

    Confucius may, however, be identified with that quintessentially European figure the King of Hearts: “Begin at the beginning, go on until you come to the end, and then stop.”

Speak Your Mind