Joel of Far Outliers has a post called “Traditional China: Multilingual” on the Hawai‘i Reader in Traditional Chinese Culture, a book that tries to “capture the complexity of the Chinese cultural mosaic” and “take into account virtually every aspect of traditional culture, including sources from the non-Sinitic ethnic minorities,” in which he quotes the introduction on the many languages spoken in China:

Take language, for example. When one thinks of what defines “China,” perhaps the first thing that comes to mind is that it is a place where the people all speak “Chinese.” But what is this “Chinese” that everyone is supposedly speaking to each other? Unfortunately, China does not today possess, nor has it ever in the past possessed, such a universally understood tongue. For starters, we have to take into account the tens of millions of speakers of non-Sinitic languages who make up a significant proportion of the population of the Chinese nation as it is currently configured…. These languages belong to such disparate groups and families as Tibeto-Burman, Austroasiatic, Austronesian, Turkic, Tungusic, Iranian, and Slavic. These are the “minorities” of the Peoples Republic of China, all of whom have roots that lie deep in the past of East Asia, West Asia, Central Asia, Southeast Asia, and Northeast Asia.

It goes on to discuss the complex linguistic facts subsumed under the rubric “the Chinese language,” succinctly and well. The post and the linked introduction deserve reading, and the book sounds like an excellent place to start learning about the diverse place that is China.
Incidentally, anyone who wants a more detailed look at the linguistic situation should get hold of S. Robert Ramsey’s The Languages of China (reviewed here).


  1. Wow. You’re quick. I really should have thought to link to Ramsey’s book–and to Danny Yee’s review of it. Kudos to you.

  2. In the movie “Big Fish” there is a scene where, in one of the father’s stories about WWII, he is shown preparing for a trip by reading an “Asian – English Dictionary”. Very funny, although I’m afraid that many viewers may actually think there is a country where people speak a language called “Asian.”

  3. There’s something odd about the concept. I guess I’m old fashioned, but while there are many spoken “Chinese” languages/dialects, there is really one written / literary Chinese language. Sure, during some periods there was more Southern influence on the literary language(e.g., probably the Chin dynasty after 200 A.D.), a few individual writers use lots of local color, during certain periods there were two different standards (Southern Sung / Chin period ca. 1200 A.D.) But Chinese is standardized at least as much as French is.
    To put it more strongly, standardization of literature and formal speech to a single national language is pretty central to Chinese culture, definitional really and one of its core values, and seems already to have been so at the time of Confucius.
    As for the “non-Han” languages — the Tibetans, Mongols and Uighurs are Chinese by conquest and international agreement. Culturally they are heavily influenced by China, but the Germans and Russians are heavily influenced by the French and English. The official story is that they are “non-Han Chinese”, and perhaps sometime that will be true, but it really isn’t now.
    This leaves mostly the hill peoples of the South, and I’d be willing to allow that in some sense they are “proto-Chinese” with a degree of belonging to Chinese culture, not only by assimilation but by archaic shared traits (it’s sort of a cliche to find among the hill tribes some of the archaic cultural forms in the Shih Ching).
    But all in all it seems like a tendentious, PC book, though if the selections are interesting that’s OK — I am not an anti-PC militant. (I am curious whether the book is framed to conform to official doctrine about non-Han Chinese, or whether it makes concessions with the Uighur and Tibetan nationalists).

  4. No offense, but you do seem to be saying “since the Chinese pretend to have only one language, it behooves us to go along.” The fact is that native speakers of Cantonese, of Hakka, of Fujianese, and of Peking Mandarin cannot understand one another without prior acquaintance with each other’s forms of speech, exactly the same situation that prevails with French, Spanish, Italian, and Romanian, and they are therefore different languages. Linguistic definitions can have no truck with official pretenses and national borders.

  5. While I sympathize with the efforts of the Hawai‘i Reader‘s authors to show the complexity and diversity of the Chinese world, I am puzzled by the following sentence in the second paragraph about fangyan 方言 (honestly, I can’t see why it would be wrong to translate the term by “dialects”):
    Despite their lack of a written form, the fangyan are still vibrant.
    Any Chinese-learned foreigner who has been in Guangdong (including Hong Kong) or is aware of so-called Lingnan 岭南 culture can easily refute this assertion. As a matter of fact, a reasonable ammount of the characters used to write Cantonese have no proper use in guoyu, or so rarely that the occurrences are almost neglectable, and some of them are such basic words as kêu (“person” radical + 巨, meaning “he/she”), gem (口 + 敢, “so, like this”), and many others. The fact that I cannot even type them in IME is pretty telling. This is one of the reasons why it is so amusing to see all the “jiang putonghua, xie guifanzi” 讲普通话, 写规范字 (speak Mandarin, write the standard characters) signs in a place like Guangzhou. Everything, from school (and sometimes college) classes taught directly in Cantonese, to the magazines imported from Hong Kong and hand-written price tags in men 蚊, show the gap between the official linguistic ideology and what is locally tolerated. And how about Yueju 粤剧 (popular Cantonese theater) libretti?
    One may think that this is a detail, but such imprecision can be devastating for students who have a hard time facing a reality which contradicts what they have been taught. Of course, nothing is wronger than judging a book from its introduction, and I do not assume that those points are not explained further.
    As an aside, Cantonese speakers often refer to their language as bag wa, that is, baihua 白话 (the excellent Guangzhouhua fangyan cidian 广州话方言词典 lists “denomination of Yue dialects spoken in both Guangxi and Guangdong” as a second, wider definition).

  6. And while I’m talking about the various denominations of “the common language”, it seems misleading to translate Huayu by “Language of the [Culturally] Florescent [People]“. I’ve always thought that Hua is taken here more for being the first part of Huaxia (one of the “traditional” names for China), than for (one of) its first meaning, but maybe I am missing some reference here. I also wonder why Victor H. Mair does not list Zhongwen 中文.

  7. Maybe I should take it back. I like a lot of Mair’s contrarian stuff.

  8. Speaking from the trenches of classical Chinese scholarship here, I should point out that the stability of written Chinese over the period of its putative standardization (i.e. from the language reforms of the First Emperor of Qin in the late third century BC to the May Fourth Movement of 1919) is definitely worth questioning. Some of Mair’s work does come across as a bit contrarian, but then he’s working against 2300 years of rhetoric that insists on reifying written Chinese as an ideal type. I’m sure many readers here are familiar with his work on written vernaculars in medieval China — he maintains that certain genres of medieval literature record spoken language of the time, in contrast with the conventional wisdom, that written and spoken Chinese parted company in the Qin dynasty, and didn’t meet up again till 1919 (and then even maybe not so much).
    Zizka wrote: To put it more strongly, standardization of literature and formal speech to a single national language is pretty central to Chinese culture, definitional really and one of its core values, and seems already to have been so at the time of Confucius. You’re right to point out that the ideology of a single national language has been an important concept in Chinese culture at many points over the years, despite evidence mustered by Mair and company to suggest that the actual situation on the ground wasn’t so simple. But it’s not correct to date this to the time of Confucius.
    Eastern Zhou and Warring States bronze inscriptions (contemporary with Confucius’ life) suggest that the various Warring States each had their own regional writing system, though the different systems were related. It was the First Emperor of Qin who standardized the writing systems of the different states he’d conquered, as part of his campaign of general standardization (weights, measures, width of roads, length of chariot axles) in the service of empire. The First Emperor was such a Legalist that I’m sure the Qin standardization was more about the exigencies of rule than about the ideology of culture; I suspect that the language ideology you’re referring to settles into place, like so many other things, in the Han dynasty. But anyway, it’s long past Confucius’ time. In my personal experience of translating early Chinese texts, the language in which the Confucian Analects (Lun Yu)are written is strongly idiosyncratic, to the extent that you get the feeling that it represents a kind of regional Lu dialect or written style. It’s a big headache to translate, actually.
    Certainly, written Chinese is strongly flavored with the conventions and vocabulary of its period. I had a strong reading ability in Han-period classical Chinese when I started grad school (i.e., I could read the Shi ji without a dictionary), but I’ve had to learn an entirely different set of reading strategies to cope with sixth-century epitaph inscriptions and Buddhist sutras. I am slowly getting a handle on Tang legal documents and Buddhist historical writings, but I stay away from poetry and other oral forms. I can still be completely flummoxed by Song inscriptions, and don’t get me started on the Yuan (Mongolian and Tibetan loanwords everywhere, sheesh). The amateur folklorist Pu Songling’s seventeenth-century collection of ghost stories, Liaozhai zhiyi, is written in a Han classical style worthy of Sima Qian himself, but the nearly contemporary romance and martial-arts novels of the time are another matter entirely.
    The point is that written Chinese over its history is an amazingly flexible system of genres and forms, but that this multivalence is widely ignored by the language ideology Zizka points out. This is the kind of thing that contrarians like Mair are trying to bring to our attention.

  9. P.S. (like you really need to hear more from me) — I agree with Jimmy that “Huayu” is best translated as just “Chinese,” for exactly the reasons he suggests. I haven’t read the book, but if Mair left out the term “Zhongwen” it is probably because technically it refers only to the written language. Similarly “Yingwen” (written English) as opposed to “Yingyu” (spoken English) and so on. It is kind of odd that for Chinese we have “Zhongwen” but not “Zhongyu,” similarly “Hanyu” but not “Hanwen.” But at any rate “wen” means written languages.

  10. As for Confucius, he said something about how one of the values of the Shih Ching was that it taught you the proper names of birds and beasts, etc. Elsewhere there’s a saying interpreted to mean that he spoke one dialect in formal occasions, and a different one on ordinary occasions. (As I remember, this interpretation is not universally accepted.) Both these sayings (granted the interpretation of the second) seem to indicate that the normative belief in a single standard correct Chinese was already there, even if not successfully attain (especially by the shrike-tngued barbarians of the south.) Of course, Brooks claims that most of the Analects comes from considerably after, or long after, Confucius’ time.
    I’ve had the same experience with developing reasonable proficiency in one area (philosophy, and poetry to a degree) and then being at a loss going to some not-very-distant genre (novels or histories). Buddhist writing shouldn’t count, though, in this context, since it’s a highly specialized technical language.
    I still would tend to argue that, the vast majority of Chinese literature is written in a single standard language developing over time. How much is there that really is written in a different language through and through, rather than just using exotic vocabulary and local color? (I’m actually asking, not rhetorically). For example, Mark Twain uses several different dialects in his dialogue, but his exposition is standard. (And all of them could be called languages rather than dialects, except that since they are not being used formally anywhere they aren’t — whereas Portuguese IS a language, even though it’s very close to Spanish).
    I guess I’m defending the political and “official literature” definition of languages. After all, people who espouse minority languages usually end up wanting to produce a written literature, get some degree of political autonomy, and produce their own prescriptive grammarians. So they really have about the same opinion as I’m arguing, in that sense, except that they think that their language has been wronged by not having those things.

  11. Xiaolongnu,
    The distinction between yu (“spoken language”, hence the yan 言 “speech” radical) and wen (which we can agree refers to any graphical sign, as in tianwen 天文) is very clear in theory, but things are much more blurred in nowadays practice: I rarely hear mandarinophone people say (about some foreigner) “ta hui jiang hanyu 他會講漢語, while “ta hui jiang zhongwen” is by far the most common sentence (I’m not giving this as a scientific evidence; this is only my personal experience, almost exclusively limited to Dalu).
    Once again, I am all for refuting stereotypes about China and refusing to transmit them under the pretext that it makes teaching simplier. However, since Mair’s statement particular statement about the various official denominations of China’s lingua franca refers explicitely to the contemporary use, I still think that there is a regrettable lack of clarity here.
    By the way, Hanwen, read kanbun is still used in Japanese for “classical Chinese”, as is *Zhongguo yu/Chuugoku go, meaning the common language.

  12. Warning: Non Expert Opinion
    language hat, you compared different Chinese dialects to French, Italian, Spanish, but my understanding was that they written langauge was the same, and only pronunciation differed (forgive me if I am wrong, I had this discussion with a Chinese friend right after she came to the US, and we never had the opportunity to discuss the subject since, so we may have misunderstood one another at the time).
    If it is only the pronunciation and not the writing that differ, doesn’t that make the difference more like someone from New Jersey trying to talk to someone from Scotland (assuming both have thick regional accents) than an Italian speaking to a Spaniard?

  13. Michelle: That’s a common misunderstanding, particularly among Chinese. If I were emperor of the world, I’d mandate that everyone everywhere must take Linguistics 101 so that we could at least start from a common understanding of the basics. At any rate, one of the first things you learn in Ling 101 is that writing is pretty much irrelevant to language in its basic sense. Of course once written literature develops, it acquires its own rules and influences the spoken language, and all of this is interesting and important, but the fact is that if everyone forgot how to write tomorrow there would be exactly the same number of languages spoken in the world. (Zizka, are you listening?) When we talk about different languages within “Chinese,” we’re not talking about different “pronunciation,” we’re talking about the same kinds of differences we find between Dutch and High German, or between Spanish and Portuguese: lots of cognates (evident when you look at them, not necessarily when you hear them), mostly similar grammar with some clear differences. To give you an example, ‘he’s in Shanghai’ would be ta zai shanghai in Mandarin, hi l@’ zonghe in Wu (Shanghainese). (I’m ignoring the tones, obviously; in the Wu I’m using @ for the central vowel [like u in butter] and ‘ for the glottal stop.) Now, shanghai is clearly historically the same as zonghe, and you could “get used to” that kind of difference and learn to recognize it, but ta and hi are a different matter: they’re historically completely distinct words for ‘he/she.’ And l@’ is not just historically distinct from zai, it’s used in incompatible ways; there’s nothing in Mandarin comparable to the use of reduplicated l@’ as a continuative: hi l@’-l@’ chhy@’ ve ‘he’s eating.’ They are different languages, and this would be perfectly obvious if it were not hidden beneath the veil of the “same written language” — which is basically written Mandarin, although as Jimmy Ho says above, it can be jiggered with extra characters to write the other “dialects.” But if you used characters to write English and German, so that He is not here and Er ist nicht hier looked exactly the same, that wouldn’t make them the same language, would it?
    like you really need to hear more from me
    xiaolongnu, the more I hear from you, the better. I always learn from your comments.

  14. language hat,
    It may also have been my misunderstanding of what she said. (i.e. I’m willing to take full blame for the misunderstanding)
    But I’m still confused–if we all lost the ability to *speak* tomorrow (really bad world-wide cold for instance), would we then have *fewer* languages?
    And how does the Internet affect this? If I am writing *more* now that I was previously, and able to communicate *more* easily with people with whom I share a common language but not necessarily dialect, does this make a difference?

  15. See, that’s one reason I didn’t get my PhD — I’m no good at teaching, and if you’re not going to teach there’s no point having a PhD. The point I was trying to make is that spoken language is basic (for the purposes of this discussion I’ll include sign language under “spoken,” since it’s used comparably); it’s what you learn when you’re a kid, it’s the basic tool of interpersonal communication. Written language is parasitic on spoken; however much it diverges, there is spoken language at its base (even if it’s pretty much dead, like Latin). But we’re so used to literacy and the importance of written communication we tend to think the opposite, that writing is primary. Certainly the Chinese, who have been literate for 3000 years and more, think so, which is why I’m sure you’re not misrepresenting your friend — I heard similar things from many Chinese when I was in Taiwan, and gave up arguing. But no matter how many Chinese think they all speak one language divided into dialects (just like English!), they’re wrong. They need to take Linguistics 101, dammit! Anyway, I’m probably not doing any better a job of explaining what I mean, but feel free to keep asking questions and we’ll see if I can stumble into an approach that works for you.

  16. Well, there is always getting a PhD just because. That’s what my mom did. Getting a PhD is pretty worthless for a 5th grade teacher, but she did it anyway. Of course it’s also a lot of work for no reason.
    So you’re saying that written language is completely dependant upon spoken language?
    I suppose that makes sense if I consider that for much of history writing and reading were limited to only certain classes, but now that writing and reading are attempting to become universal, do you think it will remain that way? That as e-mail and chat and weblogs become common means of communication, will written language remain dependant upon spoken language, or will it take on a greater significance as the written word becomes more common?
    Perhaps it is a ridiculous assumption to assume that written language will become so important–after all, computers may soon generate the ability to do away with the written word all together (although I hope not) at which point this becomes moot point, but I’m just very curious as to whether increasing education and literacy are changing the affect that written language has upon spoken.
    I’m sorry if these are basic linguistics questions, but that is one of the few areas I missed studying in all the time I have been in school. (Of course now my interest has been piqued…)

  17. Not completely dependent — once writing develops, it strikes out on its own in some ways — but dependent for its original existence and continuing health. If a written language gets too far away from its spoken base, it risks artificiality (like the so-called “purified” Greek of the last century). But it will be interesting to see what becomes of English as it gets used more and more for written communication by people whose command of the spoken language is shaky.

  18. I decided to be stubborn. Suppose we have terms language1 and language2. Language1 means spoken languages. Language2 means official languages, formal languages, written languages, languages with dictionaries and grammars.
    In Europe we have and unknown x number of Romance language1′s. The language2′s consist of Spanish. Portuguese, Catalan, French, Italian, Romansch (barely) and (maybe) Provencal. Germanic languages a similiarly undefined number of language1′s, and as language2′s English, Dutch, Low German, High German, four Scandinavian languages, and (maybe) Frisian.
    The differences between the various sorts of “German1″ or are certainly as great (probably much greater) as those between Portuguese1-Spanish1-Catalan1 or Danish1-Swedish1-Norwegian1. But there are only the two “German2″ languages. That’s just a political-cultural-historical accident, but it does have a consequence in terms of how these languages work for people. And as far as I know, throughout the German world either High or Low German2 has an authority over the local German1 languages.
    So I’m really in favor of keeping local languages alive, including nonstandard local German1 languages but especially rare, unusual, historically unique languages like Basque or Yukagir. At the same time, movements to keep languages alive usually DO have a tendency to want to transformthem into language2′s, somewhat recognizing that a purely spoken language isn’t quite there.
    And finally, the authority of some standard language (now Mandarin), for better or worse, has really had an enormous influence on Chinese culture and tends to be accepted (with exceptions; some Taiwanese refused to learn Mandarin in 1948) even by people who are monolingual in a non-standard form. So people who say that Chinese is one language with many dialects are wrong in one sense, but they’re correctly recognizing the intense domination of the high-culture dialect. Perhaps this is changing; I know that Cantonese has had advocates during the last century, but my understanding is that outside Hongkong it doesn’t challenge Mandarin.
    Incidentally, I’ve always admired the Scandinavian stubbornness. (In my home area there are three country Lutheran churches within sight of one another). Fewer than 20 million people, highly modernized, and with five standard languages (two for Norwegian — I didn’t count Finnish, of course). By Chinese standards these are odd local dialects of English, I suppose — education is heavy in English in all those countries I think.

Speak Your Mind