Spoken Hong Kong Cantonese and written Cantonese

« previous post | next post »

[This is a guest post by Robert S. Bauer, with some comments on "dialect" vs. "language" by me (VHM) at the bottom.]

1. After 1949 over the last few decades of British colonial rule, Cantonese was regarded as one more desirable/useful barrier separating HK from China.

As a consequence, the British treated Cantonese with benign neglect which allowed it to develop naturally and without interference, and this is why it has been doing as well as it has.

A couple of years ago the fact that only a handful of people showed up at a demonstration in support of Cantonese in HK shows that most HK speakers do not see it as being under imminent threat.

In Guangzhou people are told that "civilized people speak Mandarin" wénmíng rén shuō Pǔtōnghuà 文明人説普通話, which to me implies that uncivilized people speak "dialects" (topolects) such as Cantonese.

2. Some people hold the view there is only one kind of "proper" Chinese and that is standard written Chinese (i.e., Modern Standard Mandarin). The "dialects" (topolects) are nonstandard or substandard and hence unimportant and uninteresting — except to linguists and dialectologists.

Here in HK I've encountered people who have told me that Cantonese cannot be written and should not be written.

I once told students in a sociolinguistics class I was teaching some years ago that I thought written Cantonese should be standardized and promoted as another Chinese variety. One of the students who was from the mainland and a teacher of Chinese language in HK could barely suppress her outrage, telling me if that were done then written Cantonese would challenge and compete with standard written Chinese and such a situation could not be allowed or tolerated.

But in reality written Cantonese and standard written Chinese have already been in competition in HK for some time, although not on a large scale, and Cantonese could in no way be said to challenge the pre-eminent position of standard written Chinese. The standardization of written Cantonese has been evolving for some time on its own informal, ad hoc basis, but so far this phenomenon hasn't received much attention.

A couple of observations:

The other day as I was riding down the escalator in a train station I saw the phrase 啱晒 ngaam1 saai3 'thoroughly correct, absolutely right', a thoroughly Cantonese expression, written in an advertisement. I assume northerners who don't speak Cantonese would find it completely meaningless, and for the time being that doesn't seem to matter to the advertiser who rented that space, but such an attitude will likely change before long.

The other night at a linguistics conference dinner the man sitting on my left said he was from Chongqing and couldn't understand anything Cantonese speakers in HK were saying. Earlier in the day I had come across the word 萬字夾 maan6 zi6 gaap3/2 which was said to be the Cantonese word for 'paper clip' and equivalent to standard Chinese huíxíngzhēn 回形針 or huíwénzhēn 回纹针, so I showed him the Cantonese word and asked him if it meant anything to him. He said 'no'. I think it may be an old Cantonese word, since most of my Cantonese-Putonghua dictionaries published in HK haven't recorded it, although three other references have listed and defined it.

VHM:  Dialect or Language?

Cantonese is one of the main themes of our Chinese deliberations on Language Log, e.g., "Cantonese novels" and "Mandarin is weirder than Cantonese".  On the question of whether Cantonese is a language or a "dialect" (of what?), which keeps coming up in these discussions, we really do need to get beyond the trite notion that "A language is a dialect with an army and a navy."  For starters, see "Counting the Languages of the World" by Geoff Pullum, and the comments thereto about Chinese "dialects".

Quebec doesn't have an army or a navy and a national flag, but everybody would agree that French and English are different languages.  And think of all the different languages that are spoken in India, yet none of them have an army or a navy and a national flag.  Conversely, consider all the nations that DO have an army or a navy and a national flag, but speak the same language, e.g., the United States, England, Canada, Australia….  Consequently, having an army or a navy and a national flag has nothing to do with the determination of whether a linguistic entity is a language or a dialect.

Perhaps it would be useful to have a debate on the meaning of the word "dialect" and what that signifies for its relationship to "language", thinking especially of the implication of the prefix dia- ("inter; through; across; between").

< … διαλέγομαι (dialégomai, “I participate in a dialogue”) (Wiktionary)

< …dialegesthai "converse with each other," from dia- "across, between" (see dia-) + legein "speak" (Online Etymology Dictionary)

< …Latin dialectus, Greek διάλεκτος discourse, conversation, way of speaking, language of a country or district, < διαλέγεσθαι to discourse, converse, < δια- through, across + λέγειν to speak (OED)



The Free Dictionary (American Heritage)



Online Etymology Dictionary

Oxford English Dictionary


  1. Jon Lennox said,

    August 29, 2013 @ 5:31 pm

    I have always thought it would be more accurate to say "a language is a dialect with an education ministry."

    Having an army and a navy is one way to keep your education ministry intact, but not, by any means, the only one.

  2. Nathaniel said,

    August 29, 2013 @ 6:35 pm

    I think only the Indian example contradicts Max Weinreich's quip, which I think is about having state power at some point in the language's development. Quebec is a particularly poor example, since there is France and metropolitan French and its prestige. That's a huge thing for Quebec French-speakers to draw on to defend their language as something beyond the home.

  3. Nathaniel said,

    August 29, 2013 @ 6:37 pm

    And I don't know enough of Indian history to know which of the official languages had any period to be used as "ruling languages".

  4. J.W. Brewer said,

    August 29, 2013 @ 6:47 pm

    I would not advise Prof. Mair to tell a Quebecois nationalist that Quebec does not have a national flag. These are the people that call their provincial legislature the Assemblée nationale du Québec." But there are numerous Francophone nation-states that unquestionably have militaries of their own, so amending the claim to "a dialect with one or more armies/navies" would cover French nicely. Of course the Jacobin and post-Jacobin governments in Paris have often been just as bad to Provencal and other such "topolects" as the Communist (and post-Communist?) governments in Peking have been to Cantonese, so . . .

    I take it that Chinese nationalists do not by and large have any problem with the existence of multiple "languages" (Tibetan, Mongol, etc etc) spoken within the boundaries of the PRC, as long as they are spoken by members of ethnic minorities who are understood to fall outside the Han Chinese ethnicity. I expect the problem is that many Chinese nationalists are having trouble conceptualizing how the Han Chinese people, as opposed to the Han-plus-minorities population of the PRC, can manage to constitute a meaningful and metaphysically coherent group without a common language (or at least an aspirational/hypothetical common language, that the actual languages spoken by actual subgroups of Han can be categorized as deviations from). I frankly don't know how one would successfully convince such Chinese nationalists that an openly multilingual Han people could and would still form some sort of conceptual unity for ethnic/cultural/etc. purposes. (Not that I myself think such a unity is problematic, since what already exists in fact must be possible in principle and can *usually* be openly admitted to exist in fact without thereby destroying it; I just don't understand the Chinese-nationalist mindset well enough to know how to persuade it to take that view.) I do suspect that the difficulties encountered in talking about the variousness of the languages spoken by "Arabs" that are all commonly viewed as "dialects" of a single thing called "Arabic" are somewhat similar and are similarly muddled up with the self-conceptions of Arab nationalism and Arab ethnic identity. But I don't know how to talk the Arab nationalists out of their parallel quandary either.

  5. Victor Mair said,

    August 29, 2013 @ 6:50 pm

    @J.W. Brewer

    You've completely ignored what I said about India, the world's largest democracy.

    The flag of Quebec is provincial, not national:


    All the American states have flags too. They were recently featured on a series of commemorative stamps.

  6. Peter Donald said,

    August 29, 2013 @ 7:36 pm

    From the very Wikipedia article you linked: "Article 2 of An Act respecting the flag and emblems of Québec confers the status of "national emblem" to the flag of Québec." See also http://www.gouv.qc.ca/portail/quebec/pgs/commun/portrait/drapeau/?lang=en for more from the Quebec government on Quebec's national symbols and http://www.canlii.org/en/qc/laws/stat/rsq-c-d-12.1/latest/ for the full text of the Act respecting the flag and emblems of Quebec.

    For politicians, politics trump semantics.

  7. J.W. Brewer said,

    August 29, 2013 @ 7:44 pm

    Quebec nationalists think what you call their provincial flag is "national." http://fr.wikipedia.org/wiki/Drapeau_du_Qu%C3%A9bec. I assume that if the way Indian nationalists think about language was easily adaptable to the Chinese context, some clever person would have already done so and improved the situation in the PRC. It presumably helped that many/most of the "official" regional languages in India had long had their own standard written form (often with a distinctive script) and literary tradition, not to mention the phenomenon of conquerors using their own languages (first Persian and then English) rather than getting linguistically assimilated by the people they'd conquered the way the Mongols and Manchus did. But even so, India's post-independence path toward stable language policy was not entirely smooth, with e.g. numerous people killed in riots in 1965 over ultimately withdrawn proposals to promote the status of Hindi in ways objectionable to many Tamil speakers. http://en.wikipedia.org/wiki/Anti-Hindi_agitations_of_Tamil_Nadu.

    One possible parallel to the modern PRC situation is some of the history of language policy in the Russian Empire (the Soviet era ended up being even more complicated, I think). I don't know the details very well, but at various points in the czarist era, speakers of closely-related Slavic languages like Ukrainian etc. (and even Polish, which is further afield) were subjected to forced Russification because those other Slavic tongues were in effect viewed as inferior dialects/topolects of what should be the prestige national language, whereas often at the same time the not-at-all-democratic government might be reasonably tolerant of e.g. Finnish or Georgian or Buryat, because those were obviously different languages spoken by minority ethnic groups, much like the non-Han languages in the PRC. Of course, the PRC government might note that ultimate acceptance of "Ukrainian" as a fully legitimate language distinct from Russian eventually led to the existence of "Ukraine" as a real nation-state (flag, army, navy, everything) distinct from Russia, and think "see, that's exactly why we don't want to admit that Cantonese is really a language."

  8. Alexander said,

    August 29, 2013 @ 9:33 pm

    Some clarification might be helpful. What is the suggested debate about? I can't imagine it's about a stipulated technical definition of "dialect" for linguists. So is it about how ordinary people use the word? If so, then Victor's comment about India seems somewhat misplaced. People routinely refer to many languages of India as dialects, as in "this lullaby […] is in the Malayalam dialect, the main language of Kerala." Or even "Become Fluent in the Hindi Dialect Using the English Language." The question "dialect of what?" would then also be misplaced, because ordinary usage – exhibited by expressions like "the Germans never could figure out the Navaho dialect" – certainly does not require the speaker to have any linguistic genealogy in mind. So it seems to me that there may be less mystery than Victor thinks in the lay use of the term.

  9. Jake Nelson said,

    August 29, 2013 @ 9:56 pm

    I've generally been of the position that "dialect of what" is meaningless- a dialect is the basic unit.

    Dialect: An intersection of idiolects allowing communication between multiple people.
    Language: 1. A family of related dialects (such as "French language" meaning the loose set of dialects collectively considered "French"). 2. A standard or regulated form of a dialect used for official purposes by some body. (such as "French language" meaning the official standard published by the Academie Francaise.)

    Note that senses 1 and 2 are separate and distinct; the word 'language' is polysemous, but the definitions overlap enough that most people don't feel the need to distinguish which sense they're using.

  10. Victor Mair said,

    August 29, 2013 @ 11:13 pm

    On the strong feelings that the differences between Cantonese and Mandarin can evoke, see "Top U.S. Diplomat Gets Language Lessons in Hong Kong".


    (courtesy of Daniel Devaney)

  11. Stephan Stiller said,

    August 29, 2013 @ 11:55 pm

    1. Applying the word "dialect" to a particular variety of language makes no sense in a vacuum but only in relation to other varieties of language. Normally "dialect" and "language" are contrasted; this implied contrast is essential to the meaning of the word "dialect".

    2. Victor Mair's coinage "topolect" is excellent for original/historical usage of the Chinese word 方言 and in a sense it solves the problem for speakers of English, but using a more appropriate English word doesn't solve the underlying problem of many people presently using the Chinese word 方言 in an inconsistent way: One simply cannot assert in the same sentence that Romanian and Portuguese are different languages (ie: not just dialects of each other) while maintaining that Mandarin, Shanghainese, Cantonese, Taiwanese, Hakka, etc are "only dialects" (ie: with respect to each other). The implication is one of a specific comparison of distance that is at odds with reality, an assertion that these Sinitic languages are not sufficiently different from each other to be called "languages".

    3. Many linguists will assert that spoken language comes before written language, and this view is correct in many respects. But the existence of written-only idioms (Literary Sinitic, aka Classical Chinese), sign languages, and of complexities present only in written language (and more generally subtle differences in direct expressivity of the different modalities: spoken language has prosody and written language can play with letters, punctuation, formatting, layout, etc) demonstrates that this story needs to be refined – but let's not get into the details here and now. Most people use two modalities of communication that should properly be called "language" – spoken and written – and traditionally these have differed in China. Is the cognitive difference that many Chinese people are just not used to having their written variety be a reflection or derivative of their spoken variety? When they say that a Sinitic vernacular is "only" a dialect, do they mean that it is only a spoken variety in a region that all writes the same thing (MSM)? To some extent, but there is more to it. As illustrated above, a problem with the word 方言 (and with any attempt at translation) is the way it is used these days: Even if one were to translate 方言 only as "spoken variety in a region of a more or less uniform written/literary language" into English, users of the word 方言 seem to think the wrong things (eg: when comparing linguistic variation in Europe to that in China). Much of the confusion arises out of ignorance of just how different the spoken Sinitic vernaculars are from each other (this includes spoken Mandarin) and from written Mandarin. Someone who doesn't know will use the words 方言 ("dialect"? "topolect"?) and 語言 ("language") in strange ways, in a (spoken or written) Sinitic language or in any foreign language.

    4. For someone fully aware of the situation, the core of the issue is of course whether it is better to write how one speaks (我手寫我口); there are very important considerations of literacy and economy, but de facto the most decisive factor is political sentiment. And, as is well-known, the mere fact that the written domain in HK is normally occupied or dominated by and therefore associated with MSM unjustifiedly relegates Cantonese to lower status in the minds of many HKers.

  12. Victor Mair said,

    August 29, 2013 @ 11:56 pm


    "Malayalam language" 466,000 ghits

    "Malayalam dialect" 2,310 ghits, of which many, if not most, refer to dialects OF Malayalam, not the supposed Malayalam dialect

    "Hindi language" 2,310,000 ghits

    "Hindi dialect" 41,800, of which many, if not most, refer to dialects OF Hindi, not the supposed Hindi dialect

    "Bengali language" 795,000 ghits

    "Bengali dialect" 9,810 ghits, of which many, if not most, refer to dialects OF Bengali, not the supposed Bengali dialect


    It is abundantly clear from these data that the lay understanding of Malayalam, Hindi, and Bengali, etc. is that they are languages, not dialects. Likewise, the government of India refers to these different varieties as languages, not dialects.

    "Cantonese language" 565,000 ghits

    "Cantonese dialect" 60,000 ghits

    "Taiwanese language" 136,000 ghits

    "Taiwanese dialect" 63,800 ghits

    "Shanghainese language" 12,000 ghits

    "Shanghainese dialect" 8,760 ghits


    In the Chinese case, the data are not so stark as in the Indian case, but it is still obvious that the common understanding is that these are separate languages, not dialects. Unlike India, however, the government of China — for political, not linguistic, reasons — insists that these mutually unintelligible varieties of Chinese / Sinitic are dialects, not languages. This political insistence does have a limited impact in skewing the terminology slightly toward the "dialect" side, but the common opinion is still that they are languages, not dialects.

    Jake Nelson's first definition, the one for "dialect", makes sense: "An intersection of idiolects allowing communication between multiple people." That evokes the notion of mutual intelligibility between dialects, which is the lay understanding of the term, and which all of the references I cited in my original post point to. His second, two part, definition for "language", however, is more difficult to get a handle on, yet nonetheless contradicts his claim that "'dialect of what' is meaningless". When we say "dialects of French", we recognize that the dialects are subsumed under the French language.

    For those who want to learn more about the dialects OF French, here are a couple of resources to consult:



  13. Yuan said,

    August 30, 2013 @ 2:09 am

    I was a little surprised by the anecdote. I'm a non-HK Cantonese speaker, and "萬字夾" is what I and my family call paperclips.

  14. Vanya said,

    August 30, 2013 @ 3:59 am

    @Victor Mair "The flag of Quebec is provincial, not national"

    That statement may be technically true but would be considered insulting by Quebec nationalists, who consider Quebec a separate nation trapped within Canada. Just an example of why these issues are almost impossible to resolve.

    As far as American state flags go – I have met more than one Texan who considers the flag of Texas to be his (always a "his") "national" flag.

  15. Simon P said,

    August 30, 2013 @ 5:29 am

    I've always found most speakers of Chinese languages to be spectacularly misinformed about their own language. The sentiment that "Cantonese cannot be written" is spectacularly and obviously wrong. When confronted with written Cantonese, these people often claim that the characters used are "made up" and not "real" Chinese characters. Exactly what makes a character not made up is always clear: if it was made up by a person writing Literary Sinitic or MSM, it's real, otherwise it's fake.

    Other baffling ideas include, of course, "Cantonese SHOULDN'T be written down", but also things like "Cantonese doesn't have any grammar" or "Cantonese doesn't have any tones". Another common belief amongst speakers of all Chinese languages is that anyone literate in "Standard Chinese" can read and understand Literary Sinitic, which is demonstrably untrue.

  16. Stephan Stiller said,

    August 30, 2013 @ 6:06 am

    @ Simon P:

    "Cantonese doesn't have any tones" – whew! (I personally never heard that one.)

    That Chinese has no grammar is an oft-heard statement. Every time I hear it from a student in an East Asian studies (!) department, I am struck with terror. Sometimes you hear it only about the spoken language (in that form I heard it at least twice from students doing research in the field of Chinese linguistics), sometimes about "Chinese" in general. I recall someone writing somewhere that such "no grammar" statements are to be understood in the sense that there is no inflectional morphology with conjugations and declensions (after all, it's true that grammar of Sinitic languages is simpler in that way). But in my experience pretty much everyone who tells me that Chinese has no grammar will then proceed to add that you can just "put together any words and do anything you want", which contradicts the "no inflectional morphology" interpretation …

  17. peter said,

    August 30, 2013 @ 6:38 am

    I think this debate suffers from a false conflation of the concept of "nation" with the concept of "state". State and nation are not necessarily the same thing. The state of the United Kingdom of Great Britain and Northern Ireland (UK), for instance, comprises the nations of England, Wales, and Scotland, together with part of the island of Ireland. The island of Ireland may be considered itself a nation, or not, depending on one's political views. All of these four nations have flags, which are national flags, not provincial or regional flags. Their different flags reflect their very different histories and their somewhat different cultures. As a foreigner in the UK, one is struck by how often the news media refer to "the nation of Wales" or the "the nation of Scotland", etc.

    Because these 4 geographic territories (3 nations and part of another, or 4 nations) came to be part of the multi-national state of UK in different ways, there are various differences in the laws and administrative systems across the whole territory of the UK. These differences sometimes reflect the relative power balances of the constituent nations at their times of accession to the state. For example, in the state of UK, there are two parallel legal systems, one covering England, Wales, and Northern Ireland, and the other covering Scotland. Although the UK state has a single currency, there are two sets of legal paper tender, one set of currency notes being issued by the Bank of England, and the other set issued by banks in Scotland. All bank notes, of both sets, are legal tender throughout the entire territory of the state of UK.

  18. peter said,

    August 30, 2013 @ 6:51 am

    Vanya (August 30, 2013 @ 3:59 am)

    Since the flag of Texas was adopted when Texas was an independent republic and before that republic's accession to the USA, its supporters are fully justified in calling it a "national" flag.

  19. Simon P said,

    August 30, 2013 @ 7:00 am

    The "Cantonese has no tones" comment came from a Cantonese speaker in Guangdong province. Since she had learned the four tones of Mandarin in school, she knew that Mandarin has tones, but nobody had educated her in or about Cantonese (being an uncivilized dialect and all), so she never thought of Cantonese as having tones.

  20. Victor Mair said,

    August 30, 2013 @ 7:48 am


    The flag of Texas is a state flag, not a "national" (your quotation marks); Texas is a state, not a nation (see the discussion below).



    Your comment is amazingly equivocative and almost comical in its hedginess.

    Pronunciation note: http://www.howjsay.com/index.php?word=equivocative

    You admit that my statement is "technically true" but assert that it "would be considered insulting by Quebec nationalists". First of all, not all Québécois are nationalists, and secondly, though a certain portion of Québécois may wish that their province would become a nation, the fact remains that it is not. It is a province of the monarchy of Canada. Even if it should one day become an independent nation (the likelihood of that happening in the foreseeable future is very small), there would still be many other places in the world outside of France that speak one or another dialect of the French language that would not be considered nations. In common parlance, "aspiration for nationhood" ≠ "nationhood"; "language" ≠ "nation"; "dialect" ≠ "language".

    Language note: "Quebec" comes from a Mi'kmaq word k'webeq meaning "where the waters get narrow".

    "Just an example of why these issues are almost impossible to resolve."

    These political questions should not interfere with our linguistic analysis.

    "I have met more than one Texan who considers the flag of Texas to be his (always a 'his') 'national' flag."

    How many mad, male Texans have you met who consider the flag of Texas to be a "national" flag? 2? 3? 4? There are over 26,000,000 Texans, and most of them are content with Texas being what it is, a state within the federal republic of the United States of America. Furthermore, even supposing that, through some freakish combination of events, Texas would one day become an independent nation, would that make the various English dialects spoken there into languages?

  21. Rodger C said,

    August 30, 2013 @ 8:19 am

    The National Beer of Texas:


  22. Victor Mair said,

    August 30, 2013 @ 8:52 am

    @Rodger C


    The note at the bottom says: "Bring your passport–and some aspirin."

  23. Benjamin said,

    August 30, 2013 @ 9:08 am

    Mr. Mair is interpreting the saying "A language is a dialect with an army and a navy" in a way which strikes me as bizarrely literal. My understanding of the phrase has always been that it means a language is a dialect backed by a people powerful enough to enforce its status as a language, and that dialects backed by people without sufficient power will be dismissed by speakers of dominant groups who speak similar languages. It's simply a way of describing the way that social and political power determine what is called a dialect, and what is called a language.

    Certainly there are ways that linguists can determine these things on a different basis, but the quote is speaking about the power dynamics of languages and identity, not about how one actually determines what is a language and what is a dialect of a language.

    Also, coming from a Montrealer, please stop talking about Quebec politics. The fact that Wikipedia identifies the Quebec provincial flag as a provincial flag has little to nothing to do with the identification of that flag as a national flag by Quebec nationalists, or the use of other flags such as http://en.wikipedia.org/wiki/Patriote_flag. How people perceive themselves is very relevant to this discussion – the fact that you identify Canada as a monarchy, when very few Canadians perceive ourselves as such, and the monarchy has very little political or cultural influence these days, shows that you probably should refrain from comment on Canada-specific socio-political matters.

  24. Jake Nelson said,

    August 30, 2013 @ 9:19 am

    @Victor Mair: My comment was written in haste on my way out the door to work, so was perhaps not the best phrased; what I meant was:

    "dialect" can have a clear definition suitable for technical use with no mention of "language", whereas "language" is too indistinct between the 'genetic' grouping meaning and the 'official' standardized form meaning. Romanian and Moldovan are both languages because the Moldovan government says so (the 'official' language thing), but the two languages are extremely similar (if not identical) dialects- were it not for the political aspect, they'd be considered the same. Many linguists refer to them as such, ignoring the political issue, which tends to offend a number of Moldovans.

    I keep coming back to a zoological classification analogy with dialect as species, but trying to make language = genus is fraught with political problems, aside from the issue of where to draw the line. (There's plenty of western Norwegian dialects that are not truly mutually intelligible with the standard Oslo dialect, but I've never heard anyone call them separate languages, aside from those who consider Nynorsk and Bokmal separate languages, which is another kettle of fish.) As such, I prefer letting the "language" descriptor be solely the political designation, and using a different term ("dialect family" is my current favorite) for the grouping.

  25. Benjamin said,

    August 30, 2013 @ 9:21 am

    @Jake Nelson

    That better articulates what I was thinking before I got all huffy.

  26. Alexander said,

    August 30, 2013 @ 9:27 am

    Victor, I understand that the politics and so am not surprised by the usage frequencies. I also know what I hear nonlinguists and students saying to me on a daily basis. And I know the technical definitions that linguists tend to use for professional purposes – and teach to students despite their reluctance to abandon their normal usage. This is why I asked about the goal of the discussion. One goal might be to suggest a reform of ordinary usage. And in that case the question could be clarified by suggesting some desiderata for a usage that is politically sensitive, linguistically reasonable, and not utterly unnatural in relation to other ordinary usage.

    About Jake Nelson's points: to me it seems that part of his point is that, in fact, "language" is that it used sometimes as name for a class, and sometimes as a name for a distinguished member of a class. Or of course sometimes as both, somewhat loosely. On the first usage, a language is, strictly speaking, not something that anybody speaks. That's a bit odd, but it's a fact that the word is sometimes used that way. An example of this is in the following quote from a 1922 dictionary of Hawaiian: "The Hawaiian is but a dialect of the great Polynesian language, which is spoken with extraordinary uniformity over all the numerous islands of the Pacific Ocean." This is the kind of comment that seems paradoxical upon scrutiny – but that's ordinary language for you.

  27. Ran Ari-Gur said,

    August 30, 2013 @ 9:57 am

    As far as I can tell, your objection to "A language is a dialect with an army and a navy" is that it's a poor prescription for how to use the terms "language" and "dialect". When others point out that it's not meant to be prescriptive — quite the reverse, it's meant to point out that the lay use of the terms "language" and "dialect" is politically charged and mostly meaningless — you object that it's not a perfect description, either, and then you go back to pretending that it's a prescription.

    But I don't see you attacking other adages for the same reason. Do you object to "A stitch in time saves nine" on the grounds that sometimes it may save a different number? (Do you pretend that it's prescriptive? Do you read it as asserting that the correct use of a single stitch is to save exactly nine stitches later?)

  28. J.W. Brewer said,

    August 30, 2013 @ 10:43 am

    I am personally convinced that describing Cantonese as a "language" distinct from Mandarin is on balance the most sensible approach and am happy to employ that usage myself. (I think the language/dialect distinction is quite blurry but am comfortable that the differences between Cantonese and Mandarin are great enough that it's not a borderline case.) But I'm not the one who needs convincing here. So I would echo the question of who is the audience that Prof. Mair ultimately wishes to convince to either change their terminology or (perhaps more importantly) improve their perhaps muddled underlying conceptual understanding, and what if any practical change is he assuming would flow from that?

    Go back to the parallel of modern France. Unlike Provencal, presumably Basque was never conceptualized/deprecated by the Parisian government as a dialect/patois or other inferior variant form of proper French, but understanding it to be a separate language altogether rather than merely substandard French did not necessarily mean that Basque-speakers living in France did any better in their legal/social/cultural position than Provencal-speakers, so getting the government to accept conceptually that Provencal is just as much of a separate-from-French language as Basque is would not necessarily by itself change much of consequence in the real world.

    Note that not only the current regime in Peking but the U.S. Government (at least for certain purposes) treats "Chinese" as a single language. E.g. http://www.census.gov/prod/2013pubs/acs-22.pdf gives the number of "Chinese" speakers living in the U.S. with no attempt to break that group down among Mandarin/Cantonese/Taiwanese/etc. (To be fair, this is all going to be based on self-reporting, so if Chinese-Americans are mostly going to answer "Chinese" rather than "Cantonese" or "Mandarin" to an open-ended question about what languages other than English they speak, it may be hard to get better data,* and I have no idea how various Chinese-American community leaders would react to a different classification scheme.) On the other hand, I certainly hope our military and intelligence agencies are quite aware of the difference between the different language varieties and have separate personnel appropriately trained to understand intercepted communications in each of them.

    *By contrast the census numbers do now distinguish between speakers of "French" and "French Creole." I suspect a lot of those who select the latter are e.g. Haitian immigrants, but I also suspect that there many be many cases where Haitian immigrants whose actual language use is the same as each other do not check the same box (i.e. one identifies as speaking "French" and the other as "Creole"), so increasing the seeming precision of the classification scheme may decrease its level of accuracy.

  29. Jake Nelson said,

    August 30, 2013 @ 10:44 am

    Also, on the subject of Chinese, I'll just post again as I have various times here in the past, my touchstone: A few friends of mine from middle and high school grew up in Cantonese-speaking homes. They referred to "Chinese" and "Cantonese" as two languages, the former written-only with no spoken form, the latter spoken-only with no written form, imperfectly but adequately joined.
    They found "written Mandarin" (attempting to transcribe Mandarin speech using characters irrespective of meaning) extremely offensive, as an attempt to replace "Chinese" which they saw as a pan-Chinese shared language, with Mandarin, at the expense of all the other spoken languages which shared Chinese writing. (Their parents saw this as an extension of political centralization.)
    They found the idea of "written Cantonese" strange for similar reasons- if one wanted to transcribe sounds, better to use an alphabet or syllabary, and if one wanted to write meanings, there were already characters for that.
    Again, I am certainly no expert on the subject, just the formative background for my viewpoint on the matter, my friends' opinions, which I find interesting and relevant.
    I find it somewhat similar to debates over use of "eye-dialect" to transcribe the accent, etc., of Scottish people speaking English, where some find it offensive, because to them, doing so said that their way of saying the same written words was incorrect, and the way those transcribed normally spoke it was correct.

  30. peter said,

    August 30, 2013 @ 11:44 am

    Victor —

    I think again you conflate state and nation.

    The flag that is the current flag of the US state of Texas was – before it had that status – the flag of an independent nation, the Republic of Texas. That is why I said that its supporters were justified in calling it a national flag, since it was (and remains) the flag of a nation.

    The National Gallery of Victoria, in Melbourne, Australia, retains this name, despite the fact that Victoria went from being a self-governing nation to being a constituent part of another nation, Australia, in 1901.

    Ditto, the English National Opera, even though it was founded after England was already part of Great Britain.

  31. J.W. Brewer said,

    August 30, 2013 @ 11:54 am

    Surely deciding whether or not Canada is a "monarchy" by deferring to the subjective emotions of Quebecois nationalists is like deciding whether or not Cantonese is a "language" by deferring to the subjective emotions of Chinese nationalists. One can approach both questions more scientifically, in the first instance by e.g. posing the question."Who's that lady whose picture is on the coins?" The trick is figuring out how to reconcile the two propositions that (1) human emotions and subjective self-identification are often highly relevant data for many sorts of social science questions (and linguistics is inter alia a social science); while (2) if you let fear of giving offense to the subjects of your research (or politically-powerful third parties with their own reasons for interest in the topic) have dispositive control over what you say, you're just not doing science, not even social science.

  32. IDRIS said,

    August 30, 2013 @ 12:58 pm

    (This is a comment I gave at the Lingua Franca blog. It might be relevant to the discussion here.)

    I think one reason why the language/dialect distinction is a tricky business is because the degree of foreignness of a given dialect is not the same for all non-native speakers. Let me illustrate this by giving two examples:

    1- Many works on Arabic dialects describe them as rather separate languages that are not (necessarily) mutually intelligible. That might sound true on paper full of phonetic transcriptions of those dialects, but in reality the picture is remarkably different. The fact that these varieties are different does not necessarily mean that they are unintelligible, even if they are so to the non-native speaker. This is attested by, for example, the huge number of Arabs who come from Yemen, the Levant, Egypt and Sudan to work in Gulf countries and who can communicate with the locals and among themselves even though their dialects are different. Similarly, Turkish dramas dubbed in Syrian (not Egyptian, mind you) Arabic have been popular all over the Arab World in the last few years. The speeches of Colonel Gaddafi, largely in Libyan Arabic, were understood by all Arabs who were amazed by the colonel’s bizarre acts.
    On a more personal level, I myself as a speaker of Sudanese Arabic find only Algerian and Moroccan dialects (and probably Mauritanian Arabic as well) partially unintelligible, and only if their speakers talk among themselves, not to me. That’s basically because of the pronunciation (they use less vowels than speakers of most other dialects), and of course because of the use of French words. The rest of the dialects are usually easy to understand. And even when I speak with an Algerian or Moroccan, we just tone down on dialect-specific features that we think are unlikely to be found in the other dialect, with most of the toning down done by the Algerian or Moroccan, and not by me (intelligibility between Maghrebi and Mashriqi dialects is rather asymmetric). We do not use ‘Fusha’ to facilitate comprehension, maybe because the words we use are Arabic anyway, but with less vowel endings than is the case in Standard Arabic.

    2- The differences between German and Swiss German dialects are indeed significant, and it is true that “A German speaker from Bonn or Berlin will not understand Swiss German dialects like the speech of the Zurich area (sometimes known as “Züritüütsch”) without doing a lot of work.” Yet one question we should ask here is this; how much work should a German do in order to understand Swiss German? The answer is; probably not much. With enough exposure a German can understand Schweizerdeutsch quite well, even if he/she continues to use Hochdeutsch to talk to people. Were we dealing with two different languages here, the German would need to attend a Swiss German course to learn the ‘language.’ (http://personalpages.manchester.ac.uk/postgrad/jonathan.morris/Papers/CAMLing%208th%20December%202010.pdf, http://personalpages.manchester.ac.uk/postgrad/jonathan.morris/default/MA.pdf).

    So, the bottom line is; we need to take the degree of foreignness into account when dealing with the language/dialect distinction. This might raise problems with Scandinavian or Iberian languages, for example, but it might be better than succumbing to the lure of superdiversity.

  33. Victor Mair said,

    August 30, 2013 @ 1:16 pm


    Oh, no, I am being very careful to distinguish state / province / region from nation.

    @J. W. Brewer

    Thank you very much for recognizing that Cantonese is a language, not a dialect, and also for your eminently sensible remarks about the scientific method.

  34. J.W. Brewer said,

    August 30, 2013 @ 1:29 pm

    I think there may be some confusion because "nation" has multiple meanings in English and I believe Prof. Mair is using what http://en.wiktionary.org/wiki/nation#English has as sense 2 to the exclusion of sense 1.

  35. J.W. Brewer said,

    August 30, 2013 @ 1:46 pm

    There is also the risk of confusion between sense 1 and sense 2 of state

  36. Victor Mair said,

    August 30, 2013 @ 1:54 pm

    @Ran Aru-Gur

    In my original post, I said that "A language is a dialect with an army and a navy" is trite. You seem to agree with that assessment by putting it on the same level as "A stitch in time saves nine".

  37. MaryKaye said,

    August 30, 2013 @ 2:15 pm

    It seems to me that part of the confusion arises from wanting intelligibility relationships among (languages, dialects, idiolects) to be well-behaved and symmetrical. Biologists want relationships among species to be like this, too, which has been a great impediment in particular to herpetology. Amphibian species are just crazy, they do not follow the rules, and if *you* try to follow the rules you don't end up understanding what you are seeing.

    I think we'd like there to be some kind of lawful relationship between evolutionary distance between languages (how long ago did they diverge?) and intelligibility distance (how well can speakers understand each other?) And it would also be great if intelligibility were symmetrical–if I can understand you, you can understand me. But I don't think either of these is true. The assymetry was mentioned above for Arabic, and I have been told by several speakers of Portuguese that it is much easier for a Portuguese speaker to understand Spanish than vice versa.

    As a tangential question: how grammatically similar are the Chinese languages to each other? Does the written form impose grammatical constraints on the spoken forms?

  38. Stephan Stiller said,

    August 30, 2013 @ 2:47 pm

    There is a previous LL post (also by Victor Mair) whose ensuing thread overlaps in content. It is titled Mutual Intelligibility of Sinitic Languages; important arguments are here, Mark Liberman's edit to a comment here, and Victor Mair's comment here. I would encourage people to go over these to prevent a repeat discussion.

    The problem with the statement that "a language is a dialect with an army and a navy" is that many people use it to dismiss any debate of a particular language-dialect question (with the implication that it's only a political issue or at least so political as to not be worth thinking about). This happens frequently when you have a discussion about a particular topolect, commenting that it is a language or a dialect (or regarded as such by some community). Many an interlocutor will interrupt with "Ugh, a language is a dialect with an army and a navy!" and then terminate the conversation, brushing off any serious exploration. It is in my opinion this lay usage of the dictum associated with Max Weinreich that is problematic.

    If someone just says such a thing to make the point that it's so politically fraught a topic that he'd rather not get involved, that's fine. But the dictum shouldn't be used to imply that language-dialect distinctions can't or shouldn't be made for particular sets of languages.

  39. B.Ma said,

    August 30, 2013 @ 3:22 pm

    I just wanted to point out that although 啱晒 can mean 'thoroughly correct, absolutely right' as in the post, in the context of advertising it means [this product is] "perfect" [for the intended use / intended recipient].

    You would also say 啱晒 when you give someone a pop quiz and they answer every question correctly.

  40. hector said,

    August 30, 2013 @ 3:26 pm

    As another Canadian, I second what Benjamin said. Victor Mair, in his attempt to keep politics out of linguistic discussions, is injecting ill-informed, ivory-tower political opinions into a situation he clearly doesn't understand. A lot of non-separatist francophone Quebeckers consider Quebec to be a nation within the federated state of Canada.

    As to China, I was under the impression that it was well understood by historians that the reason Chinese governments insist on one national script is that the central and very real political problem in China has always been keeping the country together. There's a reason nations often insist on national languages. As a linguist, you may not like this, but, you know, politics do exist and are inescapable.

  41. B.Ma said,

    August 30, 2013 @ 3:40 pm

    @peter, August 30, 2013 @ 6:38 am

    All your assertions about the UK are incorrect.

    Here in the UK, I have never heard the media refer to the "nation of England", "nation of Scotland" or "nation of Wales".

    Northern Ireland is a separate legal entity from England and Wales. For example, abortion is not as permissive in NI as in E&W.

    Scottish banknotes are not legal tender anywhere. No banknotes are legal tender in Scotland. BoE banknotes are legal tender only in E&W. There are also 4 banks in NI which produce their own banknotes. UK coins, produced by the Royal Mint, are legal tender throughout the UK, though only £1, £2 and £5 coins are legal tender in unlimited amounts. Obviously, you will never see £5 coins in circulation unless you work in a post office.

  42. peter said,

    August 30, 2013 @ 3:56 pm

    B. Ma —

    Sitting in Britain as I write this, I respectfully and profoundly disagree with you about the use of the word "nation" in the UK media. I stand by my earlier statements.

    You misunderstood what I wrote about British law. I did not say that England, Wales and Northern Ireland have the same laws. I said that they have the same legal system, and that this system is different to the parallel system in Scotland.

    Scottish banknotes are legal tender throughout the UK. I made no statement about coins.

  43. peter said,

    August 30, 2013 @ 5:04 pm

    Another example of a state comprised of distinct nations was the Union of South Africa (1910-1961), two of whose four nations had Dutch/Afrikaans as their national language, while two had English.

  44. J.W. Brewer said,

    August 30, 2013 @ 5:11 pm

    B. Ma. may need to read the sports pages in the UK newspapers a little more closely. If Scotland were not a nation (in some commonly-used sense of the word), it presumably would not have a national soccer team. But it does http://en.wikipedia.org/wiki/Scotland_national_football_team. Indeed, so does England (which has won the World Cup – the U.K. as such can't compete for the World Cup because it lacks a national team, although it is obviously coherent to talk about the U.K. as a "nation" in various other contexts, including other sorts of international sporting competition), and those two national teams have been competing against each other since the 1870's, well before the possibility of devolving some legislative autonomy to the Scots, much less undoing the union of 1707 was a high-profile political issue. Similarly, the Six Nations Championship http://en.wikipedia.org/wiki/Six_Nations_Championship, is a big deal in the world of rugby, with England, Wales, Scotland, and Ireland constituting four of the "nations" competing for the championship, with Northern Ireland and the Republic of Ireland treated for this purpose as a single nation, fielding a single national team.

    One could also of course evaluate B. Ma's claim by picking the website of a major UK media outlet (I first tried http://www.telegraph.co.uk, as the site of the highest-circulation "quality" broadsheet) and using google to conduct a site-specific search for the string "nation of scotland." (Admittedly, use of the phrase seems rarer at the more sensationalistic http://www.dailymail.co.uk, where there's only one hit. Perhaps more research is needed.)

  45. Jake Nelson said,

    August 30, 2013 @ 5:39 pm

    @Victor Mair: "Nation" does not mean "Country", and "State" does not mean "Province". This is a common misunderstanding particularly among Americans, and globally in the post-WWI era.

    "Nation" means "tribe, race, ethic group".
    "State" means "sovereign legal entity, government, institution that controls territory".

    The fact that nearly all post-WWI states ("countries") are nation-states has led to a great deal of conflation of these words. But a nation has nothing at all to do with statehood by itself. There is a Kurdish nation ("tribe", "ethnic group"), but no Kurdish state ("country")- the homeland of the Kurdish people (Kurdistan) is divided among several states (Iraq, Iran, Syria, Turkey).

    The point of the Gettysburg Address's "Four score and seven years ago, our fathers brought forth on this continent a new nation" was not to say that on July 4, 1776 a new state/government/land was created. There was no American federal state, just an ad-hoc alliance/committee of independent states until the Articles of Confederation were approved in 1777. What Lincoln was saying was that on July 4, 1776, Americans became Americans, a new national identity, and were British no longer.

    So for the example of Quebec that has (unfortunately) been the standard one in this thread, Quebecois nationalists believe that Quebec is a separate nation from Canada, and that Quebecois national identity is separate from Canadian identity. Aside from a few with obscure legal theories, they do not argue with the fact that Quebec (the territory and the province) is part of the state of Canada, though many feel it should not continue to be. Most nationalists around the world favor an independent nation-state for their nation, though many settle for some degree of autonomy.

    Other examples of nations without states include the Basque nation, the Native American nations (see the complex treaty status, where some (but not all) are considered "sovereign nations" despite not having independent states), etc.

  46. Jake Nelson said,

    August 30, 2013 @ 5:59 pm

    The status of whether a sub-group is considered a "nation" or not is, come to think of it, extremely closely related to the "language" vs. "dialect" question.
    In the eyes of many, if the various language groups of China are considered separate nations, their distinct speech is considered a "national language", on par with all the others. If they are subgroups of a single Han Chinese nation, then Mandarin would be the "national language" and the others merely "Chinese dialects".
    This is essentially the worldview the "army and a navy" thing is trying to express, nations nearly all having a nation-state with a military these days, and the ones that don't being largely powerless and reduced to "ethnic minority" status.
    Chinese authorities would, of course, be less than fond of assertions of distinct nationality, seeing as these tend to be followed by movements for national autonomy.
    As another example, Yugoslavia was a state shared by several nations (how many is still under debate), at least some of whom had separate national languages (how many is also still under debate, Serbo-Croatian being complicated). The attempts by the Yugoslav government to try to forge a Yugoslav nation in its haphazardly-glommed-together territory, with a single Yugoslav national language, is a textbook example of failed ethnogenesis, and the typical failure mode thereof (the nation that has the most control of the power structure gives up trying to make a new nation, and instead focuses on forcing minority nationals to convert or die.)

  47. Jon said,

    August 30, 2013 @ 6:11 pm

    B Ma –

    Peter is perfectly correct that the notes issued by the private Scottish banks are not legal tender anywhere, including Scotland. They never have been, they have the same status as the local town currencies that pop up now and then. If people accept them, they work as money.

    It used to be the case 50 years ago that many shopkeepers in the holiday resorts in the north of England would only give nineteen shillings to the Scottish pound.

  48. Victor Mair said,

    August 30, 2013 @ 6:15 pm


    "…the reason Chinese governments insist on one national script…."

    Not sure what you mean by "one national script" here. Chinese characters? But — with some tinkering and adjusting — they can be used to write the other Sinitic languages / topolects as well as Mandarin.

  49. J.W. Brewer said,

    August 30, 2013 @ 6:31 pm

    @Jake Nelson: Your position is as inaccurate as that you are purporting to criticize. Or at least you are asserting that sense 2 of "nation" and sense 2 of "state" in the wiktionary links I posted above are not accurate statements about the English language, and I think you would have difficulty backing that up if you reviewed enough instances of actual usage. Or even other standard reference works. Both words have multiple meanings (with one meaning of "nation" being synonymous with one meaning of "state" but the others not so much), and the most constructive thing to do is try to make intelligent guesses using standard Gricean principles about what meaning is likely intended by a particular speaker/writer in context.

    To give them their do, over the decades the Yugoslavs did ultimately succeed in quelling Bulgarian irredentism by convincing the group of arguably-Bulgarians who spent most of the 20th century under their rule that they were a separate ethnic group speaking a separate Slavic language distinct from, rather than a dialect of, Bulgarian (i.e., "Macedonian"). My sense is that back during the Cold War after Tito had broken off from the Soviets this was quite a contentious and politicized issue among Slavicists, but now the Titoist position has mostly prevailed, possibly because you can sometimes successfully change people's self-identification if you have a couple generations worth of time to work at it. Plus now Macedonian must be a language because the Macedonians have an army – not sure about a navy since they're landlocked although there is a pretty big lake on their border with Albania they ought to have some border-security concerns about.

  50. Stephan Stiller said,

    August 30, 2013 @ 9:26 pm

    @ J.W. Brewer

    You are right when you point to different and often local usages of words that need to be contextually disambiguated.

    Most people learn the words "language" and "dialect" from contexts where they are paired up with names of linguistic varieties (topolects). The pairings often derive from considerations of identity politics and can be entirely accidental. I'd prefer if people abandoned their casual usage, but – being largely a descriptivist – I can't entirely ignore the existence of such folk usage either. Linguists need to educate people on the actual linguistic situation for any particular group of languages, so that the next time they try to make a meaningful, scientific statement about the distance of linguistic varieties (using the words "language" and "dialect"), the lay dude won't try to rationalize folk usage as linguistic truth.

  51. Dave Cragin said,

    August 30, 2013 @ 10:50 pm

    One the most interesting cultural aspects related to this is the passion many mainland Chinese have for the idea that Cantonese, Shanghainese, Fujianese etc. are dialects, not languages. Suggestion that these are dialects can cause offense to the point where the subject can't be discussed with some individuals.

    My sense is this is an off-shoot of the government’s very effective campaign to promote Mandarin to create unity within the country.

    Not mentioned above, but of similar vein is the importance of a standard accent (kouyin hen biaozhun口音很标准).

    For example, many friends from Fujian have said say they speak Mandarin poorly because their accent isn’t “standard.” However, these individuals are fully fluent and often graduated with top honors from the top universities in China (while learning in and speaking Mandarin). When being escorted across a manufacturing site in Hangzhou, workers have said to me “your accent is very standard.” To be clear, they weren’t saying I speak Chinese well, they were just saying the accent I’m using is standard which they consider “better” than their native Hangzhou accent – a perspective I can't imagine in the US.

  52. D. Sky Onosson said,

    August 30, 2013 @ 11:05 pm

    The discussion around the national status of Québec has so far omitted the significant fact that the Québécois were formally recognized as a constituent nation of Canada by the federal government in 2006.

  53. Simon P said,

    August 31, 2013 @ 1:20 am

    @Dave Cragin, I experienced the same thing in Guangdong. People would even say "Your Mandarin is better than ours!", which it of course wasn't, but my accent was closer to the official standard.

    Above someone asked about the differences in grammar. The answer is that the grammar of Cantonese and Mandarin are very similar. There are a few differences, in word order and verb usage, but the great differences are in pronunciation and vocabulary. The pronunciation difference, even between words written the same way, can be pretty startling. For example, the word 危險, "dangerous" is prounounced "weixian" in Mando and "ngaihim" in Canto.

  54. B.Ma said,

    August 31, 2013 @ 7:09 am

    It looks like people are not reading comments in this thread very carefully at all.


    I said that Scottish banknotes are not legal tender anywhere (August 30, 2013 @ 3:40 pm). I am correct, as are you.

    peter has been claiming that Scottish banknotes are legal tender throughout the UK. He has done this twice (August 30, 2013 @ 6:38 am; August 30, 2013 @ 3:56 pm). He is wrong.

    This is a basic error which keeps being made over and over again and it is very tiring to have to explain this every few days.

    Legal tender means something that must be accepted when a debt is owed. Nobody is obliged to accept Scottish banknotes, anywhere in the world. In Scotland, nobody is obliged to accept anything at all. However, they can accept anything they want to. Anybody in the world, including Scotland and England, can choose to accept Scottish banknotes if they want.

    Wikipedia explains this perfectly well, if you are still confused.


    With regards to the legal system in NI: well, it is a matter of debate as to whether it is the *same* as that in E&W. You might as well say that it is the same as the legal system of Australia.

    FWIW, let's refer to Wikipedia again: http://en.wikipedia.org/wiki/Law_of_the_United_Kingdom

    "There are three distinct legal jurisdictions in the United Kingdom: England and Wales, Northern Ireland and Scotland.[6] Each has its own legal system, distinct history and origins."


    Where did I say anything about football? I am perfectly aware that E, W, S and NI have their own national football teams.

    peter's assertion was that "As a foreigner in the UK, one is struck by how often the news media refer to "the nation of Wales" or the "the nation of Scotland"

    My response was that I have never seen these terms in the UK media, and certainly not "often". I admit that it may come up from time to time, particularly (as you say) in reference to the national football teams, but I do not read sports news.

    A google search shows that the only reference to "the nation of Scotland" is an article about Donald Trump attempting to sue it.

    peter's other posts show that he has not understood Prof. Mair's comments either.

  55. Victor Mair said,

    August 31, 2013 @ 8:14 am

    In this long discussion over the matter of nation and state, language and dialect, it may be useful for comparative purposes to say a few words about the situation with regard to Taiwan.

    The official name of Taiwan in English is The Republic of China, and in Chinese it is Zhōnghuá Mínguó 中華民國. The ROC was a sovereign state and a member of the United Nations until 1971 when the People's Republic of China replaced it in that body.

    Taiwan has many flags, the official, national flag being that of the ROC, but there are also various proposals for a new flag should Taiwan ever become a sovereign nation again.



    The mother tongue of 70% of the population of Taiwan is Hoklo, also known as Taiwanese, which is a form of Hokkien (a variant of Minnan which is very similar to Amoy [pronounced as Xiamen in Mandarin]). The official language of the ROC is Mandarin, which was imposed on the people when the island was occupied by the forces of Chiang Kai-shek after he retreated from the Mainland in 1949. Individuals who grew up in Taiwan before 1945 also would have known Japanese, since Taiwan was a colony of the Japanese empire from 1895 until September of that year, and Japanese was the language of instruction in the educational system.

    For a study of the language situation on Taiwan, see VHM, "How to Forget Your Mother Tongue and Remember Your National Language"


  56. Victor Mair said,

    August 31, 2013 @ 8:15 am

    For how far apart the pronunciation of characters can be in different Sinitic topolects / languages, allow me to relate something striking that happened in my class on Language, Script, and Society in China this past Thursday.

    The students were startled when a girl from Singapore sitting in the back of the room (who said that she was of Hokkien descent) told us that her surname, Teh, in Mandarin would be Zhèng 郑. While they were still absorbing that shocker, a girl who sat in the front row raised her hand and informed us that in Japanese it was similarly pronounced Tei. I informed the class that the Japanese pronunciation was similar to the Hokkien Teh rather than Mandarin Zheng since the Japanese would have borrowed it long before Mandarin had even come into existence, whereas Hokkien retains much older phonological features.

    This is confirmed by the Japanese form of the name of the famous late Ming loyalist and military leader, Zhèng Chénggōng 鄭成功, namely, Teiseiko. In Hokkien POJ Romanization, 鄭成功 would be Tēⁿ Sêng-kong. Zhèng Chénggōng is the Modern Standard Mandarin (MSM) pronunciation of the name, which he himself would not have used.

    Incidentally, Zhèng Chénggōng is known to Westerners as Koxinga, which is usually explained as the customary Western spelling of his name, but I believe that is incorrect. Rather, it would seem to be the customary Western spelling of his title Guoxingye 國姓爺 ("Lord of the Imperial Surname").


    References for the pronunciation of 鄭 / 郑 in various languages and topolects:




  57. Victor Mair said,

    August 31, 2013 @ 8:21 am

    I regret that the comments on this post have largely been hijacked by questions of politics. It was my original intention in adding the addendum on dialect and language to Bob Bauer's guest post to elicit discussion of the relative linguistic merits of referring to Cantonese as a dialect or as a language. It is still my hope that respondents will return to this question and to the issues about the nature of Cantonese itself raised by Bob.

    I might add that I am very grateful to B.Ma for clarifying a number of issues that have become needlessly obfuscated.

  58. julie lee said,

    August 31, 2013 @ 11:17 am

    @Victor Mair
    About Koxinga, the great pirate, admiral-conqueror (who drove out the Dutch rulers from Taiwan), Ming Dynasty loyalist, and national hero of Taiwan. Koxinga is the romanization of the characters 國姓爺 (guo xing ye in Mandarin), which literally means "National-Surname Lord" or "Lord who bears the national surname". We are told that the fugitive Ming emperor (driven south by the conquering Manchus) granted him the surname of the Ming imperial house in recognition of his victories against the Manchus.

    My question is: Is " Koxinga/Coxinga" the pronunciation of his name 國姓爺 in Taiwanese, Fujianese, or Shanghainese?

  59. julie lee said,

    August 31, 2013 @ 11:44 am


    As Victor Mair notes, Koxinga 國姓爺 (guo xing ye in Mandarin, "Lord of the Imperial Surname") was a title of the historical figure named 鄭成功 (Zheng Chenggong in Mandarin, Teiseiko in Japanese).

  60. Janet Williams said,

    August 31, 2013 @ 2:26 pm

    1) I'm fascinated by Hong Kong gossip magazines using vibrant Cantonese script. See a sample here.

    2) This song by 梁文福 – 「麻雀衔竹枝 」 (YouTube) was banned for 23 years in Singapore since 1990. The ban was lifted on 1 August this year. Reason: the Mandarin song contained a few lines of Cantonese and Hokkien/Teochew. Two Cantonese lines are: 小小麻雀衔竹枝,

    Read my post: An age with relative freedom about this song and the human suffering of banning fangyan in Singapore.

  61. Bob said,

    August 31, 2013 @ 3:24 pm

    Let's get to a lighter side matter re Mandarin/Contonese; the other day I ventured onto a website, a video record of a Shandong TV musical contest, with the participation of the famous singer Ms葉蒨文 from Hongkong. I was surprised that 2 or 3 contestants sang her popular Contonese titles: 祝福, 選擇. (other contestants chose her Putonghua numbers)…. perhaps you know also that the Hongkong songs 上海灘, 萬水千山總是情, etc., are very well known in mainland. However, even though all these songs are sung in Contonese sound, their lyrics are written in MSM…….. People in mainland can understand the lyrics' meaning, thus accept them.

  62. Bob said,

    August 31, 2013 @ 3:28 pm

    ps these songs are included in mainland's kero'OK machines' programs, and people delighted in learning to sing them in Cantonese.

  63. J.W. Brewer said,

    August 31, 2013 @ 3:45 pm

    There seems to be some consensus that a major practical obstacle to the proper classification of Cantonese is the widespread perception by some in the PRC (perhaps both governmental figures and nationalism-minded "civilians" outside the official power structure) that admitting that Cantonese is a "language" or doing various other things (promoting the use of Cantonese for literary purposes, encouraging a standardized orthography etc.) would be a danger to "national unity" or something like that. So I think one important practical question is whether that concern can be responded to. What we might call response A ("the dignity of the Cantonese language, maximum recognition of the rights of those who wish to speak and write in it, and its proper scientific classification are all more important than the unity of a nation-state with a brutal and illegitimate government; national unity can go to hell") may be appealing to outsiders but is probably at present likely to be counterproductive for convincing the people whose minds need to be changed to in fact change their minds. Response B, which seems like it could be more productive, would be something like "I appreciate your concerns, but they are misplaced, as you can see from the following list of examples of other real-world situations where recognizing the status of regional languages and promoting their use did not endanger the unity of the relevant 'nation' (possibly for multiple senses of 'nation')." What's the best case that can be made for Response B, especially given that it is quite easy to find counterexamples, i.e. historical situations in which demands for recognition/improvement of the status of regional languages were bound up with secessionist/autonomist movements and giving in to such demands only emboldened those movements? (Indeed, that Cantonese prospered in HK under the "benign neglect" of British rule and there is no reason to believe that the British were particularly interested in encouraging HK residents to feel a sense of unity with the PRC regime underscores the difficulty here.)

    I don't think India or other seemingly reasonably stable/functional multilingual polities are good counterexamples, because those tend to be places that were always understood to be multilingual and/or multicultural. (No one ever caused controversy in Swiss-nationalist circles by asserting that German/French/Italian were separate things rather than merely substandard regional dialects of the One True Swiss Language.) You ideally want examples where an empirically false myth of linguistic unity had previously been used as part of the propaganda supporting a sense of political/cultural unity, but where it turned out that subsequently admitting the falsity of the linguistic myth did not have collateral consequences for the rest. What are the best historical examples of this sort that might be invoked?

  64. J.W. Brewer said,

    August 31, 2013 @ 3:58 pm

    I'm sure VHM knows much more about the current state of play in language politics on Taiwan than I do (which I believe are still in flux a quarter-century after the end of decades of unconstructive language policy by the KMT regime), but if I am correct in my vague impression that attitudes toward the respective roles/functions of Taiwanese v. Mandarin are commonly mixed up with the "Blue" v. "Green" split in political attitudes toward the mainland that is likewise unfortunate for the issue at hand as I have tried to frame it, i.e. how to convince doubters that topolect-promotion and national disintegration do not have to go hand in hand.

  65. J.W. Brewer said,

    August 31, 2013 @ 8:25 pm

    For the final missing political piece, it would be interesting to hear from anyone who understands the motivations behind the Singapore government's promotion of Mandarin over the topolects historically spoken by the overwhelming majority of the ethnic-Chinese there — while also simultaneously pushing English competence. On the one hand, it is arguably harder to be trilingual (English/Mandarin/Hokkien) than bilingual, but why did they get so obsessed with the Mandarin bug in the first place rather than settling for English-topolect bilingualism? http://en.wikipedia.org/wiki/Speak_Mandarin_Campaign claims (I don't know how reliably. . .) that the effect has been to drive down home use of topolects among the ethnic-Chinese community from >80% to <20% in only three decades, with both English and Mandarin dramatically increasing over the same period.

  66. Victor Mair said,

    August 31, 2013 @ 11:07 pm

    Worth considering:

    "A language with too many armies and navies?"


    The article itself is excellent, and the comments overall are pertinent to our present discussion regarding "dialect" and "language" in Sinitic.

  67. J. Marshall Unger said,

    September 1, 2013 @ 12:46 am

    May I suggest that "dialect" and "language" be defined operationally with respect to the comparative method, i.e. diachronically rather than synchronically? If most linguists who know speech varieties A and B well accept that they are diachronically related without an explicit demonstration (regular sound laws and explanations for semantic divergences, etc.), then A and B ought to be called dialects of their common language. Otherwise, A and B should be called languages, which are assumed to be unrelated until proven "to have sprung from a common source."

    This definition seems pretty satisfactory to me. No honest linguist seriously doubts that the non-English variety spoken in Quebec is French. No honest linguist thinks that describing the relationship of Cantonese to Mandarin is a trivial task. There are still "gray area" cases (e.g. Okinawan, which I tend to think of as a highly aberrant dialect of Japanese, but others consider a distinct languages), but at least we know what we're arguing about in terms of data.

    Another advantage of this approach is that it doesn't get in the way of talking about pidgins and creoles as varieties of a third (or fourth) kind.

  68. Victor Mair said,

    September 1, 2013 @ 8:35 am

    @J. Marshall Unger

    Using the comparative method that you propose, is Cantonese a dialect of Mandarin or a separate language? Is Mandarin a dialect or a language? Or a large group of dialects? If Cantonese and Mandarin are both dialects, what are they dialects of? If Cantonese is a dialect and Mandarin is a language, on precisely what grounds do you make that determination?

  69. JS said,

    September 1, 2013 @ 2:36 pm

    It would be hard to find a linguist who knows Mandarin and Cantonese "well" and who does not accept that the two have emerged in large part from a common ancestor… thus, surely the two are "dialects" on the above proposal? As would be any pair of Romance languages, etc.? I am not sure where this gets us.

  70. Janet Williams said,

    September 1, 2013 @ 5:29 pm

    @Simon P

    You said "the grammar of Cantonese and Mandarin are very similar. There are a few differences, in word order and verb usage, but the great differences are in pronunciation and vocabulary.

    I recommend The Languages of China (9780691014685): S. Robert Ramsey.

    See the Yue (Cantonese) section, pg 98 – 108.

  71. Victor Mair said,

    September 1, 2013 @ 8:32 pm

    JS has suggested that, according to J Marshall Unger's proposal, Mandarin and Cantonese are both dialects. Does China, then, have no language(s)?

  72. David Morris said,

    September 1, 2013 @ 10:55 pm

    As far as I am aware, the current states of Australia were, prior to federation in 1901, self-governing *colonies* of the British Empire. I have never read or heard the term 'self-governing nations' applied to them.

  73. Victor Mair said,

    September 2, 2013 @ 6:02 am

    @David Morris

    But Australia, New Zealand, and Canada, for that matter, have armies, navies, and flags.

  74. Victor Mair said,

    September 2, 2013 @ 7:30 am

    @julie lee

    The Taiwanese for Koxinga should be Kok-Sì*-iâ in Church Romanization (* = nasal).

    As I suspected, the source of "Koxinga" must be some Minnan dialect, since I believe that he had Fujian connections, even though he was born in Japan. Not sure about that, however, because I haven't done historical research on his family background.

  75. Vanya said,

    September 2, 2013 @ 8:55 am

    Of course the statement "A language is a dialect with an army and a navy." seems trite if taken literally. I am fairly sure that Weinreich, or whoever actually came up with that, never meant it as a serious definition. Weinreich was trying to make a socio-political point about the plight of Yiddish, not draw any linguistic lines in the sand.

    It is interesting that Catalan is generally considered, even by nationalistic Spaniards, to be a "language" although mutual intelligibility between Catalan and Castilian is surely higher than that between most Chinese "dialects". Catalan has no army or navy, nor do Gallego, or Provencal for that matter. (Catalan does have a "national" flag though, some would say.) Catalan has a long standing written tradition however, which most Chinese languages, Arab "dialects" or German "dialects" lack. Couldn't you make a case that "languages", as far as the layman is concerned, only exist when someone takes the time to codify them and write it down? Not a perfect definition either, but more accurate than the Weinreich statement. This might also explain why Serb and Croat nationalists seem more concerned about establishing separate and distinct written standards for their own "languages" then actually having an independent army or navy over the long term (since Croatia is now in the EU and Serbia joining soon, presumably they are both willing to sacrifice a great deal of national sovereignty, but not their languages.).

  76. Frank Harr said,

    September 2, 2013 @ 9:45 am

    "In Guangzhou people are told that 'civilized people speak Mandarin' wénmíng rén shuō Pǔtōnghuà 文明人説普通話, which to me implies that uncivilized people speak 'dialects' (topolects) such as Cantonese."

    Anyone who says something like that deserves a punch in the nose.

  77. julie lee said,

    September 2, 2013 @ 12:23 pm

    @Victor Mair:
    Koxinga (1624-62), who died at 38, reminds me of Sir Francis Drake of England, a great pirate, naval commander. and (in Koxinga's case, free-wheeling) courtier-diplomat-international businessman. (From Jacques Gernet, _A History of Chinese Civilization_, a terrific one-volume history of China originally in French, page 470:) Born in Japan of a Chinese father and Japanese mother, Koxinga from 1650 onwards ruled the coast of Fujian province, making his base on an island near Amoy.
    By chance my son-in-law was this month in Taiwan and visited this small island next to Amoy. His hotel was a lovely old European building from the colonial days, and the island has lots of vestiges of the Europeans. It's now celebrated as the former home-base of Koxinga, and is a favorite of Chinese tourists.

  78. Victor Mair said,

    September 2, 2013 @ 12:59 pm

    @julie lee

    I should have mentioned that the Cantonese for Koxinga would be gwok3 sing3 je4.

    Because of his association with Minnan speaking areas, I still think that Koxinga derives from some Minnan dialect.

    As for the island you were talking about, was it Quemoy (Jinmen)? If so, the beautiful old hotel to which you refer may be the one owned by my friend, Prof. C. K. Wang.

  79. Victor Mair said,

    September 2, 2013 @ 1:16 pm


    I think all that you say makes good sense. I am especially pleased that somebody finally agrees with me that having an army or a navy has nothing whatsoever to do with linguistic entities being dialects or languages. What is dismaying is that journalists are constantly trotting out this hackneyed saying, to the point that it becomes nauseating (it may have been funny the first ten or so times it was repeated, but by now it is worse than annoying and contrary to serious debate on the issues, especially when professional linguists inexplicably leap to its defense).

    I also agree with you that having a written form and a literary tradition definitely helps a linguistic entity to be recognized as a separate language. On the other hand, many of the so-called endangered languages of the world have never been reduced to writing, except perhaps by missionaries.

  80. julie lee said,

    September 2, 2013 @ 3:43 pm

    @Victor Mair:

    Thanks again.
    No, Koxinga's base (1650-62) was not Quemoy (Kinmen) Island. It was the islet of Gulanyu, off Amoy (Xiamen). Wikipedia says: "[Later,] European settlements were concentrated on the islet of Gulanyu…. Today, Gulangyu is known for colonial architecture …."

    Quemoy Island is opposite Amoy in the Strait of Taiwan, and closer to Taiwan. I was there around 1953 with a troupe of Taiwan University students to "entertain" Chiang Kai-shek's troops stationed there (to defend against a possible invasion by Mao's army). Quemoy was totally pristine, undeveloped then, inhabited only by the soldiers. My group did a miserable fake flamenco dance on the stage for the soldiers. I was embarrassed doing it, clicking castanets amateurishly. Later some of them told us they'd have preferred that we hadn't come to entertain them, as they'd gotten used to not seeing women. The officers would look at Amoy through their binoculars and would fire some cannons in that direction every now and then.
    At that time Quemoy was just a semi-desert island, immaculate sand and crystal-clear water. The beaches looked heavenly, but they were mined.

  81. Victor Mair said,

    September 2, 2013 @ 8:33 pm

    @julie lee

    Shanghainese pronunciation of Koxinga: Kōq-shin-h'ia.

  82. julie lee said,

    September 2, 2013 @ 9:40 pm

    @Victor Mair,
    Thanks, Yes, the name Koxinga does seem closest to the Taiwanese or Minnan pronunciation.

  83. Dave Cragin said,

    September 3, 2013 @ 9:35 pm

    In terms of using written language to assess whether it is a “language” or a “dialect” is problematic because of the ~6000 languages in the world, only ~200 are written down. E.g., Most would agree that Navajo was a language well before it was written down.

    For an in-depth look at language vs. dialect as well as how languages change & influence each other, the linguist John McWhorter has some fascinating courses on CD/DVD (not for credit). The best is “The Story of Human Language” and another is the “Understanding Linguistics: Science of language”. He’s one of the most interesting & effective Professors I’ve ever heard.

    @F Harr 'civilized people speak Mandarin' Your note reminds me of comments from at least 3 colleagues from Gangdong who told me “Cantonese isn’t really a language, it has lots of slang.” McWhorter would point out this is the natural evolution of language, i.e., over time they change. As part of this, “slang” today becomes standard language later.

    In regards to change, English and Mandarin show a similarity in that nouns become verbs and verbs become nouns, such as the English "verb" impact. Some English speakers object to this because "impact is a noun", but they will readily use nouns such as “copy” and “worship” that are now "acceptable" verbs. The word “help” (bangzhù 帮助) in both languages is another example, serving both as a noun & verb.

    When groups speaking the same language are separated, their language grows apart evolving into dialects and ultimately languages (as did Latin in Europe). This is notable in Fujian in which travel between towns/cities was very difficult because of the terrain. As a result, Fujianese dialects differ significantly even over geographically close locations. In contrast, Mandarin was a language of trade and this interaction promoted more uniformity over large distances.

  84. Bob said,

    September 4, 2013 @ 10:15 pm

    I should write the following earlier; it is plainly wrong for the 1st paragraph of this post, e.g., the dominacy of Cantonese in Hongkong was the design of the British Colonial Government. Since most of the residents were Cantonese speakers, the colonial rulers saw no point in changing that; and, saving them from educating their lower level staffs, whom had to have contacts with the ordinary residents daily, from another set of training. –the senior staffs hired, or set from British Empire, did not know Mandarin anyway–

    As a matter of fact, PRC let people in the Guangzhou region continued to use Cantonese, including in primary and secondary education, well into the late 1990's (the late Han regions mandated to change to Putonghua)

    Another interesting point, while Cantonese was widely used in Hongkong, in the market places as well as in classrooms, the "better" popular songs and movies were in Mandarin. productions by artists and companies relocated from Shanghai after 1949;–Cantonese songs and movies were not considered with high quality until the mid-1970's– (except the Cantonese opera).

  85. Bob said,

    September 4, 2013 @ 10:43 pm

    re the reason for Mandarin being the primary Chinese in Singapore: in the formulating years (early 1950's) of this young country, where Chinese accounted for 75%+ of its residents, but were from many regions and spoke many Chinese dialects (or, languages), the government seek to unify the education system. The Republic of China (currently, common known as Taiwan) offered its helps –providing text books, teacher trainings, etc.– and were accepted and welcome. Thus, Mandarin became taught in Singapore.
    note: ROC had want to provide educational aides to southeast Asian Chinese "colonies" those years, as a mean to gain influence and loyalty from the various Chinese communities, but without great successes in this regard. –Although everyone took the donations of Chinese test books, and taught Chinese in Cantonese, mostly–

  86. David Marjanović said,

    September 8, 2013 @ 10:22 am

    I informed the class that the Japanese pronunciation was similar to the Hokkien Teh rather than Mandarin Zheng since the Japanese would have borrowed it long before Mandarin had even come into existence, whereas Hokkien retains much older phonological features.

    Of course both Hokkien and Mandarin have innovated. Middle Chinese is reconstructed as having had a retroflex plosive; Hokkien merged it into t, Mandarin changed it into the retroflex affricate zh; Japanese never had retroflexes and thus borrowed the retroflex t as a plain one. Middle Chinese had -ng, Mandarin has preserved it unchanged, Hokkien has turned it into nasalization of the preceding vowel; Japanese used to lack such a thing (nowadays, its final -n is pronounced in different ways in different dialects) and somehow regularly reflects Chinese -ng as vowel length (hence Tei and not Te); finally, Mandarin has merged a bunch of vowel phonemes.

  87. Victor Mair said,

    October 10, 2013 @ 10:27 am

    Speaking vs. writing

    From a native Pekingese student in my LSSC class:

    Last night Brendan shared some pictures from books about Pekingness through Wechat. I found it's so weird when I read them. I feel the subtle meanings of some words were weakened when they were written down. For example, 小力巴儿(伙计). Actually when I heard this word, I would feel a little humorous or disrespectful, according to context. But when I read it on book, no matter with context or not, I couldn't feel those potential emotions. I assume that may be due to the big gap between speech and writing system.
    Speaking of writing topolect, I remember the novel "海上花列传"(sing-song girls of Shanghai) in Last Qing dynasty was written with Shanghainese or Suzhounese. However, only the conversations part were written in topolect while the rest of it were written in Mandarin. I am not sure whether native speakers would have the same feeling when they read these conversations instead of hearing them.

  88. Go for aesthetic appeal said,

    November 28, 2013 @ 2:58 am

    cantonese is on its way out. its a backward and dying dialect of the chinese language. it has evolved much slower than mandarin due to the warm weather in the cantonese area and the segregation of the cantoense area from central china.

    cantonese sounds much less pleasant to ears than mandarin and is harder to learn because its less well regulated and its spoken form being incompatible with the standard written chinese. cantonese has never been well regarded by other chinese . hk's glory in the last half century might have given cantonese a twist of fate but its destined to be short-lived. hk's glory has more to do with benefitting from mainland china's ideology failure and the british rule than anything of cantonese origin.

    being backward and unpleasant is a cause for discrimination if we are intuitively aspired to improvement and progression.

  89. right said,

    December 2, 2013 @ 3:28 pm

    If ever a languagelog comment deserved to be deleted, it is Go for aesthetic appeal's above comment, one of the worst (and most misinformed) I've ever seen.

  90. Victor Mair said,

    December 2, 2013 @ 5:00 pm


    I agree with you completely, but am leaving it up as a specimen of misguided prejudice and gross ignorance.

  91. Mandy said,

    December 2, 2013 @ 7:21 pm

    The legendary "50 cents" gang (五毛)finally made its way onto Language Log!!!! The only way to "engage" them is to provoke them to keep writing, and then delete all their posts before they get paid…there's no other way to do this…

  92. julie lee said,

    December 2, 2013 @ 7:22 pm

    I disagree with Go for aesthetic appeal's comments on Cantonese. Too bad he/she has no appreciation of Cantonese. I rejoice that Cantonese is thriving.

  93. Richard W said,

    December 2, 2013 @ 7:34 pm

    peter wrote:
    "The flag that is the current flag of the US state of Texas was – before it had that status – the flag of an independent nation, the Republic of Texas. That is why I said that its supporters were justified in calling it a national flag, since it was (and remains) the flag of a nation.

    "The National Gallery of Victoria, in Melbourne, Australia, retains this name, despite the fact that Victoria went from being a self-governing nation to being a constituent part of another nation, Australia, in 1901."
    – – – – – – – – – – – – – – – – – – – – – – –
    I don't know much about Texas, but mentioning the NGV weakens the argument that the flag of Texas "remains the flag of a nation," because Victoria isn't a nation any more. "National Gallery of Victoria" is just a name (an anachronistic one that dates back to 1861, long before federation). In the 21st century, we Victorians don't think of Victoria as our nation.

  94. John Rohsenow said,

    December 3, 2013 @ 2:42 am

    Re: Nation/nationalities: from Wikipedia's entry on Zhonghua minzhu:
    "Since the late 1980s, the most fundamental rectification of People's Republic of China's nationalities and minorities policies is the renaming from "the Chinese People" (Chinese: 中国人民 or zhongguo renmin) to the Chinese Nation/Nationality" (Chinese: 中华民族 zhonghua minzu), signalling a shift from the communist statehood with people of various nationalities to a national statehood based on a single minzu (nation/nationality)……
    During the Communist period after Mao's death, the term zhonghua minzu was resurrected to include the mainstream Han Chinese and other 55 ethnic groups as a huge Chinese family."

  95. Go for aesthetic appeal said,

    December 5, 2013 @ 3:02 am

    Before 1990s, Why most of overseas chinese were of cantonese origin? Coz cantonese is the weakest link of chinese. They represented the lowest level of human resources in china, who were driven so desperate that they took whatever the risk including leaving homeland. Check on the history of chinese emigration.

  96. Go for aesthetic appeal said,

    December 5, 2013 @ 3:36 am

    Google or baidu comments/ discussion about cantonese in chinese forum, never short of negative links. Cantonese people being physically unattractive( or simply ugly, the more straight forward way) and the cantonese language being unpleasant to ears are notably the most often mentioned. To say " u don't look like cantonese" has even become a subtle compliment. Check on chinese history, cantonese has never been well regarded by other chinese. Cantonese ppl are referred to as Malay monkeys.

    Guangdong was made the guinea pig for China's economical reform in last 30yrs, just like Guangdong was made the only open port for western business to land on, coz Guangdong is dispensable.

    Warm weather slowing down evolution is no news. Warmth breeds complacency whereas coldness and harshness induce a sense of crisis. We are driven to improve bcoz of pressure to survive.

    Ever wonder why people near the tropical tend to have big and thick lips? does it affect pronouciation? Is it possible big lips also more likely to be with a big tongue? Does it affect pronouciation? Is it also possible big lip big tongue more likely coming together with less refined hearing system…..

    Very easy to label but facts speak louder.

  97. Go for aesthetic appeal said,

    December 5, 2013 @ 3:45 am

    As a language evolves, it's pronounciation becomes more polished and easier to make and thus sounds smoother and more pleasant to ears.

  98. Go for aesthetic appeal said,

    December 5, 2013 @ 3:56 am

    Beijing mandarin is often perceived as pleasant bcoz bj has been a capital city for 1000yrs . Capital city is where the elites and socialites congregate and the pressure for social refinement is greater. Btw, the cultural centres in china have always on the north side do china, along the yellow river. Cantonese area has always been the receiving end of cultural influence from the north, including the cantonese language itself. Cantonese is more expressive for strong and rough emotions such as frustration and anger, eg cantonese often being perceived as more graphic and vulgar for insult — a backward sign though if we aspire to be more civilized and polished. But cantonese is far less efficient for communicating ideas and concepts.

  99. Go for aesthetic appeal said,

    December 8, 2013 @ 11:05 pm

    Cantonese has more hard and throaty or guttural consonants and less palatalization than mandarin. Cantonese also retains the most unreleased final consonants, which are accompanied by glottal stop.

    "Throaty" sounds and glottal stop are often perceived as unpleasant or harsh to ears.

  100. Go for aesthetic appeal said,

    December 8, 2013 @ 11:19 pm

    Gluttural(from Latin guttural "throat" via Medieval Latin guttural is) is used to describe consonants articulated towards the back of the oral cavity.

    Guttural has been inexactly associated with foreign consonants that sound "throaty" to English speakers.

    Speech described as guttural May be deemed not just substandard but sublinguistic(at times even subhuman).

    The value of speech patterns labeled guttural, in other words, is already quite low in the estimation of many, even without the help of the similar sounding but etymologically unrelated gutter. Add to this the fact that gutter is often applied attributively to indicate coarse speech(gutter language, gutter talk, gutter slang etc), and the conflation of guttural and gutter to describe vulgar or distasteful forms of communication seems practically inevitable.

    From there it's a short step to Jon Corzine's "guttural politics".

  101. Go for aesthetic appeal said,

    December 8, 2013 @ 11:21 pm

    No duplication.

  102. Go for aesthetic appeal said,

    December 8, 2013 @ 11:31 pm

    " I think cantonese is one of the harshest languages around, unlike, say, mandarin or other eastern Asian languages. Cantonese is simply very cacophonous and displeasing to the ear.

    As someone who speaks both cantonese and English fluently(but not mandarin), I find these two languages are not entirely compatible. I have heard of many anecdotal reports of how people from hk who speak cantonese who cannot adjust to North American culture because of this linguistic barrier, as opposed to someone who speaks mandarin who nonetheless could acculturate relatively easier inspire of a greater cultural gulf between mainland china and North America.

    Is linguistics playing a role here?"

  103. Go for aesthetic appeal said,

    December 9, 2013 @ 12:04 am

    It's wrong to equate cantonese vs mandarin as German vs English. Instead, Cantonese is more like the Ebonics of the chinese language. Indigenous ppl in the Guangdong area first adopted chinese language when the first chinese emperor conquered the area 2000yrs ago however the accent and pronouciation and ways of expression were still much influenced by the indigenous ppl's own native languages, just like nowadays when a local Indian or Singaporean speaks English, s/he often carries a distinctive local accent.

    Imagine this: Singapore has adopted English as it's official language for 1000 years and the Singaporean English has evolved to become Singalish. The British English after 1000years has evolved to become Benglish. Bcoz of colder weather in Britain, Benglish has evolved much faster than Singalish. When ppl compare Singalish with ancient English, they find more similarities than comparing Benglish with ancient English.

    Would u conclude that Benglish comes from Singalish?

  104. Victor Mair said,

    December 14, 2013 @ 7:29 am

    On "guttural politics", see:


  105. Piyush said,

    December 19, 2013 @ 8:29 pm

    Alexander said:

    People routinely refer to many languages of India as dialects, as in "this lullaby […] is in the Malayalam dialect, the main language of Kerala."

    Disclaimer: I am not a native speaker of any of the South Indian languages, but I am a native speaker of Hindi.

    I have never met or seen anyone with the tiniest clue to linguistics refer to Malyalam as a "dialect". As Prof. Mair said somewhere, "a dialect of what", pray? People who "routinely" refer to Malyalam as a dialect probably also refer to German and French as dialects of each other.

RSS feed for comments on this post