Diglossia and digraphia in Guoyu-Putonghua and in Hindi-Urdu

« previous post | next post »

Having just returned from a month of living and teaching (in Chinese) on the Mainland (in other words, receiving an intensive dose of Putonghua), I was struck by how different Taiwan Guoyu *sounds* in this video.  It's about a subject that is dear to my heart:  the medieval caves at the Central Asian site of Dunhuang with their magnificent wall-paintings and multitudinous medieval manuscripts.

Of course, Taiwan Guoyu (National Language, i.e., Mandarin) is still basically the same language as Putonghua (Modern Standard Mandarin [MSM]) on the mainland, but the sounds and a lot of the words and typical expressions are somewhat different (judging not merely from this one short video, but from other samples, both written and spoken, as well).  And, of course, the script has radically diverged.

It's sort of like the difference between Hindi and Urdu, which are fundamentally the same language, but with the latter having a larger proportion of Islamic vocabulary (chiefly Persian and Arabic) and some differences in pronunciation.  Naturally, Devanagari and Arabic are totally separate scripts, but the *look* of written Taiwan Guoyu and of Mainland Putonghua is also noticeably dissimilar, even to someone who cannot read either of these types of written Mandarin.

W. South Coblin comments on the divergence between Taiwan Mandarin and Mainland Putonghua (PTH):

I think we are all pretty well aware of the differences between Taiwan Mandarin and Mainland PTH, in all the areas you have mentioned. I of course have no first-hand knowledge of the Hindi-Urdu situation. To our Indologist friends I would just say that Taiwan Mandarin has a distinct pronunciation that immediately distinguishes it from Mainland varieties of Modern Standard Chinese. This is because Taiwan Mandarin is most directly derived from the type of Mandarin koiné spoken in the Yangtze Watershed before 1949 and has also been influenced by the pronunciation habits of southern Min and Hakka speakers, who make up the majority of “Native Taiwanese”. In earlier times, and to a considerable extent still today, they have had to acquire Mandarin as a second language. The process of accommodation occasioned by what sociolinguists call “contact convergence”, plus the special origins of the Taiwan variety of the general Chinese koiné, have produced a unique product on the island that understandably differs from what one finds on the Mainland, where the primary basis of Standard Chinese is the speech of general north Chinese, with the sought-after ideal for pronunciation being that of Peking.

I have never heard much about pronunciation differences between Hindi and Urdu. What is usually mentioned in the literature is lexical differences of the sort you in fact allude to….  I suppose one could also talk about morphology and syntax which, again, I have not heard discussed in this connection.

And Philip Lutgendorf remarks on some of the differences between Hindi and Urdu:

Urdu has a number of sounds (such as “z” “f” and a guttural “kh”) that ultimately come from Arabic.  Since these are not represented in the Devanagari writing system, slightly modified characters have been developed to represent them (for example, “z” is “j” with a subdot, “f” is ph with a subdot, etc.). Depending on where you go in North India, some people are unable to pronounce them. Thus, in the eastern Gangetic plain, people pronounce the English word “zoo” as “jew” (joo), because they cannot say “z.” Similarly, my first name is pronounced by such people with the “ph” as a true, aspirated labial rather than the “f” sound we use in English!

There are a few other, minor pronunciation differences, but none of this has any effect on mutual intelligibility.

Hindi-Urdu and Serbo-Croatian are acknowledged instances of diglossia and digraphia (I recall that well for Hindi and Urdu from studying them both back in the late 60s and early 70s).  Considering their mutual intelligibility, yet distinct differences in pronunciation, vocabulary, orthography, and so forth, one might well ask whether Guoyu-Putonghua on Taiwan and on the Mainland have developed to the state of constituting a diglossia and / or a digraphia.  On the other hand, it seems to me that a different sort of digraphia is already emerging on the Mainland, viz., the simultaneous use of pinyin and characters to write MSM.  This was a position that was advocated by the late John DeFrancis and is still being championed by the centenarian Zhou Youguang and other script reformers in China, for which see here, here, and here.



56 Comments

  1. Zora said,

    January 1, 2012 @ 4:04 am

    On Wikipedia, I am dealing (failing to deal, actually) with an editor who insists that Urdu and Hindi are separate languages, that Bollywood movies are in HINDI, and that it is therefore inappropriate to give movie titles in Nastaliq script as well as Devanagari. He was rampaging through all the Bollywood articles removing Nastaliq. Insisting all the while that he is not prejudiced against Muslims, oh noes.

    The attempt to transform Hindustani into two separate languages, as is being done in both Pakistan and India, can have bizarre consequences. The hit Bollywood film, 3 Idiots, features a hilarious sequence in which an unpopular student delivers a prepared speech that is written in Hindi so elevated, so Sanskritized, that he doesn't understand it. Some of his fellow students have sabotaged his copy of the speech by changing some of these unfamiliar words, so that the speaker is unwittingly delivering insults and obscenities. At which the better-educated audience is laughing uproariously.

  2. Observation said,

    January 1, 2012 @ 6:39 am

    I don't think PTH and the so-called 'Guoyu' are really different. Speakers of 'Guoyu' – which is just another name for PTH in TW Province, like 粵語 is to 廣州話/廣府話 – simply can't pronounce the words right because most of the people in TW are originally Fujianese or Hakkanese speakers. Thus most of the people in TW are taught the inaccurate pronunciations of their parents, teachers, etc. This leads to a vicious cycle. As for the vocabulary, is it not just natural that different regions should develop different slang and colloquialisms? For example, 屈機 is a very popular slang term in HK, but most people who speak Cantonese don't understand it unless they're from HK.

    The situation is different, in my opinion, from Serbo-Croatian and Hindi-Urdu. Serbia, Croatia, Bosnia-Herzegovina and Montenegro are separate states and each nation-state has its own government. So are Pakistan and India. However, PTH and 'Guoyu' are both languages of China.

    Also, I think 'Mandarin' is not an accurate way of describing PTH. English isn't called 'fish and chips'.

  3. Victor Mair said,

    January 1, 2012 @ 8:38 am

    From Brian Spooker:

    This is interesting. We need more of this sort of comparative work on language divergence in the modern world. Other examples might be eastern Turkic between the various stans, and western Turkic between the Turkmen east and west of the Caspian. And what used to be Farsi to everyone is now Dari or Tojiki to some. But as you know I worry about the way the use of the term diglossia distorts the assessment of some of these situations.

  4. John Swindle said,

    January 1, 2012 @ 9:17 am

    @Observation: Why is Guoyu "so-called" and not Putonghua? The term "Guoyu" has a respectable history. And surely you didn't mean to imply that Taiwan lacks its own government?

    I agree that the Putonghua/Guoyu case is unlike the Hindi/Urdu and various Serbo-Croatian cases in that nobody says "We have our own language, and it's different from yours." The result, however, may differ from the intention.

  5. Matt_M said,

    January 1, 2012 @ 9:31 am

    @Observation: alas, English has a terrible habit of naming languages and nationalities after foods. It's not only mandarins: English-speaking waiters will also accept orders for danishes, swedes, and even turkey. When restaurateurs are confronted with this shocking lack of respect, some of them will even protest that the foods were named after the language/nationality, rather than the other way around. Not that that's a decent excuse, of course.

  6. michael farris said,

    January 1, 2012 @ 9:46 am

    All this talk of food is making me hungary.

  7. languagehat said,

    January 1, 2012 @ 11:16 am

    but with the latter having a larger proportion of Islamic vocabulary (chiefly Persian and Arabic)

    This may be picky, but it's something that irritates me: the vocabulary of Persian and Arabic is not "Islamic" unless it specifically refers to Islamic things (mimbar, masjid, etc.). The fact that the languages involved were brought by bearers of Islamic religion and Islamicate culture does not somehow make the words "Islamic."

    I don't think PTH and the so-called 'Guoyu' are really different. Speakers of 'Guoyu' … simply can't pronounce the words right because most of the people in TW are originally Fujianese or Hakkanese speakers. Thus most of the people in TW are taught the inaccurate pronunciations of their parents, teachers, etc.

    I think your use of "can't pronounce the words right" and "inaccurate pronunciations" pretty much disqualifies you as a useful observer of linguistic differences. And of course the fact that "PTH and 'Guoyu' are both languages of China" is utterly irrelevant to their linguistic relationship.

  8. Victor Mair said,

    January 1, 2012 @ 11:46 am

    @languagehat

    On your two points:

    #1 excellent observation

    #2 touché

  9. Victor Mair said,

    January 1, 2012 @ 12:01 pm

    @Observation

    You begin with an unfair characterization, continue with a politically charged assertion (a refrain which you repeat later), and end with a bizarre non-sequitur.

    To begin with, what do you mean by referring to the National Language of the Republic of China as "so-called Guoyu"? The name "Guoyu" has a respectable history; it has been around a lot longer than "Putonghua".

    Your second sentence makes the bold claim that Taiwan is a province of the People's Republic of China, but that is a PRC-centric stance, one that does not take into account the counterclaims of many people in Taiwan, Hong Kong, Singapore, and elsewhere throughout the world. The second part of your second sentence admits that — precisely as was stated in the original post — there are obvious differences in pronunciation between Guoyu and Putonghua, which you disparagingly attribute to the people on Taiwan not being able to "pronounce the words right". Your denigration of non-PTH forms of Sinitic "from different regions" is also evident when you refer to them as "slang and colloquialisms". Later on, when you say that "PTH and 'Guoyu' are both languages of China", you self-contradictorily admit that "PTH" and "'Guoyu'" are separate entities.

    The bizarre non-sequitur: "English isn't called 'fish and chips'." You state: "…I think 'Mandarin' is not an accurate way of describing PTH". "Mandarin" is actually a very precise and historically accurate designation for the language in question, and it is not derived from Mandarin oranges, if that's what you were thinking of when you mentioned "fish and chips". Here is the etymology of "Mandarin" from Online Etymological Dictionary:

    ====

    "Chinese official," 1580s, via Port. mandarim or Du. mandorijn from Malay mantri, from Hindi mantri "councilor, minister of state," from Skt. mantri, nom. of mantrin- "advisor," from mantra "counsel," from PIE base *men- "to think" (see mind). Form influenced in Portuguese by mandar "to command, order." Used generically for the several grades of Chinese officials; sense of "chief dialect of Chinese" (spoken by officials and educated people) is from c.1600. The type of small, deep-colored orange so called from 1771, from resemblance of its color to that of robes worn by mandarins.

    http://www.etymonline.com/index.php?allowed_in_frame=0&search=Mandarin&searchmode=none

    ====

    As you can see, Mandarin oranges supposedly got their name from their resemblance to the color of the robes worn by Mandarins, but the language did not get its name from the oranges. So what is the relationship between the language and Mandarins (i.e., officials)? It is very simple, namely, the language we refer to as Mandarin was originally called guanhua 官话 ("officials' speech"), since it enabled officials who came from different parts of the empire and who spoke mutually unintelligible languages to communicate with each other. Subsequently, as the empire disintegrated, guanhua 官话 ("officials' speech") was nationalistically reanalyzed as "public speech", interpreting guan 官 ("official") as gōng 公 ("public"). When linguists of the Republic of China were casting about for a suitable designation, they quite naturally hit upon guóyǔ 国语 (lit., "national language", i.e., Mandarin), an expression that was picked up from Japanese kokugo and ultimately is derived from Sanskrit desa-bhasa ("language of a region / state"; diacritical marks omitted). I will not stoop to say "so-called Putonghua", but I will say that pǔtōnghuà 普通话 is essentially the PRC renaming of guóyǔ 国语 (lit., "national language", i.e., Mandarin).

    I have written about all of this extensively in Victor H. Mair, "Buddhism and the Rise of the Written Vernacular in East Asia: The Making of National Languages," Journal of Asian Studies, 53.3 (August, 1994), 707-751 and elsewhere, for which search Google under: victor mair mandarin guoyu guanhua

    To dispel any notion that a spoken or written language must necessarily be named after a country or nation, Yiddish means the "language spoken by Jews", and there are many other language names that signify they are (or originally were) spoken by a certain group. For example, Hakka means "[Language of] the Guest People / Families". Wikipedia: "Amongst themselves, Hakka people variously called their language Hak-ka-fa (-va) 客家話, Hak-fa (-va), 客話, Tu-gong-dung-fa (-va) 土廣東話, literally, 'Native Guangdong language,' and Ngai-fa (-va) 話, 'My/our language'." For two forms of written Egyptian, we have "hieratic" (γράμματα ἱερατικά [grammata hieratika; literally "priestly writing; writing of the priests"]) and "demotic" (δημοτικός dēmotikós ["popular; of the people; folkish"]).

    For Norwegian, we have Bokmål ("book language") and Nynorsk ("New Norwegian"). Since Norwegian Bokmål and Nynorsk is a classical case of diglossia, this confutes your asseveration that diglossia can only exist when two different nations or states are involved. Another classic case of diglossia was present within traditional China, where vernacular Sinitic (the koine / Mandarin / guanhua) coexisted with literary Sinitic (classical Chinese; wényán 文言). I will not here embark upon a discussion of the exceedingly complicated relationship between vernacular Sinitic (the koine / Mandarin / guanhua) and literary Sinitic (classical Chinese; wényán 文言) on the one hand and the other Sinitic regional and local languages / topolects on the other hand.

    Finally, you completely ignore the question of the dramatic divergence between simplified and traditional scripts on Taiwan and on the Mainland.

  10. jfruh said,

    January 1, 2012 @ 1:04 pm

    Could someone indulge an English monoglot's curiosity the level of divergence and/or mutual intelligibility between (spoken) Guoyu-Putonghua or Hindi-Urdu? Is it something like British English vs. American English, or American English vs. one of the Carribean English-derived Creoles, or…?

  11. languageandhumor said,

    January 1, 2012 @ 1:07 pm

    How different are the pronunciations, vocabulary, and grammar of Putonghua as used in Taiwan and Putonghua as used in Fujian on the Mainland?

  12. arthur waldron said,

    January 1, 2012 @ 1:58 pm

    Guoyu and puthonghua are artificial languages, unlike the Chinese topolects, the product of committees. Lots of people do not distinguish the sh/s etc. not just Taiwanese. I think the person who mentioned Taiwan pronunciation as being from Yangtze valley is quite right. And from the many books and aids designed to teach the artificial language. In China the northerners assume that mandarin is simply their topolect. How do we account for 和 being "han" in guoyu and "he" in the xinhua zidian? In any case, just got finished watching the "cross straits" New Year's show on Beijing TV (here in leafy Gladwyne, PA) and I have to say when it comes to style, inflection, general soft and fuzzy feel, etc. etc. the Taiwan model is flowing powerfully into China–and into Singapore.

  13. Eskandar said,

    January 1, 2012 @ 2:12 pm

    The differences discussed between "Hindi" and "Urdu" (especially pronunciation, but also vocabulary) of course are mostly limited to the artificial standards as used in media and other formal settings such as writing (where the orthographical differences would also become salient) and otherwise are a bit more complex than the description given here. Your average Indian "Hindi" speaker and Pakistani (or Indian) "Urdu" speaker most typically speak precisely the same Hindustani language with the same vocabulary and pronunciation. Illiterate "Urdu"-speakers will often use "Hindi" phonology (ie. [z] will be realized as [j] such as pronouncing 'sabzi' as 'sabji') and well-educated "Hindi"-speakers will use "Urdu" phonology (ie. articulating words like 'qismat' with the prescriptively-correct "Urdu" pronunciation of [qismat] rather than the "Hindi" [kismat]).

  14. Jason Cullen said,

    January 1, 2012 @ 2:12 pm

    Dear Victor Mair,

    I was already to deliver an incredibly pedantic reply to Observation when I discovered your avalanche of stiletto slices.

    If only we could persuade the Taliban to make some ridiculous statement about Chinese languages, then we'd really have them on the run!

    I hope you don't mind me copying your post and sharing it (and thanks for the good reference to your article)! I still have to explain to Shanghainese that 'Chinese' (and not 'Chinaman') is a friendly term!

  15. Eskandar said,

    January 1, 2012 @ 2:24 pm

    Jfruh – didn't see your comment as I was writing mine above. The differences between "Hindi" and "Urdu" at the everyday vernacular level are smaller than the differences between different regional variations of either dialect (ie. smaller than the differences between "Hindi" as spoken in Delhi vs. Kanpur, or the differences between "Urdu" as spoken in Karachi vs. Islamabad). If you put an Urdu-speaker from Karachi and a Hindi-speaker from Delhi on the phone and didn't tell either one which country the other was from, I don't think they would realize the difference as long as the conversation remained in the realm of small talk. Different vocabulary in higher registers of speech would probably give it away eventually.

  16. Ray Girvan said,

    January 1, 2012 @ 5:39 pm

    I don't know if this is a good analogy, but I've always imagined the Hindu/Urdu division, transplanted into English terms, as being like an English that for highbrow use is polarized into an 'Anglic' stuffed with inkhorn Latinate words, and an 'Anglish' stuffed with Germanic coinages (akin to the Latin-free English forms of William Barnes or Percy Grainger).

    [(myl) The situation on both sides of the Hindi/Urdu divide seems to be complex, involving a multidimensional space of times, places, registers, classes, and so on. There's some discussion from a few years ago in "Camp language" (12/31/2007) and "Scripts, Scriptures, and Scribes" (1/3/2008).]

  17. Stuart said,

    January 1, 2012 @ 5:59 pm

    This quote from Lutgendorf intrigued me:
    " in the eastern Gangetic plain, people pronounce the English word “zoo” as “jew” (joo), because they cannot say “z.”"

    What he writes there is true of almost all the Panjabi people I know, too. I worked as a study coach/tutor with a Panjabi nursing student who once wrote "region" in an assignment when she meant "reason". Is the Panjab is considered part of the *Eastern* Gangetic Plain?

    [(myl) I can't answer the geographical question, but on the phonetic on, I'll observe that this issue came up during the Mumbai attacks a few years ago: "Our Z remains Z from Sindh to Punjab", 23/6/2008. In fact, I think that you contributed to that discussion.]

  18. Stuart said,

    January 1, 2012 @ 7:54 pm

    Since it seem to be my day for senescent repetition of old posts, I'm going to throw in another plug for Amrit Rai's "A House Divided" (ISBN 19 561643X) for anyone interested in the Hindi/Urdu digraphia

  19. Circe said,

    January 1, 2012 @ 8:32 pm

    Zora:

    I just had a look at the sequence in 3 Idiots, and it is really contrived. Most of the words involved are by no means too Sanskritized or "elevated". The difference between the meanings of "Chamarkaar" (miracle) and 'Balatkaar" (rape), (on which much of the humour in that speech is based) would be known to almost all Hindi speakers, and also to speakers of most other Indian languages (including at least Marathi, Bengali, Kannada and Telugu), which use the same Sanskrit words for those two concepts.

  20. Circe said,

    January 1, 2012 @ 8:37 pm

    well-educated "Hindi"-speakers will use "Urdu" phonology (ie. articulating words like 'qismat' with the prescriptively-correct "Urdu" pronunciation of [qismat] rather than the "Hindi" [kismat]).

    Having lived in the Hindi hinterland of eastern Uttar Pradesh, I can attest that this is quite far from the truth. In casual conversation, most "Hindi" speakers (educated or otherwise) who have not had academic exposure to Urdu would just use the "Hindi" phonology and pronounce [qismat] as [kismat] and [sabzi] as [subji]. Though, for some reason, when the word is just [sabz], rather than [sabzi], the Urdu pronunciation is more common.

  21. Circe said,

    January 1, 2012 @ 8:44 pm

    A more pronounced difference between "educated" speakers of Hindi and the rest is in the pronunciation of the three "s" sounds (dental, palatal and retroflex) in Hindi/Sanskrit. A lot of the people in eastern UP, even educated ones, would just differentiate the retroflex "s" from the palatal by pronouncing it as "kh", so that "shhaTkoN"[ʂəṭkoːɳ] (hexagon) just becomes "khaTkoN" [kʰəṭkoːɳ]. Several others would not even differentiate between the palatal and retrolflex "s" sounds.

  22. Stuart said,

    January 1, 2012 @ 9:07 pm

    "Several others would not even differentiate between the palatal and retrolflex "s" sounds"
    If this refers to not distinguishing between श and ष surely this is becoming quite normal? Snell mentioned it in the 1st edition of his "Teach Yourself Hindi" years ago.

    When it comes to श and स, is this an issue of education or of region? It certainly seems to be commonly used as an accent marker in the Hindi films I watch, and in words like विदेशी the very limited experience I have had is of native speakers with similar educational backgrounds but from different regions pronouncing it differently .

  23. John Swindle said,

    January 1, 2012 @ 10:05 pm

    languageandhumor's good question hasn't been answered. Does Putonghua on the Fujian coast sound more like the Putonghua of other regions (say North China) or more like the Guoyu of Taiwan?

  24. Jason Cullen said,

    January 2, 2012 @ 12:55 am

    @John Swindle Neither and both. There is a difference between the Guoyu spoken in Taipei and the Guoyu spoken in Kaohsiung, as Taipei is where most mainland Chinese settled after 1949. I even heard one student of mine use 在哪儿 zai4 nar3 'where', a northern pronunciation where most Taiwanese would use the southern Mandarin 在哪里 zai4 na3li3 'where', a form she probably learned–and consciously chose to employ–from her parents. Not to mention that Xiamen is the locus of a prestige form of 闽南话 or Minnanhua, from which Taiwanese is derived, but Fuzhou, the provincial capital, belongs to a completely different dialect. So to answer the question, I would say (hazard?) that the average Taiwanese 'sounds' like a Fujian speaker of Mandarin when compared with a northern speaker of Mandarin–but in terms of syntax and lexicon, that's a whole other bucket of squirrels.

  25. John Swindle said,

    January 2, 2012 @ 5:45 am

    Jason Cullen, thanks! That does answer my question and probably languagehumor's question.

    Meanwhile, to the part of jfruh's earlier question that asked how far apart spoken Taiwan Guoyu and Mainland Putonghua were: They can be very close. I dare say an educated Beijing speaker would be taken as speaking model Guoyu if she were to visit Taiwan and model Putonghua at home. Or they can be farther apart, as Professor Mair indicated in his original post; and the question of Taiwan/Mainland differences is complicated by regional differences even within Taiwan, as was just pointed out by Jason Cullen.

  26. John Swindle said,

    January 2, 2012 @ 5:52 am

    Hmm. What I meant was "…standard Guoyu, if she were to visit Taiwan, or standard Putonghua at home in the PRC."

  27. languagehat said,

    January 2, 2012 @ 9:25 am

    I'm going to throw in another plug for Amrit Rai's "A House Divided" (ISBN 19 561643X) for anyone interested in the Hindi/Urdu digraphia

    Seconded; it's an excellent book (and discusses the entire history of the separation, not just digraphia).

  28. Randy Alexander said,

    January 2, 2012 @ 10:55 am

    First I'd like to note that "guoyu" is used extensively on the mainland as a synonym for putonghua.

    @Languagehat: "pretty much disqualifies you as a useful observer of linguistic differences" Doesn't this make you sound like a (or perhaps the) self-proclaimed judge of who can be a useful observer of linguistic differences? Putonghua is a language with established standards, and being such, a speaker can either pronounce it correctly or not, based on those standards. People whose mother tongues are other Chinese languages often have heavy accents that prevent them from learning this "correct" pronunciation.

    @Victor Mair: "Your second sentence makes the bold claim that Taiwan is a province of the People's Republic of China, but that is a PRC-centric stance." I'm not going to defend Observation's overall stance at all (whatever that may be; I found his/her writing very unclear), but I just want to interject that when I went to Taiwan in 2009 I saw license plates that said 台湾省 (Taiwan Province), so I don't think that Observation was being PRC-centric with that phrase.

    Maybe Observation is just a troll that shouldn't be fed.

    Anyway, Since I moved to Xiamen about a year and a half ago after living in northeast China (Jilin) for eight years, I can comment a bit about some of the specific differences in pronunciation between standard Putonghua and the version of Putonghua that is spoken in Xiamen. There are some small differences in vowels and placement of fricatives that simply give it a different flavor, but there are two huge differences that can cause a lot of "what did you say" responses: the first is that/f/ is usually pronounced as a rounded /h/, so "fangjia" (be on vacation) becomes something like "huangjia" (the "ji" also gets moved toward "dz", but that's not as shocking). The second is that /y/ (ü) becomes /i/, so "xiayu" (rain) becomes "xiayi".

    These things still throw me off but I'm getting faster at "oh yeah, they must mean such and such". In the beginning, conversations could be challenging, but communication got across OK.

    When I went to Taiwan in 2009 it was to see Geoff Pullum, who was lecturing there. We went to see the Palace Museum, and in the cab, the driver was very talkative, asking me all sorts of questions, and Geoff was asking me how well I could communicate with him. I could, but as I told Geoff, his accent was hard to understand.

    I think the degree of difference between mainland Putonghua (a.k.a "guoyu") and Taiwanese Putonghua (a.k.a. "guoyu") is about the same amount as between US English and UK English.

    The education level of the speakers probably plays a substantial part in the level of mutual intelligibility; more educated = more intelligible. Speaking to native Southern Min speakers, I can understand well-educated people better than I can understand less educated people.

  29. Victor Mair said,

    January 2, 2012 @ 11:10 am

    @Randy Alexander

    "…when I went to Taiwan in 2009 I saw license plates that said 台湾省 (Taiwan Province), so I don't think that Observation was being PRC-centric with that phrase."

    Except that, given the context, Observation meant "province of the People's Republic of China", whereas the 2009 license plates that you mention are referring to the Republic of China. Quite a different outlook.

    When I first went to Taiwan in 1970-72, Taiwan was very much considered a province of the ROC, but in recent decades, there has been a definite downplaying of its status as a province of the ROC, and greater emphasis on its independent identity.

  30. languagehat said,

    January 2, 2012 @ 11:28 am

    @Languagehat: "pretty much disqualifies you as a useful observer of linguistic differences" Doesn't this make you sound like a (or perhaps the) self-proclaimed judge of who can be a useful observer of linguistic differences? Putonghua is a language with established standards, and being such, a speaker can either pronounce it correctly or not, based on those standards

    No. "Established standards" is a prescriptivist idea with no relevance to actual language use; it is impossible for native speakers to pronounce their own language "incorrectly." Foreign learners, of course, do often pronounce a language incorrectly (i.e., differently from native speakers).

  31. Syz said,

    January 2, 2012 @ 11:56 am

    Languagehat: not to pile on, but Randy's right. In the case of Putonghua, it's one of those "nobody's native language" situations. There exist standards that people are taught to speak to in school and, in some cases, at work. So the established standards of PTH have a lot of relevance for actual language use here in the mainland.

    Yes, PTH is "based on" the pronunciation in Beijing, but as I've discussed on Beijing Sounds (and as often comes up on Sinoglot, where — full disclosure — Randy and I are co-bloggers) it's not at all *identical* to Beijing pronunciation. As for other regions, they generally have even more differences with PTH that they need to overcome to meet standards.

  32. minus273 said,

    January 2, 2012 @ 12:16 pm

    I concur with Coblin's "Taiwan Mandarin is most directly derived from the type of Mandarin koiné spoken in the Yangtze Watershed before 1949" bit. Even acrolectal Taiwanese Mandarin, little influenced by Min dialects, has a distinct feeling for a speaker of the PRC Mandarin, much more based on the actual dialect of Beijing.

  33. Victor Mair said,

    January 2, 2012 @ 12:51 pm

    From Henning Kloeter:

    The status of a province with officially "frozen" (dong4 sheng3) following a decision in 1996. The provinicial government currently only has an insignificant advisory function, hardly anyone knows and cares who the governor is. The duties of the former provincial government were distributed to other government agencies. I assume that the remaining licence plates with "Taiwan sheng" on it were issued before the constitutional reform. Nowadays, plates should be issued by the city and county administrations; but I have to double-check on that. Apart from that, the provincial designation has largely disappeared.

    It turns out that the Wikipedia entries are quite informative:
    http://en.wikipedia.org/wiki/Taiwan_Province
    http://zh.wikipedia.org/wiki/台灣車輛號牌
    http://en.wikipedia.org/wiki/Vehicle_registration_plates_of_the_Republic_of_China

    Apparently, even after the constitutional reform leading to the quasi-abolishment of the Provincial Government took effect in 1997, the Provinical Government remained an issuing agency of licence plates, together with Taipei City, Kaohsiung City, Kinmen County, Lienchiang County. Since 2007, these names do not appear any longer on newly issued licence plates.

    My impression is that nowadays, the term "Taiwan sheng" does indeed convey a sense of PRC-centrism. Also, expressions like _quan sheng_ 'whole/entire province' (referring to Taiwan in media reports) are used to a much lesser extent than 10 or 15 years ago. Instead, saying or writing "quan Taiwan" has become more common.

    New license plates issued since 2007 have on them only numbers/letters, no references to issuing agencies.

  34. Randy Alexander said,

    January 2, 2012 @ 11:50 pm

    @Languagehat: It's a really really bad unscholarly habit to thoughtlessly slam people with knee-jerk reactions without doing your homework or thinking through something first.

    "Established standards" is a prescriptivist idea with no relevance to actual language use

    Yes, it's a prescriptivist idea, but "prescriptivist" does not equal "evil", and it often has a lot to do with actual language use. In China you cannot teach in a public school, or become a radio or TV announcer without passing a standardized test in putonghua (there are other jobs that require it also; you can read about it here (in Chinese): http://baike.baidu.com/view/21049.htm). Prescriptivism is useful in some circumstances; it allows me to be able to communicate with almost anyone no matter where I travel in China and, and in a more general sense, foreign language education in any major language would be faced with perhaps insurmountable challenges without prescriptivism providing models to emulate.

    it is impossible for native speakers to pronounce their own language "incorrectly."

    I bet that mothers the world over would disagree with you. I am a native speaker of English. If I pronounce the word "cap" as /ʃɪt/, I would be pronouncing it incorrectly. See there? It's possible.

    Descriptivism is not a cause, it's a perspective; a very useful one for scientists to determine the differences between languages and dialects. Prescriptivism has its uses as well. Of course prescriptivism can be used to stamp out minority languages and dialects, and there is a cause against that (and I support that cause).

    "Established standards" is a prescriptive fact of life for Putonghua, and it has enormous relevance to actual language use. Knee-jerk reactions are far less relevant or useful.

  35. languagehat said,

    January 3, 2012 @ 9:15 am

    "Established standards" is a prescriptive fact of life for Putonghua, and it has enormous relevance to actual language use.

    It's "a prescriptive fact of life" all over the place, not just for Putonghua, and it's still nonsense if it's supposed to apply to people actually using a language in daily life. "Established standards" are relevant for artificial languages like Esperanto or dead ones like Latin, where users are consciously trying to make sure they're using the same rulebook. To the extent that such languages start being used in daily life (which presumably is the goal for, say, Putonghua), the only "standards" are how the speakers actually use them. Your disingenuous attempt to present yourself as down with descriptivism, as long as it knows its place, reminds me of David Foster Wallace in his terrible "Tense Present" article.

    If I pronounce the word "cap" as /ʃɪt/, I would be pronouncing it incorrectly. See there? It's possible.

    LOL u r funny

  36. Victor Mair said,

    January 3, 2012 @ 10:23 am

    From Mark Swofford:

    Yes, in Taiwan one can see many license plates marked TAIWAN SHENG (in
    Hanzi). This, however, has absolutely nothing to do with the feelings or
    political sympathies of the majority of people in Taiwan. And I agree with
    your characterization that someone outside Taiwan who refers to "TW
    Province" is very likely taking a "PRC-centric stance."

    So how do I explain the apparent discrepancy? This is a bit complicated.

    And although the roots of this license plate labeling are indeed in
    Sino-centrism, it's a legacy of what Chiang Kai-shek and the Kuomintang
    forced upon Taiwan. They set up a national government to handle every
    place under their control — and supposedly to handle China, too, even
    though it wasn't under their control. But each region had its own
    government as well. So there was a national (ROC) government and a
    government for what was called Taiwan Province. Because Mazu and Jinmen
    were under ROC control but not nominally part of Taiwan Province, they had
    separate local governments. And because Taibei and Gaoxiong were at the
    level of "special municipalities" they weren't under the Taiwan Provincial
    Government but directly under the national government.

    As people might expect, there was a great deal of overlap between the
    national government and the Taiwan Provincial Government, seeing as how
    they both controlled largely the same area. This was not just ridiculous
    but wasteful and expensive. And so about thirteen or fourteen years ago
    Taiwan finally abolished the Taiwan Provincial Government. And good
    riddance.

    Getting back to license plates now, each government put out its own
    plates. So, yes, there were license plates for Taiwan Province. But there
    were also plates for Taipei (TAIBEI SHI), Kaohsiung (GAOXIONG SHI), Kinmen
    County (JINMEN XIAN), and Lienchiang County (LIANJIANG XIAN, i.e.,
    Matsu/Mazu).

    This was all rather silly. But people the world over are used to the
    oddities of governments. And in Taiwan, officially "the Republic of China"
    even though most people are uninterested in or even disdainful of that —
    people are especially used to paying little attention to such officially
    declared distinctions and going about their lives.

    Eventually, however, even politicians recognized that having a government
    that no longer officially existed issue license plates was absurd and
    pointless. So rather than come up with something new to call everyplace in
    Taiwan except for for Mazu, Jinmen, Taipei, and Kaohsiung, the officials
    came up with a simpler solution: they just eliminated the place
    designation from license plates. So for the past five years new plates are
    blank except for their tag numbers.
    http://www5.thb.gov.tw/ThbNews/anys/showsub.aspx?TitleID=99

    That's just as well, because about a year ago there was some political
    restructuring, with many other places becoming special municipalities, so
    what used to be Taiwan Province would have grown considerably smaller and
    there would have had to be a lot more place names on license plates.

    But if one really wishes to look for a significant cultural lesson to be
    gained from studying license plates in Taiwan, I think it would be far
    more fruitful to examine how superstition led to the abolition of the use
    of the number 4.
    http://pinyin.info/news/2005/taiwan-to-abandon-use-of-4-on-license-plates/

    If I had to choose just one word for the "Taiwan Sheng" situation, I'd pick "anachronism."

  37. Barbara Partee said,

    January 3, 2012 @ 5:36 pm

    I think I'm more with Randy than with Language Hat on the side issue those two got into. Maybe it would be useful to start that bit of the discussion over trying not to use the word "prescriptivism" at all — that word seems to call up a lot of extra emotion.

  38. Bob Violence said,

    January 4, 2012 @ 1:09 am

    Some figures on the differences between standard Guóyǔ and standard Pǔtōnghuà:

    A comparison was made between the Xīnhuá zìdiǎn of the mainland and the Guóyǔ cídiǎn of Taiwan with regard to the pronunciation of 3,500 most common characters. It is found that the two sides differ on the pronunciation of 789 characters, accounting for 23 per cent of the total number of items under investigation.

    (From Modern Chinese: History and Sociolinguistics, p. 46)

    789 of the top 3,500 characters would equate to thousands of words with different pronunciations, and this doesn't even take full account of Guóyǔ as actually spoken in Taiwan (as influenced by Taiwanese Hokkien and Hakka) — the writer goes on to assert (much like Coblin) that these standard Guóyǔ readings were largely established in the '30s and '40s, before the KMT even arrived in Taiwan. I assume that mainland dictionaries of the early PRC period would've hewed closer to these pronunciations than to the current Pǔtōnghuà standards.

  39. Circe said,

    January 4, 2012 @ 11:12 pm

    "When it comes to श and स, is this an issue of education or of region? It certainly seems to be commonly used as an accent marker in the Hindi films I watch, and in words like विदेशी the very limited experience I have had is of native speakers with similar educational backgrounds but from different regions pronouncing it differently ."

    I think it is a function of both region and education. If I can remember correctly, in Bhojpuri (I am not a native speaker of that dialect) विदेशी is supposed to be written (and pronounced) बिदेसी. I think the same is true of some other dialects (like Awadhi) too: the Ram Charit Manas, for example, often uses "स" or "ख" in its Awadhi parts where the Sanskrit pronunciation would require "श" or "ष".

  40. vanya said,

    January 5, 2012 @ 8:43 am

    it is impossible for native speakers to pronounce their own language "incorrectly."

    Oh, it is quite possible. If you are a 10 year old child your peers will probably make that quite clear to you. Hence the existence of speech therapists.

  41. Victor Mair said,

    January 5, 2012 @ 11:20 am

    @vanya

    But what do you make of this comment by Randy?

    ====

    "I am a native speaker of English. If I pronounce the word 'cap' as /ʃɪt/, I would be pronouncing it incorrectly. See there? It's possible."

    ====

    Does this help the debate at all? Is it pertinent? Would any child (or even adult) attempting to pronounce "cap" conceivably pronounce it as /ʃɪt/? It seems to me that Randy has offered an *intentional* mispronunciation that is grossly and ludicrously unrelated to the topic at hand. It is not germane to our discussion; rather, it drags the discussion down to an unproductive level.

  42. Boyang said,

    January 5, 2012 @ 1:53 pm

    I have tried to google sound clips of a 'genuine' ph, the aspirated labial, and hear the difference to a f, but I could not find any. Can Language Log provide anything alike? (Maybe also sound comparisons between th, aspirated t and unaspirated t).

    Thanks.

  43. Circe said,

    January 5, 2012 @ 6:59 pm

    Boyang:

    I tried looking for audios teaching Hindi phonology (As a native Hindi speaker, the "correct" Hindi pronunciation of the letter "फ" is an aspirated labial), and all I got was people teaching me how to pronounce "f" (which in HIndi would be represented as "फ़"). But there seems to be this website teaching Sanskrit pronuciation which seems to get it right: http://www.tilakpyle.com/sanskrit_alphabet.htm

    Look for the row marked "labial". The second character in that row is what (I think) you are looking for. There is a little green "play" button on the row which plays the whole row, but I think the sounds are reasonably well separated.

  44. Circe said,

    January 5, 2012 @ 7:02 pm

    However, I think in the very same row, their pronunciation of the voiced aspirated labial (भ) is certainly different from the usual HIndi pronunciation (and, in my experience, from the Sanskrit pronunciation too).

  45. Vijay John said,

    January 6, 2012 @ 12:54 am

    Sounds like the guy pronouncing the letters on that website is just some American dude (not a native-speaker of Hindi or anything).

    If you're looking for examples of the four-way aspiration contrast in Hindi, may I suggest http://hctv.humnet.ucla.edu/departments/linguistics/VowelsandConsonants/vowels/chapter12/hindi.html?

  46. Stuart said,

    January 6, 2012 @ 1:09 am

    Thanks for that link, Vijay John. It would have been even better if it had included clear examples of the difference between श and ष, the only pair I still have real trouble distinguishing.

  47. Vijay John said,

    January 6, 2012 @ 1:24 am

    Stuart, my sister-in-law is a native speaker of Hindi. She may be born and brought up in the US, but her parents are from near Delhi (her mom is from Muzaffarnagar and her dad is from Ghaziabad). Their family is Hindu.

    She makes absolutely *no* difference between श and ष. (My first language, Malayalam, does). I'm pretty sure lots of Hindi-speakers don't even make a difference between any of the three. I thought they pronounced all of them just like स (but judging from some of the above comments, it might be somewhat more complicated than that). Both श and ष are relatively new to Hindi, being introduced by loanwords that are usually from either Persian or Sanskrit.

    Wikipedia seems to suggest that those who do distinguish श and ष pronounce ष as a voiceless retroflex fricative. I don't think I've ever heard that pronunciation before, and I'm pretty sure these sounds do NOT contrast in Hindi, but here's a language where clearly they do make that distinction: http://hctv.humnet.ucla.edu/departments/linguistics/VowelsandConsonants/vowels/chapter13/toda.html. (Click on the word for 'language' and the clan name).

    Anyway, it's so nice to see that two of my posts actually helped someone! Go ahead and call me Vijay :)

  48. Circe said,

    January 6, 2012 @ 3:01 pm

    Vijay John: I am a native Hindi speaker (from eastern UP), and my mother (who used to be a high school Hindi teacher) hammered it into me that "श" and "ष" needed to be pronounced differently, with "श" (also called तालव्य, menaing "of the palate" in Hindi) pronounced with the toung touching the palate, and "ष" was to be pronounced with the toung curled back even more than in the case of "श".

    However, in my experience, most native speakers (with the exception of professors and teachers of Hindi) either do not try to differentiate at all, or (especially in parts of Eastern UP) pronounce "ष" as "ख". As I said above, this is quite common in Hindi in deriving words from Sanskrit ("शुष्क" (dry), for example, become "सूखा" in Hindi).

    In fact, I just found that the Hindi wikipedia article on "ष" explains this phenomenon very well. Here is the link, if you can read Hindi in devanagari script: http://hi.wikipedia.org/wiki/%E0%A4%B7

  49. Circe said,

    January 6, 2012 @ 3:04 pm

    The Hindi Wikipedia article also has audio demonstrating between the Sanskrit pronunciation of "ष" and how it is different from the currently popular Hindi pronunciation.

  50. Stuart said,

    January 6, 2012 @ 10:52 pm

    Thanks for the added information, Circe. It confirms what I suspected, that the distinction is being lost in actual, everyday use. That's reassuring for a learner, since a mythically shuddh Sanskritic Hindi is not what I'm trying for.

  51. Circe said,

    January 6, 2012 @ 11:37 pm

    You are welcome, Stuart.

    That's reassuring for a learner, since a mythically shuddh Sanskritic Hindi is not what I'm trying for.

    I'd recommend against it too. Speaking in "shuddh" Hindi is the best way to communicate to a native Hindi speaker that you are not really a native speaker of the language. :) This is partly is due to the fact that many other Indian languages (such as Kannada, Telugu or Bangla) have a more Sanskritized everyday vocabulary (since their contact with Persian was smaller). Native speakers of these languages who are learning Hindi therefore have a tendency to use Sanskrit (tatsam) words where a native Hindi speaker would find such usage to be extremely formal.

  52. Randy Alexander said,

    January 9, 2012 @ 5:56 am

    @Victor: I apologize for being so gross, but my point was to show how absurd Languagehat's comment was that it's "impossible for native speakers to pronounce their own language 'incorrectly'" (overlooking the fact that "pronounce their own language" has quite an indistinct meaning, and taking it as "pronounce words in their own language").

    I could (well, should have) said something more like "pronounce the word 'cap' as /bag/, or maybe even as /kan/". I specifically chose something that would be an intentional mispronunciation simply to demonstrate the falsity of his statement; if one can intentionally mispronounce things, then it is certainly possible. Once we accept that it is possible, we might begin to take this further and try to draw lines according to what could be considered correct, and what is incorrect. We would soon discover that this doing so opens a big can of worms — who draws the lines? We can accept variant pronunciations to a certain degree, but past that degree pronunciations somehow become "incorrect". For any given language (even a largely prescribed one like Putonghua), what constitutes acceptable variation in pronunciation is extremely complex — too complex to draw some kind of rule-based lines.

    On the "prescriptivist" problem, I agree with Barbara that perhaps it's best to avoid that particular term (though I am quite fascinated by the fact that it ruffles so many feathers). I suspect the main problem is the suffix "-ist", which makes it seem like an emotionally charged cause instead of a simple point of departure.

  53. LeGrand said,

    January 10, 2012 @ 3:16 pm

    An anecdote.

    There are a variety of local pronunciation differences that make Taiwan Mandarin(s) fairly distinct. The hu/f variation mentioned above is one that featured in an amusing (though tense) incident I observed in a cafeteria in Taipei some years ago.

    A fellow a few people ahead of me in line had gone down row of dishes, putting his tray out to be served (in the usual wordless way ) when he came across a dish fancied . At the end, where the rice was served, he stuck out is tray again, but the woman serving the rice was distracted by some conversation, and didn't notice his gesture. He made it again, more clearly, and she continued to be distracted. So he said,

    "Huan4 gei wo".

    Her eyes tracked over to him, in some puzzlement.

    He said again, a bit more distinctly, "Huan4 gei wo". She didn't move.

    This time, with some agitation, he said, "HUAN4 gei wo!" She looked alarmed; her eyes darted from side to side.

    "Xiaojie, baituo gei wo yi wan huan4!" So she quickly loaded up his tray with a big bowl of rice, saying apologetically,

    "Duibuqi xianxheng, wo mei ting qingchu." (Sorry, mister, I didn't hear
    what you said.) At which he humphed loudly (as we describe it in English), and went off to chow down.

    +

  54. SeekTruthFromFacts said,

    January 10, 2012 @ 4:35 pm

    @languagehat:

    "Established standards" are relevant for artificial languages like Esperanto or dead ones like Latin, where users are consciously trying to make sure they're using the same rulebook."

    Can you review Syz's commend again? If we place languages on a spectrum, isn't Mandarin a lot nearer the artificial pole than English is? An adult child growing up in rural 1960s China may have experienced Mandarin learning much like someone learning Latin. Of course, they should have been in a Mandarin immersion environment, but it didn't always work in practice…

  55. Chau H. Wu said,

    February 24, 2012 @ 6:20 pm

    arthur waldron wrote, "How do we account for 和 being "han" in guoyu and "he" in xinhua zidian?" This is the second time I saw the question raised in Language Log. I would like to share an explanation I heard when I was in high school in Taiwan. A Mandarin teacher of mine said that the "han" pronunciation was adopted during World War II. (The teacher, a Mainlander, was in college during the war.) The Chinese people suffered greatly during Japanese invasion and the hatred toward the Japanese (who called themselves 大和民族) was so much that the "he" pronunciation for 和 became a taboo. So it was changed to "han" (4th tone), becoming homophonous with 漢 "han". However, this pronunciation is restricted to the meaning of "and". In all other uses such as 和平 'peace', 和尚 'monk' etc. it is still "he". I don't know how valid the explanation is, but it seems to me quite logical.

    BTW, on the topic of diglossia, after more than 100 years of separation, the Minnan variety in Taiwan has become quite different from those varieties in Fujian. For example, the pronunciation for 艦 differs between Taiwan and Fujian (Xiamen). It is kàm in Taiwan but lâm in Xiamen, reflecting the "diglossic" nature of the phonophore 監 as noted by Bernhard Karlgren. Lâm was used in my grandfather generation, but must have changed to kàm through the influence of Japanese (Japanese 艦 kan).

  56. Daniel said,

    April 9, 2012 @ 10:30 am

    I think two entities should be distinguished: (1) standard Guóyǔ as it is officially described, and (2) Guoyu as it is popularly spoken in Taiwan. In terms of their official description, the phonology of Guoyu and Putonghua are identical, since both are based on Beijing pronunciation. The KMT government initially tried a “chimera-like” definition for Guoyu incorporating the phonology of different Mandarin topolects (e.g., combining the rusheng of Nanjing with the four tones of Beijing), but abandoned it in 1932 in favor of Beijing phonology. Of course, over the decades, the pronunciations of some words in the Mainland and in Taiwan have diverged.

    I certainly do not intend to add fuel to the “prescriptivism” debate, but there is some basis to the contention that the way Guoyu is spoken now in Taiwan can be considered “incorrect,” since it often falls short of the standards of Guoyu. This is the position taken by reference books on language published in Taiwan, such as the 國音學 Mandarin Chinese Phonetics by the National Taiwan Normal University (8th ed., published 2010). As already pointed out, this is most likely due to the influence of Hokkien, and maybe Hakka. Despite their prevalence in Taiwan, the substitution of the zh-ch-sh initials by z-c-s, the substitution of “f” by “h”, and the substitution of the schwa vowel (like in 夢, 風) by “o” are still frowned upon as “errors” by authorities (Mandarin Chinese Phonetics pp. 68, 92-94). In Taiwan TV programs up to the 90s, the actors and hosts still spoke in relatively standard Guoyu, but during the last decade more and more “Taiwan-accented Guoyu” 台灣國語 has come to be heard on TV. As for news anchors, their pronunciations of individual sounds are still relatively standard, but even they nowadays speak in a very syllable-timed manner, and code-switching to Hokkien is frequent. Does this mean that (1) the general level of Guoyu has deteriorated in Taiwan, or that (2) the language proficiency rules for broadcast media have become less stringent (meaning, in other words, that the Guoyu spoken by the common people has always had a substantial Minnan accent and lexical influence, and was just “concealed” in the past by well-trained TV actors and hosts)?

    Mandarin topolects are usually grouped into Jiaoliao, Jianghuai, etc. (if we assume that there are valid criteria for dividing the topolects into groups; please see http://languagelog.ldc.upenn.edu/nll/?p=1211#comment-186916). Perhaps the various varieties of Guoyu spoken around Taiwan can be considered a new group within Mandarin?

RSS feed for comments on this post