Speak Cantonese

« previous post | next post »

We started teaching Cantonese at Penn more than a quarter of a century ago, and it has been a very successful program.  Yale was teaching Cantonese long before us, and the title of this post is the same as that of a famous Yale textbook for the teaching of Cantonese by Parker Po-fei Huang and Gerard P. Kok.

I'm very pleased that more and more schools are offering Cantonese, and I'm hoping that the same will hold true for Taiwanese, Shanghainese, and other Sinitic languages.  Penn offers a dozen modern South Asian languages, which shows that linguistic diversity is possible in universities when there's a will to make it happen.

Cantonese language instruction is booming in Hong Kong as well, and that is entirely appropriate, since — although Cantonese is the Mother Tongue of the overwhelming majority of the population — in recent years, it has increasingly been threatened by the rise of Mandarin as the language of the central government, which has been exerting ever greater control in the SAR (Special Administrative Region), particularly after the British returned the former colony to Chinese suzerainty in 1997.

In "Can you truly understand Hong Kong if you don't speak Cantonese?  Uptick in local language study enrolment points to desire for greater social understanding" (SCMP, 6/10/16), Jessie Lau documents the growing numbers of individuals who have been enrolling in Cantonese language courses in Hong Kong.  She also raises a number of important issues about how Cantonese should be taught and other related questions as well.

For example, she makes a crucial point when she writes:

Since spoken Cantonese and written Chinese differ vastly, educators must grapple with whether to teach spoken or written first.

Any educator who tries to teach written Chinese first as a way to get at Cantonese is not worth their salt.  As we have shown repeatedly on Language Log, written Chinese is much, much closer to Mandarin than it is to Cantonese, so if you start out by teaching written Chinese, you wouldn't be teaching Cantonese at all; you'd be teaching Mandarin (and would be doing a bad job at that as well, since — in terms of pedagogical methods — teaching Mandarin via written Chinese is frustrating and ineffective.  Consequently, Lau is entirely correct when she says that "spoken Cantonese and written Chinese differ vastly", but when she states that "educators must grapple with whether to teach spoken or written first", that really should be a non-starter.

Lau is also correct when she says:

In addition to being phonetically complex, highly colloquial and nuanced, Cantonese also does not have a standard qualification system. There are currently no standardised Cantonese tests or certifications, leaving individual programmes and centres to create and offer their own.

All of that is true, as is the lack of a Romanization system, such as Pinyin for Mandarin, that is widely accepted as standard or official.  There are more than half a dozen Romanizations for Cantonese, but the educational authorities in Hong Kong are doing a terrible disservice to students, not only in Hong Kong, but everywhere, when they fail to authorize any particular Romanization for Cantonese.  In my estimation, every Hong Kong school child should be taught how to spell the sounds of their mother tongue, just as they do in China.

The same Cantonese boom that is taking place in Hong Kong is occurring elsewhere.  For instance, at the University of British Columbia, with a two million dollar donation from Alex and Chi Shum Watt, Cantonese language instruction is off to a brisk start as of fall 2015, and they are even employing an innovative device for helping students to pronounce Cantonese correctly:

"UBC researchers use ultrasound technology to teach Cantonese: UBC researchers are partnering with the Cantonese-language program to pilot a technique that combines ultrasound imagery with follow-along video clips" (Wanyee Li, Toronto Metro, 6/9/16).

Given that speakers of Cantonese and closely related languages number around 80,000,000 worldwide, and the fact that its economic and cultural impact is considerable, this is a language that fully deserves to be far more widely taught.  It would also help the language to survive and prosper if more native speakers would write in Cantonese.  It can be done; they simply have to do it.

Speak Cantonese, then write it.

[h.t. Mark Metcalf]


  1. WSM said,

    June 10, 2016 @ 4:28 pm

    All of this activity overseas, even as the Chinese government is taking steps to prevent foreign students from getting visas to study any languages other than Mandarin inside China (ref a tweet by @farwestchina on 7 June). Xinjiang and Uighur (which isn't even Sinitic) may be extremely special cases, but it's a disturbing precedent that leads one to imagine a linguistic diaspora in which topolects flee overseas as they are gradually banned at home.

    The great irony in all of this is that, as many know, spoken Cantonese/Minnan/Hakka are in some senses more traditionally "Chinese" than Mandarin, and knowledge of these dialects helps immeasurably in the appreciation of phonic properties of classical chinese poetry and prose. I've always regretted not learning more Taiwanese while I was in Taiwan; hopefully someday that shortfall may be remedied.

  2. Rubrick said,

    June 10, 2016 @ 4:31 pm

    I'm a bit surprised (out of ignorance) that the number is as low as 80,000,000. About what would it be if you replaced "closely related languages" with "languages/topolects more closely related to Cantonese than Mandarin"?

  3. JS said,

    June 10, 2016 @ 6:05 pm

    The visa policy change is disheartening (esp. as studying languages other than Mandarin in China is a pursuit I personally much enjoy), but the Tweeter you point to is speaking specifically of Xinjiang, where such a policy is no less disheartening but at least relatively unsurprising. Am interested to hear any other news on the matter…

  4. Lazar said,

    June 10, 2016 @ 6:16 pm

    "languages/topolects more closely related to Cantonese than Mandarin"

    Well, this definition might include things like Hakka that no one would consider Cantonese. A better alternative might be "languages/topolects more closely related to Cantonese than to any other of the 'big seven'".

  5. WSM said,

    June 10, 2016 @ 6:46 pm

    @JS – Right, and I doubt that such policies are in effect anywhere else in China at the moment. Still, it's very disquieting to see such tactics being used.

  6. Shannon said,

    June 10, 2016 @ 9:58 pm

    I've noticed that sometimes, Chinese Americans, as well as others in the Chinese diaspora, who come from families with non-Mandarin speaking backgrounds (eg. Cantonese, Hokkien) will learn Mandarin as the default Chinese because of Mandarin's dominance nowadays, but still perceive it as a "heritage language" even if it wasn't an actual heritage language in their family.

    I don't know if it is common for some folks to want to preferably learn their heritage Chinese language such as Cantonese or Hokkien rather than Mandarin through taking classes etc. but if said heritage language is not offered for teaching, then settling for Mandarin as the next closest language to their heritage one (I know Cantonese has been taught in Saturday schools for decades in the US though). Or if it's more the case that the justification for learning Mandarin among Chinese-Americans whose ancestors never spoke it, is mainly practical or economic rather than heritage-related.

    Nonetheless, it's interesting that often times, an American with roots in the Pearl River Delta area of China reconnects with their roots by "learning Chinese" that's not really the Chinese of their ancestors.

    I wonder if this kind of perception of what's considered reconnecting to their heritage language (or "close enough" to a heritage language) would be common for other groups — eg. a non-Hindi-speaking but Gujarati- or Punjabi-speaking American learning Hindi as a language of the old country. Or even something like an Italian-American reconnecting with their roots but learning another Romance language instead of the one spoken by their grandparents etc.

  7. Lazar said,

    June 10, 2016 @ 11:49 pm

    I think an analogy can be drawn with American Jews, for whom Israeli Hebrew has largely displaced Yiddish as their perceived heritage language. From my understanding, "gut/good Shabbos" and "Shabbat shalom" roughly serve as shibboleths for the Orthodox and non-Orthodox, respectively.

    (And for my part as a half-Jew and aspiring polyglot, I can certainly see why Hebrew is the more attractive choice. I do feel affection for Yiddish, but in the end it's a marginal and only partially standardized language largely extinct outside of insular religious communities, and I find that my interest in it can rather easily be sublimated into standard German.)

    The situation does differ from the others, of course: Hebrew was never totally absent from any branch of Jewish culture, while at the same time the two languages are from totally different families.

  8. Victor Mair said,

    June 11, 2016 @ 12:00 am

    Just received this from a correspondent:


    An American contact in XJ has informed me that foreign students (particularly Westerners, I'm sure) are no longer allowed to study minority languages in Xinjiang on a student visa. It's apparently been a law on the books since Jan 1 but no one was sure if they would enforce it. Now it's clear that they will. They're also enforcing the 4-year rule – you can't live in XJ on a student visa for more than 4 years.


  9. tsts said,

    June 11, 2016 @ 12:26 pm

    @Shannon: concerning your scenario of Italian-Americans reconnecting with their heritage by learning "another" Romance language, this is quite common, in a way, since many of their ancestors did not speak Italian, but languages such as Sicilian or Lombard. Same for descendants of immigrants from the Occitan region in Southern France who are reconnecting with their heritage by learning French rather than Occitan, and a number of other examples.

    In fact, what is happening in China in terms of languages reminds me of what happened in Europa a hundred years ago, when universal schooling, the rise of mass media, and some amount of official pressure led to many smaller languages losing most of their speakers. There are a lot of differences, but some similarities. (Note also that many Chinese immigrants in the US originally spoke Toishanese, not Cantonese. Their descendants might have a hard time finding Toishanese courses if they wanted to connect to their heritage that way.)

    Anyway, Prof. Mair's comments on learning Cantonese are spot on, from my much much more limited perspective on this issue. Though I am actually fairly optimistic about the future of Cantonese – it seems very alive and well in Guangdong, and I have seen a number of cases of non-Cantonese speakers (some Mandarin speakers in Shenzhen, Fuzhounese speakers in NYC) picking up fluent Cantonese after a few years.

  10. OHO said,

    June 11, 2016 @ 10:46 pm

    why is that in every native english speaking country, standard english is taught? why don't they teach Shakespeare English which is old?

    Now, any mentality behind the practice of referring Hong Kong and China sperately in a single sentence?

  11. AntC said,

    June 12, 2016 @ 8:29 pm

    Thank you Victor. When you say … hoping that the same will hold true for Taiwanese …

    You've written before of Taiwanese as if it were a single language/topolect, including I've read your paper of some years ago.

    But as I understand it, there's several sinolects spoken in Taiwan (including Hakka and Hokkien), as well as what are called 'aboriginal' languages — that is pre-sinolects.

    So what would "teaching Taiwanese" involve?

  12. Ajax said,

    June 12, 2016 @ 10:24 pm

    AntC: Taiwanese is Taiwanese Minnan aka Taiwanese Hokkkien. It's not Hakka and definitely not the aboriginal languages.

  13. JK said,

    June 13, 2016 @ 10:56 am

    There has a been a major push over the last year or so toward "dual-language" education in Xinjiang. While a few decades ago Mandarin was taught to Uyghurs as a "foreign language", now Mandarin is referred to as the "national common language" (国家通用语言) (I'm not sure if that is what it is referred to in Hong Kong as well). As far as I can tell, the teachers who taught exclusively in Uyghur were fired and told to reapply once they were able to teach bilingually. Xinjiang has also doubled its recruitment outside of the region of "bilingual teachers" this year. They are building boarding schools for compulsory bilingual education (parents can be imprisoned for not sending their children to compulsory education), meaning Uyghur students will not return to a home language environment in Uyghur after school each day, and they are pressing to extend "bilingual education" to multiple years of preschool.

  14. Eidolon said,

    June 13, 2016 @ 6:57 pm

    @Ajax that is what is usually accepted, but given the sensibility about not using "Chinese" to stand for all Sinitic languages in the PRC, the same ought to apply to "Taiwanese" in the ROC. There are large populations of native Hakka and Mandarin speakers in Taiwan, and if we were to talk about the most popular language, that'd actually be Taiwanese Mandarin, the official language and lingua franca. Only when restricting ourselves to "the Sinitic variety used by the majority of Taiwanese as their ancestral language," would we arrive at Taiwanese Hokkien.

    @Shannon from my experience, for immigrants "heritage language" refers just as, or more so, to the language spoken in the ancestral homeland, rather than the ancestral language itself. Thus, if the language spoken in the ancestral homeland should change, then the "heritage language" would also change. This is because, at the end of the day, language's function is communication, and immigrants learn "heritage language" not for the sake of learning it, but to be able to communicate with "people back home." Since people in the PRC increasingly speak Standard Mandarin today, while use of the other Sinitic languages is slowly decreasing, it is prudent to learn Standard Mandarin.

  15. Shannon said,

    June 13, 2016 @ 10:17 pm


    Yeah, I can see how there could be situations where the language of a diaspora or immigrant group changes to keep in touch with the linguistic changes in the homeland, and people thus update what they consider to be their "heritage language" but I can also imagine situations where if the ancestral language is gone from the homeland or only or mostly survives outside the homeland, the diaspora might no longer feels much in common with the homeland and/or may even see other members/places of the diasporic community as the ones carrying the heritage language, not the homeland.

    Also, "heritage language" as defined to mean language currently or most recently spoken (or perhaps dominant/prestigious) in the ancestral land, not ancestral language, could lead to some odd conceptions of heritage language such as the Irish diaspora's heritage language being English, or possibly some African and Asian diaspora's heritage language being a colonial European language.

  16. 艾力·黑膠 said,

    June 16, 2016 @ 3:59 am

    the lack of a Romanization system … that is widely accepted as standard or official. There are more than half a dozen Romanizations for Cantonese

    In my experience, there are effectively two Cantonese romanizations: Yale and jyutping, with usage of the former having severely dropped off in the past ~10 years (perhaps due to its ASCII-unfriendly nature?)

    the educational authorities in Hong Kong are doing a terrible disservice to students, not only in Hong Kong, but everywhere, when they fail to authorize any particular Romanization for Cantonese.

    It was my understanding that part of the reason for jyutping's relatively sudden surge also had to do with its having received the imprimatur of the LSHK.

RSS feed for comments on this post