Which is worse?

« previous post | next post »

A couple of years ago, I wrote a post with with the title "English or Mandarin as the World Language?" (5/2/14).  The purpose of that post was basically to call attention to Geoff Pullum's fine Lingua Franca article titled "There Was No Committee" (4/30/14).  It was all about English as a Medium of Instruction (EMI) in non-English-speaking countries, and included a glance at Mandarin as a possible alternative.

Now, in "The Awful Chinese Writing System" (Lingua Franca, 1/20/16), Pullum takes a more direct look at the question of whether the Chinese writing system is a bar to Mandarin's becoming a global language like English.  I will let the article speak for itself, but will say only that it is packed with convincing facts and compelling ideas.

The final, unequivocal paragraph is classic Pullum:

In consequence, this horror-show of a writing system, with its crippling memorization burden for students and malign impediment to progress in science and industry, is the focus of so much intellectual investment and cultural pride that getting rid of it is out of the question. Intolerable though it is, it will continue to be tolerated – leaving English, with a spelling system that positively stinks, smelling almost like a rose.

As for what he says about me, I will use a Mandarin idiomatic expression in response:  bù hǎoyìsi.  You don't need Chinese characters to say or write that.  In fact, if you focus on the literal meaning of the individual characters ("not good meaning / sense / gist" and so on and so forth), you'll probably just get confused about its many actual applications, for which see here, here, and here.


  1. @bnjmnbnjmn said,

    January 21, 2016 @ 8:24 pm

    Without the rest of Pullum's article, the quoted text is enjoyable ambiguous and potentially satire of English. On its own the paragraph never mentions Chinese but from our/my English language bias I assumed it was about Chinese. Pullum *couldn't* be criticizing English. The last line, though, provides a volta: it is not Chinese but English that is criticized! English's writing system is intolerable but will continue to be tolerated; English requires crippling memorization of spelling; people are attached to English's history and therefore unwilling to change it. Even the name, "The Awful Chinese Writing System" reeks of satire.

    Upon reading Pullum's article it becomes apparent that he means no satire—in context, the paragraph actually is unequivocal. It is more interesting, though, as a critique of English and my English bias, and a reminder that English isn't as superior as I tend to think.

  2. Apollo Wu said,

    January 21, 2016 @ 9:26 pm

    I agree completely with Pullum's article and your insight on the Chinese writing system. It is hard to envisage a complete abandonment of the Chinese script, but hopefully a digraphia usage will be created with the use of a enhanced (more accurate) Pinyin input method, which may evolve into an alphabetic script.

  3. Doctor Science said,

    January 21, 2016 @ 10:37 pm

    It hasn't been clear to me, from your posts and the discussion here, whether character amnesia is actually on the rise, or whether substantial literacy is much more widespread than in the past.

    Based on things I've learned here, I've formed the hypothesis that, for instance, the notorious difficulty of the imperial examinations was because most human brains can't retain the number of characters high-scoring applicants needed to read and (especially) write. A small percentage of the population can do it, of course, but for most applicants passing the test was a contest between cramming and the inevitable secondary character amnesia.

    Basically, I suppose that what makes today different is that much (most?) of the Chinese population is expected to be literate on a day-to-day basis: substantial literacy can't be left to a carefully-selected minority in whom it is constantly reinforced.

  4. Victor Mair said,

    January 22, 2016 @ 7:32 am

    @Doctor Science

    Character amnesia is a new phenomenon which refers to the loss of the ability to write large numbers of characters on the part of individuals who once knew how to write them. It is directly related to reliance on electronic devices for writing Chinese. This has all been spelled out in great detail in my various posts on this subject.

  5. leoboiko said,

    January 22, 2016 @ 8:42 am

    As usual, I find that this kind of discussion loses me when they get to part about "but why do they keep using them". The purely utilitarian POV has to resort to abstractions like "cultural pride" or "historical factors", which is basically the same as saying "because they do”. Aesthetics are left out of the discussion, or else are treated pejoratively as "attachment", "obsession", "fantasy", even "intoxication". But the interesting question is, how exactly can written symbols (of all things) be “intoxicating”—and why? To what extent are Chinese people themselves "intoxicated", and what role do characters play in their culture (e.g. like this)? Doesn't reading hanzi subjectively feel different than reading pinyin? Doesn't this feeling have to do with the structure of the writing system itself, with how it's processed (cf.)? Why do I feel that decorating a New Year's festival in pinyin, or tattooing a Chinese word in pinyin, just isn't the same thing? “Social, cultural and emotional factors”—aren’t these the most important factors of all?

    The reply could be that aesthetics are a poor reason to subject so many people to so much memorization. I happen to disagree, but at least non-utilitarian factors would be recognized in the argument.

    Re: amnesia: There may be a flip side to the increasing reliance on phonetic input methods. If on one hand people lose recall (they forget how to write characters from memory), on the other hand they may use a wider variety of characters, because now they only need recognition: the computer suggests characters that one wouldn't recall, but that one can recognize, so that one picks them where they wouldn't write them. I keep hearing that this is leading to obscure kanji popping up back to life in Japan. I'd love to see quantitative data on both effects (recall amnesia, and recognition expansion).

  6. Victor Mair said,

    January 22, 2016 @ 9:39 am


    The link within the parentheses here just references back to this post:

    "what role do characters play in their culture (e.g. like this)".

  7. leoboiko said,

    January 22, 2016 @ 9:43 am

    Oh, sorry, empty link. I meant to reference O'Neills essay on Xu Shen's rhetorical strategies: http://www.jstor.org/stable/10.7817/jameroriesoci.133.3.0413

  8. Jon Forrest said,

    January 22, 2016 @ 9:47 am

    I regrettably haven't followed the discussion about why pinyin hasn't become more popular. But, I've always wondered why the fact that Vietnamese, with a phonology that's just as "complicated" as Chinese, has been able to adopt a latin alphabet, albeit with lots of diacritic marks, as its official writing system.

    In discussing the Chinese writing system with many Chinese people, I've come to recognize that there are many non-linguistic factors that make most Chinese (and Japanese) people think that switching to a writing system without characters is the stupidest idea they've ever heard.

  9. languagehat said,

    January 22, 2016 @ 10:16 am

    The reply could be that aesthetics are a poor reason to subject so many people to so much memorization. I happen to disagree

    So would you be in favor of English switching to characters? If not, why not?

  10. Elessorn said,

    January 22, 2016 @ 10:23 am

    In consequence, this horror-show of a writing system, with its crippling memorization burden for students and malign impediment to progress in science and industry…

    …has resulted in the trifecta of chronic illiteracy, poverty, and backwardness familiar to even the most casual observer of East Asia today???

    1. There is no question that romanization would make Japanese and Chinese immensely easier for non-natives to learn.

    2. There is no question that romanization would be perfectly workable as a replacement script, and good reasons to think it would make literacy easier to acquire even for native speakers.

    3. At the same time, there is no question that the current script clearly poses no obstacle to the prosperity of those societies that continue to use it.

    4. And, finally, there is no question that even at its very best, romanization would entail some of the "catastrophic success" of the Turkish model, severing modern users from more than two-millennia's worth of written tradition.

    The first two reasons are clearly decisive for Professor Pullum. They may be decisive someday for the people of the Sinographosphere as well. But as long as reasons three and four hold, it seems to me that the current situation is not only likely but logical. Optimal? Of course not, and surely Professor Pullum is right that the writing system is a barrier to Chinese becoming a truly international language–as long as English is useful enough to be a competitor.

    But should the difficulty of L2 acquisition really be the controlling concern for native speakers? It might be too much to ask L2 learners and adepts to share the native perspective, but surely we can have the self-perspective to privilege it over our own.

  11. Ross Bender said,

    January 22, 2016 @ 10:55 am

    cf Mark Twain's "Die Schrecken der deutschen Sprache".


  12. leoboiko said,

    January 22, 2016 @ 11:00 am

    @languagehat: Interesting question! I find the entire idea a little bit imperialistic, to be frank; I feel like it's not my place to say how cultures should write. So the following is just idle speculative fiction; but if I could choose, and resources weren't a question, I'd favor universal digraphia.

    There would be some caveats. Chinese characters, specifically, are tied to the phonology, morphology and lexicon of the Chinese languages. They kinda work for Japanese or Korean in part because these languages borrowed so much vocabulary—indeed, the adoption of characters is how the vocabulary was borrowed in the first place. I don't feel like English would benefit from such a deep transformation; so I'd favor a new sort of morphogram, better suited for the common European heritage and Indo-European structures. Of course, Europe has no extant literary tradition or cultural practices involving morphograms, so it would have to be something brand-new; say, a creative school exploring the aesthetic possibilities of complex ways of expressing English (and French, etc.).

    There's the matter of calligraphy and typography, too; I remain unconvinced that Chinese characters and Latin letters can be mixed in aesthetically pleasing ways. So, for the new school, if we're going for a Chinese visual style, I'd want a Hangul-like system for the phonographic part. Alternatively, if we're keeping the Latin letters, I'd want the new morphograms to be designed from the same basic strokes, and in shapes such that could flow well within lines of Latin text.

    There's basically zero chance of this happening, but I'm not being facetious: I’d very much welcome such a development. Perhaps emoji could one day evolve to do at least a little bit of that :) (cf.)

  13. cameron said,

    January 22, 2016 @ 11:21 am

    @ Elessorn

    With regard to "the 'catastrophic success' of the Turkish model, severing modern users from more than two-millennia's worth of written tradition."

    Didn't the Chinese take that very step a century ago when they replaced the classical language with modern mandarin as the written standard?

  14. J. W. Brewer said,

    January 22, 2016 @ 11:43 am

    I don't see any particular reason to think that there would be a sudden boom of L2 Mandarin speakers outside its current home country/ies if only there were a simpler writing system. Hindi has a simple enough writing system, yet there has been no rush to adopt it as a convenient L2 even in its own region of the world (indeed, it has failed to become the default L2 even within its home country, as see the legendary riots some decades ago in Madras in favor of the right of upwardly-mobile L1 Tamil-speakers to have English as their L2 and not suffer career/educational barriers for not knowing Hindi). Indonesian/Malay is/are fully romanized, and without any of the complicated diacriticals that Vietnamese requires (and also as best as I can tell comparatively easy for "generic outsiders" to learn in terms of not having too many bits of phonology or syntax that are unusually baffling or complex), yet it/they have likewise not expanded much outside it/their home turf (maybe even contracted a bit compared to several centuries ago?). Even within the PRC it's not clear to me how much Mandarin fluency is growing in the regions that historically had other Sinitic topolects as L1's via "natural" processes of people finding it convenient to know versus the heavy hand of government imposing it. It's not like there's some basic principle that a language with a sufficient absolute number of L1 speakers is destined to become a "world language," such that some case-specific explanation (e.g. unusually difficult writing system) is necessary when it fails to do so.

  15. J. W. Brewer said,

    January 22, 2016 @ 12:22 pm

    I'm sure I've touted it before, but for those who haven't read it, Ostler's Empires of the Word is fascinating as a set of case studies as to why certain languages did prosper and become "world languages" (or the regional equivalent) and others didn't, with the most important takeaway being that there is lots of contingency and no single pattern. E.g. there are many cases where a conquered people assimilates to the conquerors' language but many other cases (not least in East Asia!) where the conquerors assimilate to the language of their new subjects. Even in a hypothetical future where by the end of this century the PRC or some successor regime thereto becomes more globally important (economically, militarily, or both) than the US or even all L1 Anglophone countries in the aggregate, there could be a historical parallel to e.g. the ancient Persian Empire which as it expanded used a pre-existing regional lingua franca (Aramaic) as its administrative language, so its own native language did not (at that point) have a geographic or demographic expansion parallel to the expansion of the power and influence of its L1 speakers.

  16. MikeA said,

    January 22, 2016 @ 12:25 pm

    For evidence of "obsession" with obscure glyphs that require memorization to understand at all, one need look no farther than the horde of UI designers who, not content with dominating and obfuscating mobile applications, have now laid siege to the desktop. ( 1/2 :-)

  17. Eidolon said,

    January 22, 2016 @ 1:31 pm

    From a strictly linguistic perspective, the question is actually not "why Chinese?" but "why English?" Pullum said it himself in the first article: there was no committee. There are certainly writing systems and languages that beat out English in efficiency & ease of learning, but no one ever seriously thought about replacing English with them.

    To this end, I'd actually quote Pullum in the final analysis:

    "People must make their language choices for themselves. If they are increasingly choosing to be educated in English, neither fans of linguistic diversity nor politicians taking pride in the national language have a right to overrule them."

    But I'd add that "fans of linguistic diversity" and "politicians taking pride in the national language" do have a say in the matter, and that the latter's say especially is liable to carry heavier weight, as it usually does, in society.

    And that is why China is still using characters.

  18. Schroduck said,

    January 22, 2016 @ 1:49 pm

    @Ross Bender. One thing that's interesting is that as languages go, German's spelling is pretty straightforward (it's almost totally phonetic, although with a couple of sounds that can be represented by two letters, and a few letters that have different pronunciations depending on context) and for a while it was a lingua franca in the worlds of science and maths. That didn't stop it getting dropped like a hot potato as soon as the English-speaking scientific community outgrew the German one. This in itself kind of proves Victor's point about there being "no committee" – at risk of running in circles, English became widely-spoken because… well, everyone was speaking it.

  19. Schroduck said,

    January 22, 2016 @ 2:06 pm

    (Oops, I mean Geoff Pullum's point)

  20. liuyao said,

    January 22, 2016 @ 2:25 pm

    What's actually remarkable is the fact that Chinese characters did manage to live on into the modern world, serving the vernacular language, if only Mandarin, very well. This debate of adopting a completely phonetic script has become so remote, even ridiculous, in the minds of most Chinese today.

    From my personal experience, seeing pinyin on book covers or merchandise is not helping at all. I always read the Chinese characters. This is "tested out" in the field, and it's not working. The obvious reason is, of course, the homophones, and from my experience with pinyin input, the most glaring example is shishi (my input gives 30+ different words, not counting shisi and sishi, which many Mandarin speakers have trouble distinguishing). They are commonly used words, not esoteric characters. One could of course cite Y. R. Chao's story "shi shi shi shi shi", but that's like citing biang without mentioning that it's a joke character, and most characters are not nearly as complicated.

  21. cameron said,

    January 22, 2016 @ 3:35 pm

    How much of the fear of abandoning the characters is due to the fear of laying bare China's true linguistic diversity? If people simply wrote phonetically then they could with relative ease write their local languages, as actually spoken in their local areas.

  22. The Other Mark said,

    January 22, 2016 @ 5:37 pm

    3. At the same time, there is no question that the current script clearly poses no obstacle to the prosperity of those societies that continue to use it.

    Just saying this does not make it true. I question it strongly.

    So do all those people who have spent a lot of time making it easier to enter ideographs on computers/phones etc which the West solved with the introduction of the typewriter and didn't even really have to invent for computers.

    Does anyone write computer code in ideographs? Or is the alphabetic system so superior that all the Asians have to learn a second way of writing to even start? How can that not impact on their economic development?

    That an obstacle has been worked around does not mean it isn't there.

  23. Riikka said,

    January 23, 2016 @ 1:29 am

    @Ross Bender – and to not forget the follow-ups like http://www.smart-words.org/humor-jokes/language-humor/english-spelling-reform.html

    Fainali, wi wud hev a lojikl, kohirnt speling in ius xrewawt xe Ingliy-spiking werld.

  24. michael farris said,

    January 23, 2016 @ 2:29 am


    The catastrophic success wasn't the switch from Arabic script to a Latin based alphabet. That was a huge success and nobody in Turkey seriously wants to revert back. Also pruning elements of Persian and Arabic grammar was a good idea.

    The catastrophic success was the language purification movement that grew out of the switch which invented thousands and thousands of new terms to replace a large portion of the vocabulary. Had it just created terms for new technology it wouldn't have been so bad but they forced usage of a lot of new (sometimes werid) words to replace vocabulary that most people knew.

    The Arabic

  25. michael farris said,

    January 23, 2016 @ 2:35 am

    "What's actually remarkable is the fact that Chinese characters did manage to live on into the modern world, serving the vernacular language, if only Mandarin, very well. This debate of adopting a completely phonetic script has become so remote, even ridiculous, in the minds of most Chinese today."

    Well part of the point is not about how Chinese perceive them, but that they comprise a major barrier to entry for learners, especially adult learners of Chinese and help guarantee that Mandarin will not become widely learned by non-Chinese.

    The not very good English spelling system is also a barrier to entry for many but not bad enough to prevent widespread learning by adults (even if they never become very proficient – especially in pronunciation).

  26. Lazar said,

    January 23, 2016 @ 2:43 am

    As Stalin would say, they are both worse!

  27. Matt_M said,

    January 23, 2016 @ 8:48 am


    "From my personal experience, seeing pinyin on book covers or merchandise is not helping at all. I always read the Chinese characters. This is "tested out" in the field, and it's not working."

    I often see this argument from native speakers of Mandarin, but I don't think it's very persuasive. Of course you find it easier to read characters — you've had far, far more practice reading characters than reading pinyin for almost all of your life. It doesn't prove anything about the superiority of characters as a writing system. I'm sure that speakers of Dungan (a language closely related to Mandarin, but written using the Cyrillic alphabet) could make a similar argument about how Cyrillic is obviously the best method of writing Chinese languages, since it's so much easier to read that way!
    As for homophones — if they're really such a problem, how do Mandarin speakers manage to communicate effectively with each other using just the spoken language?

  28. languagehat said,

    January 23, 2016 @ 9:16 am

    As for homophones — if they're really such a problem, how do Mandarin speakers manage to communicate effectively with each other using just the spoken language?

    I bring this up every time the subject comes up, and nobody ever seems to have thought about it, and there's never a good answer, but it never seems to affect the argument.

  29. leoboiko said,

    January 23, 2016 @ 9:59 am

    @Matt_M, @languagehat: You're right; any living, spoken language must be able to be expressed through phonography. But Liuyao is also right; for someone who can read morphograms, reading a phonographic transcription of the same text (n.b.: keep this in mind) feels like a chore. How come?

    Matt says it's lack of familiarity. I think this is going in the right direction, but isn't quite a satisfying answer. The Japanese learn to read in kana; how come they viscerally reject kana-only texts?

    My bet is in the word "viscerally". Again, morphography subjectively feels different than reading phonograms (cf. my link above for some evidence). It's this clash of feeling that makes people reject pure phonography. A similar rationale applies (to a lesser extent) to English spelling reform.

    This brings us to the homophone question. If the language is alive, i.e. able to be expressed through sounds, how come the universal observation that texts get confusing when you remove the characters? My answer is digraphia. As long as you have morphograms available, you don't have to care about homophones, so you use a lot more than you'd use in speech. How much more? I don't have data for Chinese, but written Japanese uses about twice as much Sinitic words than spoken (even more for formal styles); and Sino-Japanese, lacking tones and most sibilants, is even more homophonic than Literary Chinese.

    This is the point that Chao was criticizing with the shishishishi story; the use of characters lets one produce texts that would be near-unintelligible as speech. You can argue, like Unger, that this means that characters only solve a problem that characters created in the first place: excess homophones. Alternatively, you could argue that characters allow different modes of expression—and that these possibilities are actively explored by native writers. This is what I'm arguing, with examples, in my thesis.

    There's nothing mystical about this; writing everywhere does things that speech doesn't (cf. Nunberg, "Linguistics of Punctuation"; or Vachek and Halliday on written language). Think, for example, of concrete poetry, or Sir Terry Pratchett's skillful use of footnotes. It's unsurprising that East Asian writers would take advantage of systematic ability to distinguish homophones.

    To write truly fluent phonographic texts, all you have to do is to write them in pinyin in the first place; then you'd instinctively avoid excessive homophony (and you'd boost the prestige of the vernacular, to boot). I'm an ecumenic observer; I like both extremes and everything in between, and I hope more writers explore all kinds of scripts.

  30. Lazar said,

    January 23, 2016 @ 11:29 am

    I think the argument from homophony may be a bit more trenchant in Japanese, where both kana and romaji fail to mark pitch accent – thus obscuring many distinctions that do exist in speech. Though that should be easy enough to remedy with an extra diacritic or two.

  31. liuyao said,

    January 23, 2016 @ 11:38 am

    I confess that I have not thought about this question (that homophones work fine in conversation, but not when written). I'm no linguist, as you probably can tell. I'm not able to elaborate the points more fully, but here they are.

    First of all, it doesn't work as perfectly in conversation as you might think (I recall having trouble expressing deer "lu", which doesn't have a two-character compound in Mandarin, and I had to say "the animal lu"), but I suppose that's true in English too, to a lesser degree (deer/dear). If you are all for different spellings in English, then magnify the problem tenfold. Pinyin does not offer different spellings, except in special cases such as Shanxi vs Shaanxi.

    Secondly — I'm not sure if that's a sound analogy — why do you have word breaks in English when they don't exist in speech? Why not have breaks right before the stressed syllables? (Mind you that you having been reading English with conventional word breaks all your life.) Clearly we can parse speech faster than text (for the same "easy" content), and as @leoboiko says, text could offer a greater dimension of expressiveness: literature, scientific writings, etc.

    What the Chinese script poses as a major obstacle for adult learners, it compensates somewhat by its easier grammar. I'm not saying everyone should learn Chinese to communicate to each other when dealing in business or scientific research. It's very unlikely, but people in early 20th century would say that the Chinese script is unlikely to adapt to the modern world. What happened was indeed remarkable, if you would entertain to look it this way.

  32. leoboiko said,

    January 23, 2016 @ 11:52 am

    @Lazar: I don't think the lack of accent makes a lot of difference–after all, you guys don't mark it either (fínance/finánce), and besides, there are pitchless Japanese dialects.

    Rather, as I've said above, I think the difference is in wording–something around 20% kango (Sinitic) for speech, and double that for writing. If you want a kanji–heavy text to be phonetic, you should reword it to get closer to speech (what Matsunaga calls "iikae"). E.g. " shiyō" has half a dozen homophones, but "tsukau" doesn't.

  33. Elessorn said,

    January 23, 2016 @ 12:34 pm

    @The Other Mark

    Just saying this does not make it true. I question it strongly.

    Of course I don't expect anyone to take what I say on faith. On the other hand, the standard of living in Japan, Taiwan, and Hong Kong is hardly a well-kept secret. And if the rise of China over the last two decades is not proof enough that the character script "clearly poses no obstacle to the prosperity of those societies that continue to use it," I can't imagine that anything will ever be proof enough.

    Does anyone write computer code in ideographs? Or is the alphabetic system so superior that all the Asians have to learn a second way of writing to even start? How can that not impact on their economic development?

    "How can [it] not impact?" It may be understandable that users of alphabetic scripts would instinctively think so, but why would we want to stick to what feels correct in the face of overwhelming empirical evidence to the contrary? Does anyone walk through Tokyo or Beijing with the sense that here before them is a people held back from the fruits of modern progress? One doesn't have to wonder about these things in the abstract. The last hundred years have provided us with a massive natural experiment to answer precisely this question: is Sinography compatible with the technological and economic requirements of modern life as known in the richest countries of the West? I continue to think the answer is pretty obvious.

  34. Victor Mair said,

    January 23, 2016 @ 1:04 pm


    Think of how much science, technology, business, commerce, education, etc. in China, Japan, Korea, Hong Kong, Taiwan, and Singapore is done in English.


    Thanks for your good comments. I'm also grateful to everyone else who has contributed to the excellent discussion on this post.

    You wrote: "…for someone who can read morphograms, reading a phonographic transcription of the same text (n.b.: keep this in mind) feels like a chore."

    Well, yes and no.

    If a text is written in good baihua ("vernacular"), not banwenbanbai (semi-vernacular semi-literary), I can read Pinyin as easily and quickly as I can read the same text written in characters, actually faster and more easily. This is true even when the tones are not marked; in fact, having the tones marked slows me down. The same was true of my wife, Li-ching Chang, who wrote her childhood memoirs in Pinyin:

    "New book in Pinyin" (9/13/10)

    "Li-ching Chang, 1936-2010" (7/20/10)

    Li-ching worked together with me on editing and publishing "Xin Tang: a journal of romanized Mandarin" during the 80s.

    Xin Tang included articles on all manner of subjects, all without characters.

    Before long, we will sponsor an international literary contest in memory of Li-ching for writing in Pinyin (fiction, essays, poetry) with very substantial prizes. We will stress that the texts should be in good baihua, i.e., that they can be understood when read aloud.

    To close this comment, I'll just mention an example of my ability to read off Pinyin texts effortlessly (i.e., not a chore; as easy as reading English). Last semester, in my "Language, Script, and Society in China" class, I opened an issue of Xin Tang that I hadn't looked at for about 25 years. The page I turned to happened to have a story about Afanti (Effendi [–> Nasreddin]). Without any hesitation whatsoever and with natural rhythm and intonation, I read the story aloud, and the thirty students in the room (about half of whom were from China) gasped because it was all in Pinyin, but they understood everything I said.

    When I show Pinyin texts to Chinese who are literate in characters but who have had no previous exposure to Romanized Chinese, even if they are completely literate in English, their eyes usually glaze over and they insist that they can't make any sense of the Pinyin. But if I work together with them and encourage them to sound out the Romanization as if it were English, before long they are starting to make sense of what they are pronouncing.

    Having been at this business of working with publishing in Pinyin for more than three decades, I'm certain that the main barriers for Chinese "who can read morphograms" are mainly three:

    1. psychological

    2. cultural

    3. lack of habituation

    But you're right, leoboiko, if you want to make Romanized Sinitic languages, Japanese, and Korean work, the more your texts resemble speech, the more successful you'll be at it.

  35. Elessorn said,

    January 23, 2016 @ 1:56 pm

    To step back a bit, I think it's fair to say these arguments (sometimes) have a repetitive quality to them. I submit that it's because two separate theses are being advanced which seem to contradict each other:

    Thesis 1: By any isolated measure of efficiency, alphabetic writing has an obvious, empirical, probably incontrovertible edge over any form of Sinography.

    Thesis 2: By any empirical measure, the societies that use characters seem at little fundamental disadvantage against the countries of the alphabetic West, whether in literacy, economic development, technological advancement, etc.

    But if we think about it, these two theses are in fact both perfectly correct and perfectly compatible, because:

    1. A script can be BOTH monstrously inefficient and AND pose no obstacle whatsoever to its L1 users.
    2. A script can be exquisitely BOTH awful for L2 learners and AND pose no obstacle whatsoever to its L1 users.

    "No obstacle" of course hardly means "no effect", merely "no impediment." And with no penalty to staying put and clear costs to moving, why should we expect any shift? I for one always enjoy Professor Mair's posts about Chinese digraphia, but we should be clear: these are no more the signs of an "orthographic dilemma" than the seeping anglophony reported of contemporary French and German is a sign of cultural decline.

  36. J.Wapp. Brewer said,

    January 23, 2016 @ 2:26 pm

    I am fascinated by the claim that Prof Mair can read better/faster in pinyin if the tones are not marked, because I would have thought clear and consistent tone marking would significantly reduce (although not eliminate) the homophony problem that characters avoid. Indeed, one problem with these comparisons is whether one is comparing hanzi to some actually-existing/necessarily-imperfect alternative (like hanyu pinyin) or some theoretical ideal phonetic/phonemic script for the particular language. I would have thought a good desideratum for the latter in a tonal language would have been a way of denoting tone that is superior (both clearer for the reader and harder to omit for the writer) than the diacriticals used with hanyin pinyin and most other approaches to rendering Mandarin in the latin alphabet. But maybe I'm wrong. Surely Gwayou Romatzh's approach to clearer tone-disambiguation did not find much favor in practice. But maybe the more indigenous system used in the Brahmic script in which Thai is standardly written works better? Or maybe it doesn't. For all I know maybe people writing Thai in a hurry omit the bits that indicate tone and readers muddle through anyway.

  37. Elessorn said,

    January 23, 2016 @ 2:42 pm

    @Victor Mair

    Think of how much science, technology, business, commerce, education, etc. in China, Japan, Korea, Hong Kong, Taiwan, and Singapore is done in English.

    This is nothing new to you, but consider: Singapore and Hong Kong are former British colonies. The infrastructure of global science, capitalism, and politics was built by (and of course at the convenience of) Western Europe in an age when English was already well on its way to becoming the new Western French. The regnant presence of English in such professional fields in 21st-century East Asia thus seems no more surprising or normatively significant than the similar Chinese dominance that obtained there 200 years ago. And while I can't speak to the domestic usage of English by the Chinese elite, Japan does far, far less business in English than the alphabetically-advantaged countries of Western Europe.

    Of course pinyin would be absolutely serviceable as a script. Why wouldn't it be, after all? But this has nothing to do with the lack of an empirical basis for claims made about the necessity of a switch to alphabetization or digraphia. (Which is something quite different from arguments for the advantages such a switch might bring.) It especially frustrates that the column in question is by Professor Pullum, who is so often a beacon of calm sanity in a sea of dependably bad reporting on linguistics in the media. Biang, the extremes of the periodic table, and gross dictionary counts as a shorthand proxy for the lived nature of the script? And all these without any examination the useability of the script as actually observed? I cannot bring myself to believe such an evidentiary basis would pass muster for him on an issue closer to home.

  38. michael farris said,

    January 23, 2016 @ 5:15 pm

    "Or maybe it doesn't. For all I know maybe people writing Thai in a hurry omit the bits that indicate tone and readers muddle through anyway."

    They don't. neither do people writing Vietnamese (when diacritics aren't available on electronic devices they leave off tone but never when writing by hand).

  39. GH said,

    January 23, 2016 @ 6:51 pm


    The problem with your argument is that you're trying to use the economic performance of China and other countries in the region to prove that they do not face any disadvantages of the sort. This is logically unsound. All countries have advantages and disadvantages to economic development, so it could for example be that the handicap of the writing system is outweighed by some other factor. And disadvantages can be overcome, so even if the writing system does "pose an obstacle to the prosperity of those societies", that does not mean they are doomed to poverty. Still, it remains a disadvantage nonetheless.

    Perhaps the writing system shaves – let's say – a percentage point off annual GDP growth. That makes a big difference in the long term, but you wouldn't be able to tell from the kind of evidence you point to.

  40. Doctor Science said,

    January 23, 2016 @ 11:51 pm


    Sorry to have taken so much time to get back to you, I've been playing "blizzard".

    I am non-plussed when you say "Character amnesia is a new phenomenon which refers to the loss of the ability to write large numbers of characters on the part of individuals who once knew how to write them." [emphasis mine]

    For instance, in your 2012 post "Japanese survey on forgetting how to write kanji", the included graph certainly shows that kanji amnesia is getting *worse*, but it also shows that in 2001 a majority of people in their 30s and 40s were already reporting kanji amnesia, though cell-phone use was still quite a new phenomenon.

    Similarly, the anecdotes in David Moser's 20-year-old article Why Chinese Is So Damn Hard, especially:

    I have actually kept a list of characters that I have observed Chinese people forget how to write. (A sick, obsessive activity, I know.) I have seen highly literate Chinese people forget how to write certain characters in common words like "tin can", "knee", "screwdriver", "snap" (as in "to snap one's fingers"), "elbow", "ginger", "cushion", "firecracker", and so on. And when I say "forget", I mean that they often cannot even put the first stroke down on the paper.

    I could have sworn I saw (on LL or linked to it) a graph for Chinese or Japanese reading vocabulary, the equivalent to this chart for native speakers of English, showing a radical dropoff in effective literacy after schooling, but I can't find it via Google and I may have made it up in my mind.

    All of these points convinced me that character amnesia has probably *always* been a problem with Chinese, that the writing system has always imposed a very significant cognitive load. What's changed is a) cell phones and other technology make it easier for people to shed a lot of that load while remaining functionally literate, and b) a much larger percent of the population *has* to be functionally literate than in the past.

  41. Victor Mair said,

    January 24, 2016 @ 7:41 am

    @Doctor Science

    No need to be nonplussed.

    There has always been a problem with character retention, as Unger and others have shown.

    This new phenomenon of character amnesia, as I've written about extensively on Language Log, is the result of reliance on electronic devices to write the characters, leading to the loss of the ability to write them by hand. This is a dramatic development that people started to notice about fifteen years ago, as I have documented in my posts on the subject. It's different from the sort of forgetting how to write characters one once knew in the pre-electronic gadget age. The electronic devices greatly hasten handwritten character attrition. This is not something I'm making up. Chinese complain about it all the time, both for themselves and for the public at large.

  42. liuyao said,

    January 24, 2016 @ 11:41 am

    @Victor Mair

    I find it remarkable that you'd be able to read pinyin out loud without hesitations or backtracking, with proper cadence. That your students would be able to understand it is not surprising, given that your pronunciation is quite good and that's how any class or lecture in Chinese is conducted, even thrown in some literary Chinese.

    The typical native Chinese speaker's response, as I was giving here and corroborated by your observations, is not to be written off lightly though. Children's books are written in pinyin and characters, and I wonder if the teachers or parents force them to read the characters and only consult the pinyin when they have trouble, or it was a voluntary choice on the part of the children. (I imagine it's probably easier to read than English, being purely phonetic.)

    I have a question for you. When you and your wife wrote the book, did you find yourselves having to avoid certain words?

    You may have discussed this elsewhere, but would it be a problem to search for keywords in a pinyin text? Is that an issue for Korean or Vietnamese?

  43. Elessorn said,

    January 24, 2016 @ 11:57 am


    The problem with your argument is that you're trying to use the economic performance of China and other countries in the region to prove that they do not face any disadvantages of the sort. This is logically unsound. All countries have advantages and disadvantages to economic development, so it could for example be that the handicap of the writing system is outweighed by some other factor. And disadvantages can be overcome, so even if the writing system does "pose an obstacle to the prosperity of those societies", that does not mean they are doomed to poverty. Still, it remains a disadvantage nonetheless.

    Thank you for your comments. Your points about the complexity that necessarily underlies any story of successful economic development are well taken. And had I been claiming a fine-grained expertise into the economic effects of writing systems, your criticism would surely be apt. Yet this is hardly the case. The comments above will show that I have been repeatedly advancing the following claim:

    Many difficulties stemming from the non-alphabetic nature of sinography have been alleged, but no demonstration has been made of these difficulties actually disadvantaging the cultural, economic, or political lives of the script's users.

    You say "…even if the writing system does "pose an obstacle to the prosperity of those societies", that does not mean they are doomed to poverty. Still, it remains a disadvantage nonetheless." Yet where does this conviction come from? It is well enough to appreciate the merits of an alphabetic script. I certainly do. But can you actually provide a basis for your assumption that those particular merits confer a decisive comparative advantage in the cultural, economic, and political lives of phonetic-script using populations? Consider that the issue is not whether phonetic scripts are more efficient in the abstract (clearly they are), but whether this efficiency gap is an actual problem. It might seem like it obviously should be, but it also seems perfectly reasonable to most people that Eskimos have 100 words for snow. It might seem incredulous that printing was invented first by the users of the "awful Chinese writing system," but that is exactly what happened. Intuitions are not reliable.

    Recall Professor Pullum's actual conclusion:
    …this horror-show of a writing system, with its crippling memorization burden for students and malign impediment to progress in science and industry…

    I think it is perfectly reasonable to respond to such sweeping claims with simple reference to the quite uncontroversial level of cultural and material attainment in Taiwan, Hong Kong, Japan, and Mainland China. "Malign impediment to progress in science and industry"? It leaves me literally at a loss for words.

    Your own case for an undetectable drag on progress is far more modest, but given the sheer level of the empirically-observable quality of life among these peoples, I think the claim is counter-intuitive at best, and at the very least imposes a high burden of proof.

  44. Victor Mair said,

    January 24, 2016 @ 12:43 pm


    Please read my latest post:

    "Englishes in action in the Sinosphere" (1/24/16)

    It is directly relevant to the questions you have raised in your comments on this post.

  45. Victor Mair said,

    January 24, 2016 @ 12:57 pm


    "When you and your wife wrote the book, did you find yourselves having to avoid certain words?"

    Not in particular. We always had one rule of thumb, which I mentioned above in this comment, namely, that what we wrote should be intelligible when read aloud. Indeed, we would always read out anything we had written to each other to make sure that this was the case. Our test was similar to Y. R. Chao's concept of "Sayable Chinese".

    "You may have discussed this elsewhere, but would it be a problem to search for keywords in a pinyin text? Is that an issue for Korean or Vietnamese?"

    If the Pinyin is written according to the orthographic rules, that shouldn't prove to be any more of a problem than searching for key words in any other alphabetical language. As for Korean and Vietnamese, I hope that specialists in those languages will comment.

  46. Steve Harrell said,

    January 24, 2016 @ 3:00 pm

    I'm a latecomer to this conversation, namely I haven't joined until the second day. But it seems to me that there is no proof that Chinese characters pose much of an obstacle for L1 learners, or else the obstacle is easily overcome by hours and hours of practice embedded in the Chinese education system.

    And for L2 learners, we don't need any proof that the script is an obstacle, because I've only ever met one person who insisted it wasn't. And his L1 was Tigrania–I will let you real linguists figure out whether that has any significance or not. But I'll insert my own 40+ year experience anyway. For me, Chinese was originally L4 (after English, Spanish, and German), but now it's L2 in the sense that it's my second best spoken language. But I still read it about the speed I read French, which I studied for all of two months in 1966. At the same time, I sometimes have trouble understanding texts in Pinyin. This may have to do with the fact that I'm a "sub vocalizer" when I read any language, so I absorb the sound faster than I absorb the meaning. So I can read aloud even in Czech or Turkish without any trouble (though of course I butcher the pronunciation) without understanding a bit of what it means. But I can't do that in Chinese characters. My brain has to figure out both the sound and the meaning before I go on to the next character, and that takes time.

  47. Victor Mair said,

    January 24, 2016 @ 4:40 pm

    @Steve Harrell

    "…the obstacle is easily overcome by hours and hours of practice embedded in the Chinese education system."

    That's for sure, except that it's not "easily overcome". Even for L1 students, it is — as you say — "hours and hours of practice embedded in the Chinese education system"; sheer drudgery, as we have seen over and over again in Language Log posts and comments.

  48. languagehat said,

    January 24, 2016 @ 9:12 pm

    the obstacle is easily overcome by hours and hours of practice embedded in the Chinese education system.

    This is such a blatant oxymoron I'm amazed you could type it with a straight face. Do you seriously think Chinese students would not learn their native language far more quickly using pinyin, or any other alphabetic system?

  49. Steve Harrell said,

    January 24, 2016 @ 10:26 pm

    @Victor Mair, @languagehat, Of course I was being a little facetious. But only a little. What I meant is that there is a readily available solution to the problem of characters; it's easy to solve; all it takes is hours and hours of memorization. Would the Tibet issue or the Taiwan Issue or the Israeli-Palestinian conflict or money in American politics or even climate change were that easy to solve. Whenever I bring up the question of the difficulty of characters with Chinese friends, colleagues, students, they shrug and say yeah, I guess we spend a lot of time learning characters. But English is *really hard.*
    One way or the other, it's not my business to tell Chinese what they should do about writing their language. They write it a hell of a lot better than I do.

  50. Elessorn said,

    January 25, 2016 @ 2:17 am


    To take the question seriously, from my experience in Japan, it's not actually the case that learning the characters is entirely a matter of simple rote memorization at school. A lot of ambient learning goes on precisely at the ages when the mind is most plastic and receptive. A lot of school instruction is about how to write "correctly", and about understanding the principles behind characters, or is spent on drilling characters which are not actually that frequent. (And of course all schoolwork, qua text, is also necessarily also character practice.) The point of commonality here is that a student can perform poorly at all of these and still become functionally literate. They fail to wield their script with confident mastery in the way that many even in say, the US, never learn to spell or write with confidence. The burden feels roughly, though of course not exactly, equivalent to our spelling homework or grammar homework. That a student can be non-studious, never learn to write a lot of characters in the orthodox way, never develop a nice-looking hand, and remain a little uncertain about a lot of low-frequency characters, and yet still be able to get a job, read manga and sports tabloids, and fall asleep browsing on a smart phone every night, argues very strongly that the burden is not of the character imagined.

    Of course it may be argued that the burden of spelling practice books and the like are a big waste of time for the children of the anglosphere as well, and it may be true that a solid time savings could be achieved if all children benefitted from scripts as awesome as Hangul. But this is another debate entirely.

    This is why I continually stress observing the lived reality of the systems instead: it stands to reason that in any country a lot of people are going to be very poor students. If sinography was the awfulness argued for, I think you would have to expect a noticeable ceiling on literacy. Instead you see that while many people never master math or history or science, pretty much everyone can read. That should be telling.

  51. leoboiko said,

    January 25, 2016 @ 8:02 am

    Where I said "my answer is digraphia", read "diglossia". My mistakes insist on hiding from me until I click the "submit" button.

    Professor Mair, thanks for the kind words; it counts a lot. I fully believe that people can read pinyin fluently (and that digraphia will be great for China). My point is just that reading hanzi is a different aesthetic experience; and that, if a text was written relying on hanzi, then a blind pinyin transcription will feel clumsy (depending on how "reliant" are we talking about; in the case of Chinese this will likely mean percentage of Literary Sinitic and syllabic conciseness, while in Japanese it will be the percentage of kanji as well as of Sinitic words).

    Regarding the discussion on Chinese characters being an obstacle… Of course they take longer to learn, even for L1 readers; it would be hard to dispute that. The question is whether this impact matters. English spelling is, from my point of view, completely bonkers, and it clearly takes longer for schoolchildren to learn than, say, Italian or Finnish. Does this make the US at a significant disadvantage from Italy or Finland? As Elessorn said, I think the burden of proof lies on the person making the claim.

    For example, Hannas has suggested that Chinese characters set back East Asia on science. I live in Brazil, a country with a pretty regular phonographic writing system; while Japanese writing is, by common agreement, the most complex system currently in use in the world. Japan has 24 Nobel laureates; China (second most complex?) has 5; we have 1. Unger has suggested that kanji might be involved in the high Japanese rates of student suicide. But South Korea, whose writing system is a thing of beauty, has a lot more suicides than even Japan. How much of an impact does the writing system have, on the big scheme of things?

    Admittedly, I'm not a social scientist, and I'm not even a schoolteacher, so I can only argue anecdotes from my short time in the trenches as a volunteer. But I see the wonderful reader's culture that Japan have, the marvelously thriving publishing industry, the bookstore chains and vending machines; and then I consider the dismal literacy rates of my own country; and I can't bring myself to think that the writing system is a big impediment to literacy. I think the military police kicking your shack open and punching the hell out of your father because "if you don't have the papers to live here, then you're criminals" is an impediment to literacy; how disposed would you be, the day after, to listen to me drone on about spelling? I think being afraid of going to school due to the very real risk of being raped is an impediment to literacy. I think slowly falling behind for years without anyone noticing because classes are 70+ students and the teacher can't possibly know what's happening with each one, is an impediment to literacy. I think having entire subjects not being taught altogether because the inflation is bad and no one wants to be a teacher for $450/month is an impediment to literacy. I think not having proper libraries with interesting things to read, while books are priced as luxuries, is (as Krashen has argued insistently and convincingly) an impediment to literacy. I think being routinely and systemically treated by your entire society as if you're destined from birth to be either a truck driver or a drug dealer, internalizing from a young age that the educational system just isn't for you, is an impediment to literacy. All of these are real examples; either kids I've met, or that I heard about from my colleagues.

    You want people to read? Give them food, shelter, healthcare, books, and dignity. Then they'll learn Mayan if you teach them. I'm convinced that all of these things are orders of magnitude more important than the writing system.

  52. languagehat said,

    January 25, 2016 @ 8:29 am

    Very well said, leoboiko, and I'm in complete agreement with your last paragraph. While there's nothing wrong with counting the number of angels on a linguistic pin (that is, after all, what my own blog is all about), it's important to remember that there are more important things.

  53. liuyao said,

    January 25, 2016 @ 1:24 pm

    I'd like to attempt the impossible…

    "Do you seriously think Chinese students would not learn their native language far more quickly using pinyin, or any other alphabetic system?"

    I think there's a mismatch between what the Chinese/Japanese and the rest of the world regard as language, and what they think constitutes as "knowing one's language." If you mean to be able to record what one speaks, then all first graders (in Mandarin speaking regions) can do that, but they haven't even begun to learn Chinese. I hope that doesn't sound rhetorical. The word 文 has a slew of other meanings than simply characters or a piece of writing, that to be "literate" (colloquially 有文化) is more than being able to read, and is almost tantamount to being "educated" in all aspects, from being moral (traditional) to scientific literacy (modern).

    Everyone is alluding to the awful spelling system in English, though not surprisingly that's not an issue to Chinese speakers (Maybe Prof Mair could comment if his students can spell better than average Americans, if not Indians; certainly many have good handwriting). Instead, what the Chinese students find most difficult is the irregular past tense, and the use of particles in phrasal verbs. Yet I imagine others would think English as much easier in those aspects. To bring the discussion a bit closer to home, I wonder if there's any study in technical (medical, biological, etc.) terminology in Chinese compared to European languages (predominantly in Latin or Latin based words), in terms of word acquisition and/or retention. As biased as I am, I think 肺结核 "speaks" more than feijiehe or tuberculosis. Could it be that the initial "obstacle" pays off later in technical terminology or even SAT-type vocabulary?

    (I'm not trying to prove anything; in fact proofs only exist in mathematics, in a priori knowledge. I wish we could all use "evidence" or "arguments" instead of proofs.)

    This Euro-centric proselytizing reminds me of political systems, but I found another analogy. If the biologist found a species that is not based in DNA, he would treasure it, study it, instead of killing it or converting it (if he could); he may be able to argue which system is superior though. The existence of Chinese/Japanese stands as a testimony that there's another possibility to alphabetic system and phonetic spelling, and that in itself should be treasured.

    I'd like to share a dialogue (more in the style of Zhuangzi than Socrates perhaps), written in the 1920s by someone named 沈有乾, a little-known logician. I hope someone could translate it into English.







  54. Victor Mair said,

    January 25, 2016 @ 5:11 pm


    Several things.

    First of all, and most importantly, I don't think that people are proselytizing. Rather, I think that they are trying their best to describe what the situation is, how things are developing quite naturally within the Sinosphere.

    Second, do you know of, or can you imagine, a species not based on DNA? Similarly, as we've been seeing lately, Sinitic languages are processed in the brain in ways that are quite similar to non-Sinitic languages. I will be making another post about that within a few days.

    Third, as I ended a recent post about "Kongish, ch.2", one doesn't have to make a choice between characters and no characters (i.e., alphabet only). There's something called "digraphia", and that is what's emerging, without anyone actually promoting it. I will soon make another post on that subject as well.

    Finally, I think that your English and Chinese are both good enough for you to translate that dialog by Shen Youqian 沈有乾 yourself.

  55. Doctor Science said,

    January 25, 2016 @ 7:51 pm


    Thinking about the fact that we seemed to be talking past each other, I realized that you're right, the current wave of "character amnesia" is truly novel, different in kind as well as in scope, compared to the character amnesia seen in previous decades.

    The difference is that the modern form isn't actually "character amnesia", it's "calligraphy amnesia" only. Nowadays, people who forget the brushstrokes for a particular Chinese character can still write it, via the pinyin-to-hanzi system on their phones and computers. That's still writing Hanzi, at least if my use of a keyboard instead of a quill is still writing English.

    I think it's probably misleading, then, to talk about Chinese (or Japanese) people "losing the ability to write characters". They can still write them just as well, if your criterion for "well" is how easy the product is to read. Indeed, they may write them *better*: more consistently, and with all the fiddly bits clear and precise, even in very complex characters.

    This new writing technique actually undercuts Pullum's argument, because pinyin2Hanzi takes out a lot of the Awfulness of the Chinese writing system, while leaving the good parts: historical continuity, cross-language utility, and extreme concision on e.g. Twitter.

  56. Victor Mair said,

    January 25, 2016 @ 8:39 pm

    @Doctor Science

    It is not just calligraphy amnesia, since what Chinese complain about is not simply forgetting how to write characters with a brush, but with pencil and pen too. There are times when one wants to jot something down in characters but one doesn't have an electronic device handy. Those are the moments when Chinese feel constrained by forgetting how to write characters by hand.

  57. Doctor Science said,

    January 25, 2016 @ 9:35 pm


    By "calligraphy" I was including all types of writing by hand, whether with brush, pencil, or stick scratching in the dirt. "Handwriting" would have been clearer.

    I can certainly see that losing handwriting is a problem, e.g. when you want to scrawl "Out of order, biohazard, do not release the tiger" on a door, or "v. attractive person, 867-5309" on your hand. But it's not losing *writing* or characters, just one technique for producing them.

    I don't know, but it may be that thinking of pinyin2hanzi as "what we mean by writing" could produce better pedagogical methods for hanzi handwriting, and for remembering them in general. Focusing less on number of strokes, for instance, and more on visual gestalt or features.

    Hanzi is (are?) a uniquely powerful writing system, in certain ways: the way they can be used across mutually-unintelligible languages, for one, and their extreme concision, for another. Pullum could end up being 180° wrong: hanzi, written via pinyin2hanzi, could help Mandarin become more dominant, especially in situations where concision and cross-language communication are important.

  58. Victor Mair said,

    January 26, 2016 @ 3:06 pm

    @Doctor Science

    "Handwriting". That's what I was talking about all along — for all the years I've been writing about character amnesia.

    Grocery lists (e.g., dumpling ingredients), a carpenter's or other workman's notes on materials they're working with, a scholar's / student's marginalia (my wife felt very uncomfortable without writing her ideas down on the pages of whatever she was reading), comments on student papers, writing on black/green/white boards (we still use them here at Penn; it's very embarrassing if you forget how to write a character in front of 40 students), etc. — the instances when one is called upon to jot something down without resorting to an electronic device are beyond counting.

    Chinese friends, colleagues, and acquaintances complain to me all the time about forgetting how to write things by hand because of their reliance on electronic media.

    Yes, "Pullum could end up being 180° wrong", but he also could end up being completely right. We'll just have to wait and see. Meanwhile, read my latest post to see which way we're headed right now — including a look at "the way they [i.e., Chinese characters] can be used across mutually-unintelligible languages", which is untrue, by the way (people who don't know spoken Cantonese cannot read full-blown written Cantonese — a phenomenon we've discussed many times on Language Log).

  59. Tom V said,

    January 26, 2016 @ 9:01 pm

    Bù hǎoyìsi brings back memories of my childhood. In the early fifties I absorbed from my parents an operational definition of it as "something that upsets my stomach" (figuratively, of course). Mandarin was their fallback language for discussing things they didn't want us kids to understand.
    It's not an expression I use actively, but it comes to mind frequently when I read the morning paper.

  60. Dan Gilles said,

    January 27, 2016 @ 2:15 am

    Speaking as a second language learner of Chinese, I never bothered to learn the writing of the characters. I agree that starting from Pinyin and learning to read first is optimal, and could almost certainly be employed to decrease the amount of drudgery for first-language learners of acquiring the written language. However, I personally would hate to see people not learning characters at all, since I find them beautiful and fascinating; they were and continue to be the largest source of my interest in learning Chinese in the first place.

    Pullum says that Chinese has a "horror-show of a writing system". I could not disagree more strongly. Compare Chinese to Esperanto. The latter is easy to learn and would be a great world language (though as another post of Pullum's makes clear, linguae francae are chosen by political and not linguistic considerations). But as lovers of language, we must consider more than mere utility! While it would be wonderful if everyone in the world learned Esperanto, it would be tragic if we all stopped speaking other languages. Variety is the spice of life, and complexity can be beautiful. I'm sure Pullum has studied and appreciated Shakespearean or even Chaucerian English, despite its difficulty. I would suggest that he give Chinese a try. It amazes me that the elegant and beautiful Chinese written system is actually used as an everyday medium of communication. Every Chinese character has a history. For many of us, learning their histories becomes a wonderful addiction.

    There are lots of methods to make learning to read Chinese easier. Using an etymology dictionary to learn the etymologies of characters instead of just memorizing, and then using spaced repetition software to optimize the remembering process, decreases the cognitive workload hugely. And if one does want to learn to write, Remembering Simplified Hanzi is easy and works. But in any case, both first and second language learners should always learn Pinyin and the spoken language first, and then begin to pick up character recognition and perhaps writing where necessary.

    In sum, Chinese characters are hard, but hard things are sometimes worth mastering. Learning to write Chinese characters by rote is an inefficient method which makes literacy much harder than it needs to be. But learning to read them at all is not as difficult, and in my opinion is extremely worthwhile.

  61. languagehat said,

    January 27, 2016 @ 8:43 am

    But nobody's saying characters should be wiped from the face of the earth and all memory of them eliminated. (This reminds me of other debates in which one side takes an extreme, reductio-ad-absurdum version of the other side's position and reacts as if that were what their opponents were advocating, making discussions frustrating and intractable; I won't mention any specific examples because I don't want to drag politics into this site.) Of course characters are beautiful and lovable — I love them myself and am glad I had the opportunity to get acquainted with at least a fair quantity of them, and learned to write them correctly. The point is not that they are bad and should be dumped into the memory bin, the point is that they take a ridiculously long time to master and are thus an absurd medium for mass literacy. Just because one has invested that time and effort oneself does not mean future generations should be condemned to the same expenditure, although of course those who are attracted to them will learn them voluntarily, just as I have studied Georgian.

  62. leoboiko said,

    January 27, 2016 @ 8:55 am

    @languagehat: The question is to what extent would the culture of characters survive if they weren't part of general education. A large amount of the most creative kanji play I've found for my dissertation is from comics and videogames (indeed, I had to make an effort to stop picking examples from these sources, because they're that numerous). Here's one example from the manga comics Assassination Classroom: 今の殺[ころ] (where the brackets denote glosses). This reads Ima-no-Koro, now-GEN-time, a regular expression for “about now; at this time”. However, rather than the regular kanji 頃 Koro "time", it uses 殺 Koro- "assassination". In this way, without changing the sound, it implies something like "today's assassination".

    I'd like to draw attention to the fact that this kind of kanji-play is extremely frequent in shōnen-age manga, i.e. boy's comics, i.e. marketing items sold to people who haven't even memorized many characters yet (as shown by the fact that they have sō-furigana or full glosses). Of course, I find much of the same modes of expression in Classical as well as modern literature, and in technical/academic texts too. But this is not some obscure trick used only by experimental poets. It shows up all the time in advertising, song lyrics, pop-lit etc.

    If the Japanese stop learning kanji in school, all this wonderfully creative production would cease with it. Characters would be relegated to the position of Georgian, or to Tolkien's Poems for Philologists. It's that that I feel would be a loss to the world's culture—though, again, I think it's not my position to judge, and if they decide they don't want to pay the price for kanji anymore, good for them. But, luckly for me, the vast majority of Japanese are favourable to keep using kanji, as shown by MEXT polls.

  63. languagehat said,

    January 27, 2016 @ 9:06 am

    Well, of course such things would be lost (or would become the preserve of a small group of enthusiasts). That's the price that would be paid for a vast expansion of literacy among the kind of people who have no idea that manga even exist. I appreciate the value enthusiasts find in subtle puns and the like, but I draw the line at assuming that such refined pleasures are worth the price of keeping millions in a state of semi-literacy. Much was lost when the old aristocratic cultures of Europe went down before the tide of modernity, too, but most people today think it was worth it.

  64. leoboiko said,

    January 27, 2016 @ 10:00 am

    @languagehat: I see two problems with that argument. The first is that, as I've argued above, I don't think it's a given that removing characters from education would expand literacy. Before taking such a wide-reaching political decision, you better give some damn good evidence of it. Japanese literacy is already among the best of the world. I see countries with simple and complex writing systems, with good and bad literacy, in all combinations; it seems evident to me that economic and social factors are the determinant. All rich countries with high education investment and high income equality have high literacy. If the Chinese government removed characters from school tomorrow, I'm 100% convinced that Chinese literacy would still be suffering, for the very same reasons it suffers around me right now, despite our regular Latin use of the Latin alphabet. If you want to improve literacy rates, rather than imposing a top-down writing system reform, your efforts would be much more effective if you dedicated them to build better schools, and to improve general life conditions. In China, political reform is much more urgent than orthographic reform could even dream to be.

    The second problem is that, as I think it's pretty self-evident to anyone who's been to the country, fluent character reading is not a luxury of the elite in Japan, nor a pasttime of enthusiasts. It's already about as widespread as it gets. Suppose the Japanese government removed kanji from school education. Suppose even that this act had literacy-improving powers – I don't think it has, but for the sake of argument. Where would Japanese-language manga reading even expand to? Everyone who wants to read them already do! Brazil's best-selling magazine, right-wing Veja, goes for around 1 million copies/month; Japan's Shōnen Jump! comics' anthology alone sells 2.4 million per week. N.B.: 9.6M is about ten times as much, in a country with ~60% less people; per capita, we can say some 15×. And, like most pop-culture media, Jump! is a treasure-trove for creative kanji expression; I challenge you to give me any issue, and I guarantee I'll find some sort of interesting kanji use in it, be it 超 glossed as sūpā (<eng. "super") or morphological reanalysis or obscure characters resurrected for effect—they do these things all the time, and not only them.

    (For a comics-to-comics comparison, the Turma da Mônica empire would reach some 1.3M issues/month if you counted together all their many magazines—some 11× less circulation, and buying that would cost a hell lot more than a Jump! issue, i.e. would be inaccessible to most Brazilians—because here, reading is a luxury; again, the writing system is a red herring.)

    As someone from a working-class, third-world background, I take exception to notions like, "this subject is nice but too complicated, we should remove it from schools." As a local rock song goes, "we don't want just food". There's more to people's dignity than the bare minimum. Saying that characters shouldn't be taught because they're an ineffective way of representing language is an utilitarian claim that, to me, sounds like saying that painting shouldn't be taught in art classes because it's an ineffective way of representing visuals. By all means do teach photography (I'm pro-digraphia), but please do realize that even the poor do want beauty—as shown by Mao's failure in abolishing characters.

  65. Elessorn said,

    January 27, 2016 @ 10:27 am


    That's the price that would be paid for a vast expansion of literacy among the kind of people who have no idea that manga even exist. I appreciate the value enthusiasts find in subtle puns and the like, but I draw the line at assuming that such refined pleasures are worth the price of keeping millions in a state of semi-literacy.

    What country are you talking about? What people's state of semi-literacy? I hope you'll agree with me that these things have to be actually demonstrated to exist. Do you have evidence for them?

    I completely agree with you that the "Why do you want to get rid of the characters!?!" reductio ad absurdum is unfair and irritating, but this objection has hardly been made at all in this thread.

    It has, however, been argued repeatedly that while the script *is* indeed relatively more difficult, there seems to be little empirical evidence that this greater relative difficulty is actually doing the script's users harm, or even putting them at a competitive disadvantage. Clearly these are two different questions, no? Would you argue they are the same? The vast literacy deficit you hint at is completely alien to my (necessarily limited) experience: where do you see it?

  66. languagehat said,

    January 27, 2016 @ 10:30 am

    I thoroughly agree with your last point, and I'm sorry if I came off in any way as suggesting that the poor don't need or want beauty. But there are an infinite number of forms of beauty in this world, and this one is not so vital to life that it should be forced down anyone's throat. Clearly our disagreement is with the idea that the difficulty of characters is a serious impediment; this seems self-evident to me (and many others), but I can't prove it to you. Fortunately our disagreement is entirely academic, since there is zero chance of any governments actually getting rid of characters. The experiment continues!

  67. languagehat said,

    January 27, 2016 @ 10:40 am

    Er, my last comment was in response to leoboiko; Elessorn snuck in while it was being posted. I will only add the following: it is as if everyone were forced to spend years learning to bake bricks and lay them and create various forms with them, and those who had already gone through this were indignant at the idea that future generations shouldn't be forced to continue the ritual: "But I'm glad to have learned how to lay bricks, and I understand the forms of architecture so much better! It's enriched my life and the lives of millions of others! Where do you get this idea that it's held us back? Look at how advanced our civilization is!" I mean, sure, Japan is a very advanced country and everyone lives well; it's obviously not being held back in some sort of Stone Age condition by kanji, but that was not my contention. My contention is that it would be an improvement if everyone didn't have to put in years of their lives learning how to bake and lay bricks. This seems to me so obvious on the face of it that I am at a loss as to how it can seriously be denied.

  68. leoboiko said,

    January 27, 2016 @ 10:59 am

    @languagehat: There are a few problems with your analogy. The first is that brick-laying is a specialized profession, while reading is a general, everyday activity in a literate society. The second is that the word "brick-laying" denotes a utilitarian activity, a non-creative task. There's no kind of building that you can't understand without learning to lay bricks, the aesthetics of architecture lie at a different level; but there are kinds of expression that you can't make without morphography. Nothing aesthetic would be lost if people in general stopped laying bricks; something aesthetic would be lost if people in general stopped writing and reading character-based texts.

    A further problem is that, if the said brick-laying is creative and aesthetically pleasurable, then it isn't "brick-laying" or "a ritual" at all; it's a traditional art form, a kind of ceramics or sculpture. This should make it clear why people would oppose the idea of removing the transmission of an aesthetic cultural practice from a nation's general education. If I found a remote island where people had a rich tradition of expression through sculpture, and this tradition was already widespread and not causing any problems, then I'd feel it would be a cultural loss for the world if I induced them to abandon this social practice because factory production is more efficient. That's how I'd feel if foreign script reformers managed to abolish kanji expression from Japan's already existing, already thriving, already popular reader's culture. There are infinite forms of beauty, yes, but this one particular form of beauty would be killed together with the social structures that support it.

  69. languagehat said,

    January 27, 2016 @ 11:34 am

    It's an analogy, not an identity, and the point of the analogy is to emphasize precisely the hard work of learning the character system.

    Nothing aesthetic would be lost if people in general stopped laying bricks

    You say that because that's not your form of aesthetics; if you'd gone through years of training, with the aesthetic qualities endlessly emphasized by both your teachers and cultural tradition, you'd feel differently. I find your insistence on the aesthetics elitist (why should everyone suffer because you find characters so thrilling?); you find my insistence on utilitarianism elitist (why should people be denied the aesthetic thrill of characters?); we're obviously never going to convince each other. But I repeat that I understand the aesthetic thrill of characters and don't want to deprive anyone who wants to go through the years of study the pleasure of doing so; I simply don't think those years of hard study should be a prerequisite for enjoying the fruits of civilization. People said the same thing about eliminating the classics requirement: you want to deny the masses the thrill of learning Greek and Latin, reading Sophocles and Horace in the original, understanding all the subtle allusions of centuries' worth of literature? But the requirement was eliminated, and we all muddle along somehow, even though the loss was real (and I love reading Sophocles and Horace in the original).

  70. leoboiko said,

    January 27, 2016 @ 11:57 am

    I find the analogy to Latin a lot more convincing than the brick-laying one, with the caveat that Latin creative production was already a lot less thriving than East Asian character-text production (by the time Latin was taken from the schools, it wasn't common for sugar-pop bands to use it in song lyrics). But even so: Do you know the public school I've mentioned above, the one I volunteered to? The fact that Latin and Greek were reintroduced to it to great student acclaim should make it clear what I think of the topic. Yes, I think the loss of Latin education in a country of Romance heritage (however colonial it was) was a cultural loss, and a damn shame at that; and I also think that it didn't help one iota in spreading literacy, improving schools, or increasing equality (just look around). I'm told Portugal is buffing it up, while in Germany it's the 3rd most popular in schools (after English and French); I hope the trend catches.

  71. languagehat said,

    January 27, 2016 @ 12:32 pm

    Well, I'm glad to see you're consistent in your views, and may I say that this is by far the most civilized discussion of the issue I've ever had the pleasure of participating in — hearty thanks to all!

  72. leoboiko said,

    January 27, 2016 @ 1:00 pm

    It's an emotionally-charged topic, to be sure, and I hope I haven't gotten too carried away.

    I'd like to add that I have nothing but the profoundest admiration for the work of DeFrancis, Boltz, Unger, Mair and other linguists of this character-critical school. I've devoured all I could find, with wide-eyed abandon; and, while I may disagree with the extent of the conclusions, I regard them as the teachers who have taught me everything I know about this topic. I'll do my best to expose more people to the debunking of hanzi-related myths, with was sorely needed and expertly done.

  73. Doctor Science said,

    January 27, 2016 @ 7:49 pm

    I have been searching and searching for a paper (or book) talking about how statistically significant numbers of Chinese and/or Japanese readers lose the ability to read many hanzi/kanji after they leave school. In other words: gradual character amnesia (not just handwriting amnesia), leading to a decline in reading level with age. This would specifically be *before* computerized pinyin-to-hanzi/kanji conversion was possible.

    I'm pretty sure I've seen something like that linked from LL before — and there may even have been a LL post — but I'll be damned if I can come up with the right search terms now. Does anyone know what I'm thinking of? Or know the correct search terms to call it forth?

    [(myl) I think that you were right, at the start of this comments section, to ask whether current "character amnesia" is more consequential than earlier (and current) limited literacy. Some historical perspective (on the case of Japan) is sketched in "Japanese Literacy: Back to the Future again?", 12/26/2006, which quotes from J. Marshall Unger, Literacy and Script Reform in Occupation Japan, 1996:

    The first full-fledged nationwide attempt to measure literacy in Japan was the survey conducted in 1948 under the auspices of the Civil Information and Education Section … [which] involved the testing of about 17,000 Japanese men and women between the ages of 15 and 64 throughout the country. According to Ishiguro Yoshimi, who chaired the survey's Central Planning and Analysis Committee, the survey was of unprecendented scope and rigor not only by Japanese standards but by world standards as well. Although the survey is sometimes cited as proof that the level of literacy of the majority of prewar Japanese was high, it clearly shows that earlier government claims were grossly inflated. It was found that the rate of illiteracy (monmōritsu 'complete inability to read or write') was indeed very low, but it was also concluded that only 6.2 percent of the population were literate in terms of the survey definition, which was liberal…. By today's standards, all the questions were very simple. The ability to write kanji from dictation (kanji no kakitori), which was identified as the single most important skill tested, was found to be "remarkably low" in ALL groups surveyed. Performance was closely correlated with levels of formal education… Finally, the claim that the average Japanese experienced trouble dealing with the media of mass communication, a claim long made by script reform advocates, was deemed proven.

    Neustupný points out that a second, smaller survey conducted in 1955-6 by the Ministry of Education produced similar results.

    The same post also notes that the complexities of the Japanese writing system represent one of the key problems in the way of solving Japan's impending demographic train-wreck — facing a situation without enough working-age people to support the growing elderly population, a logical part of the solution would be to increase immigration. But the difficulties of making this work in an insular society are massively increased by the problems of making immigrants functionally literate in Japanese.]

  74. Victor Mair said,

    January 28, 2016 @ 8:59 am

    I wish to thank everyone who participated in this debate, especially for remaining civil in discussing what by nature is a contentious subject.

    I also want to thank my colleague Mark Liberman for helping Doctor Science find the source that he was looking for.

    Of course, there has always been a problem with character retention, but the advent of electronic devices for writing in Chinese, especially when coupled with alphabetical or other phonetic inputting (which is true most of the time nowadays), has exacerbated the rate of loss to such a degree that it threatens the very foundations of handwriting ability, even for those who are highly educated. This new kind of digitally enhanced forgetting is of such magnitude that it alarms educational authorities and literate individuals across the board.

    The existence of this qualitatively and quantitatively different kind of forgetting how to write characters by hand together with the simultaneous rise of digraphia (brought about in part through phonetic inputting) has led to a situation that is unprecedented in the history of writing with Chinese characters.

  75. J. M. Unger said,

    January 28, 2016 @ 11:31 am

    I think a distinction should be drawn here. While unofficial digraphia may be a symptom of the problem, official recognition and encouragement of digraphia would go far toward solving it. See my review of Language Policy in Japan: the Challenge of Change, by Nanette Gottlieb in the Journal of Japanese Studies, vol. 39 no. 1 pp. 229–33 (2013).

  76. Victor Mair said,

    January 28, 2016 @ 1:33 pm

    @J. M. Unger

    While "official recognition and encouragement of digraphia would go far toward solving" "the problem" that you point to, I think it is unlikely that any Chinese government would make digraphia official policy, at least not for the foreseeable future. Anyway, it's not something that I want to speculate about. That's why I always refer to digraphia as "rising", "emerging", and so forth. I think that it is happening before our very eyes, incrementally and as a matter of natural evolution.

RSS feed for comments on this post