Backward Thinking about Orientalism and Chinese Characters

« previous post | next post »

 This is a guest post by David Moser of Beijing Capital Normal University

For those of us who teach and research the Chinese language, it is often difficult to describe how the Chinese characters function in conveying meaning and sound, and it’s always a particular challenge to explain how the writing system differs from the alphabetic systems we are more familiar with. The issues are complex and multi-layered, and have important implications for basic literacy and the teaching of Chinese to both native speakers and foreign learners. Tom Mullaney, a professor of history at Stanford University, has lately been muddying these pedagogical waters in a series of articles and interviews that seriously misrepresent the merits and relative advantages of the alphabet over the Chinese script.

Mullaney’s recent article in Foreign Policy was one that particularly caught my eye. The problems with the article start with the title: “Chinese is Not a Backward Language”, which raises a red flag at the outset. For any scholar of the Chinese language, the question immediately arises: “Chinese language or Chinese script?” The two are not synonymous, after all. Reviewing the article, it is obvious that Mullaney constantly conflates, confuses, or ignores the distinction between the two. Sometimes the conflation seems careless and unintentional, but at other times I suspect that he blurs the line intentionally to bolster the straw-man argument of his piece: namely that criticisms of the Chinese characters are part of a lingering racist, “Orientalist” prejudice against the Chinese language, and by extension, Chinese culture itself.

Mullaney wants to contend that Chinese has gotten a bad rap. He says that Chinese has a reputation as an unscientific, backward language, due in part to misinformed criticisms of its grammar from early sinologists, whose arguments are now consigned to the dustbin of history. Yet he tells us that these attacks linger on in a Western disparagement of the Chinese script, which is said to be a symbolic medium quite unfit for the information age. But now, Mullaney announces, this longstanding Western “Orientalist” critique has been definitively refuted. The Chinese characters have been vindicated by their undeniable success as one of the major scripts in cyberspace:

The great irony of Orientalism 2.0 is that Chinese characters are not only going strong in the 21st century, but they are one of the fastest, most widespread, and successful languages [sic] of the digital age.

He goes on to note how China has become an IT giant in which social media and smart phones proliferate. “For better and for worse [sic?],” he says, “none of these outcomes it turns out depended upon China going the route of wholesale alphabetization.”

First of all, one has to take issue with his claims about the attacks against Chinese. Exactly who is saying that Chinese is a backward language? For the record, no reputable contemporary scholars are espousing anything like this claim, which is why Mullaney has to go back to Hegel and Social Darwinism for his examples. One can, however, find plenty of reputable linguists and experts who have much to say about the problems with the Chinese script, and many of these problems are still with us despite the relatively successful adoption of characters into cyberspace. But the claim that the Chinese language itself is under assault from chauvinistic Western linguists is a much starker and more sensationalist narrative that draws more media attention and re-tweets.

Second, the fact that Chinese script has a major presence on the Internet is no proof that historical criticisms of the inefficiency of the Chinese script were somehow invalid. After all, it took a massive, concerted effort on the part of computer programmers for more than two decades to put Chinese character input on a par with that of alphabetic systems. And despite the exciting new advances in character entry that Mullaney mentions, even now the most common input methods (straightforward pinyin, wubi entry, etc.) are still cumbersome in comparison with typing in alphabetic text. (Pinyin or handwriting input on a smart phone or pad still generally involves a two-step process, in which the user must choose the correct character candidate from a pop-up menu.) Mullaney seems to assume that, because modern computer science has come up with workable, workaround solutions to the problem of digitizing Chinese characters, there evidently was no problem to begin with!

In an interview for the LA Review of Books, Mullaney goes even further, implying that it is alphabetic writing that is that is more retrograde in the information age, whereas Chinese characters are somehow the wave of the future:

If anything, Chinese conquered the alphabet, not the other way around…. In the Western world — or really in the “Alphabetic World” — we use the computer keyboard in a dumb, what-you-type-is-what-you-get kind of way. In all but rare instances, we assume a one-to-one correspondence between the symbols on the keys we strike and the symbols that we want to appear on the screen. Press the button marked ‘Q’ and ‘Q’ appears. It’s just that simple…. Chinese “input” uses the QWERTY keyboard in an entirely different manner. In China, the QWERTY keyboard is “smart,” in the sense that it makes full use of modern-day computer power to augment and accelerate the input process…. [T]he Latin alphabetic world has spent more than two centuries congratulating itself for “Our Glorious Alphabet,” and yet at the same time has done far less to explore and push the Latin alphabet to its fullest potential.

It is an odd contention that a writing system that still mostly relies on alphabetic input methods has somehow "conquered the alphabet", but put that aside for a moment. There is some dubious characterization here, as well as subtly misleading use of the words “dumb” and “smart” to contrast the two systems. Mullaney seems to assume that computer keyboards always possessed these inherent “smart” capabilities, but the smug, complacent users of the plodding alphabetic script were unable to bring these functions into “full use.” But this is sheer nonsense. Mullaney is obviously right that the increased functionality of the traditional keyboard that came as a result of techniques developed to process Chinese characters is also a boon to input of alphabetic languages, as well. But again, this “smart” functionality of the computer keyboard was the result of decades of very hard work and ingenuity, as computer programmers had to struggle mightily to figure out a way to input thousands of complex graphs using only the limited key set of the QWERTY keyboard. The expansion of computer power he refers to was a hard-won solution to the very real problem of dealing with a Chinese writing system whose component symbols were a crippling burden on human memory and computer memories of the 1980s and 90s. For those of us who struggled through this development phase, rooting for the breakthroughs that would make character entry on the computer faster and more intuitive, this historical amnesia is simply baffling.

Mullaney is obviously correct in saying that advances in entry technology that resulted from tackling the problems of digitizing Chinese script have now opened exciting new horizons in the field. Bravo. But surely we owe our thanks not to the Chinese writing system itself, but rather to the ingenuity and cleverness of the programmers who devised the solutions. Advances in cancer drugs may produce useful spinoff drugs that can cure other diseases, but we usually don’t thank cancer itself for the breakthroughs.

Mullaney’s reversal of logic can be made clearer through analogy to the legendary inadequacies of English spelling. Scholars for generations have bemoaned the state of English orthography, noting that the system is riddled with inconsistent spellings in words such as through, though, rough, bough, cough, etc. There were multiple schemes for completely overhauling the spelling rules of English to make the conventions more efficient and consistent, as are those of Spanish or German. Imagine, if you will, a writer who pooh-poohs these criticisms of English orthography, accusing critics of English spelling as “Anglophobes”, and crowing that “The great irony of this “Anglophobia” is that English orthography is not only going strong in the 21st century, but it is one of the fastest, most widespread, and successful languages [sic] of the digital age!” No one would seriously contend that the development of things like computer spell-checkers refutes the criticism that English orthography is seriously broken. If anything, the need for such technological fixes confirms the intrinsic drawbacks of the writing system.

But the main thing wrong with Mullaney’s utopian reasoning is that he focuses exclusively on the problem of digitizing Chinese characters for processing and manipulation in cyberspace, ignoring a host of other deep-seated problems that still persist. He is surely right that computer technology has come up with exciting new character-input paradigms, which can be used to improve upon traditional one-to-one alphabetic entry. Yet he completely ignores the many age-old burdensome problems that are so much a fact of daily language life that users have simply become inured to them. Among these are:

  • The countless hours of time it takes to attain basic literacy in Chinese due to the sheer number of characters to be memorized. China is very proud of its supposed 95% literacy rate, yet what goes unspoken is the inordinate amount of time children must spend in mastering the system, in comparison to alphabetic writing systems. As long as Chinese school kids (and hapless foreign adults) are expected to master the writing of this system by hand, all the advanced input systems in the world will not lighten this burden. (One could make the case that the true contribution of computer technology to Chinese script is not improved input methods, but the option of sidestepping the inputting of characters entirely by simply using speech-to-text technology.)
  • The use of pinyin as an add-on writing system employed to teach Chinese school kids the sounds of Putonghua. It is in some sense absurd that the poor overworked Chinese children must learn pinyin, an alphabetic writing system, in addition to the burdensome character system, in order to master basic reading skills in their own language. This is especially ironic, given that pinyin alone would be sufficient to achieve basic literacy, if only Chinese books were printed in this imported phonetic writing system.
  • The “character amnesia” problem (in Chinese tibiwangzi 提笔忘字), namely, the inability to recall how to write the graph for a sound due to memory overload. (Referenced in Language Log here, and here, and here). This has been a problem for centuries, and made only worse now with the advent of the pinyin input method and voice messaging options on digital devices. The problem has gotten so bad among young people in China that Chinese state media have produced several TV shows such as Hanzi yingxiong《汉字英雄》 “Chinese Character Heroes”, and Hanzi tingxie dahui《汉字听写大会》 “Chinese Character Dictation Meeting”, in which middle school children compete in various character writing tasks, as a way of reviving the rapidly deteriorating skill of writing by hand. This character amnesia — ironically the result of relying on the less memory-intensive alphabetic pinyin input method! — results in much wasted time as users have to check their digital dictionaries to retrieve the forgotten graph, resort to pinyin – and, increasingly English! – or simply opt out of the system entirely, resorting to some kind of speech-to-text technology to get the right character on the screen.
  • The enormous cognitive burden of the reading process, which is due to the lack of phonetic information in the script. The lack of phonetic feedback for Chinese characters increases the difficulty of reading acquisition many fold, discourages non-native speakers – and even many native speakers – from even attempting to tackle the written language. Many perfectly intelligent ex-pats living in China make a very rational decision to eschew the written language altogether, realizing that the ratio of effort-to-effect is simply too abysmal. And in Singapore’s multilingual education system, Mandarin studies are increasingly shunned by ethnically Chinese students, in large part because the kids increasingly balk at the difficulty of memorizing Chinese characters. The problem has become so severe that the Singaporean government has set up a special agency, the Singapore Centre for Chinese Language (SCCL) to improve the effectiveness of teaching Mandarin in the schools.

These are all daily frustrations that are simply not a part of the digital lives of citizens in the “Alphabetic World”, as Mullaney calls it. They are problems with the Chinese script that the world of Information Technology has not yet solved, and in many cases, has exacerbated. Mullaney wants us to believe that notions of the superiority of the phonetic alphabet as a writing system are merely smug, chauvinistic prejudices of the West, but his evoking of this trendy cultural trope obscures the real pedagogical and informational challenges still presented by the Chinese script. Alphabets still hold enormous informational and ergonomic advantages over the Chinese writing system, and Mullaney seems to totally ignore these considerations.

Finally, there is Mullaney’s contention that critics who consider the Chinese characters as unsuitable for cyberspace are exhibiting vestiges of Orientalism, colonialism, and even racism. This is a talking point that has garnered much media attention, even in the Chinese-language press. Yet such claims exhibit an astonishing indifference to historical reality. Surely Mullaney is aware that the fiercest critics of the Chinese writing system were not foreigners, but the Chinese themselves. May Fourth intellectuals such as Chen Duxiu and Guo Moruo advocated the eventual abolition of Chinese characters, and even China’s most famous modern writer, Lu Xun, was quoted as saying “If the Chinese characters are not eliminated, China is doomed.” Were these patriotic May Fourth Chinese guilty of racism, or colonialist thinking?

In fact, throughout the 20th century, many Chinese writers and educators championed the Romanization of Chinese as a prerequisite to self strengthening, and in the years prior to the communist takeover in 1949, Mao Zedong and his language planners expressed every intention of abolishing Chinese characters once they came to power. (See my book A Billion Voices: China’s Search for a Common Language for more on this history.) Mao himself said to Edgar Snow in 1936:

Chinese characters are so difficult to learn that even the best system of rudimentary characters, or simplified teaching, does not equip the people with a really efficient and rich vocabulary. Sooner or later, we believe, we will have to abandon characters altogether, if we are to create a new social culture in which the masses fully participate.

Do these remarks reveal that Mao thought that Chinese was a “backward language”? Was he exhibiting a form of Orientalist thinking? Or was he merely expressing a pragmatic acknowledgement that the characters were an enormous stumbling block to universal literacy?

Despite Mullaney’s starry-eyed claims, the advanced, multi-functional Chinese input methods have not miraculously solved all the problems associated with Chinese script. Computer technology has indeed made great strides in solving the problem of character entry, putting alphabets and Chinese characters on more-or-less the same footing. Yet educators, publishers, websites, and software developers still struggle with the same kinds of inefficiencies and added burdens that have always plagued the writing system.

The calls for alphabetization of the Chinese language were quite reasonable and valid throughout most of the 20th century. Only in the last three decades has this changed. Make no mistake: The incorporation of Chinese characters into modern cyberspace is not proof that the calls for their abolition were merely misguided prejudice; rather, it is modern IT technology that has saved the characters from almost certain obsolescence in the increasingly interconnected and complex 21st century.

Note that it is not my contention – nor that of any serious scholar I’ve ever read – that the Chinese script is “backward.” This is simply not a sensible or productive category in linguistics, and Mullaney’s evoking of this culturally loaded term is an annoying distraction. The Chinese script is a marvelously complex, rich and sophisticated cultural product, one that has its own semiotic dynamism and esthetic beauty, as well as its informational strengths and weaknesses as a writing system. As such, it should be approached and evaluated on the basis of all the relevant practical considerations, and not as an excuse to engage in rhetorical displays of political correctness.

Chinese characters may have a bright future, as Mullaney suggests, but the stubborn reality of their continuing cognitive and ergonomic disadvantages require clear-headed, scientific solutions, not headline-grabbing hype about nonexistent culture clashes. We should pool our knowledge and our digital resources to tackle the very real problems of learning to read and write Chinese characters — and not merely pretend that such problems are a mirage brought on by Western delusions of cultural superiority.


  1. J.W. Brewer said,

    May 16, 2016 @ 2:06 pm

    Advocates of a more phonetic writing system for Mandarin could avoid the "orientalism" charge by promoting the use of a non-Western, indeed "oriental," script rather than hanyu pinyin. And apparently such a solution has been there all along but was foolishly overlooked by the PRC's decisionmakers:

  2. Bob Ladd said,

    May 16, 2016 @ 2:44 pm

    I was going to make a comment along similar lines on the original thread, but this is much more fully worked out. In particular, I was going to make the same point about criticisms of English orthography and the fact that no one takes those as evidence of Anglophobia. I even got as far as digging up a quote from Leonard Bloomfield: "The difficulty of our spelling greatly delays elementary education, and wastes even much time of adults. When one sees the admirably consistent orthographies of Spanish, Bohemian [sic – this was written in 1933], or Finnish, one naturally wishes that a similar system might be adopted for English. It is not true that to change our orthography would be to 'change our language': our language is the same, regardless of how we write it." Lots of people have had lots of nasty things to say about Bloomfield since he wrote, but as far as know no one has ever accused him of deprecating English culture.

  3. J.W. Brewer said,

    May 16, 2016 @ 3:31 pm

    But criticisms of English orthography are often taken (and rightly so) as evidence that one is a crackpot. Perhaps it would be better for advocates of a different writing system for Mandarin to be viewed as crackpots rather than bigots, but I doubt that's what they're aiming for.

    I would be happy to accuse Bloomfield of deprecating (or at least failing to appreciate) English culture. Finnish and "Bohemian" have simple modern orthographies because until well into the 19th century they were low-status (deprecated, one might say) languages associated with the illiterate peasantry, with other languages being used for government, education, and high culture in the areas where they were spoken. This had the side effect of making it easy to standardize orthography along rationalizing lines in the 19th century, because there was no well-established traditional-yet-messy system to overthrow. But Anglophones would hardly be better off if their L1 had been a low-status and deprecated language until the rise of 19th century nationalism.

  4. Vanya said,

    May 16, 2016 @ 4:06 pm

    Why on earth would David Moser think that citing Mao strengthens his argument? Mao was a mass murdering monster who did more damage to China in the 20th century than any individual you can think of, and certainly one of the top 5 worst human beings of that century.

  5. D.O. said,

    May 16, 2016 @ 4:32 pm

    Mr. Brewer, how about Spanish and German? French orthography is somewhat cumbersome, but by and large consistent. Difficulties of English orthography probably come from its long history of written language, multi-layered sources of large chunks of vocabulary, absence of "native" script (in form of extra letters or diacritics) and voracious habits in adopting foreign words.

  6. S Frankel said,

    May 16, 2016 @ 4:47 pm

    I haven't read Mullaney's article, but from the first part of this post, it sounds as though he's committing the common error of mixing up "language" with "script." Many, probably most, people are only dimly aware, if that much, that "language" is primarily "spoken language," with the script as sort of an add-on.

  7. WSM said,

    May 16, 2016 @ 4:48 pm

    Leaving aside the implicit comparison of Chinese characters to a cancer, I'm curious as to why there has not been a wholesale abandonment of Chinese characters on the Internet in favor of pinyin, if the advantages of pinyin (or some other phonetic system such as bopomofo) vis a vis characters are really so compelling? I am aware of interesting linguistic trends involving various forms of "digraphia", but the incidence rate of such trends seems dwarfed by the overwhelming presence of character use within the Chinese speaking world. This is to say nothing of more formal realms such as domestic scientific research, where the choices continue to be English (as the international lingua franca) or Chinese characters: if alphabetic scripts are really so much more convenient forms of communication, why aren't they being used more? Neither the Internet nor certainly Pinyin are new inventions at the point, so I'm curious when you expect pinyin / other alphabetic systems to really take off, and/or under what conditions.

    Second, And here I would appreciate some perspectives from native speakers, I am curious how you interpret the presence of Chinese character subtitles (for which I'm grateful, btw) in many TV broadcast interviews, news clips, drama, etc: is that part of state-level efforts to reinforce characters wherever possible, or does it reflect the fact that spoken Mandarin Chinese (generally all you will hear on TV, with exceptions) continues to be subject to sufficient variation, that the utility of characters' weak association with spoken sounds continues to outweigh whatever inconveniences it might pose to readers. In the latter case, use of pinyin would seem to pose significant disadvantages to communication between regions within China, *particularly* on the Internet.

  8. J.W. Brewer said,

    May 16, 2016 @ 5:26 pm

    @D.O. I do sometimes have an impractical romantic wish that English had retained its distinctive medieval graphemes (thorn, edh, yogh etc) and not sold out to the dumbed-down pan-West-European ascii of the day. (Indeed, if we were working on a blanker slate, a mildly-tweaked Cyrillic might be a better alphabet for English than the Latin script.) But what's done is done, and the other points you mention simply underscore that the peculiarities of English orthography are side effects of some of the same historical circumstances that give it such extraordinary range or expression and political/cultural power. At a minimum the divergent historical trajectories of Anglophone and Hispanophone nations (a century ago Argentina and Australia had about the same per capita amount of wealth; these days, not so much) suggest that orthographic simplicity is not the royal road to prosperity and decent government and is generally at most rounding error on top of the other factors that influence good or bad historical outcomes for different peoples.

  9. GH said,

    May 16, 2016 @ 6:09 pm

    @J.W. Brewer:

    But criticisms of English orthography are often taken (and rightly so) as evidence that one is a crackpot.

    I'm not sure I believe that. Advocating for spelling reform may perhaps be taken as such, because the practical obstacles are so overwhelming. But can anyone really disagree that English orthography is a mess, and that it would be preferable if it were not? (Assuming that could somehow be achieved without affecting all the various historical contingencies that led to the current situation.)

  10. Chris C. said,

    May 16, 2016 @ 6:27 pm

    It's hard to defend English orthography on the merits, but it at least has the advantage of universality. A phonetically-spelled English would spell whose English? Yorkshire, London, Dorset, Melbourne, New York, Houston, and Los Angeles would all end up with different systems, with varying degrees of mutual intelligibility.

    On the other hand, I think drawing an equivalence between English orthography and hanzi is perhaps going a bit too far. Most written words can be sounded out reasonably well in most dialects. Production is more difficult than consumption, but not much more so. It's only certain conventions that are troublesome, which is why we tend to see the same examples cited every time someone files a complaint. My personal favorite is the name of a certain street in San Francisco. On moving to the area thirty years ago I had no idea how to pronounce it, and I've heard different pronunciations given as "correct". How would you pronounce "Gough?"

  11. J.W. Brewer said,

    May 16, 2016 @ 6:37 pm

    GH: viewing the actually-existing facts of an actually-existing natural language, which necessarily result from a bunch of historical contingencies, as an unfortunate mess in need of reform rather than an endlessly fascinating thing in need of study, analysis (because the surface messiness often turns out to be patterned rather than simply perverse), and simple appreciation (philosophy begins in wonder etc.) is to my mind an indication that one is missing the whole point of the thing. Spelling reform, Esperanto, peevish objections to split infinitives, etc etc all seem to me to be different facets of the same basic missing of the point.

    But it seems likely that the gap between traditional Chinese script and any alphabet-like system is so different (in kind not just degree) than the variation between English and Finnish orthography that I'm not sure that the arguments either way carry over in ways that are likely to be more helpful than confusing.

  12. Eidolon said,

    May 16, 2016 @ 7:07 pm

    "Leaving aside the implicit comparison of Chinese characters to a cancer, I'm curious as to why there has not been a wholesale abandonment of Chinese characters on the Internet in favor of pinyin, if the advantages of pinyin (or some other phonetic system such as bopomofo) vis a vis characters are really so compelling?"

    I, too, was struck by the use of that metaphor; Moser will certainly take heat for it if it was ever translated as-is in China, as implicitness and political correctness is even more sensitively received in the PRC, than it is in the West. He will have a hard time professing his objectivity among Chinese audiences with such a strong and, in the end, unnecessary slight.

    But cultural sensitivity also provides an answer to your question – pinyin is not preferred to Chinese characters because "serious writing" is culturally expected to be in characters, not in pinyin. Pinyin can be used in chat rooms, instant messages, and other casual settings limited to one or two lines, but essays, journals, books, creative writing, message boards, websites, and other types of long writing – as in several sentences or more – are expected to be in characters. A Chinese individual not literate in characters is not considered literate at all. There is, as such, an immense amount of cultural prestige attached to the characters, which has a historical precedent, as character literacy was always associated with the scholar elite and upward mobility. Thus, one cannot expect that pinyin will replace Chinese characters on the internet simply because it is more expedient, just as one cannot expect that English orthographic reform will occur just because it is sensible.

    As for the critique itself, while I agree that Mullaney is basically wrong in: 1) conflating the Chinese languages with the Chinese writing system and 2) ignoring the objective issues with the Chinese writing system, I think Moser's response is also flawed, which is unfortunate because his argument is otherwise reasonable. For one, the fact that Chinese intellectuals have also criticized the Chinese writing system does not mean such criticism cannot be Orientalist/colonialist/racist. Moser must surely be aware of postcolonial concepts such as internalized racism/Orientalism, which though alien to many in the West, was a powerful phenomenon in countries that were colonized or that were under the imminent threat of colonization, such as China. Many Chinese intellectuals writing in the 19th century did, in fact, harbor strong biases against their own social and cultural traditions, and the havoc wrecked by the Communists on those traditions were in no small part due to the perception that all Chinese culture was backwards and in need of eradication. Moser would surely not want to associate himself with the rhetoric of the cultural purges that resulted in such destruction across China under the Communists, and which were motivated by self-hatred/internal Orientalism dating back to 19th century radical reformists. To this end, asking whether the Communists considered traditional Chinese characters/culture backwards is not a rhetorical question of the negative form, because the answer is an emphatic yes.

    Equally, it is both dangerous and disingenuous to speak of modern IT technology as the "savior" of Chinese characters, which was otherwise fated for a footnote on the wrong side of history. IT technology is new, and as Moser says, it took decades for efficient input methods to be developed for Chinese characters. Yet, Chinese characters continued to be used during those decades, and the PRC and ROC's literacy rates were not achieved with the aid of electronic inputs, but with vastly improved education for the masses. The "success" of Chinese characters is evidence of the malleability of the human experience, rather than the other way around, as the script that Moser imagines must certainly have become obsolete without the benefit of modern IT technology, was used, however inefficiently, by over a billion people in the world, and continues to be so.

  13. Starry Gordon said,

    May 16, 2016 @ 10:06 pm

    I once read somewhere the contention that Chinese characters could be read faster than alphabetic representations of Chinese (or other languages). I don't recall any evidence being given, but it seems possible that such a set of symbols might have higher optical or neural bandwidth, so to speak.

  14. Tom said,

    May 16, 2016 @ 10:39 pm

    As Zev Handel brought up in a discussion of Mullaney's piece elsewhere, there are many things to consider when thinking about changing a language's orthography. English writing bridges dialectal differences in very useful ways (the vowels in Mary/merry/marry, the consonants in witch/which). It also encodes etymological information that would be lost if we switched to something closer to IPA. Both of these have parallels in the Chinese writing system. On top of that, there's the actual cost of what it would take to convert all the written infrastructure (websites, textbooks, road signs, etc.) from characters to a phonetic system, which would be enormous.

    Relatedly, I've found the debate taking place between Handel and J. Marshall Unger in the pages of Scripta to be very illuminating in thinking about what it means to call something a "logographic" writing system.

    Handel, Zev. "Logography and the classification of writing systems: a response to Unger." Scripta 7:109-151, 2015.

    Response to: Unger, J. Marshall. "Empirical Evidence and the Typology of Writing Systems: A Response to Handel." Scripta 6:75-95, 2014.

    Response to a footnote in: "Can a logographic script be simplified? Lessons from the 20th century Chinese writing reform informed by recent psycholinguistic research." Scripta 5:21-66, 2013.

  15. WSM said,

    May 16, 2016 @ 11:14 pm

    @Eidolon – thank you for the long and thoughtful response. I'm aware of the importance cultural history plays in the predominance of characters on the Internet (and elsewhere, of course), but, within the admittedly narrow confines of my own experience, have seen almost *no* examples of pinyin used consistently in contexts where one might expect expedience to outweigh cultural factors: as a banal example, how many Chinese restaurants have you been to where the handwritten list of your orders is in Pinyin?

  16. Sally said,

    May 17, 2016 @ 1:21 am

    @Starry Gordon
    > I once read somewhere the contention that Chinese characters could be read faster than alphabetic representations of Chinese (or other languages).

    This is true. Chinese isn't really built for alphabetic representation. English words are so varied in length and sound that you don't need to pay attention to every letter in each word to be able to parse a sentence (otherwise put: "it deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteer be at the rghit pclae.")
    Not so for Chinese, where a single syllable could represent many possible characters and many syllables are quite similar. For example, "guan," "gan," "gang," "guang" are all possible syllabic sounds in Mandarin.

  17. Eli said,

    May 17, 2016 @ 2:16 am

    @Sally: and "swan, song, son, swum, sum, sun, sung, swung" are all possible syllabic sounds in English.

    Mandarin written in pinyin is readable; therefore, it's obviously not impossible to represent Chinese with an alphabetic script. Reading speed is based largely on how familiar people are with the script they are reading, so it wouldn't be surprising if a native Chinese speaker who is used to reading characters is quicker at reading characters than at reading an alphabetic representation.

    I don't think we have any way to objectively determine what kind of script a language is best suited for, especially since this kind of judgement includes many intangible elements. It seems like a matter of opinion. But pretty much any kind of human script can suffice.

  18. Sean M said,

    May 17, 2016 @ 2:29 am

    The difficulties of keyboarding Chinese characters strike me as red herrings, since the history of text input is one of systems designed by people who thought that German and Greek had exotic writing systems meeting people who explained that, no, Ascii plus a few special characters would not do at all for their writing system. In a world where an abjad with accents or a logosyllabary became the =lingua franca= of industrialization, there would have been struggles to adapt systems to variants of the Latin alphabet ("why do they ask for so many keys? They even want separate keys for letters and numbers! And they keep talking about this difference between capital and lowercase letters like its a big deal … we had to adopt the red letter/black letter key for that.")

  19. GH said,

    May 17, 2016 @ 2:36 am

    @J.W. Brewer:

    viewing the actually-existing facts of an actually-existing natural language, which necessarily result from a bunch of historical contingencies, as an unfortunate mess in need of reform rather than an endlessly fascinating thing in need of study, analysis (because the surface messiness often turns out to be patterned rather than simply perverse), and simple appreciation (philosophy begins in wonder etc.) is to my mind an indication that one is missing the whole point of the thing.

    There are two flaws in your reasoning. First, you are assuming that you cannot do both. Second, you are assuming that criticism is the same as advocating for reform.

    The problem with the first is that if being a result of historical contingency and "interesting" from a cultural or scholarly point of view made something exempt from criticism, you could make the same defense of any human institution or tradition: the legal system, gender roles, guild control of industry, female circumcision, any superstitious practice, etc. To say that appreciating the richness is "the whole point" is to assert that it's only a quaint custom with no real-world consequences.

    The problem with the second is that… well, it's wrong. One may argue that something is flawed while also accepting that the flaw cannot now be rectified. This week sees the centennial of the Sykes-Picot agreement that drew up what eventually became the borders of many of the Middle-Eastern states. Most retrospectives agree that it has had disastrous consequences, but none, as far as I have seen, have proposed going back to redraw the borders en masse.

    So one can argue that the English system of spelling is highly problematic while also recognizing that conditions that would make improvement (by means of a coherent, wide-reaching reform, as has been done with other languages) possible do not currently prevail, and are not likely to do so for the foreseeable future.

  20. Bob Ladd said,

    May 17, 2016 @ 3:00 am

    Several people have mentioned the fact that English spelling bridges dialect differences. True, but also true of the writing systems of many languages, irrespective of how regularly they represent pronunciation. E.g. Spanish Y vs LL or Z vs S; French UN vs IN; Dutch G vs CH; etc. etc. Dialect neutrality shouldn't be seen as a beneficial by-product of the irregularity of English spelling; substantial amounts of dialect neutrality could be built into a reformed orthography that eliminated a lot of the current irregularity.

  21. Sean M said,

    May 17, 2016 @ 3:01 am

    GH: The people of Iraq and Syria have been busy drawing their own borders for a decade or more, and I seem to recall that a terrorist organization that I won't name made a point of erasing the markers of the Sykes-Picot border in the desert! The existing borders seem to be more important to diplomats in Istanbul or Washington or Baghdad, who are worried about any threat to the agreement that borders can never move at gunpoint (which is a pillar of the current system for avoiding wars between states), or about their own minorities deciding that they would rather be part of Greater Kurdistan than the country that they are now in, or don't want to admit what they have lost, than they are to the people living there.

  22. GH said,

    May 17, 2016 @ 3:14 am

    @ Bob Ladd:

    Indeed. And the same could potentially be achieved in terms of representing etymology. Systematic spelling doesn't necessarily imply direct one-to-one phonetic spelling (see e.g. French).

    @ Sean M:

    I don't see how that contradicts anything I said. Those wars and insurrections are exactly the sort of disastrous consequences I was referring to, and I don't think these chaotic struggles (and whatever consequences they'll ultimately have on borders) are equivalent to a systematic grand reform. Regardless of that, I highly doubt that the official borders are as irrelevant on the ground as you imply: until peace is achieved and new borders and new states are recognized, they will continue to be a source of conflict.

  23. J.W. Brewer said,

    May 17, 2016 @ 9:33 am

    @GH: One common characteristic of language prescriptivists and crackpots is that they view language as the sort of social phenomenon that can be evaluated in essentially moral terms. I think moral judgment is appropriate (although it needs to be done with humility and appreciation for historical context etc) for social phenomena like legal and economic institutions, at least some cultural practices involving bodily mutilation etc. But I think it's a category error to view e.g. Hawaiian's impoverished phoneme inventory or Mandarin's crappy phonotactics or English's irregular verbs as moral defects. I should perhaps emphasize btw that essentially moralistic praise or defense of a given language is on this account as much of a category error as moralistic criticism or attack would be. (To preempt a possible counterargument, let me just state w/o further elaboration that criticizing a particular feature of a language as "inefficient" or "confusing" or "difficult for learners" as if these were empirical observations free of moralizing is in my view still a moral criticism, because implicitly privileging efficiency etc. over competing desiderata is a contestable value judgment.)

    The perhaps more difficult question here is to what extent a writing system is really a "part of the language" the way that phonotactics or verbal morphology (or let's say for the sake of argument to anticipate obvious objections specialized morphosyntactic rules used only in a specialized register and typically only mastered by the subset of the relevant language community that has practical occasion to need to function in that register) are. S Frankel way up-thread gave the conventional modern view that spoken language is the Real Thing and that a writing system is at best an "add-on." I was taught that as an undergraduate student of linguistics 30 years ago and believed it at the time, but have come to have my doubts about that as time goes on (although it was understandable in context as an overreaction to the contrary claim that the standard written form of a language is the Real Thing). Once a given language community has a long enough historical tradition of literacy plus a high enough current percentage of literacy, the conventional written form of the language and the spoken form start interacting with each other in complicated feedback-loop sorts of ways that make it difficult to completely tease them apart. Of course, the written form, like the spoken form, can and sometimes does evolve, and in situations where there is some degree of digraphia, the balance can shift — sometimes slowly and sometimes quickly.

  24. Eli said,

    May 17, 2016 @ 9:59 am

    I agree with you that written language is not just epiphenomenal, but an interactive part of the linguistic system of adults in literate societies. I don't see how saying something is "inefficient" is inherently a moral judgement. Imagine someone says something like "This writing system takes longer on average for children to learn than this other writing system. That sentence doesn't have to come from a position of "privileging efficiency over competing desiderata" or "moral criticism." The next sentence could be "Despite this inefficiency, this writing system is still the best for this language because of [some competing desideratum]." For someone doesn't value efficiency, saying that something is inefficient is not a moral criticism.

  25. JK said,

    May 17, 2016 @ 10:11 am

    It seems that all sides are making confusing, unsubstantiated statements in this argument that aren't helpful at all.

    Regarding the computing argument, I am not convinced at all that typing alphabetic text is more efficient than inputting Chinese characters given today's computing power. Moser says "And despite the exciting new advances in character entry that Mullaney mentions, even now the most common input methods (straightforward pinyin, wubi entry, etc.) are still cumbersome in comparison with typing in alphabetic text." I can input 知道 by typing "z" "d" "1" in one input method (3 keys), while typing "know" takes four keys. If there is still some inefficiency I haven't noticed, please cite some evidence.

    Moser also says "After all, it took a massive, concerted effort on the part of computer programmers for more than two decades to put Chinese character input on a par with that of alphabetic systems." Microsoft Word was first released in 1982, and I don't believe their programmers have stopped trying to help users improve their input. Just look in the autocorrect tab at how many options there are for tweaking your inputs and correcting spelling. I would also call Microsoft's efforts "massive." These arguments over which text is more suitable for computing strike me as resembling the arguments from a century ago about which great civilization is more advanced.

    Finally, with regards to the human learning literacy argument, I think it is also time for all sides to start citing some data rather than anecdotes. It may also be helpful to remember that literacy in reading pinyin will be different than literacy in reading Chinese characters, and both the gains and losses must be evaluated in making a switch. There is a ridiculous volume of texts written in Chinese characters, and not having access to this is what would be lost in the switch. Whether it is valuable for ordinary Chinese people to access these texts is the key to the issue.

  26. GH said,

    May 17, 2016 @ 10:38 am

    @J.W. Brewer:

    Well, I disagree with hardly any of that. But be that as it may, the written form is nobody's native tongue, so I think arguments about inefficiency, lack of consistency, difficulty of learning, etc. are not problematic in the way they would be for a natural spoken language; and if true (and I believe some fairly objective assessments could in principle be made) I think we have to grant them some weight.

    How much weight may be subjective and open to argument. Certainly there are other valid considerations (aesthetics, historical heritage, …), not to mention ones that we might consider less worthy but must still take into account. For example, I've heard that one of the reasons Hangul didn't take off in Korea for hundreds of years was that the literate classes saw it as a threat to their status as a privileged elite.

    While the situation is not quite parallel, we may compare developments in mathematical notation, the greatest and most familiar one probably being the switch from Roman to Arabic numerals, that have often offered obvious improvements. I don't find it a moralizing category error to say that Roman numerals are inferior to Arabic ones for most purposes.

  27. Elessorn said,

    May 17, 2016 @ 12:04 pm

    95% literacy rate

    Before this number, however heavily we discount it, most of Moser's arguments lose force. The comparative efficiencies of alphabetic writing for language tasks that alphabetic natives consider indispensable can be claimed ad infinitum, and of course with much truth. But without demonstration that these same comparative inefficiencies of Sinography are an actual problem for its native users, what does it all amount to? Twenty, even ten years ago, I could understand, but in 2016? Really, what degree of simultaneous achievement in high-volume book production, general educational attainment, scientific advancement, political development, etc., is necessary before the idea that "characters are a hindrance" stops seeming like obvious wisdom?

    I will completely agree that Mullaney's "Orientalism 2.0" is obnoxious to say the least. And while it may be out there, the allegation of garden-variety chauvinism is unfair. But on a more subtle level I still think there is something to these critiques. I would argue that the kind of confidence that balances the successful use of the character writing system by a billion people against one's own intuition as an outsider, and yet picks the latter, is not universal or natural but learned. And three centuries of unabashed undisputed cultural chauvinism certainly go a long way towards explaining why such towering xenocritical confidence is more common in some cultures than in others.

  28. Nick Kapur said,

    May 17, 2016 @ 2:59 pm

    Somehow this post manages to completely overlook/not even mention the piece to which Mullaney was actually responding to, namely, the risibly uniformed piece by Ted Chiang in the New Yorker, which clearly did smack of orientalism. Many of this post's arguments miss the mark by somehow misconstruing the target of Mullaney's attacks as present-day linguists rather than uninformed laymen.

    Moreover, this post blithely accepts at face value the self-orientalizing tendencies of Mao and Lu Xun. Indeed, it's true that they felt that the Chinese script was holding their country back, but it turned out that they were wrong about that. The Maoist regime also thought, using the same faulty logic, that simplified characters would somehow be easier to memorize, which turned out to have no scientific validity.

    The fact is that despite his at times overheated rhetoric, Mullaney is correct that in certain ways Chinese script is better suited to the digital age. It can convey far more information in far fewer characters with far fewer keystrokes than English can. You can say that choosing a character from a drop-down menu is an "extra step." But in many cases that character (or in most cases, two characters) constitutes an entire word. So by that same logic the extra letters in an English word that is say, eight letters long, constitutes 7 extra steps.

    Seriously, try typing some common Chinese phrases and counting how many keystrokes it takes, compared to typing the English translation of the same phrase. It is generally significantly fewer keystrokes in Chinese.

  29. Alex said,

    May 17, 2016 @ 3:56 pm

    Thanks, David Moser, for what seems a comprehensive and definitive response to Mullaney's rhetorical overkill.

  30. GH said,

    May 17, 2016 @ 4:07 pm

    @Nick Kapur:

    Ted Chiang's article is not being overlooked. It was (favorably) addressed here in a post a few days ago. And Chiang may be a layman, but he is clearly not uninformed; he's used ideas from linguistics in several of his stories, for example.

    A direct keystroke-to-keystroke comparison between the input methods is not applicable unless the typist knows in advance where in the list the right symbol will appear (because—to give a very loose account of the keystroke-level model of human–computer interaction—sequences of known keystrokes can be "automated", but if you need to respond to an unknown prompt, that breaks your "flow": you can't "pre-program" your fingers with the next sequence of keystrokes until you have processed the visual input; for many skilled users, the mental operation is an order of magnitude slower than performing a predicted keystroke), which I would assume is not the case for any but the most frequent symbols.

  31. liuyao said,

    May 17, 2016 @ 4:27 pm

    Many interesting points by everyone.

    Social media (Weibo and WeChat), as well as private instant messages (even between two speakers of the same topolect), are predominantly in characters, though I couldn't find hard statistics on that. (I'm afraid the frequent discussion of digraphia, while interesting, may be misleading to those who are not on Chinese social media.) Indeed typing characters is often faster than the full pinyin, not to mention tone marks. Actually, why doesn't any input system provide the simple functionality that prompts pinyin (with tone diacritics) the exact same way they do characters? I wonder if given this alternative people would be more willing to use pinyin.

    Maybe I missed it, but neither side seems to mention that the success of pinyin input would not be possible without decades of putonghua+pinyin education in primary schools all across PRC. On the other hand, people in Hong Kong and Taiwan do not find characters considerably harder to input to resort to phonetic spelling more than occasionally (like full sentences), do they?

    Subtitles – That would be evidence that Chinese can read characters as fast as they hear it spoken, or they wouldn't be able to follow. I don't believe the government is promoting the use of characters (unlike the case of putonghua), and it's all consumer driven. In fact, as Moser admits himself, there have been almost no discussion in China on alphabetization in the last three (or more) decades, just like no one takes Hegel's arguments seriously in the last century. For reference, calls for reverting to traditional characters come up on Weibo almost every year.

    Let me add to the argument a couple of things that characters can do more than alphabets.

    1. Word plays (or "character plays") such as 童鞋 for 同學, 醬紫 for 這樣子, would be dull in pinyin.

    2. Abbreviation in Chinese conveys much more information than acronyms. Consider 中科院 for 中國科學院 vs. NAS for National Academy of Sciences, both three syllables. We do use DNA and WTO, when the Chinese abbreviations are not as catchy (世貿組織 is okay).

    3. The fascination of the general public for archaeological findings is, I suppose, partly due to the familiar characters unearthed (a recent one being the Han tomb in Nanchang). Of course experts are needed to "decipher" what it is saying, everyone can read the individual characters. The average Chinese may not be more literate in classical Chinese than Europeans and Americans in Latin, yet seeing characters in recognizable form arguably has a different effect than seeing a piece of Roman inscription. (I hope this is not discounted as "high culture" such as calligraphy and reading historical documents.)

  32. Leon M. said,

    May 17, 2016 @ 5:42 pm

    I suspect Xiao'erjing was a non-starter for the PRC's decisionmakers because it was so associated with Islam, and the Hui minority.

    They didn't even put it on the paper money, even though they included plenty of pinyin and a number of minority scripts:

  33. Victor Mair said,

    May 17, 2016 @ 11:47 pm

    95% literacy rate

    Zhou Youguang (the 110-year-old "Father of Pinyin") doesn't believe that figure, nor did any of my colleagues at the Language and Script Reform Committee (Ni Haishu, Yin Binyong, Wang Jun, et al.).

    "Zhou Youguang, Father of Pinyin" (1/14/14)

    "Zhou Youguang, 109 and going strong" (1/13/15)

    The 95% figure is grossly inflated, hence is about as believable Chinese "official" economic data.

    "The cost of illiteracy in China" (3/31/12)

    "Mair on Washington Post on illiteracy in China" (5/1/07)

  34. Elessorn said,

    May 18, 2016 @ 12:59 am

    The 95% figure is grossly inflated, hence is about as believable Chinese "official" economic data.

    Thanks for the response. I'm quite sure you're right it's inflated, as I explicitly said above ("Before this number, however heavily we discount it…"). I question the implication that this is due to the writing system, when we have Hong Kong and Taiwan, not to mention Japan, as counterexamples–not to mention the growing list of countries blessed with phonetic alphabets whose educational attainments nonetheless lag even far poorer Mainland China's.

    I argue that there is a deficit of empiricism at the core of this recurring dispute. I can't recall ever disagreeing with anything David Moser or yourself have written describing the Sinographic system. And I completely understand starting from the hypothesis that a script with demonstrable comparative disadvantages at a number of language tasks (ease of accurate transcription, teachability. etc.) seems likely to be a hindrance to its users. But I cannot understand how this hypothesis survives confrontation with the empirical reality. If the contention is that what worked in Taiwan and Japan will never work in the PRC, I would like to see that claim made explicitly, and openly argued. My feeling is that as poverty is reduced, what remains of illiteracy will follow.

    There is no need to feel particularly attached to characters to see things this way, and it need not imply disagreement with your many scathing critiques of them. But while it is all well and good to point out that most people can't correctly write the characters for "sneeze", I don't think it's logically very careful to just assume that this drawback is a problem. And judging from the high literary attainments of the Sinosphere, every reason to dismiss our initial hypothesis that it must be.

    I think a far more productive turn might be, after perplexedly accepting the empirical reality, to consider why this is so. One key factor, from my experience, is that people do not learn characters only from school, but absorb a lot of them passively from constant immersion, and are highly incentivized to do so on a daily and hourly basis, helped along by the fact that they already know the target language being represented. In contrast to the case of an adult foreign student, school learning hours alone are simply not a good proxy for the ways a native to the character system becomes one.

  35. George said,

    May 18, 2016 @ 3:05 am

    J.W. Brewer wrote, way back towards the start of the thread, that "criticisms of English orthography are often taken (and rightly so) as evidence that one is a crackpot". I wouldn't go quite that far myself but calling for 'reform' of English orthography is certainly evidence that one has a tenuous grip on reality. Reform (as opposed to the sort of unplanned, organic change that has given us the minor orthographical differences between varieties of English) can only happen where there is an authority tasked with and capable of imposing that reform. No such authority exists for the English language.

  36. Vanya said,

    May 18, 2016 @ 5:44 am

    @Elessorn. Sadly Professor Mair seems unwilling to ever consider the empiric or statistical evidence. I think it is very interesting that there seems to be no obvious correlation between the internal logic of a writing system and the ability of individuals using that system to become literate or successfully navigate 21st century technologies.

    Your observation about "constant immersion" is spot on. Native speakers of languages like Chinese or Japanese do not seem to be at a significant disadvantage using character based writing systems as long as there is constant passive reinforcement in the surrounding environment. I have noticed this myself living in Japan and China. I was able to achieve decent fluency reading Japanese in 12 months through constant immersion. Living in China I was able to learn 1000 characters and combinations passively within a few months. It is not that hard if you are reinforced constantly.

    The flip-side to this, and to my mind the true empiric disadvantage of character based systems, is that without immersion, character amnesia is certainly a real problem. This means it is extraordinarily difficult for non-native speakers to learn to read Chinese or Japanese relative to learning to read, say, Vietnamese or Turkish. It is also a serious problem for Chinese or Japanese who live abroad. No question that character-based systems put you at a serious disadvantage if you want your language to become an international language of trade, technology or diplomacy. Reliance on characters is why Chinese will never challenge English as a global language no matter how dominant China becomes otherwise, and why educated Chinese will continue to have to learn English in the future. On the other hand, for many Chinese and Japanese the fact that their writing systems create large barriers preventing foreigners from seeing into their culture is a feature not a bug.

  37. George said,

    May 18, 2016 @ 5:50 am


    "[F]or many Chinese and Japanese the fact that their writing systems create large barriers preventing foreigners from seeing into their culture is a feature not a bug."

    A friend of mine has described this as a grande muraille linguistique.

  38. Victor Mair said,

    May 18, 2016 @ 6:24 am

    "Sadly Professor Mair seems unwilling to ever consider the empiric or statistical evidence."

    Huh? How many times have I talked about the hours upon hours of rote memorization Chinese children and foreign learners put into mastering the characters? How many times have I addressed the problem of character amnesia? What do you want me to do? Stand on my head?

  39. Vanya said,

    May 18, 2016 @ 7:20 am

    I would like you to explain why Chinese children spending hours upon hours on rote memorization is a problem. Apparently to you it is self-evident, but not to me. Do you have empiric evidence that by age 18 Chinese, Taiwanese or Japanese lag behind their peers from developed countries in educational, cultural or even athletic achievement? Is it your experience that Chinese people are somehow emotionally or intellectually stunted from their hours of memorization? I can only say that I have met countless people from various cultures who grew up with alphabetic writing systems and on average they are not noticeably more literate, insightful or more creative than Chinese or Japanese I have known.
    Rote memorization may even be a useful skill in and of itself, whether its characters, poetry, or the Koran.

    I don't disagree with you that character amnesia is a serious issue, but I suspect it is a real problem only for a small minority of people who truly care about expressing or reading original and complex thoughts. For 99% of the population a forgotten character here or there doesn't matter much in daily life. I am surprised that there isn't more evidence showing that Chinese intellectual thought is constrained by the difficulty of learning enough characters to join the intellectual elite, but admittedly that might be hard to demonstrate empirically.

    I agree with you completely about the ridiculous barriers characters present to foreign learners, but again, the fact that foreigners struggle may or may not be a problem for native speakers and is not necessarily a reason to reform your language.

  40. Victor Mair said,

    May 18, 2016 @ 8:25 am

    Tell all that to the Chinese script reformers of the last century and more, including many who are still as active as ever, not to me.

  41. JK said,

    May 18, 2016 @ 8:33 am

    I'm fully on the side of the empiricists here. Here are some more claims I would like to see evidence to back them up:

    Input of Chinese characters is slower than alphabetic input and having to deal with Chinese characters is more of a burden to programmers than dealing with alphabetic input (I don't buy the argument that alphabetic input is a one-step process and Chinese input is two-step. I spend much more time editing my English for spelling, punctuation, and grammar than Chinese, and right-clicking on a misspelled word to choose the correct one from the drop-down menu is a two-step process)

    "The countless hours of time it takes to attain basic literacy in Chinese due to the sheer number of characters to be memorized." — This here I think is the center of the whole debate, that is, how long it takes to learn characters and attain "basic literacy"

    "The enormous cognitive burden of the reading process, which is due to the lack of phonetic information in the script." — This is where the arguments starts slipping into Orientalism, as it suggests the act of reading Chinese characters itself even for a fully literate person is such a mental burden that Chinese people are already at a disadvantage compared to the rest of the world. This should probably be rephrased as "the enormous cognitive burden of achieving a high level of literacy"

  42. Victor Mair said,

    May 18, 2016 @ 11:51 am

    "I'm fully on the side of the empiricists here."

    So am I!

    I guess it depends upon who you think "the empiricists" are and what the nature of empirical evidence is. (See below for the definition of "empirical".)

    Somebody asked whether there is a demonstrable differential or deficit between students who receive their education via characters and those who receive their education via alphabetical scripts. One way you can measure this is by the hundreds of thousands of students who flock from character education settings to alphabetical education settings, not just in college and graduate school, but increasingly in high school and even middle school, versus the tiny handful of those who go in the opposite direction. Why are they (and their parents who foot the expensive bill) opting for an alphabetical education? Don't tell me that it's because the alphabetical societies are richer (China is plenty rich by now). There's something about the intrinsic nature of the education that is imparted in the two systems. This is quite evident to those of us who have taught hundreds of students who were educated in China, Taiwan, Hong Kong, and Singapore, vs. Chinese who received the bulk of their education in alphabetical societies. And it is not just a matter of English skills, because lately some students from China have decent ability in English. Rather, it is the way students brought up in the two systems THINK (or don't think, as some of my colleagues would put it).

    Naturally, there are all sorts of degrees of education that is partly or wholly in characters in the various countries of the Sinosphere. Students from Singapore, for example, often receive the bulk of their education via the alphabet, and it tells in the way they reason and express themselves.

    Then there's the matter of native Chinese speakers preferring to do e-mail, etc. in English because it's "easier", "faster", "less troublesome". I encounter this all the time, not just in business settings, but even among members of the same family or among friends. Often they will start out in Chinese, but then switch over to English when they encounter a stumbling block or get frustrated because they can't produce the character they're looking for.

    See "Language notes from Macao and Hong Kong" (6/22/14), especially the part where I describe my encounter with a woman at a Starbucks in Hong Kong. When you multiply that sort of experience dozens of times, you begin to amass a considerable body of evidence about the actual preferences and usages of Chinese in the cybersphere.

    I've written about such phenomena hundreds of times on Language Log, so the data are there for those with a truly empirical ("based on, concerned with, or verifiable by observation or experience rather than theory or pure logic") cast of mind.



    As for "Orientalism", usually it amounts to little more than name-calling.

  43. Guy_H said,

    May 18, 2016 @ 1:10 pm

    A rather late comment, but to what extent is the widespread use of hanyu pinyin in Mainland China an accident of history?

    If the Nationalists had won the civil war, surely we'd all be using bomomofo instead, and "romanization" wouldn't even be an issue.

  44. Elessorn said,

    May 18, 2016 @ 1:32 pm

    @Victor Mair

    I can't understand why you would feel comfortable putting much weight on testimonials from the incredibly unrepresentative sample of Chinese who are comfortable enough with English to have a choice between the two languages in settings ranging from friendly to professional. But more importantly, this is an entirely different question. The issue here has been the alphabetic representation of Chinese. And here we have a clear natural experiment taking place daily and hourly, no? What percentage of free, voluntary communication in Chinese between mainland Chinese speakers online is conducted in pinyin? Is it even 10%? 5%?

    I'm not sure what to say about the claims that the huge wave of Chinese students studying abroad is driven in part by the superior thinking skills allegedly imparted by alphabetic education. Do you really believe this? I don't want to misread, but by the same logic it seems we would be able to conclude that Hangul and Devanagari are also at a disadvantage against the Latin alphabet. And where would this stop? Given the wide diversity of linguistic backgrounds among foreign exchange students, the wide variety of countries where a foreign English plays important roles in administration and higher education, and the fact that this is all happening at the same time that the most politically and economically dominant nation happens to be the Unites States, I think the most natural takeaway would be that English occupies a superior political and economic position now, not that English is a superior medium for politics and economy. Was French also a superior medium? Or Latin before it? Certainly partisans of every dominant tongue have had no trouble finding transcendent justifications for their languages' temporary popularity in the past, but I can't believe you intended such an implication,

    Either way, the causal link offered between high levels of study-abroad and script difficulty is not empirically derived. What standard could allow one to make objective judgements about the thinking skills imparted by different scripts? What historical precedents make us believe that such a correlation is possible?

  45. Victor Mair said,

    May 18, 2016 @ 1:32 pm


    No, that "surely" would not necessarily have been the case, since the Nationalists also had a very elegant romanization system that was espoused by the best linguists of the day. I'll write about this in my next post, which will probably go up within 24 hours.

  46. Victor Mair said,

    May 18, 2016 @ 3:36 pm

    There is yet no precedent, precept, practice, or pattern for writing MSM on a broad scale in romanization, whether sanctioned by the government or not. That is gradually arising through the emerging digraphia driven by near-universal alphabetical inputting, pinyin pedagogy, and all the other usages of romanization that we have repeatedly discussed on Language Log. Meanwhile, English offers an alphabetical alternative to writing MSM in pinyin. Other Sinitic languages have long since been written in integral phonetic scripts. That will be the subject of my next post.

    Meanwhile, to take summary stock (this is by no means a complete account) of what has been happening with writing in China during the last century or so:

    1. script reformers invented hundreds of phonetic alphabets

    2. China abandoned Literary Sinitic (Classical Chinese) for the written vernacular

    3. continued proposals for phonetic scripts (will address some of these in my next post, and already have mentioned many of them in previous posts)

    4. calls for the abolition of characters and their replacement by romanization from leading intellectuals and political figures

    5. the promulgation of Hanyu Pinyin as the official romanization of the PRC

    6. extensive simplification of the character writing system

    7. calls for further simplification of the characters

    8. retraction of calls for further simplification

    9. repeated experimentation with radical reshaping of the character writing system by artists such as Xu Bing

    10. integration of the alphabet within the character writing system (as chronicled by Liu Yongquan and described by Mark Hansell)


    Please note that all of these developments in the writing system have been the result of internal dynamics within China and have been carried out by the Chinese themselves. The rest of us (we outsiders) are just observing and recording what's going on.

    To conclude this comment, what we see is a two-pronged spread in the use of the alphabet in China:

    1. with the tremendous growth of English (there are now more users [N.B.] of English in China than in any other country, including the United States)

    2. with the emerging digraphia for writing all or parts of various Sinitic languages, including, but not limited to, MSM

    More on that tomorrow.

  47. Jeff said,

    May 18, 2016 @ 3:52 pm

    I'm wondering, and this is just a hypothesis without statistical proof, if the gap between the vocabulary used in Chinese writing (especially in academic and literary works) and spoken Chinese isn't widening these days. Is Chinese writing (necessarily in characters) that is composed by elites and would-be elites taking a different, "more flowery" direction these days than during the old vernacular "literary revolution"? (There was already a divergence between, say, the vocabulary of Chinese writing in Taiwan and that of the mainland after 1949. –which is a different question from divergences in the spoken languages, which also existed.) If written Chinese is becoming more "literary" (in an old sense of the word), does the continued use of characters to write these words (seldom spoken words?) have an impact? Just a question.

  48. Victor Mair said,

    May 18, 2016 @ 8:50 pm

    Excellent thought, Jeff! I think that what you have said makes a lot of sense with regard to writing, including choice of diction and types of characters that are favored. At the same time as there are powerful forces for change and innovation, especially among young people and progressive intellectuals of all ages, there are also strong tendencies toward conservatism, particularly among older individuals who may have had more exposure and attachment to earlier styles of writing than most literate citizens.

  49. David Moser said,

    May 18, 2016 @ 9:37 pm

    Re: empiricism. This is a reasonable point. There is actually quite a lot of empirical information that can support claims about the cognitive burdens of the script (the sheer amount of time it takes Chinese kids to master the system, all the way into middle school, as just one example), and this could be the subject of another article entirely. But it's true that few scholars have carried out specific tests to measure the differences. Such surveys are tricky to devise, fraught with problems of cross-cultural incommensurability, and probably hard to find funding for. So we're stuck with a pile of observations, good inferences, telling statistics, and first-hand experience. I think it adds up to a convincing body of evidence, but of course it's not going to satisfy the most rigorous of empirical demands. Our situation is somewhat like this: Suppose someone makes the claim "Traffic jams in Beijing and other large Chinese cities are a plague. They waste countless millions of man-hours, reduce the effectiveness of the work force, and are a burden to human resources and budgets. We should take steps to reduce traffic jams." And then suppose the hard empiricists say "China has the highest GDP in the world, the world's largest car market, companies like Alibaba that rival any Western country, and a fast-rising middle class. So evidently traffic jams are not affecting efficiency and productivity. Otherwise how do you explain all this burgeoning economic activity and wealth creation? Show me some statistics that prove that traffic jams impede productivity!" It may be a flawed metaphor, but you get the point.

  50. Kratoklastes said,

    May 19, 2016 @ 12:13 am

    This was a terrific counterpoint to an abysmal article that I had already bookmarked into my 'Evidence of Dunning-Kruger Effect' folder.

    That said, it would have been useful to point out that literally every line of programming that runs literally every website in the world, depends on English language keywords. Try writing a shell script in Mandarin (or Turkish, or French, or Indonesian) and see how far you get.

    It would also be useful to point out that 'accepted' English orthography is a relatively recent invention, and that the putative watchmen of orthographyu – the major dictionaries – by and large repudiate the idea that there is such a thing as korrekt authograffy.

    And lastly but not leastly – if we nutters who deplore korrect-lyne spellyng ideologges and gramma nartzies get our way, it won't just be the orthography that changes: we can do away with some of the idiotic conjugational and other-grammatical nonsense that we retain for no good reason whatever.

    Obviously we are not as bad as the French ("vert, verts, verte, vertes"; the fact that my thigh is feminine even though its DNA isn't; and the mare's nest of the subjunctive, among other horrors), but we could simply use the infinitive for the present and nobody would care – e.g., the old rural West Country "I be" and "I bain't" (be not).

    Many verbs be like that already, and it be only the core verbs that bain't and many of those be close: get rid of 'has', 'does' and 'goes' for 2nd person present, and 'to have', 'to do' and 'to go' be fixed rightwise: imagine how hard it would be to similarly fix the French versions of those verbs!

    The beauty of English be its modularity and the fact that it evolves endogenously, not the fact that we cling to some aspects of stupid Continental/Latin-derived grammar.

    At least we dispensed with gender- and number-dependent adjectives, but we retained dependent preposition/pronoun and determiner/noun patterns (the Devil take them both: "this dogs" and "that cats" ought to be fine, since there is enough information in the plural noun).

    We also got rid of silly accents – for the simple reason that we don't have a bunch of pompous jobsworths trying to 'manage' the language (I be lookin' at you, Academie Française).

    I cannot imagine a person of median intelligence being able to learn to both read and write if the minimum alphabet for moderate literacy requires memorisation of 3000 symbols with literally zero information content.


  51. Vanya said,

    May 19, 2016 @ 3:06 am

    No, that's a decent metaphor. You are making my point. Traffic jams suck, and no one would want to create a society that includes them but also traffic jams are not as big an impediment to overall productivity and efficiency as you might think when you are sitting in one and being miserable. The economic data is pretty clear, as you say. Would it be nice to have China's GDP growth without traffic jams? Of course, in a perfect world, but Chinese cities are already "locked in" to certain street patterns and traffic flows because of the way they were built before automobiles became readily available. Everyone would rather not have traffic jams, but how much of the existing city are you willing to destroy to make peoples' commutes easier, and is traffic flow really the main point of living in a city? There is a long history of well-meaning but single-minded reformers trying to fix traffic jams in other cultures who just made things worse, such as destroying historic city centers by ramming freeways through them (see Boston, LA,Cincinnati, Cleveland, Detroit, etc. ), paving over wetlands in southern Florida with 16 lane highways, ripping up sidewalks in suburban areas, etc.. There is also a significant cost to building more public transport, and finallly individuals do place a value on having their own cars. The urge to fix things that seem obviously wasteful and inefficient is understandable, but if history of the 20th century should teache us anything, it is to beware idealistic reformers.

  52. Chris Coulouris said,

    May 19, 2016 @ 5:09 am

    "Learning the Chinese language requires bodies of iron, lungs of brass, heads of oak, hands of spring steel, eyes of eagles, hearts of apostles, memories of angels, and lives of Methuselah." The missionary Thomas Milne said this in the early 19th century. 200 years later and it is still hard to disagree with him.

  53. Alan Chin said,

    May 19, 2016 @ 5:11 am

    Surprised that no one has brought up the Chinese script as a cultural and political unifying factor through Chinese history: Because there is no one spoken language called "Chinese", but rather a group of Sinitic languages of which MSM is more or less the same as post-1919 written vernacular Chinese. You can certainly write in Cantonese, although even in HK this isn't truly accepted, especially by the older generation. But the larger point is that if Latin were a character rather than alphabetic written language, modern French, Spanish, Italian, Romanian, etc. speakers might all still share the same, or almost the same, written language called "Latin", rather than evolve as separately as they have.

    And same for "Old Slavic" and Russian, Bulgarian, Ukrainian, Serbian, and the other Cyrillic script languages. And so on.

    And then it's a fair question to ask, if the much-slower-to-change-character-based script of written Chinese hasn't help maintain a unitary political and cultural identity of "Chinese". Or is it chicken vs. egg? We see how quickly the language and script diverge of PRC vs. Taiwan vs. HK vs. Singapore with traditional/simplified, standard written Chinese/written Cantonese, and so on, when the politics are divided.

  54. Elessorn said,

    May 19, 2016 @ 5:18 am

    @Vanya (earlier comment)

    Exactly! I can heartily agree with every word. There are clear empirical disadvantages to Sinography, but they all kick in only in a non-Sinographic environment. They are also disadvantages that, from my experience, are experienced quite differently by native and non-native users of the script.

    This is why collocations like "rote memorization by Chinese and foreigners" are absolutely specious. I'm sure there are borderline cases of users who have some elements of native character facility, but not all, and not in the same degree, but even very character-fluent foreign users have a control that is different in kind from that of native speakers who may be much less knowledgeable.

    A good corollary in English might be the way that natives seem to be on average much better at reading the underlying sentences in that "scramble all the letters and you can still read it" meme a few years back. There are a lot of things that most native readers can do with little effort that it is devilishly hard for second-script readers to do with equal facility. Anyone who has lived in a Sinographic environment, for just one example, will have seen the phenomenon of speakers rapidly tracing out on the palm, or less commonly in the air, a character that the listener failed to get from pronunciation alone. And I do mean rapidly. This is expecially common with names, or with characters uncommon enough that the speaker can't use the usual standby of saying "it's the character X as in word Y." On that note, even "character amnesia" is a slightly unfortunate term, unless English speakers pick up on the nuance that the amnesiac has not actually forgotten, but has simply failed to recall, how to properly write the desired character. Unless it involves a character that is actually uncommon, the speaker would almost always be able to read the same character without effort.

    As you said, environment is everything. As far as I can tell, most character natives raised in character environments experience what has been here played up as "rote memorization" as a chore, certainly, but one on the level of handwriting or spelling classes for Ameican children. After all, in most character environments today (N. B. quite unlike in premodern times) they know from experience that (1) essentially everyone succeeds at becoming literate (2) they can enjoy reading character texts with little effort even if their handwriting is terrible and their Chinese grades are atrocious. Because characters are not learned from school alone.

  55. Ed Connelly said,

    May 19, 2016 @ 8:35 am

    "pinyin alone would be sufficient to achieve basic literacy, if only Chinese books were printed in this imported phonetic writing system." Basic literacy, perhaps,where spoken and written Chinese coalesce. Given the classical (文言文) style in which huge quantities of Chinese literature is written, however, the vast majority of Chinese literature cannot be Romanized and understood because of the differences between the spoken and written Chinese languages. For example, some twenty years after he published his masterpiece "A New Discourse on the Uniqeness of Consciousness" (新唯识论) in Classical Chinese, the philosopher Xiong Shili (熊十力) allowed his students finally to produce a vernacular version. Also, Chinese schooled only in simplified characters (简体字)have great difficulty reading works from say the Buddhist Canon (佛藏经) not reproduced in simplified characters.

  56. Victor Mair said,

    May 19, 2016 @ 9:00 am

    "rote memorization" as a chore, certainly, but one on the level of handwriting or spelling classes for Ameican children.


  57. Elessorn said,

    May 19, 2016 @ 10:24 am

    @Victor Mair


    You will disagree, of course. And of course it's a harder chore that goes on a lot longer. That said, I think this way of thinking of it is truer to the experience of native speaker pupils of the target language.

    For foreigners, every character is new, the vocabulary is weak and insecure in memory, and characters occupy only a few hours of class and homework. And while they may be forced to copy out each character dozens of times, in the end the number of times they use each character expressively is much smaller, and the number of times they need to use a character close to zero, though there may be classroom ways to simulate it.

    Natives in China or Japan are a different matter. As opposed to pre-modern times, they live in a world saturated by characters, have constant need, pressure, and incentive to be able to use them, and of course speak the language this sea of orthography is representing. In school, characters are not only taught themselves, but are used to teach everything and anything else. Characters are also a gateway to many forms of leisure–teading for pleasure, browsing the net, texting with friends, etc.: the list goes on and on, and with increasing penetration of electronic text into daily life at earlier and earlier ages, grows ever longer. And all this is amazingly successful– we teach math for ten or twelve years, but most people forget most of it, and even brilliant people are not necessarily good at doing calculations in their head. But reading of characters by literate adults in Sinographic societies, as you well know, is done as effortlessly as the reading of alphabetic texts in the U.S.

    This is not to say that script reform is wrong. That is a choice for Chinese and Japanese users to make, and right or wrong don't enter into the picture. That alphabetization would have drawbacks but also benefits is also clearly true. But that script reform is necessary for Chinese or Japanese speakers, or its lack holding them back, is clearly, empirically false.

    I would never try and argue you out of your support for script reform, and especially in the language learning classroom, agree wholeheartedly with your proposed remedies. But the empirical reality is the most important thing for any student of language, and love or hate them, the reality of characters as actually lived is not, I urge you to consider, accurately conveyed by one-sided insistence on school hours committed to character instruction, or by unqualified calls to some "cognitive burden." Would you really want non-specialists reading here to think that on a crowded train in Tokyo of people silently buried in novels, newspapers, or smart phones, the reading act is actually always accompanied by a steady cognitive strain that everyone has just tragically grown numb to?

  58. Vanya said,

    May 19, 2016 @ 10:53 am

    "rote memorization" as a chore, certainly, but one on the level of handwriting or spelling classes for Ameican children.

    Well, I agree with Professor Mair that it is more of a chore than English handwriting and spelling. My objection is that the time that the Chinese waste on characters is generally not used in other cultures to encourage creativity but in most mediocre school systems is usually just filled with some other rote activity, whether workbooks, memorizing bad poetry, state capitals, the names of rivers and provinces, copying pictures, etc. It takes a lot of effort and imagination to engage children, if Chinese teachers can easily use up several hours every day by simply forcing children to copy out characters then that is a bonus for the mediocre teachers, and they are not going to be enthusiastic proponents of reform.

  59. Elessorn said,

    May 19, 2016 @ 11:44 am


    Well, perhaps a better metaphor is needed. But the key point, that characters are not just disconnected bits of information thrown out there to be memorized for their own sake, is solid. A native might easily say, "I like reading books, but I'm terrible at kanji." For a foreign learner the same statement would be nonsensical.

  60. Bruce Kowal said,

    May 19, 2016 @ 2:41 pm

    Given the success in Vietnam with Romanization, and observing that the intelligibility of the that systems is due to the availability of more distinct morphemes – – more tone, and consonants – – I wonder if, for example, speakers of Cantonese or the Min dialects, would be able to use a Romanized alphabet with minimal problems, when contrasted with Putonghua and it cursed homonyms. Maybe, humorously, Cantonese or Fujianhua should be the new Putonghua, in which case we can make the switch and go on happily!

  61. Jeff W said,

    May 19, 2016 @ 10:28 pm


    But the empirical reality is the most important thing for any student of language, and love or hate them, the reality of characters as actually lived is not, I urge you to consider, accurately conveyed by one-sided insistence on school hours committed to character instruction, or by unqualified calls to some "cognitive burden."

    I agree wholeheartedly.

    It seems perfectly plausible to me that, even with the enormous time invested in learning Chinese characters, even forgetting a character they might want to input, even using English in texting, adult native speakers of Chinese don’t have enough of a problem with the characters that they think using an alphabetic system is necessary or even desirable. I’m not saying that is the case—I’m saying it seems to me that none of the evidence here necessarily presents a different case.

    I have friends in China, all native speakers, and they also write in English to each other because, they say, it’s “easier.” (One or two also say it’s “cooler.”) But that—to me—doesn’t mean that they would necessarily prefer some alphabetic version of Chinese. In fact, one friend once said to me flatly, “That would be horrible.” (And that doesn’t mean that, on the whole, Chinese people feel the same way. He’s one guy.)

    I’d like to know, along with Elessorn, what the empirical data says about “the reality of the characters as actually lived.” I’d like to know, from the adult native speakers’ point of view, what they think the issues raised here—of, say, the “enormous cognitive burden” of reading Chinese characters, if any, and, even if there is one, would they prefer some alphabetic orthographic system instead or, if they wouldn’t, why not?

  62. Victor Mair said,

    May 19, 2016 @ 11:24 pm

    @Alan Chin

    …the Chinese script as a cultural and political unifying factor through Chinese history and so forth

    We take all of that for granted, and have discussed it numerous times on Language Log. The question the language reformers ask is whether such conservatism also has it negative aspects, such as holding China back from timely and full modernization and from competing with the West on an equal footing.

  63. liuyao said,

    May 20, 2016 @ 12:28 am

    Would this be a better analogy?

    Despite that the US continues to use feet and pounds and gallons and miles, with all the crazy conversion factors that do not conform to the natural way we use numbers (in base 10), there does not seem to be any measurable effects on people's ability to conduct business and trade, and science and technology, in all of which the US reigns supreme. How much time is wasted in school on those mind-numbing exercises that none of the other countries have to deal with? May we say that computers (and google) saved the American units?

    Another analogy (more related to languages): Among all the languages in the world, what are some of the worst way to say numbers? 98 in French? I wonder if Chinese has the best (simplest, most logical) system, and if there have ever been "number reforms" (in languages, not switching form Roman to Arabic numerals) recorded in history.

  64. ohwilleke said,

    May 20, 2016 @ 12:47 am

    I think that it is fair to conclude that the English language itself is slowly and gradually becoming more "logographic" although a phonetic alphabetical system will almost surely remain predominant.

    I probably routinely use or read dozens or scores of logographs in my daily life, from @ to & to section and paragraph marks to the biohazard symbol and so on. Just in the last month or so, a gender neutral restroom logograph (half male on one side and half female on the other) has gone from a twinkle in someone's eye to a widely understood logograph.

    This may level off at some point (perhaps a few hundred), and some logographs either lose currency or lose old meanings (few people still use the # symbol to represent "pounds"), but I could easily see a near future where several hundred English logographs are in wide use both in English speaking communities and beyond them.

  65. Vanya said,

    May 20, 2016 @ 2:20 am

    The question the language reformers ask is whether such conservatism also has it negative aspects, such as holding China back from timely and full modernization and from competing with the West on an equal footing.

    That was a fair question in the 1890s-1950s, but the question has been answered conclusively over the last 30 years, and the answer is no. The Chinese script is certainly sub-optimal in many ways but one look at places like Brazil, Argentina, Egypt, Russia or even Vietnam demonstrates that a cumbersome writing system is just not a very serious issue compared to all the other factors that can derail economic development.

  66. Vanya said,

    May 20, 2016 @ 2:31 am

    Among all the languages in the world, what are some of the worst way to say numbers?

    The German/Dutch/Danish way. In a modern environment it is very cumbersome to say numbers in the reverse order of the way you write them (e.g. vierundzwanzig), especially when trying to write down phone numbers.

  67. Elessorn said,

    May 20, 2016 @ 5:36 am

    The focus on school-fostered character practice in particular is a bit myopic. The effortless ease with which even very overworked, underpaid laborers will–voluntarily!–spend their breaktime reading a newspaper, or their phone, should tip us off to the fact that our hypothesis of characters as cognitive handicap, though reasonable enough, is missing the mark.

    But even if we focus on school alone, if character education were such a time-costly drag on educational achievement, and if reading with them–particularly for children still learning them–were such a burden, then the negative effect should be multiply compounded, showing up in every subject, all of which are taught… in characters. To me, at least, no such headwind on Chinese or Taiwanese or Japanese attainment seems evident.

    I agree that it seems to be a miracle that children can master several bodies of abstract knowledge at once in a script which has yet to be mastered itself, all this for many of them in a foreign topolect which also has yet to be mastered. Nonetheless, it happens regularly, and rather than insist it doesn't, wouldn't it be more interesting to explore why and how it does?

  68. Victor Mair said,

    May 20, 2016 @ 6:51 am

    …collocations like "rote memorization by Chinese and foreigners" are absolutely specious.

    That is a categorical claim / verbiage without data or evidence to back it up.

  69. Victor Mair said,

    May 20, 2016 @ 7:16 am


    Very interesting observations!

    But, when you wrote "This was a terrific counterpoint to an abysmal article that I had already bookmarked into my 'Evidence of Dunning-Kruger Effect' folder", were you talking about the Orientalist critique of the NYer piece on the uninvention of characters, or David Moser's post pointing out the backwardness of the former?

  70. Victor Mair said,

    May 20, 2016 @ 9:30 am

    From a former student of mine who is now a professor of Chinese history at a major university:

    Under the subject heading: Language firestorm

    I stayed up late last night reading Mullaney's article on Chinese in "Foreign Policy," David Moser's response in Language Log, and all the discussion on Language Log and on the Sinologist's page on Facebook. Perhaps you've already seen: Mullaney posted a sarcastic "apology to the alphabet" on Facebook and his Twitter account, actually an incredibly rude response to Moser.


    I just wanted to write to you to thank you for teaching me well, and giving me a solid understanding of the Chinese language. I'm surprised that no one has mentioned your "Characters and Computers" in the debate.

    I feel sorry for readers of "Foreign Policy" who don't have a basic understanding of Chinese. Mullaney writes that you're racist if you think the idea of a giant typewriter is funny, without explaining at any point how such devices actually worked or how we should understand the Chinese language. Such a missed opportunity.

  71. liuyao said,

    May 20, 2016 @ 10:07 am

    Speaking of "backwards" numbers, everything from date format to street address, are in the "wrong" order.

    Perhaps the US partial correction of date format is like simplified characters; it doesn't solve all the problems and add to the confusion, but we are stuck with it.

  72. yien said,

    May 21, 2016 @ 3:56 am

    Some key metrics, which don't seem to be in the discussion:

    -hours of literacy instruction versus any given literacy level (aka –what could one possibly do with an extra 3000 hours of their life?),

    -hours of literacy instruction versus time to get to the “reading for pleasure level” (aka –sunken time before kids start reading Harry Potter/Condor Heroes, if at all),

    -actual reading for pleasure rates (aka –percentage of kids that enter into a “I must read Three Body Problem/Lord of the Rings” craze/phase),

    -how well does the script apply to science and technology progress? And is it a preferred script for world tertiary education.

    Given the rise of “pinyin over hanzi” reading apps such as lingq, mandarinspot or even pleco reader’s tabbing function – and not to mention the entrenchment of pinyin input as the absolute dominant writing system for Mandarin – a key choice that currently faces intelligent western adult learners of Mandarin is — why put in an extra 2000-3000 hours of study for no real added benefit in the modern world?

  73. Elessorn said,

    May 21, 2016 @ 4:38 am

    That is a categorical claim / verbiage without data or evidence

    On the contrary, the evidence is overwhelming, or one might more aptly say the situation is clear. Foreigners learn lists of characters, representing a language they don't speak, in an environment that provides no reinforcement, positive or negative, to use that knowledge outside coursework. Needless to say they don't also study other subjects in these characters, or read books in these characters, or depend on these characters to navigate every public setting, or…

    Chinese speakers use characters to do all these things, the characters are constantly reinforced, and prove immediately useful and necessary, and are passively and actively drilled in every walk of life. Rote memorization is not the proper term, I think anyone would agree, for a body of information so immediately vital and useful under current conditions. At the very least there is no comparing the learning situations of foreigners and natives in a character environment.

    Borderline cases of course exist. Foreigners learning in China will tend to pick up characters faster as a rule, and natives learning characters abroad will often end up like…Ted Chiang. In either case, these exceptions prove the empirical principle that character environment makes all the difference, and that to focus on hours spent copying alone is necessarily misleading.

  74. Victor Mair said,

    May 22, 2016 @ 8:18 am

    …to focus on hours spent copying alone is necessarily misleading.

    Wrong on two counts:

    a. Nobody is focusing on "hours spent copying alone". We have been discussing numerous other factors involved in the learning and use of Chinese characters.

    b. Chinese children do spend countless hours copying the characters over and over and over again. I know this from observation, having lived in China, Taiwan, and Hong Kong for many years, in close relationships with Chinese families. By and large, the children in Singapore devote less time to character learning, with the result that they generally do not master the script as well as the children in China, Taiwan, and Hong Kong (the latter being a mixed case, depending upon what sort of schools they go to). On the other hand, the children who spend less time learning characters tend to acquire more types of other skills, abilities, and knowledge.

    It is unfortunate that a few participants in this debate have resorted to using "empirical" as a kind of meaningless shibboleth, just as several others have been employing "Orientalism" as a form of name-calling. Neither of these usages serve to advance the discussion in useful ways.

  75. Elessorn said,

    May 22, 2016 @ 7:34 pm

    @Victor Mair

    Thank you for your response, which I would say has moved the conversation forward.

    Wrong on two counts:

    On the second count, hardly. Elsewhere and in the very phrase you quote ("…to focus on hours spent copying alone is necessarily misleading"), I unambiguously refer precisely to Chinese and Japanese students' extensive copying and drilling of characters. Who would deny it? This is purest fact, disputed by none.

    The conclusion you draw from this fact, however, that "on the other hand, the children who spend less time learning characters tend to acquire more types of other skills, abilities, and knowledge," is merely a hypothesis. It makes sense, sounds reasonable. But if common sense and intuitions were enough, I submit that we wouldn't need philology.

    In this thread and many others, you have repeatedly been entreated in respectful terms (in contrast to Mullaney's attitude) by several posters to provide evidence that the character-literate adult population constituting a significant portion of humanity seems, in contrast to their alphabetic peers, lacking, stunted, uncompetitive, unaccomplished, disadvantaged, etc. I think it is fair to say that no empirical evidence has been provided.

    On the first count, however, it is true that I did not mention your emphasis on "character amnesia." I think it is fair to say that the hours "lost" in school is the core of the argument, but it is an omission on my part nonetheless, however I may dispute its significance as a stumbling block in day-to-day life. I should have been more complete. (I hope you will allow me to pass in silence over the sometimes implication that alphabetic natives are better at thinking.)

  76. Vanya said,

    May 23, 2016 @ 3:03 am

    On the other hand, the children who spend less time learning characters tend to acquire more types of other skills, abilities, and knowledge.

    Well, that should be testable. At least anecdotally, comparing Western children to Chinese, I see no evidence that that is the case. Even comparing Singaporean children to Taiwanese, that seems like a dubious proposition. Taiwan produces world class professional baseball players, musicians, film-makers, businessmen, etc. In my experience Singapore, if anything, seems to lag behind Taiwan when it comes to producing well-rounded individuals.

  77. Mark Csikszentmihalyi said,

    May 23, 2016 @ 5:42 am

    Alan Chin's remark about the unifying effect of the script deserves more consideration. More broadly, the way this whole kerfuffle arose from a piece based on a historical counter-factual conceit has resulted in bracketing practical issues like the effects of a reductively phonetic script on diverse spoken languages. The problem is that Mullaney's response blurred a distinction between two questions. The Chiang piece encourages us to think about "what if there had never been…" but the practical impact of such a consideration quickly starts morphing into "what if we changed, or will change things to…" And to me, it is a very different question to ask: 1) what if Zhuangzi had originally written "you yi zihao ye er pi qi suo buwei zhe ye" rather than 有以自好也而吡其所不為者也, compared to 2) what if he wrote the latter but now we all will switch to alluding to it as the former? One is relevant in a context of speculative science fiction as a thought experiment, but fundamentally differs from throwing around numbers about the impact of May Fourth era or contemporary proposals on the virtual or actual lives of modern schoolchildren.

  78. Victor Mair said,

    May 23, 2016 @ 6:27 am

    …seems like a dubious proposition….

    …if anything, seems to lag behind….


  79. Victor Mair said,

    May 23, 2016 @ 7:26 am

    Mark Csikszentmihalyi has made some valuable contributions to the current debate. He's one of the few participants who understands what Ted Chiang's piece is all about and the practical implications it may or may not have for actual efforts toward language reform of the last century and more (even before the May Fourth movement there were plenty of proponents for script reform in China — I have a section on that in the Columbia Sources of Chinese Tradition).

    The unifying effect of the Chinese script is well-nigh universally accepted. In contrast, look at what happened in Europe. I encourage everyone to take a good, hard look comparing China with Europe in terms of language, writing, politics, law, religion, culture, science, and virtually all other aspects of civilization, not least literature.

    I love the Zhuang Zi and translated the whole of it with much joy and zest, but I would be the first to recognize that presenting it solely (without the characters) in a MSM transcription just wouldn't work. However, if you first rendered it into a full-blown, fluent vernacular (Mandarin, Taiwanese, Cantonese, etc.) and then transcribed that accurately with a phonetic script, it could be read as well as if someone were delivering the translation (i.e., telling the stories in the Zhuang Zi) orally.

  80. James said,

    May 23, 2016 @ 12:34 pm

    "Given the rise of “pinyin over hanzi” reading apps such as lingq, mandarinspot or even pleco reader’s tabbing function – and not to mention the entrenchment of pinyin input as the absolute dominant writing system for Mandarin – a key choice that currently faces intelligent western adult learners of Mandarin is — why put in an extra 2000-3000 hours of study for no real added benefit in the modern world?"

    As someone who is putting in the extra 2000-3000 hours I can give some reasons for wanting to read Chinese characters and of course they will not apply to everyone.

    – It seems to me that knowing the characters is necessary for a fuller insight into Chinese culture; discussion of characters and their components can be part of conversation with Chinese people; they will give insight into the history of China (just as reading heiroglyphs would for a history of Egypt); appreciation of calligraphy is based upon knowing them; etc.

    – some Chinese people respect the effort I make and respond positively to it; they are more accepting as I seen to make the full effort to integrate with the culture. It can also impress others and give one a little extra status because it's a relatively rare accomplishment for Chinese.

    – It's not always convenient to use an app. For example, when I watch Chinese TV to practice listening, having the characters underneath helps me make sense of the sounds (often I know the characters before I am capable of hearing the sound in a sentence.)

    Not all Western adult learners take a strict pragmatic approach to learning the minimum of the language possible to get by; many I've met would like to benefit from the at least one of the above. I think all three points also add value even in the modern world.

  81. Victor Mair said,

    May 23, 2016 @ 2:06 pm


    Although I would not recommend to most learners of Chinese that they spend the extra 2,000-3,000 hours to learn the characters the traditional way when there are so many much faster and easier ways available nowadays, I respect you for making the conscious choice to do so, and you have reasons for adopting the old fashioned method. What I especially appreciate about the manner in which you present your point of view is that you do so without being dogmatic, contentious, accusatory, or snide, and you recognize that others may view the matter differently. Thank you for being so civil!

RSS feed for comments on this post