I have already discussed the Sapir-Whorf hypothesis, admitting that I don’t have the faintest idea whether it’s true or not but hoping for research that might do more to answer the question than the ex cathedra statements of linguists. Imagine my surprise when the NY Times tries to settle the question by consulting the ex cathedra statements of linguists—in this case, one particular linguist, the media’s current darling, Steven Pinker. Now, Pinker is a smart guy (and a brilliant self-promoter, not that there’s anything wrong with that), but he subscribes to a linguistic theory that holds that all languages are essentially cosmetic variants, identical below the surface. Regardless of whether you think that’s plausible or not (regular readers of this site will know that I come down on the “not” side), to ask an adherent of it to judge the validity of Sapir-Whorf is like asking an atheist to judge the validity of the Pope’s infallibility. The Times might (I’m stretching here) ask an atheist for an opinion, but certainly wouldn’t allow it to carry their implicit imprimatur, as they do with Pinker. But this is no surprise; the surprise would be if the newspaper that employs William Safire as its language guru suddenly got a clue about these matters.

By the way, the main focus of the Times article is not Sapir-Whorf but a book that’s just come out, The Writing on the Wall: How Asian Orthography Curbs Creativity by William C. Hannas, “a linguist who speaks 12 languages and works as a senior officer at the Foreign Broadcast Information Service.” Mr. Hannas (to adopt the Times‘s immutable form of reference—they legendarily referred to the singer Meat Loaf as “Mr. Loaf”) claims that Asian science has suffered because the main Asian languages are written in “character-based rather than alphabetic” systems. (Note to the Times: “character-based writing system” is not the same thing as “syllabary.”) According to him (if the Times summary is correct), “because East Asian writing systems lack the abstract features of alphabets, they hamper the kind of analytical and abstract thought necessary for scientific creativity.” This seems unlikely to me. What is likely, though it goes unmentioned in the article, is that the writing systems based on characters, which take much more time and effort to learn than alphabets, hamper literacy in general, and if China (to take the biggest country involved) switched over completely to pinyin romanization, literacy rates would skyrocket (true literacy, not the minimal ability China’s official statistics are based on) and there would be a much larger base of potential scientists and inventors. But don’t hold your breath; there’s a lot of cultural capital invested in those beautiful, inefficient characters.


  1. As I remember Sapir-Whorf, it’s problematic in its implications in the same type of way that this article approaches: arguing for a narrowing (probably better here to say qualitative difference in) of perception and ability because of writing systems, or vocabulary, or past tense formation (or any of the other classic examples that I am dredging up from the memory of my antique linguistics major) is another scientific idea that becomes too easily implicated in ideas of cultural superiority and inferiority. On the other hand, the idea that all languages are essentially, under the surface, the same fed my own sentimental and idealistic notions of the unity of the human race way back when. Where’s transformational grammar and Chomsky these days on the linguistic and academic map? Out of favour, out of fashion? I’m remembering an academic climate where the big battles were between the Chomsky school and structuralists/phonologists… (Cue nostalgic free association: loved those phonological puzzles, way better than the NYT crossword.)

  2. Every so often someone in Japan suggests a reform, and of course they are shouted down. The funny thing is that the counter-arguments are rarely as simple as “We like kanji, they’re an important part of our cultural identity”, they’re more like: “If we stop using kanji, we’ll forget how to read ancient literature” (like anyone other than specialists do now) or “Japanese has too many homophones, we won’t understand what we mean” (ignoring the fact that the lack of kanji doesn’t seem to hamper spoken discourse).
    I haven’t been here long enough to personally vouch for this, but my Japanese friends tell me that mobile phones are reducing their ability to write kanji while simultaneously building their recognition skills.
    This is because on a Japanese mobile phone, you enter words phonetically, then choose the correct kanji from a list that pops up — which includes even the difficult ones that normally you wouldn’t bother to write.
    I know that I can actually write far, far fewer kanji than I can use on my mobile phone. Maybe even an order of magnitude fewer.

  3. I’m remembering an academic climate where the big battles were between the Chomsky school and structuralists/phonologists…
    Yeah, those were the days — my days. The Chomskyans won those battles (in the military sense — they drove the enemy from the field), but I believe they’ve given much ground to yet newer theories since. I’ve been out of the loop for 25 years or so, though, so you’ll have to ask one of those crazy kids still in grad school.
    “Japanese has too many homophones, we won’t understand what we mean” (ignoring the fact that the lack of kanji doesn’t seem to hamper spoken discourse).
    This drove me absolutely nuts in Taiwan. People kept making that argument to me, and I’d look at them like they were crazy and explain to them that if you can understand the sentence when you hear it, you can understand it when written phonetically, and they’d look at me like I was crazy. People have such strange ideas about language.

  4. Exactly how does this yahoo (in the Swiftian sense) explain the phenomenal advances made by *centuries* of Chinese scientists? Hmm?
    Chomsky and TG are very much alive, at least in the local linguistics department. Ask my husband. I think they do feel threatened, however. Although the last retiree from the department was a brilliant (and very much missed) historical reconstructionist, they want to hire — yet another Chomskybot, though the department already boasts no less than three.

  5. Oh yes, historical lingustics, reconstructed IE and all that. *happy memories*

  6. I find it hard to believe that that sort of book was written by someone with intimate knowledge of East Asian languages. Chinese characters are emphatically not all pictoral, despite the stereotype. (If so, what brilliant mind decided to make “to answer” “bamboo-over-fitting”? Where that kanji is a roof over a line over mouth, anyway. And, it’s hard to find a pithy meaning for au. And, yes, you can come up with weird character etymologies which may or may not have any vailidity, like the “woman, again, heart” one for “to get angry”. Or “man-woman-man” for “to taunt”. But “perish-moon-king” for “hope”? I’d have to know the Chinese tradition here to evaluate these character etymologies, but.) Characters that look alike may be pronounced alike: it’s not a phonetically neutral system necessarily, although the phonetic reality is encoded outside of the characters themselves.
    Kanji in Japanese can have a discourse use, actually. When it’s an ambiguous word (which we have in English discourse too, despite fewer homophones: comes from importing from a more phonologically rich language, like Chinese, I think), people will sometimes trace the kanji in the air, or on a table. It’s not completely common, but I’ve seen it done. And the stylistic information in books, whether an author uses the kanji for something or the kana, is a really fascinating layer on top of the rest. (In general, nouns are more often kanji than not, and verbs not completely so; and, interesetingly, verbs that are used in compounding for meaning, like kuru or naru which can add aspectual meaning, are often used in all-kana, instead of kanji-kana form, when they’re used in that sense. Or so is my naive experience.) There was an interesting poster at AAS just recently about the possible cognitive switching costs between kanji and kana reading, sort of a pilot study, this reminds me.
    This is my ignorant, irate, reaction, however. I’ll have to see what sort of rant I come up with when I read the article. One hopes that it’s not so bad as my experience is telling me it’s likely to be.
    Chomsky’s basically the reason I left linguistics, or poor supporters of his. As a scientist’s daughter, I had trouble with accepting “this is the way language must work!” without extensive acceptability/grammaticality work with a large sample group. A professor and a grad student and their friend isn’t enough to convince me at all. Fortunately, as I was leaving, it seemed like that was changing.
    As for Sapir-Whorf… as I recall from my classes, there has been a little suggestive research using color words that may support the weak linguistic relativity hypothesis, but damned if I remember all the ins and outs….

  7. I always saw sapir-whorf as a kind of mirror image of chomsky: one claiming separate languages could form minds to see the world in separate ways, the other claiming that a common grammar engine makes separate languages actually not that separate.
    Both always sounded pretty silly to me. Don’t we need a stable measure of separate and same?
    For chomsky to make sense to me I would need to see chomskians predict a plausible-sounding structure no human language could have, and establish there are no such languages. Since my crude impression is that existing human languages cover pretty much the whole combinatorical space of possible human languages, chomsky’s deep grammar seems to have zero predictive or retrodictive value.
    Verbs, adjectives and nouns are (I think) roughly what you might expect upright tool-using bipeds to use in slicing up the world. Consider dolphins – if they have ‘language’, would it necessarily distinguish verbs and nouns? Probably not. My sense of chomsky is he repackaged verbs and nouns as used in english – and there are existing languages, like the aspect-rich amerindian or west-african languages, which I hear give chomsky a lot of trouble.
    Even the one tribal-ish language I know, hungarian, makes his stuff look odd. No hungarian words for ‘he’ or ‘she’, for example. I personally have difficulty imagining a “deep grammar” that (a) does not distinguish those two or (b) makes any sense if it doesn’t need to.
    This idea that oriental languages like chinese and japanese might cramp thinking sounds quite sapir-whorfian (if chomsky’s right, they couldn’t make a difference, right?) and I have some sympathy. Learning a beautiful, elaborate writing system could cramp originality because you have to follow convention in more detail – NOT because you lack certain kinds of abstract thinkng. The lack of match between many hiragana and katakana for the same sound strikes me as very arbitrary and anti-logic: take the three, similarly constructed, hiragana signs for nu, ne and me, for example. Or the astonishing way in the supposedly-foreigner-friendly katakana the tsu is almost like the shi, and the n is almost like the so. Focussing literate children on exact detail can be very conducive to certain kinds of precise, careful thinking, but it seems pretty uncontroversial to me that it might also hamper their ability to make big jumps of creative, original, unexpected thinking.

  8. I think that the easiest way to fix this imbalance is for the West to scrap its alphabet and adopt hanzi.

  9. Jeez, I wish you’d pinged me with this one, LH. Almost missed it in my daily rounds.
    Nothing to add at the moment, other than the Korean alphabet is of course a beautifully designed and elegant one, even if it’s not terribly well-suited for spelling out the sounds many languages other than Korean itself.
    Reagardless, Mr. Hannas sounds like a complete dipshit, but I haven’t read the article yet, so I’ll hold off…oh wait, I already called him a dipshit, didn’t I?
    Ah well.

  10. From the NYTimes article : “They now recognize that the writing systems of East Asia, including Chinese, Japanese and Korean, are “syllabaries,” in which each character corresponds to a syllable of sound”
    No, damn it. Not in Korean. Now I’m annoyed.

  11. OK. My rantereriffic take on Mr Hannas is at my site. You need to switch on Trackbacks, LH.

  12. Dammit, I thought the RSS feed was supposed to take care of all this stuff. Now I have to learn how to do trackbacks as well? (Deep sigh.) Oh, all right. When the wonderchicken clucks, I listen.

  13. Well, Korean’s tricky to classify, even by professionals: I am utterly and cynically not surprised the Times got it wrong. Each character’s kind of a syllable, right? But the characters are subdivideable phonetically, so…. It’s not an alphabet, and not a syllabary. It’s just… Korean.
    The absolutely neat thing is to look at it and know that the designer(s) (the king gets the credit, but) knew their front and their back articulators. It’s really quite impressive.

  14. Actually, Christina, it’s not too complicated. Each character is a sound (sometimes modified a bit – like voiced or unvoiced or held or released depending on position in a syllable, in a much more consistent way than English), and characters are put together into twos or threes by completely consistent rules into syllables, which look complicated, but are utterly simple once one learns the basic rules. It *is* an alphabet, and an elegant one, too, and about a million times simpler than (what little I know) of Japanese or Chinese.
    The grammar on the other hand is a challenge, but it literally takes anyone of average intelligence a couple of hours to be able to read Korean, even if they don’t understand what they are reading.
    I really need to write that Hangul primer I promised Jonathon Delacour last year….

  15. Sorry, *Kristina*.
    Force of habit.

  16. But it’s not *quite* an alphabet in the same way that roman and arabic and sanskrit (and even uighur script? I’m shaky on that now) are: each part of each blog is not always in the same place, right? Sometimes you read a spacial-unit (is there terminology for this?) left to right, sometimes up to down, sometimes a combo, and then you matrix this with the general reading style (which is left to right, unless I misunderstand more than I thought). All other alphabets are fairly unidirectional, even if that direction may have options.
    And yet, of course, the cognitive cost is very little since we pretty much recgonize *shapes* of words, rather than letter-by-letter, unless it’s an unfamiliar word. That’s pretty much proven. The technical cost, however, in the pre-computer typesetting, is fascinating.
    In almost all classifications I’ve seen, it’s Character, Syllabary, Korean and Alphabet, in decreasing order of separability.

  17. You’ll get offers from everyone hereabouts, but if you do need help enabling trackbacks, holler.

  18. Sorry, *increasing* order of separability. But, I definitely admit Korean’s similarity to alphabets, in that each phoneme is represented as a symbol, organized in syllable shapes, which are then organized as words. Anyway, except for the rare Chinese character, rather not what the summary of Hannas’ book implies. (I’m tempted, again, to require credential checks before people can be called linguists, but that’s petty on my part. Everyone assumes it’s the multi-language thing….)

  19. “Sometimes you read a spacial-unit (is there terminology for this?) left to right, sometimes up to down, sometimes a combo, and then you matrix this with the general reading style (which is left to right, unless I misunderstand more than I thought). All other alphabets are fairly unidirectional, even if that direction may have options.”
    As far as reading a syllable, it’s always left to right (then down, if there’s a third character). So I guess that is a somewhat unique feature, but it’s certainly no big deal. It’s complicated sounding to describe, but simplicity itself to do. Words are left to right, as are sentences and paragraphs and so on.
    (It is true that in the archaic style, Korean chunks of text can be written top-down, right-to-left, in terms of sentences and paragraphs, but that is basically *never* done anymore, and doesn’t alter the basic way in which syllables and words are put together.)
    Hangul *is* an alphabet, with only a few more characters than this one I’m using now, and although single sound characters are marshalled into twos and threes to make syllables and look a bit like chinese characters, they are still perfectly regular (much more so than any of the other half-dozen languages and two alphabets with which I have some familiarity) in the way in which a symbol represents a sound.
    “each phoneme is represented as a symbol, organized in syllable shapes, which are then organized as words”
    This is true. But the syllable shapes are not freighted with any significance in and of themselves – they are just non-arbitrary arrangements for the sake of clarity. Hangul is no less an alphabet because of them.
    What may be confusing is the way in which they are written to form syllables (with consonant or vowel ending) (using english equivalents here to muddy the water even further) :
    But that’s hardly as complicated as you make it sound – if fact once you understand the incredibly simple rules for writing syllables, it seems like a system miles beyond what we’re used to with English (and so on) for clarity and good sense.
    I do love to diss Korea when I’m grumpy, but the writing system is truly a thing of beauty.

  20. Oh, I never ever mean to diss Korean (which I’m trying to convince my adviser makes more sense for me to learn next than Chinese, but since I do early East Asia, that’s a lost cause). I just think it’s something more special than an alphabet in my view…. Plus, trying to explain as I understand it, the rational about the typography classifications as I’ve seen them. So an amicable disagreement? Pace?
    Oh, if you know, where would the “trashy” historical fiction in Korean be? Since that seems to be what motivates me in self-study, and since my next year’s been spoken for with classical Japanese…. I’d like to get a start as I can. (Do you know a good textbook, as well?)

  21. I don’t know about the trashy historicals – they are big on the TV, though, so I’m sure they must abound.
    I’ve never actually used a textbook, although I’ve had a look at a few – what I’ve learned has been picked up of necessity, living here. I don’t know that there are many that are good.
    I’ve never seen discussion of some of the really useful stuff in any book (aspiration vs voicing for example).
    Maybe I need to write one….

  22. Ooh Dorothea, I like that term ‘Chomskybot’.

  23. It’s not original. There actually *is* such a thing. Check it out for yourself.
    (and many other places)

  24. “Japanese has too many homophones, we won’t understand what we mean” (ignoring the fact that the lack of kanji doesn’t seem to hamper spoken discourse).

    Spoken language has two things going for it to overcome homophones that written language simply does not.
    First, spoken language has more context than written. Listener and speaker are in basically the same environment and share common knowledge about the recent past of the interaction; furthermore, the listener can directly perceive the speaker’s non-verbal communication, such as body language. The greater context and more channels of communication act to reduce ambiguity.
    Second, spoken language has a dedicated error-correction channel. The speaker prompts for listener response, the listener confirms that she’s following or requests repetition, confirmation, or explanation.
    Written language is a different medium from spoken language and has different requirements. The ability to distinguish homophones is more important in a written language than it is in a spoken language.

  25. A couple of comments:
    (1) Taiwan also uses Chinese characters, ones that are actually more complicated than the simplified ones used in Mainland China. And, even though you may doubt the literacy statistics of the PRC, Taiwan is one of the most literate countries in the world. I don’t see characters as being an impediment to literacy (except for my own, since it does make it difficult to learn Chinese as a second language).
    (2) The Sapir-Whorf hypothesis is often mischaracterized. Pinker, I believe, deliberately mischaracterizes it in order to make a ‘straw-man’ for his own arguments. There are actually two flavors of the Sapir-Whorf hypothesis: strong and weak. The strong states that a specific linguistic system has a determinant effect on cognition. Interestingly, this strong version is NOT what you find in Whorf! His argument was that language structure can FACILITATE a TENDANCY to see the world in one way rather than another. One example he uses is from his days working for an insurance company. Someone would light a cigarette near “empty” gas canisters. The word “empty” conveying the sense of “null” and therefore of no danger – when in fact there are still strong fumes that pose a serious danger. His more famous example is about Hopi and the way that time is referred to. Unlike English where time is a discrete unit. “Tomorrow is another day!” Hopi sees time as something that is continuous much the same way that we refer to a “cup of water” (i.e. it is a ‘mass noun’). Each day being a ‘cup of time’ so to speak. This means, according to Whorf, that actions performed today might have an action on tomorrow. He uses this to argue that the ritual practices of the Hopi are thus influenced by their sense of time, since a ritual performed today will help ensure that an activity happening tomorrow will go off as planned.
    It is, however, a big jump to go from such a claim to saying that Hopi are UNABLE to conceive of time as a discrete unit. They do think of other things in discrete amounts and almost certainly are capable of doing so. Any such claim would have been rejected by Whorf.
    (3) I don’t understand the argument about writing, since all college educated Chinese I know can read and write in English even if they can’t speak it. A professor at my brother’s computer science graduate program in the US told him that the department’s English GRE scores would actually go up if they ONLY admitted students from China! If they can read and write English, should they be able to do everything associated with such literacy?

Speak Your Mind