Hanzi vs Pinyin, in case you hadn’t heard enough

Just procrastinated my way into a (sort of) recent Language Log post by Victor Mair. The subject is whether Chinese characters are “necessary” for writing Chinese. There are 62 comments at this writing and a frenzy of emotion. One of the key quotes from Mair:

My rule of thumb is always this: if homography were a problem in (more or less) phonetic scripts based on real, spoken languages, then homophony would be a problem in the speech of such languages.

I’m not trying to pick on words, but this looks to me more like a tautology than a rule of thumb. By definition, spoken language written in a phonemic script is not going to have homophony problems unless the spoken language has homophony problems.

So why the big debate over whether Mandarin “can” be written in Pinyin? It’s helpful to parse the question a bit. The real issue is whether Mandarin as it is currently written could be written successfully using Pinyin. That’s the only case of serious interest. The other two — (a) writing Classical Chinese in Pinyin, or (b) writing spoken Mandarin in Pinyin — should be universally acknowledged as (a) impossible, and (b) a cinch.

The difficulty is that written language, as Kevin Miller puts it, has two mommies. True, writing is primarily and originally a derivative of spoken language, but it’s also more than that. Writing is never JUST a representation of spoken language — it’s also self-referential and sometimes purely symbolic. Moreover, since it exists in a different medium (on the paper / screen) it can do things in that medium that spoken language cannot.

So what does modern written Mandarin do that spoken Mandarin doesn’t? It does the same thing that any writing system does. First, it represents speech, but then also…

  1. Writing can contain elements that have no counterpart in spoken language. Miller cites 囧 as an example. If it’s heard in speech at all (jiǒng) it’s probably a derivative of its written use.
  2. Writing can combine morphemes in ways that wouldn’t make sense in spoken language but are perfectly understandable in writing.
  3. Writing keeps alive vocabulary that has fallen into spoken disuse, therefore you may encounter words in writing that seem perfectly normal that you would have a hard time processing if you heard spoken.

That’s the list I come up with and it probably needs supplementing. One and three are pretty clear. But for two, what does it really mean to “combine morphemes in ways that wouldn’t make sense in spoken language but are perfectly understandable in writing”? If you read Mandarin, you’ll know exactly what it means because it happens all the time. The written language — with its abundance of redundance and extra-phonetic information — is chock full of shorthands that are visually clear but beg explanation if spoken. Names of programs, committees, etc. are probably the easiest examples. I’ll reserve the space below for outstanding examples from Sinoglot’s erudite commentariat:




Because the point I want to make is that #2 above, the “visual morpheme” phenomenon, is not unique to Chinese. Just the other day Evolving English discussed the apparent productivity of -lebrity as a suffix:

  • Rhee is a Grade-A edu-lebrity
  • [Y]ou’re far more likely to be picking through the sausage-makings as they just sort of spray willy-nilly out of the meat grinder of news-lebrity that has replaced the news.
  • Ever want to know what it’s like to log into Twitter as your favorite Twitter-lebrity?
  • And there are perks to being a bona fide Z-lebrity
  • Bornstein refers to herself as a “sub-lebrity

and noted that the -lebrity words…

work better in written language than said out loud. A number of them are a bit awkward to say, possibly because they end up with sound sequences that don’t entirely work

Point being: you can do stuff with written language that just doesn’t work in speech. And vice versa of course.

Bringing it back home to Mandarin: is there so much of items 1-3 above — script-play, shorthands, etc. — that modern written Mandarin would be hard to follow in Pinyin? Would the written language have to change dramatically to accommodate a script change?

Personally I see the answer depending hugely on writing style. Lots of dialog in novels would work just fine (arguably better) in Pinyin. Lots of academic writing: not so good, I guess.

It would be fun to propose an experiment for the “does it work with Pinyin” question, but given how far I’ve gotten with past experiment proposals, I’m loathe to add to the list of commitments.

My rule of thumb is always this:  if homography were a problem in (more or less) phonetic scripts based on real, spoken languages, then homophony would be a problem in the speech of such languages.

29 responses to “Hanzi vs Pinyin, in case you hadn’t heard enough”

  1. Louis says:

    Mandarin can be written phonetically. That’s been shown many times. However, I don’t think the Chinese want to change…

    It has been mentioned that the Chinese were thinking about creating a phonetically-based script with 三十六字母. We know that intellectuals thought about a phonetic script when 八思巴文 was out and about. During the 20th century, some highly effective systems were created: 新文字 and what not. Still, the Chinese have always gone back to the characters. I think that says a lot…

  2. 慈逢流 says:

    “Mandarin can be written phonetically. That’s been shown many times.”—i think one aspect here that is wildly underestimated is the step from inner-familiar and inner-communal, everyday speech (which obviously somehow ‘works’, and especially in hinterland china with its relatively many analphabets) to a form of expression that works over a wider range of topics depends very much on the written word; at least this is the seemingly the case in all societies today (i am simplifying the hell out of this. the argument seems to hold for modern china, though).

    now if we turn to modern academic and technical chinese, we see that a tremendous vocabulary has been built on top of the assumption that the spoken word can be, maybe must be, underpinned by chinese written in 漢字. one of the areas where this is most visible is chemical nomenclature; i recommend glancing over two articles: http://en.wikipedia.org/wiki/Organic_nomenclature_in_Chinese and http://en.wikipedia.org/wiki/Organic_nomenclature_in_Chinese. this is the tip of an iceberg and not very representative; other technical vocabulary is certainly less removed from everyday speech. for example, 3-丁烯-1-醇 (sanwei-ding-xi yi-chun) is the chinese expression for homoallyl alcohol, also called but-3-en-1-ol. either expression is utterly unparsable for non-expert chinese and english speakers alike. at least, however, an english uneducated speaker will be able to catch the ‘alcohol’ in ‘homoallyl alcohol’. what is an uneducated speaker of to make of ‘sanwei-ding-xi yi-chun’? is he going to understand the ‘chun’ here means ‘alcohol’? i would guess that is harder to do in chinese than in english.

    fact is, quite a bit of ‘advanced’ vocabulary has been treating the problem of homophony as if it didn’t really exist, and has relied on a written form of expression that can sort things out. korean and japanese have, at least in their borrowed vocabularies, inherited this. i mean, imagine, while there are lots of ‘minor’/’miner’ doublets in english all right, even korean with its ~50% sinitic vocabulary and overall good preservation of a slightly richer syllabary as was current 500 years ago has oodles and oodles of doublets. which is why the place where to this day chinese characters are found in the written is academic articles.

    to put it bluntly, chinese homophony paired with the tendency for terseness of expression has resulted in a language that *is* hard in places to communicate as-is-written, without the help of circumlocutions, clarifying passages, or subtitles. i remember after the kobe quake a professor appeared in tv to explain why the expressway collapsed just the way it did, and at some point used ‘tekkin’ for ‘reinforcement steel bars’. i don’t know but maybe in english he would’ve gotten away with just saying ‘rebars’; in japan, the program makers felt they had to support that utterance by showing 鐵筋 with some furigana. my japanese is really bad, and in chinese you would say 鋼筋 gāngjīn instead, but seeing the characters it was immediately clear to me what was meant.

    be it said that the celebrated chengyu also to a large part reveal part of their meaning when presented in characters. Saiwengshima, what does that mean? at least 塞翁失馬 gives a hint that a horse was lost.

  3. 慈逢流 says:

    in the language log referred to above, one commenter says that “I don’t really understand this argument that we should reductively examine the costs and benefits of something so inherently wrapped up in culture and history. What is our desired end state exactly? Maximized efficiency? But we could just as easily begin maximizing efficiency by eliminating Christmas, football, and art museums, surely?”—i cannot say how much i agree with this argument. people who do not have a grasp of any CJK language often come up with comments on how wasteful chinese writing is. calendar reforms were proposed many times, some with the argument how much efficiency could be gained if each year looked the same and calendar printing could be reduced. streetcars were banned from many american cities for their glaring inability to be helpful in matters of transportation, which helped a lot to shape the efficient cityscapes of today, with their congestion-free and beautiful flyovers. jobs got shipped from europe and the us to china, with tremendous benefits for foxcom workers and untold numbers of jobless people who can now live off of the well-developed welfare systems. for efficiency, farmers have been spraying the landscape with untold tons of herbicides that kill all life, including the colateral human, and live only the efficient cash crops grow.

    we’re so rampant in our seemingly ingrained desire to max out the hell out everything, our time on this planet, our ways of living, our production methods, the soil. what kind of machine world we’re building right now, few seem to be concerned ybout. but before we die, let us weed out all those unfathomable, inefficient writing systems. they’re preservers of millenia of human experience. but who needs history in an efficient world? who needs humans anyway? the other day, i saw a guy smile. what a despicable utterance that served no real purpose! no really i am growing bitter and have to stop here.

  4. 慈逢流 says:

    another helpful comment from over at the language log:

    “despite its perceived orthographic handicaps, Japan has for the last century and a half pretty consistently been the most economically and technologically successful society using a non-IE language out there, despite having been bombed to rubble in the middle of that period. Not a very good marketing slogan to tell them that if only they’d gone all the way to romaji earlier and thus freed up all that extra time in elementary school for other subjects, they might by now hope to be as prosperous as the Phillipines or Indonesia. (I’m leaving out Vietnam b/c of the confounding communist-brutality variable.)”

    yeah man, go go go. that’s a lot of historical ape shit in the fan, flying straight into the face of the optimizers.

  5. 慈逢流 says:

    the same commenter as referred by me above, J. W. Brewer, also said: “With all due respect, the late Prof. DeFrancis does seem to have been rather a monomaniac on this subject”.

    i could not have worded it any the better. i never met defrancis in person, but my first teacher in chinese over here in berlin did. i spent many hours reading defrancis’ books when starting to immerse myself into the chinese language. his pieces on the abstruse outcome of fictional attempts to write english in chinese characters were informative and well thought out. i am full of respect for this man who used to appear in b/w likeness wearing flowery shirts and a broad grin on the blurbs of those university of hawaii press books. i feel affection.

    but i must say i have also been feeling, over the years, that the drive to condemn chinese writing as foolish, to show that it is really 90% half-bad-ass phonemic writing and but 10% semantic or ideographic or what you’d call it writing, and finally the conclusion that the chinese had better switch over to a letter-based system rather sooner than later is foolish, uninformed, in fact backwards, eurocentric, and so on.

    i say this as someone who has invested years of learning (but not one coupled to a successful career) into chinese writing. i am at this very time attempting to build a database of chinese characters, and the past few years of work have taught me that 漢字 do form a horribly intertwined mess that is very hard to fathom. i have learned pinyin, katakana, hiragana, hangeul, hepburn, wade-giles in addition, so i believe i can write down about anything in whatever occurs to me. i also write old german hand, which makes me as hip as a failed dinosaur. i sometimes feel dumbfounded when i discover how easy it is to write down anything in any script except the chinese one. it is definitely the hardest one.

    yet i also see that all arguments ever stuffed into the all-pervasive cannon of seemingly “rational efficiency” or whatever levelled against the notion that chinese is ideographic etc get lost when you look at japanese. japanese, i hold, is in fact one example, and the only one existing in the present day, where a truly ‘ideographic’ script has been realized. this fact is conveniently ignored by defrancis, who only looks at chinese. yet in japanese, 人 is both jin and hito, and 阪 can stand for han and saka, yet the essential meaning remains the same.

    i do not claim that this system is overly efficient or practical, i just want to point out that it seems to work somewhat, in a highly copmpetitive world with competitors all relying on less time-consuming writing systems. chinese is in fact much less ideographic. in japanese, you have to read something, make out what it means, assess linguistic probabilities, and then come up with an interpretation of meaning and sense. this system has become so complicated that personal names and addresses only really ‘work’ when given in, say, parallel kanji and kana. otherwise you soon get lost on the map. that is, maybe, deplorable. japanese themselves often feel it is hard.

    ok i’m drunk by now and must stop here. i don’t intend to take over this fine forum by posting excessive spam. it’s just i’m waking up each and every morning these days thinking how horribly wrong we’ve gone, as an industrial society. industrial used to mean hard-working; these days it means you lost your job to some kind of machinery that turns out 200000 cans of beer an hour and threatens to kill earth. all for the benefit of what or whom?

    i say: no idea what those lousy hard-to-read classics talk about. guess they’re just graffiti of ol’days gone by, just as clueless as i am now. but i give damn shit to anyone who wants to erase ’em off the board, as egypt and babylonia and the aztecs got, for the sole benefit of a culture of consumption that is, in terms of consumption, so limited in expression as the present one. i do not want to take over this forum with political comments, either. but this supposedly efficient and only way of living, y’know, it’s gonna take two earths a mere twenty years from whence, y’know, according to a recent united nations report.

    is this the right time to kick a 5000 years old culture of writing into its grave? i guess not.

  6. Louis says:


    I’m not a supporter of doing away with the characters. Prof. DeFrancis wasn’t either. He proposed a system called 雙文制。In this framework, there would be two systems, one using the characters and the other a phonetic script. His underlying reasoning was to provide the peasantry with a system they could learn and use. Things like scientific essays, literature and what have you wouldn’t be affected. I think this is a practical first-step solution to the problem of wide-spread illiteracy in China.

  7. Syz says:

    @Louis: “Mandarin can be written phonetically” — the point I’m trying to make in this post is that there are different kinds of Mandarin. Some types, maybe most types, would clearly be no trouble in Pinyin (e.g. spoken language). 慈逢流 may be onto the kind of example I was looking for with the chemistry thing: a style of writing that exists today (not classical chinese) but is difficult/impossible to understand when read out loud, even with proper technical background.

    @慈逢流: Sinoglot is honored to be the recipient of alcohol-fueled analyses, even if they include digressions! Seriously, it reminds me of a math teacher I had who told us there was no better comfort than to sit down with a beer and a good math problem.

    On to substance: The chemistry thing looks interesting but tough. Maybe all the cases are tough — as in tough to show that the written form of an article, if read aloud, would be harder for a Mandarin speaker to understand than for an English speaker to understand the equivalent thing in English. Are you in possession of a passage of text I could try out on some people?

    Proofs aside, I tend to agree that “a tremendous vocabulary has been built on top of the assumption that the spoken word can be, maybe must be, underpinned by chinese written in 漢字”. Nicely put.

    成语 are an interesting case too. Any that are in common spoken usage, of course, shouldn’t need hanzi to be understood because they’re understood when spoken. However, I’ll bet there’s a slew of more obscure ones that are clear when written (or at least provide a hint of the meaning) while being obscure to the point of meaninglessness when spoken.

    I tend to take a dim view of efficiency too. Whether Hanzi lose the efficiency debate or not, the proper question is: so what? You want to switch to [insert script here] so that the kids will have more time to watch, uh, Strawberry Shortcake?

    The efficiency case I do have empathy for, when it comes to reading, is one that Mark Liberman mentioned in the comments. It’s not written very compactly, but it’s worth quoting. He’s talking about Japan, but same applies here:

    we could consider the effect of this writing system on educational outcomes for the children of immigrants, who are likely to become more numerous as Japan’s demographic changes create pressures to bring in workers.

    It’s strange to consider a national writing system as if it were a complex artistic skill like oil painting. Learning to read and write is not an optional skill that people with interest and time can choose to pursue, as part of the fulfillment of their “individual abilities and passions”. It’s an absolute requirement for progressing in the educational system and for getting a decent job.

    I guess instead of “immigrants” read “internal immigrants” or simply any who have not had the freedom to devote their childhoods to education. There are still a lot in China today.

  8. Louis says:

    0. The Chinese will find a way to adapt the phonetic writing system to their needs. But,…

    1. The characters do have their benefits. Acronyms are a good example. Acronyms in Mandarin are much clearer when written. Comparatively speaking, they are clearer than English ones as well; for example:

    (1) What does 政大 stand for?

    (2) What does NCCU stand for?

    I bet you guess (1) easily. You probably won’t figure out (2) without google.

    The characters have their disadvantage, too: they take a long time to learn, which leads to high illiteracy rates (even in Taiwan, although it won’t show up in the census).

    One nice thing about the characters is they are not phonoetically linked to a specific dialect. I think this is an important point.

  9. Syz says:

    Louis, good point about the abbreviation advantage. I’ve always thought the biggest advantage is in understanding technical vocabulary. English is full of technical words whose meanings are utterly opaque. Mandarin in hanzi seems less so for just about every technical word I’ve come across.

  10. Chris Waugh says:

    Louis: ” I think this is a practical first-step solution to the problem of wide-spread illiteracy in China.”


    “they take a long time to learn, which leads to high illiteracy rates”

    Prove it. Last I checked, which I admit was a few years ago, Hong Kong and Taiwan had literacy rates on a par with any other developed territory, and the Mainland’s literacy rates were alright for a developing country. I have seen no actual evidence, only assertions, to prove that characters present any barrier to literacy. Everything I have seen on the subject suggests quite strongly that it is access to education, and not script, that determines literacy rates.

  11. 慈逢流 says:

    @Syz “English is full of technical words whose meanings are utterly opaque. Mandarin in hanzi seems less so for just about every technical word I’ve come across.” i’ve had the same experience, overall. there are even cases where the chinese term really helped me to understand what a given term means. i mean, ‘rebars’ might be some kind of ‘bars’, but in my native german, they’re often called ‘die Armierung’ or ‘die Bewehrung’. actually ‘die Bewehrung’ is very opaque to a na(t)ive german (all the more for its similarity to ‘die Bewährung’), and as for ‘Armierung’ you have to be told or you wouldn’t know. this is one example where a german/chinese dictionary entry ‘Armierung, Bewehrung: 鋼筋, 鐵筋’ actually works the other way round for me, the chinese (and japanese) there helping me to understand my *native* vocabulary.

    of the sciences i had contact with in chinese, chemistry seems to be a bit special as so many characters have been created to signify chemical elements. these characters tend to be somewhat hard to learn and pretty hard to catch in the spoken. i cannot find anywhere any reference how to e.g. utter 釙 po in a way to make it a bit clearer to the audience that polonium is being spoken about; you’d probably say 釙(元)素 or 釙金. interestingly both japanese and korean chose to replace written terseness with spoken expressiveness, and so polonium is nowadays called ポロニウム and 폴로늄 (in this case both opting to take german instead of english pronunciation as a model, btw) in those languages.

    other than that, i have no suitable text to try out on people; i’d probably search wikipedia for one. i likewise have no idea exactly *how* to try a text on people to see whether they catch it. maybe in the form of a dictation test? that would quickly turn to a spelling contest. but of course one could instruct subjects to write a (not too fast) dictation down using whatever spelling (characters, pinyin, squiggles) comes to their mind, and then let the interviewer point to passages in the text and answer questions. so say some subject writes ‘po的化學性質與xi及liu類似,但帶有放射性。’ for ‘釙的化學性質與硒及硫類似,但帶有放射性。’ it’s easy to spot whatever was written for polonium, selenium, and sulphur, respectively, and inquire. exactly *what* you’d ask is another problem though. maybe just ask people to tell you everything they know about that particular thingie. or, rather try texts where the chemistry is not explicitly introduced. it’s hard to come up with something, but you probably know better than me how to do such things.

  12. Katie says:

    @Syz: I’ve been meaning to write a blog post on this one for a while now, but while I agree with you completely about the technical terms, I wonder whether native speakers get as big of a boost from the characters as we non-native speakers do. I think, in some perverse way, the characters actually help me learn Chinese because they make all the compound words, not just the technical ones, more transparent, but I’m wondering to what degree native speakers benefit from this. I suppose they do–I learned the word carbohydrate in Chinese the other day, and my teacher pointed out to me that it was compounded the same way that the English word is, but I had never previously thought about the fact that the English word was a compound (like) word or that it might give me some hint as to what a carbohydrate was actually constructed of. Perhaps a Chinese person approaching 碳水化合物 would think differently. My teacher seemed worried that she couldn’t better explain to me what the chemical composition of a carbohydrate was, but on the other hand, she seemed surprised that I could tell her what foods were high in them even though I didn’t have a clue about the chemical composition.

  13. Katie says:

    @Chris: I have no idea what literacy rates are in Mainland China/Taiwan. But here’s a question (to which I think there is not an obvious answer): which is easier to do–ensure that, say, 95% of your population gets at least (say) a tenth grade education or reform your writing system so that it can not only be learned but also be retained in a few years of education? (What is the average education level in rural China these days?) Of course, there is an obvious answer to which is better to do. But, jumping off from Syz’s point, there are a lot of people who would stand to benefit from a simplified writing system. It’s not just the kids who grow up as “internal immigrants” (not to mention us expats) but also the kids with learning disabilities (I’d think) or the kids who have a mental disability. In the US someone with a mental disability might never learn to read beyond a third grade level–but that would still be good enough to independently accomplish basic life tasks as well as communicate in writing with friends and family and read simple things for pleasure. How would the same person fare in China? Would they ever learn to read and write at all? I know the system here is stacked against these kids in any number of ways, but accessibility to some independent life skills might make a difference.

    Also, the ability to write characters takes, I would think, repeated usage, not just one time acquisition. Suppose you go to a Chinese school through third grade–I’m pretty sure you would have already have pretty good reading and writing skills by this time. But if you leave school then, would you retain them? A phonetic script would likely have a huge advantage here.

  14. Louis says:

    @Chris Waugh,

    Of course the long-term solution to illiteracy in China/Taiwan is to teach everyone how to read characters during their school years. This is exactly what has been happening in Taiwan for the last 30+ years. But, we should still have a solution to the current problem, which is…

    A good amount of older women from the countryside can’t read in Taiwan. For example, my wife must go to the bank with her mother because her mother can’t fill out a deposit form. This isn’t because her mom is exceptionally dumb/poor (actually the contrary). Instead, it’s a result of Taiwan’s recent history.

    The reason why 雙文制 is important now is because it gives people who are not going to learn to read characters a way to function. If the deposit slips at the bank had pinyin or zhuyin fuhao on them, I know for a fact my mother in law could handle it by herself.

    I never said the characters needed to be replaced, I just said they take a long time to learn relative to pinyin. (There is a video of Prof. DeFrancis teaching illiterate farmers in China how to read pinyin. The farmers were reading texts within a couple of hours.)

  15. Louis says:

    @慈逢流 and Syz,

    I agree with the observations on technical terms. However, sometimes the seemingly clear logic within a Chinese word can cause misunderstandings — example:舌根 is often a translation for ‘velar’. ‘Velar’ should not be confused with ‘tongue root’. 

  16. 慈逢流 says:

    @Katie: you are certainly aware of the fact that kids in china these days do learn pinyin at a very early age. to the degree that script is accepted and used outside the school they can benefit from it. the least they can use it for is private notes.

    also, you’re arguing with the hidden assumption that latin letters are ‘easier’ to learn if you have a ‘mental disability’. there are probably as many kinds of disorders as there are people in the world, and i’m afraid that latin letters (and alphabets in general) are not inherently easier for everyone, with or without a diagnosed mental whatever.

    as a matter of fact: if letters are so ‘easy’, why did none of the writing systems we know of started out as an alphabet, or quickly turned to letters? that it is possible, in east-asia, in a language severely ridden by homophony, to come up with alphabetic system and make a whole society successfully switch to that system, is evidenced by korean hangeul, an invention of 세종대왕 世宗大王, made around 1445, probably inspired by the mongolian phagspa. chinese history is full of ingeneous inventors and erudite pundits as well as illiterate masses, yet no alphabetic system gained much traction before 注音符號 and 拼音. this is certainly also due to social reasons, to be sure.

    one writer once pointed out this interesting thought: what westerners believe about writing, eg that letters are ‘simple’, that orthography should be straightforward and unified, and so on, are oftentimes rather artifacts of our own social situation. the author then went on to say that in traditional japan, the complexity, the vacillating usages, the hard-to-read calligraphic styles of writing were very much seen as status symbols; a fully literate person needed in fact to spend a lot of time (ie money) on learning and practising. today, with better education for basically everybody, definite kanji catalogs with reduced options about what a character can mean and how it is to be read, that complexity is somewhat mitigated.

    i would think that the chinese script is not necessarily much more difficult than the latin script. some aspects of it are downright simple, surely simpler than alphabetic writing, like the basics of 人目木母日月 and so on. i also experienced, when learning chinese, that the real difficult thing is to get down the first 500 or 800 or so characters; after that, you can go on at a much faster rate. some characters you simply learn because they pop up in a text and you guess at them, or because they happen to be written on a sticker in your local bus line. some you learn ten times only to forget them as often.

    asian schools are often imagined as some kind of slave labor camps for minors, and then people at parties ask you how many years those kids have to lose before they have learned all of the—how many? three thousand? fifty thousand? how many are there?—characters ‘needed to write that language’. otherwise very educated people in the west come up with bollöcks like this all the time.

    ok written too much again. let me put it this way: from my experience, chinese writing is certainly not the simplest on earth, but latin script isn’t either (in fact hangeul is).

    joking aside, i believe that in places like the US and europe the schooling systems with all their flaws are in fact good enough to be able to offer chinese in addition to what they offer today without having kids to stop every other activity except sleeping. with which i mean to say that learning chinese is in no real way harder than, say, chemistry or french or such.

    and yes, i find it difficult to write out 噴嚏 myself.

  17. 慈逢流 says:

    @Louis: i assume you are talking about ‘velar’ as used in phonetics and phonology. well, careful writers in those fields often make a point about their terminology, and one system i encountered they made clear they wanted to stick to terms like ‘dorsovelar’, which says there is an active (moving) articulator (the dorsum) articulating against a passive articulator (the velum). i just discovered the All Knowing Oracle doesn’t follow that usage; they prefer to simply say ‘velar’. but other western writers have sure used ‘dorsal’ in their terminology, and maybe in your case the author meant this with 舌根. wikipedia wants us to preach that this term is really appropriate for ‘radical consonants’ (a term to include pharyngeal, epiglottal, and epiglotto-pharyngeal articulations). all i can say is that terminology in these particular fields is about as solid as quicksand.

    in old china the autochtone (ahh never know how to write that one always get it wrong… thats 本地的 in chinese for you! 本地!!! for all of autochthon with all the ouches in the wrong places, a word i can’t even SAY!!)—the native phonological sciences used to be known as 苦學, bitter studies (in part because of the mysterious and convoluted terminological systems). that, as far as i can tell, is an understatement.

  18. Katie says:

    @ 慈逢流 –actually I specifically avoided arguing for an alphabetic system. English is the only point of comparison I have to even give examples from, but obviously a phonemic system would be easier to learn than English. I would hazard a guess that Korean kids do learn to read more quickly and with less pain than American kids do. I don’t know how the Taiwanese system works so can’t comment on that one, but if it works the way I imagine it does, that would do the trick too. I’m not thinking about letters. I’m thinking about volume. I agree with you that the initial learning curve for Hanzi is pretty steep and then it starts to level out–but you just can’t get around the fact that you’re looking at an initial learning curve of 500-800 in Hanzi vs an initial learning curve of 20-ish in pinyin. I’d be really curious to know whether any studies about learning disabilities and Hanzi have been done, actually. For some people, my basic assumption is still that the sheer volume of memorization with Hanzi is going to be too much. But what about specific learning disabilities that don’t have to do with memory, such as dyslexia? Anybody know whether you’re better off learning an alphabetic system or a character system there?

    And I am aware that kids learn pinyin early on (and that this would be the obvious choice for reform), so I suppose what I should have said is not “reform writing” but “reform the system such that basic life tasks can be completed in pinyin”. For example: my post office receipt is printed in Hanzi and French. I filled out a census form the other day that was printed in Hanzi and English. The subway system in Shanghai can be navigated in Hanzi or English (although most of the place names, for all practical purposes, could also be considered pinyin). The only place I see pinyin used regularly is on street signs (is this true everywhere in the Mainland or just in my city?) and in a minority of books for young children. The former is the type of reform that I, like Louis, would deem helpful to solve the problem. But that’s still a bit idealized, in the sense that it only does you any good if you speak reasonably standard Mandarin, or if the sounds in the dialect you do speak correspond in a fairly nice 1-1 manner to reasonably standard Mandarin.

    As to the bit about our views on writing being the result of our social situation, you’re undoubtedly right. My feelings about universal access to education are certainly biased by my own cultural views. Nevertheless, I feel pretty strongly about them, or I wouldn’t be bothering to write this post, I suppose.

    I’m all for Hanzi remaining in use, though that probably wasn’t obvious from my earlier post. I’d actually be sad to see them go, and presumably (and what is actually relevant) most Chinese people feel likewise, or they would have gone by now.

  19. Louis says:


    ‘Dorso-velar’ and ‘velar’ mean the same thing in SC phonology. The dorsum is not the tongue root. So the term 舌根 is misleading regardless of which English term you want to use.

    Here are some sources, though:

    I’m looking at “漢語音韻學” by 董同龢 and he used ‘舌根’ to to signify the place of articulation for [k, kʰ, ŋ, x] (p.15). 竺家寧 did the same in his book “聲韻學” with ‘velar’ written next to it(p.36). In the phonology of SC, this correlation is well established…

    Lin Yen-Hwei (2007): “The Sounds of Chinese.”
    Duanmu San (2006):”The Phonology of Standard Chinese.”
    竺家寧(1992): “聲韻學.”
    董同龢 (2007): “漢語音韻學.”
    林慶勳, 竺家寧 (2007): “古音學入門.”

    These are the books I have next to me right now. There are many more.

    You wrote: “but other western writers have sure used ‘dorsal’ in their terminology, ”

    —‘Radical’ and ‘dorsal’ are actually natural classes. (Peter Ladefoged (2005): “A Course in Phonetics.”)Natural classes are categories that contain language sounds according to shared features. There is a trend in linguistics for places of articulation to be refered to by their natural class. This is a bad habit (Peter Ladefoged (2005)) and doesn’t mean that we should confuse the two.

    For example, I might say “there are no voiced coronal stops in SC.” This means the same thing as “there are no voiced alveolo-dental stops in SC.” Why? Because the only coronal stops in SC are the alveolo-dental ones.

    But regardless, ‘dorsal’ is not ‘tongue root’.

    You wrote: “all i can say is that terminology in these particular fields is about as solid as quicksand. ”

    –This is simply wrong. I suggest you consult the Journal of the International Phonetic Association if you don’t believe me. A lot of work goes into standardizing phonetic terms…

  20. 慈逢流 says:

    @Louis: “‘Dorso-velar’ and ‘velar’ mean the same thing in SC phonology”. no. because there is no single thing you might want to call ‘SC phonology’. it’s not like there is a “SC phonology, Inc., Buffalo, NY”. there is a discourse with thousands of voices stretching over several millennia. an author is free and should be free to define their own technical terms. also, the chinese language (their users) are free to choose their terminology. ‘velar’ might be a convenient shortcut for ‘dorso-velar’ in one author’s writings, it might be shunned together by another, and might mean something distinct from ‘dorso-velar’ in a third book.

    so allow me to say when you insist that “’Radical’ and ‘dorsal’ are actually natural classes”, you utter a tenable statement within the context of a suitably framed discourse. however, in the discourse as framed here, you are taking the finger for the moon that was pointed at, and now you fight for it. you even supplied the quotes. for ‘radical’ and ‘dorsal’ are not ‘actually natural classes’, they’re just words. suitable words, maybe, but, then, just words, mere terms.

    it is not like the IPA or any other body has the authority or the power to absolutely define what any term, in any language, at any time is to mean. rather—and justly so, because of the very shifting nature of verbal and graphic vocabulary in the field of speech sound science—they provide a terminological standard that you can, if you will, adhere to, to whatever degree you see fits your purpose. as an example, when i wrote my thesis on the phonology of mandarin syllable finals, i chose to replace (among other things) the standard [sɕʂ] symbols by [sśš] as i felt they were better to read and more consistent.

    the class of sounds under discussion here has been variously called ‘dorsal’, ‘dorsovelar’, and ‘guttural’. wikipedia says, ‘in some definitions this is restricted to pharyngeal consonants, but in others includes some but not all velar and uvular consonants’, and it goes on: ‘In colloquial usage, the term is used for any sound pronounced in the throat or near the back of the mouth that is considered “harsh”. […] Phonologists such as Miller (2005) and Pullum & Ladusaw restrict the term guttural to sounds articulated in the throat, which include pharyngeal, epiglottal, and glottal consonants (see radical consonant), and murmured, pharyngealized, and glottalized vowels (see strident vowel).’ so you can see how usages vary.

    that said, we have yet another issue, and that is your insisting that 舌根 is the ‘tongue root’ and thus not appropriate to describe dorsovelar sounds. #1, i am in no position to proscribe to the chinese public at large, or to chinese authors in particular, how to use their language. if they happen to use 舌根 for velar (and they seem to do that a lot; also, my 1984 Das Neue Chinesisch-Deutsche Wörterbuch) corroborates: “舌根音 Velar; Gaumensegellaut” (it has, which i find funny, no entry for 舌根 itself). you might find that deplorable, but it is a fact. i myself wonder what term you want to want to use for sounds from deeper down, but i guess there will be ways to express that concept, too.

    there is also some justification for the term 舌根: in a language which has [ŋkx], but not regularly sounds like [ʕɁħ], the velars are as deep down the tongue as you will get, within that language. therefore, from a particularistic point of view, [t] is tongue tip, and [k] is tongue root. that is understandable. it also helps to explain why ‘guttural’ used to be a popular term in european phonetics for dorso-velar, as languages like like, greek, english and german lack consonants produced from a deeper place than [kgx].

    for similar considerations, in my thesis i also dismissed the ages-old obsession of writers on chinese to insist that since the initial sounds of 跑 and 包 are ‘really’ [pʰ] and [p], they ought to be written as pʿ (or p’) and p. this single, malinformed commitment to ‘correctness’ has resulted in much confusion. it severely hurts romanization efforts in taiwan and korea (even done more harm in that language with absurd orthographical conventions) to this very day. from the a comparative viewpoint, pʿ, p may be suitable symbols; from a particularistic viewpoint, p and b for the fortis and lenis bilabial plosive are much better. this is largely an artifact of the limitations of the roman alphabet, to be sure.

    “A lot of work goes into standardizing phonetic terms” exactly *because* it is such a mess, and because orientation against a well-known standard is so tremendously helpful in this area.

  21. Louis says:


    Ok — in the field of modern SC phonology, limited to academic contexts, ‘dorso-velar’ and ‘velar’ mean the same thing in reference to [k, kʰ, ŋ, x]. Show me otherwise…

    It seems you accept that “radical” and the other terms refer to natural classes. You should have known this, though. Thank you for telling me that they are words. I didn’t know that…

    [k, kʰ, ŋ, x] are standard velar sounds in SC. SC speakers do use a glottal stop and a glottal fricative. They’re not contrastive. You should know this since you’ve written a thesis in SC phonology…Standard English speakers generally use [h] (glottal fricative), not [x]. There is a glottal stop in English. It’s not contrastive. Again, you should know this considering you’re background.

    It’s clear from you’re post that you don’t know why I brought up 舌根. My point is that the characters can be misleading, because the actual meaning of this word is not ‘tongue root’ in every case I’ve seen it used.  

    By the way, do you know why we don’t distinguish p/p’, t/t’, k/k’ in the roman alphabet? You should, as you’re a phonologist.

  22. Louis says:


    喝 [xɤ]⁵⁵

    後 [hou]⁵¹

    [h] is a allophone of [x] for many speakers. /x/ –> [h] __ [ou]or [ei]

    (1) Yen-Hwei Lin (2007): “The Sounds of Chinese.”
    (2) Personal communication with Hsiao, Yu-Chau HSIAO
    (3) Duanmu San (2000): “The Phonology of Standard Chinese.”


    [p’jɘŋ.an] is sometimes said as [pʰiəŋ.ˀan]

    Duanmu San (2000): “The Phonology of Standard Chinese.”
    Chao, Yuen-ren (1968): “A Grammar of Spoken Chinese.”
    Li, Fang-kui (1966): “The zero initial and the zero syllabic.” Language 42:300-302
    Yen-Hwei Lin (2007): “The Sounds of Chinese.”

  23. Louis says:

    One last thing. Modern Chinese scholars have used 舌根 to translate the English term that refers to the place of articulation for [k, kʰ, ŋ, x].

    The term from the Tang Dynsasty 等韻圖 that modern Chinese scholars think refers to ‘velar’ is ’牙音’.

    From 《聲韻學》 by 竺家寧 (Pp.242-243):

    唐宋的學者把聲母分爲五大類,稱爲[五音],亦即唇音、舌音、齒齦、牙音、喉音。又有所謂[七音],就是加上半舌音和半齒音。至於[九音],包括重音、輕音、舌頭、舌上、齒頭、正齒、牙音、喉音、舌齒音.下面把這些傳統的聲韻術語和現代語音學的用語作一比較:(traditional term — modern term)

    1.重唇音-雙唇音 5.齒頭音-舌尖塞摖音  9.半舌音-舌尖邊飲    13.全濁-濁音塞音和塞摖音   
    2.輕唇音-唇齒音 6.正齒音-舌尖面塞摖音 10.半齒音-鼻塞摖音   14.次濁-流音
    3.舌頭音-舌尖塞音和鼻音 7.牙音-舌根音 11.全音-不送氣清音   15.又次清-清摖音
    4.舌上音-舌面塞音和鼻音  8.喉音-喉音      12.次清-送氣清音    16.又次濁-濁摖音

    He provided English glosses as well:

    1. bilabial 2. labio-dental 3. dental plosive and nasal 4. palatal plosive and nasal 5. dental affricate and fricative 6. palato-alveolar affricate and fricative 7. velar 8. glottal 9. dental lateral 10. nasal affricate 11. unaspirated voiceless 12. aspirated voiceless 13. voiced stop affricate 14. liquid sound 15. voiceless fricative 16. voiced fricative

  24. Louis says:


    I misinterpreted the paragraph that starts off as “there is some justification…” I thought you were saying [k, kʰ, ŋ, x] were articulated in the pharynx. You can disregard my paragraph on the back consonants [k, kʰ, ŋ, x,ˀ,h].

    I was a bit rude in my response as well — Sorry!

  25. 慈逢流 says:

    @Louis: “I was a bit rude in my response as well”—no offense here.

    “SC speakers do use a glottal stop and a glottal fricative”, “English speakers generally use [h] (glottal fricative), not [x]”. yeah, i was weary about that. i mean, in german the glottal stop is an important part of the phonetics. on the other hand, the glottal stop in german is not written out, and for the naive german speaker, it simply does not exist. the precise nature of [h] has been the subject of much contention, as a glancing over http://en.wikipedia.org/wiki/Voiceless_glottal_fricative quickly shows. and of course, you get to hear glottal stops by chinese speakers.

    the difficulty for the phonologist here is to deliver a reasonably systematic depiction of a language’s sound system. now what we seem to observe is that while people do make all kinds of sounds with their speech organs, only a limited number of types of sound get admitted as ‘core’ language sounds. for example, to express bewilderment or dismissal, germans utter things like [ts:], [pf:], and [ǀ] (a dental click). yet only the first two appear in any ‘words’ of that language. also, many people use [ǁ] (a lateral click) to express surprise or appreciation. try to teach the same people to use the same sound built right into a word!

    in chinese btw, the same phenomenon exist also on the syllable plane. today, there exist characters for syllables like hng, hm, yo, which traditionally were not accepted as ‘true syllables’. most phonologies of modern chinese likewise (and rightly so) put those into a group apart (so the description you get is multisystematic). i say these things to motivate that when we speak about chinese, english, or german phonology (not phonetics, that’s a whole different story), we can well dispense with the subtleties of the dorsovelar vs radical thing. this is because of the peculiar ways that sounds like [h] and [Ɂ] behave like in those languages.

    “in the field of modern SC phonology, limited to academic contexts, ‘dorso-velar’ and ‘velar’ mean the same thing in reference to [k, kʰ, ŋ, x]”—sorry, i’m afraid you have to narrow that down a bit again. with academic contexts, you probably mean the particular school that you’re studying at, or the particular authors that you’ve read. i’ve never talked about dorsovelar consonants in modern chinese with anyone from the university of pisa, or of waseda, if memory serves. no idea what they prefer over at waseda. i’m not nitpicking or trolling here; i just find it problematic when people come up with so global a claim about the validity of the particular convention they’ve learned. i mean, in china, they eat with chopsticks, and in india, they use their fingers. it’s not like it’s universally correct that you lay out the fork to the left of the knife. in many countries, a fork has no place on the the table. same with words.

    that said, and since we’re only fighting about words here, let me quote wikipedia once more, which is conveniently accessible: “舌根音是按發音部位分類的一類輔音。漢語中,「舌根音」往往指「舌面後音」(軟齶音)。但在一般語音學家的術語里,「舌根」位於「舌面後部」的後下方,與「咽壁」等被動發音部位相對。” i think that about sums it up. 舌根 is not a particularly well chosen term for dorsovelar, but it’s a matter of fact that some authors follow that usage. and yes, of course, characters in some vocabulary *can* be misleading. but come to think of it, that confusion would not be mitigated by abolishing characters, i believe? in alphabetically written languages, you have the same effect.

  26. Louis says:

    @慈逢流 —

    I agree!

  27. Chris H says:

    I would like to add that any written language be it Chinese or English, is a visual representation that links the present to the past and the future. Substituting characters for Pinyin would cut people off from their past. Characters have taken their own evolution through time, just like our western spelling has and it will keep on evolving for better or for worse, I personally feel this whole discussion on substitution of characters for pinyin to be a non-issue only persued by some people frustrated with the slow pace of acquiring any literacy in Chinese (something we all deal with) the emotional appeals are proof of that.

  28. Greg says:

    “…frustrated with the slow pace of acquiring any literacy in Chinese (something we all deal with)…”

    You know Chris, most of us writing here have learned the characters well enough to not have problems with them anymore. This is even more true for the Chinese scholars who propose shuang55wen35zhi51 and the like.

    "any written language be it Chinese or English, is a visual representation that links the present to the past and the future. "

    — How so?

    “, I personally feel this whole discussion on substitution of characters for pinyin to be a non-issue ”

    — I don’t think anyone here is saying the characters should be done away with.

    “the emotional appeals are proof of that.”

    — What emotional appeals are you referring to?

  29. […] Hanzi vs Pinyin, in case you hadn’t heard enough (Sinoglot) […]

Leave a Reply