Perso-Arabic and Sinitic Literacy

« previous post | next post »

In discussions of literacy in contemporary China, because of the unreliability of government statistics and the emotional, controversial nature of the topic, it is sometimes good to adopt a more historical perspective.  Consequently, I shall from time to time write blogs drawing on first-hand records from earlier periods.

On July 23, 1845, a British missionary named George Smith visited a mosque in the city of Ningbo, which is a major commercial city on the coast about a hundred miles south of Shanghai.  He recorded the remarkable observations he made on that occasion in his book entitled A narrative of an exploratory visit to each of the consular cities of China (London:  Seeley, Burnside, & Seeley; New York:  Harper & Brothers, 1847), pp. 154-155:


The old (Muslim) priest was a native of Shantung [VHM:  Shandong, a large province situated on the northeast coast of China; birthplace of Confucius and Mencius], having been sent for thence to Ningpo, forty years ago, according to the custom of supplying the priesthood, on a vacancy, from their original province. After we had taken some tea together, and made an exchange of some trifling presents, he sent his grandson to bring some Arabic books and portions of the Koran, which he appeared to read with great fluency. His knowledge of geographical names exceeded that of the generality of Chinese to be met with in the north of China. He mentioned the countries in which his religion prevailed, among which he named Bokhara, Madras, Turkey, and several places in Arabia. We adjourned into the temple, which was written over with sacred sentences from the Koran, and had a little ark for the sacred books, with a movable pulpit. I had previously supplied him and another Mohammedan with one of the gospels and epistles in Chinese, but was surprised to find, on asking the priest to read some Chinese inscriptions in the temple, that he was unable to decipher a single character, though he speaks the language very well, and has been during forty years a resident in Ningpo.

Later, as described on p.186 of the same book, Smith meets a Muslim shopkeeper, also a native of Shandong, who spoke Chinese but was "ignorant of the written character.  The whole sect appear to devote their studies exclusively to their own sacred language, the Arabic.  His bold features, prominent nose, and restless eye, confirmed the fact of the distinct origin of this descendant of Ishmael."

From the evidence supplied by Smith, we learn that Muslim citizens of the late Qing / Manchu period (the last dynasty [1644-1912]), who were fluent in one or another of the Sinitic languages, were illiterate in Chinese characters but literate in Perso-Arabic script.  This information is all the more stunning in light of the fact that Muslims in China developed a system of writing Sinitic languages (primarily one or another topolect of Mandarin [especially in the northwest], so far as I know), in Perso-Arabic script.  This was called XIAOERJIN or XIAOERJING.  Although the etymology of XIAOERJIN(G) remains uncertain, this Perso-Arabicization of Sinitic was important in the spread of Islam in China.

The above information would seem to indicate that it was much easier to become Sinitically literate in Perso-Arabic script than in Chinese characters.  This conclusion (that it is easier to become literate in an alphabetic script than in a morphosyllabic script) is supported by the fact that tens of thousand of Sinographically illiterate Dungans who fled from persecution in northwest China during the late 19th century acquired literacy in their own northwest Mandarin topolects first through the Perso-Arabic alphabet, then through the Latin alphabet (1928-1953), and since 1953 through the Cyrillic script.

One might also mention the quick acquisition of literacy by tens of thousand of previously illiterate Hokkien (Minnan) speakers through romanization.  But that is a separate topic, one which has been well studied, and one which is hugely relevant for the politico-linguistic situation on Taiwan, so I shall leave further discussion of it for another time.

I thank Sanping Chen for calling Smith's book to my attention.



47 Comments

  1. Jonathan Badger said,

    June 19, 2009 @ 6:52 pm

    Is this controversial? I thought it was fairly well accepted that alphabetical systems are easier to acquire — and both Japan and China have had movements which supported such a transition — but obviously, there is a lot of political reasons (abandoning tradition, fear that various dialects would want their own alphabetization perhaps leading to linguistic fragmentation and separatist movements) to oppose alphabetization.

  2. Mark Liberman said,

    June 19, 2009 @ 7:12 pm

    Jonathan Badger: Is this controversial?

    You'll find plenty of people who argue that an alphabetic system couldn't possibly work for Han languages because there are too many homonyms. See for instance the citations in John DeFrancis's discussion of the indispensibility myth.

  3. Mark F. said,

    June 19, 2009 @ 7:51 pm

    While I suspect it really is easier to learn to read a phonetic script than a logographic one like Chinese, this example doesn't strongly support that. Jews in German-speaking areas used the Hebrew alphabet to write their dialect of German (i.e., Yiddish), and Muslim countries have fairly consistently adopted Arabic script for the local language. In both cases, a lot of value was (and is) placed on reading the respective scriptures in the original language, and so it made sense if possible (given the social circumstances of the time) to use the same writing system for other languages they used.

  4. David said,

    June 19, 2009 @ 7:52 pm

    To puncture one little part of your argument, the literacy statistics for Japan should be pretty accurate, and they are very high.

    [(myl) Sadly, no. More important, probably, is the enormous amount of time taken in Kanji instruction, as well as the style of learning that it imposes.]

    I'm a Japanese (second language) speaker and I would be hugely disappointed to see the Kanji go, probably mostly for sentimental reasons.

    [(myl) I feel the same way about traditional English spelling, which is almost as much of a disaster for learners. But we shouldn't let our personal preferences (and our personal investment in intellectual capital, which is closely related) skew our opinions of what would be best for future generations. ]

    But I don't see how the homonym argument can be dismissed so easily when the dictionary lists 27 different words with the reading こうしょう (koushou). And don't forget that most pages of a South-Korean newspaper still include Hanja in paranthesis for a handful of words, to avoid ambiguity.

    [(myl) The trouble with this argument is that it proves that no one could possibly understand spoken Chinese or Japanese…]

    These and other arguments against romanization may not prove unequivocally that using the roman alphabet would be less effective for Japanese/Chinese than Chinese characters. However I do believe together with the cultural/sentimental argument and the huge effort and disruption a change would entail is enough to conclude that it's not worth it.

    [(myl) Well, it's not about to happen, any more than reform of English spelling is. But let's not pretend that this is the way it should be, if we were starting over with a blank historical slate. ]

  5. Chris said,

    June 19, 2009 @ 8:43 pm

    In the case of Japanese, which uses two syllabaries as well as Chinese characters (Kanji), there is always the possibility of putting the reading next to the kanji using the syllabary (this sort of mark-up is called "ruby" in Japanese). This is commonly done for words written in kanji whose reading is not very familiar, or for characters not found in the government-mandated list of "common use" characters. The novel Shonen H has this sort of mark-up on *all* the Chinese characters, despite being a novel not directly aimed at children. Many comics also use this mark-up on all kanji used. Also, I've found that in many older editions of novels from the Meiji period, this kind of mark-up is commonly found on all the Chinese characters used in the novel.

    It seems to me that a convention like this would be one way to increase the accessibility of written works in Chinese (and thus increase literacy rates), while also preserving the use of Chinese characters.

  6. JonJ said,

    June 20, 2009 @ 12:51 am

    I don't really know anything about Chinese, but I do know something about Japanese, and I think the failure of the Japanese to alphabetize their language (or more likely, kanaize it) is due to more than just the quaintness or the supposed "prestige" of the kanji.

    Japanese speakers conversing with each other in certain kinds of situations not infrequently cannot understand words that are written with kanji unless they stop to actually write them either on paper or with a computer or by tracing them on their palms, etc., if said words are not part of the vocabulary they normally use in conversation–for example, technical terms in a field they are unfamiliar with, etc. For example, doctors will know the words for diseases and therapies that are part of their daily work without having to see the kanji, but lay persons may very likely not know them, and may have to resort to a kanji dictionary.

    David is right that words with very common pronunciations, like "koushou," are written with several dozen kanji compounds. Only a few of them may be used in most daily conversations, of course, and context generally makes it clear which one the speaker means in a given sentence. But once one strays beyond everyday conversational topics, the situation becomes more complicated.

    I was interested to see, when I first went to Japan and began studying the language, the extent to which TV news broadcasts repeat the announcers' spoken words by showing them on the screen. I'm not sure of this, but I suspect that much of the reason for this practice is that viewers will often be unsure what words the announcers are using unless they can see the kanji.

    We English speakers are not very aware of this, but the vocabulary we use in our everyday conversations is also much smaller than the total vocabulary of all the books we need to use. If we used kanji rather than our alphabet, we would probably be doing what the Japanese do. It is much too simple to conclude that the Japanese could easily convert their writing system to a phonetic one simply from the fact that they can understand each other when they speak.

  7. Chris said,

    June 20, 2009 @ 1:04 am

    [(myl) The trouble with this argument is that it proves that no one could possibly understand spoken Chinese or Japanese…]

    One occasionally finds native speakers of Japanese resorting to finger-writing a character (or indicating it by some other means) to disambiguate a word in conversation. In my experience, this is typically only necessitated by words in the Chinese stratum of the lexicon (kango). Many of the words made ambiguous by the large amount of homophony are technical or literary words that rarely appear in oral discourse, but when they do, they are often accompanied by some indication of the characters with which they are written, at least when the listener asks for clarification about what word was intended.

    [(myl) Similarly, English speakers often spell an ambiguous or unusual word, or ask "how do you spell that?" ]

  8. Claw said,

    June 20, 2009 @ 2:04 am

    [(myl) The trouble with this argument is that it proves that no one could possibly understand spoken Chinese or Japanese…]

    This argument comes up because the current writing system seems to encourage more pithy writing, despite the switch to Baihua. For instance, it's understood in writing that 因, 如, and 但 commonly stand for 因为, 如果, and 但是, respectively, so many people just don't bother to write out the full two-character words. In addition, the monosyllabicity may give the text a bit more of a classical feel, which some people prefer. But if you alphabetize these, then you'll get yīn, , and dàn, respectively, which may result in confusion. This is what people cite when they say too many homophones would result.

    However, it's not really an argument because if the system were indeed alphabetized, then the issue wouldn't come up in the first place because people would indeed write the full form instead — yīnwèi, rúguǒ, and dànshì, respectively, which reflects what would actually occur in speech.

    [(myl) Let me also note that Chinese audiobooks seem to be fairly popular, which would be impossible if normal written Chinese were so ambiguous as to be effectively uninterpretable. ]

    So there's no technical reason why you can't alphabetize Chinese, but there are practical and political reasons why it hasn't been done, many of which were already mentioned by Jonathan Badger.

    My personal opinion is that it would necessarily form a linguistic barrier with the non-Mandarin speaking south, who would need to learn Mandarin in order to read and write alphabetized Mandarin. This is not the case with the current logographic writing system, where non-Mandarin speakers can still read and write even without knowing a word of Mandarin.

  9. Claw said,

    June 20, 2009 @ 2:29 am

    I should clarify my last statement:

    Although non-Mandarin Chinese speakers can learn to read and write without knowing a word of spoken Mandarin, they still need to be educated in Standard Written Chinese, which is based on Mandarin grammar and vocabulary. The pronunciation of the characters would still be in their local dialect.

    In my opinion, this is actually not as difficult as learning spoken Mandarin, which to the students, would be akin having to learn a foreign language (e.g., like a French speaker having to learn Italian). Although Standard Written Chinese as pronounced in their local dialects is different from how the actual local dialects are spoken, the differences are big enough to cause an undue burden. While the phonology of the various Chinese languages have diverged a lot, the vocabulary and grammar have not diverged as much, so most speakers actually just consider Standard Written Chinese a higher register of their own speech rather than a different language altogether.

  10. Claw said,

    June 20, 2009 @ 2:39 am

    Oops, I meant, "the differences aren't big enough to cause an undue burden."

  11. Nathan Myers said,

    June 20, 2009 @ 2:46 am

    … non-Mandarin speakers can still read and write even without knowing a word of Mandarin.

    I'm used to hearing this from Mandarin speakers, and to hearing it flatly contradicted by speakers of other Chinese languages, who say those "non-Mandarin speakers" are, in fact, bilingual, in their non-Mandarin language and in Mandarin. I don't know the truth of the matter, which I expect is more interesting in detail than either expression, and I hope to learn more from others here.

  12. J. W. Brewer said,

    June 20, 2009 @ 7:39 am

    Beyond the ruby characters, the pedagogical use of hiragana in Japan (and I suppose the parallel use of chu-yin/zhuyin/bopomofo in the ROC, although I know less about that) for teaching young children before phasing in kanji/hanzi would sort of seem to speak for itself. I guess what it doesn't address is the "disambiguation" facet of the "indispensibility" argument once you get to the full range of adult vocabulary. You could I suppose test that by, e.g., taking various found-in-the-wild samples of "adult" Japanese prose (newspaper stories, excerpts from popular novels, scholarly articles, documentation for software packages, what have you), transliterate them into kana-only form, and see what sort of reading comprehension you did or didn't get and in particular how frequently you got problems with homophones that were actually ambiguous in context. (You could do the same thing with romaji transliteration and see if the results were better/worse than with kana.) I don't know enough about Japanese phonology to know how precise the current kana system is at mapping to the pronunciation, or whether the addition of further diacritical marks would permit further distinctions between words with slightly different pronunciation which would be spelled identically in kana.

  13. Steve Politzer-Ahles said,

    June 20, 2009 @ 8:53 am

    I'm not sure I am understanding Dr. Mair's last claim correctly; perhaps I am missing something. Because these populations of Muslim Chinese were able to write in Arabic script but not Chinese, Arabic must be easier? That seems like a bit of a leap…there are so many other factors behind what script people choose to write (i.e., maybe they just never bothered to try learning Chinese writing, if that's not what was used in their communities). Plus, why couldn't you also claim the reverse—that since eastern Chinese people were learning Chinese rather than Arabic writing (or, for that matter, any other writing), Chinese writing must be easier? That kind of logic doesn't make much sense.

  14. Michael Maxwell said,

    June 20, 2009 @ 11:15 am

    As someone who knows neither Chinese nor Japanese, I have a couple questions.

    David wrote "the dictionary lists 27 different words with the reading こうしょう (koushou)." First, are these really different words (i.e. with different pronunciations), or are (some of) these different senses of the same word?

    Second, I'm guessing this is Japanese, but the original discussion was on Chinese, and I suppose that when people say there are lots of homonyms in Chinese, they're referring to a romanization that leaves off tone marking. But it's easy to mark tone in roman scripts (lots of languages do so); if Chinese were written with a roman script that marked tones, would it be highly ambiguous? I'm guessing not. (On the other hand, I have heard there is considerable dialectal variation in tone even within Mandarin, and a stanardized romanized script would probably choose one of those dialects, making it hard for speakers of other dialects to learn spelling; whereas the present script avoids that problem by making "spelling" hard for everyone…)

  15. Kellen said,

    June 20, 2009 @ 11:58 am

    The other issue with xiaoerjing aside from homophones is that a number of syllables which would be different in pinyin take the same form in xiaoerjing. 金 and 经, jin and jing in pinyin respectively, would both be written [دِ] in xiaoerjing. that may be one reason for the two different spellings of the system itself: both would be written [شِيَوْ عَر دِ]. That and the fact that there was never a real standardisation (to my knowledge) would make it a real mess if it had ever really survived into modern times. At one point I wrote out a quick javascript that will convert from pinyin into xiaoerjing, but I ended up having to take some liberties in order to clear up conflicts in the form of ambiguity that resulted from different ways of transcribing the characters.

    Forgive the plug, but if you have a Mac you can click on the link in my name to get a widget for the Dashboard that includes the script, for anyone interested.

  16. Kellen said,

    June 20, 2009 @ 12:13 pm

    @Michael: Yes. Even with tones included, it would still be terribly ambiguous. 九 (9) 久 (a long time) and 酒 (alcohol) are all "jiu", third tone. That's not even one of the more troublesome syllables. There are at least 5 common words for "shi" in the first tone, 5 for the second, 3 or 4 for the third tone and 10 for the fourth tone.

    The standardised spelling uses standard tones, however were you to use Pinyin to write a dialect, assuming it was one where the only notable different were tones, you'd probably not rely on the same manner of writing tones. If you tell someone a word is the 6th tone in the Suzhou dialect of Wu, that's pretty meaningless even to someone from Suzhou. Giving some representation of the contour would be much more useful, e.g. ˨˩˦ for Mandarin third tone.

  17. Mary Kuhner said,

    June 20, 2009 @ 1:11 pm

    The school of aikido (a Japanese martial art) that I study is called "Shin Shin Toitsu Aikido", aikido with oneness of mind and body. The two "Shin" are pronounced just the same as far as I can hear, and would certainly be the same hiragana. It would be a pretty puzzle, unless the phrase is an idiom, to decide if it meant "oneness of mind and body" or perhaps "oneness of [your] body and [my] body" which would make equal sense in context, or even "oneness of [your] mind and [my] mind."

    My Japanese is rudimentary: can anyone tell me if this "shin shin" is an idiom, or is it really ambiguous? Is there any way, short of writing it in kanji, to tell which shin is mind and which is body?

    (All of this is not an argument against readability of Japanese written in hiragana. I haven't the experience to make such an argument in either direction.)

  18. jfruh said,

    June 20, 2009 @ 1:12 pm

    I have no professional opinion on whether logographic scripts like that used for Chinese are harder to learn than alphabetic scripts like that used for Arabic, though common sense would seem to indicate that they are. However, I don't see what exactly about the anecdote about the missionary and Chinese Muslim would lead you to that conclusion.

    In the anecdote, we have a Chinese Muslim cleric who is fluent (presumably natively) in whatever the local Chinese dialect is, but who can't read Chinese characters. However, he does read Arabic. Presumably as a religious scholar of some standing he reads Arabic in the sense that he can understand what he reads, not just that he knows what letters denote what sounds and can sound out the words. In other words, he's become literate in his second language but not in his first.

    While this is definitely striking, it certainly doesn't mean that he chose to learn to read in Arabic because it was easier than learning to read Chinese. Arabic — particularly written Arabic — has a huge importance in Islam, because of the divine qualities attributed to the written text of the Qu'ran. Any Muslim who chooses to become a religious scholar would ex officio need to become reasonably literate in Arabic, no matter what their native language.

    In constrast, assuming that in mid-19th century China a substantial portion of the population was illiterate (I'm guessing this is the case, but please correct me if not?), there might not be any particular reason why even a learned Muslim scholar would need to be able to read Chinese. Written Chinese was no doubt important for those in the administration — both as a practical necessity and as a means of indicating one's membership in the literate, literary class that produced administrators — but since the society wasn't geared on the assumption that most of the common people were literate, our Muslim cleric didn't need to know written Chinese to get by day to day, and probably wasn't a member of the social group that valued Chinese literacy.

    It's certainly odd from a Western perspective (both to the British missionary and to us) to find an obviously educated man who couldn't read his own native language. But my guess is that that's because literacy had a different function in 19th century Chinese society than it does in ours, not because logographic scripts are harder to learn than alphabetic scripts (even if they are in fact harder).

    A point of comparison might be to see whether rabbis in Eastern Europe were literate in Russian/Polish/Lithuanian/etc. as well as Hebrew (though I suppose that they might not have even been fluent in those languages, if they lived in Yiddish-speaking enclaves.) As a side note, does anyone know if Buddhists in China tended to be literate in Sanskrit?

  19. J. W. Brewer said,

    June 20, 2009 @ 2:21 pm

    @jfruh, it's not an area I know much about, but my impression is that by some point circa the 8th or 9th century AD the canonical Buddhist texts had all been translated from Pali (not Sanskrit) into Chinese and thereafter Chinese Buddhism developed and flourished without needing or possessing any significant number of individuals who could read the original (and of course Buddhist by then was becoming increasingly marginal back in India). Not really any weirder than the dearth of even clergymen/intellectuals who could read Greek among the Christian population of Western Europe in the Middle Ages.

  20. Ryan said,

    June 20, 2009 @ 3:01 pm

    J. W. Brewer: There is a pretty close correspondence in standard Japanese to what is written and what's pronounced. There is certainly some allophonic variation but, to the extent of my knowledge, こうしょう is こうしょう, no matter what kanji one uses to write it. There may or may not be senses of the word disambiguated by pitch accent, but that would be a regional feature.

    Michael Maxwell: They're different words, pronounced the same (although pitch accent might create a phonemic contrast, I'm not too sure). You can do a search here and see for yourself: http://www.csse.monash.edu.au/~jwb/cgi-bin/wwwjdic.cgi?1C

  21. Bob Ladd said,

    June 20, 2009 @ 4:53 pm

    @jfruh and Mark F.: You definitely have a point about cultural prestige and/or religious motivation being an important factor in the case discussed in the main post: it's quite common in Africa for educated people to be much more comfortable reading French or English than reading their own native language, even though African languages are generally written in the Roman alphabet just like English and French. But that doesn't mean that the difficulty of learning a logographic system isn't also playing a role.
    @ J.W.Brewer: the experiment you suggest (seeing whether Japanese speakers have trouble with kana-only text) will not straightforwardly demonstrate whether kanji have an important disambiguating function, because the people you are talking about already read kanji fluently, and any detrimental effect of reading a text in kana could be attributed to its unfamiliarity. Anyone who has ever tried reading English written in phonetic transcription will know what I mean.

  22. jfruh said,

    June 20, 2009 @ 5:38 pm

    @Bob Ladd: "But that doesn't mean that the difficulty of learning a logographic system isn't also playing a role."

    I guess what I'm trying to get my head around, when trying to figure out why this anecdote has been cited, is — playing a role in what, exactly? The fact that our Muslim cleric doesn't read Chinese? Well, sure, but in that sense it plays the same role as it would for illiterate Chinese who can't read other langauages either.

    Is the point supposed to be that the cleric said "Learning to read Chinese is too hard, so I'm going to learn an entirely separate language and become literate in that instead?" That seems extremely unlikely to me. Or is the point just "Look, this guy was obviously capable of learning to read, so the fact that he didn't learn to read Chinese indicates that learning Chinese was really hard?" I guess just don't buy that logical leap. He had a huge motivation for learning to read Arabic that he didn't have for learning Chinese. I mean, I buy the theory that learning to read Chinese is hard, harder than learning to read an alphabet, I really do, I just don't think this anecdote serves to buttress that theory, and it appears that it's being presented as something that buttresses that theory.

  23. Suzanne McCarthy said,

    June 20, 2009 @ 5:44 pm

    This conclusion (that it is easier to become literate in an alphabetic script than in a morphosyllabic script)

    This is an odd statement in itself, contrasting morphosyllabic and alphabetic. Some argue that it is easier to become literate in a homographic syllabic script such as Cree Syllabics or hiragana. Certainly syllabic literacy is an early stage for children, preceding the alphabetic stage which requires analysis which does not come naturally to many. John Berry's work on Cree syllabics certainly suggests that literacy in this script was once widespread outside of formal education of any kind.

    On the other hand a syllabic system which is heterographic, as is Han Chinese, makes enormous demands on the memory, so it takes longer to acquire basic fluency.

    In some romance languages even now, the syllables are taught to children first, before words, or any expectation of alphabetic literacy. This used to be the case in English too, but it is not terribly efficient in English.

    Of course, the question remains, does a heterographic syllabary make too high demands on a people.

    However, this is partly overridden by the ability now to input pinyin and produce characters.

    One thing I have not seen mentioned yet, but perhaps I missed it, is that as pinyin becomes written with multisyllabic words, and not monosyllabic, there are fewer homonyms. This is, now the latest software works so that as you type in pinyin, one syllable after another disambiguating the first syllable in the sequence, it is no longer necessary to choose the correct character, but it is chosen by the software.

  24. Bob Ladd said,

    June 20, 2009 @ 5:46 pm

    Discussion of the homonym-distinguishing function of Chinese characters seldom mentions the fact that this function is actually also rather common in written English – there are hundreds of pairs like night, knight; genes, jeans; mail, male; friar, fryer; sheer, shear; foul, fowl; die, dye; weigh, way; prey, pray; a fair number of triplets like flu, flue, flew; their, there, they're; you, ewe, yew; for, four, fore; check, Czech, cheque; raise, raze, rays; hue, hew, Hugh; road, rode, rowed (though not all of those work in all varieties of English); and at least one quadruple (or, oar, ore, o'er). The same is even truer in French (sang, cent, sans, sent; vert, verre, ver, vers, vair; etc.). As with Chinese, so in French and English: once you have learned to make these written distinctions they seem important (and as I said in my earlier comment, they actually are important for a fluent reader). But learning to make them places a considerable burden on the learner, and in all cases the fact that the spoken language is perfectly intelligible makes it hard to believe that in the written language distinguishing homonyms is somehow indispensable.

  25. Suzanne McCarthy said,

    June 20, 2009 @ 5:46 pm

    I notice my link was broken so here it is corrected.

    I do think Victor Mair has a strong point here, but I feel it deserves more explicit development.

  26. Alexandra said,

    June 20, 2009 @ 9:16 pm

    @ Suzanne McCarthy:

    "In some romance languages even now, the syllables are taught to children first, before words, or any expectation of alphabetic literacy."

    This sounds fascinating, but I don't quite understand what you mean. Could you give an example?

    (I hope this isn't going too far off topic.)

  27. Matthew Stuckwisch said,

    June 20, 2009 @ 9:52 pm

    Alexandra: in Spanish (in Spain at least, and judging from reading Mafalda in Argentina too) children are taught the alphabet learning each letter paired with the vowels. So they learn m as ma, me, mi, mo, and mu, andthen learn phrases that only use those or have a high frequency of them ("mi mamá me ama" and such). In the books, the consonant-vowel grouping is often highlighted in some way together. It actually works well given Spanish's structure, but I've always thought in the back of my head it really hurts them later on for learning languages that have more consonants in clusters and even to an extent with closed syllables in their own language.

  28. Suzanne McCarthy said,

    June 20, 2009 @ 10:09 pm

    English literacy was at one time taught by getting children to recite the syllables off a hornbook. ba be bi bo bu and so on. This practice has completely disappeared in English to the point that we are no longer aware of it. However, children are taught to read French this way, even now. And we still have the wooden plates from which children learned the syllables in ancient Greece. It is a very enduring practice.

    This does not negate Mairs post.

  29. Etienne said,

    June 20, 2009 @ 11:45 pm

    To jfruh: actually, in pre-modern Europe situations such as the one described in Professor Mair's posting would have been common, indeed the norm: Medieval Jews, whose mother tongue was typically that of their co-territorial Christian neighbors, became literate in Hebrew first, and often didn't become literate in any other language: for Romance speakers in the pre-reconquista Iberian peninsula becoming literate meant becoming literate in Classical Arabic: at the other end of Romance-speaking Europe, for centuries in (what is today) Romania and Moldova literacy could ONLY mean literacy in Church Slavonic, and for all too many Western Christians (whatever their L1) literacy meant literacy in Latin.

    To thereby conclude that Medieval Hebrew, Classical Arabic, Church Slavonic and Latin were easier to acquire as written languages than any contemporary vernacular would be patently absurd, of course. Almost as much of Professor Mair's "point". After all, the persistence of Chinese characters in China and Japan to the present day (despite repeated contact with alphabetical and/or syllabic scripts) could just as easily be used as an "argument" in favor of the claim that they are a uniquely efficient/beautiful/ [insert whichever flattering adjective fits your fancy here]/ whatever/ writing system.

  30. Greg Morrow said,

    June 21, 2009 @ 12:55 am

    I assert as obvious that the spoken language will contain enough disambiguation in the speech act to achieve communication, for the simple reason that any version of the language that did not would be superseded in a generation by a version of the language that did.

    Therefore, I assume that in spoken Chinese(-group languages) that there are tactics used to distinguish, e.g., the first tone shi when more than one could be salient. In English, for example, one can clarify that one is going snow skiing in order to avoid ambiguity over which kind of skiing you mean.

    The written language tackles such problems in a quite different fashion, I think. The spoken language has a very broad error channel, in which the speaker and listener constantly signal each other to ensure that communication is successful. The written language does not have that feedback, so the writer has to put more work into making sure the context is clear and that the structure is solid. If a character announces that they're going skiing, the writer can solve the potential ambiguity by showing the character at the ski resort or on the lake, an option not available to a speaker.

    A language with a Sinitic-derived writing system uses the difference in characters to avoid homophone ambiguity, which is cheap for the writer, but expensive for the reader, who has to memorize and process thousands of characters. So if such a language switched to an alphabetic system, what you'd see is a shift of the load from the reader to the writer. In short, you'd see a distinct style change in the written language. And that change would be one with a lot more specificity and care to explain thoughts more thoroughly. And everyone experienced in the old way of writing would be appalled at how much this new way of writing was dumbing down the language.

    But writing is conventional. And one of the virtues of the alphabetic convention is that far more people can be literate on a practical basis. According to Wikipedia, one of the scholars' nicknames for Korean Hangul was something like "writing learned in an afternoon". And they meant it as a criticism!

  31. jo said,

    June 21, 2009 @ 8:02 am

    @Mary Kuhner

    The 'shinshin' in 'shinshin tooitsu' means 'mind and body' – it is two different Sino-Japanese morphemes (身 shin, 'body'; and 心 shin, 'heart/mind') which both happen to have the same sound. Usually it is written 心身, but 身心 is also attested. The whole phrase means something like 'harmony/equilibrium of mind and body'. Even fluent/native writers of Japanese can be unsure of the conventional order when writing (not saying, of course!) these kinds of Sino-Japanese words where there are two morphemes with the same sound, e.g. 収集 shuushuu, 'collection', 収拾 shuushuu, 'putting things in order'. As a side point, there are of course words that are the same Sino-Japanese morpheme repeated, such as 空空 (kuukuu, 'completely empty'), but in the case of shinshin, 心心 and 身身 do not seem to exist as potential interpretations.

  32. Mark F. said,

    June 21, 2009 @ 12:01 pm

    Bob Ladd said,

    You definitely have a point about cultural prestige and/or religious motivation being an important factor in the case discussed in the main post: it's quite common in Africa for educated people to be much more comfortable reading French or English than reading their own native language, even though African languages are generally written in the Roman alphabet just like English and French. But that doesn't mean that the difficulty of learning a logographic system isn't also playing a role.

    I still think the example of Yiddish is more to the point. Middle German dialects, I think, were already being written with the Roman alphabet, but Jews transcribed their dialect of it (or, if you prefer, their closely related language) with the Hebrew alphabet. Ladino also used to be written in Hebrew characters, even though Spanish uses the Roman alphabet. So there are good examples of religious minorities speaking a language that already had an alphabet, but choosing instead to write it with their own alphabet. So, even if Chinese had a simple phonetic alphabet (presumably encoding tones too), I imagine these Muslim clerics, under these circumstances, would have written it with the Arabic alphabet.

  33. ^love*encounter~flow said,

    June 21, 2009 @ 12:16 pm

    when will people stop to advocate the abolishment of chinese characters because of their purported relative complexity when compared to the latin letters?

    when i look at the arabic script with its many deficiencies (sorry but only three vowels expressable as, you choose, consonants and/or optional diacritics; also, really arcane orthographic rules the moment you dig a littler deeper) i wonder wether it shouldn't be replaced with latin letters on grounds of its complexity, too! i do not seriously suggest that, but at least some turks feel that atatürk for one did the right thing when he converted a non-arabic language away from using the arabic script.

    i have been 'collecting' chinese characters for years and had countless discussions with a lot of people over various aspects of these. i am a fan of chinese writing but i am here any moment to demonstrate that the system is 'broken' in more than one way. yet i do not advocate their abolishment. in my view, chinese characters work pretty well for both china and japan (and, potentially, korea, where they are not so popular these days). the world's most intricate orthography hasn't stopped japan from rising to one of the leading nations in terms of economic prowess, cultural visibility, technological innovation, and, yes, lieracy levels. it doesn't stop china from experiencing economic growth at breathtaking levels right now either. this flies right into the face of well over 90% of any romanization advocacy, be it the one made at the coffee-table or the one argued for in books, committees and parliaments.

    i take it nobody here is seriously suggesting that, of all theoretically possible ways to write chinese and japanese alphabetically, 小兒經 is one of the recommendable ways (disclaimer: I take interest in that subject and have devised several experiments, one of them using Hangŭl—just to say that I know how hard it is and I know it cannot be done "right" or "perfect", as it is logically contradictory given present-day phonological insights).

    i would have to look at it much closer than I have done up to now, but the thought of writing chinese, a language rich in vocalic nuance (yao, wai, nian) and relatively sparse consonants (and absence of the guttural wealth of arabic) using the arabic script really strikes me as… challenging. you would have to introduce a lot of extra features to arabic to make that work well. I adore the arabic script for its beauty, history and fitness for arabic, but the most promosing role model for the alphabetization of chinese speakers it is quite clearly not.

    i would like to add that ease of learning is just one of the factors that help humans to determine a measure for worthwhile-ity of doing it. if ease of learning were so all-important in the arts and crafts of humankind, for sure no-one would have ever learned to walk on a rope. that said, there sure is a potential to evolve chinese writing to make it more accessible. when the communist writing reforms mandated the use of 从 for 從 and 网 for 網 they clearly did the exactly right thing; in other cases, i find the choices less fortunate.—reading chinese sentences in pinyin is a whole different experience when compared to characters; as Greg said, doing so changes the way you express yourself in the written. it is more than a small step.

  34. ^love*encounter~flow said,

    June 21, 2009 @ 3:15 pm

    "we shouldn't let our personal preferences […] skew our opinions of what would be best for future generations"—maybe right so i say people let's just level this entire place and convert it to a vast highway. future generations will thank us for removing all the impediments and difficult left turns so they can go as the crow flies wherever they want to go in their cars. well there won't be much to go then but hey you gotta pay a price right. what's that crow thing again?

  35. hsknotes said,

    June 21, 2009 @ 3:57 pm

    the impossibility of crows?

  36. Etienne said,

    June 21, 2009 @ 5:59 pm

    To Mark F: while Middle High German and Iberian Romance dialects were indeed written in the Roman script (by a tiny minority of monks and priests who had typically first acquired literacy in Latin), from the vantage point of Jewish speakers of those forms of Middle High German which were to become Yiddish and of Iberian Romance which were to become Ladino this written tradition (bound as it was to a religion not their own) was entirely irrelevant.

    Medieval Jews in Europe, whatever their L1, first became literate in Hebrew, so that the use of Hebrew script to write Yiddish or Ladino wasn't a "rejection" or a change of script: it was simply using the most convenient script (i.e. the one they had first become literate in) to write their vernacular. Which, incidentally, is why Yiddish words of Hebrew origin are spelled etymologically (rather than as they are pronounced in Yiddish): because all writers of Yiddish in the Middle ages had first become literate in Hebrew it was natural for them, when writing the vernacular, to "fall back" upon the Hebrew spelling of those Hebrew words found in the vernacular.

    Reading the various comments, I get the impression that many people assume that vernacular literacy (i.e. literacy in one's L1) is the norm: however true this may be in most "advanced" democracies today, it most certainly is NOT the norm historically, either in the West or in pre-modern China: indeed, in Japan, Korea and Vietnam (Classical) Chinese was for centuries the first (and often sole) language of literacy, which of course says nothing about the inherent superiority/inferiority of Chinese characters or of Classical Chinese: all it says is that Classical Chinese enjoyed remarkable prestige in the above-named East Asian countries, as much prestige as Arabic and its script enjoy in the Muslim world.

  37. Dmishin said,

    June 22, 2009 @ 5:25 am

    It is hard to deny the fact that the continued use of Chinese characters has strong political and cultural reasons.
    But I would like to note that Chinese characters just fit Chinese language very well. They provide compact, expressive and unambiguous way to write Chinese. I think that the correspondence [one root] = [one syllable] = [one character] is the main force that allowed Chinese ideograms to survive while all other ideographic writing systems evolved to phonetic scripts thousand years ago.
    Yes, Chinese characters require some additional efforts from learner, but is it really different from English, with all its non-quite-phonetic orthography?
    In fact, I think that writing system like Chinese is better then phonetic. But, alas, such system can't be effective applied to the languages with inflections (and the quirky Japanese writing system, using both phonetics and ideographs is good example).

  38. vanya said,

    June 22, 2009 @ 10:48 am

    Victor Muir complains about the tremendous amount of time Japanese (and implicitly Chinese) spend learning characters and the style of learning it imposes. Is there any evidence that this time would be better spend doing something else? Is there any real evidence that characters have hurt the development of Japan or China in any way? Japan and China are both far more literate societies than Brazil or Mexico – where the Latin alphabet has been in use for 500 years. Have you ever been to a book store in Mexico? Pretty sad. Hell, Japan and China are more literate societies than Italy, where the Latin alphabet was invented. Japan and China also appear to doing pretty well moving into the modern age compared to Indonesia or the Philippines where the Latin alphabet dominates. True, Vietnam and South Korea don't seem to have suffered from moving away from characters, but they don't seem to have made any significant strides in efficiency or pedagogical techniques that Japan or China have missed. This is why Muir's rants against characters strike me as grumpy old man rhetoric. Just as all the English language spelling reformers seem to ignore the fact that somehow our inefficient system hasn't stopped the spread of English around the globe, reformers of Chinese or Japanese can't really explain how life will actually be better for Chinese or Japanese in the glorious age of Romanization. When I hear people talk about "efficiency" I'm tempted to reach for my revolver in self-protection. There's no evidence it really matters much either way – so why destroy the links to 4000 years of culture for some small almost unmeasurable gain in efficiency? Haven't the Chinese destroyed enough of their heritage over the past 60 years?

  39. k said,

    June 22, 2009 @ 1:46 pm

    So, we are only going to get irrelevant nationalistic rants by people with chips on their shoulders who didn't bother to even read the article much less the comments?

    I was hoping someone would answer Nathan's question.

  40. ^love*encounter~flow said,

    June 22, 2009 @ 1:53 pm

    …"I think that the correspondence [one root] = [one syllable] = [one character] is the main force that allowed Chinese ideograms to survive while all other ideographic writing systems evolved to phonetic scripts thousand years ago." almost looks like the chinese characters found sth like a local (or global?) energy minum, a gravitational valley, while all the other ideographic scripts rolled down the one or the other side towards some kind of more phonetic speeling, or into oblivion (cuneiform comes to mind, which did both).

  41. John Cowan said,

    June 22, 2009 @ 8:21 pm

    Vanya: A waste of time remains a deadweight loss, even if it's not enough to drag you under.

  42. vanya said,

    June 22, 2009 @ 10:46 pm

    But there's no evidence presented that learning characters is "a waste of time." Just an assumption. Or, more likely, most of the school day is a waste of time for most children. Filling it with character study is apparently no worse than social studies in the West.

  43. Ken Brown said,

    June 23, 2009 @ 1:48 pm

    Bob Ladd said: "… at least one quadruple (or, oar, ore, o'er)."

    Those are homonyms of "awe" and "aw" for me. So that might be a sixer. Though I'm not sure how I would say "o'er" if I ever had to. It might be two syllables. I think it might usually rhyme with "mower" not with "more" As its only likely to come up in verse or song I suppose it would depend on the stress pattern of the line it was in.

    Also "for", "fore", "four" and are the same as "faugh" Its a pity there is no English word "faw", though there are foreign place names with that word in.

  44. Pettis said,

    June 25, 2009 @ 5:45 pm

    Characterizing certain arguments for the continued use of Chinese characters (in China, Japan, and elsewhere) as cultural/sentimental/traditional overlooks a very important point: eliminating characters would effectively block native Chinese and Japanese speakers' access to thousands of years' worth of printed matter.Is it fair to describe a desire to retain the ability to read pre-21st century texts as "sentimentality" or "clinging to tradition'"?

    Of course, one might argue that the most important works could be transcribed to the new standard alphabet/syllabary or that only those who truly "need" to read them (scholars? lawmakers? judges? clergy?) would be forced to study the "inefficient" hanzi/kanji. From my own limited conversations with South Korean graduate students in the US, it seems that both of these are the case (namely the expanded study of hanja among university and graduate students).

    I'm not sure whether this proves that South Koreans have had difficulty fully abandoning Chinese characters or suggests that use of the characters could be significantly pared down in Chinese or Japanese. And I don't know what insights a consideration of the Vietnamese case might add. But I do think this point about retaining access to a vast body of knowledge–whether philosophic, religious, or scientific–in texts written using characters deserves more serious consideration when we debate script reform for the writing of Chinese or Japanese.

  45. Kacie Landrum said,

    July 2, 2009 @ 8:00 am

    True, we English people could probably get by if we spelled all homonyms (read and red, knight and night) the same, but I'm not so sure the same thing could be said for Japanese. I've been studying kanji intensively (a couple of hours a day) for the past year, and can testify that exam problems where the sentence is written all in hiragana and the student has to pick out the correct kanji are next to impossible. Frequently I run into problems that even my junior high students can't solve.

    In the one level I am studying (not all daily-use kanji in general, just the one level of the test I am studying for), I have to distinguish between 6 characters pronounced KAN, 13 pronounced KEN, 30 pronounced KOU, 13 pronounced SHIN, 14 pronounced SEI, etc.

    And I can personally attest that I see Japanese people write out kanji on their hands for clarification on a weekly basis, but can't think of the last time I ever had to spell anything out in English (to a native speaker, at least).

    So, yeah, attempts to replace kanji with a more 'logical' script will not be getting my vote of support.

  46. David Brophy said,

    July 12, 2009 @ 12:35 am

    I'd query the extent to which Muslims in China ever became "Sinitically literate" in Xiaoerjing. Are there any examples of sustained prose texts written in Xiaoerjing? I've only ever seen it used to gloss texts in other languages. The wikipedia entry on Xiaoerjing shows a complete facing-page translation into Xiaoerjing, but my guess is that the Chinese was used as a crutch for those reading the Arabic, rather than something to be understood as a stand-alone text. I've seen a number of manuscripts of Persian literature with words glossed in the margins in Xiaoerjing, and I understand that manuscripts of the Quran can be found which display similar glossing practices. Can someone clarify whether Xiaoerjing was ever used more widely than this? If it wasn't, it's hard to see how it serves as a good example of the alphabetization of Sinitic languages.

  47. Leah Aharoni said,

    October 14, 2010 @ 6:51 am

    @jfruh Here is some anecdotal evidence regarding local language literacy of Eastern European Jews. I posses several vital records from a Polish town circa 1870-1890. The records are in Russian and the overwhelming majority state that the rabbis officiating/witnessing the events were not literate in Russian.

RSS feed for comments on this post