Responses to objections to romanization

by Y.R. Chao (Zhao Yuanren / Zhào Yuánrèn / 趙元任) (1892-1982)

Webmaster's note: Y.R. Chao, the great linguist, wrote this in 1916 as part of other remarks by him in favor of romanization. For related essays, see Lü Shuxiang's "Comparing Chinese Characters and a Chinese Spelling Script — an evening conversation on the reform of Chinese characters" (which is also in the form of a discussion) and Lu Xun's "An Outsider's Chats about Written Language." For more such writings in English by Chao himself, see Linguistic Essays by Yuenren Chao.

Sixteen Objections to Alphabetization

Obj. 1 You advocate an alphabet for writing Chinese without being able to show me one. Until you have worked out a system that is actually practicable, I am not convinced.

Rep. 1 I spelled out a short passage a long time ago with an extemporized system. When I saw it again a year or so after. I found that I could still read it. But I grant that my estimation of the possibility of alphabetization from phonetic and other philological facts is not entirely free from subjective factors, until I can show you a satisfactory alphabet.1

Obj. 2 All changes of tradition involves economic wastes. Your alphabetization is a great change.

Rep. 2 I cannot help it. I give up the point, consoling myself with future great savings.

Obj. 3 You were advocating reforms in the language without alphabetization. Now you are advocating alphabetization. As if the next generation will not be burdened enough with one language!

Rep. 3 We have to use a pronouncing alphabet anyhow, which is a help and not a burden. The study of characters, on the other hand, helps to bring out the etymology of words, and educated persons should study the characters 古文, , to the same extent as English speaking people study Anglo Saxon, Latin, Greek, etc. We don't have to learn two languages.2

Obj. 4 Since you say the literary idiom is not auditorily intelligible, and since the bulk of existing literature is in this idiom, we should lose it by alphabetizing our language.

Rep. 4 My dear friend, there you are striking at the heart of things. A language and a nation must exist by its literature, and before a sufficient development of speakable literature is developed, the old literature must be taught and learned and lived by those whose educational opportunities can afford it. However, we can preserve a good deal of the value of existing literature by paraphrasing it.3

Obj. 5 Why should you insist on a speakable literature? Answer me, does the spoken language actually sound vulgar or does it not?

Rep. 5 Yes, it does, because not much serious literature has been written in it, and consequently it only gives associations of the vulgar. But the same objections were raised by scholars in the European countries when people began to write their literature in the vulgar tongue, that is, in their national language, instead of Latin. And now, is German vulgar? Is English vulgar?

Obj. 6 Language is a growth, it must not be tampered with by artificial devices.

Rep. 6 I already answered this in my first article. The actual tendency of our language is toward unification of spoken and written idioms, and the logical outcome (see argument(l)) is a phonetic writing. Sticking to monosyllables and "dead" words, trying to coin new monosyllabic characters for new ideas and neglecting the spoken idiom — these are ways of 扭天行事.

Obj. 7 The written language is the only common medium of communication. If we alphabetize our dialects, the result is national disintegration.

Rep. 7 We shall not use any dialect as the standard, but a standard idiom between Mandarin (as distinguished from Pekingese) and the written idiom, according to the principles I discussed under 4 and 5. We are approaching this sort of standard before, or even without our language being alphabetized.4 As to the variations in dialectal idioms, they are no better off in characters than in letters. Just see how much a Pekingese can get out of a novel written in Shanghai or Cantonese dialect.

Obj. 8 Alphabetized Chinese loses its etymology.

Rep. 8 This argument is like that often urged against simplified English spelling and is to be met similarly. In actual usage, how much attention do we give to etymology in words like 學, 暴, 發, 旋, 之, through, draught, etiquette, row, disaster?5 Of how many of these very common words do you know the original meaning? It is not to be denied, of course, that it is useful to know the etymology of words by looking them up, and our future dictionaries of alphabetized polysyllabic words should no doubt give their derivations.

Obj. 9 Your phonetic writing give the sound but not the meaning; my ideographic writing gives the meaning but not the sound. But meaning is obviously more important than the sound. Therefore the ideographic writing is intrinsically superior.6

Rep. 9 Meaning is more important than sound, and this fact would be an objection but for another fact that meaning is in most cases carried in the sounds themselves. On the one hand, the suggestiveness of ideographic characters is often greatly exaggerated. When I learned the words in 學, 暴, 發, 旋, 之, I did not get any help from their modern forms or their ancient forms, nor do I get any help from their etymology as I use them now. On the other hand, the child does not have to learn what the sound mamma means, or dog or cat, or hungry, or come, or will, or soon, or with. He learns the ideas, which are often the words themselves, by experience. His language and thought grows together as one. When in later life he sees or hears a word which he does not understand, the trouble is not with their sound, but with his not having learned them as simple or complex ideas. And by no symbolic contrivance can you get a liberal education quick by trying to go around hard and solid experience and life. This common psychological fallacy comes from the fact that we Chinese students have had the relatively abnormal condition of having to learn a new language after having acquired a rich stock of ideas.

Obj. 10 Even by using polysyllabic words, they still have ambiguous sounds. For instance 見識, 劍式; 清楚, 青 (醋); 年高, 黏糕; 十步階, 拾(布)街.7< In Pekingese, rice toast, the elbow, and pasteboard are all called kə-pa.

Rep. 10 Such cases are relatively rare. Homonyms also occur in English, such as can, count, irony8, lie, might, and those of different spellings are still more numerous.9 Yet they are not numerous enough to interfere with auditory intelligibility. In French, homonyms are still more numerous. I am not advocating the use of extra letters for making distinctions, but here again, we might apply the principle of regular variation.10 For, although in French there are more silent letters than in English, yet they come in so regularly that French spelling is much easier than English. Another way to avoid ambiguities is to choose favorable sounds (see above 5. (a)).

Obj. 11 If in the future most of our words are polysyllabic and can be learned without going to their etymology or monosyllabic roots, then since there are more combinations than roots, we should have more words to learn. This is why the English language has many times more words than the few thousand common Chinese characters.

Rep. 11 Just open one of our new dictionaries, the best being Ts'ï Yuen (辭源). Most of the combinations given there are single polysyllabic words. Do you think that as soon as you have learned the meanings of the four or five thousand important characters, you can (a) express all ideas by synthesizing them apart from conventional combinations and (b) can understand the meaning of all compounds in Ts'ï Yuen without looking up their definitions? You might as well try to understand and use the German inseparable verbs from their prefix and stem. The polysyllabic words simply have to be learned, whether they are written ideographically or phonetically being irrelevant.

Obj. 12 Continuous letters are less beautiful than the articulated characters, and the short list offer less variety of artistic expression. Compare a scroll written in Roman letters and one in Yen, Liu, Eu, Chao styles.11

Rep. 12 I appreciate this from the bottom of my heart.12 But then, we have to consider essentials. There is nothing to hinder us from writing characters as an art.

Obj. 13 Chinese surnames and other names would be ambiguous if alphabetized.

Rep. 13 This, again, is a real difficulty for the time being. There must be a special system of identifying and distinguishing the few hundred names. But if the language is growing, as it is, more and more polysyllabic, thus demanding an alphabetic writing, in a few generations the Chinese will naturally find long names no longer cumbersome.13

Obj. 14 A phonetic language like English is hard to spell. The precipitation of water might be spelled rane, rein, rene or reign; it is hard to pronounce, e.g., in bough, though, through, enough, cough, ought, what is the sound of ough?

Rep. 14 The difficulty lies precisely in its being unphonetic. The irregular spelling of English has philological reasons. Silent gh, e.g., was pronounced like ch in German ach. If the Chinese alphabet is made regular, there will be only one spelling to each syllable.

Obj. 15 Very well, if the English language has changed, so that its spelling becomes obsolete, your Chinese alphabet will also become obsolete; in time, though it may be regular to start with.

Rep. 15 We can modify our system accordingly to follow actual changes. But we can also control our growth by conscious and rational methods. The change of the European languages has been much slower than before since the beginning of educational and other controls. Remember that we are actors as well as spectators in the drama of human destiny.

Obj. 16 Romanized Chinese is not Chinese, it is foreign. It cannot give Chinese sounds accurately. For instance, spelling the names in 鍾, 趙, 上海 as Chung, Chao, Shanghai would sound like 穹, 竅, 向海 or 香海.

Rep. 16 Consider a parallel case. Imagine that you were a Frenchman, who can speak French but has never learned French spelling, but who also has learned to read and speak English. Now you are told that the French word for year should be spelled an. Since you have always used the Roman letters in English, you would object that an would sound like a girl's name in French and that Roman letters are unfit for expressing the full round nasal sound in the French word for year. But if you go home, you will find that all good Frenchmen pronounce the Roman letters in the French way. So, whatever symbols we use, of course we pronounce them in the Chinese way. If, for the sake of argument, we should see reasons for letting ch denote the sound class of "照"母 (ch), sh that of "審"母 (sh), then Chung, Chao, Shanghai are 穹, 竅, 向海 or 香海 and not 盎 (Shanghai nasal pron.), except when we are imitating foreigners. A Chinese Alphabet may he derived from any source, but it must be of the Chinese, by the Chinese, and for the Chinese.

By this time the patient reader will have observed that the real result of my arguments have been the raising of problems, the challenging of enterprises, rather than the settling of questions. The articles are concluded, the problems are not. I should consider my labors amply repaid if I have succeeded in putting a few thoughtful minds on the right track in judging and working on this most technical and apparently most popular of problems.

The Chinese Students' Monthly, 11:7-8, 1916.

Webmaster's notes

  1. Chao went on to develop just such a practicable romanization system: Gwoyeu Romatzyh. He wrote several books in this. For a sample text in this system, see Humpty Dumpty in Mandarin Chinese (with parallel Hanyu Pinyin and English).
  2. The failure to recognize the distinction between scripts (e.g., Chinese characters, the Roman alphabet, the Greek alphabet) and languages (e.g., Mandarin, English, Greek) is at the root of many, many misunderstandings about the issue of romanization.
  3. The majority of the "existing literature" Chao refers to was written in Literary Sinitic. Since Modern Standard Mandarin replaced Literary Sinitic as the standard for literacy (not too many years after Chao wrote this piece), most people who read pre-twentieth-century Chinese texts do so in translated, paraphrased, or extensively glossed modern editions. In other words, the supposed loss that the objector refers to has already occurred ... and for reasons that have nothing to do with romanization. (See, for example, Hu Shih on the Chinese renaissance.) Then there was China's switch to "simplified" characters, which, although a far less momentous change than the switch in languages, necessitated the issuing of new editions for those who did not know the old script.
  4. This happened just as Chao predicted, and without romanization: the selection of Mandarin based on but not identical to that of Beijing as the national language for China.
  5. For the curious: through, draught, etiquette, row, disaster. People should also be careful to avoid assuming that knowing the etymology of certain Chinese characters is the same as knowing the etymology of the words written with them. (Taiwan is a good example, with the whole "terraced bay" story being nothing but folk etymology.)
  6. See John DeFrancis on why the term "ideographic" is wrong to apply to the system of Chinese characters.
  7. See also William Hannas on the Homonym Problem.
    • 見識: jiànshi (experience; knowledge; sensibleness) and 劍式: jiàn shì (a style of sword use). Not a real homophone pair in Modern Standard Mandarin; thus they're not written identically in Hanyu Pinyin.
    • 清楚: qīngchu (clear; distinct; without ambiguity) and 青醋: qīng cù (a kind of vinegar). Again, not a real homophone pair in Modern Standard Mandarin. (The standard pronunciations of these words may have shifted since this was written.) Also, they're different parts of speech.
    • 年高: niángāo (venerable) and 黏糕: niángāo (sticky rice cake). Finally, an actual homophone pair, which is part of the reason that people eat niángāo during the Chinese New Year season: They want to niánnián gāoshēng 年年高升 (year by year, get better and better — such as in one's position at work). But nobody would get them confused in context, as Chao knows well.
    • 十步階: shí bù jiē (10 steps (as in stairs)) and 拾布街: Shíbù Jiē (Pick Up Cloth Street).
  8. irony: as in "made or consisting of iron"
  9. A Dictionary of English Homonyms: Pronouncing and Explanatory has page after page of examples. If some of the examples there seem obscure, keep in mind that the same is the case with most of the supposed examples in Mandarin, which are often strained in the extreme in the matter of plausibility.
  10. For more on the possibilities of "regular variation," see "Practical Contributions of Pinyin-to-Chinese Character Conversion Systems to Digraphia in China," which is part III of Pinyin-to-Chinese Character Computer Conversion Systems and the Realization of Digraphia in China, by Yin Binyong.
  11. Yen, Liu, Eu, Chao styles: the schools of Chinese calligraphy associated with Yán Zhēnqīng 顏真卿 (709–785), Liǔ Gōngquán 柳公權 (778–865), Ōuyáng Xún 歐陽詢 (557–641), and Zhào Mèngfǔ 趙孟頫 (1254–1322), respectively.
  12. Lü Shuxiang offers a contrary point of view. Much more might also be said about the state of typography, that subtle and fascinating art, when it comes to fonts for the Roman alphabet vs. those designed for Chinese characters.
  13. China already has name problems, with its enormous population and a populace who continue the tradition of giving disyllabic or even monosyllabic names to their children. Under such conditions, instances of many people sharing the same name are inevitable, regardless of the script used to write the names. In fact, romanization could actually help relieve the problems, if the PRC would allow people the basic right of having their official names in their own languages. That way, for example, the people with the Chinese character 陳 assigned to their family name would no longer have to be just Mandarin's Chen, but could also be, as appropriate to their native language, Chan, Chun, Chin, Chinn, Tan, Ding, Ting, Zen, etc.