Some Things Chinese Characters Can’t Do-Be-Do-Be-Do

by David Moser

Ph.D., University of Michigan

Here’s an odd question, but bear with me: How would one scat sing in Chinese? We all know how Ella Fitzgerald does it in English. “Doo-dee-op be-yoo-bee-yiddy-yoo-bee, yabba bip-byoo ba-di-bip-dee-YOOO-bee-op!”1 These are nonsense syllables made up of different vowels and consonants pieced together in rhythmic patterns. Most often they do not correspond to any English morphemes, and in fact, recognizable words are avoided, to retain the flow of pure musical sound free of semantic associations. My question is: What would scat singing performed by the “Chinese Ella Fitzgerald” sound like?

Jazz is, of course, an American art form, so there is no obvious cultural equivalent. Yet since scat singing involves using the human voice to imitate the sound of musical instruments, it might be instructive to compare Ella’s highly developed art to the rhythmic nonsense that Chinese verbal arts performers sing to imitate musical accompaniment patterns. I’m referring to the performers who specialize in quyi 曲艺, the storytelling forms which include Shandong kuaishu 山东快书, kuaibanr 快板, dagu 大鼓, xiangsheng 相声, and so forth. Often in the course of their narratives, a character will provide a little musical patter to simulate Peking Opera accompaniment motifs, including the ubiquitous clanging of the luo 锣, the distinctive little mini-gong that is part of the instrumental array of the wenwuchang 文武场, the group of musicians playing on the stage. The result is phrases like: deng genr li genr long genr long 噔根儿里根儿隆根儿隆 2 and A qiang, a dou, dou a, qi dou qi dou qiang 啊锵,啊豆,豆啊.起豆起豆锵. 3 Some of these phrases are relatively fixed and invariant, while others are more flexibly combined in a semi-improvisational way that bears at least some resemblance to scat singing.

A closer look at these nonsense phrases reveals a deep difference in the way the two languages are perceived and processed. What Ella Fitzgerald is doing is to mix and match English phonemes to create novel syllabic structures, the result being “legal” but non-existent syllables like “dwee”, “wap”, “yab”, “byoo”, etc. In contrast, what the Chinese performers seem to be doing is juxtaposing entire syllables of the language, choosing from the 1280 or so set of distinct syllables in Mandarin (with tonal information), and stringing them into onomatopoeic patterns. There is no manipulation of the segments, no mixing and matching of phonemes.

Isn’t it an amazing coincidence that these singers quite naturally “restrict” themselves (if that is the word) to exactly the linguistic level that their writing system represents? After all, one can easily imagine manipulation of Chinese at the phoneme level. Using a kind of quasi-fanqie assembly method to notate it, you could arrive at creations like these (which I render here in pinyin):

g as in “gan” + ing as in “ding” = ging

shu as in “shuang” + ing as in “ding” = shuing

p as in “ping” + uan as in “guan” = puan

d as in “dong” + ua as in “gua” = dua

ku as in “kuan” + en as in “men” = kuen

And so on. A Chinese Ella might sing something like “Ging-a-ging puan dua-bo KUEN-de shuing shuing!” (It don’t mean a thing if it ain’t got that “shuing”?) Those accustomed to pinyin will find these creations a little off-putting at first, but they are all perfectly readable creations based on the sound-to-orthography rules of pinyin. Note that if these syllables are “illegal” in any sense, they are only so because the Chinese characters cannot accommodate them, not because the sounds of Chinese (as represented by pinyin, Wade-Giles, IPA or whatever) cannot be recombined in these ways.

So why are the Chinese performers so accommodating to this idiosyncrasy of the Chinese writing system? Do speakers of the language “naturally” carve up their speech into neat morphosyllabic chunks, as reflected by the character set? Or do they gravitate toward this level because the writing system has coerced them into doing so?4 Which came first, the kung-pao chicken or the egg drop soup?

The problem is that Chinese speakers don’t always adhere to the principle their writing system is based on, preferring to “play it by ear” very often, and in such cases the writing system does not do a very good job of representing their speech. If the syllabic nature of the Chinese writing system precluded only Ella Fitzgerald-style scat singing, it would not be a very interesting restriction. However, this quality of the Chinese characters also effectively precludes a host of other orthographic conveniences and techniques that alphabetic systems afford. In what follows, I mention just a few.

English has numerous conventions for representing casual oral speech: “Are you kiddin’ me?” “Whaddya wanna do tonight, Marty?” “I’m gettin’ outta here!” “Gimme that.” And so on. Such spelling conventions have been employed in the literature of most alphabetic traditions for hundreds of years, and are often an invaluable link to the vernaculars of the past. English-language writers from Mark Twain to James Joyce have used the flexibility of the alphabet to vividly re-created various speech worlds in their works. It is, in fact, hard to imagine how much of the literature of the West could have been produced without recourse to such devices.

Chinese characters, by contrast, cannot reproduce the equivalent elisions and blends of colloquial Chinese, except in rare cases, and only at the level of the syllable. Chinese readers who encounter written phrases like 哥们儿 gemenr (“pal, buddy”) or 宝贝儿 baobeir (“baby, sweetie, etc.”) will drop the the /n/ sound from men 们 in the first compound, and perform a vowel shift on the character bei 贝 from the second, but this is effectively due to the fact that the er 儿suffix codes for a different syllable in each case. I have also occasionally seen one character substituted for another in writing to represent the way sounds are truncated and altered in everyday language (for example, a pop star on stage saying a perfunctory thanks to the audience after a performance written as “西西” xixi, instead of “谢谢” xiexie). Such devices are not productive, however, and are not often used, for the obvious reason that the substitution of a different character discards the original semantic information and substitutes different semantics.5 The result is that China effectively has no tradition of realistically notating vernacular speech. Wenyanwen 文言文, classical Chinese, exerted a virtual stranglehold on written literature up until the early twentieth century, and even then, most writers did not attempt to accurately represent common speech, despite the appearance of an occasional Lao She or Ba Jin. But even if such writers had so desired, working within the Chinese system of writing, they could never have notated the sounds of the language around them with the same kind of vivid verisimilitude of the following examples in English:

Mark Twain, Huckleberry Finn:

“Yo’ ole father doan’ know yit what he’s a-gwyne to do. Sometimes he spec he’ll go ‘way, en den agin he spec he’ll stay. De bes’ way is to res’ easy en let de ole man take his own way. Dey’s two angels hoverin’ roun’ ‘bout him. One uv ‘em is white en shiny, en t’other one is black. De white one gits him to go right a little while, den de black one sail in en bust it all up. A body can’t tell yit which one gwyne to fetch him at de las’. But you is all right. You gwyne to have considable trouble in yo’ life, en considable joy. Sometimes you gwyne to git hurt, en sometimes you gwyne to git sick; but every time you’s gwyne to git well agin. Dey’s two gals flyin’ ‘bout you in yo’ life. One uv ‘em’s light en t’other one is dark. One is rich en t’other is po’. You’s gwyne to marry de po’ one fust en de rich one by en by. You wants to keep ‘way fum de water as much as you kin, en don’t run no resk, ‘kase it’s down in de bills dat you’s gwyne to git hung.”

George Bernard Shaw, Pygmalion:

The Mother: How do you know that my son’s name is Freddy, pray?

The Flower Girl: Ow, eez yə-ooa san, is e? Wal, fewd dan y’də-ooty bawmz a mather should, eed now bettern to spawl a pore gel’s flahrzn than ran awy athaht pyin. Will ye-oo py me f’them?
[Oh, he’s your son, is he? Well, if you’d done your duty by him as a mother should, he’d know better than to spoil a poor girl’s flowers and then run away without paying. Will you pay me for them?]

Charles Dickens, The Cricket on the Hearth:

‘An’t he beautiful, John? Don’t he look precious in his sleep?’

‘Very precious,’ said John. ‘Very much so. He generally is asleep, an’t he?’

‘Lor, John! Good gracious no!’

‘Oh,’ said John, pondering. ‘I thought his eyes was generally shut. Halloa!’

‘Goodness, John, how you startle one!’

‘It an’t right for him to turn ‘em up in that way!’ said the astonished Carrier, ‘is it? See how he’s winking with both of ‘em at once! And look at his mouth! Why he’s gasping like a gold and silver fish!’

Walt Kelly, Pogo:

Weevil: You isn’t from China. You is mere a common ant bug.

Pogo: Why, Mr. Weevil. I sees our Oriental friend clumb outen this hole afore my very own soft brown eyes.

Ant: Sho’ nuff! I will talk some more China: “Chicken Chow Dog. Egg Foo Young. Okay boss, plenty of starch.”

Pogo: Man! What more proof is you need?

Weevil: Who can’t talk that kind Chinese? Egg Foo Young, Egg Foo Old, Egg Foo in the pot, nine days old? This hole you is say come up from China is only a inch deep.

Pogo: AY-mazin’! Din’t have no idea China was so close.

This sort of thing is quite impossible to achieve with Chinese characters. Due to the nature of the Chinese writing system, China has no Mark Twains, no Dickenses, no Faulkners, no James Joyces; that is, no literature with phonetically realistic re-creations of vernacular speech. Chinese characters effectively preclude such writing, though authors have made masterful attempts working within the Chinese system. As hypothetical as it might be, and at the risk of mixing apples and Mandarin oranges, it is interesting to imagine what kinds of literature Lao She or the contemporary writer Wang Shuo might have produced had they been able to work within an alphabetic system. How could the actual sounds of spoken Chinese be written if pinyin were the medium rather than the characters?

Those familiar with the colloquial putonghua environment know that Chinese people, like the speakers of all languages, do not utter words and phrases according to textbook standards. Just to pick a very few examples, common phrases like bu zhidao 不知道 , “[I] don’t know”, duoshao qian 多少钱, “How much money?” or zenmehuishi? 怎么回事?, “What’s this all about?” could be represented in pinyin as follows to represent the way they are usually spoken:

不知道. Bu zhidao. —> Bu’r dao.
多少钱? Duoshao qian? —> Duo’ao qian?
怎么回事? Zenmehuishi? —> Ze’m hui shi?

Note that, just as in English, one does not need to find linguistically accurate phonetic renditions of these forms, merely ones that intuitively map onto the sounds of the original. My versions above are merely suggestions, and if pinyin were used to write Chinese, native speakers would arrive at reasonable conventional renditions that would trigger the target sounds. Orthographic representation at the phoneme level would open up vast worlds of sound in written Chinese. Putonghua with various dialect accents could be represented with enough accuracy to evoke the actual phonetic flavor of the real thing. For example, the speech of southern speakers, who do not usually pronounce the retroflex initials of northern Mandarin, could be represented as:

不知道. Bu zhidao. —> Bu zidao.
多少钱? Duoshao qian? —> Duosao qian?
怎么回事? Zenmehuishi? —> Ze’m hui si?

And so on. Currently Chinese writers, if they wish to evoke the sounds of regional dialects in their writing, can only do so by the inclusion of giveaway lexical items, such as the substitution of an 俺, a dialect form, for wo 我, “me”. Given these lexical cues, the reader then mentally shifts the “voice” of the passage into the intended dialect, just as Americans, when reading “I say, old chap, what say we go get a spot of sherry?” can be expected to begin hearing the sounds of British English in their heads.

Using pinyin, foreign accents could also be represented, just as they are in English:

Eet eez easy to noteece zat I am writing wiz a Franch accent, non?

Und now I haff svitched to a Cherman accent.

Just as in these English examples, rendering foreign-accented Mandarin in pinyin might also entail bizarre pinyin spellings, but this is half the fun of it—coercing the tongue into silly or non-native sounds, something Chinese characters are incapable of.

Needless to say, nonsense words a la Lewis Carroll are also effectively blocked in Chinese orthography:

  Twas brillig, and the slithy toves
  Did gyre and gimble in the wabe:
  All mimsy were the borogoves,
  And the mome raths outgrabe. 

Zhao Yuanren did marvelous translations of Alice in Wonderland and Through the Looking Glass into Chinese, attempts that found miraculous equivalences and creative solutions to Carroll’s whimsy. In translating the above poem, The Jabberwocky, Zhao had to resort to made-up characters, such as the component 白 with the “fire” radical 灬 below it to translate “brillig”, a creation that would presumably be read bai, but “presumably” is precisely the problem. Zhao wisely used this technique sparingly, as too many such characters in the text soon results in a kind of visual game in which the whimsical play with spoken language is lost.

I won’t bore the reader with extensive counterfactual exercises imagining Chinese literature within an alphabetic system. I think one can easily imagine the possibilities. I also don’t want to suggest that works like Lao She’s Teahouse are somehow merely failed attempts to create a Dickensian linguistic world. Lao She was working brilliantly within the only system he had at his disposal, and the success he achieved was on its own terms. But given that there is a quite considerable functional overlap between Chinese characters and pinyin (that is, pinyin can fulfill almost every function performed by the characters) it is reasonable to raise issues of relative power and flexibility.

Finally, a note on the transliteration of proper names. Chinese characters isolate the Chinese texts from world community, acting like a kind of firewall against the alphabetic domain. This is nowhere more apparent than in the translation and use of foreign proper names, which is a real headache in Chinese, much more so than in any other world language. In the roman alphabet universe, proper names are common currency that can be traded freely, with only negligible tweaks, between languages. English accommodates German names like “Beethoven”, German absorbs “Sartre”, and French doesn’t bat an eye at “Eminem.” Non-native spellings are either pronounced according to indigenous spelling rules, or more-or-less sophisticated stabs are made at the actual pronunciation—“Bach” in the glottal German way or Americanized to sound like “Bok.” Occasional wholesale translations occur, such as the French penchant for “Jean-Sebastien Bach” instead of “Johann Sebastian Bach,” but these are rare. Usually the character string remains invariant, and the pronunciation takes care of itself.6 And transliteration from other alphabets, such as Russia’s Cyrillic, is usually also a rather straightforward, rule-governed process. “Chaikofsky” and “Tchaikovsky” might compete for a time, but eventually a consensus is reached.7 And even transliterations from Chinese present no serious problem, since the sounds first go through the standard romanization (formerly Wade-Giles, now universally pinyin), and then can be printed in any alphabetic language, to be pronounced willy-nilly as the natives see fit.8

Not so in Chinese, where every new proper name must go through a torturous process of sinification in order to enter the language. Though there is a relatively small set of characters that are routinely used to transliterate foreign names into Chinese, such as 斯 si, 迪 di, 拉 la, 顿 dun, 巴 ba, 里 li, 克 ke, 特 te, and so on, these are applied rather haphazardly. In addition, Taiwan and the mainland often diverge, with, for example, “Reagan” being rendered as Ligen 里根 in the mainland and Leigen 雷根 across the Strait.

To make matters worse, some brand names are transliterated with the sounds of Cantonese in mind, and others into Mandarin, resulting in puzzling clunkers like Maidanglao 麦当劳, Bishengke 必胜客, and Shashebiya 沙士比亚, all of of which sound a lot more like their targets (“McDonald’s”, “Pizza Hut” and “Shakespeare”) in Cantonese than in Mandarin. All these transliteration problems would not be automatically solved by a switch to pinyin, but significant degree of ambiguity and uncertainty would be reduced.

As for scat singing, music is a kind of universal language, and I don’t think we need a Chinese Ella Fitzgerald when the American one will do nicely. But who knows? Maybe someday someone will translate this artform into Chinese, and instead of deng genr li genr long genr long 噔根儿里根儿隆根儿隆, it will be du bi du bi du 嘟比嘟比嘟.


  1. From Ella’s classic recording of How High the Moon, from the collection Something to Live For, Verve Records 314-547-800-2, 1990.
  2. From the xiangsheng piece Chuan diao,《串调》 , in Zhongguo chuantong xiangsheng daquan《中国传统相声大全》( Beijing: Wenhua yishu chubanshe 文化艺术出版社, p. 455).
  3. From the xiangsheng piece Mai baozi《卖包子》, also from Zhongguo chuantong xiangsheng daquan, p. 470.
  4. For valuable opinions on this matter, see William Hannas’ two books, Asia’s Orthographic Dilemma, Honolulu: University of Hawaii Press (1997), and The Writing on the Wall: How Asian Orthography Curbs Creativity. Philadelphia: University of Pennsylvania Press (2003).
  5. In English one can also play the game of simply substituting homophonic syllables to simulate actual speech. I remember a website that used to parody Bill Clinton’s Arkansas accent, providing “translations” of his phrases for the uninitiated. For example, “child care center” was spoken in Clintonese as “chalk air center.” This is isomorphic to the Chinese technique, and as humorous as it is, it’s clear that it won’t get one very far, for the same reason: the interfering semantic sense of the substitution requires heavy contexting.
  6. Except for occasional breakdowns, when Americans try to pronounce a Polish name like “Szczepan Szczurowski.”
  7. There is Woody Allen’s joke to the effect that the Russian revolution broke out when the people realized that the “czar” and the “tsar” were actually the same person.
  8. Such as “Beijing”, the /j/ of which American newscasters inexplicably pronounce like the /s/ in “pleasure”. My guess is that the doggedly monolingual Americans assume that all /j/ sounds in foreign terms are to be pronounced according to the only foreign language they have any knowledge of, which is high-school French. Making the analogy with French words like je and jeter, they “sophisticatedly” avoid the more obvious pronunciation.